DRAFT DELIBERATIVE: FOR INTERAGENCY REVIEW ONLY.
DO NOT DISTRIBUTE OUTSIDE YOUR AGENCY.
e CDA	EPA/635/R-14/378
* *	Interagency Review Draft
www.epa.gov/iris
Toxicological Review of tert-Butyl Alcohol (tert-butanol)
(CASRN 75-65-0]
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
Supplemental Information
September 2014
NOTICE
This document is an Interagency Science Consultation Review Draft. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. It has not been formally disseminated by EPA. It does not represent and should
not be construed to represent any Agency determination or policy. It is being circulated for review
of its technical accuracy and science policy implications.
Integrated Risk Information System
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC

-------
Supplemental Information—tert-Butanol
DISCLAIMER
This document is a preliminary draft for review purposes only. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. It has not been formally disseminated by EPA. It does not represent and should
not be construed to represent any Agency determination or policy. Mention of trade names or
commercial products does not constitute endorsement 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ii	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert-Butanol
CONTENTS
APPENDIX A. OTHER AGENCY AND INTERNATIONAL ASSESSMENTS	A-l
A.l.	OTHER AGENCY AND INTERNATIONAL ASSESSMENTS	A-l
APPENDIX B. INFORMATION IN SUPPORT OF HAZARD IDENTIFICATION AND DOSE-
RESPONSE ANALYSIS	B-l
B.l.	TOXICOKINETICS	B-l
B.l.l. Absorption	B-l
B.l.2. Distribution	B-l
B.l.3. Metabolism	B-2
B.l.4. Excretion	B-4
B.l.5. Physiologically Based Pharmacokinetic Models	B-5
B.2. PBPK MODEL EVALUATION SUMMARY	B-9
B.2.1. Evaluation of Existing tert-Butanol Submodels	B-9
B.2.2. Modification of Existing tert-Butanol Submodels	B-ll
B.2.3. tert-Butanol Model Application	B-14
B.2.4. PBPK Model Code	B-15
B.3.	OTHER PERTINENT TOXICITY INFORMATION	B-16
B.3.1. Genotoxicity	B-16
B.3.2.	Summary	B-21
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.l.	BENCHMARK DOSE MODELING SUMMARY	C-l
C.l.l.	Noncancer Endpoints	C-l
C.l.2. Cancer Endpoints	C-30
APPENDIX D. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS AND
EPA'S DISPOSITION	D-l
REFERENCES FOR APPENDICES	1
This document is a draft for review purposes only and does not constitute Agency policy.
iii	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert-Butanol
TABLES
Table A-l. Other Agency and International Assessments	A-l
Table B-l. PBPK model physiologic parameters and partition coefficients	B-12
Table B-2. Rate constants for tert-butanol determined by optimization of the model with
experimental data	B-l4
Table B-3. Summary of genotoxicity (both in vitro and in vivo) studies of tert-butanol	B-20
Table C-l. Non-cancer endpoints selected for dose-response modeling for tert-butanol	C-2
Table C-2. Summary of BMD modeling results 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	C-3
Table C-3. Summary of BMD modeling results for kidney transitional epithelial hyperplasia in
female F344 rats exposed to tert-butanol in drinking water for 2 years (NTP, 1995);
BMR = 10% extra risk	C-6
Table C-4. Summary of BMD modeling results for relative kidney weights in male F344 rats exposed
to tert-butanol in drinking water for 15 months (NTP, 1995); BMR = 10% relative
deviation and 1 standard deviation	C-9
Table C-5. Summary of BMD modeling results for relative kidney weights in female F344 rats
exposed to tert-butanol in drinking water for 15 months (NTP, 1995); BMR = 10%
relative deviation and 1 standard deviation	C-l3
Table C-6. Summary of BMD modeling results for kidney inflammation in female rats exposed to
tert-butanol in drinking water for 2 years (NTP, 1995); BMR = 10% extra risk	C-16
Table C-7. Summary of BMD modeling results for thyroid follicular cell hyperplasia in male B6C3F1
mice exposed to tert-butanol in drinking water for 2 years (NTP, 1995); BMR = 10%
extra risk	C-l9
Table C-8. Summary of BMD modeling results for thyroid follicular cell hyperplasia in female
B6C3F1 mice exposed to tert-butanol in drinking water for 2 years (NTP, 1995); BMR =
10% extra risk	C-20
Table C-9. Summary of BMD modeling results 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	C-23
Table C-10. Summary of BMD modeling results for relative 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	C-26
Table C-ll. Summary of BMD modeling results 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	C-29
Table C-l 2. Summary of BMD modeling results for relative kidney weight in female F344 rats
exposed to tert-butanol via inhalation for 6 hrs/d, 5d/wk for 13 weeks (NTP, 1997);
BMR = 10% relative deviation from the mean	C-30
Table C-13. Cancer endpoints selected for dose-response modeling for tert-butanol	C-31
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-3 2
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 (NTP, 1995); BMR = 10% extra
risk	C-34
This document is a draft for review purposes only and does not constitute Agency policy.
iv	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert-Butanol
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 PBPK
[tert-butanol, mg/L) dose units and including all dose groups (NTP, 1995); BMR = 10%
extra risk	C-36
Table C-17. 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 PBPK
[tert-butanol, mg/L) dose units and excluding high-dose group (NTP, 1995); BMR =
10% extra risk	C-38
Table C-18. 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 PBPK
(metabolized, mg/hr) dose units and including all dose groups (NTP, 1995); BMR =
10% extra risk	C-40
Table C-19. 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 PBPK
(metabolized, mg/hr) dose units and excluding high-dose group (NTP, 1995); BMR =
10% extra risk	C-42
Table C-20. 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 (Hard et al.,
2011; NTP, 1995); BMR = 10% extra risk	C-44
Table C-21. 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; re-analyzed data (Hard et al.,
2011; NTP, 1995); BMR = 10% extra risk	C-44
Table C-22. 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 PBPK
[tert-butanol, mg/L) dose units and including all dose groups; reanalyzed data (Hard et
al., 2011; NTP, 1995); BMR = 10% extra risk	C-47
Table C-23. 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 PBPK
[tert-butanol, mg/L) dose units and excluding high-dose group; reanalyzed data (Hard
et al., 2011; NTP, 1995); BMR = 10% extra risk	C-47
Table C-24. 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 PBPK
(metabolized, mg/hr) dose units and including all dose groups; reanalyzed data (Hard
et al., 2011; NTP, 1995); BMR = 10% extra risk	C-49
Table C-25. 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 PBPK
(metabolized, mg/hr) dose units and excluding high-dose group; reanalyzed data (Hard
et al., 2011; NTP, 1995); BMR = 10% extra risk	C-49
Table C-26. Summary of BMD modeling results for thyroid follicular cell adenomas in female
B6C3F1 mice exposed to tert-butanol in drinking water for 2 years (NTP, 1995); BMR =
10% extra risk	C-52
This document is a draft for review purposes only and does not constitute Agency policy.
v	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert-Butanol
FIGURES
Figure B-l. Biotransformation of tert-butanol in rats and humans	B-3
Figure B-2. Comparison of the tert-butanol portions of existing MTBE models with tert-butanol
blood concentrations from i.v. exposure by Poet et al. 1997	B-9
Figure B-3. Schematic of the PBPK submodel for tert- butanol in rats	B-ll
Figure B-4. Comparison of the EPA model predictions with measured tert-butanol blood
concentrations for i.v., inhalation, and oral gavage exposure to tert-butanol	B-l3
Figure B-5. Comparison of the EPA model predictions with measured amounts of tert- butanol in
blood after repeated inhalation exposure to tert-butanol	B-l6
Figure C-l. Plot of mean response by dose, with fitted curve for selected model	C-4
Figure C-2. Plot of mean response by dose, with fitted curve for selected model	C-7
Figure C-3. Plot of mean response by dose, with fitted curve for selected model (10% relative
deviation)	C-10
Figure C-4. Plot of mean response by dose, with fitted curve for selected model (10% relative
deviation)	C-l 3
Figure C-5. Plot of mean response by dose, with fitted curve for selected model	C-16
Figure C-6. Plot of mean response by dose, with fitted curve for selected model	C-20
Figure C-7. Plot of mean response by dose, with fitted curve for selected model; dose shown in
mg/m3	C-24
Figure C-8. Plot of mean response by dose, with fitted curve for selected model; dose shown in
mg/m3	C-27
Figure C-9. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/kg-d	C-32
Figure C-10. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/kg-d	C-34
Figure C-ll. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/L	C-36
Figure C-12. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/L	C-38
Figure C-13. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/hr	C-40
Figure C-14. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/hr	C-42
Figure C-15. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/kg-d	C-45
Figure C-16. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/L	C-48
Figure C-17. Plot of incidence rate by dose, with fitted curve for selected model; dose shown in
mg/hr	C-50
Figure C-18. Plot of mean response by dose, with fitted curve for selected model	C-52
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-Butanol
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
CA	chromosomal aberration
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
SE	standard error
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 Butanol
1	APPENDIX A. OTHER AGENCY AND
2	INTERNATIONAL ASSESSMENTS
3 A.l. OTHER AGENCY AND INTERNATIONAL ASSESSMENTS
Table A-l. Other Agency and International Assessments.
Organization
Toxicity value
American Conference of
Governmental Industrial
Hveienists (ACGIH, 2012)
Threshold Limit Value - 100 ppm (303.1493 mg/m3) time-weighted
average (TWA) for an 8-hour workday and a 40-hour work week
National Institute of
Occupational Safety and Health
(NIOSH,2007)
Recommended Exposure Limit - 100 ppm (300 mg/m3) TWA for up to
a 10-hour workday and a 40-hour work week
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)
ferf-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).
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 Butanol
APPENDIX B. INFORMATION IN SUPPORT OF
HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS
B.l. TOXICOKINETICS
There is little information on the absorption, distribution, metabolism, or excretion of tert-
butyl alcohol (tert-butanol) in humans. The studies identified were conducted in conjunction with
methyl tert-butyl ether (MTBE) and ethyl tert-butyl ether (ETBE) because tert-butanol is a primary
metabolite of these two compounds. Several studies examining some aspect of the toxicokinetic
behavior of tert-butanol in animals have been identified. Many of the studies were carried out in
conjunction with other specific endpoints (e.g., developmental). ARCO (19831 determined that
there were no differences in the pharmacokinetics of tert-butanol following either oral (i.e., gavage)
or inhalation exposure. Although there is some information available after both oral and inhalation
exposures, many studies also 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 distribution, metabolism, and excretion.
B.l.l. Absorption
Extensive tert-butanol toxicity testing data 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) and inhalation (6,164 mg/m3 for 6 hours) exposures fARCO.
1983): however, the absorption rate after inhalation exposure could not be determined because the
blood was saturated with radioactivity after 6 hours of a 6,164-mg/m3 exposure. Based on blood
concentrations, absorption was found to be complete at 1.5 hours following repeated oral exposure,
with an apparent zero-order decline in tert-butanol concentration for the majority of the
elimination phase, thus indicating that previous exposures did not affect the absorption of tert-
butanol fFaulkner etal.. 19891.
B.l.2. Distribution
The available animal data suggest that tert-butanol is distributed throughout the body
following oral, inhalation, and injection exposures (Poetetal.. 1997: Faulkner etal.. 1989: ARCO.
1983). Nihlen etal. (1995) determined a partition coefficient for tert- butanol using blood from
human volunteers. The study was approved by an ethical review board, and informed consent was
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 Butanol
obtained from the participants. Their results indicated that tert-butanol would rapidly move from
the blood into the tissues.
tert-Butanol was found in the kidney, liver, and blood of both sexes of rat following oral
exposure, but male rats retained more tert-butanol compared with females (Williams and Borghoff.
20011. Radioactivity was found in the low-molecular-weight protein fraction from the kidney
cytosol in male rats but not female rats, indicating that tert-butanol or one of its metabolites was
bound to a2u-globulin. Further analysis determined that it was tert-butanol that was bound and not
its metabolite acetone. The majority of tert-butanol in the kidney cytosol was eluted as the free
compound in both males and females, but a small amount was also found associated with the high-
molecular-weight protein fraction in both males and females. Borghoff et al. f20011 found similar
results in rats after inhalation exposure. Male rat kidney-to-blood ratios were significantly elevated
over female ratios 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 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 below. Urinary metabolites of tert-butanol in a human male
volunteers 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 fBernauer etal.. 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; however, no
information regarding informed consent was reported. In the same study, 2-hydroxyisobutyrate,
MPD, and tert-butanol sulfate were identified as major metabolites in rats, while acetone, tert-
butanol, and tert-butanol glucuronides were identified as minor metabolites fBernauer et al.. 19981.
Baker etal. f!9821 found that tert- butanol was a source of acetone, but also may have stimulated
acetone production from other sources.
There are no studies that identify specific enzymes responsible for the biotransformation of
tert-butanol. Using purified enzymes 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 P-450 (CYP450) (Cederbaum etal.. 1983: Cederbaum
and Cohen. 19801. In the first 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 the production of formaldehyde at a concentration of approximately 25 nmoles/mg
protein/30 min. According to study authors, the amount of formaldehyde generated by tert-butanol
is approximately 30% of the amount of formaldehyde formed during the metabolism of 10 mM
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
Supplemental Information—tert Butanol
aminopyrene in a similar microsomal system. The rate of formaldehyde generation from
tert-butanol was 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 within the same study there was a major reduction of formaldehyde formation 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 was found to
be dependent 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.
cm
glucuronide-O-
-CH,
CH,
t-butyl glucuronide
HO^O
HO-
rats, humans
-CH,
[O]
CH,
HO-
-CH,
CH3
t-butanol
CYP450
	I
rats,
humans
OH
"Y
CH3 oh
2-methyl-1,2-propanediol

H,C-
-OH
CH,
CH,
2-hydroxyisobutyric acid
h2c=o
formaldehyde
o
rats
\^0 CH,
CH,
/S-
0/ \
acetone
-CH,
CH,
t-butyl sulfate
Figure B-l. Biotransformation of tert-butanol in rats and humans.
Source: NSF International (20031ATSDR (19961 Bernauer et al. (19981 Amberg et al. (19991 and
Cederbaum and Cohen (19801
In a follow-up study, tert-butanol was oxidized to formaldehyde and acetone by a variety of
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
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
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert Butanol
[ethylenediaminetetraacetic acid]) and (2) the Fenton system of chelated ferrous iron and H2O2. In
both of these Fenton-type systems, H2O2 served as a precursor of -OH. Additionally, a Haber-Weiss
enzymatic system involving oxidation of xanthine 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 upon addition of -OH-scavenging agents (e.g., benzoate, mannitol). Some of the
experiments in this study of the oxidation of tert-butanol by the liver microsomal metabolizing
system were similar to those in the previous study except that, in addition to formaldehyde,
acetone formation was also 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)
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 the extent of 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 Or' might be involved (Cederbaum etal.. 1983).
B.1.4. Excretion
Human data on the excretion of tert-butanol comes from studies of MTBE and ETBE (Nihlen
etal.. 1998a. b). Eight or ten male human volunteers were exposed to 5, 25, or 50 ppm MTBE or
ETBE 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). 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).
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
37
38
Supplemental Information—tert Butanol
Amberg etal. (20001 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,
2-hydroxyisobutyrate (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 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
both concentrations.
Amberg etal. (2000) 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 fairly constant fraction of the administered radioactivity (23-33%) was
recovered in the urine at 24 hours postdosing. However, only 9% of a 1500-mg/kg administered
dose was recovered in urine, suggesting that the urinary route of elimination is saturated following
this dose (ARCO. 1983).
B.1.5. Physiologically Based Pharmacokinetic Models
There have been no physiologically based pharmacokinetic (PBPK) models developed
specifically for administration of tert-butanol. The majority studied tert-butanol as the primary
metabolite after oral or inhalation exposure to MTBE or ETBE. The most recent models for MTBE
oral and inhalation exposure include a component for the binding of tert-butanol to a2U-globulin
(Borghoff etal.. 2010: Leavens and Borghoff. 2009).
Faulkner and Hussain (1989) used a one-compartment open model with Michaelis-Menten
elimination kinetics to fit tert-butanol blood concentrations obtained from C57BL/6J mice given i.p.
injections of 5,10, or 20 mmol/kg tert-butanol. Elimination was indistinguishable from first-order
kinetics in the range of concentrations studied. An increase in Vmax and decrease in apparent
volume of distribution with dose are consistent with this model and suggest the existence of
parallel elimination processes.
Borghoff et al. (1996) developed a PBPK model for MTBE and its metabolite tert-butanol in
rats. Doses and blood levels were taken from several published studies. The initial model included a
tissue-specific five-compartment model using blood, liver, kidney, muscle, and fat with liver
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
38
Supplemental Information—tert Butanol
metabolism rate constants. The model was able to predict the accumulation of tert-butanol in blood,
but not its clearance. A two-compartment model was better at predicting tert-butanol blood levels,
however, the volume of the total body water had to be changed to obtain an adequate fit, suggesting
dose-dependent changes in the kinetics of tert-butanol. Overall, evaluation of the tert-butanol
models suggests that the clearance of tert-butanol from the blood of rats after exposure to MTBE
involves processes beyond metabolic elimination.
Nihlen and Tohanson (19991 developed a PBPK model for evaluation of inhalation exposure
in humans to the gasoline additive ETBE. Model compartments for ETBE included lungs (with
arterial blood), liver, fat, rapidly perfused tissues, resting muscles, and working muscles. The same
set of compartments and an additional urinary excretion compartment were used for the
metabolite, tert-butanol. First-order metabolism was assumed in the model, and tissue/blood
partition coefficients were determined by in vitro methods fNihlen et al.. 19951. Estimates of
individual metabolite parameters of eight subjects were obtained by fitting the PBPK model to
experimental data from humans (5, 25, or 50 ppm ETBE; 2-hour exposure) fNihlen etal.. 1998al.
This model was applied primarily to predict levels of the biomarkers ETBE and tert-butanol in
blood, urine, and exhaled air after various scenarios, such as prolonged exposure, fluctuating
exposure, and exposure during physical activity fNihlen and Tohanson. 19991.
Rao and Ginsberg (19971 developed a PBPK model for MTBE and its principal metabolite,
tert-butanol, based on the Borghoff et al. f!9961 model. The modified model included a skin
compartment to simulate dermal absorption of MTBE during bathing or showering. A brain
compartment was added as a target organ for MTBE-induced neurological responses. MTBE
metabolism to tert-butanol was assumed to occur in the liver through two saturable pathways. The
tert-butanol portion of the model included further metabolism of tert-butanol in the liver,
exhalation in the lungs, and renal excretion (in the human model only). The model was validated
against published human and rat data and was used to help determine the contribution of tert-
butanol in the acute central nervous system effects seen after MTBE dosing.
The Rao and Ginsberg (19971 model used peak concentrations of MTBE and tert-butanol in
the blood and brain for interspecies, route-to-route, and low/high-dose extrapolations. The
MTBE/tert-butanol PBPK model was adapted to humans by adjusting physiology according to
literature values, incorporating the blood/air partition coefficient for humans reported by Tohanson
etal. f!9951. and allometrically scaling the metabolic rate based on body weight. A renal
elimination component was added to account for the small percentage of MTBE disposition that
occurs in humans via urinary excretion of tert-butanol. tert-Butanol concentrations in human blood
during and after MTBE exposure (25 or 50 ppm for 2 hours) were accurately predicted by the
human model (Tohanson etal.. 19951.
Kim etal. (20071 expanded the Borghoff et al. (19961 model to develop a multi-exposure
route model for MTBE and its primary metabolite, tert-butanol, in humans. The significant features
and advantages of the Kim etal. f20071 model are that parameters used for quantifying 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
36
37
38
Supplemental Information—tert Butanol
pharmacokinetic behavior of MTBE and tert-butanol are calibrated using time-series
measurements from controlled-exposure experiments in humans as reported by Prah et al. (20041.
MTBE partition coefficient values described in the Licata etal. f20011 model and skin compartment
parameters from the Rao and Ginsberg (19971 model were incorporated. The PBPK model for
MTBE consists of nine primary compartments representing the lungs, skin, fat, kidney, stomach,
intestine, liver, rapidly perfused tissue, and slowly perfused tissue. The tert-butanol model consists
of three compartments representing blood, liver, and other tissue.
Leavens and Borghoff (20091 developed a PBPK model for inhalation exposures in male and
female rats that expanded on Borghoff et al. (19961 and Rao and Ginsberg (19971 to include the sex-
specific effects of MTBE binding to a2U-globulin, a protein unique to male rats, and to describe the
induction of tert-butanol metabolism after repeated exposures. Although the primary purpose of
the model was to estimate MTBE and tert-butanol tissue concentrations after MTBE exposure, the
model was also parameterized to include inhalation uptake of tert-butanol. The tert-butanol portion
of the model was calibrated using data from rat exposures to tert-butanol as well as MTBE. Model
compartments included blood, brain, fat, gastrointestinal tissues, kidney, liver, poorly perfused
tissues (blood flow of <100 mL/min/100 g of tissue: bone, muscle, skin, fat), and rapidly perfused
tissues.
Distribution of MTBE and tert-butanol was assumed to be perfusion (i.e., blood-flow)
limited. This model used the same assumptions as Borghoff et al. f 19961 regarding MTBE
metabolism and kinetics, and assumed that tert-butanol was metabolized only in the liver through
one low-affinity pathway and excreted through urine. The model described binding of MTBE or
tert-butanol with a2U-globulin in the kidney, due to the high concentration of a2U-globulin in the
kidney. As chemicals bind to a2U-globulin, the rate of hydrolysis of the protein decreases and causes
accumulation in the kidney; however, there is no evidence that binding of a2U-globulin affects its
synthesis, secretion, or circulating concentrations [Borghoff et al. (1990) as cited in Leavens and
Borghoff f20091]. Equations describing this phenomenon were included in the model for male rats
only to account for the effects of the binding with a2u-globulin on metabolism of MTBE and tert-
butanol. Partition coefficient values in the model that differed from those published in previous
PBPK models included poorly perfused tissues:blood and kidney:blood values. The kidney:blood
value was based on calculated kidney:blood concentrations in female rats only because of the lack
of a2u-globulin-associated effects in female rats. The deposition of tert-butanol during inhalation in
the nasal cavity and upper airways was reflected in the high blood:air partition coefficient for tert-
butanol, and the ability of tert-butanol to induce its own metabolism after chronic exposure was
also taken into account. No differences in the induction of metabolism were reported between
males and females. The model simulated concentrations of MTBE and tert-butanol in the brain,
liver, and kidney of male and female rats following inhalation exposure at concentrations of 100,
400,1,750, or 3,000 ppm MTBE, and compared them to measured concentrations of MTBE and tert-
butanol from rats exposed at those levels.
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
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information—tert Butanol
Concentrations of MTBE and tert-butanol in the brain and liver were similar in male and
female rats during exposure and postexposure, but the concentrations of the chemicals in the
kidney were significantly different in male rats when compared with female rats. The additional
parameter accounting for a2U-globulin protein-binding in this PBPK model more accurately reflects
the metabolism of both MTBE and tert-butanol in male rat kidneys over time compared with other
PBPK models. The model highlights that binding can stimulate increased renal effects in male rats
after exposure to MTBE and tert-butanol. The assumptions made to reflect tert-butanol metabolism
induction and deposition in the nasal cavity and upper airways generally were supported by
measured data from rats exposed to 250, 450, or 1,750 ppm tert-butanol as evidenced by the fact
that the model was within one standard deviation of the mean concentrations for most data points.
However, the model overpredicted the concentration of tert-butanol in the brain, liver, and kidney
of male rats after repeated exposures.
Borghoff et al. (20101 modified the PBPK model of Leavens and Borghoff (20091 by adding
oral gavage and drinking water exposure components. This was done to compare different dose
metrics to the toxicity observed across different studies. The Borghoff et al. (20101 model assumed
first-order uptake of MTBE absorption from the gut, with 100% of the MTBE dose absorbed for
both drinking water and oral gavage exposures. They conducted a series of pharmacokinetic
studies comparing the effects of different rat strains and different dosing vehicles on the blood
concentration-time profiles of MTBE and tert-butanol following MTBE exposure. The effects of
exposure to MTBE via drinking water, oral gavage, and inhalation routes over 7 and 91 days on
male and female rats were modeled and compared with measured data collected from F344 rats
(exposed 28 days) and Wistar Han rats (exposed 14 and 93 days).
The model predicted the blood concentrations of tert-butanol that were observed after 250
or 1,000 mg/kg-day administration of MTBE in males and females, as well as the blood
concentrations of MTBE after 1,000 mg/kg-day, but was not able to predict peak concentrations of
MTBE after 250 mg/kg-day in males or females using either olive oil or 2% Emulphor as vehicles.
When comparing strains, the blood concentrations were similar across strain and sex, except in
female Sprague-Dawley rats administered 1,000 mg/kg-day MTBE. The female Sprague-Dawley
rats had a significantly (p-value not specified) higher blood concentration of both MTBE and tert-
butanol compared with the F344 and the Wistar Han females. However, the study authors
considered this an outlier and still considered the metabolic patterns similar. The model
overpredicted the amount of MTBE in the male rat kidney, but it accurately predicted the level of
tert-butanol in the male rat kidney at all exposures tested. The model did not accurately predict the
kidney concentrations of tert-butanol in the female kidney after exposure to MTBE via drinking
water, but the study authors attributed the inaccuracies to the study design as opposed to the
model formulation. All of the tert-butanol entering the submodel comes from MTBE metabolism in
the liver, and the model does not include a separate oral intake of tert-butanol.
This document is a draft for review purposes only and does not constitute Agency policy.
B-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
Supplemental Information—tert Butanol
B.2. PBPK MODEL EVALUATION SUMMARY
B.2.1. Evaluation of Existing tert-Butanol Submodels
The Blancato etal. (20071 and Leavens and Borghoff (20091 PBPK models for MTBE were
evaluated by comparing predictions from the tert-butanol portions of the models with the
tert- butanol i.v. data of Poet etal. (19971 (see Figure B-2). Neither model adequately represented
the tert-butanol blood concentrations. Modifications of model assumptions for alveolar ventilation,
explicit pulmonary compartments, and induction of metabolism of tert-butanol did not significantly
improve model fits to the data. Attempts to reoptimize model parameters in the tert-butanol
submodels of Blancato etal. f20071 and Leavens and Borghoff f20091 to match blood
concentrations from the i.v. dosing study were unsuccessful.
	300 mg/kg
~
male
o
female
- - 150 mg/kg
¦
male
~
female
	75 mg/kg
•
male
0
female
	37.5 mg/kg
A
male
a
female
10000
10000 =
1000
	300mg/kg	~	male
—	150 mg/kg	¦	male
	75 mg/kg	•	male
-	• -37.5 mg/kg	*	male
«	female
a	female
o	female
a	female
1000
0 2 4 6 8 10 12 14 16 18 20 22 24
time (hours)
4 6 8 10 12 14 16 18 20 22 24
time (hours)
Figure B-2. Comparison of the tert-butanol portions of existing MTBE models with
tert-butanol blood concentrations from i.v. exposure by Poet et al. 1997.
Neither the a] Blancato etal. (20071 nor the b] Leavens and Borghoff (20091 model adequately
represents the measured tert-butanol blood concentrations.
The PBPK submodel for tert-butanol in rats was developed in acslX (Advanced Continuous
Simulation Language, Aegis, Inc., Huntsville, Alabama) by adapting information from the many
PBPK models that were developed in rats and humans for the structurally related substance, MTBE,
and its metabolite tert-butanol fBorghoffetal.. 2010: Leavens and Borghoff. 2009: Blancato etal..
2007: Kim etal.. 2007: Rao and Ginsberg. 1997: Borghoff et al.. 19961. A brief description
comparing the Blancato etal. (20071 and (Leavens and Borghoff. 20091 models is given, followed by
an evaluation of the MTBE models and the assumptions adopted from MTBE models or modified in
the tert-butanol model.
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
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert Butanol
The Blancato etal. (20071 model is an update of the earlier Rao and Ginsberg (19971 model,
and the Leavens and Borghoff (20091 model is an update of the Borghoff et al. (19961 model. Both
the Blancato etal. f20071 and Leavens and Borghoff f20091 models are flow-limited models that
predict amounts and concentrations of MTBE and its metabolite tert-butanol in blood and six tissue
compartments: liver, kidney, fat, brain, and rapidly and slowly perfused tissues. These tissue
compartments are linked through blood flow, following an anatomically accurate, typical,
physiologically based description (Andersen. 19911. The parent (MTBE) and metabolite
(tert-butanol) models are interlinked by the metabolism of MTBE to tert-butanol in the liver. Routes
of exposure included in the models are oral and inhalation for MTBE; Leavens and Borghoff (20091
included inhalation exposure to tert-butanol. Oral doses are assumed to be 100% bioavailable and
100% absorbed from the gastrointestinal tract represented with a first-order rate constant.
Following inhalation of MTBE or tert-butanol, the chemical is assumed to directly enter the
systemic blood supply, and the respiratory tract is assumed to be at a pseudo-steady state.
Metabolism of MTBE by CYP450s to formaldehyde and tert-butanol in the liver is described with
two Michaelis-Menten equations representing high- and low-affinity enzymes, tert-Butanol is either
conjugated with glucuronide or sulfate or further metabolized to acetone through MPD and HBA;
both of these processes are described by a single Michaelis-Menten equation in the models. All of
these model assumptions are valid for tert-butanol and were applied to the EPA-developed tert-
butanol PBPK model, except for the separate brain compartment. The brain compartment was
lumped with other richly perfused tissues in the EPA tert-butanol PBPK model.
In addition to differences in parameter values between the Blancato etal. (20071 and the
Leavens and Borghoff (20091 models, there were three differences in the model structure: (1) the
alveolar ventilation was reduced during exposure, (2) the rate of tert-butanol metabolism increased
over time due to induction of CYP enzymes, and (3) binding of MTBE and tert-butanol to
a2u-globulin was simulated in the kidney of male rats. The Blancato etal. (20071 model was
configured through EPA's PBPK modeling framework, ERDEM (Exposure-Related Dose Estimating
Model), which includes explicit pulmonary compartments. The modeling assumptions related to
alveolar ventilation, explicit pulmonary compartments, and induction of metabolism of tert-butanol
are discussed in the model evaluation section.
MTBE and tert-butanol binding to a2U-globulin in the kidneys of male rats were
incorporated in the PBPK model of MTBE by Leavens and Borghoff f20091. Binding to a2U-globulin
is one hypothesized MOA for the observed kidney effects in MTBE-exposed animals. For a detailed
description of the role of a2U-globulin and other MOAs in kidney effects, see the kidney MOA section
of this document (see Section 1.1.1). Binding of MTBE to a2U-globulin was applied to sex differences
in kidney concentrations of MTBE and tert-butanol in the Leavens and Borghoff (20091 model, but
acceptable estimates of MTBE and tert-butanol pharmacokinetics in the blood are predicted in
other models that did not consider a2U-globulin binding. Given the uncertainty of tert-butanol
binding to a2U-globulin, it was not included in the tert-butanol PBPK submodel.
This document is a draft for review purposes only and does not constitute Agency policy.
B-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
Supplemental Information—tert Butanol
B.2.2. Modification of Existing tert-Butanol Submodels
To account for the tert-butanol blood concentrations after i.v. tert-butanol exposure, the
model was modified by adding a pathway for reversible sequestration of tert-butanol in the blood.
This could represent binding of tert-butanol to proteins in blood (see Figure B-3). The JPEC PK
studies showed that approximately 60% of the radiolabel in whole blood is in the plasma, providing
some limited evidence for association of tert-butanol with components in blood. The PBPK model
represented the rate of change in the amount of tert-butanol in the sequestered blood compartment
(Abiood2) with the following equation where Kon is the binding rate constant, CV is the free tert-
butanol concentration in blood, Koff is the unbinding rate constant, and Cbiood2 is the concentration
of tert-butanol bound in blood (equal to Abiood2/Vbiood).
dAblood2/dt = KoN*CV* - KoFF*Cblood2
Inhalation Exhalation
Figure B-3. Schematic of the PBPK submodel for tert-butanol in rats.
Exposure can be via multiple routes including inhalation, oral, or i.v. dosing. Metabolism of tert-
butanol occurs in the liver and is described by Michaelis-Menten equations with one pathway for
tert-butanol. tert-Butanol is cleared via exhalation and tert-butanol is additionally cleared via urinary
excretion. See Table B-l for definitions of parameter abbreviations.
This document is a draft for review purposes only and does not constitute Agency policy.
B-ll	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
Supplemental Information—tert Butanol
Table B-l. PBPK model physiologic parameters and partition coefficients.
Body weight and organ volumes as fraction of body weight
Body Weight (kg)
0.25 (Brown etal., 1977)
Body fraction that is blood perfused (Fperf)
0.8995 (Brown etal., 1977)
Liver
0.034 (Brown et al., 1977)
Kidney
0.007 (Brown et al., 1977)
Fat
0.07 (Brown et al., 1977)
Rapidly perfused
0.04 (Brown et al., 1977)
Slowly perfused
0.7485 a
Blood
0.074 (Brown et al., 1977)
Cardiac output and organ blood flows as fraction of cardiac output
Cardiac output (L/hr)
5.38 (Brown etal., 1977)b
Alveolar ventilation (L/hr)
5.38 (Brown etal., 1977)°
Liver
0.174 (Brown etal., 1977)d
Kidney
0.141 (Brown etal., 1977)
Fat
0.07 (Brown et al., 1977)
Rapidly perfused
0.279 e
Slowly perfused
0.336 (Brown etal., 1977)
Partition coefficients for tert-butanol
Blood:air
481 (Borghoff etal., 1996)
Liver:blood
0.83 (Borghoffetal., 1996)
Fat:blood
0.4 (Borghoff et al., 1996)
Rapidly perfused:blood
0.83 (Borghoffetal., 1996)
Slowly perfused:blood
1.0 (Borghoff et al., 1996)
Kidney:blood
0.83 (Borghoffetal., 1996)
a Fperf - Z(other compartments)
b 15.2*BW0'75 (bw = body weight)
c Alveolar ventilation is set equal to cardiac output
d Sum of liver and gastrointestinal blood flows

1 - Z(all other compartments).
The physiologic parameter values obtained from the literature are shown in Table B-l
fBrown etal.. 19771. tert-Butanol partition coefficients were obtained from literature in which they
were determined by the ratios of measured tissue:air and blood:air partition coefficients (Borghoff
etal.. 19961. The parameters describing rate constants of metabolism and elimination of
tert- butanol were obtained from the literature (Blancato etal.. 20071 and kept fixed because these
have been optimized to tert-butanol blood concentrations measured after MTBE exposure, which is
also metabolized to tert-butanol. The parameters describing tert-butanol absorption and
tert-butanol sequestration in blood were estimated by optimizing the model to the blood tert-
butanol time-course data for rats exposed via i.v., inhalation, and oral routes (Leavens and Borghoff.
2009: Poetetal.. 1997: ARCO. 19831. The model parameters were estimated with the acslX
This document is a draft for review purposes only and does not constitute Agency policy.
B-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
Supplemental Information—tert Butanol
optimization routine to minimize the log-likelihood function of estimated and measured
tert-butanol concentrations. The Nedler-Mead algorithm was used with heteroscedasticity allowed
to vary between 0 and 2. The predictions of the model with optimized parameters have a much
improved fit to the tert-butanol blood concentrations after tert-butanol i.v., as shown in panel A of
Figure B-4. Additionally, the model adequately estimated the tert-butanol blood concentrations
after inhalation and oral gavage exposures. The optimized parameter values are shown in Table
B-2. The ARCO (19831 study measured tert-butanol in plasma only unlike the Poetetal. (19971 and
Leavens and Borghoff (20091 studies that measured tert-butanol in whole blood. Based on the
measurements of plasma and whole blood by JPEC 2008, the concentration of tert- butanol in
plasma is approximately 60% of the concentration in whole blood. The tert-butanol plasma
concentrations measured by ARCO were increased (divided by 60%) to the expected concentration
in whole blood for comparison with the PBPK model.
A)	B)
400
350
!» 300
° 250
2
g 200
I 150
	1750 ppm m ale
~
17S0 ppm male
— — 1914 ppm female
O
1914 ppm female
450 ppm male
u
4S0 ppm male
— — 4S0 ppm female
~
4SO ppm female
	() ppm male
•
2S0 ppm male
— — 2 SO ppm female
o
2S0 ppm female
O 100
10 12 14 16 18 20 22 24
time (hours)
TBA iv exposure
¦ 1000
300 mg/kg	~male	o	female
—	— 150 mg/kg	¦ male	~	female
----75 mg/kg	• male	o	female
—	• —37.5 mg/kg	A male	A	female
6 8 10 12 14 16 18 20 22 24
time (hours)
c)
TBA gavage
— 500 mg/kg A
¦ - 1 mg/kg	•
6	9
time (hours)
Figure B-4. Comparison of the EPA model predictions with measured tert-butanol
blood concentrations for i.v., inhalation, and oral gavage exposure to tert-butanol.
A) i.v. data from Poet et al. f19971: B] inhalation data from Leavens and Borghoff f2009"l: and C] oral
gavage data from ARCO f!9831 with the optimized parameter values as shown in Table B-2.
This document is a draft for review purposes only and does not constitute Agency policy.
B-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
Supplemental Information—tert Butanol
Table B-2. Rate constants for tert-butanol determined by optimization of the
model with experimental data.
Parameter
Value
Source or Reference
Metabolism (VMTBA; mg/kg-hr)a
8.0
Blancato et al. (2007)
Metabolism (KMTBA; mg/L)
28.8
Blancato et al. (2007)
Urinary elimination (KEUM2; l/hr)
0.5
Blancato et al. (2007)
TBA sequestration rate constant (K0N; L/hr)
0.148
Optimized
TBA unsequestration rate constant (K0FF; L/hr)
0.0134
Optimized
Absorption from gastrointestinal tract (KAS2; l/hr)
0.5
Optimized
a scaled by BW0'7 (0.25°7 = 0.379), bw = body weight.
Induction of tert-butanol-metabolizing enzymes was included in the Leavens and Borghoff
(20091 model of MTBE based on their study of rats exposed for 8 days to tert-butanol via inhalation.
The enzyme induction equation and parameters developed in the Leavens and Borghoff f20091
model that were applied to the tert-butanol submodel are as follows.
Vmax tert-butanol IND = Vmax tert-butanol *INDMAX(l-exp(-KIND*t))
Vmax tert-butanol IND is the maximum metabolic rate after accounting for enzyme induction, Vmax
tert-butanol is the metabolism rate constant from Table B-2 for both tert-butanol pathways, and
INDMAX is the maximum percent increase in Vmax tert-butanol (124.9). KIND is the rate constant
for enzyme induction (0.3977/day). The increased tert-butanol metabolism better estimates the
measured tert-butanol blood concentrations as can be seen in the comparison of the model
predictions and experimental measurements shown in Figure B-5. The model better predicted
blood concentrations in female rats than male rats. The male rats had lower tert-butanol blood
concentrations after repeated exposures compared with female rats, and this difference could
indicate greater induction of tert-butanol metabolism or other physiologic changes such as
ventilation or urinary excretion in males. The current data for tert-butanol metabolism do not
provide sufficient information for resolving this difference between male and female rats.
B.2.3. Summary of the PBPK model for tert-butanol
A PBPK model for tert-butanol was developed by adapting previous models for MTBE and
tert-butanol (Blancato etal. (20071: Leavens and Borghoff (200911. Published tert-butanol models
(or sub-models) do not adequately represent the tert-butanol blood concentrations measured in
the i.v. study (Poet et al. 1997). The addition of a sequestered blood compartment for tert-butanol
substantially improved the model fit. The alternative modification of changing to diffusion-limited
distribution between blood and tissues also improved the model fit, but was considered less
biologically plausible. Physiological parameters and partition coefficients were obtained from
This document is a draft for review purposes only and does not constitute Agency policy.
B-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
Supplemental Information—tert Butanol
published measurements. The rate constants for tert-butanol metabolism and elimination were
from a published PBPK model of MTBE with a tert-butanol subcompartment (Blancato et al.
£2007}). Additional model parameters were estimated by calibrating to data sets for i.v., oral and
inhalation exposures as well as repeated dosing studies for TBA. Overall, the model produced
acceptable fits to multiple rat time-course datasets of TBA blood levels following either inhalation
or oral gavage exposures.
B.2.4. tert-Butanol Model Application
The PBPK model as described above was applied to toxicity studies to predict tert-butanol
blood concentrations (the preferred internal dose metric). For simulation studies where tert-
butanol was administered in drinking water, the consumption was modeled as episodic, based on
the pattern of drinking observed in rats fSpiteri. 19821.
B.2.5. PBPK Model Code
The PBPK acslX model code is made 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, 201#, HEROID##).
This document is a draft for review purposes only and does not constitute Agency policy.
B-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
Supplemental Information—tert Butanol
Figure B-5. Comparison of the EPA model predictions with measured amounts of tert-
butanol in blood after repeated inhalation exposure to tert-butanol.
Male rats were exposed to 239, 444, or 1726 ppm and female rats were exposed to 256,444, or
1914 ppm tert-butanol for up to 8 consecutive days (Borghoff et al.. 20011 tert-Butanol blood
concentrations are better predicted by the model after 8 days of exposure with enzyme induction
(right panels) compared to without enzyme induction (left panels).
B.3. OTHER PERTINENT TOXICITY INFORMATION
B.3.1. 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 formation, and DNA strand breaks and
adducts. The available genotoxicity data for tert-butanol are discussed below, and the summary of
the data is provided in Table B-3.
B.3.1.1. Bacterial Systems
The mutagenic potential of tert-butanol has been tested by Zeiger etal. (1987) 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,
either in the presence or absence of S9 metabolic activation.
Conflicting results have been obtained with tert-butanol-induced mutagenicity in 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
This document is a draft for review purposes only and does not constitute Agency policy.
B-16	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert Butanol
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 as to why the revertants decreased at higher concentrations. The results of
this study indicated that tester strain TA102 may be a more sensitive strain for monitoring tert-
butanol levels fWilliams-Hill etal.. 19991. However, in another study by Mcgregor etal. f20051.
experiments were conducted on Salmonella strain TA102 in two different laboratories using similar
protocols, tert-Butanol was dissolved in dimethyl sulfoxide (DMSO) or distilled water and tested
both in the presence and absence of S9 metabolic activation. No statistically significant increase in
mutants was observed in either of the solvent media. In one experiment where tert-butanol was
dissolved in water, a significant, dose-related increase in the number of revertants was produced,
reaching almost two-fold the control value at a concentration of 2,250 [ig/plate. It should be noted
that DMSO is known to be a free radical scavenger, and its presence at high concentrations might
mask a mutagenic response caused due to oxidative damage.
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.. 19491. On the other hand, tert- butanol, without exogenous metabolic activation,
significantly increased the frequency of petite mutations (the mitochondrial deoxyribonucleic acid
[DNA] deletion rho-) in Saccharomyces cerevisiae laboratory strains K5-A5, MMY1, D517-4B, and
DS8 (Timenez etal.. 19881. 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.. 19881.
B.3.1.2. 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 both in
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
increase in mutant fraction atthe highest tested concentration (5,000 [ig/mL) f McGregor etal..
19881.
To further determine potential DNA and/or chromosomal damage induced by tert-butanol
in in vitro systems, (NTP. 19951 studied sister chromatid exchanges (SCEs) and chromosomal
aberrations (CAs). Chinese hamster ovary (CHO) cells were exposed to tert-butanol both in the
This document is a draft for review purposes only and does not constitute Agency policy.
B-17	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information—tert Butanol
presence and absence of S9 activation at concentrations of 160-5,000 |J.g/mL for 26 hours, tert-
Butanol did not induce SCEs in any of the concentrations tested, although in one experiment, there
was a slight increase in percent relative change of SCEs per chromosome scored. The same authors
also studied the effect of tert-butanol on CA formation. CHO cells were exposed to four
concentrations (160, 500,1,600, or 5,000 [ig/mL) of tert-butanol both in the presence and absence
of S9. No significant increase in CAs was observed in any of the concentrations tested. It should be
noted 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. (20091 examined the effects of tert-butanol on DNA damage using 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 was then 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 were also 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 in a dose-dependent
manner as measured by the number of cells after 24 and 48 hours of exposure at IC50
concentrations, as well as with concentrations at 1/10th the IC50. There was no increase in cell
death, suggesting a reduction in cell number due to reduced replication rather than cytotoxicity.
tert-Butanol caused an accumulation in the G0/G1 phase of replication. These were related to
different effects on the expression of cyclin Dl, p27Kipl, and p53 genes. An initial increase in DNA
damage as measured by nuclear fragmentation was observed at the 30-minute timepoint 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 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 a total of 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.3.1.3. In Vivo Mammalian Studies
A limited number of 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 that were 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 formation in peripheral blood lymphocytes was
This document is a draft for review purposes only and does not constitute Agency policy.
B-18	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information—tert Butanol
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. 19951.
Male Kumming mice (8 per treatment) were administered 0, 0.099, 0.99,10,101, or
997 |ig/kg bw 14C-tert-butanol in saline via gavage with specific activity ranging from 1.60 to
0.00978 mCi/mol fYuan etal.. 20071. 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 may 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-19	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table B-3. Summary of genotoxicity (both in vitro and in vivo) studies of tert-
2	butanol.
Test system
Dose/ Cone.
Results3
Comments
Reference
Bacterial Systems

-S9
+S9

Reverse Mutation Assay
Salmonella typhimurium
(TA98, TA100, TA1535,
TA1537, TA1538)
100, 333, 1000,
3333, 10,000
Hg/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
Hg/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, 1000,
1500, 2500, 5000
Hg/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, 1000, 1250,
2000, 3000, 4000,
5000 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, 1600,
2000, 3000, 4000,
5000 Hg/mL
-
-
This study was part of the NTP
1995 testing results
Galloway, 1987;
NTP (1995)
Chromosomal Aberrations,
Chinese Hamster Ovary cells
160, 500, 1600,
2000, 3000, 4000,
5000 Hg/mL


This study was part of the NTP
1995 testing results
Galloway, 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-20	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Supplemental Information—tert Butanol
Test system
Dose/ Cone.
Results3
Comments
Reference
DNA damage, (comet
assay), human HL-60
leukemia cells
1, 5,10, 30 mmol/L
+
ND
Exposure duration - lh
Tang et al.
(1997)
In vivo Animal Studies
Micronucleus formation,
B6C3F1 mouse peripheral
blood cells
3000, 5000,
10,000, 20,000,
40,000 ppm

13-week, subchronic, drinking
water study
NTP (1995)
DNA adducts, male
Kunming mouse liver,
kidney and lung cells
0.1-1000 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.3.2. Summary
tert-Butanol has been tested for its genotoxic potential using a variety of genotoxicity
assays. Bacterial assays that detect reverse mutations have been thought to predict carcinogenicity
with accuracy up to 80%. 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 may impact 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 2-fold 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 as a result of 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 exchange or chromosomal aberrations were not observed in CHO cells in
response to tert-butanol treatment. However, DNA damage was detected using comet assay in both
rat fibroblasts and human 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 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
This document is a draft for review purposes only and does not constitute Agency policy.
B-21	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Supplemental Information—tert Butanol
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, it is not possible to discern whether DNA-repair
mechanisms would occur after a longer period of observation.
Limited in vivo animal studies have been conducted on DNA adduct formation or
micronucleus formation. 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
this method may be sensitive and not specific to the adducts in question. The method uses
14C-labeled chemical for dosing, and the isolated DNA is oxidized to carbon dioxide and reduced to
filamentous graphite, and the ratios of 14C/12C are measured. The ratio is then converted to DNA
adducts based on nucleotide content of the DNA, hence the debate for the reliability of the data
obtained. Confirmation of this data will provide assurance in understanding the mechanism of
tert-butanol-induced DNA adducts. No increase in micronucleus formation was observed in mouse
peripheral blood cells in a 13-week drinking water study conducted by the National Toxicology
Program.
Overall, there is a limited database to understand the role of tert-butanol-induced
genotoxicity for mode of action and carcinogenicity. The database is limited either in terms of 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 may impact results. Furthermore, in some studies,
the methodology used has been challenged for its specificity. 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 genotoxicty 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-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
Supplemental Information—tert Butanol
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. BENCHMARK DOSE MODELING SUMMARY
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 describes 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, it may be appropriate to use alternative methods based
on statistical judgment; exceptions are noted as necessary in the summary of the modeling results.
C.l.l. Noncancer Endpoints
C.l.1.1. Evaluation of Model Fit
For each dichotomous endpoint, BMDS dichotomous models were fitted to the data using
the maximum likelihood method. 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 were also 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 each continuous endpoint, BMDS continuous models were fitted to the data using the
maximum likelihood method, and 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 fitted to the data assuming constant variance. If
Test 2 was rejected (x2 p-value < 0.10), the variance was modeled as a power function of the mean,
and the variance model was tested for adequacy of fit using a likelihood ratio test (BMDS Test 3).
For fitting models using either constant variance or modeled variance, models for the mean
response were tested for adequacy of fit using a likelihood ratio test (BMDS Test 4, with x2
p-value <0.10 indicating inadequate fit). Other factors were also used to assess the model fit, such
as scaled residuals, visual fit, and adequacy of fit in the low-dose region and near the BMR.
C.l.1.2. Model Selection
For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
estimated by the profile likelihood method) and the Akaike's information criterion (AIC) value were
This document is a draft for review purposes only and does not constitute Agency policy.
C-l	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	used to select a best-fit model from among the models exhibiting adequate fit If the BMDL
2	estimates were "sufficiently close," that is, differed by at most 3-fold, the model selected was the
3	one that yielded the lowest AIC value. If the BMDL estimates were not sufficiently close, the lowest
4	BMDL was selected as the POD.
5	Table C-l. Non-cancer endpoints selected for dose-response modeling for
6	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
Mean relative
kidney weight
NTP (1995)
Rat (F344)
/ Male
Dose (mg/kg-d)
0
90
200
420
Mean ± SE (n)
3.68 ±0.09
(10)
3.96 ±0.13
(10)
4.22 ±0.13
(10)
4.42 ±0.15
(10)
Mean relative
kidney weight
NTP (1995)
Rat (F344)
/ Female
Dose (mg/kg-d)
0
180
330
650
Mean ± SE (n)
3.49 ± 0.08
(10)
3.99 ±0.07
(10)
4.21 ±0.08
(10)
4.95 ±0.17
(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
Thyroid follicular
cell hyperplasia
NTP (1995)
Mouse
(B6C3F!)/
Male
Dose (mg/kg-d)
0
540
1,040
2,070
Incidence /
Total
5/60
18/59
15/59
18/57
Thyroid follicular
cell hyperplasia
NTP C19951
Mouse
(B6C3F1) /
Female
Dose (mg/kg-d)
0
510
1,020
2,110
Incidence /
Total
19/58
28/60
33/59
47/59
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 relative
kidney weight
NTP (1997)
Rat (F344)
/ Male
Concentration
(mg/m3)
0
406
825
1643
3274
6369
Mean ± SD (n)
3.68 ±
0.253
(10)
3.71 ±
0.12
(9)
3.64 ±
0.126
(10)
3.76 ±
0.19
(10)
3.96 ±
0.158
(10)
4 ±
0.158
(10)
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 Butanol
Endpoint/Study
Species/
Sex
Doses and Effect Data
Increased absolute
kidney weight
NTP (1997)
Rat (F344)
/ Female
Concentration
(mg/m3)
0
406
825
1643
3274
6369
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)
Increased relative
kidney weight
NTP (1997)
Rat (F344)
/ Female
Concentration
(mg/m3)
0
406
825
1643
3274
6369
Mean ± SD (n)
4.00 ±
0.474
(10)
3.98 ±
0.190
(10)
4.03 ±
0.158
(10)
4.14 ±
0.126
(10)
4.09 ±
0.190
(10)
4.35 ±
0.095
(10)
1	C.l.1.3. Modeling Results
2	Below are tables summarizing the modeling results for the noncancer endpoints modeled.
3	The following parameter restrictions were applied, unless otherwise noted.
4	• Dichotomous models: For the log-logistic and dichotomous Hill models, restrict slope >
5	1; for the gamma and Weibull models, restrict power > 1; for the multistage models,
6	restrict beta values > 0.
7	• Continuous models: For the polynomial models, restrict beta values > 0; for the Hill,
8	power, and exponential models, restrict power > 1.
9	Table C-2. Summary of BMD modeling results for kidney transitional epithelial
10	hyperplasia in male F344 rats exposed to tert-butanol in drinking water for 2
11	years (NTP. 1995): BMR = 10% extra risk.
Model3
Goodness of fit
BMDio
(mg/kg-d)
BMDL10
(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
12	aScaled residuals for selected model for doses 0, 90, 200, and 420 mg/kg-d were -0.076, 0.147, 0.046, and -0.137,
13	respectively.
14
This document is a draft for review purposes only and does not constitute Agency policy.
C-3	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
Log-Logistic Model with 0.95 Confidence Level
dose
17:16 05/13 2011
Figure C-l. Plot of mean response by dose, with fitted curve for selected model.
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
64
65
66
67
68
69
70
71
Supplemental Information—tert Butanol
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
background =
intercept =
slope =
Parameter Values
0.5
-5.54788
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
Variable
background
intercept
slope
Estimate
0.505366
-5.58826
1	*
95.0% Wald Confidence Interval
Std. Err. Lower Conf. Limit Upper Conf. Limit
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-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
Supplemental Information—tert Butanol
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
Chi^2 = 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
Table C-3. Summary of BMD modeling results 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.
Model3
Goodness of fit
BMD10
(mg/kg-d)
BMDL10
(mg/kg-d)
Basis for model selection
p-value
AIC
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 differed by less than 3-
fold).
Logistic
0.50
92.81
461
393
LogLogistic
0.79
91.57
414
333
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
Error"
Error"
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.
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
Supplemental Information—tert Butanol
Multistage Model with 0.95 Confidence Level
dose
17:18 05/13 201 1
Figure C-2. Plot of mean response by dose, with fitted curve for selected model.
Multistage Model. (Version: 3.2; Date: 05/26/2010)
Input Data File: M:\NCEA terf-blltanol\BMD modeling\BMDS Output\20 NTP
1995b_Kidney transitional epithelial hyperplasia, female rats_Multi3_10.(d)
Gnuplot Plotting File: M:\NCEA tCTt-hutS nol\BMD modeling\BMDS Output\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
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
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
Supplemental Information—tert Butanol
( *** 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	q	*	*	*
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
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
Chi^2 = 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 = 33 8.618
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 Butanol
1	Table C-4. Summary of BMD modeling results for relative kidney weights in
2	male F344 rats exposed to tert-butanol in drinking water for 15 months (NTP.
3	1995): BMR = 10% relative deviation and 1 standard deviation.
Model3
Goodness of fit
BMD10%
(mg/kg-d)
BMDL10%
(mg/kg-d)
BMD1sd
(mg/kg-d)
BMDL1sd
(mg/kg-d)
Basis for model
selection
p-value
AIC
Hill
NA
-27.27
120
39
124
45
Exponential (M4) is
selected as the
best-fitting model
based on visual fit
at the low-dose
region.
Exponential
(M4)
0.854
-29.23
117
48
123
53
Exponential
(M5)
N/A
-27.27
121
48
126
54
Linear
0.421
-29.54
222
155
229
161
Polynomial
0.421
-29.54
222
155
229
161
Power
0.421
-29.54
222
155
229
161
Exponential
(M2)
0.365
-29.25
236
170
243
176
Exponential
(M3)
0.365
-29.25
236
170
243
176
aConstant variance case presented (BMDS Test 2 p-value = 0.466), selected model in bold; scaled residuals for
selected model for doses 0, 90, 200, 420 mg/kg-d were 0.04009, -0.1264, 0.122, and -0.03578, respectively.
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
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
Supplemental Information—tert Butanol
Exponential Model 4 with 0.95 Confidence Level
dose
Figure C-3. Plot of mean response by dose, with fitted curve for selected model (10%
relative deviation).
Exponential Model. (Version: 1.7; Date: 12/10/2009)
Input Data File: M:\NCEA t-Butanol\BMD modeling\BMDS Output\21 NTP
1995b_Mean relative kidney weight, male rats_ExpCV_10RD.(d)
Gnuplot Plotting File: M:\NCEA t-Butanol\BMD modeling\BMDS Output\21 NTP
1995b_Mean relative kidney weight, male rats_ExpCV_10RD.pit
Fri May 13 16:30:21 2011
[notes]
The form of the response function by Model:
Model 2
Model 3
Model 4
Model 5
Y[dose] = a
Y[dose] =	a
Y[dose] =	a
Y[dose] =	a
exp{sign
exp{sign
[c-(c-l)
[c-(c-l)
b * dose}
(b * dose)Ad}
exp{-b * dose}]
exp{-(b * dose)'
^d}]
Note: Y[dose] is the median response for exposure
sign = +1 for increasing trend in data;
sign = -1 for decreasing trend.
dose ;
Model 2 is nested within Models 3 and 4.
Model 3 is nested within Model 5.
Model 4 is nested within Model 5.
Dependent variable = Response
Independent variable = Dose
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
rho is set to 0.
A constant variance model is fit.
Total number of dose groups = 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
This document is a draft for review purposes only and does not constitute Agency policy.
C-10	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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 Butanol
MLE solution provided: Exact
Initial Parameter Values
Variable
lnalpha
rho(S)
a
b
Model 4
-1.93171
0
3. 496
0.00417714
1.32752
1
(S) = Specified
Parameter Estimates
Variable	Model 4
lnalpha
rho
a
b
c
d
-1.93087
3.67517
0.00469937
1.23673
1
Table of Stats From Input Data
Dose N	Obs Mean Obs Std Dev
0
90
200
420
10
10
10
10
3.68	0.2846
3.96	0.4111
4.22 0.4111
4.42 0.4743
Estimated Values of Interest
Dose Est Mean Est Std Scaled Residual
0	3.675	0.3808	0.04009
90	3.975	0.3808	-0.1264
200 4.205 0.3808 0.1221
420 4.424 0.3808 -0.03578
Other models for which likelihoods are calculated:
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma (i) A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)
Model R: Yij = Mu + e(i)
Var{e(ij)} = SigmaA2
Likelihoods of Interest
Model Log(likelihood) DF	AIC
This document is a draft for review purposes only and does not constitute Agency policy.
C-ll	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
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
Supplemental Information—tert Butanol
Al
18.63423
5
-27.26846
A2
19.91058
8
-23.82116
A3
18.63423
5
-27.26846
R
10.08355
2
-16.1671
4
18.61733
4
-29.23465
Additive constant for all log-likelihoods = -36.76. This constant added to the
above values gives the log-likelihood including the term that does not
depend on the model parameters.
Explanation of Tests
Test 1: Does response and/or variances differ among Dose levels? (A2 vs. R)
Test 2: Are Variances Homogeneous? (A2 vs. Al)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 6a: Does Model 4 fit the data? (A3 vs 4)
Test
Test 1
Test 2
Test 3
Test 6a
Tests of Interest
-2*log(Likelihood Ratio)
19. 65
2.553
2.553
0.03381
p-value
0.00319
0.4658
0.4658
0.8541
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose
levels, it seems appropriate to model the data.
The p-value for Test 2 is greater than .1. A homogeneous
variance model appears to be appropriate here.
The p-value for Test 3 is greater than .1. The modeled
variance appears to be appropriate here.
The p-value for Test 6a is greater than .1. Model 4 seems
to adequately describe the data.
Benchmark Dose Computations:
Specified Effect = 0.100000
Risk Type = Relative deviation
Confidence Level = 0.950000
BMD = 116.807
BMDL = 48.0466
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 Butanol
1	Table C-5. Summary of BMD modeling results for relative kidney weights in
2	female F344 rats exposed to tert-butanol in drinking water for 15 months
3	(NTP. 1995): BMR = 10% relative deviation and 1 standard deviation.
Model3
Goodness of fit
BMDio%
(mg/kg-d)
BMDLio%
(mg/kg-d)
BMDisd
(mg/kg-d)
BMDLisd
(mg/kg-d)
Basis for model
selection
p-value
AIC
Exponential
(M2)
Exponential
(M3)
0.48
-49.14
178
154
108
80
The linear model
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).
Exponential
(M4)
Exponential
(M5)
0.33
-47.64
154
107
90
56
Hill
0.33
-47.64
154
105
90
Errorb
Linear
Power
0.62
-49.63
158
133
92
68
Polynomial
3°
0.33
-47.63
158
133
98
68
aModeled variance case presented (BMDS Test 2 p-value = 0.0091), selected model in bold; scaled residuals for
selected model for doses 0,180, 330, and 650 mg/kg-d were -0.383, 0.887, -0.411, and -0.105, respectively.
bThe BMDL1sd computation failed for the Hill model.
Linear Model with 0.95 Confidence Level
O	10O	200	300	400	500	600
.	dose
4	16:36 05/13 2011
5	Figure C-4. Plot of mean response by dose, with fitted curve for selected model (10%
6	relative deviation).
7	====================================================================
8	Polynomial Model. (Version: 2.16; Date: 05/26/2010)
9	Input Data File: M:\NCEA te/"t"butanol\BME) modelincABMDS Output\23 NTP 1995b Mean
10 relative kidney weight, female rats_Linear_10RD.(d)
This document is a draft for review purposes only and does not constitute Agency policy.
C-13	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 Butanol
Gnuplot Plotting File: M:\NCEA te/"t"butanol\BMD modeling\BMDS Output\23 NTP
1995b_Mean relative kidney weight, female rats_Linear_10RD.plt
Mon May 09 18:34:15 2011
[notes]
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/x2 + ...
Dependent variable = Response
Independent variable = Dose
Signs of the polynomial coefficients are not restricted
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 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
Default Initial Parameter Values
lalpha = -2.14986
rho =	0
beta_0 = 3.52312
beta 1 = 0.00219613
Asymptotic Correlation Matrix of Parameter Estimates
lalpha	rho beta_0 beta_l
lalpha	1	-1	0.1 -0.14
rho	-1	1	-0.1	0.14
beta_0	0.1	-0.1	1 -0.66
beta 1 -0.14	0.14 -0.66	1
Parameter Estimates
95.0% Wald Confidence Interval
Variable	Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
lalpha	-8.78559	2.23999	-13.1759	-4.39529
rho	4.47471	1.57167	1.39429	7.55513
beta_0	3.51492 0.0580177	3.40121	3.62864
beta 1 0.00223049 0.0002217 6	0.0017 9585	0.0026 6513
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
10
3.49
180
10
3 . 99
330
10
4 .21
650
10
4 . 95
3.51	0.253
3.92	0.221
4.25	0.253
4.96	0.538
0.206	-0.383
0.262 0.887
0.315 -0.411
0.446 -0.105
This document is a draft for review purposes only and does not constitute Agency policy.
C-14	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
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 Butanol
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = SigmaA2
Likelihoods of Interest
Model	Log(likelihood) # Param's AIC
A1	25.104490	5 -40.208981
A2	30.882250	8 -45.764500
A3	29.295765	6 -46.591531
fitted	28.815603	4 -49.631206
R	-0.698257	2 5.396514
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (A1 vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 63.161	6	<.0001
Test 2	11.5555	3	0.009072
Test 3	3.17297	2	0.2046
Test 4	0.960325 2	0.6187
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect =	0.1
Risk Type = Relative risk
Confidence level =	0.95
BMD = 157.585
BMDL = 132.699
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
Supplemental Information—tert Butanol
Table C-6. Summary of BMD modeling results for kidney inflammation in
female rats exposed to tert-butanol in drinking water for 2 years (NTP. 1995):
BMR = 10% extra risk.
Model3
Goodness of fit
BMDio%
(mg/kg-d)
BMDLio%
(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.
LogProbit Model with 0.95 Confidence Level
dose
17:17 05/13 201 1
Figure C-5. Plot of mean response by dose, with fitted curve for selected model.
Probit Model. (Version: 3.2; Date: 10/28/2009)
Input Data File: M:/NCEA te/"t"butanol/BMD modeling/BMDS Output/19 NTP
19 95b_Kidney inflammation, female rats_LogProbit_10.(d)
Gnuplot Plotting File: M: /NCEA te/"t-blltanol/BMD modeling/BMDS Output/19 NTP
1995b_Kidney inflammation, female rats_LogProbit_10.pit
Fri May 13 17:17:59 2011
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
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 Butanol
[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
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
This document is a draft for review purposes only and does not constitute Agency policy.
C-17	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Supplemental Information—tert Butanol
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
-0.700
330.0000
0.1859
9.295
13.000
50
1.347
650.0000
0.3899
19.495
17.000
50
-0.724
Chi/X2 = 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-18	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-7. Summary of BMD modeling results for thyroid follicular cell
2	hyperplasia in male B6C3F1 mice exposed to tert-butanol in drinking water
3	for 2 years (NTP. 1995): BMR = 10% extra risk.
Model3
Goodness of fit
BMDio%
(mg/kg-d)
BMDLio%
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.052
254.7
702
430
No model was selected as a
best-fitting model as models
did not fit the overall
goodness-of-fit criterion.
Logistic
0.031
256.1
1,064
751
LogLogistic
0.069
254.1
586
340
LogProbit
0.012
258.2
1,320
810
Multistage 3°
Weibull
0.052
254.7
702
430
Probit
0.032
255.9
1,020
713
Dichotomous-Hill
N/Ab
253.6
0.19
Error"
aNo model was selected as a best-fitting model.
bNo available degrees of freedom to estimate a p-value.
c BMDL computation failed for the Dichotomous-Hill model.
4
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 Butanol
1	Table C-8. Summary of BMD modeling results for thyroid follicular cell
2	hyperplasia in female B6C3F1 mice exposed to tert-butanol in drinking water
3	for 2 years (NTP. 1995): BMR = 10% extra risk.
Model3
Goodness of fit
BMD io
(mg/kg-d)
BMDL10
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.63
303.1
327
154
The probit model 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
0.94
301.0
297
243
LogLogistic
0.52
303.2
375
115
LogProbit
0.48
303.3
388
277
Multistage 3°
0.81
302.9
269
155
Probit
0.95
300.9
298
246
Weibull
0.66
303.0
321
154
Dichotomous-Hill
0.66
27,226
Errorb
Errorb
aSelected model in bold; scaled residuals for selected model for doses 0, 510,1,020, and 2,110 mg/kg-d were -
0.110, 0.255, -0.174, and 0.025, respectively.
bThe BMD and BMDL computations failed for the Dichotomous-Hill model.
Pro bit Model with 0.95 Confidence Level
0.2 r
BMDU BMD
dose
4	18:36 05/09 2011
5	Figure C-6. Plot of mean response by dose, with fitted curve for selected model.
6	====================================================================
7	Probit Model. (Version: 3.2; Date: 10/28/2009)
8	Input Data File: M:\NCEA te/"t-blltanol\BMD modeling\BMDS Output\27 NTP
9	15 55b_Thyroid follicular cell hyperplasia, female mice_Probit_10.(d)
10	Gnuplot Plotting File: M:\NCEA te/"t-blltanol\BMD modeling\BMDS Output\27 NTP
11	1995b Thyroid follicular cell hyperplasia, female mice Probit 10.pit
12	Mon May 09 18:36:50 2011
This document is a draft for review purposes only and does not constitute Agency policy.
C-20	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
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 Butanol
[notes]
The form of the probability function is:
P[response] = CumNorm(Intercept+Slope*Dose),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Incidence
Independent variable = Dose
Slope parameter is not restricted
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
Default Initial (and Specified) Parameter Values
background =	0 Specified
intercept = -0.425261
slope = 0.000589168
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept slope
intercept	1 -0.76
slope -0.76	1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err.	Lower Conf. Limit Upper Conf. Limit
intercept -0.427828 0.129459	-0.681563 -0.174092
slope 0.000593756 0.00011419	0.000369947 0.000817564
Analysis of Deviance Table
Model Log(likelihood)	# Param's Deviance Test d.f. P-value
Full model -148.416	4
Fitted model -148.47	2 0.108205 2	0.9473
Reduced model -162.896	1 28.9589 3	<.0001
AIC:	300.941
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.3344 19.395 19.000 58	-0.110
510.0000 0.4503 27.015 28.000 60	0.255
1020.0000 0.5706 33.663 33.000 59	-0.174
2110.0000 0.7953 46.923 47.000 59	0.025
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
10
11
12
13
14
15
16
17
Supplemental Information—tert Butanol
ChiA2 = 0.11 d.f. = 2 P-value = 0.9473
Benchmark Dose Computation
Specified effect =	0.1
Risk Type = Extra risk
Confidence level =	0.95
BMD = 297.997
BMDL = 2 4 6.07 5
This document is a draft for review purposes only and does not constitute Agency policy.
C-22	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-9. 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.
4
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
The Hill model was selected as
Exponential (M3)
<0.0001
-203.06
9.2E+07
7094
the only adequately-fitting
model.
Exponential (M4)
<0.0001
-203.06
errorb
0
Exponential (M5)
<0.0001
-201.06
errorb
0

Hill
0.763
-226.82
1931
1705

Powerc
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

a Constant variance case presented (BMDS Test 2 p-value = 0.390), selected model in bold; scaled residuals for
selected model for doses 0, 406, 825,1643, 3274, and 6369 mg/m3 were 0.395, 0.374, -0.75, -1.96e-006, 0.381,
and -0.381, respectively.
b BMC or BMCL computation failed for this model.
c For the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear
model.
d For 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.
6 For 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.
f BMC or BMCL computation failed for this model
5
This document is a draft for review purposes only and does not constitute Agency policy.
C-23	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information—tert Butanol
Hill Model with 0.95 Confidence Level
dose
Figure C-7. Plot of mean response by dose, with fitted curve for selected
model; dose 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
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 Butanol
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
nterest
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
Tests of Interest
Test
2*log(Likelihoo
d 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
6
7
This document is a draft for review purposes only and does not constitute Agency policy.
C-25	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-10. Summary of BMD modeling results for relative 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.
4
Model3
Goodness of fit
BMCiord
(mg/m3)
BMCLiord
(mg/m3)
Basis for model selection

p-value
AIC

Exponential (M2)
Exponential (M3)b
0.168
-141.06
6356
4923
The linear model is selected
on the basis of lowest AIC.
Exponential (M4)
0.169
-140.46
5973
3386

Exponential (M5)
0.560
-142.35
errorc
0

Hill
0.612
-142.53
errorc
errorc

Powerd
Polynomial 5°e
Polynomial 4°f
Polynomial 3°g
Polynomial 2°h
Linear
0.181
-141.25
6309
4821

a Constant variance case presented (BMCS Test 2 p-value = 0.165), selected model in bold; scaled residuals for
selected model for doses 0, 406, 825,1643, 3274, and 6369 mg/m3 were 0.181, 0.282, -1.42, -0.102,1.81, and -
0.736, respectively.
b For the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
c BMC or BMCL computation failed for this model.
d For the Power model, the power parameter estimate was 1. The models in this row reduced to the Linear
model.
6 For the Polynomial 5° model, the b5, b4, and b3 coefficient estimates were 0 (boundary of parameters space).
The models in this row reduced to the Polynomial 2° model. For the Polynomial 5° model, the beta coefficient
estimates were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
f For the Polynomial 4° model, the b4 and b3 coefficient estimates were 0 (boundary of parameters space). The
models in this row reduced to the Polynomial 2° model. For the Polynomial 4° model, the b4, b3, and b2
coefficient estimates were 0 (boundary of parameters space). The models in this row reduced to the Linear
model.
8 For the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
h For the Polynomial 2° model, the b2 coefficient estimate was 0 (boundary of parameters space). The models in
this row reduced to the Linear model.
5
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
Supplemental Information—tert Butanol
Linear Model with 0.95 Confidence Level
10:24 04/30 2014
0	1000	2000	3000
dost
Figure C-8. Plot of mean response by dose, with fitted curve for selected model; dose
shown in mg/m3.
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
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 risk
BMD = 6308.98
BMDL at the 95% confidence level = 4820.9
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
alpha
0.0303258
0.0303618
rho
n/a
0
beta_0
3.67004
3.67051
beta_l
0.0000581717
0.000058076
Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std Dev
Scaled Resid
0
10
3.68
3.67
0.253
0.174
0.181
406
9
3.71
3.69
0.12
0.174
0.282
825
10
3.64
3.72
0.126
0.174
-1.42
1643
10
3.76
3.77
0.19
0.174
-0.102
This document is a draft for review purposes only and does not constitute Agency policy.
C-27	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
3274
10
3.96
3.86
0.158
0.174
1.81
6369
10
4
4.04
0.158
0.174
-0.736
1
Likelihoods of
nterest
Model
Log(likelihood)
# Param's
AIC
A1
76.753535
7
-139.507071
A2
80.677068
12
-137.354137
A3
76.753535
7
-139.507071
fitted
73.624808
3
-141.249616
R
60.931962
2
-117.863925
3
4	Tests of Interest
Test
2*log(Likelihoo
d Ratio)
Test df
p-value
Test 1
39.4902
10
<0.0001
Test 2
7.84707
5
0.1649
Test 3
7.84707
5
0.1649
Test 4
6.25745
4
0.1807
5
6
7
This document is a draft for review purposes only and does not constitute Agency policy.
C-28	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-ll. 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)
BMCL10Rd
(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
errorc
0

Exponential (M5)
0.374
-265.71
errorc
0

Hill
0.227
-265.57
errorc
errorc

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

a Modeled 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.
b For the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
c BMC or BMCL computation failed for this model.
d For the Polynomial 3° model, the b3 coefficient estimates was 0 (boundary of parameters space). The models in
this row reduced to the Polynomial 2° model. For the Polynomial 3° model, the b3 and b2 coefficient estimates
were 0 (boundary of parameters space). The models in this row reduced to the Linear model.
6 For 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
5
This document is a draft for review purposes only and does not constitute Agency policy.
C-29	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-12. Summary of BMD modeling results for relative kidney weight in
2	female F344 rats exposed to tert-butanol via inhalation for 6 hrs/d, 5d/wk for
3	13 weeks (NTP. 1997): BMR = 10% relative deviation from the mean.
Model3
Goodness of fit
BMCiord
(mg/m3)
BMCL10rd
(mg/m3)
Basis for model selection

p-value
AIC

Exponential (M2)
Exponential (M3)b
0.813
-125.14
6859
5476
No model adequately fit the
data.
Exponential (M4)c
0.660
-123.12
6846
4832

Exponential (M5)d
0.660
-123.12
6846
4832

Hill
0.00189
-123.12
6845
5380

Power
0.809
-125.12
6846
5389

Polynomial 3°
0.00210
-123.34
6853
5439

Polynomial 2°
0.00191
-123.14
6865
5393

Linear
0.00488
-125.12
6846
5389

Polynomial 5°
0.00238
-123.61
6762
5504

Polynomial 4°
0.00228
-123.51
6807
5480

a Modeled variance case presented (BMDS Test 2 p-value = <0.0001, BMDS Test 3 p-value = 0.00105), no model
was selected as a best-fitting model.
b For the Exponential (M3) model, the estimate of d was 1 (boundary). The models in this row reduced to the
Exponential (M2) model.
c The Exponential (M4) model may appear equivalent to the Exponential (M5) model, however differences exist in
digits not displayed in the table.
d The Exponential (M5) model may appear equivalent to the Exponential (M4) model, however differences exist
in digits not displayed in the table.
4
5	C.1.2. Cancer Endpoints
6	For each endpoint, multistage models were fitted to the data using the maximum likelihood
7	method. Each model was tested for goodness-of-fit using a chi-square goodness-of-fit test (x2
8	p-value < 0.051 indicates lack of fit). Other factors were used to assess model fit, such as scaled
9	residuals, visual fit, and adequacy of fit in the low dose region and near the BMR.
10	For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
11	estimated by the profile likelihood method) and AIC value were used to select a best-fit model from
12	among the models exhibiting adequate fit If the BMDL estimates were "sufficiently close," that is,
13	differed by more than 3-fold, the model selected was the one that yielded the lowest AIC value. If
14	the BMDL estimates were not sufficiently close, the lowest BMDL was selected as the POD. For the
V 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-30	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	NTP (19951 and Hard etal. (20111 data, models were run with all doses included, as well as with
2	the high dose dropped. Dropping the high dose resulted in a better fit to the data. Including the high
3	dose caused the model to overestimate the control.
4	Table C-13. Cancer endpoints selected for dose-response modeling for tert-
5	butanol
Endpoint/Study
Species / Sex
Doses and Effect Data
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
Dose (PBPK, mg/L)
0
4.6945
12.6177
40.7135
Incidence / Total
8/50
13/50
19/50
13/50
Renal tubule adenoma or
carcinoma
NTP (1995)
Rat (F344) /
Male
Dose (PBPK, mg/hr)
0
0.7992
1.7462
3.4712
Incidence / Total
8/50
13/50
19/50
13/50
Renal tubule adenoma or
carcinoma; Hard reanalysis
NTP (1995);Hard etal. (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 etal. (2011)
Rat (F344) /
Male
Dose (PBPK, mg/L)
0
4.6945
12.6177
40.7135
Incidence / Total
4/50
13/50
18/50
12/50
Renal tubule adenoma or
carcinoma; Hard reanalysis
NTP (1995);Hard etal. (2011)
Rat (F344) /
Male
Dose (PBPK, mg/hr)
0
0.7992
1.7462
3.4712
Incidence / Total
4/50
13/50
18/50
12/50
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
6
7	C. 1.2.1. Modeling Results
8	Below are tables summarizing the modeling results for the cancer endpoints that were
9	modeled. For the multistage models, the coefficients were restricted to be non-negative (beta
10	values > 0).
11
This document is a draft for review purposes only and does not constitute Agency policy.
C-31	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-14. Summary of BMD modeling results for renal tubule adenoma or
2	carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
3	years modeled with administered dose units and including all dose groups
4	fNTP. 19951: BMR = 10% extra risk.
5
Model3
Goodness of fit
BMDiopct (mg/kg-
d)
BMD Liopct
(mg/kg-d)
Basis for model
selection
P-
value
Scaled residuals
AIC
Three
Two
0.080
6
-0.989, 0.288,
1.719, and -1.010
233.9
4
294
118
Multistage 2° is
selected as the
most
parsimonious
model of
adequate fit.
One
0.080
6
-0.989, 0.288,
1.719, and -1.010
233.9
4
294
errorb
Selected model in bold.
1 BMD or BMDL computation failed for this model.
Multistage Cancer Model with 0.95 Confidence Level
10:57 04/30 2014
0	50	100	150 200 250 300 350 400
dose
9	Figure C-9. Plot of incidence rate by dose, with fitted curve for selected model; dose
10	shown in mg/kg-d.
11
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
Supplemental Information—tert Butanol
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-33	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-15. Summary of BMD modeling results for renal tubule adenoma or
2	carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
3	years modeled with administered dose units and excluding high-dose group
4	fNTP. 19951: BMR = 10% extra risk.
5
Model3
Goodness of fit
BMDiopct (mg/kg-d)
BMD Liopct
(mg/kg-d)
Basis for model
selection
P-
value
Scaled residuals
AIC
Two
N/Ab
0.000, -0.000, and -
0.000
173.6
8
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
171.6
9
70.1
41.6
a Selected model in bold.
b No available degrees of freedom to calculate a goodness of fit value.
6
Multistage Cancer Model with 0.95 Confidence Level
7	11:02 04/30 2014
8
9	Figure C-10. Plot of incidence rate by dose, with fitted curve for selected model; dose
10	shown in mg/kg-d.
11
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
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert Butanol
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 Tab
e
Model
Log(likelihoo
d)
# 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-35	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
0
1
Supplemental Information—tert Butanol
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 PBPK (tert-butanol, mg/L) dose units and including all
dose groups fNTP. 19951: BMR = 10% extra risk.
Model3
Goodness of fit
BMDiopct (mg/L)
BMDLiopct (mg/L)
Basis for model
selection
P-
value
Scaled residuals
AIC
Three
Two
One
0.051
8
-1.373,0.155,
1.889, and -0.668
234.8
3
51.8
13.9
Multistage 1° was
selected as the
most
parsimonious
model of
adequate fit.
a Selected model in bold.
Multistage Cancer Model with 0.95 Confidence Level
dose
11:19 04/30 2014
Figure C-ll. Plot of incidence rate by dose, with fitted curve for selected model; dose
shown in mg/L.
This document is a draft for review purposes only and does not constitute Agency policy.
C-36	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert Butanol
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 = 51.8357
BMDL at the 95% confidence level = 13.9404
BMDU at the 95% confidence level = error
Taken together, (13.9404, error) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = error
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
Background
0.243327
0.253053
Beta(l)
0.00203259
0.00150893
Analysis of Deviance Table
Model
Log(likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-112.492
4



Fitted model
-115.417
2
5.84883
2
0.0537
Reduced
model
-115.644
1
6.30404
3
0.09772
AIC: = 234.834
Goodness of Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.2433
12.166
8
50
-1.373
4.6945
0.2505
12.526
13
50
0.155
12.6177
0.2625
13.124
19
50
1.889
40.7135
0.3034
15.171
13
50
-0.668
ChiA2 = 5.92 d.f=2 P-value = 0.0518
This document is a draft for review purposes only and does not constitute Agency policy.
C-37	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-17. Summary of BMD modeling results for renal tubule adenoma or
2	carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
3	years modeled with PBPK (tert-butanol, mg/L) dose units and excluding high-
4	dose group fNTP. 19951: BMR = 10% extra risk.
Model3
Goodness of fit
BMDiopct (mg/L)
BMDLiopct (mg/L)
Basis for model
selection
P-
value
Scaled residuals
AIC
Two
One
0.891
-0.054,0.113, and -
0.057
171.7
0
4.33
2.54
Multistage 1° was
selected as the
most
parsimonious
model of
adequate fit.
a Selected model in bold.
5
Multistage Cancer Model with 0.95 Confidence Level
6	11:20 04/30 2014
7
8	Figure C-12. Plot of incidence rate by dose, with fitted curve for selected model; dose
9	shown in mg/L.
10
This document is a draft for review purposes only and does not constitute Agency policy.
C-38	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information—tert Butanol
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 = 4.33496
BMDL atthe 95% confidence level = 2.53714
BMDU atthe 95% confidence level = 12.8097
Taken together, (2.53714,12.8097) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.0394144
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
Background
0.162798
0.164724
Beta(l)
0.0243048
0.0238858
Analysis of Deviance Table
Model
Log(likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-83.8395
3



Fitted model
-83.8489
2
0.0187339
1
0.8911
Reduced
model
-86.9873
1
6.29546
2
0.04295
AIC: = 171.698
Goodness of Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.1628
8.14
8
50
-0.054
4.6945
0.2531
12.654
13
50
0.113
12.6177
0.3839
19.195
19
50
-0.057
ChiA2 = 0.02 d.f = 1 P-value = 0.891
This document is a draft for review purposes only and does not constitute Agency policy.
C-39	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-18. Summary of BMD modeling results for renal tubule adenoma or
2	carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
3	years modeled with PBPK (metabolized, mg/hr) dose units and including all
4	dose groups fNTP. 19951: BMR = 10% extra risk.
5
Model3
Goodness of fit
BMDiopct (mg/hr)
BMD Liopct
(mg/hr)
Basis for model
selection
P-
value
Scaled residuals
AIC
Three
Two
One
0.088
5
-0.920,0.301,
1.677, and -1.049
233.7
6
2.28
0.954
Multistage 1° was
selected as the
most
parsimonious
model of
adequate fit.
a Selected model in bold.
Data from NTP1995
6
Multistage Cancer Model with 0.95 Confidence Level
7	11:22 04/30 2014
8
9	Figure C-13. Plot of incidence rate by dose, with fitted curve for selected model; dose
10	shown in mg/hr.
11
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
Supplemental Information—tert Butanol
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 = 2.28299
BMDL at the 95% confidence level = 0.95436
BMDU at the 95% confidence level = error
Taken together, (0.95436, error) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = error
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
Background
0.21328
0.229822
Beta(l)
0.0461502
0.0349139
Analysis of Deviance Table
Model
Log(likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-112.492
4



Fitted model
-114.879
2
4.77309
2
0.09195
Reduced
model
-115.644
1
6.30404
3
0.09772
AIC: = 233.758
Goodness of Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.2133
10.664
8
50
-0.92
0.7992
0.2418
12.088
13
50
0.301
1.7462
0.2742
13.71
19
50
1.677
3.4712
0.3297
16.487
13
50
-1.049
ChiA2 = 4.85 d.f=2 P-value = 0.0885
This document is a draft for review purposes only and does not constitute Agency policy.
C-41	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-19. Summary of BMD modeling results for renal tubule adenoma or
2	carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
3	years modeled with PBPK (metabolized, mg/hr) dose units and excluding
4	high-dose group fNTP. 19951: BMR = 10% extra risk.
5
Model3
Goodness of fit
BMDiopct (mg/hr)
BMD Liopct
(mg/hr)
Basis for model
selection
P-
value
Scaled residuals
AIC
Two
N/Ab
-0.000, -0.000, and
-0.000
173.6
8
0.673
0.365
Multistage 1° was
selected on the
basis of lowest
AIC.
One
0.906
0.037, -0.096, and
0.057
171.6
9
0.614
0.364
a Selected model in bold.
b No available degrees of freedom to calculate a goodness of fit value.
Data from NTP1995
6
Multistage Cancer Model with 0.95 Confidence Level
7	11:24 04/30 2014
8
9	Figure C-14. Plot of incidence rate by dose, with fitted curve for selected model; dose
10	shown in mg/hr.
11
This document is a draft for review purposes only and does not constitute Agency policy.
C-42	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Supplemental Information—tert Butanol
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 = 0.613798
BMDL at the 95% confidence level = 0.364494
BMDU at the 95% confidence level = 1.77845
Taken together, (0.364494,1.77845) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.274353
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
Background
0.158068
0.156284
Beta(l)
0.171653
0.174305
Analysis of Deviance Table
Model
Log(likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-83.8395
3



Fitted model
-83.8465
2
0.0138544
1
0.9063
Reduced
model
-86.9873
1
6.29546
2
0.04295
AIC: = 171.693
Goodness of Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.1581
7.903
8
50
0.037
0.7992
0.266
13.3
13
50
-0.096
1.7462
0.3761
18.806
19
50
0.057
ChiA2 = 0.01 d.f = 1 P-value = 0.9064
This document is a draft for review purposes only and does not constitute Agency policy.
C-43	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1
2
3
4
5
Model3
Goodness of fit
BMDiopct (mg/kg-d)
BMDLiopct
(mg/kg-d)
Basis for model
selection
P"
value
Scaled residuals
AIC
Three
Two
One
0.011
7
-1.476, 1.100,
1.855, and -1.435
218.6
8
184
94.8
No model fit the
data.
a No model was selected as a best-fitting model.
Table C-21. 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;
re-analvzed data fHard et al.. 2011: NTP. 19951: BMR = 10% extra risk.
Model3
Goodness of fit
BMD10Pct (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
154.8
4
54.2
36.3
Multistage 1° was
selected as the
most
parsimonious
model of
adequate fit.
a Selected model in bold.
This document is a draft for review purposes only and does not constitute Agency policy.
C-44	DRAFT—DO NOT CITE OR QUOTE
Table C-20. 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.

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Supplemental Information—tert Butanol
Multistage Cancer Model with 0.95 Confidence Level
O	50	100	150	200
dose
11 :05 04/30 2014
Figure C-15. Plot of incidence rate by dose, with fitted curve for selected model; 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



Fitted model
-75.4201
2
0.315716
1
0.5742
Reduced
model
-81.4909
1
12.4574
2
0.001972
This document is a draft for review purposes only and does not constitute Agency policy.
C-45	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1
2	AIC: = 154.84
3
4	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
5
6	ChiA2 = 0.32 d.f=l P-value = 0.5715
7
8
This document is a draft for review purposes only and does not constitute Agency policy.
C-46	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-22. Summary of BMD modeling results for renal tubule adenoma or
2	carcinoma in male F344 rats exposed to tert-butanol in drinking water for 2
3	years modeled with PBPK (tert-butanol, mg/L) dose units and including all
4	dose groups; reanalyzed data fHard et al.. 2011: NTP. 19951: BMR = 10% extra
5	risk.
Model3
Goodness of fit
BMDiopct (mg/L)
BMDLiopct (mg/L)
Basis for model
selection
P"
value
Scaled residuals
AIC
Three
Two
One
0.004
8
-2.089, 0.864,
2.165, and -0.929
220.8
2
31.4
11.7
No model fit the
data.
a No model was selected as a best-fitting model.
Table C-23. 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 PBPK (tert-butanol, mg/L) dose units and excluding high-
dose eroup: reanalyzed data fHard etal.. 2011: NTP. 19951: BMR = 10% extra
risk.
Model3
Goodness of fit
BMD10Pct (mg/L)
BMDL10Pct (mg/L)
Basis for model
selection
P"
value
Scaled residuals
AIC
Two
One
0.364
-0.285,0.750, and -
0.424
155.3
3
3.35
2.21
Multistage 1° was
selected as the
most
parsimonious
model of
adequate fit.
a Selected model in bold.
This document is a draft for review purposes only and does not constitute Agency policy.
C-47	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Supplemental Information—tert Butanol
Multistage Cancer Model with 0.95 Confidence Level
Figure C-16. Plot of incidence rate by dose, with fitted curve for selected model; dose
shown in mg/L.
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 = 3.34903
BMDL at the 95% confidence level = 2.20865
BMDU at the 95% confidence level = 6.49702
Taken together, (2.20865, 6.49702) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.0452765
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
Background
0.0916116
0.110649
Beta(l)
0.03146
0.0276674
Analysis of Deviance Table
Model
Log(likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-75.2622
3



Fitted model
-75.664
2
0.803466
1
0.3701
Reduced
model
-81.4909
1
12.4574
2
0.001972
This document is a draft for review purposes only and does not constitute Agency policy.
C-48	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Supplemental Information—tert Butanol
AIC: = 155.328
Goodness of Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.0916
4.581
4
50
-0.285
4.6945
0.2163
10.817
13
50
0.75
12.6177
0.3892
19.462
18
50
-0.424
ChiA2 = 0.82 d.f = 1 P-value = 0.3643
Table C-24. 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 PBPK (metabolized, mg/hr) dose units and including all
dose groups; reanalyzed data fHard et al.. 2011: NTP. 19951: BMR = 10% extra
risk.
Model3
Goodness of fit
BMD10Pct (mg/hr)
BMDL10pct
(mg/hr)
Basis for model
selection
P-
value
Scaled residuals
AIC
Three
Two
One
0.014
2
-1.367, 1.119,
1.783, and -1.484
218.2
6
1.44
0.770
No model fit the
data.
a No model was selected as a best-fitting model.
Table C-25. 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 PBPK (metabolized, mg/hr) dose units and excluding
high-dose group: reanalyzed data (Hard etal.. 2011: NTP. 1995): BMR = 10%
extra risk.
Model3
Goodness of fit
BMD10pct
BMDL10pct
(mg/hr)
Basis for model
selection
p-value
Scaled residuals
AIC
(mg/hr)
Two
One
0.593
-0.130,0.435, and
-0.281
154.81
0.474
0.319
Multistage 1° was
selected as the most
parsimonious model
of adequate fit.
a Selected model in bold.
This document is a draft for review purposes only and does not constitute Agency policy.
C-49	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 Butanol
Multistage Cancer Model with 0.95 Confidence Level
Figure C-17. Plot of inddence rate by dose, with fitted curve for selected model; dose
shown in mg/hr.
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 = 0.474241
BMDL at the 95% confidence level = 0.318504
BMDU at the 95% confidence level = 0.882859
Taken together, (0.318504, 0.882859) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.313968
Parameter Estimates
Variable
Estimate
Default Initial
Parameter Values
Background
0.0851364
0.0969736
Beta(l)
0.222167
0.206161
Analysis of Deviance Table
Model
Log(likelihood)
# Param's
Deviance
Test d.f.
p-value
Full model
-75.2622
3



Fitted model
-75.4029
2
0.281435
1
0.5958
Reduced
model
-81.4909
1
12.4574
2
0.001972
This document is a draft for review purposes only and does not constitute Agency policy.
C-50	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1
2	AIC: = 154.806
3
4	Goodness of Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled Resid
0
0.0851
4.257
4
50
-0.13
0.7992
0.234
11.699
13
50
0.435
1.7462
0.3793
18.966
18
50
-0.281
5
6	ChiA2 = 0.29 d.f = 1 P-value = 0.5933
7
8
This document is a draft for review purposes only and does not constitute Agency policy.
C-51	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
1	Table C-26. Summary of BMD modeling results for thyroid follicular cell
2	adenomas in female B6C3F1 mice exposed to tert-butanol in drinking water
3	for 2 years (NTP. 1995): BMR = 10% extra risk.
Model3
Goodness of fit
BMD10%
(mg/kg-d)
BMDL10%
(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
a Selected (best-fitting) model shown in boldface type
bAIC = Akaike Information Criterion
c Confidence level 0.95
4
Multistage Cancer Model with 0.95 Confidence Level
0	50 100 150 200 250 300 350
dose
5	15:51 05/13 2011
6	Figure C-18. Plot of mean response by dose, with fitted curve for selected model.
7	====================================================================
8	Multistage Cancer Model. (Version: 1.5; Date: 05/26/2010)
9	Input Data File: M:\NCEA te/"t-blltanol\BME) modeling\BMDS Output\29 NTP
10	1555b_Thyroid folluclar cell andenoma, female mice (HEC)_MultiCanc3_10.(d)
11	Gnuplot Plotting File: M:\NCEA te/"t-blltanol\BME) modeling\BMDS Output\25 NTP
12	1995b Thyroid folluclar cell andenoma, female mice (HEC) MultiCanc3 10.pit
13	Fri May 13 15:51:46 2 011
14	====================================================================
15
16	[notes]
17		
This document is a draft for review purposes only and does not constitute Agency policy.
C-52	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 Butanol
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose/xl-beta2*dose/x2-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 = 0.0344951
Beta (1) =	0
Beta (2) =	0
Beta (3) = 3.4555e-009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -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 )
Background Beta (3)
Background	1 -0.54
Beta (3) -0.54	1
Parameter Estimates
95.0% Wald Confidence Interval
Variable	Estimate Std. Err.	Lower Conf. Limit Upper Conf. Limit
Background 0.0359685	*	*	*
Beta (1)	0	*	*	*
Beta (2)	0	*	*	*
Beta (3) 3.30537e-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 -54.5437	4
Fitted model -54.8422	2 0.597063 2	0.7419
Reduced model -58.5048	1 7.92235 3	0.04764
AIC:	113.684
Goodness of Fit
Scaled
This document is a draft for review purposes only and does not constitute Agency policy.
C-53	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 Butanol
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.0360
2.086
2 .000
58
-0.061
83.0000
0.0378
2 . 267
3 .000
60
0.496
164.0000
0.0499
2 . 945
2 .000
59
-0.565
334.0000
0.1477
8 . 713
9.000
59
0 . 105
Chi^2 = 0.58 d.f. = 2 P-value = 0.7482
Benchmark Dose Computation
Specified effect =	0.1
Risk Type =	Extra risk
Confidence level	=	0.95
BMD = 317.068
BMDL = 228.888
BMDU = 60 0.031
Taken together, (228.888, 600.031) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.000436894
This document is a draft for review purposes only and does not constitute Agency policy.
C-54	DRAFT—DO NOT CITE OR QUOTE

-------
Supplemental Information—tert Butanol
APPENDIX D. SUMMARY OF EXTERNAL PEER
REVIEW AND PUBLIC COMMENTS AND EPA'S
DISPOSITION	
[placeholder]
Additional Appendices:
[appendices can be us-'-i 		uppl-'inent illio III 'ii-! lb II1 analysis - the information
presented by the appendices will be chemical-specific]
Examples:
•	PBPK Modeling of [chemical] and metabolites - detailed methods and results
•	Meta-analysis of	results from epidemiological studies
•	Lifetable analysis and weighted linear regression based on results from
[reference]
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
39
40
Supplemental Information—tert Butanol
REFERENCES FOR APPENDICES
ACGIH (American Conference of Governmental Industrial Hygienists). (2012). Documentation of
the threshold limit values and biological exposure indices
tert-Butanol (7th ed.). Cincinnati, OH.
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.doi.Org/10.1093/toxsci/53.2.194
Andersen. ME. (1991). Physiological modelling of organic compounds. Ann Occup Hyg 35: 309-
321.
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. MJ; Pascoe. E; Reinhold. JG. (1971). Rat liver alcohol dehydrogenase. Purification and
properties. BiochemJ 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.atsd r.cdc.gov/toxprofiles/tp.asp?id=228&tid=41
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.llll/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.doi.org/10.1021/tx970215v
Blancato. JN; Evans. MV; Power. FW; Caldwell. JC. (2007). Development and use of PBPK
modeling and the impact of metabolism on variability in dose metrics for the risk
assessment of methyl tertiary butyl ether (MTBE). J Environ Prot Sci 1: 29-51.
Borghoff. S; Murphy. J: Medinsky. M. (1996). Development of physiologically based
pharmacokinetic model for methyl tertiary-butyl ether and tertiary-butanol in male
Fisher-344 rats. Fundam Appl Toxicol 30: 264-275.
http://dx.doi.org/10.1006/faat.1996.0Q64
Borghoff. S; Parkinson. H; Leavens. T. (2010). Physiologically based pharmacokinetic rat model
for methyl tertiary-butyl ether; comparison of selected dose metrics following various
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 Butanol
MTBE exposure scenarios used for toxicity and carcinogenicity evaluation. Toxicology
275: 79-91. http://dx.doi.Org/10.1016/i.tox.2010.06.003
Borghoff. SJ; Prescott. JS; Janszen. DB; Wong. BA; Everitt. JI. (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.
Borghoff. SJ: Short. BG; Swenberg. JA. (1990). Biochemical mechanisms and pathobiology of
alpha 2u-globulin nephropathy [Review]. Annu Rev Pharmacol Toxicol 30: 349-367.
http://dx.doi.org/10.1146/annurev.pa.30.040190.002Q25
Brown. MA: Cornell. BA: Davenport. JB. (1977). PERTURBATION OF BIOLOGICAL-MEMBRANES
WITH TERT-BUTANOL. Seikagaku 49: 1-1.
Cederbaum. Al; Cohen. G. (1980). Oxidative demethylation of t-butyl alcohol by rat liver
microsomes. Biochem Biophys Res Commun 97: 730-736.
Cederbaum. Al: 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/10.1016/0006-2952(83)90297-6
Dickey. FH; Cleland. GH; Lotz. C. (1949). The role of organic peroxides in the induction of
mutations. PNAS 35: 581-586.
Faulkner. T; Hussain. A. (1989). The pharmacokinetics of tertiary butanol in C57BL/6J mice. Res
Comm Chem Pathol Pharmacol 64: 31-39.
Faulkner. TP: Wiechart. JD; 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
Hard. GC; Bruner. RH; Cohen. SM; Pletcher. JM; 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.yrtph.2011.01.007
Jimenez. J: Longo. E; Benitez.T. (1988). Induction of petite yeast mutants by membrane-active
agents. Appl Environ Microbiol 54: 3126-3132.
Johanson. G; Nihlen. A: Lof. A. (1995). Toxicokinetics and acute effects of MTBE and ETBE in
male volunteers. Toxicol Lett 82/83: 713-718. http://dx.doi.org/10.1016/Q378-
4274(95)03589-3
Kim. D; Andersen. ME: Pleil. JD: Nylander-French. LA: Prah. JD. (2007). Refined PBPK model of
aggregate exposure to methyl tertiary-butyl ether. Toxicol Lett 169: 222-235.
http://dx.doi.Org/10.1016/i.toxlet.2007.01.008
Leavens. T; Borghoff. S. (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
42
43
Supplemental Information—tert Butanol
Licata. AC; Dekant. W; Smith. CE; Borghoff. SJ. (2001). A physiologically based pharmacokinetic
model for methyl tert-butyl ether in humans: Implementing sensitivity and variability
analyses. Toxicol Sci 62:191-204. http://dx.doi.Org/10.1093/toxsci/62.2.191
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; Caspary. WJ. (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: Johanson. G. (1999). Physiologically based toxicokinetic modeling of inhaled ethyl
tertiary-butyl ether in humans. Toxicol Sci 51: 184-194.
http://dx.doi.Org/10.1093/toxsci/51.2.184
Nihlen. A: Lof. A: Johanson. 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: Johanson. 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: Johanson. G. (1998b). Experimental exposure to methyl tertiary-butyl ether: I
Toxicokinetics in humans. Toxicol Appl Pharmacol 148: 274-280.
NIOSH (National Institute for Occupational Safety and Health). (2007). NIOSH pocket guide to
chemical hazards. (2005-149). Cincinnati, OH. http://www.cdc.gov/niosh/docs/2005-
149/
NSF International. (2003). t-Butanol: Oral Risk Assessment Document (CAS 75-65-0). Ann Arbor,
Ml.
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). 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. JL; Borghoff. SJ. (1997). Pharmacokinetics of tertiary butyl alcohol in male
and female Fischer 344 rats. Toxicol Lett 92: 179-186.
Prah. J: Ashley. D; Blount. B; Case. M; Leavens. T; Pleil. J: Cardinali. F. (2004). Dermal, oral, and
inhalation pharmacokinetics of methyl tertiary butyl ether (MTBE) in human volunteers.
Toxicol Sci 77: 195-205. http://dx.doi.org/10.1093/toxsci/kfh009
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
Supplemental Information—tert Butanol
Rao. HV; Ginsberg. GL. (1997). A physiologically-based pharmacokinetic model assessment of
methyl t-butyl ether in groundwater for a bathing and showering determination. Risk
Anal 17: 583-598.
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
Spiteri. NJ. (1982). Circadian patterning of feeding, drinking and activity during diurnal food
access in rats. Physiol Behav 28: 139-147. http://dx.doi.org/10.1016/0Q31-
9384(82)90115-9
Tang. G; Wang. J: 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.
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-draft.htm
Videla. LA: Fernandez. V; de Marinis. A: Fernandez. IN; 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/10.1016/0006-291X(82)91343-2
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/sl383-5718(99)00137-0
Williams. TM; Borghoff. SJ. (2001). Characterization of tert-butyl alcohol binding to "alpha"2u-
globulin in F-344 rats. Toxicol Sci 62: 228-235.
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
Zeiger. E; Anderson. B; Haworth. S; Lawlor. T; Mortelmans. K; Speck. W. (1987). Salmonella
mutagenicity tests: III. Results from the testing of 255 chemicals. Environ Mutagen 9:1-
60. http://dx.doi.org/10.1002/em.28600906Q2
This document is a draft for review purposes only and does not constitute Agency policy.
R-4	DRAFT—DO NOT CITE OR QUOTE

-------