-------
Supplemental Information—tert-Butyl Alcohol
1 DISCLAIMER
2 This document is a preliminary draft for review purposes only. This information is
3 distributed solely for the purpose of pre-dissemination peer review under applicable information
4 quality guidelines. It has not been formally disseminated by EPA. It does not represent and should
5 not be construed to represent any Agency determination or policy. Mention of trade names or
6 commercial products does not constitute endorsement or recommendation for use.
This document is a draft for review purposes only and does not constitute Agency policy.
ii DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert-Butyl Alcohol
CONTENTS
APPENDIX A. ASSESSMENTS BY OTHER NATIONAL AND INTERNATIONAL HEALTH AGENCIES A-l
APPENDIX B. INFORMATION IN SUPPORT OF HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS B-l
B.l. TOXICOKINETICS B-l
B.1,1. Absorption B-l
B.1.2. Distribution B-2
B.1.3. Metabolism B-2
B.1,4, Excretion B-5
B.1,5. Physiologically Based Pharmacokinetic Models B-7
B.1.6. PBPK Model Code B-8
B.2. OTHER PERTINENT TOXICITY INFORMATION B-9
B.2,1. Other Toxicological Effects B-9
B.2.2. Genotoxicity B-22
B.2.3. Summary B-26
APPENDIX C. DOSE-RESPONSE MODELING FOR THE DERIVATION OF REFERENCE VALUES FOR
EFFECTS OTHER THAN CANCER AND THE DERIVATION OF CANCER RISK
ESTIMATES C-l
C.1.1. Noncancer Endpoints C-l
C.1,2. Cancer Endpoints C-23
APPENDIX D. SUMMARY OF PUBLIC COMMENTS AND EPA's DISPOSITION D-l
REFERENCES R-l
This document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Supplemental Information—tert-Butyl Alcohol
TABLES
Table A-l. Health assessments and regulatory limits by other national and international
health agencies A-l
Table B-l. Changes in kidney weight in animals following exposure to ferf-butanol B-ll
Table B-2. Changes in liver weight in animals following exposure to ferf-butanol B-14
Table B-3. Changes in liver histopathology in animals following exposure to ferf-butanol B-16
Table B-4. Changes in urinary bladder histopathology in animals following oral exposure to
ferf-butanol B-18
Table B-5. Summary of genotoxicity (both in vitro and in vivo) studies of ferf-butanol B-25
Table C-l. Noncancer endpoints selected for dose-response modeling for ferf-butanol C-2
Table C-2. Summary of BMD modeling results for kidney transitional epithelial hyperplasia in
male F344 rats exposed to ferf-butanol in drinking water for 2 years (NTP,
1995); BMR = 10% extra risk C-3
Table C-3. Summary of BMD modeling results for kidney transitional epithelial hyperplasia in
female F344 rats exposed to ferf-butanol in drinking water for 2 years (NTP,
1995); BMR = 10% extra risk C-6
Table C-4. Summary of BMD modeling results for absolute kidney weight in male F344 rats
exposed to ferf-butanol in drinking water for 15 months (NTP, 1995); BMR =
10% rel. dev. from control mean C-9
Table C-5. Summary of BMD modeling results for absolute kidney weight in female F344 rats
exposed to ferf-butanol in drinking water for 15 months (NTP, 1995); BMR =
10% rel. dev. from control mean C-12
Table C-6. Summary of BMD modeling results for kidney inflammation in female rats
exposed to ferf-butanol in drinking water for 2 years (NTP, 1995); BMR = 10%
extra risk C-15
Table C-7. Summary of BMD modeling results for absolute kidney weight in male F344 rats
exposed to ferf-butanol via inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP,
1997); BMR = 10% relative deviation from the mean C-18
Table C-8. Summary of BMD modeling results for absolute kidney weight in female F344 rats
exposed to ferf-butanol via inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP,
1997); BMR = 10% relative deviation from the mean C-21
Table C-9. Cancer endpoints selected for dose-response modeling for ferf-butanol C-24
Table C-10. Summary of the oral slope factor derivations C-25
Table C-ll. Summary of BMD modeling results for thyroid follicular cell adenomas in female
B6C3F1 mice exposed to ferf-butanol in drinking water for 2 years (NTP, 1995);
BMR = 10% extra risk C-26
Table C-12. Summary of BMD modeling results for thyroid follicular cell adenomas or
carcinomas in male B6C3F1 mice exposed to ferf-butanol in drinking water for
2 years (NTP, 1995); BMR = 5% extra risk C-29
Table C-13. Summary of BMD modeling results for thyroid follicular cell adenomas or
carcinomas in male B6C3F1 mice exposed to ferf-butanol in drinking water for
2 years, high dose omitted (NTP, 1995); BMR = 5% extra risk C-32
Table C-14. Summary of BMD modeling results for renal tubule adenoma or carcinoma in
male F344 rats exposed to tert-butanol in drinking water for 2 years modeled
with administered dose units and including all dose groups (NTP, 1995); BMR
= 10% extra risk C-35
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Supplemental Information—tert-Butyl Alcohol
Table C-15. Summary of BMD modeling results for renal tubule adenoma or carcinoma in
male F344 rats exposed to ferf-butanol in drinking water for 2 years modeled
with administered dose units and excluding high-dose group (NTP, 1995); BMR
= 10% extra risk C-37
Table C-16. Summary of BMD modeling results for renal tubule adenoma or carcinoma in
male F344 rats exposed to ferf-butanol in drinking water for 2 years modeled
with administered dose units and including all dose groups; reanalyzed data
(Hard et al., 2011; NTP, 1995); BMR = 10% extra risk C-39
Table C-17. Summary of BMD modeling results for renal tubule adenoma or carcinoma in
male F344 rats exposed to ferf-butanol in drinking water for 2 years modeled
with administered dose units and excluding high-dose group; re-analyzed data
(Hard et al., 2011; NTP, 1995); BMR = 10% extra risk C-39
FIGURES
Figure B-l. Biotransformation of ferf-butanol in rats and humans B-4
Figure B-2. Example oral ingestion pattern for rats exposed via drinking water B-8
Figure B-3. Exposure-response array of other effects following oral exposure to ferf-butanol B-20
Figure B-4. Exposure-response array of other effects following inhalation exposure to ferf-
butanol B-21
Figure C-l. Plot of incidence by dose, with fitted curve for LogLogistic model for kidney
transitional epithelial hyperplasia in male F344 rats exposed to ferf-butanol in
drinking water for 2 years (NTP, 1995); BMR = 10% extra risk; dose shown in
mg/kg-d C-4
Figure C-2. Plot of incidence by dose, with fitted curve for Multistage 3° model for kidney
transitional epithelial hyperplasia in female F344 rats exposed to ferf-butanol
in drinking water for 2 years (NTP, 1995); BMR = 10% extra risk; dose shown in
mg/kg-d C-6
Figure C-3. Plot of mean response by dose, with fitted curve for Linear model with constant
variance for absolute kidney weight in male F344 rats exposed to ferf-butanol
in drinking water for 15 months (NTP, 1995); BMR = 10% rel. dev. from control
mean; dose shown in mg/kg-d C-10
Figure C-4. Plot of mean response by dose, with fitted curve for Exponential (M4) model
with constant variance for absolute kidney weight in female F344 rats exposed
to ferf-butanol in drinking water for 15 months (NTP, 1995); BMR = 10% rel.
dev. from control mean; dose shown in mg/kg-d C-13
Figure C-5. Plot of incidence by dose, with fitted curve for LogpPobit model for kidney
inflammation in female rats exposed to ferf-butanol in drinking water for 2
years (NTP, 1995); BMR = 10% extra risk; dose shown in mg/kg-d C-15
Figure C-6. Plot of mean response by concentration, with fitted curve for Hill model for
absolute kidney weight in male F344 rats exposed to ferf-butanol via inhalation
for 6 hr/d, 5d/wk for 13 weeks (NTP, 1997); BMR = 10% relative deviation from
the mean; concentration shown in mg/m3 C-19
Figure C-7. Plot of mean response by concentration, with fitted curve for Hill model for
absolute kidney weight in female F344 rats exposed to ferf-butanol via
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT—DO NOT CITE OR QUOTE
This document is a draft for review purposes only and does not constitute Agency policy.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Supplemental Information—tert-Butyl Alcohol
inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP, 1997); BMR = 10% relative
deviation from the mean; concentration shown in mg/m3 C-22
Figure C-8. Plot of mean response by concentration, with fitted curve for Power model for
absolute kidney weight in female F344 rats exposed to ferf-butanol via
inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP, 1997); BMR = 10% relative
deviation from the mean; concentration shown in mg/m3 C-22
Figure C-9. Plot of incidence by dose, with fitted curve for Multistage 3° model for thyroid
follicular cell adenomas in female B6C3F1 mice exposed to ferf-butanol in
drinking water for 2 years (NTP, 1995); BMR = 10% extra risk; dose shown in
mg/kg-d C-26
Figure C-10. Plot of incidence by dose, with fitted curve for Multistage 1° model for thyroid
follicular cell adenomas or carcinomas in male B6C3F1 mice exposed to ferf-
butanol in drinking water for 2 years (NTP, 1995); BMR = 5% extra risk; dose
shown in mg/kg-d C-29
Figure C-ll. Plot of incidence by dose, with fitted curve for Multistage 2° model for thyroid
follicular cell adenomas or carcinomas in male B6C3F1 mice exposed to ferf-
butanol in drinking water for 2 years, high dose omitted (NTP, 1995); BMR = 5%
extra risk; dose shown in mg/kg-d C-32
Figure C-12. Plot of incidence by dose, with fitted curve for Multistage 2° model for renal
tubule adenoma or carcinoma in male F344 rats exposed to ferf-butanol in
drinking water for 2 years modeled with administered dose units and including
all dose groups (NTP, 1995); BMR = 10% extra risk; dose shown in mg/kg-d C-35
Figure C-13. Plot of incidence by dose, with fitted curve for Multistage 1° model for renal
tubule adenoma or carcinoma in male F344 rats exposed to ferf-butanol in
drinking water for 2 years modeled with administered dose units and excluding
high-dose group (NTP, 1995); BMR = 10% extra risk.; dose shown in mg/kg-d C-37
Figure C-14. Plot of incidence by dose, with fitted curve for Multistage 1° model for renal
tubule adenoma or carcinoma in male F344 rats exposed to tert-butanol in
drinking water for 2 years modeled with administered dose units and excluding
high-dose group; re-analyzed data (Hard et al., 2011; NTP, 1995); BMR = 10%
extra risk; dose shown in mg/kg-d C-40
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
i ABBREVIATIONS
AIC Akaike's information criterion
ARCO ARCO Chemical Company
BMD benchmark dose
BMDL benchmark dose lower confidence limit
BMDS Benchmark Dose Software
BMDU benchmark dose upper confidence limit
BMR benchmark response
BW body weight
CFR Code of Federal Regulations
CHO Chinese hamster ovary
CYP450 cytochrome P450
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
EDTA ethylenediaminetetraacetic acid
EPA U.S. Environmental Protection Agency
ETBE ethyl tert-butyl ether
HBA 2-hydroxyisobutyrate
HL human leukemia
IC50 half-maximal inhibitory concentration
i.p. intraperitoneal
i.v. intravenous
MFO mixed function oxidase
MPD 2-methyl-l, 2-propanediol
MTBE methyl tert-butyl ether
NADPH nicotinamide adenine dinucleotide
phosphate
NTP National Toxicology Program
•OH hydroxyl radical
PBPK physiologically based pharmacokinetic
POD point of departure
SD standard deviation
TWA time-weighted average
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1
2 APPENDIX A. ASSESSMENTS BY OTHER NATIONAL
3 AND INTERNATIONAL HEALTH AGENCIES
4 Table A-l. Health assessments and regulatory limits by other national and
5 international health agencies
Organization
Toxicity value
National Institute of Occupational
Safetv and Health (NIOSH, 2007)
Recommended Exposure Limit - 100 ppm (300 mg/m3) time-weighted
average (TWA) for up to a 10-hour workday and a 40-hour workweek.
Occupational Safety and Health
(OSHA, 2006)
Permissible Exposure Limit for general industry - 100 ppm (300 mg/m3) TWA
for an 8-hour workday.
Food and Drug Administration
(FDA, 2011a, b)
tert-Butyl alcohol: Indirect food additive that may be safely used in surface
lubricants employed in the manufacture of metallic articles that contact
food, subject to the provisions of this section (21 Code of Federal
Regulations [CFR] 178.3910); substance may be used as a defoaming agent
(21 CFR 176.200).
6
This document is a draft for review purposes only and does not constitute Agency policy.
A-l DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Supplemental Information—tert-Butyl Alcohol
APPENDIX B. INFORMATION IN SUPPORT OF
HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS
B.l. TOXICOKINETICS
Little information is available on the absorption, distribution, metabolism, or excretion of
tert-butyl alcohol (tert-butanol) in humans. The studies identified for this Toxicological Assessment
were conducted in conjunction with methyl tert-butyl ether (MTBE) or ethyl tert-butyl ether
(ETBE), as tert-butanol is a metabolite of both compounds. Several studies examining some aspect
of the toxicokinetic behavior of tert-butanol in animals have been identified. Many were carried out
in conjunction with other specific endpoints (e.g., developmental). ARCO f 19831 determined no
differences in the pharmacokinetics of tert-butanol following either oral (i.e., gavage) or inhalation
exposure. Although some information is available for both oral and inhalation exposures, many
studies administered tert-butanol via intraperitoneal (i.p.) or intravenous (i.v.) injection. Although
these studies do not inform the absorption of tert-butanol, they can provide information on its
distribution, metabolism, and excretion.
B.l.l. Absorption
Toxicity data on tert-butanol submitted by industry to the U.S. Environmental Protection
Agency (EPA) under Section 8(e) of the Toxic Substances Control Act and other reporting
requirements indicate that tert-butanol is rapidly absorbed after oral administration. Very little of
the administered dose was excreted in the feces of rats, indicating 99% of the compound was
absorbed. Comparable blood levels of tert-butanol and its metabolites have been observed after
acute oral (350 mg/kg) or inhalation (6,060 mg/m3 for 6 hours) exposure in male Sprague-Dawley
rats (ARCO. 1983): the absorption rate after inhalation exposure could not be determined, however,
because the blood was saturated with radioactivity after 6 hours of exposure to 6,060 mg/m3. In
another study (Faulkner etal.. 1989). blood concentrations indicated that absorption was complete
at 1.5 hours following the last of six oral gavage doses of 10.5 mmoles tert-butanol/kg (twice daily)
in female C57BL/6J mice. There was an apparent zero-order decline in tert-butanol concentration
for most of the elimination phase, and no differences in absorption or elimination rates were
observed between mice on a repeated dosing regimen and mice administered equivalent volumes
of tap water every 12 hours before administration of a single dose of 10.5 mmoles tert-butanol/kg.
The study therefore concluded that previous exposures did not affect the absorption or elimination
of tert-butanol (Faulkner et al.. 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
B-l DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
B.1,2. Distribution
The available animal data suggest that tert-butanol is distributed throughout the body
following oral, inhalation, and i.v. exposures fPoet etal.. 1997: Faulkner et al.. 1989: ARCO. 19831.
Nihlen et al. f 19951 calculated partition coefficients for tert-butanol using blood from human
volunteers and available information about the relative content of water and fat in each tissue. The
calculated tissue:blood partition coefficients for tert-butanol were slightly above 1 (from 1.02 to
1.06) for most tissues, except for fat:blood, which had a partition coefficient of 0.646. The same
study evaluated the partition coefficients of three oxygenated ethers, including MTBE and ETBE,
which are metabolized to tert-butanol (see Section B.1.4). The study concluded that, although tert-
butanol preferentially distributes in body water, the ethers distribute uniformly throughout the
body with a preference for fatty tissues fNihlen etal.. 19951.
In a study aimed at determining whether tert-butanol (or metabolites) can bind to
a2u-globulin, Williams and Borghoff (20011 exposed F-344 rats to a single gavage dose of 500
mg/kg 14C-tert-butanol and evaluated tissue levels at 12 hours. They found the radiolabel in three
tissues (kidney, liver, and blood) in both sexes, but male rats retained more of the tert-butanol
equivalents than females fWilliams and Borghoff. 20011. Radioactivity was found in the low-
molecular-weight protein fraction isolated from the kidney cytosol in male rats but not in female
rats, indicating that tert-butanol or one of its metabolites was bound to a2U-globulin. Further
analysis determined that tert-butanol, and not its metabolite acetone, was bound. Most tert-butanol
in the kidney cytosol was eluted as the free compound in both males and females, but a small
amount was associated with the high-molecular-weight protein fraction in both males and females.
In another study on a2U-globulin nephropathy, Borghoff et al. (20011 found similar results after F-
344 rats were exposed to 0, 250, 450, or 1750 ppm tert-butanol by inhalation for 8 consecutive
days (with tissue levels measured at 2, 4, 6, 8, and 16 hours postexposure). Male rat tert-butanol
kidney-to-blood ratios were significantly elevated over ratios in females at all dose levels and
exposure durations. Although the female tert-butanol kidney-to-blood ratio remained similar with
both duration and concentration, the male tert-butanol kidney-to-blood ratio increased with
duration. The liver-to-blood ratios were similar regardless of exposure duration, concentration, or
sex. Both of these studies indicate distribution of tert-butanol to the liver and kidney with kidney
retention of tert-butanol in the male rat.
B.1.3. Metabolism
A general metabolic scheme for tert-butanol, illustrating the biotransformation in rats and
humans, is shown in Figure B-l. Urinary metabolites of tert-butanol in a human male volunteer who
ingested a gelatin capsule containing 5 mg/kg [13C]-tert-butanol were reported to be 2-methyl-l,2-
propanediol (MPD) and 2-hydroxyisobutyrate (HBA) fBernauer et al.. 19981. Minor metabolites of
unconjugated tert-butanol, tert-butanol glucuronides, and traces of the sulfate conjugate also were
detected. The study was approved by an ethical review board, but no information regarding
informed consent was reported. In the same study, HBA, MPD, and tert-butanol sulfate were
This document is a draft for review purposes only and does not constitute Agency policy.
B-2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Supplemental Information—tert-Butyl Alcohol
identified as major metabolites in rats, while acetone, tert-butanol, and tert-butanol glucuronides
were identified as minor metabolites fBernauer etal.. 19981. Baker etal. (1982) found that tert-
butanol was a source of acetone, but acetone production might have been stimulated from other
sources.
No studies identified specific enzymes responsible for biotransforming tert-butanol. Using a
purified enzyme from Sprague-Dawley rats or whole-liver cytosol from Wistar rats, alcohol
dehydrogenase had negligible or no activity toward tert- butanol fVidela etal.. 1982: Arslanian etal..
19711. Other in vitro studies have implicated the liver microsomal mixed function oxidase (MFO)
system, namely cytochrome P450 (CYP450) fCederbaum etal.. 1983: Cederbaum and Cohen. 19801.
In the 1983 study, incubation of tert-butanol at 35 mM with Sprague-Dawley rat liver microsomes
and a nicotinamide adenine dinucleotide phosphate (NADPH)-generating system resulted in
formaldehyde production at a rate of approximately 25 nmoles/mg protein/30 min. According to
study authors, the amount of formaldehyde generated by tert-butanol was approximately 30% of
the amount of formaldehyde formed during the metabolism of 10 mM aminopyrene in a similar
microsomal system. The rate of formaldehyde generation from tert-butanol increased to about
90 nmol/mg protein/30 min upon addition of azide, which inhibits catalase and thereby prevents
the decomposition of hydrogen peroxide (H2O2). In other experiments in the same study,
formaldehyde formation was greatly reduced when H2O2 was included but NADPH was absent or
when the microsomes were boiled prior to incubation. Additionally, the rate of formaldehyde
formation in the microsomal oxidizing system depended on the concentration of tert-butanol, with
apparent Km and Vmax values of 30 mM and 5.5 nmol/min/mg protein, respectively. The study
authors concluded that tert-butanol is metabolized to formaldehyde by a mechanism involving
oxidation of NADPH, microsomal electron flow, and the generation of hydroxyl-radical (-OH) from
H2O2, possibly by a Fenton-type or a Haber-Weiss iron-catalyzed reaction involving CYP450, which
might serve as the iron chelate fCederbaum and Cohen. 19801.
This document is a draft for review purposes only and does not constitute Agency policy.
B-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Supplemental Information—tert-Butyl Alcohol
CH,
glucuronide-O-
-CH,
CH,
t-butyl glucuronide
rats, humans
CH,
HO-
-CH,
CH3
t-butanol
CYP450
I
rats,
humans
OH
A
ch3 oh
2-methyl-1,2-propanediol
r^
-OH
CH,
rats
V° CH,
.S
O^ \
-CH,
CH,
t-butyl sulfate
HO.
HO-
-CH,
[O]
CH,
2-hydroxyisobutyric acid
formaldehyde
CH,
H3C
acetone
Sources: NSF International (2003), ATSDR (1996), Bernauer et al. (1998), Ambers et al. (1999), and
Cederbaum and Cohen (1980).
Figure B-l. Biotransformation of tert-butanol in rats and humans.
In a follow-up study, tert-butanol was oxidized to formaldehyde and acetone by various
systems known to generate -OH radical, including rat liver microsomes or other nonmicrosomal
•OH-generating systems fCederbaum etal.. 19831. The nonmicrosomal tests included two chemical
systems: (1) the iron-catalyzed oxidation of ascorbic acid (ascorbate-Fe-EDTA
[ethylenediaminetetraacetic acid]) and (2) the Fenton system of chelated ferrous iron and H2O2. In
both Fenton-type systems, H2O2 served as a precursor for -OH. Additionally, a Haber-Weiss
enzymatic system involving xanthine oxidation by xanthine oxidase in the presence of Fe-EDTA was
used. In this system, -OH is thought to be produced by the interaction of H2O2 and superoxide (02-~).
Further experiments demonstrated the involvement of -OH in either the ascorbate-Fe-EDTA or the
xanthine oxidation systems based on inhibition of formaldehyde and acetone production from tert-
butanol when -OH-scavenging agents (e.g., benzoate, mannitol) were added. Some experiments in
this study of the oxidation of tert-butanol by the microsomal metabolizing system of the liver were
similar to those in the previous study (Cederbaum and Cohen. 19801 except that, in addition to
formaldehyde, acetone formation was measured. Again, these experiments showed the dependence
of the microsomal metabolizing system on an NADPH-generating system and the ability of H2O2 to
enhance, but not replace, the NADPH-generating system. Addition of chelated iron (Fe-EDTA)
This document is a draft for review purposes only and does not constitute Agency policy.
B-4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information—tert-Butyl Alcohol
boosted the microsomal production of formaldehyde and acetone, while -OH-scavenging agents
inhibited their production. The study authors noted that neither Fe-EDTA nor -OH-scavenging
agents is known to affect the CYP450-catalyzed oxidation of typical MFO substrates such as
aminopyrene or aniline. The study also showed that known CYP450 inhibitors, such as metyrapone
or SKF-525A, inhibited the production of formaldehyde from aminopyrene but not from tert-
butanol. Finally, typical inducers of CYP450 and its MFO metabolizing activities, such as
phenobarbital or 3-methylcholanthrene, had no effect on microsomal metabolism of tert-butanol to
formaldehyde and acetone. According to the study authors, the oxidation of tert-butanol appears to
be mediated by OH (possibly via H2O2), which can be produced by any of the tested systems by a
Fenton-type reaction as follows:
H2O2 + Fe2+-chelate -> -OH + OH- + Fe3+-chelate
According to this reaction, reduction of ferric iron (Fe3+) to ferrous iron (Fe2+) is required
for continuous activity. The study authors concluded that the nature of the iron and the pathway of
iron reduction within the microsomes remain to be elucidated even though an NADPH-dependent
electron transfer or 02-_ might be involved fCederbaum etal.. 19831.
B.1.4. Excretion
Human data on the excretion of tert-butanol derives from studies of MTBE and ETBE
(Nihlen etal.. 1998a: 1998b). Eight or 10 male human volunteers were exposed to 5, 25, or 50 ppm
MTBE (18.0, 90.1, and 757 mg/m3) or ETBE (20.9,104, and 210 mg/m3) by inhalation during 2
hours of light exercise. The half-life of tert-butanol in urine following MTBE exposure was 8.1 ± 2.0
hours (average of the 25- and 50-ppm MTBE doses); the half-life of tert-butanol in urine following
ETBE exposure was 7.9 ± 2.7 hours (average of 25- and 50-ppm ETBE doses). In both studies, the
urinary excretion of tert-butanol was less than 1% of the uptake or absorption of MTBE or ETBE.
The renal clearance rate of tert-butanol was 0.67 ± 0.11 mL/hr-kg with MTBE exposure (average of
25- and 50-ppm MTBE doses); the renal clearance rate was 0.80 ± 0.34 mL/hr-kg with ETBE
exposure (average of 25- and 50-ppm ETBE doses).
Amberg etal. (2000) exposed six volunteers (three males and three females, 28 ± 2 years
old) to 18.8 and 170 mg/m3 ETBE. Each exposure lasted 4 hours, and the two concentrations were
administered to the same volunteers 4 weeks apart Urine was collected at 6-hour intervals for
72 hours following exposure, tert-Butanol and two metabolites of tert-butanol, HBA and MPD, also
were identified in urine. At an ETBE level of 170 mg/m3, tert-butanol displayed a half-life of
9.8 ± 1.4 hours. At the low-exposure ETBE concentration, the tert-butanol half-life was 8.2 ±
2.2 hours. The predominant urinary metabolite identified was HBA, excreted in urine at 5-10 times
the amount of MPD and 12-18 times the amount of tert-butanol (note: urine samples had been
treated with acid before analysis to cleave conjugates). HBA in urine showed a broad maximum at
12-30 hours after exposure to both concentrations, with a slow decline thereafter. MPD in urine
This document is a draft for review purposes only and does not constitute Agency policy.
B-5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information—tert-Butyl Alcohol
peaked at 12 and 18 hours after exposure to 170 and 18.8 mg/m3 ETBE, respectively, while tert-
butanol peaked at 6 hours after exposure to both concentrations.
Amberg etal. f20001 exposed F344 NH rats to 18.8 and 170 mg/m3 ETBE. Urine was
collected for 72 hours following exposure. Similar to humans, rats excreted mostly HBA in urine,
followed by MPD and tert-butanol. The half-life for tert-butanol in rat urine was 4.6 ± 1.4 hours at
ETBE levels of 170 mg/m3, but half-life could not be calculated at the ETBE concentration of
18.8 mg/m3. Corresponding half-lives were 2.6 ± 0.5 and 4.0 ± 0.9 hours for MPD and 3.0 ± 1.0 and
4.7 ± 2.6 hours for HBA. In Sprague-Dawley rats treated with radiolabeled tert-butanol by gavage at
1, 30, or 500 mg/kg, a generally constant fraction of the administered radioactivity (23-33%) was
recovered in the urine at 24 hours postdosing. Only 9% of a 1500-mg/kg administered dose was
recovered in urine, however, suggesting that the urinary route of elimination is saturated following
this dose fARCO. 19831. Among all tested doses, most of the urinary radiolabel was attributed to a
polar fraction that was not characterized, while only 0.3-5.5% of the administered dose was
considered tert-butanol. The saturation in urinary elimination of radioactivity with the increased
dose was considered a manifestation of saturated metabolic capacity; however, no further
information was provided on the fate or balance of the administered radiolabel at any of the tested
tert-butanol doses fARCO. 19831.
Borghoff and Asgharian f 19961 evaluated the disposition of 14C radiolabel in F344 rats and
CD-I mice after nose-only inhalation exposure to 500,1750, or 5,000 ppm 14C-ETBE for six hours.
Besides recovery of total radioactivity in urine, feces, and expired air, air and urine samples were
analyzed for ETBE and tert-butanol. Urine samples were also analyzed for tert-butanol metabolites
HBA and MPD, and 14C02 was measured in exhaled air. Results were also obtained in rats after 13
days of exposure to 500 or 5,000 ppm ETBE. Total ETBE equivalents in exhaled air and excreted
urine were found to increase linearly with exposure level, with over 90% eliminated by 48 hours
(with the majority of exhalation occurring by 8 hours postexposure). Elimination shifted from being
primarily in the urine at 500 ppm to occurring primarily by exhalation at 5,000 ppm in naive rats,
indicating a saturation of metabolism of ETBE to tert-butanol; this shift was greater in female
rats than males. In rats pre-exposed to 5,000 ppm ETBE for 13 days, most of the excretion was in
urine even at 5,000 ppm. For rats pre-exposed to 500 ppm ETBE, there also was a shift from
exhalation to urinary excretion compared with naive rats, but to a lesser degree than that elicited
by the 5,000-ppm pre-exposure group.
Like rats, the fraction of radiolabel in exhaled volatiles increased with exposure level in
mice while the fraction excreted in urine decreased. The exhalation pattern observed in rats
showed levels of ETBE falling approximately 90% in the first 8 hours postexposure, while levels of
TBA exhaled rose between 0 and 3 hours postexposure and then fell more slowly between 3 and 16
hours, particularly at 5,000 ppm ETBE. The increase in tert-butanol between 0 and 3 hours
postexposure can be explained by the continued metabolism of ETBE during that period. The
This document is a draft for review purposes only and does not constitute Agency policy.
B-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Supplemental Information—tert-Butyl Alcohol
slower decline after 3 hours likely results from a generally slower clearance of tert-butanol, which
is saturated by the higher ETBE exposure levels.
B.1.5. Physiologically Based Pharmacokinetic Models
Three physiologically based pharmacokinetic (PBPK) models have been developed
specifically for administration of tert-butanol in rats Leavens and Borghoff (2009): Salazar et al.
(2015). and Borghoff et al. (2016): other models have incorporated tert-butanol as a submodel
following MTBE administration. In Leavens and Borghoff (2009). tert-butanol is incorporated as a
metabolite of MTBE; in Salazar etal. (2015) and Borghoff et al. (2016). it is incorporated as a
metabolite of ETBE. In all three models, inhalation and oral exposure to tert-butanol can be
simulated in rats. A detailed summary of these toxicokinetic models is provided in a separate report
evaluating the PK/PBPK modeling of ETBE and tert-butanol (U.S. EPA. 2017).
The PBPK model described in Borghoff et al. (2016). with parameters modified as described
by U.S. EPA (2017). was applied to conduct oral-to-inhalation route extrapolation based on an
equivalent internal dose (the average concentration of tert-butanol in the blood). The time to reach
a consistent periodic pattern of tert-butanol blood concentrations ("periodicity"), given the
drinking water ingestion pattern described below, was much shorter than the duration of the oral
bioassay studies. To allow for possible metabolic induction, computational scripts used a simulated
time of 7 weeks, although periodicity was achieved in only a few days without metabolic induction.
The average blood concentration was calculated over the last week of the simulation and was
considered representative of the bioassays. To calculate steady state values for continuous
inhalation exposure, the simulations were run until the blood concentration had a <1% change
between consecutive days. The continuous inhalation exposure equivalent to a given oral exposure
was then selected by identifying the inhalation concentration for which the final (steady-state)
blood concentration of tert-butanol matched the average concentration from water ingestion, as
described above.
For simulating exposure to drinking water, the consumption was modeled as episodic,
based on the pattern of drinking observed in rats (Spiteri. 19821. In particular, rats were assumed
to ingest water in pulses or "bouts," which were treated as continuous ingestion, interspersed with
periods ofno ingestion. During the active dark period (12 hours/day), it was assumed that 80% of
total daily ingestion occurs (45-minute bouts with alternating 45-minute periods of other activity).
During the relatively inactive light period (12 hours/day), it was assumed that the remaining 20%
of daily ingestion occurs; during this time, bouts were assumed to last 30 minutes with 2.5 hours in
between. This resulting pattern of drinking water ingestion is thought to be more realistic than
assuming continuous 24 hours/day ingestion (see Figure B-2).
This document is a draft for review purposes only and does not constitute Agency policy.
B-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Supplemental Information—tert-Butyl Alcohol
15 I
12 -
9 -
6 -
3 -
o IIIII1111IIII111II—U—U—u—IIIII1111IIII111II—u—U—u—
0 6 12 18 24 30 36 42 48
'Time (h)
Figure B-2. Example oral ingestion pattern for rats exposed via drinking
water.
PBPK modeling was also used to evaluate a variety of internal dose metrics (daily average
TBA blood concentration, daily amount of TBA metabolized in liver, daily average of ETBE blood
concentration, and daily amount of ETBE metabolized in liver) to assess their correlation with
different endpoints following exposure to ETBE or TBA fSalazar etal.. 20151. Administering ETBE
either orally or via inhalation achieved similar or higher levels of TBA blood concentrations or TBA
metabolic rates as those induced by direct TBA administration. Altogether, the PBPK model-based
analysis by Salazar etal. (2015) [which applied a model structurally similar to Borghoff et al.
£2016}] indicates that kidney weight, urothelial hyperplasia, and chronic progressive nephropathy
(CPN) yield consistent dose-response relationships using TBA blood concentration as the dose
metric for both ETBE and TBA studies. For kidney and liver tumors, however, a consistent dose-
response pattern was not obtained using any dose metric. These data are consistent with TBA
mediating the noncancer kidney effects following ETBE administration, but additional factors
besides internal dose are necessary to explain the induction of liver and kidney tumors.
B.1.6. PBPK Model Code
The PBPK acslX model code is available electronically through EPA's Health and
Environmental Research Online (HERO) database. All model files may be downloaded in a zipped
workspace from HERO (U.S. EPA. 2016).
cm
E
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information—tert-Butyl Alcohol
B.2. OTHER PERTINENT TOXICITY INFORMATION
B.2.1. Other Toxicological Effects
B.2.1.1. Synthesis of Other Effects
Effects other than those related to kidney, thyroid, reproductive, developmental, and
neurodevelopmental effects were observed in some of the available rodent studies. These include
liver and urinary bladder effects. As previously mentioned in the Study Selection section of the
Toxicological Review, all studies discussed employed inhalation, oral gavage, or drinking water
exposures for >30 days. Studies are arranged in evidence tables by effect, species, duration, and
design. The design, conduct, and reporting of each study was reviewed, and each study was
considered adequate to provide information pertinent to this assessment.
Central nervous system effects similar to those of ethanol (i.e., animals appearing
intoxicated and having withdrawal symptoms after cessation of oral or inhalation exposure) were
observed with tert-butanol. Severity of central nervous system symptoms increased with dose and
duration of exposure. Study quality and utility concerns associated with these studies (e.g.,
inappropriate exposure durations, lack of data reporting, small number of animals per treatment
group) (Grant and Samson. 1981: Snell. 1980: Thurman et al.. 1980: McComb and Goldstein. 1979a.
b; Wood and Lavertv. 1979). preclude an understanding of potential neurotoxicity following tert-
butanol exposure; therefore, central nervous system studies are not discussed further.
Exposure-response arrays of liver and urinary bladder effects are provided in Figure B-3
and Figure B-4 for oral and inhalation studies, respectively.
Kidney effects
Absolute and relative kidney weight numerical data are presented in Table B-l.
Liver effects
Liver weight and body weight were demonstrated to be proportional, and liver weight
normalized to body weight was concluded optimal for data analysis fBailevetal.. 20041: thus, only
relative liver weight is presented and considered in the determination of hazard. Although some
rodent studies observed liver effects (organ weight changes and histopathologic lesions), the effects
were not consistent across the database. Increases in relative liver weight with tert-butanol
exposure were observed, but the results pertaining to histopathologic changes were inconsistent
(Table B-2 and Table B-3). The NTP (1995) oral subchronic and chronic studies did not observe
treatment-related effects on liver histopathology in either sex of F344 rats. In a 10-week study in
Wistar rats, several liver lesions (including necrosis) and increased liver glycogen were observed in
male rats (no females were included in the study) with the only dose used fAcharva etal.. 1997:
Acharva etal.. 1995). The study provided no incidence or severity data. The dose used in this rat
study was in the range of the lower doses used in the NTP (1995) subchronic rat study. An
increased incidence of fatty liver was observed in the male mice of the highest dose group in the 2-
This document is a draft for review purposes only and does not constitute Agency policy.
B-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Supplemental Information—tert-Butyl Alcohol
year mouse bioassay, but no histopathological changes were seen in the subchronic mouse study
fNTP. 19951. No treatment-related effects in liver histopathology were observed in rats or mice of
the NTP f 19971 subchronic inhalation study.
Urinary bladder effects
Subchronic studies reported effects in the urinary bladder (Table B-4), although the chronic
studies indicated little progression in incidence with increased exposure. Transitional epithelial
hyperplasia of the urinary bladder was observed in male rats and male mice after 13 weeks of
exposure at doses of 3,610 mg/kg-day (male rats) and >3,940 mg/kg-day (male mice). In rats, the
increase in transitional epithelial hyperplasia of the urinary bladder was not observed in the 2-year
study. Male mice exposed at the high dose (2,070 mg/kg-day) for 2 years exhibited minimal
transitional epithelial hyperplasia of the urinary bladder. Neither female rats nor female mice
showed increased incidences of this lesion. Both sexes of mice demonstrated incidence of minimal
to mild inflammation in the urinary bladder after both subchronic and chronic exposures, with a
greater incidence in males compared with females.
B.2.1.2. Mechanistic Evidence
No mechanistic evidence is available for liver and urinary bladder effects.
B.2.1.3. Summary of Other Toxicity Data
Based on lack of consistency and lack of progression, the available evidence does not
support liver and urinary bladder effects, respectively, as potential human hazards of tert-butanol
exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
B-10 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Table B-l. Changes in kidney weight in animals following exposure to
tert- butanol
Reference and study design
Results
Kidney weight (percent change as compared to control)
Lvondell Chemical Co. (2004)
Males
Sprague-Dawley rat;
12/sex/treatment
Gavage 0, 64,160, 400, or
Dose
(mg/kg-d)
Left absolute Left relative Right absolute
weight weight weight
Right relative
weight
1,000 mg/kg-d
0
0
0
0
0
Males: 9 weeks beginning
4 weeks prior to mating
64
6
8
6
8
Females: = 10 weeks (4 weeks
160
9
14*
6
11*
prior to mating through PND21)
400
12*
14*
14*
17*
1,000
18*
28*
20*
31*
Females
Dose
Left absolute Left relative Right absolute
Right relative
(mg/kg-d)
weight
weight
weight
weight
0
0
0
0
0
64
-1
-2
2
0
160
0
0
1
0
400
3
2
4
2
1,000
4
0
7
2
NTP (1995)
Males
Females
F344/N rat; 10/sex/treatment
Drinking water 0, 2.5, 5,10, 20,
40 mg/mL
Dose
Absolute
Relative Dose
Absolute
Relative
(mg/kg-d)
weight
weight (mg/kg-d)
weight
weight
M: 0, 230, 490, 840, 1,520,
0
0
0
0
0
0
3,610a mg/kg-d
F: 0, 290, 590, 850, 1,560,
230
12*
19*
290
19*
17*
3,620a mg/kg-d
490
17*
26*
590
16*
15*
13 weeks
840
16*
32*
850
29*
28*
1,520
26*
54*
1,560
39*
40*
3,610
All dead
All dead 3,620
36*
81*
This document is a draft for review purposes only and does not constitute Agency policy.
B-ll DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Reference and study design
Results
NTP (1995)
Males
Females
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20,
40 mg/mL)
Dose
Absolute
Relative
Dose
Absolute
Relative
(mg/kg-d)
weight
weight
(mg/kg-d)
weight
weight
M: 0, 350, 640, 1,590, 3,940,
0
0
0
0
0
0
8,210a mg/kg-d
F: 0, 500, 820, 1,660, 6,430,
350
1
1
500
0
-3
11,620a mg/kg-d
640
3
2
820
-3
-1
13 weeks
1,590
2
8
1,660
1
0
3,940
6
22*
6,430
6
15*
8,210
0
48*
11,620
12*
35*
NTP (1995)
Males
Females
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at
15 months)
Dose
Absolute
Relative
Dose
Absolute
Relative
(mg/kg-d)
weight
weight
(mg/kg-d)
weight
weight
Drinking water (0,1.25, 2.5, 5, or
0
0
0
0
0
0
10 mg/mL)
M: 0, 90, 200, or 420a mg/kg-d
90
4
8
180
8*
14*
F: 0,180, 330, or 650a mg/kg-d
200
11
15*
330
18*
21*
2 years
420
7
20*
650
22*
42*
Only rats sacrificed at 15 months were evaluated for organ weights.
Males
Females
Concentration
Absolute
Relative
Absolute
Relative
(mg/m3)
weight
weight
weight
weight
0
0
0
0
0
406
1
1
-4
-1
824
-2
-1
0
1
1,643
3
2
4
4
3,273
11*
8*
2
2
6,368
9.8*
9*
4
9*
NTP (1997)
F344/N rat; 10/sex/treatment
Inhalation analytical
concentration: 0,134, 272, 542,
1,080, or 2,101 ppm (0, 406, 824,
1,643, 3,273 or 6,368 mg/m3)
(dynamic whole-body chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist
Ultrasonic spray nozzle
nebulizer), analytical
concentration and method were
reported
Right kidney weights measured
This document is a draft for review purposes only and does not constitute Agency policy.
B-12 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Reference and study design
Results
NTP (1997)
Males
Females
B6C3Fi mouse; 10/sex/treatment
Inhalation analytical
concentration: 0,134, 272, 542,
Concentration
Absolute
Relative
Absolute
Relative
(mg/m3)
weight
weight
weight
weight
1,080, or 2,101 ppm (0, 406, 824,
0
0
0
0
0
1,643, 3,273 or 6,368 mg/m3)
(dynamic whole-body chamber)
406
-6
-4
1
-3
6 hr/d, 5 d/wk
824
-1
3
5
9
13 weeks
1,643
Generation method (Sonimist
4
3
1
-2
Ultrasonic spray nozzle
3,273
-10
-3
0
7
nebulizer), analytical
concentration and method were
6,368
3
6
3
15*
reported
Right kidney weights measured
1
2 aThe high-dose group had an increase in mortality.
3 * Statistically significant p < 0.05 as determined by the study authors.
4 Percentage change compared to control = (treated value - control value) -f control value x 100.
5 Conversions from drinking water concentrations to mg/kg-d performed by study authors.
6 Conversion from ppm to mg/m3 is 1 ppm = 3.031 mg/m3.
7
This document is a draft for review purposes only and does not constitute Agency policy.
B-13 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Table B-2. Changes in liver weight in animals following exposure to
tert- butanol
Reference and study design
Resu Its
Acharva et al. (1995)
No significant treatment-related effects (results were only provided in a figure)
Wistar rat; 5-6 males/treatment
Drinking water (0 or 0.5%), 0 or
575 mg/kg-d
10 weeks
Lvondell Chemical Co. (2004)
Percent change compared to control:
Sprague-Dawley rat; 12/sex/treatment
Males
Females
Gavage 0, 64,160, 400, or 1,000 mg/kg-d
Males: 9 weeks beginning 4 weeks prior to
Dose
Absolute
Relative
Dose
Absolute
mating
Females: 4 weeks prior to mating through
(mg/kg-d)
weight
weight
(mg/kg-d)
weight
Relative weight
PND21
0
-
-
0
-
-
64
-1
0
64
-4
-4
160
-3
1
160
-7
-5
400
-2
-1
400
2
1
1,000
8
16*
1,000
8
3
NTP (1995)
Percent change compared to control:
F344/N rat; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, or
Males
Females
40 mg/mL)
Dose
Absolute
Relative
Dose
Absolute
Relative
M: 0, 230, 490, 840, 1,520, 3,610a mg/kg-d
F: 0, 290, 590, 850, 1,560, 3,620a mg/kg-d
(mg/kg-d)
weight
weight
(mg/kg-d)
weight
weight
13 weeks
0
-
-
0
-
-
230
-2
4
290
11*
9*
490
1
8*
590
10*
9*
840
5
20*
850
12*
ii*
1,520
8
31*
1,560
15*
16*
3,610
All dead
All dead
3,620
9*
41*
NTP (1995)
Percent change compared to control:
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, or
Males
Females
40 mg/mL)
Dose
Absolute
Relative
Dose
Absolute
Relative
M: 0, 350, 640, 1,590, 3,940,
8,210a mg/kg-d
(mg/kg-d)
weight
weight
(mg/kg-d)
weight
weight
F: 0, 500, 820, 1,660, 6,430,
0
-
-
0
-
-
ll,620a mg/kg-d
13 weeks
350
2
3
500
-1
-4
640
-1
-2
820
-5
-3
1,590
-1
5
1,660
-8
3,940
0
14*
6,430
-2
6
8,210
-16
22*
11,620
-6
13*
This document is a draft for review purposes only and does not constitute Agency policy.
B-14 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
NTP (1995)
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at 15 months)
Drinking water (0,1.25, 2.5, 5 or 10 mg/mL)
M: 0, 90, 200, or 420a mg/kg-d
F: 0,180, 330, or 650a mg/kg-d
2 years
Percent change compared to control:
Males Females
Dose Absolute Relative Dose Absolute Relative
(mg/kg-d) weight weight (mg/kg-d) weight weight
0 0
90 2 7 180 -14* -8
200 8 11 330 -3 -1
420 1 14* 650 -6 9*
Only animals sacrificed at 15 months were evaluated for organ weights. Organ
weights were not measured in the 2-year mouse study
NTP (1997)
F344/N rat; 10/sex/treatment
Inhalation analytical concentration: 0,134,
272, 542, 1,080, or 2,101 ppm (0, 406, 824,
1,643, 3,273 or 6,368 mg/m3) (dynamic
whole body chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist Ultrasonic
spray nozzle nebulizer), analytical
concentration and method were reported
Percent change compared to control:
Males Females
Concentration Absolute Relative Absolute Relative
(mg/m3) weight weight weight weight
0 ....
406 -8 -8 0 3
824 -2 -1 0 0
1,643 1 -1 3 2
3,273 10 7 9 9*
6,368 5 5 4 8*
NTP (1997)
B6C3Fi mouse; 10/sex/treatment
Inhalation analytical concentration: 0,134,
272, 542, 1,080, or 2,101 ppm (0, 406, 824,
1,643, 3,273 or 6,368 mg/m3) (dynamic
whole body chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist Ultrasonic
spray nozzle nebulizer), analytical
concentration and method were reported
Percent change compared to control:
Males Females
Concentration Absolute Relative Absolute Relative
(mg/m3) weight weight weight weight
0 ....
406 -1 0 1 -4
824 4 9 1 5
1,643 7 5 5 1
3,273 -8 -2 2 9*
6,368 5 7 8 21*
aThe high-dose group had an increase in mortality.
* Statistically significant p < 0.05 as determined by study authors.
Conversions from drinking water concentrations to mg/kg-d performed by study authors.
Conversion from ppm to mg/m3 is 1 ppm = 3.031 mg/m3.
Percentage change compared to control = (treated value - control value) 4- control value x 100.
This document is a draft for review purposes only and does not constitute Agency policy.
B-15 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Table B-3. Changes in liver histopathology in animals following exposure to
tert- butanol
Reference and study design
Resu Its
Acharva et al. (1997)
Acharva et al. (1995)
Wistar rat; 5-6 males/treatment
Drinking water (0, 0.5%), 0, 575 mg/kg-d
10 weeks
'T* liver glycogen (~ 7 fold)*
^incidence of centrilobular necrosis, vacuolation of hepatocytes, loss of
hepatocyte architecture, peripheral proliferation, and lymphocyte
infiltration (incidences and results of statistical tests not reported)
NTP (1995)
F344/N rat; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, or 40 mg/mL)
M: 0, 230, 490, 840, 1,520, 3,610a mg/kg-d
F: 0, 290, 590, 850, 1,560, 3,620a mg/kg-d
13 weeks
No treatment-related effects observed.
NTP (1995)
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, 40 mg/mL)
M: 0, 350, 640, 1,590, 3,940, 8,210a mg/kg-d
F: 0, 500, 820, 1,660, 6,430, ll,620a mg/kg-d
13 weeks
No treatment-related effects observed.
NTP (1995)
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at 15 months)
Drinking water (0,1.25, 2.5, 5,10 mg/mL)
M: 0, 90, 200, or 420a mg/kg-d
F: 0,180, 330, or 650a mg/kg-d
2 years
No treatment-related effects observed.
NTP (1995)
B6C3Fi mouse; 60/sex/treatment
Drinking water (0, 5,10, 20 mg/mL)
M: 0, 540,1,040, or 2,070a mg/kg-d
F: 0, 510,1,020, or 2,110 mg/kg-d
2 years
Males
Dose
(mg/kg-d)
0
540
1,040
2,070
Incidence of fatty
change
12/59
5/60
8/59
29/59*
Females
Dose
(mg/kg-d)
0
510
1,020
2,110
Incidence of fatty
change
11/60
8/60
8/60
6/60
NTP (1997)
F344/N rat; 10/sex/treatment
Inhalation analytical concentration: 0,134,
272, 542, 1,080, or 2,101 ppm (0, 406, 824,
1,643, 3,273 or 6,368 mg/m3) (dynamic whole
body chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist Ultrasonic spray
nozzle nebulizer), analytical concentration and
method were reported
No treatment-related effects observed in the high dose group (only
treatment group with liver endpoints evaluated).
This document is a draft for review purposes only and does not constitute Agency policy.
B-16 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Reference and study design
Resu Its
NTP (1997)
B6C3Fi mouse; 10/sex/treatment
Inhalation analytical concentration: 0,134,
272, 542, 1,080, or 2,101 ppm (0, 406, 824,
1,643, 3,273 or 6,368 mg/m3) (dynamic whole
body chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist Ultrasonic spray
nozzle nebulizer), analytical concentration and
method were reported
Authors stated that there were no treatment-related microscopic changes,
but data were not provided.
aThe high-dose group had an increase in mortality.
* Statistically significant p < 0.05 as determined by study authors.
Conversions from drinking water concentrations to mg/kg-d performed by study authors.
Conversion from ppm to mg/m3 is 1 ppm = 3.031 mg/m3.
This document is a draft for review purposes only and does not constitute Agency policy.
B-17 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table B-4. Changes in urinary bladder histopathology in animals following
2 oral exposure to tert-butanol
Reference and study design
Resu Its
NTP (1995)
Incidence (severity):
F344/N rat; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20,
Males
Females
40 mg/mL)
Transitional
M: 0, 230, 490, 840, 1,520,
Dose epithelial
Dose
Transitional epithelial
3,610a mg/kg-d
(ms/ks-d) hyperplasia
(mg/kg-d)
hyperplasia
F: 0, 290, 590, 850, 1,560,
3,620a mg/kg-d
0 0/10
0
0/10
13 weeks
230 not evaluated
290
not evaluated
490 not evaluated
590
not evaluated
840 0/10
850
not evaluated
1,520 1/10 (3.0)
1,560
0/10
3,610 7/10* (2.9)
3,620
3/10 (2.0)
Severity: 1 = minimal, 2 = mild, 3 = moderate, 4
= marked
NTP (1995)
Incidence (severity):
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20,
Males
Females
40 mg/mL)
Transitional
Transitional
M: 0, 350, 640, 1,590, 3,940,
Dose epithelial Inflam-
Dose
epithelial Inflam-
8,210a mg/kg-d
(mg/kg-d) hyperplasia mation (mg/kg-d)
hyperplasia mation
F: 0, 500, 820, 1,660, 6,430,
ll,620a mg/kg-d
0 0/10 0/10
0
0/10 0/10
13 weeks
350 not evaluated
500
0/10 0/10
640 not evaluated
820
not evaluated
1,590 0/10 0/10
1,660
not evaluated
3,940 6/10* (1.3) 6/10* (1.3)
6,430
0/10 0/10
8,210 10/10* (2.0) 10/10*
11,620
3/9(2.0) 6/9* (1.2)
(2.3)
Severity: 1 = minimal, 2 = mild, 3 = moderate, 4
= marked
NTP (1995)
No treatment-related effects observed
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at
15 months)
Drinking water (0,1.25, 2.5, 5, or
10 mg/mL)
M: 0, 90, 200, 420a mg/kg-d
F: 0,180, 330, 650a mg/kg-d
2 years
This document is a draft for review purposes only and does not constitute Agency policy.
B-18 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Reference and study design
Resu Its
NTP (1995)
B6C3Fi mouse; 60/sex/treatment
Drinking water (0, 5,10, or
20 mg/mL)
M: 0, 540,1,040, 2,070a mg/kg-d
F: 0, 510,1,020, 2,110 mg/kg-d
2 years
Incidence (severity):
Males
Females
Severity: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked
Transitional
Transitional
Dose
epithelial
Inflam-
Dose
epithelial
Inflam-
(mg/kg-d)
hyperplasia
mation
(mg/kg-d)
hyperplasia
mation
0
1/59 (2.0)
0/59
0
0/59
0/59
540
3/59 (1.7)
3/59 (1.7)
510
0/60
0/60
1,040
1/58 (1.0)
1/58 (1.0)
1,020
0/59
0/59
2,070
17/59*
37/59* (2.0)
2,110
3/57(1.0)
4/57*
(1.8)
(2.0)
1
2 aThe high-dose group had an increase in mortality.
3 * Statistically significant p < 0.05 as determined by study authors.
4 Conversions from drinking water concentrations to mg/kg-d performed by study authors.
This document is a draft for review purposes only and does not constitute Agency policy.
B-19 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
¦ = exposures at which the endpoint: was reported statistically significant by study authors
~ = exposures at which the endpoint was reported not statistically significant by study authors
x = exposures at which all animals died and were unable to be examined for the endpoint
1
2
3
4
URINARY Transitional epithelium hyperplasia; M rat(C)
BLADDER
EFFECTS
Transitional epithelium hyperplasia; Frai(€)
Siihrhroiiic
Transitional epithelium hyperplasia; M mouse (CI
Inflammation; M mouse (C)
inflammation: Fmouse (O
TmnsUional epithelium hyperplasia; F mouse (€)
TrMsifiostaSepilteiumi hyperplasia; M mouse (C)
Inflammation; M mouse (
Chronic inflammation; Fmouse (
Transitional epithelium '\jit p is-ia; F mouse (
!r ^geii; M rat (A)
LIVER
EFFECTS
Suhchioim-
Relative weighs M rat (At
Relative weight M
Absolute weight;
Relative weight Frat(C)
Absolute weight; M rat (€)
Relative weight M mouse I'C)
Relative weight; F mouse
Absolute weight; M mouse (C)
Absolute weight; F mouse (C)
( hmnir
Relative weight M raS;(C)
Relative weight F rat (
Absolute weight M rat
Absolute weighs Fratfu
Fatly tissue; M mouse (€)
Fatly tissue; F mouse (0)
Reproductive
Relative weight M rat (8)
Absolute weight; M m{
Absolute weight; Fratf
Relative weight; F rat (u i
DO j
| o~a
! Q ¦
-0-
£3—0
¦ ¦¦ ¦—¦
¦ ¦¦ ¦—¦
~—E3-Q 0—0
Q—0 H ¦-
0-B-
~ ~ ~ B—0
' ¦ ~ D i
~—-0—0
¦ ~ ~ a
o-b-
Q—0 B
O B B~
10
Q B B B
13—0 B B
~—B B B
100
1,000 10,000 100,000
Dose fing/kg-day)
Sources: (A) Acharva et al. (1997); Acharva et al. (1995); (B) Lvondell Chemical Co. (2004); (C) NTP (1995)
Figure B-3. Exposure-response array of other effects following oral exposure
to tert-butanol.
This document is a draft for review purposes only and does not constitute Agency policy.
B-20 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
¦ = exposures at which the endpoint was reported statistically significant by study authors
~ = exposures at which the endpoint was reported not statistically significant by study authors
LIVER EFFECTS
Absolute liver weight; M rat (A)
Relative liver weight; M rat (A)
Absolute liver weight; F rat (A)
Relative liver weight; F rat (A)
Absolute liver weight; M mouse (A)
Relative liver weight; M mouse (A)
Absolute liver weight; F mouse (A)
Relative liver weight; F mouse (A)
Liver histopathology; M rat (A)
Liver histopathology; Frat(A)
Liver histopathology; M mouse (A)
Liver histopathology; F mouse (A)
100 1,000 10,000
Concentration (mg/m1)
Source: (A) NTP (1997)
Figure B-4. Exposure-response array of other effects following inhalation
exposure to tert-butanol.
This document is a draft for review purposes only and does not constitute Agency policy.
B-21 DRAFT—DO NOT CITE OR QUOTE
~ B B B B
~ B B B B
~ B B ¦ ¦
13 B B B B
~ B B B B
B B B B B
B B B ¦ ¦
~ B B B B
B B B B B
B B B B B
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information—tert-Butyl Alcohol
B.2.2. Genotoxicity
The genotoxic potential of tert-butanol has been studied using a variety of genotoxicity
assays, including bacterial reverse mutation assays, gene mutation assays, chromosomal
aberrations, sister chromatid exchanges, micronucleus induction, and deoxyribonucleic acid (DNA)
strand breaks and adducts. The available genotoxicity data for tert-butanol are discussed below,
and the data summary is provided in Table B-5.
B.2.2.1. Bacterial Systems
The mutagenic potential of tert-butanol has been tested by Zeiger etal. f 19871 using
different Salmonella typhimurium strains both in the presence and absence of S9 metabolic
activation. The preincubation assay protocol was followed. Salmonella strains TA98, TA100,
TA1535, TA1537, and TA1538 were exposed to five concentrations (100, 333,1,000, 3,333, or
10,000 [ig/plate) and tested in triplicate. No mutations were observed in any of the strains tested,
in either the presence or absence of S9 metabolic activation.
Conflicting results have been obtained with tert-butanol-induced mutagenicity in
Salmonella strain TA102, a strain that is sensitive to damage at A-T sites inducible by oxidants and
other mutagens and is excision-repair proficient In a study by Williams-Hill etal. (1999).
tert-butanol induced an increase in the number of revertants in the first three concentrations with
S9 activation in a dose-response manner. The number of revertants decreased in the last two
concentrations. No discussion was provided on why the revertants decreased at higher
concentrations. The results of this study indicated that test strain TA102 might be a more sensitive
strain for monitoring tert- butanol levels fWilliams-Hill etal.. 19991. In another study by Mcgregor
etal. (20051. however, experiments were conducted onTA102 at two different laboratories using
similar protocols, tert-Butanol was dissolved in dimethyl sulfoxide (DMSO) or distilled water and
tested in both the presence and absence of S9 metabolic activation. No statistically significant
increase in mutants was observed in either solvent medium.
Mutagenicity of tert-butanol has been studied in other systems including Neurospora crassa
and Saccharomyces cerevisiae. Yeast strain Neurospora crassa atthe ad-3Alocus (allele 38701) was
used to test the mutagenic activity of tert-butanol at a concentration of 1.75 mol/L for 30 minutes.
tert-Butanol did not induce reverse mutations in the tested strain at the exposed concentration
(Dickey etal.. 1949). tert- Butanol without exogenous metabolic activation, however, significantly
increased the frequency of petite mutations (the mitochondrial DNA deletion rho-) in
Saccharomyces cerevisiae laboratory strains K5-A5, MMY1, D517-4B, and DS8 flimenez et al.. 1988).
This effect on mitochondrial DNA, also observed with ethanol and other solvents, was attributed by
the study authors to the alteration in the lipid composition of mitochondrial membranes, and
mitochondrial DNA's close association could be affected by membrane composition flimenez etal..
1988).
This document is a draft for review purposes only and does not constitute Agency policy.
B-22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information—tert-Butyl Alcohol
B.2.2.2. In Vitro Mammalian Studies
To understand the role of tert-butanol-induced genotoxicity in mammalian systems, in vitro
studies have been conducted in different test systems and assays, tert-Butanol was tested to
evaluate its ability to induce forward mutations at the thymidine kinase locus (tk) in the L5178Y
tk+/- mouse lymphoma cells using forward mutation assay. Experiments were conducted in both
the presence and absence of S9 metabolic activation. The mutant frequency was calculated using
the ratio of mutant clones per plate/total clones per plate x 200. tert-Butanol did not reliably
increase the frequency of forward mutations in L5178Y tk+/- mouse lymphoma cells with or
without metabolic activation, although one experiment without addition of S9 yielded a small
(1.7-fold) increase in mutant fraction at the highest tested concentration (5,000 [ig/mL) (McGregor
etal.. 19881
To further determine potential DNA or chromosomal damage induced by tert-butanol in in
vitro systems, NTP (1995) studied sister chromatid exchanges and chromosomal aberrations.
Chinese hamster ovary (CHO) cells were exposed to tert-butanol in both the presence and absence
of S9 activation at concentrations of 160-5,000 |J.g/mL for 26 hours, tert-Butanol did not induce
sister chromatid exchanges at any concentration tested, although in one experiment, percent
relative change of sister chromatid exchanges per chromosome scored slightly increased. The same
authors also studied the effect of tert-butanol on chromosomal aberration formation. CHO cells
were exposed to four concentrations (160, 500,1,600, or 5,000 ng/mL) of tert-butanol in both the
presence and absence of S9. No significant increase in chromosomal aberration was observed at
any concentration tested. Of note is that, due to severe toxicity at the highest concentration
(5,000 [ig/mL), only 13 metaphase cells were scored instead of 100 in the chromosomal aberration
assay.
Sgambato etal. (2009) examined the effects of tert- butanol on DNA damage using a normal
diploid rat fibroblast cell line. Cells were treated with 0- to 100-mM tert-butanol for 48 hours to
determine the half-maximal inhibitory concentration (IC50; 0.44 ± 0.2 mM). The 48-hour IC50
concentration then was used to determine DNA content, cell number, and phases of the cell cycle
after 24 and 48 hours of exposure. Total protein and DNA oxidative damage also were measured. A
comet assay was used to evaluate DNA fragmentation at time 0 and after 30 minutes, 4 hours, or
12 hours of exposure to the IC50 concentration, tert-Butanol inhibited cell division as measured by
the number of cells after 24 and 48 hours of exposure at IC50 concentrations and with
concentrations at 1 /10th the IC50. Cell death did not increase, suggesting a reduction in cell number
due to reduced replication rather than to cytotoxicity, tert-Butanol caused an accumulation in the
G0/G1 phase of replication, related to different effects on the expression of the cyclin Dl, p27Kipl,
and p53 genes. An initial increase in DNA damage as measured by nuclear fragmentation was
observed at 30 minutes. The DNA damage declined drastically after 4 hours and disappeared
almost entirely after 12 hours of exposure to tert-butanol. This reduction in the extent of DNA
This document is a draft for review purposes only and does not constitute Agency policy.
B-23 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Supplemental Information—tert-Butyl Alcohol
fragmentation after the initial increase is likely the result of an efficient DNA repair mechanism
activated by cells following DNA damage induced by tert-butanol.
DNA damage caused by tert-butanol was determined by single-cell gel electrophoresis
(comet assay) in human promyelocytic leukemia (HL-60) cells. The cells were exposed to
concentrations ranging from 1 to 30 mmol/L for 1 hour, and 100 cells were evaluated for DNA
fragmentation. A dose-dependent increase in DNA damage was observed between 1 and
30 mmol/L. No cytotoxicity was observed at the concentrations tested (Tang etal.. 19971.
B.2.2.3. In Vivo Mammalian Studies
Few in vivo studies are available to understand the role of tert-butanol on genotoxicity. The
National Toxicology Program studied the effect of tert-butanol in a 13-week toxicity study fNTP.
19951. Peripheral blood samples were obtained from male and female B6CF1 mice exposed to tert-
butanol in drinking water at doses of 3,000-40,000 ppm. Slides were prepared to determine the
frequency of micronuclei in 10,000 normochromatic erythrocytes. In addition, the percentage of
polychromatic erythrocytes among the total erythrocyte population was determined. No increase in
micronucleus induction in peripheral blood lymphocytes was observed either in male or female
B6C3Fi mice exposed for 13 weeks to tert-butanol in drinking water at concentrations as high as
40,000 ppm (2,110 mg/kg-day) fNTP. 1997.19951. Furthermore, no induction of micronuclei in
polychromatic erythrocytes was observed in bone marrow cells of male rats receiving
intraperitoneal injections fNTP. 19971.
Male Kunming mice (8 per treatment) were administered 0, 0.099, 0.99,10,101, or
997 |ig/kg BW14C-tert-butanol in saline via gavage with specific activity ranging from 1.60 to
0.00978 mCi/mol fYuan etal.. 2007). Animals were sacrificed 6 hours after exposure, and liver,
kidney, and lung were collected. Tissues were prepared for DNA isolation with samples from the
same organs from every two mice combined. DNA adducts were measured using accelerated mass
spectrometry. The results of this study showed a dose-response increase in DNA adducts in all
three organs measured, although the methodology used to detect DNA adducts is considered
sensitive but could be nonspecific. The authors stated that tert-butanol was found, for the first time,
to form DNA adducts in mouse liver, lung, and kidney. Because this is a single and first-time study,
further validation of this study will provide certainty in understanding the mechanism of tert-
butanol-induced DNA adducts.
This document is a draft for review purposes only and does not constitute Agency policy.
B-24 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Table B-5. Summary of genotoxicity (both in vitro and in vivo) studies of tert-
butanol
Test system
Dose/ Cone.
Results3
Comments
Reference
Bacterial Systems
-S9
+S9
Reverse Mutation Assay
Salmonella typhimurium
(TA98, TA100, TA1535,
TA1537, TA1538)
100, 333, 1,000,
3333,
10,000 ng/plate
Preincubation procedure was
followed. This study was part of
the NTP 1995 testing results.
Zeiger et al.
(1987); NTP
(1995)
Reverse Mutation Assay
Salmonella typhimurium
(TA102)
1000-
4000 ng/plate
ND
+
Only tested with S9 activation
Williams-Hill et
al. (1999)
Reverse Mutation Assay
Salmonella typhimurium
(TA98, TA100, TA102,
TA1535, TA1537)
5, 15, 50, 100, 150,
200, 500, 1,000,
1,500, 2,500,
5,000 |jg/plate
Experiments conducted in two
different laboratories, two
vehicles - distilled water and
DMSO were used, different
concentrations were used in
experiments from different
laboratories
Mcgregor et al.
(2005)
Reverse mutation
Neurospora crassa, ad-3A
locus (allele 38701)
1.75mol/L
Eighty four percent cell death
was observed; note it is a 1949
study
Dickev et al.
(1949)
Mitochondrial mutation
Saccharomyces cerevisiae
(K5-5A, MMY1, D517-4B,
and DS8)
4.0% (vol/vol)
+b
ND
Mitochondrial mutations,
membrane solvent
Jimenez et al.
(1988)
In vitro Systems
Gene Mutation Assay,
Mouse lymphoma cells
L5178Y TK+/"
625, 1,000, 1,250,
2,000, 3,000,
4,000, 5,000 Hg/mL
Cultures were exposed for 4 h,
then cultured for 2 days before
plating in soft agar with or
without trifluorothymidine,
3 ng/mL; this study was part of
the NTP 1995 testing results
McGregor et al.
(1988); NTP
(1995)
Sister-chromatid exchange,
Chinese Hamster Ovary cells
160, 500, 1,600,
2,000, 3,000,
4,000, 5,000 Hg/mL
-
-
This study was part of the NTP
1995 testing results
Gallowav et al.
(1987); NTP
(1995)
Chromosomal Aberrations,
Chinese Hamster Ovary cells
160, 500, 1,600,
2,000, 3,000,
4,000, 5,000 Hg/mL
This study was part of the NTP
1995 testing results
Gallowav et al.
(1987); NTP
(1995)
DNA damage (comet assay),
Rat fibroblasts
0.44 mmol/L (IC50)
+c
ND
Exposure duration - 30 min,
4 h, 12 h; this study provides
other information on effect of
cell cycle control genes and
mechanism of action for TBA
Sgambato et al.
(2009)
This document is a draft for review purposes only and does not constitute Agency policy.
B-25 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Supplemental Information—tert-Butyl Alcohol
Test system
Dose/ Cone.
Results3
Comments
Reference
DNA damage, (comet
assay), HL-60 leukemia cells
1, 5,10, 30 mmol/L
+ ND
Exposure duration - lh
Tang et al.
(1997)
In vivo Animal Studies
Micronucleus induction,
B6C3F1 mouse peripheral
blood cells
3,000, 5,000,
10,000, 20,000,
40,000 ppm
13-week, subchronic, drinking
water study
NTP (1995)
Micronucleus induction,
male rats, bone marrow
cells
39, 78, 156, 312,
625, 1250
i.p injections - 3 times at 24 h
intervals
NTP (1997)
DNA adducts, male
Kunming mouse liver,
kidney, and lung cells
0.1-1,000 ng/kg
body weight
+
Gavage, 6-h exposure, DNA
adduct determined by
accelerator mass spectrometry
Yuan et al.
(2007)
a+ = positive; - = negative; ND = not determined.
bEffect is predicted to be due to mitochondrial membrane composition.
CDNA damage was completely reversed with increased exposure time.
B.2,3. Summary
tert-Butanol has been tested for its genotoxic potential using a variety of genotoxicity
assays. In general, a positive result in the Ames assay is 73-77% predictive of a positive result in the
rodent carcinogenicity assay (Kirkland et al.. 20051. tert-Butanol did not induce mutations in most
bacterial strains; however, when tested in TA102, a strain that is sensitive to damage at A-T sites
inducible by oxidants, an increase in mutants was observed at low concentrations, although
conflicting results were reported in another study. Furthermore, the solvent (e.g., distilled water or
DMSO) used in the genotoxicity assay could influence results. In one experiment where tert-butanol
was dissolved in distilled water, a significant, dose-related increase in the number of mutants was
observed, with the maximum value reaching almost twice the control value. DMSO is known to be a
radical scavenger, and its presence in high concentrations might mask a mutagenic response
modulated by oxidative damage. Other species such as Neurospora crassa did not produce reverse
mutations due to exposure to tert-butanol.
tert-Butanol was tested in several human and animal in vitro mammalian systems for
genotoxicity (gene mutation, sister chromatid exchanges, chromosomal aberrations, and DNA
damage). No increase in gene mutations was observed in mouse lymphoma cells (L5178YTK+/-).
These specific locus mutations in mammalian cells are used to demonstrate and quantify genetic
damage, thereby confirming or extending the data obtained in the more widely used bacterial cell
tests. Sister chromatid exchanges or chromosomal aberrations were not observed in CHO cells in
response to tert-butanol treatment. DNA damage was detected using a comet assay, however, in
both rat fibroblasts and HL-60 leukemia cells, with either an increase in DNA fragmentation at the
beginning of the exposure or dose-dependent increase in DNA damage observed. An initial increase
in DNA damage was observed at 30 minutes that declined drastically following 4 hours of exposure
This document is a draft for review purposes only and does not constitute Agency policy.
B-26 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Supplemental Information—tert-Butyl Alcohol
and disappeared almost entirely after 12 hours of exposure to tert-butanol. This reduction in the
extent of DNA fragmentation after an initial increase is likely the result of an efficient DNA repair
mechanism activated by cells following DNA damage induced by tert-butanol. A dose-dependent
increase in DNA damage was observed in human cells tested; however, because the exposure
occurred for only 1 hour in this study, whether DNA-repair mechanisms would occur after a longer
period of observation cannot be discerned.
Limited in vivo animal studies have been conducted on DNA adduct formation and
micronucleus induction. A dose-response increase in DNA adducts was observed in mouse liver,
kidney, and lung cells. The authors used accelerated mass spectrometry to detect DNA adducts, but
the identity of these adducts was not determined. The method uses 14C-labeled chemical for dosing,
isolated DNA is oxidized to carbon dioxide and reduced to filamentous graphite, and the ratios of
14C/12C are measured. The ratio then is converted to DNA adducts based on nucleotide content of
the DNA. Confirmation of these data will further the understanding of the mechanism of
tert-butanol-induced DNA adducts. No increase in micronucleus induction was observed in mouse
peripheral blood cells in a 13-week drinking water study conducted by the National Toxicology
Program.
Overall, a limited database is available for understanding the role of tert-butanol-induced
genotoxicity for mode of action and carcinogenicity. The database is limited in terms of either the
array of genotoxicity tests conducted or the number of studies within the same type of test In
addition, the results are either conflicting or inconsistent The test strains, solvents, or control for
volatility used in certain studies are variable and could influence results. Furthermore, in some
studies, the specificity of the methodology used has been challenged. Given the inconsistencies and
limitations of the database in terms of the methodology used, number of studies in the overall
database, coverage of studies across the genotoxicity battery, and the quality of the studies, the
weight-of-evidence analysis is inconclusive. The available data do not inform a definitive conclusion
on the genotoxicity of tert-butanol and thus the potential genotoxic effects of tert-butanol cannot be
discounted.
This document is a draft for review purposes only and does not constitute Agency policy.
B-27 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Supplemental Information—tert-Butyl Alcohol
APPENDIX C. DOSE-RESPONSE MODELING FOR
THE DERIVATION OF REFERENCE VALUES FOR
EFFECTS OTHER THAN CANCER AND THE
DERIVATION OF CANCER RISK ESTIMATES
This appendix provides technical detail on dose-response evaluation and determination of
points of departure (PODs) for relevant endpoints. The endpoints were modeled using EPA's
Benchmark Dose Software (BMDS), version 2.1.2. The preambles for the cancer and noncancer
parts below describe the common practices used in evaluating the model fit and selecting the
appropriate model for determining the POD as outlined in the Benchmark Dose Technical Guidance
Document (U.S. EPA. 20001. In some cases, using alternative methods based on statistical judgment
might be appropriate; exceptions are noted as necessary in the summary of the modeling results.
C.l.l. Noncancer Endpoints
C.1.1.1. Data Sets
Data sets selected for dose-response modeling are provided in Table C-l. In all cases,
administered exposure was used in modeling the response data.
C.l.l.2. Model Fit
All models were fit to the data using the maximum likelihood method. The following
procedures were used, depending on whether data were dichotomous or continuous.
• For dichotomous models, the following parameter restrictions were applied: for log-logistic
model, restrict slope >1; for gamma and Weibull models, restrict power >1; and for
multistage models, restrict beta values >0. Each model was tested for goodness-of-fit using
a chi-square goodness-of-fit test (x2 p-value <0.10 indicates lack of fit). Other factors also
were used to assess model fit, such as scaled residuals, visual fit, and adequacy of fit in the
low-dose region and near the benchmark response (BMR).
• For continuous models, the following parameter restrictions were applied: for polynomial
models, restrict beta values >0; and for Hill, power, and exponential models, restrict power
>1. Model fit was assessed by a series of tests. For each model, first the homogeneity of the
variances was tested using a likelihood ratio test (BMDS Test 2). If Test 2 was not rejected
(X2 p-value > 0.10), the model was fit to the data assuming constant variance. If Test 2 was
rejected (x2 p-value < 0.10), the variance was modeled as a power function of the mean, and
This document is a draft for review purposes only and does not constitute Agency policy.
C-l DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Supplemental Information—tert-Butyl Alcohol
the variance model was tested for adequacy of fit using a likelihood ratio test (BMDS
Test 3). For fitting models using either constant variance or modeled variance, models for
the mean response were tested for adequacy of fit using a likelihood ratio test (BMDS Test
4, with x2 p-value < 0.10 indicating inadequate fit). Other factors also were used to assess
the model fit, such as scaled residuals, visual fit, and adequacy of fit in the low-dose region
and near the BMR.
C.l.1.3. Model Selection
For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
estimated by the profile likelihood method) and the Akaike's information criterion (AIC) value were
used to select a best-fit model among the models exhibiting adequate fit. If the BMDL estimates
were "sufficiently close," that is, differed by no more than three-fold, the model selected was the
one that yielded the lowest AIC value. If the BMDL estimates were not sufficiently close, the lowest
BMDL was selected as the POD.
Table C-l. Noncancer endpoints selected for dose-response modeling for
tert- butanol
Endpoint/Study
Species/
Sex
Doses and effect data
Kidney transitional
epithelial hyperplasia
NTP (1995)
Rat (F344)/Male
Dose (mg/kg-d)
0
90
200
420
Incidence/Total
25/50
32/50
36/50
40/50
Kidney transitional
epithelial hyperplasia
NTP (1995)
Rat
(F344)/Female
Dose (mg/kg-d)
0
180
330
650
Incidence/Total
0/50
0/50
3/50
17/50
Increased absolute
kidney weight
NTP (1995)
Rat (F344)/Male
Dose (mg/kg-d)
0
90
200
420
Mean ± SD (n)
1.78 ±0.18
(10)
1.85 ±0.17
(10)
1.99 ±0.18
(10)
1.9 ±0.23
(10)
Increased absolute
kidney weight
NTP (1995)
Rat
(F344)/Female
Dose (mg/kg-d)
0
180
330
650
Mean ± SD (n)
1.07 ± 0.09
(10)
1.16 ±0.10
(10)
1.27 ±0.08
(10)
1.31 ±0.09
(10)
Kidney inflammation
NTP (1995)
Rat
(F344)/Female
Dose (mg/kg-d)
0
180
330
650
Incidence/Total
2/50
3/50
13/50
17/50
Increased absolute
kidney weight
NTP (1997)
Rat (F344)/Male
Concentration
(mg/m3)
0
406
825
1643
3274
6369
Mean ± SD (n)
1.21 ±
0.082
(10)
1.21 ±
0.096
(9)
1.18 ±
0.079
(10)
1.25 ±
0.111
(10)
1.34 ±
0.054
(10)
1.32 ±
0.089
(10)
Increased absolute
kidney weight
Rat
(F344)/Female
Concentration
(mg/m3)
0
406
825
1643
3274
6369
This document is a draft for review purposes only and does not constitute Agency policy.
C-2 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Endpoint/Study
Species/
Sex
Doses and effect data
NTP (1997)
Mean ± SD (n)
0.817 ±
0.136
(10)
0.782 ±
0.063
(10)
0.821 ±
0.061
(10)
0.853 ±
0.045
(10)
0.831 ±
0.054
(10)
0.849 ±
0.038
(10)
1 C.1.1.4. Modeling Results
2 Below are tables summarizing the modeling results for the noncancer endpoints modeled.
3 Table C-2. Summary of BMD modeling results for kidney transitional epithelial
4 hyperplasia in male F344 rats exposed to tert-butanol in drinking water for 2
5 years (NTP. 19951: BMR = 10% extra risk
Model3
Goodness of fit
BMDio
(mg/kg-d)
BMDLio
(mg/kg-d)
Basis for model selection
p-value
AIC
Log-logistic
0.976
248.0
30
16
Log-logistic model selected as best-
fitting model based on lowest AIC
with all BMDL values sufficiently
close (BMDLs differed by slightly
more than 3-fold).
Gamma
0.784
248.5
46
29
Logistic
0.661
248.8
58
41
Log-probit
0.539
249.2
84
53
Multistage, 3°
0.784
248.5
46
29
Probit
0.633
248.9
60
43
Weibull
0.784
248.5
46
29
Dichotomous-Hill
0.968
250.0
25
15
6 a Scaled residuals for selected model for doses 0, 90, 200, and 420 mg/kg-d were -0.076, 0.147, 0.046, and -0.137,
7 respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
C-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Supplemental Information—tert-Butyl Alcohol
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.9
0.7
0.6
0.5
0.4
BMDL
BMD
0
50
100
150
200
250
300
350
400
dose
17:16 05/13 2011
Figure C-l. Plot of incidence by dose, with fitted curve for LogLogistic model
for kidney transitional epithelial hyperplasia in male F344 rats exposed to
tert-butanol in drinking water for 2 years (NTP. 1995): BMR = 10% extra risk;
dose shown in mg/kg-d.
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File: M:\NCEA t-Butanol\BMD modeling\BMDS Output\17 NTP 1995b_Kidney
transitional epithelial hyperplasia, male rats_LogLogistic_10.(d)
Gnuplot Plotting File: M:\NCEA t-Butanol\BMD modeling\BMDS Output\17 NTP
1995b_Kidney transitional epithelial hyperplasia, male rats_LogLogistic_10.pit
Fri May 13 17:16:25 2011
[notes]
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Incidence
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 = 0.5
intercept = -5.54788
This document is a draft for review purposes only and does not constitute Agency policy.
C-4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Supplemental Information—tert-Butyl Alcohol
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.71
intercept -0.71 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
background 0.505366 * * *
intercept -5.58826 * * *
slope 1 * * *
- Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -121.996 4
Fitted model -122.02 2 0.048148 2 0.9762
Reduced model -127.533 1 11.0732 3 0.01134
AIC: 248.04
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.5054 25.268 25.000 50 -0.076
90.0000 0.6300 31.498 32.000 50 0.147
200.0000 0.7171 35.854 36.000 50 0.046
420.0000 0.8076 40.382 40.000 50 -0.137
ChiA2 = 0.05 d.f. = 2 P-value = 0.9762
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 2 9.6 9 67
BMDL = 15.6252
This document is a draft for review purposes only and does not constitute Agency policy.
C-5 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-3. Summary of BMD modeling results for kidney transitional epithelial
2 hyperplasia in female F344 rats exposed to tert-butanol in drinking water for
3 2 years (NTP. 1995): BMR = 10% extra risk
Goodness of fit
Model3
p-value
AIC
BMDio
(mg/kg-d)
BMDLio
(mg/kg-d)
Basis for model selection
Gamma
0.83
91.41
409
334
Multistage 3rd-order model
selected as best-fitting model
based on lowest AIC with all BMDL
values sufficiently close (BMDLs
Logistic
0.50
92.81
461
393
LogLogistic
0.79
91.57
414
333
differed by less than 3-fold).
LogProbit
0.89
91.19
400
327
Multistage 3°
0.92
89.73
412
339
Probit
0.62
92.20
439
372
Weibull
0.76
91.67
421
337
Dichotomous-Hill
N/Ab
117.89
Error0
Error0
aScaled residuals for selected model for doses 0,180, 330, and 650 mg/m3 were 0.0, -0.664, 0.230, and 0.016,
respectively.
bNo available degrees of freedom to estimate a p-value.
CBMD and BMDL computation failed for the Dichotomous-Hill model.
Multistage Model with 0.95 Confidence Level
O 10O 200 300 400 500 600
dose
17:18 05/13 2011
Figure C-2. Plot of incidence by dose, with fitted curve for Multistage 3° model
for kidney transitional epithelial hyperplasia in female F344 rats exposed to
tert-butanol in drinking water for 2 years (NTP. 19951: BMR = 10% extra risk;
dose shown in mg/kg-d.
Multistage Model. (Version: 3.2; Date: 05/26/2010)
This document is a draft for review purposes only and does not constitute Agency policy.
C-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
Supplemental Information—tert-Butyl Alcohol
Input Data File: M:\NCEA te/"t"butanol\BMD modeling\BMDS Output\20 NTP
1995b_Kidney transitional epithelial hyperplasia, female rats_Multi3_10.(d)
Gnuplot Plotting File: M:\NCEA t6rt_butanol\BMD rnodeling\BMDS Gutput\20 NTP
1995b_Kidney transitional epithelial hyperplasia, female rats_Multi3_10.pit
Mon May 09 18:31:33 2011
[notes]
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2-beta3*doseA3)]
The parameter betas are restricted to be positive
Dependent variable = Incidence
Independent variable = Dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta (1) = 0
Beta (2) = 1.51408e-007
Beta (3) = 1.29813e-009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l) -Beta (2)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta (3)
Beta (3) 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0 * * *
Beta (1) 0 * * *
Beta (2) 0 * * *
Beta (3) 1.50711e-009 * * *
- Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -43.4002 4
This document is a draft for review purposes only and does not constitute Agency policy.
C-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Supplemental Information—tert-Butyl Alcohol
Fitted model -43.8652 1 0.9301 3 0.8182
Reduced model -65.0166 1 43.2329 3 <.0001
AIC: 89.7304
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0000 0.000 0.000 50 0.000
180.0000 0.0088 0.438 0.000 50 -0.664
330.0000 0.0527 2.636 3.000 50 0.230
650.0000 0.3389 16.946 17.000 50 0.016
ChiA2 = 0.49 d.f. = 3 P-value = 0.9200
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 411.95
BMDL = 338 . 618
BMDU = 4 6 9.73
Taken together, (338.618, 469.73 ) is a 90 % two-sided confidence
interval for the BMD
This document is a draft for review purposes only and does not constitute Agency policy.
C-8 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-4. Summary of BMD modeling results for absolute kidney weight in
2 male F344 rats exposed to tert-butanol in drinking water for 15 months (NTP.
3 1995): BMR = 10% rel. dev. from control mean
Model3
Goodness of fit
BMDiord
(mg/kg-d)
BMDLiord
(mg/kg-d)
Basis for model selection
p-value
AIC
Exponential (M2)b
0.123
-86.757
661
307
Of the models that provided an
adequate fit and a valid BMDL
estimate, the linear model was
selected based on lowest AIC.
Exponential (M3)c
0.123
-86.757
661
307
Exponential (M4)
0.167
-87.041
errord
0
Exponential (M5)
N/Ae
-85.880
errord
0
Hill
0.301
-87.880
errord
errord
Power'
Polynomial 3°B
Polynomial 20h
Linear
0.126
-86.804
657
296
aConstant variance case presented (BMDS Test 2 p-value = 0.777), selected model in bold; scaled residuals for
selected model for doses 0, 90, 200, and 420 mg/kg-d were -0.78, -0.11,1.65, -0.76, respectively.
bThe Exponential (M2) model can appear equivalent to the Exponential (M3) model, however differences exist in
digits not displayed in the table.
cThe Exponential (M3) model can appear equivalent to the Exponential (M2) model, however differences exist in
digits not displayed in the table.
dBMD or BMDL computation failed for this model.
eNo available degrees of freedom to calculate a goodness-of-fit value.
fFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
gFor the Polynomial 3° model, the b3 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
hFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
This document is a draft for review purposes only and does not constitute Agency policy.
C-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information—tert-Butyl Alcohol
Linear Model, with BMR of 0.1 Rel. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
Linear
2.1
2
1.9
1.8
1.7
BMDL
BMD
0
100
200
300
400
500
600
dose
11:46 05/26 2015
Figure C-3. Plot of mean response by dose, with fitted curve for Linear model
with constant variance for absolute kidney weight in male F344 rats exposed
to tert-butanol in drinking water for 15 months fNTP. 19951: BMR = 10% rel.
dev. from control mean; dose shown in mg/kg-d.
Polynomial Model. (Version: 2.20; Date: 10/22/2014)
The form of the response function is: Y[dose] = beta_0 + beta_l*dose.
A constant variance model is fit.
Benchmark Dose Computation.
BMR = 10% Relative deviation
BMD = 656.583
BMDL at the 95% confidence level = 295.826
Parameter Estimates
Variable
Estimate
Default initial
parameter values
alpha
0.0361494
0.0362125
rho
n/a
0
beta_0
1.83173
1.83173
beta_l
0.000278979
0.000278979
This document is a draft for review purposes only and does not constitute Agency policy.
C-10 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std Dev
Scaled Resid
0
10
1.78
1.83
0.18
0.19
-0.777
90
10
1.85
1.86
0.17
0.19
-0.114
200
10
1.99
1.89
0.18
0.19
1.65
420
10
1.9
1.95
0.23
0.19
-0.763
2 Likelihoods of Interest
Model
Log(likelihood)
# Param's
AIC
A1
48.474229
5
-86.948457
A2
49.025188
8
-82.050377
A3
48.474229
5
-86.948457
fitted
46.401914
3
-86.803828
R
45.368971
2
-86.737942
3 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1
7.31243
6
0.2929
Test 2
1.10192
3
0.7766
Test 3
1.10192
3
0.7766
Test 4
4.14463
2
0.1259
This document is a draft for review purposes only and does not constitute Agency policy.
C-ll DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-5. Summary of BMD modeling results for absolute kidney weight in
2 female F344 rats exposed to tert-butanol in drinking water for 15 months
3 (NTP. 1995): BMR = 10% rel. dev. from control mean
Model3
Goodness of fit
BMDiord
(mg/kg-d)
BMDLiord
(mg/kg-d)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0594
-144.00
318
249
The Exponential (M4) model was
selected as the only model with
adequate fit.
Exponential (M4)
0.176
-145.81
164
91.4
Exponential (M5)
N/Ac
-145.65
207
117
Hill
N/Ac
-145.65
202
119
Powerd
Polynomial 3°e
Polynomial 2°f
Linear
0.0842
-144.70
294
224
aConstant variance case presented (BMDS Test 2 p-value = 0.852), selected model in bold; scaled residuals for
selected model for doses 0,180, 330, and 650 mg/kg-d were 0.21, -0.9, 0.94, -0.25, respectively.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
cNo available degrees of freedom to calculate a goodness-of-fit value.
dFor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
eFor the Polynomial 3° model, the b3 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
fFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
4
This document is a draft for review purposes only and does not constitute Agency policy.
C-12 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
Exponential 4 Model, with BMR of 0.1 Rel. Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
Exponential 4
1.4
1.35
1.3
1.25
1.2
1.15
1.1
1.05
1
0
100
200
300
400
500
600
2 Figure C-4. Plot of mean response by dose, with fitted curve for Exponential
3 (M4) model with constant variance for absolute kidney weight in female F344
4 rats exposed to tert-butanol in drinking water for 15 months (NTP. 1995):
5 BMR = 10% rel. dev. from control mean; dose shown in mg/kg-d.
6 Exponential Model. (Version: 1.10; Date: 01/12/2015)
7 The form of the response function is: Y[dose] = a * [c-(c-l) * exp(-b * dose)].
8 A constant variance model is fit.
9 Benchmark Dose Computation.
10 BMR = 10% Relative deviation
11 BMD = 163.803
12 BMDL at the 95% confidence level = 91.3614
13 Parameter Estimates
Variable
Estimate
Default initial
parameter values
Inalpha
-4.84526
-4.89115
rho
n/a
0
a
1.06808
1.0203
b
0.00258011
0.00282085
c
1.29013
1.35122
d
n/a
1
14
This document is a draft for review purposes only and does not constitute Agency policy.
C-13 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std Dev
Scaled Resid
0
10
1.07
1.07
0.09
0.09
0.2112
180
10
1.16
1.18
0.1
0.09
-0.8984
330
10
1.27
1.25
0.08
0.09
0.9379
650
10
1.31
1.32
0.09
0.09
-0.2507
2 Likelihoods of Interest
Model
Log(likelihood)
# Param's
AIC
A1
77.82307
5
-145.6461
A2
78.21688
8
-140.4338
A3
77.82307
5
-145.6461
R
62.21809
2
-120.4362
4
76.90527
4
-145.8105
3 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1
32
6
<0.0001
Test 2
0.7876
3
0.8524
Test 3
0.7876
3
0.8524
Test 6a
1.836
1
0.1755
This document is a draft for review purposes only and does not constitute Agency policy.
C-14 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-6. Summary of BMD modeling results for kidney inflammation in
2 female rats exposed to tert-butanol in drinking water for 2 years (NTP. 1995):
3 BMR = 10% extra risk
Model3
Goodness of fit
BMDior.
(mg/kg-d)
BMDLior.
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.084
169.9
231
135
LogProbit was selected on the
basis of the lowest AIC with all
BMDL values for fitting models
being sufficiently close (BMDLs
differed by less than 3-fold).
Logistic
0.082
169.7
305
252
LogLogistic
0.092
169.8
228
124
LogProbit
0.243
167.6
254
200
Multistage 3°
0.072
170.3
216
132
Probit
0.108
169.2
285
235
Weibull
0.081
170.0
226
134
Dichotomous-Hill
N/Ab
169.5
229
186
aSelected model in bold; scaled residuals for selected model for doses 0,180, 330, and 650 mg/kg-d were -0.067,
-0.700,1.347, and -0.724, respectively.
bNo available degrees of freedom to estimate a p-value.
Log Pro bit Model with 0.95 Confidence Level
LogProbit
B M DL
BMD
17:17 05/13 201 1
5
6
7
Figure C-5. Plot of incidence by dose, with fitted curve for LogpPobit model for
kidney inflammation in female rats exposed to tert-butanol in drinking water
for 2 years (NTP. 19951: BMR = 10% extra risk; dose shown in mg/kg-d.
This document is a draft for review purposes only and does not constitute Agency policy.
C-15 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
Supplemental Information—tert-Butyl Alcohol
Probit Model. (Version: 3.2; Date: 10/28/2009)
Input Data File: M:/NCEA terf-blltanol/BMD modeling/BMDS Output/19 NTP
1995b_Kidney inflammation, female rats_LogProbit_10.(d)
Gnuplot Plotting File: M:/NCEA tert-blltanol/BMD modeling/BMDS Output/19 NTP
1995b_Kidney inflammation, female rats_LogProbit_10.pit
Fri May 13 17:17:59 2011
[notes]
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+Slope*Log(Dose)),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Incidence
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 (and Specified) Parameter Values
background = 0.04
intercept = -8.01425
slope = 1.18 928
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.51
intercept -0.51 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
background 0.0381743 0.0246892 -0.0102155 0.0865642
intercept -6.82025 0.161407 -7.1366 -6.5039
slope 1 NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
This document is a draft for review purposes only and does not constitute Agency policy.
C-16 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Supplemental Information—tert-Butyl Alcohol
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -80.4502 4
Fitted model -81.8218 2 2.7432 2 0.2537
Reduced model -92.7453 1 24.5902 3 <.0001
AIC: 167.644
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.0382
1. 909
2 .000
50
0 . 067
180.0000
0.0880
4.402
3 .000
50
o
r-
o
330.0000
0.1859
9.295
13.000
50
1.347
650.0000
0.3899
19.495
17.000
50
-0 .72
Chi"2 = 2.83 d.f. = 2 P-value = 0.2427
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 254.347
BMDL = 19 9.789
This document is a draft for review purposes only and does not constitute Agency policy.
C-17 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-7. Summary of BMD modeling results for absolute kidney weight in
2 male F344 rats exposed to tert-butanol via inhalation for 6 hr/d, 5d/wk for 13
3 weeks (NTP. 1997): BMR = 10% relative deviation from the mean
Model3
Goodness of fit
BMCiord
(mg/m3)
BMCLiord
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
<0.0001
-205.06
errorb
errorb
Although the Hill model was the
only adequately fitting model
(p>0.1), the resulting fit was
essentially a step-function that
does not support interpolation
between the well-fit
observations.
Exponential (M3)
<0.0001
-203.06
9.2E+07
7094
Exponential (M4)
<0.0001
-203.06
errorb
0
Exponential (M5)
<0.0001
-201.06
errorb
0
Hill
0.763
-226.82
1931
1705
Power0
Linear
0.0607
-220.97
5364
3800
Polynomial 5°d
Polynomial 4°e
Polynomial 3°
1.44E-04
-207.06
-9999
errorf
Polynomial 2°
1.44E-04
-207.06
-9999
18436
aConstant variance case presented (BMDS Test 2 p-value = 0.390), selected model in bold; scaled residuals for
selected model for doses 0, 406, 825,1,643, 3,274, and 6,369 mg/m3 were 0.395, 0.374, -0.75, -1.96e-006,
0.381, and -0.381, respectively.
bBMC or BMCL computation failed for this model.
Tor the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear model.
dFor the Polynomial 5° model, the b5 and b4 coefficient estimates were 0 (boundary of parameters space). The
models in this row reduced to the Polynomial 3° model.
eFor the Polynomial 4° model, the b4 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 3° model.
fBMC or BMCL computation failed for this model.
This document is a draft for review purposes only and does not constitute Agency policy.
C-18 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information—tert-Butyl Alcohol
Hill Model with 0.95 Confidence Level
1 .4
1 .35
1 .3
1 .2
1 .15
1 .1
0 1000 2000 3000 4000 5000 6000
dose
10:15 04/30 2014
Figure C-6. Plot of mean response by concentration, with fitted curve for Hill
model for absolute kidney weight in male F344 rats exposed to tert-butanol
via inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP. 1997): BMR = 10%
relative deviation from the mean; concentration shown in mg/m3.
Hill Model. (Version: 2.15; Date: 10/28/2009)
The form of the response function is: Y[dose] = intercept + v*doseAn/(kAn + doseAn).
A constant variance model is fit.
Benchmark Dose Computation.
BMR = 10% Relative risk
BMD = 1931.35
BMDL at the 95% confidence level = 1704.82
Parameter Estimates
Variable
Estimate
Default initial
parameter values
alpha
0.00687349
0.00750263
rho
n/a
0
intercept
1.19966
1.21
V
0.130345
0.13
n
18
18
k
1685.82
4451.94
Hill
BMDL
BMD
This document is a draft for review purposes only and does not constitute Agency policy.
C-19 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std Dev
Scaled Resid
0
10
1.21
1.2
0.0822
0.0829
0.395
406
9
1.21
1.2
0.096
0.0829
0.374
825
10
1.18
1.2
0.0791
0.0829
-0.75
1643
10
1.25
1.25
0.111
0.0829
-0.00000196
3274
10
1.34
1.33
0.0538
0.0829
0.381
6369
10
1.32
1.33
0.0885
0.0829
-0.381
2 Likelihoods of Interest
Model
Log(likelihood)
# Param's
AIC
A1
117.992549
7
-221.985098
A2
120.600135
12
-217.20027
A3
117.992549
7
-221.985098
fitted
117.41244
4
-226.82488
R
105.528775
2
-207.05755
3 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1
30.1427
10
0.0008118
Test 2
5.21517
5
0.3902
Test 3
5.21517
5
0.3902
Test 4
1.16022
3
0.7626
This document is a draft for review purposes only and does not constitute Agency policy.
C-20 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-8. Summary of BMD modeling results for absolute kidney weight in
2 female F344 rats exposed to tert-butanol via inhalation for 6 hr/d, 5d/wk for
3 13 weeks (NTP. 1997): BMR = 10% relative deviation from the mean
Model3
Goodness of fit
BMCiord
(mg/m3)
BMCLiord
(mg/m3)
Basis for model selection
p-value
AIC
Exponential (M2)
Exponential (M3)b
0.0378
-261.52
14500
7713
No model adequately fit the data.
Exponential (M4)
0.533
-267.48
error0
0
Exponential (M5)
0.374
-265.71
error0
0
Hill
0.227
-265.57
error0
error0
Power
0.0392
-261.61
14673
7678
Polynomial 3°d
Polynomial 2°e
Linear
0.0274
-261.61
14673
7678
Polynomial 5°
0.0274
-261.61
14673
7569
Polynomial 4°
0.0274
-261.61
14673
7674
aModeled variance case presented (BMDS Test 2 p-value = 1.90E-04, BMDS Test 3 p-value = 0.374), no model
was selected as a best-fitting model.
bFor the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
CBMC or BMCL computation failed for this model.
dFor the Polynomial 3° model, the b3 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
eFor the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
Note: Graphs of the better fitting models are provided for illustration.
4
This document is a draft for review purposes only and does not constitute Agency policy.
C-21 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
0
1
Supplemental Information—tert-Butyl Alcohol
Hill Model
1000
4000
10:32 04/30 2014
Figure C-7. Plot of mean response by concentration, with fitted curve for Hill
model for absolute kidney weight in female F344 rats exposed to tert-butanol
via inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP. 1997): BMR = 10%
relative deviation from the mean; concentration shown in mg/m3.
Power Model with 0.95 Confidence Level
Power
BMDL
10:32 04/30 2014
6000 8000
dose
Figure C-8. Plot of mean response by concentration, with fitted curve for
Power model for absolute kidney weight in female F344 rats exposed to tert-
butanol via inhalation for 6 hr/d, 5d/wk for 13 weeks (NTP. 1997): BMR =
10% relative deviation from the mean; concentration shown in mg/m3.
This document is a draft for review purposes only and does not constitute Agency policy.
C-22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Supplemental Information—tert-Butyl Alcohol
C.1.2. Cancer Endpoints
C.1.2.1.Z)ataSets
The cancer data sets selected for dose-response modeling are summarized in Table C-9. In
all cases, administered exposure was used in modeling the response data. Due to the significant
difference in survival in the high-dose male mice compared with the concurrent control, the Poly-3
procedure (Bailer and Portier. 19881 for adjusting tumor incidence rates for intercurrent mortality
was used. The procedure is based on the observation that the cumulative incidence of tumors tends
to increase with time raised to the second through the fourth powers for a large proportion of
cases. In the Poly-3 procedure, for a study of T weeks' duration, an animal that is removed from the
study after t weeks (t < T) without a specified type of tumor of interest is given a weight of (t/T)3.
An animal that survives until the terminal sacrifice at T weeks is assigned a weight of (T/T)3 = 1. An
animal that develops the specific type of tumor of interest obviously lived long enough to develop
the tumor, and is assigned a weight of 1. The Poly-3 tumor incidence, adjusted for intercurrent
mortality up to time T, is the number of animals in a dose group with the specified type of tumor
divided by the sum of the weights (the effective number of animals at risk). The tumor incidences,
adjusted using this procedure, also are provided in Table C-9.
C.1.2.2. Model Fit
The multistage model was fit to the cancer data sets. Model coefficients were restricted to
be non-negative (beta values > 0) to estimate a monotonically increasing function. Each model was
fit to the data using the maximum likelihood method, and was tested for goodness of fit using a chi-
square goodness-of-fit test (x2 p-value < 0.051 indicates lack of fit). Other factors were used to
assess model fit, such as scaled residuals, visual fit, and adequacy of fit in the low dose region and
near the BMR.
For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
estimated by the profile likelihood method) and AIC value were used to select a best-fit model from
among the models exhibiting adequate fit For the NTP Q9951 and Hard etal. f20111 data, models
were run with all doses included, as well as with the high dose dropped. Dropping the high dose
resulted in a better fit to the data. Including the high dose caused the model to overestimate the
control.
*A significance level of 0.05 is used for selecting cancer models because the model family (multistage) is
selected a priori fU.S. EPA. 20001.
This document is a draft for review purposes only and does not constitute Agency policy.
C-23 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-9. Cancer endpoints selected for dose-response modeling for tert-
2 butanol
Endpoint/Study
Species/Sex
Doses and effect data
Thyroid
Thyroid follicular cell
adenoma
NTP (1995)
B6C3Fi
mice/female
Dose (mg/kg-d)
0
510
1,020
2,110
Incidence/Total
2/58
3/60
2/59
9/59
Thyroid follicular cell
adenoma
NTP (1995)
B6C3Fi
mice/male
Dose (mg/kg-d)
0
540
1,040
2,070
Incidence/Total
1/60
0/59
4/59
2/60
lncidence/Poly-3
adjusted Total
1/50
0/50
4/51
2/35
Kidney3
Renal tubule adenoma or
carcinoma
NTP (1995)
Rat (F344) /
Male
Dose (mg/kg-d)
0
90
200
420
Incidence /Total
8/50
13/50
19/50
13/50
Renal tubule adenoma or
carcinoma
NTP (1995)
Rat (F344) /
Male
Incidence /Total
8/50
13/50
19/50
13/50
Renal tubule adenoma or
carcinoma
NTP (1995)
Rat (F344) /
Male
Incidence /Total
8/50
13/50
19/50
13/50
Renal tubule adenoma or
carcinoma; Hard
reanalysis
NTP (1995);Hard et al.
(2011)
Rat (F344) /
Male
Dose (mg/kg-d)
0
90
200
420
Incidence /Total
4/50
13/50
18/50
12/50
Renal tubule adenoma or
carcinoma; Hard
reanalysis
NTP (1995);Hard et al.
(2011)
Rat (F344) /
Male
Incidence /Total
4/50
13/50
18/50
12/50
Renal tubule adenoma or
carcinoma; Hard
reanalysis
NTP (1995);Hard et al.
(2011)
Rat (F344) /
Male
Incidence /Total
4/50
13/50
18/50
12/50
3 aEndpoint presented if kidney tumors are acceptable for quantitation.
This document is a draft for review purposes only and does not constitute Agency policy.
C-24 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Table C-10. Summary of the oral slope factor derivations
Tumor
Species/Sex
Selected Model
BMR
BMD
(mg/kg-
d)
POD=
BMDL
(mg/kg-d)
BMDLhed3
(mg/kg-d)
Slope factor13
(mg/kg-day)1
Thyroid
Thyroid follicular cell
adenoma
Female
B6C3F1
mouse
3° Multistage
10%
2002
1437
201
5 x 10"4
Kidneyc
Renal tubule
adenoma or
carcinoma
Male F344
rat; dose as
administered
1° Multistage
(high dose
dropped)
10%
70
42
10.1
1 x 10"2
Renal tubule
adenoma or
carcinoma [Hard et
al. (2011) reanalvsisl
Male F344
rat; dose as
administered
1° Multistage
(high dose
dropped)
10%
54
36
8.88
1 x 10"2
2 aHED PODs were calculated using BW3/4scaling (U.S. EPA, 2011).
3 bHuman equivalent slope factor = 0.1/BMDLiohed.
4 Alternative endpoint if kidney tumors are acceptable for quantitation.
5
This document is a draft for review purposes only and does not constitute Agency policy.
C-25 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Supplemental Information—tert-Butyl Alcohol
C.1.2.1. Modeling Results
Table C-ll. Summary of BMD modeling results for thyroid follicular cell
adenomas in female B6C3F1 mice exposed to tert-butanol in drinking water
for 2 years fNTP. 19951: BMR = 10% extra risk
Model3
Goodness of fit
BMDio%c
(mg/kg-d)
BMDLio%c
(mg/kg-d)
Basis for model selection
p-value
AICb
Three
0.75
113.665
2002
1437
Multistage 3° was selected on the basis of
the lowest AIC with all BMDL values for
fitting models being sufficiently close
(BMDLs differed by less than 3-fold).
Two
0.36
115.402
2186
1217
One
0.63
114.115
1987
1378
aSelected (best-fitting) model shown in boldface type.
bAIC = Akaike Information Criterion.
Confidence level = 0.95.
0.3
Multistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
0.25
0.2
0.15
0.1
0.05
15:22 05/13 2011
BMDL
500
1000
dose
1500
BMP
2000
Figure C-9. Plot of incidence by dose, with fitted curve for Multistage 3° model
for thyroid follicular cell adenomas in female B6C3F1 mice exposed to tert-
butanol in drinking water for 2 years fNTP. 19951: BMR = 10% extra risk; dose
shown in mg/kg-d.
Multistage Cancer Model. (Version: 1.9; Date: 05/26/2010)
Input Data File: M:\NCEA t-Butanol\BMD modeling\BMDS Output\29 NTP 1995b_Thyroid
folluclar cell andenoma, female mice_MultiCanc3_10.(d)
Gnuplot Plotting File: M:\NCEA t-Butanol\BMD modeling\BMDS Output\29 NTP
1995b_Thyroid folluclar cell andenoma, female mice_MultiCanc3_10.pit
Fri May 13 15:22:18 2011
This document is a draft for review purposes only and does not constitute Agency policy.
C-26 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Supplemental Information—tert-Butyl Alcohol
[notes]
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2-beta3*doseA3)]
The parameter betas are restricted to be positive
Dependent variable = Incidence
Independent variable = Dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0347373
Beta (1) = 0
Beta (2) = 0
Beta (3) = 1.36917e-Gll
Asymptotic Correlation Matrix of Parameter Estimates
user,
Limit
( *** The model parameter(s) -Beta(l) -Beta(2)
have been estimated at a boundary point, or have been specified by the
and do not appear in the correlation matrix )
Background Beta(3)
Background 1 -0.53
Beta (3) -0.53 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Background 0.0361209 * * *
Beta (1) 0 * * *
Beta (2) 0 * * *
Beta (3) 1.313Gle-Gll * * *
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
#
Log (likelihood)
-54.5437
-54.8326
-58.5048
Param's
4
2
1
Deviance Test d.f.
0.577881
7 . 92235
P-value
0.7491
0.04764
AIC:
113.665
This document is a draft for review purposes only and does not constitute Agency policy.
C-27 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Supplemental Information—tert-Butyl Alcohol
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.0361
2 . 095
2 .000
58
-0.067
510.0000
0.0378
2.268
3 .000
60
0.496
1020.0000
0.0495
2 . 918
2 .000
59
-0.551
2110.0000
0.1480
8 .730
9.000
59
0.099
Chi^2 = 0.56 d.f. = 2 P-value = 0.7544
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 2 0 02.03
BMDL = 14 3 6.6 9
BMDU = 3 8 02.47
Taken together, (1436.69, 3802.47) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 6.96G43e-GG5
This document is a draft for review purposes only and does not constitute Agency policy.
C-28 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Supplemental Information—tert-Butyl Alcohol
Table C-12. Summary of BMD modeling results for thyroid follicular cell
adenomas or carcinomas in male B6C3F1 mice exposed to tert-butanol in
drinking water for 2 years (NTP. 1995): BMR = 5% extra risk
Model3
Goodness of fit
BMDsr.
(mg/kg-d)
BMDLs%c
(mg/kg-d)
Basis for model selection
p-value
AICb
One, Two,
Three
0.202
61.6
1788
787
Multistage 1° was selected. Only form of
multistage that resulted; fit adequate.
aSelected (best-fitting) model shown in boldface type.
bAIC = Akaike Information Criterion.
Confidence level = 0.95.
Multistage Cancer Model, with BMR of 5% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
0.2
0.15
0.05
0
Multistage Cancer
Linear extrapolation
1000
dose
11:02 06/05 2015
Figure C-10. Plot of incidence by dose, with fitted curve for Multistage 1°
model for thyroid follicular cell adenomas or carcinomas in male B6C3F1 mice
exposed to tert-butanol in drinking water for 2 years (NTP. 1995): BMR = 5%
extra risk; dose shown in mg/kg-d.
Multistage Model. (Version: 3.4; Date: 05/02/2014)
Input Data File: C:/Users/KHOGAN/BMDS/BMDS260/Data/msc_TBA NTP1995 MMthyroid tumors
poly3_Mscl-BMR05.(d)
Gnuplot Plotting File: C:/Users/KHOGAN/BMDS/BMDS260/Data/msc_TBA NTP1995 MMthyroid
tumors poly3_Mscl-BMR05.pit
Fri Jun 05 11:02:14 2015
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl)]
This document is a draft for review purposes only and does not constitute Agency policy.
C-29 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Supplemental Information—tert-Butyl Alcohol
The parameter betas are restricted to be positive
Dependent variable = Effect
Independent variable = Dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 500
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0164855
Beta (1) = 2.58163e-005
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.56
Beta(1) -0.56 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0.0149284 0.0144833 -0.0134584 0.0433151
Beta(1) 2.86952e-005 1.99013e-005 -1.03105e-005 6.7701e-005
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-26.5891
-28.808
-29.8255
Param's
4
2
1
Deviance Test d.f.
4.43785
6.47273
P-value
0.1087
0.09074
AIC:
61.616
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
540.0000
1040.0000
2070.0000
0.0149
0.0301
0.0439
0.0717
0.746
1.504
2 . 238
2 . 511
1.000
0 .000
4 .000
2 .000
50.000
50.000
51.000
35.000
0.296
-1.245
1.204
-0.335
ChiA2
3 .20
d.f.
P-value
0.2019
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
This document is a draft for review purposes only and does not constitute Agency policy.
C-30 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Supplemental Information—tert-Butyl Alcohol
Confidence level = 0.95
BMD = 1787.52
BMDL = 7 87.153
BMDU did not converge for BMR = 0.050000
BMDU calculation failed
BMDU = Inf
This document is a draft for review purposes only and does not constitute Agency policy.
C-31 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Supplemental Information—tert-Butyl Alcohol
Table C-13. Summary of BMD modeling results for thyroid follicular cell
adenomas or carcinomas in male B6C3F1 mice exposed to tert-butanol in
drinking water for 2 years, high dose omitted (NTP. 1995): BMR = 5% extra
risk
Model3
Goodness of fit
BMD5%
(mg/kg-d)
BMDI_5%c
(mg/kg-d)
Basis for model selection
p- value
AIC b
One stage
0.105
46.0
1341
538
Multistage 2° was selected based on lowest AIC.
Two stage
0.174
44.9
1028
644
aSelected (best-fitting) model shown in boldface type.
bAIC = Akaike Information Criterion.
Confidence level = 0.95.
Multistage Cancer Model, with BMR of 5% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
0.2
0.15
0.05
0
Multistage Cancer
Linear extrapolation
11:18 06/05 2015
Figure C-ll. Plot of incidence by dose, with fitted curve for Multistage 2°
model for thyroid follicular cell adenomas or carcinomas in male B6C3F1 mice
exposed to tert-butanol in drinking water for 2 years, high dose omitted (NTP.
1995): BMR = 5% extra risk; dose shown in mg/kg-d.
Multistage Model. (Version: 3.4; Date: 05/02/2014)
Input Data File: C:/Users/KHOGAN/BMDS/BMDS260/Data/msc_TBA NTP1555 MMthyroid tumors
poly3 -h_Msc2-BMR05.(d)
Gnuplot Plotting File: C:/Users/KHOGAN/BMDS/BMDS260/Data/msc_TBA NTP1555 MMthyroid
tumors poly3 -h_Msc2-BMR05.pit
Fri Jun 05 11:18:05 2015
BMDS Model Run
This document is a draft for review purposes only and does not constitute Agency policy.
C-32 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Supplemental Information—tert-Butyl Alcohol
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2)]
The parameter betas are restricted to be positive
Dependent variable = Effect
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 500
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.00347268
Beta (1) = 0
Beta (2) = 6.65923e-GG8
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background Beta(2)
Background 1 -0.34
Beta (2) -0.34 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable
Background
Beta (1)
Beta (2)
Estimate
0.011558
0
4 . 84 624e-008
Std. Err.
0.0114911
NA
3.15009e-008
Lower Conf. Limit
-0.010964
-1.327 8le-0 0 8
Upper Conf. Limit
0.0340801
1.10203e-007
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
Full model
Fitted model
Reduced model
Log(likelihood) #
-18.
-20.
-21.
9229
4481
9555
Param's
3
2
1
Deviance Test d.f.
3.05031
6.0651
P-value
0.08072
0.04819
AIC:
44 . 8962
Goodness of Fit
This document is a draft for review purposes only and does not constitute Agency policy.
C-33 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Supplemental Information—tert-Butyl Alcohol
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.0116
0 . 578
1.000
50.000
0 . 558
540.0000
0.0254
1.271
0 .000
50.000
-1.142
1040.0000
0.0620
3 .164
4 .000
51.000
0.485
ChiA2 = 1.85 d.f. = 1 P-value = 0.1735
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 1028.79
BMDL = 644.475
BMDU did not converge for BMR = 0.050000
BMDU calculation failed
BMDU = 14 661.6
Cancer Slope Factor = 7.75825e-005
This document is a draft for review purposes only and does not constitute Agency policy.
C-34 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information—tert-Butyl Alcohol
Table C-14. Summary of BMD modeling results for renal tubule adenoma or
carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
years modeled with administered dose units and including all dose groups
fNTP. 19951: BMR = 10% extra risk
Model3
Goodness of fit
BMDiopct (mg/kg-d)
BMDLiopct (mg/kg-
d)
Basis for model
selection
P-
value
Scaled residuals
AIC
Three
Two
0.0806
-0.989, 0.288,1.719,
and -1.010
233.94
294
118
Multistage 2° is
selected as the most
parsimonious model
of adequate fit.
One
0.0806
-0.989, 0.288, 1.719,
and -1.010
233.94
294
errorb
aSelected model in bold.
bBMD or BMDL computation failed for this model.
Multistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
BMD
BMD
10:57 04/30 2014
Figure C-12. Plot of incidence by dose, with fitted curve for Multistage 2°
model for renal tubule adenoma or carcinoma in male F344 rats exposed to
tert-butanol in drinking water for 2 years modeled with administered dose
units and including all dose groups (NTP. 1995): BMR = 10% extra risk; dose
shown in mg/kg-d.
This document is a draft for review purposes only and does not constitute Agency policy.
C-35 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert-Butyl Alcohol
Multistage Cancer Model. (Version: 1.9; Date: 05/26/2010)
The form of the probability function is: P[response] = background + (l-background)*[l-EXP(-
betal*doseAl-beta2*doseA2...)]
The parameter betas are restricted to be positive
Benchmark Dose Computation.
BMR = 10% Extra risk
BMD = 293.978
BMDL atthe 95% confidence level = 117.584
BMDU atthe 95% confidence level = 543384000
Taken together, (117.584, 543384000) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.000850453
Parameter Estimates
Variable
Estimate
Default initial
parameter values
Background
0.217704
0.2335
Beta(l)
0.000358397
0.000268894
Beta(2)
0
0
Analysis of Deviance Table
Model
Log( likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-112.492
4
Fitted model
-114.97
2
4.95502
2
0.08395
Reduced
model
-115.644
1
6.30404
3
0.09772
AIC: = 233.94
Goodness of
Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.2177
10.885
8
50
-0.989
90
0.2425
12.127
13
50
0.288
200
0.2718
13.591
19
50
1.719
420
0.327
16.351
13
50
-1.01
ChiA2 = 5.04 d.f = 2 P-value = 0.0806
This document is a draft for review purposes only and does not constitute Agency policy.
C-36 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information—tert-Butyl Alcohol
Table C-15. Summary of BMD modeling results for renal tubule adenoma or
carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
years modeled with administered dose units and excluding high-dose group
fNTP. 19951: BMR = 10% extra risk.
Model3
Goodness of fit
BMDiopct (mg/kg-d)
BMDLiopct
(mg/kg-d)
Basis for model
selection
P-
value
Scaled residuals
AIC
Two
N/Ab
0.000, -0.000, and -
0.000
173.68
75.6
41.6
Multistage 1° was
selected as the only
adequately-fitting
model available
One
0.924
0.031, -0.078, and
0.045
171.69
70.1
41.6
aSelected model in bold.
bNo available degrees of freedom to calculate a goodness of fit value.
Multistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
BMDL
100
dose
11 :02 04/30 2014
Figure C-13. Plot of incidence by dose, with fitted curve for Multistage 1°
model for renal tubule adenoma or carcinoma in male F344 rats exposed to
tert-butanol in drinking water for 2 years modeled with administered dose
units and excluding high-dose group (NTP. 1995): BMR = 10% extra risk.; dose
shown in mg/kg-d.
This document is a draft for review purposes only and does not constitute Agency policy.
C-37 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert-Butyl Alcohol
Multistage Cancer Model. (Version: 1.9; Date: 05/26/2010)
The form of the probability function is: P[response] = background + (l-background)*[l-EXP(-
betal*doseAl-beta2*doseA2...)]
The parameter betas are restricted to be positive
Benchmark Dose Computation.
BMR = 10% Extra risk
BMD = 70.1068
BMDL at the 95% confidence level = 41.5902
BMDU at the 95% confidence level = 203.311
Taken together, (41.5902, 203.311) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.00240441
Parameter Estimates
Variable
Estimate
Default initial
parameter values
Background
0.158399
0.156954
Beta(l)
0.00150286
0.0015217
Analysis of Deviance Table
Model
Log( likelihood
)
# Param's
Deviance
Test d.f.
p-value
Full model
-83.8395
3
Fitted model
-83.8441
2
0.00913685
1
0.9238
Reduced
model
-86.9873
1
6.29546
2
0.04295
AIC: = 171.688
Goodness of
Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.1584
7.92
8
50
0.031
90
0.2649
13.243
13
50
-0.078
200
0.3769
18.844
19
50
0.045
ChiA2 = 0.01 d.f = 1 P-value = 0.9239
This document is a draft for review purposes only and does not constitute Agency policy.
C-38 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1
2
3
4
5
6
7 Table C-17. Summary of BMD modeling results for renal tubule adenoma or
8 carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
9 years modeled with administered dose units and excluding high-dose group;
10 re-analyzed data fHard et al.. 2011: NTP. 19951: BMR = 10% extra risk
11
Model3
Goodness of fit
BMDiopct (mg/kg-d)
BMDLiopct
(mg/kg-d)
Basis for model
selection
P-
value
Scaled residuals
AIC
Two
One
0.572
-0.141, 0.461, and -
0.296
154.84
54.2
36.3
Multistage 1° was
selected as the most
parsimonious model
of adequate fit.
aSelected model in bold.
Table C-16. Summary of BMD modeling results for renal tubule adenoma or
carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
years modeled with administered dose units and including all dose groups;
reanalyzed data fHard et al.. 2011: NTP. 19951: BMR = 10% extra risk
Model3
Goodness of fit
BMDiopct (mg/kg-d)
BMDLiopct
(mg/kg-d)
Basis for model
selection
P-
value
Scaled residuals
AIC
Three
Two
One
0.0117
-1.476, 1.100, 1.855,
and-1.435
218.68
184
94.8
No model fit the
data.
aNo model was selected as a best-fitting model.
This document is a draft for review purposes only and does not constitute Agency policy.
C-39 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert-Butyl Alcohol
Multistage Cancer Model with 0.95 Confidence Level
0.5
0.4
o
o
as
ul 0.2
0.1
0
0 50 100 150 200
dose
11 :05 04/30 2014
Figure C-14. Plot of incidence by dose, with fitted curve for Multistage 1°
model for renal tubule adenoma or carcinoma in male F344 rats exposed to
tert-butanol in drinking water for 2 years modeled with administered dose
units and excluding high-dose group; re-analyzed data fHard et al.. 2011: NTP.
1995): BMR = 10% extra risk; dose shown in mg/kg-d.
Multistage Cancer Model. (Version: 1.9; Date: 05/26/2010)
The form of the probability function is: P[response] = background + (l-background)*[l-EXP(-
betal*doseAl-beta2*doseA2...)]
The parameter betas are restricted to be positive
Benchmark Dose Computation.
BMR = 10% Extra risk
BMD = 54.1642
BMDL atthe 95% confidence level = 36.3321
BMDU atthe 95% confidence level = 101.125
Taken together, (36.3321,101.125) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.00275239
Parameter Estimates
Variable
Estimate
Default initial
parameter values
Background
0.0855815
0.0981146
Beta(l)
0.00194521
0.00179645
Analysis of Deviance Table
Model
Log( likelihood)
# Param's
Deviance
Test d.f.
p- value
Full model
-75.2622
3
BMDL
BMD
This document is a draft for review purposes only and does not constitute Agency policy.
C-40 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Supplemental Information—tert-Butyl Alcohol
Fitted model
-75.4201
2
0.315716
1
0.5742
Reduced
model
-81.4909
1
12.4574
2
0.001972
AIC: = 154.84
Goodness of
Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.0856
4.279
4
50
-0.141
90
0.2324
11.622
13
50
0.461
200
0.3803
19.015
18
50
-0.296
ChiA2 = 0.32 d.f = 1 P-value = 0.5715
This document is a draft for review purposes only and does not constitute Agency policy.
C-41 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Supplemental Information—tert-Butyl Alcohol
APPENDIX D. SUMMARY OF PUBLIC COMMENTS
AND EPA's DISPOSITION
The Toxicological Review of tert-Butyl alcohol (tert-Butanol) was released for a 60-day
public comment period on May 16, 2016. Public comments on the assessment were submitted to
EPA by:
• Japan Petroleum Energy Center (posted June 24, 2016);
• Exponent, Inc. on behalf of LyondellBasell (posted June 24, 2016);
• Samuel M. Cohen (posted July 7, 2016);
• Lawrence H. Lash (posted June 24, 2016);
• LyondellBasell (posted June 24, 2016 and July 19, 2016);
• American Chemistry Council (posted July 7, 2016);
• Tox-Logic Consulting on behalf of ExxonMobil Biomedical Sciences (posted July 19,
2016);
• Tox Strategies on behalf of LyondellBasell (posted June 24, 2016); and
• American Petroleum Institute (posted July 19, 2016).
A summary of major public comments provided in these submissions and EPA's response to
these comments are provided in the sections that follow. The comments have been synthesized and
paraphrased. Because several commenters often covered the same topic, the comment summaries
are organized by topic. Editorial changes and factual corrections offered by public commenters
were incorporated in the document as appropriate and are not discussed further. All public
comments provided were taken into consideration in revising the draft assessment prior to
releasing for external peer review.
Comments Related to the Preface, Preamble and Executive Summary
Comment [LyondellBasell]: The Salazar etal. (2015) model should be verified. This includes the
model structure, code, and data sets used.
EPA Response: This peer-review draft uses a PBPK model based on Borghoff etal. (2016).
This document is a draft for review purposes only and does not constitute Agency policy.
D-l DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
Comment [LyondellBasell]: The Executive Summary does not adequately capture:
1) The key uncertainties and high degree of conservatism associated with selecting rat kidney
transitional epithelial hyperplasia as the key response for the RfD derivation in that this response is
a recognized element of CPN and thus not relevant to human risk;
2) Uncertainties regarding the excessively high dose(s) used in the mouse thyroid tumor
assessment, which exceed both the EPA and OECD test guidance for selection of a Limit Dose and
other EPA and OECD guidance addressing the limitations of toxicity responses observed at dose
levels saturating metabolic saturation with resulting nonlinear toxicokinetics of TBA; and
3) An acknowledgment that the oral SF should be clearly annotated with the conclusion that the
overall "suggestive evidence" of TBA carcinogenicity does not allow for its use in quantitative
human risk analyses.
EPA Response: The previous Executive Summary did not include the commenter's points because
they had not been conclusions of the public-comment draft The Executive Summary has been
revised to reflect the conclusions of this peer-review draft, and the IRIS program has proposed a
charge question for the SAB/CAAC to comment on whether the Executive Summary appropriately
presents the major conclusions of the assessment.
Comments Related to the Literature Search and Study Quality
Comment [LyondellBasell]: Despite statements in preamble, there is no evidence that toxicity data
from TBAc, MTBE or ETBE was robustly searched, despite clear toxicokinetic bridging from these
studies to TBA.
EPA Response: The literature search was focused on tert-butanol as the primary chemical of
interest Toxicity reported on chemicals extensively metabolized to tert-butanol (i.e., TBAc, MTBE,
and ETBE) are summarized in 1.1.4, and cross-compound comparisons for non-cancer and cancer
effects are discussed in Sections 1.3.1 and 1.3.2, respectively.
Comment [LyondellBasell]: Clarity is needed regarding how the primary references were selected
for "Sources of Health Effects Data" versus "Supporting Studies" and consistent application of
decision criteria.
EPA Response: Table LS-3, in the row labeled "Outcome" provides a list of the health effects that
cause a study to be considered a "source of health effects data." Other pertinent studies, including
This document is a draft for review purposes only and does not constitute Agency policy.
D-2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
mechanistic studies, are considered "supporting studies."
Comments Related to Data Presented in Evidence Tables
Comment [LyondellBasell]: Clarification is needed on what determined the endpoints selected for
inclusion in evidence tables. Also, the exclusion of mechanistic key events (e.g., hyaline droplet
accumulation) is not consistent with the intent of the EPA IRIS program and does not lead to the
development of useful hazard evaluation.
EPA Response: Primary health effects information are included in evidence tables. Hyaline droplets
do not constitute primary health effects information, but it is included in a table of mechanistic
events used to evaluate the a2u-globulin mode of action (Table 1-4).
Comments Related to Kidney Effects
Comment [Dr. Bogen on behalf of American Petroleum Institute]: tert- Butanol induced male rat
kidney tumors are not relevant to humans because tert-butanol-associated male rat kidney tumors
were exacerbated by a CPN mode of action that is specific to rats. CPN has no human counterpart
and is not considered relevant for human health risk assessment.
EPA Response: CPN is a common and well-established constellation of age-related lesions in the
kidney of rats, and there is no known counterpart to CPN in aging humans. However, CPN is not a
specific diagnosis on its own. These individual lesions or processes (tubular
degeneration/regeneration and dilatation, glomerular sclerosis and atrophy, interstitial fibrosis
and inflammation, etc.) could certainly occur in a human kidney. Because they happen to occur as a
group in the aged rat kidney does not necessarily make them rat-specific individually if there is a
treatment effect for one or more of them. In addition, exacerbation of one or more of these
processes likely reflects some type of cell injury/cytotoxicity, which is relevant to the human
kidney. One potential confounder is the alpha-2u globulin nephropathy in males which could also
exacerbate CPN but would not be considered relevant to human risk.
In a recent draft proposal for public comment (2015), FDA used CPN in their calculation of PDEs for
MIBK f http://www.fda.gov/ucm/groups/fdagov-public/@fdagov-drugs-
gen/documents/document/ucm467089.pdfl Similarly, EPA considered the rat kidney tumors to be
relevant to human health risk assessment.
Comment [Dr. Cohen, Dr. Hard, and LyondellBasell]: All the kidney changes identified in the
assessment associated with tert-butanol exposure are associated with a2u-globulin nephropathy
or/and CPN, except for the cortical-medullary calcification that is common in F344 rats. Cortical-
This document is a draft for review purposes only and does not constitute Agency policy.
D-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
medullary calcification also is not relevant to humans based on a long series of articles published
over the past 3 0 years or more. They asserted that transitional epithelial hyperplasia is not a valid
endpoint for dose response because the lesion is a component of CPN, and that tert-butanol male
rat renal tumors are adequately explained by alpha2u-globulin nephropathy combined with
advanced CPN.
EPA Response: Section 1.2.1 shows that, although renal tumors are correlated with CPN in male
rats, this correlation is weak in female rats and so the renal tumors cannot be attributed to solely to
CPN. The peer-review draft does not consider cortical-medullary calcification relevant to humans,
nor does it use TEH as an endpoint for dose-response assessment.
Comment [LyondellBasell]: Hyaline droplet accumulation was not given adequate importance and
more discussion of the hyaline droplet pathology should be included.
EPA Response: Additional text describing hyaline droplets would not affect overall conclusion of
the MOA nor would it serve to increase the clarity of the decision.
Comment [LyondellBasell]: An increase in kidney weights is a non-specific endpoint and should
not be a candidate for potential use in BMD modeling.
EPA Response: Changes in kidney weights are a sensitive, non-specific endpoint that is often used
for dose-response modeling. Although CPN and a2u-globulin nephropathy are occurring and can
influence kidney weights, MOA analysis determined that they are only partially responsible for the
observed kidney effects. Therefore, kidney weight changes are relevant for human health risk
assessment.
Comment [LyondellBasell]: Suppurative inflammation in the rat kidney is a result of bacterial
infection, and therefore this lesion cannot be used as an indicator of chemically-induced renal
toxicity for the purposes of characterizing human hazard associated with tert-butanol. The
suppurative inflammation would have been associated with either advanced CPN or an ascending
infection probably related to urinary tract calculi, or both processes.
EPA Response: Suppurative inflammation is often but not always associated with bacterial
infection. The bacteria may not be apparent on H&E staining though; culture and/or special
staining are often needed. Suppurative inflammation may be part of the spectrum of CPN lesions
and additional analysis indicates that suppurative inflammation is correlated with CPN in females.
According to p39S of INHAND document: "Solitary proximal tubules affected with
microabscessation often occur in advanced stages of CPN, in which setting they need not be
This document is a draft for review purposes only and does not constitute Agency policy.
D-4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
diagnosed separately (Frazier et al., 2012)." The text in the Executive Summary, Section 1.2.1, and
1.3.1 of the assessment has been clarified to reflect this information.
Comments Related to Thyroid Effects
Regarding the potential for high dose effects
Comment [Dr. Bus on behalf of LyondellBasell]: The dose levels at which thyroid tumors were
identified in male and female mice was a major concern, and discussion of the toxicity and
carcinogenicity findings at these high exposure concentrations would be informative, proposing
that the MOA most likely operates under nonlinear kinetics. If the NTP (1995) tert- butanol bioassay
was designed according to current dose selection guidance of EPA and OECD, mouse thyroid tumors
likely would not have emerged as a significant cancer concern. In addition to referencing passages
from OECD Guidance Document 116 (2012) describing the importance of considering rodent
toxicokinetic or ADME data, and suggesting that the presence of toxicokinetic inflection point could
be used as a surrogate for traditional target organ effects, the EPA Cancer Guidelines (2005) were
referenced, noting that changes in toxicokinetics with increasing dose may result "...in important
differences between high and low dose levels in disposition of the agent or generation of its active
forms. These studies play an important role in providing a rationale for dose selection in
carcinogenicity studies."
EPA Response: The discussion of the thyroid follicular cell tumors (adenomas and carcinoma) as
well as the follicular cell hyperplasias, considered by both NTP and EPA to be pre-neoplastic lesions
and thus not suitable candidates for non-cancer reference value derivation, is presented in Section
1.2.2, while considerations and uncertainties pertaining to dose-response evaluation are discussed
in Sections 2.3.2 and 2.3.4, respectively. As discussed in Section 1.2.2, incidence of thyroid follicular
cell hyperplasia was significantly elevated in all male mice treatment groups (i.e. doses of 540,1040
or 2070 mg/kg-d), in both mid and high-dose female mice groups (i.e. 1020 and 2110 mg/kg-d),
and was increased in low-dose female mice as well (510 mg/kg-d). As the hyperplasia was
considered to be a pre-neoplastic lesion, and would be a key precursor step in the progression of
initiated thyroid follicular cells towards neoplasia, the presence of increased hyperplasia and/or
neoplasia incidence in all treatment groups in both sexes of mice does not support the assertion
that a kinetic non-linearity exists which is responsible for a tumor-relevant response within the
experimental treatment range (i.e. 510 - 2110 mg/kg-d). Furthermore, no MOA was identified for
thyroid tumorigenesis, and no mouse PBPK model is available; as such, the available information
appears insufficient to clearly describe the kinetics of mouse thyroid tumorigenesis following tert-
butanol exposure. Therefore, there is insufficient information to predict with confidence what
exposure level, if any, may result in metabolic saturation of tert-butanol oxidative metabolism in
B6C3Fi mice, especially considering the highly inducible nature of the cytochrome p450 system
following repeated substrate exposure. The mouse follicular cell thyroid tumors were determined
This document is a draft for review purposes only and does not constitute Agency policy.
D-5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
to be qualitatively and quantitatively relevant to human cancer hazard characterization following
recommendations from the EPA guidance on assessment of thyroid follicular cell tumors (EPA,
1998) and the EPA Cancer Guidelines (2005).
Comment [Dr. Bus and Dr. Borghoff on behalf of LyondellBasell]: Dose selection in these studies
did not follow recommendations of EPA and OECD testing guidelines, including long-established
limit dose guidelines of 1000 mg/kg bw/day and more recent considerations of saturated
metabolism and nonlinear toxicokinetics in selection of appropriate doses for carcinogenicity and
other test bioassays. Both male and female top dose groups reported in the NTP (1995) bioassay
(2,070 and 2,100 mg/kg-d, respectively) exceeded the Limit Dose by 2-fold, questioning dose-
relevance of thyroid tumor findings. Regarding the limit dose of 1,000 mg/kg-d, statements from a
OECD dose selection guidance were provided "A limit of 1000 mg/kg body weight/day may apply
except when human exposure indicates the need for a higher dose level to be used" (OECD, 2009),
along with a 1998 guideline document from the EPA Office of Prevention, Pesticides and Toxic
Substances (OPPTS): "The highest dose tested need not exceed 1,000 mg/kg/day." (EPA 870.4300,
Combined Chronic Toxicity/Carcinogenicity, 1998).
EPA Response: The 1998 guideline document referenced from the EPA Office of Prevention,
Pesticides and Toxic Substances (OPPTS), informs the design of chronic animal bioassays in the
context of chemical testing and assessment prioritization, but does not provide instructions
regarding the evaluation or interpretation of similar bioassays already planned or conducted, such
as the NTP rodent bioassay reporting thyroid effects in mice following 2 years of oral tert-butanol
exposure (NTP, 1995), and so referencing any specific statement from the document without noting
the larger context for which it was intended can be misleading. Furthermore, the public comment
appears to interpret "need not" as "shall not", whereas another section from the same EPA guidance
document suggests that this was not the intended interpretation, and that bioassays evaluating
doses higher than 1,000 mg/kg-d could be informative: "If a test at one dose level of at least 1,000
mg/kg body weight (expected human exposure may indicate the need for a higher dose level), using
the procedures described for this study, produces no observable toxic effects or if toxic effects
would not be expected based upon data of structurally related compounds, then a full study using
three dose levels might not be necessary." Lastly, the paragraph immediately preceding the
sentence from EPA OPPTS guidance quoted in the public comment describes the various
considerations informative to selecting dose levels for a chronic rodent bioassay, which are
analogous to the process that NTP employs. Notably, NTP does not appear to observe an arbitrary
high-dose level limit, such as 1,000 mg/kg-d, as part of recent 2-year rodent bioassay study design;
5 NTP 2-year oral exposure technical reports (TR) published since 2010 have evaluated one or
more dose levels > 1,000 mg/kg-d in mice and rats (TR-578, TR-567, TR-565, TR-562 and TR-556).
This document is a draft for review purposes only and does not constitute Agency policy.
D-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
Comment [Dr. Borghoffon behalf of LyondellBasell and Dr. Fowles on behalf of Exxon Mobil
Biomedical Sciences]: The fact that rats did not exhibit thyroid effects may likely be due to fact that
rats were administered substantially lower TBA doses relative to mice (i.e. high doses of 650
mg/kg-d in female rats versus 2,110 mg/kg-d in female mice). Reliability of the NTP (1995)
bioassay is questionable because the dose levels used for rats were significantly lower than used for
mice.
EPA Response: The assertion is not entirely correct: treatment-related pre-neoplastic and/or
neoplastic thyroid lesions were observed in mouse treatment groups exposed to 510 -2,110
mg/kg-d, an exposure range which overlaps with the upper end of rat treatment doses (90 - 650
mg/kg-d). The lowest doses administered to male and female B6C3Fi mice (540 and 510 mg/kg-d,
respectively) are comparable to the highest doses administered to male and female F344 rats (420
and 650 mg/kg-d, respectively). Therefore, the presence of treatment-associated thyroid toxicity in
mice, and the absence of thyroid toxicity in rats, is not due to rats being administered lower doses:
if rats were at least similarly sensitive, which is the general assumption (Section 1.2.2; EPA, 1998),
then thyroid effects should have been observed in the high dose rat groups.
It is unclear how the reliability of a rodent bioassay is directly related to the dose levels
employed; however, uncertainties in dose-response evaluations are discussed in Section 2.3.4. As
dose levels in NTP chronic rodent bioassays are selected based upon toxicity findings from
subchronic and acute range-finding studies, and renal toxicity from these studies was found to be
dose-limiting in rats but not mice, the higher dose-levels evaluated for mice versus rats is not
unexpected, and does not represent any flaw in study design or methodology pertaining to study
reliability.
Regarding the possible Mode of Action
Comment [Dr. Fowles on behalf of Exxon Mobil Biomedical Sciences and Dr. Bogen on behalf of
American Petroleum Institute]: tert-butanol is a weak CYP and SULT liver enzyme inducer in
female B6C3Fi mice, sharing some PB- and CAR- like induction elements (Blanck et al., 2010).
Furthermore, observed thyroid effects are likely secondary to a high dose hepatic enzyme induction
effect on thyroid hormone elimination and homeostasis in mice. While the magnitude of effects
were generally similar after 14 days in both the 2 and 20 mg/mL treatment groups (receiving 418
and 1616 mg/kg-d, respectively), early events such as the selective induction of SULT1A1, an
important enzyme regulating mouse thyroid hormone clearance rates, were consistent with the
slight reductions in T3/T4 as reported in the 14-day study, and would be expected to magnify with
time at the high dose. While agreeing that no liver pathology was reported in the NTP study in rats
or mice, as described in the tert-butanol draft, a statistically significant increase in relative liver
weight was seen in high dose male and female mice, and that a fatty change was observed in male
This document is a draft for review purposes only and does not constitute Agency policy.
D-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information—tert-Butyl Alcohol
high dose mice (but not females), the commenters concluded that TBA is not a general liver enzyme
inducer but selectively induces specific CYPs and SULT1A1, and therefore evidence of
histopathology associated with generalized liver enzyme induction would not be necessarily
expected.
The decreases in T3 andT4 levels reported by Blanck etal., (2010), along with increases in
liver enzyme levels and mRNA induction, particularly in CYP2B10, demonstrates effects which were
"virtually certain to have been associated with CAR activation". CAR-activation mediated anti-
thyroid MOA is not relevant to humans due to differences in T3/T4 hormone half-lives in humans
versus rodents.
EPA Response: As discussed in Section 1.2.2, based upon recommendations from the EPA guidance
document regarding the assessment of rodent thyroid follicular cell tumors (EPA, 1998), the
available evidence was found to be inadequate to determine if any anti-thyroid MOA was operative
in mice, including the suggested mechanism of nuclear-receptor stimulated induction of hepatic
enzyme expression and increased thyroid hormone metabolism. The conclusion of a high dose
effect is inconsistent both with the results reported in the single mechanistic study available
(Blanck et al., 2010), which noted similar decreases in female B6C3F1 mouse serum T3 and T4
levels following 14 days of exposure to either 418 or 1616 mg/kg-d, as well as with the single
chronic mouse bioassay available (NTP, 1995), which reported increased incidence of thyroid
follicular cell hyperplasia and/or neoplasia in male and female B6C3F1 mice following 2 years of
exposure to 540 - 2110 mg/kg-d. The effects on thyroid hormones in mice following short-term
exposure to > 418 mg/kg-d, and on thyroid histology following chronic exposure to > 540 mg/kg-d,
suggests that tert-butanol induces thyroid effects at the lowest doses evaluated in both studies.
Inter-species differences in thyroid hormone metabolism is discussed in the EPA thyroid follicular
cell tumor guidance (EPA, 1998), which concludes that despite these and other uncertainties,
rodent thyroid tumors should be considered relevant to human cancer hazard characterization.
As discussed in the thyroid cancer MOA analysis in Section 1.2.2, if the decreases in T3/T4
levels observed following 14 days of exposure to 418 or 1616 mg/kg-d tert-butanol would be
expected to magnify with time, then the sustained liver enzyme induction should have resulted in
some treatment-associated increase in liver histopathology (such as centrilobular hyprotrophy)
after subchronic or chronic exposures to comparable doses, i.e. 510 - 2110 mg/kg-d. However, as
noted in the public comments, no such liver effects were reported in male or female B6C3Fi mice.
Furthermore, while it is unclear whether or not a 15 or 22% increase in liver SULT1A1 mRNA levels
reported following 14 days to 418 or 1616 mg/kg-d exposure (Blanck etal., 2010) is sufficient to
increase T3/T4 catabolism, and therefore cause the decrease in serum T3/T4 levels as reported at
14 days, it does provide further support to the identification of thyroid effects associated with
thyroid carcinogenesis following exposure to low and high doses.
This document is a draft for review purposes only and does not constitute Agency policy.
D-8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
Regarding liver effects in mice following subchronic or longer exposure, the comment is
assumed to reference the increased mouse relative liver weight in the 13 week subchronic
component of the bioassay reported by NTP (1995) since a 15 month sacrifice was not collected
from mice due to higher than anticipated early mortality, and organ weights were not reported at
the terminal harvest (i.e. 2 years). Indeed, after 13 weeks of oral exposure, relative liver weight was
induced in the two highest dose groups of male B6C3F1 mice (8210 and 3240 mg/kg-d), and in the
highest dose group of female B6C3F1 mice (11620 mg/kg-d). However, in the groups receiving
administered doses similar to those evaluated by Blanck et al. (2010), i.e. 418 and 1616 mg/kg-d,
there was no significant change in relative liver weights in male mice exposed to 640 and 1590
mg/kg-d, or in female mice exposed to 820 mg/kg-d, while the relative liver weights in female mice
exposed to 1660 mg/kg-d were significantly decreased by treatment, not increased. Notably, the
relative liver weights of male and female F344 rats were increased following 13 weeks of exposure
to 290 - 3620 mg/kg-d, and were significantly increased after 15 months of exposure in the high
dose males and females administered 420 and 650 mg/kg-d, respectively. However, there were no
liver or thyroid histopathological effects associated with tert-butanol exposure in rats, so changes
in relative liver weight does not appear to be linked to either liver or thyroid pathology in either
rats or mice following tert-butanol exposure.
The fatty change noted by the public comment was observed in the liver of only high dose
male mice after 2 years of exposure, but no such effect was present in the livers of female mice,
which were more sensitive to the thyroid toxicity induced by chronic tert-butanol exposure (80% of
high dose females had thyroid lesions versus 30% of high dose males; NTP, 1995). Also, thyroid
hyperplasia was induced in male mice in all treatment groups, despite fatty liver being only induced
in the high dose group. Because of this, the induction of fatty change in the livers of males was
considered to not be related to thyroid toxicity, as discussed in Section 1.2.2.
Comment [Dr. Fowles on behalf of Exxon Mobil Biomedical Sciences]: The lack of a statistically
significant elevation in TSH in the short-term study does not invalidate the MOA, as TSH is a
notoriously variable parameter, strongly influenced by stress and diurnal factors, and that low
magnitude decreases in thyroid hormones in rodents may or may not trigger a measurable increase
in circulating TSH. After such a short-term exposure to TBA and mild reductions in T3/T4, TSH
levels would not be expected to be induced, as even the positive control phenobarbitol failed to
induce an increase in TSH (Blanck et al., 2010).
EPA Response: While detecting small, treatment-associated changes in thyroid hormone and
related pituitary hormone levels such as TSH maybe be experimentally difficult, complicated by
several factors including diurnal variation in background levels, inter-animal differences, and
analytical variability as pointed out by the public comment, the presence of such complications
does not in and of itself constitute positive evidence supporting the effect. To address these
This document is a draft for review purposes only and does not constitute Agency policy.
D-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
challenges, alternative endpoints (e.g. pituitary TSH subunit mRNA expression) have been
evaluated in studies of other compounds as an alternative measure to blood TSH levels. As
discussed in Section 1.2.2, changes in TSH levels was one of numerous factors evaluated in the
analysis of support for an anti-thyroid MOA, as described in the EPA guidance on rodent thyroid
follicular cell tumorigenesis (EPA, 1998).
Comments Related to Reproductive, Developmental and Neurotoxic Effects
Comment [LyondellBasell]: The assessment should draw a conclusion regarding reproductive
toxicity that indicates that there is a low concern for TBA reproductive toxicity and there is no need
for further reproductive testing. This conclusion is supported by studies in MTBE (Bevan et al,
1997) and ETBE (CIT, 2004; JPEC 2008; Fujii et al 2010; de Peyster, 2010)
EPA Response: The reproductive toxicity studies for MTBE and ETBE provide evidence for some
reproductive effects at higher doses, but are not consistent across studies or doses. For example, a
one-generation reproductive study in rats given 0,100, 300, or 1,000 mg/kg-day ETBE for 16-17
weeks reported a 13.6% total incidence of total litter loss in the high dose group (control data for
whole litter loss ranged from 0-4.8%, mean value 0.7%), but was confounded by evidence of
systemic toxicity in two of the dams. There was also a slight but significant prolongation of
gestation in the high dose group. In another study, ETBE administration at 0, 250, 500, or 1,000
mg/kg-day during the pre-mating, mating, gestation, and lactation periods, and no effects on
reproductive endpoints were reported.
Data for maternal body weight gain after exposure to ETBE were also inconsistent across studies. In
Asano et al. (2001) and JPEC (2008i), New Zealand White rabbits were exposed to 0,100, 300, or
1,000 mg/kg-day ETBE and showed a significant decrease in maternal weight gain at the highest
dose (although interpretation of maternal weight in rabbits should reviewed with caution due to
the high variability in weight in rabbits during pregnancy). Similarly, decreased maternal weight
gain was also reported by Gaoua (2004a) in Sprague Dawley rats exposed to ETBE at 1,000 mg/kg-
day. In contrast, ETBE induced an increase in maternal weight the same dose and rat strain in
another study (Fujii et al 2010 and JPEC (2008e). In other studies with similar dose
administrations, maternal weight was not affected (Aso etal. 2014; Asano et al 2011; Fujii etal
2010; Gaoua, 2004b). Taken together, these studies do notprovide compelling evidence of no effect
in animals, therefore we cannot draw conclusions about the reproductive toxicity of ETBE. Further,
the inconsistent findings across these studies do not strengthen the TBA database to the level that
would allow a different conclusion to be drawn.
Comment [LyondellBasell]: The assessment should include a separate section on neurotoxicity
that includes relevant data from well-conducted TBA studies as well as available neurotoxicity and
This document is a draft for review purposes only and does not constitute Agency policy.
D-10 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
developmental neurotoxicity studies of structurally related chemicals such as other butanol
isomers and chemicals that are metabolized to TBA.
EPA Response: While we appreciate the comment about including other chemicals, this assessment
is focused on the toxicity resulting from TBA exposure as the parent compound.
Comment [LyondellBasell]: Contrary to the statement in the draft assessment on page 1-53 line
19-22, Nelson etal. (1991) provided detailed information on exposure methods and results to
indicate the studies were conducted similarly. Taken together, the data from these two studies
indicate that no pattern of developmental neurotoxicity was observed in studies at very high dose
levels. These results also support the current statement in Section 2.1.3 on database uncertainty
factor that further studies are unlikely to lead to identification of a more sensitive endpoint or a
lower point of departure than those selected by EPA for oral and inhalation exposures. This is
further supported by a series of studies conducted on other butanols and short-chain aliphatic
alcohols that should be considered relevant read-across data (Nelson etal. 1989,1990)(note this
Nelson et al., 1989 is a different publication from that reported in the draft assessment, the citation
for this present reference is found in our reference list to this document).
EPA Response: The statement that the studies were not conducted similarly was removed.
However, as noted by the authors, there were still limitations in the study design and the groups
were exposed nonconcurrently which prevented direct comparisons between the two
concentrations of TBA. In regard to the read-across for other butanols, while it is true that some
studies point toward little to no evidence for neurotoxicity (e.g. Nelson et al, 1989) still other more
recent publications have concluded thatthere is evidence of neurotoxicity (Bale et al., 2016).
Therefore, it is difficult to make the conclusion that there is compelling evidence of no neurotoxicity
or developmental neurotoxicity due to exposure to TBA.
Comment [LyondellBasell]: "...conclusion fails to mention that the developmental toxicity
observed occurred only in the presence of significant maternal toxicity and it is not possible to
determine if the maternal toxicity observed played a role in the developmental toxicity." In
addition, "there is no examination or consideration of the dose levels used in several of the studies
presented as causing developmental toxicity". Also, "While developmental effects should not be
"discounted" because of the maternal toxicity observed, it is the responsibility of the document to
inform the reader of all possible explanations for the observed effects. Inclusion of the maternal
toxicity endpoints, in the same level of detail and accuracy as the developmental toxicity endpoints
is necessary for this assessment to be complete." Finally, "In considering the fetal and maternal
toxicity data following tert-butanol exposure, the severity of the maternal effects were minimal and
therefore the developmental effects in the fetuses should not be discounted (U.S. EPA, 1991b)."
This document is a draft for review purposes only and does not constitute Agency policy.
D-ll DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Supplemental Information—tert-Butyl Alcohol
EPA Response: Although this was mentioned in the current draft we have strengthened the
language in Sections 1.2.3 and 1.2.4 to clearly address this issue. The draft now calls out the higher
doses at which both maternal toxicity and developmental effects are observed. The draft contains
maternal toxicity in the developmental evidence tables. The maternal data, if collected, is presented
in similar fashion to that of the other data. This language has been altered and the draft no longer
states that the maternal effects were minimal.
Comment [LyondellBasell]: The draft assessment is incorrect in concluding that the data provide
inadequate information to draw conclusions regarding neurodevelopmental toxicity of TBA. The
Nelson et al. (1991) study is a comprehensive developmental neurotoxicity study that included
multiple tests of motor activity, motor coordination, and cognitive behavior including schedule
controlled operant behavior.
EPA Response: We appreciate the commenter's position on the neurodevelopmental toxicity of
TBA. However, as noted by the authors, there were still limitations in the Nelson study including
not running the two concentrations of TBA concurrently and often only reporting data for the
significant changes. As noted by the commenters and in the draft, there are limitations with the
Daniel and Evans study as well. Taken together, it is difficult to make the strong conclusions about
the evidence of neurotoxicity or developmental neurotoxicity due to exposure to TBA. For this
reason, EPA concludes that the available evidence is inadequate.
Comments Related to Cancer Weight of Evidence
Comment [LyondellBasell]: The descriptor suggestive evidence represents a highly conservative
assessment The overall weight of evidence indicates that tert-butanol induced rat renal cancer is
qualitatively not relevant to humans based on robust mode of action evidence (a2u-globulin and
CPN). In addition, the mouse thyroid tumors are not quantitatively relevant to humans due the
observation that the high-dose used in mouse oral bioassay was substantially above EPA guidance
recommendations for a Limit Dose, as well as being above the dose at which tert-butanol
metabolism was saturated with associated onset of nonlinear toxicokinetics.
EPA Response: Although the evidence suggests that tert-butanol induces a2U-globulin nephropathy,
the data indicate that tert-butanol is a weak inducer of a2U-globulin and that this process is not
solely responsible for the renal tubule nephropathy and carcinogenicity observed in male rats. The
lack of compensatory cell proliferation in male rats and evidence of nephrotoxicity in female rats
suggest that other processes, in addition to the a2U-globulin process, are operating. Furthermore,
the accumulation of hyaline droplets and the induction of renal tubule hyperplasia were affected at
higher doses compared to those inducing renal tubule tumors. EPA conducted a MOA analysis
This document is a draft for review purposes only and does not constitute Agency policy.
D-12 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert-Butyl Alcohol
under EPA's cancer guidelines using the proposed criteria from Hard and Khan (2004) and Hard et
al. (2013). This analysis is presented in 1.2.1. under mode of action analysis- kidney effects. In
summary, considering discrepant patterns in the dose-response relationships for CPN, hyperplasia,
and renal tubule tumors and the lack of relationships between CPN grades and renal tubule tumors
in female rats, together with the lack of a generally accepted MOA for CPN, the renal tubule tumors
in rats cannot be attributed to CPN. Regarding the relevance of the mouse thyroid follicular cell
tumors, please see the EPA response to a similar comment above in the "Comments Related to
Thyroid Effects" Section.
Comments Related to Dose-Response
Comment [ACC]: The following analogy was not clear: "A 10% relative change from control was
used as a BMR for absolute kidney weight by analogy with a 10% change in body weight as an
indicator of toxicity." Is a 10% change in absolute kidney weight known to be adverse? How would
a 10% extra risk calculation compare?
EPA Response: Extra risk and relative deviation are not alternatives; extra risk is for dichotomous
data and relative deviation is for continuous data. Whatever is selected as the BMR (including a
10% relative deviation in kidney weight) doesn't have to be adverse per se, since the point of POD
selection is to identify an exposure level without adverse effects (minimally biologically significant
effects).
Comment [LyondellBasell]: While correct for female mice, a BMR of 5% was selected for males "to
represent the observed response for low-dose extrapolation", likely because the responses for the
control and 3 treatment groups were 2, 0, 7 and 2 percent, respectively. The use of the 5% BMR, for
apparent statistical reasons only, should be clearly indicated in Table 2-9 as an additional source of
uncertainty in the derivation of the SF. In addition, it should be noted that NTP guidance for
statistical evaluation of tumors responses with a high background incidence should be evaluated
against a p<0.01 to reduce the potential for false positive tumor determinations.
EPA Response: The text in the Toxicological Review has been clarified on page 1-42. The trend test
that was used tested for a linear trend in the mortality-adjusted incidences, so the apparent non-
monotonicity mostly reflects noisy data (i.e., insufficient to conclude that the responses in the two
highest groups differ). The statement that the data were non-monotonic is an over-interpretation,
given the reduced effective size of the high dose group. Contrary to the comment, follicular cell
thyroid adenomas and carcinomas overall are not common (3.4% in females for years 1984-1994)
and NTP does not rely solely on p-value cut-offs for interpreting whether there is a positive
response. The NTP specifically noted that the adenomas were uncommon, that related hyperplasia
This document is a draft for review purposes only and does not constitute Agency policy.
D-13 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information—tert-Butyl Alcohol
was increased in higher dose groups, and that the only carcinoma occurred in the highest dose
group. Support also comes from the concordance with female mice. Although none of the NTP's
reported trend tests had p<0.05, the trend test EPA applied was developed by NTP after this NTP
report was issued.
Comments Related to the Physiologically-Based Pharmacokinetic Model and Toxicokinetics
Comment [LyondellBasell]: Given that kidney toxicity is being considered following inhalation and
oral exposure, a more appropriate dose metric to evaluate would be the AUC for tert-butanol in the
kidney.
EPA Response: There are effects of tert-butanol in both male and female rats, showing these effects
are not attributable to a2u-globulin binding alone. For the females the tert-butanol blood
concentrations are very representative of kidney concentrations because the blood:kidney partition
coefficient is 1.1. For the exposure levels evaluated for route-to-route extrapolation, the ratio of the
AUC of the free tert- butanol concentration in kidney to blood AUC is 0.828 in female and male rats
(slight variation in 4th decimal place) for oral exposures. For inhalation exposures, at steady state
the ratio of free TBA in kidney to blood ranged from 0.832-0.834 in both male and female rats, but
the ratio differed by no more than 0.04% for any exposure level. Thus, use of AUC in the kidney vs.
blood would result in less than a 0.1% change in the resulting HECs. Given that there are many
more data points to inform blood AUC, there is greater confidence in the model's ability to predict
blood AUC, so it was selected as a dose metric.
Comment [LyondellBasell]: A variety of questions and concerns are related to specific aspects of
PBPK modeling as implemented in the Salazar etal. (2015) model.
EPA Response: EPA has adopted the newly available Borghoff et al. (2016) model.
Comment [LyondellBasell]: Prior to using a model for extrapolation to derive an RfC, the model
and data sets used to develop and verify the model need to be confirmed. There is a lack of
identification of data sets from unpublished reports used for model development and review of the
model code is needed before its use for deriving an RfC.
EPA Response: A table has been included in the PBPK evaluation (U.S. EPA 2017), which replaces
most of Appendix B, summarizing all of the data sets used in the modeling, with details on specific
data sets provided. All citations should now be correct. All of the sources are available in HERO.
This document is a draft for review purposes only and does not constitute Agency policy.
D-14 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Supplemental Information—tert-Butyl Alcohol
Comment [LyondellBasell]: The differential responses in the 13 week inhalation studies versus the
oral studies could be attributed to the discontinuous nature of the inhalation exposures. Thus, use
of "similar blood TBA" is only appropriate if AUC values are the basis of the route comparisons.
EPA Response: Using a 6 h/d, 5 d/w exposure pattern, the average blood concentrations predicted
by the rat model are 84 and 190 mg/L for the two highest inhalation exposure levels. The
corresponding levels from the oral NTP bioassay range from 74 to 1900 mg/L, with the 2nd lowest
dose predicted to yield 200 mg/L in male rats. So when accounting for the pulsatile pattern, the
highest inhalation exposures are predicted to yield blood levels comparable to the two lowest oral
doses. So the internal dose ranges for the two overlap, though are much lower overall for
inhalation.
AUCs are the basis being used for route comparisons. But tissue levels will track with blood levels
and even using a model with alpha-2-u binding, when exposures are selected to match blood AUCs,
the free concentrations of TBA in kidney will also be similar for males exposed by DW vs.
inhalation, and likewise females exposed by DW vs. inhalation. So incorporation of this metric
and/or using alternate internal dose metrics is not likely to change the conclusion.
Comment [LyondellBasell]: Figure 4F and FG in the Salazar et al. (2015) were produced by
assuming only 67% of the amount of ETBE or TBA in exhaled air was collected as a cumulative
amount in exhaled breath following ETBE nose only inhalation exposure. While this improves the
model fit to the data, it is not supported by experimental evidence.
EPA Response: The 67% correction was mistakenly applied to account for the difference between
expired alveolar air and total expired air, which includes air that only enters the conducting
airways ("dead space"). We agree that it is not appropriate and the term has been removed from
the calculation.
Comment [LyondellBasell]: The ARCO (1983) study reported toxicokinetics evaluations in rats
using radiolabeled TBA. Although this study provides useful information, the data collected in this
study needs to be described clearly as to how it was recalculated to provide the actual
concentration of tert-butanol in blood.
EPA Response: The values from ARCO (1983) were calculated by combining the % tert-butanol and
tert-butanol equivalents from Tables 15 and 24 in the Arco report for 1 mg/kg and from Tables 37
and 59 for 500 mg/kg. A table is included in the PBPK evaluation (U.S. EPA, 2017), which replaces
Appendix B, showing the calculations. A table has been included with details on all data sources
used.
Comment [LyondellBasell]: The Leavens and Borghoff (2009) tert-butanol PBPK model does in
fact represent the tert-butanol blood levels measured in the Poet et al. (1997) study, as shown in
This document is a draft for review purposes only and does not constitute Agency policy.
D-15 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Supplemental Information—tert-Butyl Alcohol
the figure presented in these comments. As also noted, this PBPK model predicted tert-butanol
concentration in tissues of male and female rats exposed to tert-butanol via inhalation.
EPA Response: The U.S. EPA attempted to reconstruct the tert-butanol submodel from the Leavens
and Borghoff (2009) publication; the results of using that model to simulate the i.v. exposures of
Poetetal. (1997) are shown in the PBPK evaluation (U.S. EPA, 2017), and were deemed inadequate.
The Borghoff etal. (2016) model differs in several details from thatof Leavens and Borghoff (2009).
In particular, Leavens and Borghoff (2009) shows urinary clearance as coming from the kidney
compartment, while Borghoff et al. (2016) describes it as coming from (mixed) venous blood.
Hence we cannot conclude that Leavens and Borghoff (2009) fits the Poet et al. (1997) data, though
it is possible that EPA's attempt to reproduce that model was erroneous. In any case, the Borghoff et
al. (2016) model does fit the data.
Comment [LyondellBasell]: The metabolism of tert-butanol suggests that it is reasonable to predict
potential high-dose specific metabolic saturation. There is reasonable toxicokinetic data indicating
that the top male/female doses of greater than 2000 mg/kg bw/day, and possibly even the
female/male mid-doses of 1020 and 1040 mg/kg bw/day, exceeded saturation of tert-butanol
metabolism resulting in onset of nonlinear plasma tert-butanol toxicokinetics. Mice (C56BL6)
administered single intraperitoneal doses of tert-butanol at doses of 5,10 and 20 mmol/kg bw
(370, 741 and 1482 mg/kg bw) resulted in respective AUC values of 28, 96 and 324 mmol.hrs/L
(Faulkner and Hussain, 1989). Thus, a 4-fold increase in dose (5 to 20 mmol/kg bw) resulted in an
11.6-fold increase in systemic AUC; metabolic saturation may have been present even at the next
lowest dose of 10 mmol/kg bw in which a 2-fold increase in dose (5 to 10 mmol/kg bw) resulted in
a 3.4-fold increase in AUC.
EPA Response: Saturation of tert-butanol metabolism is reasonable and the PBPK model
incorporated Michaelis-Menten kinetics which account for saturation. The KM used is 0.379 mM.
Considering that this value is for rats rather than mice, it appears reasonably consistent with the
range reported by Faulker and Hussain (1989), 0.56-0.92 mM, from fitting a one-compartment TK
model separately to each dose level. What matters is that the PBPK model adequately fits the PK
data across the range of exposures. Since this is a PK nonlinearity which is consistent with the
(fairly standard) model structure, that may or may not indicate nonlinearity in pharmacodynamic
mechanisms, the EPA does not consider it particularly useful to point it out for each data set where
it occurs. One of the reasons for using a PBPK model is that it allows one to appropriately account
for metabolic saturation, and the impact that has on the internal dose(s) across the entire dose
range.
Comment [LyondellBasell]: Use of toxicokinetic data to provide a data-informed selection of the
appropriate top dose in animal toxicity tests has recently been described as a Kinetically Derived
This document is a draft for review purposes only and does not constitute Agency policy.
D-16 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information—tert-Butyl Alcohol
Maximum (KMD) dose selection strategy (Saghir etal., 2012). The KMD dose selection strategy
specifically emphasizes that toxicokinetic data, when available, can and should be used as an
alternative to conventional top dose selection strategies based on Maximum Tolerated Dose (MTD).
tert-Butanol would have been a strong candidate for a KMD-based dose selection strategy.
EPA Response: While the KMD approach may provide better study designs in the future, the NTP
studies were conducted more than 15 years prior to Saghir etal. (2012). But the U.S. EPA does not
agree that toxicity data collected at exposure levels which saturate a metabolic pathway are not
usable or relevant for estimating human health risk, in particular when part of a dose-response
array that spans high and low exposure levels.
Comments Related to Genotoxicity
Comment [LyondellBasell]: It is rather surprising that the draft assessment concluded that "...a
limited database is available for understanding the role of tert-butanol-induced genotoxicity for
mode of action and carcinogenicity." LyondellBasell commented that the statement that only
limited animal studies were conducted to investigate micronucleus formation is inaccurate.
Contrary to the above statement, the WoE from a large database informs that TBA does not have the
potential to be an in vivo genotoxicant and a mutagenic MOA in the etiology of animal tumors can
thus be excluded with a reasonable degree of certainty. Finally, the NTP conducted a total of 3
micronucleus studies, two in the mouse and one in the rat and all three studies were clearly
negative.
EPA Response: As indicated in the summary of the genotoxicity section, the database is rather small
for both the array of genotoxicity tests conducted as well as the number of studies within the same
type of test category. In addition, sometimes, the data is either conflicting or inconsistent Since
there are a few studies that are positive, tert-butanol cannot be considered nongenotoxic with
complete certainty, therefore, the conclusion presented in Section B.3.2 of the Supplementary
Materials will remain unchanged. Finally, the two mouse studies referred to by LyondellBasell are
one study published both in NTP (1995) and NTP (1997).
Comment [LyondellBasell]: There are two key studies missing in Section B.3.2.3 under in vivo
mammalian studies. The first study is a rat bone marrow micronucleus test reported by the NTP
(1997). The second missing study is a mouse bone marrow micronucleus test reported by the NTP.
The two missing, negative NTP mouse and rat bone marrow micronucleus studies that were
discussed above should be included under "In vivo Animal Studies" section of this table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-17 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 EPA Response: The NTP (1997) rat study is now included in the assessment. However, both
2 micronucleus tests referenced in the comments are the same study. With respect to the mouse
3 study, the NTP (1995) and NTP (1997) studies are from the same set of experiments and is already
4 present in the current assessment.
This document is a draft for review purposes only and does not constitute Agency policy.
D-18 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Supplemental Information—tert-Butyl Alcohol
REFERENCES
Acharva. S: Mehta. K: Rodrigues. S: Pereira. T: Krishnan. S: Rao. CV. (1995). Administration of
subtoxic doses of t-butyl alcohol and trichloroacetic acid to male Wistar rats to study the
interactive toxicity. Toxicol Lett. 80: 97-104. http://dx.doi.org/10.1016/Q378-
4274C9S103340-Q.
Acharva. S: Mehta. K: Rodriguez. S: Pereira. T: Krishnan. S: Rao. CV. (1997). Ahistopathological study
of liver and kidney in male Wistar rats treated with subtoxic doses of t-butyl alcohol and
trichloroacetic acid. Exp Toxicol Pathol. 49: 369-373. http://dx.doi.org/10.1016/S094Q-
2993C97180119-4.
Amberg. A: Rosner. E: Dekant. W. (1999). Biotransformation and kinetics of excretion of methyl-
tert-butyl ether in rats and humans. Toxicol Sci. 51:1-8.
Amberg. A: Rosner. E: Dekant. W. (2000). Biotransformation and kinetics of excretion of ethyl tert-
butyl ether in rats and humans. Toxicol Sci. 53: 194-201.
http: / /dx. do i. o r g /10.109 3 /toxs ci /5 3.2.194.
ARCO (ARCO Chemical Company). (1983). Toxicologist's report on metabolism and
pharmacokinetics of radiolabeled TBA 534 tertiary butyl alcohol with cover letter dated
03/24/1994. (8EHQ86940000263). Newton Square, PA.
Arslanian. MT: Pascoe. E: Reinhold. TG. (1971). Rat liver alcohol dehydrogenase. Purification and
properties. Biochem J. 125: 1039-1047.
ATSDR (Agency for Toxic Substances and Disease Registry). (1996). Toxicological profile for
methyl-tert-butyl ether [ATSDR Tox Profile], Atlanta, GA: U.S. Department of Health and
Human Services, Public Health Service. http: //www.atsdr.cdc.gov/ToxProfiles/tp91.pdf.
Bailer. AT: Portier. CT. (1988). Effects of treatment-induced mortality and tumor-induced mortality
on tests for carcinogenicity in small samples. Biometrics. 44: 417-431.
Bailey. SA: Zidell. RH: Perry. RW. (2004). Relationships between organ weight and body/brain
weight in the rat: What is the best analytical endpoint? Toxicol Pathol. 32: 448-466.
http://dx.doi.Org/10.1080/01926230490465874.
Baker. RC: Sorensen. SM: Deitrich. RA. (1982). The in vivo metabolism of tertiary butanol by adult
rats. Alcohol Clin Exp Res. 6: 247-251. http://dx.doi.org/10.1111 /i. 1530-
0277.1982.tb04970.x.
Bernauer. U: Amberg. A: Scheutzow. D: Dekant. W. (1998). Biotransformation of 12C- and 2-13C-
labeled methyl tert-butyl ether, ethyl tert-butyl ether, and tert-butyl alcohol in rats:
Identification of metabolites in urine by 13C nuclear magnetic resonance and gas
chromatography/mass spectrometry. Chem Res Toxicol. 11: 651-658.
http: / /dx. do i. o r g /10.10 21 /tx9 7 0 215v.
Borghoff. ST: Asgharian. B. (1996). Ethyl tertiary-butyl ether (ETBE): Pharmacokinetic study in male
and female CD-I mice after single inhalation exposure and male and female F-344 rats after
single and repeated inhalation exposure. (CUT Protocol 95026). La Palma, CA: ARCO
Chemical Company.
This document is a draft for review purposes only and does not constitute Agency policy.
R-l DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Supplemental Information—tert-Butyl Alcohol
Borghoff. ST: Prescott. IS: lanszen. DB: Wong. BA: Everitt. II. (2001). alpha2u-Globulin nephropathy,
renal cell proliferation, and dosimetry of inhaled tert-butyl alcohol in male and female F-
344 rats. Toxicol Sci. 61: 176-186. http: //dx.doi.org/10.1093/toxsci/61.1.176.
Borghoff. ST: Ring. C: Banton. MI: Leavens. TL. (2016). Physiologically based pharmacokinetic model
for ethyl tertiary-butyl ether and tertiary-butyl alcohol in rats: Contribution of binding to
a2u-globulin in male rats and high-exposure nonlinear kinetics to toxicity and cancer
outcomes. J Appl Toxicol, http://dx.doi.org/10.1002/iat.3412.
Cederbaum. AT: Cohen. G. (1980). Oxidative demethylation of t-butyl alcohol by rat liver
microsomes. Biochem Biophys Res Commun. 97: 730-736.
Cederbaum. AT: Qureshi. A: Cohen. G. (1983). Production of formaldehyde and acetone by hydroxyl-
radical generating systems during the metabolism of tertiary butyl alcohol. Biochem
Pharmacol. 32: 3517-3524. http://dx.doi.org/10.1016/0006-2952r83190297-6.
Dickey. FH: Cleland. GH: Lotz. C. (1949). The role of organic peroxides in the induction of mutations.
PNAS. 35: 581-586.
Faulkner. TP: Wiechart. TP: Hartman. DM: Hussain. AS. (1989). The effects of prenatal tertiary
butanol administration in CBA/J and C57BL/6J mice. Life Sci. 45: 1989-1995.
FDA (U.S. Food and Drug Administration). (2011a). Indirect food additives: Adjuvants, production
aids, and sanitizers. Surface lubricants used in the manufacture of metallic articles. 21 CFR
178.3910.
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=178.3910.
FDA (U.S. Food and Drug Administration). (2011b). Indirect food additives: Paper and paperboard
components. Defoaming agents used in coatings. 21 CFR 176.200.
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=176.200.
Galloway. SM: Deasv. DA: Bean. CL: Kravnak. AR: Armstrong. MT: Bradley. MO. (1987). Effects of
high osmotic strength on chromosome aberrations, sister-chromatid exchanges and DNA
strand breaks, and the relation to toxicity. Mutat Res. 189: 15-25.
http://dx.doi.Org/10.1016/0165-1218f87190029-2.
Grant. KA: Samson. HH. (1981). Development of physical dependence on t-butanol in rats: An
examination using schedule-induced drinking. Pharmacol Biochem Behav. 14: 633-637.
http: //dx.doi.org/10.1016/0091 -3057f81190124-6.
Hard. GC: Bruner. RH: Cohen. SM: Pletcher. TM: Regan. KS. (2011). Renal histopathology in toxicity
and carcinogenicity studies with tert-butyl alcohol administered in drinking water to F344
rats: A pathology working group review and re-evaluation. Regul Toxicol Pharmacol. 59:
430-436. http://dx.doi.org/10.1016/i.vrtph.2011.01.007.
limenez. T: Longo. E: Benitez. T. (1988). Induction of petite yeast mutants by membrane-active
agents. Appl Environ Microbiol. 54: 3126-3132.
Kirkland. D: Aardema. M: Henderson. L: Muller. L. (2005). Evaluation of the ability of a battery of
three in vitro genotoxicity tests to discriminate rodent carcinogens and non-carcinogens: I.
Sensitivity, specificity and relative predictivity. Mutat Res. 584: 1-256.
http://dx.doi.Org/10.1016/i.mrgentox.2005.02.004.
Leavens. TL: Borghoff. ST. (2009). Physiologically based pharmacokinetic model of methyl tertiary
butyl ether and tertiary butyl alcohol dosimetry in male rats based on binding to alpha2u-
globulin. Toxicol Sci. 109: 321-335. http://dx.doi.org/10.1093/toxsci/kfp049.
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Supplemental Information—tert-Butyl Alcohol
Lvondell Chemical Co. (Lyondell Chemical Company). (2004). Reproductive and developmental
toxicity screening test in rats by oral gavage. (Document Control Number: 89-040000106).
McComb. I: Goldstein. D. (1979a). Additive physical dependence: evidence for a common
mechanism in alcohol dependence. J Pharmacol Exp Ther. 210: 87-90.
McComb. 1: Goldstein. D. (1979b). Quantitative comparison of physical dependence on tertiary
butanol and ethanol in mice: Correlation with lipid solubility. J Pharmacol Exp Ther. 208:
113-117.
Mcgregor. D: Cruzan. G: Callander. R: May. K: Banton. M. (2005). The mutagenicity testing of
tertiary-butyl alcohol, tertiary-butyl acetate and methyl tertiary-butyl ether in Salmonella
typhimurium. Mutat Res. 565: 181-189. http://dx.doi.Org/10.1016/i.mrgentox.2004.10.002.
McGregor. DB: Brown. A: Cattanach. P: Edwards. I: Mcbride. D: Casparv. Wl. (1988). Responses of
the L5178Y tk+/tk- mouse lymphoma cell forward mutation assay II: 18 coded chemicals.
Environ Mol Mutagen. 11: 91-118.
Nihlen. A: Lof. A: lohanson. G. (1998a). Controlled ethyl tert-butyl ether (ETBE) exposure of male
volunteers: I Toxicokinetics. Toxicol Sci. 46: 1-10.
http://dx.doi.org/10.1006/toxs.1998.2516.
Nihlen. A: Lof. A: lohanson. G. (1995). Liquid/air partition coefficients of methyl and ethyl t-butyl
ethers, t-amyl methyl ether, and t-butyl alcohol. J Expo Anal Environ Epidemiol. 5: 573-582.
Nihlen. A: Lof. A: lohanson. G. (1998b). Experimental exposure to methyl tertiary-butyl ether: I
Toxicokinetics in humans. Toxicol Appl Pharmacol. 148: 274-280.
http://dx.doi.org/10.1006/taap.1997.8333.
NIOSH (National Institute for Occupational Safety and Health). (2007). NIOSH pocket guide to
chemical hazards. (DHHS-2005-149. CBRNIAC-CB-112149). Cincinnati, OH.
http://www.cdc.gov/niosh/docs/20Q5-149/.
NSF International. (2003). t-Butanol: Oral Risk Assessment Document (CAS 75-65-0). Ann Arbor,
MI.
NTP (National Toxicology Program). (1995). Toxicology and carcinogenesis studies of t-butyl
alcohol (CAS no 75-65-0) in F344/N rats and B6C3F1 mice (Drinking water studies) (pp. 1-
305). (NTPTR436). Research Triangle Park, NC.
NTP (National Toxicology Program). (1997). NTP technical report on toxicity studies of t-butyl
alcohol (CAS no 75-65-0) administered by inhalation to F344/N rats and B6C3F1 mice (pp.
1-56, A51-D59). Research Triangle Park, NC.
http://ntp.niehs.nih.gov/ntp/htdocs/ST rpts/tox053.pdf.
OSHA (Occupational Safety & Health Administration). (2006). Table Z-l: Limits for air
contaminants. Occupational safety and health standards, subpart Z, toxic and hazardous
substances. (OSHA standard 1910.1000, 29 CFR). Washington, DC: U.S. Department of
Labor.
http://www.osha.gov/pls/oshaweb/owadisp.show document?p table=STANDARDS&p id=
9992.
Poet. TS: Valentine. IL: Borghoff. SI. (1997). Pharmacokinetics of tertiary butyl alcohol in male and
female Fischer 344 rats. Toxicol Lett 92: 179-186.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Supplemental Information—tert-Butyl Alcohol
Salazar. KD: Brinkerhoff. CI: Lee. IS: Chiu. WA. (2015). Development and application of a rat PBPK
model to elucidate kidney and liver effects induced by ETBE and tert-butanol. Toxicol Appl
Pharmacol. 288: 439-452. http://dx.doi.Org/10.1016/i.taap.2015.08.015.
Sgambato. A: lavicoli. I: De Paola. B: Bianchino. G: Boninsegna. A: Bergamaschi. A: Pietroiusti. A:
Cittadini. A. (2009). Differential toxic effects of methyl tertiary butyl ether and tert-butanol
on rat fibroblasts in vitro. Toxicol Ind Health. 25: 141-151.
http://dx.doi.org/10.1177/0748233709104867.
Snell. D. (1980). Impairment of avoidance behavior following short-term ingestion of ethanol,
tertiary-butanol, or pentobarbital in mice. Psychopharmacology. 69: 53-57.
http://dx.doi.org/10.1007/BF00426521.
Spiteri. Nl. (1982). Circadian patterning of feeding, drinking and activity during diurnal food access
in rats. Physiol Behav. 28: 139-147. http://dx.doi.org/10.1016/0031-9384r82190115-9.
Tang. G: Wang. 1: Zhuang. Z. (1997). [Cytotoxicity and genotoxicity of methyl tert-butyl ether and its
metabolite to human leukemia cells], Zhonghua Yufang Yixue Zazhi. 31: 334-337.
Thurman. RG: Winn. K: Urquhart. B. (1980). Rat brain cyclic AMP levels and withdrawal behavior
following treatment with t-butanol. Adv Exp Med Biol. 126: 271-281.
U.S. EPA (U.S. Environmental Protection Agency). (2000). Benchmark dose technical guidance
document [external review draft] [EPA Report], (EPA/630/R-00/001). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
http://www.epa.gov/raf/publications/benchmark-dose-doc-drafthtm.
U.S. EPA (U.S. Environmental Protection Agency). (2011). Recommended use of body weight 3/4 as
the default method in derivation of the oral reference dose. (EPA/100/R11/0001).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://www.epa.gov/raf/publications/interspecies-extrapolation.htm.
U.S. EPA (U.S. Environmental Protection Agency). (2016). Model files for tert-butanol and ETBE.
U.S. EPA (U.S. Environmental Protection Agency). (2017). PK/PBPK model evaluation for the IRIS
assessments of ethyl tertiary butyl ether (CASRN 637-92-3) and tert-butyl alcohol (CAS No.
75-65-0) (Draft) [EPA Report], Washington, DC: U.S. Environmental Protection Agency,
Pharmacokinetics Working Group.
Videla. LA: Fernandez. V: de Marinis. A: Fernandez. N: Valenzuela. A. (1982). Liver lipoperoxidative
pressure and glutathione status following acetaldehyde and aliphatic alcohols
pretreatments in the rat. Biochem Biophys Res Commun. 104: 965-970.
http:/ /dx.doi.org/10.1016/0006-291 Xf82191343-2.
Williams-Hill. D: Spears. CP: Prakash. S: Olah. GA: Shamma. T: Moin. T: Kim. LY: Hill. CK. (1999).
Mutagenicity studies of methyl-tert-butylether using the Ames tester strain TA102. Mutat
Res. 446: 15-21. http: //dx.doi.org/10.1016/sl383-5718f99100137-0.
Williams. TM: Borghoff. SI. (2001). Characterization of tert-butyl alcohol binding to "alpha"2u-
globulin in F-344 rats. Toxicol Sci. 62: 228-235. http://dx.doi.org/10.1093/toxsci/62.2.228.
Wood. I: Lavertv. R. (1979). Physical dependence following prolonged ethanol or t-butanol
administration to rats. Pharmacol Biochem Behav. 10: 113-119.
Yuan. Y: Wang. HF: Sun. HF: Du. HF: Xu. LH: Liu. YF: Ding. XF: Fu. DP: Liu. KX. (2007). Adduction of
DNA with MTBE and TBA in mice studied by accelerator mass spectrometry. Environ
Toxicol. 22: 630-635. http://dx.doi.org/10.1002/tox.20295.
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT—DO NOT CITE OR QUOTE
-------
Supplemental Information—tert-Butyl Alcohol
1 Zeiger. E: Anderson. B: Ha worth. S: Lawlor. T: Mortelmans. K: Speck. W. (19871. Salmonella
2 mutagenicity tests: III. Results from the testing of 255 chemicals. Environ Mutagen. 9: 1-
3 109. http://dx.doi.Org/10.1002/em.2860090602.
4
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT—DO NOT CITE OR QUOTE
------- |