vvEPA
EPA/635/R-20/400FC
www.epa.gov/iris
Toxicological Review of Ethyl Tertiary Butyl Ether
(CASRN 637-92-3]
August 2021
Integrated Risk Information System
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
EXECUTIVE SUMMARY
Summary of Occurrence and Health Effects
Ethyl tertiary -butyl ether (ETBE) does not occur naturally; it is a man-made ether
oxygenate used primarily as a gasoline additive. It was used until 2006 in the United
States and is still used in Japan and the European Union. ETBE is released into the
environment through gasoline leaks, evaporation, and spills. Exposure to ETBE can
occur by drinking contaminated groundwater or by inhaling off-gases containing
ETBE. Dermal exposure is possible in occupational settings where ETBE is
manufactured. The magnitude of human exposure to ETBE depends on factors such
as the distribution of ETBE in groundwater and the extent of the contamination.
Animal studies demonstrate that exposure to ETBE is associated with noncancer
kidney effects following oral and inhalation exposure. Evidence is suggestive of
carcinogenic potential for ETBE based on liver tumors in rats following inhalation
exposure.
EFFECTS OTHER THAN CANCER OBSERVED FOLLOWING ORAL EXPOSURE
Kidney effects are identified in this assessment as a potential human hazard of ETBE
exposure. Although no human studies are available to evaluate the effects of ETBE, oral exposure
studies in animals have consistently reported increased kidney weight in male and female rats
accompanied by increased chronic progressive nephropathy (CPN), urothelial hyperplasia of the
renal pelvis (in males), and increased blood concentrations of total cholesterol, blood urea nitrogen
(BUN), and creatinine. Overall, there was consistency across multiple measures of potential kidney
toxicity, including organ-weight increases, exacerbated CPN, urothelial hyperplasia of the renal
pelvis, and increases in serum markers of kidney function. Additionally, effects were also observed
across routes of exposure, and sex (with the exception of urothelial hyperplasia of the renal pelvis
which was observed only in male rats).
The relevance of the kidney findings to humans was evaluated with respect to alpha
2u-globulin nephropathy, a disease process that occurs exclusively in the male rat kidney (Capen et
al.. 1999: U.S. EPA. 1991al. While ETBE binds to alpha 2u-globulin and meets some criteria of the
alpha 2u-globulin U.S. Environmental Protection Agency (EPA) and International Agency for
Research on Cancer (IARC) frameworks (Capen etal.. 1999: U.S. EPA. 1991a). it does not meet all.
With respect to male rats, U.S. EPA (1991a) noted that "[i]f a compound induces a2u-globulin
accumulation in hyaline droplets, the associated nephropathy in male rats is not an appropriate
endpoint to determine noncancer (systemic) effects potentially occurring in humans." However, as
alpha 2u-globulin nephropathy is strictly a male rat phenomenon, the dose-related kidney effects in
female rats are not confounded by alpha 2u-globulin nephropathy.
2

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
It has been observed that chemicals that bind to alpha 2u-globulin also exacerbate the
incidence and/or severity of background CPN in male rats fFrazier etal.. 2012: Travlos etal.. 2011:
U.S. EPA. 1991al. While the etiology of CPN is unknown fNIEHS. 2019: Hard and Khan. 2004: Peter
etal.. 19861 and it has no known analogue in the aging human kidney fNIEHS. 2019: Hard etal..
20091. it cannot be ruled out that a chemical that exacerbates CPN in rats could also exacerbate
disease processes in the human kidney [e.g., chronic kidney disease, diabetic nephropathy,
glomerulonephritis, interstitial nephritis, etc.; NIEHS f20191]. Therefore, increased incidence of
kidney effects with ETBE exposure in the female rat (but not the male rat) are considered
appropriate for identifying a hazard to the kidney.
Evidence is suggestive that liver toxicity follows oral ETBE exposure. The strongest
supporting evidence is the increased liver weights and centrilobular hypertrophy in exposed male
and female rats consistently reported across oral-exposure studies. No additional histopathological
findings were observed, however, and only one serum marker potentially indicative of liver toxicity
(y-glutamyl transferase [GGT]) was elevated, while other markers (aspartate aminotransferase
[AST], alanine aminotransferase [ALT], and alkaline phosphatase [ALP]) were unchanged. The
magnitude of change for these noncancer effects was minimal and, except for organ-weight data, no
consistent dose-response relationships were observed. Mechanistic data suggest that ETBE
exposure leads to activation of several nuclear receptors, but there is inadequate evidence to
establish a relationship between receptor activation and liver toxicity resulting from ETBE
exposure. In addition, mechanistic data suggest possibly greater susceptibility of toxic effects
related to reduced clearance of acetaldehyde, a metabolite of ETBE. Thus, even with the
consistently observed increases in rat liver weight and centrilobular hypertrophy, the evidence
remains suggestive that liver toxicity follows ETBE exposure because of the relatively small
magnitude of effects and inconsistent dose-response relationships.
Inadequate information exists to draw conclusions regarding reproductive, developmental,
or immune system effects. The ETBE database does include developmental, reproductive, and
multigenerational studies, which are generally null and do not appear to indicate an area of concern
(see hazard discussions in Sections 1.2.3 and 1.2.4). However, this body of evidence is not
sufficiently robust to conclude that ETBE is not likely to be a reproductive or developmental hazard.
Regarding immune system effects, the ETBE database contains no evidence of altered immune
function that correlate with modest T cell population reductions and altered splenic organ weights
(see Appendix B of the Supplemental Information); thus, the available immune data are inadequate
to draw conclusions as a human hazard of ETBE exposure.
ORAL REFERENCE DOSE (RFD) FOR EFFECTS OTHER THAN CANCER
Kidney toxicity, represented by increased absolute kidney weight in female rats, was chosen
as the basis for the overall RfD (see Table ES-1). The chronic study by TPEC f2010al [with selected
data published as Suzuki etal. f20121] and the observed kidney effects were used to derive the RfD.
The endpoint of increased kidney weight was selected as the critical effect because it is a specific
3

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
and sensitive indicator of kidney toxicity and was induced in a dose-responsive manner. The
increases in kidney weight seen in male and female rats are likely a result of the increases in the
severity of CPN seen with ETBE exposure. A characteristic feature of CPN in the rat is increasing
kidney size and weight fHard etal.. 20131. Benchmark dose (BMD) modeling was used to derive the
benchmark dose lower confidence limit corresponding to 10% extra risk (BMDLio) of
120 mg/kg-day. The BMDL was converted to a human equivalent dose (HED) of 28.8 mg/kg-day
using body weight3/4 scaling (U.S. EPA. 20111. and this value was used as the point of departure
(POD) for RfD derivation.
The overall RfD was calculated by dividing the POD for increased absolute kidney weight by
a composite uncertainty factor (UFc) of 30 to account for extrapolation from animals to humans (3)
and interindividual differences in human susceptibility (10).
Table ES-1. Organ/system-specific RfDs and overall RfD for ETBE
Hazard
Basis
Point of departure3
(mg/kg-d)
UFC
Chronic RfD
(mg/kg-d)
Study exposure
description
Confidence
Kidney
Increased absolute
kidney weight
28.8
30
lx 10°
Chronic
High
Overall RfD
Kidney
28.8
30
1 x 10°
Chronic
High
aHuman equivalent dose PODs were calculated using body weight scaling to the 3/4 power (BW3/4) (U.S. EPA,
2011).
EFFECTS OTHER THAN CANCER OBSERVED FOLLOWING INHALATION EXPOSURE
Kidney effects are a potential human hazard of inhalation exposure to ETBE. Although no
human studies are available to evaluate the effects of exposure, studies in animals have observed
increases in kidney weight, altered kidney histopathology, as well as alterations in clinical
chemistry including serum cholesterol, BUN, and creatinine. While the histological lesion of
urothelial hyperplasia of the renal pelvis was a sensitive endpoint in male rats, it was not observed
in female rats or mice of either sex, whereas increases in kidney weight were observed in multiple
studies in rats of both sexes and in mice. Changes in kidney weight in female rats were dose-
dependent, consistent across multiple studies and are not confounded by alpha 2u-globulin
nephropathy, and therefore considered appropriate for identifying a hazard to the kidney.
INHALATION REFERENCE CONCENTRATION (RFC) FOR EFFECTS OTHER THAN CANCER
Kidney toxicity, represented by increased absolute kidney weight, was chosen as the basis
for the overall RfC (see Table ES-2). The chronic study by TPEC (2010b) [selected data published as
Saito etal. (2013)] and the observed kidney effects were used to derive the RfC. The endpoint,
increased absolute kidney weight, was selected as the critical effect because it is a specific and
4

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
sensitive indicator of kidney toxicity and was induced in a dose-responsive manner. The increases
in kidney weight are likely a result of the increases in the severity of CPN seen with ETBE exposure,
as CPN is characterized by cell proliferation and chronic inflammation that results in increased
kidney weight fMelnick et al.. 2012: Travlos etal.. 20111.
Benchmark concentration modeling of increased kidney weight (in female rats) was
attempted, but an adequate fit was not achieved. Therefore, a no-observed-adverse-effect level
(NOAEL) was used to derive the POD of 6,270 mg/m3. The NOAEL was adjusted for continuous
exposure and converted to a human equivalent concentration (HEC) of 1,110 mg/m3.
The overall RfC was calculated by dividing the POD for increased absolute kidney weight by
a UF of 30 to account for toxicodynamic differences between animals and humans (3) and
interindividual differences in human susceptibility (10).
Table ES-2. Organ/system-specific RfCs and overall RfC for ETBE
Hazard
Basis
Point of departure3
(mg/m3)
UF
Chronic RfC
(mg/m3)
Study exposure
description
Confidence
Kidney
Increased absolute
kidney weight
1,110
30
4x 101
Chronic
Medium
Overall RfC
Kidney
1,110
30
4x 101
Chronic
Medium
Continuous inhalation HEC was adjusted for continuous daily exposure and calculated by adjusting the
duration-adjusted POD (PODadj) by the dosimetric adjustment factor (DAF = 0.992) for a Category 3 gas.
EVIDENCE OF HUMAN CARCINOGENICITY
Under EPA Cancer Guidelines (U.S. EPA. 2005a). the evidence of carcinogenic potential for
ETBE is suggestive for inhalation exposure but inadequate for oral exposure. ETBE induced liver
tumors in male (but not female) rats in a 2-year inhalation exposure study fSaito etal.. 2013: TPEC.
2010b). No significant effects were observed in two chronic oral studies in male and female rats
[one of high quality; see Section 1.2.5; TPEC f2010al: Maltoni etal. f 19991], Data on tumorigenicity
in mice following ETBE exposure were not available. However, supplementary evidence from
two-stage initiation-promotion oral carcinogenesis bioassays indicate increased mutagen-initiated
liver tumors, as well as increased tumor incidence in the thyroid, colon, and urinary bladder.
QUANTITATIVE ESTIMATE OF CARCINOGENIC RISK FROM ORAL EXPOSURE
A quantitative estimate of carcinogenic potential from oral exposure to ETBE was not
derived because an increase in tumors was not observed in the two available chronic oral cancer
bioassays (TPEC. 2010a: Maltoni et al.. 1999). A route-to-route extrapolation of cancer risk from the
inhalation-to-oral route was not carried out because there was no consistent dose-response
5

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
relationship observed for liver tumors when compared across oral and inhalation studies on the
basis of physiologically based pharmacokinetic (PBPK) modeled internal dose.
QUANTITATIVE ESTIMATE OF CARCINOGENIC RISK FROM INHALATION EXPOSURE
Although ETBE was considered to have suggestive evidence of carcinogenic potential, the
main study (Saito etal.. 2013: TPEC. 2010b) was conducted according to well-established guidelines
for examining potential carcinogenicity and was suitable for quantitative analyses. Thus, while
recognizing the uncertainty in the data and the suggestive nature of the weight of evidence, the
analysis may be useful for some purposes, such as providing a sense of the magnitude of potential
risks, ranking potential hazards, or setting research priorities fU.S. EPA. 2005a).
A quantitative estimate of carcinogenic potential from inhalation exposure to ETBE was
based on the increased incidence of hepatocellular adenomas and carcinomas in male F344 rats
following 2-year inhalation exposure (Saito etal.. 2013: TPEC. 2010b). The study included
histological examinations for tumors in many different tissues, contained three exposure levels and
controls, contained adequate numbers of animals per dose group (~50/sex/group), treated the
animals for up to 2 years, and included detailed reporting of methods and results.
An inhalation unit risk was derived for liver tumors in male F344 rats. The modeled ETBE
POD was scaled to an HEC according to EPA guidance based on inhalation dosimetry for a
Category 3 gas fU.S. EPA. 19941. Using linear extrapolation from the benchmark concentration
lower confidence level corresponding to 10% extra risk (BMCLio), a human equivalent inhalation
unit risk was derived using inhalation unit risk = 0.1/BMCLio and calculated to be
8 x 10"5 per mg/m3.
SUSCEPTIBLE POPULATIONS AND LIFESTAGES FOR CANCER AND NONCANCER OUTCOMES
ETBE is metabolized to tert-butanol and acetaldehyde. Evidence is suggestive that genetic
polymorphism of aldehyde dehydrogenase (ALDH)—the enzyme that oxidizes acetaldehyde to
acetic acid—could affect ETBE toxicity. The virtually inactive form, ALDH2*2, is found in about
one-half of all East Asians [and by extension people of East Asian ancestry; Brennan etal. (2004)].
Evidence is strong in humans that this ALDH2 variant increases the internal dose of acetaldehyde
and the cancer risks from acetaldehyde, especially in the development of ethanol-related cancers
fEriksson. 2015: IARC. 20101. Several in vivo and in vitro genotoxicity assays in Aldh2 knockout
(KO) and heterozygous mice reported that genotoxicity was significantly increased compared with
wild-type controls following ETBE exposure to similar doses associated with cancer and noncancer
effects in rodents (Weng etal.. 2019: Weng etal.. 2014: Weng etal.. 2013: Weng etal.. 2012: Weng
etal.. 2011). Inhalation ETBE exposure increased blood concentrations of acetaldehyde in Aldh2
KO mice compared with wild type (Weng etal.. 2013). Thus, exposure to ETBE in individuals with
the ALDH2*2 variant would be expected to increase the internal dose of acetaldehyde and
potentially increase risks associated with acetaldehyde produced by ETBE metabolism in the liver.
6

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
Collectively, these data present evidence that people with diminished ALDH2 activity could
be considered a susceptible population that could be more sensitive to liver toxicity from ETBE
exposure.
KEY ISSUES ADDRESSED IN ASSESSMENT
Dose-related kidney effects were observed in male and female rats. The human relevance of
these effects, particularly as they relate to alpha 2u-globulin nephropathy and the exacerbation of
chronic progressive nephropathy, is a key issue analyzed in this assessment. An evaluation of
whether ETBE causes alpha 2u-globulin-associated nephropathy was performed using the EPA and
IARC frameworks (Capen etal.. 1999: U.S. EPA. 1991a). ETBE induced an increase in hyaline
droplet accumulation and increased alpha 2u-globulin deposition in male rats; however, most of the
subsequent steps in the pathological sequence were not observed. U.S. EPA f!991al states that "[i]f
a compound induces a2u-globulin accumulation in hyaline droplets, the associated nephropathy in
male rats is not an appropriate endpointto determine noncancer (systemic) effects potentially
occurring in humans." However, because alpha 2u-globulin nephropathy is strictly a male rat
phenomenon, the dose-related kidney effects in female rats are not confounded by alpha
2u-globulin nephropathy. CPN also plays a role in exacerbating nephropathy in rats; however, the
mode of action (MOA) is unknown. Given that there is no definitive pathogenesis for CPN, it cannot
be fully ruled out that chemicals that exacerbate CPN in rats may have the potential to exacerbate
other disease processes in the human kidney (NIEHS. 2019). Dose-related changes in several
indicators of kidney toxicity were observed in male and female rats, including increased absolute
kidney weight, histological changes, and increased blood biomarkers (Saito etal.. 2013: Suzuki et
al.. 2012: TPEC. 2010a. b). These specific effects are considered relevant to humans, particularly the
endpoints observed in female rats, because they are not confounded by alpha 2u-globulin related
processes.
In addition, the human relevance of the observed liver tumors is discussed in the
assessment (see Sections 1.2.2 and 1.3.2). Briefly, a well-conducted inhalation study demonstrated
a significant, positive exposure-response for hepatocellular adenomas and carcinomas in male rats
(Saito etal.. 2013: TPEC. 2010b). While the majority of liver tumors occurred at the highest
exposure, statistical tests conducted by the study authors found a significant dose-response trend
by both the Peto (incidental tumor test) and the Cochran-Armitage tests. However, two chronic
oral exposure studies (one with unrelated mortality were negative for liver tumors flPEC. 2010a:
Maltoni et al.. 1999). The integration of relevant carcinogenic evidence is discussed in Section 1.3.2.
The potential MOA for the observed liver tumors was accessed in the Section 1.2.2. The
available evidence base for the nuclear hormone receptor MOAs (i.e., peroxisome
proliferator-activated receptor a [PPARa], pregnane X receptor [PXR], and the constitutive
androstane receptor [CAR]) was inadequate to determine the role these pathways play, if any, in
ETBE-induced liver carcinogenesis. Acetaldehyde-mediated genotoxicity also was evaluated as a
possible MOA, and although evidence suggests that ALDH2 deficiency enhanced ETBE-induced
7

-------
Toxicological Review of Ethyl Tertiary Butyl Ether
genotoxicity in exposed mice, the available database was inadequate to establish
acetaldehyde-mediated mutagenicity as an MOA for ETBE-induced liver tumors. No other MOAs for
liver carcinogenesis were identified. Because an MOA for liver carcinogenicity could not be
established, in the absence of data to indicate otherwise, the rat liver tumors observed following
inhalation exposure are considered relevant to humans fU.S. EPA. 2005al
8

-------