Peer Review Report
External Peer Review of the
1995 Koch Industries Study Report
90-Day Oral Gavage Toxicity Study of
1,3,5- Trimethyl benzene in Rats with a Recovery Group
April 29,2013
Peer Reviewers:
M. Christopher Newland, Ph.D.
Carol S. Wood, Ph.D., DABT
Raymond G. York, Ph.D., DABT, ATS, ERT
Contract No. EP-C-12-045
Task Order 11
* A •
I#
VPRO^°
Prepared for:
J. Allen Davis, MSPH & John Cowden, Ph.D.
U.S. Environmental Protection Agency
National Center for Environmental Assessment
109 T.W. Alexander Drive (B242-01)
Research Triangle Park, NC 27711

VERSAR
Prepared by:
Versar, Inc.
6850 Versar Center
Springfield, VA 22151

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
CONTENTS
I.	INTRODUCTION	 1
II.	CHARGE TO REVIEWERS	2
III.	INDIVIDUAL REVIEWER COMMENTS	3
Review by M. Christopher Newland	4
Review by Carol S. Wood	11
Review by Raymond G. York	 15
l

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
I. INTRODUCTION
The U.S. Environmental Protection Agency's (EPA) National Center for Environmental
Assessment (NCEA) is currently developing a human health assessment of trimethylbenzenes
(CAS No. 25551-13-7, 95-63-6, 526-73-8, and 108-67-8). In the 1990s, IIT Research Institute
performed a 90-day oral gavage toxicity study of 1,3,5-trimethylbenzene (1,3,5-TMB) on the
behalf of Koch Industries, Inc. The results of this subchronic, oral toxicity study were submitted
to EPA in June 1995. The 1995 Koch Industries report, "90-Day Oral Gavage Toxicity Study of
1,3,5 Trimethylbenzene in Rats with a Recovery Group," is a potential principal or influential
study for the IRIS assessment of trimethylbenzenes that is currently under development.
However, this report has not been subjected to a formal peer review process. Such a peer review
process is important in establishing the appropriateness, validity, and robustness of the study
design, conduct, and interpretation of the reported findings. The purpose of the requested letter
review is for EPA to receive written comments from individual experts.
Versar selected three senior scientists with expertise in the following disciplines to serve as peer
reviewers: (1) neurotoxicology, (2) human health risk assessment, and (3) general laboratory
animal toxicology studies
Peer Reviewers:
M. Christopher Newland, Ph.D.
Auburn University
Auburn, Alabama 36849
Carol S. Wood, Ph.D., DABT
Oak Ridge National Laboratory
Oak Ridge, TN 37831
Raymond G. York, Ph.D., DABT, ATS, ERT
R.G. York and Associates, LLC
Manlius, NY 13104
1

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
II. CHARGE TO REVIEWERS
The IRIS Program has a strong preference for use of peer-reviewed, published studies as
principal or influential studies. Such a peer review process is important to establishing the
appropriateness, validity, and robustness of the study design, conduct, and interpretation of
findings of the reported investigation. The purpose of the requested peer review is for EPA to
receive comments from individual experts. It is important that selected outside experts evaluate
the accuracy of the content and interpretation of the findings presented in this report.
Charge Questions:
1.	Study Design - Based on your knowledge of toxicological protocols, please comment on the
experimental design of the 90-day oral gavage toxicity study described in the Koch Industries
report.
a.	Please comment on any significant issues with the test system or test article employed,
controls employed, endpoints recorded, terminal procedures, statistical analyses, and quality
assurance?
b.	In consideration of the toxicological properties of trimethylbenzenes reported in the
provided contextual references (Wiaderna et al., 2002; Gralewicz and Wiaderna, 2001;
Korsak et al., 2000a, b; Wiaderna et al., 1998; Gralewicz et al., 1997a; Gralewicz et al.,
1997b; Korsak et al., 1997; Korsak and Rydzynski, 1996; Korsak et al., 1995), please
comment on whether there are key physiological/toxicological endpoints that should have
been assessed that were not part of the investigation.
2.	Study Results - Please comment on the strength, credibility, and relevance of the
toxicological results of the Koch Industries study.
3.	Study Conclusions - Please comment on the discussion and conclusion sections of the Koch
Industries report. Were there critical results or issues that were not addressed? Were there any
contradictory statements or observations made? Do you agree with the final conclusions of the
Koch Industries report?
4.	Study Reliability - Describe the reliability of the subject Koch Industries study for
consideration in the qualitative characterization of noncancer risk and quantitative derivation of
human health reference doses. Describe any major strengths or uncertainties with the study
described in this report that might preclude them from being used as consideration for derivation
of a noncancer reference dose.
2

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
III. INDIVIDUAL REVIEWER COMMENTS
3

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Review by
M. Christopher Newland, Ph.D.
4

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Peer Review Comments on the 1995 Koch Industries Study Report:
90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
M. Christopher Newland, Ph.D.
Auburn University
April 22, 2013	
I.	GENERAL IMPRESSIONS
This 1995 report described a study designed to examine the effects of chronic (90-day) exposure
to a single timethylbenzene compound, 1,3,5-trimethylbenzene (mesitylene). Multiple doses
representing a broad range of exposures were employed. The study appears to have followed the
design as specified by the sponsor (Koch Industries) as described in the first appendix. The
report is clearly and succinctly written. The effects reported in the summary table accurately
reflect the results as presented in the detailed tables. Some minor protocol deviations were
reported, but these are not judged to present a challenge to the overall conclusion and their
reporting is consistent with GLP requirements.
The test compound was analyzed at weekly intervals throughout the study and the actual
concentrations of the corn-oil solution prepared for dosing was within 10% of the specified dose.
The purpose of the report is to provide data to support a NOEL for mesitylene. While the
summary and conclusions are strictly accurate in describing where there were, and were not,
effects of exposure, the overall design cannot support a NOEL for human exposure to
mesitylene, in my opinion. There are two principle reasons for this conclusion. First, the route of
administration, oral gavage, is not the principle route of human exposure and relating the oral
route to the inhalation route is difficult under the best of circumstances. This difficulty is
compounded by the absence of data relating biomarkers of exposure (e.g., blood concentrations)
between the two routes. The second reason is that the effect-markers presented are insensitive to
exposure. There is a substantial peer-reviewed literature linking trimethylbenzene (including
mesitylene) exposure to functional deficits involving behavior, electrophysiology, and
respiratory function. This literature provides strong evidence that effects are readily detected on
these important functions during chronic, low-level exposure, many effects persist for many
weeks after exposure has ended, and these effects appear at concentrations below those that
affect body weight gain, important clinical signs, blood chemistry, or organ weights, markers
used in the Koch Industries report. Thus, the "NOEL" identified in the report is a "no effect"
level only because insensitive endpoints were investigated. If more sensitive measures had been
taken then the NOEL would certainly be much lower.
II.	RESPONSE TO CHARGE QUESTIONS
1. Study Design - Based on your knowledge of toxicologicalprotocols, please comment on the
experimental design of the 90-day oral gavage toxicity study described in the Koch
Industries report.
This was a multi-dose study conducted under GLP standards with doses ranging from 50 to
600 mg/kg/day, plus a corn-oil control. Dosing was accomplished daily, five days/week, as a
5

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
single bolus delivered intragastrically. With the 5 day/week dosing regimen there was a two-
day washout period every weekend. The range of doses did include an exposure level that
produced detectable toxicity using the some of the methods employed in this study.
The route of administration was oral gavage in a constant volume. This route of
administration is of questionable relevance since human exposure would be via inhalation.
The absence of blood concentrations, pharmacokinetic data, or citations to studies containing
such data makes it difficult to compare these doses to those likely to be experienced by
humans.
There were 10 males and 10 females in each group. No power analysis was provided, so it is
difficult to ascertain whether this was an adequate sample size to detect effects of many of
the endpoints used. An additional high-dose (600 mg/kg/day) group was examined 28 days
after dosing ended in an effort to detect long-lasting effects. The effects detected at the end of
exposure had largely disappeared by this follow-up test.
A single species, the Sprague Dawley rat, was used as a test subject. The animals were fed a
standard chow diet and purified drinking water. They were housed in wire-bottom cages,
which are thought to produce some stress on the animal due to lesions on the bottom of the
feet; these cages are no longer used (at least in most academic laboratories).
The storage conditions of the dosing solution were specified, except it was not stated whether
it was covered (trimethylbenzenes are volatile). It can probably be assumed that it was
covered and data indicate that an adequate concentration was always present.
a. Please comment on any significant issues with the test system or test article employed,
controls employed, endpoints recorded, terminal procedures, statistical analyses, and
quality assurance?
The Sprague Dawley rat is a widely used experimental model. I cannot comment specifically
on its use in testing laboratories in the 90s but I know of nothing to suggest that this would be
an inappropriate model. A stronger assessment would have used multiple species. Terminal
procedures used do not raise any special concerns.
Effects that were reported include increased phosphorus levels in the blood, decreased body
weight, increased phosphorus levels in the blood, discolored inguinal fur, salivation, and
increased liver and kidney weights. These effects were not detected in the 28-day post
exposure group, so they are interpreted as being reversible.
Clinical signs, which involve subjective judgment, were apparently not conducted by an
observer blind to treatment, so the possibility that bias appeared by knowing an animal's
dose group cannot be ruled out conclusively.
Statistical analyses were conducted using ANOVA followed by Dunnett's tests to determine
which dose was effective. It is not clear whether Dunnett's tests were conducted routinely or
whether they were conducted only following a statistically significant main effect on the
6

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
ANOVA. This is an important issue since routinely conducting Dunnett's tests will
undermine the protection against false positives ("Type 1 error") provided by the ANOVA.
Statistically significant results are shown only as asterisks in the tables or narrative
statements: no F tests or degrees of freedom are provided.
Apparently no statistical analysis was conducted on the qualitative clinical observations or
pathology. The narrative notes that 18 males and 13 females from the high-dose group
showed discolored inguinal fur. Since there were only 10 males and 10 females in this group,
one must conclude that some animals showed discolored fur on more than one observation,
and this is confirmed in the data tables. While these results are not confirmed with a
statistical assessment, I do not dispute their conclusion that this effect was seen only in the
high-dose groups. There was no mention of the problem of multiple comparisons, but this is
unlikely to be a concern since the effects were confined to the high-dose group, which would
be expected. There was no power analysis and no positive control for most endpoints so the
ability of this overall test system to detect known effects is difficult to ascertain.
b. In consideration of the toxicological properties of trimethylbenzenes reported in the
provided contextual references (Wiaderna et al., 2002; Gralewicz and Wiaderna, 2001;
Korsak et al, 2000a, b; Wiaderna et al, 1998; Gralewicz et al, 1997a; Gralewicz et al,
1997b; Korsak et al, 1997; Korsak andRydzynski, 1996; Korsak et al, 1995), please
comment on whether there are key physiological/toxicological endpoints that should have
been assessed that were not part of the investigation.
A series of studies conducted by scientists at the Nofer Institute of Occupational Medicine
has been published in peer-reviewed journals in the toxicology literature. These provide a
comprehensive neurotoxicological profile of three trimethylbenzenes, including mesitylene,
when administered by inhalation. The other two trimethylbenzenes examined were
pseudocumene (1,2,4-trimethylbenzene) and hemimellitene (1,2,3-trimethylbenzene).
The exposure levels used in these studies did not produce changes in food consumption, body
weight, body weight gain, and lethality, and no changes were reported on "toxicologically
significant" clinical signs. Specific effects on fur discoloration are unknown since this was
not mentioned explicitly. The effects of acute, subchronic (30 day), chronic (90 day), and
post chronic (one to two months after chronic exposure ended) were assessed. Therefore,
effects of acute and chronic exposure, as well as irreversible effects of chronic exposure,
were described. Several functional domains, including motor (rotarod), pain sensitivity (paw-
lick latency on a hot plate), overall activity, cognitive (radial arm maze and passive/active
avoidance), electrophysiological (EEG), and respiratory endpoints were examined.
A broad range of effects were noted on several functional domains with the
trimethylbenzenes. Mesitylene was examined in some, but not all of these studies. The three
trimethylbenzenes differed somewhat in potency but the effects profile was generally similar
for these three compounds. Therefore, it is possible draw reasonable general inferences about
mesitylene from the effects seen with the other two trimethylbenzenes.
Overall, these three trimethylbenzenes altered rotarod performance, impaired passive and
active avoidance, produced deficits on the rotarod, and caused significant respiratory effects.
7

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
There were effects, though less consistently, on radial arm maze performance. Many of these
effects were irreversible when tested one to two months after exposure ended. As noted
(Korsak and Rydzynski, 1996), trimethylbenzene's toxicity resembles that of other organic
solvents such as toluene, but it is more potent than toluene. This is of interest here because
the extensive literature on toluene suggests that even more refined behavioral tests detect
effects of inhaled toluene at concentrations that are similar to those reported here, or even
lower. This is important because it suggests that the NOEL or LOEL for trimethylbenzenes
might even be lower than in these peer-reviewed studies with more advanced testing.
In these studies, mesitylene in particular produced the following effects:
•	Deficits on rotarod after acute exposure (EC50 = 963 ppm) (Korsak and Rydzynski,
1996).
•	Deficit on hotplate after acute exposure (EC50 = 1212 ppm) (Korsak and Rydzynski,
1996).
•	Decreased respiration rate after acute exposure (EC50 = 519 ppm)(Korsak et al.,
1997).
•	Passive avoidance after 30 day exposure to 25 ppm, and higher concentrations
(Wiaderna et al., 2002).
•	Diminished pain sensitivity after 30 day exposure to 100 ppm (Wiaderna et al., 2002).
•	Changes in open field activity 25 days after 4 weeks of exposure to 100 ppm
(Gralewicz and Wiaderna, 2001).
•	Deficits in passive avoidance 39 to 40 days after 4 weeks of exposure to 100 ppm
(Gralewicz and Wiadera, 2001)
•	Decreased pain sensitivity 50 to 51 days after 4 weeks of exposure to 100 ppm.
(Graleeicz and Wiaderna, 2001)
•	Impaired acquisition of active avoidance 54 to 60 days after 4 weeks of exposure to
100 ppm. (Graleeicz and Wiaderna, 2001)
These studies are difficult to compare directly with the Koch report because of differences in
the dosing regimen and, in particular, the route of exposure, but some points can be noted:
•	The exposure levels used in the Nofer Institute studies did not change body weight or
body weight gain; yet consistent and reproducible effects were detected in functional
domains, including behavioral, electrophysiological, and respiratory endpoints.
•	The route of administration in the Nofer Institute studies is more directly relevant to
human exposure than that used in the Koch study.
8

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
• The route of administration used in the Nofer Institute studies likely produces a fairly
stable blood concentration throughout exposure, whereas the oral gavage studies used
in the Koch study would likely produce a brief, high blood concentration followed
quickly by a return to baseline levels. This inference is not based on data provide in
any of the studies reviewed but instead is based on a general understanding of the
pharmacokinetics following inhalation or gavage administration of organic solvents.
2.	Study Results - Please comment on the strength, credibility, and relevance of the
toxicological results of the Koch Industries study.
The Koch Industries study appears to have been competently conducted and the conclusions
are supported by the data presented. However, the relevance of the results to an evaluation of
human exposure is highly questionable. This is because the route of administration is likely
to produce a kinetic profile that is quite different from the chronic, low-level exposure
experienced by humans. In addition, the toxicological endpoints selected for study are
generally insensitive ones. No relevant behavioral, respiratory, or electrophysiological
endpoints were examined. The pathology tests performed were generally of a gross nature so
would be unlikely to detect effects of chronic low-level exposure.
3.	Study Conclusions - Please comment on the discussion and conclusion sections of the
Koch Industries report. Were there critical results or issues that were not addressed? Were
there any contradictory statements or observations made? Do you agree with the final
conclusions of the Koch Industries report?
The discussion was strictly limited to the results of the study and the discussion described the
results accurately. It did not identify limitations such as the absence of a power analysis or
the statistical analysis of the qualitative endpoints. There was no attempt to link these results
to a broader (and extensive) literature on other organic solvents. In fairness, it must be noted
that there was little literature on trimethylbenzenes at the time that this study was conducted,
so it would have been difficult to relate the study under review to a broader literature on
trimethylbenzenes in particular. However, there was an extensive literature on other organic
solvents that could have been used for comparison.
I cannot disagree with the narrow conclusion that there is a NOEL of 200 mg/kg based
narrowly on the studies reported. However, if endpoints more pertinent to human health and
a more relevant dosing regimen were employed, then I am very confident that a lower NOEL
would have been detected.
In addition, the statistical analysis employed and overall strategy relies on the detection of a
NOAEL. As has been noted many times, this approach is of questionable value because an
experiment can easily be designed to produce a NOAEL by making it underpowered (using
too few subjects) or by using insensitive endpoints. The latter seems to be at issue here.
Alternatively, had a benchmark dosing analysis been attempted, a different conclusion might
also have been reached.
9

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
4. Study Reliability — Describe the reliability of the subject Koch Industries study for
consideration in the qualitative characterization of noncancer risk and quantitative
derivation of human health reference doses. Describe any major strengths or uncertainties
with the study described in this report that might preclude them from being used as
consideration for derivation of a noncancer reference dose.
The study seems to be reliable as far as it goes. The implementation of the study appears to
have been competently performed and the documentation is extensive. However, its validity
(the ability to predict human toxicity) is questionable. I would have significant misgivings
about using this study as a basis for the derivation of a noncancer reference dose.
•	The dosing regimen and route of administration are of questionable relevance.
•	No data on blood concentrations of mesitylene are presented.
•	The endpoints used are insensitive.
•	There is an extensive and reproducible peer-reviewed literature available that can be
used to identify a LOEL or a NOEL (probably a LOEL). That literature would
identify such a level in units (air concentration) that are directly relevant to human
exposure and using more sensitive and relevant endpoints. While it can be noted that
there are flaws in that literature, the strengths for identifying a LOEL or NOEL from
the peer-reviewed literature far outweigh the weaknesses, especially in comparison
with the Koch Industries report.
III. SPECIFIC OBSERVATIONS
All of my comments are noted above.
10

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Review by
Carol S. Wood, Ph.D., DABT
11

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Peer Review Comments on the 1995 Koch Industries Study Report:
90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Carol S. Wood, Ph.D., DABT
Oak Ridge National Laboratory
April 18, 2013	
I.	GENERAL IMPRESSIONS
The study report is well written and accurately reflects the findings in the study. This study was
conducted appropriately based on standard guidelines for a 90-day subchronic oral toxicity
study, with the exception of dose selection. Clinical signs of toxicity, including salivation and fur
staining, were observed in high-dose males and females. These clinical findings were general in
nature, mild, not observed in all high-dose animals, and not observed until after about three
weeks of dosing. Other statistically significant findings are not considered by the reviewer to be
adverse or biologically relevant. Taken together, the lack of findings indicating marked toxicity
suggests that the animals could have tolerated a higher dose and that dosing was not adequate. A
dose selection rationale was not given beyond the statement that the doses were chosen by the
sponsor.
Potential neurotoxicity was not evaluated by the oral route in this study. Considering long-
lasting, post-exposure effects observed in several of the contextual references, this endpoint is
critical to evaluating the full toxicity of TMB by any route. Delayed onset of the clinical signs
observed in the Koch Industries study supports potential long-lasting effects of TMB that were
not evaluated. Similar to the current study, the standard subchronic inhalation studies (Korsak et
al., 2000a,b) found very few effects when neurotoxicity endpoints were not evaluated. Based on
the known neurotoxic potential of TMB, shown conclusively by the inhalation route, the clinical
signs were too general to be predictive of neurotoxicity. The overt clinical signs were not
observed during the recovery period, but this does not mean more subtle neurotoxic effects did
not occur.
This reviewer does not think that the Koch Industries study is reliable for assessing noncancer
risk because the endpoint of concern for TMB exposure, neurotoxicity, was not evaluated.
II.	RESPONSE TO CHARGE QUESTIONS
1. Study Design - Based on your knowledge of toxicologicalprotocols, please comment on
the experimental design of the 90-day oral gavage toxicity study described in the Koch
Industries report.
As conducted, the study followed a standard accepted protocol for a 90-day oral toxicity
study. The experimental design was adequate for a subchronic study with the exception of
dose selection, which may not have been adequate.
12

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
a.	Please comment on any significant issues with the test system or test article employed,
controls employed, endpoints recorded, terminal procedures, statistical analyses, and
quality assurance?
No problems with the study conducted as a subchronic oral toxicity study.
b.	In consideration of the toxicological properties of trimethylbenzenes reported in the
provided contextual references (Wiaderna et al., 2002; Gralewicz and Wiaderna, 2001;
Korsak et al., 2000a, b; Wiaderna et al., 1998; Gralewicz et al., 1997a; Gralewicz et al.,
1997b; Korsak et al., 1997; Korsak andRydzynski, 1996; Korsak et al., 1995), please
comment on whether there are key physiological/toxicological endpoints that should have
been assessed that were not part of the investigation.
Potential neurotoxicity was not evaluated by the oral route in this study. Considering the
long-lasting effects observed in several of the contextual references, this endpoint is critical
to evaluating the full toxicity of TMB by any route. Delayed onset of the clinical signs
observed in the Koch Industries study supports potential long-lasting effects of TMB that
were not evaluated. Similar to the current study, the standard subchronic inhalation studies
(Korsak et al., 2000a,b) found very few effects when neurotoxicity endpoints were not
evaluated. The reviewer acknowledges different isomers of TMB were used in these
inhalation studies, but considers the toxicity of each to be similar.
2. Study Results - Please comment on the strength, credibility, and relevance of the
toxicological results of the Koch Industries study.
The clinical signs of toxicity in high-dose males and females were the only adverse findings
in this study. These clinical findings were general in nature, mild, and not observed in all
high-dose animals. Based on the known neurotoxic potential of TMB, shown conclusively by
the inhalation route, the clinical signs were too general to be predictive of neurotoxicity. The
main findings of salivation and fur staining could have been due to other properties of the
chemical such as irritation or bad taste. In addition, the clinical findings were not observed
until after about three weeks of dosing (earlier in a few animals) and generally persisted until
termination of dosing either by sacrifice or start of the recovery period. In several of the
contextual references, neurotoxicity was shown to be long-lasting after cessation of exposure.
The overt clinical signs were not observed during the recovery period, but this does not mean
more subtle neurotoxic effects did not occur.
Other statistically significant findings are not considered by the reviewer to be adverse or
biologically relevant. The increased relative kidney weight observed in the high-dose males
is not a direct effect of TMB administration, but was entirely due to the slightly lower final
body weight of these animals. Absolute liver weight was slightly increased in high-dose
males and significantly increased in high-dose females, but the magnitude was small and not
accompanied by histopathological correlates. Lack of hepatocellular hypertrophy suggests
that the liver weight increases were marginally adaptive and not adverse. Increased
phosphorus levels were consistent in both high-dose males and females, but the magnitude of
change was not biologically significant.
13

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Taken together, the lack of findings indicating toxicity (except clinical observations) suggests
that the animals could have tolerated a higher dose and that dosing was not adequate. A dose
selection rationale was not given beyond the statement that the doses were chosen by the
sponsor. Conversely, dosing may have been adequate, but potential neurotoxicity endpoints
were not evaluated and, therefore, could have been missed.
3.	Study Conclusions - Please comment on the discussion and conclusion sections of the
Koch Industries report. Were there critical results or issues that were not addressed? Were
there any contradictory statements or observations made? Do you agree with the final
conclusions of the Koch Industries report?
Conclusions given in the report were consistent with the results as presented and no critical
findings were omitted. This reviewer disagrees that increased relative kidney weight
observed in the high-dose males should be part of the LOAEL because this effect was
entirely due to the slightly lower final body weight of these animals. Increased phosphorus
levels were consistent in both high-dose males and females, but the magnitude of change was
not biologically significant and thus, of questionable relevance to the LOAEL. The clinical
signs observed in high-dose males and females were the best evidence that dosing had
occurred.
4.	Study Reliability - Describe the reliability of the subject Koch Industries study for
consideration in the qualitative characterization of noncancer risk and quantitative
derivation of human health reference doses. Describe any major strengths or uncertainties
with the study described in this report that might preclude them from being used as
consideration for derivation of a noncancer reference dose.
This reviewer does not think that the Koch Industries study is reliable for assessing
noncancer risk because the endpoint of concern for TMB exposure, neurotoxicity, was not
evaluated. Thus, qualitative characterization of noncancer risk cannot be assessed from this
study. A NOAEL was identified that could be quantitatively used to derive an RfD, but the
endpoint of concern may not be protected against.
II. SPECIFIC OBSERVATION
None found.
14

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Review by
Raymond G. York, Ph.D., DABT, ATS, ERT
15

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Peer Review Comments on the 1995 Koch Industries Study Report:
90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Raymond G. York, Ph.D., DABT, ATS, ERT
R.G. York and Associates, LLC
April 21, 2013	
I.	GENERAL IMPRESSIONS
The 90-day oral gavage study of 1,3,5-trimethylbenzene (TMB) in rats with a recovery group
was conducted by IIT Research Institute for Koch Industries in 1994. William Johnson, Ph.D.,
DABT, was the study director and the study was conducted according to the 1979 EPA Toxic
Substances Control Act (TSCA) testing guideline 40 CFR 798.2650, and testing as required by
CFR 799.5075 (Fed. Reg. Vol. 58 No. 216, pp. 59681-82).
In reviewing the study report, adequate numbers of animals were included in the experimental
protocol and the study, for the most part, was adequately designed, conducted, and reported.
The bulk chemical and dosing solution analyses, animal maintenance, study design (10
animals/sex/group for three dose groups plus one high dose group of 20 animals/sex, 10/sex for a
recovery group; 100 total) and dosing 5 days/week are consistent with other subchronic 90-day
studies conducted at that time. This study was conducted according to GLP regulations 40 CFR
Part 792.
The conclusions reached for this subchronic study were supported by the rendered data. The 600
mg/kg/day high dose group clinical signs (including discolored and wet inguinal fur and
salivation), cumulative decreased body weight gain in male rats, adverse clinical chemistry
parameter of increased phosphorus blood levels in male and female rats, increased absolute liver
weight for female rats, increased relative liver weights for male and female rats, and increased
relative kidney weight in male rats were all considered treatment-related. All of the treatment-
related effects were reversible, as none were present in the recovery rats after 28 days of
cessation of treatment.
II.	RESPONSE TO CHARGE QUESTIONS
1. Study Design - Based on your knowledge of toxicologicalprotocols, please comment on the
experimental design of the 90-day oral gavage toxicity study described in the Koch
Industries report.
This study was adequately conducted under the toxicity guidelines and GLPs in effect during
the mid-1990s. It employed appropriate experimental procedures, including animal model
selection of appropriate age, sex, body weight, animal identification method, acclimation
period, randomization, group size, husbandry, dose levels, dose level selection, route and
period of administration, study endpoints measured, test and control substance preparation,
characterization, storage, and sampling, as well as statistical evaluations of study results,
quality assurance inspections and compliance. There were no significant issues with the
experimental design of the study.
16

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
•	Protocol: The experimental design for the study followed the then current
EPA798.2650 Oral Toxicity Guidelines.
•	Test Article Employed: 1,3,5-Trimethylbenzene (TMB) with the required purity
of 99% and stored properly. The identity, composition, purity and stability of the
test article (TMB) were confirmed by the testing facility. The identity, purity,
composition, and stability of the vehicle (corn oil) were confirmed by the supplier
(deviation noted). Reserve samples of each were retained by the Study Sponsor.
•	Stock Dosing Solutions: Each concentration was prepared weekly and was not
sequentially diluted from one stock solution (sequential dilution has the potential
to multiple a preparation error). Aliquots from the top, middle and bottom of
dosing solutions were determined to be homogeneous and within 10% of the
mean concentration for each of the three dose levels. Concentration and stability
analyses of dosing solutions were determined to be within specifications. No
TMB was detected in any vehicle control samples.
•	Route of Exposure: No comments except it could be important to know why the
route of administration for the early 14 day study with the same test article
(Project Number L08512, Study Number 1) was oral gavage. Were the diet or
drinking water routes excluded due to palatability issues? Why was whole body
inhalation, the route used in all the Nofer Institute studies and a known human
route of exposure, not considered? The health effects of TMB toxicity were being
required to be assessed because it was a drinking water contaminant. Drinking
water and inhalation exposure (TMB is a volatile aromatic hydrogen) routes have
more human relevancy and give much different absorption, metabolism and
elimination profiles than bolus gavage administration.
•	End Points Recorded: EPA798.2650 Oral Toxicity Guidelines require a measure
of clotting potential such as clotting time or prothrombin time and should have
been measured. Platelet counts only give current coagulation status. This was
clarified in the 1998 EPA OPPTS 870.3100 and the 1998 OECD 408 Guidelines
for a 90-day oral toxicity study in rodents by requiring both a measure of clotting
time as well as platelet counts.
•	Terminal Procedures: Procedures were followed according to the protocol. The
report stated that the absence of significant ophthalmic lesions during the final
week of dosing obviated the examination of the recovery rats 28 days later. This
is not specifically defined in the guidelines and, if one of the aims of the
guidelines is to "observe delayed occurrence of toxic effects during the post-
treatment period," this omission should not have occurred.
•	Statistical Analysis: Statistical analyses were appropriate. One shortcoming was
the lack of a vehicle recovery control group for proper comparison. This has been
addressed in the 1998 EPA OPPTS 870.3100 and the 1998 OECD 408 Guidelines
for a 90-day oral toxicity study in rodents by requiring such a control group.
17

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
•	Quality Assurance: Quality assurance procedures were correct.
•	Report: All the elements required by the EPA 798.2650 Guidelines were
included.
a. Please comment on any significant issues with the test system or test article employed,
controls employed, endpoints recorded, terminal procedures, statistical analyses, and
quality assurance?
Most of the physiological and toxicological endpoints collected were standard for protocols
for this period of time. A current EPA OCSPP Harmonized Test Guideline 870.3100 study
protocol should include:
•	7-day/week exposures for a continuous 91 days (if by gavage, 5 days/week still
acceptable).
•	Functional observation battery (FOB), if the two-week repeat study had clinical
signs of depression of the CNS.
•	Complete and contemporary statistical analyses.
•	Keeping the volume of corn oil at 2 mL/kg or less may reduce the scattered and
intermittent discolored and wet inguinal fur observed.
•	Detailed clinical observations outside the home cage each week in the same
standard arena and at the same approximate time.
•	After week 11, assessment of motor activity, grip strength and reactivity to
different sensory stimuli (visual, auditory, proprioceptive).
•	Measure of clotting potential (prothrombin time or activated partial
thromboplastin).
•	Urinalysis determinations during the last week of study (appearance, volume,
osmolality or specific gravity, pH, protein, glucose and blood/blood cells) may be
indicated since the liver was a target organ in the 14-day study (L08512) and
kidney weights were increased in the 90-day study conducted.
•	The study design is not as robust as it could have been since no vehicle control
rats (extra 10/sex) were included in the core study to be retained for recovery
observations and statistical comparisons to the high dose animals. The 1998
OPPTS 870.3100 Guidelines of a 90-day oral toxicity study in rodents require an
additional satellite control group of 20 animals (10 animals of each sex) for
comparison to the high dose satellite group.
•	A dietary, drinking water, or inhalation route of exposure would have had more
human relevancy.
18

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
•	If gavage was determined to be the best route of exposure, 7 days/week
administration would have produced a more robust study since Huo et al. (1989)
demonstrated that 99% of TMB and its metabolites are eliminated in 24 hours,
leaving 48 hours of non-exposure and extended clearance every 5 days.
b. In consideration of the toxicological properties of trimethylbenzenes reported in the
provided contextual references (Wiaderna et al, 2002; Gralewicz and Wiaderna, 2001;
Korsak et al, 2000a, b; Wiaderna et al, 1998; Gralewicz et al, 1997a; Gralewicz et al,
1997b; Korsak et al, 1997; Korsak andRydzynski, 1996; Korsak et al, 1995), please
comment on whether there are key physiological/toxicological endpoints that should have
been assessed that were not part of the investigation.
The Contextual Reference articles provided convincing evidence that 4 or 13 week inhalation
exposures at low-levels to TMB isomers results in: a) long-lasting neurobehavioral
alterations in rats several weeks after exposure to low levels of the test article, b) neurotoxic
effects occurring in a nonlinear concentration-effect relationship, and c) an altered emotional
fear response.
However, there is difficulty in considering these studies as key in the development of the
IRIS assessment of TMB:
•	The exposure route for all the Contextual References was whole body inhalation
at non-toxic levels, making it difficult to compare to any bolus oral gavage study.
•	The Contextual Reference studies were all conducted at Nofer Institute of
Occupational Medicine, Lodz, Poland. These were research studies and did not
follow standardized USEPA or OECD guideline protocols.
•	The studies were not conducted under any kind of GLPs and used non-validated
methods and equipment.
•	There were no negative or positive historical control data for adequate
comparisons.
•	The investigators measured an emotional fear response by using classical
conditioning. Usually a relatively neutral stimulus is associated with an
unconditioned stimulus. To measure an emotional fear response, they associated a
painful stimulus with a fear-inducing experience. As a result, the formerly neutral
stimulus elicits fear. For example, if seeing a dog (a neutral stimulus) is paired
with the pain of being bitten by the dog (unconditioned stimulus), seeing a dog
may become a conditioned stimulus that elicits fear (conditioned response). In
the Nofer studies, a persistent foot-shock (non-neutral stimulus) was used to
measure the paw-lick response to heat from a hot-plate test (unconditioned
stimulus). This may not be a valid or equivalent paradigm usage for classical
conditioning.
19

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
One more reference could be added for completeness: "Contrasting effects of 4-week
inhalation exposure to pseudocumene or hemimellitene on sensitivity to amphetamine and
propensity to amphetamine sensitization in the rat." Lutz, P., Gralewicy, S.,Wiaderna, D.,
Swiercz, R., Grzelinska, Z., Majcherek, W. Inter. J. of Occup. Med, and Environ. Hlth. 2010;
23(l):85-94. The study used two of the isomers of TMB, exposed rats by the same route,
length of time, and the research was conducted at the same institute. The institution's
previous work with TMB demonstrated long-lasting changes in behavior, so they were
interested in determining if TMB exposure increased behavioral sensitivity to amphetamine.
2. Study Results - Please comment on the strength, credibility, and relevance of the
toxicological results of the Koch Industries study.
Ophthalmology. Ophthalmology exams were not conducted on the recovery rats 28 days
post-exposure, because the absence of significant ophthalmic lesions during the final week of
dosing obviated the examination. This is not specifically defined in the guidelines. However,
as one of the aims of the guidelines is to "observe delayed occurrence of toxic effects during
the post-treatment period," this omission should not have occurred.
Body Weights. Fasted terminal body weights for the male rats in all the exposure groups
were reduced, albeit none significantly (584g, 576g, 562g vs. 602g for the low, mid high vs.
controls, respectively), and should have been mentioned. A trend analysis would most likely
have been significant.
Clinical Chemistry. Multiple clinical chemistry endpoints for the male rats in the 600
mg/kg/day dose group were significantly altered at the day 30 interim blood draw and
reported. However, the study director did not considered them to be related to treatment
because: a) the same parameters were not significantly altered at the day 90 termination
(which allows time for more complete enzyme adaptation to TMB exposure) and b) the
values were within or approximately within (emphasis added) the range of the in-house
historical control values. It needs to be noted that the in-house historical data do not give the
dates of the information collected, or the sex (male, female or combined) of the rats. The
significant increase in alkaline phosphate (ALK P) for the male rats in the 600 mg/kg/day
dose group at termination was considered by the study director to be due to rats (61 and 63)
that had exceptionally high values (257 and 213 IU/I). However, ALK P was increased
almost 41% for the male rats (albeit not significantly) at the day 30 interim blood draw and it
was the same two rats that had the highest levels. It is not clear why the significant increase
in ALK P for the male rats in the 600 mg/kg/day dose group at termination was not included
in determination of the NOEL by the study director.
Hematology. Mean monocyte counts (MONO) were significantly increased in the male rats
at 200 and 600 mg/kg/day when compared to the vehicle control group at termination of
exposure, but not considered by the study director as treatment-related. No reason was given.
An increased number of monocytes in the blood (monocytosis) occur in response to chronic
infections, in autoimmune disorders, in blood disorders, and in certain cancers. An
explanation on why these findings were being excluded should have been included in the
results section.
20

-------
External Peer Review of 90-Day Oral Gavage Toxicity Study of 1,3,5-Trimethylbenzene in Rats with a Recovery Group
Organ Weights. Male relative kidney weights in the 600 mg/kg/day dose group were noted
as being significantly increased; female relative kidney weights were also increased (7%) at
this dose level, albeit not significantly, but should have been mentioned in the results section.
3.	Study Conclusions - Please comment on the discussion and conclusion sections of the
Koch Industries report. Were there critical results or issues that were not addressed? Were
there any contradictory statements or observations made? Do you agree with the final
conclusions of the Koch Industries report?
The increased phosphorus levels in the 600 mg/kg/day dose group for the male and female
rats should have been discussed in more detail. The kidneys excrete phosphate; therefore, the
most common cause of increased phosphate levels (or hyperphosphatemia) is the kidney's
inability to eliminate phosphate. This may have been related to the significantly reduced
blood urea nitrogen (BUN) levels for female rats in the 600 mg/kg/day dose group.
The conclusions reached for this subchronic study were supported by the rendered data. The
600 mg/kg/day high dose group showed clinical signs (including discolored and wet inguinal
fur and salivation), cumulative decrease in body weight gain in male rats, adverse clinical
chemistry parameter of increased phosphorus blood levels in male and female rats, increase
in absolute liver weight for female rats, increased relative liver weights for male and female
rats, and increased relative kidney weight in male rats. In the last line of the Summary, the
adverse clinical signs in the 600 mg/kg/day dose group should have been included in the
basis for establishing the observed toxicity level.
4.	Study Reliability - Describe the reliability of the subject Koch Industries study for
consideration in the qualitative characterization of noncancer risk and quantitative
derivation of human health reference doses. Describe any major strengths or uncertainties
with the study described in this report that might preclude them from being used as
consideration for derivation of a noncancer reference dose.
The conclusions reached in this 90-day oral gavage study of 1,3,5-trimethylbenzene (TMB)
in rats with a recovery group study were supported by the data. The adverse clinical signs,
increase in blood phosphorus levels, and increase in liver and kidney weights in the 600
mg/kg/day dose group were clearly caused by the test substance and are the lowest-observed-
effect-level (LOAEL). The 200 mg/kg/day dose group was established as the no-observed-
adverse-level (NOAEL) in this study.
III. SPECIFIC OBSERVATIONS
Page
Paragraph
Comment or Question
86
Appendix 4
Control female rat # 14 was observed to have a swollen limb on day 92
(pathology confirmed a fractured tibia), which may account for the high
platelet (p 126; Appendix 14) and monocyte counts (p. 135; Appendix
16) at termination and should have been deleted from analysis.
21

-------