DRAFT DO NOT CITE OR QUOTE
oEPA EPA/635/R-16/001b
Final Agency/Interagency Science Discussion Draft
www.epa.gov/iris
Toxicological Review of Trimethylbenzenes
(CASRNs 25551-13-7, 95-63-6, 526-73-8, and 108-67-8]
Supplemental Information
June 2016
NOTICE
This document is a Final Agency/Interagency Science Discussion 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.
Integrated Risk Information System
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Supplem en tal Information—Trim ethylbenzenes
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ii DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
CONTENTS
APPENDIX A. RESPONSE TO EXTERNAL PEER REVIEW COMMENTS PROVIDED BY THE CHEMICAL
ASSESSMENT ADVISORY COMMITTEE OF THE SCIENCE ADVISORY BOARD A-l
APPENDIX B. HEALTH ASSESSMENTS AND REGULATORY LIMITS BY OTHER NATIONAL AND
INTERNATIONAL HEALTH AGENCIES B-l
APPENDIX C. INFORMATION IN SUPPORT OF HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS C-l
C.l. TOXICOKINETICS C-l
C.1.1. Absorption C-l
C.l.2. Distribution C-2
C.l.3. Metabolism C-3
C.1.4. Excretion C-7
C.2. PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELS C-7
C.2.1. Summary of Available Physiologically Based Pharmacokinetic
(PBPK) Models for 1,2,4-TMB C-7
C.2.2. 1,2,4-TMB PBPK Model Selection C-14
C.2.3. Details of Hissink et al. (2007) Model Analysis C-15
C.2.4. Summary of Available PBPK models for 1,3,5-TMB or 1,2,3-TMB C-49
C.3. HUMAN STUDIES C-49
C.4. ANIMAL TOXICOLOGY STUDIES C-61
C.5. HUMAN TOXICOKINETIC STUDIES C-187
C.6. ANIMAL TOXICOKINETIC STUDIES C-201
C.7. ANIMAL AND HUMAN TOXICOKINETIC STUDIES C-228
APPENDIX D. DOSE-RESPONSE MODELING FOR THE DERIVATION OF REFERENCE VALUES FOR
EFFECTS OTHER THAN CANCER AND THE DERIVATION OF CANCER RISK
ESTIMATES D-l
D.l. BENCHMARK DOSE (BMD) MODELING SUMMARY D-l
D.l.l. Noncancer Endpoints D-l
REFERENCES FOR APPENDICES R-l
This document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
TABLES
Table B-l. Other national and international health agency assessments for trimethylbenzenes
(TMBs) B-l
Table C-l. Measured and calculated partition coefficients for TMB isomers at 37°C C-9
Table C-2. PBPK model parameters for 1,2,4-TMB toxicokinetics in humans using the Jarnberg
and Johanson (1999) model structure C-10
Table C-3. Comparison of rat anatomical and physiological parameters in Hissink et al. (2007) to
those of Brown et al. (1997) C-18
Table C-4. 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) C-19
Table C-5. Comparison of chemical-specific parameters in Hissink et al. (2007) to literature data C-20
Table C-6. Parameter values for the rat and human PBPK models for 1,2,4-TMB used by EPA C-26
Table C-7. Rat 1,2,4-TMB kinetic studies used in model development and testing C-27
Table C-8. Model simulated and experimental measured venous blood concentrations of
1,2,4-TMB in male Wistar rats exposed to 1,2,4-TMB C-30
Table C-9. Model simulated and experimental measured tissue concentrations of 1,2,4-TMB in
male Wistar rats exposed to 1,2,4-TMB C-32
Table C-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 hours/day, for
3 days) at the end of exposure or 12 hours after the last exposure C-33
Table C-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 C-34
Table C-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 hours/day, for 14 days) at the end of exposure C-35
Table C-13. Human kinetic studies of 1,2,4-TMB used in model validation C-36
Table C-14. Parameter sensitivity for venous blood 1,2,4-TMB concentration in rats exposed to
1,2,4-TMB via inhalation C-45
Table C-15. Parameter sensitivity for steady-state venous blood 1,2,4-TMB concentration in
humans exposed to 1,2,4-TMB via inhalation C-47
Table C-16. Characteristics and quantitative results for epidemiologic studies of TMB and
related compounds and mixtures C-50
Table C-17. Characteristics and quantitative results for Adenuga et al. (2014) C-61
Table C-18. Characteristics and quantitative results for Battig et al. (1958) C-68
Table C-19. Characteristics and quantitative results for Carrillo et al. (2014) C-73
Table C-20. Characteristics and quantitative results for Clark et al. (1989) C-79
Table C-21. Characteristics and quantitative results for Douglas et al. (1993) C-85
Table C-22. Characteristics and quantitative results for Gralewicz et al. (1997b) C-90
Table C-23. Characteristics and quantitative results for Gralewicz et al. (1997a) C-94
Table C-24. Characteristics and quantitative results for Gralewicz and Wiaderna (2001) C-96
Table C-25. Characteristics and quantitative results for Janik-Spiechowicz et al. (1998) C-99
Table C-26. Characteristics and quantitative results for Juran et al. (2014) C-103
Table C-27. Characteristics and quantitative results for Koch Industries (1995b) C-106
Table C-28. Characteristics and quantitative results for Korsak et al. (1995) C-115
Table C-29. Characteristics and quantitative results for Korsak and Rydzyriski (1996) C-118
Table C-30. Characteristics and quantitative results for Korsak et al. (1997) C-123
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Tab
e C-31.
Characteristics
and
quantitative
resu
ts for Korsak et al. (2000a)
C-125
Tab
e C-32.
Characteristics
and
quantitative
resu
ts for Korsak et al. (2000b)
C-130
Tab
e C-33.
Characteristics
and
quantitative
resu
ts for Lammers et al. (2007)
C-135
Tab
e C-34.
Characteristics
and
quantitative
resu
ts for Lutz et al. (2010)
C-138
Tab
e C-35.
Characteristics
and
quantitative
resu
ts for Maltoni et al. (1997)
C-141
Tab
e C-36.
Characteristics
and
quantitative
resu
ts for Mckee et al. (1990)
C-143
Tab
e C-37.
Characteristics
and
quantitative
resu
ts for Mckee et al. (2010)
C-150
Tab
e C-38.
Characteristics
and
quantitative
resu
ts for Saillenfait et al. (2005)
C-153
Tab
e C-39.
Characteristics
and
quantitative
resu
ts for Schreiner et al. (1989)
C-158
Tab
e C-40.
Characteristics
and
quantitative
resu
ts for Tomas et al. (1999a)
C-166
Tab
e C-41.
Characteristics
and
quantitative
resu
ts for Tomas et al. (1999b)
C-168
Tab
e C-42.
Characteristics
and
quantitative
resu
ts for Tomas et al. (1999c)
C-169
Tab
e C-43.
Characteristics
and
quantitative
resu
ts for Wiaderna et al. (1998)
C-172
Tab
e C-44.
Characteristics
and
quantitative
resu
ts for Wiaderna et al. (2002)
C-176
Tab
e C-45.
Characteristics
and
quantitative
resu
ts for Wiglusz et al. (1975b)
C-178
Tab
e C-46.
Characteristics
and
quantitative
resu
ts for Wiglusz et al. (1975a)
C-182
Tab
e C-47.
Characteristics
and
quantitative
resu
ts for Jarnberg et al. (1996)
C-187
Tab
e C-48.
Characteristics
and
quantitative
resu
ts for Jarnberg et al. (1997a)
C-191
Tab
e C-49.
Characteristics
and
quantitative
resu
ts for Jarnberg et al. (1997b)
C-193
Tab
e C-50.
Characteristics
and
quantitative
resu
ts for Jarnberg et al. (1998)
C-194
Tab
e C-51.
Characteristics
and
quantitative
resu
ts for Jones et al. (2006)
C-195
Tab
e C-52.
Characteristics
and
quantitative
resu
ts for Kostrzewski et al. (1997)
C-197
Tab
e C-53.
Characteristics
and
quantitative
resu
ts for Dahl et al. (1988)
C-201
Tab
e C-54.
Characteristics
and
quantitative
resu
ts for Eide and Zahlsen (1996)
C-202
Tab
e C-55.
Characteristics
and
quantitative
resu
ts for Huo et al. (1989)
C-203
Tab
e C-56.
Characteristics
and
quantitative
resu
ts for Mikulski and Wiglusz (1975)
C-205
Table C-57 Characteristics and quantitative results for Swiercz et al. (2002) C-206
Table C-58. Characteristics and quantitative results for Swiercz et al. (2003) C-208
Table C-59. Characteristics and quantitative results for Swiercz et al. (2006) C-210
Table C-60. Characteristics and quantitative results for Swiercz et al. (2016) C-213
Table C-61. Characteristics and quantitative results for Tsujimoto et al. (2000) C-221
Table C-62. Characteristics and quantitative results for Tsujimoto et al. (2005) C-222
Table C-63. Characteristics and quantitative results for Tsujino et al. (2002) C-223
Table C-64. Characteristics and quantitative results for Zahlsen et al. (1990) C-225
Table C-65. Characteristics and quantitative results for Zahlsen et al. (1992) C-227
Table C-66. Characteristics and quantitative results for Meulenberg and Vijverberg (2000) C-228
Table D-l. Noncancer endpoints selected for dose-response modeling for 1,2,3-TMB,
1,2,4-TMB, and 1,3,5-TMB D-2
Table D-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) (Korsakand Rydzynski, 1996) D-4
Table D-3. 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 (modeled variance) (Korsak and Rydzynski, 1996) D-5
Table D-4. 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) D-6
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table D-5. 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) (Korsakand Rydzynski, 1996) D-8
Table D-6. 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) (Korsak and Rydzynski, 1996) D-9
Table D-7. 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, high dose dropped) (Korsakand Rydzynski, 1996) D-10
Table D-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 (modeled variance, high dose dropped) (Korsak and Rydzynski, 1996) D-ll
Table D-9. Summary of BMD modeling results for decreased RBCs 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) D-13
Table D-10. 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 variance) (Korsak et al., 2000a) D-16
Table D-ll. 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 (modeled variance) (Korsak et al., 2000a) D-17
Table D-12. 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 variance, high dose dropped) (Korsak et al., 2000a) D-18
Table D-13. 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 (modeled variance, high dose dropped) (Korsak et al., 2000a) D-19
Table D-14. 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., 2000a) D-20
Table D-15. 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., 2000a) D-23
Table D-16. Summary of BMD modeling results for increased reticulocytes 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., 2000a) D-26
Table D-17. Summary of BMD modeling results for decreased fetal weight in male Sprague-
Dawley rat pups exposed to 1,2,4-TMB by inhalation on GDs 6-20; BMR = 1 SD or
5% change from control mean (constant variance) (Saillenfait et al., 2005) D-29
Table D-18. Summary of BMD modeling results for decreased fetal weight in male Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (constant variance) (Saillenfait et al., 2005) D-32
Table D-19. Summary of BMD modeling results for decreased fetal weight in male Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (modeled variance) (Saillenfait et al., 2005) D-33
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table D-20. Summary of BMD modeling results for decreased fetal weight in male Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (constant variance, high dose dropped) (Saillenfait et al.,
2005) D-34
Table D-21. Summary of BMD modeling results for decreased fetal weight in male Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (modeled variance, high dose dropped) (Saillenfait et al.,
2005) D-35
Table D-22. Summary of BMD modeling results for decreased fetal weight in female Sprague-
Dawley rat pups exposed to 1,2,4-TMB by inhalation on GDs 6-20; BMR = 1 SD or
5% change from control mean (constant variance) (Saillenfait et al., 2005) D-36
Table D-23. Summary of BMD modeling results for decreased fetal weight in female Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (constant variance) (Saillenfait et al., 2005) D-39
Table D-24. Summary of BMD modeling results for decreased fetal weight in female Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (modeled variance) (Saillenfait et al., 2005) D-40
Table D-25. Summary of BMD modeling results for decreased fetal weight in female Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (constant variance, high dose dropped) (Saillenfait et al.,
2005) D-41
Table D-26. Summary of BMD modeling results for decreased fetal weight in female Sprague-
Dawley rat pups exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (modeled variance, high dose dropped) (Saillenfait et al.,
2005) D-42
Table D-27. Summary of BMD modeling results for decreased dam weight gain in female
Sprague-Dawley rats exposed to 1,2,4-TMB by inhalation on GDs 6-20; BMR = 1 SD
or 10% change from control mean (constant variance) (Saillenfait et al., 2005) D-43
Table D-28. Summary of BMD modeling results for decreased dam weight gain in female
Sprague-Dawley rats exposed to 1,2,4-TMB by inhalation on GDs 6-20; BMR = 1 SD
or 10% change from control mean (modeled variance) (Saillenfait et al., 2005) D-44
Table D-29. Summary of BMD modeling results for decreased dam weight gain in female
Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (constant variance) (Saillenfait et al., 2005) D-47
Table D-30. Summary of BMD modeling results for decreased dam weight gain in female
Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (modeled variance) (Saillenfait et al., 2005) D-48
Table D-31. Summary of BMD modeling results for decreased dam weight gain in female
Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (constant variance, high dose dropped) (Saillenfait et al.,
2005) D-49
Table D-32. Summary of BMD modeling results for decreased dam weight gain in female
Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on GDs 6-20; BMR = 1 SD
change from control mean (modeled variance, high dose dropped) (Saillenfait et al.,
2005) D-50
Table D-33. Summary of BMD modeling results for increased monocytes in male Wistar rats
exposed to 1,3,5-TMB by gavage for 13 weeks; BMR = 1 SD change from control
mean (constant variance) (Adenuga et al., 2014) D-51
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table D-34. Summary of BMD modeling results for increased monocytes in male Wistar rats
exposed to 1,3,5-TMB by gavage for 13 weeks; BMR = 1 SD change from control
mean (modeled variance) (Adenuga et al., 2014) D-52
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
FIGURES
Figure C-l. Metabolic scheme for 1,2,4-TMB C-5
Figure C-2. Metabolic scheme for 1,2,3-TMB C-6
Figure C-3. Metabolic scheme for 1,3,5-TMB C-6
Figure C-4. Physiologically based toxicokinetic model for 1,2,4-TMB in humans C-8
Figure C-5. Schematic of human model structure for 1,2,4-TMB using the NLE-based model
approach C-12
Figure C-6. Schematic of rat and human PBPK model structure C-14
Figure C-l. Simulated and measured blood concentrations of 1,2,4,-TMB in rats exposed to 600,
2,400, or 4,800 mg/m3 white spirit for 8 hours C-22
Figure C-8. Simulated and measured brain concentrations of 1,2,4-TMB in rats exposed to 600,
2,400, or 4,800 mg/m3 white spirit for 8 hours C-23
Figure C-9. Simulated and measured exhaled air concentrations of 1,2,4-TMB in three
volunteers exposed to 600 mg/m3 white spirit for 4 hours C-24
Figure C-10. Comparisons of model predictions to measured blood concentrations in rats
exposed to 1,2,4-TMB in white spirit (Hissink et al., 2007) (a) before and (b) after
numerical optimization C-28
Figure C-ll. Comparisons of model predictions to measured brain concentrations in rats
exposed to 1,2,4-TMB in white spirit (Hissink et al., 2007) using model parameters
optimized for fit to Hissink et al. (2007) rat blood data C-29
Figure C-12. Comparisons of model predictions to measured venous blood concentrations by
Swiercz et al. (2003) in rats repeatedly exposed to 1,2,4-TMB (a) before and (b) after
numerical optimization C-30
Figure C-13. Comparisons of model predictions to measured rat venous blood concentrations by
Swiercz et al. (2002) in acutely exposed rats (a) during and (b) after exposure C-31
Figure C-14. Comparisons of model predictions to measured human venous blood
concentrations of Kostrzewski et al. (1997) in volunteers exposed to 154 mg
1,2,4-TMB/m3 for 8 hours C-37
Figure C-15. Comparisons of model predictions to measured human venous blood
concentrations 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) (Jarnberg et al., 1998, 1997a; Jarnberg et
al., 1996) C-3 8
Figure C-16. Comparisons of model predictions to measured (a) human venous blood and (b)
end of exposure exhaled air 1,2,4-TMB in volunteers exposed to 100 ppm white
spirit with 7.8% 1,2,4-TMB (38.4 mg/m3 1,2,4-TMB) (Hissink et al., 2007) C-38
Figure C-17. Time course of NSCs 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 (Swiercz et al.,
2003; Swiercz et al., 2002) C-46
Figure C-18. Effect of route of exposure and dose rate on steady-state venous blood
concentration (t = 1,200 hours) for continuous human exposure to 1,2,4-TMB C-49
Figure D-l. Plot of mean response by dose for increased latency to paw-lick in male Wistar rats,
with fitted curve for Linear model with constant variance (Korsak and Rydzyriski,
1996) D-6
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Figure D-2. Plot of mean response by dose for increased latency to paw-lick in male Wistar rats,
with fitted curve for Linear model with constant variance (Korsak and Rydzyriski,
1996) D-ll
Figure D-3. Plot of mean response by dose for decreased RBCs in male Wistar rats, with fitted
curve for Exponential 2 model with constant variance (Korsak and Rydzynski, 1996) D-14
Figure D-4. Plot of mean response by dose for decreased segmented neutrophils in male Wistar
rats, with fitted curve for Exponential M2 model with constant variance (Korsak et
al., 2000a) D-21
Figure D-5. Plot of mean response by dose for decreased segmented neutrophils in female
Wistar rats, with fitted curve for Hill model with constant variance (Korsak et al.,
2000a) D-24
Figure D-6. Plot of mean response by dose for increased reticulocytes in female Wistar rats,
with fitted curve for Linear model with constant variance (Korsak et al., 2000a) D-27
Figure D-7. Plot of mean response by dose for decreased fetal weight in male Sprague-Dawley
rat pups, with fitted curve for Linear model with constant variance (Saillenfait et al.,
2005) D-30
Figure D-8. Plot of mean response by dose for decreased fetal weight in male Sprague-Dawley
rat pups, with fitted curve for Linear model with constant variance (Saillenfait et al.,
2005) D-30
Figure D-9. Plot of mean response by dose for decreased fetal weight in female Sprague-Dawley
rat pups, with fitted curve for Linear model with constant variance (Saillenfait et al.,
2005) D-37
Figure D-10. Plot of mean response by dose for decreased fetal weight in female Sprague-
Dawley rat pups, with fitted curve for Linear model with constant variance
(Saillenfait et al., 2005) D-37
Figure D-ll. Plot of mean response by dose for decreased dam weight gain in female Sprague-
Dawley rats, with fitted curve for Polynomial 3 model with modeled variance
(Saillenfait et al., 2005) D-45
Figure D-12. Plot of mean response by dose for decreased dam weight gain in female Sprague-
Dawley rats, with fitted curve for Polynomial 3 model with modeled variance
(Saillenfait et al., 2005) D-45
Figure D-13. Plot of mean response by dose for increased monocytes in male Wistar rats, with
fitted curve for Exponential M4 model with modeled variance (Adenuga et al.,
2014) D-53
This document is a draft for review purposes only and does not constitute Agency policy.
x DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
ABBREVIATIONS
AAQC
Ambient air quality criterion
Hb/g-A
animal blood:gas partition coefficient
ABR
amount of 1,2,4-TMB in the brain
Hb/g-H
human blood:gas partition coefficient
ADME
absorption, distribution, metabolism,
HEC
human equivalent concentration
and excretion
HED
human equivalent dose
AEGL
Acute Exposure Guideline Level
HERO
Health and Environmental Research
AIC
Akaike Information Criterion
Online
ALT
alanine aminotransferase
HFAN
High-Flash Aromatic Naphtha
ANCOVA
analysis of covariance
HLVOC
highly lipophilic volatile organic
ANOVA
analysis of variance
chemical
AP
alkaline phosphatase
HSDB
Hazardous Substances Data Bank
AST
aspartate aminotransferase
IL-8
interleukin-8
AUC
area under the curve
i.p.
intraperitoneal
BAL
bronchoalveolar lavage
IRIS
Integrated Risk Information System
BMCL
lower confidence limit on the
JP-8
jet propulsion fuel 8
benchmark concentration
KCCT
kaolin-cephalin coagulation time
BMD
benchmark dose
Km
Michaelis-Menten constant
BMDL
lower confidence limit on the
LLF
log-likelihood function
benchmark dose
LOAEL
lowest-observed-adverse-effect level
BMDS
benchmark dose software
MCH
mean corpuscular hemoglobin
BMR
benchmark response
MCHC
mean corpuscular hemoglobin
BrdU
5-bromo-2'-deoxyuridine
concentration
BUN
blood urea nitrogen
MCV
mean cell volume
BW
body weight
MMS
methyl methanesulfate
CAAC
Chemical Assessment and Advisory
MOE
Ministry of the Environment
Committee
NIOSH
National Institute for Occupational
CASRN
Chemical Abstracts Service Registry
Safety and Health
Number
NLE
neutral lipid equivalent
CE
cloning efficiency
NLM
National Library of Medicine
CHO
Chinese hamster ovary
NMDA
N- methyl-D-aspartate
CI
confidence interval
NOAEL
no-observed-adverse-effect level
CMIX
average of arterial and venous blood
NOEL
no-observed-effect level
concentrations
NRC
National Research Council
CNS
central nervous system
NSC
normalized sensitivity coefficient
CV
concentration in venous blood
OSHA
Occupational Safety and Health
CVS
concentration in venous blood exiting
Administration
slowly perfused tissues
p-value
probability value
CXEQ
concentration in exhaled breath
PBPK
physiologically based pharmacokinetic
CYP450
cytochrome P450
(model)
DAF
dosimetric adjustment factor
PCV
packed cell volume
df
degree of freedom
Pg
picogram
DMBA
dimethylbenzoic acid
PMR
proportional mortality ratio
DMHA
dimethylhippuric acid
PND
postnatal day
DMSO
dimethylsulfoxide
POD
point of departure
DNA
deoxyribonucleic acid
PODadj
duration-adjusted POD
ECso
half maximal effective concentration
ppm
parts per million
EEG
electroencephalogram
QPC
alveolar ventilation rate
EPA
U.S. Environmental Protection Agency
OR
odds ratio
fMRI
functional magnetic resonance imaging
QRTOTC
sum of fractional flows to rapidly
GABA
gamma-aminobutyric acid
perfused tissues, liver, and brain
GD
gestational day
QSTOTC
sum of fractional flows to slowly
GGT
gamma-glutamyl transpeptidase
perfused tissues
This document is a draft for review purposes only and does not constitute Agency policy.
xi DRAFT-DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
RBC
red blood cell
TOXLINE
Toxicology Literature Online
RD
relative deviation
TWA
time-weighted average
RD50
50% respiratory rate decrease
UF
uncertainty factor
REL
recommended exposure limit
UFa
interspecies uncertainty factor
RfC
reference concentration
UFh
intraspecies uncertainty factor
RfD
reference dose
UFs
subchronic-to-chronic uncertainty
ROS
reactive oxygen species
factor
SAB
Science Advisory Board
UFl
LOAEL-to-NOAEL uncertainty factor
SCE
sister chromatid exchange
UFd
database deficiency uncertainly factor
SCI
Science Citation Index
VEP
visual evoked potential
SD
standard deviation
Vmax
% maximal enzyme rate
SDH
sorbitol dehydrogenase
voc
volatile organic compound
SE
standard error
w
watt
SMR
standardized mortality ratio
WBC
white blood cell
SOA
secondary organic aerosol
WOS
Web of Science
SVEP
short-latency visual evoked potential
I2
chi-squared
SWD
spike-wave discharge
TLV
threshold limit value
TMB
trimethylbenzene
This document is a draft for review purposes only and does not constitute Agency policy.
xii 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
Supplem en tal Information—Trim ethylbenzenes
APPENDIX A. RESPONSE TO EXTERNAL PEER
REVIEW COMMENTS PROVIDED BY THE CHEMICAL
ASSESSMENT ADVISORY COMMITTEE OF THE
SCIENCE ADVISORY BOARD
The Toxicological Review of Trimethylbenzenes (TMBs) has undergone a formal external
peer review by the Chemical Assessment Advisory Committee (CAAC) of the U.S. Environmental
Protection Agency (EPA) Science Advisory Board (SAB). An external peer-review workshop was
held June 14-16, 2014. The CAAC Panel was tasked with providing written answers to general
questions on the overall assessment and on chemical-specific questions in areas of scientific
controversy or uncertainty; these comments and answers were then provided to EPA in the form of
a Peer Review Report The following sections present the CAAC Panel's comments on the external
peer review draft of the Toxicological Review of Trimethylbenzenes; in most cases, the CAAC Panel
comments were paraphrased for presentation, but in some situations, the Appendix uses direct
language from the CAAC. Each CAAC Panel comment is followed by an EPA response reflecting
consideration of the comment and revisions made to the Toxicological Review in light of that
comment. Given the overall nature of the CAAC comments, based on EPA policy guidance, no
additional review by the CAAC is warranted.
General Charge Questions
SAB Comment 1: In providing comments on the first four charge questions related to how
the Agency has implemented recommendations provided by the National Research Council (NRC),
the SAB noted that the Agency was implementing a phased approach to address the NRC
recommendations for several assessments that were under review. The SAB recognized that the
Agency was implementing the first phase of the Agency's efforts to enhance the Integrated Risk
Information System (IRIS) process in the TMB draft assessment and the SAB acknowledged the
improvement in the new format for IRIS assessments and commended the Agency for its progress
in addressing the NRC recommendations. The SAB noted that it used the peer review of the
Toxicological Review of Trimethylbenzenes as a case study to provide advice and comments on
improving IRIS toxicological assessments by further addressing the NRC recommendations.
Specific comments on developing the Preamble and Executive Summary for future assessments, as
well as the TMB assessment, were provided in the SAB's report The SAB noted that it anticipates
that after several IRIS reviews are completed, the CAAC will compare the reviews to provide the
Agency, through the Chartered SAB, with advice and comments on the Agency's progress to
enhance IRIS assessments.
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
37
38
Supplem en tal Information—Trim ethylbenzenes
EPA Response 1: The SAB noted that it uses the review of the draft TMB assessment as a
case study to provide recommendations on strategies to implement the NRC's recommendations
regarding improvements to the IRIS document structure. Although SAB noted that these
recommendations are intended for future assessments, EPA has implemented some
recommendations, where possible, in order to facilitate the rapid improvement of IRIS products.
Other recommendations, such as full implementation of systematic review methods, are not
implemented in order to prevent undue delays in posting the final IRIS TMB assessment. In
comments below, it is noted that SAB acknowledges and supports this rationale for the phased
implementation of the NRC recommendations.
General Charge Question 1: NRC (2011) indicated that the introductory section of IRIS
assessments needed to be expanded to describe more fully the methods of the assessment. NRC stated
that they were "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 [toxicity values]." Please comment on whether the new Preamble provides a
clear and concise description of the guidance and methods that EPA uses in developing IRIS
assessments.
SAB Comment GC.1-1: The SAB noted that"[t]o a substantial degree, the Preamble as
currently written provides a concise and clear description of the process that is followed, its steps,
the places in the process where decisions or judgments are made, the guidance that applies to
making those judgments (with explanation of the main considerations and available choices), and
the process by which the results of each step feed into the next" The SAB further noted that it
presumed that the Preamble "will change from one assessment to the next to reflect newly adopted
procedures" and recommended that the current assessment note where it has not fully
implemented procedures outlined in the Preamble and planned for subsequent assessments. The
SAB also recommended that Section 2 on the IRIS Process include further discussion, as part of the
problem formulation step, on issues needing to be addressed in assessments, including how these
issues will be addressed with the available data and how uncertainties and alternative
interpretations will be considered. The SAB also recommended that the EPA make clear that the
Preamble itself is not guidance and ensure that the Preamble refer users to the appropriate
guidance documents taking care to not imply that it supersedes policy existing guidance. The SAB
helpfully pointed out a number of instances where it might be construed that the Preamble
contradict current guidance. The SAB also noted that Section 5.5 could be confusing as to what
guidelines for assessing causality were used in the TMBs assessment and advised that discussing
the intent of weight-of-evidence descriptors was more advisable.
EPA Response GC.1-1: In the time since the SAB External Review meeting for the
Trimethylbenzenes Toxicological Review, the IRIS program has substantially revised the Preamble
based on a number of considerations, including: 1) experience with implementing the new
document structure and systematic review procedures after the trimethylbenzenes assessment was
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
submitted for SAB review in 2013; 2) recommendations from SAB reports on other draft
assessments (such as ammonia) and; 3) comments from EPA's program and regional offices, other
federal agencies and the Executive Office of the President, and the public.
The revised Preamble reflects recommendations for a shorter Preamble, and some
information previously in the Preamble is now discussed in the Toxicological Review (e.g.,
literature searching, screening, and study evaluation) or in the upcoming IRIS Handbook of
Operating Procedures for Systematic Review being developed by the IRIS Program. The Preamble
begins with a new statement that it summarizes general principles and systematic review
procedures, and specifically states in Section 1 that the "... Preamble summarizes and does not
change IRIS operating procedures or EPA guidance". Consistent with SAB recommendations, new
text was also added to the Preface to describe where approaches in the trimethlybenzenes
assessment differ from those outlined in the Preamble. Additionally, Section 2 of the Preamble has
been rewritten to elaborate that through the Problem Formulation step of the IRIS Process, EPA
identifies the science questions that will be addressed in an IRIS assessment and that Problem
Formulation includes input from the scientific community and public. Problem formulation further
includes multiple systematic reviews of the literature. Section 2 in the updated Preamble also
delineates that protocols will be established and used by EPA to conduct its literature searches,
considerations for evaluating study quality, and extracting data. It is through the Problem
Formulation step and application of protocols that EPA will determine how to address the science
issues covered by the assessment and how to appropriately consider any uncertainties and
plausible alternative interpretations. As stated above, the Preamble now clearly states that it does
not change existing EPA guidance and that IRIS assessments follow existing EPA guidance
documents. The shortened format of the Preamble no longer includes specific citations to guidance
documents, but rather directs users to IRIS's guidance website. With a shorter, refocused Preamble,
specific instances were it seemed that the Preamble superseded existing guidance have been
removed. Section 5 of the revised Preamble (Integrating the Evidence of Causation for Each Health
Outcome) has been rewritten to report that EPA uses standardized hazard descriptors for cancer
endpoints and that the "objective is to promote clarity and consistency of conclusions across
assessments." EPA still describes briefly what level of evidence is generally required for
determination of the individual descriptors. The Preamble further reports that IRIS is currently
discussing the potential for development of a causality framework for non-cancer effects.
General Charge Question 2: NRC (20111 provided comments on ways to improve the
presentation of steps used to generate IRIS assessments and indicated key outcomes at each step,
including systematic review of evidence, hazard identification, and dose-response assessment. Please
comment on the new IRIS document structure and whether it will increase the ability for assessment to
be more clear, concise and easy to follow.
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
SAB Comment GC.2-1: The SAB recommended that the revised structure for IRIS
assessments should allow for three different modes of reading the document: (1) quickly to get the
main qualitative and quantitative conclusions; (2) more thoroughly, but still rapidly, to get a
complete idea of the types of data and toxicity information that were considered, the main features
and issues involved in the interpretation of those data, and the choices that were made and their
rationale; and (3) in detail in order to find the particulars of individual study features, data, and
analyses. The SAB found that, in general, the structure of the TMB assessment has markedly
improved compared to previous IRIS assessments, and the current document structure facilitates
all three modes of recommended reading.
EPA Response GC.2-1: No response necessary.
Consistent Presentation of the Studies Considered
SAB Comment GC.2-2: The SAB recommended that each study used in the assessment
should be in a consistently formatted table. The table should be in an appropriate appendix and
present the study-specific considerations that bear on evaluation of study quality and pertinence,
including shortcomings and assumptions that are needed to interpret the study's outcomes.
Consistency of format is important within each document, but it would also be a useful goal to
achieve from one IRIS assessment to another.
EPA Response GC.2-2: Currently, a study summary table is included for each study cited in
the assessment. These tables are formatted consistently to the extent possible given the varying
type, amount, and detail of information provided in the individual studies. Information is provided
at the head of each table regarding additional study details important to interpretation of study
findings.
As EPA moves forward with implementing systematic review methodology, the SAB's
recommendations to include study-specific information such as evaluations of study quality and
strengths and weaknesses will be more fully implemented. In the current assessment, the study
summary tables provide some information that can be used to judge the overall quality of the study
(including numbers of animals, dosing schemes, etc.).
SAB Comment GC.2-3: The SAB suggested that it would be useful for each study to have a
short overview section (also in its appendix listing, not repeating tabulated details) of the nature of
the study, its examined endpoints, and relevant findings. The goal of the overview is to provide
context for the tabulated details, so that the details need not be read in full to gain an idea of the
general nature of the study and its importance to the assessment as a whole. This overview should
not discuss interpretations.
EPA Response GC.2-3: This information is provided at the head of each study summary
table included in Appendix C. Specifically, general information about what effects were observed
and at what dose levels those effects occurred are provided in the "Additional study details" section
in each study summary table provided in Appendix C. For example, for Gralewicz and Wiaderna
£20011 Table C-24, it is noted in a bullet that "1,2,3-TMB-, 1,2,4-TMB-, and 1,3,5-TMB-exposed rats
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
38
Supplem en tal Information—Trim ethylbenzenes
showed alterations in performance in spontaneous locomotor activity, passive avoidance learning,
and paw-lick latencies."
SAB Comment GC.2-4: SAB recommended that as IRIS makes enhancements to the
systematic review process, the overriding issue is transparency regarding study selection criteria.
Studies that support a hypothesized human hazard should be included, but studies that are
contrary to these hypotheses should also be included as they result in alternative, scientifically
supportable conclusions regarding human risk.
EPA Response GC.2-4: The revised Preamble includes discussion of criteria for study
selection. In the TMB assessment, studies most relevant to hazard identification and dose-response
analyses have been included in the main body of the text, including those data that may seem
inconsistent. For example, while an argument of sufficient similarity is used in the assessment to
support adopting reference concentrations (RfCs) derived for one isomer as the RfC for another
isomer when lacking sufficient-isomer specific data, instances where the toxicities or toxicokinetics
appear to differ between isomers are clearly discussed. Additionally, information contained in
appendices in the draft TMB assessment regarding the C9 fraction studies, including differences
between these studies and isomer-specific studies, have been included in the main body of the
assessment consistent with the recommendation of the SAB.
Describing the Literature Search
SAB Comment GC.2-5: SAB commented that the Literature Search Strategy section is brief
and focuses only on identification of pertinent studies from the literature. The SAB was concerned
that the general description of the process and the specific implementation for TMBs may be too
exclusive, missing potentially informative ancillary studies that could help in interpretation or
evaluation of those studies strictly observing toxicity outcomes of the TMBs alone in controlled
settings. SAB recommended a more inclusive literature search in which evidence from related
compounds are incorporated in order to provide context to evidence gleaned from the chemicals
under assessment (i.e., TMBs).
EPA Response GC.2-5: The "primary" (initial) TMB literature search has been re-tagged in
the Health and Environmental Research Online (HERO) database such that all of the identified
studies are tagged more thoroughly, including those references determined to not be relevant to
the assessment. For example, there are now exclusion tags that identify which studies were
excluded based on being published in non-relevant journals (e.g., chemical engineering journals)
and which studies were excluded based on title and abstract screenings. The "primary" (initial)
literature search has also been updated to November, 2015 and the results of this literature search
update are reported in a similar fashion.
A secondary, targeted literature search for information pertaining to the effects and
properties of similar chemicals has been conducted, and the results of this literature search are also
reported. Briefly, the literature search was limited to integrated reviews of the toxicological effects
of related compounds (see SAB Comment GC.2-6 below for further details).
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
SAB Comment GC.2-6: SAB recommended that the primary literature search should be
comprehensive and subjected to an orderly process of systematic review, and further commented
that the secondary search is for literature that is useful to provide context, in terms of what might
be expected given the knowledge of other chemicals and of the potential pathways of toxic action.
SAB recommended that the secondary search need not be comprehensive and could include
reviews as well as original experimental studies in order to provide information that can potentially
fill data gaps that exist in the primary TMB literature.
EPA Response GC.2-6: In response to the SAB recommendation, a secondary literature
search was conducted to identify studies on related compounds focused primarily on review
articles in order to assess a large body of literature for the pertinent pieces of information that
could serve to fill data gaps in the primary TMB literature. The related chemicals included in this
targeted, secondary literature search were toluene, xylene, styrene, and ethylbenzene; specific
toxicity endpoints included in the secondary literature search included neurotoxicity,
developmental neurotoxicity, respiratory toxicity, developmental toxicity, and hematotoxicity. The
literature search was set up as: (at least one chemical) + (at least one toxicity endpoint) + (review
article). The secondary literature search resulted in approximately 70 review articles that were
manually screened for relevance to provide context for the TMB assessment, and to identify
additional relevant primary literature. The final TMB assessment includes both relevant review
articles and new primary literature identified through the secondary literature search. Information
from the secondary, targeted literature search were used to fill in gaps in the existing TMB
database, and to help inform decisions in setting the value of the database uncertainty factor.
Describing the Hazard Identification Step
SAB Comment GC.2-7: The SAB recommended that the individual endpoint sections of the
Hazard Identification section have some discussion about interpretation across studies and
evaluations of bearing and relevance, though further discussion of interpretation rationales and
consideration of alternatives would be beneficial. The SAB made this recommendation in the
context of the larger process of a systematic review of the literature, stating that it is the middle
section of systematic review—after the studies are chosen but before the interpretation of their
overall bearing gets considered—that does not have a clear home in the current document
structure. The SAB recognized that the implementation of systematic review methods have not
been fully implemented and recommended that the Agency further develop its approach for
systematic review so that the ways for abstracting data, judging study quality, documenting factors
bearing on interpretation and its limits, and considering the impact of related studies have discrete
locations in the updated IRIS document structure.
EPA Response GC.2-7: EPA agrees with the SAB's comments regarding the evolving
structure of the systematic review of the literature. It is EPA's intention that, moving forward, the
NRC recommendations will be fully implemented in future assessments and that specific comments
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
received from SAB on current assessments will be invaluable in the implementation of those
recommendations.
In the final TMB assessment, EPA has partially addressed this SAB comment by
strengthening the discussion of the interpretation of studies, including the consideration of
alternative explanations or conflicting evidence, in the synthesis sections at the end of each organ
section. For example, in the write-up for the neurotoxic effects observed in animal toxicology
studies, full discussions of the Douglas etal. (1993) neurotoxicity study have been included.
Instances where the results of the Douglas etal. (1993) C9 study and individual isomer studies
differ in observed effects have been exhaustively discussed, and possible interpretations of those
differences are included in the text. This discussion of differing results and possible
interpretational issues across studies is also included in other health effects sections, and in
Sections 1.2.7 (Similarities among TMB Isomers Regarding Observed Inhalation and Oral Toxicity)
and 1.3.1 (Weight of Evidence for Effects Other than Cancer).
SAB Comment GC.2-8: The SAB noted that Preamble has a section (Section 5) on evaluation
of causality, which depends on the existence of such a documented review and evaluation process,
but that the TMB assessment has no particular place where the Preamble's named considerations—
strength, consistency, specificity, temporal relationship, biologic plausibility, coherence, natural
experiments, and analogy—are systematically considered or documented.
EPA Response GC.2-8: Although the Preamble lays out the precepts by which human or
animal evidence can be evaluated systematically for causality, a systematic causality framework has
not been fully implemented in this assessment. However, the evidence was more clearly
characterized with respect to the various considerations affecting causality determinations (e.g.,
strength, consistency, specificity, temporal relationship, biologic plausibility, coherence, natural
experiments, and analogy). For example, in evaluating the evidence in the neurotoxicity database,
the TMB assessment notes that "[neurotoxicity is strongly and consistently (emphasis added)
associated with exposure to TMBs in multiple studies, and these associations are coherent in
human populations exposed to mixtures containing TMBs and in laboratory animals exposed to
individual TMB isomers." Additionally, the TMB assessment notes that "TMBs are neurotoxic
following inhalation or oral exposure, based on strong and consistent effects in experimental
animals that are coherent with observations in exposed humans; biological plausibility based
primarily on similarities to findings from related chemicals; evidence of effects that worsen with
increasing duration of exposure; delayed-onset and/or latent neurological effects in animals several
weeks following exposure; and observed exposure-response relationships in animals tested
immediately after exposure." The considerations that relate to evaluation of causality are also
applied to the other health effect domains throughout the document
SAB Comment GC.2-9: The SAB recommended adding a brief summary of the main features
of the assessment—in this case, pharmacokinetics and metabolism—before the section on Hazard
Identification. The SAB noted that the aim of this section would not be to replace the fuller
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
treatment of these issues in an appendix, but rather to set the context for the interpretation of
studies bearing on hazard, and the main presentation of pharmacokinetic details should continue to
reside in an appendix. The SAB suggested that the main text's section would note such things as
extent of absorption, rapidity of elimination, main metabolic processes, main means of clearance
(and what part of that is by metabolism), indications of whether metabolic saturation or enzyme
induction might play a relevant role in toxicity studies, and any notable unusual differences
between experimental animals and humans.
EPA Response GC.2-9: Previously, all information on the toxicokinetic properties of the
TMB isomers was located in Appendix B of the External Peer Review draft Supplemental
Information document Given CAAC's recommendation, this section has been moved to Section
1.1.1 of the main body of the final assessment. Section 1.1.2 was added to provide a brief overview
of the available physiologically based pharmacokinetic (PBPK) models for TMB isomers.
SAB Comment GC.2-10: The SAB noted that the current IRIS document structure in which
the Hazard Identification section is separated into assessments of each endpoint, with relevant data
for that endpoint being reviewed within the section is a great improvement over the past practice
of summarizing study by study. SAB was also impressed that the endpoint-by-endpoint analysis
permits the examination of consistency and sufficiency of data to draw hazard conclusions about
each effect. The SAB commented that there were possible overarching ties among endpoints that
would help in evaluation of the hazard characterization of each that should be discussed in an
appropriate place. SAB further recommended that it would be useful to include considerations that
might indicate a study as the critical study.
EPA Response GC.2-10: A short discussion of commonalities between endpoints regarding
possible modes of action has been added to Section 1.3.1. Discussions of important considerations
that might help indicate a study a potential critical study, especially extensive discussions on study
design and its effect on the observation of particular endpoints, have been added throughout
Section 1.
SAB Comment GC. 2-11: The SAB commented that the tabulation of studies into Evidence
Tables is useful, noting that the inclusion of dose levels and dose-specific responses are important
details to provide. The SAB also noted that providing hyperlinks to the study summary tables in the
Supplemental Information document makes finding relevant data easier, and that the Exposure-
Response arrays provide a valuable overview of the data.
EPA Response GC.2-11: No response necessary.
Describing the Dose-Response Steps
SAB Comment GC.2-12: The SAB noted that the tabulation of points of departure (PODs),
human equivalent concentrations (HECs), and applied uncertainty factors (UFs) is useful and allows
for the comparison of endpoints and the distinction between a low POD with few UFs and a high
POD and many UFs.
EPA Response GC.2-12: No response necessary.
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
37
38
Supplem en tal Information—Trim ethylbenzenes
SAB Comment GC.2-13: SAB noted that the inclusion of discussions of consistencies and
inconsistencies among data, relevance of studies for human risk evaluation, knowledge of mode of
action (even if it must say that little is known), and alternative interpretations of the available data
on potential causation for each endpoint represents an important advance in the Hazard
Identification sections. SAB further noted concern that these interpretation passages are too
concise and recommended that a consistent way be developed to document these arguments
without unduly distracting from the main Hazard Identification discussions.
EPA Response GC.2-13: Discussions in the interpretations of the organ-specific TMB-
induced toxicities have been augmented where appropriate to highlight commonalities across
effects. As IRIS continues to implement NRC- and SAB-recommended changes to the documents, a
more consistent way to present summaries and interpretations will be developed.
Presenting Outcomes
SAB Comment GC.2-14: SAB noted that the both the Hazard Identification and Dose-
Response Analysis sections simply dive in to the first endpoint or analysis to be considered, and
then have separate sections on each. SAB commented that there is little overview to prepare a
reader for what is coming or to point to the parts that are critical versus those that are there for
completeness. In general, to help enable a reader to grasp the main lines of argument and only go
into detail when needed, the SAB recommended that both the Hazard Identification and the Dose-
Response Analysis sections have an initial paragraph setting out the main issues that will be
considered and indicating which considerations (to be developed in the subsequent text) are the
most notable for the larger assessment process. SAB also recommended a parallel paragraph at the
end of each of these chapters to summarize what its contents have provided to the larger
assessment process. The aim of these paragraphs would be to make it possible to not only read the
document in more detail than provided in the Executive Summary, but also still quickly see the
deeper structure of the report and where to focus for more information on particular aspects.
EPA Response GC.2-14: An introductory paragraph has been added to the beginning of the
Hazard Identification section. This paragraph summarizes the broad scope and purpose of the
Hazard Identification section and analysis/interpretations therein, including highlighting particular
sections most important for the assessment conclusions (i.e., the neurotoxicity section, similarities
in toxicity between isomers, and the differing results observed in the C9 studies). No new
concluding paragraph was added to the Hazard Identification section as such a paragraph would be
largely duplicative of Section 1.3 (Summary and Evaluation). An introductory paragraph has also
been added to the Dose-Response Analysis section, briefly highlighting what types of benchmark
dose (BMD), PBPK, and/or default dosimetric adjustment analyses were performed and the major
conclusions of the dose-response section.
General Charge Question 3: NRC f2011) states that "all critical studies need to be
thoroughly evaluated with standardized approaches that are clearly formulated" and that
"strengthenedmore integrative, and more transparent discussions of weight of evidence are needed."
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
NRC also indicated that the changes suggested would involve a multiyear process. Please comment on
EPA's success thus far in implementing these recommendations.
SAB Comment GC. 3-1: The SAB found that, in general, a great deal of progress has been
made in restructuring the document to focus the main body on documenting and explaining the
interpretations, choices, and analyses, and relegating the supporting information to appendices.
However, the SAB also noted that the process of systematic review still needs development.
Documentation of the process of identifying literature has progressed, but further development is
needed in establishing standard practices for abstracting relevant data, evaluating study quality,
strengths, and shortcomings, and integrating evidence across studies. In making this
recommendation, the SAB recognized that an important challenge facing the Agency is that
assessments must go ahead even as this further development proceeds and before all aspects are
complete. Ultimately, the SAB recommended that a good principle to follow in conducting
assessments during the process of revision is to consider the reasons behind the recommendations
for change, and to make efforts to address the issues and to explain how the chosen approaches
seek to reflect the NRC recommendations, although the methods may not yet be fully developed and
agreed upon.
EPA Response GC.3-1: The SAB acknowledged and agreed with EPA's phased
implementation of the NRC recommendations for improving the IRIS process. As such, EPA is fully
implementing systematic review methods (e.g., including methods to systematically judge study
quality and the consistent application of study exclusion/inclusion criteria) in new IRIS
assessments that are in the Problem Formulation or Draft Development steps. Assessments that
are further along in the IRIS process, such as the TMB assessment, are incorporating elements of
systematic review methods, as well as other document improvements such as streamlining the
document structure and increased incorporation of tables, figures, and exposure-response arrays
for the efficient presentation of data, in order to keep the program at large on track.
General Charge Question 4: EPA solicited public comments on the draft IRIS assessment
oBof trimethylbenzenes [May 2012] and has revised the assessment to respond to the scientific issues
raised in the comments. A summary of the public comments and EPA's responses are provided in
Appendix F of the Supplemental Information to the Toxicological Review of Trimethylbenzenes. Are
there scientific issues that were raised by the public as described in Appendix F that may not have been
adequately addressed by EPA?
SAB Comment GC.4-1: While the SAB felt that Appendix F (External Peer Review draft)
addressed issues raised in public comments in a transparent manner, the panel was divided on the
adequacy and dispositions that were made as presented in the appendix. Most importantly, the
SAB panel expressed a number of opinions on the role that the C9 fraction studies should play in the
assessment and whether or not the possible reversibility of the critical effect of decreased pain
sensitivity was discussed adequately.
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
37
38
Supplem en tal Information—Trim ethylbenzenes
EPA Response GC.4-1: The Agency appreciates that the SAB found that Appendix F in the
External Peer Review draft assessment was generally responsive to public comments. Regarding
the adequacy and disposition of comments regarding the C9 fraction studies, in the final TMB
assessment, the C9 studies are covered more extensively below in EPA Responses C.l (Synthesis of
Evidence)-6 and -8. The issues surrounding the possible reversibility of decreased pain sensitivity
are covered below in EPA Responses E.l-5 and E.4-4; briefly, it was concluded that when the entire
pain sensitivity database was taken into consideration (short-term TMB and subchronic TMB or C9
studies), the data clearly indicated that decreased pain sensitivity was not a transient effect, and
that exposure TMB isomers resulted in persistent alterations in an organism's ability to correctly
process painful stimuli.
Chemical-Specific Charge Questions
Charge Question A.1: The major conclusions of the assessment pertaining to the hazard
identification and dose-response analysis have been summarized in the Executive Summary. Please
comment on whether the conclusions have been clearly and sufficiently described for purposes of
condensing the Toxicological Review information into a concise summary.
SAB Comment A.l-1: While the SAB commented that the Executive Summary did an
adequate job at condensing a large amount of information presented in the TMB assessment, the
panel provided a number of recommendations for improving the presentation and flow of
information included. The SAB recommended that the Executive Summary be shortened to
emphasize the major conclusions of the assessment. Specifically, the panel recommended removing
all citations and combining the duplicative sections on "Confidence" into a single succinct section.
The SAB also recommended that information not be duplicated in tables and the text of the
Executive Summary. Finally, the SAB noted that much of Section 15 of the Executive Summary
seemed speculative and should not be included.
EPA Response A.l-2: All recommendations made regarding the Executive Summary have
been incorporated. The Executive Summary has been shortened to emphasize major conclusions of
the assessments: the available information in the inhalation and oral toxicity databases and the
derivation of the RfC and reference dose (RfD). Citations have been removed. The structure of the
executive summary has changed to consolidate discussions of particular issues (confidence, etc.)
into one section covering all isomers; this follows the restructuring of the Dose-Response Analysis
section in the main body of the assessment. All of the discussion regarding Susceptible Populations
and Lifestages has been removed from the Executive Summary other than to state "No chemical-
specific data that would allow for the identification of populations or lifestages with increased
susceptibility to TMB exposure exist"
Charge Question B.l: The process for identifying and selecting pertinent studies for
consideration in developing the assessment is detailed in the Literature Search Strategy/Study
Selection section. Please comment on whether the literature search approach, screening, evaluation,
and selection of studies for inclusion in the assessment are clearly described and supported. Please
This document is a draft for review purposes only and does not constitute Agency policy.
A-11 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
Supplem en tal Information—Trim ethylbenzenes
identify any additional peer-reviewed studies from the primary literature that should be considered in
the assessment of noncancer and cancer health effects of 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB.
SAB Comment B.1-1: The SAB found that the search strategy was clearly articulated and
that the databases and search terms were clearly defined. However, the SAB noted some concerns
that the way that studies were selected for use in the assessment was not transparent. Specifically,
the SAB noted that while it was clear which papers were included in the assessment, there were no
means of determining which papers were excluded from the assessment and for what reasons. The
SAB recommended that the EPA provide citations for all studies identified via the literature search
and group them according to reasons why they were excluded from consideration.
EPA Response B.l-1: As noted above in EPA Response GC.2-5, EPA has provided all of the
identified studies in the HERO database, and has re-tagged all of the references such that all of the
identified studies are tagged more thoroughly, including those references determined to not be
relevant to the assessment. For example, there are now exclusion tags that identify which studies
were excluded based on being published in non-relevant journals (e.g., chemical engineering
journals) and which studies were excluded based on title and abstract screenings. The "primary"
(initial) literature search has also been updated to November 2015 and the results of this literature
search update are reported in a similar fashion.
SAB Comment B. 1-2: The SAB further commented that in the External Peer Review Draft,
65 references were excluded "based upon manual review of papers/abstracts," but these papers
were not individually identified. The SAB also commented that excluding papers because they were
not available in English is not a valid reason for exclusion. Lastly, SAB noted that reporting some
papers as being excluded based on being in vitro reports, but including other in vitro reports
elsewhere in the document, was inconsistent
EPA Response B.l-2: The entire "primary" (initial) literature search has been re-tagged in
the HERO database. As such, all studies found via the literature search are now included in the
database, and users can now determine which individual studies were excluded for which reasons
at what step in the process (i.e., some references were excluded based on which journals they were
published in, and some were excluded based on manual screening of titles/abstracts based on
whether they were exposure studies, in nonrelevant in vitro systems [e.g., bacterial systems], etc.).
A number of papers were previously excluded based on being published in foreign language
journals; these foreign language journal articles were re-screened based on their title and/or
abstract If it was judged that any non-English reference should be excluded on content or subject,
it was binned in the appropriate exclusion bin. If a non-English reference was judged to possibly be
relevant to the assessment, it was placed in the "Considered" bin and reviewed further to determine
whether it should be translated into English. Ultimately, no non-English references were judged to
be critical to the needs of the assessment and correspondingly no references were translated into
English. In re-tagging all of the references in the TMB database, any decision to exclude in vitro
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
Supplem en tal Information—Trim ethylbenzenes
studies has been tagged such that it is clear that the study was excluded because it was unrelated
and uninformative to the purposes of the TMB assessment, not for simply being an in vitro study.
SAB Comment B. 1-3: The SAB noted that the search strategy did not mention compounds
structurally related to TMB isomers, including xylenes or ethylbenzenes, and that this may have
resulted in important studies being excluded from the assessment. The SAB recommended a
number of human occupational studies investigating the effects of exposure to complex mixtures of
volatile organic compounds (VOCs) that should be added to the assessment in order to strengthen
its conclusions:
1. Chapter 8 on TMBs fNRC. 20131
2. Health hazards of solvents exposure among workers in paint industry fEl Hamid Hassan et
al.. 20131
3. Xylene-induced auditory dysfunction in humans fFuente etal.. 20131
4. Hearing loss associated with xylene exposure in a laboratory worker fFuente etal.. 20121
5. Visual dysfunction in workers exposed to a mixture of organic solvents f Gong etal.. 20031
6. Ototoxicity effects of low exposure to solvent mixture among paint manufacturing workers
fluarez-Perez etal.. 20141
7. Short latency visual evoked potentials (SLVEPs) in occupational exposure to organic
solvents f Pratt etal.. 20001
8. Auditory brainstem response in gas station attendants fOuevedo etal.. 20121
EPA Response B.l-3: The studies recommended by the SAB for inclusion have been added
to the TMB assessment where appropriate. However, it should be noted that these studies either
involve human exposures to complex organic solvent mixtures or related alkylbenzene compounds.
Therefore, while these studies provide further qualitative support that exposure to TMBs and/or
related compounds as part of complex solvent mixtures result in adverse health effects, caveats
regarding their interpretations still apply. Namely, it's not possible to attribute the observed effects
completely to one specific component of the mixture, and there is some uncertainty that related
alkylbenzenes would elicit the exact same health effects as TMBs. Other shortcomings of the human
studies involved imprecision in effect estimates due to low statistical power and lack of quantitative
exposure assessment. As discussed above in EPA Responses GC.2-5 and GC.2-6, EPA also conducted
a targeted secondary literature search of review papers on related compounds in order to identify
additional data that would potentially strengthen the conclusions of the assessment
SAB Comment B. 1-4: The SAB recommended that a summary table be included for each
human health effect that reports study design, inclusion/exclusion criteria, results, etc. in
Appendix B.
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
30
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
EPA Response B.l-4: Instead of including a summary table covering all of the human
studies included in the assessment, EPA replaced all of the individual human study summary tables
with Table C-16 that provides all of the pertinent study details requested by SAB, as well as study
details previously reported in the individual tables.
Charge Question C.l (Synthesis of Evidence): A synthesis of the evidence for
trimethylbenzene toxicity is provided in Chapter 1, Hazard Identification. Please comment on whether
the available data have been clearly and appropriately synthesized for each toxicological effect.
Please comment on whether the weight of evidence for hazard identification has been clearly
described and scientifically supported.
SAB Comment C.l fSynthesis of Evidencel-1: The SAB noted that the synthesis of evidence
for the three TMB isomers was efficiently divided up into sections corresponding to the various
target organs or forms of toxicity, and then by human versus animal studies and route of exposure
when possible. The SAB noted that the studies chosen for review were clearly described and that
the evidence tables and exposure-response arrays augmented the text effectively. The SAB
recommended that an introductory paragraph describing the section layout, including the summary
tables for each endpoint, would improve readability.
EPA Response C.l fSvnthesis of Evidencel-1: As noted above in EPA Response GC.2-14, an
introductory paragraph has been added to the beginning of the Hazard Identification section. This
paragraph briefly outlines the structure of the Hazard Identification section and what types of data
are presented.
SAB Comment C.l (Synthesis of Evidence)-2: The SAB expressed concern that the
discussion of individual endpoints was flawed by questionable statistical statements or inferences.
Several instances in the document were provided as evidence of these flawed statistical statements.
For example, the TMB document notes, regarding decreased performance on the rotarod, that "This
impaired function [i.e., failures on the rotarod apparatus] was still evident at 2 weeks post-
exposure and, while not statistically significant for 1,2,4-TMB, may indicate long-lasting
neuromuscular effects of subchronic exposures to 1,2,4-TMB and 1,2,3-TMB." The SAB
recommended that descriptions of results more closely adhere to the rule that statistical
significance provides the criterion of whether an effect has occurred.
EPA Response C.l (Synthesis of Evidencel-2: It is EPA's practice that evaluation of evidence
should first consider biological significance to the extent possible. The purpose of this evaluation is
to understand the extent to which individuals could demonstrate some adverse effect in response
to exposure. It is important to note that at the population level, even small changes in the average
of a response parameter can result in an increase in the number of people in the "abnormal" or
"impaired" range for the particular endpoint Thus, a relatively small difference can be considered
biologically significant When biological significance is uncertain or understood less clearly (e.g., no
suitable normal range), statistical significance testing has been used to augment this evaluation.
When suitable, well-designed studies are used, a pattern of statistically significant results for an
This document is a draft for review purposes only and does not constitute Agency policy.
A-14 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
39
40
41
42
43
Supplem en tal Information—Trim ethylbenzenes
effect, or related effects, across such studies generally increases the confidence that the effect is
associated with the exposure. It is important to note, however, that statistical significance testing,
while a useful tool for the systematic evaluation of data, has limitations, that, when overlooked, can
lead to flawed conclusions. Specifically, lack of statistical significance should not automatically be
interpreted as evidence of no effect For example, if an exposure at a particular level leads to a
measurable effect, studies with low statistical power are unlikely to produce statistically significant
results. It is important to examine patterns in results across all studies that report data for the
same endpoint, taking into account relative exposure ranges and variability of effects. The final
TMB assessment has been revised such that discussions of observed health effects appropriately
note cases of both statistical and biological significance, taking particular care to note trends across
studies and isomers. Using the example above (failures on the rotarod apparatus), EPA notes that:
Significant decreases in rotarod performance were observed at 1,230 mg/m3
1,2,4-TMB (40% response) and >493 mg/m31,2,3-TMB (50-70% response) when
tested immediately after exposure for 13 weeks fKorsak and Rvdzvriski. 19961: a
clear exposure duration-dependency for this effect was observed, with less robust,
but statistically significant, decreases in performance also reported at 1,230 mg/m3
after 4 (40 and 30% response) or 8 (60 and 40% response) weeks of exposure to
1.2.3-TMB or 1,2,4-TMB, respectively. This impaired function was still evident at
2 weeks post-exposure, indicating a persistence of this effect Specifically, failures in
70 and 40% of animals after 13 weeks of exposure to 1,230 mg/m31,2,3-TMB and
1.2.4-TMB, respectively (compared to 0% of animals in control groups at any time),
were 50 and 30% at 2 weeks post-exposure, although 30% failures at 15 weeks for
1,2,4-TMB was no longer significantly different from controls (note: statistical
comparisons did not appear to include a repeated measures component and
comparisons to the 13-week time-point were not performed). The observation of
substantial decrements in rotarod performance is interpreted as a biologically
relevant response in light of the lack of failures in controls and the similarities in
response magnitude across isomers.
It is important to note that this discussion of nonstatistically significant, but possibly
biologically significant, decreases in rotarod performance was included in the context of other
statistically significant decrements of neuromuscular performance. All discussions of biologically
significant, but not statistically significant, effects are included in that context. In other words,
when nonstatistically significant effects are included in the discussion, they are used to compare
results across studies and isomers in order to provide a fuller account of the pattern of TMB-
induced toxicity.
SAB Comment C.l (Synthesis of Evidence)-3: The SAB recommended that the discussion of
respiratory effects should be strengthened by further consideration of the relevance to humans of
the effects observed in the high-dose animal studies. SAB noted that while it is clear that
respiratory effects are observed and are a relevant endpoint in humans, the distinction between the
high-dose animal effects and the human effects could have been made more clearly. The SAB also
recommended that the limitations of the human evidence for hematological and clinical chemistry
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
31
32
33
34
35
36
37
Supplem en tal Information—Trim ethylbenzenes
effect, based on the uncertainties in exposures (mixture components, doses) also be more clearly
described. The SAB noted that the TMB assessment clearly communicates the inadequacy of the
cancer toxicity database, including the minimal genotoxicity database.
EPA Response C.l (Synthesis of Evidencel-3: The discussions regarding the human
relevance of respiratory effects observed in high-dose animals and the limitations of the human
hematological evidence have been augmented in the final TMB assessment.
SAB Comment C.l (Synthesis of Evidencel-4: The SAB noted that the summary table (page
1-49, Table 1-7 in the External Peer Review draft; page 1-60, Table 1-8 in the current document)
was very helpful in understanding the points made with regard to the toxicological similarities
across TMB isomers, and recommended that a summary table or scheme regarding toxicokinetics
and metabolism would also be useful.
EPA Response C.l fSvnthesis of Evidencel-4: A summary table presenting the similarities in
toxicokinetics (absorption, distribution, metabolism, and excretion [ADME]) has been added to
Section 1.1.1 (Toxicokinetics ofTMB Isomers).
SAB Comment C.l (Synthesis of Evidence)-5: The SAB noted that the synthesis section that
provides weight-of-evidence determinations for the noncancer and cancer effects would be a good
place for a separate subsection that describes the major uncertainties and gaps present in the TMB
toxicological database.
EPA Response C.l fSvnthesis of Evidencel-5: A discussion of the major gaps and
uncertainties in the TMB toxicological database has been added to Section 1.3.1 (Weight of
Evidence for Effects Other than Cancer).
SAB Comment C.l (Synthesis of Evidence)-6: The SAB noted that the current synthesis
discussions are brief and do not weigh the value of evidence from related chemicals or from studies
done on the C9 fraction. The SAB further noted that structurally related alkylbenzenes such as
toluene, xylene, ethylbenzene, and styrene have similarities in neurotoxic effect and metabolic
disposition and that use of such information is clearly supported in the External Peer Review draft
version of the IRIS Preamble, Section 3.1 (lines 11-15) "[s]earches for information on mechanisms
of toxicity are inherently specialized and may include studies on other agents that act through
related mechanisms" and in Section 5.4, p. xxiii (lines 18-21), "Pertinent information may also
come from studies of metabolites or of compounds that are structurally similar or that act through
similar mechanisms." SAB therefore recommended that additional animal and human studies on
related aromatic solvents be considered in the qualitative and mechanistic interpretations ofTMB
toxicity. A list of such studies are included in SAB Comment 3 of Charge Question B.l. SAB
suggested that these data be used in multiple fashions, including the determination of whether
effects seen in TMB-only studies are consistent across related compounds and to inform potential
modes of action. The SAB noted that perfect consistency is not required, but major discrepancies
should be noted.
This document is a draft for review purposes only and does not constitute Agency policy.
A-16 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
Supplem en tal Information—Trim ethylbenzenes
EPA Response C.l (Synthesis of Evidencel-6: As noted above in EPA Response B.l-3, the
human studies investigating the health effects of related compounds or mixtures containing those
substances have been added to the TMB assessment where appropriate. Additionally, a targeted
literature search has been conducted to identify review articles on related compounds in order to
assess a large body of literature for the pertinent pieces of information that could serve to fill data
gaps in the primary TMB literature. Information gleaned from these review articles, and from
additional primary literature identified through the evaluation of the review articles, has been
included in the TMB assessment to make informed assumptions regarding TMB isomers' potential
mode of action and whether it can be reasonably anticipated that TMB isomers could cause certain
types of toxicity when isomer-specific data are missing (e.g., developmental neurotoxicity) (see EPA
Responses GC.2-5 and GC.2-6).
SAB Comment C.l fSynthesis of Evidencel-7: SAB noted that the data gaps for the TMB
database appear to be the lack of a developmental neurotoxicity study, the lack of a multi-
generational reproduction study, and the lack of a chronic noncancer (neurotoxicity) study. The
SAB recommended that the EPA could potentially utilize data from these analogous alkylbenzenes
to inform these data gaps and inform the selection of the value for the database UF.
EPA Response C.l fSvnthesis of Evidencel-7: EPA agrees with the SAB regarding the major
limitations in the TMB toxicity database. Information obtained through the secondary literature
search has been used to fill in data gaps in the TMB toxicological database, especially regarding the
potential mode of action of TMBs and the possibility that gestational exposure to TMB isomers
affect neurodevelopment Consideration of the fuller database, TMB isomer, related alkylbenzene,
and C9 fraction studies helped further support EPA's selection of a database uncertainty factor of 3
(see EPA Response E.4-5 below for complete details).
SAB Comment C.l fSvnthesis of Evidencel-8: SAB recommended that the discussion of the
existing C9 mixtures studies be brought into the main document describing their strengths and
weaknesses and relevance to the setting of RfDs/RfCs for individual TMB isomers, with particular
emphasis on whether they provide evidence to inform the aforementioned data gaps. For example,
regarding the developmental neurotoxicity data gap, the SAB noted that a Hungarian study
(Lehotzkv et al.. 1985) tested a C9 mixture containing TMBs (Aromatol) for developmental
neurotoxicity in rats. SAB reported that study had minimal reporting of results, simply stating that
there were no effects of Aromatol on dams or offspring at any time point in spite of the fact that the
high dose of Aromatol was 2,000 mg/m3, a dose that one would expect to have a neurotoxic effect in
dams during and after exposure, based upon results of other testing. SAB concluded that the lack of
any toxicity in dams or offspring, combined with the lack of reporting of any data (including
Aromatol treatment group neurological testing or Aromatol composition), and the fact that it was a
mixture and not a specific TMB, makes this study of limited utility for filling the developmental
neurotoxicity data gap. The SAB further noted that other issues relevant to the interpretation of the
C9 faction studies be discussed in the TMB assessment, including issues related to possible
This document is a draft for review purposes only and does not constitute Agency policy.
A-17 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
Supplem en tal Information—Trim ethylbenzenes
differences in metabolic clearance and distribution between TMB isomers and the C9 fraction. SAB
noted that considering this information is relevant for the evaluation of individual TMB isomers
and would help strengthen the Agency's decisions regarding the role of the C9 fraction in the
current assessment
EPA Response C.l fSvnthesis of Evidencel-8: Information on the C9 studies has been
brought into the main body of the text and discussed in the relevant subsections of the Hazard
Identification section. Discussions regarding the utility of the C9 studies for deriving reference
values has also been expanded in the Dose-Response Analysis section, with a particular focus on
whether these studies are suitable for derivation of reference values and whether or not
consideration of these studies and other studies on related compounds (i.e., toluene, etc.) help
inform decisions related to selecting the value for the database UF for TMB isomer-specific
reference values. Ultimately it was determined that the C9 fraction studies were not suitable for
derivation of reference values. However, consideration of the related alkylbenzenes data was
judged to be useful for supporting EPA's selection of the database uncertainty factor (see EPA
Response E.4-5 below for complete details).
Two other industry reports regarding the toxicity of mixtures containing the isomers (IBT
Labs. 1992: Chevron. 19851. however, were carefully considered but not included in the
Toxicological Review. There were multiple rationales for the exclusion of these studies. Of note,
these studies were not peer-reviewed and did not investigate the toxicity of individual TMB
isomers. EPA generally only includes studies that are peer-reviewed, and will seek out a peer-
review for a non-peer-reviewed reference if it appears to be critical for the needs of the assessment.
Neither of these references were deemed critical for the assessment The reasons for excluding the
Chevron study included deficiencies in reporting the composition of the test substance, the
conclusion that there was no need for a 1 generation reproduction C9 fraction study when a full
multigenerational reproduction C9 fraction study was already included in the database (Mckee et
al.. 19901. and that it was a dermal toxicity study. The main rationale for the exclusion of the IBT
Labs study was that it was a short-term inhalation study of a complex mixture containing TMB
isomers not likely to be critical to the needs of the assessment. As such, peer-review was not
sought for either of these references. Another 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 high oral doses of either 0.5 or 2.0 g/kg
1,2,4-TMB daily for 28 days. All rats in the high-dose group 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 in
this report (other than total mortality in the high-dose group), this report was not included in the
Toxicological Review.
Charge Question C.l (Summary and Evaluation): Does EPA's hazard assessment of
noncancer human health effects oftrimethylbenzenes clearly integrate the available scientific
This document is a draft for review purposes only and does not constitute Agency policy.
A-18 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
Supplem en tal Information—Trim ethylbenzenes
evidence (i.e., human, experimental animal and mechanistic evidence) to support the conclusions that
trimethylbenzenes pose potential hazards to the nervous system, respiratory system, the developing
fetus, and the circulatory system (i.e., blood)?
SAB Comment C.l (Summary and Evaluationl-1: The SAB noted that, while Section 1.3.1
(Weight of Evidence for Effects Other than Cancer) contains a summary description of the
toxicological evidence of effects of the TMBs on the nervous, respiratory, circulatory, and
developmental systems, the section does not adequately describe the limitations and uncertainties
within the database or how the results of the hazard assessment will be utilized in the subsequent
dose-response evaluation. The SAB recommended that Section 1.3.1 be revised to include the
following: (1) a short summary of the toxicokinetic similarities and differences among the three
isomers early in the section to provide context to the subsequent effect summaries; (2) a short
summary of the neurological effects database limitations and accompanying uncertainties such as
lack of subchronic data for some isomers, lack of chronic data for all isomers, questions of
reversibility, and lack of mechanistic data; (3) statement(s) regarding the confidence in the hazard
identification results given the limitations of the available database; and (4) inclusion of a
concluding paragraph(s) that states how the results of the hazard identification will be utilized in
the subsequent dose-response evaluation.
EPA Response C.l (Summary and Evaluationl-1: All of the SAB-recommended additions to
Section 1.3.1 have been incorporated into the text.
Charge Question C.2 (Summary and Evaluation): Does EPA's hazard assessment of the
carcinogenicity of trimethylbenzenes clearly integrate the available scientific evidence to support the
conclusions that under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005). there is
"inadequate information to assess the carcinogenic potential" of trimethylbenzenes?
SAB Comment C.2 fSummarvand Evaluationl-1: The SAB agreed with the EPA's
determination that there was "inadequate information to assess the carcinogenic potential" of TMB
isomers and concluded that EPA's hazard assessment of the carcinogenicity of TMB isomers did
integrate all available scientific evidence. The SAB recommended that EPA incorporate data on
related compounds qualitatively to fill data gaps if possible.
EPA Response C.2 (Summary and Evaluationl-1: Information on related alkylbenzene
compounds has been incorporated into the cancer hazard assessment to the extent possible.
Charge Question D.l: Data characterizing the toxicokinetics of 1,2,3-TMB, 1,2,4-TMB, and
1,3,5-TMB following inhalation and oral exposures in humans and experimental animals support the
use of physiologically-based pharmacokinetic (PBPK) models for 1,2,4-TMB. For the purposes of this
assessment, the Hissink et al. (2007) model, originally describing 1,2,4-TMB toxicokinetics following
exposure to white spirit (a complex mixture of volatile organic compounds), was modified by EPA to
calculate internal dose metrics following exposure to 1,2,4-TMB alone for the derivation of an
inhalation RfCfor 1,2,4-TMB. Additionally, the model was further modified by the addition of an oral
route of exposure for use in a route-to-route extrapolation for the derivation of an oral RfDfor
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
30
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
1,2,4-TMB. Please comment on whether the selected PBPKmodel (Hissink et al.. 2007) with EPA's
modifications adequately describe the toxicokinetics of 1,2,4-TMB (Appendix B [of the TMB
Assessment]). Was the PBPK modeling appropriately utilized and clearly described? Are the model
assumptions and parameters scientifically supported and clearly described? Are the uncertainties in
the model structure adequately characterized and discussed?
SAB Comment D. 1-1: The SAB found that the selected model did an adequate job of
simulating the time-course of TMB in the blood of human subjects during and following acute
inhalation exposures. The SAB noted that there was excellent agreement between predicted and
measured blood TMB levels, both during and following 4-hour exposures, for the subjects of Hissink
etal. f20071 inhaling 100 ppm white spirit The SAB noted that the model modestly, but
consistently, under-predicted blood levels in volunteers inhaling 30 ppm TMB for 8 hours
fKostrzewski etal.. 19971 and that the model also consistently under-predicted blood levels in
persons inhaling 2 or 25 ppm TMB for 2 hours (larnbergetal.. 1998.1997a: larnbergetal.. 1996).
but to a larger degree. The SAB noted that these subjects exercised during exposure, which would
increase their systemic uptake of TMB.
EPA Response D.l-1: It should be noted that while exercise will increase systemic uptake,
as stated by the reviewers (by increasing respiration rate and cardiac output), the accompanying
increase in cardiac output would also increase TMB's distribution to the liver, which would
therefore also increase the rate of metabolic clearance. It is unclear how the respective increases in
both respiration and cardiac output, as well as distribution to the liver due to exercise would
influence the ultimate model predictions of TMB blood levels following exercise in humans.
However, given that the model did an adequate job of simulating the time-course of TMB in the
blood of human subjects, EPA determined there was no need to further investigate the "modest"
under-predictions of some of the human data.
SAB Comment D.1-2: The SAB concluded that, in most instances, the model over-predicted
blood TMB levels in rats subjected to single exposures to white spirit fHissink et al.. 20071 and TMB
(Swiercz etal.. 2003). The differences between predicted and empirical levels typically increased
from 1.5-2-fold at lower inhaled concentrations to 4-6-fold at >100 ppm. The accuracy of
predictions of brain levels was similar to those for blood. The SAB found that the model reasonably
simulated blood and brain levels in rats after repeated TMB exposures, and that disparity between
simulated and empirical data also increased with increasing vapor concentration. With the
repeated exposure data of Swiercz et al. (2003). there were ~2- and 3-fold differences for the
25 and 50 ppm exposures, respectively. Differences in brain levels after 606 hours were somewhat
greater. SAB found that there was more disparity (4-5-fold) for blood and brain levels in the rats of
Zahlsen et al. (1992) inhaling 100 ppm TMB for 3 days.
EPA Response D.l-2: In considering these comments on the model fit to the Swiercz et al.
f 20031 data, further attention was given to the discrepancy between the results in Table C-9 and
the model fits in Figure C-12. The data in Figure C-12 come from Table 2 of Swiercz etal. f20031
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
30
31
32
33
34
35
36
37
Supplem en tal Information—Trim ethylbenzenes
and the data in Table C-9 come from Table 4 of that paper, but the results are significantly different
For example. Swiercz etal. (2003) Table 2 lists the 1,2,4-TMB (venous) blood concentration at
3 minutes post-exposure (end of 4th week) as 4.06 ± 0.46 mg/L, while Table 4 lists the (arterial)
blood concentration after "4 weeks" as 1.54 ± 0.32 mg/L. The model calibration used time-course
data from tail-vein sampling, such as in Swiercz etal. f20031 T able 2, and the internal dose being
used is venous concentration, so Table C-8 has been updated to provide a numerical comparison of
these two. At 25 and 100 ppm, the model results are within 30% of the tail-vein data, mostly within
10%, all within 1 standard deviation (SD). At 250 ppm, the discrepancy ranges from a factor of 1.5
(50% over-prediction) to 6-fold.
In the experimental methods, Swiercz et al. f20031 only state that the samples for T able 4
were collected "after decapitation." During the time, or range of times, between removal of animals
from the exposure chamber and decapitation, and until a tissue sample is chilled, evaporative loss
of TMB could occur. Therefore, the table has been revised to compare the data for model results
30-60 minutes post-exposure, rather than immediately after exposure. In contrast, Zahlsen et al.
(1992) state that animals were removed from the exposure chamber and tissues were collected
within 3 minutes.
SAB Comment D.1-3: SAB noted that the poor model prediction for inhaled concentrations
>100 ppm in rats is acknowledged by the EPA authors. SAB further noted that EPA uses the PBPK
model to provide simulations for exposures outside its application domain. This is necessitated by
the fact that the 100 ppm dose is in the middle of the rat dose-response range used for BMD
modeling. SAB concluded that over-predicting rat dosimetry in this range thus has the potential to
influence the results of dose-response modeling and extrapolation of potency to humans. Marked
over-prediction of high-dose data necessitated the omission of the highest dose for BMD modeling.
The SAB recommended two possible options for alleviating this issue. The first option is to
refine the rat PBPK model to improve fits or conduct BMD modeling first using inhaled
concentration to identify the POD, and then using the rat and human PBPK models to determine the
HEC. SAB noted that refining the PBPK model may require recalibration of some type, such as the
addition of a first-order metabolic pathway consistent with the PBPK model of larnberg and
lohanson (1999). or changing hepatic blood flow to 25% instead of 17% of cardiac output
The second option proposed by SAB is for EPA to conduct BMD modeling of the Korsak and
Rvdzvriski f 19961 data using air TMB concentration as the dose metric to derive the POD.
Subsequently, the PBPK model would be used to convert the POD to the weekly average blood
concentration.
EPA Response D.l-3: EPA has chosen to pursue the second option offered by the SAB.
When implementing this option, EPA ensured that the resulting lower confidence limits on the BMD
(BMDLs) used for HEC estimation were below the 100 ppm (492 mg/m3) threshold of model
validation.
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
SAB Comment D.1-4: SAB noted that they conducted a quality control/quality assurance
review and confirmed that the model simulations presented in Appendix B of the IRIS document
draft were accurate. SAB noted that aside from a couple of minor technical issues that were
identified, no fundamental flaws or issues were found.
EPA Response D.l-4: No response necessary.
SAB Comment D.1-5: The SAB found that the EPA's assumptions, in modifying the Hissink
etal. (2007) model to predict the kinetics of inhaled TMB for repeated exposure scenarios, were
reasonable and appropriate. The major caveats, however, were not identified up-front on
page B-20 (e.g., that the original model and its parameters were for TMB and white spirit, lack of
parameters for the oral route, lack of parameters for pregnancy). The SAB recommended that the
EPA expand the explanation and justification for the modifications of model parameters.
Specifically, the discussion of the input parameters (e.g., human tissue:blood partition coefficients,
cardiac output, liver blood flow) should be justified. Additionally the use of scaled-up rat Vmax
values, when human values were available, requires further explanation. Metabolic constants could
be questioned, as they summarily reflect the rate of TMB metabolism during mixed exposures to
white spirits, rather than exposure to TMB alone. The EPA did not attempt any re-estimation or
adjustment of parameters for chronic exposure (e.g., enzyme induction, dose-dependency, growth
dilution). Results of sensitivity analyses can be used to indicate whether the choice of liver blood
flow substantially impacts the model predictions and thus warrants revisiting. It was noted that
human tissue:blood partition coefficients used in modeling were twice those for rats. Meulenberg
and Viiverberg (2000) estimated human brain:blood, fat:blood, and kidney:blood partition
coefficients that were higher for rats than for humans. It was suggested that first-order and
saturable metabolism be incorporated into the model, and the model be run to explore the impact
of the change.
EPA Response D.l-5: As recommended by the SAB, the major caveats and concerns for the
Hissink etal. f20071 model have been added to Section C.2.2. Additional points on specific items
have been added at appropriate points in Section C.2.3. A justification statement for revising model
parameters (i.e., to address the caveats and concerns identified above) was added at the beginning
of Section C.2.3.2, with further justification provided at appropriate points in the section. A
sentence was added to the description of the human model fits, and a brief paragraph was added to
the "Summary of Optimization and Validation," to explain that because the scaled Vmax (i.e., rat-
derived VmaxC) and rat-derived Km were found to adequately predict the human data, and numerical
optimization did not provide a significant improvement in the fit, the scaled Vmax and rat Km were
used for the human model.
Regarding SAB's comment related to fractional blood flow to the liver: if the fractional
blood flow to the liver was increased and no other parameters were changed, then the predictable
result is that the net rate of metabolism would increase. However, the metabolic rate constant Vmax
was calibrated using the fractional hepatic blood flow set in model. So to fully evaluate the model
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
behavior, if hepatic blood flow were increased to 25% of cardiac output for example, there is a need
to first make that change and then re-calibrate the Vmax to the available data. It seems likely that
doing both of these things in combination would for the most part cancel out any impact of
increasing hepatic blood flow alone. The chance of obtaining a significant improvement is uncertain
and this sensitivity analysis would entail considerable effort. As the SAB's overall conclusion was
that the model was adequate describing TMB blood concentrations as currently parameterized, the
significant, additional effort required for this type of sensitivity analysis was not undertaken.
Although Meulenberg and Viiverberg (2000) reported tissue:blood partition coefficients that were
higher in rats than in humans, the original partition coefficients (as identified by Hissink etal.
£2007}) used in the original model fitting were retained in the current PBPK model.
SAB Comment D.1-6: The SAB did not find a specific discussion of the uncertainties in the
model's structure. While these uncertainties may be implicitly included in the uncertainties
discussion, SAB recommended that they should be specifically discussed in reference to the PBPK
model.
EPA Response D.l-6: An extensive discussion of modeling uncertainties was added to
Section C.2.3.2.
SAB Comment D.1-7: One SAB Panelist noted that there is a published human PBPK model
(larnberg and lohanson. 1999). The SAB acknowledged that the EPA requested the model code
through email and was unable to obtain the model. The SAB noted that the model is for TMB alone,
and suggested that using this model may have the following benefits over the Hissink et al. (2007)
model: (1) it avoids the complications and uncertainties of concurrent exposure to other
components in white spirit and necessary species-to-species extrapolations; (2) empirical human
kinetic data are available from the same laboratory for model parameterization and validation; and
(3) human neurobehavioral data are also available in the literature from other research groups.
The SAB noted that the results of these studies identify human no-observed-adverse-effect levels
(NOAELs)/lowest-observed-adverse-effect levels (LOAELs) for acute irritation and central nervous
system (CNS) effects by TMB and white spirit. The SAB noted that EPA policy is to use and consider
human data and validated human models when available. Because the EPA could not obtain the
larnberg and lohanson (1999) model, the SAB provided recommendations to improve the use of the
Hissink etal. (2007) model and encouraged the EPA to, at a minimum, be more transparent in its
discussion of available models and model selection in this and future assessments.
EPA Response D.l-7: The EPA has followed its practices for using human toxicokinetic data,
including data from larnberg and lohanson f 19991 and previous studies by these authors, and of
using a validated human model (i.e., Hissink etal. (2007)) in the TMB assessment. The toxicokinetic
data generated from the Jarnberg and Johanson studies were used in the validation of the human
Hissink etal. (2007) model: these validations are extensively reported and discussed in
Section C.2.3.2. Discussions of the other PBPK models (Section C.2.1) were expanded, specifically
addressing the lack of availability of the larnberg and lohanson f 19991 model and that the EPA
This document is a draft for review purposes only and does not constitute Agency policy.
A-23 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
Supplem en tal Information—Trim ethylbenzenes
generally prefers to use model structures that have been shown to fit both animal and human data,
as this consistency is considered a validation of the model structure.
Charge Question D.2: The internal dose metric selected for use in the derivation of the RfC
and RfDfor 1,2,4-TMB was the steady-state weekly average venous blood concentration (mg/L) of
1,2,4-TMB for rats exposed for 6 h/day, 5 days/week. Please comment on whether the selection of this
dose metric is scientifically supported and clearly described. If a different dose metric is recommended
for deriving the RfC\ please identify this metric and provide scientific support for this choice. Are the
uncertainties in the selected dose metric adequately characterized and discussed?
SAB Comment D.2-1: The SAB stated that the use of any dose metric should be guided by
the mode of action of the chemical being examined. For the TMBs, the SAB acknowledged that there
is a paucity of information on their mode of action, and that the Agency has inferred the mode of
action to be similar to that for chemicals such as toluene. Given the uncertainties in the mode of
action, the SAB found that the selection of the internal dose metric of the venous blood
concentration averaged over a week of exposure is reasonable.
EPA Response D.2-1: No response necessary.
SAB Comment D.2-2: SAB stated that clarification is needed on how the average weekly
venous concentration was determined given that the longer phase half-life of the TMB isomers
indicates that an exposure period longer than a week is required for blood levels to achieve a
steady state. In addition, the SAB noted that the experimental data for both rats and humans show
that steady state is not achieved with only a single week of exposure. Executing the PBPK model
over a 4-week period shows that the average blood levels are still continuing to rise slightly. The
SAB recommended that the model should be run long enough to come to a weekly steady state and
then the associated venous blood concentration should be used as the internal dose metric.
EPA Response D.2-2: This discussion has been added to the relevant section (where
internal metrics are described). The average weekly venous concentration was calculated by
simulating 3 weeks of exposure (6 hours/day, 5 days/week) and calculating the area under the
curve (AUC) during the 3rd week, divided by 168 hours. Extending the simulation to 4 weeks and
using the 4th week for the calculation changed the results by <0.02%.
SAB Comment D.2-3: The SAB noted that the multiple tissues of interest for derivation of an
RfC are primarily extrapulmonary tissues. However, the Agency has a goal to establish RfCs for
multiple endpoints beyond the critical effect endpoint currently being addressed. If an effect in the
respiratory tract is established (such as a change in bronchial alveolar lavage fluid composition)
and an RfC is to be determined, then the appropriate dose metric would be based on the mass
deposited per unit surface area of the lung rather than on the average venous blood concentration.
A mass per unit lung surface area dose metric enables species with significantly different lung sizes
than humans to be used in the derivation of the RfC.
This document is a draft for review purposes only and does not constitute Agency policy.
A-24 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
Supplem en tal Information—Trim ethylbenzenes
EPA Response D.2-3: A dose metric of mass of TMB deposited per unit surface area of the
lung was used in the derivation of RfC values for respiratory effects (i.e., increased inflammatory
lung lesions) (see Section 2.1.2).
SAB Comment D.2-4: The SAB noted that using the PBPK model-estimated internal dose
metrics as the dose inputs for BMD modeling required the Agency to drop the high-dose exposures
from all modeling efforts because the venous blood dose metrics consistently over-predicted
experimental results for high exposures. This overestimation may be due, in part, to the Agency
using minute ventilation as the driver function for internal dose rather than decomposing minute
ventilation into its two components, namely tidal volume and breathing frequency. The SAB noted
that while the exposure level is high, which may lead to a 50% reduction in respiratory rate,
respiratory irritants such as the TMBs cause subtle shifts in the breathing pattern while
maintaining the same overall minute ventilation. Shallower breathing leads to a shift upward in the
respiratory tract for the site of deposition. In addition, the PBPK modeling for humans did not take
into account the periods of exercise that the subjects underwent, which may explain the model's
greater deviations from experimental results at high exposure levels. Consistent with previous
comments, the SAB noted that external air can be used as the dose metric and then the PBPK model
can be used to back-calculate the appropriate venous blood level. If the SAB's suggestions for
improvements in the PBPK model do not lead to a better agreement with the high-dose exposures,
the SAB recommended that the Agency include the external air dose metric and corresponding
venous blood back-calculations.
EPA Response D.2-4: None of the existing PBPK models specifically account for the impact
of varying tidal volume versus breathing frequency on regional deposition and uptake in the
respiratory tract While compartmental models exist that do so (e.g., for acetaldehyde), such a
revision in model structure would be a very large effort and is beyond the scope of what EPA would
consider for this assessment Given this decision, EPA has redone all of the BMD modeling using the
external air concentrations as the dose inputs and then calculated the HEC based on the BMDL
values, consistent with SAB recommendations in SAB Comment 3 of Charge Question D.l.
SAB Comment D.2-5: The SAB noted that, while uncertainties concerning model
parameters, potential for kinetic changes with repeated exposures, and model estimates of internal
dose are discussed, the uncertainties in the selected dose metric (weekly average venous blood
concentration) are not adequately characterized or discussed in the TMB assessment.
EPA Response D.2-5: This discussion was added to Appendix C (Section C.2.3.2).
Charge Question E.l: A 90-day inhalation toxicity study of 1,2,4-TMB in male rats (Korsak
and Rydzynski. 19961 was selected as the basis for the derivation of the RfC. Please comment on
whether the selection of this study is scientifically supported and clearly described. If a different study
is recommended as the basis for the RfC\ please identify this study and provide scientific support for
this choice.
This document is a draft for review purposes only and does not constitute Agency policy.
A-25 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
Supplem en tal Information—Trim ethylbenzenes
SAB Comment E.l-1: The SAB generally agreed with the choice of the Korsak and Rydzvhski
(1996) study as the basis for derivation of the RfC for 1,2,4-TMB. The study utilized a 90-day
exposure period and, thus, the longest duration exposure study available in the literature; in
addition, it included multiple exposure levels. It was well-conducted and utilized adequate sample
sizes of rats. In addition, it was based on widely-used behavioral assays. An examination of the
study indicates that these behavioral studies were carefully carried out and data from control
animals were consistent with previously published observations. However, the SAB recommended
several ways in which the clarification for this choice could be strengthened (see SAB
Comments E. 1-2-E. 1-8 below for specifics).
EPA Response E.l-1: See EPA Responses E.l-2 through E.l-8 below for detailed responses
to the individual recommendations.
SAB Comment E.l-2: The SAB noted that the rationale for the choice of Korsak and
Rydzvhski (1996) is not specifically described and recommends that the reasons for its choice over
other studies (e.g., the 4-week exposure studies) need to be more clearly stated.
EPA Response E.l-2: An increased justification for selection of the Korsak and Rydzvhski
(1996) study was added to Section 2.1.5, including the rationale for selection of that study over the
other neurotoxicity studies that utilized a short-term exposure protocol fWiaderna etal.. 2002:
Gralewicz and Wiaderna. 2001: Wiadernaetal.. 1998: Gralewicz etal.. 1997b).
SAB Comment E.1-3: The SAB expressed concern that the TMB assessment, as currently
written, is confusing regarding the chronicity of exposure versus effects. The SAB recommended
that it would be helpful to modify the terminology particularly related to outcome measures,
perhaps as acute effects versus long-term effects/irreversible effects and retain the use of the word
chronic/subchronic etc. to descriptions of statements related specifically to exposure.
EPA Response E.l-3: The Hazard Identification and Dose-Response Analysis sections have
been edited to increase clarity with respect to language describing either the chronicity of exposure
or the nature of the described effects (i.e., acute or long-term/latent effects).
SAB Comment E.1-4: The SAB recommended that EPA separate the dose-response write-up
into sections that specifically elaborate on the acute effects and provide a separate section related
to effects observed post-exposure. The SAB also recommended that, given the commonality of the
trends in data across these studies, some mention of the biological significance in the absence of
statistical significance should be mentioned.
EPA Response E.l-4: The discussion of acute and post-exposure effects has been
reorganized in the Dose-Response Analysis section to the extent possible. A discussion of the
biological significance of the post-exposure data was also included in the assessment. However,
acute and long-term/latent/post-exposure effects have not been separated into distinct sections as
each type of data, when considered in tandem, informs the larger decisions made in the assessment
regarding the suitability of the decreased pain endpoint for derivation of the RfC. As such, EPA
concluded that, while more clearly delineating the types of effects was possible within a single
This document is a draft for review purposes only and does not constitute Agency policy.
A-26 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
Supplem en tal Information—Trim ethylbenzenes
section, separating the effects into individual sections would possibly obscure the rationales behind
EPA's conclusions.
SAB Comment E.1-5: The SAB recommended that the text, where applicable, could include
additional qualifications as to "reversibility of effects" at the 2-week post-exposure time point. This
assessment of reversible effects of failures on the rotarod is based on the finding of lack of
statistical difference between treated and control groups at 1 week post-exposure following a
13-week exposure period for one of two isomers. Some TMB Panel members felt that this was
sufficient evidence for reversibility, while other members did not feel that this provided sufficient
evidence. Specifically, this interpretation of a reversal relied on a reduction from 40% rotarod
failure during the final week of exposure compared to 35% 1 week post-exposure, as compared to
0% rates for controls. There was no such statistical reversal for the other isomer, and for both
isomers, the magnitude of the reduction post-exposure was minimal. Further, it was not clear that
the statistical analyses of these data incorporated a repeated measures component that would be
required by the experimental design. Thus, while a case was stated for a statistically significant
reversal, several TMB Panel members felt that it was not consistent nor did it appear to be
biologically meaningful.
EPA Response E.l-5: Additional qualifications on the determination of whether the
decreased pain sensitivity endpoint was reversible have been added to Sections 1.2.1,1.3.1, and
2.1.5. In particular, it is noted throughout the section that all of the available evidence, especially
considering information from the short-term studies, strongly indicates that the pain sensitivity
endpoint is not immediately reversible upon termination of exposure, and that persistent changes
to the nervous system occur due to TMB exposure. It should be noted that the SAB focused solely
on decreased rotarod performance in their comment, which is not used in the RfC derivation. The
data for decreased rotarod performance, as a measure of decreased neuromuscular function, were
determined by EPA to not be appropriate for consideration for derivation of the RfC (Section 2.1.1)
due to the manner in which the data were reported. Failures on the rotarod were recorded as
quantal data (percent of animals "failing" on the rotarod due to latencies of up to 119 seconds)
rather than being recorded as a continuous variable (i.e., latency to falling off rotarod apparatus).
Therefore, as the rotarod data were not considered for derivation of the RfC, extensive discussions
regarding the possible reversibility of this endpoint were not added to the assessment. However,
where possible, evidence from all effects has been discussed in the context of overall alterations of
neurological function due to TMB exposure.
SAB Comment E.1-6: SAB recommended that the EPA re-calculate the RfC as if the study
were subchronic (i.e., UF converts to 1 from 3) and report these subchronic RfC values as well.
EPA Response E.l-6: EPA has calculated and included these subchronic RfC values in
Section 2.1.8 of the Dose-Response Analysis section.
This document is a draft for review purposes only and does not constitute Agency policy.
A-27 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
Supplem en tal Information—Trim ethylbenzenes
SAB Comment E.1-7: SAB recommended that more specific mention of the potential
cumulative neurotoxicity that is suggested by the repeated measurement finding of rotarod
performance failures across the course of exposure be included in the document.
EPA Response E.l-7: As the rotarod data were not considered for derivation of the RfC, this
discussion was not added to the assessment Data on decreased pain sensitivity were not provided
in the same manner as rotarod data (i.e., measures of effect provided at multiple intervening time
points during the period of exposure) and therefore, a discussion of the possible cumulative effects
regarding decreased pain sensitivity was likewise not added to the document.
SAB Comment E.1-8: The SAB recommended including more specific descriptions of the
similarity of the animal behavioral endpoints to what has been observed in humans.
EPA Response E.l-8: A discussion of the similarity of animal neurobehavioral endpoints to
the measures of neurotoxicity observed in human studies has been added to Section 2.1.5.
Charge Question E.2: Decreased pain sensitivity (measured as an increased latency to
pawlick response after a hotplate test) in male Wistar rats was concluded by EPA to be an adverse
effect on the nervous system and was selected as the critical effect for the derivation of the RfC. Please
comment on whether the selection and characterization of this critical effect is scientifically supported
and clearly described. If a different endpoint(s) is recommended as the critical effect(s)for deriving
the RfC\ please identify this effect and provide scientific support for this choice.
SAB Comment E.2-1: The SAB agreed that the reduction in pain sensitivity as indicated by
an increased latency to paw-lick response in a hot plate test, is a valid adverse nervous system
effect and appropriately selected as a critical effect for the derivation of the RfC. This effect was
variously seen in response to short-term, 4-week, and 90-day studies. The associated U-shaped
dose-effect curves seen with these isomers, moreover, are highly consistent with the effects of
various other pharmacological agents (e.g., opioids) on this response and likely reflective of the
mechanisms by which these isomers act. This assay is widely used in the behavioral pharmacology
literature and particularly in the study of pain nociception and opioid pharmacology.
EPA Response E.2-1: No response necessary.
SAB Comment E.2-2: The SAB agreed that the observation of prolonged latency in the hot
plate test 24-hour post-footshock delivery that was observed in studies by Gralewicz and
colleagues (Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997b) also constitutes an adverse
effect. The administration of footshock immediately after the hot plate test trial strains the
capabilities of the nervous system and, thus, provides a type of nervous system probe that then
unmasks a prolonged latency to a hot plate stimulus 24 hours later. It shows that when the nervous
system is maximally stressed, it cannot respond/recover in a normal timeframe.
EPA Response E.2-2: No response necessary.
SAB Comment E.2-3: SAB, in addition to making the recommendations above for the
document related to the nervous system effects, also noted that this section could benefit from
some additional description of the hot plate procedures, including the rationale/approach for using
This document is a draft for review purposes only and does not constitute Agency policy.
A-28 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
Supplem en tal Information—Trim ethylbenzenes
the footshock intervention in the post-exposure behavioral assessments carried out after the
4-week exposures.
EPA Response E.2-3: Additional details on the hot plate procedure have been added to the
Hazard Identification and Dose-Response Analyses sections. Additional rationale for the inclusion
of the footshock challenge in the short-term studies has also been added to the assessment.
Charge Question E.3: In order to characterize the observed dose-response relationship
comprehensively, benchmark dose (BMD) modeling was used in conjunction with dosimetric
adjustments for calculating the human equivalent concentration (HEC)from a rat and human PBPK
model (Hissink et al.. 20071 to identify the point of departure (POD) for derivation of the RfC. Please
comment on whether this approach is scientifically supported for the available data, and clearly
described.
A. Has the modeling been appropriately conducted and clearly describedbased on EPA's
Benchmark Dose Technical Guidance U.S. EPA (20121?
B. Has the choice of the benchmark response (BMRJfor use in deriving the POD (i.e., a BMR
equal to 1 standard deviation change in the control mean for the latency to pawlick response) been
supported and clearly described?
SAB Comment E.3-1: SAB expressed concern over EPA's decision to omit the high-dose
group from the Korsak and Rydzvhski (1996) study before BMD modeling. However, an BMD
analysis conducted by the SAB on the same dataset using air concentration as the dose metric
results in the same POD air concentration as BMD modeling based on internal dose and using the
low- and mid-dose groups. As a result, the SAB agreed that the overall results for the POD
generated by the EPA are adequate, but strongly suggested that the Agency provide a more robust
explanation of any analyses. The SAB also considered Appendix C-2 in the TMB Assessment
(External Review Draft) as inappropriate and recommended deleting it. If the EPA is so inclined,
they could replace it with the BMD analysis using air concentration as the dose metric.
EPA Response E.3-1: In SAB's analysis above, they ran one model (Exponential M4) against
the data as that was the model that was selected in External Peer Review draft It is true that this
model returns the same POD regardless of if air concentrations or internal dose is used. However,
the method SAB used doesn't take into account other model fits and the model selection protocols
EPA uses in BMD modeling. When all available continuous models were run agains the decreased
pain sensitivity endpoint, the HEC generated for decreased pain sensitivity due to exposure to 1,2,4-
TMB using the SAB-suggested modeling method (model air concentrations and then convert to HEC
using the PBPK model) differs slightly from the POD included in the External Peer Review Draft of
the TMB assessment (18.1 versus 15.8 mg/m3). However, SAB's larger point stands in that it is
appropriate to model the TMB toxicity endpoints using the external air concentrations as the dose
inputs and then convert the resultant BMDLs into HECs using the available PBPK model. This
methodology obviates the need for extensive revisions to the PBPK model code, and ensures that
any HECs generated from the PBPK model originate from BMDLs that fall within the model's range
This document is a draft for review purposes only and does not constitute Agency policy.
A-29 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
Supplem en tal Information—Trim ethylbenzenes
of validation. As such, all BMD modeling has been redone according to SAB's recommendations.
EPA also agrees with the SAB regarding Appendix C-2 (External Peer Review draft); this appendix
has been removed from the document
SAB Comment E.3-2: The SAB recommended that the EPA provide better justification for
applying the "one standard deviation" from the mean of the control group for the neuro-
toxicological endpoint than using the Agency default value. The EPA should also provide better
explanation of the issues associated with the homogeneity of variance across dose groups in the
Korsak and Rydzvhski (1996) study, its implications for BMD modeling, and how the EPA
addressed this in their BMD modeling.
EPA Response E.3-2: A more robust justification for the selection of 1 control group SD as
the BMR for modeling some continuous endpoints has been added to Section 2.1.2, and a brief
discussion regarding the uncertainty around the BMR selection has been added to Section 2.1.6.
The observation of differential variance estimates across dose groups, and how this was handled
when performing BMD modeling, was also discussed more extensively in Section 2.12. For
example, the variances reported for decreased pain sensitivity were clearly non-constant, with the
reported variances at 492 mg/m3 being lower (1,2,4-TMB) or higher (1,2,3-TMB) compared to
other dose groups. This heteroscedasticity could reflect measurement error (e.g., different lab
technicians recording responses differently), experimental error (e.g., the hotplate apparatus may
not have held a constant temperature), or may reflect that the latency response may be log-
normally distributed rather than the assumed normal distribution. The latter possibility does not
seem to be the case as the approximation of geometric means and SDs from the reported arithmetic
means and SDs did not reduce the heterogeneity in reported variances. In order to account for data
with reported heteroscedasticity, BMD modeling was performed using variance estimates that were
modeled as a power function of the reported mean value.
Charge Question E.4: Please comment on the rationale for the selection of the uncertainty
factors (UFs) applied to the POD for the derivation of the RfCfor 1,2,4-TMB. Are the UFs appropriate
based on the recommendations described in Section 4.4.5 of A Review of the Reference Dose and
Reference Concentration Process U.S. EPA (20021. and clearly described? If changes to the selected
UFs are proposedplease identify and provide scientific support for the proposed changes.
SAB Comment E.4-1: The SAB agreed with the UFa of 3 and its rationale. The default UFa of
10 can be divided into two half-log UF components of 3 each to account for species differences in
toxicokinetics and toxicodynamics, respectively. In developing the RfC for 1,2,4-TMB, the EPA used
PBPK modeling to convert estimated internal doses in rats in toxicity studies of 1,2,4-TMB to
corresponding applied doses in humans. PBPK modeling substantially reduces uncertainty
associated with extrapolating animal exposures to humans based upon toxicokinetic differences,
justifying elimination of one of the half-log components of the default UFA of 10 fU.S. EPA. 20021.
Uncertainty regarding possible toxicodynamic differences among species (i.e., different sensitivity
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
to toxicity at equivalent internal doses) remains, justifying keeping the other half-log component
of 3.
EPA Response E.4-1: No response necessary.
SAB Comment E.4-2: The SAB agreed with the UFh of 10 and its rationale, although one
TMB Review Panel member thought that a UFh of 3 would be adequate. This UF is intended to
account for potential differences among individuals in susceptibility to toxicity. The EPA concluded
that no information on potential variability in human susceptibility to 1,2,4-TMB toxicity exists with
which to justify using a value other than the default of 10. It was noted during discussion that
numerous clinical studies have demonstrated that humans, including pediatric and geriatric
patients, differ by only about 2-fold in their susceptibility/sensitivity to inhaled lipophilic
anesthetics (e.g., chloroform, halothane), indicating to one Panel member that a UFh of 3 would be
scientifically defensible given the neurotoxicity endpoint used to establish the POD. Other TMB
Panel members disagreed, stating that the mode of action of neurotoxicity of 1,2,4-TMB is unknown
and that the actions of general anesthetics may have little or no bearing on variability in TMB
susceptibility. In their opinion, the full UFh of 10 is warranted.
EPA Response E.4-2: EPA agrees with the majority of the SAB Panel members in that, given
the lack of information regarding TMB's mode of action, limited information exists that could
predict the potential for variation in human susceptibility to TMB exposure. Therefore, the value of
UFh = 10 is retained in the TMB assessment
SAB Comment E.4-3: The SAB agreed with the EPA's choices for UFl values (i.e., a UFl of
1 for all endpoints except increased bronchoalveolar lung cells, for which a UFl of 10 was selected).
However, the SAB suggested that the justification for the UFl be strengthened. This UF is intended
to be used when the POD is a LOAEL rather than a NOAEL. In conducting BMD modeling, a BMD
equal to 1 SD change in the control mean for modeled endpoints was selected. The document
would be improved by adding an explanation of the reasoning for selection of 1 SD (versus 0.5 SD)
along with a clearer discussion of why this is expected to lead to a POD for which a UFl of 1 is
appropriate.
EPA Response E.4-3: A stronger justification for selection of a BMR = 1 control SD has been
added to the text. The LOAEL to NOAEL UF, UFl, of 1 was applied for endpoints modeled with the
EPA Benchmark Dose Software (BMDS) because the current approach is to address this factor as
one of the considerations in selecting a BMR for BMD modeling. In other words, when selecting a
BMR value, care should be taken to select a response level that constitutes a minimal, biologically
significant change so that the estimated BMDLs can be assumed to conceptually correspond to a
NOAEL. In the case of TMBs, BMRs were preferentially selected based on biological information on
what constitutes a biologically significant change for these effects. For example, a 5% reduction in
fetal body weight was selected as the BMR for that endpoint based on the fact that a 10% reduction
in adult body weight is considered adverse, the assumption that fetuses are a susceptible
population and thus more vulnerable to body weight changes, and the fact that decreases in fetal
This document is a draft for review purposes only and does not constitute Agency policy.
A-31 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
Supplem en tal Information—Trim ethylbenzenes
weight in humans are associated with a number of chronic diseases such as hypertension and
diabetes. For endpoints for which there was no information available to make assumptions about
what constitutes a minimal, biologically significant response, a BMR equal to a 1 SD change in the
control mean was selected. In both cases, the BMR selected was assumed to return BMDL values
that conceptually correspond to a NOAEL, thus obviating the need for a LOAEL to NOAEL UF.
SAB Comment E.4-4: The SAB agreed with the UFs of 3, although one TMB Panel member
thought that a UFs of 10 would be more appropriate. When the data used to generate a chronic RfC
are from subchronic studies, a UFs is used to address uncertainty around whether longer exposures
might lead to effects at lower doses. The EPA justified using less than a full default factor of 10 for
this UF based on evidence suggesting possible reversibility of neurotoxicity and hematotoxicity
endpoints. Most of the SAB Panel members were satisfied with this justification, but some
members of the TMB Panel disputed the evidence for reversibility of effects. In addition, several
TMB Panel members noted that reversibility following cessation of exposure was irrelevant since
the chronic RfC is applicable to lifetime exposure (i.e., there is no post-exposure period). The
discussion regarding reversibility of neurotoxic effects is presented in response to the RfC for
1,2,4-TMB (see Section 2.2.5). The TMB Review Panel discussed that some hematologic effects
considered by the EPA appeared to resolve when exposure ceased, but other effects did not resolve,
and that inflammatory pulmonary effects can lead to persistent injury. The SAB noted that factors
other than reversibility could contribute to selection of a UFs less than 10, such as evidence from
PBPK modeling that 1,2,4-TMB does not accumulate in the body over time and empirical evidence
that the POD does not appear to decrease when results from subchronic studies are compared with
studies of shorter duration. One TMB Panel member thought that none of these considerations had
sufficient merit to justify using less than the full default UFs of 10.
EPA Response E.4-4: Upon reconsideration of the neurotoxicity, hematological toxicity, and
respiratory toxicity data contained in the TMB database, EPA agrees with members of the SAB
Panel recommending a UFs of 3. Given that the adaptive responses of the nervous system appear to
be impaired several weeks after short-term exposure, including prolongation of decreased pain
sensitivity phenotypes following environmental challenge using a footshock, the concern that
chronic exposures may more thoroughly overwhelm adaptive responses in the nervous system, and
thus lead to more severe responses, remains. In addition, there is evidence that neurotoxicity
worsens with continued exposure, and thus, effects are expected to be more severe following
chronic exposure. For example, decrements in rotarod function were shown to increase in
magnitude as a function of exposure duration, worsening from 4 to 8 weeks of exposure, and
worsening further from 8 to 13 weeks of exposure (Korsak and Rvdzyhski. 19961. Although a
similar time-course is not available for reduced pain sensitivity, reduced pain sensitivity is
observed at approximately 5-fold lower concentrations following subchronic exposure, as
compared to acute exposure (see discussion in Section 1.2.1 of the Toxicological Review). However,
there does not seem to be an exacerbation of other neurotoxic effects at lower doses when
This document is a draft for review purposes only and does not constitute Agency policy.
A-32 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
Supplem en tal Information—Trim ethylbenzenes
comparing subchronic exposures to short-term exposures. Further, evidence from toxicokinetic
studies indicates that blood and organ concentrations of TMBs are similar following repeated vs.
acute exposures (approximately 600 hours vs. 6 hours, respectively; see Table C-9) and the PBPK
model predicts less than a 5% increase between the first day and subsequent days of repeated
exposures. By extension, it can be reasonably assumed that TMB isomers would not accumulate to
an appreciably greater degree following a longer chronic exposure and thus may not lead to effects
at lower doses compared to shorter duration studies. Taken together, the toxicokinetic and
toxicological data support the application of a UFs of 3 for neurotoxic, hematological, and
respiratory endpoints. The text regarding the selection of the UFs has been revised in Section 2.1.3
to reflect these conclusions, and a UFs of 3 has been applied to all endpoints other than fetal weight
Additionally, as previously discussed in EPA Response E.l-5, a more extensive discussion of the
possible reversibility of the decreased pain endpoint has been added to Sections 1.2.1,1.3.1, and
2.1.5.
SAB Comment E.4-5: The SAB was divided on whether the UFd should be 3, as selected by
the Agency, or 10. The purpose of this UF is to account for overall deficiencies in the database of
studies available to assess potential toxicity. The EPA cited strengths in the database in terms of
availability of information on multiple organ/systems from three well-designed subchronic toxicity
studies in justifying not using the full default factor of 10. In retaining a half-log factor of 3, the EPA
noted the absence of a multi-generation reproductive/developmental toxicity study as a weakness
in the database, and specifically concern for the absence of a developmental neurotoxicity study for
1,2,4-TMB given the importance of neurotoxicity in establishing the RfC. Among those who agreed
with a UFd of 3, some found the justification provided by the EPA to be satisfactory, while others
thought that toxicity data available for C9 mixtures should contribute to the rationale to lower the
value from the default of 10. Others disagreed with including C9 mixture data as relevant to the
database UF. Panel members who thought that the UFd should be 10 cited various reasons,
including the absence of data in other species and the absence of a multi-generational reproductive
study, as well as the opinion that the absence of a developmental neurotoxicity study alone
warranted a full factor of 10. One TMB Panel member pointed out that analogy with toluene
suggests that the perinatal exposure could lead to neurodevelopmental effects at doses 10-fold
lower than the NOAEL for effects in adults. An additional point made by another Panel member
was that the RfCs for all of the isomers are being set at the same value, whereas the database is
severely limited for the 1,2,3- and 1,3,5-TMB isomers and the latter two compounds deserve a UFd
of 10. Therefore, for consistency, a factor of 10 should be used for all the isomers.
EPA Response E.4-5: After careful consideration of the available TMB toxicity database, and
the database for mixtures containing TMB isomers (i.e., the C9 fraction) and information pertaining
to related alkylbenzenes, EPA determined that a UFd of 3 was the most appropriate value. This
decision was further supported by the restructuring of the TMB RfC derivation section into an
overarching section covering all three TMB isomers, rather than three individual RfC sections
This document is a draft for review purposes only and does not constitute Agency policy.
A-33 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
Supplem en tal Information—Trim ethylbenzenes
covering a single isomer. In this manner, the entirety of the TMB toxicity database for all isomers
could be considered in total. Strengths of this database include three well-conducted subchronic
studies that investigated effects in multiple organ/systems in Wistar rats (nervous, respiratory, and
hematological systems) and a well-conducted developmental toxicity study that investigated
maternal and fetal toxicity in a different strain of rats (Sprague-Dawley). Consideration of
developmental toxicity studies investigating the effects of mixtures containing TMB isomers fMckee
etal.. 1990: U ngvarv and Tatrai. 19851 supports the general observation of the developmental
toxicity of individual TMB isomers. In these studies, developmental toxicity was observed in rats,
mice, and rabbits, but only at doses >500 mg/m3, which is higher than the lowest LOAEL for
neurotoxicity effects in rats (i.e., 123 mg/m3 for decreased pain sensitivity following exposure to
1,2,3-TMB). Identified gaps in the TMB database include the lack of a multi-generational
reproductive study and the lack of a developmental neurotoxicity study. Regarding the lack of a
reproductive study, information from a C9 fraction study investigating reproductive and
developmental toxicity in rats provided suggestive evidence of reproductive toxicity (decreased
male fertility in the Fi generation and a possible intergeneration effect on body weight in which
fetal/pup/adultbody weights were decreased at lower doses in later generations compared to
earlier generations) fMckee etal.. 19901. However, the lowest concentration of TMB isomers that
elicited these results was 1,353 mg/m3 (as part of the total mixture), which is much greater than
TMB concentrations that elicit neurotoxicity in adult animals (123 mg/m3 for 1,2,3-TMB and 492
mg/m3 for 1,2,4-TMB).
Another gap in the TMB database is the lack of a developmental neurotoxicity study.
Current U.S. EPA (20021 guidance, EPA's A Review of the Reference Dose and Reference
Concentration Processes, recommends that the database UF take into consideration where there is
concern from the available toxicity database that the developing organism may be particularly
susceptible to effects in any organ/system. Given the observations that exposure to all three TMB
isomers elicits strong and consistent markers of neurotoxicity, that exposure to TMB isomers
results in developmental toxicity, as well as explicit information that TMB isomers can cross the
placenta, there exists a concern that exposure to TMB isomers may result in developmental
neurotoxicity. However, evidence from the toluene literature indicates that, while toluene does
cross the placenta and that toluene levels in the placenta, amniotic fluid, and fetal brains increased
with increasing exposures, concentrations in the amniotic fluid were less than those in maternal
tissues. Although this fails to account for potential differences in sensitivity of the developing
organism to induced effects, or for differences in metabolism, it does suggest that gestational
exposure to TMBs might result in lower exposure concentrations to the fetus, which raises
uncertainty in the TMB and related compound database regarding whether sufficient amounts of
the toxic agent crosses the placenta to elicit effects, and whether the concentrations necessary to
elicit effects are lower than those that result in neurotoxicity in the adult organism. Further, while
there is clear evidence from the human and animal literature that exposure to related
This document is a draft for review purposes only and does not constitute Agency policy.
A-34 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
Supplem en tal Information—Trim ethylbenzenes
alkylbenzenes results in developmental neurotoxicity, much of this evidence comes from
epidemiological studies of inhalant abuse or animal studies using exposure paradigms intended to
approximate inhalant abuse patterns (i.e., high exposure concentrations and intermittent and
noncontinuous exposures). Therefore, there is some uncertainty whether the concentrations
necessary to cause developmental neurotoxicity are lower than those that result in neurotoxicity in
the adult organism.
However, evidence from perinatal exposures (during a period of postnatal brain
development that continues processes begun early in embryogenesis, including synaptogenesis and
myelination) indicates that the developing organism is at some risk of early life exposures (and
possibly prenatal exposures). These studies fWin-Shwe etal.. 2012: Win-Shwe and Fuiimaki. 20101
demonstrated that low-level exposures early in life (5 ppm toluene, postnatal days [PNDs] 4-12)
altered the expression of neurotransmitter receptors and increased the expression of
neuroimmune markers in the hippocampus of mice. Additionally, early postnatal exposure to
5 ppm toluene produced decrements in spatial learning compared to higher adult doses (50 ppm)
that induced the same effect Ultimately, it is difficult to parse out exactly how the database UF
should account for this. Sensitive subpopulations, including children, are protected against the
effects of exposure to environmental toxicity through the application of the human variability UF.
However, as the processes that are perturbed in the Win-Shwe studies fWin-Shwe etal.. 2012: Win-
Shwe and Fuiimaki. 20101 begin during gestation, residual uncertainty exists concerning
developmental susceptibility to the neurotoxic effects of TMB isomers. As such, EPA determined
that a 3-fold database UF should be applied to the PODhec in order to account for the lack of a
developmental neurotoxicity study in the available toxicity database for TMB isomers.
Charge Question F.l: A 90-day inhalation toxicity study ofl,2,3-TMB in male rats (Korsak
and Rydzynski. 19961 was selected as the basis for the derivation of the RfC. Please comment on
whether the selection of this study is scientifically supported and clearly described. If a different study
is recommended as the basis for the RfC\ please identify this study and provide scientific support for
this choice.
SAB Comment F.l-1: The SAB agreed with the EPA's conclusion not to base the RfC
derivation for 1,2,3-TMB on isomer-specific data. The justification for this conclusion is supported
and clearly described. The SAB was not aware of chronic or subchronic studies that could be used
to support an RfC derivation for 1,2,3-TMB with neurotoxicity as the critical endpoint, similar to the
Korsak and Rydzvhski (19961 study used to develop the 1,2,4-TMB RfC. As with 1,2,4-TMB, the SAB
found that the clarification of this choice, however, could be greatly improved by expanding the
assessment on the same points discussed for 1,2,4-TMB (see SAB Comments 2-8 under Charge
Question E.l).
EPA Response F.l-1: Contrary to SAB's statement regarding the RfC for 1,2,3-TMB, the EPA
did use isomer-specific data on decreased pain sensitivity observed in Korsak and Rvdzvriski
f 19961 to derive the RfC for 1,2,3-TMB in the External Peer Review Draft for TMBs (i.e., both
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplem en tal Information—Trim ethylbenzenes
1,2,4-TMB and 1,2,3-TMB isomer-specific data were available in this study). This analysis is
retained in the current assessment In the revised TMB assessment, the RfC derivation sections for
all isomers have been combined into a unified section. Therefore, in responding to SAB
Comments 2-8 under Charge Question E.l (and most other comments made under Charge
Questions F and G), the recommendations made under this comment have been achieved in
reorganizing the overall section (see EPA Responses E.l-2 through E.l-8).
Charge Question F.2: Decreased pain sensitivity (measured as an increased latency to
pawlick response after a hotplate test) in male Wistar rats was concluded by EPA to be an adverse
effect on the nervous system and was selected as the critical effect for the derivation of the RfC. Please
comment on whether the selection and characterization of this critical effect is scientifically supported
and clearly described. If a different endpoint(s) is recommended as the critical effect(s)for deriving
the RfC\ please identijy this effect and provide scientific support for this choice.
SAB Comment F.2-1: The SAB agreed that reduction in pain sensitivity as indicated by an
increased latency to paw-lick response in a hot plate test was a valid adverse nervous system effect
and was appropriately selected as a critical effect for RfC derivation of 1,2,3-TMB. The SAB noted
that the Agency appropriately uses the same rationale to derive the RfC for 1,2,4-TMB, and as such,
the comments provided under Charge Question E.2 pertain to the derivation of the RfC for
1,2,3-TMB.
EPA Response F.2-1: In the revised TMB assessment, the RfC derivation sections for all
isomers have been combined into a unified section. Therefore, in responding to the comments
under Charge Question E.2, the recommendations made under this comment have been achieved in
reorganizing the overall section (see responses to Charge Question E.2).
Charge Question F.3: In order to characterize the observed dose-response relationship
comprehensively, benchmark dose (BMD) modeling was used in conjunction with default dosimetric
adjustments (U.S. EPA. 1994b) for calculating the human equivalent concentration (HEC) to identijy
the point of departure (POD) for derivation of the RfC. Please comment on whether this approach is
scientifically supported for the available data, and clearly described.
A. Has the modeling been appropriately conducted and clearly describedbased on EPA's
Benchmark Dose Technical Guidance U.S. EPA (2012)?
B. Has the choice of the benchmark response (BMR)for use in deriving the POD (i.e., a BMR
equal to a 1 standard deviation change in the control mean for the latency to pawlick response) been
supported and clearly described?
SAB Comment F.3-1: The SAB response to this charge question deals with the same issues
as charge question for 1,2,4-TMB and did not identify any issues specific to 1,2,3-TMB; see Charge
Question E.3 for specific comments.
EPA Response F.3-1: See EPA Response E.3-2 for details regarding providing a more robust
justification for use of 1 SD change as the BMR for BMD modeling purposes. SAB Comment 1 to
This document is a draft for review purposes only and does not constitute Agency policy.
A-36 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
Supplem en tal Information—Trim ethylbenzenes
Charge Question E.3 does not pertain to 1,2,3-TMB, as the available PBPK model was not used
generate HEC values for 1,2,3-TMB; default dosimetric methods were employed.
Charge Question F.4: Please comment on the rationale for the selection of the uncertainty
factors (UFs) applied to the POD for the derivation of the RfCfor 1,2,3-TMB. Are the UFs appropriate
based on the recommendations described in Section 4.4.5 of A Review of the Reference Dose and
Reference Concentration Process U.S. EPA f2002). and clearly described? If changes to the selected UFs
are proposedplease identify and provide scientific support for the proposed changes.
SAB Comment F.4-1: The SAB noted that the UF values selected by the EPA for 1,2,3-TMB
are identical to those selected for 1,2,4-TMB, and that the justifications are the same. Thus, all
recommendations made by SAB under Charge Question E.4 pertain to the derivation of the RfC for
1,2,3-TMB as well.
EPA Response F.4-1: As all of the individual RfC sections for each isomer have been
combined into a unified RfC section; please refer to EPA Responses E.4-1 through E.4-5 for full
details on EPA's response.
Charge Question G.l: One developmental toxicity study (Saillenfait et al.. 20051 following
inhalation exposure to 1,3,5-TMB was identified in the literature and was considered as a potential
principal study for the derivation of the RfCfor 1,3,5-TMB. However, the candidate RfC derived for
1,3,5-TMB based on this study (and the critical effect of decreased maternal weight gain) was 20-fold
higher than the RfC derived for 1,2,4-TMB (based on decreased pain sensitivity). Given the available
toxicological database for 1,2,4-TMB and 1,3,5-TMB, there are several important similarities in the
two isomers' neurotoxicity that support an RfCfor 1,3,5-TMB that is not substantially different than
the RfC derived for 1,2,4-TMB. Additionally, the available toxicokinetic database for the two chemicals
indicates that internal dose metrics would be comparable. Thus, EPA concluded that deriving such
disparate RfCsfor these two isomers was not scientifically supported. Rather, EPA concluded that
given the similarities in toxicokinetics and toxicity between the two isomers, there was sufficient
evidence to support adopting the RfCfor 1,2,4-TMB as the RfCfor 1,3,5-TMB.
Please comment on EPA's conclusion to not base the RfC derivation for 1,3,5-TMB on isomer-
specific data. Is the scientific justification for not deriving an RfC based on the available data for
1,3,5-TMB supported and has been clearly described?
SAB Comment G.l-1: The SAB agreed with the EPA conclusion not to base the RfC
derivation for 1,3,5-TMB on isomer-specific data. The justification for this conclusion is supported
and clearly described. The SAB was not aware of chronic or subchronic studies that could be used
to support an RfC derivation for 1,3,5-TMB with neurotoxicity as the critical endpoint, similar to the
Korsak and Rydzvhski (1996) study used to develop the 1,2,4-TMB RfC. The candidate inhalation
values for 1,3,5-TMB, based on maternal and fetal toxicity from the study of Saillenfait etal. (2005).
are presented by EPA, but were not chosen as the overall RfC. Although the SAB took issue with the
PODs selected by EPA in their analysis of the Saillenfait et al. f20051 study, as discussed below in
SAB Comments G.l-2 and G.l-3, it nevertheless agreed with the decision notto use this study to
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplem en tal Information—Trim ethylbenzenes
derive the overall RfC for 1,3,5-TMB. The SAB concurred with EPA that the best approach under the
circumstances is to adopt the RfC for 1,2,4-TMB, based on decreased pain sensitivity, as the overall
RfC for 1,3,5-TMB.
EPA Response G.l-1: As detailed above, EPA has significantly restructured the RfC
derivation section for the three TMB isomers. Whereas before, a single RfC section was provided
for each individual TMB isomer, the revised draft includes a unified RfC derivation section that
covers all three TMB isomers. EPA restructured the RfC section in this way to reduce the difficulty
of reading three separate RfC sections, and to make more apparent the scientific decisions that
were reached in deriving RfCs for the individual TMBs. In the old RfC section structure, a final RfC
value was selected in each RfC section for the individual RfC isomers. This led to the situation
where the "final" RfC for 1,3,5-TMB, based on isomer-specific data on decreased maternal weight
gain, was 20-fold higher than the "final" RfC for 1,2,4-TMB (based on decreased pain sensitivity). In
this situation, EPA made the justification that the toxicokinetic and toxicological databases for 1,2,4-
TMB and 1,3,5-TMB did not support such disparate RfCs for the two isomers. Thus, EPA provided a
justification for adopting the RfC for 1,2,4-TMB as the RfC for 1,3,5-TMB. However, the structure for
the new RfC section in the revised draft is streamlined such that all of the RfCs for the TMB isomers
are presented together, and then one final RfC value is selected that applies to all three isomers.
SAB Comment G. 1-2: SAB noted that EPA incorrectly identified the appropriate effects for
maternal toxicity and the NOAEL values for decreased maternal weight gain in the External Peer
Review Draft TMB assessment Saillenfait et al. (2005) selected 100 ppm (492 mg/m3) for the
maternal NOAEL for 1,3,5-TMB with 300 ppm (1,476 mg/m3) as the maternal LOAEL based on
decreased maternal weight gain and food intake. In the External Peer Review Draft TMB
Assessment, the EPA set the maternal NOAEL at 300 ppm (1,476 mg/m3) and the maternal LOAEL
at 600 ppm (2,952 mg/m3) based on decreased corrected body weight gain and higher exposure
levels than Saillenfait etal. (2005). The SAB found that this is not a correct interpretation of a
maternal NOAEL for the Saillenfait etal. (2005) paper. Decreased corrected body weight gain was
measured only at one time point (C-section) 1 day after cessation of exposure. Statistically
significant decreased maternal weights were observed at gestational days (GDs) 13-21 when the
fetuses would be contributing far less to the mother's weight and at GDs 6-21 (entire treatment
period). Reduced maternal body weights correspond exactly with the statistically significant
decreased food consumption values recorded at GDs 6-13,13-21, and 6-21 (entire treatment
period). The SAB recommended that EPA use decreased maternal body weight gain data from
GDs 6-13 and 6-21 as the basis of the maternal endpoint POD and RfC rather than corrected
maternal weight gain data. If BMD modeling is unsuccessful, the SAB recommended that EPA use
the maternal NOAEL of 492 mg/m3 as the POD.
EPA Response G.l-2: EPA agrees with the SAB comments and has revised the RfC
derivations for 1,3,5-TMB. In the revised draft, EPA selected decreased maternal weight gain from
This document is a draft for review purposes only and does not constitute Agency policy.
A-38 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
Supplem en tal Information—Trim ethylbenzenes
GDs 6-21 as the basis for the maternal endpoint, and used a NOAEL of 497 mg/m3 (measured
concentration) as the basis for derivation of the RfC.
SAB Comment G. 1-3: SAB found that EPA incorrectly identified 2,974 mg/m3 as the NOAEL
for decreased male fetal weight. Saillenfait et al. (2005) identified the developmental NOAEL in the
study as 300 ppm (1,476 mg/m3) and the developmental LOAEL as 600 ppm (2,952 mg/m3) based
on decreased mean male fetal body weights. The SAB recommended using the NOAEL of
1,476 mg/m3 as the POD for derivation of a developmental endpoint RfC. The SAB also suggested
that EPA consider increasing the UFd from 3 to 10, to address the lack of neurodevelopmental
testing, in the derivation of the developmental RfC. The SAB noted that this approach may not fully
address neurological effects that serve as the basis for the other isomers. However, the revised
developmental endpoint RfC calculation will be based on a more appropriate POD and improve the
justification for using the extrapolation from the lower neurological-based RfC from 1,2,4-TMB.
EPA Response G.l-3: EPA used the correct NOAEL of 1,471 mg/m3 (measured
concentration) as the basis for derivation of the RfC for decreased male fetal weight. As stated
above (see EPA Response E.4-5 for details), EPA revised the RfC section for TMB isomers to cover
all three isomers simultaneously rather than have three separate RfC sections for each individual
isomer. This allows the whole TMB toxicity database to be considered holistically. As such, EPA
determined that a UFd of 3 was appropriate to account for the lack of a developmental neurotoxicity
study in the TMB toxicity database.
SAB Comment G.l-4: In addition to the above analysis and considerations, the SAB noted
the following minor errors in the description of the 1,3,5-TMB inhalation data: (1) in Table 2-12,
the female fetal body weight average for the 100 ppm (492 mg/m3) group should be 5.47 ± 0.21 and
not 5.74 ± 0.21 (it is correct in other tables of the document); (2) the level of significance for
decreased maternal body weight gain for the 600 ppm (2,952 mg/m3) group should have two (**)
and not one (*) asterisk to indicate p < 0.01; and (3) the table also states with a footnote (b) that
numbers of live fetuses were not explicitly reported. However, Saillenfait etal. f20051 did report
them in Table 3 of their manuscript The total numbers of fetuses were 297, 314, 282, 217, and 236,
for the control and exposure groups, respectively, and should be included in Tables 2-2 and 2-12 of
the draft TMB Review document.
EPA Response G.l-4: The minor errors in Tables 2-2 and 2-12 have been corrected; the
correct information is now presented in Table 2-3 in the unified RfC section.
Charge Question G.2: Please comment on whether EPA's approach to developing the RfC for
1,3,5-TMB is scientifically supported for the available data and clearly described.
SAB Comment G.2-1: The SAB acknowledged that the Agency's approach to developing the
overall RfC (based on neurological effects) for 1,3,5-TMB based on a structurally and toxicologically
related isomer is scientifically appropriate. However, the SAB recommended that the Agency
strengthen the justification for using this approach for 1,3,5-TMB by: (1) following the
recommendations provided above regarding recalculating the maternal- and developmental-based
This document is a draft for review purposes only and does not constitute Agency policy.
A-39 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
Supplem en tal Information—Trim ethylbenzenes
RfCs from Saillenfait et al. (2005): and (2) discussing the differences as well as similarities in
physical and toxicological parameters (i.e., Henry's Law constant and toxicokinetics) for 1,3,5-TMB
as compared with the other isomers.
EPA Response G.2-1: As noted above (EPA Responses F.2-1 and G.l-1), EPA has completely
restructured the RfC section for the TMB assessment This restructuring has, in a large part,
removed the necessity to set RfCs for one isomer as that for other data-poor isomers. In the new
structure, RfCs are derived for each isomer-endpoint combination, and then a single, overarching
RfC is selected for TMBs as a whole (this is detailed in Section 2.1.5 in the assessment). However,
following SAB's recommendations above, EPA has: (1) recalculated all of the maternal- and
developmental-based RfCs for 1,3,5-TMB; and (2) discussed the similarities and differences
between the physical and toxicokinetic properties for the individual isomers (see Section 1.1.1)
Charge Question H.l: The oral database for 1,2,4-TMB was considered inadequate for
derivation of an RfD. However¦, available evidence demonstrates similar qualitative profiles of
metabolism and patterns of parent compound distribution across exposure routes (i.e., oral and
inhalation). Furthermore, there is no evidence that would suggest the toxicity profiles would differ to
a substantial degree between oral and inhalation exposures. Therefore, route-to-route extrapolation,
from inhalation to oral using the modified Hissink et al. f2007) PBPKmodel was used to derive a
chronic oral RfD for 1,2,4-TMB. In order to perform the route-to-route extrapolation, an oral
component was added to the model, assuming a constant infusion rate into the liver. Specifically, in
the absence of isomer-specific information, an assumption was made that 100% of the ingested
1,2,4-TMB would be absorbed by constant infusion of the oral dose into the liver compartment. The
contribution of first-pass metabolism was also evaluated. Please comment on whether EPA's
conclusion that the oral database for 1,2,4-TMB is inadequate for derivation of an RfD is scientifically
supported and clearly described. Please comment on whether oral data are available to support the
derivation of an RfD for 1,2,4-TMB. If so, please identify these data.
SAB Comment H.l-1: The SAB agreed that the primary toxicological endpoints for
1,2,4-TMB (neurotoxicity, hematotoxicity) can be extrapolated across dose routes from the
inhalation data with the assistance of PBPK modeling. There is ample precedent with IRIS
assessments to use this approach to derive a reference value for a chemical with missing data by a
particular dose route.
EPA Response H.l-1: No response necessary.
SAB Comment H.1-2: The SAB noted that they were not aware of adequate repeat-dose
studies for 1,2,4-TMB via the oral dose route. The available acute exposure studies offer limited
support in developing an RfD. The SAB recognized that this represents a data gap and that one
potential way to fill this data gap is to use oral data for a closely related TMB isomer such as the
subchronic gavage toxicology data available for 1,3,5-TMB (Adenuga etal.. 2014: Koch Industries.
1995b). The SAB disagreed with EPA's decision to not use the Koch Industries f!995bl/Adenuga et
al. f 20141 study for derivation of an RfD due to the lack of neurotoxicity data. The SAB
This document is a draft for review purposes only and does not constitute Agency policy.
A-40 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
Supplem en tal Information—Trim ethylbenzenes
recommended that the EPA derive RfD(s) for endpoints observed in the oral 1,3,5-TMB study, such
as liver and kidney weight changes. The SAB commented that this would be consistent with EPA's
goal to derive RfD values for multiple endpoints (such as what was done with the RfC). The SAB
then stated that these RfDs could then be considered for extrapolation to other TMB isomers. The
SAB commented that the EPA should consider the appropriateness of applying a database UF to the
oral POD to compensate for the data gap of not having an oral neurotoxicity endpoint in the current
approach. Finally, the SAB noted that by comparing the RfD(s) generated from the oral studies and
from the extrapolation from the RfC through using route-to-route extrapolation, the EPA can
provide a clear explanation for why the use of the PBPK route-to-route-based RfD for 1,2,4-TMB
may be preferable to application of a database UF to an oral POD.
EPA Response H.l-2: Upon further consideration of the Adenuga et al. (2014) study, EPA
agreed with SAB that it was suitable for derivation of a candidate oral value for increased
monocytes. This is a hematological effect that is consistent with effects seen following inhalation
exposures to 1,2,4-TMB and 1,2,3-TMB. A full discussion of the appropriateness of this endpoint for
derivation of an RfD has been included in Section 2.2.1. However, the EPA further determined that
the changes in kidney and liver weight would not support RfD derivations, as no accompanying
histopathological changes were noted in these organs following examination. Given that organ
weight changes occurring in the absence of histopathological lesions or other evidence of clear
adversity may be compensatory or adaptive changes, the liver and kidney weight changes observed
in subchronic inhalation studies for 1,2,4-TMB and 1,2,3-TMB were similarly discounted; no RfD
values were derived for these endpoints. To support the decision to not consider the organ weight
changes as suitable for reference value derivations text was added in multiple places in the
assessment First, Section 1.2.5 (General Toxicity) was added to the Hazard Identification section to
discuss the observation of organ weight changes. Secondly, Sections 2.1.1 and 2.2.1 in the Dose-
Response Analysis section more thoroughly covered the Agency's rationale behind the
determination that these endpoints were not suitable for reference value derivations.
After consideration of the oral TMB toxicity data, and by extension the inhalation database
as well, EPA determined that the application of a 3-fold database UF was suitable to account for the
lack of an oral neurotoxicity or developmental neurotoxicity study. EPA's rationale for this decision
regarding the lack of developmental neurotoxicity study is the same as was used for the derivation
of the RfC for TMB isomers (see EPA Response E.4-5 for details). EPA determined that there was no
need to increase the UFd to 10-fold to account for the lack of an oral neurotoxicity study as the
derived RfCs for neurotoxicity and hematotoxicity endpoints were equal, indicating that RfDs
calculated for these endpoints might also be assumed to be equivalent However, in order to fully
explore this possibility, EPA used the available PBPK model to perform a route-to-route
extrapolation on the decreased pain sensitivity endpoint for 1,2,4-TMB. In doing so, EPA
subsequently derived an RfD of 1 x 10~2 mg/kg-day for decreased pain sensitivity, equal to the RfD
derived for decreased monocytes. As with the RfC derivations, this result indicates that some
This document is a draft for review purposes only and does not constitute Agency policy.
A-41 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
Supplem en tal Information—Trim ethylbenzenes
endpoints in the hematological system are equivalently as sensitive to exposure to TMB isomers as
endpoints in the nervous system. This determination is further supported by the derivation of an
RfD of 1 x 10"2 mg/kg-day for 1,2,4-TMB based on decreased clotting time via a route-to-route
extrapolation. Ultimately, EPA decided to select the RfD based on the route-to-route extrapolation
of the decreased pain sensitivity endpoint given the confidence in the PBPK model extrapolations
and that neurotoxicity endpoints are the most consistently observed effects in the TMB toxicity
database.
SAB Comment H.1-3: The SAB noted that there were limitations in the Koch Industries
study (primarily that it didn't involve neurotoxicity endpoints) and that use of the study would
involve an extrapolation across congeners. Presented with those limitations, the SAB determined
that the Koch Industries study does not provide a superior alternative to the PBPK approach for
dose route extrapolation that the EPA implemented. As discussed in SAB Comment 1 of Charge
Question H.l, the SAB noted that the Koch Industries study may provide a means to derive RfDs for
several additional endpoints (e.g., liver, kidney) for 1,3,5-TMB. The SAB recommended that EPA
consider such additional RfDs and whether they are potentially useful for 1,2,4-TMB based upon
extrapolation across congeners.
EPA Response H.l-3: EPA agrees that the Adenuga etal. f 20141 study does not provide a
clearly superior alternative to the route-to-route extrapolation that has been used to derive the RfD
for TMB isomers. However (as discussed above in EPA Response H.l-1), the EPA derived an RfD
from data on increased monocytes reported in Adenuga etal. (2014). and has compared this
isomer-route-specific RfD to the RfD derived from the route-to-route extrapolation. As thoroughly
discussed in Section 2.2.3, use of the monocyte data results in an RfD of 1 x 10~2 mg/kg-day,
compared to an RfD of 1 x 10~2 mg/kg-day for decreased pain sensitivity when using the route-to-
route approach. Ultimately, the EPA chose the RfD based on the route-to-route extrapolation given
the increased confidence in using the validated PBPK model to conduct the route-to-route
extrapolation and numerous lines of evidence indicating the similarities in the toxicological and
toxicokinetic properties of the TMB isomers.
Charge Question H.2: A route-to-route extrapolation from inhalation to oral exposure using
the modified Hissink et al. (2007) PBPK model has been used to derive an oral RfD for 1,2,4-TMB.
Please comment on whether the PBPK modeling been appropriately utilized and clearly described. Are
the model assumptions and parameters scientifically supported and clearly described? Are the
uncertainties in the model structure adequately characterized and discussed? Please comment on
whether this approach is scientifically supported and clearly described in the document.
SAB Comment H.2-1: The SAB noted that the EPA adapted the modified Hissink et al.
(2007) model for dose route extrapolation of internal dose by adding an oral delivery component
(continuous gastric infusion, instantaneous and complete absorption). The Hissink etal. (2007)
inhalation human model is a reasonable starting point as it simulated the available human
toxicokinetic data fairly well. The SAB concluded that, while the incorporation of the oral dose
This document is a draft for review purposes only and does not constitute Agency policy.
A-42 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
Supplem en tal Information—Trim ethylbenzenes
route is simplistic, it is acceptable for the current purposes in that the dose metric used for dose-
response modeling (parent compound average weekly venous concentration) is not sensitive to
peaks and valleys of a more normal oral intake pattern. A constant infusion averages out the
exposure over the course of the day, thus creating an average venous concentration that is
compatible with the dose metric without further calculation. Overall, the SAB determined that the
modified Hissinketal. f20071 model adapted for the oral route is likely to adequately predict
human oral exposures and be useful for dose-response modeling and the derivation of the RfD.
EPA Response H.2-1: Although the SAB concluded that an assumption of constant infusion
was acceptable, albeit simplistic, for the route-to-route extrapolation, EPA, upon further
consideration of the data, implemented a more realistic pattern of human oral exposure. In this
new scenario, ingestion was simulated as an idealized pattern of six events, each lasting
30 minutes. Twenty-five percent of the total daily dose was assumed to be ingested at each of three
events beginning at 7 am, 12 pm (noon), and 6 pm (total of 75%). Ten percent of the daily dose was
assumed to be ingested at events beginning at 10 am and 3 pm (total of 20%). The final 5% was
assumed to be ingested in an event beginning at 10 pm.
Charge Question H.3: Please comment on the rationale for the selection of the uncertainty
factors (UFs) applied to the POD for the derivation of the RfD for 1,2,4-TMB. Are the UFs appropriate
based on the recommendations described in Section 4.4.5 of A Review of the Reference Dose and
Reference Concentration Processes; and clearly described? If changes to the selected UFs are proposed
please identify and provide scientific support for the proposed changes.
SAB Comment H.3-1: The SAB agreed with the UFs selected in the development of the oral
RfD for 1,2,4-TMB. As discussed in the SAB Comment 1 of Charge Question H.2, the oral RfD for
1,2,4-TMB was derived by incorporating an oral intake component into the PBPK model for
1,2,4-TMB to obtain a human equivalent oral dose POD and then used the same UFs for the oral RfD
as were used in the development of the inhalation RfC. Given that the oral RfD was based upon the
same endpoint and derived from the same study as the RfC, the SAB agreed that it is logical to use
the same UFs. Thus, the comments and recommendations regarding UFs for the RfC derivations
(Charge Questions E.4 and F.4) are applicable to this charge question as well.
EPA Response H.3-1: No response necessary.
SAB Comment H.3-2: The SAB discussed whether there is additional uncertainty associated
with incorporation of the oral intake component in the PBPK model, and specifically regarding
assumptions made with that component regarding oral absorption of 1,2,4-TMB and first-pass
metabolism. Unlike modeling of internal concentrations from inhalation exposure that can be
verified with existing experimental data, there are no data with which to assess model predictions
of internal doses following oral 1,2,4-TMB exposures. The SAB ultimately did not consider this
additional uncertainty sufficient to increase the composite UF for the oral RfD, largely because the
nature of the uncertainty (possible lower absorption by the oral route) would add extra health
This document is a draft for review purposes only and does not constitute Agency policy.
A-43 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
Supplem en tal Information—Trim ethylbenzenes
protection. The SAB recommended that the potential uncertainties associated with oral
bioavailability of 1,2,4-TMB be discussed more clearly in the document
EPA Response H.3-2: A discussion of the uncertainty surrounding the assumption of 100%
bioavailability of ingested TMB isomers has been added to Section 2.2.4.
Charge Question 1.1: The oral database for 1,2,3-TMB was considered to be inadequate for
derivation of an RfD. Based on the similarities in chemical properties, toxicokinetics, and toxicity
profiles between the 1,2,4-TMB and 1,2,3-TMB isomers, EPA concluded that there was sufficient
evidence to support adopting the 1,2,4-TMB RfD as the RfD for 1,2,3-TMB. Please comment on whether
EPA's conclusion that the oral database for 1,2,3-TMB is inadequate for derivation of an RfD is
scientifically supported and clearly described. Please comment on whether oral data are available to
support the derivation of an RfD for 1,2,3-TMB. If so, please identify these data. Please comment on
whether EPA's approach to developing the RfD for 1,2,3-TMB is scientifically supported and clearly
described.
SAB Comment 1.1-1: The SAB was not aware of adequate repeat-dose studies for 1,2,3-TMB
via the oral dose route. The available acute exposure studies offer limited support in developing an
RfD. The SAB agreed that the primary toxicological endpoints used for 1,2,4-TMB (neurotoxicity,
hematotoxicity) and extrapolated across dose routes from the inhalation data with the assistance of
PBPK modeling are appropriate for 1,2,3-TMB. There is ample precedent within the IRIS system for
this approach to derive a reference value for a chemical with missing data by a particular dose
route. The SAB noted that the Agency appropriately uses the same rationale to derive the RfD for
1,2,4-TMB.
EPA Response 1.1-1: It should be noted that, as with the RfC section, the individual isomer
RfD sections have been combined into a unified RfD section for all of the isomers. As such, given
SAB comments on both the 1,2,4-TMB and 1,3,5-TMB RfD sections, the unified RfD section covers
extensive discussion and quantitation of RfDs based on increased monocytes (1,3,5-TMB oral-
specific data) and decreased pain sensitivity (1,2,4-TMB route-to-route extrapolation), including
the ultimate adoption of the route-to-route-derived RfD as the RfD for TMBs. Thus, while an
explicit discussion of adoption of 1,2,4-TMB's RfD as the RfD for 1,2,3-TMB no longer is included in
the document, the discussion regarding the ultimate adoption of 1,2,4-TMB's RfD as the RfD for all
isomers still covers the issues identified by SAB above.
Charge Question J.l: The oral database for 1,3,5-TMB was considered to be inadequate for
derivation of an RfD. EPA concluded that given the similarities in the chemical properties,
toxicokinetics, and toxicity profiles between the two isomers, there was sufficient evidence to support
adopting the RfD for 1,2,4-TMB as the RfD for 1,3,5-TMB. Please comment on whether EPA's
conclusion that the oral database for 1,3,5-TMB is inadequate for derivation of an RfD is scientifically
supported and clearly described. Please comment on whether oral data are available to support the
derivation of an RfD for 1,3,5-TMB. If so, please identify these data.
This document is a draft for review purposes only and does not constitute Agency policy.
A-44 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
Supplem en tal Information—Trim ethylbenzenes
SAB Comment T. 1-1: The SAB agreed with the EPA's approach to extrapolating the RfD of
1,2,4-TMB to 1,3,5-TMB. However, the SAB was aware of an isomer-specific study fKoch Industries.
1995b) and the recently released data on 1,3,5-TMB (Adenuga etal.. 20141 provided by public
commenters.
EPA Response 1.1-1: EPA incorporated data from Adenuga et al. f20141 in the RfD
derivation section as outlined below.
SAB Comment 1.1-2: The SAB commented that the Koch Industries f!995bl study was the
only isomer-specific and route-specific study available in the peer-reviewed literature for oral
exposure to 1,3,5-TMB when the TMB assessment was drafted in 2013. Although EPA's rationale
for not using this study for RfD derivation is clearly described (i.e., it did not assess the potential for
neurological effects and "presented limited toxicological information"), the SAB disagreed and
considered the Koch Industries Q995bl study suitable for development of one or more candidate
oral values for 1,3,5-TMB.
EPA Response 1.1-2: The Koch Industries (1995bl/Adenuga etal. f 20141 study has been
used in the current draft to derive an RfD based on increased monocytes.
SAB Comment 1.1-3: The SAB found that the Koch Industries study of 1,3,5-TMB toxicity
after subchronic (90-day) gavage treatment was consistent with good laboratory practices and
requirements and, when submitted for an EPA Office of Water test rule, was peer-reviewed by three
senior scientists fVersar. 20131. Although the study does not include neurological endpoints, it
does provide information on toxicity to other organs such as liver and kidney. The SAB concluded
that this study is suitable for providing candidate oral values for one or more endpoints in the same
way that, for example, candidate values based upon a variety of endpoints were developed and
presented for 1,2,4-TMB (see Table 2-4 of the draft TMB Toxicological Review).
EPA Response 1.1-3: As noted above, the Koch Industries fl995bl/Adenuga etal. f 20141
study has been used to derive an RfD for increased monocytes in the current draft. One note of
clarification, the Koch Industries study was not peer-reviewed when submitted for an EPA Office of
Water test rule, but was peer-reviewed in order to include it in the IRIS Toxicological Review of
Trimethylbenzenes.
SAB Comment 1.1-4: The SAB noted that, given the importance of neurotoxicity as a critical
endpoint for inhalation exposure to TMB isomers, there should be confidence that any value
selected as the RfD for 1,3-5-TMB is adequately protective of this type of effect In order to produce
an RfD protective of neurotoxicity using PODs from the Koch Industries study, a large UFd (e.g., 10)
could be used to account for the absence of isomer- and route-specific neurotoxicity data. However,
the SAB concluded that there is stronger scientific support for use of a PBPK-extrapolated RfD for
1,2,4-TMB based on a neurotoxic endpoint as the overall RfD for 1,3,5-TMB. Thus, while the SAB
recommended use of the Koch Industries data and Adenuga etal. (20141 to develop candidate oral
values for comparison purposes, it agrees with the overall RfD for 1,3,5-TMB as proposed by EPA.
EPA Response 1.1-4: No response necessary.
This document is a draft for review purposes only and does not constitute Agency policy.
A-45 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
Supplem en tal Information—Trim ethylbenzenes
Charge Question K.1: The draft Toxicological Review ofTrimethylbenzenes did not conduct
a quantitative cancer assessment for any isomer due to the lack of available studies. Please comment
on whether data are available to support the derivation of a quantitative cancer risk estimate.
SAB Comment K.l-1: The SAB found that the evidence for carcinogenicity of TMBs is
limited and that this fact was well presented by the EPA in the draft toxicological review.
EPA Response K.l-1: No response necessary.
SAB Comment K. 1-2: The SAB agreed with the Agency that TMBs do not appear to be
genotoxic when assessed in a standard battery of genotoxicity assays. The one exception was
1,2,3-TMB in the Ames assay in the absence of S9. The SAB concluded that the significance of the
finding was uncertain because it was not clear what mechanism could lead to such a response.
EPA Response K.l-2: No response necessary.
SAB Comment K. 1-3: The SAB was not aware of any human studies on carcinogenicity of
TMBs, but noted that a number of biomarker studies and their association with cancer of various
sites have been published. These biomarker studies should be reviewed and included. Some
examples are: (1) solid-phase microextraction, mass spectrometry and metabolomic approaches
for detection of potential urinary cancer biomarkers—a powerful strategy for breast cancer
diagnosis fSilva etal.. 20121: (2) investigation of urinary volatile organic metabolites as potential
cancer biomarkers by solid-phase microextraction in combination with gas chromatography-mass
spectrometry fSilva etal.. 20111: and (3) cellular responses after exposure of lung cell cultures to
secondary organic aerosol particles fGaschen etal.. 20101.
EPA Response K.l-3: Information gleaned from studies on biomarkers of exposure and
their association with cancers at various sites in humans has been added the Carcinogenicity
section (Section 1.2.6) of the Hazard Identification section where applicable.
SAB Comment K. 1-4: Based upon the deficiencies of the Maltoni etal. Q9971 study, the lack
of bioassays with 1,2,3-TMB and 1,3,5-TMB, and the lack of human studies, the SAB agreed that the
EPA could not conduct a quantitative cancer assessment for any isomer due to the lack of
appropriate studies.
EPA Response K.l-4: No response necessary.
Additional SAB Recommendations
1. Candidate Reference Values
SAB Comment AR. 1-1: The SAB noted that Section 7.6 of the Preamble (External Peer
Review draft version) describes how IRIS assessments derive candidate values for each suitable
data set and effect that is credibly associated with an agent. These results are arrayed, using
common dose metrics, to show where effects occur across a range of exposures using guidance on
methods to derive RfCs and RfDs. The assessment process develops an organ- or system-specific
reference value for each organ or system affected by the agent and selects an overall RfD and an
overall RfC for the agent to represent lifetime human exposure levels where effects are not
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
35
36
37
Supplem en tal Information—Trim ethylbenzenes
anticipated to occur. Providing these organ/system-specific reference values, IRIS assessments
may facilitate subsequent risk assessments that consider the combined effect of multiple agents
acting at a common site or through common mechanisms.
EPA Response AR.1-1: No response necessary.
SAB Comment AR.1-2: The SAB encountered an issue where further clarification by EPA is
strongly encouraged. Interest by the EPA in developing PODs and RfCs/RfDs for multiple endpoints
in new IRIS profiles is noted. As shown in this toxicological review, one of the uses of RfCs/RfDs for
various endpoints is as candidates for selection as the overall toxicity value. The overall toxicity
value is one that is intended to be protective of toxicity of all types, and this is taken into
consideration when selecting the UFd. Another use of these RfCs/RfDs is to better understand the
effects of combined chemical exposures. Risks from combined or cumulative exposures to
chemicals is generally of greatest concern when the chemicals affect the same targets organs. While
an overall RfC or RfD is based upon one effect chosen as the critical effect, that chemical may
produce other types of toxicity at doses that are only marginally higher than the selected overall
toxicity value. To illustrate the problem, consider the situation in which individuals are exposed to
three chemicals, each with an RfC based upon a different endpoint, but all have the potential to
affect the liver. For the risk assessor, the combined effect of the three chemicals on the liver may be
greater concern than the effects of the individual chemicals on other organ/systems. In order to
evaluate the risk of liver injury from combined exposure, the risk assessor needs a liver RfC for each
compound. Conceivably, this information could come from RfCs for the chemicals, if available for
the liver, but there is a difference in the way that an RfC for this use would be developed versus an
RfC suitable for selection as the overall RfC. The difference is in the way that the UFd is selected—
on one hand to ensure that the RfC is protective against all forms of toxicity and on the other that it
is reliably protective of toxicity to a specific target organ. Conceivably, the UFd values selected for
those two purposes, and the resulting RfC/RfD values, could be quite different The SAB was
unaware of any discussion of this issue by EPA or clear description of how organ/system-specific
RfC/RfD values are to be developed and used. As the IRIS process moves forward, it will be
important to provide much greater clarity on this subject.
EPA Response AR.1-2: EPA agrees that as the IRIS Program moves forward, the process by
which organ/system-specific RfCs/RfDs are derived must be clearly defined and presented
transparently to the public. In the current assessment, however, the RfCs/RfDs were derived via
the application of a composite UF that took into account database uncertainties (UFd = 3 for lack of
developmental neurotoxicity information). Calculation of RfCs/RfDs associated with systems that
are likely not affected by the lack of additional developmental neurotoxicity information could use a
composite UF = 100 (UFa = 3, UFh = 10, UFs = 3, UFl = 1, UFd = 1 [hematological, respiratory, or
maternal endpoint]) or UF = 30 (UFA = 3, UFH = 10, UFS = 1, UFL = 1, UFD = 1 [developmental
endpoints]).
This document is a draft for review purposes only and does not constitute Agency policy.
A-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
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
2. Sensitive Lifestages and Vulnerable Populations
SAB Comment AR.2-1: The draft TMB assessment provided only one paragraph on this
subject While the SAB found that it correctly identified various types of immaturity (metabolism,
renal clearance) as potentially leading to greater vulnerability in early life, the Panel felt that this
section could provide a better outline of the kinds of information needed to understand the
potential vulnerabilities in early life, including key aspects of TMB mode of action and key
developmental features.
EPA Response AR.2-1: This section was expanded according to the specific comments that
SAB provided below.
SAB Comment AR.2-2: Regarding mode of action, the SAB noted that it is important to
know: (1) whether it is the parent compound or metabolites (or both) that contribute to toxic
effect; (2) which metabolic systems are responsible for removing the parent compound and
creating important metabolites; and (3) the role of distributional phenomena (e.g., uptake into
brain; partitioning into fat) and other clearance mechanisms in determining chemical fate and
access to target sites. Based upon the available mode-of-action information, the developmental
factors that may influence toxicokinetics can be discussed in this section. For TMBs, the draft
document assumes that the parent compound is responsible for toxicity with modeling assuming
that a saturable Phase I oxidative cytochrome P450 (CYP450) process is responsible for decreasing
parent compound levels in venous blood. This section should state whether it is known which
CYP450(s) are responsible for TMB saturable metabolism, as different CYP450s have different
developmental patterns. Analogy may be drawn with other alkylbenzenes that do have
toxicokinetic modeling data in early life such as toluene. Toluene has already been referred to in
the mode-of-action section of the document; it is also neurotoxic and its mode of action is based
upon parent compound, with the level getting to the brain determined by saturable CYP450
metabolism. If the EPA determines these parallels to provide a useful analogy, then early life
modeling papers for toluene by Pelekis etal. (2001) and Nong etal. (2006) may be useful for
describing the degree of toxicokinetic uncertainty presented by early lifestage exposure to TMBs.
EPA Response AR.2-2: A more detailed discussion of what is known regarding the mode of
action for TMB isomers and whether information exists on what CYP450 isozyme is responsible for
metabolizing parent compound has been added to Section 1.3.3 (Susceptible Populations and
Lifestages). Information from early-life modeling on toluene was also incorporated into the
discussion to support the conclusion that early life may be a susceptible lifestage for the neurotoxic
effects of TMB exposure.
SAB Comment AR.2-3: The SAB concluded that some discussion was warranted concerning
what is known about early life vulnerability to aromatic solvent neurotoxicity. Several studies are
available suggesting a vulnerable window of brain development in mice to the neurotoxic effects of
toluene fWin-Shwe etal.. 2012: Win-Shwe etal.. 20101. The SAB recommended that the EPA
evaluate this evidence relative to other developmental neurotoxicity studies that may be available
This document is a draft for review purposes only and does not constitute Agency policy.
A-48 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
Supplem en tal Information—Trim ethylbenzenes
for toluene and other related alkylbenzenes to determine whether this data gap represents a large
uncertainty.
EPA Response AR.2-3: A discussion of the possible developmental neurotoxicity of toluene
as a surrogate for TMB was added to Section 1.3.3 (Susceptible Populations and Lifestages) to
support the decision that early life is a window of susceptibility for the neurotoxic effects of TMB
exposure.
SAB Comment AR.2-4: The SAB noted that this section should conclude with a statement as
to whether any specific data exist for TMBs that would show the extent of early life vulnerability
based upon toxicokinetic and toxicodynamics considerations and the degree to which such data for
related alkylbenzenes help to fill these data gaps.
EPA Response AR.2-4: A concluding statement was added to this section.
3. Developing Subchronic RfCs and RfDs
SAB Comment AR.3-1: The SAB noted that the EPA and other environmental regulatory
agencies are frequently required to address the risks associated with exposures lasting less than a
lifetime. Because the toxic endpoint(s) of concern for a given chemical, as well as threshold doses
or concentrations for toxicity, can change with exposure duration, the toxicity value used in risk
assessment should be matched to the extent possible to the length of exposure associated with the
scenario of interest Recognizing the need for toxicity values for less-than-lifetime exposures, the
EPA Risk Assessment Forum recommended that the Agency develop such values and incorporate
them into the IRIS database (U.S. EPA. 20021.
EPA Response AR.3-1: No response necessary.
SAB Comment AR.3-2: In the case of the TMBs, the SAB noted that the principal studies
used to create the proposed RfCs and RfDs are all subchronic in duration, and the analysis needed
to support a robust set of subchronic toxicity values has, in effect, already been done for these
chemicals. The SAB acknowledged that the derivation of subchronic RfCs and RfDs may not always
be appropriate. However, the toxic endpoints and dose-response relationships for the TMBs in the
draft report are clearly relevant for subchronic exposure, and the same PODs and the same UFs—
except UFs, which is used to generate a chronic toxicity value from subchronic study data—would
apply to the development of a set of subchronic RfCs and RfDs.
EPA Response AR.3-2: No response necessary.
SAB Comment AR.3-3: Given the potential usefulness of these toxicity values for risk
assessment, the importance of having the values available on IRIS, and the very small amount of
additional work required to add them to the TMB assessment, the SAB suggested that the EPA
consider including subchronic RfCs and RfDs for 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB. These
values would be calculated using the same inputs as for the chronic toxicity values, but omitting the
UFs. The SAB anticipated that incorporation of these values would require minimal edits to existing
tables and text.
This document is a draft for review purposes only and does not constitute Agency policy.
A-49 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
EPA Response AR.3-3: EPA has provided a set of subchronic RfCs and RfDs (both the
candidate and final values) for the TMB isomers in Sections 2.1.8 and 2.2.6 (respectively).
This document is a draft for review purposes only and does not constitute Agency policy.
A-50 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
APPENDIX B. HEALTH ASSESSMENTS AND
REGULATORY LIMITS BY OTHER NATIONAL AND
INTERNATIONAL HEALTH AGENCIES
1 Table B-l. Other national and international health agency assessments for
2 trimethylbenzenes (TMBs)
Agency
Toxicity value
National Institute for
Occupational Safety and Health
(NIOSH, 1992, 1988)
Recommended Exposure Limit (REL) for TMBs: 25 ppm (123 mg/m3) time-
weighted average (TWA) for up to a 10-hr workday and a 40-hr work week,
based on the risk of skin irritation, central nervous system (CNS) depression,
and respiratory failure (Battig et al., 1956)
National Advisory Committee for
Acute Exposure Guideline Levels
(AEGLs) 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
Rvdzvriski, 1996)
AEGL-2 (disabling): -460 ppm (2,300 mg/m3) to 150 ppm (740 mg/m3) (10 min
to 8 hrs, respectively) (Gage, 1970)
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
Supplem en tal Information—Trim ethylbenzenes
APPENDIX C. INFORMATION IN SUPPORT OF
HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS
C.l. TOXICOKINETICS
There has been a significant amount of research conducted on the toxicokinetics of
1.2.4-trimethylbenzene (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 (Swiercz
etal.. 2006: Tsuiimoto etal.. 2005: Swiercz etal.. 2003: Swiercz etal.. 2002: Tsuiino etal.. 2002:
Tsuiimoto etal.. 2000: Eide andZahlsen. 1996: Zahlsen etal.. 1990: Huo etal.. 1989: Dahl etal..
1988: Mikulski and Wiglusz. 19751 and volunteers (Swiercz etal.. 2016: Tanasik etal.. 2008: Tones et
al.. 2006: Tarnbergetal.. 1997a: Tarnbergetal.. 1997b: Kostrzewski etal.. 1997: Tarnbergetal..
1996: Kostrewski and Wiaderna-Brvcht. 1995: Fukavaetal.. 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 C-47-C-66 in Appendices C.6-C.8.
C.l.l. 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]) (Tarnberg etal..
1998.1997a: Tarnbergetal.. 19961. Kostrzewski etal. (19971 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, 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 standard devation [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. 19991 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
37
Supplem en tal Information—Trim ethylbenzenes
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).
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 fSwiercz etal..
2003: Eide and Zahlsen. 1996: Zahlsen etal.. 1990). Zahlsen etal. (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 etal. f2003I Using a four-comparment toxicokinetic
model, Yoshida (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 |J.g/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 (59.1 [56.9-61.3], 66.5 [63.7-69.3], and 43.0 [40.8-45.2], 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 veiy similar for all three isomers: 1,2,4-TMB (59.1 versus 57.7), 1,2,3-TMB
(66.5 versus 62.6), and 1,3,5-TMB (43.0 versus 55.7) (Meulenberg and Viiverberg. 2000). This
further indicates that patterns of absorption would be similar across species.
C.1.2. 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
Viiverberg. 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. TMBs (unspecified isomer) have
also been detected in cord blood, and can therefore 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) fHuo et
al.. 1989). Following inhalation exposures, 1,2,4-TMB, 1,2,3-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.. 2016: Swiercz etal.. 2006: Swiercz etal.. 2003: Eide and
Zahlsen. 1996). 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,
1,2,3-TMB, and 1,3,5-TMB and liver concentrations of 1,2,4-TMB were observed to be significantly
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
Supplem en tal Information—Trim ethylbenzenes
lower in repeatedly exposed animals versus animals exposed only once to higher concentrations
fSwiercz etal.. 2016: Swiercz etal.. 2006: Swiercz etal.. 2003: Swiercz etal.. 20021. 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. However, kidney
concentrations of 1,2,3-TMB were observed to be higher in repeatedly exposed animals versus
those exposed only once at low and medium doses, but not high doses fSwiercz etal.. 20161. 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 cytochrome P450 (CYP450)
enzyme induction in the livers, kidneys, and lungs of rats exposed to 1,200 mg/kg-day 1,3,5-TMB
for 3 days fPvvkko. 19801.
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 etal..
20031. Zahlsen etal. (19901 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). The only studies to investigate distribution following dermal exposure
utilized kerosene as the test agent In one study, 1,2,4-TMB preferentially distributed to the
kidneys fTsuiino etal.. 20021. 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.
Other studies simply reported that 1,2,4-TMB was detected in blood following dermal exposure to
kerosene (Kimlira etal.. 1991: Kimura etal.. 19881.
C.1.3. 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 flarnberg etal.. 1997b).
3,4-Dimethylhippuric acid (DMHA) comprised 82% of the DMHAs 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 flarnberg etal.. 1997b). 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 that it
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
Supplem en tal Information—Trim ethylbenzenes
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 flarnberg et
al.. 1997bl. 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 fKostrzewski etal.. 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
(lones etal.. 2006: Fukavaetal.. 1994: Ichibaetal.. 1992).
Following oral exposures in animals, the quantitative metabolic profiles of the three
isomers appears to differ. Mikulski and Wiglusz f!9751 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 DMHAs (23.9 ± 2.3 and 59.1 ± 5.2% total dose, respectively). DMHA
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), and 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 (TMPs), or dimethylbenzoic acids (DMBAs). Huo etal. (1989) reported that the
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 the administered dose
( 100 g/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 fLaham and Potvin. 1989: Cerf
etal.. 1980). 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 (Laham and Potvin. 1989). 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
(~59% 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) (Bakke
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
Supplem en tal Information—Trim ethylbenzenes
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 fHuo etal.. 19891. 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 hindered metabolites fTsuiimoto
etal.. 2005: Huo etal.. 19891. The proposed metabolic schemes for 1,2,4-TMB, 1,2,3-TMB, and
1,3,5-TMB are shown in Figures C-l, C-2, and C-3, respectively.
2,3,5 -trimethylphenol
ch3
HO^ ^CH3
2,3,6-trimethylphenol
CHs
tt
2,4-dimethyl-
benzyl
alcohol
3,4-dimethyl-
benzyl
alcohol
2,4,5-trimethylphenol
CH^OH
V
1,2,4-trimethyl
benzene
C HPH
2,5-dimethyl
benzyl alcohol
CHoR
CHoR
2,4-dimethyl-
benzyl
mercapturic acid
2,4-dimethyl M
benzoic acid
CH3
2,5 -dimethylbenzyl
.OH mercapturic acid
2,4-dimethyl-
hippuric acid
2,5-dimethyl
benzoic acid
Ok Q
3,4-dimethyl-
benzyl
mercapturic acid
3,4-dimethyl
benzoic acid
1 1
-,3 2,5-dimethyl-
hippuric acid
HO O
OH
3,4-dimethyl-
hippuric acid
Figure C-l. Metabolic scheme for 1,2,4-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
C-5 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
3,4,5-trimethyl-
phenol
1,2,3-trimethyl-
benzene
2,3,4-trimethyl-
phenol
CH-,R
CHtOH
ch2r
ch2oh
2,6-dimethyl-
mercapturic acid
2,6-dimethyl-
benzoic acid
2,6-dimethyl-
benzyl alcohol
CH,
2,3-dimethyl-
benzyl alcohol
2,3-dimethyl-
mercapturic acid
2,6-dimethyl-
hippuric acid
H
0„V.
2,3-dime thy 1-
benzoic acid
CH;,
2,3-dimethyl-
^>^rH hippuric acid
Figure C-2. Metabolic scheme for 1,2,3-TMB.
CH,
'CH,
I J,5-trimeth)l-
lien/coc
CH,
H,C
CH,
OH
2.
triirKttnlpIicnoI
CH
CH,
h3c
¦ch2oh
3.5-dimetli\ Ihcn/vl
akvhol
HjC
3,5 -dcmel hy I bcnzy 1
mcrcapturic acid
Jk OH
OH
3.5-dimeth> Ilvn/oic
arid
HN
3,5-dimcthylhippuric ]
acid
O OH
Figure C-3. Metabolic scheme for 1,3,5-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
Supplem en tal Information—Trim ethylbenzenes
C.1.4. 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 minutes,
21 ± 5 minutes, 3.6 ± 1.1 hours), 1,2,3-TMB (1.5 ± 0.9 minutes, 24 ± 9 minutes, 4.7 ± 1.6 hours), and
1,3,5-TMB (1.7 ± 0.8 minutes, 27 ± 5 minutes, 4.9 ± 1.4 hours) (larnberg et al.. 19961. 1,3,5-TMB
had a higher total blood clearance value compared with 1,2,4-TMB or 1,2,3-TMB (0.97 ± 0.06 versus
0.68 ± 0.13 or 0.63 ± 0.13 L/hour/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 versus
87 ± 27 and 78 ± 22 hours, respectively). Urinary excretion of unchanged parent compound was
extremely low (<0.002%) in humans (N = 6-10, male) for all three isomers flanasik etal.. 2008:
larnberg 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 versus 3.8-5.8 and
4.8-8.1 hours, respectively) (larnbergetal.. 1997b).
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
(larnberg etal.. 1996) and elimination of 1,3,5-TMB via breath was biphasic with an initial half-life
of 60 minutes, and a terminal half-life of 600 minutes (lones etal.. 2006). Following exposure of
rats to 25 ppm (123 mg/m3) 1,2,4-TMB, 1,2,3-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) or 1,2,3-TMB (3.1 hours) fSwiercz etal.. 2016: Swiercz 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 (4.1 hours) or 1,2,3-TMB (5.3 hours).
Following repeated-dose experiments (4 weeks), the terminal half-lives of elimination of TMB
isomers in venous blood were similar for 1,2,4-TMB and 1,2,3-TMB (9.9 and 8.0 hours,
respectively), but larger than that of 1,3,5-TMB (4.6) fSwiercz etal.. 2016: Swiercz etal.. 2006:
Swiercz etal.. 2003: Swiercz etal.. 20021.
C.2. PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELS
C.2.1. Summary of Available Physiologically Based Pharmacokinetic (PBPK) Models for
1,2,4-TMB
larnberg and lohanson (1999)
larnberg and lohanson (1999) described 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
This document is a draft for review purposes only and does not constitute Agency policy.
C-7 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 distribution) for the metabolite, 3,4-DMHA (see Figure C-4). The lung compartment includes lung
2 tissue and arterial blood. Excretion of parent compound is assumed to occur solely by ventilation.
3 As 1,2,4-TMB has a pronounced affinity to adipose tissue, a separate compartment for fat is
4 incorporated into the model. Remaining non-metabolizing compartments are rapidly perfused
5 tissues, comprising the brain, kidneys, muscles, and skin.
6
* Q ,
^alv
alv
alv
art
rap
cu
/K,
other
metabolites
I 3,4-dimethyl |
I hippuric acid |
Rapidly perfused
tissues
Working muscles
Adipose tissue
Liver
Resting muscles
Lungs and arterial
blood
urine urine
8 C = concentration of 1,2,4-TMB; Cair = concentration in ambient air; Can = concentration in arterial blood;
9 Cven = concentration in venous blood; Qaiv = alveolar ventilation; Qco = cardiac output; Qi = blood flow to
10 compartment i (where i = rap = rapidly perfused tissues; f = adipose tissue; w = working muscles,
11 r = resting muscles, h = liver); Vmax = maximum rate of metabolism, pathway I; Km = Michaelis-Menten
12 constant for metabolic pathway I; CL1 = intrinsic hepatic clearance of metabolic pathway II; ke = excretion
13 rate constant of 3,4-DMHA.
14
15 Source: Jarnberg and Johanson (1999).
16 Figure C-4. Physiologically based toxicokinetic model for 1,2,4-TMB in
17 humans.
This document is a draft for review purposes only and does not constitute Agency policy.
C-8 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Because previous experimental data were gathered during exercise flarnberg etal.. 1997a:
2 larnberg etal.. 19961. the muscle compartment was divided into two equally large compartments,
3 resting and working muscles. Two elimination pathways (a saturable Michaelis-Menten pathway
4 for all metabolites other than 2,4-DMHA [pathway I] and a first-order pathway [pathway II] for
5 formation of 3,4-DMHA) from the hepatic compartment were included. Metabolism was assumed
6 to occur only in the liver compartment. Tissue:blood partition coefficients of 1,2,4-TMB were
7 calculated from experimentally determined blood:air, water:air, and olive oil:air partition
8 coefficients (larnberg and lohanson. 19951 (Table C-l).
9 Table C-l. Measured and calculated partition coefficients for TMB isomers at
10 37°C
Substance
Measured values3
Calculated values
P saline:air
N =42
P oikair
N = 25
Human P blood:air
N = 39
Human P bloodiair^
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
11
12 aMean values and 95% confidence interval (CI).
13 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
14 the relative content of fat in blood (Fiserova-Bergerova, 1983).
15
16 Source: Jarnberg and Johanson (1995).
17
18 The model was used to investigate how various factors (work load, exposure level,
19 fluctuating exposure) influence potential biomarkers of exposure (end-of-shift and prior-to-shift
20 concentrations of parent compound in blood and 3,4-DMHA in urine). Biomarker levels estimated
21 at end-of-shift remained fairly constant during the week, whereas biomarker levels prior-to-shift
22 gradually increase throughout the week. This indicates that end-of-shift values represent the same
23 day's exposures, whereas prior-to-shift values reflect cumulative exposure during the entire work
24 week. Increased work load increased uptake of 1,2,4-TMB. For example, a workload of 150 W over
25 an exposure period of 8 hours increased the level of 1,2,4-TMB in the blood more than 2-fold,
26 compared to levels of 1,2,4-TMB in the blood after an 8-hour exposure at rest Simulated 8-hour
27 exposures at air levels of 0-100 ppm (0-492 mg/m3) shows that overall metabolism is saturable,
28 and that the metabolic pathway yielding 3,4-dimethylbenzene becomes more important as
29 exposure concentrations increase.
30 Previously performed experimental human exposures to 1,2,4-TMB were used to estimate
31 the metabolic parameters and alveolar ventilation flarnberg etal.. 1997a: larnberg etal.. 19961.
32 Individual simulated arterial blood concentrations and exhalation rates of 1,2,4-TMB, as well as the
This document is a draft for review purposes only and does not constitute Agency policy.
C-9 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 urinary excretion rate of 3,4-DMHA, were simultaneously adjusted to the experimentally obtained
2 values by varying the alveolar ventilation at rest One individual's compound-specific and
3 physiological parameters were then used for subsequent model predictions (Table C-2).
4 Table C-2. PBPK model parameters for 1,2,4-TMB toxicokinetics in humans
5 using the larnberg and lohanson (1999) model structure
Parameters
Rest
Both3
50 W
Body height (m)
1.78
Body weight (kg)
75.5
Vmax (nmol/min)
3.49
Km (nM)
4.35
CL1 (L/min)
0.149
Elimination rate constant (min-1)
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
Working muscles
0.55
Resting muscles
0.55
Partition coefficients
Blood:air
59
Fat:blood
125
Liver:blood
5
Rapidly perfused tissues:blood
5
Muscle:blood
5
6
7 Parameters used for both working and resting conditions.
8
9 Source: Jarnberg and Johanson (1999).
10
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
Supplem en tal Information—Trim ethylbenzenes
While based on the published results, the larnberg and lohanson (1999) model appears to
provide a good description of 1,2,4-TMB kinetics in humans, the model code could not be obtained
from the authors. Based on previous experience with other PBPK models, the U.S. Environmental
Protection Agency (EPA) has determined that attempting to reproduce (and thereby validate) a
model based only on the published description is nearly impossible. Therefore, because the model
code is not available, this model is not considered further in the Integrated Risk Information System
(IRIS) TMB Assessment.
Emond and Krishnan f2006)
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 the use
of the model for the prediction of 1,2,4-TMB kinetics and distribution.
This model consisted of five compartments (see Figure C-5) with systemic circulation,
where the tissue volumes corresponded to the volumes of the neutral lipids (i.e., their NLEs), 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 larnberg and lohanson T19991 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 25 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.
C-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
Supplem en tal Information—Trim ethylbenzenes
Lungs
Adipose tissue
y 1 (Lipid Fraction) * A
E ¦ j R
N l | L T
® ' ! Richly perfused tissues I ] ' t
U (Lipid Fraction) | «======j P R
S d| j d I
| j A
B F i Resting muscles • j F . I.
L r (Lipid Fraction) r
O aj | ® |j
c j j c "
O 11 | t L
^ ' j j Working muscles I j ' ®
(Lipid Fraction) ° ' '
1 ' U D
Hepatic tissue
(Lipid Fraction)
Metabolism
Note: Arrows represent blood flows, gas exchange, and metabolism as indicated.
Source: Emond and Krishnan (2006).
Figure C-5. Schematic of human model structure for 1,2,4-TMB using the NLE-
based model approach.
The PBPK model developed by Emond and Krishnan (2006) is used to test the hypothesis
that a model could be developed for HLVOCs using the NLE content of tissues and blood as the
basis. To test this NLE-based approach, the uptake and distribution kinetics in humans for several
chemicals, including 1,2,4-TMB, were simulated. The model appeared to accurately reflect
experimental data; however, a rodent model is needed for this assessment for animal-to-human
extrapolation, and no known rodent NLE model for 1,2,4-TMB is available. The EPA generally
prefers to use a consistent model structure for both experimental animals and humans when
conducting animal-to-human extrapolation, since this consistency is considered a validation of the
model structure. Therefore, use of the Emond and Krishnan (2006) model for human predictions
alone was considered less preferable than use of a model that has been developed for, and shown to
describe, dosimetry in both rats and humans.
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
37
38
Supplem en tal Information—Trim ethylbenzenes
Hissink et al. (2007)
This model was developed to characterize internal exposure following white spirit
inhalation. Since white spirit 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, and the
rat model was then 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 are: liver, fat, slowly perfused
tissues, rapidly perfused tissues, and brain (Figure C-6).
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 were scaled to human body weight (BW°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 white spirit. 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 and 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 in both rats and
humans, the model was valid for the purpose of interspecies extrapolation of blood and brain
concentrations of 1,2,4-TMB as a component of white spirit Moreover, the fact that the model was
demonstrated to adequately fit or predict both rat and human data with a single model structure is
considered a degree of validation of the model structure that does not exist for the other published
models described above.
This document is a draft for review purposes only and does not constitute Agency policy.
C-13 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Gas Exchange
Slowly perfused
Richly perfused
Brain
]
Fat
Liver
(metabolism)
Boxes represent tissue compartments, while solid arrows represent blood flows, gas exchange, and
1
2 Boxes represent tissue compartments, while solid arrows represent blood flows, gas exchange, and
3 metabolism as indicated.
4
5 Source: Hissink et al. (2007).
6 Figure C-6. Schematic of rat and human PBPK model structure.
C.2.2. 1,2,4-TMB PBPK Model Selection
7 All available 1,2,4-TMB PBPK models were evaluated for potential use in this assessment.
8 Of the three deterministic PBPK models available for 1,2,4-TMB fHissink et al.. 2 0 0 7: Emond and
9 Krishnan. 2006: larnberg and lohanson. 19991. the Hissink et al. (2007) model was chosen to utilize
10 in this assessment because it was the only published 1,2,4-TMB model that included
11 parameterization for both rats and humans, for which the model code was available, and for which
12 the model adequately predicted experimental data in the dose range of concern. The Hissink et al.
This document is a draft for review purposes only and does not constitute Agency policy.
C-14 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
Supplem en tal Information—Trim ethylbenzenes
(2007) model was thoroughly evaluated, including a detailed computer code analysis (details
follow in Section C.2.3).
While the Hissink etal. f20071 model had the noted advantages, it did have the following
shortcomings and sources of uncertainty that EPA needed to address:
1) the model was developed and calibrated only for inhalation exposure;
2) the rat model used a different value for the maximum metabolic capacity, Vmax, for each
exposure level, which makes extrapolation or interpolation of the model problematic;
3) the model describes a typical adult and is not parameterized for pregnancy;
4) some physiological parameter values were not consistent with published sources, in
particular, values more commonly used today; and
5) data used to calibrate the model were from inhalation exposure to white spirits, a
complex mixture, and the model does not include all of the resulting potential
interactions.
In particular, the metabolic parameters calibrated against white-spirit data could reflect
metabolic interactions from the mixture, and not accurately predict dosimetry for exposure to
1,2,4-TMB alone. For this reason, model predictions were compared to additional pharmacokinetic
data, a single value of Vmax was identified and used for consistency across the dose range, and some
other model parameters were revised to better match those data, or make better use of existing
biochemical and physiological data. The changes made and specific justifications are detailed in the
following sections, including more minor issues not mentioned here.
C.2.3. Details of Hissink etal. f20071 Model Analysis
C.2.3.1. Review and Verification of the Hissink etal. (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 etal. (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 hour, 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 hours, 8 hours 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. Further, this type of time-dependence is not
based on a predictable or verifiable factor (e.g., dose-dependent metabolic induction); hence, it is
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
31
32
33
34
35
36
37
38
Supplem en tal Information—Trim ethylbenzenes
inconsistent with the intention to extrapolate the model to bioassay conditions. The impact of this
deviation from the published Vmax value is described below with regard to the verification of the
Hissink etal. f20071 model.
Other minor issues were identified by examining the code and comparing it to the model
documentation in Hissink et al. f2007I The code contained some elements that were not necessary
(e.g., intravenous 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 etal. (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 the mixing of air
exiting the alveolar region with air that has only entered the pulmonary dead space.
Comparisons between the computer .m files and published descriptions fHissink etal..
2007) 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 white spirit (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 242-296 g fHissink et al.. 2007). but the .m files used values of 210.01, 204.88,
and 209.88 g in the low-, mid-, and high-exposure groups, respectively. 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 etal.. 2007). 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
This document is a draft for review purposes only and does not constitute Agency policy.
C-16 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
Supplem en tal Information—Trim ethylbenzenes
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. 2016al
Verification of model parameter plausibility
Anatomical and physiological parameters
The anatomical physiological parameters used by Hissink etal. (2007) were taken from U.S.
EPA (1988). but the more current convention is to use the parameters in Brown etal. (1997).
Comparisons of the rat anatomical and physiological parameters in these sources are found in
Table C-3. 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 etal. (1997) do not sum to 100%, which is both a physiological requirement
and a computational requirement to ensure that conservation of mass holds for the model.
Perfusion rates of various depots of fat may differ, so the single value or fractional blood flow to fat
given by Brown etal. (1997) of 7% may be deemed sufficiently uncertain that the Hissink et al.
f20071 value of 9% is considered acceptable. Brown et al. f 19971 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 etal. (2007) assigned a larger
volume fraction of tissue to this compartment Hissink etal. (2007) also assigned a higher
percentage of blood flow to the liver than indicated by Brown etal. (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 etal. f20071
and Brown etal. (1997) are found in T able C-4. In general, the agreement was better for humans
than it was for rats. Brown etal. (1997) proposed a higher default body fat percentage than was
used by Hissink etal. (2007). but Hissink etal. (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.
C-17 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-3. Comparison of rat anatomical and physiological parameters in
2 Hissink et al. (2007) to those of Brown etal. (1997)
Parameter
Hissink et al. (2007)"
Range from Brown
etal. (1997)
Values in agreement?
Alveolar ventilation rate (L/hr/kg07)
20
12-54b
Yes
Total cardiac output (L/hr/kg0 7)
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
Acceptablec
Muscle
35.36-45.5
Skin
15.8-23.6
Total
91
3
4 aValues from U.S. EPA (1988).
5 bAssuming a standard 250-g rat.
6 cHissink et al. (2007) value outside of literature range, but acceptable (see discussion in text).
7
This document is a draft for review purposes only and does not constitute Agency policy.
C-18 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-4. Comparison of human anatomical and physiological parameters in
2 Hissink et al. (2007) to those of Williams and Leggett (1989) as reported by
3 Brown et al. (1997)
Parameter
Hissink et al. (2007)"
Ranee from Brown et
al. (19971
Values in agreement?
Alveolar ventilation rate (L/hr/kg0-7)
20
15
Acceptable
Total cardiac output (L/hr/kg0 7)
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 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 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
4
5 aThe Hissink et al. (2007) value differs from Brown et al. (1997), but is acceptable (see discussion in text).
6
7 Chemical-specific parameters
8 The chemical-specific model parameters, partition coefficients, and metabolic parameters
9 are summarized in Table C-5.
This document is a draft for review purposes only and does not constitute Agency policy.
C-19 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-5. Comparison of chemical-specific parameters in Hissink et al.
2 (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
3
4 aJarnberg and Johanson (1995).
5 bZahlsen et al. (1990).
6 °Jarnberg and Johanson (1999).
7
8 Source: Hissink et al. (2007).
9
10 Where data were available, the agreement is generally acceptable. While the rat-derived Km
11 is less than the lower 95% confidence interval (CI) value for the human Km, the human VmaxC/Km
12 ratio is in acceptable agreement with the published range. When considering sufficiently low
13 exposure concentrations, the performance of the Hissink et al. (2007) human model metabolism
14 parameters would be consistent with the larnberg and lohanson (1999) value.
15 Verification that the model can reproduce all figures and tables in the publication by Hissink
16 et al. C20071
17 The experimental data in Hissink et al. (2007) were estimated by use of Plot Digitizer
18 (version 2.4.1) to convert the symbols on the relevant figures into numerical estimates. The model
This document is a draft for review purposes only and does not constitute Agency policy.
C-20 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplem en tal Information—Trim ethylbenzenes
code provided (adapted for acslX), with a variable value for Vmax, does not appear to perfectly
reproduce the rat simulations in Hissink etal. (2007) (Figures C-7a and b and C-8a and b) (please
note that the Hissink etal. f20071 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 etal. f20071 simulations, but the post-
exposure clearance appears faster in EPA's calculations (see, for example, the 16-hour time points
for the high exposures). When the simulations were run with Vmax constant (Figures C-7c and C-8c),
as documented in Hissink etal. (2007). the rat simulations yield higher blood and tissue
concentrations than depicted in Hissink etal. (2007). most notably at the high exposure
concentration. Similar results were obtained for the rat brain concentrations (Figure C-8). The
human simulations of blood and exhaled air appear to be faithfully reproduced by the model
(Figure C-9). The predicted brain concentration for humans exposed to 600 mg/m3 white spirit
(45 mg/m31,2,4-TMB) for 4 hours was reported as 721 ng/g (0.721 mg/L) in Hissink etal. (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.
C-21 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
I
1 100
(a)
y' a
/ 0
—J
1o 8
N. ;
' A
8 12
Time (h)
Vtfirte sprit exposure of rats to G.Q5,0.19, and 0.37 mo/l 1,2,4-TMB (ttssir* tt al.. 2007)
cs
£
——i
!
p
. A
I
1
*
k. 'N
v
1
A
\ l\
\ \
^ _
1
3— Inel
r;— lr»2
0— Sr»3
f RATT, RAT CVLO W
/ • RATT.RATCVMID
¦ RATT.RATCVHl
(b)
1.0 ^0 3.0 4.0 SO 6-0 7.0 8.0 9,0 lttO 11.0 120 13.0 14.0 15.0 16.0 17.0 iaO 19.0 20.0 21-0 22.0 2a0 24.0
Time 0ir)
spirit «xposur* of rat«to 0-05,0.19, and 0 37 mo/l 1,2,4-TMB (Hifsnk »t al., 2007)
£
i
¦
i
I .
r\ f
. : -
V •
4 ' ¦
v
\
A
:
\
y— i»i
0
pjjjj Srw3
v RATT.RATCVLOW
y f RATT.RATCVMIO
~ ¦ RATT.RATCVHl
(c)
0L0 1.0 20 ao 4.0 5.0 6.0 7.0 S.O 9,0 10.0 11,0 12.0 13.0 14.0 15,0 16.0 17.0 ISO 1910 30.0 210 23.0 2310 24.0
Time (ho
(a) Hissink et al. (2007), Figure 2, lower panel (b) variable Vmax, (c) constant Vmax.
Figure C-7. Simulated and measured blood concentrations of 1,2,4,-TMB in
rats exposed to 600, 2,400, or 4,800 mg/m3 white spirit for 8 hours.
This document is a draft for review purposes only and does not constitute Agency policy.
C-22 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
100000
S» 10000
U)
•O 1000
(a)
8 12
Time (h)
White spirit exposure of rate at 0,05, CL19 and 0,37 rh(|IL l,2>#tTMB(HissinketaL,20B7)
-i
k
kl
r
i
j.
-Hi/ ,
1
r~
^ '
10 12 14
Time (f*)
f
l«»l
*
line 2
-------
Supplem en tal Information—Trim ethylbenzenes
1000
Time (h)
White spirit exposure of humans at 45 mg/L (100 ppm) (Hissink et al., 2007)
t — t, cmix
y ~ Iine2
01234567
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
White spirit exposure of humans at45 mg/L (100 ppm)
(c)
(7 — t, cxeq ||
0 ~ Ime2 ||
V
~ Cj ~ \
0 D r,
? 81 tv.
0
KcLj
:
10 12 14 16 18 20 22 24
(a) Hissink et al. (2007), Figure 4 (b) mode! simulation during exposure, and (c) model simulation after
exposure.
4
5
Figure C-9. Simulated and measured exhaled air concentrations of 1,2,4-TMB
in three volunteers exposed to 600 mg/m3 white spirit for 4 hours.
This document is a draft for review purposes only and does not constitute Agency policy,
C-24' 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
Supplem en tal Information—Trim ethylbenzenes
C.2.3.2. PBPKModel Optimization and Validation
Because of the various issues described above for the Hissink etal. f20071 model, including
inconsistency of physiological parameters, non-mechanistic dose-dependence in metabolic
parameters, and the inability to exactly reproduce the model simulation figures in Hissink et al.
f20071. model parameters were revised as described below. The EPA attempted to minimize the
number of parameters that were changed, focusing on those which were most discrepant from
other published literature or to which model predictions were most sensitive.
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.
(1999). All kinetic studies were conducted with adult animals or adult volunteers. In many cases,
blood and tissue concentration data in a numerical form were available from the literature (Swiercz
etal.. 2003: Swiercz etal.. 2002: Kostrzewski etal.. 1997: Eide and Zahlsen. 1996: Zahlsen et al..
1992: Dahl etal.. 19881. The 1,2,4-TMB blood, brain, and exhaled breath concentration data in
Hissink etal. (20071 were published in graphical format and a colleague of Dr. Hissink also
provided these in numerical form to EPA for use in this analysis.
Average estimates of the blood concentrations of 1,2,4-TMB (average and SD) in humans
exposed only to 1,2,4-TMB as presented in graphs (see larnbergetal.. 1998.1997a: larnbergetal..
19961 were used in this evaluation. Estimates of the blood and tissue 1,2,4-TMB concentrations in
rats presented in graphs in Zahlsen etal. (19901 were also used in this evaluation. Prior to model
optimization, physiological parameters were modified from those in Hissink etal. f20071 to better
reflect a more recent literature compilation (Brown etal.. 19971 than the references cited by
Hissink etal. f20071 (Table C-6). 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. 2016a). For the Zahlsen etal. (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.
C-25 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-6. Parameter values for the rat and human PBPK models for
2 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/kg0 70)
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
3
4 aStudy-specific.
5
This document is a draft for review purposes only and does not constitute Agency policy.
C-26 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Rat Model Optimization
2 The rat studies considered in model optimization and model testing (validation) are
3 summarized in Table C-7.
4 Table C-7. Rat 1,2,4-TMB kinetic studies used in model development and
5 testing
Reference
Strain
Sex
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 white
spirit (7.8%
1,2,4-TMB [39.1,
157.3, 314.7
mg/m3])
8 hrs
Mixed blood
time course
Optimization
(1,2,4-TMB in
mixture)
Figure C-10
Brain time
course
Testing
Figure C-ll
Swiercz et
al. (2003)
Wistar
Male
25, 100, 250 ppm
(123, 492,
1,230 mg/m3)
6 hrs/d,
5 d/wk
4 wks
Venous blood
time course
Optimization
(1,2,4-TMB
only)
Figure C-12
Arterial blood,
liver, brain
Testing
Tables C-8
and C-9
6 hrs
Arterial blood,
liver, brain
Testing
Tables C-8
and C-9
Swiercz et
al. (2002)
Wistar
Male
25, 100, 250 ppm
(123, 492,
1,230 mg/m3)
6 hrs
Venous blood
time course
Testing
Figure C-13
Zahlsen et
al. (1990)
Sprague-
Dawley
Male
1,000 ppm
(4,920 mg/m3)
12 hrs/d
14 d
Blood, brain,
perirenal fat on
d 1, 3, 7,10,
and 14
Testing
Table C-12
Zahlsen et
al. (1992)
Sprague-
Dawley
Male
100 ppm
492 mg/m3)
12 hrs/d
3d
Blood, brain,
liver, kidney,
and perirenal
fat at end of
exposures and
after 12-hr
recovery
Testing
Table C-10
Eide and
Zahlsen
(1996)
Sprague-
Dawley
Male
75, 150, 300,
450 ppm (369,
738, 1,476,
2,214 mg/m3)
12 hrs
Blood, brain,
liver, kidney,
and perirenal
fat
Testing
Table C-ll
Dahl et al.
(1988)
F344/N
Male
100 ppm
(492 mg/m3)
80 min
Inhalation
uptake
Testing
Text
6
This document is a draft for review purposes only and does not constitute Agency policy.
C-27 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
Supplem en tal Information—Trim ethylbenzenes
In order to demonstrate that the model could adequately fit the data used by Hissink et al.
(20071 with appropriate physiological parameters (Table C-6) and a single, constant value for
VmaxC, and to provide an initial condition for subsequent optimization (see below), the metabolic
parameters were re-fitted to the data of Hissink etal. (20071. Specifically, 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 etal. f20071 for rats exposed once to one of three
concentrations of 1,2,4-TMB as a component of white spirit The optimized value of VmaxC was only
modestly different from the value determined by Hissink etal. f20071 (initial: 3.5 versus optimized:
3.08 mg/hour/kg0-7) from visual optimization (with slightly different physiological parameters), but
the Km value differed by 5-fold (initial: 0.25 versus 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 C-10) obtained by re-optimization is statistically significant. This provides quantitative
justification for using the re-optimized values over the original values. 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 over-predicting) 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 C-ll).
White spirit exposure of rats to 0.05, 0.19, and 0.37 mg/L 1,2,4-TMB {Hissink et al., 2007)
~ 10
a
E
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.C
Time (br)
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 $.0 7.0 80 9.0 100 110 120 130 14.0 150 ISO
Time (for)
Figure C-10. Comparisons of model predictions to measured blood
concentrations in rats exposed to 1,2,4-TMB in white spirit f Hissink et al..
20071 (a) before and (b) after numerical optimization.
This document is a draft for review purposes only and does not constitute Agency policy.
C-28 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
Supplem en tal Information—Trim ethylbenzenes
~ 100
—I
o>
£
CO
? 10
o 1
c
Figure C-ll. Comparisons of model predictions to measured brain
concentrations in rats exposed to 1,2,4-TMB in white spirit (Hissink etal..
2007) using model parameters optimized for fit to Hissink etal. (2007) rat
blood data.
Because the model will be applied by estimating 1,2,4-TMB blood levels in rats under
bioassay conditions, it is particularly important that it accurately describe those levels after
repeated exposures. Pharmacokinetic parameters can change after repeated exposures, for
example by metabolic induction. For 1,2,4-TMB, repeated exposure data are available from Swiercz
etal. (20031. Therefore, the VmaxC and Km values derived from optimization to the Hissink et al.
(20071 rat data were used as the starting values for optimizing fit to the venous blood data of
Swiercz etal. (20031. 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/hour/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 C-12, Table C-8). Model simulations matched the observations at 25 and 100 ppm
excellently, while predictions were 1.5-6-fold greater than the 250 ppm data (Table C-8). The
change in the LLF provides justification for using these revised metabolic parameters for simulating
repeated exposure studies versus the original values. The deviation between the model and
experimental data is primarily exhibited 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/hour/kg0-7 and 0.11 mg/L,
respectively, with 96.7% of variation explained (simulations not shown).
White spirit exposure of rats at 0.05, 0.19 and 0.37 mg/L 1,2,4-TMB (Hissink et al., 2007)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Time (hr)
This document is a draft for review purposes only and does not constitute Agency policy.
C-29 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
~ 10
* 1
o
o
m
* 0.1
2
v
0.01
606 607 608 609 610 611 612
1 (a) 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)
_ 10
Oi
E
m
s:
J 1
fN
O
CO
w 0.1
3
0.01
606 607 608 609 610 611 612
2 (b) Time (hr)
3 Figure C-12. Comparisons of model predictions to measured venous blood
4 concentrations by Swiercz et al. (2003) in rats repeatedly exposed to
5 1,2,4-TMB (a) before and (b) after numerical optimization.
6 Table C-8. Model simulated and experimental measured venous blood
7 concentrations of 1,2,4-TMB in male Wistar rats exposed to 1,2,4-TMB
Venous blood 1,2,4-TMB In rats repeatedly exposed to 25, 100 or 250 ppm 1,2,4-TMB (Swlerc etal., 2003)
¦
^ ¦
— i I
i
—
T i
i
i
—
—
Exposure concentration
Time
3 min
30 min
1 hr
3 hrs
6 hrs
25 ppm
Experiment (mg/L)a
0.56 ±0.18
0.33 ±0.03
0.22 ±0.02
0.11 ±0.04
0.06 ± 0.02
Model (mg/L)
0.51
0.29
0.22
0.12
0.06
Ratio (model/experiment)
0.9
0.9
1.0
1.1
1.0
100 ppm
Experiment (mg/L)a
4.06 ± 0.46
3.02 ± 1.43
2.62 ±0.82
0.88 ±0.24
0.37 ±0.14
Model (mg/L)
4.47
2.80
1.95
0.98
0.47
Ratio (model/experiment)
1.1
0.9
0.7
1.1
1.3
250 ppm
Experiment (mg/L)a
13.77 ±3.34
8.28 ± 2.07
6.27 ± 1.72
3.17 ±0.76
1.25 ±0.22
Model (mg/L)
20.44
16.61
14.43
10.80
7.41
Ratio (model/experiment)
1.5
2.0
2.3
3.4
5.9
8
9 aData from Swiercz et al. (2003), Table 2.
10
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
22
Supplem en tal Information—Trim ethylbenzenes
Rat Model Validation
The parameters derived from the Swiercz et al. f20031 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 (Figure C-13)
was very good. In fact, the fit to the acute, high-exposure blood concentrations was superior to the
fit to the repeated, high-exposure data (Figure C-12b). This may reflect adaptation (induction of
metabolism) resulting from repeated, high concentration exposures.
Venous blood 1,2,4-TMB during acute exposure to 25,100, or 250 ppm 1,2,4-TMB (Swiercz et al., 2002)
100
10
1
0.1
0.01
0
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)
100 1 1 1 1
0.01 1 1 1 I I .
6 7 8 9 10 11 12
(b) Time (hrs)
Figure C-13. Comparisons of model predictions to measured rat venous blood
concentrations by Swiercz et al. (2002) in acutely exposed rats (a) during and
(b) after exposure.
Besides the venous blood data to which the model was fit (Figure C-12, Table C-8), Swiercz
etal. (2003) also measured arterial blood and tissue concentrations in animals sacrificed at the end
of the 4-week study (Table 4 in that paper). However, model predictions did not match those post-
sacrifice data very accurately (Table C-9), which is surprising considering that the venous blood
data from the same study were used for optimization. The discrepancies between seemingly
contemporaneous venous and arterial blood measurements were noted by the authors of the
original study and may be due to collection delays (i.e., tail vein for venous blood, decapitation for
arterial samples). Volatilization can also occur from tissue samples until they are significantly
cooled from body temperature, and likewise, metabolism can continue in the liver. Since the
This document is a draft for review purposes only and does not constitute Agency policy.
C-31 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 venous blood data (Table C-8) had specific times post-exposure identified, but the timing of the
2 arterial blood and tissue data was not stated by Swiercz etal. (2003). model simulations were
3 conducted assuming a 0.5-1 hour delay between the end of exposure and sample collection, and are
4 compared to the data in Table C-9. Under these assumptions, most model simulations were within
5 a factor of 2 or 3 of the data, with the largest discrepancy being 5-fold. Differences in PBPK model
6 predictions for single vs. repeated exposures in Table C-9 are primarily due to differences in actual
7 exposure levels used in those predictions.
8 Table C-9. Model simulated and experimental measured tissue concentrations
9 of 1,2,4-TMB in male Wistar rats exposed to 1,2,4-TMB
Exposure concentration
Model
(mg/L)
Experiment
(mg/L)a
Model:
experiment ratio
Repeated exposure (Model t = 606.5-607 hr)
Arterial blood
25 ppm (123 mg/m3)
0.30-0.22
0.33 ±0.11
0.9-0.7
100 ppm (492 mg/m3)
2.8-2.0
1.54 ±0.32
1.8-1.3
250 ppm (1,230 mg/m3)
17.6-15.4
7.52 ±2.11
2.3-2.0
Brain
25 ppm (123 mg/m3)
0.81-0.59
0.45 ± 0.05
1.8-1.3
100 ppm (492 mg/m3)
8.1-5.7
2.82 ± 0.40
2.9-2.0
250 ppm (1,230 mg/m3)
44.1-38.2
18.6 ±4.3
2.4-2.1
Liver
25 ppm (123 mg/m3)
0.14-0.10
0.45 ±0.15
0.3-0.2
100 ppm (492 mg/m3)
4.3-2.3
3.00 ± 0.49
1.4-0.8
250 ppm (1,230 mg/m3)
39.5-33.8
22.5 ±4.1
1.8-1.5
Acute exposure (Model t = 6.5-7 hr)
Arterial blood
25 ppm (123 mg/m3)
0.25-0.19
0.31 ±0.12
0.8-0.6
100 ppm (492 mg/m3)
4.4-3.2
1.24 ±0.41
3.5-2.6
250 ppm (1,230 mg/m3)
14.0-12.0
7.76 ± 1.64
1.8-1.5
Brain
25 ppm (123 mg/m3)
0.91-0.66
0.49 ± 0.06
1.9-1.3
100 ppm (492 mg/m3)
12.5-9.3
2.92 ±0.73
4.3-3.2
250 ppm (1,230 mg/m3)
46.1-40.0
18.3 ± 1.9
2.5-2.2
Liver
25 ppm (123 mg/m3)
0.16-0.11
0.44 ± 0.01
0.35-0.2
100 ppm (492 mg/m3)
8.3-5.3
7.13 ± 1.31
1.2-0.7
250 ppm (1,230 mg/m3)
41.5-35.5
28.2 ±5.3
1.5-1.3
10
11 aData from Swiercz et al. (2003), Table 4.
12
13 Zahlsen and co-workers fEide and Zahlsen. 1996: Zahlsen etal.. 1992: Zahlsen et al.. 19901
14 conducted studies in which male Sprague-Dawley rats were exposed to 1,2,4-TMB by inhalation for
This document is a draft for review purposes only and does not constitute Agency policy.
C-32 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 12 hours/day. For the studies conducted at concentrations similar to those in the Swiercz et al.
2 (2002) and Swiercz etal. (2003) studies, the model error was similar to that of the arterial blood
3 and tissue measurements in the Swiercz etal. (2002) and Swiercz etal. (2003) studies (geometric
4 mean error of 3.3 for Zahlsen et al. (1990). and 2.9 for Eide and Zahlsen (1996)) (Tables C-10 and
5 C-l 1). Since Zahlsen etal. (1992) specifically stated that animals were sacrificed and tissues were
6 collected within 3 minutes of removal from the exposure chamber, the model results in Tables C-10
7 and C-11 do not assume any delay.
8 Table C-10. Model simulated and experimental measured concentrations of
9 1,2,4-TMB in male Sprague-Dawley rats exposed to 100 ppm (492 mg/m3)
10 1,2,4-TMB (12 hours/day, for 3 days) at the end of exposure or 12 hours after
11 the last exposure
Day
Model
(mg/L)
Experiment
(mg/L)a
Model:
experiment ratio
Venous blood
1
8.52
1.70
5.0
2
8.71
1.51
5.8
3
8.72
2.05
4.2
Recovery15
1.08
0.024
7.6
Brain
1
22.6
4.57
4.9
2
23.1
4.19
5.5
3
23.1
4.38
5.3
Recovery15
0.46
Nondetect
Not calculated
Liver
1
18.2
4.92
3.7
2
18.7
3.66
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
17.0
1.4
3
23.1
12.4
1.9
Recovery15
0.46
0.24
1.9
Fat
1
491
210
2.3
2
503
165
3.1
3
504
128
3.9
Recovery15
29.1
14.4
2.0
12
13 aData from Zahlsen et al. (1992).
14 bRecovery period is designated as 12 hours after the last exposure.
15
This document is a draft for review purposes only and does not constitute Agency policy.
C-33 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-ll. Model simulated and experimental measured concentrations of
2 1,2,4-TMB in male Sprague-Dawley rats exposed to 1,2,4-TMB at the end of
3 12-hour exposure
Exposure concentration
Model
(mg/L)
Experiment
(mg/L)a
Model:
experiment ratio
Venous blood
75 ppm (369 mg/m3)
4.21
1.69
2.5
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
Brain
75 ppm (369 mg/m3)
11.5
2.83
4.1
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
Liver
75 ppm (369 mg/m3)
7.39
6.41
1.2
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
Kidney (compared
to rapidly
perfused)
75 ppm (369 mg/m3)
11.5
6.41
1.8
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
Fat
75 ppm (369 mg/m3)
255
61.9
4.1
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
4
5 aData from Eide and Zahlsen (1996).
6
7 There was essentially no difference in the measured venous blood concentration of
8 1,2,4-TMB in the Zahlsen et al. (1992) study at 100 ppm (492 mg/m3) and at 75 ppm (369 mg/m3)
9 in the Eide and Zahlsen (1996) study (1.70 andl.69 mg/L, respectively), so there is evidently some
10 inter-study variability or subtle differences in how the studies were conducted, perhaps in the
11 rapidity of sample collection. The Zahlsen etal. Q9901 study, which used a higher nominal
12 concentration of 1,000 ppm (4,920 mg/m3), exhibited greater deviation between predicted and
13 measured blood and tissue 1,2,4-TMB concentrations (Table C-12), which generally increased with
14 a greater number of exposure days and then plateaued (geometric mean errors of 2.7, 8.4,12.6,
15 13.9, and 12.1 on exposure days 1, 3, 7,10, and 14, respectively). 1,2,4-TMB is also a known
This document is a draft for review purposes only and does not constitute Agency policy.
C-34 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 respiratory irritant, with an RD50 of 519-578 ppm in mice fKorsak etal.. 19971. so it is possible that
2 the 1,000 ppm exposure elicited some sort of avoidance behavior in the rats.
3 Table C-12. Model simulated and experimental measured concentrations of
4 1,2,4-TMB in male Sprague-Dawley rats exposed to 1,000 ppm (4,920 mg/m3)
5 1,2,4-TMB (12 hours/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
6
7 aData from Zahlsen et al. (1990).
8
9 Da hi etal. f 19881 exposed male F344 rats to 1,2,4-TMB at 100 ppm (492 mg/m3) for
10 80 minutes and monitored the total uptake. Under the conditions of the experiment, it was
11 determined that the average rat took up 3.28 (trial 1) or 3.89 (trial 2) mg 1,2,4-TMB. In a model
12 simulation, the predicted uptake was 3.61 mg. The geometric mean model error for the two trials
13 was 1.2.
14 Human Model Validation
15 Kinetic parameters derived from optimal fit for rat venous blood data (described above)
16 were tested for the applicability to human kinetics by comparison to studies in which humans were
17 exposed to 1,2,4-TMB alone or 1,2,4-TMB in co-exposures with white spirit (Table C-13). The key
18 data set for validation in humans was deemed to be Kostrzewski etal. fl9971 because these
19 volunteers were exposed to 1,2,4-TMB alone (no co-exposure, as in Hissink etal. (2007)) under
This document is a draft for review purposes only and does not constitute Agency policy.
C-35 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 sedentary conditions (i.e., level of effort was not elevated, as in the studies by Jarnberg and
2 colleagues flarnberg et al.. 1998.1997a: larnberg etal.. 19961.
3 Table C-13. Human kinetic studies of 1,2,4-TMB used in model validation
Reference
Ethnicity
Sex
Nominal
concentration
Exposure
regimen
1,2,4-TMB
measurements
Use in
model
evaluation
Form of
comparison
Kostrzewski
etal. (1997)a
Not stated;
conducted
in Poland
Sex not
stated;
assumed
male
30 ppm
(147.6 mg/m3)
8 hrs
Venous blood
time course
Testing
Figure C-14
Jarnberg et
al. (1997a);
Jarnberg
and
Johanson
(1999);
Jarnberg et
al. (1998);
Jarnberg et
al. (1996)b
Caucasian;
conducted
in Sweden
Male
2 and 25 ppm
(~10 and 123
mg/m3)
2 hrs at
50 W
(bicycle)
Venous blood
and exhaled air
time course
Testing
(blood data
only)
Figure C-15
Hissink et al.
(2007)°
Not stated;
spoke
Dutch as
"native
language"
Male
100 ppm white
spirit 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 C-16
4
5 aFive volunteers, ages 24-37 years, with no known occupational exposure to 1,2,4-TMB. Height of 1.70-1.86 m
6 and body weight of 70-97 kg. The average of the high and low values for age, height, and weight plus assumed
7 gender (male) were used to calculate central tendency estimate of 22.44% for volume of body fat (VFC), per
8 Deurenberg et al. (1991). Alveolar ventilation rate (QPC) estimated from the midpoint of the range for total
9 ventilation (0.56-1 m3/hour), average of high and low body weights, BW°74 scaling, and an assumption that
10 alveolar ventilation was 2/3 of total ventilation.
11 bTen volunteers, average age 35 (range 26-48) years, with no known occupational exposure to solvents; volunteers
12 were instructed to avoid contact with organic solvents and to refrain from taking drugs or drinking alcoholic
13 beverages for 2 days before exposure. Average body weight was 76.5 kg. QPC estimated from the mean value
14 for total ventilation rate during exposure, average body weights, BW0 74 scaling, and an assumption that alveolar
15 ventilation was 2/3 of total ventilation. Digitized blood data (group averages) extracted from figures.
16 Three volunteers, ages 23-26 years, body weight was 69-82 kg, mean body fat of 14.6% (skin caliper
17 measurement); alcohol consumption 10-15 drinks/week (all subjects), one smoker (four cigarettes per day).
18
19 Using the VmaxC and Km derived from the Swiercz etal. (2003) rat repeated-exposure data,
20 the simulated blood concentration underestimated those measured during exposure of volunteers
21 by Kostrzewski et al. f 19971. then over-predicted blood concentrations up to 7 hours post-
22 exposure, and under-predicted subsequent measured blood concentrations (Figure C-14). Of
23 21 blood measurements, only two differed from the simulated value by more than a factor of
This document is a draft for review purposes only and does not constitute Agency policy.
C-36 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
Supplem en tal Information—Trim ethylbenzenes
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/hour/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 and subsequently used for the human model (see the section regarding rat
model optimization).
Blood 1,2,4-TMB in human volunteers exposed to 154 mg/m3 1,2,4-TMB (Kostrzewski et al., 1997)
0.01
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)
Figure C-14. Comparisons of model predictions to measured human venous
blood concentrations of Kostrzewski et al. (1997) in volunteers exposed to
154 mg 1,2,4-TMB/m3 for 8 hours.
For comparisons between the data in the studies by Jarnberg and colleagues (Tarnberg and
Tohanson. 1999: Tarnberg etal.. 1998.1997a: Tarnberg etal.. 19961 and the model, simulations were
conducted with alveolar ventilation rate (QPC; calculated as described in footnote to Table C-13) at
the elevated (working) level throughout the simulation, but with no other adjustments made for
exercise conditions. The model consistently under-predicted the measured venous blood
concentrations of 1,2,4-TMB (Figure C-15). At 25 ppm (123 mg/m3), blood concentrations were
under-predicted by a factor of 2.1-3.5 during exposure and by a factor of 1.04-1.5-fold in the post-
exposure period, for a geometric mean discrepancy of 1.7 for this concentration. At 2 ppm
(~10 mg/m3), blood concentrations were under-predicted by factors of 1.7-2.7 during exposure
and 1.01-1.2 in the post-exposure 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.
C-37 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
|
e
0
? 0.1
S
v
1
8
m
Z
Jo.oi
"g
i
to
0.001
1
2 Figure C-15. Comparisons of model predictions to measured human venous
3 blood concentrations in volunteers exposed to 2 or 25 ppm (~1Q or
4 123 mg/m3) 1,2,4-TMB for 2 hours while riding a bicycle (50 W) flarnbere et
5 al.. 1998.1997a: larnber^ et al.. 19961.
6 Comparisons of model predictions and experimental data were also made for the human
7 study described in Hissink et al. f20071 in which volunteers inhaled 100 ppm white spirit with
8 7,8% 1,2,4-TMB (38.4 mg/m31,2,4-TMB) for 4 hours (Figure C-16], The agreement between
9 simulated and measured concentrations of 1,2,4-TMB in blood during exposure was excellent The
10 agreement between the modeled and measured 1,2,4-TMB in end-exhaled air during the post-
11 exposure period was very good.
12
(a)
13
(b)
14
15
16
17
Blood concentrations of 1,2,4 -TMB in volunteers exposed to 2 or 25 ppm 1,2,4-TMB (Jamberg and coworkers)
2
3
S
1
4
Time (hr)
White ipln< exposure of human* at 45 mg/L {100 ppm) (hissink et al., 2007)
0.16
| 0.14
I 0.12
I "
I 0.08
0
(E 0.06
»
<\ o.w
1 0.02
_ J _
T 1 °D
~
AT ° o
\ D
i-b ~
ocK
k
D
~
i
s i.oxiae-3
I
IB
?
- i.oxioe-4
*
I
|
I 1.0x10 £-5
10 11 12 13 14
o iP
~
D I
5 F
3 i
r -
-Q-
10 11 12 13 14
Figure C-16. Comparisons of model predictions to measured (a) human
venous blood and (b) end of exposure exhaled air 1,2,4-TMB in volunteers
exposed to 100 ppm white spirit with 7.8% 1,2,4-TMB (38.4 mg/m3
1,2.4-TMBl fHissink et al. 20071.
This document is a draft for review purposes only and does not constitute Agency policy,
C-38 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
Supplem en tal Information—Trim ethylbenzenes
Summary of Optimization and Validation
Numerical optimization of the fit to the rat data in Hissink et al. f20071 produced a similar
VmaxC, but smaller Km, than the values determined by Hissink et al. f20071 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 etal. f20071 study were co-exposed to other
components of white spirit, 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 etal. (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 reference
concentration (RfC) and reference dose (RfD) development The VmaxC and Km values from the
numerical optimization to the Hissink etal. (2007) rat data were used as starting values for
optimization of the fit to the Swiercz etal. (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 volunteers were acceptable. The
measured Wistar rat arterial blood and tissue concentrations were consistently over-predicted by
the model, suggesting collection delays in the studies. The model also consistently over-predicted
the measured Sprague-Dawley rat tissue and blood concentrations, including the "recovery"
(12 hours 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.
Using the VmaxC and Km values obtained by fitting the PBPK model to the Swiercz et al.
(2003) rat data and appropriate human physiological parameters (Table C-6), model predictions of
the human pharmacokinetic data were found to be adequate, and were not significantly improved
by numerical re-optimization. Therefore, the VmaxC and Km from the rat were used for the human
model (i.e., allometric scaling).
Overall, it was concluded 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
This document is a draft for review purposes only and does not constitute Agency policy.
C-39 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
Supplem en tal Information—Trim ethylbenzenes
using VmaxC and Km parameters optimized for repeated, high concentration exposures (e.g.,
250 ppm (1,230 mg/m3) from Swiercz etal. (2003)1.
Uncertainties in Model Structure
All PBPK models are a simplification of physical reality, and a full discussion of the resulting
uncertainties is beyond the scope of this review. For example, this model uses the typical
assumption of perfusion-limited transport between circulating blood and tissues, but a more
realistic representation that also requires more data and parameters is diffusion-limited transport.
If model predictions systematically over-predicted the rate of change of 1,2,4-TMB in blood, then
diffusion-limited transport could have been evaluated as a more accurate model structure, but
given the overall agreement in model predictions and measured kinetics, such an evaluation was
not considered a valuable use of existing resources.
A simplification in the model structure used in Hissink etal. (2007) versus that of larnberg
and lohanson (1999) is that larnberg and lohanson (1999) included working versus resting muscle
compartments, which effectively allowed a higher fraction of cardiac output to go to the muscle
compartment under working conditions versus resting. When simulating the corresponding human
exposure data (larnberg etal.. 1998.1997a: larnberg etal.. 1996). the Hissink etal. (2007) model
was adjusted for the working conditions by increasing cardiac output, but that adjustment would
increase blood flow to all tissues proportionally, including hepatic blood flow, which then can
increase the predicted rate of metabolism (more so than larnberg and lohanson. 1999). This
simpler approach offers an explanation of why the blood-levels are under-predicted in Figure C-15
by ~2-3 fold. This difference suggests a comparable uncertainty in the model for predicting blood
levels during working conditions, but the model matched the post-exposure data in Figure C-15
quite well, within a factor of 1.5 beyond the first couple of time-points. Hence, while the model
might be improved by adding a working muscle compartment and appropriate work-level
parameterization, the impact for predictions of 30 working hours in a 168-hour week are expected
to be less than a factor of 1.5. (Assuming an error of 2.5-fold for 30/168 hours, the average error is
2.5*30/168 = 0.45-fold.)
Another place where systematic differences between model predictions and data suggest
model structure errors is that the model over-predicted the 250 ppm rat venous blood data of
Swiercz etal. (2003) after 4 weeks of exposure, although it did fit the 2 5 and 100 ppm data
(Figure C-12, panel (b)), and it fit the acute-exposure data Swiercz etal. f20021 at all three
concentrations (Figure C-13). The over-prediction of 1,000 ppm, 14-day rat data fTable C-12:
Zahlsen et al.. 1990) was significantly greater than the over-prediction of 75-450 ppm acute-
exposure data. One possible explanation for the dose-dependence of the errors is that a first-order
(or high-Km) metabolic pathway was operative only significantly at higher exposure levels.
However, in that regard, one would have expected optimization of the single Km in the existing
model to have identified an intermediate value that better-predicted the 250 ppm 4-week data from
Swiercz etal. (2003). Identifying more complex metabolic schemes is difficult using only parent-
This document is a draft for review purposes only and does not constitute Agency policy.
C-40 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
Supplem en tal Information—Trim ethylbenzenes
concentration in vivo data. The hypothesis of multiple metabolic pathways with differing dose-
dependence would best be evaluated by careful in vitro metabolic studies, but the possibility is
certainly suggested given the multiple routes of metabolism shown in Figure C-l.
A second structural possibility suggested by these discrepancies between rat model
predictions and data (which is not exclusive of multiple pathway kinetics discussed in the
preceding paragraph) is metabolic induction, which would be both time-dependent (i.e., would not
occur, or occur to a lesser extent, with acute exposures) and concentration-dependent The results
in Table C-12, where measured blood and tissue levels decline and hence model:data ratios
increase with exposure days, are particularly suggestive of this possibility. However there was not
a clear time-dependent change in the 3-day study of Zahlsen etal. Q9921 fTable C-10), at 100 ppm.
So this hypothetical mechanism may not be relevant at exposures near the point of departure
(POD) (benchmark dose [BMD] levels). In any case, verification of this hypothesis would require a
combination of in vivo and in vitro studies, where liver samples are collected from rats after
different exposure levels and durations, and evaluated for metabolic capacity.
A third possible explanation for the discrepancies is that, given 1,2,4-TMBs irritancy
(Korsak etal.. 1997). rats exposed in open cages may be reducing their activity level or otherwise
finding ways to reduce their exposure. For example, by huddling or tucking their noses into their
fur, the rats could be re-breathing a portion of expired air, which would then have a lower
1,2,4-TMB concentration than in the rest of the exposure chamber. Testing of this hypothesis could
be performed by observation of rat behavior in open exposure chambers as a function of exposure
level and duration, and comparison of results to nose-only exposures, in conjunction with
plethysmography to determine any changes in respiration rates.
In summary, based on comparisons of model predictions to various data sets, it appears that
the most significant structural uncertainty for the human PBPK model is the lack of realism in
predicting physiological changes due to work/physical activity, but the overall impact of this
uncertainty is less than a factor of 1.5. Discrepancies between the rat model and reported data
suggest two model structure uncertainties (the presence of multiple metabolic pathways with
significantly different concentration-dependence, and metabolic induction) and one possibility
related to exposure levels or specification (avoidance behavior, which is not a part of the model
itself). In the range of application, these uncertainties in the rat model for estimating venous blood
levels represent a factor of 2-3-fold, though the lack of fit of the model to the data becomes more
severe at higher exposure levels.
Uncertainties Due to Choice of Dose Metric
The use of the average, parent-chemical venous blood concentration as the internal dose for
predicting systemic effects of 1,2,4-TMB is based on the following assumptions/general
expectations:
1) the parent chemical, and not a metabolite, is the causative agent for systemic effects;
This document is a draft for review purposes only and does not constitute Agency policy.
C-41 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
Supplem en tal Information—Trim ethylbenzenes
2) average concentration (equivalent to the area under the curve [AUC] calculated over
comparable total time in rats and humans) is a good predictor of risk;
3) the ratio of 1,2,4-TMB's concentration in the target tissue to the venous blood is
approximately the same in humans as in rats; and
4) while target-tissue concentrations are generally expected to be better predictors than
blood concentration, this expectation is counter-balanced by the lack of target-tissue
dosimetry in humans, leading to greater uncertainty in human target tissue estimates.
As discussed in the mode-of-action section, little is known about the mechanisms of action
for 1,2,4-TMB, in particular whether the parent or a metabolite is responsible for the hematological
or neurological effects. One might assume that if a metabolite is causative, then the concentration
of the metabolite would vary in proportion to the parent. However, if two individuals have similar
exposures, and thus absorb 1,2,4-TMB at a similar rate, but metabolism to the toxic compound is
twice as fast in the second individual, then the venous concentration of 1,2,4-TMB in that individual
would be lower than the first (because it's being metabolized faster), but the rate of toxic
metabolite production is higher. Likewise, the blood:air concentration ratio of 1,24-TMB in humans
might be lower than in rats, but the concentration of the toxic metabolite in humans could be
higher. But for this lack of proportionality to occur, the scaling of the metabolic conversion of
1,2,4-TMB to the toxic metabolite, between rats and humans, would have to be significantly
different from the scaling for the rate at which the toxic metabolite is cleared from the body. Such a
difference can occur, but the general expectation is that metabolism and other physiological
processes that affect clearance (including blood-flow) scale allometrically, as BW°75. In fact, for
1,2,4-TMB, the metabolism in humans was found to be fairly consistent with this scaling. Therefore,
a lack in proportionality of a subsequent (toxic) metabolite would only occur if the clearance of that
metabolite does NOT scale allometrically. In summary, it is possible that misidentification of the
toxic metabolite could result in a very large error in the predicted human risk, but the fact that most
metabolic and clearance processes scale similarly (allometrically) makes this possibility unlikely.
Quantifying the resulting uncertainty is beyond the scope of this assessment.
The use of average concentration, or AUC, calculated over a similar time-frame (1 week) in
rats and humans reflects the assumption that the observed hematological and neurological effects
result from an accumulation of cellular or tissue damage, that the damage accumulates in
proportion to 1,2,4-TMB concentration, and that clearance or repair of the damage is relatively slow
(i.e., requires weeks or longer). Testing of this hypothesis would require a set of experiments
where exposure level and duration were varied independently (i.e., C x t experiments), and damage
was assessed at multiple recovery times. Such data are mostly not available for 1,2,4-TMB.
However, the hematological effects are likely the result of cytotoxicity, which is expected to
increase with both concentration and duration. So the uncertainty for using average concentration
for this endpoint is considered low.
This document is a draft for review purposes only and does not constitute Agency policy.
C-42 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
Supplem en tal Information—Trim ethylbenzenes
Since the dose-dependent delayed recovery from a sensory challenge (footshock/paw-lick
experiments) show a persistent effect, 50+ days after exposure ended, that effect is also assumed to
result from cumulative damage, rather than a single day's exposure. Whether the same effect level
would have been seen after a single week's exposure, or if chronic exposure might have resulted in
a more severe effect at a given exposure level, is simply not known. The uncertainty in using
subchronic exposure data to set a reference level is mitigated by application of the subchronic-to-
chronic uncertainty factor (UF). The use of the weekly average (blood) concentration is still
appropriate, even if the effect only takes 1-2 weeks to develop, since the damage is still likely to
accumulate within that time-frame according to the number of hours/week of exposure. For a
presumed continuous (24 x 7) inhalation exposure to the general human population, use of weekly
average concentration results in a more appropriate reference level than use of peak concentration.
If the effect is not cumulative for exposure beyond several hours (i.e., can be better predicted from
peak concentration), then use of the weekly average would over-predict human risk by a factor of
5-6 (~168 hours/30 hours).
The use of venous blood versus tissue concentrations creates some uncertainty, but this
uncertainty is counterbalanced by uncertainties in the exact tissues where effects occur and the
partitioning of 1,2,4-TMB into those tissues. The tissue:blood partition coefficients of Hissink etal.
(2007) are obtained by combining a correlation for tissue:air partition coefficients, developed
previously using data for a single representative tissue from a single species, against oil:air and
saline:air partition coefficients (which have been measured for 1,2,4-TMB), with values for the
blood:air partition coefficient measured separately with rat and human blood. So there is
considerable uncertainty in the use of these partition coefficients for human versus rat bone
marrow, for example (assuming that this is the site for hematological effects), given that species-
and chemical-specific values for bone marrow are not available. The measured blood:air partition
coefficients for 1,2,4-TMB indicate that its affinity for human blood is 1.74 times lower than for rat
blood, so if the typical assumption was made that the affinity for other tissues does not vary across
species, then use of tissue versus venous blood concentration would result in an approximately
1.7-fold increase in the estimated human risk. However, such use would also increase the level of
uncertainty because there are no human tissue data to validate those model predictions, and
because the site of action is uncertain. For example, it's not known if the neurological effects occur
primarily due to effects in the brain or to effects on peripheral nerves, and, if the latter, whether the
partition coefficient for "brain" versus "slowly perfused" tissue (which differ ~2-fold) should be
used. As with other aspects of uncertainty, a full quantitation of the uncertainty resulting from the
use of venous blood versus tissue concentrations is beyond the scope of this assessment. But the
identifiable uncertainty is less than a factor of 2. The direction of this uncertainty is the opposite of
that from using average versus peak concentration for continuous human exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
C-43 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
Supplem en tal Information—Trim ethylbenzenes
C.2.3.3. 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 etal. (20031 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 hours of
exposure, 6 hours post-exposure), conditions that are essentially identical to those in Swiercz et al.
f20021. 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 C-13b), the agreement between the model and the experimental data atthe 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 C-14) as relatively
insensitive (maximum|NSC| < 0.2 for 0.25 hours < t < 12 hours), 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. It was concluded that Km and VmaxC are distinctly
identifiable using the Swiercz etal. f20031 and Swiercz etal. f20021 data.
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 Zahlsen etal. (199011. The accuracy of
the predicted net uptake in the Da hi etal. f 19881 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 tissue compartments are also moderately important
parameters (with time courses similar to those of the corresponding partition coefficients shown in
Figure C-17). The volume of the fat compartment in particular is known to vary with age and strain
f Brown etal.. 19971. 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.
C-44 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplem en tal Information—Trim ethylbenzenes
Table C-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
PF
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 [1,230 mg/m3]).
BW = body weight; CONC = concentration of 1,2,4-TMB in the air; Vmax = Michaelis-Menten maximum rate of
metabolism; VmaxC = Michaelis-Menten constant: concentration where Vmax is half-maximal (Vmax); PB = blood:air
partition coefficient; PF = fat:blood partition coefficient; PS = slowly perfused:blood partition coefficient;
PR = rapidly perfused:blood partition coefficient; PL = liver:blood partition coefficient; PBR = brain:blood partition
coefficient; VFC = volume of fat; VSTOTC = volume of slowly perfused tissues; VRTOTC = volume of rapidly
perfused tissues; VLC = volume of liver; VBRC = volume of brain; QCC = cardiac output; QFC = blood flow to fat;
QRTOTC = blood flow to slowly perfused tissues; QLC = blood flow to liver; QBRC = blood flow to brain.
This document is a draft for review purposes only and does not constitute Agency policy.
C-45 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Sensitivity analysis: rat CV, low concentration
exposure
(Swiercz et al.f 2002, 2003)
cv:km
cv:vmaxc
cv:pb
cv;pf
cv:ps
-0.C
(a)
Time (hr)
Sensitivity analysis: ratCV, high concentration
exposure
(Swiercz et al., 2002, 2003)
0.8
§
o
(b)
cv:km
cv:vmaxc
cv:pb
cv:pf
cv:ps
Time (hr)
1 Figure C-17. Time course of NSCs of moderately sensitive chemical-specific
2 parameters (response: venous blood concentration) in rats exposed to (a) 25
3 ppm (123 mg/m3) or (b) 250 ppm (1,230 mg/m3) of 1,2,4-TMB via inhalation
4 for 6 hours fSwiercz etal.. 2003: Swiercz et al.. 20021.
C.2.3.4. Sensitivity Analysis of Human Model Predictions
5 A sensitivity analysis for human model predictions to all parameters was conducted for
6 continuous inhalation exposures, and results are shown in Table C-15. The results are presented as
7 NSCs (i.e., percent change in output/percent change in input, calculated using the central difference
This document is a draft for review purposes only and does not constitute Agency policy,
C-46 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 method; NSC). Similar to analyses performed for the rat, parameters are noted as relatively
2 insensitive (|NSC| < 0.2), moderately sensitive (0.2 < |NSC| < 1.0), or highly sensitive (|NSC| > 1.0).
3 To bracket the range of human equivalent concentrations (HECs), inhalation sensitivities were
4 evaluated at 10 and 150 ppm (49.2 and 738 mg/m3) concentration. The resulting coefficients
5 (Table C-15) are not surprising. The two fitted metabolic parameters, VmaxC and Km, both influence
6 model predictions. The VmaxC sensitivity is higher at 150 ppm (738 mg/m3) (10.88731) than at
7 10 ppm (49.2 mg/m3) (|0.238|) due to the slight metabolic saturation.
8 Table C-15. Parameter sensitivity for steady-state venous blood 1,2,4-TMB
9 concentration in humans 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
L, H
PB
L, H
PF
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
10
11 L = low exposure concentration (10 ppm [49.2mg/m3]), H = high exposure concentration (150 ppm [738 mg/m3]).
12
13 Body weight (BW), concentration of 1,2,4-TMB in the air (CONC), alveolar ventilation rate (QPC), Michaelis-Menten
14 maximum rate of metabolism (VmaxC), Michaelis-Menten constant: concentration where Vm,ax is half-maximal
15 (V max ), blood:air partition coefficient (PB), fat:blood partition coefficient (PF), slowly perfused:blood partition
16 coefficient (PS), rapidly perfused:blood partition coefficient (PR), liver:blood partition coefficient (PL), brain:blood
17 partition coefficient (PBR), volume of fat (VFC), volume of slowly perfused tissues (VSTOTC), volume of rapidly
18 perfused tissues (VRTOTC), volume of liver (VLC), volume of brain (VBRC), cardiac output (QCC), blood flow to fat
19 (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.
C-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
25
26
27
28
Supplem en tal Information—Trim ethylbenzenes
C.2.3.5. Modification of the Hissink et ah (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.
f20071 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. To more accurately approximate patterns of human oral ingestion, ingestion was simulated
as an idealized pattern of six events, each lasting 30 minutes. Twenty-five percent of the total daily
dose was assumed to be ingested at each of three events beginning at 7 am, 12 pm (noon), and 6 pm
(total of 75%). Ten percent of the daily dose was assumed to be ingested at events beginning at
10 am and 3 pm (total of 20%). The final 5% was assumed to be ingested in an event beginning at
10 pm. After the daily blood concentration profile achieved a repeating pattern, or periodicity, the
weekly average blood concentration was then used to determine the human equivalent dose (HED).
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 of continuous exposure) for a
standard human at rest (70 kg) for a range of concentrations and doses. For ease of visual
comparison (Figure C-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.
C-48 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
Supplem en tal Information—Trim ethylbenzenes
T3
{10
OH
£
aj
a.
—i
txfl
E.
O
T3
_>¦
'rn
>¦
w
Figure C-18. Effect of route of exposure and dose rate on steady-state venous
blood concentration (t = 1,200 hours) for continuous human exposure to
1,2,4-TMB.
C.2.3.6. Conclusions
Several changes were made to the model for use in this assessment: (1) updated
physiological parameters were implemented fBrownetal.. 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 Hissinketal. (20071 data and resulted in a
VmaxC of 4.17 mg/hour/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 fSwiercz etal.. 2003:
Swiercz etal.. 2002: Eide andZahlsen. 1996: Zahlsenetal.. 1992: Zahlsen etal.. 1990: Dahl etal..
19881. Using the optimized values, the model adequately predicted the data and lower
concentrations. Human data fHissink etal.. 2007: Tarnbergand Tohanson. 1999: Tarnbergetal..
1998.1997a: Kostrzewski etal.. 1997: Tarnbergetal.. 19961 were also utilized to validate model
predictions.
C.2.4. 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.
C.3. HUMAN STUDIES
Table C-16 provides study details for epidemiology studies.
This document is a draft for review purposes only and does not constitute Agency policy.
C-49 DRAFT—DO NOT CITE OR QUOTE
0.1
0.01
0.1
•Oral
Inhalation
10 100 1000 10000 100000
Daily dose (mg/kg/d)
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-16. Characteristics and quantitative results for epidemiologic studies of TMB and related compounds and
2 mixtures
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
Respiratory/irritative effects
Battig et al.
(1956), as
reviewed
bv Battig et
al. (1958)
Cross-sectional.
Exposed: 27TMB-
exposed workers who
worked primarily in the
painting shop of a
transportation plant.
Controls: 10 unskilled
workers from the same
plant that were not
exposed to TMB vapors.
Various respiratory
and hematological
endpoints were
assessed via worker
interviews and
clinical assessments.
Exposure level: 10-60 ppm
(49.2-295 mg/m3) in
working rooms.
Exposure duration:
approximately lOyrs.
Compounds exposed to:
Fleet-X DV-9, a solvent
containing 1,2,4-TMB and
1,3,5-TMB (50 and 30%,
respectively). Fleet X DV-9
also potentially contained
1,2,3-TMBand numerous
methylethyl benzenes.
No statistical analyses were reported.
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.
Billionnet
et al.
(2011)
Cross-sectional survey in
a national population-
based sample of
residences in France.
Final sample consisted
of 567 residences and
1,612 individuals.
Asthma and rhinitis,
determined via
standardized self-
administered
questionnaire.
Diagnosis of asthma
or rhinitis not
confirmed by
physician.
Pollutants measured for
1 wk in the bedroom of the
home.
Exposure level: For
1,2,4-TMB, exposure varied
from undetectable to
111.7 ng/m3, with median
concentration 4.0 ng/m3.
Median tests were used for continuous endpoints, x2 test for
categorical variables.
Pollutant correlations tested by Spearman's rank correlation
coefficient.
Generalized estimating equation approach was 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.
C-50 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
OR for association of asthma to 1,2,4-TMB statistically
significantly increased (OR = 2.1). OR of the 95th percentile
compared to 75th percentile = 3.13 (95% CI: 1.6-6.12).
Norseth et
al. (1991)
Cross-sectional study of
road repair and
construction workers in
Norway exposed to
asphalt.
First group: 79 workers.
Second group:
254 workers with
247 controls.
A number of
neurological and
irritative symptoms
were recorded by
standard
questionnaire on
last day.
Exposure to 14 groups of
organic compounds during
5 d was assessed in the
various groups. 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 5 d
of monitoring.
Exact two-sided Fisher-Irving test was used to analyze differences
in symptom frequency.
Mean difference between groups was calculated via two-sided
Wilcoxon rank-sum test with a significance level of 5%.
Spearman's correlation coefficient was used to estimate
correlation between symptoms and possible confounders.
Among workers reporting at least 1 d of experiencing a symptom,
asphalt workers were observed to have increased incidences of
abnormal fatigue, reduced appetite, laryngeal/pharyngeal
irritation, eye irritation, and other unspecified symptoms,
compared to non-asphalt workers (all differences reported to be
statistically significant).
Neurological effects
Chen et al.
(1999)
Retrospective mortality
cohort study: included
all 1,292 men who had
worked at the paint
shop of a dockyard in a
Scottish dockyard for
>12 mo from 1950 to
Mortality, cause of
death coded
according to ICD-9.
Questionnaire
recorded self-
reported symptoms
Exposure level: Specific
concentrations not
discussed.
Exposure duration: at least
1 yr; range 1-41 yrs.
Intra-cohort PMRs were calculated, as were SMRs for comparison
with all Scottish males; 95% CIs were calculated assuming a
Poisson distribution.
X2 test was used to assess differences in neuropsychological
symptoms between painters and non-painters.
This document is a draft for review purposes only and does not constitute Agency policy.
C-51 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
1992 (followed up from
12/1/60 to 12/31/94);
205 deceased workers
Included In analysis.
Cross-sectional study:
953 painters not
identified as dead as of
12/31/95 and 953 age-
matched male controls.
875 subjects returned
questionnaire:
302 painters,
573 controls;
260 painters and
539 controls included in
final analysis.
of psychological or
neurological
disorders.
Questionnaire also
recorded
information on
potential
confounders:
educational level,
smoking status, and
alcohol
consumption.
Compounds to which study
participants were exposed:
white spirit (1,2,4-TMB),
xylene, TMB (unspecified),
n-butanol, trichlorethylene,
naptha, and cumene.
Breslow-Cox model was used to adjust for covariates including
educational level, smoking, alcohol consumption, and social
conformity.
Log-regression model was used for case-control study.
Mortality was not generally increased among painters; the only
statistical significant increase was for ischemic heart disease
(PMR = 132, 95% CI: 105-164)
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: for
painters with total symptom score >12: 2.27,1.20-4.30 (1-4 yrs);
2.42, 1.18-4.94 (5-9 yrs); 2.89, 1.42-5.88 (10-14 yrs); and 3.41,
1.81-6.36 (15-41 yrs). No apparent decrease in symptoms was
observed when investigating time since stopping painting: 3.71,
1.66-8.29 (1-10 since stopping); 3.53,1.79-6.96 (11-18 yrs since
stopping); and 2.98,1.06-8.53 (>19 yrs since stopping).
Multivariate-adjusted ORs showed the same relationship.
Gong et al.
(2003)
Cross-sectional study;
exposed workers
(N = 251) worked in
53 furniture factories in
Japan. A control group
(N = 147) was drawn
from un-exposed
workers in different
factories.
Questionnaire
recorded
information
pertaining to work
history and lifestyle
habits,
occupational/
vocational solvent
exposure, alcohol
consumption,
The exposure
concentrations of solvents
were assessed via
environmental sampling
and biomonitoring.
Exposures included toluene,
xylene, styrene,
ethylbenzene; urinary
metabolites included xylene
and hippuric acid. Neither
TMBs nor TMB metabolites
The Wilcoxon rank sum test was used to compare color vision and
color contrast between exposed workers and controls.
Multiple regression analysis was used to assess the association
between exposure and visual dysfunction outcomes, with age,
alcohol, smoking, educational experience, and duration of
exposure as independent variables.
Color vision and color contrast were statistically significantly
altered in exposed workers compared to controls (p-values
<0.05).
This document is a draft for review purposes only and does not constitute Agency policy.
C-52 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
cigarette smoking,
and medical usage.
A variety of visual
dysfunction tests
(color vision
assessment, visual
contrast sensitivity,
and VEP) were
administered to
exposed workers
and controls.
were listed as explicit
exposures.
The total exposure index
was 0.35 compared to
Japanese threshold limit
values, indicating low
exposures.
Multiple regression revealed that color vision was significantly
negatively correlated with age, and that methylhippuric acid
metabolites were correlated with decreased color contrast
sensitivity. Smoking was also significantly associated with
increased color contrast sensitivity.
Tang et al.
(2011)
Cross-sectional study of
133 solvent exposed
workers and 78 non-
exposed controls. All
participants underwent
a medical evaluation and
screening for smoking
and drug use;
27 exposed and controls
were ultimately selected
for fMRI study to
compare
pathophysiological
changes in brain
function.
An N-back task
(identifying letters
in a sequence) was
performed during
fMRI scans.
A cumulative lifetime
exposure index was
calculated for each subject
who reported solvent
exposure. The duration and
time spent performing
specific job tasks was
determined via
questionnaire.
Representative solvent
exposures were determined
via field samples. Historic
solvent exposures and
information on protective
equipment usage were used
to adjust exposure
estimates.
fMRI scans were analyzed via ANCOVA to compare activity levels
in specific brain regions.
Solvent-exposed workers were more likely to be African-American
compared to controls, and had lower reading test scores and
higher blood lead levels. Performance scores for the N-back task
was significantly lower than controls (p = 0.005).
After correcting for verbal IQ and lead, Caucasian exposed
workers had reduced activity in the anterior cingulate cortex and
dorsolateral prefrontal cortex. ANCOVA revealed significantly
reduced activity in the dorsolateral prefrontal cortex and left
parietal regions in exposed workers.
El Hamid
Hassan et
al. (2013)
Cross-sectional study of
Egyptian paint factory
workers. The exposed
group (N = 92) included
Questionnaire
recorded self-
reported symptoms
of psychological or
No explicit exposure
analysis were conducted.
Analyses were based on
comparisons of exposed
X2 test was used to investigate pair-wise differences in neuro-
psychological symptoms in exposed workers, compared to
controls.
This document is a draft for review purposes only and does not constitute Agency policy.
C-53 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
workers exposed to
organic solvents as part
of their job. These
solvents included
mixtures of aliphatic and
aromatic solvents
(xylene, toluene, methyl
iosbutyl and methyl
ethyl ketone, mineral
spirits, etc. TMB
isomers not specifically
mentioned). The control
group (N = 95) consisted
of members of the
faculty of medicine at a
nearby university not
exposed to these
solvents.
neurological
disorders.
Questionnaire also
recorded
information on
potential
confounders:
educational level,
smoking status, and
alcohol
consumption.
groups (determined by job
type) to controls. Duration
of exposure was also used
in some analyses
Highly significant differences (p < 0.001) between exposed
workers and controls were noted for most psychological (short
memory, problems concentrating, abnormally tired, headache),
neuropsychological (painful tingling, trouble buttoning/
unbuttoning), and neurological (dizziness, hand tremble,
weakness in arms/legs) symptoms.
63.0% of workers demonstrated neuropsychological symptoms,
compared to 2.1% of controls (p-value < 0.001, OR = 79.3; 95% CI:
18.73-688.3).
Smoking (>15 versus <15 yrs), level of education (illiterate or
read/write versus school education), age (40-60 versus
20-40 yrs), type of job (production versus packing), and duration
of work (>15 versus <15 yrs) were all observed to be highly
associated (p-values < 0.001; OR > 4.4) with increased
neuropsychological symptoms.
Logistic regression revealed that the strongest predictors of
neuropsychological symptoms were type of job performed
(production or packing) and duration of work (>15 yrs).
Not clear whether any confounders were taken into account in
the logistic regression analysis.
Juarez-
Perez et al.
(2014)
Cross-sectional study of
77 solvent exposed paint
factory workers in
Mexico and 84 control
subjects drawn from
donors at a local blood
bank. All exposed
participants were male.
Exposed workers were
given a questionnaire to
Hearing
assessments were
conducted for each
participant and
hearing loss
prevalence was
calculated in
exposed and
unexposed
populations.
134 workplaces at various
production sites were
examined; air samples from
the worker's respiratory
zone were collected from
workers during all shifts of a
single workday. Toluene,
xylene, and benzene were
listed as exposures, but not
TMB isomers.
Univariate analysis of quantitative variables was performed.
Mean differences were analyzed via Student's t and X2 tests.
Robust multiple linear regression was used and were adjusted for
age, environmental noise, diabetes, hypertension/hyperlipidemia,
ototoxic drugs, and alcohol.
19.5% of solvent-exposed workers had hearing loss.
This document is a draft for review purposes only and does not constitute Agency policy.
C-54 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
determine demographic
characteristics, hearing
pathologies, chronic
disease status, ototoxic
medication usage, and
other factors
(alcohol/drug usage,
motorcycle usage, etc.).
Controls were
questioned regarding
solvent exposure.
Brainstem auditory-
evoked potentials
were also recorded.
Noise measurements were
also collected at each
worksite
Robust multiple linear regression showed that hearing loss (low,
high, and all frequencies) was significantly increased in left and
right ears in exposed workers, and that age and chronic pathology
were also related to hearing loss; 24-39% of hearing loss
variability was explained by the regression model. Exposure to
environmental noise did not appear to increase hearing loss.
Multiple linear regression also revealed increased latencies in
brainstem auditory-evoked potentials, although the R2 values
were much lower (0.2-12.4).
Maule et al.
(2013)
Cross-sectional study of
37 male and female
active duty Air Force
personnel (N = 23 with
occupational exposure
to JP-8 exposure, N = 14
with little to no JP-8
exposure). Each
participant completed a
questionnaire regarding
demographic data, work
history, and other
lifestyle and/or physical
characteristics.
Postural sway was
analyzed in all
participants.
Evaluations were
conducted pre- and
post-shift.
Breathing zone sampling
was conducted on all
participants; total
hydrocarbons and
naphthalene were reported.
Pre- and post-shift urine
samples were taken and
analyzed for metabolites of
naphthalene. TMB isomers
were not explicitly noted in
the study results.
Multiple linear regression were used to investigate associations
between JP-8 exposure and postural sway. Measures of postural
sway (total angular area and mean path velocity) were used as
the dependent variables in four models of stance tasks: eyes
open, eyes closed, eyes open, foam support, and eyes closed,
foam support. Covariates considered included age, smoking
status, and body mass index.
The high exposure group was more likely to be male than the low
exposure group (p < 0.05). Increased sway was noted in tests
involving foam support versus no foam for both eyes open and
eyes closed tasks. Regression models using total hydrocarbons,
naphthalene, 1-naphthol, or 2-naphthol did not demonstrate
statistically significant associations between exposure and sway.
Pre-shift measures of sway were positivity associated with post-
shift measures. Younger age was also predictive of balance
control. Although the regression models did not indicate an
association between sway and exposure metrics, they explained
39-62% of variance in the outcome measurements.
This document is a draft for review purposes only and does not constitute Agency policy.
C-55 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
Pratt et al.
(2000)
Cross-sectional study of
48 male subjects with no
history of neurological
or ophthalmological
impairment; 31 subjects
were occupationally
exposed to gasoline in
the workplace and
17 had no occupational
exposure to gasoline.
Participants were
tested for pattern-
reversal VEPs and
SVEPs.
Exposure levels of each
participant were
determined using personal
samplers. No participants
were reported to be
exposed to levels of
benzene, xylenes, toluene,
carbon tetrachloride, or
methyl-te/t-butyl ether
above legal exposure levels
(which exposure values
used were not noted). TMB
isomers were not explicitly
noted in study.
The effect of gasoline on latencies of SVEP or VEP was assessed
via ANOVA, with subject group as a factor (N = controls, L = low,
laboratory exposure, A & B = high exposure groups).
Latencies corresponding to retinal activity, optical nerve activity,
scalp distribution with optic radiation, and cortical activity were
increased when comparing gasoline-exposed workers to
unexposed workers (p-value < 0.05).
Ruiiten et
al. (1994)
Cross-sectional study of
28 shipyard painting
employees exposed to
solvents and 25 control
workers with no
exposure to solvents.
Participants were
screened on education
(higher education
excluded, control only),
alcohol consumption,
and occupational
exposure to neurotoxic
substances (control
group only).
Symptoms were
assessed via a
questionnaire
concerning various
neurotoxic
symptoms
(including mood
changes, fatigue,
sleep disturbances,
etc.). Neuro-
physiological
examinations were
also conducted
(sensory and motor
nerve conduction
velocity). A
psychometric
examination
consisting of
computerized tasks
An individual cumulative
exposure index was
calculated for each
participant. Environmental
monitoring (all solvents)
and biological monitoring
(methylhippuric acid) were
used to estimate exposure
levels. Cumulative
exposure indices were
calculated for five broad
categories of painting tasks.
Cumulative exposure for all
painters was 495 mg
methylhippuric acid/g
creatinine.
Differences in effects between painters and controls were
investigated using ANCOVA, with age and alcohol used as
confounders. The association between the cumulative exposure
index and neurological effects was investigated using multiple
linear regression.
Mood changes, equilibrium complaints, sleep disturbances, and
solvent-related complaints were increased in painters compared
to controls (p < 0.05).
Differences in peripheral nerve function was statistically
significant between painters and controls, particularly in the
peroneal nerve (p < 0.05).
Neurobehavioral test performance indicated a detrimental effect
of solvent exposure on color word vigilance, symbol digit
substitution, and hand-eye coordination (p < 0.05).
This document is a draft for review purposes only and does not constitute Agency policy.
C-56 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
was also
administered.
Lee et al.
(2005)
Cross-sectional study of
workers at a shipyard in
Ulstan, Korea;
180 workers included in
study along with
60 randomly selected
non-exposed controls.
Workers were pre-
screened for educational
level, absence of
alcohol/drug
dependency, and lack of
existing neurological
disease.
Questionnaire was
administered to pre-
screen workers and
to collect additional
data on age and
work duration.
A number of tests
were administered
to judge
neurological
function: simple
reaction time,
symbol digit
substitution, and
finger tapping speed
(dominant and non-
dominant hand).
Data on exposure were
collected from 61 workers
who wore passive
dosimeters on 3 work days.
Workers exposed to
3.71 ±3.95 ppm 1,2,4-TMB
(geometric mean,
18.25 mg/m3, geometric
standard deviation = 19.43),
range = 0.2-57.0 ppm.
Average exposure duration:
16.5 ± 9 yrs in exposed
workers.
A cumulative exposure index was calculated for each worker.
Student t-test was used to determine statistical significance of
results in exposed workers compared to non-exposed workers.
Multiple regression analysis was performed to ascertain and
control for confounders.
Exposure had a significant effect on symbol digit substitution and
finger tapping speed in multiple regression analysis of all subjects.
Age and education were observed to be statistically significant
confounders.
After adjusting for age and education, painters were observed to
have statistically significantly slower symbol digit substitution and
finger tapping speeds (dominant and non-dominant) compared to
controls.
Symbol digit substitution and finger tapping speed also
statistically significantly slower in subjects when comparing
workers with >20 yrs of exposure to workers with <10 yrs of
exposure.
Sulkowski
et al.
(2002)
Cross-sectional study of
Polish workers in a
factory in which paints
and varnishes were
produced; 61 exposed
workers were included
in the final analysis
following a
questionnaire and
otolayrngological
examination. Subjects
Comprehensive
evaluation of
hearing: air and
bone pure tone
audiometry,
impedance
audiometry with
tympanometry,
acoustic reflex
threshold,
otoacoustic
Exposure was assessed via
individual dosimeters and
biological monitoring of
blood and urine. TMB
isomers were reported to
be the most commonly
detected contaminants in
air. Blood levels of TMB
isomers ranged from 0.60 to
70.14 ng/dL.
Student's t-test was to analyze differences between groups.
Linear regression was used to investigate the association of
exposure to single contaminants with specific effects.
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-57 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
with middle ear damage,
previous ear surgery,
head injury, ototoxic
drug treatment,
diabetes, hypertension,
neurological disease,
alcohol/drug abuse, and
a history of noise
exposure were
excluded; 40 non-
exposed workers were
included as controls.
emissions, and
electronystagmo-
graphic
investigations.
Average duration of
exposure: 15.8 ± 9.1 yrs.
High frequency hearing loss, as indicated by pure tone
audiometry was detected in 42% of exposed individuals versus 5%
of the control population.
All three TMB isomers (measured in subjects' breathing zones)
were observed to be statistically significantly associated with
distortion product otoacoustic emissions (p-values < 0.05). These
associations were reported as the strongest amongst the
detected contaminants.
Fuente et
al. (2013)
Cross-sectional study in
Santiago, Chile:
30 participants each
(15 males/15 females) in
the xylene-exposed and
control groups.
Otoscopy was
performed to exclude
participants with
external ear damage, a
questionnaire was
provided to collect data
on participants' history
of neurological,
metabolic,
cardiovascular disease,
otitis media, or previous
excessive noise
exposure. A report of
one or more of the
previous was used to
Comprehensive
evaluation of
hearing:
audiological
assessments,
masking level
difference test,
pitch pattern
sequence test, and
dichotic digit test.
Workers were interviewed
to collect self-reports of
occupational xylene
exposure; mean duration of
exposure to xylene in the
workplace was 11.8 ±
10.5 yrs.
Air samples were also
collected at different work
stations of the xylene-
exposed workers; mean air
concentration was
36.5 ± 66.6 mg/m3.
Urine samples were
collected post-shift on the
last day of the working
week and analyzed for
methylhippuric acid: mean
concentration was
Student's t test, ANCOVA (with age and hearing levels as
covariates), and Spearman rank correlations (for stratified
analyses) were used to analyze the differences in hearing
between xylene-exposed workers and controls.
Xylene-exposed workers consistently had increased measures of
auditory dysfunction compared to controls: worse audiometric
thresholds; greater latency in the auditory brainstem response;
and decreased performance in the pitch pattern sequence,
dichotic digits test, and hearing in noise test (p-value < 0.01).
Simple linear regression demonstrated that increasing levels of
methylhippuric acid are positively correlated with binaural
hearing thresholds (R2 = 0.32, p-value < 0.01).
When stratifying participants based on cumulative exposure
(low = 96.8 ± 26.36 mg*yr, medium = 434.9 ± 289.9 mg*yr, and
high = 5,630.2 ± 3,150 mg*yr), the high exposure group had
statistically significantly higher binaural hearing threshold
compared to low and medium exposure groups (p-value < 0.05).
There was also a statistically significant difference between the
This document is a draft for review purposes only and does not constitute Agency policy.
C-58 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
exclude participants
from the study.
216.3 ± 44.2 mg per g
creatinine.
Cumulative exposure was
calculated by multiplying
methylhippuric acid
concentration by duration
of exposure.
low and high exposure groups regarding hearing in noise tests
(p-value < 0.05).
Quevedo et
al. (2012)
Cross-sectional study of
gas station workers in
Santa Maria, Brazil:
21 participants
(18 males/3 females).
Otoscopy was
performed to identify
conditions that would
alter test results.
Exclusion criteria for
participants were:
history of ear problems,
abnormal auditory
thresholds, age >40 yrs,
exposure to noise,
organic solvents, or
pesticides, and use of
ototoxic medications.
Threshold tonal
audiometry,
brainstem auditory
evoked potential
testing, and acoustic
reflex testing.
No explicit exposure
analysis was conducted.
Analyses based on
comparisons of exposed
group (i.e., gas station
workers) to the normal
range of response for the
various tests. Duration of
exposure was also used in
some analyses.
Binomial test was used to test differences in absolute latency and
interpeak differences in the brainstem auditory evoked potential
test.
Right ear: 19 and 29% of participants had abnormal Wave 1 and III
absolute latencies; no difference was noted for Wave V. Only the
difference in Wave 1 latency was statistically significant
(p = 0.025). None of the latencies in the interpeak intervals (l-lll,
lll-V, 1—V) were statistically different.
Left ear: 14 and 5% of participants had altered Wave 1 and V
latencies (p = 0.015 and 0.0001, respectively). Although 38% of
participants had altered Wave III latencies, these alterations
failed to achieve statistical significance. None of the latencies in
the interpeak intervals were statistically different.
Duration analysis: Among workers exposed for <3 yrs, no
statistically significant differences were noted for absolute
latencies in the right ear. However, the interpeak interval change
for Waves lll-V was statistically significant. A statistically
significant alteration in the absolute latency of Wave V was
observed in the left ear (p = 0.0257).
For workers exposed between 3 and 5 yrs, no statistically
significant effects were noted in either ear for absolute latencies
or interpeak interval changes.
This document is a draft for review purposes only and does not constitute Agency policy.
C-59 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Study
citation
Study design/study
population
Outcome
measured
Exposure assessment
Results
For workers exposed >5 yrs, statistically significant effects were
noted for the l-V interpeak difference in the right ear, the
absolute latency in Wave 1 in the left ear, and the 1II—V interpeak
interval in the left ear.
1 ANCOVA = analysis of covariance; ANOVA= analysis of variance; fMRI = functional magnetic resonance imaging; JP-8 = jet propulsion fuel 8; OR = odds
2 ratio; PMR = proportional mortality ratio; SMR = standardized mortality ratio; SVEP = short-latency visual evoked potential; VEP = visual evoked
3 potential; VOC = volatile organic compound.
This document is a draft for review purposes only and does not constitute Agency policy.
C-60 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
C.4. ANIMAL TOXICOLOGY STUDIES
Tables C-17 through C-46 provide study details for animal toxicology studies.
Table C-17. Characteristics and quantitative results for Adenuga et al. (2014)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats
M & F
10 rats/dose group
Gavage
0, 50, 200, and
600 mg/kg
body weight/d
1,3,5-TMB
Single exposure, once a day,
5 d/wk, for 90-91 d, for
65-66 doses
Additional study details:
• Rats were given one oral dosage of 1,3,5-TMB each day for 5 d/wk, for 90-91 d.
• Rats were randomized and assigned to five groups according to sex and body weight.
• Two deaths were reported, but were considered to have resulted from dosing errors and not related
to treatment.
• No statistically significant effects on mean body weight were observed in any of the treated groups as
compared to the vehicle control group.
• Liver and kidney weights increased, but were considered adaptive effects.
• All histopathology findings at termination of dosing were determined to be unrelated to treatment
but typical of spontaneous lesions common to the rat strain.
• The NOAEL was 600 mg/kg-d.
Analysis of dosing solutions of 1,3,5-TMB in corn oil
wk r
Wk 7b
Wk 13b
0
Below detection limit
-
-
10 (50 mg/kg)
9.78
9.60
9.92
40 (200 mg/kg)
39.04
-
-
120 (600 mg/kg)
120.4
128.2
114.6
aValues represent means of duplicate analysis for 0 mg/mL and six replicates for 10, 40, and 120 mg/mL
bValues represent means of duplicate analysis.
Experimental design
Group
Dose (mg/kg-d)
Number of rats (M & F)
1
0 (corn oil vehicle control)
10+10
2
50
10+10
3
200
10+10
4
600
10+10
5
600 (28-d recovery group)
10+10
This document is a draft for review purposes only and does not constitute Agency policy.
C-61 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Growth curves for male
rats following 90-d gavage exposure (including 28-d recovery period) to 1,3,5-TMB.
700
i
64M1
1
_ SOD
S
1
8
* 300
I
| Recovery
] Ph»w?
'' Control
~ 200
5 0 ji'day
1
-------
Supplem en tal Information—Trim ethylbenzenes
Values obtained at a terminal sacrifice in a 90-d gavage study of 1,3,5-TMB with 28-d recovery
Mean clinical chemistry
Exposure (mg/kg-d)
Observation
0 (control)
50
200
600
600 (recovery)
Males
Protein (g/dL)
6.0 ±0.38
5.9 ±0.24
6.0 ±0.31
6.1 ±0.42
6.0 ±0.25
Albumin (g/dL)
3.6 ±0.23
3.6 ±0.19
3.7 ±0.19
3.8 ±0.22
3.7 ±0.09
Glucose (mg/dL)a
150.2 ± 22.80
134.6 ± 15.11
136.9 ± 15.76
121.1 ± 13.14*
168.4 ± 26.39
Cholesterol (mg/dL)
38.2 ±6.83
33.1 ±9.13
31.6 ±9.93
45.3 ± 15.99
35.3 ± 10.10
Sodium (meq/L)
142.4 ± 1.49
142.7 ±0.65
143.0 ± 1.40
142.4 ± 1.32
141.6 ± 1.30
Potassium (meq/L)
4.32 ±0.397
4.51 ±0.339
4.37 ±0.328
4.54 ±0.270
4.33 ±0.240
Chloride (meq/L)
105.3 ± 2.59
105.3 ± 2.33
106.0 ± 1.72
106.2 ±2.18
104.7 ± 0.88
Phosphorus (mg/dL)
6.5 ±0.64
6.7 ±0.80
7.0 ±0.68
7.6 ±0.58*
5.8 ±0.59
Total bilirubin (mg/dL)
0.4 ±0.12
0.4 ±0.10
0.5 ± 0.09
0.5 ±0.14
0.5 ±0.09
AP (IU/1)
107 ±28.1
112 ±26.5
121 ±33.7
156 ± 56.2*
77 ± 20.5
ALT (IU/1)
29 ± 6.4
30 ± 9.8
25 ±7.0
33 ±9.1
25 ± 4.4
AST (IU/1)
72 ± 18.9
91 ±31.9
86 ± 25.5
85 ± 25.0
89 ± 16.7
Females
Protein (g/dL)
6.2 ±0.44
6.3 ±0.41
6.6 ±0.69
6.5 ±0.68
6.3 ±0.66
Albumin (g/dL)
4.1 ±0.29
4.3 ±0.36
4.5 ±0.58
4.5 ±0.56
4.3 ±0.51
Glucose (mg/dL)
131.8 ±7.65
136.4 ± 11.72
140.1 ± 14.48
132.8 ± 15.91
150.7 ± 19.18
Cholesterol (mg/dL)b
36.2 ±8.83
35.2 ±6.64
38.8 ±6.24
51.2 ± 17.84*
28.7 ± 12.93
Sodium (meq/L)c
142.1 ± 1.10
141.6 ±0.96
141.7 ± 2.07
138.9 ± 2.83*
140.9 ± 1.47
Potassium (meq/L)
3.94 ±0.195
4.13 ±0.200
4.01 ±0.119
3.86 ±0.292
4.06 ±0.259
Chloride (meq/L)d
105.9 ± 2.32
106.2 ± 1.63
106.1 ± 1.05
103.0 ± 3.81*
107.0 ± 1.68
Phosphorus (mg/dL)
6.1 ± 1.08
6.1 ± 1.27
6.4 ± 1.18
7.5 ± 1.24*
5.3 ±0.80
Total bilirubin (mg/dL)
0.5 ± 0.08
0.5 ±0.10
0.4 ± 0.08
0.5 ±0.07
0.5 ±0.07
AP (IU/L)
59 ± 14.8
57 ± 10.3
55 ± 14.9
78 ±24.5
38 ± 10.1
ALT (IU/L)
21 ±2.3
22 ± 4.0
23 ±7.3
24 ±4.1
27 ±7.1
AST (IU/L)
60 ± 16.5
75 ± 18.6
62 ± 15.2
60 ± 15.0
77 ±21.4
*p < 0.05.
aGlucose historical control range: 97.4-155.7 mg/dL (N = 20).
bCholesterol historical control range: 32-112 mg/dL (N = 20).
cSodium historical control range: 141-148meq/L (N = 20).
dChloride historical control range: 105-111 meq/L (N = 20).
AP = alkaline phosphatase; ALT = alanine aminotransferase; AST = aspartate aminotransferase.
This document is a draft for review purposes only and does not constitute Agency policy.
C-63 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Values obtained at terminal sacrifice in a 90-d gavage study of 1,3,5-TMB with 28-d recovery
Mean hematology
Exposure (mg/kg-d)
Observation
0 (control)
50
200
600
600 (recovery)
Males
WBCs (xl06/mm3)
9.1 ±2.70
8.1 ±2.50
8.1 ± 1.74
7.7 ± 1.76
7.8 ± 1.24
RBCs (xl06/mm3)
8.94 ±0.375
8.50 ± 0.4,863
8.98 ±0.565
8.72 ±0.275
8.51 ±0.423
Hemoglobin (g/dL)
15.6 ±0.52
15.3 ±0.76
15.8 ±0.77
15.4 ±0.53
15.4 ±0.58
Hematocrit (%)
43.9 ± 1.65
42.2 ±2.72
44.1 ±2.12
43.3 ± 1.60
41.6 ± 1.99
MCV (xlO-15 L)
49.1 ± 1.17
49.7 ± 1.09
49.2 ± 1.76
49.6 ± 1.66
49.0 ± 1.62
MCH (pg)
17.5 ±0.45
18.0 ±0.73
17.7 ±0.85
17.7 ±0.68
18.2 ±0.61
MCHC (%)
35.6 ±0.67
36.3 ± 1.07
35.9 ±0.60
35.6 ±0.67
37.1 ±0.60
Platelet count (xlO6 /mm3)
1,092 ± 134.1
1,098 ± 120.8
1,041 ± 100.9
1,125 ± 145.9
1,083 ± 112.6
Females
WBCs (xl06/mm3)
5.5 ±2.05
5.6 ± 1.53
5.4 ± 1.64
5.7 ± 1.99
4.6 ± 1.55
RBCs (xl06/mm3)
7.88 ±0.729
8.01 ±0.354
7.90 ±0.578
8.34 ± 0.548
7.70 ±0.423
Hemoglobin (g/dL)
14.8 ± 0.88
15.0 ± 0.48
15.2 ±0.82
15.3 ±0.78
15.1 ±0.57
Hematocrit (%)
41.0 ±3.15
41.4 ± 1.91
41.9 ±2.93
43.3 ±2.33
39.9 ± 1.67
MCV (xlO-15 L)
52.1 ± 1.65
51.7 ± 1.18
53.0 ± 1.03
52.0 ± 1.24
51.9 ± 1.33
MCH (pg)
18.9 ±0.89
18.7 ±0.67
19.2 ±0.53
18.4 ±0.68
19.6 ±0.78
MCHC (%)
36.2 ±0.79
36.2 ±0.86
36.3 ±0.83
35.4 ±0.54
37.7 ±0.64
Platelet count (xl06/mm3)
1,094 ± 153.3
1,089 ± 132.0
1,011 ±97.2
1,053 ± 125.7
1,008 ± 105.7
WBC = white blood cell; RBC = red blood cell; MCV = mean cell volume; MCH = mean corpuscular hemoglobin;
MCHC = mean corpuscular hemoglobin concentration.
This document is a draft for review purposes only and does not constitute Agency policy.
C-64 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Differentials obtained at terminal sacrifice in a 90-d gavage study of 1,3,5-TMB with a 28-d recovery
Mean absolute WBC
Exposure (mg/kg-d)
Observation
0 (control)
50
200
600
600 (recovery)
Males
Polynuclear neutrophils
(xl06/mm3)
1.8 ± 1.07
1.7 ± 1.10
1.4 ±0.36
1.5 ±0.75
1.0 ±0.29
Lymphocytes (xl06/mm3)
7.1 ±2.78
6.2 ±2.16
6.4 ± 1.59
6.0 ±2.16
6.6 ± 1.23
Monocytes (xl06/mm3)
±0.09
± 0.09
0.3 ±0.17*
0.2 ±0.18*
0.2 ±0.10
Eosinophils (xl06/mm3)
±0.06
0.1 ±0.09
0.0 ±0.07
0.0 ± 0.05
0.1 ±0.07
Females
Polynuclear neutrophils
(xl06/mm3)
0.8 ± 0.48
0.7 ±0.32
0.9 ±0.69
1.0 ±0.39
0.7 ± 0.45
Lymphocytes (xl06/mm3)
4.6 ± 1.93
4.7 ± 1.52
4.2 ± 1.52
4.4 ± 2.08
3.7 ± 1.34
Monocytes (xl06/mm3)
±0.14
0.1 ±0.10
0.1 ±0.08
0.2 ±0.17
0.2 ±0.11
Eosinophils (xl06/mm3)
±0.07
0.1 ±0.07
0.1 ±0.09
0.1 ±0.09
0.0 ± 0.07
*p < 0.05.
Weights obtained at terminal sacrifice in a 90-d gavage study of 1,3,5-TMB with 28-d recovery
Mean absolute and relative kidney and liver weights
Exposure (mg/kg-d)
Observation
0 (control)
50
200
600
600 (recovery)
Males
Mean absolute (g)
Kidney
3.92 ±0.326
3.95 ±0.262
4.10 ±0.610
4.16 ±0.464
4.05 ±0.491
Liver
19.28 ± 1.843
8.91 ±3.074
18.38 ± 2.885
20.90 ±3.313
17.38 ±2.222
Mean relative (g)
Kidney
0.65 ±0.052
0.68 ±0.052
0.71 ±0.082
0.74 ± 0.045*
0.68 ±0.039
Liver
3.20 ±0.158
3.23 ±0.336
3.19 ±0.402
3.71 ±0.288*
2.93 ±0.274
Females
Mean absolute (g)
Kidney
2.34 ±0.314
2.23 ±0.228
2.38 ±0.116
2.51 ±0.264
2.38 ±0.248
Liver
9.44 ± 1.60
9.13 ±0.77
10.05 ± 0.96
11.78 ± 1.44
9.71 ± 1.41
This document is a draft for review purposes only and does not constitute Agency policy.
C-65 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mean relative (g)
Kidney
0.76 ±0.059
0.71 ±0.088
0.76 ±0.051
0.82 ± 0.059
0.71 ±0.040
Liver
3.04 ±0.365
2.90 ±0.330
3.19 ±0.357
3.82 ±0.223
2.88 ±0.207
*p < 0.05.
Gross necropsy observations obtained at terminal sacrifice in a 90-d gavage study of 1,3,5-TMB with 28-d
recovery (10 rats/sex/group)
Male (mg/kg-d)
Female (mg/kg-d)
Observation
0 (vehicle
controls)
50
200
600
600 (recovery
rats)
0 (vehicle
controls)
50
200
600
600 (recovery
rats)
Mandibular lymph nodes
Red/dark red
0
0
1
0
1
1
0
0
0
0
Enlarged
1
0
1
0
1
0
0
0
0
0
Liver
Pale
0
0
0
1
0
0
0
0
0
0
Lung
Enlarged
0
0
r
0
0
0
0
0
0
0
Thymus
Focus, red
0
0
0
0
0
0
1
0
0
0
Mottled
0
0
0
1
0
0
0
0
0
0
Adrenals
Small, unilateral
0
1
0
0
0
0
0
0
0
0
Accidental death due to gavage error.
Histopathological findings in the kidney and liver obtained at terminal sacrifice in a 90-d gavage study of
1,3,5-TMB
Male (mg/kg-d)
Female (mg/kg/-d)
Observation
0
50
200
600
0
50
200
600
Liver/chronic inflammation
Incidence (%)
40
_a
-
30
50
-
-
50
Mean grade
0.40
-
-
0.30
0.50
-
-
0.60
Liver/necrosis
Incidence (%)
0
-
-
0
10
-
-
0
Mean grade
0
-
-
0
0.10
-
-
0
Kidney mineralization
Incidence (%)
0
-
-
0
70
-
-
70
Mean grade
0
-
-
0
0.80
-
-
0.70
This document is a draft for review purposes only and does not constitute Agency policy.
C-66 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Kidney nephropathy
Incidence (%)
30
-
-
10
0
-
-
0
Mean grade
0.30
-
-
0.10
0
-
-
0
aDose group not examined.
Histopathological findings in the liver of rats obtained at terminal sacrifice in a 14-d gavage study of 1,3,5-TMB
Male (mg/kg-d)a
Female (mg/kg-d)a
Observation
0
50
200
600
Rb
0
50
200
600
Rb
Liver/chronic inflammation
Incidence (%)
30
20
10
20
20
60
20
10
30
20
Mean grade
0.23
0.20
0.10
0.05
0.20
0.25
0.20
0.10
0.13
0.20
Liver/necrosis
Incidence (%)
0
0
0
10
0
0
0
0
0
0
Mean grade
0
0
0
0.15
0
0
0
0
0
0
Liver/centrilobular hypertrophy
Incidence (%)
0
0
0
100
0
0
0
0
30
0
Mean grade
0
0
0
1.00
0
0
0
0
0.30
0
aTotal of 10 rats examined per group.
bRecovery rat (600 mg/kg body weight; rats sacrificed 14 d after the last treatment).
NOAEL
LOAEL
LOAEL Effect
600 mg/kg-d (NOAELhed =
105 mg/kg-d)
Not identified
Not applicable
Comments: The highest dose was considered the no-observed-adverse-effect level (NOAEL), as the systemic
effects were regarded as adaptive responses to chemical exposure and not relevant to human health hazard. A
lowest-observed-adverse-effect level (LOAEL) could not be inferred from the study.
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-67 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-18. Characteristics and quantitative results for Battigetal. £1958}
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Rats
M
8 rats/dose
Intraperitoneal (i,p.)
injection
0, 200, 500, and 1,700 ppm
(0, 984, 2,460, 8,364 mg/m3)
TMB mixture
4 mo; 8 hrs/d, 5/wks
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, RBC 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.
U)
U)
'o
£
<
y [Control rats
i\\W?1 Exposed rats
Effect of iong-term exposure to TMB (about 1,700 ppm [8,364 mg/m3]) on the growth of rats.
340
330
320
310
300
290
280
270
260
250
240
230
220
1
Days of Exposure
¦ ¦¦¦¦¦¦¦¦!
1 Dec 1 Jan 1 Feb 1 Mar
Dates in Treatment
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.
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,
C-68 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Behavior of the relative number of lymphocytes in TMB-exposed rats (exposure: about 1,700 ppm [8,364 mg/m3]).
90
80 I- i
70
60
a>
>»
o
o
g- 50
a>
o
i_
CD
CL
40
30
20
10
Exposed rats
Control rats
Days of exposure
Date: 1 Nov 1 Dec 1 Jan 1 Feb 1 Mar
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958).
1 April
Average intake of food by the rats during experimental exposure to TMB mixture
Month
Number of days
exposed per month
Average daily food intake
(g/100 g body weight 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
Source: Reproduced with permission of Springer-Verlag (Battie et al.. 1958).
This document is a draft for review purposes only and does not constitute Agency policy.
C-69 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
CO
a>
o
o
JsC
3
CD
O.
o
CD
CD
O
>—
CD
Q_
I | Exposed rats
Behavior of the relative number of neutrophil leukocytes in TMB-exposed rats (exposure: about 1,700 ppm
[8,364 mg/m3]).
60
50
40
30
20
10
p = 0.05
±
Date: 1 Nov 1 Dec
Days of exposure
!_L
1 Jan 1 Feb
1 Mar 1 April
Source: Reproduced with permission of Springer-Verlag (Battig et al„ 1958).
Average intake of drinking water by rats during experimental exposure to TMB
Month
Number of days
exposed per month
Average daily food intake
(g/100 g body weight per 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
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,
C-70 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Specific gravity of spontaneous and dilution urines in TMB-exposed rats (exposure: about 1,700 ppm [8,364 mg/m3]).
1
2 3 4
Time in hours after dilution test
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958).
Effect of TMB inhalation on urinary phenol excretion in the rat
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
Free
Bound
1,700
1,700
1,700
15
15
15
10
3
9
Total
Free
Bound
500
500
500
21
21
21
8
8
21
Total
Free
Bound
200
200
200
10
10
10
10
10
Not increased
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.
C-71 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Health effect at LOAEL
NOAEL
LOAEL
Increased urinary excretion of free
and total phenols
0 ppm
200 ppm (984 mg/m3)
Comments: Battig et al. (1956) is published in German. However, Battig et al. (1958) presents an English-translation of
the results originallv presented in Battig et al. (1956). As such, a separate studv summarv table is not provided for Battig
et al. (1956). Four of the eight rats in the long-term inhalation experiment died and were subseauentlv replaced within
the first 2 wks. 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.
C-72 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-19. Characteristics and quantitative results for Carrillo etal. (2014)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure
duration
Wistar rats
M & F
18 males and
18 females by
weight/dose group
Inhalation
2,000, 4,000, or
3,000 mg/m3
white spirit
6 hrs/d, 5 d/wk
for 13 wks
Additional study details:
• Rats were exposed to nominal concentrations of 2,000, 4,000, or 8,000 mg/m3 white spirit for 6 hrs/d,
5 d/wk, for a total of 13 wks.
• Rats were distributed into groups by weight between 10 and 13 wks of age.
• All rats survived treatment.
• Terminal body weights of high-exposure group animals were significantly below control values.
• Clinical and hematological observations were statistically different, were small, and were within
normal physiological limits.
• The NOAEL was 4,000 mg/m3.
Approximate hydrocarbon composition of white spirit over the past 40 yrs in terms of carbon number and
hydrocarbon constituents: normal and n- and iso-paraffins (naphthenics iso-alkanes, cyclo-alkanes) and aromatics.
Hydrocarbon
constituents by carbon
number
Pre-1980
Post-1980
Kuwait sample
Arabian light
sample
EU sample 1982
EU sample 1985
EU sample 2011
Paraffins (n + iso)
Approximate constituent concentrations in % w/w
C8
<0.5
<0.5
<0.5
<0.5
-
C9
13
13
10
12
7
C10
33
33
24
24
20
Cll
13
12
16
15
17
C12
2
2
3
3
3
C13
-
-
-
-
<0.1
Sum P
61
60
53
54
47
Naphthenes
C8
<0.5
<0.5
<0.5
<0.5
<0.1
C9
5
5
7
8
8
C10
8
8
11
10
14
Cll
4
4
8
7
10
C12
1
1
2
2
2
Sum N
18
18
28
27
34
Aromatics
C8
1
1
1
2
<1
C9
11
11
9
9
8
C10
6
6
7
6
6
Cll
2
2
3
2
3
C12
-
-
-
-
<1
Sum A
20
20
20
19
18
This document is a draft for review purposes only and does not constitute Agency policy.
C-73 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Carbon number range
C7
<0.1
<0.1
-
-
-
C8
2
2
2
3
<0.1
C9
29
30
26
29
23
CIO
48
48
41
40
40
Cll
18
18
26
23
31
C12
3
3
5
4
5
C13
<1
<1
<1
<1
<0.1
*Predominantly branched mono-aromatics.
Physical and chemical properties of white spirit used in this study
Property
White spirit
Hydrocarbons, C9-C14 (2-25%
aromatics)
Physical state at 20°C and 1,013 hPa
Clear colorless liquid with pungent
odor
Clear colorless liquid with pungent
odor
Melting/freezing point (°C)
<-15 °C
<-20 (ASTM 5950)
Boiling range (°C)
150-200 (ASTM D1078)
110-270 (ASTM D86)
Relative density (g/cm3) at 15°C)
0.78 (ASTM D4052)
0.70-0.87 (ISO 12185)
Vapor pressure (kPa @ 20°C)
0.37
0.02-0.5
Flash point (°C)
44 (IP 170)
>23 (ASTM D56)
Flammability (% v/v)
0.7
0.6-0.7
Self-ignition temperature (°C)
293 (ASTM E659)
>200
Surface tension (mN/m)
26 (Du Novy ring)
22-28 (Wilhelmy plate method)
Viscosity (mm2/s)
1.1 (ASTM D445)
0.7-3.5
Odor threshold (mg/m3)
5-158 mg/m3
5-158 mg/m3
Additional descriptors for the white spirit test sample
Parameter
Value
Specific gravity (15.6/15.6°C)
0.777
Color (Saybolt)
+30
Aniline point (°C)
56
Total sulfur (% w/w)
<0.0005
Kauri-butanol value
37
Copper corrosion
No. 1 strip
Molecular weight (g/mol)
~140
Hydrocarbon constituents of white spirit test sample
Constituent
Carbon range (at >5%)
Content (% w/w)
Paraffins (n- + iso)
C9-C11
56.0
Naphthenes
C9-C11
25.0
Aromatics
C9-C10
19.0
* Predominantly branched mono-aromatics.
This document is a draft for review purposes only and does not constitute Agency policy.
C-74 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Overall weekly mean vapor concentrations throughout the experimental period
Nominal concentration
Measured concentrations
(mg/m3)
(mg/m3)
ppm (v/v)
8,000
7,500 ± 395
1,293 ± 68
4,000
4,000 ± 119
690 ±21
2,000
2,000 ± 52
345 ±9
Mean clinical chemistry values after 13-wk exposure to white spirit
Exposure concentration (mg/m3)
Observation
Control
2,000
4,000
7,500
SD of single
observation
Males
Protein (g/L)
66.1
64.7
65.9
65.2
2.77
Urea (mm/L)
8.4
8.5
8.4
8.2
0.94
AP (IU)
76
75
79
13.8
ALT (III)
25
29
27
30
12.0
AST (IU)
40
41
44
46*
8.6
Na (mm/L)
146
147
146
147
1.2
K (mm/L)
5.5
5.7
6.1
5.9
0.73
CI (mm/L)
103
102
101
101
2.67
Albumin (g/L)
36.5
36.8
35.7
37.3
2.64
Bilirubin (mm/L)
2.83
3.06
3.28
3.06
0.76a
Glucose (mm/L)
3.26
n.d.
3.40
3.82
0.82a
Females
Protein (g/L)
65.6
67.7
69.2**
68.7**
3.45
Urea (mm/L)
10.1
9.7
9.7
9.3
1.89a
AP (IU)
54
58
60
-j
15.2
ALT (IU)
22
20
23
22
6.7
AST (IU)
43
39
42
42
12.4
Na (mm/L)
146
146
146
146
2.0C
K (mm/L)
5.5
5.0
5.9
5.9
l.llc
CI (mm/L)
105
106
105
105
2.0
Albumin (g/L)
39.7
40.3
41.4
42.3*
3.03
Bilirubin (mm/L)
3.28
3.25
3.56
3.33
0.47
Glucose (mm/L)
4.05
n.d.
3.84
3.87
0.20
*p < 0.05.
**p< 0.01.
aCage effect.
n.d. = not determined.
This document is a draft for review purposes only and does not constitute Agency policy.
C-75 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mean hematology values of male rats after 13-wk exposure to white spirit
Observation
Exposure concentration (mg/m3)
SD of single
observation
Control
2,000
4,000
7,500
Hemoglobin (g/100 mL)
15.2
14.9
14.5
14.6
1.00a
PCV (%)
42.4
41.2*
40.6**
40.3**
1.67
RBCs (xl06/cmm)
8.28
7.94*
7.76**
7.70**
0.37a
WBCs (xl03/cmm)
4.2
4.5
5.7
5.6
1.30a
MCV (n3)
50.9
52.1*
52.4*
52.0*
1.54
MCH (pg)
18
19*
19*
19*
0.6
MCHC (g/100 mL)
36
36
36
36
0.5
Prothombin time (sec)
16.0
16.0
16.0
16.2
0.62
KCCT (sec)
21.5
21.7
20.2
20.6
2.56
*p < 0.05.
**p< 0.01.
aCage effect.
PCV = packed cell volume; pg = picogram; KCCT = kaolin-cephalin coagulation time.
Organ weights after 13-wk exposure to white spirit
Exposure concentration (mg/m3)
SD of a single
observation
Observation
Control
2,000
4,000
7,500
Males
Absolute organ weights (g)
Kidney
2.84
3.25**
3.31**
3.40**
0.335
Liver
15.82
16.48
17.11
17.11
1.892
Spleen
0.89
0.94
1.10*
0.97*
0.22
Heart
1.20
1.27
1.25
1.23
0.107
Organ weights adjusted for terminal body weights
Kidney
2.74
3.20**
3.33**
3.53**
0.27
Liver
15.12
16.17
17.25**
17.98**
1.64a
Spleen
0.86
0.93
1.11**
1.00**
0.21
Heart
1.17
1.25
1.26
1.28*
1.13a
Females
Absolute organ weights
Kidney
1.80
1.82
1.90*
1.87*
0.130
Liver
8.69
9.33*
10.57**
0.775
Spleen
0.65
0.65
0.67
0.67
0.078
Heart
0.80
0.82
0.81
0.80
0.05
This document is a draft for review purposes only and does not constitute Agency policy.
C-76 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Organ weights adjusted for terminal body weights
Kidney
1.79
1.80
1.90*
1.90*
o.i;
Liver
8.67
9.16*
9.87**
10.79**
0.65
Spleen
0.65
0.64
0.67
0.69
0.07
Heart
0.80
0.81
0.81
0.81
0.02
*p < 0.05.
**p< 0.01.
aCage effect.
Statistically significant toxicological findings after 13-wk exposure to white spirit
Exposure concentration (mg/m3)
Males
Females
Observation
2,000
4,000
7,500
2,000
4,000
7,500
Body weight gain
-
-
D
-
-
D
Water intake
-
-
1
-
-
1
Clinical chemistry
AP
-
-
1
-
-
1
AST
-
-
1
-
-
Albumin
-
-
-
-
-
1
Protein
-
-
-
-
1
1
Hematology
PCV
D
D
D
-
-
-
RBC
D
D
D
-
-
-
MCV
1
1
1
-
-
1
MCH
1
1
1
-
-
-
WBC
-
-
-
-
1
1
Relative organ weights
Kidney
1
1
1
-
1
1
Liver
-
1
1
1
1
1
Spleen
-
1
1
-
-
-
Heart
-
-
1
-
-
-
Kidney
Hyaline droplets
1
1
1
NE
-
-
Tubular basophilia
1
1
1
NE
-
-
Spleen
Extramedulary
hematopoesis
NE
1
1
NE
1
Hemosiderin deposition
NE
1
1
NE
1
1 = increased compared to control; D = decreased compared to control; NE = not examined.
This document is a draft for review purposes only and does not constitute Agency policy.
C-77 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mean body weights of male (A) and female (B) rats at each of 13 wks of exposure; exposure levels were low
(2,000 mg/m3), medium (4,000 mg/m3), and high (7,500 mg/m3).
.103 -|
El = Ferrules
. 1
as
_c
¦P
"D
£
C
ra
250
1
"
, t s8!!
¦ " X X
t X x X
x *
I
X
1
X
* * ' *
X x
* Control
5
¦ Luw
A Mr-nium
200 ¦
X Him
¦
12 3 4 5 6 7
a
9
1G 11 12 13
Week ot exposure
550
3
1
»¦
:¦
5
D
C
?
5C0
4ca
1
1
A = Ma es
, « '
• • * X
¦ « *
v A
X *
¦
A
X
¦
A
X
« i ; !
A v
* v X X
X x
~Control
ISO
-
¦ Low
A. N'ed lih
300
XHigh
1
13 4 5 6 7
a
9
10 11 12 13
Week of esposure
Source: Carrillo et al.
(2014).
NOAEL
LOAEL
LOAEL effect
4,000 mg/m3
7,500 mg/m3
Lethargy, reduced weight in males
and females, increased male and
female AP, male AST, and female
albumin, increased male and female
kidney weights, increased female liver
weight, increased male spleen, liver,
and heart weights, increased male
erythropoetic activity, increased
female deposition of hemosiderin
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-78 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-20. Characteristics and quantitative results for Clark etal. (1989)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure
duration
Wistar rats
M & F
50 males/group
50 females/group
Inhalation
0, 450, 900, or
1,800 mg/m3
SHELLSOL A/
SOLVESSO 100
(1,2,4-TMB,
1,3,5-TMB, and
1,2,3-TMB)
6 hrs/d,
5 d/wk,
12 mo
Additional study details:
• Rats were exposed by inhalation to 50:50 SHELLSOL A/SOLVESSO 100, a mixture containing 1,2,4-TMB,
1,3,5-TMB, and 1,2,3-TMB for 6 hrs/d, 5 d/wk for 12 mo.
• Rats were sorted into two groups of 50 animals by sex.
• Animals were placed into stainless steel chambers with volumes of at least 8 m3 with ventilation of air
drawn from the laboratory by means of a fan to remove particulate and organic vapor impurities.
• Two male and two female control animals, and two male medium exposure animals died.
• Seven rats were removed during the exposure period and 30 rats were removed during the recovery
period due to sore hocks.
• No apparent biological significance of hematological changes were seen in males; however, they were
statistically significant. Mean cell hemoglobin concentration was increased in males up 2%.
• Animals tested at 1,800 mg/m3 had increased kidney and liver weights at 6 and 12 mo, but were
considered to be physiological adaptive responses.
• Male rats at the 1,800 mg/m3 appeared to be more aggressive/irritable.
• The NOAEL was 0 mg/m3.
Target concentrations and actual concentrations expressed as the overall means of the daily atmosphere
analyses
Exposure group
Concentration (mg/m3)
Target
Actual
Mean
SD
Control
0
0
-
Low
450
470
29
Medium
900
970
70
High
1,800
1,830
130
This document is a draft for review purposes only and does not constitute Agency policy.
C-79 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Inhalation exposure to SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Mean hematological values of cardiac blood
Observation
Male
Female
0
450
900
1,800
SD of a single
observation
0
450
900
1,800
SD of a single
observation
Hemoglogin
(g/100 mL)
14.4
14.6
13.9
14.5
0.80
14.0
14.0
13.9
14.0
0.71
HCT (%)
39.7
40.3
38.4
39.9
2.14
39.1
38.6
38.3
38.4
1.91
RBCs (xl06/cmm)
7.49
7.51
7.06
7.52
0.449
6.86
6.78
6.71
6.81
0.356
WBCs (xl03/cmm)
3.3
3.2
3.6
4.2
1.07
2.3
2.1
1.9
2.3
0.61
MCV (nm3)
53
54
55
53
1.7
57
57
57
56
1.1
MCH (pg)
19.5
19.7
20.0
19.6
0.47
21.1
21.0
21.0
20.9
0.43
MCHC (g/100 mL)
36.4
36.3
36.3
36.4
0.44
36.6
36.4
36.6
36.6
0.42
Prothombin time
(sec)
14.0
14.5
14.0
14.3
0.63
14.0
14.0
14.0
14.0
0.58
KCCT (sec)
20.8
21.1
20.3
19.7
2.35
22.4
21.5
22.0
22.5
2.47
Reticulocytes (%)
5.68
-
-
4.31
2.111
3.30
-
-
3.66
0.951
Osmotic fragility3
0% hemolysis
0.62
0.64
0.63
0.65
0.038
0.65
0.63
0.62
0.64
0.039
50% hemolysis
0.42
0.40*
0.40*
0.40*
0.015
0.44
0.42
0.41
0.44
0.026
100% hemolysis
0.29
0.26
0.27
0.26
0.027
0.30
0.28
0.29
0.30
0.030
aValues reported are % saline at which 0, 50, or 100% hemolysis occurred.
Inhalation exposure to SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Mean differential leucocyte values of cardiac blood mg/m3
Male
Female
Observation
0
450
900
1,800
SD of a
single
observation
0
450
900
1,800
SD of a single
observation
WBCs
(xl03/cmm)
3.3
3.2
3.6
4.2
1.07
2.3
2.1
1.9
2.3
0.60
Polymorph
neutrophils (%)
32
27
39
35
9.4
36
45
40
38
12.1
Lymphocytes
(%)
63
67
59
62
8.8
59
51
54
56
11.3
Monocytes (%)
3
3
2
3
1.6
3
3
4
3
2.3
Eosinophils (%)
3
3
1
1
1.7
2
2
2
3
1.7
Absolute value
neutrophils
(xl03/cmm)
1.1
0.9
1.5
1.3
0.63
0.8
1.0
0.8
0.9
0.38
Absolute value
lymphocytes
(xl03/cmm)
2.1
2.2
2.0
2.7*
0.62
1.4
1.1
1.0
1.3
0.47
This document is a draft for review purposes only and does not constitute Agency policy.
C-80 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Inhalation exposure to SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Mean clinical chemistry values of cardiac blood mg/mB
Male
Female
Observation
0
450
900
1,800
SD of a single
observation
0
450
900
1,800
SD of a single
observation
Protein (g/L)
63
64
64
64
1.9
66
69
68
66
3.7
Urea (mm/L)
8.6
8.6
8.8
9.0
1.13
8.7
8.4
8.4
9.0
1.63
Uric acid (mm/L)
0.14
0.11
0.11
0.13
0.068
0.11
0.11
0.08
0.09
0.049
AP (IU)
93
82
81
82
16.7
58
55
52
48
14.2
AST (IU)
65
57
45
60
24.1
68
58
66
78
37.8
ALT (IU)
56
49
44
52
21.5
61
48
62
66
23.7
Creatinine (nm/L)
68
68
73
74*
6.5
64
60
63
63
5.9
Bilirubin (nm/L)
2
2
2
1
1.4
2
2
2
2
0.6
Na+ (mm/L)
146
146
146
146
0.7
146
146
147
148**
1.4
K+ (mm/L)
5.5
5.9
5.5
5.5
0.71
5.9
5.4
5.4
5.6
0.99
CI" (mm/L)
107
105
105
105
1.8
104
105
105
105
1.9
Ca++(mm/L)
2.67
2.70
2.67
2.70
0.089
2.66
2.63
2.64
2.61
0.127
Inorganic P (mm/L)
1.89
1.45
1.40
1.51
0.168
1.46
1.29
1.46
1.45
0.198
Glucose (mm/L)
3.5
3.4
3.5
3.4
0.66
3.7
3.5
3.7
3.3
0.63
Albumin (%)
64.4
60.7
63.5
61.3
3.57
55.9
56.5
53.0
51.5*
4.18
*p < 0.05 = significance of the difference between treatment and control means.
**p< 0.01.
Inhalation exposure to SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Mean organ weights (g)
Male
Female
Observation
0
450
900
1,800
SD of a single
observation
0
450
900
1,800
SD of a single
observation
Initial body
weight
280
280
283
280
11.2
181
183
182
183
5.9
Brain
2.29
2.27
2.28
2.29
0.065
2.05
2.04
2.02
2.08
0.059
Heart
1.48
1.54
1.50
1.52
0.193
1.06
1.06
1.06
1.08
0.091
Liver
21.23+
20.23+
21.62+
23.51*+
2.447
12.89
12.40
12.63
13.20
1.232
Spleen
1.36
1.27
1.34
1.32
0.216
0.87
0.80
0.84
0.86
0.125
Kidneys
3.99
3.78
3.97
4.38*
0.488
2.51+
2.47+
2.49+
2.49+
0.214
Testes
3.79
3.76
3.77
3.78
0.238
-
-
-
-
-
+Adjusted for initial body weight.
*p < 0.05 = Significance of the difference between treatment and control means.
This document is a draft for review purposes only and does not constitute Agency policy.
C-81 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Inhalation exposure to SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Summary of gross necropsy findings of major organs
Male
Female
Observation
0
450
900
1,800
0
450
900
1,800
Liver
Exaggerated
lobular pattern
2
6
4
3
2
1
0
2
Red or
haemorrhagic
areas
0
0
0
0
0
0
0
2
Enlarged
1
0
0
0
0
0
0
0
Kidneys
Hydronephrosis
1
1
1
0
0
1
0
0
Granular surface
3
6
1
5
4
0
0
3
Enlarged
0
1
0
1
0
0
0
1
Patchy or pale
areas
0
1
0
0
1
0
0
1
Cyst
0
0
0
0
1
1
2
2
Lungs
Patchy or pale
areas
3
9
5
8
3
2
9
3
Red or
haemorrhagic
areas
3
3
1
4
1
0
0
4
Spleen
Patchy or pale
areas
1
0
0
0
0
0
0
0
Granular surface
0
0
0
0
1
1
0
0
Enlarged
0
0
0
1
0
0
0
0
Uterus
Dilated
-
-
-
-
0
3
1
0
Mass
-
-
-
-
0
0
0
1
Gonads
Cyst
0
0
0
0
4
6
5
5
Values are numbers of rats/group of 25 males, 25 females showing the lesion.
This document is a draft for review purposes only and does not constitute Agency policy.
C-82 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Inhalation of SHELLSOL A/SOLVESSO 100 after 6 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) mg/m3
Incidence and severity of histopathological lesions of kidney and lung
Observation
Male
Female
0
450
900
1,800
0
450
900
1,800
Kidney nephrosis
Normal (grade 0)
7
8
10
5
10
10
10
Increased
(grades 1-5)
3
2
0
5
0
0
0
0
Mean grade
0.4
0.4
0
0.6
0
0
0
0
Kidney mineralisation
Normal (grade 0)
10
10
10
10
1
2
1
1
Increased
(grades 1-5)
0
0
0
0
9
8
9
9
Mean grade
0
0
0
0
0.8
2.1
2.3
2.4
Pulmonary macrophage infiltration
Normal (grade 0)
6
8
5
5
8
5
7
5
Increased
(grades 1-5)
4
2
5
5
2
5
3
5
Mean grade
0.6
0.4
1.1
1.0
0.4
0.9
0.7
0.8
Alveolar wall thickening
Normal (grade 0)
5
5
5
2
4
0
4
4
Increased
(grades 1-5)
5
5
5
8
6
10
6
6
Mean grade
0.9
0.7
1.0
1.4
1.2
2.2
1.3
1.4
Values are numbers of rats/group of 10 males, 10 females affected at each grade.
Inhalation of SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Incidence and severity of histopathological lesions of the kidney and lung
Male
Female
Observation
0
450
900
1,800
0
450
900
1,800
Kidney nephrosis
Normal (grade 0)
1
3
1
1
14
8
10
7
Increased
(grades 1-5)
23
22
24
24
10
16
14
17
Mean grade
2.0
1.9
2.2
2.5
0.8
0.9
0.9
1.4
Kidney mineralisation
Normal (grade 0)
24
25
25
25
1
1
2
1
Increased
(grades 1-5)
0
0
0
0
23
23
22
23
Mean grade
0
0
0
0
2.0
2.2
1.8
2.0
This document is a draft for review purposes only and does not constitute Agency policy.
C-83 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Pulmonary macrophage infiltration
Normal (grade 0)
18
9
9
11
12
12
20
15
Increased
(grades 1-5)
7
16
16
14
12
12
4
9
Mean grade
0.5
1.3
1.3
1.3
1.1
1.1
0.4
0.8
Alveolar wall thickening
Normal (grade 0)
9
7
8
6
4
5
11
7
Increased
(grades 1-5)
16
18
17
19
20
19
13
17
Mean grade
1.3
1.6
1.5
1.6
1.9
1.8
1.5
1.6
Values are numbers of rats/groups of 25 males, 24 females affected at each grade (1 control male kidney
autolysed).
Inhalation of SHELLSOL A/SOLVESSO 100 after 12 mo (1,3,5-TMB, 1,2,4-TMB, 1,2,3-TMB) (mg/m3)
Incidence of neoplasia
Observation
Male
Female
0
450
900
1,800
0
450
900
1,800
Pituitary
2
0
0
0
7
7
4
3
Spleen
0
0
0
1
0
0
0
0
Uterus
-
-
-
-
0
0
0
1
Brain
0
1
0
0
0
0
0
0
Values are numbers of rats/group of 25 males, 24 females with a tumor.
NOAEL
LOAEL
LOAEL Effects
0 mg/m3
450 mg/m3
Male osmotic fragility, liver and
kidney lesions
This document is a draft for review purposes only and does not constitute Agency policy.
C-84 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-21. Characteristics and quantitative results for Douglas et al. (1993)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure
duration
Sprague-Dawley
rats
Male
20 rats/dose group
Inhalation
0, 100, 500, or
1,500 ppm High-
Flash Aromatic
Naphtha (HFAN)
6 hrs/d, 5 d/wk,
for 90 d
Additional study details:
• Rats were exposed to a mixture of 0,100, 500 or 1,500 ppm HFAN (1,3,5- TMB, 1,2,4-TMB, and
1,2,3-TMB) for 6 hrs/d, 5 d/wk for 90 d in 16 m3 glass and stainless steel chambers.
• Rats were randomly divided into four equal weight groups of 20 animals.
• Animals were sacrificed and tissues removed for histopathological examination after 13 wks.
• Exposure level measurements were taken on an hourly basis and accuracy confirmed by vapor
standards.
• Increases in motor activity in the 100 and 1,500 ppm group appear to be aberrant and are not
considered to have biological significance.
• Compared to the control group, the 1,500 ppm dose group gained 12% less weight.
• No signs of neurotoxicity were seen in any evaluation.
• The NOAEL was 100 ppm.
Composition of HFAN
Compound
Weight percent
o-Xylene
3.20
Cumene
2.74
n-Propylbenzene
3.97
4-Ethyltoluene
7.05
3-Ethyltoluene
15.1
2-Ethyltoluene
5.44
1,3,5-TMB
8.37
1,2,4-TMB
40.5
1,2,3-TMB
6.18
>C10s
6.19
Total
98.74
Mean chamber concentrations (ppm)
Target concentrations
Nominal concentrations mean (SD)
Actual concentrations mean (SD)
0
-
-
100
94 (1.0)
101 (2.5)
500
481 (5.1)
432 (2.8)
1,500
1,334 (17)
1,320 (13)
This document is a draft for review purposes only and does not constitute Agency policy.
C-85 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Average (SD) body weights (g) of male ratsa
Study wk
HFAN exposure level (ppm)
0
100
500
1,500
0
280 (15)
283 (13)
280 (13)
281 (13)
1
316 (18)
322 (16)
313 (15)
301 (17)*
2
346 (23)
352 (21)
338 (18)
314 (21)**
3
373 (27)
281 (23)
356 (19)
331(22)**
4
401 (32)
406 (30)
374 (20)*
347 (26)**
5
414 (33)
424 (34)
392 (24)
361(25)**
6
424 (34)
441 (33)
413 (25)
367 (32)**
7
436 (39)
455 (42)
426 (26)
383 (29)**
8
448 (38)
469 (39)
437 (28)
390 (30)**
9
459 (37)
484 (41)
449 (40)
401(32)**
10
462 (38)
484 (46)
455 (35)
410 (30)**
11
467 (39)
491 (54)
469 (32)
412 (32)**
12
476 (41)
504 (55)
481 (36)
418 (32)**
13
483 (42)
508 (56)
491 (37)
425 (34)**
a20 animals per group.
^Significantly different from control; p < 0.05.
**Significantly different from control; p < 0.01.
Average motor activity counts (SD) of male ratsa
Study wk
Time interval
(min)
HFAN
concentration
(ppm)
Horizontal activity (H)
Vertical activity (V)
Total activity (H + V)
5
0-10
0
1,548
(1,163)
269
(243)
1,818
(1,391)
100
1,511
(856)
287
(279)
1,298
(1,106)
500
1,701
(1,143)
229
(156)
1,930
(1,287)
1,500
1,395
(699)
219
(157)
1,614
(819)
10-20
0
882
(800)
124
(144)
1,006
(931)
100
1,142
(569)
204
(148)*
1,346
(689)
500
1,202
(772)
178
(156)
1,381
(915)
1,500
862
(546)
130
(102)
992
(640)
20-30
0
732
(664)
116
(113)
848
(766)
100
690
(497)
138
(117)
829
(579)
500
772
(485)
100
(98)
872
(575)
1,500
555
(357)
72
(57)
626
(407)
9
0-10
0
1,327
(1,018)
227
(197)
1,554
(1,192)
100
996
(811)
133
(125)
1,129
(917)
500
1,454
(1,051)
235
(236)
1,689
(1,274)
1,500
1,624
(1,027)
249
(195)
1,872
(1,205)
This document is a draft for review purposes only and does not constitute Agency policy.
C-86 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
10-20
0
589
(614)
105
(152)
694
(754)
100
758
(653)
115
(154)
873
(783)
500
647
(735)
104
(158)
752
(887)
1,500
1,138
(746)*
165
(153)
1,303
(887)*
20-30
0
458
(487)
85
(113)
543
(593)
100
517
(584)
83
(140)
600
(719)
500
463
(516)
79
(116)
542
(627)
1,500
556
(455)
91
(108)
646
(547)
13
0-10
0
1,618
(1,053)
270
(217)
1,889
(1,252)
100
1,356
(1,071)
260
(277)
1,616
(1,320)
500
1,579
(950)
317
(271)
1,895
(1,193)
1,500
1,882
(773)
288
(188)
2,170
(925)
10-20
0
814
(807)
140
(173)
955
(961)
100
634
(637)
165
(202)
808
(832)
500
887
(798)
198
(198)
1,085
(966)
1,500
945
(678)
188
(175)
1,133
(836)
20-30
0
518
(500)
85
(96)
603
(586)
100
552
(654)
116
(170)
667
(787)
500
593
(429)
110
(109)
703
(496)
1,500
511
(314)
77
(62)
588
(366)
aAnimal group size was between 18 and 20.
^Significantly different from control; p < 0.05.
Average total motor activity counts
SD) of male rats
Study wk
Time interval
(min)
HFAN
concentration
(ppm)
Horizontal activity (H)
Vertical activity (V)
Total activity (H + V)
5
0-30
0
3,162
(2,332)
509
(457)
3,671
(2,759)
100
3,343
(1,533)
629
(462)
3,972
(1,923)
500
3,675
(1,849)
507
(329)
4,182
(2,152)
1,500
2,812
(1,269)
421
(254)
3,233
(1,478)
9
0
2,467
(1,960)
437
(436)
2,903
(2,362)
100
2,271
(1,843)
331
(374)
2,602
(2,191)
500
2,646
(2,078)
433
(465)
3,079
(2,524)
1,500
3,364
(1,663)
515
(376)
3,879
(2,004)
13
0
2,950
(1,813)
496
(363)
3,446
(2,142)
100
2,605
(2,173)
519
(606)
3,152
(2,729)
500
3,136
(1,859)
641
(509)
3,777
(2,295)
1,500
3,338
(1,315)
553
(346)
3,891
(1,619)
aAnimal group size between 18 and 20.
This document is a draft for review purposes only and does not constitute Agency policy.
C-87 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Average (SD) grip strength (g) of male rats
Exposure
period
Limb
tested
HFAN exposure level
(ppm) 0
HFAN exposure
level (ppm) 100
HFAN exposure
level (ppm) 500
HFAN exposure level
(ppm) 1,500
0
Forelimb
558
(118)
538
(151)
586
(130)
592
(161)
5
Forelimb
580
(117)
622
(176)
578
(167)
590
(157)
9
Forelimb
385
(117)
433
(140)
492
(173)
448
(124)
13
Forelimb
440
136)
458
(166)
498
(148)
457
(148)
0
Forelimb
399
(63)
421
(82)
394
(80)
424
(90)
5
Forelimb
255
(63)
269
(55)
250
(44)
248
(55)
9
Forelimb
404
(89)
471
(120)
393
(107)
401
(116)
13
Forelimb
423
(85)
455
(143)
415
(70)
429
(114)
a20 animals per group.
Average (SD) auditory startle response of male rats
Exposure
period (wks)
Parameter
measured
(msec or kg)
HFAN exposure level
(ppm) 0
HFAN exposure
level (ppm) 100
HFAN exposure
level (ppm) 500
HFAN exposure
level (ppm) 1,500
0
Latency
27
(4.9)
28
(6.2)
28
(6.2)
26
(6.3)
5
Latency
23
(5.9)
24
(6.1)
26
(6.1)
25
(3.3)
9
Latency
23
(6.9)
23
(5.1)
26
(5.1)
25
(4.9)
13
Latency
23
(4.1)
24
(4.6)
25
(4.6)
23
(3.6)
0
Amplitude
0.17
(0.1)
0.16
(0.1)
0.17
(0.1)
0.17
(0.1)
5
Amplitude
0.42
(0.3)
0.35
(0.2)
0.28
(0.2)
0.38
(0.3)
9
Amplitude
0.52
(0.3)
0.35
(0.2)*
0.27
(0.2)*
0.37
(0.3)
13
Amplitude
0.47
(0.3)
0.36
(0.3)
0.32
(0.3)
0.44
(0.2)
a20 animals per group.
^Significantly different from control; p < 0.01.
Average (SD) thermal response (sec) of male rats
Exposure
period (wks)
HFAN exposure level (ppm)
0
HFAN expo
(ppm
sure level
100
HFAN exposure level
(ppm) 500
HFAN expc
(ppm)
jsure level
1,500
0
8.0
(2.7)
12.2
(4.6)*
10.7
(3.4)*
9.5
(4.0)
5
12.2
(4.8)
16.0
(7.7)
11.6
(4.6)
17.9
(12.2)
9
10.2
(3.8)
10.2
(3.0)
9.8
(3.9)
11.1
(2.9)
13
10.9
(4.2)
11.3
(3.9)
10.8
(13.0)
12.8
(4.9)
a20 animals per group.
^Significantly different from control; p < 0.01.
Average (SD
hindfoot splay distance (mm) of male rats
Exposure
period (wks)
HFAN exposure level (ppm)
0
HFAN expo
(ppm
sure level
100
HFAN expc
(ppm
>sure level
500
HFAN exposure level
(ppm) 1,500
0
109
(16)
107
(16)
114
(10)
108
(14)
5
128
(20)
125
(22)
126
(15)
113
(17)
9
131
(19)
122
(14)
124
(19)
126
(14)
13
120
(23)
121
(19)
127
(18)
124
(17)
a20 animals per group.
This document is a draft for review purposes only and does not constitute Agency policy.
C-88 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
NOAEL
LOAEL
LOAEL effects
100 ppm
500 ppm
Decreased body weight
Source: Douglas et al. (1993).
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-89 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-22. Characteristics and quantitative results for Gralewicz et al.
f!997bl
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
15 rats/
dose
Inhalation (6 hrs/d,
5 d/wk)
0, 25, 100, or 250 ppm (0,
123, 492, or 1,230 mg/m3)
1,2,4-TMB
4 wks
Additional study details
• Animals were exposed to 1,2,4-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/d,
5 d/wk for 4 wks. Food and water were 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 d 14-54 following exposure.
• Rats displayed decreased performance on several tests at the 100 and 250 ppm (492 and
1,230 mg/m3) exposure levels.
• CNS disturbances were observed up to 2 mo after termination of exposure, indicating the persistence
of effects after the metabolic clearance of 1,2,4-TMB from the test animals.
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.
ISO
- lOO
JE
m SO
£
e
5
E
©O
40
20
O
20
1 8
1 6
1 4
1 2
10
8
6
4
2
0
7
** 6
| S
I *
« 3
I.
3
2S£ 1
o
TMBG TMB25 TMB1QO TMB2SO
TMBO TMB2S TMBIOO TMB25D
TMBQ TMB25 TMBtOO TMB2SO
The test was performed 25 d after a 4-wk exposure to TMB. The bars represent group means and standard error
(SE) (N = 15 for each group). *p < 0.05 compared with TMBO group (0 ppm control group).
This document is a draft for review purposes only and does not constitute Agency policy.
C-90 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
8
120
100
3 80
c
*
o
T 60
a.
40
20
Diagrams illustrating the effect of a 4-wk exposure to 1,2,4-TMB on the step-down passive avoidance learning in
rats.
1 B0
140
¦ trial 1
rnn - trial 2
• trial 3 («hook)
- trio I 4
- trial S
¦ trial 6
7MB0
TMB2S
TMB100 TM8250
The test was performed on d 35-45 after exposure. Trials 1, 2, and 3 were performed at 24-hr intervals. The
step-down response was punished by a 10-sec footshock only in trial 3. Trials 4, 5, and 6 were performed 24 hrs,
3 d, and 7 d after trial 3, respectively. The maximum step-down latency was 180 sec. The bars represent group
means and SE (N = 15 for each group).
***p < 0.001 compared with respective data from group TMB0 (0 ppm control group).
This document is a draft for review purposes only and does not constitute Agency policy.
C-91 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Hot plate behavior tested in rats on d 50 (trials 1 and 2) and d 51 (trial 3) after 4-wk exposure to 1,2,4-TMB.
60
o
ffl
50
«
v«<»
>»
40
o
c
«
o
30
u
20
1
*
o
10
a.
0
o
o
K>
©
3
O
>
©
>
©
L.
•O
- L1
300
250
200
150
100
50
0
~ **
X
***
x
_L
***
§
***
§
T
TMBO TMB25 TMB100 TMB250
X
J **
TMBO TMB25 TMB100 TMB250
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 d 50. LI:
paw-lick latency in trial 1 performed before a 2 min intermittent footshock. L2: paw-lick latency in trial 2
performed several sec after the footshock.
***p < 0.001 compared with LI in the same group.
Lower diagram: A comparison of the change in the paw-lick latency noted 24 hrs after footshock (trial 3).
***p < 0.001, **p < 0.01 when compared to TMBO (0 ppm control group).
This document is a draft for review purposes only and does not constitute Agency policy.
C-92 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
A comparison of the active avoidance performance increment during a single 30-trial training session in
consecutive groups of rats.
80
70
60
« 50
M
C
o
CL
m
£ 40
j,.
o
5
20
10
block
block 8
TMBO
TMBZ5
TMB100 TMB250
The testing was performed on d 54 after 4-wk 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; therefore, blocks 1 and 2 were omitted in the analysis.
Health effect at LOAEL
NOAEL
LOAEL
Open field grooming
significantly increased, lower
than expected step down
latency
25 ppm (123 mg/m3
100 ppm (492 mg/m3
Comments: CNS disturbances were observed up to 2 mo 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 was only 4 wks.
Generally, short-term exposure studies have limited utility in quantitation of human health reference values.
This document is a draft for review purposes only and does not constitute Agency policy.
C-93 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-23. Characteristics and quantitative results for Gralewicz et al.
f!997al
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
9 rats/
dose
Inhalation (6 hrs/d,
5 d/wk)
0, 25, 100, or 250 ppm (0,
123, 492, or 1,230 mg/m3)
1,2,4-TMB
4 wks
Additional study details
• Animals were exposed to 1,2,4-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/d,
5 d/wk for 4 wks. Food and water were 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 (SWDs).
• 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.
Diagrams showing the effect of a 4-wk 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
electroencephalogram (EEG) during successive 1-hr recording periods.
before exposure
K«vsS ZA h after e*p.
KSSi 30 days after «xp.
on 120 days after axp.
TMBO TMB25 TMBIOO TMB260
*> 25
TMBO TMB25 TMBIOO TMB250
TMBO TMB23 TMBIOO TMB2S0
The bars represent group means and SE.
This document is a draft for review purposes only and does not constitute Agency policy.
C-94 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Diagram showing the effect of a 4-wk inhalation exposure to 1,2,4-TMB on the SWD burst occurrence (upper
diagram) and on the percent contribution of SWD activity within TRANS state (lower diagram) during successive
1-hr recording periods.
w,
3
O
JC
•X
0
«
3
o
Q
w
B
ja
E
3
c
140
120
100
80
60
40
S23before exposure
S3 24 h after «xp.
§£§ 30 days after exp.
OH 120 days after exp.
• - p<0.6 compared
to the preexposure
~alma
TMB0 TMB25 TMB1O0 TMB250
K
to
z:
c
o
3
JO
c
o
o
S
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-24. 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/
dose
Inhalation (6 hrs/d,
5 d/wk)
0 or 100 ppm (0 or
492 mg/m3) 1,2,3-, 1,2,4-,
or 1,3,5-TMB
4 wks
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/d, 5 d/wk for 4 wks. Food and water were 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 wks 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.
• CNS disturbances were observed up to 2 mo after termination of exposure, indicating the persistence
of effects after the metabolic clearance of 1,2,4-TMB from the test animals.
Radial maze performance of rats exposed for 4 wks to m-xylene or a TMB isomer at a concentration of 100 ppm
(492 mg/m3).
I
- day 1
day 2'
day 3
day 4
day &¦
Control XYL PS ME
5-5
8 6'°
0 4,5
e 4,0
1 3-6
.1 3.0
I 2.8
"3 2.0
I 1-5
I 1.0
8 0,6
0.0
Control XYL PS MIS HM
The test (one trial a day) was performed on d 14-18 after exposure. The diagrams illustrate the number of
perseveration (upper diagram) and omission (lower diagram) errors in successive daily trials. Bars represent group
means and SE.
Control = sham-exposed group (N = 10); XYL = m-xylene-exposed group (N = 11); PS = 1,2,4-TMB exposed group
(N = 11); MES = 1,2,3-TMB exposed group (N = 11); HM = hemimellitene exposed group (N = 11).
This document is a draft for review purposes only and does not constitute Agency policy.
C-96 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
A comparison of open-field locomotor activity in sham-exposed and solvent-exposed rats.
M
B
m
m
o
u
o
u
m
XI
a
a
120
100
80
60
40
20
* - p<0.05 compared
to control
o - p<0.05 compared
to PS and MES
Control XYL PS
MES HM
The test was performed on d 25 after a 4-wk exposure to m-xylene or a TMB isomer at concentration of 100 ppm
(492 mg/m3). Bars represent group means and SE.
Diagram illustrating the effect of a 4-wk inhalation exposure to m-xylene or a TMB isomer at concentration of
100 ppm (492 mg/mB) on the step-down response latency in the passive avoidance test.
5 120
trial 1
trial 0
trial 3
trial 4
trial 5
trial 6
(shock)
- p<0.05 compared
to control
Control XYL
The test was performed on d 39-48 after exposure. Trials 1, 2, and 3 were performed at 24-hr intervals. The
step-down response was punished by a 10 sec footshock in trial 3 only. Trials 4, 5, and 6 were performed 24 hrs,
3 d, and 7 d after trial 3, respectively. The maximum time of staying on the platform was 180 sec. Bars represent
means and SE.
This document is a draft for review purposes only and does not constitute Agency policy.
C-97 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 sec after (L2), and 24 hrs after a 2-min intermittent footshock.
60
50
8
I 30
S
3
S
i
g
CL
20
10
rfrn- 12
EB~
p<.0.06 corop*r«d
to control
Control XYL
The test was performed on d 50 and 51 after a 4-wk inhalation exposure to m-xylene or a TMB isomer at a
concentration of 100 ppm (492 mg/m3). Bars represent group means and SE.
Active avoidance learning in rats after a 4-wk inhalation exposure to m-xylene or a TMB isomer at a
concentration of 100 ppm (492 mg/m3).
pcO.OS compared
to control
Control XYL PS
MIS MM
In one massed-trial session (inter-trial interval 20-40 sec; maximum number of trials 60) the rats learned to
shuttle between two neighboring compartments in order to avoid a footshock. The test was performed on
d 54-60 after exposure. Bars represent group means and SE of the number of trials.
Health effect at LOAEL
NOAEL
LOAEL
Deleterious effects on
locomotor activity, passive
avoidance learning, and paw-
lick latencies
N/A
100 ppm (492 mg/m3) 1,2,3-TMB,
1,2,4-TMB, or 1,3,5-TMB
Comments: CNS disturbances were observed up to 2 mo 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 was only 4 wks.
Generally, short-term exposure studies have limited utility in quantitation of human health reference values.
This document is a draft for review purposes only and does not constitute Agency policy.
C-98 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-25. Characteristics and quantitative results for lanik-Spiechowicz et
al. f!998)
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
0,1,470, 2,160, and
2,940 mg/kg body weight
Single exposure, or two i.p.
injections spaced out over
24 hrs
Additional study details
• Animals were given one or two i.p. injections of 1,2,3-TMB.
• Animals were randomized and assigned to the experimental groups.
• Most deaths occurred within the first 2 d 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.
Dose-related increase in the number of His+ revertants for 1,2,3-TMB in Salmonella typhimurium strains.
TA102
TA100
TA98
TA97a
1200 1000 800 600 400 200 0 200 400
-59 Revertants I plate +S9
* - mutagenic ellecr ,a 2-loid or greater increase in the number of revertants
per plate, at- compared with the solvent control number)
Spontaneous revpiunlx; TA37a (-39) ; 141±17 {+S9) ;
TA98 (-S9) ; 351-6 C+S3J,-
TA100 x26x' {-S3); 113iB (+S9);
TAX02 2B2.13J (-S9) ; 32b±32 {+SS|
dose [ jul/plate]
-S9 +S9
20
Solvent
control
This document is a draft for review purposes only and does not constitute Agency policy.
C-99 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Observation
Exposure to 1,2,4-TMB (|ig or |iL)
0
100
(solvent
control)
1
5
10
20
30
TA97a (-S9)
212 ±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
TOXa
TA98 (+S9)
31 ±3
31 ±5
35 ±4
28 ± 1
29 ±4
30 ±3
29 ±6
TA100 (-S9)
123 ± 71
125 ± 41
138 ± 15
148 ± 18
143 ±9
124 ±7
118 ±4
TA100 (+S9)
25 ±4
21 ± 10
126 ± 62
125 ±5
112 ±4
108 ±3
110 ±4
TA102 (-S9)
258 ±6
280 ± 12
290 ± 33
262 ± 16
273 ± 20
214 ±8
TOX
TA102 (+S9)
294 ± 11
315 ± 14
279 ± 24
276 ± 11
276 ± 11
236 ± 32
TOX
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
1,470
2,160
2,940
Percentage 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
-
0.21 ±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
-
This document is a draft for review purposes only and does not constitute Agency policy.
C-100 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Observation
Exposure to 1,2,4-TMB (mg/kg body weight
0
2,000
3,280
4,000
Percentage 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
-
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
1,800
2,960
3,600
Percentage 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
-
^Significant difference versus control at p < 0.05.
aTOX = toxic effects (background growth reduced).
bNT = not tested.
This document is a draft for review purposes only and does not constitute Agency policy.
C-101 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Sister chromatid exchanges (SCEs) induced in bone marrow cells of Imp:Balb/c mice.
8
uj
o
03
Jt I}
t
solvent.
control
C> hemimellitene 730
O pseudocumene 900
V mesitylene 900 -
1470-
t
I
T
T
o
o
X
X
i
1
c
c
i
i
t
t
y
y
2200
2940
1800-
-1800-
2700
- 2700 -
3600 dose
3600 [mg/kg b.W.]
~ - significant difference vs. control at p<0 .05
as positive control in dose of 2,0 mg/kg b.w.
Mitomycin C administered
gave 21,34 ± 1.36 SCE/cell
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 hrs) and therefore less germane to the 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-102 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-26. Characteristics and quantitative results for iuran etal. (2014)
Species
Humans
Study design
Sex
N
Exposure route
Dose range
Exposure duration
M & F
6 rats/sex/dose
group
Inhalation
100 or
300 mg/m3
aromatic or
dearomatized
white spirit
4 hrs for 5 consecutive d
Additional study details:
• Males and females were exposed to 100 or 300 mg/m3 of aromatic or dearomatized white spirit in an
inhalation chamber for 4 hrs/d for 5 d.
• White spirit concentrations were determined via gas chromatography at 5-min intervals.
• Humans exhibited weak and inconsistent neurobehavioral impairment after 4-hr exposures.
• No significant effect on response inhibition was observed.
• Exposure concentrations correspond to recommendations from the Scientific Committee on
Occupational Exposure Limits of the European Commission.
• The NOAEL was 100 mg/m3.
Performance in vigilance task is given as (a) absolute omission errors and (b) reaction times.
34 x,# jr
#' •# ^
* if ^
~ TOTS min
1Z3 TOT 10 rwn
(55 TOT 15 mil*
TO lOT2®r»n
U'."l TOT 25 min
Source: Juran et al. (2014).
Performance in the divided attention task
Time
Reaction time (ms)
False
alarm (%)
Misses (%)
Visual
Auditory
Mean
SE
Mean
SE
Mean
SE
Mean
SE
53 min
725
18.7
517
22.2
2.1
0.4
4.2
1.4
3 hrs 48 min
741
22.2
506
23
2.2
0.4
3.9
1.4
Comments: Mean and SEs given for 12 participants in two modalities and two task performances. Visual reaction
times significantly slower than auditory ones (p < 0.001), and only visual reaction times showed a significant
increase over time. Accuracy in the form of false alarms and misses did not vary over time or between visual and
auditory trials.
This document is a draft for review purposes only and does not constitute Agency policy.
C-103 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Speed (reaction times; upper panel) accuracy (hits; middle panel) and false alarms (lower panel) of the 2- and
3-back condition of the working memory task.
a 700
* Back
b?oo
3 Back
19.it
3h3
.1.
iiii
^ y *
&
&
WA
© 15
a***
Source: Juran et al. (2014).
Performance in the response shifting task is given as (a) correct responses and (b) reaction times.
too
9a
\ih
CHJ Rfrpetitsx*
wm srnmm
N « -J/
...
$ 15
550
¦v n
^ ^ ^ ^ ^
I 1 I M
I
II
in :i; ii i i .iI
_j Rep
-------
Supplem en tal Information—Trim ethylbenzenes
Performance in the response inhibition task is given for (a) amount of errors and (b) reaction times.
a »
— 40
\2
30
20
10
Compslijle
Srnn
// 3> h «?!> mn
AA i*iA
Nogo
lncorrv«
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-27. Characteristics and quantitative results for Koch Industries
2 f1995b)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley CD
M/F
20 rats/dose
Gavage
0, 50, 200, and
600 mg/kg-d 1,3,5-TMB
90 d
Additional study details
• Rats were treated with 0, 50, 200, or 600 mg/kg-d of 1,3,5-TMB (5 d/wk) and were observed daily for
adverse clinical signs.
• Hematology and serum chemistry was analyzed after 30 d, at the end of the exposure period, and
after a 28-d recovery period (in an additional 600 mg/kg-d "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 d 1,3,5-TMB dosing period
Males
Dose (mg/kg-d)
0
50
200
600
Mean
624
607
602
585
SD
48.2
62.0
40.8
66.4
Number of rats
10
10
9
20
Females
Mean
327
335
334
330
SD
24.8
37.6
21.2
29.3
Number of rats
10
10
10
20
Mean clinical chemistry parameters, terminal and recovery in males
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Sodium, mean
142.4
142.7
143.0
142.4
141.6
Sodium, SD
1.49
0.65
1.40
1.32
1.30
Sodium, number of rats
10
10
9
10
10
Potassium, mean
4.32
4.51
4.37
4.54
4.33
Potassium, SD
0.397
0.339
0.328
0.270
0.240
Potassium, number of rats
10
10
9
10
10
Chloride, mean
105.3
105.3
106.0
106.2
104.7
Chloride, SD
2.59
2.33
1.72
2.18
0.88
Chloride, number of rats
10
10
9
10
10
Creatine kinase, mean
594
962
934
595
884
Creatine kinase, SD
340.4
929.8
799.2
389.1
353.4
Creatine kinase, number of rats
10
10
9
10
10
AP, mean
107
112
121
156*
77
AP, SD
28.1
26.5
33.7
56.2
20.5
This document is a draft for review purposes only and does not constitute Agency policy.
C-106 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
AP, number of rats
10
10
9
10
10
ALT, mean
29
30
25
33
25
ALT, SD
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, SD
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, SD
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, SD
1.45
1.87
1.22
1.30
1.53
BUN, number of rats
10
10
9
10
10
Creatinine, mean
0.42
0.43
0.42
0.47
0.48
Creatinine, SD
0.092
0.079
0.110
0.065
0.067
Creatinine, number of rats
10
10
9
10
10
Total protein, mean
6.0
5.9
6.0
6.1
6.0
Total protein, SD
0.38
0.24
0.31
0.42
0.25
Total protein, number of rats
10
10
9
10
10
Albumin, mean
3.6
3.6
3.7
3.8
3.7
Albumin, SD
0.23
0.19
0.19
0.22
0.09
Albumin, number of rats
10
10
9
10
10
Globulin, mean
2.4
2.3
2.3
2.3
2.3
Globulin, SD
0.27
0.18
0.16
0.24
0.24
Globulin, number of rats
10
10
9
10
10
Albumin/globulin ratio, mean
1.6
1.6
1.6
1.7
1.7
Albumin/globulin ratio, SD
0.19
0.17
0.11
0.15
0.17
Albumin/globulin ratio, number of rats
10
10
9
10
10
Glucose, mean
150.2
134.6
136.9
121.1*
168.4
Glucose, SD
22.80
15.11
15.76
13.14
26.39
Glucose, number of rats
10
10
9
10
10
Cholesterol, mean
38.2
33.1
31.6
45.3
35.3
Cholesterol, SD
6.83
9.13
9.93
15.99
10.10
Cholesterol, number of rats
10
10
9
10
10
Calcium, mean
10.2
10.2
10.2
10.2
9.9
Calcium, SD
0.22
0.29
0.37
0.23
0.24
Calcium, number of rats
10
10
9
10
10
Phosphorus, mean
6.5
6.7
7.0
7.6*
5.8
Phosphorus, SD
0.64
0.80
0.68
0.58
0.59
Phosphorus, number of rats
10
10
9
10
10
Total bilirubin, mean
0.4
0.4
0.5
0.5
0.5
This document is a draft for review purposes only and does not constitute Agency policy.
C-107 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Total bilirubin, SD
0.12
0.10
0.09
0.14
0.09
Total bilirubin, number of rats
10
10
9
10
10
Mean clinical chemistry parameters, terminal and recovery in females
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Sodium, mean
142.1
141.6
141.7
138.9*
140.9
Sodium, SD
1.10
0.96
2.07
2.83
1.47
Sodium, number of rats
10
10
10
10
10
Potassium, mean
3.94
4.13
4.01
3.86
4.06
Potassium, SD
0.195
0.200
0.119
0.292
0.259
Potassium, number of rats
10
10
10
10
10
Chloride, mean
105.9
106.2
106.1
103.0*
107.0
Chloride, SD
2.32
1.63
1.05
3.81
1.68
Chloride, number of rats
10
10
10
10
10
Creatine kinase, mean
404
574
381
362
532
Creatine kinase, SD
172.6
346.4
228.3
242.5
369.7
Creatine kinase, number of rats
10
10
10
10
10
AP, mean
59
57
55
78
38
AP, SD
14.8
10.3
14.9
24.5
10.1
AP, number of rats
10
10
10
10
10
ALT, mean
21
22
23
24
27
ALT, SD
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, SD
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, SD
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, SD
1.34
2.57
1.49
4.61
2.31
BUN, number of rats
10
10
10
10
10
Creatinine, mean
0.53
0.51
0.53
0.56
0.55
Creatinine, SD
0.106
0.085
0.099
0.110
0.099
Creatinine, number of rats
10
10
10
10
10
Total protein, mean
6.2
6.3
6.6
6.5
6.3
Total protein, SD
0.44
0.41
0.69
0.68
0.66
Total protein, number of rats
10
10
10
10
10
Albumin, mean
4.1
4.3
4.5
4.5
4.3
Albumin, SD
0.29
0.36
0.58
0.56
0.51
Albumin, number of rats
10
10
10
10
10
This document is a draft for review purposes only and does not constitute Agency policy.
C-108 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Globulin, mean
2.1
2.0
2.1
2.1
2.0
Globulin, SD
0.21
0.17
0.19
0.20
0.18
Globulin, number of rats
10
10
10
10
10
Albumin/globulin ratio, mean
2.0
2.1
2.1
2.1
2.1
Albumin/globulin ratio, SD
0.16
0.22
0.26
0.23
0.18
Albumin/globulin ratio, number of rats
10
10
10
10
10
Glucose, mean
131.8
136.4
140.1
132.8
150.7
Glucose, SD
7.65
11.72
14.48
15.91
19.18
Glucose, number of rats
10
10
10
10
10
Cholesterol, mean
36.2
35.2
38.8
51.2*
28.7
Cholesterol, SD
8.83
6.64
6.24
17.84
12.93
Cholesterol, number of rats
10
10
10
10
10
Calcium, mean
10.1
10.2
10.4
10.5
10.0
Calcium, SD
0.35
0.24
0.42
0.63
0.36
Calcium, number of rats
10
10
10
10
10
Phosphorus, mean
6.1
6.1
6.4
7.5
5.3
Phosphorus, SD
1.08
1.27
1.18
1.24
0.80
Phosphorus, number of rats
10
10
10
10
10
Total bilirubin, mean
0.5
0.5
0.4
0.5
0.5
Total bilirubin, SD
0.08
0.10
0.08
0.07
0.07
Total bilirubin, number of rats
10
10
10
10
10
Mean male hematology parameters terminal and recovery
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
WBCs, mean
9.1
8.1
8.1
7.7
7.8
WBCs, SD
2.70
2.50
1.74
1.76
1.24
WBCs, number of rats
10
10
9
10
10
RBCs, mean
8.94
8.50
8.98
8.72
8.51
RBCs, SD
0.375
0.483
0.565
0.275
0.423
RBCs, number of rats
10
10
9
10
10
Hemoglobin, mean
15.6
15.3
15.8
15.4
15.4
Hemoglobin, SD
0.52
0.76
0.77
0.53
0.58
Hemoglobin, number of rats
10
10
9
10
10
Hematocrit, mean
43.9
42.2
44.1
43.3
41.6
Hematocrit, SD
1.65
2.72
2.12
1.60
1.99
Hematocrit, number of rats
10
10
9
10
10
MCV, mean
49.1
49.7
49.2
49.6
49.0
MCV, SD
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, SD
0.45
0.73
0.85
0.68
0.61
This document is a draft for review purposes only and does not constitute Agency policy.
C-109 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
MCH, number of rats
10
10
9
10
10
MCHC, mean
35.6
36.3
35.9
35.6
37.1
MCHC, SD
0.67
1.07
0.60
0.67
0.60
MCHC, number of rats
10
10
9
10
10
Platelet, mean
1,092
1,098
1,041
1,125
1,082
Platelet, SD
134.1
120.8
100.9
145.9
112.6
Platelet, number of rats
10
10
9
10
10
Mean female hematology parameters terminal and recovery
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
WBCs, mean
5.5
5.6
5.4
5.7
4.6
WBCs, SD
2.05
1.53
1.64
1.99
1.55
WBCs, number of rats
10
10
10
10
10
RBCs, mean
7.88
8.01
7.90
8.34
7.70
RBCs, SD
0.729
0.354
0.578
0.548
0.423
RBCs, number of rats
10
10
10
10
10
Hemoglobin, mean
14.8
15.0
15.2
15.3
15.1
Hemoglobin, SD
0.88
0.48
0.82
0.78
0.57
Hemoglobin, number of rats
10
10
10
10
10
Hematocrit, mean
41.0
41.4
41.9
43.3
39.9
Hematocrit, SD
3.15
1.91
2.93
2.33
1.67
Hematocrit, number of rats
10
10
10
10
10
MCV, mean
52.1
51.7
53.0
52.0
51.9
MCV, SD
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, SD
0.89
0.86
0.83
0.54
0.64
MCH, number of rats
10
10
10
10
10
MCHC, mean
36.2
36.2
36.3
35.4
37.7
MCHC, SD
0.79
0.86
0.83
0.54
0.64
MCHC, number of rats
10
10
10
10
10
Platelet, mean
1,094
1,089
1,011
1,053
1,008
Platelet, SD
153.3
132.0
97.2
125.7
105.7
Platelet, number of rats
10
10
10
10
10
Mean male absolute differential WBC counts (terminal and recovery)
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Nucleated RBCs, mean
0
0
0
0
0
Nucleated RBCs, SD
0
0
0.7
0
0
Nucleated RBCs, number of rats
10
10
9
10
10
Mature neutrophils, mean
1.8
1.7
1.4
1.5
1.0
This document is a draft for review purposes only and does not constitute Agency policy.
C-110 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mature neutrophils, SD
1.07
1.10
0.36
0.75
0.29
Mature neutrophils, number of rats
10
10
9
10
10
Lymphocytes, mean
7.1
6.2
6.4
6.0
6.6
Lymphocytes, SD
2.78
2.16
1.59
2.16
1.23
Lymphocytes, number of rats
10
10
9
10
10
Monocytes, mean
0.1
0.2
0.3*
0.2*
0.2
Monocytes, SD
0.09
0.09
0.17
0.18
0.10
Monocytes, number of rats
10
10
9
10
10
Eosinophils, mean
0.1
0.1
0.0
0.0
0.1
Eosinophils, SD
0.06
0.09
0.07
0.05
0.07
Eosinophils, number of rats
10
10
9
10
10
Basophils, mean
0
0
0
0
0
Basophils, SD
0
0
0
0
0
Basophils, number of rats
10
10
9
10
10
Immature neutrophils, mean
0
0
0
0
0
Immature neutrophils, SD
0
0
0
0
0
Immature neutrophils, number of rats
10
10
9
10
10
Mean female absolute differential WBC counts (terminal and recovery)
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Nucleated RBCs, mean
0
0
0
0
0
Nucleated RBCs, SD
0
0
0
0
0
Nucleated RBCs, number of rats
10
10
10
10
10
Mature neutrophils, mean
0.8
0.7
0.9
1.0
0.7
Mature neutrophils, SD
0.48
0.32
0.69
0.39
0.45
Mature neutrophils, number of rats
10
10
10
10
10
Lymphocytes, mean
4.6
4.7
4.2
4.4
3.7
Lymphocytes, SD
1.93
1.52
1.52
2.08
1.34
Lymphocytes, number of rats
10
10
10
10
10
Monocytes, mean
0.1
0.1
0.1
0.2
0.2
Monocytes, SD
0.14
0.10
0.08
0.17
0.11
Monocytes, number of rats
10
10
10
10
10
Eosinophils, mean
0.1
0.1
0.1
0.1
0
Eosinophils, SD
0.07
0.07
0.09
0.09
0.07
Eosinophils, number of rats
10
10
10
10
10
Basophils, mean
0
0
0
0
0
Basophils, SD
0
0
0.03
0
0
Basophils, number of rats
10
10
10
10
10
Immature neutrophils, mean
0
0
0
0
0
Immature neutrophils, SD
0
0
0
0
0
Immature neutrophils, number of rats
10
10
10
10
10
This document is a draft for review purposes only and does not constitute Agency policy.
C-lll DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mean male absolute organ weights (g)
Parameter
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Adrenal glands, mean
0.062
0.059
0.058
0.063
0.060
Adrenal glands, SD
0.010
0.015
0.011
0.010
0.008
Adrenal glands, number of rats
10
10
9
10
10
Brain, mean
2.25
2.28
2.23
2.19
2.24
Brain, SD
0.073
0.090
0.094
0.084
0.112
Brain, number of rats
10
10
9
10
10
Kidneys, mean
3.92
3.95
4.10
4.16
4.05
Kidneys, SD
0.326
0.262
0.610
0.464
0.491
Kidneys, number of rats
10
10
9
10
10
Liver, mean
19.28
18.91
18.38
20.90
17.38
Liver, SD
1.843
3.074
2.885
3.313
2.222
Liver, number of rats
10
10
9
10
10
Lung, mean
2.19
2.19
2.20
2.06
2.04
Lung, SD
0.299
0.292
0.134
0.158
0.229
Lung, number of rats
10
10
9
10
10
Testes, mean
4.15
3.78
4.04
4.00
3.91
Testes, SD
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)
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Adrenal glands, mean
0.075
0.078
0085
0.082
0.084
Adrenal glands, SD
0.007
0.012
0.013
0.015
0.015
Adrenal glands, number of rats
10
10
10
10
10
Brain, mean
2.06
2.06
2.11
2.06
2.11
Brain, SD
0.080
0.083
0.094
0.050
0.059
Brain, number of rats
10
10
10
10
10
Kidneys, mean
2.34
2.23
2.38
2.51
2.38
Kidneys, SD
0.314
0.228
0.116
0.264
0.248
Kidneys, number of rats
10
10
10
10
10
Liver, mean
9.44
9.13
10.05
11.78*
9.71
Liver, SD
1.601
0.774
0.967
1.444
1.411
Liver, number of rats
10
10
10
10
10
Lung, mean
1.63
1.73
1.66
1.60
1.63
Lung, SD
0.187
0.140
0.106
0.150
0.140
Lung, number of rats
10
10
10
10
10
This document is a draft for review purposes only and does not constitute Agency policy.
C-112 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Ovaries, mean
0.128
0.123
0.122
0.142
0.142
Ovaries, SD
0.023
0.039
0.042
0.058
0.036
Ovaries, number of rats
10 10 10
10
9
Mean male relative13 organ weights (g)
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Fasted body weight, mean
602
584
576
562
595
Fasted body weight, SD
46.4
60.4
40.1
52.2
81.8
Fasted body weight, number of rats
10
10
9
10
10
Adrenal glands, mean
0.011
0.010
0.010
0.011
0.010
Adrenal glands, SD
0.002
0.002
0.002
0.001
0.001
Adrenal glands, number of rats
10
10
9
10
10
Brain, mean
0.38
0.39
0.39
0.39
0.38
Brain, SD
0.033
0.032
0.035
0.035
0.044
Brain, number of rats
10
10
9
10
10
Kidneys, mean
0.65
0.68
0.71
0.74*
0.68
Kidneys, SD
0.052
0.052
0.082
0.045
0.039
Kidneys, number of rats
10
10
9
10
10
Liver, mean
3.20
3.23
3.19
3.71*
2.93
Liver, SD
0.158
0.336
0.402
0.288
0.274
Liver, number of rats
10
10
9
10
10
Lung, mean
0.37
0.38
0.38
0.37
0.34
Lung, SD
0.045
0.052
0.027
0.038
0.042
Lung, number of rats
10
10
9
10
10
Testes, mean
0.69
0.65
0.71
0.72
0.67
Testes, SD
0.060
0.101
0.092
0.089
0.136
Testes, number of rats
10
10
9
10
10
Mean female relative13 organ weights (g)
Parameter3
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Fasted body weight, mean
309
317
316
308
336
Fasted body weight, SD
23.4
34.8
20.0
28.2
33.9
Fasted body weight, number of rats
10
10
10
10
10
Adrenal glands, mean
0.025
0.025
0.027
0.027
0.025
Adrenal glands, SD
0.003
0.005
0.005
0.004
0.005
Adrenal glands, number of rats
10
10
10
10
10
Brain, mean
0.67
0.66
0.67
0.68
0.63
Brain, SD
0.067
0.075
0.047
0.065
0.059
Brain, number of rats
10
10
10
10
10
Kidneys, mean
0.76
0.71
0.76
0.82
0.71
Kidneys, SD
0.059
0.088
0.051
0.059
0.040
This document is a draft for review purposes only and does not constitute Agency policy.
C-113 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Kidneys, number of rats
10
10
10
10
10
Liver, mean
3.04
2.90
3.19
3.82*
2.88
Liver, SD
0.365
0.330
0.357
0.223
0.207
Liver, number of rats
10
10
10
10
10
Lung, mean
0.53
0.55
0.53
0.52
0.49
Lung, SD
0.071
0.059
0.052
0.047
0.079
Lung, number of rats
10
10
10
10
10
Ovaries, mean
0.041
0.040
0.039
0.046
0.043
Ovaries, SD
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)
Dose (mg/kg-d)
0
50
200
600
600 (recovery)
Tissue and observation
M
F
M
F
M
F
M
F
M
F
Number of 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
-
-
-
-
ld
-
-
-
-
-
Large intestine, cecum; focus, red
-
-
-
-
1
-
-
-
-
-
Liver; pale
-
-
-
-
-
-
1
-
-
-
^Significantly different from vehicle control, p < 0.05.
aSUnits of measure: sodium (mE/litter serum); potassium (mE/litter serum); chloride (mE/litter serum); creatine
kinase (Ill/liter serum); AP (Ill/liter serum); ALT (Ill/liter serum); AST (Ill/liter serum); GGT (Ill/liter serum);
BUN (mg N/dL serum); creatinine (mg/dL serum); total protein (g protein/dL serum); albumin (g/dL serum);
globulin (g/dL serum); albumin/globulin ratio; glucose (mg/dL serum); cholesterol (mg/dL serum); total
bilirubin (mg/dL serum); WBC (103/mm3); RBC (106/mm3); hemoglobin (g/dL blood); hematocrit (%); MCV
(femoliter); MCH (picogram); MCHC (%); Platelet (103/mm3); nucleated RBCs (number/100 WBCs); mature
neutrophils (103/mm3); lymphocytes (103/mm3); monocytes (103/mm3); eosinophils (103/mm3); basophils
(103/mm3); immature neutrophils (103/mm3)
bRelative organ weight = [absolute organ weight (g)/fasted body weight (g)] x 100.
cZero incidence.
dAnimal died due to gavage error (accidental death).
BUN = blood urea nitrogen; GGT = gamma-glutamyl transpeptidase (lU/liter serum).
Comments: 1,3,5-TMB was the only isomer tested in this study. Effects reported in study appeared reversible in
the recvery group, which was observed for 28 d following cessation of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
C-114 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-28. 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
250-2,000 ppm
(1,230-9,840 mg/m3)
1,2,4-TMB
4 hrs—neurotoxicity tests
6 min—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 min on 10 consecutive d were used.
• Rotarod, hot plate, and respiratory tests were conducted to measure effects on neuromuscular
activity, pain sensitivity, and respiratory rate respectively.
Rotarod performance of rats exposed to 1,2,4-TMB (i.e., pseudocumene).
1 8
| 7-
!
8.
I 3
*'*~
o y
EC^o = 4693 mg/m3
(954 ppm)
i it i i r
t—r
IMIWI
10000
Concentration of pseudocumene, mg/m3
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-115 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Hot plate behavior in rats exposed to 1,2,4-TMB (i.e., pseudocumene).
70
65
* 60-
I 55
8 50
6 45
g 40
35
30
25
EQo = 5682 mg/m3
(1155 ppm)
~
a
a
1000
"t*——i 1 1 r
Concentration of pseudocumene, mg/m3
10000
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.
Time-response relationship for the effect of 1,2,4-TMB (i.e., pseudocumene) on respiratory rate in mice.
o
u
•—*
o
*
100*
*4
t-
o
a
*4
ft
B3
m
pisadocuiniB*
— 1244 mfl/ms(253 ppm)
—t-' 2686 mg/ui3(S4a ppm)
518S m9/m3(1054 pprrO
-3- 6391 mg/m3(1299 ppm)
-x- 9485 mo/m3a»2e ppm)
12
Each point represents the mean value in 8-10 mice. After termination of 6 min exposure, recovery of respiratory
rate was observed.
This document is a draft for review purposes only and does not constitute Agency policy.
C-116 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Respiratory rate of mice exposed to 1,2,4-TMB (i.e., pseudocumene) in 8-10 mice.
100
a,
<33
XS
IS
&
1
n
%
DC
10;
n
[ i
_#*
m¥w
„#*
ROje = 2844 mg/m3
(578 ppm)
1000
10000
Concentration of pseudocumene, mg/m3
The decrease of respiratory rate observed in the 1st min of exposure was taken for consideration. The regression
line was determined by the least squares procedure.
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 values cannot be
determined. Exposures were of an acute duration, and were 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-117 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-29. Characteristics and quantitative results for Korsak and Rvdzvnski
f!996)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP: Wistar
rats
M
9-10/dose
(1,2,4-TMB)
10-30/dose
(1,2,3-TMB)
Inhalation (4 hrs or
6 hrs/d, 5 d/wk, for
3 mo)
Acute exposure:
250-2,000 ppm
1,230-9,840 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 mo
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 16 air changes/hr.
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 immediately after termination of exposure.
Normal neuromuscular function was indicated by the rats' ability to remain on a rod rotating at
12 rotations/min for 2 min.
Hot plate behavior was tested immediately after termination of exposure.
Latency of 60 sec 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 2 wks after the termination of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
C-118 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Rotarod performance of rats exposed to 1,2,3-TMB (hemimellitene), 1,2,4-TMB (pseudocumene), or 1,3,5-TMB
(mesitylene).
QL
J?
"5
I
s
3
SS
m
%
I
o
OL
!
PSEUDOCUMENE 0
y''
0/
y
/•'b
X
/ 0
o
ECa® 4S93 mg/m3
|954 ppm)
1000
10000
1 1 1—I ITT
100000
MESITYLENE
EC,,-- 4738 mg/m®
(963 ppm)
1000
10000
100000
I
3
I
"5
4™
3-
HEMIMELLITENE 0
O
y'6
a 377S mg/m3
{768 ppm)
1000
10000
Concentration, mg/m3
100000
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 rotations/min for 2 min. The rotating
rod was suspended 20 cm above metal bars connected to a 80 V/2 mA power source.
Source: Reproduced from Korsak and Rvd/.vii.ski (19961.
This document is a draft for review purposes only and does not constitute Agency policy.
C-119 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Hot plate behaviors in rats exposed to 1,2,3-TMB (hemimellitene), 1,2,4-TMB (pseudocumene), or 1,3,5-TMB
(mesitylene).
Effects of exposure to trtmothythenzone isomers
PSEUDOCUMENE
EC.„ = 56ing'm '
{1155 ppnij
1000
100-
10000
MESITYLENE
ECW = 5683 mg/rn3
(1212 ppm)
/
30-
1000
100
10000
.9 90-
§ 80
70-
60
50-
.# 40-
1 3®.
1 20
C 10
-6
aS
-------
Supplem en tal Information—Trim ethylbenzenes
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/m3).
40
35
1
~ 20
a
o
Pseudocumene
¦X" control
4-25 ppm
*100 ppm
H • 2S0 ppm
- exposure -
13 18
weeks
60
8
3
5 40
o
45 20
Hemimellitene
*- control
-4-25 ppm
*100 ppm
-» 250 ppm
- exposure
Rats were exposed to vapors of solvents for 6 hrs/d, 5 d/wk, 3 mo. Statistical significance marked by asterisks,
p < 0.005.
Source: Reproduced from Korsakand Rvdzvnski (1996).
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 wks after termination of
exposure
17.3 ±3.9
11.0 ±2.4
^Statistically significant from controls at p < 0.05.
**Statistically significant from controls at p < 0.01.
***Level of significance not reported in Table 1 from Korsak and Rvdzvnski (1996); however, the results of an ad-
hoc t-test (performed by EPA) indicated significance at p < 0.01.
Health effect at LOAEL
NOAEL
LOAEL
Decreased pain sensitivity
N/Afor 1,2,3-TMB
25 ppm (123 mg/m3)
1,2,4-TMB
for
25 ppm (123 mg/m3
1.2.3-TMB
100 ppm (492 mg/m3
1.2.4-TMB
for
for
This document is a draft for review purposes only and does not constitute Agency policy.
C-121 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 paw-lick and rotarod tests were
performed sequentially in the same cohort of animals.
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-122 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-30. Characteristics and quantitative results for Korsak et al. (1997)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP:DAK
Wistar rats
and Balb/C
mice
M
Acute:
8/dose
Subchronic
6-7/dose
Acute: Inhalation
Subchronic:
Inhalation
Acute: 250-2,000 ppm
(1,230-9,840 mg/m3)
1.2.4-TMB, 1,2,3-TMB, or
1.3.5-TMB
Subchronic: 0,123, 492, or
1,230 mg/m31,2,4-TMB
Acute: 6 min
Subchronic: 6 hrs/d, 5 d/wk
for 90 d
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 lavage (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/m3)
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*
^Statistically significant from control at p < 0.05.
**Statistically significant from control at 0.01.
***Statistically significant from control at 0.001.
aJonckheere's test for trend: total protein, p = 0.0577; mucroprotein, p = 0.3949; lactate dehydrogenase,
p = 0.2805; and acid phosphatase, p = 0.0164.
This document is a draft for review purposes only and does not constitute Agency policy.
C-123 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Time-response relationship for the effect of 1,2,4-TMB (i.e., pseudocumene) on respiratory rate in mice.
120
pseudocumene
2 60-
253 ppm
648 ppm
1054 ppm
-o-
1299 ppm
-%r
1828 ppm
mesitylene
120
b 100'
O
O
5^
I
t
cc
0
hemimellitene
air
»—
J^35^
/ / /
J / / 286 ppm
/ Cj
/ / 511 ppm
/ ~Xr
/ 842 ppm
/ -a-
~ 1581 ppm
12 3 4
5 8 7 8 9 10 11 12 13
min
Each point represents the mean value in 8-10 mice. After termination of 6 min exposure, recovery of respiratory rate
was observed.
Source: Reproduced from Korsak et al. (1997).
Health effect at LOAEL
NOAEL
LOAEL
Increased total BAL cells
N/A
123 mg/m3
Comments: The observed markers of inflammation are coherent with the observed respiratory irritative effects
observed in mice exposed to 1,2,4-TMB acute (i.e., 6 min). The authors did not report at which dose groups the
numbers of polymorphonuclear leucocytes and lymphocytes were significantly elevated relative to control.
This document is a draft for review purposes only and does not constitute Agency policy.
C-124 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-31. Characteristics and quantitative results for Korsak et al. (2000a)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP:
Wistar rats
M & F
10/dose
Inhalation
(6 hrs/d, 5
d/wk)
0, 123, 492,
or
1,230 mg/m
3
90 d
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 (five 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 wk prior to termination of
exposure, and for the 1,230 mg/m3 exposure group, also evaluated 2 wks 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/m3)
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 ± 0.0050
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.
C-125 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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
0.0028 ±
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/m3)
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)
9.98 ± 1.68
9.84 ± 1.82
8.50 ± 1.11
7.70 ± 1.38**
7.61 ± 1.6
0.0004
WBCs (x 107mm3)
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
0.0589
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
76.8 ±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.
C-126 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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)
8.22 ± 1.16
7.93 ± 2.04
8.51 ± 1.13
7.71 ± 1.58
6.99 ± 1.8
0.1891
WBCs (x 107mm3)
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
16 ± 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/m3)
0
123
492
1,230
Trend testb
Clinical chemistry parameters (mean ± SD)
Males
AST (U/dL)
138.7 ± 20.6
141.3 ±21.0
134.5 ± 27.0
138 ±35.0
0.2223
ALT (U/dL)
51.7 ±5.9
48.3 ±7.8
49.7 ±9.1
46.8 ±5.1
0.0637
ALP (U/dL)
80.4 ± 12.0
86.2 ±22.0
84.9 ± 21.0
90.5 ± 19.0
0.1518
SDH (U/dL)
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
1,393 ± 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.
C-127 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Females
AST (U/dL)
139.4 ± 16.6
136.7 ±27.1
145.5 ± 22.7
141.4 ± 15.6
0.2118
ALT (U/d L)
49.8 ±6.3
51.4 ±8.2
50.4 ± 9.0
55.1 ±9.5
0.1844
ALP (U/dL)
41.2 ±7.8
37.2 ±6.8
39.8 ± 11.0
49.8 ± 15.5
0.1740
SDH (U/dL)
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/m3)
[dose group ID]
0
[1]
123
[2]
492
[3]
1,230
[4]
Comparison to
controlsc
Trend testb
Males
Proliferation of peribronchial
lymphatic tissue (0—4)d
16.0e
15.6
30.6
17.4
1-3*
0.13
Formation of lympho-
epithelium in bronchii (0-4)
18.1
15.6
27.9
18.2
22
Bronchitis and broncho-
pneumonia (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.
C-128 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Females
Proliferation of peribronchial
lymphatic tissue (0—4)k
19.4
21.7
21.2
17.5
0.36
Formation of lympho-
epithelium in bronchii (0-4)
18.3
20.1
25.1
16.1
0.48
Bronchitis and broncho-
pneumonia (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
^Statistically significant from controls at p < 0.05.
**Statistically significant from controls at p < 0.01.
aEffects measured in rats exposed to 1,230 mg/m3 2 wks after termination of exposure.
bp-value reported from Jonckheere's trend test.
cReports the results of pair-wise statistical significance of exposure groups compared to controls (i.e., 1-3 would
indicate that the 492 mg/m3 was statistically significantly different from controls).
dGrading system (0-4, 0-3; see Additional study details above).
eResults presented as ranges of the Kruskal-Willis test.
SDH = sorbitol dehydrogenase.
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
Dooulations in BAL fluid in Korsak et al. (1997). The authors did not report the incidences of pulmonarv 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-129 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-32. 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
1,230 mg/m3
group
Inhalation (6
hrs/d, 5 d/wk)
0,123, 492, or 1,230 mg/m3
1,2,3-TMB
90 d
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 (five 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 wk prior to termination of
exposure, and for the 1,230 mg/m3 exposure group, also evaluated 2 wk 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/m3)
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.
C-130 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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/m3)
0
123
492
1,230
l,230a
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), 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), 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), 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), 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
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-131 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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/m3)
0
123
492 1,230
Trend
testb
Clinical chemistry parameters (mean ± SD)
AST (U/dL), males
107.8 ± 14.2
102.9 ± 15.1
103.6 ± 14.5
119.6 ±27.3
0.2223
AST (U/dL), females
96.1 ±9.4
96.9 ±9.9
117.1 ±23.9
104.6 ± 15.7
0.2118
ALT (U/dL), males
41.3 ±2.0
40.7 ±3.1
41.5 ±5.5
45.5 ±5.6
0.0637
ALT (U/dL), females
39.7 ±3.5
39.5 ±6.4
36.2 ±3.3
30.5 ± 9.9**
0.1844
ALP (U/dL), males
70.5 ± 15.2
70.6 ± 11.7
66.5 ± 10.8
63.7 ± 15.7
0.1518
ALP (U/dL), females
21.5 ±2.7
25.8 ±8.4
31.1 ±8.6*
30.5 ±9.9*
0.1740
SDH (U/dL), males
1.6 ±0.7
2.3 ± 1.3
2.5 ±0.9
2.7 ±0.7*
0.0083
SDH (U/dL), 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), 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
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-132 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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/m3)
[Dose group ID]
0
[1]
123
[2]
492
[3]
1,230
[4]
Comparison to
controlsc
Trend testb
Proliferation of peribronchial
lymphatic tissue (0—3)d, males
2.0e (23.4)f
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
24 (22.8)
1.3 (12.1)
1.5 (16.4)
L3 (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
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-133 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Cumulative score of all
Individual males
Cumulative score of all
individual females
7.1 (19.8)
8.4 (24.9)
4.8(11.2)
5.7 (13.5)
7.7 (24.2)
6.5 (16.8)
8.7 (25.8)
8.2 (24.6)
1-2*
p = 0.01
p = 0.4
^Statistically significant from controls at p < 0.05.
**Statistically significant from controls at p < 0.01.
aEffects measured in rats exposed to 1,230 mg/m3 2 wks after termination of exposure.
bp-value reported from Jonckheere's trend test.
cReports the results of pair-wise statistical significance of exposure groups compared to controls (i.e., 1-3 would
indicate that the 492 mg/m3 was statistically significantly different from controls).
dGrading system (0-4, 0-3; see Additional study details above).
eMean.
'Results presented as ranges of the Kruskal-Willis test.
Health Effect at LOAEL
NOAEL
LOAEL
Pulmonary lesions
492 mg/m3
1,230 mg/m3
Comments: The observed inflammatory lesions are coherent with observations of increased inflammatory cell
populations in BAL 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-134 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-33. 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 hrs/d
for 3 consecutive d)
0, 600, 2,400, or
4,800 mg/m3
1,2,4-TMB (as a
constituent of white
spirit)
3d
Additional study details
• Rats were exposed to 1,2,4-TMB as a constituent of white spirit at concentrations of 0, 600, 2,400, or
4,800 mg/m3 for 3 d. Several tests were conducted to evaluate impact of white spirit 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 white spirit (exposure concentration mg/m3)
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 response15
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.
C-135 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Effects of white spirit on total distance run during motor activity assessment in rats.
sr 3000
£0
+
E
o
c
2500
~ 2000
2
§
5 1500
3
O 1000
control
600 mg/m3
2400 mg/m3
4800 mg/m3
**
pre-test
1 x 8 hr
lime of testing
3 x 8 hr
Observation
Visual discrimination performance in rats exposed to white spirit for
3 consecutive d (exposure concentration in mg/m3)c
600
2,400
4,800
Lever response latency (sec)
Before first 8-hr exposure
1.93 ±0.34
2.09 ±0.24
1.70 ±0.15
2.29 ± 0.31*
After first 8-hr exposure
2.44 ±0.56
2.66 ±0.29
3.24 ±0.21
12.00 ± 2.37*
After second 8-hr exposure
2.17 ±0.41
2.32 ±0.29
2.10 ±0.18
4.88 ± 1.53*
After third 8-hr exposure
3.21 ± 1.22
2.68 ±0.41
3.86 ±0.65
6.31 ± 1.35*
One day after third 8-hr
exposure
2.27 ±0.52
1.93 ±0.16
1.88 ±0.16
2.34 ±0.31*
Number of lever response latencies <2 sec
Before first 8-hr exposure
68.00 ± 5.46
67.38 ±2.58
77.12 ±4.32*
71.25 ±4.00*
After first 8-hr exposure
70.38 ±2.93
61.88 ±3.92
58.75 ±2.58*
45.62 ±4.87*
After second 8-hr exposure
70.62 ± 3.60
68.00 ± 3.81
69.00 ± 2.98*
61.50 ± 5.00*
After third 8-hr exposure
71.50 ±3.38
66.38 ±3.34
63.75 ± 5.04*
55.62 ±5.12*
One day after third 8-hr
exposure
72.50 ±3.58
69.75 ±2.90
73.38 ±2.93*
64.88 ±4.23*
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*
This document is a draft for review purposes only and does not constitute Agency policy.
C-136 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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
0.36 ±0.03
0.31 ±0.02
0.34 ± 0.02
0.33 ± 0.04
exposure
^Statistically significant from controls at p < 0.05.
**Statistically significant from controls at p < 0.01.
***Statistically significant from controls at p < 0.001.
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.
Health effect at LOAEL
NOAEL
LOAEL
N/A
N/A
N/A
Comments: Exposure to 1,2,4-TMB was via white spirit, which is comprised of additional substances. LOAEL and
NOAEL values cannot be extracted from this study because other constituents of the white spirit mixture may
confound results.
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-137 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-34. Characteristics and quantitative results for Lutz etal. (2010)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
6-8 rats/
dose
Inhalation (6 hrs/d,
5 d/wk)
0, 25, 100, or 250 ppm (0,
123, 492, or 1,230 mg/m3)
1,2,3- or 1,2,4-TMB
4 wks
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/d, 5 d/wk for 4 wks. Food and water were 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.
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 d after (session 2) a repeated (2.5 mg/kg, 1/d x 5 d)
amphetamine treatment.
Session 1
(before AMPH sensitization)
Session 2
(after AMPH sensitization)
5*2.5 mg/kg AMPH
lllMlli fli
Contra
I I Block 1
HEM HEM HEM
25 ppm 100 ppm 250 ppm
Control
I Block2 j [ BiockS Block 4
HEM HEM HEM
25 ppm 100 ppm 250 ppm
.5 I I Block6
Block 1 — control (preinjcction) activity, block 2 — activity after the SAL injection, blocks 3,4,5 and 6 — activity during successive 30 rain 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 < Q 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, !/dayx5 days) AMPH treatment
This document is a draft for review purposes only and does not constitute Agency policy.
C-138 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 d after (session 2) a repeated
amphetamine treatment (2.5 mg/kg, 1/d x 5 d).
: 280
>
»24Q
200
160
120
80
40
0
Session 1
(before AMPH sensitization)
Session 2
(after AMPH sensitization)
x
_L
5*2,5 mg/kj AMPH
X
T
Control hem HEM HEM
25 ppm 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.
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 d after (session 2) a repeated (2.5 mg/kg, 1/d x 5 d)
amphetamine treatment. Remaining notations are the same as in the figures above.
Session 1
(before AMPH sensitization)
Session 2
(after AMPH sensitization)
5x2,5 mg/kg AMPH
PS
100 ppm
PS
100 ppm
Control
Control
250 ppm
I I Block 1
Hock 2 Lj Block 3
Block 4
Blocks
PS
250 ppm
Blocks
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-139 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 d after (session 2) a repeated
amphetamine treatment (2.5 mg/kg, 1/d x 5 d).
> £ 280
i'E
240
|!
200
160
120
40
0
Session 1
(before AMPH sensitization)
Session 2
(after AMPH sensitization)
5x2,5 mg/kg AMPH
X
T
*
X
T
I
Control
PS
25ppm
PS PS
100 ppm 250 ppm
Control
PS
25 ppm
PS
100 ppm
PS
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.
Health effect at LOAEL
NOAEL
LOAEL
Increased sensitivity to
amphetamine as measured by
open-field locomotion
0 ppm
25 ppm (123 mg/m3) 1,2,4-TMB or
1,2,3-TMB
Comments: This study observed increased amphetamine sensitization, particularly in rats exposed to 100 ppm
(492 mg/m3) 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-140 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-35. Characteristics and quantitative results for Maltoni etal. (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 body
weight 1,2,4-TMB
4 d/wk for 104 wks
Additional study details
• Rats were exposed to 1,2,4-TMB for 2 yrs via stomach tube administration 4 d/wk.
• Animals were 7 wks 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
Number of malignant tumors/100 rats
26.0
34.0
Females
Total benign and malignant tumors
70.0
66.0
Malignant tumors
22.0
24.0
Number of 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
Number of malignant tumors/100 rats
24.0
33.0
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-141 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-142 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-36. Characteristics and quantitative results for Mckee etal. (1990)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
CD-I mice,
rats
Developmental
toxicity: Female
Reproductive
toxicity: M & F
Developmental
toxicity:
30 mice/dose
Reproductive
toxicity:
Fo: 30 rats/sex/dose
Fi: 30 rats/sex/dose
F2: 40 rats/sex/dose
Inhalation
0, 100, 500, or
1,500 ppm
HFAN
(1,3,5-TMB,
1,2,4-TMB,
and
1,2,3-TMB)
6 hrs/d on gestational
days (GDs) 6-15 - mice
Fo: M & F: 6 hrs/d,
10 wks; F: GDs 0-20:
6 hrs/d, 7 d/wk
Additional study details:
• Mice in the developmental toxicity test were exposed to HFAN (1,3,5-TMB, 1,2,4-TMB, and 1,2,3-TMB)
6 hrs/d between GDs 6 and 15.
• Rats in the reproductive toxicity test were exposed to HFAN (1,3,5-TMB, 1,2,4-TMB, and 1,2,3-TMB), in
Fo, Fi, and F2 generations for 6 hrs/d for 10 wks.
• 1,500 ppm was an adverse effect level for both maternal and developmental toxicity.
• In the developmental study, maternal and fetal weight gain was slightly reduced at 500 ppm, while the
100 ppm group did not exhibit maternal or developmental toxicity.
• In the reproductive study, the parental generation had reduced weight gain, but did not exhibit
reproductive toxicity, and birth weights as well as postnatal survival were similar to control values at
1,500 ppm.
• The three generation experiment demonstrated that high level exposures was toxic but had little
effect on reproductive organs.
• The NOAEL was 100 ppm.
Composition of HFAN
Compound
Weight percent
o-Xylene
3.20
Cumene
2.74
n-Propylbenzene
3.97
4-Ethyltoluene
7.05
3-Ethyltoluene
15.1
2-Ethyltoluene
5.44
1,3,5-TMB
8.37
1,2,4-TMB
40.5
1,2,3-TMB
6.18
>C10s
6.19
Total
98.74
This document is a draft for review purposes only and does not constitute Agency policy.
C-143 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mean chamber concentrations (ppm)
Target concentrations
Nominal concentrations
Actual concentrations
Mean
SD
Mean
SD
Developmental toxicity study
0
-
-
-
-
100
102
3.5
102
2.6
500
463
5.3
500
3.7
1,500
1,249
16.5
1,514
22.9
Reproductive toxicity Study
0
-
-
-
-
100
107
2.4
103
2.1
500
513
12.8
495
8.0
1,500
1,483
33.0
1,480
20.5
Reproductive parameters after HFAN exposure
Observation
0 ppm
100 ppm
500 ppm
1,500 ppm
Number of deaths/
number females
0/30
0/30
2/30
14/32a
Number pregnant/
number mated
26/30
26/30
27/30
22/30b
Number of litters with
viable fetuses
24
21
23
13°
Corpora lutea/dam
12.9 ± 1.8d
12.6 ± 1.8
12.7 ±2.3
13.8 ±2.6
Implantations/dam
11.6 ± 1.5d
11.0 ± 1.9
11.3 ± 1.6
12.3 ± 1.8
Live fetuses/litter
10.7 ± 1.8d
8.7 ±4.6*
9.3 ±3.1
7.9 ±4.3*
Postimplantation
loss/dam
0.9 ±0.9d
2.3 ±4.1
2.0 ±3.1
4.3 ±3.7**
Fetal body weight
(grams)
1.25 ±0.14d
1.24 ± 0.08
1.16 ±0.11*
0.82 ±0.17**
Fetal sex ratio, males:
females
57: 41
51:49
54:46
52:48
*p < 0.05.
**p< 0.01.
includes two replacement dams added on GD 6.
bTwo mice died on day 6G; pregnancy could not be determined.
Three litters had resorptions only.
dMean ± SD.
This document is a draft for review purposes only and does not constitute Agency policy.
C-144 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Weights of pregnant mice after HFAN exposure
Maternal body weight
0 ppm
100 ppm
500 ppm
1,500 ppm
GD:
0
25 ± 2.1 (26)a
24 ± 1.7 (24)
25 ±2.2 (27)
25 ±2.8 (27)
6
25 ±2.2 (26)
25 ± 1.9 (24)
26 ±2.3 (27)
26 ±3.3 (21)
15
39 ±3.3 (26)
35 ±7.6 (24)*
36 ±4.9 (25)*
33 ±6.0 (13)**
18
47 ± 3.4 (22)
43 ± 9.6 (24)
44 ± 7.0 (24)
40 ±8.7 (12)*
Maternal body weight gain
Gestational intervals:
Days 0-6
1± 1.7 (26)
1 ± 1.1 (24)
1± 1.12 (27)
1± 1.2 (21)
Days 6-15
16 ±2.2 (26)
14 ± 6.5 (24)
14 ±4.1 (25)**
10 ±5.0 (13)**
Days 0-18
23 ±2.7 (23)
19 ± 8.8 (24)
19 ±5.6 (24)*
14 ±6.8 (12)**
Maternal organ weights (gm)
Lung
0.26 ±0.03 (25
0.27 ±0.04 (26)
0.27 ±0.03 (25)
0.28 ±0.04 (16)
Liver
2.39 ±0.34 (25)
2.35 ±0.51 (26)
2.51 ±0.43 (25)
2.43 ±0.53 (16)
Kidney
0.40 ± 0.06 (25)
0.41 ±0.06 (25)
0.42 ± 0.05 (25)
0.42 ±0.05 (16)
*p < 0.05.
**p< 0.01.
aMean ± SD, number examined given in parentheses.
Fetal alterations in fetuses after HFAN exposure
Observation
0 ppm
100 ppm
500 ppm
1,500 ppm
External examination
280 (26)
226 (21)
232 (24)
128 (13)
Visceral examination
139 (26)
112 (21
112 (24)
68 (13)
Skeletal examination
141 (26)
114 (21)
120 (24)
60 (13)
Malformations observed
Ablepharia
-
1(1)
1(1)
-
Folded retina
7(26)
5(5)
4(3)
1(1)
Cleft palate
1(1)
-
1(1)
14 (7)
Mandibular micrognathia
-
-
1(1)
-
Thoracogastroschisis
-
1(1)
-
-
Syringomyelocele
-
-
-
1(1)
Sternoschisis
-
1(1)
-
-
Interrupted ossification of an
arch
1(1)
1(1)
-
-
Vertebrae malformation
(with or without an
associated rib malformation)
5(4)
3(3)
4(4)
2(2)
Rib malformation
1(1)
-
1(1)
1(1)
Interrupted ossification of a
rib
1(1)
-
-
Total fetuses (litters) with
malformations
15 (10)
11 (8)
11(8)
19 (7)
This document is a draft for review purposes only and does not constitute Agency policy.
C-145 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Developmental variations
Tarsal flexure
-
1(1)
-
-
Skull reduced in ossification
-
-
-
18 (6)
Accessory skull bone
-
-
4(3)
-
Hyoid unossified
-
-
1(1)
-
14th Rudimentary rib(s)
25 (15)
18 (12)
18 (12)
10 (7)
More than 13 pairs of full
ribs
17 (9)
21(12)
26 (9)
27 (10)
7th cervical ribs
37 (16)
25 (15)
19 (11)
12 (7)
Sternebrae #5 and/or #6
unossified
-
1(1)
3(2)
25 (10)
Fused sternebrae
3(3)
-
-
-
Misaligned sternebrae
7(5)
7(7)
6(6)
1(1)
Extra sternebrae
3(3)
1(1)
-
1(1)
Other sternebrae unossified
-
-
-
4(3)
Total fetuses (litters) with
variations
78 (24)
63 (20)
67 (22)
48 (13)
Fertility indices after HFAN exposure
Observation
0 ppm
100 ppm
500 ppm
1,500 ppm
Pregnant females/number of females mated
Parental generation:
Fo
93.3 (30)
96.7 (30)
93.3 (30)
92.6 (27)
Fi
80 (30)
76.7 (30)
96.7 (30)
88.9 (27)
F2
96.7 (30)
93.3 (30)
96.7 (30)
83.3 (6)
Females delivering a live litter/number identified pregnant females (%)
Parental generation:
Fo
103.6 (28)
100 (29)
89.3 (28)
92.0(25)
Fi
125 (30)
104.3 (30)
96.7 (30)
87.0 (24)
F2
96.6 (30)
100 (30)
96.6 (30)
120 (6)
Females delivering a live litter/number of females delivering a litter (%)
Parental generation:
Fo
100 (29)
100 (29)
96.1 (26)
100 (23)
Fi
100 (30)
100 (24)
96.7 (30)
90.5 (21)
F2
100 (28)
100 (28)
100 (28)
100 (6)
Fertile males/number of males mated (%)
Parental generation:
Fo
86.7 (30)
96.7 (30)
83.3 (30)
84.6 (26)
Fi
89.7 (30)
86.7 (30)
93.3 (30)
64.3 (28)*
F2
93.3 (30)
83.3 (30)
80.0 (30)
100 (4)
This document is a draft for review purposes only and does not constitute Agency policy.
C-146 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Cohabitation time (days) required for mating3
Parental generation:
Fo
2.9 ±2.3
2.1 ± 1.6
3.8 ±2.5
4.2 ±2.7
Fi
3.3 ±2.4
2.5 ±2.1
2.6 ±2.2
4.5 ±2.9
F2
2.3 ± 1.1
3.0 ±2.4
3.4 ±2.9
3.4 ± 1.3
Litter size at birth"
Parental generation:
Fo
12.1 ±3.4
12.9 ± 1.5
12.2 ±3.1
11.3 ±3.0
Fi
12.4 ±2.0
11.1 ±2.9
11.7 ±3.0
8.7 ±4.3**
F2
12.6 ±2.7
11.8 ±2.3
11.4 ±2.1
12.2 ± 1.3
*p < 0.05.
**p< 0.01.
aAverage number of male/female cohabitation days required to produce a sperm-positive vaginal smear.
bMean (± SD) number of live offspring delivered.
Gestation and postnatal survival among litters after HFAN exposure
0 ppm
100 ppm
500 ppm
1,500 ppm
Gestational survival index3 (%)
Generation:
Fo
95.9 (366)
97.9 (382)
94.9 (333)
92.8 (279)
Fi
97.4 (383)
95.4 (280)
91.6 (371)
85.1(215)**
F2
97.8 (361)
98.2 (335)
98.5 (325)
100 (73)
Postnatal survival index, 4-d
(%)
Generation:
Fo
93.7 (351)
93.3 (374)
98.7 (316)
94.2 (260)
Fi
95.4 (373)
96.3 (267)
97.6 (340)
87.4 (183)
F2
97.5 (353)
96.4 (329)
97.5 (320)
97.3 (73)
Postnatal survival index, 21-dc(%)
Generation:
Fo
99.1 (214)
99.6 (225)
100 (200)
95.1 (164)
Fi
96.2 (234)
98.9 (179)
98.6 (216)
99.2 (119)
F2
100 (215)
99.1 (216)
99.1 (220)
97.9 (48)
**Significantly different from control.
aPups alive at birth/number of pups born (%).
bPups surviving for 4 days/total number of liveborn pups.
cPups surviving for 21 days/total number of live pups after culling on day 4.
Body weights of pups exposed to HFAN
0 ppm
100 ppm
500 ppm
1,500 ppm
Day 0 body weights
Generation:
Fo
6.1 ±0.5
6.2 ±0.5
6.5 ±0.6
6.1 ± 1.0
Fi
6.0 ±0.5
6.1 ±0.5
6.0 ±0.5
5.7 ±0.7
F2
6.0 ±0.5
6.0 ±0.4
6.1 ±0.6
5.7 ±0.2
This document is a draft for review purposes only and does not constitute Agency policy.
C-147 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Day 4 body weights
Generation:
Fo
9.7 ±0.9
9.8 ±0.6
10.1 ± 1.0
9.2 ± 1.3
Fi
9.5 ± 1.4
10.0 ± 1.2
9.9 ± 1.0
9.3 ± 1.0
F2
9.7 ± 1.1
10.0 ± 0.7
9.8 ± 1.0
9.2 ±0.6
Day 7 body weights
Generation:
Fo
13.7 ± 1.3
13.2 ± 1.1
14.0 ± 1.7
12.0 ± 1.8
Fi
13.3 ± 1.8
13.3 ± 1.6
13.5 ± 1.4
11.7 ± 1.3
F2
14.0 ± 2.0
14.1 ± 1.2
13.4 ± 1.5
12.0 ± 1.0
Day 14 body weights
Generation:
Fo
24.9 ±2.7
23.2 ± 1.8
23.9 ±2.4
19.6 ±2.7
Fi
24.3 ±2.5
23.5 ±2.8
23.7 ±2.7
19.3 ± 1.8
F2
26.2 ±4.0
25.6 ± 1.9
23.2 ±2.7
20.8 ± 1.3
Day 21 male body weights
Generation:
Fo
39.5 ±5.1
37.2 ±5.9
40.0 ± 4.9
29.9 ±3.6
Fi
40.9 ±5.5
39.3 ±5.5
39.7 ±5.6
30.4 ± 4.2
F2
42.9 ±7.6
42.7 ±3.8
38.7 ±5.1
32.8 ±3.0
Day 21 female body weights
Generation:
Fo
38.0 ±5.0
35.7 ±5.7
38.0 ± 5.0
29.4 ±4.3
Fi
39.6 ±5.1
37.9 ±4.8
38.6 ±5.5
29.1 ±4.2
F2
41.4 ±6.2
41.2 ±3.6
37.2 ±4.8
31.8 ±3.6
Effect of prolonged exposure to HFAN on gestation and postnatal survival (f2 generation)
0 ppm
100 ppm
500 ppm
1,500 ppm
Litter size3
Total
12.4 ± 2.0 (30)
11.1 ±2.9 (24)
11.7 ± 3.0 (30)
8.7 ±4.3** (21)
Prolonged exposure
11.3 ± 1.8 (6)
11.0 ± 2.4 (8)
4.0 (l)b
4.9 ±5* (7)
Exposure stopped on
GD 20
12.7 ± 1.9 (24)
11.2 ±3.2 (16)
12.0 ±2.7 (29)
10.6 ± 2.2 (14)
Birth weight3
Total
6.0 ±0.5 (30)
6.1 ±0.5 (24)
6.0 ±0.5 (30)
5.7 ±0.7 (21)
Prolonged exposure
6.0 ±0.6 (6)
5.9 ±0.4 (8)
5.4 (l)b
5.1 ±0.7
Exposure stopped on
GD 20
6.0 ±0.5 (24)
6.2 ±0.5 (16)
6.0 ±0.5 (29)
5.9 ±0.6 (14)
Gestation survival index0
Total
97.4 (383)
95.4 (280)
91.6 (371)
85.1**(215)
Prolonged exposure
91.9 (74)
91.7 (96)
30.8 (13)
63.0 (54)
Exposure stopped on
GD 20
98.7 (309)
97.2 (184)
93.8 (358)
92.5**(161)
This document is a draft for review purposes only and does not constitute Agency policy.
C-148 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Postnatal survival index, d 4C
Total
95.4 (373)
96.3 (267)
97.6 (340)
87.4 (183)
Prolonged exposure
82.3 (68)
90.9 (88)
100 (4)
44.1(34)
Exposure stopped on
GD 20
98.4 (305)
98.9 (179)
97.6 (336)
97.3 (149)
Postnatal survival index, d 21°
Total
99.6 (226)
99.4 (178)
99.1 (215)
99.2 (119)
Prolonged exposure
100 (34)
98.3 (61)
100 (4)
91.7 (12)
Exposure stopped on
GD 20
99.5 (192)
100 (117)
99.0 (211)
100 (107)
*p < 0.05.
**p< 0.01.
aNumber of live born offspring/litter; number of litters given.
Statistics not conducted because of small sample size,
initial number of offspring for evaluation interval given.
NOAEL
LOAEL
LOAEL effects
100 ppm, fetal weight gain (F3
generation)
500 ppm
Fetal weight gain, and maternal
weight gain reduced
This document is a draft for review purposes only and does not constitute Agency policy.
C-149 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-37. Characteristics and quantitative results for Mckee etal. (2010)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
8 rats/
group
Inhalation (
C
), 125, 1,250, or
5,000 mg/m31,2,4-TMB
8 hrs/d for 3 consecutive d
Additional study details
• Animals were exposed to 1,2,4-TMB for 8 hrs/d for 3 d 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.
Observation
Exposure concentration 1,2,4-TMB (mg/m3)
0
125
1,250
5,000
Results of functional and motor activity observations
Forelimb grip strength (g)
1 d 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)
1 d 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
1 d 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*
Observation
Exposure concentration 1,2,4-TMB (mg/m3)
0
125
1,250
5,000
Visual discrimination performance testing (means ± SD)
Trials3
1 d 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
1 d post-exposure
100 ± 0.0
100 ± 0.0
100 ± 0.0
100 ± 0.0
This document is a draft for review purposes only and does not constitute Agency policy.
C-150 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Percentage reinforcements obtained15
1 d 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
1 d post-exposure
99.63 ±0.26
99.88 ±0.13
99.88 ±0.13
100 ± 0.0
Discrimination ratio0
1 d 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
1 d post-exposure
0.87 ± 0.03
0.89 ±0.03
0.92 ±0.02
0.88 ±0.03
Percentage inter-trial intervals responded tod
1 d 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
1 d post-exposure
10.88 ± 1.39
10.63 ± 1.81
11.25 ±0.92
8.50 ± 1.40
Repetitive errors6
1 d 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
1 d post-exposure
4.75 ±2.81
2.75 ± 1.35
4.63 ± 3.09
4.13 ± 1.38
Repetitive inter-trial responses'
1 d 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*
1 d post-exposure
6.63 ± 1.94
2.88 ±0.83
5.13 ± 1.54
2.63 ±0.68
Trial response latency8
1 d 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
1 d post-exposure
1.68 ±0.16
2.70 ±0.60
2.18 ±0.73
1.45 ± 0.06
SD of response latency
1 d 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
1 d post-exposure
1.84 ±0.38
5.95 ±2.40
5.88 ±4.21
1.81 ±0.38
This document is a draft for review purposes only and does not constitute Agency policy.
C-151 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Latency <2 sech
1 d 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
1 d post-exposure
69.38 ±2.98
67.63 ±3.20
78.13 ±3.05
78.00 ± 2.34
Latency >6 sec1
1 d 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
1 d post-exposure
2.13 ±0.67
6.00 ± 1.68
3.38 ± 1.40
1.88 ±0.35
Drink response latency*
1 d 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
1 d post-exposure
0.27 ±0.01
0.34 ± 0.03
0.36 ±0.03
0.30 ± 0.02
^Statistically significant from controls at p < 0.05.
aTotal number of trials completed during each session, maximum = 100.
bNumber of reinforcements obtained divided by the number of reinforcements delivered (x 100).
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
(xl00).
eThe total number of incorrect trial responses following an initial incorrect response.
The total number of ITI responses following an initial ITI response.
gThe latency (sec) to make a correct trial response.
hThe number of responses within 2 sec.
The number of responses taking more than 6 sec.
The mean latency (sec) to obtain reinforcement.
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
values 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 d 3. Acute duration of exposure (exposure on 3 consecutive d).
Generally, acute exposure studies have limited utility in quantitation of human health reference values.
This document is a draft for review purposes only and does not constitute Agency policy.
C-152 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-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/
dose
Inhalation (6 hrs/d
GDs 6-20)
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
GDs 6-20
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/d
starting on GD 6 and ending on GD 20.
• Animals were randomized and assigned to the experimental groups.
• After GD 20, dams were sacrificed and weighed, as were their uteri and any fetuses.
• Decreases in maternal body weight and fetal toxicity were observed.
Observation
Exposure concentration to 1,3,5-TMB
0 ppm
100 ppm
(492 mg/m3)
300 ppm
(1,476 mg/m3)
600 ppm
(2,952 mg/m3)
1,200 ppm
(5,904 mg/m3)
Maternal parameters
Number of treated
24
24
24
24
24
Number of (%) pregnant at
euthanization
21 (87.5)
22 (91.7)
21(87.5)
17 (70.8)
18 (75.0)
Number of deaths
0
0
0
0
0
Body weight (g) on d 6
274 ± 17g
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/d)
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.
C-153 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Observation
Exposure concentration to 1,3,5-TMB
0 ppm
100 ppm
(492 mg/m3)
300 ppm
(1,476 mg/m3)
600 ppm
(2,952 mg/m3)
1,200 ppm
(5,904 mg/m3)
Gestational parameters
All litters"
21
22
21
17
18
Number 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 number 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 number 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/m3)
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.
C-154 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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
(492 mg/m3)
300 ppm
(1,476 mg/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 d 6
271 ± 18g
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/d)
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**
This document is a draft for review purposes only and does not constitute Agency policy.
C-155 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Observation
Exposure concentration to 1,2,4-TMB
0 ppm
100 ppm
(492 mg/m3)
300 ppm
(1,476 mg/m3)
600 ppm
(2,952 mg/m3)
900 ppm
(4,428 mg/m3)
Gestational parameters
All litters"
24
22
22
22
24
Number of corpora lutea per
dam
15.4 ± 2.1g
15.2 ± 1.3
15.2 ±2.1
15.8 ± 1.7
15.7 ±2.5
Mean number 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
Live littersd
23
22
22
22
24
Mean number 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
(492 mg/m3)
300 ppm
(1,476 mg/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.
C-156 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Skeletal variations
Third sternebrae,
0
1(1)
0
0
0
incomplete ossification
Fifth sternebrae
1(1)
0
4(4)
5(4)
6(6)
incomplete ossification or
unossifiedf
Extra ossification site
0
1(1)
0
0
0
Cervical rib, rudimentary
1(1)
2(2)
0
3(2)
2(2)
Fourteenth rib,
25 (10)
13 (8)
18 (12)
21 (10)
34 (16)
supernumerary
Thirteenth rib, short
1(1)
0
0
0
0
(unilateral)
Thoracic vertebral centra,
8(6)
4(4)
7(4)
6(6)
7(5)
incomplete ossification
^Statistically significant from controls at p < 0.05.
**Statistically significant from controls at p < 0.01.
aBody weight gain during GDs 6-21 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.
Health effect at LOAEL
NOAEL
LOAEL
Maternal toxicity: decrease in
Maternal toxicity: 300 ppm (1,476 mg/m3)
Maternal toxicity: 600 ppm (2,952 mg/m3)
maternal body weight and
for 1,3,5-TMB and 1,2,4-TMB
for 1,3,5-TMB and 1,2,4-TMB
food consumption
Developmental toxicity:
Fetal toxicity: 300 ppm (1,476 mg/m3) for
Fetal toxicity: 600 ppm (2,952 mg/m3) for
significant reduction in fetal
1,2,4-and 1,3,5-TMB
1,2,4-and 1,3,5-TMB
body weight
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 was not investigated.
This document is a draft for review purposes only and does not constitute Agency policy.
C-157 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-39. Characteristics and quantitative results for Schreiner et al. (1989)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-Dawley
Rats
M & F
15 male and 15
female/dose
group
Inhalation
150, 500, 1,500
ppm HFAN
(1,2,4-TMB,
1,3,5-TMB, and
1,2,3-TMB)
6 hrs/d for 5
consecutive d
Additional study details:
• Rats were exposed by inhalation to mixture HFAN (1,2,4-TMB, 1,3,5-TMB, and 1,2,3-TMB) 6 hrs/d for
5 d
• The positive control contained 10 rats total (5 M & 5 F), while the experimental and negative controls
each contained 30 rats total (15 M & 15 F)
• All animal groups were exposed in 16 m3 glass and stainless steel chambers
• There were no increases in SCE or chromosomal aberration frequency in Chinese hamster ovary (CHO)
cells
• Both male and females exhibited a 10% reduction in body weight gain at 1,500 ppm
• HFAN was not clastogenic at levels up to and including 1,500 ppm
• The NOAEL was 500 ppm
Physical and chemical properties of HFAN (CAS 64742-95-6)
ASTM D-3734 specifications
Composition (weight percent)3
Appearance
Clear and free of suspended
matter and undissolved water
o-Xylene
3.20
Color
Not darker than +25 Saybolt
Cumene [isopropylbenzene]
2.74
Aromatics, volume %
90 minimum
n-Propylbenzene
3.97
Copper corrosion, l/2h at
100°C
No iridescence, discoloration, or
gray or black deposits
4-Ethyltoluene
7.05
Distillation, °F
3-Ethyltoluene
15.1
Initial boiling point
300 minimum
2-Ethyltoluene
5.44
10%
-
1,3,5-TMB
8.37
50%
335 maximum
1,2,4-TMB
40.5
90%
-
1,2,3-TMB
6.18
Dry point
335 maximum
>C10s
6.19
Flash Point, °F
100 minimum
Total
98.74
Kauri-butanol value
87 minimum
Mixed aniline point, °F
60 maximum
Odor
Characteristic, as agreed
Specific gravity
0.864 minimum
60/60°F
0.884 maximum
This document is a draft for review purposes only and does not constitute Agency policy.
C-158 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Exposure to HFAN (nL/plate) (CAS 64742-95-6)
Mutagenic response
Observation
DMSO
Control
Positive
Control
C9 Aromatics
Dose
(nL/plate)
50
A
0.0025
0.0050
0.0100
0.0250
0.0500
0.1000
0.2500
0.5000
TA1535 (-S9)
12.3 ±
2.5
1,075.0
± 31.4b
11.0±3.6
10.7±3.1
13.3±3.5
14.0±2.6
12.3±0.6
12.0±1.7
10.7±0.6
5.3±1.
2
TA1535 (+S9)
10.3 ±
1.5
209.3±
17.8b
8.7 ±0.6
8.7 ± 1.5
9.0 ± 1.7
8.7 ±2.3
9.7 ±4.2
6.7 ±4.0
9.7 ±0.6
6.3 ±
2.1
TA1538 (-S9)
11.7 ±
2.9
1,269.7
± 51.6b
13.7 ±2.5
16.3 ±
0.6
12.7 ±
1.5
13.7 ±
7.4
13.0 ±
2.0
11.7 ±
2.1
12.7 ± 1.2
10.3 ±
1.5
TA1538 (+S9)
22.3 ±
4.7
981.0 ±
28.6b
17.0 ±2.6
17.3 ±
0.6
15.7 ±
5.1
17.3 ±
2.9
13.0 ±
2.6
17.0 ±
3.6
16.7 ±6.4
14.7 ±
1.2
TA98 (-S9)
21.0 ±
2.6
1,088.3
± 73.3b
22.3 ±6.1
24.0 ±
1.7
21.3 ±
6.5
23.0 ±
2.6
18.3 ±
1.5
19.0 ±
5.6
19.0 ±4.6
11.0 ±
2.6
TA98 (+S9)
27.7 ±
8.3
1,486.0
± 78.5b
24.3 ±4.5
30.7 ±
4.0
29.3 ±
1.5
26.3 ±
2.3
24.7 ±
0.6
26.3 ±
4.0
25.0 ±3.5
24.7 ±
3.1
TA100 (-S9)
106.7 ±
4.9
1,053.7
± 22.8 b
116.0 ±9.6
103.7 ±
4.6
102.0 ±
10.5
107.7 ±
8.4
109.3±
14.2
106.3 ±
12.7
86.0 ± 14.4
66.3 ±
10.2
TA100 (+S9)
102.7 ±
15.0
1,761.0
± 60.2 b
104.3 ±
11.9
94.7 ±
7.6
90.7 ±
4.0
111.0 ±
18.0
102.3 ±
3.8
86.0 ±
14.1
82.0± 3.5
94.0 ±
6.1
TA1537 (-S9)
10.0 ±
2.6
1,008.7
± 21. lb
7.3 ±0.6
7.0 ±2.0
9.0 ±2.0
10.7 ±
3.2
9.3 ± 1.2
10.3 ±
3.2
5.3 ±5.0
5.0 ±
2.0
TA1537 (+S9)
10.7 ±
3.8
159.3 ±
6.8 b
10.3 ±2.1
9.3 ± 1.5
10.0 ±
2.0
11.3 ±
2.1
11.3 ±
0.6
12.3 ±
2.3
6.7 ± 1.5
10.7 ±
2.3
aPositive control:
(1) Activation (+S9): all strains: 2-anthramine (2.5 ng/plate)
(2) Nonactivation (-S9): TA1538: 2-nitrofluorene (10 ng/plate)
TA98: 2-nitrofluorene (10 ng/plate)
TA1534: sodium azide (10 ng/plate)
TA100: sodium azide (10 ng/plate)
TA1537: quinacrine mustard (5 ng/plate)
bResult equal to or greater than 3 times the spontaneous reversion frequency.
DMSO = dimethylsulfoxide.
This document is a draft for review purposes only and does not constitute Agency policy.
C-159 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Exposure to HFAN without metabolic activation (CAS 64742-95-6)
CHO/HGPRT forward mutation suspension assay
Observation
Vehicle controls:
Positive controls:
Cg aromatics
DMSO
DMSO
Brdll
MMS
MMS
Dose nL/mL
10
10
50
15
20
0.01
0.02
0.04
0.06
0.07
0.08
0.1
0.13
Mean colony
number ± SD
202.7 ±
7.6
190.0 ±
17.8
161.3
± 11.2
83.0 ±
7.0
41.0 ±
7.2
185.0
± 10.6
204.7
± 1.5
204.3
±2.5
202.7
±20.5
77.3 ±
7.0
0.08
UL/mL
0.0 ±
0.0
0.0 ±
0.0
Percent
vehicle
control
103.2
96.8
82.1
42.3
20.9
94.2
104.3
104.0
103.2
39.4
2.9
0.0
0.0
Relative
population
growth (% of
control)
111.0
89.0
114.1
63.5
38.3
176.6
148.6
147.5
107.5
35.2
10.7
NDb
ND
Total mutant
colonies in
12 dishes
2
4
27d
95
88
2C
0C
4
r
3
2
ND
ND
Absolute CE
± SD (%)
80.7 ±
6.2
77.5±
3.1
87.5 ±
4.5
66.9 ±
3.2
61.7 ±
3.6
94.2 ±
7.6
93.5 ±
4.1
86.5 ±
6.3
86.5 ±
4.8
91.0 ±
10.1
94.0 ±
3.1
ND
ND
Mutant
frequency in
10"6 units3
1.0
2.2
14.0e
59.2e
59.4e
1.1
0.0
1.9
0.6
1.4
0.9
ND
ND
aMutant frequency = total mutant colonies/(number of dishes x 2 x 105 x absolute CE).
bND = not determined due to excessive toxicity.
Total number of dishes = 10.
dTotal number of dishes = 11.
Significant increase, p < 0.01.
Brdll = 5-bromo-2'-deoxyuridine; CE = cloning efficiency; MMS = methyl methanesulfonate.
Exposure to HFAN with metabolic activation (CAS 64742-95-6) (uL/mL)
CHO/HGPRT forward mutation suspension assay
Vehicle controls
DMSO
Positive
control
Cg aromatics
Dose uL/mL
10
10
5 uL/mL
3-MCe
0.02
0.04
0.06
0.08
0.1
0.13
0.16
0.2
Mean colony
number ± SD
203.7 ±
16.9
201.0 ±
12.5
201.0 ±7.8
185.3 ±
3.5
205.3 ±
21.1
196.7
±22.0
3.3 ±
1.5
0.0 ±
0.0
0.0 ±
0.0
0.0 ±
0.0
0.0 ±0.0
Percent
vehicle
control
100.7
99.3
99.3
91.6
101.5
97.2
1.6
0.0
0.0
0.0
0.0
Relative
population
growth (% of
control)
90.5
109.5
77.1
119.7
111.7
110.0
4.0
NDb
ND
ND
ND
Total mutant
colonies in
12 dishes
2C
8d
245f
6
7C
8
3
ND
ND
ND
ND
This document is a draft for review purposes only and does not constitute Agency policy.
C-160 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Absolute CE
± SD (%)
99.7 ±
6.4
90.9 ±
7.1
84.4 ± 7.4
97.9 ±
2.9
92.4 ±
9.9
99.2 ±
9.0
98.4 ±
12.6
ND
ND
ND
ND
Mutant
frequency in
10"6 units3
0.9
4.4
161.3s
2.6
3.4
3.4
1.3
ND
ND
ND
ND
aMutant frequency = total mutant colonies/(number of dishes x 2 x 105 x absolute CE).
bND = not determined due to excessive toxicity.
Total number of dishes = 11.
dTotal number of dishes = 10.
e3-methylcholanthrene.
Total number of dishes = 9.
Significant increase, p < 0.01.
SCE in CHO cells exposed to HFAN in the absence of metabolic activation
Assay 1
Controls
Observation
Negative:
none
Solvent:
DMSO
Positive:
MMCa
C9 aromatics
Dose ng/mL
(nL/mL)
-
11
0.005
2.00
6.67
20.00
66.70
200.00
Total cells
scored
50
50
20
50
50
50
50
-
Number of
chromosomes
1,044
1,038
420
1,037
1,038
1,044
1,038
Toxic
Number of SCE
443
536
570
530
474
480
524
-
SCE
chromosomes
0.42
0.52
1.36
0.51
0.46
0.46
0.50
-
SCE/cell (mean ±
SE)
8.86 ±0.36
10.72 ± 0.45
28.50 ±
1.13b
10.60 ±
0.43
9.48 ±0.51
9.60 ± 0.44
10.48 ±
0.39
-
Cell cycle stages
(%): Ml
1.5
2.5
1.0
2.5
2.0
3.5
6.5
-
M1+
12.5
39.0
22.5
36.5
48.0
30.0
57.0
-
M2
86.0
58.5
76.5
61.0
50.0
66.5
36.5
-
% SCE increase
over solvent
-
163
Confluence %
solvent control
-
100
100
100
100
100
100
100
Assay 2
Dose ng/mL
(nL/mL)
-
11
0.005
35.0
50.1
66.7
90.1
Total cells
scored
50
50
20
50
50
50
Number of
chromosomes
1,038
1,047
417
1,043
1,042
1,041
10x1 (
Number of SCE
399
432
547
428
461
443
-
SCE
chromosomes
0.38
0.41
1.31
0.41
0.44
0.43
-
This document is a draft for review purposes only and does not constitute Agency policy.
C-161 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
SCE/cell (mean ±
SE)
7.98 ±0.38
8.64 ±0.50
27.35 ±
1.49b
8.56 ±0.49
9.22 ±0.36
8.86 ± 0.44
-
Cell cycle stages
(%): Ml
0.5
2.0
1.5
3.0
7.0
11.0
-
M1+
6.0
16.0
9.5
27.0
47.5
45.0
-
M2
93.5
82.0
89.0
70.0
45.5
44.0
-
% SCEincrease
over solvent
218
7
3
-
Confluence %
solvent control
100
100
100
100
63
6
aMitomycin C.
Significant increase versus solvent controls.
SCE in CHO cells exposed to HFAN in the presence of metabolic activation
Assay 1
Controls
Observation
Negative:
None
Solvent:
DMSO
Positive:
CPa
C9 Aromatics
Dose ng/mL
(nL/mL)
-
11
1.5
0.667
2.00
6.67
20.0
66.7
Total cells
scored
50
50
20
50
50
50
50
-
Number of
chromosomes
1,037
1,032
415
1,038
1,034
1,045
1,040
Toxic
Number of SCE
443
430
379
449
484
474
441
-
SCE
chromosomes
0.43
0.42
0.91
0.43
0.47
0.45
0.42
-
SCE/cell (mean ±
SE)
8.86 ± 0.43
8.60 ± 0.49
18.95 ±
1.20b
8.98 ±0.34
9.68 ±0.43
9.48 ± 0.46
8.82 ±0.45
-
Cell cycle stages
(%): Ml
-
1.5
0.5
-
-
1.5
-
-
M1+
18.5
15.5
24.0
20.0
16.5
19.5
18.5
-
M2
81.5
83.0
75.5
80.0
83.5
79.0
81.5
-
M2+
-
-
-
-
-
-
-
-
% SCE increase
over solvent
-
119
4
12
9
2
-
Confluence %
solvent control
-
100
100
100
100
100
100
7
Assay 2
Dose ng/mL
(nL/mL)
-
11
1.5
15.0
20.0
35.0
50.1
66.7
Total cells
scored
50
50
20
50
50
50
50
-
Number of
chromosomes
1,048
1,046
418
1,043
1,048
1,055
1,047
Toxic
Number of SCE
417
398
457
372
444
400
420
-
This document is a draft for review purposes only and does not constitute Agency policy.
C-162 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
SCE
chromosomes
0.40
0.38
1.09
0.36
0.42
0.38
0.40
-
SCE/cell (mean ±
SE)
8.34 ± 0.43
7.96 ±0.38
22.85 ±
0.91b
7.44 ± 0.40
8.88 ± 0.44
8.00 ± 0.46
8.40 ± 0.48
-
Cell cycle stages
(%): Ml
-
-
0.5
0.5
0.5
1.5
0.5
-
M1+
10.5
20.0
15.0
8.0
14.0
14.5
23.0
-
M2
89.5
80.0
84.5
90.0
85.5
82.5
76.5
-
M2+
-
-
-
1.5
-
1.5
-
-
% SCEincrease
over solvent
187
11
5
-
Confluence %
solvent control
100
100
100
100
100
63
6
Cyclophosphamide.
Significant increase versus solvent controls.
Chromosome aberrations in CHO cells exposed to HFAN in the absence of metabolic activation
Assay 1
Controls
C9 Aromatics
Negative and
solvent
Positive:
MMC
Dose (ng/mL)
1.0
45.0
60.0
75.0
90.0
Cells scored
200
25
200
200
200
200
Number of
aberrations per
cell
0.03
0.32
0.02
0.01
0.00
0.01
% cells with
aberrations
2.5
24.0a
2.0
0.5
0.0
1.0
% cells with
>1 aberration
0.0
00
o
qj
0.0
0.0
0.0
0.0
Assay 2
Dose (ng/mL)
1.0
15.0
30.1
60.1
90.2
Cells scored
200
25
200
200
200
Toxic
Number of
aberrations per
cell
0.01
0.32
0.02
0.04
0.02
% cells with
aberrations
0.5
24.0a
1.0
2.0
1.5
-
% cells with
>1 aberration
0.0
00
o
QJ
0.5
0.5
0.0
-
Significantly greater than the pooled negative and solvent controls, p < 0.01.
This document is a draft for review purposes only and does not constitute Agency policy.
C-163 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Chromosome aberrations in CHO cells exposed to HFAN in the presence of metabolic activation
Assay 1
Controls
C9 aromatics
Negative and
solvent
Positive: CPa
Dose (ng/mL)
50.0
25.0
37.5
50.0
70.0
Cells scored
200
25
200
200
200
200
Number of
aberrations per
cell
0.03
0.28
0.03
0.02
0.01
0.01
% cells with
aberrations
2.5
24.0a
3.0
2.0
1.0
0.5
% cells with
>1 aberration
0.0
4.0
0.0
0.0
0.0
0.0
Assay 2
Dose (ng/mL)
50.0
20.0
40.1
60.1
80.2
100
Cells scored
200
25
200
200
14
100
Toxic
Number of
aberrations per
cell
0.03
0.28
0.01
0.02
0.00
0.01
% cells with
aberrations
2.0
24.0a
1.0
1.5
0.0
1.0
% cells with
>1 aberration
0.5
4.0
0.0
0.5
0.0
0.0
Cyclophosphamide.
Significantly greater than the pooled negative and solvent controls, p < 0.01.
Chromosome aberrations due to exposures of 6 hrs/d on 5 consecutive d
6-Hr post-exposure interval
Exposure group
Number and sex
Number of
spreads
Number of
aberrations
% Aberrations
per metaphase
>1
Aberration
>2
Aberrations
Air
3 M
150
0
0
0
0
3 F
250
0
0
0
0
8 Combined
400
0
0
0
0
150 ppm
5 M
250
0
0
0
0
4 F
200
0
0
0
0
9Combined
450
0
0
0
0
500 ppm
5 M
250
0
0
0
0
5 F
237
0
0
0
0
10 Combined
487
0
0
0
0
1,500 ppm
5 M
250
0
0
0
0
4 F
200
0
0
0
0
9 Combined
450
0
0
0
0
This document is a draft for review purposes only and does not constitute Agency policy.
C-164 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
24-Hr post-exposure interval
Air
4 M
200
0
0
0
0
5 F
250
1
0.4
0.4
0
9 Combined
450
1
0.2
0.2
0
150 ppm
5 M
250
0
0
0
0
5 F
432
0
0
0
0
10 Combined
482
0
0
0
0
500 ppm
5 M
250
0
0
0
0
5 F
250
0
0
0
0
10 Combined
500
0
0
0
0
1,500 ppm
5 M
250
1
0.4
0.4
0
5 F
250
0
0
0
0
10 Combined
500
1
0.2
0.2
0
Cyclophosphamide
4 M
203
70b
34.5b
16.3b
10.3b
5 F
250
60b
24b
13.2b
6.4b
9 Combined
453
130b
28.7b
14.6b
8.2b
48-Hr exposure interval
Air
2 M
100
0
0
0
0
2 F
100
0
0
0
0
4 Combined
200
0
0
0
0
150 ppm
2 M
100
0
0
0
0
2 F
100
0
0
0
0
4 Combined
200
0
0
0
0
500 ppm
3 M
150
0
0
0
0
1 F
20
0
0
0
0
4 Combined
200
0
0
0
0
1,500 ppm
2 M
100
0
0
0
0
1 F
20
0
0
0
0
3 Combined
150
0
0
0
0
aData were evaluated only under the following conditions:
(1) Animal had at least 30 readable metaphase spreads.
(2) At least three animals (of either sex) with adequate data at any time point.
Statistical increase.
NOAEL
LOAEL
LOAEL effects
500 ppm
1,500 ppm
Reduced body weight gain in males
and females
This document is a draft for review purposes only and does not constitute Agency policy.
C-165 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-40. Characteristics and quantitative results for Tomas et al. f!999a)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
WAG/Rij
Rats
M
6 rats/
dose
Oral (gavage, in olive
oil)
0, 2, 8, or 32 mmol/kg
body weight (240, 960, or
3,840 mg/kg body weight)
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 arousal by an
electrocardiogram before and after oral administration (in olive oil) of 0, 0.002, 0.008, or 0.032 mol/kg
body weight 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.
Changes in total duration of high-voltage spindle episodes following acute exposure to toluene and 1,2,3-,
1,2,4-, or 1,3,5-TMB at doses of 0.002, 0.008, and 0.032 mol/kg.
450-
I500
!«¦
TOUIEttE
0.002 BWiftl
V *
^
10.901 nttslkf
0.011 iiq*8
S, S, S S S, S, S, S, S, S, S, S,
S5 S, % S,
45#-
Jpoo-
1150 -
I.
450-
re'
1156.
0
450-
l
-11S0 •
HEMMELUTENE
-> ^,iTl
I 1
S^f * *
PSEUOOCUKNE
ri!
MESiTYLEHE
I * ! T
_]1L
S,-prstojaet»n- m
S,-» mm p38ttii|itcfofl - O
S,-« mm pwtiijtction • O
S, &0 mm postinjectioa - O
- comp»r» to oil group
- pcfl.081 compart to cental
Source: Reproduced from Tomas et al. (1999a).
This document is a draft for review purposes only and does not constitute Agency policy.
C-166 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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
1 20
oil
TOLUENE
0.002 mol/kg
II *
i 'I .JL, r^"l
0,008 mol/kg 0.032 mol/kg
* * *
* t *
40-
1 20
JS
£
S, S, S2 S, S, S, S; S, S, S, S2 S3 S, S, S: S,
HEMIMELUTENE
* *
i * ;
lili
40 ¦
1 20
S, s, S, Ss s, s, s, s, s, s, s2 s,
PSEUDOCUMENE
n
T * Fi
m *
m
£2
m
40'
•5
1
E
20
S, S, Sj S: 3, S, S: S3 S, S, Ss S,
MESITYLENE
* *
rM
» * *
s, s, s, s, s, s, sa s3 s, s, s, s,
Sa - preinjection - ES3
S, - 20 min postinjection - O
S;-40 min postinjection - a
Sj-60 min postinjection - ~
Source: Reproduced from Tomas et al. (1999a).
* - 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 were 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-167 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-41. Characteristics and quantitative results for Tomas etal. (1999b)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
WAG/Rij rats
M
10 rats/dose
Oral (in olive oil)
0, 8,16, or 32 mmol/kg
body weight (960,1,920,
or 3,850 mg/kg body
weight) 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 body weight of all isomers.
• All three TMB isomers were found to cause a slight increase in locomotor activity.
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.
i£
0.008 mol/kg
- pcO.OOQt compare to timo point 1, 2
injection
Time points
i
I
y
a
£
0.01 S mol/kg
- p<0.0001 compare to timer point 1. 2, 3
injection
I T Ja
Tim® points
0.032 mol/kg
- p<0.000t compare to time point 1
B'i -I'l i l 11. 1 ill
injection T
! ill i...
0 12 3
Time points
control group (oil) pseudocwmene mm hamimelliten» fret mesMtytane r~i toluene ¦
Source: Reproduced from Tomas et al. (1999b).
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 were 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-168 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-42. 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/
dose
i.p. injection
6.6 mmol/kg body weight
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 body weight 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.
Amplitude abnormalities of the cortical N1 wave 30 and 60 min after i.p. solvent injection.
15
110
MESITYLENE
JL
PSEUDOCUMENE HEMIMELLITENE
%
10
Reproduced from Tomas et al. (1999c).
TOLUENE
30 min
~ 80 min
Source:
Amplitude abnormalities of the cortical Pl-Nl wave 30 and 60 min after i.p. solvent injection.
TOLUENE MESITYLENE PSEUDOCUMENE HEMIMELLITENE
T
12
10
%
I 6
¦ 30 min
~ 60 min
W
Source: Reproduced from Tomas et al. (1999c).
This document is a draft for review purposes only and does not constitute Agency policy.
C-169 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Amplitude abnormalities of the hippocampal N1 wave 30 and 60 min after i.p. solvent injection.
TOLUENE MESITYLENE PSEUDOCUiENE HEMIMELLITENE
%
~ 60 mm
Source: Reproduced from Tomas et al. (1999c).
The effect of i.p. solvent injection on the cortical EEG in the 13-20.75 Hz frequency band.
TOLUENE
MESITYLENE
PSEUDOCUMENE HEMIMELLITENE
On
5 "
10™
15-
20-
25-
30 -
35-
40-
45-
50.
T
~ 30 min
~ 60 min
T
Source: Reproduced from Tomas et al. (1999c).
This document is a draft for review purposes only and does not constitute Agency policy.
C-170 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
The effect of i.p. solvent injection on the hippocampal EEG in the 1-3.75 Hz frequency band.
80 -
50 -
40
30
20
10
% 0
TOLUENE
~ 30 min
~ 60 min
IESITYLENE PSEUDOCUMENE HEilMELLITENE
J I
Source: Reproduced from Tomas et al. (1999c).
The effect of i.p. solvent injection on the hippocampal EEG in the 7-9.75 Hz frequency band.
TOLUENE
MESITYLENE PSEUDOCUMENE HEilMELLITENE
10 H
20
301
40
50 H
80
T
~ 30 min
o 60 min
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 was used.
Exposures were of an acute duration, and were 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-171 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-43. Characteristics and quantitative results for Wiaderna etal. (1998)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
13 or
14 rats/
dose
Inhalation (6 hrs/d,
5 d/wk)
0 or 25,100, or 250 ppm
(0, 123, 492, or
1,230 mg/m3) 1,2,3-TMB
4 wks
Additional study details
• Animals were exposed to 1,2,3-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/d,
5 d/wk for 4 wks. Food and water were 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 d 14-18 following exposure.
• Neurobehavioral effects were observed at 25 and 100 ppm (123 and 492 mg/m3) concentrations, but
not at 250 ppm (1,230 mg/m3).
Radial maze performance of rats exposed for 4 wks to 1,2,3-TMB.
day 1
day 4
LZj "
-------
Supplem en tal Information—Trim ethylbenzenes
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.
Locomotion
Exploration
Groomm
tYci fiu»p ?fy?r.
PS3 ptanp HM'Of?
I i* J, UMi^O
The test was performed 25 d after a 4-wk exposure to 1,2,3-TMB. Denotation of groups as in previous figure
above. The bars represent group means and SE.
Diagrams illustrating the effect of a 4-wk exposure to 1,2,3-TMB on the step-down passive avoidance learning.
140
120
100
£
o
x?
f
trio'
sock)
HMO HM2S HM1Q0 HM2S0
in rats.
The test was performed on d 39-48 after exposure. Trials 1, 2, and 3 were performed at 24-hr intervals. The
step-down response was punished by a 10-sec footshock in trial 3 only. Trials 4, 5, and 6 were performed 24 hr,
3 d, and 7 d after trial 3, respectively. The maximum step-down latency was 180 sec. Denotations of groups as in
previous figures above. The bars represent group means and SE.
*, ***p < 0.05 and p < 0.001, compared with respective data from control group.
This document is a draft for review purposes only and does not constitute Agency policy.
C-173 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Hot plate behavior tested in rats on d 50 (trials 1 and 2) and d 51 (trial 3) after a 4-wk exposure to 1,2,3-TMB.
4ft
/»o
to
5
O
,5.^
HMtOO HM2SO
IH =1
_ J-r.c
h
K/1
';V
Kk
•
J,
f b
s S*
^ r1
*
Oj ft
* N
1
<1
A ^
j
1 t.
V * 1 1 V !
Denotation of groups as in previous figures above. The bars represent group means and SE. Upper diagram: A
comparison of the latency of the paw-lick response to a thermal stimulus (54.5°C) on d 50. Ll-paw-lick latency in
trial 1 performed before a 2 min intermittent footshock. L2-paw-lick latency in trial 2 performed several sec after
the footshock. L3-paw-lick latency in trial 3 performed 24 hr after the footshock.
*p < 0.05 compared to L2/L1 of the same group.
Active avoidance learning and retention in rats after a 4-wk exposure to 1,2,3-TMB.
Trmnmg
50
4 0
HO
10
0
50
! «
t so
\ lO
o
too
BO
00
4G
20
0
^ c~:
a
[st A
' y*
Ret ra i ning
r*" T1
y iy
ItUO HM2& HUISQ
Reteniioxj
r
tekZL.
HM21A
's/S
Huw> nu 100 nuiebu
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 previous figures above. The bars represent group means and SE.
*p < 0.05 compared to control group.
This document is a draft for review purposes only and does not constitute Agency policy.
C-174 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Health effect at LOAEL
NOAEL
LOAEL
Impaired learning of passive
avoidance
N/A
25 ppm (123 mg/m3)
Comments: CNS disturbances were observed up to 2 mo 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 was only 4 wks. Generally, short-term exposure studies
have limited utility in quantitation of human health reference values.
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-175 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-44. Characteristics and quantitative results for Wiaderna etal. (2002)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
LOD: Wistar
rats
MM
12 rats/
dose
Inhalation (6 hrs/d,
5 d/wk)
0 or 25,100, or 250 ppm
(0, 123, 492, or
1,230 mg/m3) 1,2,3-TMB
4 wks
Additional study details
• Animals were exposed to 1,3,5-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/d,
5 d/wk for 4 wks. Food and water were 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.
Passive avoidance; the comparison of the time of staying on the platform in the consecutive test trials.
~ p< 0,001 compared to MESQ
~ ma 1
MESO
MES25
MES1Q0
MES250
The test was performed between d 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 d after trial 3, respectively. The bars represent group means and SE.
This document is a draft for review purposes only and does not constitute Agency policy.
C-176 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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.
* p < 0.02 compared to MESO and MES25 in the same trial
60
50'
10
1
~L1
~L2
~ L3
MESO MES25 MES1Q0 MES250
The test was performed on d 50 and 51 after the exposure. The bars represent group means and SE.
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 (four shock avoidances) in five consecutive trials (attempts) during the
training session.
~ p < 0.02 compered to MESO
+ *
_L
43 20
MESO
ME525
ME810Q
ME8250
The test was performed on d 54 (training) and d 60 (retraining) after the exposure. The bars represent group
means and SE.
Health effect at LOAEL
NOAEL
LOAEL
Shorter retention of passive
avoidance reaction
N/A
25 ppm (123 mg/m3
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 mo 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 wks. Generally, short-term exposure
studies have limited utility in quantitation of human health reference values.
This document is a draft for review purposes only and does not constitute Agency policy.
C-177 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-45. Characteristics and quantitative results for Wiglusz et al. (1975h)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
5-8/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/d,
6 d/wk for 5 wks
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/d, 6 d/wk,
for 5 wks.
• 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 d before exposure then on d 1, 7,14, and 28.
Observation
1,3,5-TMB exposure concentration (mg/L)—hematological parameters following
single 6-hr exposure
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.
C-178 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 ±4.13
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
0.5 ±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.
C-179 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Percentage of segmented neutrophilic granulocytes after 6 hrs exposure to 1,3,5-TMB.
S?
J"
4?
50
%
e *>
io
• control
¦ me$liglene f.Omgfl
mesitylene 3.0 mg/l
mesitylc-ne. 6.0 rngfl
0 1
14 days after exposure
__
¦28 »
Percentage of segmented neutrophilic granulocytes during exposure to 1,3,5-TMB 3.0 mg/L for 6 hrs/d, 6 d/wk,
for 5 wks.
a
s-
I
«?
•««*
•f. if
I
3 to
*
o
control
mesitytene 3.0mq/l
PI i b i | |
O i
11 n i ¦ ¦ i ¦ b ¦
« t*
¦ II ill
days of exposure 1
Observation
Hematological parameters during 5-wk 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 mm3 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 mm3 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
This document is a draft for review purposes only and does not constitute Agency policy.
C-180 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
% 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-TMBgroup
14.0 ± 5.0
17.0 ±9.4
16.6 ±5.0
21.5 ±7.4
18.4 ±8.6
% 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-TMBgroup
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-TMBgroup
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-TMBgroup
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-TMBgroup
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: Slight increases in percent segmented neutrophilic granulocytes on d 14 of the short-term exposure
study. Authors do not report statistical significance of results. Only one dose group was used in chronic study.
This document is a draft for review purposes only and does not constitute Agency policy.
C-181 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-46. 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/d,
6 d/wk for 5 wks
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/d, 6 d/wk,
for 5 wks.
• 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 d before exposure then on d 1, 7,14, and 28.
Observation
1,3,5-TMB exposure concentration (mg/L)—hematological parameters following
single 6-hr exposure (means ± SE)
0
0.3
1.5
3.0
AST 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
ALT 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
308 ± 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
AP 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
-
^Statistically significant in relation to initial values (p < 0.05).
This document is a draft for review purposes only and does not constitute Agency policy.
C-182 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Serum activity of AST after 6 hrs of exposure to 1,3,5-TMB; values are expressed in % of initial values.
em trot
— — me$itylerte. 0.3 mg}l
mesitylene 1.5 mqlt
mesitylene 3.0 mgjt
m
no
H 4ays after exposure
Serum activity of ALT after 6 hrs of exposure to 1,3,5-TMB; values are expressed in % of initial values.
— 1- control
„—. mesitt/iene 0,3 mg/i
mesityle/ts. i.S my/£
mesitylene 3.0 mgfi
2 no
2
.ts
I' 1 11 I I
Jtf days after exposure £81
This document is a draft for review purposes only and does not constitute Agency policy.
C-183 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Serum activity of AP after 6 hrs of exposure to 1,3,5-TMB; values are expressed in % of initial values.
*5
S.
e
C;
£40
iZO
mo
¦80
_ •*
o ¦
control
mesitytene O.i mglt
mesiiylene 1,0 mgfi
mesitylene J.o mg(L
4-
0 2
H days after exposure £g~
Observation
Hematological parameters during 5-wk exposure to 1,3,5-TMB (means ± SD)
Day 0
Day 1
Day 3
Day 7
Day 14
Day 28
AST 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
ALT 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
AP 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
^Statistically significant in relation to initial values (p < 0.05).
This document is a draft for review purposes only and does not constitute Agency policy.
C-184 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Serum activity of AST during exposure to 1,3,5-TMB at 3.0 mg/Lfor 6 hrs/d, 6 d/wk, for 5 wks; values are
expressed in % of initial values.
c&n ir-Ql
j» y M f »
M a'ayr sf e>xposisrm
Serum activity of ALT during exposure to 1,3,5-TMB at 3.0 mg/Lfor 6 hrs/d, 6 d per wk, for 5 wks; values are
expressed in % of initial values.
** days of exposure
23
Serum activity of AP during exposure to 1,3,5-TMB at 3.0 mg/Lfor 6 hrs/d, 6 d/wk, for 5 wks; values are
expressed in % of initial values.
rtc
,o
** of exposure
This document is a draft for review purposes only and does not constitute Agency policy.
C-185 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Health effect at LOAEL
NOAEL
LOAEL
Increase in AP activity
1.5 mg/L
3.0 mg/L
Comments: This study observed increases in AP activity on d 7 of the short-term exposure study. Only one dose
group was used in chronic study. Data were not recorded daily; significant gaps exist between sampling days.
1
2
This document is a draft for review purposes only and does not constitute Agency policy.
C-186 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
C.5. HUMAN TOXICOKINETIC STUDIES
1 Tables C-47 through C-52 provide study details for human toxicokinetic studies.
2 Table C-47. Characteristics and quantitative results for larnberg et al. (1996)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Caucasian
humans
M
9/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 of exposure, followed by
4 hrs of 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 exposures.
• 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-hr inhalation exposure
mean ± 95% CI)
Exposure
25 ppm
(123 mg/m3)
1,2,3-TMB
25 ppm
(123 mg/m3)
1,3,5-TMB
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
Kinetic values of TMB isomers following 2-hr inhalation exposure (mean ± 95% CI)
Kinetic parameter
25 ppm
(123 mg/m3)
1,2,3-TMB
25 ppm
(123 mg/m3)
1,3,5-TMB
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)g
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-187 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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(nMxhrs)
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
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.
ePost-exposure, percentage of respiratory uptake.
'Calculated from respiratory uptake.
Calculated from net respiratory uptake.
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).
b
124TMB in blood
(MM)
9
8
7
6
5
4
3
2
1
0
60
0
1 20
180
240
300
360
Time (tnln)
This document is a draft for review purposes only and does not constitute Agency policy.
C-188 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Concentration of 1,3,5-TMB in capillary blood during and after 2-hr exposure to 25 ppm (123 mg/m3) 1,3,5-TMB
(mean values ± 95% CI).
13STMB In blood
i *
120 1 BO 240
Time (min)
300 360
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
(HM)
9
120 180 240 300 360
Time (mint
This document is a draft for review purposes only and does not constitute Agency policy.
C-189 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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
(t#M)
101
- 25 ppm
- 2 ppm
0.01
360
0
60
240
120
180
Time (min)
Comments: Exposure duration possibly not sufficient to detect metabolic changes. Metabolites were not
measured.
This document is a draft for review purposes only and does not constitute Agency policy.
C-190 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-48. Characteristics and quantitative results for larnherg et al. (1997a)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Caucasian
humans
M
9
Inhalation
11 mg/m31,2,4-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 white spirit.
• Exposure lasted 2 hrs, during which time, 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.
• High performance liquid chromatography 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.
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 white spirit.
1,2,4-TMB in blood (iiM)
1,0i
0,8
0,6-
0.4-
0,2-
0,0
Exposure to white spirit
Exposure to 1,2,4-TMB
exposure
—i ¦ —i 1
60 120 180
Time (mill)
This document is a draft for review purposes only and does not constitute Agency policy.
C-191 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Results from 2-hr exposure to 1,2,4-TMB alone or 1,2,4-TMB as a component of white spirit (mean ± SD)
Exposure
1,2,4-TMB alone
1,2,4-TMB in white spirit
p-value
Net respiratory uptake (mmol)
0.15 ±0.01
0.14 ±0.02
0.5a
AUC (nM x min), 0-3 hrs
53 ±4
86 ±9
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-49. Characteristics and quantitative results for larnherg et al. (1997h)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Caucasian
humans
M
10
Inhalation
25 ppm (123 mg/m3)
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 exposures.
• Isomers of all DMHA metabolites in urine were detected via high performance liquid chromatography.
• 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 (hrs)
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-193 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-50. Characteristics and quantitative results for larnherg 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 white spirit,
25 ppm (123 mg/m3)
1,2,4-TMB
2 hrs of exposure, followed by
6 hrs of observation
Additional study details
• Caucasian males were exposed to 2 ppm (~ 10 mg/m3) 1,2,4-TMB, 2 ppm (~10 mg/m3) in white spirit,
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 white spirit
exposure group suggesting that components of white spirit 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-hr inhalation exposure to 1,2,4-TMB and 1,2,4-TMB in white spirit—mean values
(95% CI)
Kinetic parameter
2 ppm (~10 mg/m3)
group
2 ppm (~10 mg/m3)
in white spirit
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)
AUCbiood (nM x min)
95 (54-137)
157 (136-178)*
1,286 (1,131-1,441)
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)
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)
*p < 0.05, **p < 0.01, compared to 2 ppm (~ 10 mg/m3) alone by repeated measures ANOVA.
aNet respiratory uptake + amount cleared by exhalation during exposure.
b% of net respiratory uptake.
cNot determined.
Comments: Multiple exposure concentrations were tested and multiple tissues were analyzed. Study of
1,2,4-TMB as a component of white spirit. Toxicokinetics of 1,2,3- and 1,3,5-TMB not studied.
This document is a draft for review purposes only and does not constitute Agency policy.
C-194 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-51. Characteristics and quantitative results for lones etal. (2006)
Study design
Species
Sex
Exposure route
Dose range
Exposure duration
Humans
M & F
2/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 cm2 gauze patch soaked with liquid 1,3,5-TMB and reported mild
itching, erythema, and edema where gauze contacted skin.
Authors conclude that urinary DMBA and breath TMB are suitable markers of TMB exposure, and that
repeated exposures during the work week can result in significant accumulation in tissues.
The study was approved by the Health and Safety Executive's Research Ethics Committee.
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/m3) 1,3,5-TMB.
0
100
Timt (hours)
This document is a draft for review purposes only and does not constitute Agency policy.
C-195 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Mean ± SD blood levels of 1,3,5-TMB during and after 4-hr exposure to 25 ppm (1,2,3-TMB mg/m3) 1,3,5-TMB.
1,40
Exposure Time
1.20
| 1.00
| 0.80
k 0.60
T3
J 0.40
£D
0.20
0.00
0 1 2 3 4 5 6
Time (hours)
Mean ± SD breath levels of 1,3,5-TMB during and after 4-hr exposure to 1,3,5-TMB.
1KU
"5 120
s
c
"Smif
M
I
T rre (ho ji:«
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 were not measured.
This document is a draft for review purposes only and does not constitute Agency policy.
C-196 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-52. Characteristics and quantitative results for Kostrzewski et al.
fl997)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Humans
M & F
5
Inhalation
Between 5 and 150 mg/m3
1,2,4-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
concentration
(|ig/dm3)
SD
Blood
concentration
(|ig/dm3)
SD
Blood
concentration
(Hg/dm3)
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.
C-197 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Excretion rate [V, mg/hr) of 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
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-198 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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
-
-
-
-
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-199 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-200 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
C.6. ANIMAL TOXICOKINETIC STUDIES
Tables C-53 through C-65 provide study details for animal toxicokinetic studies.
Table C-53. Characteristics and quantitative results for Dahl etal. (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 min/d for 5 consecutive d
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 400 mL/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 min during each 80 min exposure.
Concentrations were increased each day. Day 1-5 concentrations were 1,10,100,1,000, and
5,000 ppm 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 d). 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-201 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1
2
Table C-54. Characteristics and quantitative results for Eide and Zahlsen
(1996)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats
M
4/dose
Inhalation
0, 75, 150, 300, or 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 were given ad libitum except during exposure, and animal weight ranged between
200 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-hr 1,2,4-TMB inhalation exposure
Exposure
Blood
(|imol/kg)
Brain (|imol/kg)
Liver (|imol/kg)
Kidneys
(|imol/kg)
Fat (|imol/kg)
75 ppm (369 mg/m3)
14.1
23.6
53.4
53.4
516
150 ppm (738 mg/m3)
57.5
97.5
123.1
168.5
3,806
300 ppm (1,476 mg/m3)
115.5
220.9
256.3
282.4
12,930
450 ppm (2,214 mg/m3)
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-202 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-55. Characteristics and quantitative results for Huo etal. (1989)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
3 rats/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
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-203 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 hrs 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-TMP
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-Dimethylbenzoic
alcohol
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-Dimethylbenzoic
alcohol
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-Dimethylbenzoic
alcohol
-
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-DMBA
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-DMBA
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-DMBA
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-DMHA
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 metabolites
37.1 ± 11.4
49.5 ± 13.0
86.4 ± 23.0
30.4
27.2
59.1
32.4
89.5
58.4
TMP = trimethylphenol
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 hrs.
This document is a draft for review purposes only and does not constitute Agency policy.
C-204 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-56. Characteristics and quantitative results for Mikulski and Wiglusz
2 f!9751
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
9 rats/dose
Unspecified
1.2 g/kg body weight
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 d.
• 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 DMHAs.
Urinary excretion of glycine, glucuronic, and sulphuric acid conjugates of TMBs
Not treated
% of dose
mean ± SD)
Glycine conjugates
Glucuronides
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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-205 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-57 Characteristics and quantitative results for Swiercz et al. (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, or 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, or 1,230 mg/m3)
1,2,4-TMB in an inhalation chamber for 6 hrs.
• 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)
Blood during 6-hr exposure
25 ppm (123 mg/m3)
25 ±2
200 ± 10
100 ppm (492 mg/m3)
109 ± 10
228 ± 10
250 ppm (1,230 mg/m3)
262 ±21
190 ± 12
Blood after 6-hr exposure
25 ppm (123 mg/m3)
26 ±3
349 ±6
100 ppm (492 mg/m3)
101 ±3
333 ± 18
250 ppm (1,230 mg/m3)
238 ±9
336 ±5
Urine after 6-hr exposure
25 ppm (123 mg/m3)
27 ±3
355 ± 10
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-hr 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.
C-206 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Blood concentrations of 1,2,4-TMB following 6-hr 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
DMBA urine concentrations after 6-hr 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
Comments: Metabolites (DMBAs) measured in urine. Appropriate number of animals per dose group (N = 4).
Exposure duration possibly not sufficient to detect other metabolic changes.
This document is a draft for review purposes only and does not constitute Agency policy.
C-207 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-58. Characteristics and quantitative results for Swiercz et al. (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, or 1,230 mg/m3)
1,2,4-TMB
6 hrs or 4 wks
Additional study details
• Male Wistar rats were exposed to either 25,100, or 250 ppm (123, 492, or 1,230 mg/m3)
pseudocumene (1,2,4-TMB) in an inhalation chamber for either 6 hrs or 4 wks.
• 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 wks
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 wks
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-wk 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
Venous blood 1,2,4-TMB concentrations after 4-wk 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
This document is a draft for review purposes only and does not constitute Agency policy.
C-208 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 wks (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 wks (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 wks (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 wks (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 wks)
0.38 ±0.23
2.33 ± 1.24
21.95 ±3.81
Temporal cortex (4 wks)
0.25 ±0.07
2.03 ±0.66
15.71 ±3.54
Hippocampus (4 wks)
0.41 ±0.27
3.03 ± 0.48
12.44 ± 2.63*
Cerebellum (4 wks)
0.33 ±0.05
3.20 ± 0.40
10.85 ± 2.47*
*p < 0.05 in comparison to brainstem.
Comments: Adipose tissue was not examined for 1,2,4-TMB content. Metabolite concentration was not
measured. No control group.
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-209 DRAFT—DO NOT CITE OR QUOTE
-------
1
Supplem en tal Information—Trim ethylbenzenes
Table C-59. 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, or 1,230 mg/m3)
1,3,5-TMB
6 hrs or 4 wks
Additional study details
• Male Wistar rats were exposed to either 0, 25,100, or 250 ppm (123, 492, or 1,230 mg/m3) mesitylene
(1,3,5-TMB) in an inhalation chamber for either 6 hrs or 4 wks.
• 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
Control
0
246 ±9
homogenates after 6-hr
25 ppm (123 mg/m3)
25 ±2
254 ± 11
exposure
100 ppm (492 mg/m3)
97 ± 14
242 ± 14
250 ppm (1,230 mg/m3)
254 ± 20
249 ±7
Liver, lung, and kidney
Control
0
331 ± 17
homogenates after 4-wk
25 ppm (123 mg/m3)
23 ±2
311 ±26
exposure
100 ppm (492 mg/m3)
101 ±8
320 ± 38
250 ppm (1,230 mg/m3)
233 ± 16
328 ±21
Blood collected after 6-hr
Control
0
251 ±7
exposure
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-wk
Control
0
310 ±9
exposure
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
Control
0
280 ±9
exposure
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-wk
Control
0
310 ± 10
exposure
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.
C-210 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
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 (|ig/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
This document is a draft for review purposes only and does not constitute Agency policy.
C-211 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
5
ND
0.38 ±0.14
2.31 ±0.67
6
ND
0.20 ± 0.04
0.76 ±0.06
Venous blood 1,3,5-TMB concentration following 4-wk 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
*p < 0.05 in comparison to brainstem.
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.
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-212 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-60. Characteristics and quantitative results for Swiercz et al. (2016)
Study design
Species
Sex
N
Exposure
Route
Dose Range
Exposure Duration
Wistar rats
Male
5 rats /dose
group
Inhalation
0, 25, 100, 250 ppm
(0-1,230 mg/m3
hemimellitene) (1,2,3-TMB)
6 hrs (single exposure) or
4 wks (6 hrs/d, 5 d/wk)
Additional study details:
• Rats were exposed to hemimellitene (1,2,3-TMB) in an inhalation exposure duration for 6 hrs or 4 wks
(6 hrs/d, 5 d/wk).
• Rats were randomized into groups of five animals with body weights between 200 and 360 g.
• All rats survived inhalation exposure of hemimellitene.
• There weren't any statistically significant changes found in tissue masses or body mass during 4-wk
exposure compared with controls.
• Highest levels of hemimellitene were found in kidneys after single and repeated exposures.
• Significantly lower concentrations of hemimellitene were detected in the blood and tissues of animals
after repeated inhalation exposure which may point to reduced hemimellitene retention in the lungs
of rats.
Body mass of rats and air concentrations
Observation
Hemimellitene
target
concentration in
inhaled air
[ppm]
Hemimellitene
concentration in
inhaled air [ppm]
(mean ± SD)
Animals treated [N]
Body weight [g] (mean
± SD)
Liver, lung, and kidney homogenates
6-Hr exposure
Control
0
5
226 ±4
25
25 ±5
5
207 ±5
100
105 ± 10
5
215 ± 20
250
242 ± 10
5
205 ±5
4-Wk exposure
Control
0
5
309 ± 26
25
25 ±2
5
280 ± 17
100
97 ±7
5
323 ± 28
250
246 ± 16
5
310 ± 13
Blood
6-Hr exposure
Control
0
5
210 ±7
25
28 ±2
5
223 ± 10
100
110 ±9
5
214 ± 11
250
234 ± 26
5
208 ±5
4-Wk exposure
Control
0
5
311 ± 10
25
24 ±3
5
333 ± 23
100
104 ±6
5
321 ±22
250
243 ± 13
5
292 ± 20
This document is a draft for review purposes only and does not constitute Agency policy.
C-213 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Urine
6-Hr exposure
Control
0
5
250 ±9
25
21 ± 1
5
243 ± 10
100
99 ±3
5
251 ± 15
250
225 ± 13
5
238 ± 14
4-Wk exposure
Control
0
5
310 ± 10
25
25± 2
5
305 ± 15
100
97 ±7
5
317 ± 22
250
246 ± 16
5
284 ± 23
Absolute and relative weight of liver, lung, and kidney
Hemimellitene target concentration in inhaled air (ppm)
6-Hr exposure
Observation
Control 0
25
100
250
Absolute organ weight (mean ± SD)
Liver
9.48 ±0.63
9.25 ±0.46
13.37 ±2.37
13.15 ± 1.12
Lung
1.31 ±0.13
1.17 ±0.30
1.34 ±0.29
1.21 ±0.20
Kidney
1.83 ±0.19
1.93 ±0.15
1.82 ±0.11
1.87 ±0.16
Relative organ weight (g/100 g body weight; mean ± SD)
Liver
4.50 ±0.41
4.47 ±0.26
4.27 ±0.72
4.57 ±0.35
Lung
0.62 ±0.08
0.57 ±0.14
0.63 ±0.17
0.59 ±0.09
Kidney
0.87 ±0.10
0.93 ±0.07
0.85 ± 0.04
0.91 ±0.08
4-Hr exposure
Absolute organ weight (mean ± SD)
Liver
12.63 ± 1.02
11.61 ± 1.62
13.37 ±2.37
13.15 ± 1.12
Lung
1.47 ± 0.24
1.63 ±0.32
1.54 ±0.33
1.43 ±0.33
Kidney
2.28 ±0.19
2.07 ± 0.08
2.51 ±0.32
2.49 ±0.17
Relative organ weight (g/100 g body weight; mean ± SD)
Liver
4.09 ±0.27
4.14 ±0.50
4.11 ±0.42
4.24 ±0.31
Lung
0.47 ± 0.06
0.58 ±0.10
0.48 ± 0.09
0.46 ± 0.09
Kidney
0.74 ± 0.08
0.74 ±0.01
0.77 ± 0.04
0.80 ± 0.05
Concentration of hemimellitene in liver, lung, and kidney homogenates and venous blood
6-Hr exposure
Hemimellitene target concentration in inhaled air (ppm)
25
100
250
Hemimellitene concentration (mean ± SD):
Liver (ng/g tissue)
1.66 ±0.48
4.20 ±0.85
20.75 ± 3.30
Lung (ng/g tissue)
0.62 ±0.08
2.57 ±0.40
18.73 ±2.81
Kidney (ng/g tissue)
2.81 ±0.40
7.78 ±3.17
31.16 ±3.84
Blood (ng/mL)
0.76 ±0.09
3.82 ±0.94
10.73 ± 1.30
This document is a draft for review purposes only and does not constitute Agency policy.
C-214 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
4-Hr exposure
Hemimellitene concentration (mean ± SD):
Liver (ng/g tissue)
1.18 ±0.28
2.68 ±0.76*
11.30 ±3.42**
Lung (ng/g tissue)
0.83 ±0.11**
2.17 ±0.24
17.28 ± 6.02
Kidney (ng/g tissue)
4.55 ±0.32***
10.07 ± 0.67
29.99 ± 8.00
Blood (ng/mL)
0.58 ±0.08**
3.14 ±0.61
6.87 ± 1.05***
*p < 0.05; significantly different from the single exposure.
**p < 0.01; significantly different from the single exposure.
***p < 0.001; significantly different from the single exposure.
Statistics of hemimellitene concentration in liver, lung, kidney homogenates and venous blood
Statistics
p-value
Liver
Lung
Kidney
Blood
Main effects
Exposure
<0.001
n.s
n.s.
<0.001
Concentration
<0.001
<0.001
<0.001
<0.001
Interaction effects
Exposure x cone.
<0.001
n.s.
n.s.
<0.001
Simple effects
Concentration within 6-hr
exposure
<0.001
<0.001
<0.001
<0.001
Concentration within 6-hr
exposure
n.s
<0.001
<0.010
<0.050
Venous blood hemimellitene concentrations
Hemimellitene concentration (|ig/mL) (mean ± SD)
Time
25 ppm
100 ppm
250 ppm
6-Hr exposure
0(3)
0.76 ±0.09
3.82 ±0.94
10.73 ± 1.30
0(15)
0.75 ±0.08
3.21 ±0.91
9.56 ± 1.40
0 (30)
0.67 ±0.14
2.83 ±0.35
7.09 ± 1.70
0(45)
0.52 ±0.14
2.76 ±0.47
6.73 ± 1.16
1(0)
0.50 ± 0.03
2.29 ±0.34
7.71 ±0.58
2(0)
0.45 ±0.15
1.63 ±0.16
5.10 ±0.62
3(0)
0.26 ±0.06
1.32 ±0.23
3.50 ±0.71
4(0)
0.18 ±0.08
0.87 ± 0.03
3.13 ±0.45
5(0)
0.12 ±0.10
0.55 ±0.10
1.51 ±0.39
6(0)
0.07 ± 0.05
0.48 ±0.14
1.25 ±0.30
4-Wk exposure
0(3)
0.58 ±0.09
3.14 ±0.70
6.87 ± 1.05
0(15)
0.40 ± 0.07
2.77 ±0.50
6.04 ± 0.80
0 (30)
0.42 ±0.10
2.03 ±0.15
4.56 ±0.73
0(45)
0.43 ±0.10
1.78 ±0.18
4.02 ±0.91
1(0)
0.43 ±0.13
1.80 ± 0.24
3.45 ±0.74
This document is a draft for review purposes only and does not constitute Agency policy.
C-215 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
2(0)
0.30 ± 0.06
1.38 ±0.30
3.04 ±0.32
3(0)
0.30 ± 0.03
1.03 ±0.15
2.43 ±0.37
4(0)
0.25 ±0.03
0.85 ±0.10
2.04 ±0.67
5(0)
0.19 ±0.06
0.82 ±0.16
1.66 ±0.36
6(0)
0.18 ±0.07
0.75 ±0.21
1.56 ±0.37
Toxicokinetics of hemimellitene elimination from blood
6-Hr inhalation exposure (ppm)
25
100
250
Elimination (E) equation
E = 0.60e"3 04t+0.52e("a23t)
E = 3.05e"Z23t+ 2.00e"°19t
E = 9.00e"1-28t+4.00e"°13t
AUC (mgh/L)
1.89
8.53
23.70
Half-life:
Phase 1 (min)
14
19
32
Phase II [h(min)]
3(4)
3(42)
5(20)
4 Wk exposure
Elimination (E) equation
E = 0.58e"2335t+0.40e"012t
E = 2.70e"5 09t+ 1.80e"015t+
E = 7.00e"3 24t+ 3.00e"°°9t
AUC (mgh/L)
1.75
7.66
16.09
Half-life:
Phase 1 (min)
2
8
13
Phase II [h(min)]
5(52)
4(34)
7(58)
Concentration of 2, 3-D MBA after exposure to hemimellitene
6-Hr exposure
Hemimellitene target concentration in inhaled air (ppm)
25
100
250
2,3-DMBA concentration (ng/g tissue) (mean ± SD)
Liver
7.68 ± 1.64
21.19 ±0.59
27.66 ±3.62
Lung
n.d.
n.d.
3.23 ±0.56
Kidney
5.52 ±0.77
23.59 ±3.33
28.69 ±6.55
4-Wk exposure
Liver
8.54 ± 1.17
13.78 ± 2.84
17.93 ±4.33
Lung
n.d.
n.d.
2.82 ±0.44
Kidney
6.84 ±0.76
11.19 ± 1.58
18.53 ±2.31
n.d. = not detected
Statistics of 2,3-DMBA concentration in liver, lung, and kidney of rats after exposure of hemimellitene
Statistics
p-value
Liver
Lung
Kidney
Main effects
Exposure
<0.001
-
<0.001
Concentration
<0.001
-
<0.001
Interaction effects
Exposure by concentration
<0.001
-
<0.001
Simple effects
This document is a draft for review purposes only and does not constitute Agency policy.
C-216 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Cone. w/L 6-hr exposure
<0.001
-
<0.001
Cone. w/L 4-hr exposure
n.s.
-
n.s.
Exposure w/L concentration
25 ppm
n.s.
-
n.s.
100 ppm
n.s.
-
<0.050
250 ppm
<0.050
-
n.s.
Urinary excretion after exposure to hemimellitene
6-Hr exposure
Hemimellitene target concentration in inhaled air (ppm)
25
100
250
Urine (mg/18 hrs) (mean ± SD):
2,6-DMBA
n.d.
0.17 ±0.03
0.59 ±0.26
2,3-DMBA
0.07 ±0.01
0.58 ±0.06
2.19 ±0.66
4-Wk exposure
Urine (mg/18 hrs) (mean ± SD):
2,6-DMBA
n.d.
0.39 ±0.13
0.58 ±0.14
2,3-DMBA
0.11 ±0.005
1.60 ± 0.40
2.79 ±0.76
Statistics of urinary excretion of DMBA isomers after exposure
Statistics
p-value
2,6-DMBA
2,3-DMBA
Main effects
Exposure
n.s.
<0.005
Concentration
<0.001
<0.001
Interaction effects
Exposure by concentration
n.s.
n.s.
Simple effects
Cone. w/L 6-hr exposure
<0.050
<0.050
Cone. w/L 4-hr exposure
n.s.
<0.001
Exposure w/L concentration
25 ppm
-
n.s.
100 ppm
n.s.
n.s.
250 ppm
n.s.
n.s.
n.s. = not significantly significant
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-217 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Changes of TMB isomers in tissues and blood of rats after 6-hr versus 4-wk exposure to isomers of TMB
Changes of TMB isomers concentration (%)
25 ppm
100 ppm
250 ppm
Lung
Blood
Liver
Lung
Blood
Liver
Lung
Blood
Liver
TMB isomer:
Pseudocumene
104,
24^
584,
19 ^
34,
204,
Mesitylene
35^
0
274,
314,
254,
34,
354,
434,
244,
Hemimellitene
341^
244,
294,
29^
184,
364,
44,
364,
464,
Toxicokinetics of TMB isomers elimination from venous blood after 6-hr or 4-wk exposure to isomers of TMB
Toxicokinetics of TMB Isomers
25 ppm
100 ppm
250 ppm
6-Hr
4-Wk
6-Hr
4-Wk
6-Hr
4-Wk
Pseudocumene
AUCo->6h [mgh/L]
1.25
0.92
7.02
8.14
53.74
23.33
Half-life [h(min)]
Phase 1
0(10)
0(9)
0(28)
0 (32)
0(57)
1(8)
Phase II
3(51)
2(53)
5(20)
5(47)
17 (20)
9(54)
Mesitylene
AUCo->6h [mgXh/L]
0.33
0.40
5.72
4.84
32.46
15.67
Half-life [h(min)]
Phase 1
0(12)
0(23)
0(11)
0(8)
0(16)
0(10)
Phase II
2(40)
2 (23)
3(9)
4(37)
4(5)
4(37)
Hemimellitene
AUCo->6h [mgXh/L]
1.89
1.75
8.53
7.66
23.70
16.09
Half-life [h(min)]
Phase 1
0(14)
0(2)
0(19)
0(8)
0(32)
0(13)
Phase II
3(4)
5(52)
3(42)
4(34)
5(20)
7(58)
1
This document is a draft for review purposes only and does not constitute Agency policy.
C-218 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Changes of DMBA isomers in tissues and urine of rats after 6-hr versus 4-wk exposure to isomers of TMB
Changes of TMB isomers concentration (%)
25 ppm
100 ppm
250 ppm
Lung
Liver
Kidney
Urine
Lung
Liver
Kidney
Urine
Lung
Liver
Kidney
Urine
DMBA isomer:
Pseudocumene (%)
2,5-DMBA
n.d
n.d.
494/4/
624/4/
374/
344/
34^
464/4/
204/
174/
50^
101^
2,4-DMBA
214/
154/
614/4/
64/
264/
104/
19 ^
334/4/
224/
134/
39^
134/
3,4-DMBA
424/
47 4/4/
444/4/
344/
394/4/
434/4/
151^
334/4/
254/
434/4/
148^
20^
Mesitylene (%)
3,5-DMBA
291
484/4/
26^
601^
621^
174/4/
15 ^
19 ^
<-
00
44^^
35^
8^
Hemimellitene (%)
2,6-DMBA
n.d.
n.d
n.d.
n.d.
n.d.
n.d.
n.d.
129
n.d.
n.d.
n.d.
24^
2,3-DMBA
n.d.
11^
24^
571^
n.d.
LO
m
m
LO
176
134/
LO
m
LO
m
27^
1" = insignificant increase; 1^1" = significant increase; 4^ = insignificant decrease; 4/4/ = significant decrease
Mean body weights
¦* *
*l|-> 300 •
1
;
-
200 ¦
t
Df rats exposed to hemimellitene at 0 ppm (N = 5), 25 ppm (N = 5), 100 ppm (N = 5), and
250 ppm (N = 5) for 4 wks.
—' — —CD— • , s| —•— i
1
Source: Swiercz et a
I- (2
7 14 21 28
8f
016).
This document is a draft for review purposes only and does not constitute Agency policy.
C-219 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Kinetics of hemimellitene elimination from venous blood of rats after termination of 6-hr and 4-wk exposures to
hemimellitene vapors at nominal concentration of (a) 25 ppm (N = 5), (b) 100 ppm (N = 5), and (c) 250 ppm
(N = 5).
~
—Q"
This document is a draft for review purposes only and does not constitute Agency policy.
C-220 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-61. Characteristics and quantitative results for Tsuiimoto et al.
f20001
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sic Wistar
rats
M
4/dose
i.p. in corn oil
0, 0.3,1, and 3 mmol/kg
body weight 1,2,4-TMB
2d
Additional study details
• Groups of four male Wistar rats dosed with 0, 0.3,1, or 3 mmol/kg body weight 1,2,4-TMB.
• Urine samples collected for 2 d.
• High performance liquid chromatography 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 hrs following exposure. Authors do not report specific speciation data for 2,4-, 2,5-, or 3,4-dimethylbenzyl
mercapturic acid.
This document is a draft for review purposes only and does not constitute Agency policy.
C-221 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-62. Characteristics and quantitative results for Tsuiimoto et al.
2 f2005)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar
rats
M
4/dose
i.p. in corn oil
0, 0.3,1, and 3 mmol/kg
body weight given
1,2,3- or 1,3,5- TMB
2d
Additional study details
• Groups of four male Wistar rats were given 1,2,3- or 1,3,5-TMB i.p. in doses of 0, 0.3,1, or 3 mmol/kg
body weight.
• Urine samples collected for 2 d, then analyzed for TMPs via gas chromatography-mass spectrometry.
Urinary excretion (% of dose ± SD) of phenolic metabolites in 1,2,3-TMB treated rats
Dose
(mmol/kg)
2,3,4-TMP
3,4,5-TMP
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
ND = not detected.
Urinary excretion (% of dose ± SD) of phenolic metabolites in 1,3,5-TMB-treated rats
2,4,6-TMP
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 hrs following exposure.
This study does not include data for 1,2,4-TMB and phenolic metabolites. Variation between rats (high SD) within
exposure groups.
This document is a draft for review purposes only and does not constitute Agency policy.
C-222 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table C-63. Characteristics and quantitative results for Tsuiino etal. (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)
Dermal (via
saturated cotton)
1 mL kerosene
0,1, 3, or 6 hrs
Additional study details
• In the first experiment, rats were dermally exposed to kerosene on a saturated, sealed piece of cotton
for 1 hr to analyze TMB and aliphatic hydrocarbon dermal absorption.
• In the 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 aliphatic hydrocarbons, and were only detected in traces
following post-mortem exposure.
• Trace concentrations of TMBs following post-mortem exposure suggest that TMB must circulate in
blood before being distributed to organs.
1 -Hr exposure and ratio of TMBs to internal standard
o-xylene dio) (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
3.6 ± 1.6
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*
* p < 0.05.
**p< 0.01.
***p < 0.001.
This document is a draft for review purposes only and does not constitute Agency policy.
C-223 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1,2,4-TMB in various tissues following 1 hr of exposure and ante- versus post-mortem exposure.
O
o
¦O
13
o
CO
Q.
600
O)
"cp
CD
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-64. Characteristics and quantitative results for Zahlsen etal. (1990)
Study design
Species
Sex
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 d 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 d 1, 3, 7,10, and 14.
• Food and water were given ad libitum except during exposure, and animal weight ranged between
150 g and 200 g prior to exposure on d 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.
Blood concentrations (+SD) of n-nonane, 1,2,4-TMB, and 1,2,4-trimethylcyclohexane following 12-hr exposures
on d 1, 3, 7,10, and 14.
3, 600 -
Z 400
Q 300
g 100
#TMB
* n~C9
*TMCH
5 10
TIME OF EXPOSURE (days)
This document is a draft for review purposes only and does not constitute Agency policy.
C-225 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Brain concentrations (+SD) of n-nonane, 1,2,4-TMB, and 1,2,4-trimethylcyclohexane following 12-hr exposures
on d 1, 3, 7,10, and 14.
*S 1600 r
1 1400-
S 1200 -
c
£ 1000
1= 800
n-C9
-c
QC
600 -
400-,
LU
o 200
O
o
Jtmch
TMB
TIME OF EXPOSURE (days)
Perirenal fat concentrations (+SD) of n-nonane, 1,2,4-TMB, and 1,2,4-trimethylcyclohexane following 12-hr
exposures on d 1, 3, 7,10, and 14.
o>
2
70000r
60000
50000
40000
g 30000
f 20000
z
o 10000
o
o
TMB
n-C9
TMCH
TIME OF EXPOSURE (days)
Brain:blood and fat:blood TMB distribution after 12-hr exposure at end of d 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 d 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-226 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table C-65. 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 1,2,4-TMB
12 hrs/d for 3 d
Additional study details
• Food and water were given ad libitum, except during exposure.
• Rats weighed between 150 and 200 g and were between 40 and 50 d of age.
• Four rats were housed in each cage, and each exposure chamber contained four cages; 16 rats were
present at the beginning of exposure.
• At each time point, four rats were sacrificed and their tissues analyzed for 1,2,4-TMB presence.
Observation
1,2,4-TMB concentration in rat tissues at various time points (mean ± SD)
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
1,741 ± 329
Fat day 2
1,375 ± 88
Fat day 3
1,070 ± 93
Fat reca
120 ± 52
arec = after 12-hour recovery.
Comments: Data were collected immediately following exposure and 12 hrs following exposure, providing insight
into metabolic clearance and excretion. Study duration was short term (5 d), making it difficult to determine if
tissue concentration changes following chronic exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
C-227 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
C.7. ANIMAL AND HUMAN TOXICOKINETIC STUDIES
1 Table C-66 provides study details for an animal and human toxicokinetic study.
2 Table C-66. Characteristics and quantitative results for Meulenberg and
3 Viiverberg f2000)
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 VOCs from multiple studies.
• 1,2,3-, 1,2,4-, and 1,3,5-TMB were among the VOCs 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
P oil :air
10,900a
10,200a
9,880a
Psaline:air
2.73a
1.61a
1.23a
Reported and predicted Ptissue:air values for various human tissues
Blood
66.5a
59.1a
43a
Fat
4,879b
4,566
4,423
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
6,484
6,068
5,878
Brain
591
552
535
Liver
288
269
260
Muscle
111
104
100
Kidney
1,064
995
963
aAveraged values as reported bv Jarnberg and Johanson (1995).
bAII other values predicted bv Meulenberg and Viiverberg (2000).
Comments: 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.
This document is a draft for review purposes only and does not constitute Agency policy.
C-228 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
Supplem en tal Information—Trim ethylbenzenes
APPENDIX D. DOSE-RESPONSE MODELING FOR
THE DERIVATION OF REFERENCE VALUES FOR
EFFECTS OTHER THAN CANCER AND THE
DERIVATION OF CANCER RISK ESTIMATES
D.l. BENCHMARK DOSE (BMD) MODELING SUMMARY
This appendix provides technical detail on dose-response evaluation and determination of
points of departure (PODs) for relevant neurological, hematological, and developmental toxicity
endpoints in the trimethylbenzene (TMB) database. The endpoints were modeled using the U.S.
Environmental Protection Agency (EPA) Benchmark Dose Software (BMDS, version 2.6.0.1).
Sections D.l.1.1 and D.l.1.2 (noncancer) 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 (U.S. EPA. 20121. 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.
D.l.l. Noncancer Endpoints
D.l.1.1. Evaluation of Model Fit
For each continuous endpoint (see Table D-l), 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, with x2 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 benchmark response (BMR).
D.l.1.2. Model Selection
For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
estimated by the profile likelihood method) and Akaike Information Criterion (AIC) value were
This document is a draft for review purposes only and does not constitute Agency policy.
D-l DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 used to select a best-fit model from among the models exhibiting adequate fit If the BMDL
2 estimates were "sufficiently close," (i.e., differed by at most 3-fold), the model selected was the one
3 that yielded the lowest AIC value. If the BMDL estimates were not sufficiently close, the lowest
4 BMDL was selected as the POD.
5 Table D-l. Noncancer endpoints selected for dose-response modeling for
6 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB
Species (strain)/sex
endpoint
Internal doses, external exposure concentrations, and effect data
Korsak and Rvdzvriski (1996)
1,2,4-TMB
Rat (Wistar)/male
Concentration (mg/m3)
0
123
492
1,230
CNS: paw-lick (seconds)
Number of animals
Mean ± SD
9
15.4 ±5.8
10
18.2 ±5.7
9
27.6 ±3.2
10
30.1 ±7.9
1,2,3-TMB
Rat (Wistar)/male
Concentration (mg/m3)
0
123
492
1,230
CNS: paw-lick (seconds)
Number 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)-l,2,4-TMB
Rat (Wistar)/male
Concentration (mg/m3)
0
129
492
1,207
Decreased RBC (106/cm3)
Number 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)
Number 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/m3)
0
128
523
1,269
Decreased segmented
neutrophils (%)
Number 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
(%)
Number 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/m3)
0
128
523
1,269
Decreased segmented
neutrophils (%)
Number of animals
Mean ± SD
10
23.1 ±6.1
10
19.7 ±3.4
10
16.4 ±4.2
10
11.9 ±7.1
Saillenfait et al. (2005)
1,2,4-TMB
Rat (Sprague-Dawley), Fi
pups and dams
Concentration (mg/m3)
0
492
1,471
2,913
4,408
Male fetal weight (g)
Number of liters
Mean ± SDa
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 ± SDa
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
This document is a draft for review purposes only and does not constitute Agency policy.
D-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
Supplem en tal Information—Trim ethylbenzenes
Species (strain)/sex
endpoint
Internal doses, external exposure concentrations, and effect data
Maternal weight gain (g)
Number of dams
Mean ± SD
24
131 ± 33
22
124 ± 18
22
126 ± 24
22
116 ± 23
24
95 ± 19
1,3,5-TMB
Fi rat pups and dams
(Sprague-Dawley)
Concentration (mg/m3)
0
497
1,471
2,974
5,874
Male fetal weight (g)
Number of litters
Mean ± SDa
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 litters
Mean ± SDa
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
135 ± 15
22
138 ± 11
21
118 ± 24
17
95 ±24
18
73 ±28
a SD reported for fetal weights represent variability among reported litter means, not among fetuses. In any
subsequent BMD analyses of these endpoints, the BMDs and BMDLs estimated using 1 SD as the compariative
BMR corresponding to the SD among litter means.
CNS = central nervous system; RBC = red blood cell; SD = standard deviation
For all endpoints, BMD modeling was conducted using the reported external exposure
concentrations as the dose inputs, except when actual concentrations were not provided. In these
cases, the target concentrations were used. In cases where the poor model fit to the mean or
variance was evident due mainly to poor fit in the high dose, the high dose was dropped and the
truncated dataset was re-modeled. Comprehensive modeling results for all endpoints are provided
on EPA's Health Effects Research Online (HERO) database (U.S. EPA. 2016b).
D.l.1.3. Modeling Results
Tables D-2 to D-34 and Figures D-l to D-13 summarize the modeling results for the
noncancer endpoints modeled. The following continuous model parameter restrictions were
applied, unless otherwise noted: (1) polynomial model (B coefficients were restricted with respect
to the appropriate direction of effect (i.e., >0 for responses that increase with dose, and <0 for
responses that decrease with dose); and (2) Hill, power, and exponential power parameters were
restricted to be >1. A1 SD change in the control mean was used as the BMR for all endpoints except
decreased fetal weight, for which a 5% RD BMR was used. However, as recommended by EPA's
Benchmark Dose Technical Guidance fU.S. EPA. 20121. a BMR equal to a 1 SD change in the control
mean was presented for decreased fetal weight to facilitate comparisons across assessments.
This document is a draft for review purposes only and does not constitute Agency policy.
D-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Table D-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) (Korsak and
Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0115
184.29
674
531
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-3).
Exponential (M4)
0.376
178.14
161
84.0
Exponential (M5)
N/Ac
179.36
211
92.5
Hill
N/Ac
179.36
195
90.2
Powerd
Polynomial 3°e
Polynomial 2°f
Linear
0.0293
182.42
535
396
aConstant variance case presented (BMDS Test 2 p-value = 0.0765, BMDS Test 3 p-value = 0.0765); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
eFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
fFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-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
Supplem en tal Information—Trim ethylbenzenes
Table D-3. 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 (modeled variance) (Korsak and
Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0172
185.21
572
400
No model selected as Test 3
p-value was <0.1. The data were
remodeled after dropping the
high dose (see Table D-4)
Exponential (M4)
0.406
179.78
154
78.4
Exponential (M5)
N/Ac
181.09
202
85.6
Hill
N/Ac
181.09
189
82.9
Powerd
Polynomial 3°e
Polynomial 2°f
0.0500
183.08
425
274
Linear8
0.0500
183.08
425
274
aModeled variance case presented (BMDS Test 2 p-value = 0.0765, BMDS Test 3 p-value = 0.0371); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dThe Power model may appear equivalent to the Linear model; however, differences exist in digits not displayed in
the table.
eFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model.
the Polynomial 2° model may appear equivalent to the Linear model; however, differences exist in digits not
displayed in the table.
gThe Linear model may appear equivalent to the Power model; however, differences exist in digits not displayed in
the table. This also applies to the Polynomial 3° model. This also applies to the Polynomial 2° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Table D-4. 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)
fKorsak and Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.854
121.80
231
181
Of the models that provided an
adequate fit, the linear model
was selected, based on lowest
AIC (BMDLS differed by <3-fold)
Exponential (M4)
N/Ac
123.79
192
84.7
Power
N/Ac
123.77
204
141
Polynomial 2°
N/Ac
123.77
206
141
Linear
0.899
121.79
192
141
aConstant variance case presented (BMDS Test 2 p-value = 0.169), selected model in bold; scaled residuals for
selected model for doses 0,123, and 492 were 0.08, —0.1, and 0.03, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
Linear Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
30
25
20
15
10
Bfr/ID
0
100
200
300
400
500
12:13 08/25 2015
BMR = 1 SD change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-l. Plot of mean response by dose for increased latency to paw-lick in
male Wistar rats, with fitted curve for Linear model with constant variance
(Korsak and Rvdzynski. 1996).
This document is a draft for review purposes only and does not constitute Agency policy.
D-6 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1
2 Polynomial Model (Version: 2.20; Date: 10/22/2014)
3 The form of the response function is: Y[dose] = beta_0 + beta_l*dose
4 A constant variance model is fit
5
6 Benchmark Dose Computation
7 BMR = 1 Estimated SD from the control mean
8 BMD = 192.088
9 BMDL atthe 95% confidence level = 140.537
10
11 Parameter Estimates
Variable
Estimate
Default initial parameter values
alpha
22.9935
25.738
rho
N/A
0
beta_0
15.277
15.2846
beta_l
0.0249633
0.0249531
12
13 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
9
15.4
15.3
5.8
4.8
0.0769
123
10
18.2
18.3
5.7
4.8
-0.0973
492
9
27.6
27.6
3.2
4.8
0.0256
14
15 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
-57.884957
4
123.769915
A2
-56.10689
6
124.213781
A3
-57.884957
4
123.769915
Fitted
-57.89298
3
121.785961
R
-68.599682
2
141.199364
16
17 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
24.9856
4
<0.0001
Test 2
3.55613
2
0.169
Test 3
3.55613
2
0.169
Test 4
0.0160462
1
0.8992
This document is a draft for review purposes only and does not constitute Agency policy.
D-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Table D-5. 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) (Korsak and
Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.00570
262.21
701
566
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-6).
Exponential (M4)
0.546
254.24
192
107
Exponential (M5)
N/Ac
255.87
201
111
Hill
N/Ac
255.87
186
110
Powerd
Polynomial 3°e
Polynomial 2°f
Linear
0.0173
259.99
578
443
aConstant variance case presented (BMDS Test 2 p-value = 1.15 x 10"4, BMDS Test 3 p-value = 1.15 x 10"4); no
model was selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
eFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
fFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Table D-6. 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) (Korsak and
Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
<0.0001
259.53
497
329
No model selected as Test 3
p-value was <0.1. The data were
remodeled after dropping the
high dose (see Table D-7)
Exponential (M4)
0.301
241.42
86.2
46.7
Exponential (M5)
N/Ac
242.59
113
52.0
Hill
N/Ac
242.59
120
Errord
Power6
Polynomial 3°f
Polynomial 2°g
Linear
3.25 x
10"4
254.41
320
196
aModeled variance case presented (BMDS Test 2 p-value = 1.15 x 10"4, BMDS Test 3 p-value = 0.0708); no model
was selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
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
15
Supplem en tal Information—Trim ethylbenzenes
Table D-7. 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, high dose dropped)
fKorsak and Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.445
218.88
301
237
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-8).
Exponential (M4)
N/Ac
220.30
223
112
Exponential (M5)
Hill
Polynomial 3°
Error
Error
Error6
Error6
Powerf
Polynomial 2°g
Linear
0.645
218.51
266
196
aConstant variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = <0.0001); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eBMD or BMDL computation failed for this model
fFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-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
Supplem en tal Information—Trim ethylbenzenes
Table D-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 (modeled variance, high dose dropped)
fKorsak and Rvdzvnski. 19961
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0745
203.27
192
132
Of the models that provided an
adequate fit, the linear model
was selected, based on lowest
AIC (BMDLS differed by <3-fold)
Exponential (M4)
N/Ac
202.08
105
52.6
Powerd
Polynomial 20e
Linear
0.202
201.71
152
97.2
aModeled variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = 0.5008), selected model
in bold; scaled residuals for selected model for doses 0,123, and 492 were -0.1, 0.32, and -0.35, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
eFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
Linear Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
22
20
8
6
4
2
O
8
15:27 08/25 2015
BMR = 1 SD change from control mean; dose shown in mg/m31,2,3-TMB.
Figure D-2. Plot of mean response by dose for increased latency to paw-lick in
male Wistar rats, with fitted curve for Linear model with constant variance
(Korsak and Rvdzynski. 1996).
This document is a draft for review purposes only and does not constitute Agency policy.
D-ll DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Polynomial Model (Version: 2.20; Date: 10/22/2014)
The form of the response function is: Y[dose] = beta_0 + beta_l*dose
A modeled variance is fit
Benchmark Dose Computation
BMR = 1 Estimated SD from the control mean
BMD = 152.065
BMDL at the 95% confidence level = 97.1911
Parameter Estimates
Variable
Estimate
Default initial parameter values
lalpha
-7.3421
2.58956
rho
3.94293
0
beta_0
9.74214
9.90769
beta_l
0.0148851
0.0131332
Table of Data and Estimated Va
ues of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
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
Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
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
-116.95626
2
237.91252
Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
42.2818
4
<0.0001
Test 2
20.665
2
<0.0001
Test 3
0.453187
1
0.5008
Test 4
1.63087
1
0.2016
This document is a draft for review purposes only and does not constitute Agency policy.
D-12 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Table D-9. Summary of BMD modeling results for decreased RBCs 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
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.611
76.397
695
452
Of the models that provided an
adequate fit, the Exponential 2
model was selected, based on
lowest AIC (BMDLS differed by
<3-fold)
Exponential (M4)
0.530
77.805
477
178
Exponential (M5)
N/Ac
79.411
482
191
Hill
N/Ac
79.411
480
Errord
Power6
Linear'
0.540
76.642
752
516
Polynomial 3g
Polynomial 2h
0.540
76.642
752
516
aConstant variance case presented (BMDS Test 2 p-value = 0.433), selected model in bold; scaled residuals for
selected model for doses 0,129, 492, and 1,207 were 0.08, 0.41, -0.83, and 0.34, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model,
the Linear model may appear equivalent to the Polynomial 3° model; however, differences exist in digits not
displayed in the table. This also applies to the Polynomial 2° model.
gFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model.
hThe Polynomial 2° model may appear equivalent to the Power model; however, differences exist in digits not
displayed in the table. This also applies to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-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
Supplem en tal Information—Trim ethylbenzenes
Exponential 2 Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
Exponential 2
11
10
9
8
7
0
200
400
600
800
1000
1200
dose
12:36 08/25 2015
BMR = 1SD change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-3. Plot of mean response by dose for decreased RBCs in male Wistar
rats, with fitted curve for Exponential 2 model with constant variance (Korsak
and Rvdzvnski. 19961.
Exponential Model (Version: 1.10; Date: 01/12/2015)
The form of the response function is: Y[dose] = a * exp(sign * b * dose)
A constant variance model is fit
Benchmark Dose Computation
BMR = 1.0000 Estimated SDs from control
BMD = 695.431
BMDL atthe 95% confidence level = 451.511
Parameter Estimates
Variable
Estimate
Default initial parameter values
Inalpha
0.759919
0.735269
rho
N/A
0
a
9.94081
8.08952
b
0.000228786
0.000222126
c
N/A
0
d
N/A
1
This document is a draft for review purposes only and does not constitute Agency policy.
D-14 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
10
9.98
9.94
1.68
1.46
0.08476
129
10
9.84
9.65
1.82
1.46
0.4072
492
10
8.5
8.88
1.11
1.46
-0.8273
1,207
10
7.7
7.54
1.38
1.46
0.3414
2
3 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
-34.70537
5
79.41075
A2
-33.33353
8
82.66706
A3
-34.70537
5
79.41075
R
-41.88886
2
87.77771
2
-35.19837
3
76.39674
4
5 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
17.11
6
0.008885
Test 2
2.744
3
0.4329
Test 3
2.744
3
0.4329
Test 4
0.986
2
0.6108
6
This document is a draft for review purposes only and does not constitute Agency policy.
D-15 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Table D-10. 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 variance) (Korsak et al.. 2000a)
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0102
205.39
1,466
691
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-ll).
Exponential (M4)
0.300
199.29
111
0.531
Exponential (M5)
N/Ac
201.25
122
0.532
Hill
N/Ac
201.25
127
Errord
Power6
Polynomial 3°f
Polynomial 2°g
Linear
0.00852
205.74
1,585
835
aConstant variance case presented (BMDS Test 2 p-value = 0.0229, BMDS Test 3 p-value = 0.0229); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-16 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
Supplem en tal Information—Trim ethylbenzenes
Table D-ll. 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 (modeled variance) (Korsak et al.. 2000a)
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)b
5.75 x
10"4
206.81
1,962
721
No model was selected as the
only possibly fitting models
(Exponential models 4 and 5 and
the Hill model) returned
implausibly low BMDL values.
The data were remodeled after
dropping the high dose (see
Table D-12)
Exponential (M3)c
5.75 x
10"4
206.81
1,962
721
Exponential (M4)
0.0922
196.72
299
0.680
Exponential (M5)
N/Ad
198.72
201
0.590
Hill
N/Ad
198.72
164
2.56 x 10"6
Power6
4.95 x
10"4
207.11
2,046
875
Polynomial 3f
Polynomial 2g
Linearh
4.95 x
10"4
207.11
2,046
875
aModeled variance case presented (BMDS Test 2 p-value = 0.0229, BMDS Test 3 p-value = 0.200); no model was
selected as a best-fitting model.
bThe Exponential (M2) model may appear equivalent to the Exponential (M3) model; however, differences exist in
digits not displayed in the table.
cThe Exponential (M3) model may appear equivalent to the Exponential (M2) model; however, differences exist in
digits not displayed in the table.
dNo available degrees of freedom to calculate a goodness-of-fit value.
eThe Power model may appear equivalent to the Polynomial 3° model; however, differences exist in digits not
displayed in the table. This also applies to the Polynomial 2° model. This also applies to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
hThe Linear model may appear equivalent to the Power model; however, differences exist in digits not displayed in
the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-17 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Table D-12. 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 variance, high dose dropped) (Korsak
etal.. 2000al
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.167
150.26
294
171
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-13).
Exponential (M4)
N/Ac
150.34
114
0.484
Exponential (M5)
Hill
Polynomial 3°
Error
Error
Errord
Errord
Power6
Linear'
0.123
150.73
340
222
Polynomial 2g
0.123
150.73
340
222
aConstant variance case presented (BMDS Test 2 p-value = 0.00849, BMDS Test 3 p-value = 0.00849); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model,
the Linear model may appear equivalent to the Polynomial 2° model; however, differences exist in digits not
displayed in the table.
gThe Polynomial 2° model may appear equivalent to the Power model; however, differences exist in digits not
displayed in the table. This also applies to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-18 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplem en tal Information—Trim ethylbenzenes
Table D-13. 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 (modeled variance, high dose dropped) (Korsak
etal.. 2000al
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.0276
148.13
413
227
No model was 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 (M3)
N/Ab
154.45
495
165
Exponential (M4)
N/Ab
145.28
149
0.431
Power0
Lineard
0.0197
148.72
447
275
Polynomial 2°e
0.0197
148.72
447
275
aModeled variance case presented (BMDS Test 2 p-value = 0.00849, BMDS Test 3 p-value = 0.116); no model was
selected as a best-fitting model.
bNo available degrees of freedom to calculate a goodness-of-fit value.
Tor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
dThe Linear model may appear equivalent to the Polynomial 2° model; however, differences exist in digits not
displayed in the table.
eThe Polynomial 2° model may appear equivalent to the Power model; however, differences exist in digits not
displayed in the table. This also applies to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-19 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Table D-14. 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
etal.. 2000al
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.716
189.11
916
535
Of the models that provided an
adequate fit, the Exponential M2
model was selected, based on
lowest AIC (BMDLS differed by
<3-fold)
Exponential (M4)
0.448
191.01
815
262
Exponential (M5)
N/Ac
192.49
548
138
Hill
N/Ac
192.49
564
Errord
Power6
Polynomial 3°f
Polynomial 2°g
Linear
0.671
189.23
979
633
aConstant variance case presented (BMDS Test 2 p-value = 0.269), selected model in bold; scaled residuals for
selected model for doses 0,128, 523, and 1,269 were -0.24, 0.57, -0.5, and 0.18, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-20 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Exponential 2 Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
Exponential 2
30
25
20
15
10
0
200
400
600
800
1000
1200
dose
08:32 08/26 2015
BMR = 1 std. dev. change from control mean; dose shown in mg/m31,2,3-TMB.
Figure D-4. Plot of mean response by dose for decreased segmented
neutrophils in male Wistar rats, with fitted curve for Exponential M2 model
with constant variance (Korsak et al.. 2000a).
Exponential Model (Version: 1.10; Date: 01/12/2015)
The form of the response function is: Y[dose] = a * exp(sign * b * dose)
A constant variance model is fit
Benchmark Dose Computation
BMR = 1.0000 Estimated SD from control
BMD = 915.77
BMDL at the 95% confidence level = 534.809
Parameter Estimates
Variable
Estimate
Default initial parameter values
Inalpha
3.57763
3.56089
rho
N/A
0
a
25.2579
19.0843
b
0.000295164
0.00028845
c
N/A
0
d
N/A
1
This document is a draft for review purposes only and does not constitute Agency policy.
D-21 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table o
Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
10
24.8
25.26
4.5
5.98
-0.242
128
10
25.4
24.32
5.8
5.98
0.5701
523
10
20.7
21.64
5.8
5.98
-0.4994
1,269
10
17.7
17.37
8.3
5.98
0.176
2
3
Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
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
2
-91.55261
3
189.1052
4
5 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
13.82
6
0.03172
Test 2
3.929
3
0.2692
Test 3
3.929
3
0.2692
Test 4
0.6696
2
0.7155
6
This document is a draft for review purposes only and does not constitute Agency policy.
D-22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Table D-15. 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
etal.. 2000al
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.640
177.65
517
335
Of the models that provided an
adequate fit, the Hill model was
selected, based on lowest BMDL
(BMDLS differed by >3-fold)
Exponential (M4)
Exponential (M5)c
0.521
179.17
365
134
Hill
0.569
179.08
337
99.2
Polynomial 3°d
0.453
178.34
646
465
Polynomial 2°e
Linear'
0.453
178.34
646
465
aConstant variance case presented (BMDS Test 2 p-value = 0.0925), selected model in bold; scaled residuals for
selected model for doses 0,128, 523, and 1,269 were 0.21, -0.41, 0.31, and -0.11, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
Tor the Exponential (M5) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M4) model.
dThe Polynomial 3° model may appear equivalent to the Polynomial 2° model; however, differences exist in digits
not displayed in the table. This also applies to the Linear model.
eFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
the Linear model may appear equivalent to the Polynomial 3° model; however, differences exist in digits not
displayed in the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-23 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Hill Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
25
20
15
10
5
0
200
400
600
800
1000
1200
dose
08:37 08/26 2015
BMR = 1SD change from control mean; dose shown in mg/m31,2,3-TMB.
Figure D-5. Plot of mean response by dose for decreased segmented
neutrophils in female Wistar rats, with fitted curve for Hill model with
constant variance (Korsak et al.. 2000a).
Hill Model (Version: 2.17; Date: 01/28/2013)
The form of the response function is: Y[dose] = intercept + v*doseAn/(kAn + doseAn)
A constant variance model is fit
Benchmark Dose Computation
BMR = 1 Estimated SD from the control mean
BMD = 337.444
BMDL atthe 95% confidence level = 99.2111
Parameter Estimates
Variable
Estimate
Default initial parameter values
alpha
26.4982
29.205
rho
N/A
0
intercept
22.76
23.1
V
-17.5026
-11.2
n
1
1.05772
k
809.904
391.333
This document is a draft for review purposes only and does not constitute Agency policy.
D-24 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Table o
Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
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
1,269
10
11.9
12.1
7.1
5.15
-0.108
2
3
Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
-85.379588
5
180.759176
A2
-82.165225
8
180.33045
A3
-85.379588
5
180.759176
Fitted
-85.541569
4
179.083138
R
-95.409822
2
194.819645
4
5 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
26.4892
6
0.0001804
Test 2
6.42873
3
0.09252
Test 3
6.42873
3
0.09252
Test 4
0.323961
1
0.5692
6
This document is a draft for review purposes only and does not constitute Agency policy.
D-25 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Table D-16. Summary of BMD modeling results for increased reticulocytes 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.. 2000a)
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.273
89.084
1,112
807
Of the models that provided an
adequate fit, the Linear model
was selected, based on lowest
AIC (BMDLS differed by <3-fold)
Exponential (M4)
0.140
90.670
900
308
Exponential (M5)
N/Ac
91.370
540
141
Hill
N/Ac
91.370
554
Errord
Power®
Polynomial 30f
Polynomial 2°B
Linear
0.311
88.829
1,025
653
aConstant variance case presented (BMDS Test 2 p-value = 0.522), selected model in bold; scaled residuals for
selected model for doses 0,128, 523, and 1,269 were 0.56, -1.14, 0.79, and -0.21, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-26 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Linear Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
6
5
4
3
2
1
0
200
400
600
800
1000
1200
dose
08:45 08/26 2015
BMR = 1SD change from control mean; dose shown in mg/m31,2,3-TMB.
Figure D-6. Plot of mean response by dose for increased reticulocytes in
female Wistar rats, with fitted curve for Linear model with constant variance
(Korsak et al.. 2000a).
Polynomial Model (Version: 2.20; Date: 10/22/2014)
The form of the response function is: Y[dose] = beta_0 + beta_l*dose
A constant variance model is fit
Benchmark Dose Computation
BMR = 1 Estimated SD from the control mean
BMD = 1025.1
BMDL at the 95% confidence level = 652.898
Parameter Estimates
Variable
Estimate
Default initial parameter values
alpha
2.91747
3.0575
rho
N/A
0
beta_0
2.50021
2.50021
beta_l
0.00166623
0.00166623
This document is a draft for review purposes only and does not constitute Agency policy.
D-27 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
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
1,269
10
4.5
4.61
1.8
1.71
-0.212
2
3 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
-40.244741
5
90.489483
A2
-39.119955
8
94.23991
A3
-40.244741
5
90.489483
Fitted
-41.414322
3
88.828645
R
-45.600613
2
95.201226
4
5 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
12.9613
6
0.04365
Test 2
2.24957
3
0.5223
Test 3
2.24957
3
0.5223
Test 4
2.33916
2
0.3105
6
This document is a draft for review purposes only and does not constitute Agency policy.
D-28 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-17. Summary of BMD modeling results for decreased fetal weight in
2 male Sprague-Dawley rat pups exposed to 1,2,4-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD or 5% change from control mean (constant variance)
4 fSaillenfait et al.. 20051
Model3
Goodness of fit
BMD
(mg/m3)
BMDL
(mg/m3)
Basis for model selection
p-value
AIC
BMR = 1 SD change from control mean
Exponential (M2)
0.571
-84.273
2,803
2,140
Of the models that provided an
adequate fit, the Linear model
was selected, based on lowest
AIC (BMDLS differed by <3-fold)
Exponential (M3)
0.833
-83.913
3,440
2,349
Exponential (M4)
0.571
-84.273
2,803
2,052
Exponential (M5)
0.546
-81.913
3,440
2,349
Hill
0.559
-81.936
3,441
2,367
Power
0.843
-83.937
3,441
2,368
Polynomial 3°
0.952
-84.180
3,444
2,408
Polynomial 2°
0.883
-84.029
3,399
2,383
Linear
0.622
-84.509
2,839
2,202
BMR = 5% change from control mean
Exponential (M2)
0.571
-84.273
2,009
1,577
Of the models that provided an
adequate fit, the Linear model
was selected, based on lowest AIC
(BMDLS differed by <3-fold)
Exponential (M3)
0.833
-83.913
2,861
1,716
Exponential (M4)
0.571
-84.273
2,009
1,428
Exponential (M5)
0.546
-81.913
2,861
1,716
Hill
0.559
-81.936
2,858
1,750
Power
0.843
-83.937
2,857
1,751
Polynomial 3°
0.952
-84.180
2,841
1,777
Polynomial 2°
0.883
-84.029
2,799
1,761
Linear
0.622
-84.509
2,057
1,640
5
6 aConstant variance case presented (BMDS Test 2 p-value = 0.101), selected model in bold; scaled residuals for
7 selected model for doses 0,492,1,471, 2,913, and 4,408 were -0.34, -0.32, 0.49, 0.91, and -0.69, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
D-29 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Linear Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
1
2
3
4
5
EjSMpL , BlVip
1500 2000 2oOO 3000
15:52 08/27 2015
4000 4500
BMR = 1SD change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-7. Plot of mean response by dose for decreased fetal weight in male
Sprague-Dawley rat pups, with fitted curve for Linear model with constant
variance (Saillenfait et al.. 2005).
Linear Model, with BMR of 0.05 Rel. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
7
8
9
10
11
ryiDL , bmd
1500 2000 2500 3000
15:57 08/27 2015
4000 4500
BMR = 5% change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-8. Plot of mean response by dose for decreased fetal weight in male
Sprague-Dawley rat pups, with fitted curve for Linear model with constant
variance fSaillenfait et al.. 20051.
This document is a draft for review purposes only and does not constitute Agency policy.
D-30 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Polynomial Model (Version: 2.20; Date: 10/22/2014)
2 The form of the response function is: Y[dose] = beta_0 + beta_l*dose
3 A constant variance model is fit
4
5 Benchmark Dose Computation
6 BMR = 5% Relative deviation
7 BMD = 2057.05
8 BMDL at the 95% confidence level = 1640.07
9
10 Parameter Estimates
Variable
Estimate
Default initial parameter values
alpha
0.165139
0.170101
rho
N/A
0
beta_0
5.88846
5.88821
beta_l
-0.000143129
-0.000142292
11
12 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
23
5.86
5.89
0.34
0.41
-0.336
492
22
5.79
5.82
0.3
0.41
-0.324
1,471
22
5.72
5.68
0.49
0.41
0.486
2,913
22
5.55
5.47
0.48
0.41
0.906
4,408
24
5.2
5.26
0.42
0.41
-0.694
13
14 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
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
16 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
42.0882
8
<0.0001
Test 2
7.7582
4
0.1008
Test 3
7.7582
4
0.1008
Test 4
1.76897
3
0.6217
17
18
19
This document is a draft for review purposes only and does not constitute Agency policy.
D-31 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Supplem en tal Information—Trim ethylbenzenes
Table D-18. Summary of BMD modeling results for decreased fetal weight in
male Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (constant variance)
fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.693
-66.941
3,397
2,560
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-19).
Exponential (M4)
0.698
-65.678
2,605
1,341
Exponential (M5)
0.397
-63.679
2,603
1,341
Hill
0.409
-63.716
2,572
1,275
Power0
Polynomial 3°d
Polynomial 2°e
Linear
0.650
-66.753
3,513
2,695
aConstant variance case presented (BMDS Test 2 p-value = 0.00237, BMDS Test 3 p-value = 0.00237); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
Tor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
dFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
eFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-32 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Supplem en tal Information—Trim ethylbenzenes
Table D-19. Summary of BMD modeling results for decreased fetal weight in
male Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (modeled variance)
fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.521
-73.291
2,523
1,779
No model selected as Test 3
p-value was <0.1. The data were
remodeled after dropping the
high dose (see Table D-20)
Exponential (M4)
0.430
-71.859
2,042
1,125
Exponential (M5)
0.388
-70.799
2,045
1,238
Hill
0.458
-70.996
1,984
1,235
Power0
Polynomial 3°d
Polynomial 2°e
Linear
0.479
-73.067
2,636
1,890
aModeled variance case presented (BMDS Test 2 p-value = 0.00237, BMDS Test 3 p-value = 0.0603); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
Tor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
dFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
eFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-33 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplem en tal Information—Trim ethylbenzenes
Table D-20. Summary of BMD modeling results for decreased fetal weight in
male Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (constant variance, high
dose dropped) fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.557
-68.864
2,536
1,720
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-21).
Exponential (M4)
0.395
-67.312
2,232
971
Exponential (M5)
N/Ac
-66.037
1,961
530
Hill
N/Ac
-66.037
2,182
551
Powerd
Polynomial 3°e
Polynomial 2°f
Linear
0.539
-68.798
2,563
1,768
aConstant variance case presented (BMDS Test 2 p-value = 0.00872, BMDS Test 3 p-value = 0.00872); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
eFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
fFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-34 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-21. Summary of BMD modeling results for decreased fetal weight in
2 male Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD change from control mean (modeled variance, high
4 dose dropped) fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.454
-70.868
2,049
1,327
No model was 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 (M3)
0.272
-69.242
2,226
1,364
Exponential (M4)
0.454
-70.868
2,049
1,130
Exponential (M5)
N/Ab
-68.255
1,549
1,204
Hill
N/Ab
-68.255
1,568
1,156
Power
0.266
-69.213
2,236
1,390
Polynomial 3°c
Polynomial 2°
0.233
-69.024
2,218
1,372
Linear
0.462
-70.905
2,067
1,360
5
6 aModeled variance case presented (BMDS Test 2 p-value = 0.00872, BMDS Test 3 p-value = 0.0269); no model was
7 selected as a best-fitting model.
8 bNo available degrees of freedom to calculate a goodness-of-fit value.
9 Tor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
10 this row reduced to the Polynomial 2° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-35 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-22. Summary of BMD modeling results for decreased fetal weight in
2 female Sprague-Dawley rat pups exposed to 1,2,4-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD or 5% change from control mean (constant variance)
4 fSaillenfait et al.. 20051
Model3
Goodness of fit
BMD
(mg/m3)
BMDL
(mg/m3)
Basis for model selection
p-value
AIC
BMR = 1 SD change from control mean
Exponential (M2)
0.506
-101.65
2,651
2,045
Of the models that provided an
adequate fit, the Linear model
was selected, based on lowest
AIC (BMDLS differed by <3-fold)
Exponential (M3)
0.654
-101.14
3,313
2,212
Exponential (M4)
0.506
-101.65
2,651
1,948
Exponential (M5)
0.357
-99.136
3,313
2,212
Hill
0.370
-99.180
3,312
2,241
Power
0.669
-101.18
3,312
2,242
Polynomial 3°
0.832
-101.62
3,322
2,307
Polynomial 2°
0.725
-101.34
3,259
2,264
Linear
0.555
-101.90
2,692
2,109
BMR = 5% change from control mean
Exponential (M2)
0.506
-101.65
1,951
1,549
Of the models that provided an
adequate fit, the Linear model
was selected, based on lowest AIC
(BMDLS differed by <3-fold)
Exponential (M3)
0.654
-101.14
2,779
1,663
Exponential (M4)
0.506
-101.65
1,951
1,398
Exponential (M5)
0.357
-99.136
2,779
1,663
Hill
0.370
-99.180
2,774
1,702
Power
0.669
-101.18
2,773
1,704
Polynomial 3°
0.832
-101.62
2,765
1,747
Polynomial 2°
0.725
-101.34
2,703
1,719
Linear
0.555
-101.90
2,001
1,613
5
6 aConstant variance case presented (BMDS Test 2 p-value = 0.394), selected model in bold; scaled residuals for
7 selected model for doses 0,492,1,471, 2,913, and 4,408 were -0.31, -0.19, 0.14,1.16, and -0.76, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
D-36 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Linear Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
1
2
3
4
5
BMDL
BMD
1000 15UO 2000 25UO 3000 3500 4000 4500
16:02 08/27 2015
BMR = 1SD change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-9. Plot of mean response by dose for decreased fetal weight in
female Sprague-Dawley rat pups, with fitted curve for Linear model with
constant variance (Saillenfait et al.. 2005).
Linear Model, with BMR of 0.05 Rel. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
6
7
8
9
10
[V1PL BMP
1000 1500 2000 2500 3000 3500 4000 4500
16:08 08/27 2015
BMR = 5% change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-10. Plot of mean response by dose for decreased fetal weight in
female Sprague-Dawley rat pups, with fitted curve for Linear model with
constant variance (Saillenfait et al.. 2005).
11
This document is a draft for review purposes only and does not constitute Agency policy.
D-37 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Polynomial Model (Version: 2.20; Date: 10/22/2014)
2 The form of the response function is: Y[dose] = beta_0 + beta_l*dose
3 A constant variance model is fit
4
5 Benchmark Dose Computation
6 BMR = 5% Relative deviation
7 BMD = 2001.36
8 BMDL atthe 95% confidence level = 1612.89
9
10 Parameter Estimates
Variable
Estimate
Default initial parameter values
alpha
0.141584
0.14543
rho
N/A
0
beta_0
5.59423
5.59388
beta_l
-0.000139761
-0.000138886
11
12 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
23
5.57
5.59
0.33
0.38
-0.309
492
22
5.51
5.53
0.31
0.38
-0.193
1,471
22
5.4
5.39
0.45
0.38
0.142
2,913
22
5.28
5.19
0.4
0.38
1.16
4,408
24
4.92
4.98
0.4
0.38
-0.757
13
14 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
54.992554
6
-97.985109
A2
57.03888
10
-94.07776
A3
54.992554
6
-97.985109
Fitted
53.949538
3
-101.899075
R
36.10487
2
-68.20974
15
16 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
41.868
8
<0.0001
Test 2
4.09265
4
0.3936
Test 3
4.09265
4
0.3936
Test 4
2.08603
3
0.5547
17
18
This document is a draft for review purposes only and does not constitute Agency policy.
D-38 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
Supplem en tal Information—Trim ethylbenzenes
Table D-23. Summary of BMD modeling results for decreased fetal weight in
female Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (constant variance)
fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.911
-61.962
3,582
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,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-24).
Exponential (M4)c
0.766
-59.962
3,573
1,916
Exponential (M5)d
0.766
-59.962
3,573
1,916
Hill
0.766
-59.963
3,570
1,866
Power6
Polynomial 3°f
Polynomial 2°g
Linear
0.909
-61.950
3,677
2,794
aConstant variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = <0.0001); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cThe Exponential (M4) model may appear equivalent to the Exponential (M5) model; however, differences exist in
digits not displayed in the table.
dThe Exponential (M5) model may appear equivalent to the Exponential (M4) model; however, differences exist in
digits not displayed in the table.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-39 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplem en tal Information—Trim ethylbenzenes
Table D-24. Summary of BMD modeling results for decreased fetal weight in
female Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (modeled variance)
fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0193
-67.537
2,693
1,828
No model selected as Test 3
p-value was <0.1. The data were
remodeled after dropping the
high dose (see Table D-25)
Exponential (M4)
0.0510
-69.499
1,482
798
Exponential (M5)
0.533
-73.064
1,469
1,070
Hill
0.782
-75.064
1,469
1,023
Power
0.0155
-67.061
2,841
1,970
Polynomial 3°c
Polynomial 2°d
Linear
0.0148
-67.061
2,841
1,970
aModeled variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = 0.0130); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
Tor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
dFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-40 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Table D-25. Summary of BMD modeling results for decreased fetal weight in
female Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (constant variance, high
dose dropped) fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.769
-50.212
3,703
2,222
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-26).
Exponential (M4)
0.565
-48.406
4,626
1,518
Exponential (M5)
N/Ac
-46.738
Errord
0
Hill
N/Ac
-46.738
Errord
Errord
Power6
Polynomial 3°f
Polynomial 2°g
Linear
0.759
-50.187
3,688
2,258
aConstant variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = <0.0001); no model was
selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dBMD or BMDL computation failed for this model.
eFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-41 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-26. Summary of BMD modeling results for decreased fetal weight in
2 female Sprague-Dawley rat pups exposed to 1,3,5-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD change from control mean (modeled variance, high
4 dose dropped) fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.310
-68.515
2,083
1,198
No model was 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 (M3)
0.159
-66.872
2,156
1,237
Exponential (M4)
0.310
-68.515
2,083
1,104
Exponential (M5)
N/Ab
-68.570
1,527
1,210
Hill
N/Ab
-68.570
1,555
Error0
Power
0.153
-66.809
2,171
1,255
Polynomial 3°d
Polynomial 2°
0.0181
-66.546
2,122
1,227
Linear
0.0608
-68.532
2,093
1,226
5
6 aModeled variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = 0.0609); no model was
7 selected as a best-fitting model.
8 bNo available degrees of freedom to calculate a goodness-of-fit value.
9 CBMD or BMDL computation failed for this model.
10 dFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
11 this row reduced to the Polynomial 2° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-42 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-27. Summary of BMD modeling results for decreased dam weight gain
2 in female Sprague-Dawley rats exposed to 1,2,4-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD or 10% change from control mean (constant variance)
4 fSaillenfait et al.. 20051
Model3
Goodness of fit
BMD
(mg/m3)
BMDL
(mg/m3)
Basis for model selection
p-value
AIC
BMR = 1 SD change from control mean
Exponential (M2)
0.221
844.93
3,204
2,312
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-28).
Exponential (M3)
0.613
843.50
3,839
2,967
Exponential (M4)
0.221
844.93
3,204
2,299
Exponential (M5)
0.322
845.50
3,839
2,967
Hill
0.324
845.49
3,850
2,943
Power
0.615
843.49
3,851
2,940
Polynomial 3°
0.664
843.34
3,813
2,924
Polynomial 2°
0.771
841.65
3,734
3,266
Linear
0.292
844.25
3,231
2,444
BMR = 10% change from control mean
Exponential (M2)
0.221
844.93
1,683
1,273
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-28).
Exponential (M3)
0.613
843.50
2,994
1,791
Exponential (M4)
0.221
844.93
1,683
1,185
Exponential (M5)
0.322
845.50
2,994
1,791
Hill
0.324
845.49
2,991
1,736
Power
0.615
843.49
2,990
1,729
Polynomial 3°
0.664
843.34
2,906
1,714
Polynomial 2°
0.771
841.65
2,753
2,451
Linear
0.292
844.25
1,781
1,406
5
6 aConstant variance case presented (BMDS Test 2 p-value = 0.0215, BMDS Test 3 p-value = 0.0215); no model was
7 selected as a best-fitting model.
8
This document is a draft for review purposes only and does not constitute Agency policy.
D-43 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-28. Summary of BMD modeling results for decreased dam weight gain
2 in female Sprague-Dawley rats exposed to 1,2,4-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD or 10% change from control mean (modeled variance)
4 fSaillenfait et al.. 20051
Model3
Goodness of fit
BMD
(mg/m3)
BMDL
(mg/m3)
Basis for model selection
p-value
AIC
BMR = 1 SD change from control mean
Exponential (M2)
0.0996
843.22
3,458
2,516
Of the models that provided an
adequate fit, the Polynomial 3
model was selected, based on
lowest AIC (BMDLS differed by
<3-fold)
Exponential (M3)
Exponential (M5)b
0.218
842.00
3,935
3,116
Exponential (M4)
0.0996
843.22
3,458
2,515
Hill
0.0827
843.97
3,941
Error0
Power
0.222
841.97
3,941
3,078
Polynomial 3°
0.274
841.55
3,899
3,094
Polynomial 2°
0.219
842.00
3,851
3,025
Linear
0.144
842.38
3,474
2,649
BMR = 10% change from control mean
Exponential (M2)
0.0996
843.22
1,581
1,232
Of the models that provided an
adequate fit, the Polynomial 3
model was selected, based on
lowest AIC (BMDLS differed by
<3-fold)
Exponential (M3)
Exponential (M5)b
0.218
842.00
2,910
1,664
Exponential (M4)
0.0996
843.22
1,581
1,152
Hill
0.0827
843.97
2,891
1,799
Power
0.222
841.97
2,889
1,573
Polynomial 3°
0.274
841.55
2,734
1,631
Polynomial 2°
0.219
842.00
2,655
1,567
Linear
0.144
842.38
1,694
1,380
5
6 aModeled variance case presented (BMDS Test 2 p-value = 0.0215), selected model in bold; scaled residuals for
7 selected model for doses 0, 492,1,471, 2,913, and 4,408 were 0.29, -0.73, 0.29, 0.22, and -0.09, respectively.
8 bFor the Exponential (M5) model, the estimate of c was 0 (boundary). The models in this row reduced to the
9 Exponential (M3) model.
10 CBMD or BMDL computation failed for this model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-44 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
Polynomial Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
50
Polynomial
40
30
20
0
00
90
2000 2500
dose
07:44 10/26 2015
2 BMR = 1SD change from control mean; dose shown in mg/m31,2,4-TMB.
3 Figure D-ll. Plot of mean response by dose for decreased dam weight gain in
4 female Sprague-Dawley rats, with fitted curve for Polynomial 3 model with
5 modeled variance (Saillenfait et al.. 2005).
Polynomial Model, with BMR of 0.1 Rel. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
50
Polynomial
40
30
20
0
00
90
0 500 1000 1500 2000 2500 3000 3500 4000 4500
dose
^ 14:45 08/25 2015
7 BMR = 10% change from control mean; dose shown in mg/m31,2,4-TMB.
8 Figure D-12. Plot of mean response by dose for decreased dam weight gain in
9 female Sprague-Dawley rats, with fitted curve for Polynomial 3 model with
10 modeled variance fSaillenfait et al.. 20051.
11
This document is a draft for review purposes only and does not constitute Agency policy.
D-45 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Polynomial Model (Version: 2.20; Date: 10/22/2014)
2 The form of the response function is: Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ...
3 A modeled variance is fit
4
5 Benchmark Dose Computation
6 BMR = 1 Estimated SD from the control mean
7 BMD = 3898.99
8 BMDL atthe 95% confidence level = 3094.13
9
10 Parameter Estimates
Variable
Estimate
Default initial parameter values
lalpha
-4.72235
6.36522
rho
2.31145
0
beta_0
129.446
129.55
beta_l
-0.00285669
-0.00648229
beta_2
-1.02802 x 10"17
0
beta_3
-0.000000000251312
-0.000000000702052
11
12 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
24
131
129
33
26
0.292
492
22
124
128
18
25.7
-0.732
1,471
22
126
124
24
24.9
0.293
2,913
22
116
115
23
22.7
0.225
4,408
24
95
95.3
19
18.3
-0.0881
13
14 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
-417.261306
6
846.522613
A2
-411.512361
10
843.024723
A3
-414.479759
7
842.959518
Fitted
-415.773389
5
841.546778
R
-432.234922
2
868.469844
15
16 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
41.4451
8
<0.0001
Test 2
11.4979
4
0.0215
Test 3
5.9348
3
0.1148
Test 4
2.58726
2
0.2743
This document is a draft for review purposes only and does not constitute Agency policy.
D-46 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplem en tal Information—Trim ethylbenzenes
Table D-29. Summary of BMD modeling results for decreased dam weight gain
in female Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on
GDs 6-20; BMR = 1 SD change from control mean (constant variance)
fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.276
705.72
1,414
1,142
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-30).
Exponential (M3)
0.153
707.61
1,520
1,147
Exponential (M4)
0.149
707.66
1,349
930
Exponential (M5)
0.281
707.01
1,634
1,126
Hill
0.341
706.76
1,611
1,131
Powerb
Polynomial 3°c
Polynomial 2°d
Linear
0.128
707.53
1,825
1,537
aConstant variance case presented (BMDS Test 2 p-value = 2.83 x 10"4, BMDS Test 3 p-value = 2.83x 10"4); no
model was selected as a best-fitting model.
bFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
Tor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
dFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-47 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-30. Summary of BMD modeling results for decreased dam weight gain
2 in female Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD change from control mean (modeled variance)
4 fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.0503
697.91
1,058
816
No model selected as Test 3
p-value was <0.1. The data were
remodeled after dropping the
high dose (see Table D-31)
Exponential (M3)
0.0234
699.62
1,180
827
Exponential (M4)
0.0209
699.84
1,011
690
Exponential (M5)
0.0675
697.45
1,266
891
Hill
0.114
696.61
1,248
Errorb
Power
Polynomial 3°c
Polynomial 2°
0.0200
699.94
1,359
1,075
Linear
0.0200
699.94
1,359
Errorb
5
6 aModeled variance case presented (BMDS Test 2 p-value = 2.83 x 10~4, BMDS Test 3 p-value = 0.0575); no model
7 was selected as a best-fitting model.
8 bBMD or BMDL computation failed for this model.
9 Tor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
10 this row reduced to the Polynomial 2° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-48 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-31. Summary of BMD modeling results for decreased dam weight gain
2 in female Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD change from control mean (constant variance, high
4 dose dropped) fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.120
564.09
1,187
910
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-32).
Exponential (M3)
0.177
563.66
1,571
1,063
Exponential (M4)
0.120
564.09
1,187
881
Exponential (M5)
N/Ab
564.12
1,471
1,132
Hill
N/Ab
564.12
1,471
1,118
Power
0.149
563.92
1,596
1,088
Polynomial 3°c
Polynomial 2°
0.112
564.36
1,595
1,064
Linear
0.188
563.18
1,288
1,028
5
6 aConstant variance case presented (BMDS Test 2 p-value = 0.00105, BMDS Test 3 p-value = 0.00105); no model was
7 selected as a best-fitting model.
8 bNo available degrees of freedom to calculate a goodness-of-fit value.
9 Tor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
10 this row reduced to the Polynomial 2° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-49 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table D-32. Summary of BMD modeling results for decreased dam weight gain
2 in female Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on
3 GDs 6-20; BMR = 1 SD change from control mean (modeled variance, high
4 dose dropped) fSaillenfait et al.. 20051
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
0.0128
559.00
978
717
Although Test 3 p-value was
approximately 0.10, indicating
appropriate fit of the variance
model, no model was selected as
Test 4 p-value was <0.10.
Therefore, this endpoint cannot
be modeled in BMDS and the
NOAEL/LOAEL approach is
recommended.
Exponential (M3)
0.0127
558.50
1,275
853
Exponential (M4)
0.0128
559.00
978
698
Exponential (M5)
N/Ab
555.51
1,410
966
Hill
0.269
553.51
1,397
Error0
Power
0.00946
559.02
1,297
858
Polynomial 3°d
Polynomial 2°
0.00618
559.78
1,256
820
Linear
0.0181
558.31
1,053
798
5
6 aModeled variance case presented (BMDS Test 2 p-value = 0.00105, BMDS Test 3 p-value = 0.0996); no model was
7 selected as a best-fitting model.
8 bNo available degrees of freedom to calculate a goodness-of-fit value.
9 CBMD or BMDL computation failed for this model.
10 dFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
11 this row reduced to the Polynomial 2° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-50 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
Supplem en tal Information—Trim ethylbenzenes
Table D-33. Summary of BMD modeling results for increased monocytes in
male Wistar rats exposed to 1,3,5-TMB by gavage for 13 weeks; BMR = 1 SD
change from control mean (constant variance) (Adenuga et al.. 2014)
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)b
0.00910
-106.57
1,600
640
No model selected as Test 2
p-value was <0.10. Therefore, as
suggested in the Benchmark Dose
Technical Guidance (U.S. EPA,
2012), the data were remodeled
using a non-homogenous
variance model (see Table D-32).
Exponential (M3)c
0.00910
-106.57
1,600
640
Exponential (M4)
0.0917
-111.12
99.3
0.410
Exponential (M5)
N/Ad
-109.20
71.7
0.329
Hill
N/Ad
-109.20
58.0
6.86 x 10"7
Power6
0.00969
-106.69
1,645
582
Polynomial 3°f
Polynomial 2°g
Linearh
0.00969
-106.69
1,645
582
aConstant variance case presented (BMDS Test 2 p-value = 0.0402, BMDS Test 3 p-value = 0.0402); no model was
selected as a best-fitting model.
bThe Exponential (M2) model may appear equivalent to the Exponential (M3) model; however, differences exist in
digits not displayed in the table.
cThe Exponential (M3) model may appear equivalent to the Exponential (M2) model; however, differences exist in
digits not displayed in the table.
dNo available degrees of freedom to calculate a goodness-of-fit value.
eThe Power model may appear equivalent to the Polynomial 3° model; however, differences exist in digits not
displayed in the table. This also applies to the Polynomial 2° model. This also applies to the Linear model.
fFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
gFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
hThe Linear model may appear equivalent to the Power model; however, differences exist in digits not displayed in
the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-51 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplem en tal Information—Trim ethylbenzenes
Table D-34. Summary of BMD modeling results for increased monocytes in
male Wistar rats exposed to 1,3,5-TMB by gavage for 13 weeks; BMR = 1 SD
change from control mean (modeled variance) (Adenuga et al.. 2014)
Model3
Goodness of fit
BMDisd
(mg/m3)
BMDLisd
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.00313
-107.32
772
334
Of the models that provided an
adequate fit, the Exponential M4
model was selected as the only
appropriately fitting model.
Exponential (M4)
0.231
-115.41
52.0
13.9
Exponential (M5)
N/Ac
-113.92
56.1
17.3
Hill
N/Ac
-113.92
51.8
33.9
Power
<0.0001
-62.935
60,000
5.87 x 10"12
Polynomial 3°d
Polynomial 2°e
Linear
0.00553
-108.45
453
161
aModeled variance case presented (BMDS Test 2 p-value = 0.0402); selected model in bold; scaled residuals for
selected model for doses 0, 50, 200, and 600 were -0.27, 0.44, 0.98, and -1.15, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dFor the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
eFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-52 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplem en tal Information—Trim ethylbenzenes
Exponential 4 Model, with BMR of 1 Std. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
.45
Exponential 4
.4
.35
.3
.25
.2
15
1
.05
300
dose
15:07 11/02 2015
BMR = 10% change from control mean; dose shown in mg/m31,2,4-TMB.
Figure D-13. Plot of mean response by dose for increased monocytes in male
Wistar rats, with fitted curve for Exponential M4 model with modeled
variance (Adenuga et al.. 2014).
Exponential Model (Version: 1.10; Date: 01/12/2015)
The form of the response function is: Y[dose] = a * [c-(c-l) * exp(-b * dose)]
A modeled variance is fit
Benchmark Dose Computation
BMR = 1.0000 Estimated SD from control
BMD = 51.9881
BMDL at the 95% confidence level = 13.9214
Parameter Estimates
Variable
Estimate
Default initial parameter values
Inalpha
-1.3469
-2.24702
rho
1.67291
1.1326
a
0.106615
0.095
b
0.0137132
0.00238203
c
2.44253
3.31579
d
N/A
1
This document is a draft for review purposes only and does not constitute Agency policy.
D-53 DRAFT—DO NOT CITE OR QUOTE
-------
Supplem en tal Information—Trim ethylbenzenes
1 Table of Data and Estimated Values of Interest
Dose
N
Observed mean
Estimated mean
Observed SD
Estimated SD
Scaled residuals
0
10
0.1
0.11
0.09
0.08
-0.2668
50
10
0.2
0.18
0.09
0.12
0.4381
200
10
0.3
0.25
0.17
0.16
0.977
600
10
0.2
0.26
0.18
0.17
-1.154
2
3 Likelihoods of Interest
Model
Log(likelihood)
Number of parameters
AIC
A1
60.98264
5
-111.9653
A2
65.13368
8
-114.2674
A3
63.4237
6
-114.8474
R
55.94043
2
-107.8809
4
62.70505
5
-115.4101
4
5 Tests of Interest
Test
-2*log(likelihood ratio)
Test df
p-value
Test 1
18.39
6
0.005336
Test 2
8.302
3
0.04016
Test 3
3.42
2
0.1809
Test 6a
1.437
1
0.2306
6
This document is a draft for review purposes only and does not constitute Agency policy.
D-54 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
Supplem en tal Information—Trim ethylbenzenes
REFERENCES FOR APPENDICES1
Adenuga. D: Carrillo. TC: Mckee. RH. (2014). The sub-chronic oral toxicity of 1,3,5-trimethylbenzene
in Sprague-Dawley rats. Regul Toxicol Pharmacol 69: 143-153.
http://dx.doi.Org/10.1016/i.yrtph.2014.03.006
Bakke. OM: Scheline. RR. (1970). Hydroxylation of aromatic hydrocarbons in the rat Toxicol Appl
Pharmacol 16: 691-700. http://dx.doi.org/10.1016/0041-008Xr70190074-8
Battig. K: Grandiean. E: Rossi. L: Rickenbacher. 1. (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.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchQuery=OTS00004600
Brown. RP: Delp. MD: Lindstedt. SL: Rhomberg. LR: Beliles. RP. (1997). Physiological parameter
values for physiologically based pharmacokinetic models [Review], Toxicol Ind Health 13:
407-484. http://dx.doi.Org/10.1177/074823379701300401
Carrillo. TC: Adenuga. MD: Mckee. RH. (2014). The sub-chronic toxicity of regular White Spirit in
rats. Regul Toxicol Pharmacol 70: 222-230. http ://dx. doi.org/10.1016 /i .vrtph. 2014.07.007
Cerf. 1: 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.
Clark. DG: Butterworth. ST: Martin. TG: Roderick. HR: Bird. MG. (1989). Inhalation toxicity of high
flash aromatic naphtha. Toxicol Ind Health 5: 415-428.
Collins. AS: Sumner. SCI: Borghoff. ST: Medinskv. 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. T: Zahm. SH.
(2001). Prenatal exposure to pesticides: a feasibility study among migrant and seasonal
farmworkers. Am J Ind Med 40: 578-585.
1Multiple references published in the same year by the same author(s) have been assigned a letter
(e.g., 1986a, 1986b) in Volume 1 of the Toxicological Review, based on which publication's 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
-------
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
39
40
41
42
43
44
45
46
47
48
Supplem en tal Information—Trim ethylbenzenes
Dahl. AR: Damon. EG: Maudeiiv. IL: Rothenberg. SI: Seiler. FA: Mcclellan. RO. (1988). Uptake of 19
hydrocarbon vapors inhaled by F344 rats. Fundam Appl Toxicol 10: 262-269.
http://dx.doi.Org/10.1016/0272-0590r88190310-7
Deurenberg. P: Weststrate. 1A: Seidell. 1C. (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. BT: Laseter. IL: Storer. 1. (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
El Hamid Hassan. AA: El Moez Elnagar. SA: El Taveb. IM: El Halim Bolbol. SA. (2013). Health hazards
of solvents exposure among workers in paint industry. OJSST 3: 87-95.
http://dx.doi.org/10.4236/oisst2013.340ri
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.
Fuente. A: McPherson. B: Cardemil. F. (2013). Xylene-induced auditory dysfunction in humans. Ear
Hear 34: 651-660. http://dx.doi.org/10.1097/AUD.0b013e31828d27d7
Fuente. A: McPherson. B: Hood. LI. (2012). Hearing loss associated with xylene exposure in a
laboratory worker. J Am Acad Audiol 23: 824-830. http://dx.doi.Org/10.3766/iaaa.23.10.7
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/BF004Q5692
Gage. IC. (1970). The subacute inhalation toxicity of 109 industrial chemicals. Br J Ind Med 27: 1-18.
http://dx.doi.org/10.ri36/oem.27.1.1
Gaschen. A: Lang. D: Kalberer. M: Savi. M: Geiser. T: Gazdhar. A: Lehr. CM: Bur. M: Dommen. I:
Baltensperger. U: Geiser. M. (2010). Cellular responses after exposure of lung cell cultures
to secondary organic aerosol particles. Environ Sci Technol 44: 1424-1430.
h ttp: / /dx. do i. o r g /10.1021/es9 02261 m
Gong. Y: Kishi. R: Kasai. S: Katakura. Y: Fuiiwara. K: Umemura. T: Kondo. T: Sato. T: Sata. F:
Tsukishima. E: Tozaki. S: Kawai. T: Mivama. Y. (2003). Visual dysfunction in workers
exposed to a mixture of organic solvents. Neurotoxicology 24: 703-710.
http:/ /dx.doi.org/10.1016/SO 161 -813Xf03100034-2
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/SO 161 -813Xf00100003-6
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
exposure to an industrial solvent, pseudocumene (1,2,4-trimethylbenzene). Int J Occup Med
Environ Health 10: 213-222.
This document is a draft for review purposes only and does not constitute Agency policy.
R-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
39
40
41
42
43
44
45
46
47
48
49
Supplem en tal Information—Trim ethylbenzenes
Gralewicz. S: Wiaderna. D: Tomas. T: Rvdzynski. 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/S0892-0362r97100001-9
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 tvpe/outbred rats/sprague dawley sd (accessed June 4, 2012).
Hissink. AM: Kriise. 1: Kulig. BM: Verwei. M: Muiiser. H: Salmon. F: Leenheers. LH: Owen. DE:
Lammers. 1H: 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
Huo. 1Z: 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/10.3109/00498258909Q34688
IBT Labs (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.
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
lanasik. B: lakubowski. M: lafowiecki. 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: Wvszvriska. K: Dziubaltowska. E. (1998). Genotoxicity evaluation of
trimethylbenzenes. MutatRes GenetToxicol Environ Mutagen 412: 299-305.
http://dx.doi.Org/10.1016/S1383-5718f97100202-7
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.Q223
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-
9697f97105482-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/s00204005Q532
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
tones. K: Meldrum. M: Baird. E: Cottrell. S: Kaur. P: Plant. N: Dyne. D: Cocker. I. (2006). Biological
monitoring for trimethylbenzene exposure: A human volunteer study and a practical
example in the workplace. Ann Occup Hyg 50: 593-598.
http://dx.doi.org/10.1093/annhvg/mel016
This document is a draft for review purposes only and does not constitute Agency policy.
R-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
39
40
41
42
43
44
45
46
47
48
49
50
Supplem en tal Information—Trim ethylbenzenes
luarez-Perez. CA: Torres-Valenzuela. A: Haro-Garci'a. LC: Boria-Aburto. VH: Aguilar-Madrid. G.
(2014). Ototoxicity effects of low exposure to solvent mixture among paint manufacturing
workers. Int J Audiol 53: 370-376. http://dx.doi.org/10.3109/14992027.2014.888597
luran. SA: lohanson. G: Ernstgard. L: Iregren. A: van Thriel. C. (2014). Neurobehavioral performance
in volunteers after inhalation of white spirits with high and low aromatic content Arch
Toxicol 88: 1127-1140. http://dx.doi.org/10.1007/s00204-014-1236-4
Kimura. K: Nagata. T: Hara. K: Kageura. M. (1988). Gasoline and kerosene components in blood-a
forensic analysis. Hum Toxicol 7: 299-305.
http://dx.doi.Org/10.1177/096032718800700401
Kim ura. K: Nagata. T: Kudo. K: Imamura. T: Hara. K. (1991). Determination of kerosene and light oil
components in blood. Biological Mass Spectrometry 20: 493-497.
http://dx.doi.Org/10.1002/bms.1200200810
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.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchQuery=OTS0558836
Koch Industries (Koch Industries, Incorporated). (1995b). 90-day oral gavage toxicity study of
I,3,5-trimethylbenzene in rats with a recovery group. (44618). Wichita, KS: Koch Industries,
Inc.
Korsak. Z: Rydzvhski. 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: Rydzvhski. K: laite. I. (1997). Respiratory irritative effects of trimethylbenzenes: An
experimental animal study. Int J Occup Med Environ Health 10: 303-311.
Korsak. Z: Stetkiewicz. I: Maicherek. W: Stetkiewicz. I: laite. I: Rydzvhski. 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: Rydzvhski. 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: Rydzvhski. 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/10.1016/0378-4274f95103305-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.
Lehotzkv. 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.
This document is a draft for review purposes only and does not constitute Agency policy.
R-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
38
39
40
41
42
43
44
45
46
47
48
49
Supplem en tal Information—Trim ethylbenzenes
Lutz. P: Gralewicz. S: Wiaderna. D: Swiercz. R: Grzelinska. 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/vl0001-010-00Q5-8
Maltoni. C: Ciliberti. A: Pinto. C: Soffritti. M: Belpoggi. F: Menarini. L. (1997). Results of long-term
experimental carcinogenicity studies of the effects of gasoline, correlated fuels, and major
gasoline aromatics on rats. Ann N Y Acad Sci 837: 15-52. http://dx.doi.org/10.1111 /i. 1749-
6632.1997.tb56863.x
Maule. AL: Heaton. KT: Rodrigues. E: Smith. KW: Mcclean. MP: Proctor. SP. (2013). Postural sway
and exposure to jet propulsion fuel 8 among US Air Force personnel. J Occup Environ Med
55: 446-453. http:/ /dx.doi.org/10.1097/IQM.OhOl3e31827db94b
Mckee. RH: Lammers. 1H: 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. 1L. (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
Mikulski. PI: Wiglusz. R. (1975). The comparative metabolism of mesitylene, pseudocumene, and
hemimellitene in rats. Toxicol Appl Pharmacol 31: 21-31. http://dx.doi.org/10.1016/0041-
008Xf 75190048-4
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 CFRPart 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/N iosh/9 2-10 O.html
Nong. A: Mccarver. DG: Hines. RN: Krishnan. K. (2006). Modeling interchild differences in
pharmacokinetics on the basis of subject-specific data on physiology and hepatic CYP2E1
levels: a case study with toluene. Toxicol Appl Pharmacol 214: 78-87.
http://dx.doi.Org/10.1016/i.taap.2005.12.001
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 (pp. 194). Washington, DC: National Academies Press.
http://www.nap.edu/catalog/13142.html
NRC (National Research Council). (2013). Acute exposure guideline levels for selected airborne
chemicals. Washington, DC: National Academies Press.
http://www.nap.edu/catalog/15852/acute-exposure-guideline-levels-for-selected-
airborne-chemicals-volume-13
This document is a draft for review purposes only and does not constitute Agency policy.
R-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
39
40
41
42
43
44
45
46
47
48
49
Supplem en tal Information—Trim ethylbenzenes
Pelekis. M: Gephart. LA: Lerman. SE. (2001). Physiological-model-based derivation of the adult and
child pharmacokinetic intraspecies uncertainty factors for volatile organic compounds.
Regul Toxicol Pharmacol 33: 12-20. http://dx.doi.org/10.1006/rtph.2000.1436
Pratt. H: Karim. N: Bleich. N: Mittelman. N. (2000). Short latency visual evoked potentials in
occupational exposure to organic solvents. Neurophysiol Clin 30: 306-312.
http:/ /dx.doi.org/10.1016/S0987-7053f00100230-6
Pvvkko. K. (1980). Effects of methylbenzenes on microsomal enzymes in rat liver, kidney and lung.
Biochim Biophys Acta 633: 1-9. http://dx.doi.org/10.1016/0304-4165r80190032-X
Ouevedo. L: Tochetto. T: Siqueira. MA: Machado. MS. (2012). Auditory brainstem response in gas
station attendants. Braz J Otorhinolaryngol 78: 63-68. http://dx.doi.org/10.5935/1808-
8694.20120035
Ruiiten. MWM. M: Hooisma. 1: Brons. IT: H abets. CEP: Emmen. HH: Muiiser. H. (1994).
Neurobehavioral effects of long-term exposure to xylene and mixed organic solvents in
shipyard spray painters. Neurotoxicology 15: 613-620.
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.
Silva. CL: Passos. M: Camara. IS. (2011). Investigation of urinary volatile organic metabolites as
potential cancer biomarkers by solid-phase microextraction in combination with gas
chromatography-mass spectrometry. Br J Cancer 105: 1894-1904.
http://dx.doi.org/10.1038/bic.20ri.437
Silva. CL: Passos. M: Camara. IS. (2012). Solid phase microextraction, mass spectrometry and
metabolomic approaches for detection of potential urinary cancer biomarkers-a powerful
strategy for breast cancer diagnosis. Talanta 89: 360-368.
http://dx.doi.Org/10.1016/i.talanta.20ri.12.041
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: Maicherek. W: Wasowicz. W. (2016). Hemimellitene (1,2,3-trimethylbenzene) in the
liver, lung, kidney, and blood, and dimethylbenzoic acid isomers in the liver, lung, kidney
and urine of rats after single and repeated inhalation exposure to hemimellitene. Int J Occup
Med Environ Health 29: 113-128. http://dx.doi.org/10.13075/iiomeh.1896.00599
Swiercz. R: Rvdzvhski. 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: Rvdzvhski. 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.
Tang. CY: Carpenter. DM: Eaves. EL: Ng. I: Ganeshalingam. N: Weisel. C: Oian. H: Lange. G: Fiedler.
NL. (2011). Occupational solvent exposure and brain function: an fMRI study. Environ
Health Perspect 119: 908-913. http://dx.doi.org/10.1289/ehp.1002529
To mas. 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.
This document is a draft for review purposes only and does not constitute Agency policy.
R-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
39
40
41
42
43
44
45
46
47
48
49
Supplem en tal Information—Trim ethylbenzenes
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.
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/S0045-6535f99100467-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 of
kerosene 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. 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.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchQuery=PB88196019
U.S. EPA (U.S. Environmental Protection Agency). (1994a). Chemical summary for 1,2,4-
trimethylbenzene [EPA Report], (EPA/749/F-94/022A). Washington, DC.
http://www.epa.gOv/chemfact/s trimettxt
U.S. EPA (U.S. Environmental Protection Agency). (1994b). Methods for derivation of inhalation
reference concentrations (RfCs) and application of inhalation dosimetry [EPA Report],
(EPA/600/8-90/066F). Washington, DC: U.S. Environmental Protection Agency, Office of
Research and Development, Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993
U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and
reference concentration processes. (EPA/630/P-02/002F). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
http://www.epa.gov/osa/review-reference-dose-and-reference-concentration-processes
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment
[EPA Report], (EPA/630/P-03/001F). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum, http://www2.epa.gov/osa/guidelines-carcinogen-risk-
assessment
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/opptinti7aegl/pubs/123 %20124 %20135 trimethylbenzenes %20i
nterim 11 2007.vl.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2012). Benchmark dose technical guidance.
(EPA/100/R-12/001). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum, http: //www.epa.gov/raf/publications/benchmarkdose.htm
U.S. EPA (U.S. Environmental Protection Agency). (2016a). PBPK data files for TMB (updated
February 2016) [PBPK], Research Triangle Park, NC.
This document is a draft for review purposes only and does not constitute Agency policy.
R-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
39
40
41
42
43
44
45
Supplem en tal Information—Trim ethylbenzenes
U.S. EPA (U.S. Environmental Protection Agency). (2016b). Results of the BMD analyses for 1,2,4-
TMB, 1,2,3-TMB, and 1,3,5-TMB [BMDS], Research Triangle Park, NC.
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.
http://ofmpub.epa.gov/eims/eimscomm.getfile7p download id=517403
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 InstMaritTrop Med Gdynia 26: 303-313.
Wiglusz. R: Kienitz. M: Delag. G: Galuszko. E: Mikulski. P. (1975b). Peripheral blood of mesitylene
vapour treated rats. Bull InstMaritTrop 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
Win-Shwe. TT: Fuiimaki. H. (2010). Neurotoxicity of toluene [Review], Toxicol Lett 198: 93-99.
http://dx.doi.Org/10.1016/i.toxlet.2010.06.022
Win-Shwe. TT: Kunugita. N: Yoshida. Y: Nakaiima. D: Tsukahara. S: Fuiimaki. H. (2012). Differential
mRNA expression of neuroimmune markers in the hippocampus of infant mice following
toluene exposure during brain developmental period. J Appl Toxicol 32: 126-134.
http://dx.doi.org/10.1002/iaLl643
Win-Shwe. TT: Yoshida. Y: Kunugita. N: Tsukahara. S: Fuiimaki. H. (2010). Does early life toluene
exposure alter the expression of NMDA receptor subunits and signal transduction pathway
in infantmouse hippocampus? Neurotoxicology 31: 647-653.
http://dx.doi.Org/10.1016/i.neuro.2010.08.006
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 CIO aliphatic,
aromatic and naphthenic hydrocarbons in rat after repeated exposures. Pharmacol Toxicol
71: 144-149. http: //dx.doi.org/10.1 1 1 1 /i.l 600-0773.1992.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.org/10.1 1 1 1 /i.l600-0773.1990.tb00859.x
This document is a draft for review purposes only and does not constitute Agency policy.
R-8 DRAFT—DO NOT CITE OR QUOTE
------- |