EPA 600/R-04/123
February, 2005
Final
An Evaluation of the Human Carcinogenic Potential of
Ethylene Glycol Butyl Ether
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
-------
TABLE OF CONTENTS
AUTHORS, CONTRIBUTORS, AND REVIEWERS iv
EXECUTIVE SUMMARY 1
ATTACHMENT 1: Forestomach Tumors in Female Mice Al-1
ATTACHMENT 2: Liver Hemangiosarcoma and Hepatocellular Carcinoma in
Male Mice A2-1
ATTACHMENT 3: Benchmark Dose Assessment of Forestomach Lesions in
Female Mice Using PBPK Models to Estimate Human
Equivalent Exposures A3-1
ATTACHMENT 4: External Peer ReviewsSummary of Comments and Disposition . . A4-1
11
-------
AUTHORS AND REVIEWERS
Primary Author
Jeff Gift, Ph.D.
U.S. EPA National Center for Environmental Assessment
Internal EPA Reviewers
Jennifer Jinot, Ph.D.
U.S. EPA National Center for Environmental Assessment
Jean Parker, Ph.D.
U.S. EPA Office of Prevention, Pesticides and Toxic Substances
Paul White, Ph.D.
U.S. EPA National Center for Environmental Assessment
John Lipscomb, Ph.D.
U.S. EPA National Center for Environmental Assessment
Larry Valcovic, Ph.D.
U.S. EPA National Center for Environmental Assessment
Michel Stevens, Ph.D.
U.S. EPA National Center for Environmental Assessment
External Peer Reviewers (2003 Letter Review)
Burhan I. Ghanayem, Ph.D.
National Institute of Environmental Health Sciences
Dale Hattis, Ph.D.
Center for Technology, Environment, and Development (CENTED)
Philip Leber, Ph.D.
The Goodyear Tire & Rubber Co.
Andrew Gale Salmon M.A., D.Phil., C.Chem., M.R.S.C.
California EPA Office of Environmental Health Hazard Assessment
Michael D. Shelby, Ph.D.
Director, NTP Center for the Evaluation of Risks to Human Reproduction
in
-------
AUTHORS AND REVIEWERS
(Cont'd)
External Peer Reviewers (2004 Panel Review)
Henry C. Pitot, M.D., Ph. D., Chair
McArdle Laboratory for Cancer Research
University of Wisconsin
Xi Haung, Ph. D.
Nelson Institute of Environmental Medicine
New York University School of Medicine
Lisa Kamendulis, Ph. D.
Indiana University School of Medicine
Hazel B. Matthews, Ph.D.
Matthews Toxicology Consulting
Abraham Nyska, DVM
National Institutes of Health
Torka Poet, Ph.D.
Pacific Northwest Laboratories
Frank Welsch, Ph. D.
Orbitox
Summaries of the external peer reviewers' comments and the disposition of their
recommendations are in Attachment 4.
IV
-------
An Evaluation of the Human Carcinogenic Potential of Ethylene Glycol Butyl Ether
EXECUTIVE SUMMARY
Since the publication of NTP's draft report (NTP, 1998) on their 2-year inhalation
bioassay of ethylene glycol butyl ether (EGBE; 2-butoxyethanol), there has been continued
discussion among scientists from government, industry, and academia concerning the human
carcinogenic potential of EGBE. NTP (1998; 2000) reported that their study results indicate no
evidence of carcinogenic activity in male F344/N rats, equivocal evidence of carcinogenic activity
in female F344/N rats based on increased combined incidence of benign and malignant
pheochromocytomas, some evidence of carcinogenic activity in male B6C3F1 mice based on
increased incidence of hemangiosarcomas of the liver, and some evidence of carcinogenic activity
in female B6C3F1 mice based on increased incidence of forestomach squamous cell papillomas
or carcinomas. The U.S. Environmental Protection Agency (EPA) IRIS (Integrated Risk
Information System) assessment (U.S. EPA, 1999a) concluded that, in accordance with the
proposed Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1996), the human
carcinogenicity of EGBE "cannot be determined at this time, but suggestive evidence exists from
rodent studies." Under the pre-existing EPA guidelines (U.S. EPA, 1986), EGBE was judged to
be a possible human carcinogen." These findings by EPA and NTP prompted investigators,
largely supported by the Glycol Ethers Panel of the American Chemistry Council, to design
research projects aimed at determining the mode of action for the formation of the forestomach
and liver tumors observed in mice. In this paper, recent findings reported in scientific
publications and meetings and EPA interim (U.S. EPA, 1999b) and draft final (U.S. EPA, 2003)
cancer guidelines are used to provide an up-to-date evaluation of the mode of action involved in
the origin of these tumors1 in mice and their human relevance.
Establishing the mode of action is critical for determining relevance to humans and for
choosing the approach most appropriate for dose-response modeling (i.e., whether to use a linear
or nonlinear approach). As is extensively discussed in the Agency's interim and draft cancer
guidelines (U.S. EPA, 1999b; 2003), in order to determine a chemical's mode of action, one must
consider the full range of key influences a chemical or its metabolites might have as an initiator
or promoter of the complex carcinogenic process. With this in mind, EGBE's role in the
formation of female mouse forestomach (Attachment 1) and male mouse liver (Attachment 2)
tumors observed following two-years of inhalation exposure (National Toxicology Program,
2000) were evaluated. These assessments are summarized below.
lrThe increased incidence of pheochromocytomas reported by NTP is not addressed here because these
tumors were reported to have been difficult to distinguish, were not statistically increased over chamber controls,
and were not a contributing factor in the IRIS assessment of human carcinogenic risk from EGBE exposure (U.S.
EPA, 1999a).
1
-------
Forestomach papillomas and carcinoma in female mice
Table 1 summarizes the dose-response data for key tumor types observed in female mice
in the NTP (2000) inhalation study of EGBE. At the 250 ppm exposure level, the 10% incidence
of squamous cell papillomas and 12% combined incidence of squamous cell papillomas or
carcinomas were significantly increased over study controls and exceeded the ranges for
historical controls of 0-2% and 0-3%, respectively. NTP (2000) reports that 8% is the highest
incidence of forestomach neoplasms that has been observed in contemporary historical controls.
NTP (2000) did not observe significant increases in forestomach papillomas and carcinomas at
other exposure levels in female mice, nor at any exposure level in male mice or either sex of rats.
Recent reviews of available in vitro and in vivo genotoxicity assays are in agreement that
EGBE is not likely to be genotoxic (Commonwealth of Australia, 1996; Elliot and Ashby, 1997;
U.S. EPA, 1999a; NTP, 2000). NTP (2000) suggested that EGBE caused chronic irritation
leading to forestomach injury including penetrating ulcers and that the observed "neoplasia
[papillomas and one carcinoma] was associated with a continuation of the injury/degeneration
process." Table 2 provides a summary of the strength of the evidence and the relevance to
humans of this nongenotoxic and nonlinear mode of action for EGBE's role in the formation of
these forestomach papillomas and carcinomas (Cantox, 2000; NTP, 2000; Green et al., 2002; Poet
et. al., 2003).
The Agency believes that a nonlinear mode of action similar to that which is represented
in Table 2 and described further in Attachment 1, is principally responsible for the increased
incidence of female mouse forestomach tumors reported by NTP (2000). While the steps
involved in the true mechanism of action may differ somewhat from those described, recent
pharmacokinetic and genotoxicity investigations indicate that a linear mode of action requiring
direct interaction of EGBE or an EGBE metabolite with cellular DNA is unlikely (see discussion
in Attachment 1 under "Other Possible Modes of Action for Forestomach Tumor Development in
Female Mice ").
Liver Tumors in Male Mice
Table 3 summarizes the data for tumor types which were significantly increased in male
mice exposed to EGBE by NTP (2000). Particular focus has been placed on hemangiosarcomas
of the liver because this was the only tumor type that was increased over both concurrent and
historical controls and because a mode of action involving EGBE has been proposed for this
tumor (Siesky et al., 2002). Though the incidence of hepatocellular carcinomas was within the
range of historical controls for male mice, consideration was given to this tumor because the
dose-response trend is significant and because a similar mode of action has been suggested for
this tumor.
-------
Table 4 summarizes the strength of the evidence and the relevance to humans of the mode
of action described in Attachment 2 for EGBE's potential role in the formation of
hemangiosarcomas and hepatocellular carcinomas in the livers of male mice. A metabolite of
EGBE, butoxyacetic acid (BAA), has long been known to cause hemolysis in rodents (Carpenter
et al, 1956). This hemolysis leads to the accumulation of hemosiderin (iron) in phagocytic
Kupffer cells of the liver of both rats and mice (NTP, 2000). Recent research in mice and rats
indicates that the increased iron levels associated with EGBE-induced hemolysis produces liver
oxidative damage that is more severe in mice and increased DNA synthesis in both endothelial
cells and hepatocytes that is unique to mice (Siesky et al., 2002). It is hypothesized that these
events can contribute to the transformation of the endothelial cells to hemangiosarcomas and
hepatocytes to hepatocellular carcinomas in male mice. Two recent analyses of carcinogenicity
studies of B6C3F1 mice at NTP found a highly significant (p<0.001) association between liver
hemangiosarcoma and Kupffer cell hemosiderin pigmentation, particularly when pigmentation is
observed subchronically, that is limited to male mice (Nyska et al., 2004; Gift, 2005). Given the
high background rate of these two tumor types in male mice (2.9% and 24%) relative to female
mice (0.9% and 14%) and rats (0% and 0.4%; combined male and female) (NTP, 2002), it is
reasonable to hypothesize that the endothelial cells and hepatocytes in the livers of male mice are
more susceptible to oxidative stress resulting from iron buildup in local Kupffer cells. While
additional research would be informative with respect to mechanistic issues such as the relative
susceptibility of endothelial cells and hepatocytes to oxidative stress caused by the hemolytic
effects of EGBE and the apparent resistance of female mice to the development of
hemangiosarcomas despite experiencing similar hemolytic effects, there is enough evidence at
this time to support an EPA determination that events associated with hemolysis contributed to
the increased incidence of these tumors in male mice exposed to EGBE.
The Agency believes that a nonlinear mode of action similar to that which is represented
in Table 4 and described further in Attachment 2, is principally responsible for the increased
incidence of male mouse liver tumors reported by NTP (2000). While the steps involved in the
true mechanism of action may differ somewhat from those described, recent pharmacokinetic and
genotoxicity investigations indicate that a linear mode of action requiring direct interaction of
EGBE or an EGBE metabolite with cellular DNA is unlikely (see discussion in Attachment 2
under "Other Possible Modes of Action for Liver Tumor Development in Male Mice ").
Risk to Humans
Forestomach tumors in female mice - Available data establish a plausible nonlinear,
nongenotoxic mode of action for the moderate increase observed by NTP (2000) in the incidence
of forestomach tumors in female mice following chronic inhalation exposure to EGBE. EGBE
appears to be one of a group of non-genotoxic compounds that can indirectly cause forestomach
tumors through the sustained cytotoxicity and cell regeneration brought about by irritation and
breakdown of the forestomach's gastric mucosal barrier. While this mode of action may be of
-------
qualitative relevance to humans, the exposure concentrations that would be necessary to cause
hyperplastic effects and tumors in humans, if attainable, are likely to be much higher than the
concentrations necessary to cause forestomach effects in mice, primarily because humans lack a
comparable organ for storage and long term retention of EGBE. However, even if this fact is
ignored, the analysis in Attachment 3 indicates that the exposure concentrations necessary to
cause hyperplastic effects in humans would be much higher than the existing RfD and RfC for
EGBE. Given these considerations, it appears reasonable to assume that the RfC and RfD
developed for EGBE (EPA, 1999a) are sufficient for the prevention of hyperplasia and associate
tumors in humans, including potentially sensitive subpopulations such as children.2
Liver tumors in male mice - Available data establish a plausible nonlinear, nongenotoxic
mode of action for the moderate increase observed by NTP (2000) in the incidence of liver
tumors in male mice following chronic inhalation exposure to EGBE. The proposed mode of
action suggests that the endothelial cells and hepatocytes of male mice are sensitive to the
formation of the subject neoplasms (as evidenced by the relatively high background rate of these
tumors in male mice) and that excess iron from EGBE-induced hemolysis can result in sufficient
iron-induced oxidative stress to cause the observed, marginal increase in the incidence of liver
hemangiosarcomas and hepatocellular carcinomas in these animals (NTP, 2000). Given the
relatively low sensitivity of humans, including subpopulations such as children, to the hemolytic
effects of EGBE, it appears reasonable to assume that the EGBE RfC and RfD (EPA, 1999a) are
sufficient for the prevention of hemolysis and associate tumors in humans.3
Conclusion Concerning EGBE's Cancer Risk - Information available to the Agency at this
time indicate that nonlinear modes of action are likely responsible for the increased incidence of
tumors observed by NTP (2000) in mice following chronic EGBE exposure. Application of
nonlinear quantitative assessment methods indicate that the noncancer RfD (0.5 mg/kg/day) and
RfC (13 mg/m3) values developed for EGBE (EPA, 1999a) are adequately protective of these
carcinogenic effects.
These analyses are consistent with the nonlinear assessment approach described in existing interim (U.S.
EPA, 1999a) and draft (2003) cancer guidelines.
3These analyses are consistent with the nonlinear assessment approach described in existing interim (U.S.
EPA, 1999a) and draft (2003) cancer guidelines.
-------
Table 1 - Key Tumors Observed in Female Mice Exposed to EGBE (NTP, 2000)
Squamous Cell Papilloma - Forestomach
Overall rate
Rate adjusted for intercurrent mortality4
First incidence (days)
Squamous Cell Papilloma or Carcinoma - Forestomach
Overall rate
Rate adjusted for intercurrent mortality
First incidence (days)
Control
0/50
0%
0%
NA
0/50
0%
0%
NA
62.5
ppm
1/50
2%
2.4%
731
1/50
2%
2.4%
731
125 ppm
2/50
4%
4.8%
731
2/50
4%
4.8%
731
250 ppm
5/50
10%
11.2%
582
6/50
12%
13.4%
582
Poly-3 test estimate taken from NTP, 2000 technical report series 484
5
-------
Table 2 - Concordance Table Showing the Relationship of Proposed Mode of Action for Formation of
Forestomach Tumors in Female Mice to Humans
Event
1. Deposition of EGBE/BAA
in stomach and forestomach
via consumption or
reingestion of EGBE laden
mucous, salivary excretions
and fur material
2. Retention of EGBE/BAA
in food particles of the
forestomach long after being
cleared from other organs
3. Metabolism of EGBE to
BAA systemically and in
forestomach
4. Irritation of target cells
leading to hyperplasia and
ulceration
5. Continued injury and
degeneration leading to high
cell proliferation and turnover
6. High cell proliferation
and turnover leads to clonal
growth of initiated
forestomach cells
Relation to Animal Tumors
Specificity6
Moderate7
Moderate4
High
High
High
Moderate
Dose/
Temporal
Relation
Lower/
Earlier
Lower/
Earlier
Lower/
Earlier
Lower/
Earlier
??
??
Biological
plausibility
Moderate
Moderate
High
High
High
High
Overall
Weight
of
Evidence5
Strong
Moderate
Strong
Moderate
Weak
Moderate
Qualitative
Relation to
Humans
Moderate
Low
??
Not Likely
Not Likely
Not Likely
Quantitative
Relation to
Humans
Moderate
Not Likely
??
Not Likely
Not Likely
Not Likely
Weight of Evidence is either strong, moderate or weak; a strong weight of evidence is defined as having
several studies which support the proposed mode of action, preferably from multiple laboratories with limited
evidence of contradiction. Weak evidence is normally defined as a single study from a single laboratory or with a
significant amount of contradiction in the literature between reports.
^Specificity to the proposed mode of action. Specificity is high if an event is unique to this particular mode
of action. Specificity is low if the event may be seen in many other modes of action.
7This step is confirmed by recent research in mice (Green et al., 2002; Poet et al., 2003) but has not been
investigated in rats, which do not develop forestomach tumors following EGBE exposure.
-------
Table 3 - Key Tumors Observed in Male Mice Exposed to EGBE (NTP, 2000)
Hemangiosarcomas - All organs
Overall rate
Rate adjusted for intercurrent mortality8
First Incidence (days)
Hemangiosarcomas - Liver only
Overall rate
Rate adjusted for intercurrent mortality
First Incidence (days)
Hemangiosarcomas/hemangiomas - All organs
Overall rate
Rate adjusted for intercurrent mortality
First Incidence (days)
Hepatocellular Carcinoma
Overall rate
Rate adjusted for intercurrent mortality
First Incidence (days)
Hepatocellular Adenoma or Carcinoma
Overall rate
Rate adjusted for intercurrent mortality
First Incidence (days)
Control
1/50
2%
2.2%
729
0/50
0%
0%
NA
1/50
2%
2.2%
NA
10/50
20%
20.8%
374
30/50
60%
61.9%
374
62.5 ppm
1/50
2%
2.1%
670
1/50
2%
2.1%
670
1/50
2%
2.1%
670
11/50
22%
22.9%
621
24/50
48%
48.9%
549
125 ppm
2/50
4%
5.0%
704
2/49
4%
5.0%
704
4/50
8%
10%
704
16/49
33%
35.9%
430
31/49
63%
67.5%
430
250 ppm
5/50
10%
12.4%
454
4/49
8%
10%
454
5/50
10%
12.4%
454
21/49
43%
45.9%
312
30/49
61%
64.8%
312
Poly-3 test estimate taken from NTP, 2000 technical report series 484
7
-------
Table 4 - Concordance Table Showing the Relationship of Proposed Mode of Action for Formation of
Liver Hemangiosarcomas & Hepatocellular Carcinomas in Male Mice to Humans9
Event
1. EGBE metabolism to BAA
by alcohol dehydrogenase
2. RBC hemolysis by BAA
3. Buildup of Hemosiderin
in Kupffer cells of liver
4a. Production of reactive
oxygen species by Fenton or
Haber-Weiss reactions
4b. Kupffer cells activated
and release cytokine
5. Reactive oxygen species
results in oxidative DNA
damage to endothelial cells
6. Modulation of endothelial
cell gene expression
7. Endothelial cell
proliferation
8. Promotion of initiated
endothelial cells
9. Neoplasm formation
Relation to Animal Tumors
Specificity6
High
High for
Hemolysis
Low
Low
Low
Low
Low
Low
Low
Low
Dose/
Temporal
Relation
Lower/
Earlier
Lower/
Earlier
Lower/
Earlier
Higher/
Earlier
??
Higher/
Earlier
??
??
??
??
Biological
Plausibility
Moderate
Moderate
Moderate
High
Moderate
High
Moderate
Moderate
High
High
Overall
Weight of
Evidence5
Moderate
to Strong10
Moderate
to Strong11
Moderate
to Strong12
Weak13
Weak
Weak12
Weak
Strong
Weak
Strong
Qualitative
Relation to
Humans
High
Low
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Quantitative
Relation to
Humans
Moderate
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Not Likely
Event, weight of evidence and specificity columns were adopted from Klaunig and Kamendulis (2005)
10EGBE is metabolized to BAA in rats and mice, but the tumor is only increased in male mice.
nHemolysis is observed in both sexes of rats and mice, but the tumor is only increased in male mice.
12-
Hemosiderin is observed in Kupffer cells of both sexes of rats and mice, but the tumor is only increased in
male mice. Early (subchronic) hemosiderin buildup is only observed in male mice, however.
13,
'These effects have been observed to be more pronounced in mice (Siesky et al, 2002), which are also
more susceptible to EGBE induced liver tumor formation than rats (NTP, 2000).
-------
REFERENCES
Cantox Environmental Inc. (2000) EGBE: review of the relevance of forestomach lesions in the
female B6C3F1 mouse. Final report; October.
Carpenter, CP; Pozzani, UC; Wiel, CS; et al. (1956) The toxicity of butyl cellosolve solvent. AMA
Arch Ind Health 14:114-131.
Commonwealth of Australia. (1996) 2-Butoxyethanol. Priority existing chemical number 6.
Canberra: Australian Government Publishing Service. Available:
http://www.nicnas.gov.au/publications/car/pec/pecindex.htm [27 January, 2003].
Corley, R. A.; Markham, D. A.; Banks, C.; Delorme,P.; Masterman, A.; Houle, J. M. (1997)
Physiologically-based pharmacokinetics and the dermal absorption of 2-butoxyethanol vapors by
humans. Fundam. Appl. Toxicol. 39: 120-130.
Elliot, B. M. and Ashby, J. (1997) Review of the genotoxicity of 2-butoxyethanol. Mutat. Res. 387:
89-96.
Gift, J. S. (2005) U.S. EPA's IRIS assessment of 2-butoxyethanol: the relationship of noncancer to
cancer effects. Toxicology Letters 156: 163-178.
Green, T.; Toghill, A.; Lee, R.; Moore, R.; Foster, J. (2002) The development of forestomach
tumors in the mouse following exposure to 2-butoxyethanol by inhalation. Studies on the mode of
action and relevance to humans. Toxicology 180: 257-273.
Grice, H. C. (1988) Safety evaluation of butylated hydroxyanisole from the perspective of effects
on forestomach and oesophageal squamous epithelium. Food Chem. Toxicol. 26: 717-723.
Klaunig, J. E. (2002) Mode of action of butoxyethanol for liver hemangiosarcomas in the male
mouse. Washington, DC: Presented to U.S. Environmental Protection Agency, National Center for
Environmental Assessment; December 3.
Klaunig, I.E. and Kamendulis, L.M. (2005) Mode Of Action Of Butoxyethanol Induced Mouse
Liver Hemangiosarcomas And Hepatocellular Carcinomas. Toxicol. Lett. 156(1): 107-15.
National Toxicology Program (NTP). (1998) Toxicology and carcinogenesis studies of 2-
butoxyethanol (CAS no. 111-76-2) in F344/N rats and B6C3F1 mice (inhalation studies).
Washington, DC: U.S. DHHS, National Institutes of Health. (National Toxicology Program
technical report series no. 484; NIH draft publication no. 98-3974)
National Toxicology Program (NTP). (2000) Toxicology and carcinogenesis studies of
2-butoxyethanol (CAS no. 111-76-2) in F344/N rats and B6C3F1 mice (inhalation
9
-------
studies). Washington, DC: U.S. Department of Health and Human Services, National
Institutes of Health. (National Toxicology Program technical report series no. 484).
NTP (2002). NTP Historical Control Information for the NIH-07 Diet. Available at
http://ehp.niehs.nih.gov/ntp/docs/ntp_hcrs.html. Updated 25-March, 2002.
Nyska A, Haseman JK, Kohen R, Maronpot RR (2004). Association of Liver Hemangiosarcoma
and secondary iron overload in B6C3F1 Mice - The National Toxicology Program Experience.
Toxicologic Pathology. 32:222-228.
Poet, T. S.; Soelberg, J. I; Weitz, K. K.; Mast, T. J.; Miller, R. A.; Thrall, B. D.; Corley, R. A.
(2003) Mode of action and pharmacokinetic studies of 2-butoxyethanol in the mouse with an
emphasis on forestomach dosimetry. Toxicol. Sci. 71: 176-189.
Siesky, A. M.; Kamendulis, L. M.; Klaunig, J. E. (2002) Hepatic effects of 2-
butoxyethanol in rodents. Toxicol. Sci. 70: 252-260.
U.S. Environmental Protection Agency (1986) Guidelines for carcinogen risk assessment. Federal
Register 51: 33992-34003.
U.S. Environmental Protection Agency (1996) Proposed guidelines for carcinogen risk assessment.
Washington, DC: Office of Research and Development; report no. EPA/600/P-92/003C. Available:
http://www.epa.gov/ORDAVebPubs/carcinogen/ [27 January, 2003].
U.S. Environmental Protection Agency. (1999a) Toxicological Review of ethylene
glycol monobutyl ether (EGBE) (CASRN 111-76-2) in Support of Summary
Information on Integrated Risk Information System (IRIS). National Center for
Environmental Assessment, Washington, DC. Available online from:
http://www.epa.gov/iris.
U.S. Environmental Protection Agency. (1999b) Guidelines for carcinogen risk
assessment [draft]. Washington, DC: Risk Assessment Forum; report no. NCEA-F-
0644. Available: http://www.epa.gov/ncea/raf/pdfs/cancer_gls.pdf.
U.S. Environmental Protection Agency. (2003) Draft Final Guidelines for Carcinogen Risk
Assessment (External Review Draft, February 2003). Other Ncea-F-0644a. 03 Mar 2003. U.S.
Environmental Protection Agency, Risk Assessment Forum, Washington, DC, 125 pp. Available:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55445.
10
-------
ATTACHMENT 1
Forestomach Tumors in Female Mice
Background
A significant increase over controls (experimental and historical) of papillomas and one
carcinoma of the forestomach (6/50; 12%) were reported by NTP (2000) in female mice exposed
for two years to 250 ppm EGBE by inhalation. Significant increases in forestomach papillomas and
carcinomas were not observed at other exposure levels in female mice, nor at any exposure level in
male mice or in rats of either sex. The historical range of forestomach tumors in female control
mice from NTP inhalation carcinogenicity studies averaged 0.3% (3/1,092) and ranged from 0% to
2% for carcinomas and averaged 1.6% (17/1,092) and ranged from 0% to 6% for papillomas
(Haseman et al. 1998). NTP (2000) reports that 8% is the highest incidence of forestomach
neoplasms that has been observed in any contemporary historical control group.
Mode of Action
Researchers from several laboratories, including the National Institute for Environmental
Health Studies (Ghanayem et al., 1987a,b; Ghanayem and Sullivan, 1993; Lee et al., 1998; Dill et
al., 1998; NTP, 2000), Battelle (Poet et al., 2003; Corley et al., 1994; 1997), Cantox Environmental
Inc. (Cantox, 2000), and Syngenta Central Toxicology Laboratory (Green et al., 2002; Bennette,
2001) have made significant contributions towards resolving the mode of action for the
development of forestomach tumors (papillomas and one carcinoma) observed in female mice
following chronic inhalation exposure to EGBE. The following is a seven step summary of a
proposed mode of action that is consistent with the research and reports of these authors.
1. Deposition ofEGBE/BAA in stomach and fore stomach via consumption or
reingestion of EGBE laden mucous, salivary excretions, and fur material.
2. Retention ofEGBE/BAA in food particles of the forestomach long after being cleared
from other organs.
3. Metabolism of EGBE to 2-butoxyacetaldehyde (BAL), which is rapidly metabolized
to BAA systemically and in the forestomach.
4. Irritation of target cells leading to hyperplasia and ulceration.
5. Continued injury and degeneration leading to high cell proliferation and turnover.
6. High cell proliferation and turnover leads to clonal growth of spontaneously
initiated forestomach cells.
Al-1
-------
Strength, Consistency and Specificity of Association of Tumor Response with Key Events
The proposed nonlinear MOA is consistent with the lack of direct genotoxicity that has been
demonstrated for EGBE and its metabolites. Of the steps listed above, the event with the strongest
association to the tumor response is step 5, continued injury and degeneration leading to high cell
proliferation and turnover. The following is a discussion of the strength and consistency of the
database supporting each step in the proposed MOA.
Several studies in rats and mice indicate that EGBE and its principle metabolite BAA
deposit in the forestomach and are selectively retained there following whole body inhalation (Poet
et al., 2002; Green et al., 2002) and nose-only inhalation (Poet et al., 2002), intravenous (iv) (Poet
et al., 2002; Green et al., 2002; Bennette, 2001), intraperitoneal (ip) (Corley et al., 1999; Poet et al.,
2002), subcutaneous (sc) (Corley et al., 1999) and gavage (Poet et al., 2002; Ghanayem et al.,
1987a,b; Green et al., 2002) exposures. Upon iv administration to mice, EGBE metabolites rapidly
accumulate in salivary secretions and are swallowed (Bennette, 2001; Green et al., 2002), and the
same phenomena likely occurs following inhalation and ip injection, which cause forestomach
lesions similar to those observed in gavage studies (Corley et al., 1999; Green et al. 2002).
Findings which strengthen the case for deposition through swallowing EGBE laden material and
retention of EGBE and BAA in the contents of the mouse forestomach (Steps 1 and 2) include (1)
EGBE or a metabolite rapidly distributes to the oral cavity (buccal and oesophagus) following ip, iv
and gavage dosing (Poet et al., 2003; Green et al., 2002); (2) a small but significant amount (9-10
mg/kg) of neat EGBE is available for oral consumption via daily grooming following the NTP
inhalation exposures (Corley et al., 1999; Green, 2000); (3) direct, neat exposure to EGBE without
first-pass liver metabolism can cause forestomach lesions similar to those observed by the NTP
(2000) following inhalation exposure (Corley et al., 1999); (4) the forestomach is poorly
vascularized and the cells of the epithelium are separated from capillaries by substantial diffusion
distances (Bueld and Netter, 1993; Browning et al., 1983); (5) several hours after ip dosing EGBE
levels were two orders of magnitude higher in forestomach contents than forestomach tissue, and
(6) 24 hours after ip and oral dosing levels of both EGBE and BAA remained high in forestomach
tissue, but were nondetectable in any other tissue, including blood, after 30 minutes (Poet et al.,
2003).
The metabolism of EGBE to BAA (Step 3) is well established in multiple in vivo and in
vitro tests involving both sexes of several species, including rats, mice, rabbits, guinea pigs, dogs,
monkeys, and humans (U.S. EPA, 1999). In addition, the irritation and the hyperplastic effects
observed in the forestomach following EGBE exposure are more severe when BAA is administered
directly (Green et al., 2002). Poet et al. (2003) have suggested that the small amount of food
remaining in the forestomach acts as a storage compartment for EGBE providing a continual source
for EGBE to forestomach tissue where it is locally metabolized to BAA. At 3-6 hours after ip
injection of 250 mg EGBE/kg, EGBE levels were 3-fold higher than BAA levels in stomach
contents, but an order of magnitude lower than BAA levels in forestomach tissue. At 9 hours after
ip injection, the levels of EGBE in stomach contents were reduced to approximately the same levels
Al-2
-------
as BAA, supporting the hypothesis that food stored in the forestomach serves as a source of EGBE
for forestomach tissue where EGBE is locally metabolized to BAA. Further support for the
importance of local metabolism is provided by the work of Corley (2003), who extended a
previously described EGBE PBPK model (Corley et al., 2003) to include the metabolism of EGBE
to BAL via alcohol dehydrogenase and the subsequent metabolism of BAL to BAA via aldehyde
dehydrogenase as an intermediate step in the metabolism of EGBE to BAA (Figure Al-1 and Al-
2). Using rate constants derived from mouse stomach fractions (Green et al., 2002) and making
several assumptions about the use of these enzyme activity data (see discussion below under
"Biological Plausibility and Coherence of the Database"}, Corley (2003) estimated that 250 ppm
EGBE would result in peak Cmax concentrations of 48 EGBE, 1.1 BAL and 3,200 BAA |^M in GI
tissue of female mice at the end of a 6 hour exposure period (see Figure Al-3). These estimates are
supported by a recent gavage study (see discussion below of Deisinger and Boatman, 2004).
Irritation of target cells leading to hyperplasia and ulceration (Step 4) is well documented
following EGBE exposure to both sexes of B6C3F1 mice (Poet et al., 2003; Green et al., 2002;
NTP, 2000). NTP (2000) reported a dose-related increase in epithelial hyperplasia (1/50, 7/50,
16/49 and 21/48 in males; 0/50, 6/50, 27/50 and 42/49 in females) and ulceration (1/50, 2/50, 9/49
and 3/48 in males; 1/50, 7/50, 13/49 and 22/50 in females) following chronic inhalation exposure to
0, 62.5, 125 and 250 ppm EGBE. However, NTP (2000) only observed forestomach tumors in
female mice (see Table 1 of Executive Summary).
An indication of the importance of continued damage and cell turnover (step 5) towards
tumor formation following EGBE exposure is given by the fact that tumors were only observed to
increase in female mice, which had more extensive and severe forestomach lesions than male mice
and rats (NTP (2000) observed epithelial hyperplasia and ulcers in rats, but the incidence in
exposed groups of both sex were not significantly increased over controls). Green et al. (2002)
reported that the number of cells in S-phase (an indication of cell turnover) increased in a dose
dependent fashion within dose-groups following EGBE (7.71 ± 2.50, 9.33 ± 2.55 and 12.88 ± 2.60)
and BAA (8.72 ± 4.97, 9.01 ± 2.32 and 16.22 ± 5.61) exposure at 50, 150 and 500 mg/kg, though
none of the changes were significantly increased because of the high value reported for the control
group (12.06 ± 2.41).14 The fact that similar frequencies of H-Ras gene mutations were detected in
DNA isolated from forestomach neoplasms from treated (8/14) and untreated (5/11) mice (Sill et
al., 2000) is consistent with a nongenotoxic mechanism such as increased cell turnover and does not
support a mechanism involving direct mutation by EGBE or its metabolite(s).
14The authors did not speculate or provide any reason for what they refer to as a"high value" in the control
group. However, only four or five animals were used per dose group and all but the high dose responses were within
a standard deviation of any other value.
Al-3
-------
Butoxyethanol Butoxyacetic Acid
Inhalation IT Exhalation IVInfusion
Gavage
Urine BAA-Conj & CO2
Figure Al-1: EGBE published model
(Corley et al., 2003)
Butoxyethanol Butoxyacetic Acid
BAA
EG BE-Gluc
CO2 BAA-Conj
Figure Al-2: Modified version of Corley et al. (2003) model, incorporating
BAL intermediate in liver and GI tissues
Al-4
-------
BE & Metabolites in Gl Tissues
BE & Metabolites in Gl Tissues
3,500-
3,000
f 2,500
n.
I 2,000-
1
§ 1,500
c
0
0 1,000
500
n
/\
/ \
/ \
/ \ BE
/ \ BAL
/ \ -- BAA
/ \
/ _ \_ ^
^^^ ,^^<^
12
40 -
30 -
20 -
10 -
n
r
\
\
\
i
i
1
i
i
\
\
\
V ^
12
Time(Hr)
Time (hr)
18
Cmax (uM)
BE 48
BAL 1.1
BAA 3,200
Note: Gl tissues include BE and BAA in saliva that is swallowed,
but not BE from groorr\\r\g of fur or from rr\uco-c\\\ary c\earar\ce
Figure Al-3: Concentrations of BE, BAL and BAA
in Gl tissues of female mice exposed to 250 ppm
Over the past two decades, nonlinear modes of action involving cytotoxicity and increased
cell turnover have been proposed for the carcinogenic activity of a number of other nongenotoxic
forestomach carcinogens (Ghanayem et al., 1994; Ghanayem et al., 1986; Hirose et al., 1986) and
for chemicals that cause cell proliferation in other organ systems (Cohen and Ellwein, 1990; Popp
and Marsman, 1991). Like many of these compounds, mutagenicity studies generally indicate that
EGBE is nonmutagenic with or without activation systems (see discussion under "Other Possible
Modes of Action..." below) and is not considered to be a tumor-initiator. In support of this
contention is the temporal relationship between the noncancerous forestomach lesions which
indicate persistent cell damage (i.e., epithelial hyperplasia and ulceration) and tumor formation.
Temporal Association
All of the steps in the proposed mode of action have been observed to occur in female mice
prior to tumor formation. NTP (2000) reported that female mice experienced epithelial hyperplasia
(1/10, 5/10, 9/10 and 10/10) after just 13 weeks of exposure at the same exposure levels used in the
Al-5
-------
chronic study, 0, 62.5, 125, and 250 ppm. The first reported incidence of a forestomach papilloma
or carcinoma in female mice was 731, 731 and 582 days in the 62.5, 125 and 250 ppm exposure
groups, respectively. This is consistent with the findings of Ghanayem et al. (1986; 1993; 1994),
who have investigated the temporal relationship between the induction of this type of forestomach
lesion by another nongenotoxic irritant, ethyl acrylate (EA), and the development of squamous cell
papillomas and carcinomas. They observed cell proliferation/hyperplasia in the forestomach of all
rats that received EA by gavage (200 mg/kg, 5 days/wk) for 6 or 12 months. All of these
potentially precancerous forestomach lesions regressed in animals treated with EA for 6 months
and allowed 2 or 15 months of recovery, and no forestomach neoplasms were observed. However,
treatment of rats with EA for 12 months followed by 2 months of recovery resulted in the
development of forestomach papillomas in 2 of 5 rats, and treatment of 13 rats for 12 months
followed by 9 months of recovery resulted in 1 rat developing a papilloma and 3 rats developing
carcinomas of the forestomach. This is an indication that some of these foci or lesions had already
developed into the stage of progression (Pitot, 2002). Although EA, an unsaturated aldehyde, is not
a metabolite of EGBE, it is an analog of BAL and a much more potent carcinogen (Gold et al.,
1993).
The high incidence of forestomach hyperplasia, the relatively lower incidence of papillomas
and the late occurrence of a single carcinoma in the high, 250 ppm exposure group is suggestive of
a temporal relationship and tumor progression following EGBE exposure to female mice. The
presence of a single carcinoma merely probably indicates the spontaneous transition from cells in
the stage of promotion (papilloma) to those in the stage of progression (carcinoma).
Male mice may show the beginnings of tumorogenic effects as the incidence of papillomas
was observed to increase, but not significantly over concurrent or historical controls. No
hyperplasia and no tumors were observed in inhalation studies of rats (NTP, 2000) and in drinking
water studies of mice (NTP, 1993 a), supporting the need for these steps prior to tumor formation.
Dose-Response Relationships
Forestomach tumors were only increased over controls at dose levels above those that
caused significant hyperplasia. The dose-response curve for the tumor response was nonlinear. All
key events and tumor effects depend on the dose rate.
Noncancer forestomach effects observed in EGBE exposed female mice, epithelial
hyperplasia (6/50, 27/50, 42/49, 44/50) and ulceration (1/50, 7/50, 13/49, 22/50), were dose related,
and were significantly increased over controls (concurrent and historical) at lower dose levels than
forestomach tumors were observed (Table 1). The incidence of epithelium hyperplasia and
ulceration were increased in male mice at all exposure levels, but not as severely.
Al-6
-------
Biological Plausibility and Coherence of the Database
Both genotoxic and nongenotoxic chemicals have been shown to induce forestomach
tumors in rodents (Kroes and Wester, 1986; Huff et al., 1991; NTP, 2000; Ghanayem et al, 1986;
1993, 1994). Nongenotoxic substances that cause such tumors appear to require long term contact
with the forestomach epithelium leading to irritation, cell proliferation and neoplasia. The
overstimulation of repair processes and enhancement of growth promoting factors are believed to
be involved (Harrison, 1992). Promotion and other activities associated with the stimulation of
cell proliferation have been observed for many of these compounds (Clayson et al., 1991;
Ghanayem et al., 1994). High concentrations of EGBE and its BAA metabolite sequestered in the
forestomach are assumed to be the cause of chronic irritation and the more serious damage
observed in the forestomach lining of female mice. Incidence of ulcers consisting of a defect in the
forestomach wall that penetrated the full thickness of the forestomach epithelium were significantly
increased in all exposed groups of females. NTP (2000) suggests that EGBE exposure-induced
irritation caused inflammatory and hyperplastic effects in the forestomach and that "the neoplasia
[papillomas and 1 carcinoma] was associated with a continuation of the injury/degeneration
process."
Other substances that have caused forestomach hyperplasia in male and female mice
following inhalation exposure include acetonitrile (NTP, 1996), 1,3-butadiene (NTP, 1993b) and
chloroprene (NTP, 1998). Both propionic and butyric acid have been shown to induce proliferative
responses in forestomach epithelium after only seven days, and long-term propionic acid exposure
has produced papillomas in the rat forestomach (Kroes and Wester, 1986). Since high levels of
EGBE and BAA have been observed in the stomachs of mice following iv, ip, oral gavage and
inhalation, it is apparent that the chemical partitions to the forestomach via multiple routes,
including grooming of fur, systemic blood circulation, ingestion of salivary excretions and
respiratory tract mucus and possibly repartitioning from the stomach contents (Poet et al., 2003;
Green et al., 2002).
Because the forestomach functions as a storage organ, there is a reduced requirement for
vascularization. The planar capillary network within the epithelial layers of the rodent forestomach
contrasts strongly with the thick mucosal network of capillaries in the glandular stomach of rodents
(Browning et al., 1983). The cells of the forestomach epithelium, especially the more superficial
squamous cells, are separated from capillaries by substantial diffusion distances (Bueld and Netter,
1993; Browning et al., 1983). In addition, the glandular stomach contains a complex mucosal
protection and buffering system necessary to withstand the high acidity of the digestion process.
As a result, acidic substances that concentrate in the forestomach tend to act as irritants to the
forestomach, but not to the glandular stomach or other gastrointestinal tissue (Browning, 1983;
Cantox, 2000). The reduced vasculature of the forestomach also suggests that EGBE and its
metabolites are delivered to the forestomach by ingestion rather than systemically.
Al-7
-------
While there is a significant amount of recent laboratory research that supports the proposed
mode of action, several questions remain. Regardless of whether these questions are resolved,
however, the EPA should have enough information from the current literature to make the
important determinations relative to EGBE induced forestomach tumors, their relevance to humans
and the application of a linear or nonlinear assessment.
Is BAA the toxic moiety responsible for the irritant effects of EGBE? EGBE is metabolized to
BAL via alcohol dehydrogenase, which is quickly metabolized to BAA via aldehyde
dehydrogenase (Green et al., 2002; Ghanayem et al., 1987a). While some have suggested that
some of the cytotoxic effects of EGBE may be attributable to BAL (Dartsch et al., 1999;
Ghanayem et al., 1987b), recent studies indicate that BAA is largely responsible for the irritant
effects observed in the forestomachs of mice following EGBE exposure. In a 10-day gavage
study, Green et al. (2002) showed that BAA was significantly more potent than EGBE at
inducing hyperkeratosis of the forestomach lining of female mice. EGBE induced minimal
hyperkeratosis in 0/5 and 2/5 mice at 150 and 500 mg/kg, respectively. However, BAA
exposure caused minimal hyperkeratosis in 3/5 mice at 150 mg/kg and more severe
hyperkeratosis in 4/4 mice at 500 mg/kg. In addition, in vitro studies indicate that BAA is
known to have other effects on cells that are thought to be important for tumor formation such
as altered cell membrane permeability (Udden, 2002; Ghanayem, 1989).
Available pharmacokinetic information also suggests that BAA is the principal
metabolite responsible for the irritant effects of EGBE. Using glandular and forestomach
tissue from mice and rats, Green et al. (2002) measured the kinetic constants Km and Vmax for
enzymes that metabolize EGBE to BAA (an alcohol dehydrogenase) and BAL to BAA (an
aldehyde dehydrogenase) (see Table Al-1). For both rats and mice, the rate constant for
conversion (Vmax/Km) of BAL to BAA via the aldehyde dehydrogenase was significantly higher
than the rate constant for conversion of EGBE to BAL via the alcohol dehydrogenase in rats
and mice, suggesting that far more BAA would accumulate in the stomach than BAL.
Table Al-1: Dehydrogenase enzyme activities in rat and mouse stomach fractions (Green et al. (2002)
Species/Tissue
Rat Forestomach
Rat glandular Stomach
Mouse Forestomach
Mouse glandular Stomach
Alcohol Dehydrogenase
Km
(mM)
0.29
0.73
46.59
87.01
* max
(nmol/min/mg
)
1.627
2.170
17.094
13.986
* max'
Km
5.61
2.97
0.37
0.16
Aldehyde Dehydrogenase
Km
(mM)
0.037
0.029
0.056
0.135
* max
(nmol/min/mg
)
3.738
5.624
8.576
8.950
* max'
Km
101
194
153
66.3
Al-8
-------
A recent attempt has been made to quantify the amount of EGBE/BAA/BAL that would
be present in the liver and GI tissues of female mice following a 250 ppm inhalation exposure
to EGBE (Corley, 2003) using the forestomach rate constants provided by Green et al. (2002)
(Table Al-1). This work extends an earlier model developed by Dr. Corley (Corley et al.,
2003) to include the intermediate formation of BAL in target tissues (liver and GI tract). Given
the limitations of the available data, Corley was required to make several assumptions,
including:
Rate constants for metabolism of EGBE to BAL in the forestomach of rats are similar to
in vitro publications of liver metabolism (Aasmoe et al., 1998; Johanson et al., 1986)
Forestomach rate constants apply to the entire GI tract
Forestomach rate constants apply to the liver
In vivo rate constants from current PBPK model correspond to the first, rate-limiting
step in metabolism (EGBE to BAL)
Ratio of in vitro BAL to BAA/EGBE to BAL can be used (parallelogram approach) to
estimate in vivo Vmax for BAL to BAA
BAL does not leave tissue where formed while BE and BAA circulate in the body
Given these assumptions and the rate constants derived from mouse stomach fractions
(Green et al., 2002), Corley (2003) estimated that 250 ppm EGBE would result in peak Cmax
concentrations of 48 EGBE, 1.1 BAL and 3,200 BAA |J,M in GI tissue of female mice at the end
of a 6 hour exposure period (see Figure Al-3). A recent gavage study performed by Deisinger
and Boatman (2004) provides support for the Corley (2003) model and the predicted low levels
of the BAL metabolite in GI tissue.15
Why are there no effects in the glandular stomach of rodents? If BAA is indeed the toxic
moiety, the higher dehydrogenase activity per volume of forestomach tissue versus glandular
stomach tissue (Table Al-1) could largely explain this difference in susceptibility. In addition,
the nature and the function of the forestomach must be considered. The rodent forestomach is
separated from the glandular stomach by a prominent limiting ridge. Material entering the
forestomach is stored without being digested prior to entering the glandular stomach.
Residence time in the forestomach is considerable. In fact, during rodent bioassays it is rarely
empty (Poet et al., 2003; Green et al., 2002). In contrast, the increased vascularity and digestive
processes of the glandular stomach cause it to empty relatively quickly. The combination of
high dehydrogenase activity and prolonged contact with the substrate is considered to be the
principle reason for the forestomach specificity of EGBE (Green et al., 2002).
15 The Corley (2003) model predicts that the concentrations of BAL in gastrointestinal tract tissues of male
and female mice would be 19 and 33 |_iM, respectively, following oral gavage exposure to 600 mg/kg EGBE. This
compares remarkably well with the levels of BAL actually observed in forestomach tissue of male and female mice,
18 and 33 |_iM, respectively, following oral gavage exposure to EGBE at 600 mg/kg (Deisinger and Boatman, 2004).
Al-9
-------
Why were there no fore stomach effects observed in the NTP (1993a) subchronic drinking water
study of mice? There is no clear answer to this question at this point in time. No signs of
forestomach irritation were observed in mice at dose levels as high as 1400 mg/kg/day in 2-
week and 13-week drinking water studies conducted by NTP (NTP, 1993a). It has been
suggested that such oral non-bolus dosing of EGBE does not result in high enough local
concentrations of EGBE and BAA (Poet et al., 2003). Studies with other nongenotoxic
forestomach carcinogens have demonstrated that forestomach effects are dependent not only on
the dose but also on the chemical concentration in the dosing solution (Ghanayem et al., 1985)
and other effects of EGBE appear to be highly dependent on the concentration attained (Nyska
et al., 1999a; Long et al., 2000; Ghanayem et al., 2000; 2001). In addition, first-pass liver
metabolism of orally administered EGBE may effect the extent to which EGBE reaches the
forestomach via the route that has been proposed following iv injection, distribution to salivary
glands followed by the swallowing of EGBE laden saliva (Green et al., 2002; Poet et al., 2003).
Why are mice more sensitive to forestomach effects of EGBE than rats? Green et al. (2002)
have suggested that, since the rate limiting step in the metabolism of EGBE to BAA is the
slower metabolism of EGBE to BAL via alcohol dehydrogenase, the order of magnitude higher
alcohol dehydrogenase activity (Vmax) in mice versus rats provides a possible explanation for
this relative mouse sensitivity. However, while the higher Vmax for this enzyme indicates that
mice may have a higher maximal capacity for metabolizing EGBE to BAL, the higher affinity
constant (Km) in mice indicates that this enzyme is less "efficient" in mice versus rats. A
comparison of the metabolic rates for this step in rats and mice is a more valid approach. Table
Al-2 compares predicted rates of BAL formation between rats and mice under substrate
concentrations ranging from 0.001 mM to 150 |J,M. Rates were estimated using the Michaelis-
Menten rate equation (Vmax * [substrate] / (Km + [substrate]) and kinetic constant values for
forestomach tissue alcohol dehydrogenase provided in Table 1 of Green et al. (2002).
The results in Table Al-2 indicate that similar rates of BAL formation would be
observed at approximately 4.5 mM EGBE. At EGBE concentrations below 4.5 mM, rates of
BAL formation in the rat will occur at higher rates, roughly 11-fold higher at an EGBE
concentration of 0.1 mM. Concentrations of EGBE in the stomach above 4.5 mM are
considered unlikely based on Corley (2003) and Figure Al-1. Thus, it does not appear that the
apparent greater sensitivity of mice to the forestomach effects of EGBE can be explained by
species differences in metabolism. It should be noted, however, that mice have a faster
breathing rate than rats and were exposed to 2-fold higher concentrations of EGBE (high of 250
ppm) than rats (high of 125 ppm) in the NTP study (NTP, 2000), factors that would likely lead
to higher target organ doses in mice.
Al-10
-------
Table Al-2: Alcohol Dehydrogenase-catalyzed conversion of EGBE to BAL in rats and mice
rates predicted from published kinetic constants (Green et al., 2002)
[EGBE]
(mM)
0.001
0.01
0.05
0.1
0.5
1
4.5
5
50
100
150
Predicted Metabolic Rate
Rats
0.005
0.054
0.239
0.417
1.029
1.261
1.528
1.538
1.618
1.622
1.623
Mice
0.0004
0.004
0.018
0.037
0.182
0.359
1.505
1.657
8.849
11.661
13.043
RatMouse
15.24
14.78
13.06
11.39
5.67
3.51
1.02
0.93
0.18
0.14
0.12
Why were no tumors in male mice observed? As has been discussed, it is likely that the mode
of action for the formation of these tumors involves chronic tissue injury as forestomach effects
were preceded or accompanied by marked signs of irritation, including hyperplasia and
ulceration. Female mice in the NTP study (NTP, 2000) were clearly more susceptible to
inflammation and hyperplasia than male mice at similar exposure levels. The incidence of
hyperplasia of the epithelium in the forestomach was 54% at 62.5 ppm, 86% at 125 ppm and
88% at 250 ppm in female mice, and 14% at 62.5 ppm, 32% at 125 ppm and 42% at 250 ppm in
male mice (NTP, 2000). Female rats and mice were also more sensitive than males to the
hematological effects of EGBE. The reasons for the apparent female rodent sensitivity to these
lesions are not clear. However, the hemolytic sensitivity of female rats is well correlated with
the longer residence time for BAA in the blood of female versus male rats (Dill et al., 1998).
Other Possible Modes of Action for Forestomach Tumor Development in Female Mice
Though the evidence favors the hypothesis that BAA is the principle toxic metabolite of EGBE,
roles for BAL (Dartsch et al., 1999; Ghanayem et al., 1987b) and butyric acid (Cantox, 2000) have
been suggested. Neither BAL nor butyric acid have been identified in vivo following EGBE
Al-11
-------
exposure, but they are recognized as required metabolites of EGBE (NTP, 2000; Green et al.,
2002).
It is not likely that butyric acid plays a significant role in the toxicity of EGBE, particularly at
environmentally relevant concentrations. High concentrations of butyric acid have caused
ulceration and other preneoplastic lesions in mice (Harrison et al., 1991). However, low
concentrations of butyric acid do not appear to be harmful as it naturally occurs in the diet through
the fermentation of fiber and starch and constitutes a significant (up to 10 mol%) portion of total
bovine milk fatty acid (Smith and German, 1995).
In vitro studies without enzyme (dehydrogenase) activation have shown that BAL is a less
potent hemolytic agent, but causes similar hemolytic effects at 0.5 mM (Ghanayem et al., 1989) and
is more cytotoxic (Dartsch et al., 1999; Ghanayem, 2003) than EGBE or BAA. Also, EGBE was as
toxic to rat and more toxic to mouse hepatocyte cell cultures (no enzyme activation) than BAA as
measured by LDH release at comparable concentrations (25 and 50 jiM) (Park et al., 2002a).
However, the analysis performed by Corley (2003) indicates that BAL concentrations in the liver
and stomach would not have risen much above 1 jiM, even in the highest exposure group, and that
BAA concentrations would likely have been at least 3 orders of magnitude higher.
In vivo studies have indicated that pretreatment of rats with an alcohol dehydrogenase inhibitor,
pyrazole, prior to a single 125 mg EGBE/kg gavage exposure protected against hemolysis
(Ghanayem et al., 1987b), presumably by blocking the production of both BAL and BAA.
Pretreatment of rats with an aldehyde dehydrogenase inhibitor, cyanamide, prior to a single 125 mg
EGBE/kg gavage exposure reduced hemolytic responses, but did increased RBC swelling,
increased mortality, decreased BAA formation and excretion in the urine, and increased the urinary
excretion of EGBE conjugates with glucuronide and sulfate (Ghanayem et al., 1987b). This
hematotoxicity in the presence of cyanamide may be due to BAL, but it may also be due to residual
BAA. Inhibitors such as cyanamide and pyrazole are not very specific and may cause other effects.
In addition, Ghanayem et al., (1990) found that while EGBE + cyanamide decreased BAA
concentrations in rats, some BAA was formed and the BAA half-life was increased. Further, when
Ghanayem et al. (1987) administered a gavage dose of 125 mg BAL/kg + cyanamide to rats they
observed almost no hemolytic activity (Ghanayem et al., 1987). Also, gavage administration to rats
of 125 mg EGBE/kg and the molar equivalent of BAL and BAA resulted in no significant
difference between the hemolytic effects of the three chemicals between 2 and 24 hours after
exposure (Ghanayem et al., 1987). These facts suggest that EGBE's hemolytic activity (without
co-exposures) is due to BAA, and that the metabolism of EGBE and BAL to BAA takes place
rapidly and completely.
Another possible alternative mode of action could exist if EGBE or one of its metabolites were
to have the capability of damaging a cell through direct interaction with its DNA. EGBE has been
extensively tested in a variety of short-term genotoxicity assays. Recent reviews of available in
vitro and in vivo genotoxicity assays are in agreement that EGBE is not likely to be genotoxic
(Commonwealth of Australia, 1996; Elliot and Ashby, 1997; U.S. EPA, 1999; NTP, 2000). Known
Al-12
-------
or proposed structure-activity relationships do not suggest that EGBE would be expected to be
genotoxic (Tennant and Ashby, 1991). EGBE was non-genotoxic in bacteria with or without
microsomal activation. Cytogenetic assays on mammalian cells in vitro and assays that
demonstrate the ability for a chemical to introduce mutations into mammalian cells in in vitro
culture systems showed no evidence of activity associated with EGBE. EGBE also showed no
evidence of clastogenicity in several in vitro chromosome aberration assays (NTP, 1993a;
Villalobos-Pietrini et al., 1989). Sister chromatid exchange assay results were mixed. EGBE did
not affect CHO cells (NTP, 1993a; Slesinski and Weil, 1989), but was reported to be "weakly
positive " in V79 cells (Elias et al., 1996) and positive in peripheral human lymphocytes
(Villalobos-Pietrini et al., 1989). These positive results may be secondary effects associated with
cell cycle delay induced by the cytotoxic effects of EGBE (Elliot and Ashby, 1997). The Syrian
hamster embryonic cell transformation (SHE) assay has provided negative (Elias et al., 1995; 1996;
Park et al., 2002b) or uncertain (Kerckaert et al., 1996) results for EGBE. In vivo tests of EGBE's
genotoxicity, including a mouse and rat micronucleus assay and a 32P-postlabelling assay (Kieth et
al., 1996) were negative. Similar frequencies ofH-ras mutations were detected in DNA derived
from forestomach neoplasms from treated (57%, 8/14) and untreated (45%, 5/11) mice, prompting
the authors to suggest that EGBE may act as a promoter that stimulates clonal growth of initiated
cells present spontaneously in forestomach tissue (Sill et al., 2000).
The major metabolite of EGBE, BAA, was negative in the Salmonella/microsomQ assay with or
without activation by rat liver homogenate (S9) (Hoflack et al., 1995). A number of tests were
reported by Elias et al. (1996), but the authors generally provided insufficient data to confirm their
conclusions (Elliot and Ashby, 1997). BAA was negative in an in vivo bone marrow micronucleus
assay in the CD-I mouse; but there was evidence of toxicity and a reduction in polychromatic
erythrocytes (Elias et al., 1996), a finding that is consistent with BAA's known ability to cause
erythrocyte membrane fragility. Overall, there is no clear evidence that BAA is genotoxic, but the
database is very limited.
BAL, a short-lived metabolite of EGBE, has also been studied in several genotoxicity assays.
B AL was negative for bacterial mutation in several strains in the Salmonella/microsomQ assay
(Hoflack et al., 1995) and did not induce mutations in the CHO AS52 cell line at up to 0.2% (v/v)
(Chiewchanwit and Au, 1995). A recent Comet Assay performed by Klaunig and Kamendulis
(2004; 2005) found that BAL did not induce DNA single strand breaks at concentrations three
orders of magnitude higher than BAL concentrations estimated to occur by PBPK modeling in liver
and forestomach (see discussions of Corley, 2003 above). However, it has been reported to be
clastogenic in in vitro assays without enzyme activation at concentrations ranging from 0.2 to 1
mM (Elliot and Ashby, 1997; Ghanayem, 2003). In Chinese hamster lung (V79), BAL and other
alkoxyacetaldehydes (methoxtacetaldehyde and ethoxyacetaldehyde) appear to be mitotic poisons
which seem to interfere with several different components responsible for cell division in a dose-
dependent fashion at concentrations ranging from 0.08 to 0.69 mM (Elliot and Ashby, 1997). The
specific cellular targets, the mechanisms and biochemical bases of their interactions are not known.
Al-13
-------
Unpublished in vitro data submitted by Ghanayem (2003) indicates that 0.5 mM BAL (no enzyme
activation) can cause sister chromatid exchange (twice control levels) in human lymphocyte cells.
Ghanayem (2003) also reported that BAL was cytotoxic to human lymphocyte cells, causing a 50%
reduction in cell number and viability at 0.5 mM.
As discussed previously, Green et al. (2002) and Corley (2003) have suggested that BAL is very
short-lived, being metabolized further to BAA due to a high aldehyde dehydrogenase activity in the
mouse forestomach. There are other lines of evidence that indicate that direct interaction of BAL
with the DNA molecules does not play a significant role in the carcinogenic activity of EGBE.
First, BAL causes cytotoxicity at levels associated with chromosome effects and cytotoxicity itself
can have effects which result in chromosome damage such as reduction in the repair of SCE.
Second, acetaldehyde is recognized as "weakly mutagenic" and structural comparisons of
acetaldehydes demonstrate that a longer-chain aldehyde such as BAL would be less likely to
interact with DNA than a shorter chain aldehyde such as acetaldehyde (Dellarco, 1988). Third, if
BAL was a stable mutagenic metabolite in any of the in vitro assays exposed to butoxyethanol, one
would expect them to give positive results; the results were generally negative. The Elias et al.
(1996) paper suggests that the V79 cells possess neither alcohol dehydrogenase nor aldehyde
dehydrogenase. The relevance of these studies, or any systems that lack these enzymes, are of
limited value in elucidating the mode of action of toxicity in biological systems which possess these
enzymes. Finally, chemicals for which mutagenesis/genotoxic effects play a significant role
generally induce more tumors at earlier time points, rather than near the end of the conducted
bioassays, due to their ability to both initiate and promote tumor pathogenesis. The mutagenic
compound ethylene dibromide, for instance, was reported to induce forestomach tumors in all dose
groups 168 to 280 days from the start of exposure (NCI, 1978). As was discussed above under
"Temporal Association" EGBE is consistent with other nongenotoxic forestomach carcinogens
such as EA in that observed tumors were not as severe (generally did not progress to carcinoma)
and were not observed until well into the study, after long periods of forestomach cell damage and
repair. The first reported incidence of forestomach papilloma or carcinoma in female mice was
731, 731 and 582 days in the 62.5, 125 and 250 ppm EGBE exposure groups, respectively. In
summary, evidence from in vivo and in vitro genotoxicity assays do not support the idea that BAL
would have any significant genotoxicity in vivo.
In general, aldehydes such as formaldehyde and acetaldehyde are irritants and have been found
to possess some genotoxic activity. The former has been found to be carcinogenic in rat nasal
tissue (Swenberg et al., 1983). With increasing carbon length, the primary aldehydes appear to
exhibit less to no genotoxic potential. While the opinion is not universal, mutagenicity is not
believed to be the driving force in the toxicity of two other analogous aldehydes, formaldehyde and
acetaldehyde. For acetaldehyde, the apparent cytogenetic damage is best described as a
intracellular reduction in pH by acetic acid. The chromosome damaging effect of lowered pH has
been clearly demonstrated by Morita et al. (1992) and Morita, (1995).
Al-14
-------
It does not appear that EGBE, BAL or BAA preferentially bind to stomach tissue
macromolecules (Poet et al., 2003; Green et al., 2002). Poet et al. (2003) found that high levels of
EGBE concentrate in the food content of the forestomach following ip exposure (Poet et al., 2003),
indicating that the observed sequestering of EGBE in the forestomach is related to its retention in
the food that remains there, not to preferential binding to proteins within forestomach tissue.
In summary, a nonlinear mode of action involving forestomach damage and cell epithelial
proliferation is likely to be responsible for the increased forestomach tumor incidence reported by
NTP (2000). A metabolite of EGBE, BAL, has been found to cause chromosome damage in some
in vitro studies at cytotoxic levels of exposure. However, available evidence from a published
EGBE PBPK model that has been modified to include kinetics for the metabolism of the BAL
intermediate (Corley, 2003) and dosimetry data from gavage studies that confirm these model
estimates (Deisinger and Boatman, 2004) suggest that the conditions of these in vitro assays (e.g.,
no metabolic activation; high, cytotoxic concentrations of BAL) are of little relevance to expected
target organ (forestomach) environment (e.g., high metabolic activity; low concentrations of BAL).
Relevance of Female Mouse Forestomach Tumors to Humans
Pending a definitive determination concerning the role of BAL genotoxicity, EGBE appears to
be one of a group of non-genotoxic compounds that can indirectly cause forestomach tumors
through the sustained cytotoxicity and cell regeneration brought about by irritation and breakdown
of the forestomach's gastric mucosal barrier. According to the EPA's Science Advisory Panel,
quantification of the cancer risk for such compounds should be a nonlinear threshold approach
based on the forestomach tumors. Ideally, such an approach would take into account differences in
the tissue doses and pharmacokinetics of EGBE in humans versus rodents (see also "Relevance of
Mouse Liver Hemangiosarcomas to Humans" in Attachment 2 of this paper; and U.S. EPA, 1999)
and certain unique characteristics of the rodent forestomach (Green et al., 2002; Poet et al., 2003) in
order to make a reasonable approximation of the human exposure that could result in a target organ
dose roughly equivalent to the dose presented to the mouse forestomach in the NTP study (NTP,
2000). Selection of a target organ is made difficult by the fact that humans do not have an organ
directly comparable to the forestomach. However, there are histological similarities between
rodent forestomach tissue and the lower part of the human esophagus, and humans do suffer from
conditions (e.g., Barrett's esophagus) where chronic irritation caused by acid reflux or other
pathological influences can cause severe histological damage that may progress to a neoplastic
result.
Relevance to Susceptible Subpopulations, Including Children
Differences in susceptibility to gastric irritation - Infants (especially those less than three months
in age) do not have fully developed digestive systems. This can lead to problems related to the
infant stomach's high pH and inability to destroy certain stomach bacteria (U.S.EPA, 1991).
Adults have low pH (high acidity) stomach acid that tends to destroy bacteria. The impact this
Al-15
-------
would have on the susceptibility of the infant stomach to irritation caused by EGBE is unknown at
this time. However, the reduced pH in adults might increase their susceptibility to the genotoxic
effects of EGBE or its metabolites due to the chromosome damaging effect of lowered pH (Morita
et al., 1992; Morita, 1995). Chronic irritation at the gastroesophageal junction induced by acid
reflux is increasing in incidence (Voutilainen et al., 1999) as is Barrett's Esophagus and the causes
of these human conditions are similar to that seen in the rodent forestomach following EGBE
exposure, i.e. chronic inflammation and induced cell proliferation. However, for individuals so
affected, it is unlikely that the small levels of EGBE in question would significantly exacerbate this
condition.
Genetic differences - Other potentially susceptible subpopulations include individuals with
enhanced metabolism or decreased excretion of BAA. Polymorphisms in alcohol and aldehyde
dehydrogenases could lead to differences in the metabolism and elimination of EGBE. Human
genetic polymorphisms in alcohol dehydrogenase and aldehyde dehydrogenase are prevalent in
certain ethnic groups (Chan, 1986) and these polymorphisms have been shown to alter rates of
metabolism and elimination of ethanol and acetaldehyde (Agarwal and Goedde, 1992). For
instance, native Americans and approximately 50% Asian people are deficient in aldehyde
dehydrogenases. Aldehyde dehydrogenases comprises more than nine isoforms in humans (Hsu et
al., 1994). A deficiency or loss of one of them (ALDH2) can lead to a nearly complete loss of
enzymatic activity (Crabb et al., 1989; Kitagawa et al., 2000). Individuals with atypical and/or
deficient alcohol dehydrogenase appear to be more susceptible to adverse effects from increased
levels of acetaldehyde including facial flushing, general discomfort, acetaldehyde-protein adducts
and alcohol-induced liver diseases (Agarwal and Goedde, 1992). However, the PBPK model
developed by Corley et al. (2004) indicates that individuals with low aldehyde dehydrogenase
activity (1A> Vmax) would not accumulate significant BAL in the liver or forestomach,16 even
following inhalation exposure to a theoretical maximum of 1160 ppm EGBE for 6 hours.
Haufroid et al. (1997) conducted a human study on workers exposed to EGBE to test the
possible influence of genetic polymorphism for CYP 2E1 on urinary BAA excretion rate. One
exposed individual exhibited a mutant allele with increased cytochrome P450 oxidative activity that
coincided with a very low urinary BAA excretion. However, the researchers did not measure BAA
conjugated to glutamine, an alternative pathway for BAA excretion in humans. Further
investigations on the influence of genetic polymorphism for CYP 2E1 on urinary BAA excretion
rate are needed before any firm conclusions can be drawn.
Gender differences - As discussed, Dill et al. (1998) have reported that female rats and mice
metabolize EGBE to BAA faster and female mice clear BAA slower than males. A number of
effects on the rat liver, kidneys, spleen, and bone marrow and, to a lesser extent, the thymus,
Predicted BAL concentrations were below 0.001 mM in the liver and 0.004 mM in the GI tract following
inhalation exposure to saturated air concentrations of EGBE. These concentrations are considerably lower than
concentrations of BAL shown to be clastogenic (0.2 mM) orhemolytic (0.5 mM: Ghanayem et al., 1989) in vitro.
Al-16
-------
particularly secondary effects of hemolysis such as anemia, infarctions, and thrombosis, are more
pronounced in females (NTP, 1993a; Nyska et al., 1999a,b; Ghanayem et al., 2000; 2001; long et
al., 2000) and female mice experienced a higher incidence and higher severity of forestomach
lesions following chronic exposure (NTP, 2000). Slight gender differences have been noted in
other rodent (Carpenter et al., 1956; Dodd et al., 1983; NTP, 1993a; NTP, 2000), rabbit (Tyler,
1984), dog, monkey, and human studies (Carpenter et al., 1956), with females being consistently
more susceptible to the primary hemolytic effects of EGBE. In the process of studying and
comparing the metabolic and cellular basis of EGBE-induced hemolysis, Ghanayem (1989)
observed that the blood from female human volunteers showed a slightly greater sensitivity to
hemolysis following incubation with BAA than male blood.
Age differences - A number of factors may differentially affect children's responses to toxicants.
The only information available on the toxicity of EGBE to children is from the case study by Dean
and Krenzelok (1991), who observed 24 children, age 7 mo to 9 years, subsequent to oral ingestion
of at least 5 mL of glass window cleaner containing EGBE in the 0.5% to 9.9% range (potentially
25 to 1500 mg EGBE exposures). No symptoms of EGBE irritation, poisoning or hemolysis were
reported. While the effects reported in adult poisonings have been more severe than those reported
in these children, adults tended to consume larger volumes and different concentrations of EGBE,
making a comparison of toxic effects observed to age sensitivity of the human extremely difficult.
The effect of age on EGBE-induced hematotoxicity was studied in male F344 rats by
Ghanayem and co-workers (1987a, 1990). These studies also demonstrated the time course for the
onset and resolution of hematological and histopathological changes accompanying hemolysis.
Adult (9-13 wk) male F344 rats were significantly more sensitive to the hemolytic effects of EGBE
than were young (4-5 wk) male rats following administration of a single gavage dose of EGBE at
32, 63, 125, 250, or 500 mg/kg. In concurrent metabolism studies, increased blood retention of
EGBE metabolite BAA (as measured by increased Cmax, AUC, and T,/2) was observed. Additionally,
young rats eliminated a significantly greater proportion of the administered EGBE dose as exhaled
carbon dioxide (CO2) or as urinary metabolites as well as excreting a greater proportion of the
EGBE conjugates (glucuronide and sulfate) in the urine (Ghanayem et al., 1987a,b; 1990). These
researchers suggested that the pharmacokinetic basis of the age-dependent toxicity of EGBE may
be due to a reduced ability by older rats to metabolize the toxic metabolite BAA to CO2 and a
diminished ability to excrete BAA in the urine.
NTP (2000) also found that young mice (6-7 weeks) eliminated BAA 10-times faster than
aged (19 months) following a 1-day of inhalation exposure to 125 ppm EGBE. This difference was
not as apparent after 3 weeks of exposure, suggesting that factors other than age may have been
involved (Dill et. al., 1998).
Developmental studies, which may also be of possible relevance to this issue, have been
conducted using rats, mice, and rabbits dosed orally, by inhalation or, in one study, dermally
(Hardin et al., 1984; Heindel et al., 1990; Nelson et al., 1984; NTP, 1993a; Sleet et al., 1989; Tyler,
Al-17
-------
1984; Wier et al., 1987). Maternal toxicity related to the hematologic effects of EGBE and
relatively minor developmental effects such as delayed skeletal ossification were reported in most
studies. No teratogenic toxicities were noted in any of the studies. It can be concluded from these
studies that EGBE is not significantly toxic to developing fetuses of laboratory animals.
Older rats have been shown to have a reduced ability to metabolize the toxic metabolite
BAA to CO2 and a diminished ability to excrete BAA in the urine (Ghanayem et al.,1987a, 1990).
However, the relevance of this finding to the possible susceptibility of elderly humans is uncertain
due to the fact that humans have conjugation pathways for the excretion of BAA (BAA-Glutamine
and BAA-Glycine) that are not available to the rat.
Summary
Available data establish a plausible nonlinear, nongenotoxic mode of action for the
moderate increase observed by NTP (2000) in the incidence of forestomach tumors in female mice
following chronic inhalation exposure to EGBE. EGBE appears to be one of a group of non-
genotoxic compounds that can indirectly cause forestomach tumors through the sustained
cytotoxicity and cell regeneration brought about by irritation and breakdown of the forestomach's
gastric mucosal barrier. While this mode of action may be of qualitative relevance to humans, the
exposure concentrations that would be necessary to cause hyperplastic effects and tumors in
humans, if attainable, are likely to be much higher than the concentrations necessary to cause
forestomach effects in mice, primarily because humans lack a comparable organ for storage and
long term retention of EGBE. However, even if this fact is ignored, the analysis in Attachment 3
indicates that the exposure concentrations necessary to cause hyperplastic effects in humans would
be much higher than the existing RfD and RfC for EGBE. Given these considerations, it appears
reasonable to assume that the RfC and RfD developed for EGBE (EPA, 1999a) are sufficient for the
prevention of hyperplasia and associate tumors in humans.17
17These analyses are consistent with the nonlinear assessment approach described in existing interim (U.S.
EPA, 1999a) and draft (2003) cancer guidelines.
Al-18
-------
REFERENCES - ATTACHMENT 1
Aasmoe, L., Winberg, J.O. ; Aarbakke, J. (1998) The role of liver alcohol dehydrogenase
isoenzymes in the oxidation of gluycol ethers in male and female rats. TAP 150: 86-90.
Ali A., Zhang Q., Dai J., Huang X. (2003) Calcein as a fluorescent iron chemosensor for the
determination of low molecular weight iron in biological fluids. Biometals 16: 285-293.
Bennette, D. (2001) The distribution of radioactivity in the female B6C3fl mouse following a
single intravenous injection of 2-butoxy[l-14c]ethanol. Alderley Park, Macclesfield, Cheshire,
United Kingdom: Central Toxicology Laboratory; report no: Ctl/r/1446.
Bogdanffy, M. S.; Randall, H. W.; Morgan, K. T. (1987) Biochemical quantitation and
histochemical localization of carboxylesterase in the nasal passages of the Fischer-344 rat and
B6C3F1 mouse. Toxicol. Appl. Pharmacol. 88: 183-194.
Browning, J.; Gannon, B. J.; O'Brien, P. (1983) The microvasculature and gastric luminal pH of the
forestomach of the rat: a comparison with the glandular stomach. Int. J. Microcirc.: Clin. Exp. 2:
109-118.
Bueld, J. E.; Netter, K. J. (1993) Factors affecting the distribution of ingested propionic acid in the
rat forestomach. Food Chem. Toxicol. 31: 169-176.
Cantox Environmental Inc. (2000) EGBE: review of the relevance of forestomach lesions in the
female B6C3F1 mouse. Final report; October.
Carpenter, C. P.; Pozzani, U. C.; Wiel, C. S.; Nair, J. H., Ill; Keck, G. A.; Smyth, H. F., Jr. (1956)
The toxicity of butyl cellosolve solvent. AMA Arch. Ind. Health 14: 114-131.
Chiewchanwit, T.; Au, W. W. (1995) Mutagenicity and cytotoxicity of 2-butoxyethanol and its
metabolite, 2-butoxyacetaldehyde, in Chinese hamster ovary (CHO-AS52) cells. Mutat. Res. 334:
341-346.
Clay son, D. B.; Iverson, F.; Nera, E. A.; Lok, E. (1991) Early indicators of potential neoplasia
produced in the rat forestomach by nongenotoxic agents: the importance of induced cellular
proliferation. Mutat. Res. 248: 321-331.
Cohen and Ellwein, 1990, cited in Ghanayem et al. (1994)
Commonwealth of Australia. (1996) 2-Butoxyethanol. Priority existing chemical number 6.
Canberra: Australian Government Publishing Service. Available:
http://www.nicnas.gov.au/publications/car/pec/pecindex.htm [27 January, 2003].
Al-19
-------
Corley, R. A.; Bormett, G. A.; Ghanayem, B. I. (1994) Physiologically-based pharmacokinetics of
2-butoxyethanol and its major metabolite, 2-butoxyacetic acid, in rats and humans. Toxicol. Appl.
Pharmacol. 129: 61-79.
Corley, R. A.; Markham, D. A.; Banks, C.; Delorme,P.; Masterman, A.; Houle, J. M. (1997)
Physiologically-based pharmacokinetics and the dermal absorption of 2-butoxyethanol vapors by
humans. Fundam. Appl. Toxicol. 39: 120-130.
Corley, R. A.; Weitz, K. K.; Mast, T. J.; Miller, R. A. and Thrall, B. D. (1999) Short-term studies to
evaluate the dosimetry and modes of action of EGBE in B6C3F1 mice [final report]. Battelle
project no. 29753; 1999.
Corley, R.A.; Grant, D.M.; Farris, E.; Weitzl, K.K; Soelberg, JJ; Thrall, K.D.; and Poet, T.S.
(2003) Impact of age and gender on the toxicokinetics of 2-Butoxyethanol and its hemolytic
metabolite, butoxyacetic Acid in rats and mice: improvements in pbpk modeling. Submitted for
publication in Tox. Letters.
Corley, R. A. (2003) Preliminary Incorporation of the Intermediate Metabolite,
Butoxyacetaldehyde, in PBPK Model for Butoxyethanol. Personally communicated to Jeff Gift,
EPA National Center for Environmental Assessment - RTF, NC.
Crabb DW, Edenberg HJ, Bosron WF, Li TK. 1989. Genotypes for aldehyde dehydrogenase
deficiency and alcohol sensitivity. The inactive ALDH2(2) allele is dominant. J Clin Invest 83:
314-316.
DartschP.C.; Hildenbrand S.; Gfrorer W.; Kimmel R.; Schmahl F.W. (1999) Cytotoxic effects of
2-butoxyethanol in vitro are related to butoxyacetaldehyde, an intermediate oxidation product.
Environ Toxicol Pharmacol 7: 135-142.
Dean, B. S.; Krenzelok, E. P. (1991) Clinical evaluation of pediatric ethylene glycol monobutyl
ether poisonings. Vet. Hum. Toxicol. 33: 362.
Deisinger, PJ. and Boatman, RJ. (2004). In vivo metabolism and kinetics of ethylene glycol
monobutyl ether and its metabolites, 2-butoxyacetaldehyde (BAL) and 2-butoxyacetic acid (BAA),
as measured in blood, liver and forestomach homogenates from mice. R&D Report of Eastman
Kodak Company, Rochester, NY to the American Chemistry Council.
Dill, J. A. ; Lee, K. M.; Bates, D. J.; Anderson, D. J.; Johnson, R. E.; Chou, B. J.; Burka, L. T.;
Roy croft, J. H. (1998) Toxicokinetics of inhaled 2-butoxyethanol and its major metabolite,
2-butoxyacetic acid, in F344 rats and B6C3F1 mice. Toxicol. Appl. Pharmacol. 153: 227-242.
Al-20
-------
Dodd, D. E.; Snelling, W. M.; Maronpot, R. R.; Ballantyne, B. (1983) Ethylene glycol monobutyl
ether: acute, 9-day, and 90-day vapor inhalation studies in Fischer 344 rats. Toxicol. Appl.
Pharmacol. 68: 405-414.
Elias, Z.; Poirot, O.; Daniere, M. C.; Terzetti, F.; Marande, A. M. (1995) Evaluation of the
transforming potential of the ethylene glycol monobutyl ether (EGBE) in Syrian hamster embryo
(SHE) cell system. Cell. Biol. Toxicol. 11: 223.
Elias, Z.; Daniere, M. C.; Marande, A. M.; Poirot, O.; Terzelti, F.; Schneider, O. (1996) Genotoxic
and/or epigenetic effects of some glycol ethers: results of different short-term tests. Occup. Hyg. 2:
187-212.
Elliot, B. M.; Ashby, J. (1997) Review of the genotoxicity of 2-butoxyethanol. Mutat. Res. 387:
89-96.
Ghanayem, B.I, Maronpot, R.R. and Matthews, H.B. (1985). Ethyl Aery late-Induced Gastric
Toxicity: II. Structure-Toxicity Relationships and Mechanism. Toxicol. Appl. Pharmacol. 80:336-
344.
Ghanayem, B.I, Maronpot, R.R. and Matthews, H.B. (1986). Association of Chemically-Induced
Forestomach Cell Proliferation and Carcinogenesis. Cancer Letters 32:271-278.
Ghanayem, B. I.; Burka, L. T.; Sanders, J. M.; Matthews, H. B. (1987a) Metabolism and disposition
of ethylene glycol monobutyl ether (2-butoxyethanol) in rats. Drug Metab. Dispos. 15: 478-484.
Ghanayem, B. I; Burka, L. T.; Matthews, H. B. (1987b) Metabolic basis of ethylene glycol
monobutyl ether (2-butoxyethanol) toxicity: role of alcohol and aldehyde dehydrogenases. J.
Pharmacol. Exp. Ther. 242: 222-231.
Ghanayem, B. I. (1989) Metabolic and cellular basis of 2-butoxyethanol-induced hemolytic anemia
in rats and assessment of human risk in vitro. Biochem. Pharmacol. 38: 1679-1684.
Ghanayem, B. L; Sanders, J. M.; Clark, A. M.; Bailer, J.; Matthews, H. B. (1990) Effects of dose,
age, inhibition of metabolism and elimination on the toxicokinetics on the toxicokinetics of 2-
butoxyethanol and its metabolites. J. Pharmacol. Exp. Ther. 253: 136-143.
Ghanayem, B. L, Sanchez, I. M., Maronpot, R. R., Elwell, M. R., and Matthews, H. B. (1993).
Relationship Between the Time of Sustained Ethyl Acrylate Forestomach Cell Proliferation and
Carcinogenicity. Env. Health Persp. 101 (Sppl. 5):277-280.
Al-21
-------
Ghanayem, B. I.; Sullivan, C. A. (1993) Assessment of the hemolytic activity of 2-butoxyethanol
and its major metabolite, butoxyacetic acid, in various mammals including humans. Hum. Exp.
Toxicol. 12:305-311.
Ghanayem, B. I, Sanchez, I. M., Matthews, H B., and Elwell, M. R. (1994). Demonstration of a
temporal relationship between ethyl acrylate induced forestomach hyperplasia and carcinogenesis.
Toxicologic Pathology 22:497-509.
Ghanayem, B. I, Ward, S M, Chanas, B, and Nyska, A. (2000). Comparison of the acute
hematotoxicity of 2-butoxyethanol in male and female F344 rats. Human Exp. Toxicol. 19:185-
192.
Ghanayem, B. I, Long, P., Ward, S M, Chanas, B, and Nyska, A. (2001). Hemolytic Anemia,
Thrombosis, and Infarction in Male and Female F344 Rats Following Gavage Exposure to 2-
Butoxyethanol. Experimental and Toxicologic Pathology 53:97-105.
Gold, L. S., Manley, N. B., Slone, T. H., Garfmkel, G B., Rohrbach, L, Ames, B. N. (1993) The
fifth plot of the carcinogenic potency database: results of animal bioassays published in the general
literature through 1988 and by the National Toxicology Program through 1989. Environ. Health
Perspect. 100: 65-135.
Green, T. (1996) The metabolism of methyl methacrylate in the nasal tissues of rat and human.
Zeneca/Central Toxicology Laboratory Report No. CTL/R/1290, issued December 4, 1996.
Sponsor: CEFIC
Green, T. (2000) The Distribution of Radioactivity in The Female B6c3fl Mouse Following a
Single 6 Hour Exposure to 2-butoxy[l-14C]ethanol by Inhalation. Alderley Park, Macclesfield,
Cheshire United Kingdom: Central Toxicology Laboratory; report no: Ctl/r/1444; May.
Green, T.; Toghill, A.; Lee, R.; Moore, R.; Foster, J. (2002) The development of forestomach
tumors in the mouse following exposure to 2-butoxyethanol by inhalation. Studies on the mode of
action and relevance to humans. Toxicology 180: 257-273.
Grice, H. C. (1988) Safety evaluation of butylated hydroxyanisole from the perspective of effects
on forestomach and oesophageal squamous epithelium. Food Chem. Toxicol. 26: 717-723.
Hardin, B. D.; Goad, P. T.; Burg, J. R. (1984) Developmental toxicity of four glycol ethers applied
cutaneously to rats. Environ. Health Perspect. 57: 69-74.
Harrison, P. T. C. (1992) Propionic acid and the phenomenon of rodent forestomach tumorigenesis:
a review. Food Chem. Toxicol. 30: 333-340.
Al-22
-------
Harrison, P. T. C.; Grasso, P.; Badescu, V. (1991) Early changes in the forestomach of rats, mice
and hamsters exposed to dietary propionic and butyric acid. Food Chem. Toxicol. 29: 367-371.
Haseman, J.K.; Hailey, J.R. and Morris, R.W. (1998). Spontaneous neoplasm incidences in
Fsicher 344 rats and B6C3F1 mice in two-year carcinogenicity studies: a National Toxicology
Program update. Toxicologic Pathology. May-Jun, 1998, 26(3):428-41 .
Haufroid, V.; Thirion, F.; Mertens, P.; Buchet, J. P.; Lison, D. (1997) Biological monitoring of
workers exposed to low levels of 2-butoxyethanol. Int. Arch. Occup. Environ. Health 70: 232-236.
Heindel, J. J.; Gulati, D. K.; Russell, V. S.; Reel, J. R.; Lawton, A. D.; Lamb, J. C., IV. (1990)
Assessment of ethylene glycol monobutyl and monophenyl ether reproductive toxicity using a
continuous breeding protocol in Swiss CD-I Mice. Fundam. Appl. Toxicol. 15: 683-696.
Hirose M; Inoue, T.; Asamoto, M.; Tagawa, Y.; and Ito, N. (1986) Comparison of the effects of 13
phenolic compounds in induction of proliferative lesions of the forestomach and increase in the
labeling indices of the glandular stomach and urinary bladder epithelium of Syrian golden
hamsters. Carcinogenesis. 7:1285-1289.
Hoflack, J. C.; Lambolez, L.; Elias, Z.; Vasseur, P. (1995) Mutagenicity of ethylene glycol ethers
and of their metabolites in Salmonella typhimurium his". Mutat. Res. 341: 281-287.
Huff, J.; Cirvello, J.; Haseman, J.; Bucher, J. (1991) Chemicals associated with site-specific
neoplasia in 1394 long-term carcinogenesis experiments in laboratory rodents. Environ. Health
Perspect. 93:247-270.
Hsu LC, Chang WC, Yoshida A. 1994. Cloning of a cDNA encoding human ALDH7, a new
member of the aldehyde dehydrogenase family. Gene 151: 285-289.
Johanson, G. (1986) Physiologically-based pharmacokinetic modeling of inhaled 2-butoxyethanol
in man. Toxicol Lett 34:23-31.
Kerckaert, G. A.; Brauninger, R.; Le Boef, R. A.; Isfort, R. J. (1996) Use of the Syrian hamster
embryo cell transformation assay for carcinogenicity prediction of chemicals currently being tested
by the National Toxicology Program in rodent bioassays. Environ. Health Perspect. 104: 1075-
1084.
Keith, G.; Coulais, C.; Edorh, A.; Bottin, C.; Rihn, B. (1996) Ethylene glycol monobutyl ether has
neither epigenetic nor genotoxic effects in acute treated rats and in subchronic treated v-Ha-ras
transgenic mice. Occup. Hyg. 2: 237-249.
Al-23
-------
Kitagawa K, Kawamoto T, Kunugita N, Tsukiyama T, Okamoto K, Yoshida A, Nakayama K. 2000.
Aldehyde dehydrogenase (ALDH) 2 associates with oxidation of methoxyacetaldehyde; in vitro
analysis with liver subcellular fraction derived from human and Aldh2 gene targeting mouse. FEES
Klaunig, J. and Kamendulis, L. (2004). Effect Of 2-Butoxyacetaldehyde On The Induction Of DNA
Damage (Comet) In Rodent Endothelial Cells. R&D Report of Indiana University School of
Medicine to the American Chemistry Council.
Klaunig, I.E. and Kamendulis, L.M. (2005) Mode Of Action Of Butoxyethanol Induced Mouse
Liver Hemangiosarcomas And Hepatocellular Carcinomas. Toxicol. Lett. 156(1): 107-15.
Kroes, R.; Wester, P. W. (1986) Forestomach carcinogens: possible mechanisms of action. Food
Chem. Toxicol. 24: 1083-1089.
Lee, K. M.; Dill, J. A.; Chou, B. J.; Roy croft, J. H. (1998) Physiologically based pharmacokinetic
model for chronic inhalation of 2-butoxyethanol. Toxicol. Appl. Pharmacol. 153: 211-226.
Long P. H, Maronpot, R. R., Ghanavem. B. I.. Roycroft, J H., and Nyska A (2000). Dental pulp
infarction in female rats following inhalation exposure to 2-butoxyethanol. Toxicol. pathol.
28:246-252.
Mattes, P. M. and Mattes, W. B. (1992) Alpha-naphthyl butyrate carboxylesterase in human and rat
nasal tissue. Toxicol. Appl. Pharmacol. 114, 71-76.
Medinsky, M. A.; Singh, G.; Bechtold, W. E.; Bond J. A.; Sabourin P. J.; Birnbaum L. S.;
Henderson R. F. (1990) Disposition of three glycol ethers administered in drinking water to male
F344/Nrats. Toxicol. Appl. Pharmacol. 102: 443-455.
Morita,T., Nagaki,T., Fukuda,!., and Okumura,K. (1992) Clastogenicity of low pH to various
cultured mammalian cells. Mutat. Res 268: 297-305.
Morita,T. (1995) Low pH leads to sister-chromatid exchanges and chromosomal aberrations, and its
clastogenicity is S-dependent. Mutat. Res 334: 301-308.
National Toxicology Program (NTP). (1993a) Toxicology and carcinogenesis studies of
1,3-butadiene (CAS No. 106-99-0) in B6C3F1 mice (inhalation studies). Washington, DC: U.S.
Department of Health and Human Services, National Institutes of Health. (NTP technical report
series no. TR-434). Available: http://ntp-server.niehs.nih.gov/htdocs/LT-studies/tr434.html.
Al-24
-------
National Toxicology Program (NTP). (1993b) Technical report on toxicity studies of ethylene
glycol ethers 2-methoxyethanol, 2-ethoxyethanol, 2-butoxyethanol administered in drinking water
to F344/N rats and B6C3F1 mice. Washington, DC: U.S. Department of Health and Human
Services, National Institutes of Health. (National Toxicology Program toxicity report no. 26; NIH
publication no. 93-3349). Available from: NTIS, Springfield, VA; PB94-118106.
National Toxicology Program (NTP). (1996) Toxicology and carcinogenesis studies of acetonitrile
(CAS No. 75-05-8) in F344/N rats and B6C3F1 mice (inhalation studies). Washington, DC: U.S.
Department of Health and Human Services, National Institutes of Health. (NTP technical report
series no. TR-447). Available: http://ntp-server.niehs.nih.gov/htdocs/LT-studies/tr447.html.
National Toxicology Program (NTP). (1998) Toxicology and carcinogenesis studies of chloroprene
(CAS No. 126-99-8) in F344/N rats and B6C3F1 mice (inhalation studies). Washington, DC: U.S.
Department of Health and Human Services, National Institutes of Health. (NTP technical report
series no. TR-467). Available: http://ntp-server.niehs.nih.gov/htdocs/LT-studies/tr467.html.
National Toxicology Program (NTP). (2000) Toxicology and carcinogenesis studies of 2-
butoxyethanol (Cas No. 111-76-2) in F344/N rats and B6C3Fj mice (inhalation studies).
Washington, DC: U.S. Department of Health and Human Services, National Institutes of Health.
(National Toxicology Program technical report series no. TR-484). Available:
http://ntp-server.niehs.nih.gov/htdocs/LT-studies/tr484.html.
Nelson, B. K.; Setzer, J. V.; Brightwell, W. S.; Mathinos, P. R.; Kuczuk, M. H.; Weaver, T. E.;
Goad, P. T. (1984) Comparative inhalation teratogenicity of four glycol ether solvents and an
amino derivative in rats. Environ. Health Perspect. 57: 261-271.
Nyska, A., Maronpot, R. R., and Ghanayem. B. I. (1999a). Ocular Thrombosis and Retinal
Degeneration in Female Rats by 2-Butoxyethanol. Human and Exp. Toxicol. 18:577-582.
Nyska, A., Maronpot, R. R., Long, P. H., Roycroft, J. H., Haliy, J. R., Travlos, G. S., and
Ghanayem. B. I. (1999b). Disseminated Thrombosis and Bone Infarction in Female Rats
Following Inhalation Exposure to 2-Butoxyethanol. Toxicol. Pathol. 27:287-294
Park, J.; Kamendulis, L. M.; Klaunig, J. E. (2002a) Effects of 2-butoxyethanol on hepatic oxidative
damage. Toxicol. Lett. 126: 19-29.
Park, J.; Kamendulis, L. M.; Klaunig, J. E. (2002b) Mechanisms of 2-butoxyethanol
carcinogenicity: studies on Syrian hamster embryo (SHE) cell transformation. Toxicol. Sci. 68:
43-50.
Al-25
-------
Pitot, H. C. (2002) Fundamentals of Oncology, 4th edition, 998 pp. New York: Marcel Dekker.
Poet, TS; Soelberg, JJ; Curry, TL; Studniski, KG and Corley, RA. (2002). In Vivo kinetic studies
with 2-butoxyethanol Part 1: Target tissue dosimetry. Battelle Memorial Institute, Battelle Project
No. 40974, Final Report (Draft) for the Ethylene Glycol Ethers Panel of the American Chemistry
Council, Arlington, VA.
Poet, T. S.; Soelberg, J. I; Weitz, K. K.; Mast, T. 1; Miller, R. A.; Thrall, B. D.; Corley, R. A.
(2003) Mode of action and pharmacokinetic studies of 2-butoxyethanol in the mouse with an
emphasis on forestomach dosimetry. Toxicol. Sci. 71: 176-189.
Popp and Marsman, 1991 cited in Ghanayem et al. (1994)
Sills, R. C.; Ton, T.; Hong, L.; Roycroft, J.; Devereux, T. R. (2000) H-ras codon 61 mutation
spectra in forestomach neoplasms from B6C3F1 mice exposed to 2-butoxyethanol for 2 years. In:
National Toxicology Program (NTP). (2000) Toxicology and carcinogenesis studies of 2-
butoxyethanol (Cas No. 111-76-2) in F344/N rats and B6C3Fj mice (inhalation studies).
Washington, DC: U.S. Department of Health and Human Services, National Institutes of Health;
pp. 286-290. (National Toxicology Program technical report series no. TR-484).
Sleet, R. B.; Price, C. J.; Marr, M. C. (1989) Teratologic evaluation of ethylene glycol monobutyl
ether (CAS no. 111-76-2) administered to Fischer-344 rats on either gestational days 9 through 11
or days 11 through 13 (final report). Research Triangle Park, NC: National Toxicology Program;
report no. NTP-89-058. Available from: NTIS, Springfield, VA; PB89-165849.
Slesinski, R. S.; Weil, C. S. (1989) In vitro mutagenesis studies: 3-test battery. Export, PA: Union
Carbide Corporation; report no. 43-26; EPA/OTS docket no. 86-0000946.
Smith, J. G.; German, J. B. (1995) Molecular and genetic effects of dietary derived butyric acid.
Food Technol. 49: 87-90.
Swenberg, J. A.; Barrow, C. S.; Boreiko, C. J.; Heck, H.d'A.; Levine, R. J.; Morgan, K. T.; Starr, T.
B. (1983) Non-linear biological responses to formaldehyde and their implications for carcinogenic
risk assessment. Carcinogenesis 4: 945-952.
Tennant, R.; Ashby, J. (1991) Classification according to chemical structure, mutagenicity to
Salmonella and level of carcinogenicity of a further 39 chemicals tested for carcinogenicity by the
U.S. National Toxicology Program. Mutat. Res. 257: 209-227.
Tyler, T. R. (1984) Acute and subchronic toxicity of ethylene glycol monobutyl ether. Environ.
Health Perspect. 57: 185-191.
Al-26
-------
Udden, M. M. (1994) Hemolysis and decreased deformability of erythrocytes exposed to
butoxyacetic acid, a metabolite of 2-butoxyethanol: II. Resistance in red blood cells from humans
with potential susceptibility. J. Appl. Toxicol. 14: 97-102.
Udden, M. M., 2000. Rat erythrocyte morphological changes after gavage dosing with 2-
butoxyethanol: a comparison with the in vitro effects of butoxyacetic acid on rat and human
erythrocytes. J. Appl. Toxicol. 20: 381-387.
Udden, M. M. (2002) In vitro sub-hemolytic effects of butoxyethanol acid on human and rat
erythrocytes. Toxicol. Sci. 69: 258-264.
U.S. Environmental Protection Agency. (1999) Toxicological Review of ethylene
glycol monobutyl ether (EGBE) (CASRN 111-76-2) in Support of Summary
Information on Integrated Risk Information System (IRIS). National Center for
Environmental Assessment, Washington, DC. Available online from:
http://www.epa.gov/iris.
U.S. Environmental Protection Agency (EPA). (2004) External Peer Review - Ethylene Glycol
Butyl Ether (EGBE). Final Report prepared for the EPA National Center for Environmental
Assessment, Office of Research and Development, Integrated Risk Information System under
interagency agreement No. DW-89939822-01-0 between EPA and the U.S. Department of Energy,
Oak Ridge Institute for Scientific Education (ORISE).
Villalobos-Pietrini, R.; Gomez-Arroyo, S.; Altamirano-Lozano, M.; Orozco, R.; Rios, P. (1989)
Cytogenic effects of some cellosolves. Rev. Int. Contam. Ambiental 5: 41-48.
Voutilainen, M.; Sipponen, P.; Mecklin, J.-P.; Juhola, M.; Farkkila, M. (1999) Gastroesophageal
reflux disease: prevalence, clinical, endoscopic and histopathological findings in 1,128 consecutive
patients referred for endoscopy due to dyspeptic and reflux symptoms. Digestion 61: 6-13.
Wier, P. J.; Lewis, S. C.; Traul, K. A. (1987) A comparison of developmental toxicity evident at
term to postnatal growth and survival using ethylene glycol monoethyl ether, ethylene glycol
monobutyl ether and ethanol. Teratog. Carcinog. Mutagen. 7: 55-64.
Al-27
-------
ATTACHMENT 2
Liver Hemangiosarcoma and Hepatocellular Carcinoma in Male Mice
The framework for a cancer mode of action analysis proposed by EPA (U.S.
EPA, 1999a; 2003) will be used here to summarize what is currently known about
EGBE's mode of action. In general, the current draft of the cancer guidelines requires
discussion of the proposed mode of action, the strength of its supporting database, whether it is
believed to be operative in humans, and whether any human subpopulation is apt to "qualitatively
respond to the mode of action differently than the general population." The framework for this
approach is applied here to assess the mode of action for EGBE-induced liver hemangiosarcomas
and hepatocellular carcinomas in male mice and their relevance to humans.
Postulated Mode of Action
In the EGBE IRIS assessment (U.S. EPA, 1999b), EPA determined that available
information did not allow for a definitive statement regarding the mode of action for the increase in
the incidence of hemangiosarcomas in the livers of male mice. Since that time, considerable
research has been completed in this area, and several scientists have postulated a hemolysis-
mediated mode of action for the formation of these tumors (Klaunig et al., 1998; Kamendulis et al,
1999; Xue et al., 1999; Siesky et al., 2002; Foster, 2000; Boatman, 2000; Park et al., 2002a,b;
Klaunig, 2002). In addition, the Agency's 1999 assessment did not consider a role for EGBE in the
formation of hepatocellular carcinomas, largely due to the statement in the NTP (2000a) report that
"the increased incidence of hepatocellular carcinoma in 250 ppm males and the decreased
incidences of hepatocellular adenoma in 125 and 250 ppm females were interpreted as normal
variations based upon chance rather than effects associated with exposure to 2-butoxyethanol
[EGBE]." However, NTP also described these findings as "uncertain," and given the dose-response
trend observed( 10/50, 11/50, 16/49, 21/49) along with increased early mortality in the mid and
high-exposure groups, it was considered prudent to consider EGBE's potential role in the induction
or promotion of this tumor type as well. The following is a nine step summary of the mode of
action that has been proposed for the formation of these tumors.
1. EGBE is metabolized to 2-butoxyacetaldehyde (BAL) which is subsequently oxidized to 2-
butoxyacetic acid (BAA).
2. BAA causes hemolysis of red blood cells and an increase in hemoglobin levels.
3. Hemosiderin released from the excess hemoglobin is taken up by and stored in phagocytic
(e.g., Kupffer) cells of the spleen and liver.
4. IncreasedDNA synthesis of endothelial cells18 occurs due to one or more of the following:
18Endothelial cells are epithelial cells that line the sinusoidal cavities of blood and are proximate to Kupffer
cells; hemangiosarcomas are derived from these cells.
A2-1
-------
a. Generation of reactive oxygen species (ROS) from iron within Kupffer cells (Siesky
et al., 2002; Park et al, 2002a,b) and perhaps from within hepatocytes and sinusoidal
endothelial cells (Foster, 2000)19 by Fenton or Haber-Weiss reactions.
b. Activation of Kupffer cells to produce cytokines/growth factors which suppress
apoptosis and promote cell proliferation (Siesky et al., 2002).
5. Oxidative DNA damage in endothelial cells is produced by reactive oxygen species.
6. Modulation of endothelial cell gene expression occurs.
7. Proliferation of endothelial cells occurs.
8. Promotion of endothelial cells to tumor forming cells.
9. Neoplasms form.
This hypothesized mode of action would suggest that the pathology of hemangiosarcoma and
hepatocellular carcinoma development from EGBE exposure is nonlinear and dependent upon the
level of hemolysis, the amount of iron build-up within the target cell population, and the DNA
repair capacity of that cell population.
Strength, Consistency, and Specificity of Association of Tumor Response with Key Events
Steps 1 and 2, the metabolism of EGBE to BAA and the association of BAA with hemolytic
effects, have been clearly established in multiple in vivo and in vitro tests involving both sexes of
several species, including rats, mice, rabbits, guinea pigs, dogs, monkeys, and humans (U.S. EPA,
1999b). A possible indication of the importance of hemolysis to tumor formation is given by the
fact that splenic hematopoietic cell proliferation in the mice with liver hemangiosarcomas was
reported to be more severe (average severity grade of 3.2 for all dose groups) than in mice that did
not develop hemangiosarcomas (average severity grade of 2.3).
Step 3 has been verified through the observation of hemosiderin (iron) within Kupffer cells
(phagocytic cells that line the walls of the sinusoids), and hepatocytes (epithelial cells of the liver)
following prolonged breakdown of the red blood cells in both sexes of rats and mice exposed to
EGBE (NTP, 2000a; Ghanayem and Sullivan, 1993; Ghanayem et al., 1987a,b; Krasavage, 1986;
Kamendulis et al., 1999; Siesky et al., 2002). Of the steps listed above, this one has been found to
have the strongest association with tumor response. Two recent analyses of carcinogenicity studies
of B6C3F1 mice at NTP found a highly significant (p<0.001) association between liver
hemangiosarcoma and Kupffer cell hemosiderin pigmentation, particularly when pigmentation is
observed subchronically, that is limited to male mice (Nyska et al., 2004; Gift, 2005).
19It is possible that a continued cycle of hemolysis and hemoglobin release can eventually overwhelm
normal phagocytic mechanisms causing iron to appear within cells which normally would not contain significant
amounts of hemosiderin, such as the hepatocytes and sinusoidal endothelial cells. However, the necessity and extent
of such build-up are not clear (see discussion under "Strength, Consistency, and Specificity of Association of Tumor
Response with Key Events'").
A2-2
-------
A recent gavage study (Siesky et al., 2002) provides strong support for steps 4 and 5 by
providing evidence of increased oxidative DNA damage (8-hydroxyguanosine, OHSdG), increased
lipid peroxidation (malonaldehyde) and decreased antioxidant (Vitamin E) levels and increased
endothelial cell DNA synthesis in rats and mice subchronically exposed to EGBE. The more
pronounced response to these effects in mice observed by these authors is consistent with the NTP
(2000) observation of hemangiosarcomas in mice but not rats.20 Other investigators (Kamendulas
et al., 1999; Park et al., 2002a) provide additional in vitro evidence that hepatic oxidative stress by
EGBE is mediated through reactive oxygen species formed from increased iron deposition resulting
from hemolysis, rather than through a compound- or metabolite-specific mechanism. The role of
oxidative damage resulting from iron exposure was also supported by results demonstrating an
inhibition of endothelial cell DNA damage by supplementation with the antioxidant vitamin E
(Reed et al., 2003; Klaunig and Kamandulis, 2005; Siesky et al., 2002). The activation of Kupffer
cells (step 4b), either through red blood cell hemolytic components and/or iron accumulation in the
Kupffer cell, is proposed to result in the production of cytokines, possibly including vascular
endothelial growth factor, a growth factor whose importance has recently been implicated in the
induction of hemangiosarcomas in rodents (Klaunig and Kamendulis, 2005).
Step 6, has not been shown directly for endothelial cells, but the induction of oxidative
stress and oxidative damage has been shown to modify gene expression in mammalian cells. In
addition to inducing DNA damage and lipid peroxidation, the production of reactive oxygen species
can alter gene expression, resulting in stimulation of cell proliferation and/or inhibition of apoptosis
(Klaunig and Kamandulis, 2005; Nyska et al., 2002; Muller et al., 1997; Manna et al., 1998).
Step 7, endothelial cell proliferation is required prior to the formation of
hemangiosarcomas. The induction of endothelial cell proliferation by 2-butoxyethanol has been
demonstrated in vivo in the mouse but not in the rat (Siesky et al., 2002) at doses that produced
hemangiosacromas in the mouse liver (NTP, 2000).
The final two steps (8 and 9) are consistent with the lack of direct genotoxicity
demonstrated for EGBE and the high background of spontaneous endothelial neoplasms in the male
mouse liver relative to the rat (Klaunig and Kamendulis, 2005). The premise that the effects seen
in the male mouse are the result of tumor promotion mechanisms is also supported by data
suggesting a relationship between early onset hemosiderin buildup in the Kupffer cells and tumor
formation in male mice (see discussion of Temporal Association below).
20 The fact that NTP (2000) exposed rats to lower concentrations of EGBE could also be a factor.
A2-3
-------
Table A2-1: Incidence of Liver Hemangiosarcomas in NTP Chemicals Causing Increased Hemosiderin in Kupffer Cells
SEX/SPECIES/CHEMICAL/NTP TR
HEMOSIDERIN |SC3
HEMANGIOSARC.
HEPATO. CARC.
H. CARC. or
TYPE
MALE RATS (F344)
2-Butoxyethanol (EGBE) - TR-484
Butyl benzyl phthalate - TR-458
p-Chloroaniline Hydrochloride - TR-351
o-Nitroanisole - TR-416
Pyridine - TR-470
23/50, 30/50, 34/50, 42/50
2/60, 1/60, 6/60, 6/60
1/49, 0/50, 0/49, 26/49
0/20, 1/20, 18/20
4/50, 11/49, 20/50, 25/501
yes
no
no
no
no
0/50, 0/50, 1/50, 0/50
None Reported
None reported
None Reported
None Reported
0/50, 1/50, 0/50, 1/50
0/60, 0/60, 1/60, 0/60
1/49, 1/50, 1/49, 0/49
1/20, 0/20, 0/20
0/50, 0/49, 1/50, 0/50
1/50, 3/50, 0/50, 2/50
2/60, 1/60, 1/60, 4/60
1/49, 1/50, 1/49, 0/49
1/20, 0/20, 0/20
1/50, 1/49, 1/50, 3/50
Inh.
Feed
Gav.
Feed
Water
FEMALE RATS (F344)
2-Butoxyethanol (EGBE) - TR-484
Butyl benzyl phthalate - TR-458
CI Pigment Red 3 - TR-407
o-Nitroanisole - TR-416
Pyridine - TR-470
15/50, 19/50, 36/50, 47/50
4/60, 1/60, 6/60, 10/60
0/50, 3/50, 14/50, 41/50
8/20, 2/20, 20/20
6/50, 2/50, 6/50, 17/50
yes
no
no
no
no
None Reported
None Reported
None Reported
None Reported
None Reported
None Reported
None Reported
None Reported
None Reported
None Reported
None Reported
0/60, 0/60, 2/60, 2/60
0/50, 0/50, 1/50, 10/50
None Reported
None Reported
Inh.
Feed
Feed
Feed
Water
MALE MICE (B6C3F1)
2-Butoxyethanol (EGBE)
)-Chloroaniline Hydrochloride - TR-351
)-Nitroaniline - TR-418
'entachloroanisole - TR-414
CI Pigment Red 3 - TR-407
o-Nitroanisole - TR-416
0/50, 0/50, 8/49, 30/49
0/50, 0/49, 0/50, 50/50
1/50, 1/50, 8/50, 50/50
1/50, 50/50, 50/502
0/50, 5/50, 30/50, 41/50
0/50, 0/50, 3/50, 16/50
yes
yes
yes
yes
no
no
0/50, 1/50, 2/49, 4/49
2/50, 2/49, 1/50, 6/50
0/50, 1/50, 2/50, 4/50
2/50, 8/50, 10/50
0/50, 1/50, 1/50, 0/50
2/50, 2/50, 1/50, 0/50
10/50, 11/50, 16/49, 21/49
3/50, 7/49, 11/50, 17/50
10/50, 12/50, 13/50, 6/50
9/50, 16/50, 12/50
5/50, 10/50, 8/50, 4/50
7/50, 12/50, 11/50, 7/50
30/50, 24/50, 31/49, 30/49
11/50, 21/49, 20/50, 21/50
25/50, 26/50, 25/50, 13/50
26/50, 34/50, 24/50
12/50, 16/50, 16/50, 19/50
21/50, 32/50, 45/50, 32/50
Inh.
Gav.
Gav.
Gav.
Feed
Feed
FEMALE MICE (B6C3F1)
2-Butoxyethanol (EGBE) - TR-484
p-Chloroaniline Hydrochloride - TR-351
CI Pigment Red 3 - TR-407
p-Nitroaniline - TR-418
Pentachloroanisole - TR-414
0/50, 5/50, 25/49, 44/50
0/50, 0/50, 1/50, 46/50
2/50, 1/50, 1/49, 29/50
1/50, 1/50, 4/50, 39/50
0/50, 37/50, 48/50
yes
yes
no
no
yes
0/50, 0/50, 1/49, 0/50
1/50, 0/50, 0/50, 1/50
0/50, 1/50, 0/49, 0/50
1/50, 1/50, 0/50, 0/50
0/50, 0/50, 1/50
10/50, 12/50, 13/49, 10/50
1/50, 2/50, 0/50, 3/50
4/50, 8/50, 2/49, 1/50
7/52, 6/50, 10/51, 9/51
4/50, 2/50, 2/50
10/50, 12/50, 13/49, 10/50
6/50, 8/50, 6/50, 11/50
10/50, 14/50, 4/49, 9/50
17/52, 17/50, 21/51, 16/51
11/50, 10/50, 14/50
Inh.
Gav.
Feed
Gav.
Gav.
1 Pigment identified as hemosiderin, but not explicitly reported in Kupffer Cells
2 Authors reported that Kupffer cell pigment did not contain iron, bile or PAS-positive materials according to "appropriate staining
procedures." They speculate that it may consist of porphyrins known to be produced from exposure to chlorinated hydrocarbons, the
possibility that it may have consisted of hemosiderin "was not entirely eliminated."
3 Chemicals that caused hemosiderin accumulation in Kupffer cells following subchronic exposure are identified with a "yes" in this
column.
A2-4
-------
Temporal Association
Key steps in the proposed mode of action - hemolysis, hemosiderin build-up, and oxidative
damage - have all been observed in subchronic or shorter duration rat and mouse studies (NTP, 2000a;
Kamendulis et al., 1999; Siesky et al., 2002) well in advance of tumor formation. In mice, increased
endothelial cell DNA synthesis was observed at exposure days 7 and 14 in mice, and increased
hepatocyte DNA synthesis was observed at 90 days. No increase in the DNA synthesis of either cell
type was observed in rats at any time point. This may be an indication of the importance of early life
stage damage to the DNA of these cell types. In addition, mice show evidence of a more sustained
hemolytic response to EGBE than rats. Mice experienced an increase in liver and splenic
hematopoietic cell proliferation throughout the 2-year NTP (2000a) study, while rats did not show
evidence of a sustained hemolytic response21 and do not develop hemangiosarcomas.
Table A2-1 presents the incidence of liver hemangiosarcoma, hepatocellular carcinoma and
hepatocellular carcinoma or adenoma for all chemicals identified in NTP studies22 as causing
hemosiderin pigmentation of the Kupffer cells following exposure to EGBE.23 A potentially important
difference between the four chemicals that caused liver hemangiosarcoma and/or hepatocellular
carcinoma in male mice (EGBE; p-chloroaniline; p-nitroaniline; and pentachloroanisole24) and the two
that did not (CI pigment red 3 and o-nitroanisole) is that only the former four chemicals appear to have
induced hemosiderin buildup in the Kupffer cells early, by week thirteen. In the studies of the latter
two chemicals, hemosiderin buildup was not as prominent and was not observed until the end of the 2-
year study. In the case of EGBE, it appears that early buildup of hemosiderin combined with early
increases in endothelial cell and hepatocyte DNA synthesis results in a longer exposure of cells to
oxidative damage via iron-generated radicals (step 4). This would be consistent with a mechanism
involving a continuing cycle of damage and repair and accumulation of DNA mutations (steps 5 and
6). Thus, it is possible that the two chemicals did not induce liver tumors because they are not as
potent hemolytic agents and because of temporal differences in their initiation of hemosiderin buildup.
Dose-Response Relationships
Liver tumors were only increased over controls at doses that caused significant
hemosiderin buildup in Kupffer cells within the first 13 weeks of exposure. The
dose-response curves for the tumor and hemosiderin endpoints are both nonlinear. All
21Though rats are initially more sensitive to hemolysis than mice, and were used to derive the EPA RfC for
EGBE, they tend to compensate for the effects of EGBE after a few months. This increased tolerance is documented
in the IRIS file for EGBE and is evidenced by a lack of induction of splenic hematopoiesis at the end of the 2 year
NTP (2000a) study.
22Data for Tables A2-1 were obtained via a search of NTP's Post-PWG TDMS database
(http://ntp-apps.niehs.nih.gov/postpwg/webapp/open.cfm).
23Chemicals were not included if they caused pigmentation that was not believed to be hemosiderin, such as
bile or porphyrin pigmentation, and that was not observed in Kupffer cells. Chemicals that caused Kupffer cell
pigmentation were included if an association with hemosiderosis could not be ruled out.
24It is not clear that the "yellow-brown granules" found in Kupffer cells of rats and mice exposed to
pentachloroanisole contained hemosiderin (NTP, 1993a). "Appropriate staining procedures" did not reveal iron, and
there was no evidence of hemolytic activity. The authors suggested that they may consist of one or more porphyrins,
but none were identified. Thus, the possibility that the pigmentation consisted of hemosiderin "was not entirely
eliminated."
A2-5
-------
key events and tumor effects depend on the dose rate.
Hemangiosarcomas - As can be seen from the EGBE-specific data in Table A2-1, an increase in liver
hemangiosarcoma was observed in chronic studies of male mice for EGBE and the three other
chemicals (p-chloroaniline, p-nitroaniline and pentachloroanisole) that have been shown to cause a
hemosiderin buildup in the Kupffer cells of mice after just 13 weeks of exposure. The tumor response
data for these four chemicals were modeled using the Agency's benchmark dose software (BMDS),
and each represents a significant dose-response trend. In each case, the highest response was
significantly different from both concurrent and historical controls. For all but pentachloroanisole,
only the highest dose was significantly increased over controls. The dose-responses for endpoints
describing alleged precursor effects, splenic hematopoietic cell proliferation, and liver hemosiderin
accumulation are coincident to these tumor effects and dose-related (U.S.EPA, 1999b), as would be
expected if these endpoints are representative of precarcinogenic effects.
Hepatocellular carcinomas - An increasing dose-response trend was also observed for the incidence of
hepatocellular carcinomas following EGBE chronic exposure. In the EGBE study, hemangiosarcomas
and hepatocellular carcinomas were only increased over concurrent controls at the high dose level.
Dose-responses for several hemolytic effects were observed in rats, but liver tumors were not
increased in rats at any dose. However, the high dose in the rat study was half the high dose in the
mouse study.
Biological Plausibility and Coherence of the Database
Oxidative damage is increasingly recognized as playing an important role in the pathogenesis
of several diseases, including cancer and cardiovascular disease (Lesgards et al., 2002). In support of
the proposed hypothesis, increased reactive oxidative stress is known to accompany the release of
large amounts of iron from hemolysis (Ziouzenkova et al 1999). If EGBE causes oxidative stress by
this mode of action, one would expect to observe the production of protein and DNA damage, some of
which will occur via the production of adducts of OHSdG, and a decrease in antioxidant (e.g., Vitamin
E) levels following EGBE exposure (Yamaguchi et al 1996; Wang et al 1995; Houglum et al 1997).
This was verified by both Kamendulas et al. (1999) and Siesky et al. (2002) who measured a dose-
dependent increase in levels of OHSdG and MDA and a decrease in vitamin E levels in the livers of
mice acutely and sub chronically exposed to EGBE.
Liver hemangiosarcomas develop from the endothelial cell component of the vascular
sinusoidal cells of the liver (Frith and Ward, 1979). The fact that iron (hemosiderin), which is known
to accumulate in cells of rodent livers following EGBE exposure, can produce hydroxyl radicals in
combination with oxidative by-products via the Fenton reaction (Kamendulis et al., 1999) has led
several scientists (Xue et al., 1999; Siesky et al., 2002; Foster, 2000, Boatman, 2000; Kamendulis et
al., 1999; Bachowski et al, 1997; Klaunig et al., 1995) to suggest that male mouse liver
hemangiosarcomas result from increased oxygen radical damage and an associated increase in
endothelial cell DNA synthesis caused by excess iron from hemolysis. The damaging effects of iron
overload to liver sinusoidal cells in rats following a single ip injection of 200 mg iron/kg (Junge et al.,
2001) lend support to this hypothesis. Additional support for this hypothesis is provided by the fact
that endothelial cells do appear to be more susceptible than other cells (e.g., hepatocytes or Kupffer
cells) to oxidative stress (DeLeve, 1998; Spolarics, 1999).
A2-6
-------
While largely supported by recent laboratory research, several questions remain concerning the
postulated mode of action. The questions outlined below may be answerable through further research.
Is the severity and incidence of observed Kupffer cell pigmentation at terminal sacrifice sufficient
to suggest that oxidative stress from hemosiderin buildup in these cells is solely responsible for the
increased incidence of liver hemangiosarcoma in the male mouse? NTP (2000a) has suggested
that hemosiderin buildup is not related to the formation of liver hemangiosarcomas because
Kupffer cell pigmentation was only of minimal severity and only observed in three of the four
high-dose male mice that developed this tumor. However, given the apparent susceptibility of
male mouse endothelial cells, it is reasonable to hypothesize that the reported minimal
pigmentation observed over such a large percentage of high-dose animals (61% in the 250 ppm
exposure group versus 0% in controls) could have caused the marginal increase in
hemangiosarcomas reported for this dose group (8% versus 0% in controls, with a 2.5% mean
historical background rate). Additionally, the fact that Kupffer cell pigmentation was not reported
for one of the four mice with liver hemangiosarcomas does not preclude the proposed mechanism
because (a) the Perls Prussian blue stain for iron performed to determine the presence of
hemosiderin is not a particularly sensitive test (Ohio, 2002; Ali et al, 2003) and (b) the observation
of a liver hemangiosarcoma in 1/50 without hemosiderin buildup is not inconsistent with the rate
of incidence in historical controls, reported to be 0-6% (mean of 2.5%) over all NTP studies of
male mice (NTP, 2000b).
What is the relation between Kupffer cell pigmentation and the hepatocyte and sinusoidal
endothelial cells, that demonstrate no pigmentation and from which the hemangiosarcomas arise?
While the Siesky et al. (2002) study does suggest a temporal association between liver DNA
adduct formation and increased endothelial cell DNA synthesis (both having been observed on the
7th exposure day), it does not confirm whether observed endothelial cell DNA synthesis was the
result of direct DNA damage or Kupffer cell activation leading to increased endothelial cell
proliferation, both of which can lead to unscheduled DNA synthesis. Indicators of oxidative
damage (increased OHSdG and malondialdehyde levels) were only recorded for total liver, and
iron was not observed within the endothelial cells. Siesky et al. (2002) and Rose et al. (1997;
1999) have suggested that increased hepatocyte and endothelial cell DNA synthesis and oxidative
damage can occur without direct iron involvement in the endothelial cells through indirect Kupffer
cell activation leading to the release of cytokines and growth regulatory molecules that contribute
to the induction of DNA synthesis and damage. In addition, reactive oxygen species generated in
the Kupffer cells may impacting endothelial cell DNA due to the proximity of these two cells
within the liver. Given the apparent predisposition of male mouse hepatocytes and endothelial
cells to carcinogenic transformation, these indirect factors may be enough to trigger mutations
within endothelial cell DNA and tumor formation. However, Perl's iron staining is a relatively
insensitive technique (Ohio, 2002)25, and redox active iron can associate with protein, often in one
of the forms of ferritin, in many different kinds of cells (Knutson and Wessling-Resnick, 2003),
leaving open the possibility that iron stores located directly within endothelial cells may also be
contributing to the observed DNA synthesis and damage.
If hemosiderin buildup in Kupffer cells is important for the development of EGBE-induced liver
hemangiosarcomas, why was the incidence of this tumor not increased in the female mouse for
which hemosiderin pigmentation was at least as high as in the males? This observation is not
25A newly developed fluorescent calcein method is a more sensitive measure of LMW iron in biological
fluids (Ali etal., 2003).
A2-7
-------
unique to EGBE. In fact, all NTP chemicals (Table A2-1) that induced early onset hemosiderin
buildup in conjunction with an increased incidence of this tumor exhibited this apparent male
specificity. Inherent differences between the antioxidant capacity of male versus female mice
related to hormonal modulation of antioxidative enzymes has been offered (Foster, 2000; Nyska et
al., 2004) as a possible explanation for the lack of a hepatic tumorogenic response in the females.
Though there is no direct evidence to support this hypothesis, the historical incidence of
hemangiosarcomas does appear to be considerably higher in male versus female mice. NIH/NTP
has observed liver hemangiosarcomas in 105 of 4183 (2.51%) male versus just 35 of 4177 (0.84%)
female historical controls (NTP, 2000b; Klaunig, 2002). A more definitive answer to this question
could come from additional research, perhaps involving measurement of the relative capabilities of
the antioxidant systems of male and female mouse liver endothelial cells.
Is the proposed relationship between the buildup of hemosiderin in Kupffer cells and the
development of liver hemangiosarcoma and hepatocellular carcinoma supported by studies of
other chemicals? NTP (2000a) reported that they did not find an association between hemosiderin
deposition in the liver and "liver neoplasms (adenomas, carcinomas, or hemangiosarcomas)" in the
79 male and 103 female mice that had chemically related liver neoplasms at the end of an NTP
study. However, the NTP analysis did not focus on the specific relationship between hemosiderin
buildup in the Kupffer cells of male mice and liver hemangiosarcoma and/or hepatocellular
carcinoma. A more recent analysis of 130 two-year carcinogenicity studies of B6C3F1 mice at
NTP found a highly significant (p<0.001) association between liver hemangiosarcoma and Kupffer
cell pigmentation that is limited to male mice (Nyska et al., 2004). Nyska et al. (2004) report that
a chemical associated with Kupffer cell pigmentation related to hemosiderosis has a high
likelihood (3/4 or 75%) of producing liver hemangiosarcoma, compared with a very low likelihood
(3/126 or 2%) in the absence of such pigmentation. The four chemicals identified by Nyska et al.
(2004) as causing this Kupffer cell pigmentation are EGBE, p-chloroaniline hydrochloride, p-
nitroaniline and o-nitroanisole. Table A2-1 identifies two other chemicals pentachloroanisole and
CI pigment red 3, for which Kupffer cell pigmentation has been reported, but the relationship to
hemosiderosis has been questioned.26 As is indicated in Table A2-1, there appears to be no
association between hemosiderin in Kupffer cells and liver hemangiosarcomas for rats and female
mice.
If hemosiderin deposition and increased hematopoietic cell proliferation can lead to tumor
formation, why were no neoplasms noted in the spleen, where these effects were also observed?
Cells of the spleen, such as phagocytes, serve to rid the peripheral circulation of damaged red
blood cells and would be expected to encounter high levels of hemoglobin and hemosiderin in the
process, thereby requiring the need for greater protection against any harmful effects from such
buildup. The fact that phagocytic cells seem to have a higher antioxidant capacity than sinusoidal
endothelial cells lends some support to this hypothesis (DeLeve 1998; Foster, 2000). Thus, the
proposed mode of action is not inconsistent with the fact that hemosiderin buildup in splenic cells
did not lead to the type of tumor formation observed in the liver, particularly given the very slight
increase in liver tumor incidence observed.
26Authors of the NTP Pentachloroanisole study (TR-414) did not identify iron in the Kupffer cell
pigmentation, but stated that an association with hemosiderosis could not be ruled out. Author of the NTP CI
pigment red 3 study (TR-407) stated that Kupffer cell pigmentation was related to hemosiderosis, but Nyska et al.
(2004) attributed it to a compound metabolite. Pentachloroanisole exposure was associated with an increased
incidence of hemangiosarcomas, but CI pigment red 3 did not cause an increased incidence of hemangiosarcomas.
A2-8
-------
Other Possible Modes of Action for Liver Tumor Development in Male Mice
Although certain key events in EGBE's mode of action for the development of liver
tumors in male mice are fairly well described and verified, (induction of hemolysis by BAA, induction
of cell proliferation, and clonal expansion), some alternate considerations (also supported by scientific
literature) may be involved. Reactive oxygen species can potentially be derived from two sources:
iron overloading in the liver (through Fenton and Haber-Weiss reactions) and/or from Kupffer cell
activation. Via either source, oxygen radicals can induce oxidative damage to DNA and lipids as
documented in liver following EGBE treatment (Seisky et al., 2002). The activation of Kupffer cells
(through phagocytosis of red blood cell hemolytic components and/or iron in the Kupffer cell), results
in the production of cytokines, possibly including vascular endothelial growth factor that may elicit a
growth response on endothelial cells. In addition to the production of oxidative DNA damage, reactive
oxygen species, whether derived from Kupffer cell activation or other biological processes, can alter
gene expression (e.g. MAP kinase/AP-1, andNFKB) resulting in stimulation of cell proliferation
and/or inhibition of apoptosis (Klaunig and Kamendulis, 2004).
Because of the high background rate for hemangiosarcomas in male mice and the fact that the
relationship between hemosiderin buildup and hemangiosarcomas is only apparent for male mice, it is
reasonable to hypothesize that endothelial cell proliferation arise from the promotion of preexisting
(spontaneously) initiated cells. However, this has not been established. The increased DNA synthesis
and/or oxidative DNA damage caused by EGBE could result in the acquisition of new mutations in
endothelial cells (tumor initiation) rather than a selective clonal expansion of initiated endothelial cells
(tumor promotion).
Another well recognized mechanism for the development of chemically induced liver
hemangiosarcomas involves direct interaction with DNA. This is the mode of action that is recognized
for vinyl chloride and thorotrast, two chemicals that are known to induce hemangiosarcomas in
humans. As discussed in Attachment 1, BAL is the EGBE metabolite considered to have the greatest
potential to interact with DNA as it has been shown to cause in vitro SCE at concentrations ranging
from 0.2 to 1 mM. However, high aldehyde dehydrogenase activity in the liver, as in the forestomach,
is expected to results a very short residence time and low Cmax liver tissue concentrations of BAL.
The Corley (2003) model discussed in Attachment 1 includes the metabolism of EGBE to BAL via
alcohol dehydrogenase and the subsequent metabolism of BAL to BAA via aldehyde dehydrogenase in
both the liver and forestomach. Using rate constants derived from mouse stomach fractions (Green et
al., 2002) and making several assumptions about the use of these enzyme activity data (see discussion
in Attachment 1 under "Biological Plausibility and Coherence of the Database"), Corley (2003)
estimated that 250 ppm EGBE would result in peak Cmax concentrations of 7 EGBE, 0.5 BAL and
3,250 BAA |J,M in liver tissue of male mice at the end of a 6 hour exposure period (Figure A2-1).
Thus, the Corley (2003) PBPK model suggests that the conditions of the in vitro assays which show
BAL to be clastogenic (e.g., no metabolic activation; high, cytotoxic concentrations of BAL) are of
little relevance to expected target organ (liver) environment (e.g., high metabolic activity; low
concentrations of BAL). A recent gavage study performed by Deiringer and Boatman (2004) provides
support for the Corley (2003) model and the predicted low levels of the BAL metabolite in liver
A2-9
-------
BE & Metabolites in Liver
BE & Metabolites in Liver
3,500 -i
3,000
f 2,500-
n.
I 2,000
2
§ 1,500
o
c
0
0 1,000
500
n
A
/ \
/ \
/ \
/ \
/ \
/ \
/ \
/ \
/ \
/ \
,/_____ v_
12
Time (hr)
50
40
i"
centratior
NJ C.
O C
0
o
10-
(
I
I
I
I
I
| BE
| BAL
| BAA
\
\
\
\ V
) 6 12 18
Time (hr)
Cmax (uM)
BE 7
BAL 0.5
BAA 3,250
Figure A2-1: Concentrations of BE, BAL and BAA
in liver tissues of female mice exposed to 250 ppm
tissue.27 Also, as is discussed in detail in Attachment 1 under Other Possible Modes of Action for
Forestomach Tumor Development in Female Mice, evidence from in vivo and in vitro genotoxicity
assays do not support the idea that BAL would have any significant genotoxicity in vivo.
Further, the mode of action for hemangiosarcoma induction by genotoxic chemicals such as
vinyl chloride and thorotrast involves the initiation of hepatocellular and sinusoidal cell hyperplasia
and sinusoidal compression leading to the development of fibrous septa, generally in the peri-portal
area, out of which eventually develop multiple areas of angiosarcomas (Foster, 2000). EGBE exposure
does not generate this same pattern of effects prior to the development of cancer in mice. The main
non-neoplastic effect of EGBE on the liver is an accumulation of hemosiderin (iron) within the
Kupffer cells following a chronic elevated breakdown of the red blood cells with few other precursor
lesions reported.
27 The Corley (2003) model predicts that the concentrations of BAL in liver tissues of male and female
mice would be 17 and 29 |_iM, respectively, following oral gavage exposure to 600 mg/kg EGBE. The levels of BAL
actually observed in the liver tissue of male and female mice, following oral gavage exposure to EGBE at 600 mg/kg
3 and 4 |_iM, respectively, were even lower than the predicted values (Deiringer and Boatman, 2004).
A2-10
-------
Relevance of Mouse Liver Hemangiosarcomas and Hepatocellular Carcinomas to Humans
According to the latest research findings, it is quite likely that hemolysis represents a key event
in the formation of liver tumors in male mice following EGBE exposure. If the increased incidence of
male mouse liver hemangiosarcoma and hepatocellular carcinoma are indeed secondary effects of
hemolysis, they are not likely to be relevant to humans. Iron overload in humans either exposed to
excess iron or genetically susceptible due to defects in iron metabolism has been associated with the
induction of liver tumors (Mandishona et al., 1998; Stevens et al., 1994). However, humans are not
likely to experience significant iron buildup from EGBE exposure because they are comparatively less
sensitive to the hemolytic toxicity of EGBE than rats and mice. Human volunteers experienced no
hemolysis from controlled laboratory acute inhalation exposures (up to 195 ppm) that caused
significant erythrocyte fragility in rats (Carpenter et al., 1956); and only mild, reversible hemolytic
effects have been observed in humans acutely exposed to oral doses EGBE (400 to 1500 mg/kg) that
have been shown to cause marked, and in some cases irreversible, hemolytic effects in rats (Ghanayem
et al., 1987b; Grant et al., 1985; U.S. EPA, 1999b). In vitro testing suggests that blood concentrations
of the hemolytically active metabolite BAA must reach levels in human blood in excess of 7.5 mM for
even minimal prehemolytic changes to occur. This blood level of EGBE is 15-fold higher than the
blood level at which comparable effects occur in rats (Udden, 1995; 2002) and is significantly higher
than the maximum blood concentrations of approximately 2 mM predicted by PBPK modeling for the
highest, theoretically saturated (~1160 ppm), air concentrations of EGBE (Corley et al, 1994).
As is discussed below, some attempts have been made to identify susceptible human
subpopulations (Udden, 1994; 1995, 2002). While no sensitive human subpopulation has been
identified to date, there is some reason to believe that humans may vary considerably in their ability to
metabolize and excrete EGBE and in their response to EGBE exposure . Both alcohol and aldehyde
dehydrogenase enzymes are polymorphically distributed in humans. As discussed below under
Metabolic and genetic differences., these polymorphisms have been shown to alter rates of metabolism
and elimination of other alcohols and aldehydes (Agarwal and Goedde, 1992). As work in this area
continues, further information on the metabolic or structural differences that result in the lower
sensitivity of human RBCs compared to rat RBCs may eventually illuminate characteristics in the
human population that may indicate increased susceptibility. However, existing work indicates that
the usual human subpopulations of concern, including aged, sickle-cell anemia, hereditary
spherocytosis patients (Udden, 1994) and children (Udden, 2002) are not be sensitive to the hemolytic
effects of EGBE. It should be recognized, however, that effects in humans from chronic exposure to
EGBE have not been studied, and the case reports of acute exposures and short-term (4 hour on
average) in vitro tests discussed above may not accurately reflect the potential for cumulative damage
from low-level chronic EGBE exposure.
In addition to the pharmacodynamic differences described above, there are also
pharmacokinetic differences between the rodent and human response to EGBE exposure (Figure A2-
2). The two main oxidative pathways of EGBE metabolism observed in rats are alcohol
dehydrogenase and O-dealkylation by a cytochrome P450 dealkylase (CYP 2E1) (Medinsky et al.,
1990). The former pathway, which involves the production of the toxic metabolite BAA, is applicable
to both rats and humans. However, unlike rats, approximately two-thirds of the BAA formed by
humans is conjugated with glutamine and glycine (Corley et al., 1997; Rettenmeier et al., 1993). Thus,
the glutamine and glycine detoxification pathways may provide humans with some measure of added
protection from the harmful effects of BAA.
A2-11
-------
CO,
carboligase oxidase dehydrogenase
CH3CH2CH2CH2OH
(Butanol)
HOCH2CH2OH
(Ethylene Glycol)
A
CH3CH2CH2CH2OCH2CH20-Gluc
(EGBE - Glucuronide)
(Rats Only?)
(Rats Only?)
\
CH3CH2CH2CH2OCH2CH20-S03H
(EGBE - Sulfate)
1
(Rats Only?)
CH3CH2CH2CH2OCH2CH2OH
(EGBE)
alcohol dehydrogenase
CH3CH2CH2CH2OCH2CHO
(BAL)
A
CH3CH2CH2CH2OCH2C02-Glu
(BAA-Glutamine)
(Humans Only)
aldehyde dehydrogenase
CH3CH2CH2CH2OCH2C02-Gly
(BAA-Glycine)
(Humans Only)
CH3CH2CH2CH2OCH2C02H
(BAA)
I
CO,
dealkyl carboligase
Figure A2-2: Proposed metabolism of EGBE in rats and humans
(Adapted from Medinsky et al., 1990 and Corley et al., 1997)
A2-12
-------
Relevance to Susceptible Subpopulatiom, Including Children
Differences in susceptibility to hemolysis - The primary noncancer effect of EGBE is hemolysis
of red blood cells, which is caused by its primary metabolite, BAA. There is no evidence for the
existence of human subpopulations with increased susceptibility to red blood cell lysis caused by
BAA. However, Udden (1994) has shown that the RBCs of normal and aged patients and patients
with sickle-cell anemia and hereditary spherocytosis are all equally resistant to the hemolytic effects of
BAA.
Metabolic and genetic differences - Other potentially susceptible subpopulations include
individuals with enhanced metabolism or decreased excretion of BAA. Polymorphisms in alcohol and
aldehyde dehydrogenases could lead to differences in the metabolism and elimination of EGBE in
some humans. Human genetic polymorphisms in alcohol dehydrogenase and aldehyde dehydrogenase
are prevalent in certain ethnic groups (Chan, 1986) and these polymorphisms have been shown to alter
rates of metabolism and elimination of ethanol and acetaldehyde (Agarwal and Goedde, 1992). For
instance, native Americans and approximately 50% Asian people are deficient in aldehyde
dehydrogenases. Aldehyde dehydrogenases comprises more than nine isoforms in humans (Hsu et al.,
1994). A deficiency or loss of one of them (ALDH2), can lead to a nearly complete loss of enzymatic
activity for structurally similar aldehydes (e.g., methoxyacetaldehyde), in human liver mitochondria
fractions (Crabb et al., 1989; Kitagawa et al., 2000). Individuals with atypical alcohol dehydrogenase
and/or deficient aldehyde dehydrogenase appear to be more susceptible to adverse effects from
increased levels of acetaldehyde including facial flushing, general discomfort, acetaldehyde-protein
adducts and alcohol-induced liver diseases (Agarwal and Goedde, 1992). However, the PBPK model
developed by Corley et al. (2004) indicates that individuals with low aldehyde dehydrogenase activity
(l/2 Vmax) would not be expected to accumulate significant BAL levels in the liver or forestomach,28
even following inhalation exposure to a theoretical maximum of 1160 ppm EGBE for 6 hours (U.S.
EPA, 2004).
Haufroid et al. (1997) conducted a human study on workers exposed to EGBE to test the
possible influence of genetic polymorphism for CYP 2E1 on urinary BAA excretion rate. One
exposed individual exhibited a mutant allele with increased cytochrome P450 oxidative activity that
coincided with a very low urinary BAA excretion. However, the researchers did not measure BAA
conjugated to glutamine, an alternative pathway for BAA excretion in humans. Further investigations
on the influence of genetic polymorphism for CYP 2E1 on urinary BAA excretion rate are needed
before any firm conclusions can be drawn.
Gender differences - Slight gender differences have been noted in rodent (Carpenter et al.,
1956; Dodd et al., 1983; NTP, 1993b; NTP, 2000a), rabbit (Tyler, 1984), dog, monkey, and human
studies (Carpenter et al., 1956), with females being consistently more susceptible to the primary
hemolytic effects of EGBE. A number of secondary effects resulting from the hemolytic toxicity of
EGBE, such as effects on the rat liver, kidneys, spleen, and bone marrow and, to a lesser extent, the
thymus, are more pronounced in females (NTP, 1993b). In the process of studying and comparing the
metabolic and cellular basis of EGBE-induced hemolysis, Ghanayem (1989) observed that the blood
from female human volunteers showed a slightly greater sensitivity to incubation with BAA than male
blood.
Predicted BAL concentrations were below 0.001 mM in the liver and 0.004 mM in the GI tract following
inhalation exposure to saturated air concentrations of EGBE. These concentrations are considerably lower than
concentrations of BAL shown to be clastogenic (0.2 mM) or hemolytic (0.5 mM: Ghanayem et al., 1989) in vitro.
A2-13
-------
Age differences - It is generally recognized that children have fewer risk factors for anemia
than adults (Berliner, et al. 1995; Hord and Lukens, 1999), and Udden (2002) has shown that the red
blood cells of children are no more sensitive to EGBE induced hemolytic effects than adult red blood
cells. Childhood exposures can be important determinants of certain cancers (Anderson et al., 2000),
including early childhood bacterial infections (e.g. Helicobacter pylori) (Rowland and Drumm, 1998);
and because bacterial infection (Helicobacter hepaticus) has also been associated with the
development of liver hemangiosarcomas in mice (Nyska et al., 1997), it is not unreasonable to
consider whether early childhood exposure to EGBE might also be more important than adult
exposures towards the formation of EGBE induced liver hemangiosarcomas. However, if the
proposed mode of action for EGBE's involvement in the formation of these tumors in male mice is
correct, there is no reason to believe that exposed human children would experience significant
buildup of iron from hemolysis unless large toxicokinetic or toxicodynamic differences between adults
and children are identified.
The only human toxicity information available on the toxicity of EGBE to children is from the
case study by Dean and Krenzelok (1991), who observed 24 children, age 7 mo to 9 years, subsequent
to oral ingestion of at least 5 mL of glass window cleaner containing EGBE in the 0.5% to 9.9% range
(potentially 25 to 1500 mg EGBE exposures). No symptoms of EGBE irritation, poisoning, or
hemolysis were reported.
Adult (9-13 wk) male F344 rats were significantly more sensitive to the hemolytic effects of
EGBE than were young (4-5 wk) male rats following administration of a single gavage dose of EGBE
at 32, 63, 125, 250, or 500 mg/kg. In concurrent metabolism studies, increased blood retention of
EGBE metabolite BAA (as measured by increased Cmax, AUC, and T1/2) was also found. Additionally,
young rats eliminated a significantly greater proportion of the administered EGBE dose as exhaled
carbon dioxide (CO2) or as urinary metabolites as well as excreting a greater proportion of the EGBE
conjugates (glucuronide and sulfate) in the urine (Ghanayem et al., 1987c; 1990). These researchers
suggested that the pharmacokinetic basis of the age-dependent toxicity of EGBE may be due to a
reduced ability by older rats to metabolize the toxic metabolite BAA to CO2 and a diminished ability
to excrete BAA in the urine.
NTP (1998) also found that young mice (6-7 weeks) eliminated BAA 10-times faster than aged
(19 months) following a 1-day exposure to 125 ppm EGBE. This difference was not as apparent after
3 weeks of exposure. Dill et al. (1998) have suggested that this may be related to a greater sensitivity
to the acute toxicity of EGBE in older animals that appears to be compensated within 2-3 weeks.
Developmental studies, which may also be of possible relevance to this issue, have been
conducted using rats, mice, and rabbits dosed orally, by inhalation or, in one study, dermally (Hardin
et al., 1984; Heindel et al., 1990; Nelson et al., 1984; NTP, 1993b; Sleet et al., 1989; Tyler, 1984; Wier
et al., 1987). Maternal toxicity related to the hematologic effects of EGBE and relatively minor
developmental effects were reported in most studies. No teratogenic toxicities were noted in any of
the studies. It can be concluded from these studies that EGBE is not significantly toxic to developing
fetuses of laboratory animals.
Older rats have been shown to have a reduced ability to metabolize the toxic metabolite BAA
to CO2 and a diminished ability to excrete BAA in the urine (Ghanayem et al.,1987b, 1990). However,
the relevance of this finding to the possible susceptibility of elderly humans is uncertain due to the fact
that humans have conjugation pathways for the excretion of BAA (BAA-Glutamine and BAA-
Glycine) that are not available to the rat.
A2-14
-------
Summary
Available data establish a plausible nonlinear, nongenotoxic mode of action for the moderate
increase observed by NTP (2000) in the incidence of liver tumors in male mice following chronic
inhalation exposure to EGBE. The proposed mode of action suggests that the endothelial cells and
hepatocytes of male mice are sensitive to the formation of the subject neoplasms (as evidenced by the
relatively high background rate of these tumors in male mice) and that excess iron from EGBE-
induced hemolysis can result in sufficient iron-induced oxidative stress to cause the observed,
marginal increase in the incidence of liver hemangiosarcomas and hepatocellular carcinomas in these
animals (NTP, 2000). Given the relatively low sensitivity of humans, including subpopulations such
as children, to the hemolytic effects of EGBE, it appears reasonable to assume that the EGBE RfC and
RfD (EPA, 1999a), which were based on hemolytic effects in female rats, are sufficient for the
prevention of hemolysis and associate tumors in humans.29 However, this determination assumes a
nonlinear mechanism that requires exposure levels to be high enough to cause certain lesions that are
considered to be precancerous. Information is currently inadequate to dismiss the potential
contribution of a linear mechanism associated with the possible mutagenic metabolite B AL. A
definitive determination regarding the appropriateness of a nonlinear approach can not be made until
questions regarding the role of BAL are resolved. As discussed above, additional research (e.g.,
verification of existing PBPK modeling results and improved genotoxicity assays) would assist the
Agency in making a more informed decision concerning the potential for BAL to contribute to the
adverse effects seen in animals following EGBE exposure and use of the proposed nonlinear
assessment approach.
29These analyses are consistent with the nonlinear assessment approach described in existing interim (U.S.
EPA, 1999a) and draft (2003) cancer guidelines.
A2-15
-------
REFERENCES - ATTACHMENT 2
Agarwal, D.P. (1997) Molecular genetic aspects of alcohol metabolism and alcoholism.
Pharmacopsychiatry 30: 79-84.
Agarwal, D.P. and Goedde, H.W. (1992) Pharmacogenetics of alcohol metabolism and alcoholism.
Pharmacogenetics, 2: 48-62.
Ali A., Zhang Q., Dai J., Huang X. (2003) Calcein as a fluorescent iron chemosensor for the
determination of low molecular weight iron in biological fluids. Biometals 16: 285-293.
Anderson, L. M.; Diwan, B. A.; Fear, N. T.; Roman, E. (2000) Critical windows of exposure for
children's health: cancer in human epidemiological studies and neoplasms in experimental animal
models. Environ. Health Perspect. 108(suppl. 3): 573-594.
Bachowski, S.; Kolaga, K. L.; Xu, Y.; Ketcham, C. A.; Stevenson, D.E.; Walborg, E. F. Jr.; Klaunig, J.
E. (1997) Role of oxidative stress in the mechanism of dieldrin's hepatotoxicity. Ann. Clin. Lab. Sci.
27: 196-209.
Berliner, N.; Duffy, T. P.; Abelson, H. T. (1995) Approach to adult and child with anemia. In:
Hoffman, R., ed. Hematology: basic principles and practice. 2nd ed. New York, NY: Churchill
Livingstone; pp. 468-483.
Boatman, R. J. (2000) The significance of mouse liver hemangiosarcomas to the risk of cancer in man
from exposures to ethylene glycol mono-n-butyl ether. Eastman Kodak Company; March 28.
Carpenter, C. P.; Pozzani, U. C.; Wiel, C. S.; Nair, J. H., Ill; Keck, G. A.; Smyth, H. F., Jr. (1956) The
toxicity of butyl cellosolve solvent. AMA Arch. Ind. Health 14: 114-131.
Chan, A.W. (1986) Racial differences in alcohol sensitivity. Alcohol Alcohol 21: 93-104.
Corley, R. A.; Bormett, G. A.; Ghanayem, B. I. (1994) Physiologically-based pharmacokinetics of 2-
butoxyethanol and its major metabolite, 2-butoxyacetic acid, in rats and humans. Toxicol. Appl.
Pharmacol. 129: 61-79.
Corley, R. A.; Markham, D. A.; Banks, C.; Delorme,P.; Masterman, A.; Houle, J. M. (1997)
Physiologically-based pharmacokinetics and the dermal absorption of 2-butoxyethanol vapors by
humans. Fundam. Appl. Toxicol. 39: 120-130.
Corley, R. A. (2003) Preliminary Incorporation of the Intermediate Metabolite, Butoxyacetaldehyde,
in PBPK Model for Butoxyethanol. Personally communicated to Jeff Gift, EPA National Center for
Environmental Assessment - RTF, NC.
Crabb DW, Edenberg HJ, Bosron WF, Li TK. 1989. Genotypes for aldehyde dehydrogenase
deficiency and alcohol sensitivity. The inactive ALDH2(2) allele is dominant. J Clin Invest 83:
314-316.
Dean, B. S.; Krenzelok, E. P. (1991) Clinical evaluation of pediatric ethylene glycol monobutyl ether
poisonings. Vet. Hum. Toxicol. 33: 362.
A2-16
-------
DeLeve, L. D. (1998) Glutathione defence in non-parenchymal cells. Semin. Liver Dis. 18: 403-413.
Dill, J. A.; Lee, K. M.; Bates, D. J.; Anderson, D. J.; Johnson, R. E.; Chou, B. J.; Burka, L. T.;
Roy croft, J. H. (1998) Toxicokinetics of inhaled 2-butoxyethanol and its major metabolite, 2-
butoxyacetic acid, in F344 rats and B6C3F1 mice. Toxicol. Appl. Pharmacol. 153: 227-242.
Dodd, D. E.; Snelling, W. M.; Maronpot, R. R.; Ballantyne, B. (1983) Ethylene glycol monobutyl
ether: acute, 9-day, and 90-day vapor inhalation studies in Fischer 344 rats. Toxicol. Appl. Pharmacol.
68: 405-414.
Foster, J. R. (2000) EGBE (ethylene glycol butyl ether, EGBE): the proposed mode of action in the
induction of hemangiosarcomas in male mouse liver and its significance for
man. Macclesfield, Cheshire, United Kingdom: Central Toxicology Laboratory, Alderley Park; report
no: Ctl/r/1465.
Frith, C. H.; Ward, J. M. (1979) A morphologic classification of proliferative and
neoplastic hepatic lesions in mice. J. Environ. Pathol. Toxicol. 3: 329-351.
Fujimura, M. and Myou, S. (2001) Alcohol-induced asthma. Intern. Med. 40: 557-8.
Ghanayem, B. L; Burka, L. T.; Matthews, H. B. (1987a) Metabolic basis of ethylene glycol monobutyl
ether (2-butoxyethanol) toxicity: role of alcohol and aldehyde dehydrogenases. J. Pharmacol. Exp.
Ther. 242:222-231.
Ghanayem, B. L; Blair, P. C.; Thompson, M. B.; Maronpot, R. R.; Matthews, H. B. (1987b) Effect of
age on the toxicity and metabolism of ethylene glycol monobutyl ether (2-butoxyethanol) in rats.
Toxicol. Appl. Pharmacol. 91: 222-234.
Ghanayem, B. L; Burka, L. T.; Sanders, J. M.; Matthews, H. B. (1987c) Metabolism and disposition of
ethylene glycol monobutyl ether (2-butoxyethanol) in rats. Drug Metab. Dispos. 15: 478-484.
Ghanayem, B. I. (1989) Metabolic and cellular basis of 2-butoxyethanol-induced hemolytic anemia in
rats and assessment of human risk in vitro. Biochem. Pharmacol. 38: 1679-1684.
Ghanayem, B. L; Sanders, J. M.; Clark, A. M.; Bailer, J.; Matthews, H. B. (1990) Effects of dose, age,
inhibition of metabolism and elimination on the toxicokinetics on the toxicokinetics of 2-
butoxyethanol and its metabolites. J. Pharmacol. Exp. Ther. 253: 136-143.
Ghanayem, B. L; Sullivan, C. A. (1993) Assessment of the hemolytic activity of 2-butoxyethanol and
its major metabolite, butoxyacetic acid, in various mammals including humans. Hum. Exp. Toxicol.
12: 305-311.
Ghanayem, B. L, Ward, S M, Chanas, B, and Nyska, A. (2000). Comparison of the acute
hematotoxicity of 2-butoxyethanol in male and female F344 rats. Human Exp. Toxicol. 19:185-192.
Ghanayem. B. L. Long, P., Ward, S M, Chanas, B, and Nyska, A. (2001). Hemolytic Anemia,
Thrombosis, and Infarction in Male and Female F344 Rats Following Gavage Exposure to 2-
Butoxyethanol. Experimental and Toxicologic Pathology 53:97-105.
A2-17
-------
Ohio, A. (2002) Personal communication. Research Triangle Park, NC: U.S. Environmental Protection
Agency, National Health and Environmental Effects Research Laboratory.
Gift, J. S. (2005) U.S. EPA's IRIS assessment of 2-butoxyethanol: the relationship of noncancer to
cancer effects. Toxicology Letters 156: 163-178.
Grant, D; Sulsh, S; Jones, HB; et al. (1985) Acute toxicity and recovery in the hemopoietic system of
rats after treatment with ethylene glycol monomethyl and monobutyl ethers. Toxicol Appl Pharmacol
77:187-200.
Hardin, B. D.; Goad, P. T.; Burg, J. R. (1984) Developmental toxicity of four glycol ethers applied
cutaneously to rats. Environ. Health Perspect. 57: 69-74.
Haufroid, V.; Thirion, F.; Mertens, P.; Buchet, J. P.; Lison, D. (1997) Biological monitoring of
workers exposed to low levels of 2-butoxyethanol. Int. Arch. Occup. Environ. Health 70: 232-236.
Heindel, J. J.; Gulati, D. K.; Russell, V. S.; Reel, J. R.; Lawton, A. D.; Lamb, J. C., IV. (1990)
Assessment of ethylene glycol monobutyl and monophenyl ether reproductive toxicity using a
continuous breeding protocol in Swiss CD-I mice. Fundam. Appl. Toxicol. 15: 683-696.
Hord, J. D.; Lukens, J. N. (1999) Anemias unique to infants and young children. In: Lee, R. G., ed.
Wintrobe's clinical hematology, volume 2. 10th ed. Baltimore, MD: Williams & Wilkins; pp.
1518-1537.
Houglum, K.; Ramm, G. A.; Crawford, D. H.; Witztum, J. L.; Powell, L. W.; Chojkier, M. (1997)
Excess iron induces hepatic oxidative stress and transforming growth factor P1 in genetic
hemochromatosis. Hepatology 26: 605-610.
Hsu LC, Chang WC, Yoshida A. 1994. Cloning of a cDNA encoding human ALDH7, a new member
of the aldehyde dehydrogenase family. Gene 151: 285-289.
Imlay JA. Linn S. (1988). DNA damage and oxygen radical toxicity. Science. 240:1302-1309.
Jiang X. Baldwin CL (1993). Iron augments macrophage-mediated killing of Brucella abortus alone
and in conjunction with interferon-gamma. Cell. Immunol. 148:397-407.
Junge, B.; Carrion, Y.; Bosco, C.; Galleano, M.; Puntarulo, S.; Tapia, G; Videla, L. (2001) Effects of
iron overload and lindane intoxication in relation to oxidative stress, Kupffer cell function, and liver
injury in the rat. Toxicol. Appl. Pharmacol. 170: 23-28.
Kamendulis, L. M.; Park, J. J.; Klaunig, J. E. (1999) Potential mechanisms of rodent liver toxicity by
EGBE: oxidative stress studies. Indianapolis, IN: Indiana University, School of Medicine; final report
to Ethylene Glycol Ethers Panel, Chemical Manufacturers Association; December.
Kitagawa K, Kawamoto T, Kunugita N, Tsukiyama T, Okamoto K, Yoshida A, Nakayama K. 2000.
Aldehyde dehydrogenase (ALDH) 2 associates with oxidation of methoxyacetaldehyde; in vitro
analysis with liver subcellular fraction derived from human and Aldh2 gene targeting mouse. FEES
Lett 476: 3 06-311.
A2-18
-------
Klaunig, J. E.; Xu, Y.; Bachowski, S.; Ketcham, C. A.; Isenberg, J. S.; Kolaja, K. L.; Baker,
T. K.; Walborg, E. F., Jr.; Stevenson, D. E. (1995) Oxidative stress in non-genotoxic
carcinogenesis. Toxicol. Lett. 82-83: 683-691.
Klaunig, J. E.; Xu, Y.; Isenberg, J. S.; Bachowski, S.; Kolaja, K. L.; Jiang, J.; Stevenson, D. E.;
Walborg, E. F., Jr. (1998) The role of oxidative stress in chemical carcinogenesis. Environ. Health
Perspect. 106(suppl. 1): 289-295.
Klaunig, J. E. (2002) Mode of action of butoxyethanol for liver hemangiosarcomas in the male mouse.
Washington, DC: Presented to U.S. Environmental Protection Agency, National Center for
Environmental Assessment; December 3.
Klaunig, JE and Kamendulis, LM (2004). The role of oxidative stress in carcinogenesis.
Ann. Rev. Pharm. Toxicol 44: 239-267.
Klaunig, J.E. and Kamendulis, L.M. (2005) Mode Of Action Of Butoxyethanol Induced Mouse Liver
Hemangiosarcomas And Hepatocellular Carcinomas. Toxicol. Lett. 156(1): 107-15.
Knutson, M.; Wessling-Resnick, M. (2003) Iron metabolism in the reticuloendothelial system. Crit.
Rev. Biochem. Mol. Biol. 38: 61-88.
Krasavage, W. J. (1986) Subchronic oral toxicity of ethylene glycol monobutyl ether in male rats.
Fundam. Appl. Toxicol. 6: 349-355.
Lesgards, J.-F.; Durand, P.; Lassarre, M.; Stocker, P.; Lesgards, G.; Lanteaume, A.; Prost, M.;
Lehucher-Michel, M.-P. (2002) Assessment of lifestyle effects on the overall antioxidant capacity of
health subjects. Environ. Health Perspect. 110: 479-486.
Long P. H, Maronpot, R. R., Ghanavem. B. I.. Roycroft, J H., and Nyska A (2000). Dental pulp
infarction in female rats following inhalation exposure to 2-butoxyethanol. Toxicol. pathol. 28:246-
252.
Mandishona, E.; MacPhail, A. P.; Gordeuk, V. R.; Kedda, M. A.; Paterson, A. C.; Rouault, T. A.;
Kew, M. C. (1998) Dietary iron overload as a risk factor for hepatocellular carcinoma in Black
Africans. Hepatology 27: 1563-1566.
Medinsky, M. A.; Singh, G.; Bechtold, W. E.; Bond J. A.; Sabourin P. J.; Birnbaum L. S.; Henderson
R. F. (1990) Disposition of three glycol ethers administered in drinking water to male F344/N rats.
Toxicol. Appl. Pharmacol. 102: 443-455.
National Toxicology Program (NTP). (1993a) Toxicology and carcinogenesis studies of
pentachloroanisole (CAS no. 1825-21-4) in F344/N rats and B6C3F1 mice (gavage studies).
Washington, DC: U.S. Department of Health and Human Services, National Institutes of Health.
(National Toxicology Program technical report series no. 414).
A2-19
-------
National Toxicology Program (NTP). (1993b) Technical report on toxicity studies of ethylene glycol
ethers 2-methoxyethanol, 2-ethoxyethanol, 2-butoxyethanol administered in drinking water to F344/N
rats and B6C3F1 mice. Washington, DC: U.S. Department of Health and Human Services, National
Institutes of Health. (National Toxicology Program toxicity report no. 26; NIH publication no. 93-
3349). Available from: NTIS, Springfield, VA; PB94-118106.
National Toxicology Program (NTP). (1998) Toxicology and carcinogenesis studies of 2-
butoxyethanol (CAS no. 111-76-2) in F344/N rats and B6C3F1 mice (inhalation studies). Washington,
DC: U.S. Department of Health and Human Services, National Institutes of Health. (National
Toxicology Program technical report series no. 484; NIH draft publication no. 98-3974)
National Toxicology Program (NTP). (2000a) Toxicology and carcinogenesis studies of 2-
butoxyethanol (CAS no. 111-76-2) in F344/N rats and B6C3Fj mice (inhalation studies). Washington,
DC: U.S. Department of Health and Human Services, National Institutes of Health. (National
Toxicology Program technical report series no. 484).
National Toxicology Program (NTP). (2000b). Toxicology data management system historical control
tumor incidence summaries. Available at http://ehpnetl .niehs.nih.gov/docs/ehpsearch.html.
Nelson, B. K.; Setzer, J. V.; Brightwell, W. S.; Mathinos, P. R.; Kuczuk, M. H.; Weaver, T. E.; Goad,
P. T. (1984) Comparative inhalation teratogenicity of four glycol ether solvents and an amino
derivative in rats. Environ. Health Perspect. 57: 261-271.
Nyska, A.; Maronpot, R. R.; Eldridge, S. R.; Haseman, J. K.; Hailey, J. R. (1997) Alteration in cell
kinetics in control B6C3F1 mice infected whhHelicobacter hepaticus. Toxicol. Pathol. 25: 591-596.
Nyska, A., Maronpot, R. R., Long, P. H., Roycroft, J. H., Haliy, J. R., Travlos, G. S., and Ghanayem.
B. I. (1999). Disseminated Thrombosis and Bone Infarction in Female Rats Following Inhalation
Exposure to 2-Butoxyethanol. Toxicol. Pathol. 27:287-294
Nyska, A., Maronpot, R. R., and Ghanayem. B. I. (1999). Ocular Thrombosis and Retinal
Degeneration in Female Rats by 2-Butoxyethanol. Human and Exp. Toxicol. 18:577-582.
Nyska A, Haseman JK, Kohen R, Maronpot RR (2004). Association of Liver Hemangiosarcoma and
secondary iron overload in B6C3F1 Mice - The National Toxicology Program Experience.
Toxicologic Pathology. 32:222-228.
Park, J.; Kamendulis, L. M.; Klaunig, J. E. (2002a) Effects of 2-butoxyethanol on hepatic oxidative
damage. Toxicol. Lett. 126: 19-29.
Park, J.; Kamendulis, L. M.; Klaunig, J. E. (2002b) Mechanisms of 2-butoxyethanol carcinogenicity:
studies on Syrian hamster embryo (SHE) cell transformation. Toxicol. Sci. 68: 43-50.
Reed, JM, Kamendulis, LM, Klaunig, JE (2003). Examination of DNA damage in endothelial cells
following treatment with 2-butoxyethanol using the single cell gel electrophoresis (Comet) assay.
Toxicologist 72s: 206.
A2-20
-------
Rettenmeier, A. W.; Hennigs, R.; Wodarz, R. (1993) Determination of butoxyacetic acid andN-
butoxyacetylglutamine in urine of lacquerers exposed to 2-butoxyethanol. Int. Arch. Occup. Environ.
Health 65(suppl. 1): S151-S153.
Riveros-Rosas, H., Julian-Sanchez, A., and Pina, E. (1997) Enzymology of ethanol and acetaldehyde
metabolism in mammals. Archives of Medical Research, 28 (4): 453-471.
Rose, M. L.; Germolec, D. R.; Schoonhoven, R.; Thurman, R. G. (1997) Kupffer cells are causally
responsible for the mitogenic effect of peroxisome proliferators. Carcinogenesis 18: 1453-1456.
Rose, M. L.; Rivera, C. A.; Bradford, B. U.; Graves, L. M.; Cattley, R. C.; Schoonhoven, R.;
Swenberg, J. A.; Thurman, R. G. (1999) Kupffer cell oxidant production is central to the mechanism
of peroxisome proliferators. Carcinogenesis 20: 27-33.
Rowland M.; Drumm B. (1998) Clinical significance of Helicobacter infection in children. Br. Med.
Bull. 54: 95-103.
Siesky, A. M.; Kamendulis, L. M.; Klaunig, J. E. (2002) Hepatic effects of 2-butoxyethanol in rodents.
Toxicol. Sci. 70: 252-260.
Sleet, R. B.; Price, C. J.; Marr, M. C. (1989) Teratologic evaluation of ethylene glycol monobutyl ether
(Cas no. 111-76-2) administered to Fischer-344 rats on either gestational days 9 through 11 or days 11
through 13. Final report. Research Triangle Park, NC: National Institute of Environmental Health
Sciences, NTP-89-058.
Spolarics, Z. (1999) A carbohydrate-rich diet stimulates glucose-6-phosphate dehydrogenase
expression in rat hepatic sinusoidal endothelial cells. J. Nutr. 129: 105-108.
Stevens, R. G.; Graubard, B. J.; Micozzi, M. S.; Neriishi, K.; Blumberg, B. S. (1994) Moderate
elevation of body iron level and increased risk of cancer occurrence and death. Int. J. Cancer 56: 364-
369.
Swenberg, J. A.; Barrow, C. S.; Boreiko, C. J.; Heck, H.d'A.; Levine, R. J.; Morgan, K. T.; Starr, T. B.
(1983) Non-linear biological responses to formaldehyde and their implications for carcinogenic risk
assessment. Carcinogenesis 4: 945-952.
Tyler, T. R. (1984) Acute and subchronic toxicity of ethylene glycol monobutyl ether. Environ. Health
Perspect. 57: 185-191.
U.S. Environmental Protection Agency. (1999a) Guidelines for carcinogen risk assessment [draft].
Washington, DC: Risk Assessment Forum; report no. NCEA-F-0644. Available:
http://www.epa.gov/ncea/raf/pdfs/cancer_gls.pdf
U.S. Environmental Protection Agency. (1999b) Toxicological Review of ethylene glycol
monobutyl ether (EGBE) (CASRN 111-76-2) in Support of Summary Information on
Integrated Risk Information System (IRIS). National Center for Environmental
Assessment, Washington, DC. Available online from: http://www.epa.gov/iris.
A2-21
-------
U.S. Environmental Protection Agency. (2003) Draft Final Guidelines for Carcinogen Risk
Assessment (External Review Draft, February 2003). Other Ncea-F-0644a. 03 Mar 2003. U.S.
Environmental Protection Agency, Risk Assessment Forum, Washington, DC, 125 pp. Available:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55445.
Udden, M. M. (1994) Hemolysis and decreased deformability of erythrocytes exposed to butoxyacetic
acid, a metabolite of 2-butoxyethanol: II. Resistance in red blood cells from humans with potential
susceptibility. J. Appl. Toxicol. 14: 97-102.
Udden, M. M. (1995) Effects of butoxyacetic acid on human red cells. Occup. Hyg. 2: 283-292.
Udden, M. M. (2002) In vitro sub-hemolytic effects of butoxyethanol acid on human and rat
erythrocytes. Toxicol. Sci. 69: 258-264.
U.S. Environmental Protection Agency (EPA) (2004) External Peer Review - Ethylene Glycol Butyl
Ether (EGBE). Final Report prepared for the EPA National Center for Environmental Assessment,
Office of Research and Development, Integrated Risk Information System under interagency
agreement No. DW-8993 9822-01-0 between EPA and the U.S. Department of Energy, Oak Ridge
Institute for Scientific Education (ORISE).
Wang, Y. J.; Ho, Y. S.; Lo, M. J.; Lin, J. K. (1995). Oxidative modification of DNA bases in rat liver
and lung during chemical carcinogenesis and ageing. Chem.-Biol. Interact. 94: 135-145.
Wier, P. J.; Lewis, S. C.; Traul, K. A. (1987) A comparison of developmental toxicity evident at term
to postnatal growth and survival using ethylene glycol monoethyl ether, ethylene glycol monobutyl
ether, and ethanol. Teratog. Carcinog. Mutagen 7: 55-64.
Xue, H.; Kamendulis, L. M.; Klaunig, J. E. (1999) A potential mechanism for 2-butoxyethanol (2-BE)
induced mouse liver neoplasia. lexicologist 48(suppl. 1): 231.
Yamaguchi, R.; Hirano, T.; Asami, S.; Chung, M. H.; Sugita, A.; Kasai, H. (1996) Increased 8-
hydroxyguanine levels in DNA and its repair activity in rat kidney after administration of a renal
carcinogen, ferric nitrilotriacetate. Carcinogenesis 17: 2419-2422.
Ziouzenkova, O., Asatryan, L. and Sevanian, A. (1999) Oxidative stress resulting from
hemolysis and formation of catalytically active hemoglobin: protective strategies. Int. J. Clin.
Pharmacol. Ther. 37: 125-132.
A2-22
-------
ATTACHMENT 3
Benchmark Dose Assessment of Forestomach Lesions in Female Mice
Using PBPK Models to Estimate Human Equivalent Exposures
Several PBPK models have been developed for EGBE, all of which are capable of estimating
internal doses. These models are summarized briefly in the EGBE IRIS file (EPA, 1999). Consistent
with the EPA (1999) IRIS assessment, the Lee et al. (1998) was used to estimate internal dose levels
from inhalation exposures to the experimental animal, in this case the female mouse.
Cmax (peak concentration in the blood during the exposure period) of BAA was used in the
existing IRIS file (EPA, 1999) for the derivation of the RfD and RfC from hematological endpoints
because of (a) convincing evidence that BAA is the causative agent for EGBE-induced hemolysis and
(b) EGBE-induced hemolysis appears to be highly dependent upon the BAA concentration attained.
Similarly, BAA is believed to be the toxic moiety responsible for the forestomach effects observed
following EGBE exposure (Attachment 1) and concentration appears to critical in the development of
these effects as well. No signs of forestomach irritation were observed in mice at very high dose
levels of 1400 mg/kg/day in 2-week and 13-week drinking water studies conducted by NTP (NTP,
1993b). It has been suggested that such oral non-bolus dosing of EGBE does not result in high enough
local concentrations of EGBE and BAA (Poet et al., 2003). Previous studies with other nongenotoxic
forestomach carcinogens demonstrated that forestomach effects are dependent not only on the dose but
also on the chemical concentration in the dosing solution (Ghanayem et al., 1985) and other effects of
EGBE appear to be highly dependent on the concentration attained (Nyska et al., 1999; Long et al.,
2000; Ghanayem et al., 2000; 2001). Use of blood concentrations as a common dose metric is also
justifiable because, as is discussed in Attachment 1, systemic distribution of inhaled EGBE via the
blood to the salivary glands makes an important contribution to target organ dose in mice (via the
swallowing of saliva) and because humans would not be expected to receive exposure via the other
major distribution route in the mouse, oral ingestion through the grooming of fur.
The endpoint used in this analysis was epithelial hyperplasia of the female mouse forestomach
as it was the most sensitive forestomach effect observed in the NTP (2000) study. Consistent with the
1999 IRIS assessment, four steps were employed to estimate human equivalent oral and inhalation
benchmark exposures from this endpoint: (1) a BMDL10 value was estimated using modeled "end of
the week" internal dose (Cmax BAA in blood) levels; (2) verify that steady state was achieved (e.g., no
change in BAA Cmax as a result of prolonging the exposure regimen); (3) simulate the internal dose
surrogate (Cmax BAA in blood) for humans (continuous air exposure; drinking water assumption was
that a 70-kg human consumes an average of 2 liters of water during a 12-hour awake cycle); and (4)
calculate the human equivalent dose/concentration that resulted in the same internal dose (Cmax BAA)
simulated for the animal in Step 1.
STEP 1: ESTIMATION OF BMDL10 (Cmax) DOSE
Cmax for BAA in arterial blood was determined using the PBPK model of Lee et al. (1998). The results
of this model and incidence data for the endpoint of concern are summarized in Table A3-1.
A3-1
-------
Table A3-1. PBPK Model estimates of BAA Cmax blood levels and incidence of forestomach
epithelial hyperplasia in female mice.
Water Cone.
(ppm)
0
62.5
125
250
CmaxBAA
(uM/L)
0
529
1200
2620
Incidence of
Forestomach
Hyperplasia
0/50
6/50
27/50
42/49
BMD and BMDL10 estimates were derived using the available models in version 1.3.2 of the
EPA benchmark dose software.30 The estimates for each model, along with statistical goodness-of-fit
information are provided in Table A3-2.
Table A3-2. BMDS model estimates of C
epithelial hyperplasia in female mice.
BMD10 and BMDL10 values for forestomach
BMDS Model
Gamma
Logistic
Log-Logistic
Multistage (1st degree)
Multistage (2nd degree)
Multistage (3rd degree)
Probit
Log-Probit
Weibull
BMD
(uM/L)
420.56
544.757
462.513
177.442
338.483
338.485
525.521
470.876
376.085
BMDL
(uM/L)
266.87
444.896
329.04
145.713
202.437
197.436
430.612
344.412
238.952
AIC
(Lowest »best fit)
151.16
162.191
150.153
156.244
152.681
152.681
161.304
150.163
151.855
P-value
(>0.1 = adequate
fit)
0.5287
0.0067
0.8717
0.0648
0.0976
0.2535
0.0086
0.8673
0.3807
Considering all goodness-of-fit parameters, including chi-square residuals at low doses and
visual inspection of plots, the Log-Logistic model was chosen as the model that best describes the
dose-response for this endpoint. Graphical results of this model are provided in Figure A3-1. A
textual description (model output) of these results is provided in Appendix A. The BMDL10 was
determined to be 329 |j,M/L, using the 95% lower confidence limit of the dose-response curve
expressed in terms of the Cmax for BAA in blood.
30
A copy of the BMDS can be obtained from the Internet at www.epa.gov/ncea/bmds.htm.
A3-2
-------
Log-Logistic Model with 0.95 Confidence Level
0.8
1 0.6
I
I °'4
03
0.2
Log-Logistic
500
1000 1500 2000 2500
dose
20:0607/11 2003
Figure A3-1. BMD plot of fraction of female mice with forestomach epithelial hyperplasia
following inhalation exposure(NTP, 2000) vs. internal dosimetric (BAA Cmax, |J,M/L).
Step 2: Verification of Steady State
As can be seen from Table A3-3, Cmax levels are relatively constant through 6 months, then
increase at and beyond 12 months, presumably due to clearance problems in aging animals.
However, the earlier steady state levels are appropriate for use in this assessment because that is
the more conservative approach and because similar effects were observed during the subchronic
portion of the NTP (2000) study at the same dose levels, indicating that the higher internal doses at
and beyond 12 months were not required for the effects to appear.
Table A3-3: Female Mouse Cmax Levels for Various Time Points of NTP (2000) Study
Estimated by the Lee et al. (1998) Model
Months
on Studv
1
O
6
12
16
18
62.5 ppm
Male
403
402
399
484
643
756
Female
529
527
523
639
849
995
125
Male
921
925
914
1079
1443
1625
ppm
Female
1200
1202
1184
1414
1839
2102
250
Male
2080
2120
2071
2349
2798
3067
ppm
Female
2620
2652
2582
2951
3501
3803
A3-3
-------
Step 3: Simulation of Internal Human Doses
The tables below summarize the results of model simulations of the internal dose surrogate (Cmax
BAA in blood) for a 70-kg human who consumes an average of 2 liters of drinking water during a
12-hour awake cycle (Table A3-4) or is continuously exposed to air concentrations (Table 5) of
EGBE.
Table A3-4: Estimated Cmax for BAA in blood for humans continuously exposed to varying
drinking water concentrations of EGBE (Corley et al, 1994; 1997).
EGBE concentration in
water (ppm)
24
48
94
188
375
750
Calculated dose of EGBE
from drinking water (mg/kg/d)
0.7
1.4
2.7
5.4
10.7
21.4
Cmax BAA in blood
(HM/L)
9
18
36
73
147
299
Table A3-5: Estimated Cmax for BAA in blood for humans continuously exposed to varying
concentrations of EGBE (Corley et al, 1994; 1997).
Concentration of EGBE
in air (ppm)
1
5
10
20
50
100
200
Cmax BAA in blood
(jiM/L)
2.6
13.0
26.1
52.9
137.1
295.0
733.7
Step 4: Calculate the Human Equivalent Dose/concentration
The Corley et al. (1994; 1997) PBPK model was used to "back-calculate" a human equivalent oral
dose of 23.6 mg/kg-day from the Cmax BMDL10 of 320 \iM/L estimated in Step 1, assuming that rats
and humans receive their entire dose of EGBE from drinking water over a 12-hr period each day.
The Corley et al. (1997) PBPK model was used to "back-calculate" human equivalent air
concentration of 113 ppm (551 mg/m3) from the Cmax BMDL10 of 320 \iM/L estimated in Step 1,
assuming continuous exposure (24 hr/day).
These results indicate that the RfD and RfC values for EGBE, which were based on hemolytic effects
in female rats and lower human equivalent BMDL10 estimates for both water (5.1 mg/kg/day) and air
(380 mg/m3), should be adequate for the prevention of gastrointestinal hyperplastic effects.
A3-4
-------
REFERENCES
Corley, R. A.; Bormett, G. A.; Ghanayem, B. I. (1994) Physiologically-based pharmacokinetics of 2-
butoxyethanol and its major metabolite, 2-butoxyacetic acid, in rats and humans. Toxicol. Appl.
Pharmacol. 129: 61-79.
Corley, R. A.; Markham, D. A.; Banks, C.; Delorme,P.; Masterman, A.; Houle, J. M. (1997)
Physiologically-based pharmacokinetics and the dermal absorption of 2-butoxyethanol vapors by
humans. Fundam. Appl. Toxicol. 39: 120-130.
Ghanayem, B.I, Maronpot, R.R. and Matthews, H.B. (1985). Ethyl Acrylate-Induced Gastric
Toxicity: II. Structure-Toxicity Relationships and Mechanism. Toxicol. Appl. Pharmacol. 80:336-
344.
Nyska, A., Maronpot, R. R., and Ghanayem, B. I. (1999). Ocular Thrombosis and Retinal
Degeneration in Female Rats by 2-Butoxyethanol. Human and Exp. Toxicol. 18:577-582.
Ghanayem, B. I, Ward, S M, Chanas, B, and Nyska, A. (2000). Comparison of the acute
hematotoxicity of 2-butoxyethanol in male and female F344 rats. Human Exp. Toxicol. 19:185-192.
Long P. H, Maronpot, R. R., Ghanayem, B. I, Roycroft, J H., and Nyska A (2000). Dental pulp
infarction in female rats following inhalation exposure to 2-butoxyethanol. Toxicol. pathol. 28:246-
252.
Ghanayem, B. I, Long, P., Ward, S M, Chanas, B, and Nyska, A. (2001). Hemolytic Anemia,
Thrombosis, and Infarction in Male and Female F344 Rats Following Gavage Exposure to 2-
Butoxyethanol. Experimental and Toxicologic Pathology 53:97-1
Lee, K. M.; Dill, J. A.; Chou, B. J.; Roycroft, J. H. (1998) Physiologically based pharmacokinetic
model for chronic inhalation of 2-butoxyethanol. Toxicol. Appl. Pharmacol. 153: 211-226.
National Toxicology Program (NTP). (1993) Technical report on toxicity studies of ethylene glycol
ethers 2-methoxyethanol, 2-ethoxyethanol, 2-butoxyethanol administered in drinking water to F344/N
rats and B6C3F1 mice. Washington, DC: U.S. Department of Health and Human Services, National
Institutes of Health. (National Toxicology Program toxicity report no. 26; NIH publication no. 93-
3349). Available from: NTIS, Springfield, VA; PB94-118106.
National Toxicology Program (NTP). (2000a) Toxicology and carcinogenesis studies of 2-
butoxyethanol (CAS no. 111-76-2) in F344/N rats and B6C3Fj mice (inhalation studies). Washington,
DC: U.S. Department of Health and Human Services, National Institutes of Health. (National
Toxicology Program technical report series no. 484).
Poet, T. S.; Soelberg, J. J.; Weitz, K. K.; Mast, T. J.; Miller, R. A.; Thrall, B. D.; Corley, R. A. (2003)
Mode of action and pharmacokinetic studies of 2-butoxyethanol in the mouse with an emphasis on
forestomach dosimetry. Toxicol. Sci. 71: 176-189.
A3-5
-------
APPENDIX A
Logistic Model SRevision: 2.1 $ $Date: 2000/02/26 03:38:20 $
Input Data File: F:\BMDS\DATA\EGBE\F_MOUSE_HYP_LOG-LOGIST.(d)
Gnuplot Plotting File: F:\BMDS\DATA\EGBE\F_MOUSE_HYP_LOG-LOGIST.plt
Fri Jul 11 19:53:31 2003
BMDS MODEL RUN
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-slope*Log(dose))]
Dependent variable = F_Hyperplasia
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0
intercept = -16.768
slope = 2.36735
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -1
slope -1 1
A3-6
-------
Parameter Estimates
Variable Estimate Std. Err.
background 0 NA
intercept -16.7132 2.64108
slope 2.36545 0.372243
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model Log(likelihood) Deviance Test DF P-value
Full model -72.9391
Fitted model -73.0765 0.274637 2 0.8717
Reduced model -131.841 117.804 3 <.0001
AIC: 150.153
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0000 0.000 0 50 0
529.0000 0.1324 6.622 6 50 -0.2596
1200.0000 0.5145 25.725 27 50 0.3608
2620.0000 0.8705 42.653 42 49 -0.2777
Chi-square= 0.27 DF = 2 P-value = 0.8717
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD= 462.513
BMDL = 329.04
A3-7
-------
ATTACHMENT 4
External Peer ReviewSummary of Comments and Disposition
The following are summaries of the recent external reviews that this position paper has
undergone, a 2003 Letter Review of the draft position paper and a 2004 panel review of the interim
final position paper and technical reports submitted in response to the Agency's November, 2003
proposal to delist EGBE from the list of CAA hazardous air pollutants. The focus of the summaries
and EPA responses is on the comments which stated an opinion contrary to the draft or interim final
position paper.
2003 Letter Review Comments and Responses (Blue)
Genotoxicity of EGBE or its Metabolites
1. Has the position paper drawn appropriate conclusions from the available literature on the
genotoxicity of EGBE and its metabolites?
Three reviewers agreed that Butoxyacetaldehyde (BAL) is the most likely of the metabolites to pose
genotoxic risks. However, they did not consider this risk to be significant for the following stated
reasons:
1. Rapid conversion of BAL to BAA in vivo
2. Lack of evidence for BAL in rodent tissues
3. Studies indicating that aldehydes are rapidly cleared by metabolism
4. Negligible macromolecular adducts formed in vivo following EGBE exposures
5. Negative results for EGBE in in vitro w/ metabolic activation & in vivo genetox assays
6. General lack of genetox activity by other members of the glycol ether class
7. Largely negative results of in vitro genetox assays of BAL
8. Studies suggesting that DNA adducts formed by aldehydes are relatively easily repaired
9. Historical evidence of forestomach tumor formation by other non-genotoxic agents
10. Historical evidence that aldehydes carcinogenic by direct application to sensitive tissues
(e.g. in gas phase, respiratory system) are not carcinogenic when metabolically generated
One reviewer suggested expanding the statement on page Al-13, line 7 to say "... genotoxicity assays
are in agreement that neither EGBE nor its metabolites are likely to be genotoxic in vivo."
One reviewer agreed with conclusions regarding EGBE and BAA, but was not certain about BAL.
One reviewer felt that the likely modest genotoxicity of BAL was not fairly and adequately evaluated.
This reviewer felt that EPA cannot exclude that some portion of the observed carcinogenic responses
in the forestomach and liver is attributable to a genetic mechanism and recommended a combined
"linear/nonlinear" approach.
RESPONSE: All reviewers agreed that the evidence supports the contention that EGBE and
BAA are not genotoxic. Reviewers disagreed with respect to the primary genotoxicity issue
identified, the potential for the EGBE metabolite BAL to directly interact with DNA following
in vivo EGBE exposure. The position paper has been enhanced to address the potential role of
A4-1
-------
BAL, and an effort to model the concentrations of EGBE, BAL and BAA in the stomach and
liver of mice is included. Corley (2003) estimated that 250 ppm EGBE would result in peak
Cmax concentrations of 48 |J,M EGBE, 1.1 |iM BAL and 3,200 BAA |J,M in GI tissue and 7 |J,M
EGBE, 0.5 p,M BAL and 3,250 BAA p,M in liver tissue of mice at the end of a 6 hour exposure
period (see Figures Al-1 and A2-1). This modeling effort is limited by the minimal metabolic
rate information available for the formation and distribution of BAL, but does indicate that
expected in vivo BAL concentrations in the subject target organs would be significantly lower
than BAL concentrations found to be clastogenic in vitro assays (without metabolic activation).
Nevertheless, the Agency agrees with two of the five reviewers that information is currently
inadequate to completely dismiss the potential contribution of a linear mechanism associated
with the possible mutagenic metabolite BAL. A definitive determination regarding the
appropriateness of a nonlinear approach can not be made until questions regarding the role of
BAL are resolved. Additional research (e.g., verification of existing PBPK modeling results and
improved genotoxicity assays) would assist the Agency in making a more informed decision
concerning the potential for BAL to contribute to the adverse effects seen in animals following
EGBE exposure and use of the proposed nonlinear assessment approach. The role of BAL
towards the effects observed in mice following EGBE exposure is addressed further below.
2. Is there additional information (e.g., from other studies or studies of related compounds) that
would suggest an alternative conclusion?
Three reviewers said no. One of these reviewers stated that the only direct evidence to the contrary is
the small number of equivocally positive results in clastogenicity experiments [with EGBE in vitro],
but that this type of experiment is notoriously susceptible to direct toxic effects at the cellular level,
which are not reflective of the situation in vivo.
One reviewer suggested that the mutagenic/clastogenic potency of butoxyacetaldehyde needs to be
studied in relation to analogous activities by other aldehydese.g. ethoxyacetaldehyde,
methoxyacetaldehyde, formaldehyde, acetaldehyde, propionaldehyde, and possibly butyraldehyde
One reviewer offered that BAL causes cytotoxicity (30% increase) and SCE (2-fold increase) in vitro
in human lymphocyte cells at 0.5 mM (Ghanayem and Thompson, unpublished data).
RESPONSE: The new data concerning the cytotoxicity and clastogenicity of BAL to human
lymphocyte cells in vitro (obtained from Ghanayem, 2003) and its relevance to the effects
observed following in vivo exposure to EGBE are discussed in the revised position paper. Some
additional discussion of the expected toxicity of BAL relative to other aldehydes has been added
as well as it is recognized that similar issues have been raised regarding the extent to which
other aldehydes such as the acetaldehyde metabolite of ethanol contribute to genotoxicity
(Lipscomb, 2003; Dewoskin, 2003). BAL has been shown to be cytotoxic and clastogenic to
various cells in vitro and it is possible that the weak positive results seen in some of the in vitro
assays of EGBE (Elias et al., 1996, Table 1) are due to a low rate of metabolism of the alcohol to
BAL. However, cytotoxicity itself is a recognized contributor (promoter) to carcinogenesis, and
can effect natural events in the cell cycle and cause a reduction in the repair of SCE, increases
in SCE. Given the low concentrations of BAL that are expected in vivo following EGBE
exposure (see discussions in revised position paper and below) and the fact that increases in
SCE were only observed following cytotoxic, in vitro exposures to BAL, it is unlikely that the
above hypothesis regarding the role of BAL in vitro would hold true for in vivo exposures to
EGBE. However, the available data do not allow for a definitive statement in this regard. An
A4-2
-------
understanding of the potential role of BAL in the formation of the noncancer and cancer lesions
observed in mice (NTP, 2000) is key to this determination and is discussed further below and in
the revised position paper.
Role of Butoxyacetaldehyde (BAL)
Elliot andAshby (1997) concluded that EGBE and butoxyacetic acid (BAA) are not genotoxic, but
that BAL, an intermediate oxidation product, shows some evidence of clastogenicity in vitro. Dartsch
et al. (1999) attributed the cytotoxic effects of EGBE observed in vitro to BAL. However, BAL has
not been identified in vivo, and studies of EGBE in forestomach and glandular stomach tissue
supernatant (Green et al., 2002) and liver (Medinsky et al., 1990; Corley et al, 1997; Ghanayem et
al., 1987) shows that it is rapidly metabolized to BAA by aldehyde dehydrogenase. In addition,
despite the fact that metabolism of EGBE to BAL is reversible and metabolism of BAL to BAA is
irreversible, BAA is significantly more toxic to the blood and fore stomach of rodents than EGBE
(U.S. EPA, 1999; Green et al, 2002).
3. Is the position paper justified in determining that BAL did not significantly contribute to the
EGBE induced noncancer and cancer effects observed in rats and mice (NTP, 2000)?
Three reviewers responded yes. One reviewer stated while this conclusion is not strictly supported by
the available data, the weight of evidence clearly supports this conclusion.
Two reviewers responded no. These reviewers offered the following facts as support of the potential
contribution of BAL to the noncancer and cancer effects associated with EGBE exposure.
1. Pretreatment of rats with cyanamide, an aldehyde dehydrogenase inhibitor, reduced
EGBE-induced hemolytic responses, but increased RBC swelling, increased the mortality
caused by EGBE exposure, decreased BAA formation and excretion in the urine, and
increased the urinary excretion of EGBE conjugates with glucuronide and sulfate.
2. Pretreatment of rats with pyrazole, an alcohol dehydrogenase inhibitor, protected against
EGBE-induced hemolysis
3. In vitro studies demonstrated that:
a. While BAA is more potent, the effects of BAA and BAL on RBCs are qualitatively
similar.
b. At 0.25-4.0 mM, BAL decreased viability and survival of peripheral human
lymphocytes in a dose-dependent manner.
c. At 0.5 mM, BAL doubled the number of SCE/cell vs. the vehicle (DMSO)
d. No significant effects were observed using BAA at equimolar concentrations.
4. EGBE was more cytotoxic than BAA (as measured by LDH release) at comparable
concentrations (25 and 50 |iM) in cell culture (Park et al., 2002).
They also offered the following observations/suggestions:
1. Individuals with genetic polymorphisms in EGBE metabolizing enzymes may
accumulate BAL and represent a sensitive subpopulations to EGBE exposure.
2. Extend existing PBPK models for EGBE to at least address the quantitative issue of how
much concentration X time of the two metabolites can be expected to have been present
under the conditions of dosage of the bioassay experiments.
A4-3
-------
RESPONSE: The limitations and value of the in vitro and inhibition studies are addressed
extensively in the revised position paper. The Elliot and Ashby (1996) review of the genetic
toxicity studies relevant to EGBE published through 1996 concluded that EGBE is
nongenotoxic, but that "the corresponding aldehyde [BAL] presents some evidence for
clastogenicity in vivo, and that the derived acid [BAA] is non-genotoxic." This conclusion is
supported by what is know about chemicals of similar structure (e.g. ethanol, acetaldehyde and
acetic acid) and unpublished data showing that BAL causes cytotoxicity (30% increase) and
sister-chromatid exchanges (2-fold increase) in vitro in human lymphocyte cells at 0.5 mM
(Ghanayem and Thompson, unpublished data).
The primary issue of greatest relevance to this position paper is BAL's contribution to the
carcinogenic effects of EGBE. However, the concern expressed by two reviewers regarding the
contribution of BAL to noncancer effects (hemolytic and forestomach irritation) will be
addressed first. The paper by Ghanayem et al. (1987) does show that some significant
hemolytic activity remains when animals exposed to EGBE are pretreated with an aldehyde
dehydrogenase inhibitor, cyanamide. However, several factors suggest that this activity is more
likely due to residual BAA rather than BAL. Inhibitors such as cyanamide and pyrazole are
not very specific. Thus, some BAA will be formed and, in fact, Ghanayem et al., (1990) found
that while EGBE + cyanamide decreased BAA concentrations in rats, it also increased the half-
life of BAA. In addition, when Ghanayem et al. (1987) administered a gavage dose of 125 mg
BAL/kg + cyanamide to rats they observed almost no hemolytic activity (Ghanayem et al.,
1987). Also, gavage administration to rats of 125 mg EGBE/kg and the molar equivalent of
BAL and BAA resulted in no significant difference between the hemolytic effects of the three
chemicals between 2 and 24 hours after exposure (Ghanayem et al., 1987). These facts suggest
that EGBE's hemolytic activity is due almost entirely to BAA, and that the metabolism of
EGBE and BAL to BAA takes place rapidly and completely. The ATSDR (1998) profile for
EGBE states that:
Incubation of rat erythrocytes with butoxyacetaldehyde or 2-butoxyacetic acid caused
time- and concentration-dependent swelling of red blood cells followed by hemolysis,
which has also been observed in vivo (Ghanayem et al. 1990b). Butoxyacetaldehyde was
less effective in causing hemolysis than was 2-butoxyacetic acid, suggesting that, although
whole rat blood may contain enough aldehyde dehydrogenase to cause some conversion of
the aldehyde to 2-butoxyacetic acid, the 2-butoxyacetic acid is the hemolytic agent.
Addition of aldehyde dehydrogenase and its cofactors, followed by butoxyacetaldehyde,
resulted in a significant increase in hemolysis, which could be decreased with the addition
of cyanamide, an aldehyde dehydrogenase inhibitor, thus supporting the evidence that
production of 2-butoxyacetic acid is the important step in 2-butoxyethanol-specific
hemolysis (Ghanayem et al. 1989).
A recent attempt has been made to quantify the amount of EGBE/BAA/BAL that would be
present in the liver and GI tissues of female mice following a 250 ppm inhalation exposure to
EGBE (Corley, 2003) using the forestomach rate constants provided by Green et al. (2002).
This work extends an earlier model developed by Dr. Corley (Corley et al., 2003) to include the
intermediate formation of BAL in target tissues (liver and GI tract). Given the limitations of
the available data (outlined in Attachment 1), Corley (2003) estimated that 250 ppm EGBE
would result in peak Cmax concentrations of 48 |J,M EGBE, 1.1 |J,M BAL and 3,200 BAA |J,M in
GI tissue and 7 p,M EGBE, 0.5 \iM BAL and 3,250 BAA |j.M in liver tissue of mice at the end of
a 6 hour exposure period (see Figures Al-1 and A2-1). This limited analysis does not prove
A4-4
-------
that BAA is the responsible for the observed forestomach effects, but it strengthens the
hypothesis considerably.
The data available for the determination of BAL's impact on the carcinogenicity of EGBE
are limited. BAL does not appear to cause gene mutations at in vitro concentrations below 2
mM (Elliot and Ashby, 1997). However, Elias et al. (1986) and (Ghanayem and Thompson,
unpublished data) determined that BAL is more cytotoxic and clastogenic than BAA, and can
cause SCE in Chinese hamster lung (V79) and human lymphocyte cells at lower (.1 to 1 mM) in
vitro concentrations. Clastogenic effects were not observed at concentrations below 0.1 mM for
both acetaldehyde and BAL in V79 cells (Elliott and Ashby, 1997; Elias et al., 1986).
The relevance of in vitro data to in vivo exposure to EGBE is questionable. None of the in
vitro studies of BAL have involved application of activating enzymes encountered in vivo and,
as is discussed above, data from Green et al. (2002) and Corley (2003) suggest that BAL is
rapidly metabolized to BAA in vivo due to a high aldehyde dehydrogenase activity in the mouse
forestomach. Further, the BAA that is produced would tend to build up in vitro without the
detoxification/elimination mechanisms present in vivo, resulting in a decreased pH level in the
cell culture. Low pH is known to reduce the DNA repair capacity of cells (Xiao et al., 2003;
Elliot and Ashby, 1997), and this could enhance the impact of a clastogen or a weak mutagen
such as BAL. This is a mechanism that has been proposed for the SCE caused by structurally
related acetaldehyde, which has also been determined to be weakly mutagenic (Dellarco, 1988;
He and Lambert, 1995; Grafstrom et al., 1994), and produces chromosome damage in test
systems using cells from humans and research animals.
Other lines of evidence that suggest that direct interaction of BAL with the DNA molecules
would not play a significant role in the carcinogenic activity of EGBE are discussed further in
Attachment 1 of the position paper. They include the fact that BAL causes cytotoxicity at levels
associated with chromosome effects and cytotoxicity itself can have effects which result in
chromosome damage, acetaldehyde is recognized as "weakly mutagenic" and structural
comparisons of acetaldehydes demonstrate that a longer-chain aldehyde such as BAL would be
less likely to interact with DNA than a shorter chain aldehyde such as acetaldehyde, the fact
that in vitro and in vivo assays of EGBE were generally negative and that chemicals for which
mutagenesis/genotoxic effects play a significant role generally induce more tumors at earlier
time points, rather than near the end of the conducted bioassays, due to their ability to both
initiate and promote tumor pathogenesis. While the existing evidence is certainly suggestive, it
contains important data gaps. Additional research (e.g., verification of existing PBPK modeling
results and improved genotoxicity assays) would assist the Agency in making a more informed
decision concerning the potential for BAL to contribute to the adverse effects seen in animals
following EGBE exposure and use of the proposed nonlinear assessment approach.
One reviewer suggests the need for quantitative PBPK modeling combined with data on
the relative potency for the toxic effects observed. A limited analysis has been done and is
discussed in Attachment 1 (Corley, 2003), but the appropriateness of the use of this information
for predicting the corresponding liver and systemic activity of this enzyme is questionable.
Kinetic information for aldehyde dehydrogenase activity on BAL exists only in the forestomach
(Green et al., 2002). Research to determine the aldehyde activity in organs other than the
stomach and a paper on the relative potencies of aldehydes would be useful and informative.
A4-5
-------
Another reviewer mentioned the importance of considering the impact of genetic
polymorphism towards variance in the human response to EGBE exposure. The issue and
impact of polymorphism is recognized as being important to the consideration of potentially
sensitive human subpopulations and is discussed in the revised position paper in this context.
Potential for Hemolysis in Humans
4. Has the position paper drawn appropriate conclusions from the available literature on the
potential for EGBE induced hemolysis in humans?
All five reviewers said yes. However,
1. A role of BAL in the toxicity of EGBE in vivo must be considered.
2. Human cases of hemolysis should be assessed via PBPK modeling, and reported human
interindividual variability (e.g., Ghanayem and Sullivan, 1993; Ghanayem, 1989) should
be considered.
3. A broad study of the quantitative variations in sensitivity to BAA among a substantial
number of representative humans (in the hundreds, if that could be arranged) would help
prevent EPA from making its choices in ignorance of the extent of additional sensitivity
that might be seen in a minority of people.
4. Definitive data supporting such a conclusion for chronic exposure are not available.
However, the data available on both inter-species differences in pharmacokinetics and
sensitivities to hemolysis indicate that humans are more resistant than rodents to
hemolysis, iron buildup, oxidative stress, and cell proliferation as a result of EGBE
exposure.
RESPONSE: Regarding the first suggestion, the potential role of BAL in the toxicity of EGBE is
now extensively discussed in the position paper. No cases of human hemolysis have ever been
observed, so the suggestion in #2 above would not be possible to accomplish. Suggestion #3 is
not likely to make a significant impact on the Agency's determination with respect to EGBE's
ability to cause hemolysis in humans and a broad (in vitro or in vivo) study in hundreds of
humans would be costly and time consuming. While polymorphisms have been shown to alter
rates of metabolism and elimination of alcohols (Agarwal and Goedde, 1992) there is no EGBE-
specific information available, and it is not clear how such polymorphisms would influence
human susceptibility to EGBE. As described in this paper and the 1999 EGBE IRIS file, case
studies and early controlled exposure studies (Carpenter et al., 1956) and in vitro studies
involving aged, sickle-cell anemia, hereditary spherocytosis patients (Udden, 1994) and children
(Udden, 2002) strongly suggest that human RBCs are considerably less sensitive than rodent
RBCs to the hemolytic effects of EGBE/BAA. In addition, if the reviewer is suggesting
controlled human exposure experiments, considerable legal problems could forestall this type of
experimentation/approach. For the purposes of this assessment paper, the existing studies
provide sufficient information to determine the human cancer risk from a mode of action
involving EGBE induced hemolysis.
A4-6
-------
5. Is there additional information that would suggest an alternative conclusion?
Two reviewers said yes. See the responses to question 4 above.
Three reviewers said no.
RESPONSE: See above.
Relevance of in vitro studies
In vitro studies referred to in the position paper as indicating human insensitivity to hemolysis were
of short (generally 4 hour) duration, and the slight effects on RBC deformability were still increasing
at the end of the assay (Udden, 1994; 2000; Ghanayem et al, 1989). While hemolytic effects have not
been observed in humans at high acute doses, the human effects of long term exposure to EGBE are
not well studied (U.S. EPA, 1999).
6. Is the position paper justified in concluding that humans would not be expected to experience
hemolysis leading to oxidative stress and cell proliferation, key events in the proposed mode
of action for EGBE's role in the formation of the observed mouse liver tumors (NTP, 2000)?
Two reviewers said maybe, and indicated that:
1. Humans should be referred to as"comparatively less sensitive" to hemolysis.
2. Unless the molecular mechanisms responsible for the hemolytic effects of EGBE are fully
characterized and the molecular basis of the lower sensitivity of human erythrocytes are
characterized, it is not possible to conclude that humans are not sensitive.
3. Again, these issues need to be quantitatively analyzed with the aid of both PBPK modeling
and more extensive observations of the human interindividual variability.
Three reviewers responded yes, that the studies in vitro are in concordant with the lack of observed
hemolytic toxicity in humans following acute exposures. Longer-term exposures at lower levels of
EGBE are ubiquitous in industry (and even in the general population) and the effects of hemolytic
toxicity are fairly distinctive and well-known to occupational physicians. Therefore, it is not
unreasonable to expect that they would have been noted if present. IF hemolysis leading to oxidative
stress is the mechanism for the formation of liver hemangiosarcomas and other tumors, these would
not be expected to occur in humans.
RESPONSE: With respect to the first concern, humans are identified as comparatively less
sensitive in the revised position paper (as much as 150-fold according to Udden 1994; 1995 in
vitro data). The primary question is whether the level of hemolysis that humans can reasonably
be anticipated to experience could trigger a carcinogenic effect. Studies have already been
performed to examined the interindividual variability between humans with blood diseases,
elderly and children (Udden, 1994; Udden et al., 2002). No susceptible subpopulation has been
identified and, as AS points out above, it seems unlikely that such a subpopulation would not
have been identified in the literature by now. While a better idea of the molecular mechanism
may help to identify or exclude sensitive human subpopulations, complete knowledge of the
mechanisms involved is not necessary to make a reasoned judgment that humans exposed to
EGBE would not be expected to experience the level of hemolysis necessary to trigger a
secondary carcinogenic effect from EGBE induced hemolysis. It is not clear why a full PBPK
A4-7
-------
model is necessary to compare sensitivities. A simple one-compartment model might be
adequate since the effect of interest is in the blood compartment, not a peripheral (solid) tissue.
In any case, existing literature on the metabolism of EGBE in humans versus rodents suggests
that humans have pharmacokinetic mechanisms to help remove BAA that are not present in
rodents (U.S. EPA, 1999).
Mode of Action for Formation of Forestomach Tumors in Female Mice
7. Has the position paper drawn appropriate conclusions from the available literature on the
potential for EGBE to contribute to the formation offorestomach tumors reported in female
mice?
One reviewer said no, and felt that the document has not done a full quantitative
pharmacokinetic/pharmacodynamic analysis of the possible contributions of genotoxic processes via
the butoxyacetaldehyde (BAL) metabolite. Surely, as with formaldehyde, some local toxic/irritant
cell proliferation response is also involved. However, the likely genetic action via aldehyde reactions
with DNA, combined with some ongoing rate of background cell replication, should allow some rate
of fixation of DNA lesions into permanent chromosomal changes. The contributions from this
process would be expected to be linear at low doses, requiring risk assessment treatment via a
linear/nonlinear paradigm. (Simple mathematics yields the result that when a highly nonlinear
process with an upward turning dose response relationship is combined with a fundamentally linear
process, the linear process will dominate the shape of the dose response relationship at low doses.
The dose at which the linear mode of action takes over depends on the contribution of the linear
process at high doses where the observations are made, and the rapidity with which the "nonlinear"
process declines with dose relative to the linear process.)
Four reviewers said yes, but
1. The 7-step summary of the proposed mode of action must include BAL in steps 1, 2 and 3.
2. Lack offorestomach tumors in male mice are not fully explainable at this time.
3. Extrapolation from measured impacts on the cell membranes of erythrocytes is a stretch
because erythrocytes are noticeably lacking in cellular apparatus other than a cell membrane;
so the fact that the membrane is the target in those cells does not exclude the possibility of
other cellular targets in more broadly functional cells.
RESPONSE: The reviewers make some good points, most of which have been incorporated into
the position paper. However, the evidence for a nonlinear process in the formation of the
EGBE induced forestomach tumors significantly outweighs existing evidence for a linear
process. The relationship between the irritation effects/ulcers and tumors was stressed by NTP
(2000) and makes sense when examined in relation to other compounds that cause similar
forestomach effects. Also, no hyperplasia and no tumors were observed in inhalation studies of
rats (NTP, 2000) and in drinking water studies of mice (NTP, 1993a), supporting the need for
these steps prior to tumor formation. In addition, if a direct interaction of EGBE or a
metabolite were involved and irritation events were not necessary, one might expect to see
tumors earlier in the mice rather than near the end of the conducted bioassays as was observed
(NTP, 2000). The mutagenic compound ethylene dibromide, for instance, was reported to
induce forestomach tumors in all dose groups 168 to 280 days from the start of exposure (NCI,
1978). No biochemical event at low doses is known that would suggest linearity, and it is not
clear how a full PBPK model would clarify the contribution of BAL, given the available kinetic
A4-8
-------
data for EGBE and BAL metabolism. The available data support the necessity of a prior
irritation event, which suggests a nonlinear mode of action for the formation of these tumors.
However, the contribution of a linear component (e.g., BAL interaction with DNA) can not be
completely ruled out at this time.
8. Is there additional information (e.g., from other studies or studies of related compounds) that
would suggest an alternative conclusion?
All five reviewers said no, but the following comments were provided,
1. Inclusion of similar findings for other nongenotoxic carcinogens (e.g. ethyl acrylate)
(Ghanayem et al., 1985, 1993, and 1994) is recommended to support the proposed mode of
action.
2. Comparisons should be made with the local responses to other aldehydes.
3. Bring in information on clearly genetically-acting compounds such as epichlorhydrin, which
has been the subject of PBPK models with detailed modeling of local effects such as
glutathione reduction (e.g. Ginsberg, G. L., Pepelko, W. E., Goble, R. L., and Hattis, D. B.
"Comparison of Contact Site Cancer Potency Across Dose Routes: Case Study with
Epichlorohvdrin." Risk Analysis Vol. 16, pp. 667-681, 1996.).
4. The extensive discussion in the scientific literature of the significance of forestomach tumors
in rodents as indicators of potential human carcinogenicity should be addressed.
5. The paper presents a thorough and relatively convincing analysis of the "preferred"
hypothesis, but does not adequately address alternatives.
6. Lack of a forestomach in humans is not a valid justification for suggesting that EGBE induced
forestomach effects are not relevant to humans.
7. The fact that humans do not have an overall coat of hair, or extensive grooming behavior is a
more convincing argument that the experimental carcinogenicity observation is an unlikely
predictor of human risk in this particular case.
RESPONSE: It was not the paper's intention to suggest that EGBE-induced forestomach effects
are not relevant to humans. This was stated in the summary, but was not emphasized in
attachment 1. Additional language has been added to Attachment 1 to emphasize the potential,
qualitative relevance of these effects to humans. The main point that the document makes is
that due to a lack of grooming, the lack of a forestomach and a reduced sensitivity to the
cytotoxic effects of EGBE, humans are not likely to achieve the dose necessary to trigger the
irritation events necessary for the proposed mode of action.
9. The position paper offers an explanation for the relative insensitivity of male mice to the
formation of forestomach tumors following EGBE exposure. Is there an alternative
explanation?
Two reviewers said maybe, but:
1. If BAA is the responsible metabolite and the greater activity of alcohol dehydrogenase in
female mice leads to more BAA formation, then one may conclude that more BAL was
formed in female mice prior to conversion to BAA.
2. Need a quantitative analysis aided by PBPK modeling, and a theory of background interaction
to provide a quantitative estimate of likely response for comparison with the observations.
A4-9
-------
3. Further assessments could be conducted to investigate root causes (e.g., endocrine roles of
estrogens & androgens), but there are clear gender differences for many of the potential
etiological factors proposed to be responsible for tumor formation (e.g., greater sensitivity to
irritation/hyperplasia, higher levels of alcohol dehydrogenase, greater female RBC
sensitivities to the hemolytic effects, and lower plasma clearance rates of BAA).
4. The pharmacokinetic theory holds up as the principal available explanation, but this does not
exclude the possibility that other unexplored effects contribute (perhaps substantially) to the
difference.
RESPONSE: Comment #1 is certainly true. Some BAL will be formed in female mice following
EGBE exposure. Unfortunately, we do not know the relative aldehyde dehydrogenase activity
in the forestomachs of male versus female mice. The activity of hepatic alcohol dehydrogenase
enzyme is greater in female mice than in males (Dill et al, 1998), but the relevance of this has
not been fully assessed. The question of the basis for the sex specific sensitivity would benefit
from further research and analysis.
Mode of Action for Formation of Liver Tumors in Male Mice
10. Has the position paper drawn appropriate conclusions from the available literature on the
potential for EGBE to contribute to the formation of liver tumors in male mice?
Four reviewers said yes, but:
1. Add BAL in addition to BAA as causes of hemolysis because in vivo and in vitro evidence
suggested that BAL causes hemolysis (Ghanayem 1989 and 1996).
2. Although contrary to the NTP's position in its cancer bioassay report, the rationale is reasoned
and plausible. In addition, several studies that were conducted subsequent to the release of the
NTP report support the conclusions of the position paper, as does the publication of an NTP
scientist (Cunningham, 2002).
One reviewer was not sure and pointed out that:
1. Need a quantitative analysis of how much hemolysis is expected on an ongoing basis in the
different species/sexes, how much resulting accumulation of iron in affected cell types results
from this hemolysis, and then how much additional free radical generation and
clastogenic/mutagenic/carcinogenic change is likely to result from this pathway.
2. An attempt needs to be made to either qualitatively or quantitatively analyze the alternative
theory of genetic action via BAA[L] with the aid of knowable rates of generation and
metabolism of this compound and the hepatotoxic/carcinogenic activities of this and related
aldehydes.
3. Hemolysis/iron accumulation theory could be strengthened by experiments in which the
proliferative or early carcinogenic responses to EGBE were inhibited by administration of a
suitable chelator to tie up and facilitate the excretion of iron and thereby reduce the
accumulation in Kupffer cells or other cells of the liver.
RESPONSE: The Agency agrees with the majority of these comments. The potential
contributions of BAL to the hemolytic effects of EGBE will be discussed and the papers
provided by Dr. Ghanayem will be cited. With respect to comment #1, a quantitative analysis
of the "expected" relative hemolytic activity and iron buildup in the different species/sexes
A4-10
-------
could be done but would not be particularly informative as a number of in vivo studies have
confirmed that rats and mice of both sexes experience these effects to a similar degree of
severity, and occupational studies, case studies and in vitro studies indicate that humans would
not be expected to experience significant iron buildup from EGBE induced hemolysis. In
addition, predicting the amount of expected "free radical generation and
clastogenic/mutagenic/carcinogenic change" is not considered feasible at this time because the
actual mechanism of action for the induction of DNA damage (e.g., oxidative stress or induced
endothelial cell replication) is not clear.
Regarding comment #2, a recent attempt has been made to quantify the amount of
EGBE/BAA/BAL that would be present in the liver and GI tissues of female mice following a
250 ppm inhalation exposure to EGBE (Corley, 2003) using forestomach rate constants for
EGBE -> BAL provided by Green et al. (2002). This work is an attempt to extend an earlier
model developed by Dr. Corley (Corley et al., 2003) to include the intermediate formation of
BAL in target tissues (liver and GI tract). The limiting assumptions and results of this analysis
are described in the revised position paper.
Suggestion #3 might be an interesting and informative extension of the experiments
performed by Siesky et al. (2002) and could strengthen the hemolysis/iron accumulation theory.
While not directly answering the charge question, the reviewer appears to agree that the
position paper has drawn appropriate conclusions from the available literature regarding the
relationship between hemolysis, iron accumulation and oxidative stress.
11. Is there additional information (e.g., from other studies or studies of related compounds) that
would suggest an alternative conclusion?
One reviewer said yes. See question 10.
Four reviewers said no. One pointed out that although various alternatives could be proposed,
including the possibility of a direct genotoxic action via some undiscovered alternative metabolic
activation, they were not aware of any experimental data to support such alternative explanations.
RESPONSE: See responses question 10 above.
12. Is there any evidence in the literature that would suggest that the increased incidence of liver
hemangiosarcomas and hepatocellular carcinomas can be explained by mechanisms other
than those proposed or described?
One reviewer responded that they were unsure.
One reviewer responded yes, and stated that they believed one could glean some fodder for a possible
analogy with acetaldehyde from a careful review of the mechanisms by which alcohol causes
cirrhosis (and liver cancer??).
Three reviewers said no.
RESPONSE: The mechanism of action for other aldehydes such as acetaldehyde was considered
and included in the position paper's discussion of possible alternative mechanisms for the
induction of these liver carcinomas.
A4-11
-------
13. The position paper offers an explanation for the relative insensitivity of female mice to the
formation of liver tumors follow ing EGBE exposure. Is there an alternative explanation?
Two reviewers responded maybe, but stated that:
1. A comparison of the effects of EGBE on DNA synthesis in male and female mice could
confirm the hypothesis, and confirm the role of liver antioxidants in male vs. female mice.
2. A quantitative analysis aided by PBPK modeling, and a theory of background interaction is
needed for comparison with the observations.
Three reviewers said no, but one reviewer stated that since female mice are apparently more liable to
generate high intracellular levels of BAA than males (according to investigators seeking to explain
the forestomach results), pharmacokinetic differences between the sexes appear less plausible as an
explanation (although special situations and alternative routes of metabolism specific to the liver
could be involved).
RESPONSE: Male mice appear to be more sensitive than female mice to the induction of liver
hemangiosarcomas by all four chemicals that NTP has found to cause early onset hemosiderin
buildup in liver Kupffer cells (EGBE, p-chloroaniline, p-nitroaniline, and pentachloroanisole).
If, as proposed, these chemicals induce liver hemangiosarcomas through a mechanism involving
oxidative stress, the documented high sensitivity of male mice to liver oxidative stress (Klaunig
et al., 1995; 1998; 2000) offers a highly plausible explanation for this sensitivity. Further
support for this explanation is provided by the fact that the background incidence of liver
hemangiosarcomas is clearly higher in male versus female control mice. The proposed
hypothesis that males have an increased sensitivity to oxidative stress over females is not well
supported by direct experimental evidence, but does offer a reasonable explanation for both the
historical data and the apparent increased sensitivity of male mice to the induction of liver
hemangiosarcomas via an EGBE induced oxidative stress mechanism. This question would
benefit from additional research. However, the explanation suggested in the position paper is
the most reasonable explanation given the available data. No alternative explanation was
offered by any of the reviewers and it is not clear how additional PBPK modeling could result in
a more reasonable, alternative explanation.
Other
One reviewer commented that gender differences should be expanded to include most recent studies
which suggested that female rats are more sensitive than males and mice and suffered from
thrombosis, and infarction after inhalation and gavage exposure to EGBE and offered the following:
Nyska, A., Maronpot, R. R., Long, P. H., Roycroft, J. H., Haliy, J. R., Travlos, G. S., and
Ghanayem. B. I. (1999). Disseminated Thrombosis and Bone Infarction in Female Rats
Following Inhalation Exposure to 2-Butoxyethanol. Toxicol. Pathol. 27:287-294
Nyska, A., Maronpot, R. R., and Ghanayem. B. I. (1999). Ocular Thrombosis and Retinal
Degeneration in Female Rats by 2-Butoxyethanol. Human and Exp. Toxicol. 18:577-582.
Ghanayem, B. I, Ward, S M, Chanas, B, and Nyska, A. (2000). Comparison of the acute
hematotoxicity of 2-butoxyethanol in male and female F344 rats. Human Exp. Toxicol. 19:185-
192.
A4-12
-------
Long P. H, Maronpot, R. R., Ghanavem. B. I.. Roycroft, J H., and Nyska A (2000). Dental pulp
infarction in female rats following inhalation exposure to 2-butoxyethanol. Toxicol. pathol.
28:246-252.
Ghanavem. B. I. Long, P., Ward, S M, Chanas, B, and Nyska, A. (2001). Hemolytic Anemia,
Thrombosis, and Infarction in Male and Female F344 Rats Following Gavage Exposure to 2-
Butoxyethanol. Experimental and Toxicologic Pathology 53:97-105.
Another reviewer offered that similar observations with respect to temporal observations were made
with other nongenotoxic forestomach carcinogens and will strengthen the present discussion.
Ghanavem. B.L Maronpot, R.R. and Matthews, H.B. (1986). Association of Chemically-Induced
Forestomach Cell Proliferation and Carcinogenesis. Cancer Letters 32:271-278.
Ghanavem. B. I. Sanchez, I. M., Maronpot, R. R., Elwell, M. R., and Matthews, H. B. (1993).
Relationship Between the Time of Sustained Ethyl Acrylate Forestomach Cell Proliferation and
Carcinogenicity. Env. Health Persp. 101 (Sppl. 5):277-280.
Ghanavem. B. I. Sanchez, I. M., Matthews, H B., and Elwell, M. R. (1994). Demonstration of a
temporal relationship between ethyl acrylate induced forestomach hyperplasia and carcinogenesis.
Toxicologic Pathology 22:497-509.
Regarding the question "why were there no forestomach effects observed in the NTP (1993b)
subchronic drinking water study of mice, one reviewer pointed out that previous studies with other
nongenotoxic forestomach carcinogens demonstrated that forestomach effects are dependent not only
on the dose but also on the chemical concentration in the dosing solution (Ghanayem et al., 1985).
This may explain the observed effect in mice but not in rats (mice were exposed to higher
concentrations of EGBE than rats).
Ghanayem. B.L Maronpot, R.R. and Matthews, H.B. (1985). Ethyl Acrylate-Induced Gastric
Toxicity: II. Structure-Toxicity Relationships and Mechanism. Toxicol. Appl. Pharmacol.
80:336-344.
Nyska, A., Maronpot, R. R., and Ghanayem. B. I. (1999). Ocular Thrombosis and Retinal
Degeneration in Female Rats by 2-Butoxyethanol. Human and Exp. Toxicol. 18:577-582.
Ghanayem. B.L. Ward, S M, Chanas, B, and Nyska, A. (2000). Comparison of the acute
hematotoxicity of 2-butoxyethanol in male and female F344 rats. Human Exp. Toxicol.
19:185-192.
Long P. H, Maronpot, R. R., Ghanavem. B. I. Roycroft, J H., and Nyska A (2000). Dental
pulp infarction in female rats following inhalation exposure to 2-butoxyethanol. Toxicol.
pathol. 28:246-252.
Ghanayem. B.L. Long, P., Ward, S M, Chanas, B, and Nyska, A. (2001). Hemolytic Anemia,
Thrombosis, and Infarction in Male and Female F344 Rats Following Gavage Exposure to 2-
Butoxyethanol. Experimental and Toxicologic Pathology 53:97-1
A4-13
-------
One reviewer commented that, for a more comprehensive understanding of the report's message that
humans would need substantially higher exposures to EGBE to reach hemolytic risks, quantitative
information should be provided in the report. Show how the RfC and RfD were estimated originally
from rodent data, what were the safety factors used, what are the differences in BAA levels in rodents
vs. humans following comparable exposures to EGBE, what would be the additional margin-of-safety
if it is assumed that human RBCs are 100 times less sensitive to BAA, etc. While there is agreement
with the conclusions made in the report, the exposure, disposition, and metabolism stories would be
bolstered substantially by adding some quantitative information.
Another reviewer felt that it would be helpful to be more specific about the nature of the "increased
DNA synthesis" proposed and observed in the studies cited. Does it result from DNA repair or from
increased cell proliferation?
RESPONSE: In general, EPA agrees with these "other" comments made by the reviewers and
the position paper has been revised to take them into consideration and to incorporate the
suggested references.
A4-14
-------
2004 Panel Review Comments and Responses (Blue)
1. Liver hemangiosarcomas observed by NTP (2000) in male mice exposed to EGBE.
a. Does enough information now exist to support an informed decision concerning the
significance of the BAL metabolite to the formation of EGBE induced liver tumors?
All seven panel members agreed that enough information now exists to support an informed decision
concerning the significance of the BAL metabolite to the formation of EGBE induced liver tumors.
The following suggestions/comments were made regarding some additional information that could be
included in the position paper:
Whether ALDH deficient people constitute a subpopulation that is more susceptible to BAL
resulting from EGBE metabolism should be investigated or considered.
The additional data supplied with the EPA position paper (Klaunig and Kamendulis, 2005;
Deisinger and Boatman, 2004) should be included in the report as they provide important
evidence that the BAL metabolite is not likely to contribute to the formation of EGBE induced
liver tumors.
Nyska et al. (2004), retrospectively evaluating the results of 130 two-year carcinogenicity
studies conducted in B6C3FI mice at the NTP, have shown an overall association between
liver hemangiosarcoma and Kupffer cell pigmentation to be highly significant (p<0.001) and
limited to males.
A comparison of the Corley et al (2004) EGBE model results and the gavage study of
Deisinger and Boatman (2004) revealed that the predicted levels of BAL in both GI and liver
tissue are within 101% of the measured values.
The Corley et al. (2004) model has been used to predict BAL concentrations in the liver
following oral and inhalation exposures in mice and in mice with aldehyde dehydrogenase
metabolic rates (Vmax) set at !/2 the initial values (R. Corley, personal communication) and for
inhalation exposures up to the theoretical maximum of 1160 ppm for 6 hr, the prediction BAL
liver levels are not going achieve concentrations above 0.001 mM in such low metabolizing
individuals.
RESPONSE: EPA has addressed these concerns and suggestions in the revised position paper. As is
discussed in the position paper, while it is recognized that individuals with atypical alcohol
dehydrogenase and/or deficient aldehyde dehydrogenase may experience some increase, the Corley et
al. (2004) model suggests that BAL levels are not be expected to accumulate significantly in the liver
or forestomach, even if ALDH activity is half its normal rate and EGBE inhalation exposure levels
are assumed to be at the theoretical maximum of 1160 ppm. The data from the recent gavage study
by Deisinger and Boatman (2004) support the Corley et al. (2004) model results. These data, as well
as other recent reports relevant to the genotoxicity (Klaunig and Kamendulis, 2005) and mode of
action (Nyska et al, 2004) of EGBE have been considered in the revised position paper.
A4-15
-------
b. Is the current information adequate to support the mode of action described in the
position paper for the EGBE induced formation of hemangiosarcomas in male mice and the
potential relevance of this finding to humans?
All seven panel members agreed that the current information is adequate to support the mode of
action described in the position paper for the EGBE induced formation of hemangiosarcomas in male
mice and the potential relevance of this finding to humans. The following suggestions/comments
were made regarding some additional information that could be included in the position paper:
Iron bound to low molecular weight (LMW) chelators is redox active and, thus, is capable of
producing oxidants. It could be of great importance if this fraction of iron has been measured
in the animal studies. A newly developed fluorescent calcein method could measure LMW
iron in biological fluids of the EGBE-exposed animals (Ali etal., 2003).
Protection by various antioxidants, such as vitamin E, supports the oxidative stress
mechanism. However, a protection by specific iron chelators, such as deferoxamine, would
greatly strengthen the role of iron in the mode of action induced by EGBE.
It is difficult for this reviewer to believe that the iron deposition in the livers of these animals
did not occur in Kupffer and endothelial cells, although the report suggests (footnote Figure
A2-1) that there was not explicit mention of iron deposition in these cell types. If this point
becomes important, it is readily checked by a reexamination of the histology and if there is
any question, ultrastructural studies may be carried out, even on formalin-fixed tissues.
If further studies were to be done in this area, they may be oriented towards the possible role
of female sex hormones as a possible explanation for the fact that hemangiosarcomas were
observed in males only and the neoplastic response to the chronic inflammation and cell
proliferation in the forestomach was seen only in females.
Another aspect that needs to be considered is that BAA is an acid, which may release iron
from transferrin.
Some alternate considerations (also supported by scientific literature) may be involved in the
mode of action. In particular:
o in addition to the production of oxidative DNA damage, reactive oxygen species can
alter gene expression (e.g. MAP kinase/AP-1, and NFkB) resulting in stimulation of
cell proliferation and/or inhibition of apoptosis.
o If Kupffer cells are activated by iron as well as by phagocytosis of EGBE-induced
hemolyzed RBC's, and the production of reactive oxygen species occurs through
Kupffer cell derived reactions, the necessity for identifying iron in endothelial cells
(presumably to produce reactive oxygen within the target cell?) is lessened.
RESPONSE: The first four bulleted comments above refer to additional research that could be
conducted to confirm the mode of action and the explanation for the male mouse specificity that have
been presented in the position paper. While these types of studies would likely help to clarify the
proposed mode of action, they are not believed to be necessary for EPA to make an informed decision
regarding the human risk from EGBE exposure. Key steps that indicate a nonlinear mode of action
(e.g., the importance of hemolysis leading to hemosiderin buildup in liver Kupffer cells) are well
A4-16
-------
documented. What is know about these key steps and their relevance to mice and humans is
sufficient for the determination of the overall risk to humans and the characterization and, while other
potential contributing factors such as the acidic nature of BAA, the alteration of gene expression, the
necessity of iron directly in endothelial cells are considered in the revised position paper, verification
of additional factors contributing to the promotion or activation of endothelial cells is not considered
necessary at this time.
c. Does the available information support a nonlinear cancer assessment approach for the
male mouse liver tumors observed following EGBE exposure (i.e., is it reasonable to expect
that the prevention ofhemolytic effects in humans would prevent the formation of liver
tumors in humans) ?
All seven panel members agreed that the available information supports a nonlinear cancer
assessment approach for the male mouse liver tumors observed following EGBE exposure, and
therefore it is reasonable to expect that a lack ofhemolytic effects in humans would prevent the
formation of liver tumors in humans. The following suggestions/comments were made regarding
some additional information that could be included in the position paper:
The sex dependent effects might be explained by the fact that male mouse liver is well known
to be more sensitive to the development of tumors than female mouse liver. Thus exposure to a
similar concentration of EGBE may not have been quite sufficient to induce a similar increased
incidence in female mice.
The only possible group that might be affected (to this reviewer's knowledge) might be the
hemochromatosis heterozygote that comprises some 12% of the human population (Barton and
Bertoli, 1996). Smaller amounts of hemolysis in these individuals could lead, over extended
periods, to some chronic iron deposition in hepatocytes, but it would seem unlikely that even
such individuals would have any problem.
RESPONSE: None of the reviewers offered a linear mechanism for the formation of
hemangiosarcomas. Some offered possible explanations for male mouse specificity and one offered
that hemochromatosis heterozygotes might represent a sensitive subpopulation because smaller
amounts of hemolysis in these individuals could lead, over extended periods, to some chronic iron
deposition in hepatocytes. However, this same reviewer did not believe that these individuals could
have an EGBE related iron overload problem.
2. Forestomach tumors observed by NTP (2000) in female mice following EGBE exposure.
a. Does enough information now exist to support an informed decision concerning the
significance of the BAL metabolite to the formation of EGBE induced forestomach tumors?
All seven panel members agreed that enough information now exists to support an informed decision
concerning the significance of BAL genotoxicity to the formation of EGBE induced forestomach
tumors. The following suggestions/comments were made regarding some additional information that
could be included in the position paper:
A4-17
-------
Levels of BAL in ALDH-deficient people may be a concern (see la).
Available information indicates that BAL does not result in genetic alteration that leads to an
increased incidence of forestomach tumors. However, it has not been determined that this
relatively reactive metabolite does not account, at least in part, for the chronic irritation that
results in an increased incidence of tumors.
The female mice had particularly severe irritation in the forestomach relative to males treated
with the same dose, which supports a non-genotoxic mode of action for the female
forestomach carcinogenesis.
The model of Corley et al (2004) has been used to predict BAL concentrations in the GI tract
following oral and inhalation exposures in mice and in mice with aldehyde dehydrogenase
metabolic rates (Vmax) set at 1A the initial values. Only after oral doses of 300 mg/kg are BAL
levels comparable to those found to have an effect in vitro. For inhalation exposures up to the
theoretical maximum of 1160 ppm for 6 hr, the prediction BAL liver levels are not going to
achieve concentrations above 0.01 mM in these low metabolizing individuals (Table A4-1).
This predicted maximal concentration is considerably lower than concentrations of BAL
shown to be clastogenic (0.2 mM) or hemolytic (0.5 mM: Ghanayem et al., 1989) in vitro.
RESPONSE: It is recognized that BAL may be partially responsible for the irritation effects
following EGBE exposure and that ALDH deficient populations may experience higher BAL levels.
However, as one reviewer pointed out, while BAL has only been detected at very low levels in blood
following high oral EGBE doses, BAA has been detected over time and at relatively high levels in the
forestomach of female mice following in vivo exposures to EGBE through multiple routes. Another
reviewer also pointed out that since it is not possible to administer EGBE or BAL to intact animals
without their rapid metabolism to BAA conclusive determination of the contribution of BAL to the
irritating effects of EGBE is probably not feasible. Nevertheless, the Corley et al. (2004) model and
the Deisinger and Boatman (2004) gavage studies make it clear that BAL levels in the GI tract would
be very low.
A4-18
-------
Table A4-1. Dose-response simulations of the peak tissue concentrations (Cmax) of
butoxyacetaldehyde (BAL) in female mice following either oral gavage or 6-hr inhalation exposures.
To simulate a heterozygous population with lower aldehyde dehydrogenase activity, simulations with
l/2 the Vmax rate are also shown.
Cmax BAL
Liver @ 1A
Vmax
Cmax BAL
Liver
Cmax BAL
GI Tract
Cmax BAL
GI Tract @
: Vmax
Route or
Exposure (ppm)
Oral
Inhalation
Dose
(mg/kg)
1
10
25
50
100
150
300
500
600
900
1
5
10
25
50
63
100
125
150
200
250
500
750
950
1160
(HM)
0.002
0.021
0.056
0.123
0.305
0.584
2.241
4.211
4.586
4.991
0.000
0.002
0.003
0.008
0.016
0.020
0.032
0.040
0.048
0.064
0.081
0.164
0.249
0.320
0.395
0.004
0.043
0.112
0.247
0.615
1.187
4.773
9.502
10.47
11.53
0.001
0.003
0.006
0.016
0.032
0.040
0.064
0.080
0.096
0.129
0.162
0.329
0.502
0.645
0.799
0.068
0.686
1.755
3.644
7.85
12.57
25.07
31.61
32.99
34.96
0.003
0.015
0.030
0.076
0.153
0.193
0.307
0.384
0.462
0.618
0.775
1.576
2.404
3.086
3.820
0.136
1.399
3.686
8.090
19.95
38.22
160.5
419.2
525.3
725.6
0.006
0.030
0.061
0.153
0.307
0.388
0.619
0.776
0.935
1.257
1.583
3.292
5.141
6.732
8.519
b. Is the current information adequate to support the mode of action described in the
position paper for the EGBE induced formation offorestomach tumors in female mice and
the potential relevance of this finding to humans?
All seven panel members agreed that the current information is adequate to support the mode of
action described in the position paper for the EGBE induced formation of forestomach tumors in
female mice and the potential relevance of this finding to humans. The following
suggestions/comments were made regarding some additional information that could be included in
the position paper:
Whether the increased DNA synthesis results in acquisition of new mutations or results in a
selective clonal expansion of initiated cells (i.e. functions at the tumor promotion stage of
carcinogenesis), has not been established.
A4-19
-------
More careful studies of the fate of EGBE in forestomach might prove helpful in determining
just how much of each of the three chemicals, EGBE, BAL and BAA, are present in
forestomach cells, how long they persist and why EGBE in drinking water did not induce
similar lesions/tumors.
The presence of a single carcinoma merely indicates the spontaneous transition from cells in
the stage of promotion (papilloma) to those in the stage of progression (carcinoma).
If further studies were to be done in this area, they may be oriented towards the possible role
of female sex hormones in the neoplastic response to the chronic inflammation and cell
proliferation in the forestomach seen in females.
The lack of effect observed following drinking water studies may indicate a buffering of this
irritating effect from the water vehicle or, more likely, a dose-rate effect. In addition, neither
EGBE nor its major metabolite binds to stomach macromolecules.
RESPONSE: Though the data seem to be more consistent with forestomach effects being the result of
clonal expansion of initiated cells, it is realized that this is not the only possible explanation.
However, cell turnover resulting from irritation appears to play a key role for EGBE and other
chemicals that cause an increase in forestomach tumors. Studies of the distribution of EGBE and its
metabolites in the gut have been done and, as the reviewer points out, additional studies would
probably still not conclusively resolve the question of whether EGBE or a metabolite accounts for
most of the chronic irritation that lead to an increased incidence of tumors.
c. Does the available information support a nonlinear cancer assessment approach for the
female mouse forestomach tumors observed following EGBE exposure (i.e., is it reasonable
to expect that the prevention of hyperplastic effects in humans would prevent the formation
of gastrointestinal tumors in humans) ?
All seven panel members agreed that the available information supports a nonlinear cancer
assessment approach for the female mouse forestomach tumors observed following EGBE exposure
and therefore making it reasonable to expect that a lack of hyperplastic effects in the region of
gastroesophageal junction in humans would prevent the formation of gastroesophageal tumors in
humans. The following suggestions/comments were made regarding some additional information that
could be included in the position paper:
Esophageal and gastric emptying occur relatively rapidly within the human (a matter of
minutes to a few hours) unlike the mouse and rat under the conditions of the assay.
A potential alternative mode of action involving direct DNA reactivity by BAL was
previously postulated. However, as the comments outlined under question 2a above indicate,
the contribution of BAL to the induction of forestomach tumors in female mice is not likely to
contribute to the observed neoplasia based on pharmacokinetic factors.
The mode of action for the induction of forestomach tumors in mice would be expected to
apply to humans (i.e., the key events could occur in humans). However, taking into account
kinetic and dynamic factors, the key events in the mode of action is not likely to occur in
humans.
A4-20
-------
There is a clear association between chronic irritation and an increased incidence of
forestomach tumors, but it can not be conclusively determined if the parent chemical, EGBE,
or one of its major metabolites, BAL and BAA, accounted for the chronic irritation.
In the absence of intentional consumption, humans will not encounter similar exposures to
EGBE and even then the exposures would be acute rather than chronic. Thus, it would appear
that the forestomach tumors observed in female mice are not relevant to humans.
RESPONSE: EPA agrees with the reviewers on all points and these considerations are included in the
revised position paper.
3. Additional Comments and References
A description of how mice were housed would be helpful.
It is a possible that an increased incidence of forestomach tumors were observed in the
inhalation studies, but not in drinking water studies because the drinking water studies did
not permit consumption of neat EGBE as a result of condensation on the airways and/or
grooming. The inhalation study may have achieved both more concentrated and more
prolonged exposure of the forestomach to EGBE than did the drinking water study.
RESPONSE: The Agency agrees with these additional comments provided by the reviewers. The
new references listed below that were provided by the reviewers were obtained and evaluated, and the
position paper has been updated accordingly.
References Cited by Reviewers
Ali A, Zhang Q, Dai J, Huang X. 2003. Calcein as a fluorescent iron chemosensor for the
determination of low molecular weight iron in biological fluids. Biometals 16: 285-293.
Crabb DW, Edenberg HJ, Bosron WF, Li TK. 1989. Genotypes for aldehyde dehydrogenase
deficiency and alcohol sensitivity. The inactive ALDH2(2) allele is dominant. J Clin Invest 83:
314-316.
Gold, L. S., Manley, N. B., Slone, T. H., Garfmkel, G B., Rohrbach, L, Ames, B. N. (1993) The fifth
plot of the carcinogenic potency database: results of animal bioassays published in the general
literature through 1988 and by the National Toxicology Program through 1989. Environ. Health
Perspect. 100: 65-135.
Hsu LC, Chang WC, Yoshida A. 1994. Cloning of a cDNA encoding human ALDH7, a new member
of the aldehyde dehydrogenase family. Gene 151: 285-289.
Kitagawa K, Kawamoto T, Kunugita N, Tsukiyama T, Okamoto K, Yoshida A, Nakayama K. 2000.
Aldehyde dehydrogenase (ALDH) 2 associates with oxidation of methoxyacetaldehyde; in vitro
analysis with liver subcellular fraction derived from human and Aldh2 gene targeting mouse. FEES
Lett 476: 306-311.
A4-21
-------
Klaunig, JE and Kamendulis, LM (2004). The role of oxidative stress in carcinogenesis. Ann. Rev.
Pharm. Toxicol 44: 239-267.
Klaunig, I.E. and Kamendulis, L.M. (2005) Mode Of Action Of Butoxyethanol Induced Mouse Liver
Hemangiosarcomas And Hepatocellular Carcinomas. Toxicol. Lett. 156(1): 107-15.
Knutson, M.; Wessling-Resnick, M. (2003) Iron metabolism in the reticuloendothelial system. Crit.
Rev. Biochem. Mol. Biol. 38: 61-88.
Nyska A, Haseman JK, Kohen R, Maronpot RR (2004). Association of Liver Hemangiosarcoma and
secondary iron overload in B6C3F1 Mice - The National Toxicology Program Experience.
Toxicologic Pathology. 32:222-228.
Pitot, H. C. (2002) Fundamentals of Oncology, 4th edition, 998 pp. New York: Marcel Dekker.
Swenberg, J. A.; Barrow, C. S.; Boreiko, C. I; Heck, H.d'A.; Levine, R. I; Morgan, K. T.; Starr, T. B.
(1983) Non-linear biological responses to formaldehyde and their implications for carcinogenic risk
assessment. Carcinogenesis 4: 945-952.
Voutilainen, M.; Sipponen, P.; Mecklin, J.-P.; Juhola, M.; Farkkila, M. (1999) Gastroesophageal
reflux disease: prevalence, clinical, endoscopic and histopathological findings in 1,128 consecutive
patients referred for endoscopy due to dyspeptic and reflux symptoms. Digestion 61:6-13.
A4-22
------- |