EPA/635/R-16/079a
                                                                  Public Comment Draft
                                                                      www.epa.gov/iris
          Toxicological Review of tert-Butyl Alcohol (tert-Butanol)

                                 (CAS No. 75-65-0)
                                      April 2 016

                                       NOTICE
This document is a Public Comment 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.
                      National Center for Environmental Assessment
                           Office of Research and Development
                          U.S. Environmental Protection Agency
                                    Washington, DC

-------
                                                  Toxicological Review of tert-Butyl Alcohol

                                      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                                           ii              DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review of tert-ButyI Alcohol
CONTENTS
AUTHORS | CONTRIBUTORS | REVIEWERS	viii
PREFACE	x
PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS	xiv
EXECUTIVE SUMMARY	ES-1
LITERATURE SEARCH STRATEGY | STUDY SELECTION AND EVALUATION	LS-1
1.   HAZARD IDENTIFICATION	1-1
    1.1. Overview of Chemical Properties and Toxicokinetics	1-1
       1.1.1.  Chemical Properties	1-1
       1.1.2.  Toxicokinetics	1-2
       1.1.3.  Description of Toxicokinetic Models	1-3
       1.1.4.  Chemicals Extensively Metabolized to tert-Butanol	1-4
    1.2. PRESENTATION AND SYNTHESIS OF EVIDENCE BY ORGAN/SYSTEM	1-4
       1.2.1.  Kidney Effects	1-4
       1.2.2.  Thyroid Effects	1-38
       1.2.3.  Developmental Effects	1-46
       1.2.4.  Neurodevelopmental Effects	1-53
       1.2.5.  Reproductive Effects	1-56
       1.2.6.  Other Toxicological Effects	1-61
    1.3. INTEGRATION AND EVALUATION	1-61
       1.3.1.  Effects Other Than Cancer	1-61
       1.3.2.  Carcinogenicity	1-63
       1.3.3.  Susceptible Populations and Lifestages for Cancer and Noncancer Outcomes	1-65
2.   DOSE-RESPONSE ANALYSIS	2-1
    2.1. ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER	2-1
       2.1.1.  Identification of Studies and Effects for Dose-Response Analysis	2-1
       2.1.2.  Methods of Analysis	2-2
       2.1.3.  Derivation of Candidate Values	2-4
       2.1.4.  Derivation of Organ/System-Specific Reference Doses	2-8
       2.1.5.  Selection of the Overall Reference Dose	2-8

          This document is a draft for review purposes only and does not constitute Agency policy.
                                            iii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review of tert-Butyl Alcohol


       2.1.6. Confidence Statement	2-9

       2.1.7. Previous IRIS Assessment	2-9

    2.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER THAN CANCER	2-9

       2.2.1. Identification of Studies and Effects for Dose-Response Analysis	2-9

       2.2.2. Methods of Analysis	2-10

       2.2.3. Derivation of Candidate Values	2-13

       2.2.4. Derivation of Organ/System-Specific Reference Concentrations	2-16

       2.2.5. Selection of the Overall Reference Concentration	2-16

       2.2.6. Confidence Statement	2-17

       2.2.7. Previous IRIS Assessment	2-17

       2.2.8. Uncertainties in the Derivation of the Reference  Dose and Reference
             Concentration	2-17

    2.3. ORAL SLOPE FACTOR FOR CANCER	2-18

       2.3.1. Analysis of Carcinogenicity Data	2-18

       2.3.2. Dose-Response Analysis—Adjustments and Extrapolations Methods	2-19

       2.3.3. Derivation of the Oral Slope Factor	2-20

       2.3.4. Uncertainties in the Derivation of the Oral Slope  Factor	2-21

       2.3.5. Previous IRIS Assessment: Oral Slope Factor	2-23

    2.4. INHALATION UNIT RISK FOR CANCER	2-23

       2.4.1. Previous IRIS Assessment: Inhalation Unit Risk	2-24

    2.5. APPLICATION OF AGE-DEPENDENT ADJUSTMENT FACTORS	2-24

REFERENCES	R-l
          This document is a draft for review purposes only and does not constitute Agency policy.
                                             iv              DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review of tert-Butyl Alcohol
TABLES

Table ES-1. Organ/system-specific RfDs and overall RfD for tert-butanol	ES-2
Table ES-2. Organ/system-specific RfCs and overall RfCfor tert-butanol	ES-3
Table LS-1. Details of the search strategy employed for tert-butanol	LS-4
Table LS-2. Summary of additional search strategies for tert-butanol	LS-4
Table LS-3. Inclusion-exclusion criteria	LS-5
Table LS-4. Considerations for evaluation of experimental animal studies	LS-8
Table LS-5. Summary of experimental animal database	LS-8
Table 1-1. Physicochemical properties and  chemical identity of tert-butanol	1-1
Table 1-2. Changes in kidney histopathology in animals following exposure to tert-butanol	1-12
Table 1-3. Changes in kidney tumors in animals following exposure to tert-butanol	1-15
Table 1-4. Summary of data on the a2u-globulin process in male  rats exposed to tert-butanol	1-22
Table 1-5. Proposed empirical criteria for attributing renal tumors to CPN	1-33
Table 1-6. Evidence pertaining to thyroid effects in animals following oral exposure to tert-
               butanol	1-39
Table 1-7. Evidence pertaining to developmental effects in animals following exposure to tert-
               butanol	1-48
Table 1-8. Evidence pertaining to neurodevelopmental effects in animals following exposure to
               tert-butanol	1-54
Table 1-9. Evidence pertaining to reproductive effects in animals following exposure to tert-
               butanol	1-56
Table 2-1. Summary of derivations of points of departure following oral exposure for up to 2
               years	2-4
Table 2-2. Effects and corresponding derivation of candidate values	2-6
Table 2-3. Organ/system-specific RfDs and  overall RfD for tert-butanol	2-8
Table 2-4. Summary of derivation of PODs following inhalation exposure	2-12
Table 2-5. Summary of derivation of inhalation points of departure derived from route-to-route
               extrapolation from oral exposures	2-13
Table 2-6. Effects and corresponding derivation of candidate values	2-14
Table 2-7. Organ/system-specific RfCs and  overall RfCfor tert-butanol	2-16
Table 2-8. Summary of the oral slope factor derivation	2-21
Table 2-9. Summary of uncertainties in the derivation of the oral slope  factor for tert-butanol	2-22
FIGURES

Figure LS-1. Summary of literature search and screening process for tert-butanol	LS-3
Figure 1-1. Biotransformation of tert-butanol in rats and humans	1-3
Figure 1-2. Comparison of absolute kidney weight change in male and female rats across oral
               and inhalation exposure based on internal blood concentration. Spearman rank
               correlation coefficient (rho) was calculated to evaluate the direction of a
               monotonic association (e.g., positive value = positive association) and the
               strength of association	1-10
Figure 1-3. Comparison of absolute kidney weight change in male and female mice following oral
               exposure based on administered concentration. Spearman rank correlation

           This document is a draft for review purposes only and does not constitute Agency policy.
                                              v              DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review of tert-Butyl Alcohol


              coefficient (rho) was calculated to evaluate the direction of a monotonic
              association (e.g., positive value = positive association) and the strength of
              association	1-11
Figure 1-4. Comparison of absolute kidney weight change in male and female mice following
              inhalation exposure based on administered concentration. Spearman rank
              correlation coefficient (rho) was calculated to evaluate the direction of a
              monotonic association (e.g., positive value = positive association) and the
              strength of association	1-11
Figure 1-5. Exposure response array for kidney effects following oral exposure to te/t-butanol	1-17
Figure 1-6. Exposure-response array of kidney effects following inhalation exposure to tert-
              butanol (13-week studies, no chronic studies available)	1-18
Figure 1-7. Temporal pathogenesis of a2u-globulin-associated nephropathy in male rats. a2u-
              Globulin synthesized in the livers of male rats is delivered to the kidney, where
              it can accumulate in hyaline droplets and be retained by epithelial cells lining
              the S2 (P2) segment of the proximal tubules. Renal pathogenesis following
              continued te/t-butanol  (TBA) exposure and increasing droplet accumulation can
              progress step-wise from increasing epithelial cell damage, death and
              dysfunction leading to the formation of granular casts in the corticomedullary
              junction, linear mineralization of the renal papillae, and carcinogenesis of the
              renal tubular epithelium (Adapted from Swenberg  and Lehman-McKeeman
              (1999) and U.S. EPA (1991a)	1-21
Figure 1-8. Exposure-response array for effects potentially associated with a2u-globulin renal
              tubule nephropathy and tumors in male rats after oral exposure to te/t-butanol	1-24
Figure 1-9. Exposure-response array for effects potentially associated with a2u-globulin renal
              tubule nephropathy and tumors in male rats after inhalation exposure to
              te/t-butanol	1-25
Figure 1-10. Exposure-response array of thyroid follicular cell effects following chronic oral
              exposure to te/t-butanol. (Note: Only one carcinoma was observed in male mice
              at the high-dose group.)	1-41
Figure 1-11. Exposure-response array of developmental effects following oral exposure to te/t-
              butanol	1-51
Figure 1-12. Exposure-response array of developmental effects following inhalation exposure to
              te/t-butanol	1-52
Figure 1-13. Exposure-response array of reproductive effects following oral exposure to te/t-
              butanol	1-59
Figure 1-14. Exposure-response array of reproductive effects following inhalation exposure to
              te/t-butanol	1-60
Figure 2-1. Candidate values with corresponding POD and composite UF. Each bar corresponds
              to one data set described in Table 2-1 and Table 2-2	2-7
Figure 2-2. Candidate RfC values with corresponding POD and composite UF	2-15
           This document is a draft for review purposes only and does not constitute Agency policy.
                                              vi              DRAFT—DO NOT CITE OR QUOTE

-------
ABBREVIATIONS
                                                     Toxicological Review of tert-ButyI Alcohol
AIC      Akaike's information criterion                MNPCE
ALD     approximate lethal dosage
ALT     alanine aminotransferase                    MTD
AST     aspartate aminotransferase                  NAG
atm     atmosphere                                NCEA
ATSDR   Agency for Toxic Substances and
         Disease Registry                            NCI
BMD     benchmark dose                            NOAEL
BMDL    benchmark dose lower confidence limit        NTP
BMDS    Benchmark Dose Software                   NZW
BMR     benchmark response                        OCT
BW      body weight                                ORD
CA       chromosomal aberration                     PBPK
CASRN   Chemical Abstracts Service Registry           POD
         Number                                   POD[ADj]
CBI      covalent binding index                       QSAR
CHO     Chinese hamster ovary (cell line)
CL       confidence limit                            RDS
CNS     central nervous system                      RfC
CPN     chronic progressive nephropathy              RfD
CYP450  cytochrome P450                           RGDR
DAF     dosimetric adjustment factor                 RNA
DEN     diethylnitrosamine                          SAR
DMSO    dimethylsulfoxide                           SCE
DNA     deoxyribonucleic acid                       SD
EPA     Environmental Protection Agency             SDH
FDA     Food and Drug Administration                SE
FEVi     forced expiratory volume of 1 second          SCOT
GD       gestation day
GDH     glutamate dehydrogenase                    SGPT
GGT     yglutarnyl transferase
GSH     glutathione                                SSD
GST     glutathione-S-transferase                    TCA
Hb/g-A  animal blood:gas partition coefficient          TCE
Hb/g-H  human blood:gas partition coefficient          TWA
HEC     human equivalent concentration              UF
HED     human equivalent dose                      UFA
i.p.       intraperitoneal                             UFn
IRIS     Integrated Risk Information System           UFi
IVF      in vitro fertilization                         UFs
LCso     median lethal concentration
LDso     median lethal dose                          UFD
LOAEL   lowest-observed-adverse-effect level          U.S.
MN      micronuclei
micronucleated polychromatic
erythrocyte
maximum tolerated dose
N-acetyl-p-D-glucosaminidase
National Center for Environmental
Assessment
National Cancer Institute
no-observed-adverse-effect level
National Toxicology Program
New Zealand White (rabbit breed)
ornithine carbamoyl transferase
Office of Research and Development
physiologically based pharmacokinetic
point of departure
duration-adjusted POD
quantitative structure-activity
relationship
replicative DNA synthesis
inhalation reference concentration
oral reference dose
regional gas dose ratio
ribonucleic acid
structure activity relationship
sister chromatid exchange
standard deviation
sorbitol dehydrogenase
standard error
glutamic oxaloacetic transaminase, also
known as AST
glutamic pyruvic transaminase, also
known as ALT
systemic scleroderma
trichloroacetic acid
trichloroethylene
time-weighted average
uncertainty factor
animal-to-human uncertainty factor
human variation uncertainty factor
LOAEL-to-NOAEL uncertain factor
subchronic-to-chronic uncertainty
factor
database deficiencies uncertainty factor
United States
           This document is a draft for review purposes only and does not constitute Agency policy.
                                              vii              DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review of tert-ButyI Alcohol
AUTHORS  |  CONTRIBUTORS  |  REVIEWERS
 Assessment Team
     Janice S. Lee, Ph.D. (Chemical Manager)
     Keith Salazar, Ph.D.* (Co-Chemical
     Manager)
     Chris Brinkerhoff, Ph.D.
U.S. EPA
Office of Research and Development
National Center for Environmental Assessment
Research Triangle Park, NC
*Washington, DC

ORISE Postdoctoral Fellow at U.S.
EPA/ORD/NCEA
Currently with U.S. EPA, Office of Chemical Safety
and Pollution Prevention, Office of Pollution
Prevention and Toxics
Washington, DC
 Contributors
     Andrew Hotchkiss, Ph.D.
     Channa Keshava, Ph.D.
     Amanda Persad, Ph.D.
     Vincent Cogliano, Ph.D.*
     Jason Fritz, Ph.D.*
     Catherine Gibbons, Ph.D. *
     Samantha Jones, Ph.D. *
     Kathleen Newhouse *
     Karen Hogan *
U.S. EPA
Office of Research and Development
National Center for Environmental Assessment
Research Triangle Park, NC
*Washington, DC
 Production Team
     Maureen Johnson
     Vicki Soto
U.S. EPA
Office of Research and Development
National Center for Environmental Assessment
Washington, DC
 Contractor Support
     Robyn Blain, Ph.D.
     Michelle Cawley*
     William Mendez, Jr., Ph.D.
     Pam Ross
 ICF International
 Fairfax, VA
 *Research Triangle Park, NC
 Executive Direction
     Kenneth Olden, Ph.D., Sc.D., L.H.D. (Center Director)
     John Vandenberg, Ph.D,# (National Program Director, Human
     Health Risk Assessment)
     Lynn Flowers, Ph.D., DABT (Associate Director for Health)
     Vincent Cogliano, Ph.D. (IRIS Program Director)
     Samantha Jones, Ph.D. (IRIS Associate Director for Science)
     Weihsueh A. Chiu, Ph.D. (Branch Chief, Toxicity Pathways
     Branch) formerly with the U.S. EPA
                  U.S. EPA/ORD/NCEA
                  Washington, DC
                  *Cincinnati, OH
                  # Research Triangle Park, NC
          This document is a draft for review purposes only and does not constitute Agency policy.
                                           viii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review of tert-ButyI Alcohol
          Andrew Hotchkiss, Ph.D.# (Acting Branch Chief, Toxicity
          Pathways Branch)
          Jason Lambert, Ph.D., DABT* (Acting Branch Chief, Biological
          Risk Assessment Branch)
          Ted Berner (Assistant Center Director)
          Karen Hogan (former Acting Branch Chief, Toxicity Effects
          Branch)
       Internal Review Team
          General Toxicology Workgroup         U.S. EPA
          Inhalation Workgroup                 Office of Research and Development
          Neurotoxicity Workgroup              National Center for Environmental Assessment
          Pharmacokinetics Workgroup          Washington, DC
          Reproductive and Developmental       Research Triangle Park, NC
          Toxicology Workgroup                Cincinnati, OH
          Statistical Workgroup
          Toxicity Pathways Workgroup
          Executive Review Committee
 2    	
       Reviewers

 3     This assessment was provided for review to scientists in EPA's Program and Region Offices.
 4     Comments were submitted by:

 5           Office of the Administrator/Office of Children's Health Protection
 6           Office of Land and Emergency Management
 7           Region 2, New York, NY
 8           Region 8, Denver, CO

 9     This assessment was provided for review to other federal agencies and the Executive Office of the
10     President Comments were submitted by:

11           Department of Health and Human Services/Agency for Toxic Substances and Disease Registry,
12           Department of Health and Human Services/National Institute of Environmental Health
13           Sciences/National Toxicology Program,
14           Executive Office of the President/Office of Management and Budget,
15           Executive Office of the President/Office of Science and Technology Policy
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   ix              DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol
     PREFACE
 3          This Toxicological Review critically reviews the publicly available studies on tert-butyl
 4   alcohol (tert-butanol) to identify its adverse health effects and to characterize exposure-response
 5   relationships. The assessment examined all effects by oral and inhalation routes of exposure and
 6   includes an oral noncancer reference dose (RfD), an inhalation noncancer reference concentration
 7   (RfC), a cancer weight of evidence descriptor, and a cancer dose-response assessment. It was
 8   prepared under the auspices of the U.S. Environmental Protection Agency's (EPA's) Integrated Risk
 9   Information System (IRIS) program. This is the first IRIS assessment for this chemical.
10          Toxicological Reviews for tert-butanol and ethyl tert-butyl ether (ETBE) were developed
11   simultaneously because they have several overlapping scientific aspects.
12       •   tert-Butanol is one of the primary metabolites  of ETBE, and some  of the toxicological effects
13          of ETBE are attributed to tert-butanol. Therefore, data on ETBE are considered informative
14          for the hazard identification and dose-response assessment of tert-butanol, and vice versa.
15       •   The scientific literature for the two chemicals includes data on (X2u-globulin-related
16          nephropathy; therefore, a common approach was employed to evaluate these data as they
17          relate to the mode of action for kidney effects.
18       •   A combined physiologically based pharmacokinetic (PBPK) model for tert-butanol and
19          ETBE  in rats was modified to support the dose-response assessments for these chemicals
20          [Salazaretal.. 2015].
21          A public meeting was held in December 2013 to obtain input on preliminary materials for
22   tert-butanol, including draft literature searches and associated search strategies, evidence tables,
23   and exposure-response arrays prior to the development of the IRIS assessment All public
24   comments provided were taken into consideration in developing the draft assessment. The
25   complete set of public comments is available on the docket at http: / /www.regulations.gov (Docket
26   ID No. EPA-HQ-ORD-2013-0111).
27          Organ/system-specific reference values are calculated based on kidney and thyroid toxicity
28   data. These reference values could be useful for cumulative risk assessments that consider the
29   combined effect of multiple agents acting on the same  biological system.
30          This assessment was conducted in accordance  with EPA guidance, which is cited and
31   summarized in the Preamble to IRIS Toxicological Reviews. The findings of this assessment and
32   related documents produced during its development are available on the  IRIS website
33   (http: //www. epa. gov/iris). Appendices for toxicokinetic information, PBPK modeling, genotoxicity
34   study summaries, dose-response modeling, and other  information are provided as Supplemental
35   Information to this Toxicological Review. For additional information about this assessment or for
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 x             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   general questions regarding IRIS, please contact EPA's IRIS Hotline at 202-566-1676 (phone), 202-
 2   566-1749 (fax), or hotline.iris@epa.gov.

 3   Uses
 4          tert-Butanol primarily is an anthropogenic substance that is produced in large quantities
 5   (HSDB, 2007) from several precursors, including 1-butene, isobutylene, acetyl chloride and
 6   dimethylzinc, and tert-butyl hydroperoxide. The domestic production volume of tert-butanol,
 7   including imports, was approximately 4 billion pounds in 2012 (U.S. EPA. 2014).
 8          tert-Butanol has been used as a fuel oxygenate, an octane booster in unleaded gasoline, and
 9   a denaturant for ethanol. From 1997 to 2005, the annual tert-butanol volume found in gasoline
10   ranged from approximately 4 million to 6 million gallons. During that time, larger quantities were
11   used to make methyl tert-butyl ether (MTBE) and ETBE. MTBE and ETBE are fuel oxygenates that
12   were used in the United States prior to 2007 at levels of more than 2 billion gallons annually.
13   Current use levels of MTBE and ETBE in the United States are much lower, but use in Europe and
14   Asia remains strong.1
15          tert-Butanol has been used for a variety of other purposes, including as a dehydrating agent
16   and solvent. As such, it is added to lacquers, paint removers, and nail enamels and polishes.
17   tert-Butanol also is used to manufacture methyl methacrylate plastics and flotation devices.
18   Cosmetic and food-related uses include the manufacture of flavors, and, because of its camphor-like
19   aroma, it also is used to create artificial musk, fruit essences, and perfume (HSDB, 2007). It is used
20   in coatings on metal and paperboard food containers (Cal/EPA. 1999) and industrial cleaning
21   compounds, and can be used for chemical extraction in pharmaceutical applications (HSDB. 2007).

22   Fate and Transport
23   5oi7
24          tert-Butanol is expected to be highly mobile in soil due to its low affinity for soil organic
25   matter. Rainwater or other water percolating through soil is expected to dissolve and transport
26   most tert-butanol present in soil, potentially leading to groundwater contamination. Based on its
27   vapor pressure, tert-butanol's volatilization from soil surfaces is expected to be an important
28   dissipation process (HSDB. 2007). As a tertiary alcohol, tert-butanol is expected to degrade more
29   slowly in the environment compared to primary (e.g., ethanol) or secondary (e.g., isopropanol)
30   alcohols. In anoxic soil conditions, the half-life of tert-butanol is estimated to be months
31   (approximately 200 days). Microbial degradation rates are increased in soils supplemented with
32   nitrate and sulfate nutrients (HSDB. 2007).
     1 http://www.ihs.com/products/chemical/planning/ceh/gasoline-octane-improvers.aspx.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 xi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   Water
 2          tert-Butanol is expected to volatilize from water surfaces within 2 to 29 days and does not
 3   readily adsorb to suspended solids and sediments in water [HSDB. 2007). Biodegradation in
 4   aerobic water occurs over weeks to months and in anaerobic aquatic conditions, the biodegradation
 5   rate decreases. Bioconcentration of tert-butanol in aquatic organisms is low [HSDB. 2007).

 6   Air
 7          tert-Butanol primarily exists as a vapor in the ambient atmosphere. Vapor-phase tert-
 8   butanol is degraded in the atmosphere by reacting with photochemically produced hydroxyl
 9   radicals with a half-life of 14 days [HSDB. 20071.

10   Occurrence in the Environment
11          The Toxics Release Inventory (TRI) Program National Analysis Report estimated that more
12   than 1 million pounds of tert-butanol has been released into the soil from landfills, land treatment,
13   underground injection, surface impoundments, and other land disposal sources. The TRI program
14   also estimated that 476,266 pounds of tert-butanol was released into the atmosphere from  fugitive
15   emissions and point sources [U.S. EPA. 2012c]. In California, air emissions of tert-butanol from
16   stationary sources are estimated to be at least 27,000 pounds per year, based on data reported by
17   the state's Air Toxics Program [Scorecard, 2014].  The TRI National Analysis Report estimated 7,469
18   pounds of tert-butanol was released into surface waters from point and nonpoint sources in 2011
19   [U.S. EPA. 2012c).
20          tert-Butanol has been identified in drinking water wells throughout the United  States
21   [HSDB. 20071. California's Geotracker Database2 lists 3,496 detections of tert-butanol in
22   groundwater associated with contaminated sites in that state since 2011.  tert-Butanol also has been
23   detected in drinking water wells in the vicinity of landfills [U.S. EPA. 2012c]. Additionally, tert-
24   Butanol leaking from underground storage tanks could be a product of MTBE and ETBE, which can
25   degrade to form tert-butanol in soils [HSDB. 2007]. The industrial chemical tert-butyl acetate also
26   can degrade to form tert-butanol in animals post exposure and in the environment
27          Ambient outdoor air concentrations of tert-butanol vary according to proximity to urban
28   areas [HSDB. 20071.

29   General  Population Exposure
30          tert-Butanol exposure can occur in many different settings. Releases from underground
31   storage tanks could potentially result in exposure for people who get their drinking water from
32   wells. Due to its high environmental mobility and resistance to biodegradation, tert-butanol has the
33   potential to contaminate and persist in groundwater and soil [HSDB. 2007].
     2 http://geotracker.waterboards.ca.gov/.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 xii              DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1          Ingestion of contaminated food can be a source of tert-butanol exposure through its use as a
 2   coating in metallic and paperboard food containers [Gal/EPA, 1999], and tert-butanol has been
 3   detected in food [HSDB. 2007). Internal exposure to tert-butanol also can occur as a result of
 4   ingestion of MTBE or ETBE, as tert-butanol is a metabolite of these compounds [NSF International,
 5   20031
 6          Other human exposure pathways include inhalation, lactation and, to a lesser extent, dermal
 7   contact. Inhalation exposure can occur due to the chemical's volatility and release from industrial
 8   processes, consumer products, and contaminated sites [HSDB. 2007]. tert-Butanol has been
 9   identified in mother's milk [HSDB. 2007]. Dermal contact is a viable route of exposure through
10   handling consumer products containing tert-butanol [NSF International. 2003].

11   Assessments by Other National and International Health Agencies
12          Toxicity information on tert-butanol has been evaluated by the National Institute for
13   Occupational Safety and Health [NIOSH, 2007], the Occupational Safety and Health Administration
14   [OSHA. 2006], and the Food and Drug Administration [FDA. 201 la. b]. The results of these
15   assessments are presented in Appendix A of the Supplemental Information to this Toxicological
16   Review. Of importance to recognize is that these earlier assessments could have been prepared for
17   different purposes and might use different methods. In addition, newer studies have been included
18   in the IRIS assessment.

                       4                                 ^
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                xiii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review of tert-ButyI Alcohol
    PREAMBLE TO  IRIS TOXICOLOGICAL REVIEWS
 3 Note: The Preamble to IRIS assessments is
 4 being revised based on comments received
 5 from external peer reviewers and the
 6 public, and based on IRIS Program
 1 experience with the implementation of
 8 systematic review methods. Subsequent
 9 drafts of th e tert-b utan ol assessm en t will
10 include the revised Preamble.


11 1.  Scope of the IRIS Program

12     Soon after  the EPA was established in
13 1970, itwas atthe forefront of developing risk
14 assessment as  a science and applying  it in
15 decisions to  protect human health and the
16 environment. The Clean Air Act, for example,
17 mandates that  the EPA provide "an ample
18 margin of safety to protect  public health"; the
19 Safe Drinking Water Act,  that "no adverse
20 effects  on  the  health   of  persons   may
21 reasonably be anticipated  to occur, allowing
22 an  adequate  margin of safety." Accordingly,
23 the  EPA uses  information on the adverse
24 effects of chemicals and on exposure levels
25 below which these effects are not anticipated
26 to occur.
27     IRIS  assessments  critically review the
28 publicly available studies to identify adverse
29 health effects from exposure to chemicals and
30 to      characterize     exposure-response
31 relationships.  In  terms set  forth  by the
32 National Research  Council  fNRC. 19831 IRIS
33 assessments cover the  hazard identification
34 and dose-response assessment steps of risk
35 assessment, not the exposure assessment or
36 risk characterization steps that are conducted
37 by the EPA's program and regional offices and
38 by  other  federal, state,  and  local  health
39 agencies  that  evaluate   risk  in   specific
40 populations  and  exposure  scenarios.  IRIS
41 assessments  are distinct  from and do not
42  address  political,  economic,  and  technical
43  considerations that influence the design and
44  selection of risk management alternatives.
45     An IRIS  assessment may  cover a single
46  chemical,   a  group   of  structurally  or
47  toxicologically related chemicals, or a complex
48  mixture.  These  agents may be found in  air,
49  water, soil,  or sediment Exceptions  are
50  chemicals  currently  used  exclusively  as
51  pesticides,    ionizing    and    non-ionizing
52  radiation, and  criteria air pollutants  listed
53  under Section 108 of the Clean Air Act (carbon
54  monoxide,   lead,   nitrogen oxides,  ozone,
55  particulate matter, and sulfur oxides).
56     Periodically, the IRIS Program asks other
57  EPA  programs  and  regions,  other federal
58  agencies,  state  health  agencies,  and  the
59  general public  to  nominate  chemicals and
60  mixtures    for    future    assessment   or
61  reassessment Agents may be  considered for
62  reassessment as significant new studies  are
63  published. Selection is based on program and
64  regional office priorities and on availability of
65  adequate information to evaluate the potential
66  for adverse effects. Other agents may also be
67  assessed in response to  an  urgent public
68  health need.

69  2.  Process for developing and peer-
70     reviewing IRIS assessments

71     The   process   for   developing   IRIS
72  assessments  (revised  in  May  2009  and
73  enhanced in July  2013)  involves critical
74  analysis of the pertinent studies, opportunities
75  for  public  input,  and  multiple  levels of
76  scientific  review.  The  EPA  revises  draft
77  assessments after each review, and external
78  drafts and  comments  become  part  of  the
79  public record (U.S. EPA. 2009).
80     Before beginning an assessment, the IRIS
81  program discusses the scope with other EPA
82  programs and  regions to  ensure  that  the
             This document is a draft for review purposes only and does not constitute Agency policy.
                                             xiv            DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review of tert-ButyI Alcohol
 1  assessment will meet their needs. Then a
 2  public  meeting  on  problem  formulation
 3  invites discussion of the key issues and the
 4  studies and analytical approaches that might
 5  contribute to their resolution.
 6  Step 1.    Development    of    a   draft
 7     Toxicological    Review.   The    draft
 8     assessment  considers   all   pertinent
 9     publicly available  studies  and  applies
10     consistent  criteria  to  evaluate  study
11     quality,  identify  health  effects, identify
12     mechanistic   events   and   pathways,
13     integrate the evidence of causation for
14     each effect, and derive toxicity values. A
15     public meeting prior to the integration of
16     evidence and derivation of toxicity values
17     promotes   public   discussion  of  the
18     literature search, evidence, and key issues.
19  Step 2. Internal review by scientists in EPA
20     programs   and  regions.   The   draft
21     assessment  is  revised to  address the
22     comments from within the EPA.
23  Step 3.  Interagency science consultation
24     with other  federal agencies and the
25     Executive Offices of the President. The
26     draft assessment is revised to address the
27     interagency  comments.   The   science
28     consultation draft, interagency comments,
29     and  the  EPA's  response   to   major
30     comments become  part  of the public
31     record.
32  Step 4.  Public review  and   comment,
33     followed by external peer review. The
34     EPA  releases the draft assessment for
35     public  review  and  comment  A public
36     meeting  provides  an  opportunity  to
37     discuss  the  assessment  prior  to  peer
38     review. Then the EPA releases a draft for
39     external peer review. The  peer review
40     meeting is open to the public and includes
41     time  for oral public  comments. The peer
42     reviewers  assess  whether the evidence
43     has  been  assembled  and  evaluated
44     according  to guidelines and whether the
45     conclusions are justified by the evidence.
46     The  peer  review draft, written public
47     comments,  and  peer  review  report
48     become part of the public record.
49  Step 5.  Revision  of draft Toxicological
50     Review and development of draft IRIS
51     summary. The draft assessment is revised
52     to reflect the  peer  review  comments,
53     public  comments, and  newly published
54     studies that are  critical to the conclusions
55     of the assessment. The disposition of peer
56     review comments and public comments
57     becomes part of the public record.
58  Step 6. Final EPA review and interagency
59     science discussion with other federal
60     agencies and the Executive Offices of
61     the President The draft assessment and
62     summary are revised to address the EPA
63     and interagency comments. The science
64     discussion  draft,  written   interagency
65     comments, and  EPA's response to  major
66     comments become  part of the  public
67     record.
68  Step 7.  Completion and  posting.  The
69     Toxicological Review and IRIS summary
70     are  posted  on  the   IRIS  website
71     [http://www.epa.gov/iris/].
72  The remainder of this Preamble addresses step 1,
73     the  development  of a draft  Toxicological
74     Review. IRIS assessments follow standard
75     practices of evidence evaluation and peer
76     review,  many of which are discussed in
77     EPA     guidelines      [U.S.      EPA.
78     2005a. b, 2000b. 1998b. 1996. 1991b. 198
79     6a,  b)   and  other methods  [U.S. EPA.
80     2012a.  b,  2011. 2006a.  b, 2002.  19941
81     Transparent  application  of  scientific
82     judgment is of paramount importance. To
83     provide a harmonized  approach  across
84     IRIS   assessments,    this    Preamble
85     summarizes   concepts    from   these
86     guidelines and emphasizes  principles of
87     general applicability.

88  3.  Identifying and selecting
89     pertinent studies

90  3.1.  Identifying studies
91     Before beginning an assessment, the EPA
92  conducts  a  comprehensive  search  of  the
93  primary scientific  literature. The  literature
             This document is a draft for review purposes only and does not constitute Agency policy.
                                              xv             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
 1  search follows standard practices and includes
 2  the  PubMed and ToxNet databases  of the
 3  National Library of Medicine, Web of Science,
 4  and other databases listed in the EPA's HERO
 5  system (Health and Environmental Research
 6  Online, http://hero.epa.gov/].  Searches for
 7  information on mechanisms of toxicity are
 8  inherently  specialized  and may  include
 9  studies on other  agents  that  act through
10  related mechanisms.
11     Each  assessment  specifies   the  search
12  strategies, keywords, and cut-off dates of its
13  literature searches. The EPA posts the results
14  of the literature search on the IRIS web site
15  and requests information from the public on
16  additional studies and ongoing research.
17     The EPA also considers studies received
18  through the IRIS Submission Desk and studies
19  (typically unpublished) submitted under the
20  Toxic Substances Control Act or  the Federal
21  Insecticide, Fungicide, and Rodenticide  Act
22  Material submitted  as Confidential Business
23  Information is  considered only if it includes
24  health  and safety data that  can  be publicly
25  released. If a study that may be critical to the
26  conclusions of the  assessment has  not been
27  peer-reviewed, the  EPA  will have it peer-
28  reviewed.
29     The EPA also examines the toxicokinetics
30  of the agent to identify other chemicals (for
31  example, major metabolites of the  agent) to
32  include  in  the   assessment  if  adequate
33  information is available, in order to more fully
34  explain the toxicity of the agent and to suggest
35  dose metrics for subsequent modeling.
36     In  assessments of  chemical  mixtures,
37  mixture studies are preferred for their ability
38  to reflect interactions among components.
39     The literature search seeks, in decreasing
40  order of preference (U.S. EPA.
41  2000b. 32.2: 1986b. 32.11:
42     -   Studies of the mixture being assessed.
43     -   Studies   of  a  sufficiently  similar
44         mixture. In evaluating similarity, the
45         assessment considers the alteration of
46         mixtures in the environment through
47         partitioning and transformation.
48     -   Studies   of   individual   chemical
49         components of the mixture, if there are
50         not adequate studies of sufficiently
51         similar mixtures.

52  3.2.  Selecting pertinent epidemiologic
53       studies
54     Study design is the  key consideration for
55  selecting pertinent epidemiologic studies from
56  the results of the literature  search.
57     -   Cohort studies,  case-control studies,
58         and some population-based surveys
59         (for example, NHANES) provide the
60         strongest   epidemiologic   evidence,
61         especially if they collect  information
62         about   individual   exposures   and
63         effects.
64     -   Ecological    studies     (geographic
65         correlation  studies) relate exposures
66         and effects  by geographic area.  They
67         can provide strong evidence if there
68         are large exposure  contrasts between
69         geographic   areas,   relatively  little
70         exposure variation within study areas,
71         and population migration is limited.
72     -   Case reports of high  or accidental
73         exposure  lack   definition  of  the
74         population  at risk  and the expected
75         number  of cases. They can provide
76         information  about  a  rare  effect or
77         about the  relevance  of analogous
78         results in animals.
79     The assessment briefly  reviews ecological
80  studies and case reports but reports details
81  only  if they suggest effects not identified by
82  other studies.

83  3.3.  Selecting pertinent experimental
84       studies
85     Exposure    route   is   a  key  design
86  consideration   for    selecting    pertinent
87  experimental animal studies or human clinical
88  studies.
89     -   Studies of oral,  inhalation, or dermal
90         exposure involve passage through an
91         absorption barrier and are considered
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xvi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-Butyl Alcohol
 1
 2

 3
 4
 5
 6
 7
 8
 9
10
most    pertinent     to
environmental exposure.
human
45  quality  characteristics
46  similar design.
across  studies  of
Injection or implantation studies are
often considered less  pertinent  but
may provide valuable toxicokinetic or
mechanistic information.  They  also
may be useful for identifying effects in
animals if deposition or absorption is
problematic (for example, for particles
and fibers).
11     Exposure  duration  is also a key design
12  consideration   for    selecting    pertinent
13  experimental animal studies.
14     -   Studies  of  effects   from  chronic
15         exposure  are   most  pertinent  to
16         lifetime human exposure.
17     -   Studies  of  effects  from  less-than-
18         chronic  exposure are pertinent but
19         less preferred for identifying effects
20         from lifetime human exposure.  Such
21         studies may be indicative of effects
22         from    less-than-lifetime     human
23         exposure.
24     Short-duration studies involving animals
25  or humans  may  provide  toxicokinetic  or
26  mechanistic information.
27     For    developmental    toxicity    and
28  reproductive toxicity, irreversible effects may
29  result from a brief exposure  during a critical
30  period   of   development.    Accordingly,
31  specialized study  designs are used for these
32  effects [U.S. EPA. 2006b. 1998b. 1996.1991b).

33  4. Evaluating the quality of
34     individual studies

35     After   the   subsets    of    pertinent
36  epidemiologic and experimental studies have
37  been selected from the literature searches, the
38  assessment evaluates   the  quality  of  each
39  individual study. This evaluation considers the
40  design, methods, conduct, and documentation
41  of each study, but  not whether the results are
42  positive, negative, or null. The objective is to
43  identify the   stronger,  more  informative
44  studies  based on a uniform  evaluation of
         47  4.1. Evaluating the quality of
         48       epidemiologic studies
         49
         50
         51
         52
         53

         54
         55
         56

         57
         58

         59
         60

         61
         62

         63
         64
         65

         66
         67

         68
         69
       The  assessment evaluates design  and
    methodological aspects that can increase or
    decrease  the   weight   given  to   each
    epidemiologic study in the overall evaluation
    fU.S. EPA. 2005a. 1998b. 1996.1994.1991bl:

       -  Documentation   of  study   design,
          methods, population characteristics,
          and results.

       -  Definition and selection of the study
          group and comparison group.

       -  Ascertainment of  exposure  to  the
          chemical or mixture.

          Ascertainment of disease or health
          effect

       -  Duration of exposure and follow-up
          and  adequacy  for  assessing  the
          occurrence of effects.

          Characterization  of exposure during
          critical periods.

       -  Sample size and statistical power to
          detect anticipated effects.
                                      70     -  Participation rates  and potential for
                                      71        selection  bias as  a  result  of  the
                                      72        achieved participation rates.
                                      73     -  Measurement  error   (can  lead  to
                                      74        misclassification of exposure, health
                                      75        outcomes, and other factors) and other
                                      76        types of information bias.
                                      77     -  Potential   confounding  and   other
                                      78        sources of bias addressed in the study
                                      79        design or in the analysis of results. The
                                      80        basis for consideration of confounding
                                      81        is a reasonable expectation that the
                                      82        confounder is related to both exposure
                                      83        and  outcome  and   is   sufficiently
                                      84        prevalent to result in bias.
                                      85     For developmental toxicity, reproductive
                                      86 toxicity,  neurotoxicity, and cancer there is
                                      87 further guidance on the nuances of evaluating
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xvii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
 1  epidemiologic  studies of these effects [U.S.
 2  EPA. 2005a. 1998b. 1996.1991bl

 3  4.2. Evaluating the quality of
 4      experimental studies
 5     The  assessment evaluates design  and
 6  methodological aspects that can increase or
 7  decrease   the   weight   given  to   each
 8  experimental animal study, in-vitro study, or
 9  human    clinical    study    [U.S.    EPA.
10  2005a.  1998b.  1996.   1991bl   Research
11  involving human subjects is considered only if
12  conducted according to ethical principles.
13     -  Documentation   of  study   design,
14        animals or study population, methods,
15        basic data, and results.
16     -  Nature  of the assay and validity for its
17        intended purpose.
18     -  Characterization  of the nature  and
19        extent of impurities and contaminants
20        of  the  administered  chemical  or
21        mixture.
22     -  Characterization  of dose and dosing
23        regimen (including age at exposure)
24        and their adequacy to  elicit adverse
25        effects,  including latent effects.
26     -  Sample sizes and statistical power to
27        detect  dose-related  differences  or
28        trends.
29     -  Ascertainment of survival, vital signs,
30        disease or effects, and cause of death.
31     -  Control of other  variables that could
32        influence the occurrence of effects.
33     The assessment uses statistical tests to
34  evaluate whether the observations may be due
35  to  chance. The standard  for determining
36  statistical significance of  a response is a trend
37  test or comparison of outcomes in the exposed
38  groups against those of concurrent controls. In
39  some  situations,  examination  of  historical
40  control data from the same laboratory within
41  a few years of the study may improve the
42  analysis. For an uncommon effect that is not
43  statistically  significant   compared   with
44  concurrent controls, historical controls may
45  show that the effect is unlikely to be due to
46  chance. For a response that appears significant
47  against a concurrent control response that is
48  unusual,  historical  controls   may  offer  a
49  different    interpretation     [U.S.     EPA.
50  2005a. §2.2.2.1.31
51     For developmental toxicity, reproductive
52  toxicity, neurotoxicity, and cancer  there  is
53  further guidance on the nuances of evaluating
54  experimental studies of these effects [U.S. EPA.
55  2005a.  1998b.  1996.  1991bl   In multi-
56  generation  studies,  agents  that  produce
57  developmental effects at doses that are not
58  toxic to the maternal animal are of special
59  concern. Effects that occur at doses associated
60  with mild maternal toxicity are not assumed to
61  result only from maternal toxicity. Moreover,
62  maternal  effects may be reversible,  while
63  effects  on the  offspring may be permanent
64  [U.S.   EPA.   1998b. §3.1.2.4.5.4:    1991b.
65  §3.1.1.41.

66  4.3.  Reporting study results
67     The  assessment uses evidence tables to
68  present the design and key results of pertinent
69  studies. There may be separate tables for each
70  site of toxicity or type of study.
71     If a large number of studies observe the
72  same  effect, the assessment considers the
73  study quality characteristics in this section to
74  identify the strongest studies or types of study.
75  The tables present details from these studies,
76  and the assessment explains the  reasons  for
77  not reporting  details of other  studies  or
78  groups  of studies  that do  not add  new
79  information.   Supplemental   information
80  provides references to all studies considered,
81  including those not summarized in the tables.
82     The  assessment discusses  strengths and
83  limitations that affect the  interpretation of
84  each study. If the interpretation of a study in
85  the assessment differs from that of the study
86  authors, the assessment discusses the basis for
87  the difference.
88     As a check on the selection and evaluation
89  of pertinent studies, the  EPA  asks  peer
90  reviewers to identify studies that were not
91  adequately considered.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xviii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
 i  5.  Evaluating the overall evidence of
 2     each effect

 3  5.1. Concepts of causal inference

 4     For each  health  effect, the assessment
 5  evaluates  the  evidence  as   a   whole  to
 6  determine whether it is reasonable to infer a
 7  causal association between exposure to the
 8  agent and the occurrence of the effect This
 9  inference  is  based  on  information  from
10  pertinent human studies, animal studies, and
11  mechanistic   studies  of  adequate  quality.
12  Positive, negative, and null results are given
13  weight according to study quality.
14     Causal  inference   involves   scientific
15  judgment, and the considerations are nuanced
16  and complex.  Several health  agencies  have
17  developed frameworks  for causal  inference,
18  among them the U.S. Surgeon General [CDC.
19  2004: HEW. 1964). the  International Agency
20  for Research  on  Cancer [IARC. 2006]. the
21  Institute of Medicine (IOM. 2008). and the EPA
22  (2010. S1.6: 2005a. 52.51 Although  developed
23  for different  purposes,  the frameworks are
24  similar in nature and provide an established
25  structure and language for causal  inference.
26  Each considers aspects of an association that
27  suggest  causation, discussed by  Hill  (Hill.
28  1965]  and elaborated on by Rothman and
29  Greenland (Rothman and Greenland. 1998].
30  and U.S. EPA f2005a. §2.2.1.7: 1994. Appendix
31  C].
32  Strength of association: The finding of a large
33     relative risk  with   narrow  confidence
34     intervals   strongly   suggests   that an
35     association is not due to chance, bias, or
36     other  factors.  Modest  relative  risks,
37     however,  may reflect a  small range  of
38     exposures, an agent of low potency, an
39     increase  in an effect that is  common,
40     exposure   misclassification,   or   other
41     sources of bias.
42  Consistency of association: An inference of
43     causation  is strengthened if elevated risks
44     are observed in independent  studies of
45     different   populations  and   exposure
46     scenarios.  Reproducibility  of findings
47     constitutes one of the strongest arguments
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xix             DRAFT—DO NOT CITE OR QUOTE
48     for   causation.   Discordant    results
49     sometimes reflect differences  in study
50     design, exposure, or confounding factors.
51  Specificity  of  association:  As  originally
52     intended,  this  refers  to  one   cause
53     associated  with   one  effect   Current
54     understanding  that many  agents cause
55     multiple  effects and many effects have
56     multiple   causes  make  this   a  less
57     informative aspect of causation, unless the
58     effect is rare or unlikely to have multiple
59     causes.
60  Temporal    relationship:    A    causal
61     interpretation  requires  that  exposure
62     precede development of the effect.
63  Biologic   gradient   (exposure-response
64     relationship):       Exposure-response
65     relationships strongly suggest causation. A
66     monotonic increase is not the only pattern
67     consistent with causation. The presence of
68     an   exposure-response  gradient  also
69     weighs against bias and confounding as
70     the source of an association.
71  Biologic   plausibility:  An  inference   of
72     causation  is   strengthened   by  data
73     demonstrating     plausible    biologic
74     mechanisms, if available. Plausibility may
75     reflect subjective  prior beliefs if there  is
76     insufficient understanding of the biologic
77     process involved.
78  Coherence:  An  inference  of  causation  is
79     strengthened by supportive results from
80     animal experiments, toxicokinetic studies,
81     and short-term tests. Coherence may also
82     be found in other lines of evidence,  such as
83     changing   disease   patterns   in   the
84     population.
85  "Natural experiments": A change in exposure
86     that  brings  about a change  in  disease
87     frequency provides strong evidence, as  it
88     tests  the hypothesis  of causation.  An
89     example  would be  an intervention  to
90     reduce  exposure in the  workplace  or
91     environment   that  is   followed  by  a
92     reduction of an adverse effect

-------
                                                      Toxicological Review of tert-Butyl Alcohol
 1  Analogy: Information on structural analogues
 2     or  on  chemicals  that  induce  similar
 3     mechanistic events can  provide insight
 4     into causation.
 5     These considerations are consistent with
 6  guidelines  for  systematic  reviews   that
 7  evaluate the quality and weight of evidence.
 8  Confidence  is increased if the magnitude of
 9  effect is  large, if  there  is  evidence  of an
10  exposure-response  relationship,  or  if  an
11  association  was observed and the plausible
12  biases would tend to decrease the magnitude
13  of the reported effect Confidence is decreased
14  for study limitations, inconsistency of results,
15  indirectness of  evidence,  imprecision,  or
16  reporting bias [Guyatt et al., 2008b: Guyatt et
17  al.. 2008a].

is  5.2.  Evaluating evidence in humans
19     For each effect, the assessment evaluates
20  the evidence from the epidemiologic studies as
21  a whole. The objective is to determine whether
22  a credible association has been observed and,
23  if so, whether  that association is consistent
24  with causation. In doing this, the assessment
25  explores  alternative explanations (such as
26  chance, bias, and confounding) and draws a
27  conclusion about whether these  alternatives
28  can  satisfactorily   explain   any  observed
29  association.
30     To    make   clear  how   much   the
31  epidemiologic  evidence contributes to the
32  overall weight of the evidence, the assessment
33  may   select   a   standard   descriptor  to
34  characterize the epidemiologic evidence of
35  association between exposure to the agent and
36  occurrence of a health effect
37  Sufficient epidemiologic  evidence of  an
38     association consistent with  causation:
39     The   evidence   establishes   a  causal
40     association   for   which    alternative
41     explanations  such  as  chance, bias,  and
42     confounding  can   be  ruled  out  with
43     reasonable  confidence.
44  Suggestive  epidemiologic evidence of an
45     association consistent with  causation:
46     The evidence suggests a causal association
47     but chance, bias, or confounding cannot be
48     ruled out as explaining the association.
49  Inadequate epidemiologic evidence to infer
50     a  causal  association:  The  available
51     studies  do  not  permit  a  conclusion
52     regarding the presence  or absence  of an
53     association.
54  Epidemiologic evidence consistent with no
55     causal  association:  Several  adequate
56     studies covering the full range of human
57     exposures  and  considering  susceptible
58     populations,  and  for which alternative
59     explanations such as bias and confounding
60     can be ruled  out, are mutually consistent
61     in not finding an association.

62  5.3.  Evaluating evidence in animals

63     For each effect, the assessment evaluates
64  the evidence from the animal experiments as a
65  whole to determine the extent to which they
66  indicate a potential  for  effects  in humans.
67  Consistent results across various species and
68  strains increase confidence that similar results
69  would  occur  in  humans. Several  concepts
70  discussed by Hill [Hill. 1965] are pertinent to
71  the   weight   of    experimental    results:
72  consistency   of   response,  dose-response
73  relationships, strength of response, biologic
74  plausibility,   and   coherence   [U.S.   EPA,
75  2005a. 52.2.1.7: 1994. Appendix C1.
76     In  weighing  evidence   from   multiple
77  experiments,    U.S.    EPA   [2005a. §2.5]
78  distinguishes:
79  Conflicting evidence (that is, mixed positive
80     and negative  results in the same sex and
81     strain using a similar study protocol] from
82  Differing results (that is, positive results and
83     negative results are in different sexes  or
84     strains or use different study protocols].
85     Negative or null results  do not invalidate
86  positive results in a different experimental
87  system.  The   EPA  regards  all  as  valid
88  observations and looks to  explain differing
89  results using  mechanistic  information (for
90  example, physiologic or metabolic differences
91  across   test  systems]  or   methodological
              This document is a draft for review purposes only and does not constitute Agency policy.
                                                xx              DRAFT—DO NOT CITE OR QUOTE

-------
                                        Toxicological Review of tert-ButyI Alcohol
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


differences (for example, relative sensitivity of
the tests, differences in dose levels,
insufficient sample size, or timing of dosing or
data collection).
It is well established that there are critical
periods for some developmental and
reproductive effects (U.S. EPA.

2006b. 2005a. b, 1998b. 1996. 1991b).
Accordingly, the assessment determines
whether critical periods have been adequately
investigated. Similarly, the assessment
determines whether the database is adequate
to evaluate other critical sites and effects.
In evaluating evidence of genetic toxicity:

- Demonstration of gene mutations,
chromosome aberrations, or
aneuploidy in humans or experimental
mammals [in vivo] provides the
strongest evidence.
This is followed by positive results in
lower organisms or in cultured cells
[in vitro] or for other genetic events.
- Negative results carry less weight,
partly because they cannot exclude the
possibility of effects in other tissues
(IARC. 2006).
For germ-cell mutagenicity, The EPA has
defined categories of evidence, ranging from
positive results of human germ-cell
mutagenicity to negative results for all effects
of concern (U.S. EPA. 1986a. §2.3).

5.4. Evaluating mechanistic data
Mechanistic data can be useful in
answering several questions.
- The biologic plausibility of a causal
interpretation of human studies.
The generalizability of animal studies
to humans.
The susceptibility of particular
populations or lifestages.
The focus of the analysis is to describe, if
possible, mechanistic pathways that lead to a
health effect. These pathways encompass:


44
45
46
47
48
49
50
51

52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
U /
68
UO
CQ
by
70
71
/ -L
77
1 £-
73
/O
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91










th(
etc.
be
pr
ste
ev
on
WF
re;
m;
of
th;
In!
foi
rel
nn
qu
Q("*1
dLl
1)











2)






                                             Toxicokinetic processes of absorption,
                                             distribution,     metabolism,    and
                                             elimination that lead to the formation
                                             of an active agent and its presence at
                                             the site of initial biologic interaction.
                                             Toxicodynamic processes that lead to a
                                             health effect at this  or another site
                                             (also known as a mode of action}.
                                         For each effect, the assessment discusses
                                      the  available information  on  its  modes  of
                                      action and associated key events (key events
                                             empirically  observable,   necessary
                                      precursor steps or biologic markers of such
                                      steps; mode of action being a  series of key
                                           ;   involving  interaction  with  cells,
                                      operational  and  anatomic changes,  and
                                      resulting in disease). Pertinent information
                                      may also come from studies of metabolites or
                                      of compounds that are structurally similar or
                                            act   through  similar  mechanisms.
                                      Information on mode of action is not required
                                         a conclusion that the  agent is  causally
                                      related to an effect (U.S. EPA. 2005a. 52.5).
                                         The  assessment   addresses   several
                                      questions about each hypothesized mode of
                                      action fU.S. EPA. 2005a. 32.4.3.41.
                                         Is the  hypothesized  mode  of action
                                         sufficiently supported in test animals?
                                         Strong support  for  a  key event  being
                                         necessary to a mode of action can come
                                         from experimental  challenge  to the
                                         hypothesized mode  of action, in which
                                         studies that suppress a key event observe
                                         suppression of the effect Support for a
                                         mode   of   action   is   meaningfully
                                         strengthened  by consistent  results  in
                                         different  experimental  models,  much
                                         more so than by replicate experiments in
                                         the  same model. The  assessment may
                                         consider various aspects of causation in
                                         addressing this question.
                                         Is the  hypothesized  mode  of action
                                         relevant to  humans?  The assessment
                                         reviews the key events to identify critical
                                         similarities and differences between the
                                         test  animals   and    humans.   Site
                                         concordance  is  not assumed  between
                                         animals and humans, though it may hold
This document is a draft for review purposes only and does not constitute Agency policy.
                                 xxi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-Butyl Alcohol
 1     for certain  effects  or modes of action.
 2     Information   suggesting   quantitative
 3     differences in doses where effects would
 4     occur in animals or  humans is considered
 5     in the dose-response analysis.  Current
 6     levels of human exposure are not used to
 7     rule   out  human  relevance,   as  IRIS
 8     assessments  may be  used in  evaluating
 9     new or  unforeseen circumstances that
10     may entail higher exposures.
11  3)  Which populations or lifestages can be
12     particularly   susceptible   to    the
13     hypothesized  mode  of  action?  The
14     assessment  reviews the key  events  to
15     identify populations and  lifestages that
16     might be susceptible to their occurrence.
17     Quantitative  differences  may result  in
18     separate  toxicity values  for susceptible
19     populations or lifestages.
20     The assessment discusses the likelihood
21  that  an  agent  operates  through  multiple
22  modes of action. An uneven level  of support
23  for  different modes of  action can reflect
24  disproportionate       resources      spent
25  investigating them fU.S. EPA. 2005a. 32.4.3.3).
26  It should be noted that in clinical reviews, the
27  credibility of a series of studies is reduced if
28  evidence is limited to studies funded by one
29  interested sector [Guyattetal.. 2008a).
30     For  cancer,  the assessment evaluates
31  evidence  of  a mutagenic  mode of action to
32  guide extrapolation to  lower  doses  and
33  consideration of susceptible lifestages. Key
34  data include the ability  of the agent  or a
35  metabolite to  react with or bind  to DNA,
36  positive  results in multiple test systems, or
37  similar  properties   and  structure-activity
38  relationships to mutagenic carcinogens [U.S.
39  EPA. 2005a.32.3.51.

40  5.5.  Characterizing the overall weight
41       of the evidence

42     After evaluating the human, animal, and
43  mechanistic  evidence pertinent to an effect,
44  the  assessment answers the question: Does
45  the  agent cause the adverse  effect?  [NRG,
46  2009. 1983). In doing this,  the assessment
47  develops  a  narrative  that  integrates  the
48  evidence pertinent to causation. To provide
49  clarity and consistency, the narrative includes
50  a standard hazard descriptor. For example, the
51  following  standard   descriptors   combine
52  epidemiologic, experimental, and mechanistic
53  evidence of carcinogenicity [U.S. EPA. 2005a.
54  §2,5).
55  Carcinogenic to humans: There is convincing
56     epidemiologic  evidence  of  a  causal
57     association (that is,  there  is reasonable
58     confidence that the association cannot be
59     fully  explained  by  chance,  bias,  or
60     confounding); or there  is strong human
61     evidence  of  cancer  or its  precursors,
62     extensive animal evidence, identification
63     of key precursor events in animals, and
64     strong evidence that they are anticipated
65     to occur in humans.
66  Likely to be carcinogenic to humans: The
67     evidence demonstrates a potential hazard
68     to humans but does not meet the criteria
69     for carcinogenic. There may be a plausible
70     association in humans,  multiple positive
71     results in animals, or a combination of
72     human,  animal,  or  other  experimental
73     evidence.
74  Suggestive   evidence   of   carcinogenic
75     potential: The evidence raises concern for
76     effects in humans but is not sufficient for a
77     stronger  conclusion.   This   descriptor
78     covers a range of evidence, from a positive
79     result in the only available study to a single
80     positive  result in an  extensive database
81     that includes negative  results in other
82     species.
83  Inadequate    information   to    assess
84     carcinogenic  potential:    No   other
85     descriptors apply. Conflicting evidence can
86     be classified  as inadequate  information if
87     all  positive  results  are  opposed  by
88     negative  studies  of equal quality in the
89     same sex and strain.  Differing  results,
90     however, can be  classified as suggestive
91     evidence or as likely to be carcinogenic.
92  Not likely to be carcinogenic to humans:
93     There is robust evidence for concluding
94     that there is  no basis for concern. There
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review oftert-Butyl Alcohol
 1     may be no effects in both sexes of at least
 2     two  appropriate animal species; positive
 3     animal  results  and  strong,  consistent
 4     evidence that each  mode of action in
 5     animals does not operate in humans; or
 6     convincing evidence that effects are not
 7     likely by a particular exposure route or
 8     below a defined dose.
 9     Multiple descriptors may be used if there
10  is evidence that carcinogenic effects differ by
11  dose range or exposure route [U.S. EPA. 2005a.
12  §2,5).
13     Another example of standard descriptors
14  comes  from the  EPA's  Integrated  Science
15  Assessments, which evaluate causation for the
16  effects of the criteria pollutants in ambient air
17  fU.S. EPA. 2010. 31.61.
18  Causal relationship:  Sufficient evidence to
19     conclude   that   there   is   a   causal
20     relationship. Observational studies cannot
21     be explained by plausible alternatives, or
22     they  are  supported  by other  lines of
23     evidence, for example, animal studies or
24     mechanistic information.
25  Likely to be a causal relationship: Sufficient
26     evidence that a causal relationship is
27     likely, but important uncertainties remain.
28     For example, observational studies show
29     an  association  but  co-exposures  are
30     difficult to  address  or  other  lines of
31     evidence are limited  or  inconsistent; or
32     multiple  animal  studies from different
33     laboratories  demonstrate   effects  and
34     there are limited or no human data.
35  Suggestive of a causal relationship: At least
36     one  high-quality  epidemiologic  study
37     shows an association but other studies are
38     inconsistent
39  Inadequate to infer a causal relationship:
40     The  studies do not permit a  conclusion
41     regarding the presence or absence of an
42     association.

43  Not likely to be a causal relationship: Several
44     adequate studies, covering the full range of
45     human   exposure   and   considering
46     susceptible  populations,  are  mutually
47     consistent in not showing an effect at any
48     level of exposure.
49     The EPA is investigating and may on a trial
50  basis use these or other standard descriptors
51  to  characterize the  overall weight of the
52  evidence for effects other than cancer.

53  6.  Selecting studies for derivation of
54     toxicity values

55     For each effect where there is credible
56  evidence of an association with the agent, the
57  assessment derives toxicity values if there are
58  suitable  epidemiologic or experimental data.
59  The decision to  derive toxicity values may be
60  linked to the hazard descriptor.
61     Dose-response analysis requires quantitative
62  measures  of dose  and  response. Then, other
63  factors being equal:
64     -  Epidemiologic  studies are  preferred
65        over animal studies, if quantitative
66        measures  of exposure  are  available
67        and effects can be attributed to the
68        agent.
69     -  Among experimental animal  models,
70        those that respond most like humans
71        are preferred, if the comparability of
72        response can be determined.
73     -  Studies  by  a  route   of  human
74        environmental     exposure     are
75        preferred,   although  a   validated
76        toxicokinetic model  can be used to
77        extrapolate across exposure  routes.
78     -  Studies  of longer exposure duration
79        and  follow-up  are  preferred,  to
80        minimize uncertainty about whether
81        effects are  representative of  lifetime
82        exposure.
83     -  Studies with multiple exposure levels
84        are preferred  for  their  ability  to
85        provide  information about the shape
86        of the exposure-response curve.
87     -  Studies with adequate power to detect
88        effects at lower exposure levels are
89        preferred,  to minimize  the  extent of
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxiii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
           extrapolation to  levels found in the
           environment
 3     Studies with  non-monotonic  exposure-
 4  response relationships are  not necessarily
 5  excluded from  the analysis. A diminished
 6  effect at higher  exposure  levels may  be
 7  satisfactorily  explained by factors such as
 8  competing toxicity, saturation of absorption or
 9  metabolism, exposure  misclassification,  or
10  selection bias.
11     If a large number of studies are suitable for
12  dose-response  analysis,   the   assessment
13  considers  the  study characteristics  in  this
14  section to focus on the most informative data.
15  The assessment explains the reasons  for not
16  analyzing other groups of studies. As a check
17  on the selection of studies for dose-response
18  analysis,  the  EPA asks peer  reviewers to
19  identify studies  that  were  not adequately
20  considered.

21  7. Deriving toxicity values

22  7.1.  General framework for  dose-
23       response analysis

24     The  EPA uses a two-step approach that
25  distinguishes analysis of the observed dose-
26  response data from inferences  about lower
27  doses fU.S. EPA. 2005a. §31
28     Within the observed range, the preferred
29  approach is to use modeling to incorporate a
30  wide range of data into  the analysis.  The
31  modeling yields   a  point  of departure  (an
32  exposure level near the  lower end  of the
33  observed   range,    without    significant
34  extrapolation  to lower doses) (Sections  7.2-
35  7.3).
36     Extrapolation  to lower doses considers
37  what is known about the modes of action for
38  each  effect (Sections 7.4-7.5).  If  response
39  estimates at lower doses are not required, an
40  alternative is to derive reference  values, which
41  are calculated by applying factors to the point
42  of departure in order to account for sources of
43  uncertainty and variability (Section 7.6).
44     For a group of agents that induce an effect
45  through a common mode  of action, the dose-
46  response  analysis  may  derive  a  relative
47  potency factor for each agent  A full dose-
48  response analysis is conducted for one well-
49  studied index chemical in the group, then the
50  potencies of other members are expressed in
51  relative terms based on relative toxic effects,
52  relative  absorption   or  metabolic  rates,
53  quantitative structure-activity relationships,
54  or receptor binding characteristics (U.S. EPA.
55  2005a. §3.2.6: 2000b. §4.4).
56     Increasingly,  the  EPA  is basing toxicity
57  values on combined analyses of multiple data
58  sets  or multiple  responses.  The EPA also
59  considers multiple dose-response approaches
60  if they can be supported by robust data.

61  7.2.  Modeling dose to sites of biologic
62       effects

63     The preferred  approach for analysis of
64  dose is toxicokinetic modeling because of its
65  ability to incorporate a wide range of data. The
66  preferred dose metric would refer to  the
67  active agent at the site of its biologic effect or
68  to a  close,  reliable surrogate measure. The
69  active agent may be the administered chemical
70  or a  metabolite.  Confidence in the use of a
71  toxicokinetic   model   depends   on  the
72  robustness of its validation process and on the
73  results of  sensitivity  analyses   (U.S.  EPA.
74  2006a: 2005a. §3.1: 1994. §4.3).
75     Because  toxicokinetic  modeling  can
76  require many parameters and more data than
77  are typically available, the EPA has developed
78  standard approaches that  can be applied to
79  typical data sets. These standard approaches
80  also  facilitate comparison  across exposure
81  patterns and species.
82     -   Intermittent  study  exposures  are
83         standardized to a daily average over
84         the duration of exposure. For chronic
85         effects, daily exposures are averaged
86         over the lifespan. Exposures during a
87         critical  period,  however,  are  not
88         averaged over a longer duration (U.S.
89         EPA. 2005a. §3.1.1: 1991b.§3.2).
90     -   Doses are standardized to equivalent
91         human terms to facilitate comparison
92         of results from different species.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxiv             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
 1     -   Oral  doses are scaled allometrically
 2         using mg/kg3/4-day as the equivalent
 3         dose metric across species. Allometric
 4         scaling pertains to equivalence across
 5         species,  not across lifestages, and is
 6         not used to scale  doses from  adult
 7         humans  or mature  animals to infants
 8         or      children      [U.S.      EPA.
 9         2011: 2005a. 33.1.3).
10     -   Inhalation exposures are scaled  using
11         dosimetry models that apply species-
12         specific  physiologic   and  anatomic
13         factors  and consider  whether the
14         effect occurs at the site of first contact
15         or after systemic circulation [U.S. EPA.
16         2012a: 1994.331.
17     It can  be informative  to  convert doses
18  across  exposure routes. If this is done, the
19  assessment  describes  the underlying  data,
20  algorithms,  and  assumptions  [U.S.  EPA.
21  2005a. 33.1.41.
22     In the  absence of study-specific  data  on,
23  for example, intake rates or body weight, the
24  EPA has developed recommended values  for
25  use  in  dose-response analysis  [U.S.  EPA.
26  19881

27  7.3.  Modeling response in the range of
28       observation
29     Toxicodynamic   ("biologically  based")
30  modeling  can incorporate data  on biologic
31  processes  leading to an effect. Such models
32  require sufficient data to ascertain a mode of
33  action  and to quantitatively  support model
34  parameters  associated with  its key events.
35  Because   different  models   may  provide
36  equivalent fits  to the  observed data but
37  diverge substantially at lower doses, critical
38  biologic parameters should be  measured from
39  laboratory studies, not  by  model fitting.
40  Confidence in the use of a  toxicodynamic
41  model  depends  on  the  robustness of  its
42  validation  process  and on   the  results   of
43  sensitivity  analyses.  Peer review  of the
44  scientific basis and performance of a model is
45  essential [U.S. EPA. 2005a. §3.2.2).
46     Because  toxicodynamic   modeling  can
47  require    many   parameters   and   more
48  knowledge   and  data  than  are  typically
49  available, the EPA has developed a standard
50  set  of  empirical ("curve-fitting")   models
51  (http://www.epa.gov/ncea/bmds/) that can
52  be applied to typical data sets, including those
53  that are nonlinear. The EPA has also developed
54  guidance on  modeling  dose-response  data,
55  assessing model fit, selecting  suitable models,
56  and reporting modeling results  (U.S.  EPA.
57  2012b).  Additional judgment or alternative
58  analyses are used if the procedure fails to yield
59  reliable results, for example,  if the fit is poor,
60  modeling may be restricted to the lower doses,
61  especially if there is competing toxicity at
62  higher doses (U.S.  EPA. 2005a. §3.2.3).
63     Modeling is used to derive a point of
64  departure fU.S. EPA. 2012b: 2005a. 33.2.41.
65  (See Section 7.6 for alternatives if a point of
66  departure cannot be derived by modeling.):
67     -  If   linear   extrapolation  is  used,
68        selection   of  a  response   level
69        corresponding   to   the  point  of
70        departure  is not highly influential, so
71        standard values  near the low end of
72        the  observable range  are generally
73        used (for example, 10% extra risk for
74        cancer   bioassay  data,   1%   for
75        epidemiologic data,  lower for rare
76        cancers).
77     -  For   nonlinear  approaches,   both
78        statistical and biologic considerations
79        are taken into account.
80     -  For dichotomous data, a response level
81        of 10% extra risk is generally used for
82        minimally  adverse   effects,  5%  or
83        lower for more severe effects.
84     -  For  continuous data,  a response level
85        is ideally  based on  an established
86        definition  of  biologic significance. In
87        the  absence  of such definition,  one
88        control standard deviation from the
89        control  mean  is  often  used   for
90        minimally  adverse effects,  one-half
91        standard  deviation for more severe
92        effects.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxv             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
 1     The point of departure is the 95% lower
 2  bound  on  the dose associated with  the
 3  selected response level.

 4  7.4.  Extrapolating to lower doses and
 5       response levels

 6     The purpose of  extrapolating  to lower
 7  doses is to estimate  responses at exposures
 8  below   the   observed   data.   Low-dose
 9  extrapolation, typically used for cancer data,
10  considers  what is known  about modes of
11  action [U.S. EPA. 2005a. 33.3.1 and 33.3.21.
12  1)  If a biologically  based  model has been
13     developed  and validated  for the agent,
14     extrapolation may use  the fitted model
15     below  the  observed range if significant
16     model  uncertainty can be  ruled out with
17     reasonable confidence.
18   2)  Linear extrapolation is  used if the dose-
19      response  curve  is expected to  have a
20      linear  component below the point of
21      departure. This includes:
22     -  Agents  or  their metabolites that are
23        DNA-reactive    and   have    direct
24        mutagenic activity.
25     -  Agents  or their metabolites for which
26        human exposures or  body burdens are
27        near doses associated with key events
28        leading to an effect
29     Linear extrapolation is  also  used when
30     data are insufficient to establish mode of
31     action and when scientifically plausible.
32     The  result of linear  extrapolation is
33     described by an  oral slope  factor or an
34     inhalation unit risk, which is the slope of
35     the dose-response curve at lower doses or
36     concentrations, respectively.
37  3)  Nonlinear   models    are   used   for
38     extrapolation if there are sufficient data to
39     ascertain  the  mode  of action  and to
40     conclude that it  is not linear at lower
41     doses, and the agent does not demonstrate
42     mutagenic  or other  activity  consistent
43     with linearity  at  lower  doses. Nonlinear
44     approaches generally should not be used
45     in cases where mode of action has  not
46     ascertained. If nonlinear extrapolation is
47     appropriate but no model is developed, an
48     alternative is to calculate reference values.
49  4)  Both linear and nonlinear approaches may
50     be used if there a multiple modes of action.
51     For example, modeling to a low response
52     level  can be  useful for estimating the
53     response at doses where a high-dose mode
54     of action would be less important.
55     If linear  extrapolation  is  used,  the
56  assessment develops a candidate slope factor
57  or unit risk for each suitable data set. These
58  results are  arrayed,  using common  dose
59  metrics, to show the distribution of relative
60  potency   across   various    effects    and
61  experimental systems. The assessment then
62  derives or selects an overall  slope factor and
63  an overall unit risk for the  agent, considering
64  the various dose-response analyses, the study
65  preferences discussed in Section 6,  and the
66  possibility of basing a more  robust result on
67  multiple data sets.

68  7.5.  Considering susceptible
69       populations and lifestages

70     The  assessment analyzes  the  available
71  information on populations and lifestages that
72  may be particularly susceptible to each effect.
73  A  tiered approach  is   used  [U.S.  EPA.
74  2005a. 33.51.
75  1)  If an epidemiologic or experimental study
76     reports   quantitative   results   for   a
77     susceptible population  or lifestage, these
78     data  are  analyzed  to  derive  separate
79     toxicity values for susceptible individuals.
80  2)  If data on risk-related  parameters allow
81     comparison of the general population and
82     susceptible  individuals,  these  data are
83     used  to adjust the general-population
84     toxicity   values   for    application  to
85     susceptible individuals.
86  3)  In the absence of chemical-specific data,
87     the EPA  has developed age-dependent
88     adjustment factors for early-life exposure
89     to potential  carcinogens  that  have  a
90     mutagenic mode  of  action.  There  is
91     evidence  of early-life  susceptibility to
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxvi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol
 1     various carcinogenic agents, but  most
 2     epidemiologic   studies    and   cancer
 3     bioassays  do   not  include  early-life
 4     exposure. To  address the potential  for
 5     early-life    susceptibility,    the   EPA
 6     recommends fU.S. EPA. 2005b. 551
 7     -   10-fold  adjustment  for  exposures
 8         before age 2 years.
 9     -   3-fold  adjustment   for   exposures
10         between ages 2 and 16 years.

11  7.6.  Reference values and uncertainty
12       factors
13     An oral reference  dose  or  an inhalation
14  reference concentration is an estimate  of an
15  exposure (including in susceptible subgroups)
16  that is likely to be without an appreciable risk
17  of adverse health effects over a lifetime [U.S.
18  EPA.  2002. §4.2].   Reference   values  are
19  typically  calculated for effects  other  than
20  cancer and for suspected carcinogens if a well
21  characterized mode  of action indicates that a
22  necessary key  event does not occur below a
23  specific  dose.  Reference values  provide  no
24  information about risks at higher exposure
25  levels.
26     The assessment  characterizes effects that
27  form the basis for reference values as adverse,
28  considered to be adverse, or a precursor to an
29  adverse effect  For developmental toxicity,
30  reproductive toxicity, and neurotoxicity there
31  is  guidance on  adverse effects and  their
32  biologic      markers       [U.S.	EPA,
33  1998b. 1996.1991b).
34     To account for uncertainty and variability
35  in the derivation of a lifetime human exposure
36  where adverse effects are not  anticipated to
37  occur,  reference  values are  calculated  by
38  applying a series of uncertainty factors to the
39  point of  departure.  If a point of departure
40  cannot  be  derived by modeling,  a no-
41  observed-adverse-effect  level  or a  lowest-
42  observed-adverse-effect level is used instead.
43  The    assessment    discusses    scientific
44  considerations  involving several  areas   of
45  variability or uncertainty.
46  Human variation. The assessment accounts
47     for variation in  susceptibility across the
48     human population and the possibility that
49     the   available   data   may   not   be
50     representative  of individuals  who  are
51     most susceptible to the effect A factor of
52     10 is generally used  to account for this
53     variation. This factor is reduced only if the
54     point of departure is derived or adjusted
55     specifically  for  susceptible  individuals
56     (not for a general population that includes
57     both  susceptible  and  non-susceptible
58     individuals)          (U.S.	EPA.
59     2002.54.4.5: 1998b.54.2: 1996.54: 1994.
60     54.3.9.1: 1991b. 53.4).
61  Animal-to-human extrapolation.  If animal
62     results are used to make inferences about
63     humans, the assessment adjusts for cross-
64     species differences. These may arise from
65     differences    in     toxicokinetics    or
66     toxicodynamics. Accordingly,  if the point
67     of departure is standardized to equivalent
68     human terms or is based on toxicokinetic
69     or dosimetry modeling, a factor of 101/2
70     (rounded to 3) is applied to account for the
71     remaining     uncertainty     involving
72     toxicokinetic     and     toxicodynamic
73     differences.  If a biologically based model
74     adjusts   fully  for   toxicokinetic   and
75     toxicodynamic differences across species,
76     this factor is not used. In most other cases,
77     a  factor  of 10  is  applied  (U.S.  EPA.
78     2011: 2002.54.4.5: 1998b. 54.2: 1996.54:
79      1994.54.3.9.1: 1991b. 53.41
80  Adverse-effect  level   to   no-observed-
81     adverse-effect  level.  If a   point of
82     departure is based on a lowest-observed-
83     adverse-effect level, the assessment must
84     infer a dose where such  effects are not
85     expected. This can  be a  matter of great
86     uncertainty,  especially  if there  is  no
87     evidence available at lower doses. A factor
88     of 10 is  applied  to account for  the
89     uncertainty in making this inference. A
90     factor  other  than  10  may  be used,
91     depending on the magnitude and nature of
92     the response and the shape of the dose-
93     response      curve      (U.S.	EPA.
94     2002.54.4.5: 1998b.54.2: 1996.54: 1994.
95     54.3.9.1: 1991b. 53.41
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxvii            DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
 1  Subchronic-to-chronic exposure. If a point
 2     of departure  is  based  on  subchronic
 3     studies, the assessment considers whether
 4     lifetime exposure  could  have effects at
 5     lower levels of exposure. A factor of 10 is
 6     applied to account for the uncertainty in
 7     using  subchronic   studies   to  make
 8     inferences  about lifetime  exposure. This
 9     factor   may   also   be   applied    for
10     developmental or reproductive effects if
11     exposure covered less than the full critical
12     period. A factor other than 10 may be used,
13     depending  on  the duration of the studies
14     and the nature of the  response [U.S. EPA,
15     2002.54.4.5: 1998b.54.2: 1994.34.3.9.11.
16  Incomplete  database.   If  an  incomplete
17     database  raises  concern  that further
18     studies might identify a more sensitive
19     effect,  organ  system, or lifestage,  the
20     assessment  may   apply  a   database
21     uncertainty    factor    [U.S.     EPA.
22     2002.54.4.5: 1998b.54.2: 1996.54: 1994.
23     54.3.9.1:  1991b. 53.4). The  size of the
24     factor  depends on  the   nature  of the
25     database deficiency. For example, the EPA
26     typically  follows  the suggestion that a
27     factor of 10 be applied if both a prenatal
28     toxicity  study and  a  two-generation
29     reproduction  study  are  missing and a
30     factor of 101/2 if  either is missing  (U.S.
31     EPA. 2002.  §4.4.5).

32     In this  way,  the  assessment derives
33  candidate values for each suitable data set and
34  effect that is  credibly associated  with the
35  agent These  results  are   arrayed,  using
36  common dose metrics, to  show where effects
37  occur across a range of exposures [U.S. EPA.
38  1994.54.3.91
39     The  assessment  derives  or  selects  an
40  organ- or system-specific  reference value  for
41  each organ or  system affected by the agent.
42  The assessment explains the rationale for each
43  organ/system-specific reference value (based
44  on, for example, the highest quality studies,
45  the most sensitive outcome, or a clustering of
46  values).  By providing these  organ/system-
47  specific reference values, IRIS assessments
48  facilitate   subsequent   cumulative    risk
49  assessments that consider the combined effect
50  of multiple agents acting at a common site or
51  through common mechanisms [NRG, 2009].
52     The assessment then selects an  overall
53  reference  dose  and  an overall reference
54  concentration  for  the agent  to represent
55  lifetime human exposure levels where effects
56  are not anticipated to occur. This is generally
57  the  most  sensitive  organ/system-specific
58  reference value, though consideration of study
59  quality and confidence in each value may lead
60  to a different selection.

61  7.7. Confidence and uncertainty in the
62      reference values
63     The   assessment  selects   a  standard
64  descriptor  to  characterize  the  level  of
65  confidence in each reference value, based on
66  the likelihood that the value would  change
67  with further testing. Confidence in reference
68  values is based on quality of the studies used
69  and completeness of the database, with more
70  weight given  to the  latter.  The level  of
71  confidence is increased for  reference values
72  based on human data supported by animal
73  data fU.S. EPA. 1994. 54.3.9.21
74  High confidence: The reference value is not
75     likely to  change  with   further  testing,
76     except for mechanistic studies that might
77     affect the  interpretation of prior test
78     results.
79  Medium  confidence: This  is a matter of
80     judgment,  between   high  and  low
81     confidence.
82  Low  confidence:  The  reference  value is
83     especially vulnerable  to  change  with
84     further testing.
85     These   criteria  are  consistent   with
86  guidelines  for   systematic  reviews  that
87  evaluate the quality of evidence. These also
88  focus on whether further research would be
89  likely to change confidence in the estimate of
90  effect fGuyatt et al.. 2008bl
91     All  assessments discuss the significant
92  uncertainties encountered in the analysis. The
93  EPA provides guidance on characterization of
94  uncertainty  [U.S.  EPA.  2005a. 53.6].  For
95  example,  the discussion distinguishes  model
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              xxviii           DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review of tert-Butyl Alcohol


1  uncertainty (lack of knowledge aboutthe most    7  susceptibility or in exposures that modify the
2  appropriate experimental or analytic model)    8  effects of the agent).
3  and parameter uncertainty (lack of knowledge    9
4  about  the   parameters   of  a   model).   10
5  Assessments  also discuss human  variation   11  August2013
6  (interpersonal   differences   in  biologic
             This document is a draft for review purposes only and does not constitute Agency policy.
                                             xxix            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol
     EXECUTIVE SUMMARY
 3                           Summation of Occurrence and Health Effects
 4                 tert-Butanol does not occur naturally; it is produced by humans for multiple
 5          purposes, such as a solvent for paints, a denaturant for ethanol and several other alcohols, a
 6          dehydrating agent, and in the manufacture of flotation agents, fruit essences, and perfumes.
 7          tert-Butanol also is a primary metabolite of methyl tert-butyl ether (MTBE) and ethyl tert-
 8          butyl ether (ETBE). Exposure to tert-butanol primarily occurs through breathing air
 9          containing tert-butanol vapors and consuming contaminated water or foods. Exposure can
10          also occur through direct skin contact.
11                 Animal studies demonstrate that chronic oral exposure to tert-butanol is associated
12          with kidney and thyroid effects. Developmental effects (e.g., reduced fetal viability) have
13          been observed in short-term exposure to high levels of tert-butanol (via oral or inhalation
14          exposure) in animals. Neurodevelopmental effects also have been observed, but results
15          were inconsistent. No chronic inhalation exposure studies have been conducted. There is
16          suggestive evidence that tert-butanol is carcinogenic to humans based on renal tumors in
17          male rats and thyroid tumors in female mice.

18   Effects Other Than Cancer Observed Following Oral Exposure
19          Kidney effects are  a potential human hazard of oral exposure to tert-butanol. Kidney toxicity
20   was observed in males and females in two strains of rats. Kidney weights were increased in male
21   and female rats after 13 weeks or 15 months of treatment. Histopathological examination in male
22   and female rats observed increased incidence or severity of nephropathy after 13 weeks of oral
23   exposure, increased severity of nephropathy after a 2-year oral exposure, and increased
24   transitional epithelial hyperplasia after 2 years of oral exposure. Additionally, increased
25   suppurative inflammation was noted in females after 2 years of oral exposure. In one strain of mice,
26   the only kidney effect observed was an increase in kidney weight (absolute or relative) in female
27   mice after 13 weeks, but no treatment-related histopathological lesions were reported in the
28   kidneys of male or female  mice at 13 weeks or 2 years. A mode of action (MOA) analysis determined
29   that tert-butanol exposure induces a male rat-specific (X2u-globulin-associated nephropathy. tert-
30   Butanol, however, is a weak inducer of (X2u-globulin-nephropathy, and is not the sole process
31   contributing to renal tubule nephropathy. Chronic progressive nephropathy (CPN) may also be
32   involved in some of the noncancer effects, but the evidence is inconclusive. Endpoints specifically
33   related to either a2u-globulin-nephropathy or CPN were not considered for kidney hazard
34   identification. Changes in kidney weights, transitional epithelial hyperplasia, suppurative
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               ES-1            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
 1
 2
 3
 4
 5
 6
 7

 8
 9
10
11
12
13
14
15
16
17
18
19
20

21
inflammation, and severity and incidence of nephropathy, however, are considered to result from
tert-butanol exposure and are appropriate for identifying a hazard to the kidney.
       There is suggestive evidence of developmental toxicity following tert-butanol exposure.
Developmental effects include increased fetal loss, decreased fetal body weight, and increased
skeletal variations. At this time, no conclusions were drawn in regard to reproductive system
toxicity. There is inadequate information at this time to draw conclusions regarding
neurodevelopmental toxicity, liver, and urinary bladder toxicity.

Oral Reference Dose (RfD) for Effects Other Than Cancer
       Kidney toxicity, represented by kidney transitional epithelial hyperplasia, was chosen as the
basis for the overall oral reference dose (RfD)  (see Table ES-1). The chronic study by NTP T19951
and the observed kidney effects were used to derive the RfD. The endpoint of transitional epithelial
hyperplasia was selected as the critical effect because it was observed in both rat sexes
consistently, it is a specific and sensitive  indicator of kidney toxicity, and was induced in a dose-
responsive manner. Benchmark dose (BMD) modeling was used to derive the benchmark dose
lower confidence limit (BMDLio%) of 16 mg/kg-day. The BMDL was converted to a human
equivalent dose using body weight3/4 scaling, and the value of 3.84 mg/kg-day was used as the
point of departure (POD) for RfD derivation (U.S. EPA. 2011).
       The overall RfD was calculated by dividing the POD for kidney transitional epithelial
hyperplasia by a composite uncertainty factor (UF) of 30 to account for the extrapolation from
animals to humans (3) and for interindividual differences in human susceptibility (10).

       Table ES-1. Organ/system-specific RfDs and overall RfD for tert-butanol


Hazard
Kidney
Overall RfD


Basis
Transitional epithelial
hyperplasia
Kidney
Point of
departure*
(mg/kg-day)
3.8
3.8


UF
30
30

Chronic RfD
(mg/kg-day)
1 x 10'1
1 x ID'1
Study
exposure
description
Chronic
Chronic


Confidence
High
High
22   *HED PODs were calculated using BW3/4scaling (U.S. EPA, 2011).
23   Effects Other Than Cancer Observed Following Inhalation Exposure
24          Kidney effects are a potential human hazard of inhalation exposure to tert-butanol.
25   Although no effects were observed in mice, kidney weights were increased in male and female rats
26   following 13 weeks of inhalation exposure. In addition, nephropathy severity increased in male
27   rats. No human studies are available to evaluate the effects of inhalation exposure. As discussed
28   above for oral effects, endpoints specifically related to either (X2uglobulin nephropathy or CPN were

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                ES-2            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
 1
 2
 3

 4
 5
 6
 7
 8
 9
10
11
12
13
14

15
not considered for kidney hazard identification. Changes in kidney weights and severity of
nephropathy, however, are considered a result of tert-butanol exposure and are appropriate for
identifying a hazard to the kidney.
Inhalation Reference Concentration (RfC) for Effects Other Than Cancer
       Kidney toxicity, represented by transitional epithelial hyperplasia, was chosen as the basis
for the inhalation reference concentration (RfC) (see Table ES-2). Although endpoints from a route-
specific study were considered, the availability of a physiologically based pharmacokinetic (PBPK)
model for tert-butanol in rats [modified by Salazar etal. (2015]] allowed for more specific and
sensitive equivalent inhalation PODs derived from a route-to-route extrapolation from the PODs of
the  oral NTP (1995) study.  The POD adjusted for the human equivalent concentration (HEC) was
26.1 mg/m3 based on transitional epithelial hyperplasia.
       The RfC was calculated by dividing the POD by a composite UF of 30 to account for
toxicodynamic differences between animals and humans (3) and interindividual differences in
human susceptibility (10).
       Table ES-2. Organ/system-specific RfCs and overall RfC for tert-butanol


Hazard
Kidney
Overall RfC


Basis
Transitional epithelial
hyperplasia
Kidney
Point of
departure*
(mg/m3)
26.1
26.1


UF
30
30

Chronic RfC
(mg/m3)
9 x 10'1
9 x ID'1

Study exposure
description
Chronic
Chronic


Confidence
High
High
16
17

18
19
20
21
22
23
*Continuous inhalation human equivalent concentration that leads to the same average blood concentration of
 tert-butanol as continuous oral exposure at the BMDL
Evidence of Human Carcinogenicity
       Under the EPA's cancer guidelines (U.S. EPA, 2005a], there is suggestive evidence of
carcinogenic potential for tert-butanol. tert-Butanol induced kidney tumors in male (but not female)
rats and thyroid tumors (primarily benign) in male and female mice following long-term
administration in drinking water (NTP, 1995). The potential for carcinogenicity applies to all routes
of human exposure.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                ES-3            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   Quantitative Estimate of Carcinogenic Risk from Oral Exposure
 2          A quantitative estimate of carcinogenic potential from oral exposure to tert-butanol was
 3   based on the increased incidence of thyroid follicular cell adenomas in female B6C3Fi mice, and
 4   thyroid follicular cell adenomas and carcinomas in male B6C3Fi mice [NTP, 1995]. The study
 5   included histological examinations for tumors in many different tissues, contained three exposure
 6   levels and controls, contained adequate numbers of animals per dose group (~50/sex/group),
 7   treated animals for up to 2 years, and included detailed reporting of methods and results.
 8          Although tert-butanol was considered to have "suggestive evidence of carcinogenic
 9   potential," the NTP study was well conducted and quantitative analysis could be useful for
10   providing a sense of the magnitude of potential carcinogenic risk [U.S. EPA. 2005a). A slope factor
11   was derived for thyroid tumors in female and male mice. The modeled tert-butanol POD was scaled
12   to HEDs according to EPA guidance by converting the BMDLio on the basis of (body weight)3/4
13   scaling [U.S. EPA. 2011. 2005a). Using linear extrapolation from the BMDLio, a human equivalent
14   oral slope factor was derived (slope factor = 0.1/BMDLio). The resulting oral slope factor is 5 x 1Q-*
15   per mg/kg-day.

16   Quantitative Estimate of Carcinogenic Risk from Inhalation Exposure
17          No chronic inhalation exposure studies to tert-butanol are available. Lifetime exposure to
18   tert-butanol has been associated with increased renal tubule adenomas and carcinomas as well as
19   thyroid follicular cell adenomas and carcinomas. As stated above, the rat kidney tumors are
20   unsuitable for quantitative analysis as there is not enough data to determine the relative
21   contribution of azu-globulin nephropathy and other processes to the overall kidney tumor
22   response. Although the mouse thyroid tumors served as the basis for the oral slope factor, route-to-
23   route extrapolation is not possible for these thyroid effects in  mice because the only PBPK model
24   available is for rats. Therefore, no quantitative estimate of carcinogenic risk could be determined
25   for inhalation exposure.
26   Susceptible Populations and Lifestages for Cancer and Noncancer
27          In vitro studies suggest that cytochrome P-450 (CYP450) (Cederbaum etal., 1983:
28   Cederbaum and Cohen, 1980], plays a role in the metabolism of tert-butanol. No studies, however,
29   have identified the specific CYPs responsible for the biotransformation of tert-butanol. Various
30   CYPs are under-expressed in the mouse fetus and neonate (Lee etal.. 2011] and decreased in older
31   mice (Lee etal., 2011] and rats (Lee etal., 2008). Decreased ability to detoxify and transport tert-
32   butanol out of the body could result in increased susceptibility to tert-butanol.
33          With regard to cancer, differences in lifestage sensitivity to chemically induced thyroid
34   carcinogenesis are unknown (U.S. EPA. 1998a]. An increased incidence of thyroid tumors was
35   identified in mice after tert-butanol exposure, and human studies have demonstrated that children
36   are more sensitive than adults are to thyroid carcinogenesis resulting from ionizing radiation.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                ES-4            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   Collectively, there is little evidence on tert-butanol itself to identify any populations or lifestages
 2   that may be especially susceptible.

 3   Key Issues Addressed in Assessment
 4          An evaluation of whether tert-butanol caused (X2u-globulin-associated nephropathy was
 5   performed. The presence of (X2u-globulin in the hyaline droplets was confirmed in male rats by
 6   (X2u-globulin immunohistochemical staining. Linear mineralization and tubular hyperplasia were
 7   reported in male rats, although only in the chronic study. Other subsequent steps in the
 8   pathological sequence, including necrosis, exfoliation, and granular casts, either were absent or
 9   inconsistently observed across subchronic or chronic studies. None of these effects occurred in
10   female rats or in either sex of mice, although these endpoints were less frequently evaluated in
11   these models. Evidence implies an a2u-globulin MOA is operative, although it is relatively weak in
12   response to tert-butanol and is not solely responsible for the renal tubule nephropathy observed in
13   male rats. CPN also is instrumental in renal tubule nephropathy, in both male and female rats.
14   Several other effects in the kidney unrelated to (X2u-globulin or CPN, however, were observed in
15   female or male rats [U.S. EPA. 1991a). including suppurative inflammation in female rats,
16   transitional epithelial hyperplasia in male and female rats, and increased kidney weights in both
17   sexes of rats [NTP. 1997.1995). These specific effects are considered the result of tert-butanol
18   exposure and therefore, relevant to humans.
19          Concerning cancer, (X2u-globulin accumulation is indicated as relatively weak in response to
20   tert-butanol exposure and not the sole mechanism responsible for the renal tubule carcinogenicity
21   observed in male rats. Although CPN and other effects induced by both (X2u-globulin processes and
22   tert-butanol play a role in renal tubule nephropathy, the evidence indicates that CPN does not
23   induce the renal tubule tumors associated with tert-butanol exposure in male rats, suggesting that
24   other, unknown processes contribute to renal tumor development Based on this analysis of
25   available MOA data, these renal tumors are considered relevant to humans [U.S. EPA. 1991a).
26          In addition, an increase in the incidence of thyroid follicular cell adenomas was observed in
27   male and female mice in a 2-year drinking water study [NTP. 1995]. Thyroid follicular cell
28   hyperplasia was considered a preneoplastic effect associated with the thyroid tumors, and the
29   incidences of follicular cell hyperplasias were elevated in both male and  female B6C3Fi mice
30   following exposure. U.S. EPA [1998a] describes the procedures the Agency uses in evaluating
31   potential human cancer hazard and dose-response assessments for chemicals that are animal
32   thyroid carcinogens. The available database is inadequate in four of the five required areas [U.S.
33   EPA. 1998a]. suggesting that an antithyroid MOA is not operating in mouse thyroid follicular cell
34   tumorigenesis. No other MOAs for thyroid tumors were identified, and the  mouse thyroid tumors
35   are considered relevant to humans [U.S. EPA. 2005a. 1998a].
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                ES-5            DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
     LITERATURE  SEARCH STRATEGY  |  STUDY
     SELECTION AND EVALUATION
 4          A literature search and screening strategy were used to identify literature characterizing
 5   the health effects of tert-butanol. This strategy consisted of a broad search of online scientific
 6   databases and other sources to identify all potentially pertinent studies. In subsequent steps,
 7   references were screened to exclude papers not pertinent to an assessment of the health effects of
 8   tert-butanol, and remaining references were sorted into categories for further evaluation. This
 9   section describes the literature search and screening strategy in detail.
10          The chemical-specific search was conducted in four online scientific databases, including
11   PubMed, Web of Science, Toxline, and TSCATS through May 2015, using the keywords and limits
12   described in Table LS-1. The overall literature search approach is shown graphically in Figure LS-1.
13   An additional seven citations were obtained using additional search strategies described in Table
14   LS-2. After electronically eliminating duplicates from the citations retrieved through these
15   databases, 2,648 unique citations were identified.
16          The resulting 2,648 citations were screened for pertinence and separated into categories as
17   presented in Figure LS-1 using the title and either abstract or full text, or both, to examine the
18   health effects of tert-butanol exposure. The inclusion and exclusion criteria used to screen the
19   references and identify sources of health effects data are provided in Table LS-3.

20      •   12 references were identified as "Sources of Health Effects Data" and were considered for
21          data extraction to evidence tables and exposure-response arrays.

22      •   200 references were identified as "Sources of Mechanistic and Toxicokinetic Data" and
23          "Sources of Supplementary Health Effects Data"; these included 39 studies describing
24          physiologically based pharmacokinetic (PBPK) models and other toxicokinetic information,
25          73 studies providing genotoxicity and other mechanistic information, 1 human case report,
26          74 irrelevant exposure paradigms (including acute, dermal, eye irritation, and injection
27          studies), 6 preliminary toxicity studies, and 7 physical dependency studies. Information
28          from these studies was not extracted into evidence tables; however, these studies were
29          considered as support for assessing tert-butanol health effects, for example, evaluation of
30          mode of action and extrapolation of experimental animal findings to humans. Additionally,
31          although still considered sources of health effects information, studies investigating the
32          effects of acute and direct chemical exposures are generally less pertinent for characterizing
33          health hazards associated with chronic oral and inhalation exposure. Therefore, information
34          from these studies was not considered for extraction into evidence tables. Nevertheless,
35          these studies were still evaluated as  possible sources of supplementary health effects
36          information.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              LS-1             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1       •  63 references were identified as "Secondary Literature and Sources of Contextual
 2          Information" (e.g., reviews and other agency assessments); these references were retained
 3          as additional resources for development of the Toxicological Review.

 4       •  2,373 references were identified as not being pertinent (not on topic) to an evaluation of
 5          the health effects of tert-butanol and were excluded from further consideration (see Figure
 6          LS-1 for exclusion categories and Table LS-3 for exclusion criteria). For example, health
 7          effect studies of gasoline and tert-butanol mixtures were not considered pertinent to the
 8          assessment because the separate effects of the gasoline or other chemical components could
 9          not be determined. Retrieving a large number of references that are not on topic is a
10          consequence of applying an initial search strategy designed to cast a wide net and to
11          minimize the possibility of missing potentially relevant health effects data.

12          The complete list of references and the sorting of these materials can be found on the tert-
13   butanol project page of the HERO website at
14   https://hero.epa.gov/index.cfm/project/page/project id/1543.

15   Selection of Studies for Inclusion in Evidence Tables
16          To summarize the important information systematically from the primary health effects
17   studies in the tert-butanol database, evidence tables were constructed in a standardized tabular
18   format as recommended by NRG (2011). Studies were arranged in evidence tables by effect, species,
19   duration, and design, and not by quality. Of the studies that were retained after the literature search
20   and screen, 12 studies were identified as "Sources of Health Effects Data" and were considered for
21   extraction into evidence tables for hazard identification in Chapter 1. Initial review found two
22   references fCirvello etal.. 1995: Lindamoodetal..  19921 to be publications of the NTP T19951 data
23   prior to the release of the final National Toxicology Program (NTP) report One publication
24   (Takahashietal.. 1993) in the "Supplementary Studies" category also was based on data from the
25   NTP report The interim publications and the final NTP report differed. The finalized NTP (1995)
26   report was considered the more complete and accurate presentation of the data; therefore, this
27   report was included in evidence tables and Cirvello et al.  (1995), Takahashietal. (1993), and
28   Lindamoodetal. (1992) were not Data from the remaining 10 references in the "Sources of Health
29   Effects Data" category were extracted into evidence tables.
30          Supplementary studies that contain pertinent information for the toxicological review and
31   augment hazard identification conclusions, such as genotoxic and mechanistic studies, studies
32   describing the kinetics and disposition of tert-butanol absorption and metabolism, pilot studies,
33   and one case report were not included in the evidence tables. Short-term and acute studies
34   (including an 18-day study and a 14-day study by NTP) using oral and inhalation exposures were
35   performed primarily in rats and did not differ qualitatively from the results of the longer studies
36   (i.e., >30-day exposure studies). These were grouped as supplementary studies, however, because
37   the database of chronic and subchronic rodent studies was considered sufficient for evaluating
38   chronic health effects of tert-butanol exposure. Additionally, studies of effects  from chronic
39   exposure are most pertinent to lifetime human exposure (i.e., the primary characterization
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               LS-2              DRAFT—DO NOT CITE OR QUOTE

-------
                                                                   Toxicological Review of tert-Butyl Alcohol
1    provided by IRIS assessments) and are the focus of this assessment. Such supplementary studies
2    may be discussed in the narrative sections of Chapter 1 and are described in sections such as the
3    "Mode of Action Analysis" to augment the discussion or presented in appendices, if they provide
4    additional information.
                                                    Database Searches
                                            (See Table LS-1 for keywords and limits)
                          PubMed
                          n = 1,741
                             Web of Science
                                 n = 489
   Toxline
(incl. TSCATS)
   n = 964
TSCATS 2
  n = 2
                                           (After duplicates removed electronically)
                                                         n=2,641
                        Additional Search Strategies
                    (See Table LS-2 for methods and results)
                                  n = 7
                                                    Combined Dataset
                                                (After all duplicates removed)
                                                        n = 2,648
                                              Manual Screening for Pertinence
                                                 (Title/Abstract/Full Text)
                    Excluded/Not on Topic (n = 2,310)

                    62    Abstract only/comment/society
                          abstracts
                    87    Biodegradation/environmental fate
                    85    Chemical analysis/fuel chemistry
                    1,286  Other chemical/non t-butanol
                    87    Method of detection/exposure and
                          biological monitoring
                    703   Methodology/solvent
                                                    Secondary Literature and Sources of
                                                    Contextual Information (n = 126}

                                                    41    Not relevant species/matrix (e.g.,
                                                          amphibians, fish)
                                                    14    QSAR
                                                    8     Mixtures
                                                    37    Reviews/editorials
                                                    13    Other agency assessments
                                                    13    Book chapter/section
                                                                     Supporting Studies
                     Sources of Health Effects Data (n = 12)
                     0      Human health effects studies
                     12     Animal studies
                                              Sources of Supporting Health Effects Data
                                              {n = 88)

                                              1      Human case reports
                                              74     Not relevant exposure paradigms (e.g.,
                                                    dermal, eye irritation, acute)
                                              6      Preliminary data
                                              7      Physical dependency studies
                                                           Sources of Mechanistic and Toxicokinetic
                                                           Data (n = 112)

                                                           39    PBPK/ADME
                                                           22    Genotoxicity
                                                           51    Other mechanistic studies
5
6
Figure LS-1. Summary of literature search and screening process for
tert-butanol.
                  This document is a draft for review purposes only and does not constitute Agency policy,
                                                        LS-3                DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol
       Table LS-1. Details of the search strategy employed for tert-butanol
    Database
  (Search Date)
           Keywords
                    Limits
PubMed
(12/20/2012)
(4/17/2014)
(5/13/2015)
tert-butanol OR 75-65-0[rn] OR "t-
butyl hydroxide" OR "2-methyl-2-
propanol" OR "trimethyl carbinol"
OR "t-butyl alcohol" OR tert-butanol
OR "tert-butyl alcohol" OR tert-butyl
alcohol[mesh]
None
Web of Science
(12/20/2012)
(4/17/2014)
(5/13/2015)
Topic = (tert-butanol OR 75-65-0 OR
"t-butyl hydroxide" OR "2-methyl-2-
propanol" OR "trimethyl carbinol"
OR "t-butyl alcohol" OR "tert-
butanol" OR "tert-butyl alcohol")
Refined by: Research Areas = (cell biology OR
respiratory system OR microscopy OR biochemistry
molecular biology OR gastroenterology OR hepatology
OR public environmental occupational health OR
oncology OR physiology OR cardiovascular system
cardiology OR toxicology OR life sciences biomedicine
other topics OR hematology OR pathology OR
neurosciences neurology OR developmental biology)
Toxline (includes
TSCATS)
(1/11/2013)
(4/17/2014)
(5/13/2015)
tert-butanol OR 75-65-0 [rn] OR t-
butyl hydroxide OR 2-methyl-2-
propanol OR trimethyl carbinol OR t-
butyl alcohol OR tert-butanol OR
tert-butyl alcohol OR tert-butyl
alcohol
Not PubMed
TSCATS2
(1/4/2013)
(4/17/2014)
(5/13/2015)
75-65-0
None
       Table LS-2. Summary of additional search strategies for tert-butanol
Approach used
Manual search of
citations from
reviews
Manual search of
citations from
reviews conducted
Source(s)
Review article: McGregor (2010).
Tert/o/y-butanol: A toxicological
review. Crit RevToxicol 40(8): 697-
727.
Review article: Chen (2005). Amended
final report of the safety assessment
of t-butyl alcohol as used in
cosmetics. Int J Toxicol 24(2): 1-20.
IPCS (1987a). Butanols: Four isomers:
1-butanol, 2-butanol, tert-butanol,
isobutanol [WHO EHC]. Geneva,
Date
performed
1/2013
1/2013
1/2013
Number of additional references
identified
5
2
None
          This document is a draft for review purposes only and does not constitute Agency policy,
                                             LS-4               DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
Approach used
by other
international and
federal agencies
Source(s)
Switzerland: World Health
Organization.
OSHA (1992). Occupational safety and
health guideline for te/t-butyl alcohol.
Cincinnati, OH: Occupational Safety
and Health Administration.
Date
performed

1/2013
Number of additional references
identified

None
       Table LS-3. Inclusion-exclusion criteria
                       Inclusion criteria
                Exclusion criteria
Population
            •  Humans
            •  Standard mammalian animal models,
              including rat, mouse, rabbit, guinea pig,
              monkey, dog
• Ecological species*
• Nonmammalian species*
Exposure
• Exposure is to te/t-butanol
• Exposure is measured in an
  environmental medium (e.g., air, water,
  diet)
• Exposure via oral, inhalation, or dermal
  routes
  Study population is not exposed to te/t-butanol
  Exposure to a mixture only (e.g., gasoline containing
  te/t-butanol)
  Exposure via injection (e.g., intravenous)
  Exposure pattern less relevant to chronic health
  effects (e.g., acute)
Outcome
              Study includes a measure of one or
              more health effect endpoints, including
              effects on the nervous, musculoskeletal,
              cardiovascular, immune, hematological,
              endocrine, respiratory, urinary, and
              gastrointestinal systems; reproduction;
              development; liver; kidney; eyes; skin;
              and cancer
              Physical dependency studies where
              withdrawal symptoms were evaluated
              after removal of te/t-butanol treatment
Other
Not on topic, including:
 • Abstract only, editorial comments were not
   considered further because study was not
   potentially relevant
 • Bioremediation, biodegradation, or environmental
   fate of te/t-butanol, including evaluation of
   wastewater treatment technologies and methods
   for remediation of contaminated  water and soil
 • Chemical, physical, or fuel chemistry studies
 • Analytical methods  for measuring/detecting/
   remotely sensing te/t-butanol
 • Use of te/t-butanol  as a solvent or methodology for
   testing unrelated to te/t-butanol
           This document is a draft for review purposes only and does not constitute Agency policy,
                                               LS-5               DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review of tert-ButyI Alcohol
                           Inclusion criteria
Exclusion criteria
                                                      •  Not chemical specific: Studies that do not involve
                                                        testing of te/t-butanol
                                                      •  Foreign language studies that were not considered
                                                        further because, based on title or abstract, judged
                                                        not potentially relevant
                                                      •  QSAR studies
      *Studies that met this exclusion criterion were not considered a source of health effects data or supplementary
       health effects data/mechanistic and toxicokinetic data, but were considered as sources of contextual
       information.

 1    Database Evaluation
 2           For this draft assessment, 12 references reported on experimental animal studies that
 3    comprised the primary sources of health effects data; no studies were identified that evaluated
 4    humans exposed to tert-butanol (e.g., cohort studies, ecological studies). The animal studies were
 5    evaluated using the study quality considerations outlined in the Preamble, considering aspects of
 6    design, conduct, or reporting that could affect the interpretation of results, overall contribution to
 7    the synthesis of evidence, and determination of hazard potential as noted in various EPA guidance
 8    documents [U.S. EPA. 2005a. 1998b. 1996.1991b). The objective was to identify the stronger, more
 9    informative studies based on a uniform evaluation of quality characteristics across studies of
10    similar design. As stated in the Preamble, studies were evaluated to identify the suitability of the
11    study based on:

12       •   Study design

13       •   Nature  of the assay and validity for its intended purpose

14       •   Characterization of the nature and extent of impurities and contaminants of tert-butanol
15           administered, if applicable

16       •   Characterization of dose and dosing regimen (including age at exposure) and their
17           adequacy to elicit adverse effects, including latent effects

18       •   Sample sizes and statistical power to detect dose-related differences or trends

19       •   Ascertainment of survival, vital signs, disease or effects, and cause of death

20       •   Control of other variables that could influence the occurrence of effects
21           Additionally, several general considerations, presented in Table LS-4, were used in
22    evaluating the animal studies. Much of the key information for conducting this evaluation can be
23    determined based on study methods and how the study results were reported. Importantly, the
24    evaluation at this stage does not consider the direction or magnitude of any reported effects.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 LS-6             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol

 1          EPA considered statistical tests to evaluate whether the observations might be due to
 2   chance. The standard for determining statistical significance of a response is a trend test or
 3   comparison of outcomes in the exposed groups against those of concurrent controls. Studies that
 4   did not report statistical testing were identified and, when appropriate, statistical tests were
 5   conducted by EPA.
 6          Information on study features related to this evaluation is reported in evidence tables and
 7   documented in the synthesis of evidence. Discussion of study strengths and limitations were
 8   included in the text where relevant If EPA's interpretation of a study differs from that of the study
 9   authors, the draft assessment discusses the basis for the difference.

10   Experimental Animal Studies
11          The  experimental animal studies, comprised entirely of studies performed in rats and mice,
12   were associated with drinking water, oral gavage, liquid diets (i.e., maltose/dextrin), and inhalation
13   exposures to tert-butanol. With the exception of neurodevelopmental studies, these sources were
14   conducted according to Organisation for Economic Co-operation and Development Good
15   Laboratory Practice (GLP) guidelines, presented extensive histopathological data, or clearly
16   presented their methodology; thus, these studies are considered high quality. These studies include
17   2-year bioassays using oral exposures in rats and mice; two subchronic drinking water studies in
18   rats and one in mice; an inhalation subchronic study in rats and mice; a reevaluation of the NTP
19   [1995] rat data; two oral developmental studies; two inhalation developmental studies; and a
20   single one-generation reproductive study that also evaluates other systemic effects (Table LS-5). A
21   more detailed discussion of any methodological concerns that were identified precedes each
22   endpoint evaluated in the hazard identification section. Overall, the experimental animal studies of
23   tert-butanol involving repeated oral  or inhalation exposure were considered to be of acceptable
24   quality, and whether yielding positive, negative, or null results, were considered in assessing the
25   evidence for health effects associated with chronic exposure to tert-butanol.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                LS-7              DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review of tert-ButyI Alcohol
Table LS-4. Considerations for evaluation of experimental animal studies
Methodological
feature
Test animal
Experimental design
Exposure
Endpoint evaluation
Results presentation
Considerations
(relevant information extracted into evidence tables)
Suitability of the species, strain, sex, and source of the test animals
Suitability of animal age/lifestage at exposure and endpoint testing; periodicity and
duration of exposure (e.g., hr/day, day/week); timing of endpoint evaluations; and
sample size and experimental unit (e.g., animals, dams, litters)
Characterization of test article source, composition, purity, and stability; suitability of the
control (e.g., vehicle control); documentation of exposure techniques (e.g., route,
chamber type, gavage volume); verification of exposure levels (e.g., consideration of
homogeneity, stability, analytical methods)
Suitability of specific methods for assessing the endpoint(s) of interest
Data presentation for endpoint(s) of interest (including measures of variability) and for
other relevant endpoints needed for results interpretation (e.g., maternal toxicity,
decrements in body weight relative to organ weight)
Table LS-5. Summary of experimental animal database
Study category
Chronic
Subchronic
Reproductive
Developmental
Neurodevelopmental
Study duration, species/strain, and administration method
2-year study in F344 rats (drinking water) NTP (1995)
2-year study in B6C3Fi mice (drinking water) NTP (1995)

13-week study in B6C3Fi mice (drinking water) NTP (1995)
13-week study in F344 rats (drinking water) NTP (1995)
13-week study in F344 rats (inhalation) NTP (1997)
13-week study in B6C3Fi mice (inhalation) NTP (1997)
10-week study in Wistar rats (drinking water) Acharya et al. (1997), Acharya et

One-generation reproductive toxicity study in Sprague-Dawley rats (gavage) L^
Chemical Co. (2004)


al. (1995)

ondell
Developmental study (GD 6-20) in Swiss Webster mice (diet) Daniel and Evans (1982)
Developmental study (GD 6-18) in CBA/J mice (drinking water) Faulkner et al. (1989)
Developmental study (GD 6-18) in C57BL/6J mice (drinking water) Faulkner et
Developmental study (GD 1-19) in Sprague-Dawley rats (inhalation) Nelson et

al. (1989)
al. (1989)

Neurodevelopmental study (GD 6-20) in Swiss Webster mice (diet) Daniel and Evans
(1982)
Neurodevelopmental study (GD 1-19) in Sprague-Dawley rats (inhalation) Nelson et al.
(1991)

   This document is a draft for review purposes only and does not constitute Agency policy,
                                 LS-8             DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review of tert-ButyI Alcohol
    1.   HAZARD IDENTIFICATION
2   1.1.  Overview of Chemical Properties and Toxicokinetics

3   1.1.1.  Chemical Properties
4         tert-Butanol is a white crystalline solid or colorless, highly flammable liquid (above 25.7°C)
5   with a camphor-like odor [NIOSH. 2005: IPCS. 1987a). tert-Butanol contains a hydroxyl chemical
6   functional group; is miscible with alcohol, ether, and other organic solvents; and is soluble in water
7   [IPCS. 1987a). Selected chemical and physical properties of tert-butanol are presented in Table 1-1.
          Table 1-1. Physicochemical properties and chemical identity of tert-butanol
Characteristic
Chemical name

Synonyms/Trade names


Chemical formula

CASRN

Molecular weight

Melting point

Boiling point

Vapor pressure

Density/Specific gravity

Flashpoint

Water solubility at 25°C

Octanol/Water Partition
Coefficient (Log Kow )
Henry's Law Constant

Odor threshold

Conversion factors

Information
tert-Butanol

t-Butyl alcohol; tert-Butanol; tert-Butyl alcohol; t-
Butyl hydroxide; 1,1-Dimethylethanol; NCI-C55367;
2-Methyl-2-propanol; tertiary Butanol; Trimethyl
carbinol; Trimethyl methanol; t-butyl alcohol; TBA
C4HioO

75-65-0

74.12 ^^^ ^S

25.7°C

82.41°C

40.7 mm Hg @ 25°C

0.78581

11°C (closed cup)

1 x 10s mg/L

0.35

9.05 x 10"6 atm-m3/mole

219 mg/m3

1 ppm = 3.031 mg/m3
1 mg/m3 - 0.324 ppm
Reference
HSDB (2007)

HSDB (2007)
IPCS (1987b)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

HSDB (2007)

IPCS (1987b)

             This document is a draft for review purposes only and does not constitute Agency policy,
                                           1-1          DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol
Characteristic

Chemical structure
Information
C
HO
(
;HS
mi
*Hg
Reference

HSDB (2007)

 1   1.1.2.  Toxicokinetics
 2          tert-Butanol is rapidly absorbed following exposure by oral and inhalation routes (see
 3   Appendix B, Section B.I.I). Studies in experimental animals indicate that 99% of the compound was
 4   absorbed after oral administration. Comparable blood levels of tert-butanol and its metabolites
 5   have also been observed after acute oral or inhalation exposures in rats [ARCO. 1983]. In another
 6   study [Faulkner etal., 1989], blood concentrations indicated that absorption was complete at 1.5
 7   hours following oral gavage doses of tert-butanol in female mice.
 8          tert-Butanol is distributed throughout the body following oral, inhalation, and i.v. exposures
 9   [Poet etal.. 1997: Faulkner etal.. 1989: ARCO. 1983]. Following exposure to tert-butanol in rats,
10   tert-butanol was found in kidney, liver, and blood, with male rats retaining more tert-butanol than
11   female rats [Williams and Borghoff. 2001].
12          A general metabolic scheme for tert-butanol, illustrating the biotransformation in rats and
13   humans, is shown in Figure 1-1 (see Appendix B.I.3].
14          Human data on the excretion of tert-butanol comes from studies of methyl tert-butyl ether
15   (MTBE) and ethyl tert-butyl ether (ETBE] fNihlenetal.. 1998a. b). The half-life of tert-butanol in
16   urine following MTBE exposure was 8.1 ± 2.0 hours (average of the 90.1- and 757-mg/m3 MTBE
17   doses]; the half-life of tert-butanol in urine following ETBE exposure was 7.9 ±2.7 hours (average
18   of 104- and 210-mg/m3 ETBE doses]. These studies reported urinary levels of tert-butanol (not
19   including downstream metabolites] to be less than 1% of administered MTBE or ETBE
20   concentrations (Nihlenetal., 1998a, b]. Ambergetal. (2000] also observed a similar half-life of
21   9.8 ± 1.4 hours after human exposure to ETBE of 170 mg/m3. The half-life for tert-butanol in rat
22   urine was 4.6 ± 1.4 hours at ETBE levels of 170 mg/m3.
23          A more detailed summary of tert-butanol toxicokinetics is provided in Appendix B,
24   Section B.I.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                 1-2            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
                   CH3
                   I
      glucuronide-O — — CH3
                   CH3
         t-butyl glucuronide
                   rats, humans
                                                                  _
                                                                      HO-
                                                                [O]
                                                                      -CH,
          CH3
     HO——CH3
          CH3
       t-butanol
                  rats
                        CYP450
                        rats,
                        humans
                                    QU   OH
                            2-methyl-1 ,2-propanediol
                                                                          CH,
                                                                    2-hydroxyisobutyric acid
                                                                   formaldehyde
oA
                CH,
                                                                  acetone
                    -CH,
                CH3
           t-butyl sulfate
 1          Source: NSF International (2003), ATSDR (1996), Bernauer et al. (1998), Amberg et al. (1999),
 2          and Cederbaum and Cohen (1980).

 3          Figure 1-1. Biotransformation of tert-butanol in rats and humans.

 4   1.1.3.  Description of Toxicokinetic Models
 5          No physiologically based pharmacokinetic (PBPK) models have been developed specifically
 6   for administration of tert-butanol. Some models have been used to study tert-butanol as the
 7   primary metabolite after oral or inhalation exposure to MTBE or ETBE. The most recent models for
 8   MTBE oral and inhalation exposure include a component for the binding of tert-butanol to
 9   (X2u-globulin [Borghoff etal., 2010: Leavens and Borghoff, 2009]. A more-detailed summary of the
10   toxicokinetic models is provided in Appendix B, Section B.I.5.
11          A PBPK model for tert-butanol was modified by adapting previous models for MTBE and
12   tert-butanol [Leavens and Borghoff, 2009: Blancato etal., 2007]. The addition of a sequestered
13   blood compartment for tert-butanol substantially improved the model fit. The alternative
14   modification of changing to diffusion-limited distribution between blood and tissues also improved
15   the model fit, but was considered less biologically plausible. Physiological parameters and partition
16   coefficients were obtained from published measurements. The rate constants for tert-butanol
17   metabolism and  elimination were from a published PBPK model of MTBE with a tert-butanol
18   subcompartment [Blancato etal.. 2007]. Additional model parameters were estimated by
19   calibrating to data sets for i.v., oral, and inhalation exposures as well as repeated dosing studies for
20   tert-butanol. Overall, the model produced acceptable fits to multiple rat time-course datasets of
21   tert-butanol blood levels following either inhalation or oral gavage exposures.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-3            DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-Butyl Alcohol

 1   1.1.4.  Chemicals Extensively Metabolized to tert-Butanol
 2          tert-Butanol is a metabolite of other compounds, including ETBE, MTBE, and tert-butyl
 3   acetate. Some of the toxicological effects observed in these compounds are attributed to tert-
 4   butanol. There are no assessments by national or international health agencies for ETBE. Animal
 5   studies demonstrate that chronic exposure to ETBE is associated with noncancer kidney effects,
 6   including increased kidney weights in male and female rats accompanied by increased chronic
 7   progressive nephropathy (CPN), urothelial hyperplasia (in males), and increased blood
 8   concentrations of total cholesterol, blood urea nitrogen, and creatinine [Saito etal.. 2013: Suzuki et
 9   al.. 2012). In these studies, increased liver weight and centrilobular hypertrophy also were
10   observed in male and female rats exposed to ETBE. Liver adenomas and carcinomas were increased
11   in male rats following 2-year inhalation exposure [Saito etal., 2013].
12          In 1996, the U.S. Agency for Toxic Substances and Disease Registry's (ATSDR) Toxicological
13   Profile for MTBE [ATSDR. 1996] identified cancer effect levels of MTBE based on data on
14   carcinogenicity in animals. ATSDR reported that inhalation exposure was associated with kidney
15   cancer in rats and liver cancer in mice. ATSDR concluded that oral exposure to MTBE might cause
16   liver and kidney damage, and nervous system effects in rats and mice. The chronic inhalation
17   minimal risk level was derived based on incidence and severity of chronic progressive nephropathy
18   in female rats [ATSDR. 1996]. In 1997, EPA's Office of Water concluded that MTBE is carcinogenic
19   to animals and poses a potential carcinogenic potential to humans based on an increased incidence
20   of Leydig cell adenomas of the testes, kidney tumors, lymphomas, and leukemia in exposed rats
21   [U.S. EPA. 1997]. In 1998, the International Agency for Research on Cancer (IARC] found "limited
22   evidence" of MTBE carcinogenicity in animals and placed MTBE in Group 3 (i.e., not classifiable as
23   to carcinogenicity in humans]  (IARC. 1999]. IARC reported that oral exposure in rats resulted in
24   testicular tumors in males and lymphomas and leukemias (combined] in females; inhalation
25   exposure in male rats resulted in renal tubule adenomas; and inhalation exposure in female mice
26   resulted in hepatocellular adenomas (IARC. 1999].
27          No assessments by national or international agencies or chronic studies for tert-butyl
28   acetate are available.

29   1.2.   PRESENTATION AND SYNTHESIS OF EVIDENCE BY ORGAN/SYSTEM

30   1.2.1.  Kidney Effects
31   Synthesis of Effects in Kidney
32          This section reviews the studies that investigated whether subchronic or chronic exposure
33   to tert-butanol can affect kidneys in humans or animals. The database examining kidney effects
34   following tert-butanol exposure contains eight studies from 5 references performed in rats or mice
35   (Lyondell Chemical Co.. 2004: Acharyaetal.. 1997: NTP. 1997: Acharya etal.. 1995: NTP. 1995]. and
36   a reevaluation of the rat data from NTP (1995]. published by Hard etal. (2011]: no human data are

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-4           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   available. Studies using short-term and acute exposures that examined kidney effects are not
 2   included in the evidence tables; they are discussed in the text, however, if they provide data to
 3   inform mode of action (MOA) or hazard identification. tert-Butanol exposure resulted in kidney
 4   effects after both oral (drinking water) and inhalation exposure in both sexes of rats (Table 1-1,
 5   Table 1-2, Figure 1-1, and Figure 1-2); studies are arranged in the evidence tables first by effect,
 6   then by route, and then duration.
 7          The design, conduct, and reporting of each study were reviewed, and each study was
 8   considered adequate to provide information pertinent to this assessment. Interpretation of non-
 9   neoplastic kidney endpoints in rats, however,  is somewhat complicated by the common occurrence
10   of age-related, spontaneous lesions characteristic of chronic progressive nephropathy (CPN) (NTP.
11   2015: HardetaL 2013: MelnicketaL 2012: U.S. EPA.
12   1991a]:http://ntp.niehs.nih.gov/nnl/urinary/kidney/necp/index.htm]. CPN is more severe in male
13   rats than in females and is particularly common in the Sprague-Dawley and Fischer 344 strains.
14   Dietary and hormonal factors play a role in modifying CPN, although the etiology is largely
15   unknown (see further discussion below).
16          Kidney weight. Changes in kidney weight (absolute and relative to body weight) were
17   observed in male and female F344 rats following exposures of 13 weeks (oral and inhalation) (NTP.
18   1997) and 15 months (oral) (NTP. 1995). Lyondell Chemical Co. (2004) also reported increases in
19   absolute and relative kidney weight in Sprague-Dawley rats administered tert-butanol orally for
20   approximately 10 weeks (tabular data presented in Supplemental Information to this Toxicological
21   Review). Changes were observed in both male and female rats, which exhibited strong dose-related
22   increases in absolute kidney weight (Spearman's rank coefficient > 0.78) following either oral or
23   inhalation exposures (Figure 1-3). Of the oral  (Figure 1-4 and inhalation (Figure 1-5) mouse
24   studies, only inhalation exposure in female mice induced a strong dose-related increase
25   (Spearman's rank coefficient = 0.9) in absolute kidney weights.
26          Measures of relative, as opposed to absolute, organ weight are sometimes preferred
27   because they account for changes in body weight that might influence changes in organ weight
28   (Bailey et al.. 2004). although potential impact should be evaluated. For tert-butanol, body weight in
29   exposed animals noticeably decreased at the high doses relative to controls in the oral 13-week and
30   2-year studies NTP (1995). In this case, the decreased body weight of the animals affects the
31   relative kidney weight measures, resulting in an artificial exaggeration of changes. Thus, absolute
32   weight was determined the more reliable measure of kidney weight change for this assessment
33   Additionally, a recent analysis indicates that increased absolute, but not relative, subchronic kidney
34   weights are significantly correlated with chemically induced histopathological findings in the
35   kidney in chronic and subchronic studies (Craig etal.. 2014). Although relative and absolute kidney
36   weight data are both presented in exposure-response arrays (and in evidence tables in
37   Supplemental Information), the absolute measures were considered more informative for
38   determining tert-butanol hazard potential.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-5            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1          Kidney histopathology. Treatment-related histopathological changes were observed in the
 2   kidneys of male and female F344 rats following 13-week and 2-year oral exposures [NTP, 1995]
 3   and male F344 rats following a 13-week inhalation exposure [NTP. 1997}. Similarly, male Wistar
 4   rats exposed for approximately 10 weeks exhibited an increase in histopathological kidney lesions
 5   [Acharya et al.. 1997: Acharya et al.. 1995). B6C3Fi mice, however, did not exhibit histopathological
 6   changes when exposed for 13 weeks and 2 years via the oral route [NTP. 1995] and 13 weeks via
 7   the inhalation route [NTP, 1997]. More specific details on the effects observed in rats, reported by
 8   NTP [1997.1995] and Acharya etal. [1997]: [1995] are described below.
 9          Nephropathy and severity of nephropathy were reported in male and female rats in the
10   13-week oral studies [NTP. 1995]. The nephropathy was characterized as "...a spontaneous
11   background lesion...typically consisting] of scattered renal tubules lined by basophilic
12   regenerating tubule epithelium." [NTP. 1995]. NTP [1995] noted that the increase in severity of
13   nephropathy was related to tert-butanol and "characterized by an increase in the number and size
14   of foci of regeneration." The severity of nephropathy increased, compared with controls, in the
15   13-week male rats, which exhibited nephropathy in 94% of all exposed animals and 70% of
16   controls. Conversely, lesion severity was unchanged in the females, although nephropathy
17   incidence significantly increased with  tert-butanol exposure. In the 13-week inhalation study [NTP.
18   1997], nephropathy was present in all but two male rats, including controls. NTP [1997]
19   characterized the reported chronic nephropathy in control male rats as "1 to 3 scattered foci of
20   regenerative tubules per kidney section. Regenerative foci were characterized by tubules with
21   cytoplasmic basophilia, increased nuclear/cytoplasmic ratio, and occasionally thickened basement
22   membranes and intraluminal protein casts." In exposed groups, the severity generally increased
23   from minimal to mild with increasing dose as "evidenced by an increased number of foci." No
24   treatment-related kidney histopathology was reported in the female rats exposed through
25   inhalation [NTP. 1997].
26          In the 2-year oral study by NTP [1995]. nephropathy was reported at 15 months and 2
27   years. The NTP [1995] characterization of nephropathy following chronic exposure included
28   multiple lesions: "thickened tubule and glomerular basement membranes, basophilic foci of
29   regenerating tubule epithelium, intratubule protein casts, focal mononuclear inflammatory cell
30   aggregates within areas of interstitial fibrosis and scarring, and glomerular sclerosis." At 15
31   months, male and female rats (30/30 treated;  10/10 controls] had nephropathy, and the severity
32   scores ranged from minimal to mild. At 2 years, male and female rats (149/150 treated; 49/50
33   controls] also had nephropathy, and although  the severity was moderate in the control males and
34   minimal to mild in the  control females, severity increased with tert-butanol exposure in both sexes
35   [NTP. 1995].
36          The lesions collectively described by NTP [1997,1995] as nephropathy and noted to be
37   common spontaneous  lesions in rats, are consistent with CPN. The effects characterized as CPN are
38   related to age and not considered histopathological manifestations of chemically induced toxicity
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-6           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   [see U.S. EPA [1991a]. p. 35 for further details and a list of the typical observable histopathological
 2   features of CPN]. These lesions, however, are frequently exacerbated by chemical treatment [NTP,
 3   1997}. as evidenced by the dose-related increases in severity of the nephropathy compared to
 4   female and male rat controls. The chemical-related changes in nephropathy severity are included in
 5   the consideration of hazard potential.
 6          NTP [1995] observed other kidney lesions, described as being associated with nephropathy
 7   but diagnosed separately. Renal mineralization is defined by NTP [1995] as "focal mineral deposits
 8   primarily at the corticomedullary junction." This mineralization is distinct from linear
 9   mineralization, which is considered a lesion characteristic of azu-globulin nephropathy (for further
10   discussion of this particular lesion, see Mode of Action Analysis—Kidney Effects]. The mineralization
11   is characterized as distinct linear deposits along radiating medullary collecting ducts. An increased
12   incidence of linear mineralization was limited to exposed males in the 2-year oral study [NTP.
13   1995].
14          Renal mineralization was observed in essentially all female rats at all reported treatment
15   durations. A dose-related, increased incidence of mineralization was reported in male rats at the
16   13-week, 15-month, and 2-year oral evaluations [NTP. 1995]. NTP [1995] describes focal,
17   medullary mineralization as being associated with CPN but notes that focal mineralization is
18   "usually more prominent in untreated females than in untreated males," which is consistent with
19   the widespread appearance of this lesion in females. This description, however, is inconsistent with
20   the observation in this and other databases that age-related nephropathy (i.e., CPN] is generally
21   more prevalent and more severe in male rats compared to females [U.S. EPA. 1991a]. The
22   association of mineralization with CPN is unclear, considering the lack of spontaneous lesions in the
23   control and low-dose groups of 13-week males and the dose-response  relationships the tert-
24   butanol-exposed males exhibited in the 13-week [NTP. 1997.1995]  and 2-year studies [NTP.
25   1995]. Furthermore, due to the overwhelming presence of mineralization in the control and treated
26   female rats, the contribution, if any, of tert-butanol to the formation  of this lesion in females could
27   not be determined. Thus, the mineralization could be related to both aging of the animals and tert-
28   butanol exposure.
29          Two other histological kidney lesions observed in male and female rats are suppurative
30   inflammation and transitional epithelial hyperplasia. These lesions were observed in the 2-year
31   oral NTP [1995] study. Although NTP [1995] describes these lesions as related to the nephropathy
32   (characterized above as common and spontaneous, and considered CPN], that suppurative
33   inflammation and transitional epithelial hyperplasia exhibited incidence patterns different from
34   those reported for nephropathy is notable. Incidence of suppurative inflammation in female rats
35   was low in the control group and increased with dose, with incidences >24% in the two highest
36   dose groups, compared with controls. In comparison, 20% of the control males exhibited
37   suppurative inflammation, and the changes in incidence were not dose related (incidences ranging
38   from 18 to 36%]. The data for males suggest that CPN plays a role in the induction of suppurative
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-7            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   inflammation; considering the responses in the females, however, the effect appears to be
 2   predominantly treatment related. Suppurative inflammation was not observed in the animals of the
 3   13-week oral fNTP. 19951 or inhalation study fNTP. 19971 which both reported nephropathy (as
 4   CPN), providing further support that this lesion is not specifically related to the nephropathy.
 5          Transitional epithelial hyperplasia was observed in both male and female rats exposed
 6   orally [NTP. 1995). In the control males, 50% of the animals exhibited transitional epithelial
 7   hyperplasia and the incidence and severity increased with dose. Only the mid- and high-dose
 8   females, however, exhibited dose-related increases in incidence and severity of transitional
 9   epithelial hyperplasia; this lesion was not reported in the control or low-dose females. NTP [1995]
10   described transitional epithelial hyperplasia as increased layers of the transitional epithelial lining
11   of the renal pelvis; study authors noted no progression of this hyperplastic lesion to neoplasia. The
12   relatively high background in male controls (i.e., 50%) suggests some potential influence, other
13   than tert-butanol treatment, on this effect The absence of this effect in female control and low-dose
14   animals and the dose-related increases in both males and females, however, indicate that similar to
15   the suppurative inflammation, the transitional epithelial hyperplasia is predominantly treatment
16   related. Transitional epithelial hyperplasia should not be confused with another lesion noted at the
17   2-year evaluation, renal tubule hyperplasia, which was considered preneoplastic (for further details
18   regarding this type of hyperplasia, see the discussion under kidney tumors below).
19          Additional histopathological changes, including increased tubular degeneration,
20   degeneration of the basement membrane of the Bowman's capsule, diffused glomeruli, and
21   glomerular vacuolation were noted in a 10-week study in male Wistar rats (Acharyaetal.. 1997:
22   Acharyaetal., 1995). A decrease in glutathione in the kidney accompanied these changes, which the
23   study authors noted as potentially indicative of oxidative damage. Acharyaetal. (1997): Acharya et
24   al. (1995) used one dose and a control group and did not report incidences. The increased tubule
25   degeneration and glomerular vacuolation could be characterized as tubular atrophy and glomerular
26   hyalinization, respectively, consistent with CPN; however, without quantitative information,
27   examining the differences between the control and treated animals to determine if CPN plays a role
28   in development of these effects is not possible. Although based on the noted appearance of the
29   effects in the treated animals compared with controls, the effects likely are treatment related.
30          Serum or urinary biomarkers informative of kidney toxicity were not measured in the
31   studies discussed above. Some changes occurred in urinalysis parameters (e.g., decreased urine
32   volume and increased specific gravity), accompanied by reduced water consumption, and thus
33   might not be related to an effect of kidney function (NTP. 1995).
34          Kidney tumors. The kidney is also a target organ for cancer effects (Table 1-3, Figure 1-1).
35   Male F344 rats had an increased incidence of combined renal tubule adenomas or carcinomas in
36   the 2-year oral bioassay (Hardetal.. 2011: NTP. 1995). The increase in tumors from control was
37   similar in the low- and high-dose groups and highest in the mid-dose group. Overall, tumor
38   increases were statistically significant in trend testing, which accounted for mortality (p < 0.018).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-8            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   Mortality increased with increasing exposure (p = 0.001); increased mortality alone, however, does
 2   not account for the highest tumor incidence occurring at the middle dose.
 3          Increases in incidence and severity of renal tubule hyperplasia also were observed in male
 4   rats. NTP [1995] stated that "[t]he pathogenesis of proliferative lesions of renal tubule epithelium is
 5   generally considered to follow a progression from hyperplasia to adenoma to carcinoma [Hard.
 6   1986]." Similarly, EPA considered the renal tubule hyperplasia to be a preneoplastic effect
 7   associated with the renal tubule tumors. Renal tubule hyperplasia was found in one high-dose
 8   female [NTP, 1995]:  no increase in severity was observed. This effect in females, which was not
 9   considered toxicologically significant, is not discussed further. Two renal tubular adenocarcinomas
10   in male mice also were reported [NTP. 1995]. one each in the low- and high-dose groups, but were
11   not considered by NTP to be "biologically noteworthy changes"; thus the  tumors in mice are not
12   discussed further.
13          A Pathology Working Group, sponsored by Lyondell Chemical Company, reevaluated the
14   kidney changes in the NTP 2-year study to determine if additional histopathological changes could
15   be identified to inform the MOA for renal tubule tumor development [Hardetal., 2011]. In all cases,
16   working group members were blinded to treatment groups, and used guidelines published by Hard
17   and Wolf T19991 and refinements reported by (Hard and Seely. 2006): Hard and Seely f20051 and
18   Hard (2008). The group's report and analysis by Hardetal. (2011) confirmed the NTP findings of
19   renal tubule hyperplasia and renal tubule tumors in male rats at 2 years.  In particular, they
20   reported similar overall tumor incidences in the exposed groups. Hardetal. (2011). however,
21   reported fewer renal tubule adenomas and carcinomas in the control group than in the original NTP
22   study. As a result, all treated groups had statistically significant increases in renal tubule adenomas
23   and carcinomas (combined) when compared to controls. Additionally, Hardetal. (2011) considered
24   fewer tumors to be carcinomas than did the original  NTP study.  Results of both NTP  (1995) and the
25   reanalysis by Hardetal. (2011) are included in Table 1-3 and Figure 1-1.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-9            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-Butyl Alcohol
        50
     s>  40 -
     I
     o
     ^  30H
     0
     ;S  10-

     <
        o -
                            Male rats
                                                                     Female rats
            rho= 0.78 (all)
            rho= 0.89 (oral)
            rho= 0.80 (inhalation)
                                          rho= 0.78 (all)
                                          rho= 0.72 (oral)
                                          rho= 0.9 (inhalation)
                     10
                                100         1000
                                         tert-butanol blood concentration (mg/l)
•
o
Oral exposure
Inhalation exposure
3
4
5
6
7
Figure 1-2. Comparison of absolute kidney weight change in male and female
rats across oral and inhalation exposure based on internal blood
concentration. Spearman rank correlation coefficient (rho) was calculated to
evaluate the direction of a monotonic association (e.g., positive value =
positive association) and the strength of association.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-10           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
                            Male mice
                                                                   Female mice
     3>
        12 -
        10 -
         4 -

         2 -
             rho=-0.1
                                                     rho= 0.9
                  2000      4000      6000

                      Administered dose (mg/kg-day)
                                 8000   0     2000    4000    6000   8000   10000   12000

                                                 Administered dose (mg/kg-day)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
        -10 -

        -12
Figure 1-3. Comparison of absolute kidney weight change in male and female
mice following oral exposure based on administered concentration. Spearman
rank correlation coefficient (rho) was calculated to evaluate the direction of a
monotonic association (e.g., positive value = positive association) and the
strength of association.
                           Male mice
                                                                  Female mice
            mo= 0.2
                                                    rho=-0.1
   1000   2000    3000    4000    5000   6000

           Administered dose (mg/m3)
                                                        1000   2000   3000    4000   5000   6000    7000

                                                               Administered dose (mg/m3)
Figure 1-4. Comparison of absolute kidney weight change in male and female
mice following inhalation exposure based on administered concentration.
Spearman rank correlation coefficient (rho) was calculated to evaluate the
direction of a monotonic association (e.g., positive value = positive
association) and the strength of association.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-11           DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol
       Table 1-2. Changes in kidney histopathology in animals following exposure to
       tert-butanol
 Reference and study design
                               Results
Acharyaetal. (1997)
Acharyaetal. (1995)
Wistar rat; 5-6 males/treatment
Drinking water (0 or 0.5%), 0 or
575 mg/kg-d
10 weeks
1" tubular degeneration, degeneration of the basement membrane of the
Bowman's capsule, diffused glomeruli, and glomerular vacuolation (no
incidences reported)

4, kidney glutathione (~40%)*
NTP (1995)
F344/N rat; 10/sex/treatment
Drinking water (0, 2.5, 5, 10, 20,
or 40 mg/mL)
 M: 0, 230, 490, 840, 1,520,
3,610a mg/kg-d
 F: 0, 290, 590, 850, 1,560,
3,620a mg/kg-d
13 weeks
Incidence (severity):
   Males
                                                                   Females
   Dose      Minerali-     Nephro-      Dose      Minerali-     Nephro-
 (mg/kg-d)     zation^      pathy£     (mg/kg-d)    zation^      pathy£
     0

    230

    490
            0/10      7/10 (1.0)
            0/10
10/10
(1.6*)
                                           2/10 (1.5)     10/10
                                                        (2.6*)
 0

290


590
10/10 (1.7)   2/10 (1.0)

10/10 (2.0)   3/10 (1.0)
                                  840
                                            10/10 (2.0)   5/10 (1.0)


                                    850     10/10 (2.0)   7/10* (1.0)


                                   1,560    10/10 (2.0)   8/10* (1.0)


3,610a     4/10*(1.0)   7/10(1.1)     3,620a     6/10(1.2)   7/10* (1.0)
            8/10* (1.4)     10/10
                          (2.7*)
                                 1,520     4/10* (1.0)     10/10
                                                        (2.6*)
NTP (1995)
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20,
or 40 mg/mL)
 M: 0, 350, 640, 1,590, 3,940,
8,210a mg/kgd
 F: 0, 500, 820, 1,660, 6,430,
ll,620a mg/kgd
13 weeks
Study authors indicated no treatment-related changes in kidney
histopathology (histopathological data not provided for the 13-week study)
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-12            DRAFT—DO NOT CITE OR QUOTE

-------
                                         Toxicological Review of tert-ButyI Alcohol
Reference and study design
NTP (1995)

F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at
15 months interim)
Drinking water (0, 1.25, 2.5, 5, 10
mg/mL)
M: 0, 90, 200, 420a mg/kg-d
F: 0, 180, 330, 650a mg/kg-d
2 years























NTP (1995)
B6C3Fi mouse; 60/sex/treatment
Drinking water (0, 5, 10, or 20
mg/mL)
M: 0, 540, 1,040, or 2,070a
mg/kg-d
Results
Incidence (severity):
Males

Dose
(mg/kg-d)
0
90
200
420a

Dose
(mg/kg-d)
0
90
200
420a
Females

Dose
(mg/kg-d)
0
180
330
650a

Dose
(mg/kg-d)
0
180
330
650a
No treatment-related








Mineralization^
(interim)
1/10 (1.0)
2/10 (1.0)
5/10 (1.8)
9/10* (2.3)
Transitional
epithelial
hyperplasia
25/50 (1.7)
32/50 (1.7)
36/50* (2.0)
40/50* (2.1)


Mineralization^
Interim

10/10 (2.8)
10/10 (2.9)
10/10 (2.9)
10/10 (2.8)
Transitional
epithelial
hyperplasia
0/50
0/50
3/50 (1.0)
17/50*(1.4)
changes in kidney








Mineralization^ n
(terminal)
26/50 (1.0)
28/50 (1.1)
35/50 (1.3)
48/50* (2.2)

Nephropathy£
severity
3.0
3.1
3.1
3.3*


Mineralization^
Terminal

49/50 (2.6)
50/50 (2.6)
50/50 (2.7)
50/50 (2.9)

Nephropathy£
severity
1.6
1.9*
2.3*
2.9*
related histopathology








Lin 63 r
mineralization^
(terminal)
0/50
5/50* (1.0)
24/50* (1.2)
46/50* (1.7)








nflammation
[suppurative)
incidence

2/50
3/50
13/50*
17/50*







observed





This document is a draft for review purposes only and does not constitute Agency policy,
                                  1-13           DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review of tert-ButyI Alcohol
       Reference and study design
                                                               Results
      F: 0, 510,1,020, or 2,110 mg/kg-
      d
      2 years
      NTP (1997)
      F344/N rat; 10/sex/treatment
      Inhalation analytical
      concentration: 0,134, 272, 542,
      1,080, or 2,101 ppm (0, 406, 824,
      1,643, 3,273 or 6,368 mg/m3)
      (dynamic whole-body chamber)
      6 hr/d, 5 d/wk
      13 weeks
      Generation method (Sonimist
      Ultrasonic spray nozzle
      nebulizer), analytical
      concentration and method were
      reported
                                Male


                                  Concentration
                                    (mg/m3-)

                                       0

                                      406

                                      824

                                      1,643

                                      3,273

                                      6,368
 Incidence of
   chronic
nephropathy-

    9/10

    8/10

    9/10

    10/10

    10/10

    10/10
Average severity
   of chronic
  nephropathy

      1.0

      1.4

      1.4

      1.6

      1.9

      2.0
                                     Females: no treatment-related changes in kidney related histopathology
                                     observed
                                     Severity categories: 1= minimal, 2= mild. No results from statistical tests
                                     reported
      NTP (1997)
                                No treatment-related changes in kidney related histopathology observed
      B6C3Fi mouse; 10/sex/treatment
      Inhalation analytical
      concentration: 0,134, 272, 542,
      1,080, or 2,101 ppm (0, 406, 824,
      1,643, 3,273 or 6,368 mg/m3)
      (dynamic whole-body chamber)
      6 hr/d, 5 d/wk
      13 weeks
      Generation method (Sonimist
      Ultrasonic spray nozzle
      nebulizer), analytical
      concentration and method were
      reported
1
2
3
4
5
* Statistically significant p < 0.05 as determined by the study authors.
a The high-dose group had an increase in mortality.
b Mineralization defined in NTP (1995) as focal mineral deposits, primarily at the corticomedullary junction. Linear
 mineralization was defined as foci of distinct linear deposits along radiating medullary collecting ducts; linear
 mineralization not observed in female rats.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                       1-14            DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review of tert-Butyl Alcohol
 1
 2
 3
 4
 5
 6

 7
 8
c Nephropathy defined in NTP (1995) as lesions including thickened tubule and glomerular basement membranes,
 basophilic foci of regenerating tubule epithelium, intratubule protein casts, focal mononuclear inflammatory cell
 aggregates within areas of interstitial fibrosis and scarring, and glomerular sclerosis.
d Nephropathy characterized in NTP (1997) as scattered foci of regenerative tubules (with cytoplasmic basophilia,
 increased nuclear/cytoplasmic ratio, and occasionally thickened basement membranes and intraluminal protein
 casts).

Note: Conversions from drinking water concentrations to mg/kg-day performed by study authors.
Conversion from ppm to mg/m3 is 1 ppm = 3.031 mg/m3.
 9
10
       Table 1-3. Changes in kidney tumors in animals following exposure to
       tert-butanol
Reference and study design
NTP (1995)
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at
15 months)
Drinking water (0, 1.25, 2.5, 5, or
10 mg/mL)
M: 0, 90, 200, or 420a mg/kg-d
F: 0, 180, 330, or 650a mg/kg-d
2 years




^^





Results
Male
Dose
(mg/kg-d)
0
90
200
420a

Dose
(mg/kg-d)
0
90
200
420a
Female
Dose
(mg/kg-d)
0
180
330
650a
Renal tubule
hyperplasia
(standard and
extended
evaluation
combined)
14/50 (2.3)
20/50 (2.3)
17/50 (2.2)
25/50* (2.8)

Renal tubule
carcinoma
0/50
2/50
1/50
1/50
Renal tubule
hyperplasia
0/50
0/50
0/50
1/50 (1.0)
Renal tubule
adenoma (single)
7/50
7/50
10/50
10/50
Renal tubule
adenoma (single
or multiple) or
carcinoma
8/50
13/50
19/50*
13/50
Renal tubule
adenoma (single)
0/50
0/50
0/50
0/50
Renal tubule
adenoma
(multiple)
1/50
4/50
9/50*
3/50




Renal tubule
adenoma
(multiple)
0/50
0/50
0/50
0/50
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                      1-15            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review of tert-ButyI Alcohol
Reference and study design

Hard etal. (2011)
Reanalysis of the slides from
male rats (all slides in controls
and high-dose groups of males
and females, and slides from all
other males with renal tumors) in
the NTP (1995) study (see above)

NTP (1995)
B6C3Fi mouse; 60/sex/treatment
Drinking water (0, 5, 10, or
20 mg/mL)
M: 0, 540, 1,040, or 2,070a
mg/kg-d
F: 0, 510, 1,020, or 2,110 mg/kg-
d
2 years
Results
Dose
(mg/kg-d)
Renal tubule
adenoma (single
Renal tubule or multiple) or
carcinoma carcinoma
0 0/50 0/50
180 0/50 0/50
330 0/50 0/50
650a 0/50 0/50
Based on standard and extended evaluations (combined). Results
include the animals sacrificed at 15 months.
Male
Dose
(mg/kg-d)
0
90
200
420
Renal tubule Renal tubule
adenoma adenoma Renal tubule
(single) (multiple) carcinoma
3/50 1/50 0/50
9/50 3/50 1/50
9/50 9/50 0/50
9/50 3/50 1/50
do not
Renal tubule
adenoma
(single or
multiple) or
carcinoma
4/50
13/50*
18/50*
12/50*
No increases in kidney-related tumors. Two renal tubule adenocarcinomas,
one in the low-dose and one in the high-dose groups, were observed in male
mice. These tumors were not considered treatment related.
^L
1
2
3
4
* Statistically significant p < 0.05 as determined by the study authors.
a The high-dose group had an increase in mortality.

Note: Conversions from drinking water concentrations to mg/kg-d performed by study authors.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-16           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
• = exposures at which the endpoint was reported statistically significant by study authors
D = exposures at which the endpoint was reported not statistically significant by study authors
x = exposures at which all animals were dead and unable to be examined for the endpoint
Kidney Absolute weight; M Rat; Reproductive (C]
Weight Relative weight; M Rat; Reproductive (C]
Absolute weight; F Rat; Reproductive [C]
Relative weight; F Rat; Reproductive fC]
Absolute weight; M Rat; 13wk (D]
Relative weight; M Rat; 13wk [D]
Absolute weight; F Rat; 13wk CD]
Relative weight; F Rat; 13wk CD]
Absolute weight; M Mouse; 13wk CD]
Relative weight; M Mouse; 13wk CD]
Absolute weight; F Mouse; 13wk (D]
Relative weight; F Mouse; 13wk CD]
Absolute weight; M Rat; 15mo CD]
Relative weight; M Rat; 15mo CD]
Absolute weight; F Rat; 15mo CD]
Relative weight; F Rat; 15mo (D]
Kidnev Decreased glutathione; M Rat; lOwk (A]
Histopathology Inflammation; F Rat; 2yr(D]
Ncphropathy severity; M Rat; 13wk (D)
Nephropathy incidence; F Rat; 13wk CD]
Mineralization; M Rat; 13wk [D]
Mineralization; F Rat; 13wk (D]
Nephropathy severity; M Rat; 2yr CD]
Nephropathy severity; F Rat; 2yr (D]
Linear mineralization; M Rat; 2yr CD)
Interim/terminal mineralization; M Rat; 2yrfD]
Interim/terminal mineralization; F Rat; 2yr (D]
Transitional epithelium hyperplasia; M Rat; 2yr (D]
Transitional epithelium hyperplasia; F Rat; 2yr (D]
Renal tubular hyperplasia; M Rat; 2yr (D]
Renal tubule hyperplasia; F Rat; 2yr CD]
Kidnev ^ena' tubular adenoma or carcinoma; M Rat; 2yr CD]
Tumors Renal tubular adenoma or carcinoma; M Rat; 2yr CB]
Renal tubular adenomaor carcinoma; F Rat; 2yr CD]
Renal tubular adenoma or carcinoma; M Mouse; 2yr (D)
Renal tubular adenoma or carcinoma; F Mouse; 2yr [D]





E
E
E
•
E
E
E
E
•







Dn



1 D D
1 • •
•
D-«-«
BDH
Qn •

) D •
1 • •
1 D •
B-B-0
1 • •
D D •
1 D •
Q-B-O
1 • D
1 • •
D-B-FJ
B-E
B-E


]
]



Bn










3-D
3-E



-

                                                         10     100    1,000   10,000  100,000

                                                            Dose (mg/kg-day)
1
2

3
4
Sources: (A) Acharya et al. (1997); (1995); (B) Hard etal. (2011)*; (C) Lyondell Chemical Co. (2004) (D) NTP
(1995); *reanalysis of NTP (1995).

Figure 1-5. Exposure response array for kidney effects following oral exposure
to tert-butanol.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                   1-17           DRAFT—DO NOT CITE OR QUOTE

-------
                                           Toxicological Review of tert-ButyI Alcohol
• = exposures at which the endpoint was reported statistically significant by study authors
D = exposures at which the endpoint was reported not statistically significant by study authors
Absolute weight; M Rat
Relative weight; M Rat
Absolute weight; F Rat
Relative weight; F Rat
tf^
4
Absolute/relative weight; M Mouse
Absolute weight; F Mouse
Relative weight; F Mouse -


Qi— i


Qi— i




BB •


B|— i i— i



                                100               1,000              10,000
                                        Concentration (mg/m3)
Source: NTP (1997).
Figure 1-6. Exposure-response array of kidney effects following inhalation
exposure to tert-butanol (13-week studies, no chronic studies available).
   This document is a draft for review purposes only and does not constitute Agency policy,
                                    1-18           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol
 1   Mode of Action Analysis—Kidney Effects
 2       a)  a?,,-Globulin-Associated Renal Tubule Nephropathy and Carcinogenicity
 3          One disease process to consider when interpreting kidney effects in rats is related to the
 4   accumulation of (X2u-globulin protein. (X2u-Globulin, a member of a large superfamily of low-
 5   molecular-weight proteins, was first characterized in male rat urine. Such proteins have been
 6   detected in various tissues and fluids of most mammals (including humans), but the particular
 7   isoform of (X2u-globulin commonly detected in male rat urine is considered specific to that sex and
 8   species. Exposure to chemicals that induce a2u-globulin accumulation can initiate a sequence of
 9   histopathological events leading to kidney tumorigenesis. Because a2u-globulin-associated renal
10   tubule nephropathy and carcinogenicity occurring in male rats are presumed not to be relevant for
11   assessing human health hazards [U.S. EPA. 1991a). evaluating the data to determine if a2u-globulin
12   plays a role is important. The role of (X2u-globulin accumulation in the development of renal tubule
13   nephropathy and carcinogenicity observed following tert-butanol exposure was evaluated using the
14   U.S. EPA[1991a] Risk Assessment Forum Technical panel report, Alpha2u-Globulin: Association with
15   Chemically Induced Renal Toxicity and Neoplasia in the  Male Rat. This report provides specific
16   guidance for evaluating renal tubule tumors in male rats that are related to chemical exposure for
17   the purpose of risk assessment, based on an examination of the potential involvement of
18   (X2u-globulin accumulation.
19          Studies in the tert-butanol database evaluated  and reported effects on the kidney, providing
20   some evidence to evaluate this MOA. Additionally, several studies were identified that specifically
21   evaluated the role of (X2u-globulin in tert-butanol-induced renal tubule nephropathy and
22   carcinogenicity fBorghoffetal.. 2001: Williams and Borghoff. 2001: TakahashietaL 19931 Because
23   the evidence reported in these studies is specific to a2u-globulin accumulation, it is presented in this
24   section; it was not included in the animal evidence tables in the previous section.
25          The hypothesized sequence of (X2u-globulin renal tubule nephropathy, as described by U.S.
26   EPA [1991a]. is as follows. Chemicals that induce (X2u-globulin accumulation do so rapidly.
27   a2u-Globulin accumulating in hyaline droplets is deposited in the S2 (P2) segment of the proximal
28   tubule within 24 hours  of exposure. Hyaline droplets are a normal constitutive feature of the
29   mature male rat kidney; they are particularly evident in the S2 (P2) segment of the proximal tubule
30   and contain (X2u-globulin [U.S. EPA. 1991a). Abnormal increases in hyaline droplets have more than
31   one etiology and can be associated with the accumulation of different proteins. As hyaline droplet
32   deposition continues, single-cell necrosis occurs in the S2 (P2) segment, which leads to exfoliation
33   of these cells into the tubule lumen within 5 days of chemical exposure. In response to the cell loss,
34   cell proliferation occurs in the S2 (P2) segment after 3  weeks and continues for the duration of the
35   exposure. After 2 or 3 weeks of exposure, the cell debris accumulates in the S3 (P3) segment of the
36   proximal tubule to form granular casts. Continued chemical exposure for 3 to 12 months leads to
37   the formation of calcium hydroxyapatite in the papillae which results in linear mineralization. After
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-19          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   1 or more years of chemical exposure, these lesions can result in the induction of renal tubule
 2   adenomas and carcinomas (Figure 1-7).
 3          U.S. EPA(1991a) identified two questions that must be addressed to determine the extent
 4   to which a2u-globulin-mediated processes induce renal tubule nephropathy and carcinogenicity.
 5   First, whether the a2u-globulin process is occurring in male rats and is involved in renal tubule
 6   tumor development must be determined. Second, whether the renal effects in male rats exposed to
 7   tert-butanol are solely due to the (X2u-globulin process also must be determined.
 8          U.S. EPA (1991a) stated the criteria for answering the first question in the affirmative are as
 9   follows:

10       1)  hyaline droplets are increased in size and number in treated male rats,

11       2)  the protein in the hyaline droplets in treated male rats is (X2uglobulin (i.e.,
12          immunohistochemical evidence), and

13       3)  several (but not necessarily all) additional steps in the pathological sequence appear in
14          treated male rats as a function of time, dose, and progressively increasing severity
15          consistent with the understanding of the underlying biology, as described above, and
16          illustrated in Figure 1-7.

17          The available data relevant to this first question are summarized in Table 1-4, Figures 1-8
18   and 1-9, and are evaluated below.
19
                        4                                    k
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-20          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
Male rat liver
Male rat kidney
Synthesis of a2u-globulin
^< 	 TBA binding
Resorption of poorly digestible
protein-chemical complex
1
Hyaline droplet accumulation
within lysosomes
                                      Cell death and exfoliation
                                                            > 1 days
                                                            1 - 150 days
                                                5 days - 48 weeks
                                                                      3-48 weeks
                                                                      > 12 months
                                                                      > 12 months
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
Figure 1-7. Temporal pathogenesis of a2u-globulin-associated nephropathy in
male rats. a2U-Globulin synthesized in the livers of male rats is delivered to the
kidney, where it can accumulate in hyaline droplets and be retained by epithelial
cells lining the S2 (P2) segment of the proximal tubules. Renal pathogenesis
following continued tert-butanol exposure and increasing droplet accumulation can
progress step-wise from increasing epithelial cell damage, death and dysfunction
leading to the formation of granular casts in the corticomedullary junction, linear
mineralization of the renal papillae, and carcinogenesis of the renal tubular
epithelium. Adapted from Swenberg and Lehman-McKeeman [1999] and U.S. EPA
C1991a1.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-21           DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review of tert-ButyI Alcohol
1            Table 1-4. Summary of data on the ct2u -globulin process in male rats exposed
2            to tert-butanol
           Duration               Dose           Results         Comments             Reference
      1)  Hyaline droplets are increased in size and number
          10 d (inhalation)   0, 758,1,364, 5,304        +      stat sig at 5,304 mg/m3;   Borghoff et al. (2001)
                           mg/m3                          stat sig trend
          13 wk (inhalation)  0,3,273,6,368 mg/m3      -                              NTP (1997)a
          13 wk (oral)       0,230,490,840,           (+)     observed in all but        NTP (1995)
                           1,520, 3,610 mg/kg-d              highest dose group
      2)  The protein in the hyaline droplets is 
-------
                                                               Toxicological Review of tert-ButyI Alcohol
Duration
Dose
d) Linear mineralization of tubules in the
13 wk (oral)
2 yr (oral)
0, 230, 490, 840,
1,520, 3,610 mg/kg-d
0, 90, 200, 420
mg/kg-d
Results Comments
renal papilla
+; (+) all doses stat sig
Reference

NTP (1995); Hard et al.
(2011)c
NTP (1995); Hard et al.
(2011)d
e) Foci of tub ular hyperplasia
2 yr (oral)
0, 90, 200, 420
mg/kg-d
+ stat sig trend at all
doses; stat sig at 420
mg/kg-d
NTP (1995)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
+ = Statistically significant change reported in one or more treated groups.
(+) = Effect was reported in one or more treated groups, but statistics not reported.
- = No statistically significant change reported in any of the treated groups.
a  NTP (1997) did not observe any effects consistent with cuu-globulin nephropathy.
b Precursors to granular casts reported.
c Reanalysis of hematoxylin and eosin-stained kidney sections from all male control and 1,520 mg/kg-d groups, as
 well as a representative sample of kidney sections stained with Mallory Heidenhain stain, from the 13-wk study
 from NTP (1995).
d Reanalysis of slides for all males in the control and 420 mg/kg-day dose group and all animals with renal tubule
 tumors from 2-vr NTP (1995). Protein casts reported, not granular casts.
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                        1-23             DRAFT—DO NOT CITE OR QUOTE

-------
                                                               Toxicological Review of tert-ButyI Alcohol
                • = exposures at which the endpoint was reported statistically significant by study authors
                D = exposures at which the endpoint was reported not statistically significant by study authors
                 x = exposures at which all animals were dead and unable to be examined for the endpoint
                 • = exposures at which effect was observed but statistics not reported	

size/number
Identification of a2u- Williams and Borghoff [2001);
globulin in hyaline j 2 hr after singie dose
droplets
Acharya et al. [1997); 10 wk
Cytotoxicity/single-cell
necrosis of tubule epithelium;
epithelial cell exfoliation

Acharya et al. (1997); 10 wk
Tubule cell
proliferation



Granular
casts/tubule
dilation
NTP [1995); Hard et al. [2011); 2 yr
t
Linear papillary
mineralization

Foci of
tubular NTP 11995); 2 yr
hyperplasia











B

m
•




•
•


D




	 B 	 Q












• y









                 * Hard et al. [2011] reported presence of "precursor
                 granular casts"
                 "NTP(1995) 13-wkstudy reported kidney
                 mineralization but not linear mineralization
                                              10
                                                         100        1,000

                                                         Dose (mg/kg-day)
10,000
1
2

3
4
5
*Hard et al. (2011) reported presence of "precursor granular casts."
**NTP (1995) 13-wk study reported kidney mineralization but not linear mineralization.

Figure 1-8. Exposure-response array for effects potentially associated with
a2u-globulin renal tubule nephropathy and tumors in male rats after oral
exposure to tert-butanol.
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                        1-24            DRAFT—DO NOT CITE OR QUOTE

-------
                                              Toxicological Review of tert-ButyI Alcohol
   • = exposures at which the endpoint was reported statistically significant by study authors
   D = exposures at which the endpoint was reported not statistically significant by study authors
Borghoffetal. (2001)-10d -
t Hyaline
droplet
size/number
NTP (1997) - 13 wks -
Identification
ofo2u-
globulinin Borghoff ctal. (2001)-10d
hyaline
droplets
Tubule cell Borghoff ct al. (2001) - lOd -
proliferation




•

BB



• H

100 1,000 10,t
Exposure Concentration (mg/m3)
Figure 1-9. Exposure-response array for effects potentially associated with
«2u-globulin renal tubule nephropathy and tumors in male rats after
inhalation exposure to tert-butanol.
   This document is a draft for review purposes only and does not constitute Agency policy,
                                       1-25            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   Question One: Is the a2uglobulin process occurring in male rats exposed to tert-butanol?
 2          (1) The first criterion to consider is whether hyaline droplets are increased in size and
 3   number in male rats. As noted above, the excessive accumulation of hyaline droplets can appear
 4   quickly, within 1 or 2 days, and persist throughout chronic exposures, although the severity begins
 5   to decline around 5 months [U.S. EPA. 1991a]. A statistically significant positive trend in the
 6   accumulation of large protein droplets with crystalloid protein structures was observed in kidneys
 7   of male rats exposed to inhalation concentrations of 758,1,364, and5,304mg/m3 tert-butanol for 6
 8   hr/day for 10 days [Borghoff etal.. 2001]. These droplets were small and minimally present in
 9   control male rats and were not observed in female rats. Similarly, data from the 13-week NTP oral
10   study [NTP. 1995: Takahashi etal.. 1993: Lindamoodetal.. 1992] demonstrated an increase in the
11   accumulation of hyaline droplets. The lowest dose of 230 mg/kg-day had minimal hyaline droplet
12   formation compared to controls, although the next three doses (490, 840, and 1,520 mg/kg-day]
13   had a higher accumulation of droplets with angular, crystalline structures that was similar in
14   incidence and severity among these dose groups. No droplets were observed in female rats or in
15   mice.
16          NTP [1997]. however, found no difference between the control and treatment groups
17   stained for hyaline droplet formation in male rats exposed to 0, 3,273, or 6,368 mg/m3 tert-butanol
18   via inhalation for 13 weeks; in fact, this study did not report any other lesions that could be
19   specifically associated with (X2u-globulin nephropathy in male rats. These results from NTP [1997],
20   which are inconsistent with the findings of both Borghoff etal.  [2001] and NTP [1995]. do not
21   appear to be due to differences in dose. Comparison of the oral and inhalation studies on the basis
22   of tert-butanol blood concentration (see Supplemental Information] showed that a 13-week
23   exposure in the range of the NTP [1995] doses of 490-840 mg/kg-day leads to the same average
24   blood concentration as 6-hr/day, 5 day/week inhalation exposures to 3,273-6,368 mg/m3. The
25   absence of similar histopathological findings in the 13-week inhalation NTP [1997] study compared
26   to those reported in the two oral studies is not understood, but might be indicative of the strength
27   of tert-butanol to induce, consistently, (X2u-globulin nephropathy. The results from the two other
28   studies [Borghoff etal.. 2001: NTP. 1995] indicate that hyaline  droplets  increase in size and number
29   in male rats following tert-butanol exposures. Therefore, the available data are sufficient to fulfill
30   the first criterion that hyaline droplets are increased in size and number in male rats.
31          (2] The second criterion to consider is whether the protein in the hyaline droplets in male
32   rats is (X2u-globulin. Accumulated hyaline droplets with an (X2u-globulin etiology can be confirmed
33   by using immunohistochemistry to identify the (X2u-globulin protein. Two short-term studies
34   measured (X2u-globulin immunoreactivity in the hyaline droplets of the renal proximal tubular
35   epithelium [Borghoff etal.. 2001: Williams and Borghoff. 2001]. Following 10 days of inhalation
36   exposure, Borghoff etal. [2001] did not observe an exposure-related increase in a2u-globulin using
37   immunohistochemical staining. When using an enzyme-linked immunosorbent assay [ELISA], a
38   more sensitive method of detecting a2u-globulin, however, a statistically significant positive

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-26           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review oftert-Butyl Alcohol

 1   correlation of (X2u-globulin concentration with dose of tert-butanol (determined by correlating with
 2   cell proliferation labeling indices) was observed, with accumulation of a2u-globulin protein
 3   statistically significant by pairwise comparison only in the highest dose group. No positive staining
 4   for a2u-globulin was observed in exposed female rats. In a follow-up study, Williams and Borghoff
 5   [2001] used a single gavage dose of 500 mg/kg [selected on the basis of results by NTP [1995] for
 6   induction of hyaline droplet accumulation], and reported a statistically significantly higher renal
 7   concentration of (X2u-globulin (by ELISA) in treated male rats than in controls 12 hours after
 8   exposure. Further, equilibrium dialysis methods determined that the binding of tert-butanol to
 9   (X2u-globulin was reversible. These data indicate the presence of azu-globulin in tert-butanol-treated
10   male rats, although requiring a more sensitive method of detection of azu-globulin than is typically
11   used could indicate that tert-butanol is not a strong inducer of (X2u-globulin accumulation.
12   Therefore, the available data are sufficient to fulfill the second criterion for (X2u-globulin present in
13   the hyaline droplets, but suggest weak induction of (X2u-globulin by tert-butanol.
14          (3) The third criterion considered is whether several (but not necessarily all) additional
15   events in the histopathological sequence associated with (X2u-globulin nephropathy appear in male
16   rats  in a manner consistent with the understanding of (X2u-globulin pathogenesis. Evidence of
17   cytotoxicity and single-cell necrosis of the tubule epithelium subsequent to the excessive
18   accumulation of hyaline droplets, with exfoliation of degenerate epithelial cells, should be
19   observable after five days of continuous exposure, peaking at 19 days [reviewed in U.S. EPA
20   (1991a)]. The formation and accumulation of granular casts from the exfoliated cellular debris
21   would follow, causing tubule dilation at the junction of the S3 (P3) segment of the proximal tubule
22   and the descending thin loop  of Henle, and the commencement of compensatory cell proliferation
23   within the S2 (P2) segment, both occurring after three weeks of continuous exposure. Following
24   chronic exposures, this regenerative proliferation could result in focal tubular hyperplasia, and
25   eventually progress to renal adenoma and carcinoma (Figure 1-7).
26          Several of these steps were observed following tert-butanol exposure in male rats, most
27   notably linear papillary mineralization and foci of tubular hyperplasia, consistent with the expected
28   disease progression. Some lack of consistency and dose-related concordance, however, was evident
29   across the remaining steps in the histopathological sequence. First, the accumulation of hyaline
30   droplets  and the concentrations of (X2u-globulin in the hyaline droplets at doses that induced
31   significant tumor formation in male rats were not significant  Next, necrosis or cytotoxicity was
32   absent and only precursors to granular casts at stages well within the expected timeframe of
33   detectability were present Finally, a 13-week inhalation study found no evidence of a2u-globulin
34   nephropathy (NTP. 1997). despite evaluating exposure concentrations predicted to result in similar
35   blood tert-butanol levels as for the 13-week oral study (NTP. 1995). which reported increases in
36   droplet accumulation and sustained regenerative tubule cell proliferation. A detailed evaluation
37   and analysis of all the evidence relevant to this criterion follows.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-27           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol


 1   Detailed evaluation of the available evidence supporting the third criterion
 2           Single cell death and exfoliation into the renal tubules was inconsistently observed. Single
 3   cell death or necrosis was not associated with tert-butanol exposure in male rat kidneys after 10 or
 4   13 weeks [Acharyaetal.,  1997: NTP, 1995]. Acharya et al. [1997] reported degeneration of renal
 5   tubules, one pathological consequence of single cell necrosis, however, in male rats exposed to tert-
 6   butanol in drinking water for 10 weeks. As renal tubule epithelial cell death and epithelial
 7   degeneration should occur as early as 5 days post exposure and persist for up to 48 weeks
 8   [Swenberg and Lehman-McKeeman. 1999: Short etal.. 1989]. the lack of consistency in these
 9   observations could be the result of both weak induction of a2u-globulin and a lack of later
10   examinations.

11           a.  Sustained regenerative cell proliferation also was not observed. Acharya et al. [1997]
12              did not observe tert-butanol-induced proliferation following 10 weeks of oral exposure,
13              but renal tubule proliferation was observed following another chemical exposure
14              (trichloroacetic acid] in the same study. Therefore, the inference is that tert-butanol
15              treatment did not induce regenerative tubule cell proliferation in male rats from this
16              study. Borghoff etal. [2001] reported a dose-related increase in epithelial cell
17              proliferation within the proximal tubule as measured by BrdU labeling indices in all
18              male rats exposed to tert-butanol via inhalation for 10 days. The study did not report
19              cytotoxicity, however, which,  combined with the early time point makes unlikely that
20              the cell proliferation was compensatory. NTP [1995] also observed increased cell
21              proliferation in the renal tubule epithelium following  13-week oral exposures in male
22              rats [only male rats were studied in the retrospective  analysis by Takahashi et al.
23              [1993] reported in NTP [1995]]. Proliferation was elevated at 840-1,520 mg/kg-day, a
24              range higher than the single 575-mg/kg-day dose eliciting no such proliferative effect
25              [Acharya et al.. 1997]. as described above. NTP [1995] reported, however, that no
26              necrosis was observed, suggesting the proliferation was not regenerative.

27           b.  Granular cast formation was not observed, although one study noted precursors to cast
28              formation. NTP [1995] did not observe the formation of granular casts or tubular
29              dilation; however, Hard etal.  [2011] reanalyzed the 13-week oral NTP data from male
30              rats treated with 0 or 1,520 mg/kg-day and identified precursors to  granular casts in
31              5/10 animals in the treated group. The significance of these granular cast precursors,
32              described as sporadic basophilic tubules containing cellular debris, is unknown, because
33              13 weeks of exposure is within the expected timeframe of frank formation and
34              accumulation of granular casts (>3 weeks]. Granular cast formation,  however, might not
35              be significantly elevated with weak inducers of (X2u-globulin [Short etal., 1986], which is
36              consistent with the reported difficulty in measuring (X2u-globulin in hyaline droplets
37              associated with tert-butanol exposure.

38           c.  Linear mineralization of tubules within the renal papillae was consistently observed in
39              male rats. This lesion typically appears at chronic time points, occurring after exposures
40              of 3 months up to 2 years [U.S. EPA. 1991a]. Consistent with this description, 2-year oral
41              exposure to tert-butanol induced a dose-related increase in linear mineralization, but
42              not following 13-week exposure [[NTP. 1995]: Table 1-2].
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-28           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1          d.  Renal tubule hyperplasia was observed in the only available 2-year study. Renal tubule
 2              hyperplasia is the preneoplastic lesion associated with a2u-globulin nephropathy in
 3              chronic exposures that leads to renal tubule tumors [U.S. EPA, 1991a]. A dose-related
 4              increase in renal tubule hyperplasia was observed in male rats following 2-year oral
 5              exposures [NTP. 1995]. By comparison, renal tubule hyperplasia was observed in only
 6              one high-dose female.

 7          The progression of histopathological lesions for (X2u-globulin nephropathy is predicated on
 8   the initial response of excessive hyaline droplet accumulation (containing (X2u-globulin) leading to
 9   cell necrosis and cytotoxicity, which in turn cause the accumulation of granular casts, linear
10   mineralization, and tubular hyperplasia. Therefore, observations of temporal and dose-response
11   concordance for these effects are informative for drawing conclusions on causation.
12          As mentioned above, most steps in the sequence of a2u-globulin nephropathy are observed
13   at the expected time points following exposure to tert-butanol. Accumulation of hyaline droplets
14   was observed early, at 12 hours following a single bolus exposure [Williams and Borghoff. 2001).
15   and at 10 days [Borghoff etal.. 2001] or 13 weeks [NTP. 1995] following continuous exposure;
16   (X2u-Globulin was identified as the protein in these droplets [Borghoff etal., 2001: Williams and
17   Borghoff. 2001]. Lack of necrosis and exfoliation might be due to the weak induction of (X2u-globulin
18   and a lack of later examinations. Granular cast formation was not reported by any of the available
19   studies, which could also indicate weak (X2u-globulin induction. Regenerative cell proliferation,
20   which was not observed, is discussed in more detail below. Observations of the subsequent linear
21   mineralization of tubules  and focal tubular hyperplasia fall within the expected timeframe of the
22   appearance of these lesions. Overall,  no explicit inconsistencies are present in the temporal
23   appearance of the histopathological lesions associated with (X2u-globulin nephropathy; however, the
24   dataset would be bolstered by measurements at additional time points to lend strength to the MOA
25   evaluation.
26          Inconsistencies do occur in the dose-response among lesions  associated with the
27   a2u-globulin nephropathy progression. Hyaline droplets were induced in the proximal tubule of all
28   surviving male rats in the 13-week NTP oral study [NTP. 1995: Takahashietal.. 1993: Lindamood
29   etal.. 1992]. although the incidence at the lowest dose was minimal, while the incidence at the
30   three higher doses was more prominent These results are discordant with the tumor results, given
31   that all treated groups of male rats in the NTP 2-year oral bioassay had increased kidney tumor
32   incidence, including the lowest dose of 90 mg/kg-day [according to the reanalysis by Hard et al.
33   [2011]]. This lowest dose was less than the 230 mg/kg-day in the 13-week oral study that had only
34   minimal hyaline droplet formation. Furthermore, although the incidence of renal tubule
35   hyperplasia had a dose-related increase [NTP. 1995]. a corresponding dose-related increase in the
36   severity of tubular hyperplasia did not result Severity of tubule hyperplasia was increased only at
37   the highest dose, which was not consistent with renal tumor incidence.
38          Although the histopathological sequence has data gaps, such as the lack of observable
39   necrosis or cytotoxicity or granular casts at stages within the timeframe of detectability, overall, a

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-29           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol

 1   sufficient number of steps (e.g., linear papillary mineralization, foci of tubular hyperplasia) were
 2   observed to fulfill the third criterion.

 3   Summary and Conclusions for Question One:
 4          Oral exposure to male F344 rats resulted in an increased incidence of renal tubule tumors in
 5   a 2-year oral bioassay [Hardetal., 2011: NTP, 1995]. Several histopathological observations in
 6   exposed male rats were consistent with an a2u-globulin MOA. This evidence includes the increased
 7   size and number of hyaline droplets and the accumulated (X2u-globulin protein in the hyaline
 8   droplets. Additionally, several subsequent steps in the histopathological sequence were observed.
 9   Overall, available data are sufficient for all three required criteria, suggesting that the a2u-globulin
10   process is operative. Although the evidence indicates a role for a2u-globulin accumulation in the
11   etiology of kidney tumors induced by exposure to tert-butanol in male rats, that tert-butanol is a
12   weak inducer of a2u-globulin is plausible, considering the limited histopathological observations
13   and uncertainty regarding the temporal and dose concordance of the lesions.

14   Question Two: Are the renal effects in male rats exposed to tert-butanol solely due to the a2U-globulin
15   process?
16          If the a2u-globulin process is operative, U.S. EPA [1991a] identifies a second question that
17   must be answered regarding whether the renal effects are (a) solely due to the a2u-globulin
18   process, (b) a combination of the a2u-globulin process and other carcinogenic processes, or (c)
19   primarily due to other processes. U.S. EPA[1991a] states that additional data can help inform
20   whether the a2u-globulin process is the sole contributor to renal tubule tumor development in male
21   rats. These additional considerations are highlighted and discussed, where possible, in detail below.
22          Hypothesis-testing of the a2U-globulin sequence of effects and structure-activity relationships
23   that might suggest the chemical belongs in a different class of suspected carcinogens: No data are
24   available to evaluate these considerations.
25          Biochemical information regarding binding of the chemical to the a2U-globulin protein:
26   Williams and Borghoff [2001] report that tert-butanol reversibly and noncovalently binds to
27   a2u-globulin in the kidneys of male rats. This provides additional support to the involvement of the
28   a2u-globulin process.
29          Presence of sustained cell replication in the S2 (P2] segment of the renal tubule at doses
30   used in the cancer bioassay and a dose-related increase in hyperplasia of the renal tubule:
31   Sustained cell division in the proximal tubule of the male rat is consistent with, although not
32   specific to, the a2u-globulin process. Cell proliferation was observed in two studies [13-week, NTP
33   [1995] and 10-day, Borghoff etal. [2001]] but whether the proliferation was compensatory is
34   unknown, as cytotoxicity was not observed in these studies. Although the data do not support
35   sustained cell division occurring subsequent to cytotoxic cell death, renal tubule hyperplasia in
36   male rats was reported after 2 years of exposure [NTP, 1995]. Thus, although some evidence of
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-30           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol

 1   sustained cell replication is available, it does not specifically support (X2u-globulin protein
 2   accumulation.
 3          Covalent binding to DMA or other macromolecules, suggesting another process leading to
 4   tumors andgenotoxicity (a2u-globulin-inducers are essentially nongenotoxic): One study [Yuanetal.,
 5   2007] observed a dose-related increase in tert-butanol-DNA adducts in liver, kidney, and lung of
 6   mice administered a single low dose of tert-butanol (<1 mg/kg) in saline via gavage (see Appendix
 7   B.3 in Supplemental Information for further details). An extremely sensitive method of detection
 8   was used (accelerator mass spectrometry), but the DNA adduct species were not identified, and no
 9   validation of these results has been identified in the literature. The few studies available to assess
10   the genotoxic potential of tert-butanol primarily are negative, although a few studies report DNA
11   damage induced by oxidative stress. DNA damage induced by oxidative stress is consistent with the
12   decreased levels of glutathione in male rat kidneys reported by Acharyaetal.  (1995) after 10 weeks
13   of tert-butanol exposure. This type of genetic damage would not necessarily preclude a role for
14   (X2u-globulin, but not enough information is available to determine whether oxidative stress could
15   initiate or promote kidney tumors in concert with (X2u-globulin accumulation in male rat kidneys.
16          Nephrotoxicity not associated with the a2u-globulin process or CPN, suggesting the possibility
17   of other processes leading to renal tubule nephrotoxicity and carcinogenicity: Nephropathy reported
18   in the 13-week oral and inhalation and 2-year oral studies was considered CPN,  but these effects
19   were exacerbated by treatment with tert-butanol. At 13 weeks (NTP. 1997.1995] and 2 years (NTP.
20   1995], oral and inhalation exposure increased the severity of nephropathy in male rats (NTP,
21   1995]. Similarly, the severity of nephropathy was increased in females at 2 years, but only the
22   incidence of nephropathy was increased in females following a 13-week oral exposure (NTP. 1995].
23   Increased incidences of suppurative inflammation and kidney  transitional epithelial hyperplasia
24   were observed in female rats orally exposed to tert-butanol for 2 years. Although NTP (1995]
25   characterized these endpoints as associated with CPN, the low background incidence in the controls
26   combined with the dose-related increase in incidences indicate that these effects were not related
27   to an age-associated, spontaneous induction of nephropathy. At 2 years, the male rats also exhibited
28   dose-related increases in focal mineralization and transitional epithelial hyperplasia, although the
29   background incidence in the controls was high (i.e., approximately 50%] (NTP. 1995]. Neither
30   endpoint in males can be attributed to CPN or (X2u-globulin.
31          Kidney weights also were increased in male and female rats in the 13-week oral and
32   inhalation evaluations (NTP. 1997.1995] and 15-month oral evaluation (NTP. 1995]. The dose-
33   related increases  observed in both male and female rats suggest that the kidney weight changes are
34   indicative of treatment-related molecular processes primarily unrelated to either (X2u-globulin
35   protein accumulation or CPN. The exacerbation of CPN and the appearance of kidney effects in
36   female (i.e., suppurative inflammation, transitional epithelial hyperplasia] and male rats (i.e., focal
37   mineralization, transitional epithelial hyperplasia] that are not attributed to CPN or (X2u-globulin
38   indicate that tert-butanol induces renal tubule nephrotoxicity partially independently of
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-31           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol

 1   (X2u-globulin. The evidence that other processes might be responsible for the renal tubule
 2   nephrotoxicity thereby decreases the likelihood that a2u-globulin accumulation is solely
 3   responsible for the renal tubule tumors.
 4          Positive tubule tumor responses in female rats and other species implying that a2u-globulin-
 5   related processes alone do not account for the renal tubule tumor response: No increase in renal
 6   tubule tumor incidence was reported in tert-butanol-exposed female rats or mice compared with
 7   concurrent controls. Renal tubule tumors were observed only in male rats, providing support for an
 8   (X2u-globulin process in tumor development.

 9   Summary and Conclusions for Question Two:
10          Although the evidence suggests that tert-butanol induces a2u-globulin nephropathy, the
11   data indicate that tert-butanol is a weak inducer of a2u-globulin and that this process is not solely
12   responsible for the renal tubule nephropathy and carcinogenicity observed in male rats. The lack of
13   compensatory cell proliferation in male rats and evidence of nephrotoxicity in female rats suggest
14   that other processes, in addition to the (X2u-globulin process, are operating. Furthermore, the
15   accumulation of hyaline droplets and the induction of renal tubule hyperplasia were affected at
16   higher doses compared to those inducing renal tubule tumors. Collectively, these data suggest that
17   tert-butanol induces the (X2u-globulin pathway at high doses (>420 mg/kg-day), which results in
18   tumor formation. Other, unknown pathways, however, could be operative at lower doses
19   (<420 mg/kg-day), which contribute to renal tumor induction.

20   b)  Chronic Progressive Nephropathy and Renal Carcinogenicity
21          There is scientific disagreement regarding the extent to which CPN can be characterized as
22   a carcinogenic MOA suitable for analysis under the EPA's cancer guidelines. Proponents of CPN as
23   an MOA have developed an evolving series of empirical criteria for attributing renal tubule tumors
24   to CPN. Hard and Khan [2004] proposed criteria for concluding that a chemical is associated with
25   renal tubule tumors through an interaction with CPN. Hard etal. [2013] slightly revised and
26   restated their criteria for considering exacerbation of CPN as an MOA for renal tubule tumors in
27   rats. Table 1-5 lists these sets of proposed empirical criteria for attributing renal tubule tumors to
28   CPN.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-32           DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review of tert-ButyI Alcohol
             Table 1-5. Proposed empirical criteria for attributing renal tumors to CPN
          •   First and foremost, the chemical must have been
             shown to exacerbate CPN to very advanced
             stages of severity, especially end-stage kidney
             disease, in comparison to control rats in a 2-year
             carcinogenicity study.

          •   The tumors should occur in very low incidence
             and, for the most part, be minimal grade lesions
             conforming to small adenomas or lesions
             borderline between atypical tubule hyperplasia
             (ATM) and adenoma.

          •   Such tumors should be associated only with the
             highest grades of CPN severity.

          •   The tumors and any precursor foci of ATM must
             be restricted to CPN-affected parenchyma and
             are usually observed only toward the end of the
             2-year studies.

          •   Careful microscopic examination of renal
             parenchyma not involved in the CPN process
             should reveal no evidence of compound-induced
             cellular injury or other changes that would
             suggest alternative modes of action.
          Source: Hard and Khan (2004)
•   Lack of genotoxic activity based on overall
    evaluation of in vitro and in vivo data.

•   Tumor incidence is low, usually <10%.

•   Tumors are found toward the end of 2-year
    studies.

•   Lesions are usually ATM or adenomas (carcinomas
    can occasionally occur).

•   Chemical exacerbates CPN to most advanced
    stages, including end-stage kidney disease.

•   ATM and tumors occur in rats with advanced CPN
    and in CPN-affected tissue.

•   Absence of cytotoxicity in CPN-unaffected
    tubules, in rats with lower grades of CPN, and in
    subchronic studies.

Source: Hard et al. (2013)
 2           Hardetal. [2013] maintain knowing the detailed etiology or underlying mechanism for CPN
 3    is not necessary. Instead, identifying increased CPN with its associated increase in tubule cell
 4    proliferation as the key event is adequate. Nonetheless, Hardetal. [2013] also postulated a
 5    sequence of key events for renal tumorigenesis involving exacerbation of CPN:

 6       •   Exposure to chemical (usually at high concentrations];

 7       •   Metabolic activation (if necessary];

 8       •   Exacerbated CPN, including increased number of rats with end-stage renal disease;

 9       •   Increased tubule cell proliferation because more kidney is damaged due to CPN
10           exacerbation;

11       •   Hyperplasia; and

12       •   Adenoma (infrequently carcinoma].

13           In contrast to Hardetal. [2013]: Hard and Khan [2004].Melnicketal. [2013]: Melnick et al.
14    [2012] concluded, based on an analysis of 60 NTP studies, no consistent association exists between
15    exacerbated CPN and the incidence of renal tubule tumors in rats. Without a consistent association

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                    1-33            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   and an understanding of its key events, they maintain that determining the human relevance of
 2   processes that might be occurring in rats is not possible. An earlier analysis of 2 8 NTP studies
 3   [Seely etal.. 2002] found a slight but statistically significant increase in CPN severity in animals
 4   with renal tubule tumors, without determining that this relationship is causal. They suggested that
 5   the number of tumors due to chemically exacerbated CPN would be few.

 6   Evaluation of the MOA Proposed by Hard etal. (2013)
 7          Setting aside the question of whether CPN is [Hard etal.. 2013: Hard and Khan. 2004] or is
 8   not [Melnick et al.. 2013: Melnicketal.. 2012] an MOA suitable for analysis, this section provides an
 9   analysis of the mechanistic data pertinent to CPN. EPA's cancer guidelines [U.S. EPA, 2005a] define
10   a framework for judging whether available data support a hypothesized MOA; the analysis in this
11   section follows the structure presented in the cancer guidelines.
12          Description of the hypothesized MOA. Under the EPA framework, toxicokinetic studies are
13   important for identifying the active agent, but toxicokinetic events per se are not key events of an
14   MOA. Thus, the EPA analysis of the MOA proposed by Hard etal. [2013] begins with
15   (1] exacerbated CPN, including increased number of rats with end-stage renal disease, and
16   proceeds via (2] increased tubule cell proliferation, (3] hyperplasia, and (4] adenoma, or
17   infrequently, carcinoma.
18          Strength, consistency, specificity of association. The relationship between exacerbated CPN
19   and renal tumors is moderate to strong in male rats in the NTP [1995] study. According to the NTP
20   [1995] analysis, the mean CPN grades (same as "severity of nephropathy" reported by NTP]
21   presented on a scale 1-4 for male rats with renal tumors were 3.5, 3.6, 3.7, and 3.4 at doses 0,1.25,
22   2.5, and 5 mg/mL. The mean CPN grades for male rats without renal tumors were 2.9, 2.8, 2.8,  and
23   3.2 for the same dose groups. The  reanalysis of the NTP data by Hard etal. [2011] yielded similar
24   numbers. The relationship between CPN and renal tumors, however, is neither consistent nor
25   specific in the NTP [1995] study: No female rats developed renal tumors regardless of the presence
26   of relatively low-grade or relatively high-grade CPN. For example, in female rats surviving more
27   than 700 days, the mean CPN grades were 1.7 and 3.2 at doses of 0  and 10 mg/mL, respectively, but
28   no tumors developed in either group.
29          Dose-response concordance. The dose-response relationships for CPN, renal tubule
30   hyperplasia, and renal tubule tumors somewhat differ. According to the NTP [1995] analysis, at
31   doses of 0,1.25, 2.5, and 5 mg/mL, the mean CPN grades for all male rats were 3.0, 3.1, 3.1, and 3.3;
32   the incidences of renal tubule hyperplasia (standard and extended evaluation combined] were
33   14/50, 20/50,17/50, and 25/50; and the incidences of renal tubule adenomas or carcinomas were
34   8/50,13/50,19/50, and 13/50 (Table 1-3]. The reanalysis by Hard etal. [2011]  reported similar
35   tumor incidences (4/50,13/50,18/50, and 12/50], except that four fewer rats in the controls  and
36   one fewer rat in the group exposed to 2.5 mg/mL had tumors. The lower control  incidence
37   observed in this reanalysis accentuates the differences in these dose-response relationships. In
38   examining the various lesions at the mid-dose—the dose with the greatest increase  in renal tubule
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-34           DRAFT—DO NOT  CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   tumors in male rats—a minor increase (14/50 in controls versus 17/50 in the mid-dose group) in
 2   renal tubule hyperplasia incidence was observed, with a marginal change in CPN severity (i.e.,
 3   group average of 3.0 to 3.1). That a minor increase in hyperplasia and marginal increase in CPN
 4   severity would be associated with significant tumor induction seems inconsistent Furthermore,
 5   CPN severity is nearly as great in the female rats, yet no females developed tumors, as noted above.
 6          Temporal relationship. The severity of CPN progressed over time. According to the NTP
 7   (1995) analysis, the mean CPN grades in the 13-week study of male rats were 1.0,1.6, 2.6, 2.7, 2.6,
 8   and 1.1 at doses of 0, 2.5, 5,10, 20, and 40 mg/mL. At the 15-month interim evaluation of the 2-year
 9   study, the mean CPN grades were 2.4, 2.8, 2.7, and 2.6 at doses of 0,1.25, 2.5, and 5 mg/mL and at
10   2 years, increased to 3.0, 3.1, 3.1, and 3.3. Similarly, the severity of neoplastic lesions increased at
11   the end of life. At the 15-month interim evaluation, only two rats had developed renal tubule
12   hyperplasia and one other had a renal tubule adenoma; at 2 years, the incidences of these two
13   lesions were much higher in all dose groups (see previous paragraph). These results are consistent
14   with CPN as an age-related disease and with hyperplasia and tumors appearing near the end of life.
15          Biological plausibility and coherence. In general, the relationship between exacerbated CPN
16   and renal tubule tumors in male rats appears plausible and coherent. Some patterns in the dose-
17   response relationships for CPN, hyperplasia, and tumors are discrepant. Perhaps more importantly,
18   the patterns also are discrepant for the relationships between CPN grades and renal tubule tumors
19   in male and female rats. In addition, the increased incidences in renal tubule tumors in all exposed
20   male rats exceed the 10% criterion proposed by Hardetal. (2013) (Table 1-5), even more so when
21   making comparisons with the lower control tumor incidence from the Hardetal. (2011) reanalysis.

22   Conclusions about the hypothesized CPN-related MOA
23          As recommended by EPA's cancer guidelines (U.S.  EPA, 2005a), conclusions about the
24   hypothesized MOA can be clarified by answering three questions presented below.
25          (a) Is the hypothesized MOA sufficiently supported in the test animals? Exacerbated  CPN
26   leading to renal tubule tumors in male rats late in life appears to have some support There is lack
27   of consistency, however, between males and females and in the dose-response relationships
28   between CPN, hyperplasia, and adenomas. These inconsistencies make difficult attributing all renal
29   tumors to either CPN or to (X2u-globulin-related nephropathy (see previous section on a2U-globulin),
30   raising the likelihood of another, yet unspecified MOA.
31          (b) Is the hypothesized MOA relevant to humans? There is scientific disagreement on this
32   question. Hard etal. (2013): Hard et al. (2009)cite several differences in pathology between rat CPN
33   and human nephropathies in their arguments that CPN-related renal tumors in rats are not relevant
34   to humans.  On the other hand, Melnicketal. (2013): Melnicketal. (2012) argue that the etiology of
35   CPN and the mechanisms for its exacerbation by chemicals are unknown and fail to meet
36   fundamental principles for defining an MOA and for evaluating human relevance. This issue is
37   unresolved.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-35          DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1          (c) Which populations or lifestages can be particularly susceptible to the hypothesized MOA?
 2   There are no indications of a human population or lifestage that is especially susceptible to tumors
 3   induced through exacerbated CPN.
 4          In summary, considering discrepant patterns in the dose-response relationships for CPN,
 5   hyperplasia, and renal tubule tumors and the lack of relationships between CPN grades and renal
 6   tubule tumors in male and female rats, together with the lack of a generally accepted MOA for CPN,
 7   the renal tubule tumors in rats cannot be attributed to CPN.
 8          This position can be reconciled with that of Melnicketal. [2013]: Melnicketal. [2012], who
 9   argued against dismissing renal tubule tumors in rats that can be related to exacerbated CPN. It also
10   can be reconciled with Hardetal. [2013]. who, while maintaining that these tumors are not
11   relevant to humans, also allow that there is no generally accepted MOA for CPN akin to that for ct2U-
12   globulin-related nephropathy. Hardetal. [2013] made this statement after reporting on the
13   collective experience of national and international health agencies worldwide with the use of CPN
14   as an MOA. Of 21 substances that exacerbated CPN and caused renal tumors, most were multisite
15   carcinogens, and other tumor sites contributed to the evaluations. Only two assessments explicitly
16   considered CPN as a renal tumor mechanism. One was the assessment of ethylbenzene by the
17   German Federal Institute  for Occupational Safety and Health, in which the agency concluded that
18   the kidney tumors were associated with the high, strain-specific incidence of CPN that is unknown
19   for humans [as discussed in Hardetal. [2013]]. The other was the IRIS assessment of
20   tetrahydrofuran, for which EPA found the evidence insufficient to conclude that the kidney tumors
21   are mediated solely by the hypothesized MOAs [U.S. EPA. 2012d]. Hardetal. [2013] attributed
22   these different conclusions to either different data for the two chemicals or the lack of a generally
23   accepted MOA akin to (X2u-globulin-related nephropathy.
24          Relevant to this last point, IARC [1999] developed a consensus statement that listed
25   considerations for evaluating (X2u-globulin-related nephropathy in rats, which was based on the
26   work of 22 scientists, including three who were co-authors of Hardetal. [2013]  and two who were
27   co-authors of Melnicketal. [2013]: Melnicketal. [2012]. A similar broad-based  consensus that
28   defines a sequence of key events for exacerbated CPN, distinguishes it more clearly from a2u-
29   globulin-related nephropathy, and evaluates its relevance to humans would be helpful in advancing
30   the understanding of these issues.

31   Overall Con elusions on MOA for Kidn ey Effects
32          tert-Butanol increases (X2uglobulin deposition and hyaline droplet accumulation in male rat
33   kidneys, as well as several of the subsequent steps in that pathological sequence. These data
34   provide sufficient evidence (albeit minimal] that the (X2uglobulin process is operating, although
35   based on further analysis this chemical appears to be a weak inducer of a2uglobulin-nephropathy
36   and this induction is not the sole contributor to renal tubule nephropathy and carcinogenicity. CPN
37   and the exacerbation of CPN (likely due to both a2u-globulin and tert-butanol] play a role in renal
38   tubule nephropathy. Although CPN was indicated in the induction of renal tubule nephropathy, the
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-36           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol

 1   available evidence indicates that it does not induce the renal tubule tumors in male rats.
 2   Additionally, several endpoints indicate renal tubule nephrotoxicity and increased kidney weights
 3   related to tert-butanol exposure cannot be explained by the a2uglobulin or CPN processes.
 4   Collectively, the evidence indicates other, unknown processes contribute to renal tubule
 5   nephrotoxicity and carcinogenicity.

 6   Integration of kidney effects
 7          Kidney effects (increases in nephropathy, severity of nephropathy, hyaline droplets, linear
 8   mineralization, suppurative inflammation, transitional epithelial hyperplasia, mineralization, and
 9   kidney weight) were observed, predominantly in male and female rats across the multiple tert-
10   butanol studies. The available evidence indicates that multiple processes induce the noncancer
11   kidney effects. The group of lesions generally reported as "nephropathy," is related to CPN. Because
12   this disease is considered to be spontaneous and age-related in rats, the endpoints associated with
13   CPN would not be relevant to humans for purposes  of hazard identification. Additionally, two
14   endpoints in male rats (hyaline droplets, linear mineralization) are components of the (X2u-globulin
15   process. U.S. EPA(1991a) states that if the (X2u-globulin process is occurring in male rats, the renal
16   tubule effects associated with this process in male rats would not be relevant to humans for
17   purposes of hazard identification.  In cases such as these, the characterization of human health
18   hazard for noncancer kidney toxicity would rely on  effects not specifically associated with CPN or
19   the a2u-globulin-process in male rats.
20          Several other noncancer endpoints resulted from tert-butanol exposure and are appropriate
21   for consideration of a kidney hazard, specifically: suppurative inflammation in female rats,
22   transitional epithelial hyperplasia in male and female  rats, severity of nephropathy in male and
23   female rats, incidence of nephropathy in female rats, incidence of mineralization in male rats, and
24   increased kidney weights in rats but not mice. Based on dose-related increases in these noncancer
25   endpoints in rats, kidney effects are a potential human hazard of tert-butanol exposure. The hazard
26   and dose-response conclusions regarding these noncancer endpoints associated with tert-butanol
27   exposure are discussed further in  Section 1.3.1.
28          The carcinogenic effects observed following tert-butanol exposure include increased
29   incidences of renal tubule hyperplasia (considered a preneoplastic effect) and tumors in male rats.
30   EPA concluded that the three criteria were met to indicate that an (X2u-globulin process is operating.
31   Because renal tubule tumors in male rats did not arise solely due to the (X2u-globulin process and
32   some of the tumors are attributable to other carcinogenic processes, such tumors remain relevant
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-37           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol

 1   for purposes of hazard identification U.S. EPA (1991a).3 The hazard and dose-response conclusions
 2   regarding the renal tubule hyperplasia and tumors associated with tert-butanol exposure are
 3   further discussed as partof the overall weight of evidence for carcinogenicity in Section 1.3.2.

 4   1.2.2.  Thyroid Effects

 5   Synthesis of Effects in Thyroid
 6           The database on thyroid effects following tert-butanol exposure contains no human data,
 7   two oral subchronic and two oral chronic studies (one of each duration in rats and in mice) [NTP.
 8   1995], and two inhalation subchronic studies (one in rats and one in mice) [NTP, 1997}. Studies
 9   employing short-term and acute exposures that examined thyroid effects are not included in the
10   evidence table; they are discussed, however, in the text if they provide data informative of MOA or
11   hazard identification. No gross thyroid effects were reported in the 13-week evaluations of mice or
12   rats following oral or inhalation exposure [NTP. 1997.1995). and therefore subchronic studies
13   were not included in the evidence table. The two available chronic studies are arranged in the
14   evidence table by effect and then by species. The design, conduct, and reporting of each study were
15   reviewed, each study was considered adequate to  provide information pertinent to this assessment
16   (Table 1-6 and  Figure 1-10).
17           Thyroid effects, specifically follicular cell hyperplasia and adenomas, were observed in mice
18   of both sexes after 2 years of oral exposure via drinking water (NTP, 1995). NTP (1995) noted that
19   "[proliferation of thyroid gland follicular cells is generally considered to follow a progression from
20   hyperplasia to adenoma and carcinoma." Similarly, EPA considered the thyroid follicular cell
21   hyperplasia to be a preneoplastic effect associated with the thyroid tumors. Both male and female
22   mice  exhibited  a dose-related increase in the incidence of hyperplasia, and the average severity
23   across all dose groups was minimal to mild with scores ranging from 1.2 to 2.2 (out of 4).  Increased
24   incidence of adenomas were also observed in the tert-butanol-treated mice, with the only
25   carcinoma observed in high-dose males. No treatment-related thyroid effects were reported in rats
26   of either sex following 2 years of oral exposure (NTP. 1995).
27           Although the tumor response in male mice showed a statistically significant increasing
28   trend (Cochran-Armitage trend test, p = 0.041) (analysis performed by EPA using the mortality-
29   adjusted rates), the response was non-monotonic, with a slightly lower response at the high-dose
     3 When the 
-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   level than at the mid-dose level. The reason for the non-monotonicity is unclear, although it could
 2   be related to the increased mortality in the high-dose group (17/60 animals survived compared
 3   with 27/60 animals in the control group). The decreased survival of male mice might have affected
 4   the thyroid tumor incidences because animals could have died before tumors could develop. High
 5   mortality in the high-dose group occurred before tumors appeared; about 40% of the high-dose
 6   males died before the first tumor (a carcinoma) appeared in this group at week 83. By comparison,
 7   only ~10% of the control group had died by this time, and the single tumor in the control group
 8   was observed at study termination. Mortality in the exposed female mice was similar to controls.
 9
10
Table 1-6. Evidence pertaining to thyroid effects in animals following oral
exposure to tert-butanol
Reference and study design
Follicular cell hyperplasia
NTP (1995)
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at 15
months)
Drinking water (0, 1.25, 2.5, 5, or 10
mg/mL)
M: 0, 90, 200, or 420a mg/kg-d
F: 0, 180, 330, or 650a mg/kg-d
2 years

NTP (1995)
B6C3Fi mouse; 60/sex/treatment
Drinking water (0, 5, 10, or 20 mg/mL)
M: 0, 540, 1,040, or 2,070a mg/kg-d
F: 0, 510, 1,020, or 2,110 mg/kg-d
2 years



Follicular cell tumors
NTP (1995)


Incidence15
Males

Dose
(mg/kg-d)
0

90
200
420a
Incidence (severity)
Males

Dose
(mg/kg-d)
0
540
1,040
2,070 a

Incidence15
Result




Follicular cell
hyperplasia
3/50

0/49
0/50
0/50



Follicular cell
hyperplasia
5/60 (1.2)
18/59* (1.6)
15/59* (1.4)
18/57* (2.1)


S


Females

Dose
(mg/kg-d)
0

180
330
650 a

Females

Dose
(mg/kg-d)
0
510
1,020
2,110







Follicular cell
hyperplasia
0/50

0/50
0/50
0/50



Follicular cell
hyperplasia
19/58 (1.8)
28/60 (1.9)
33/59* (1.7)
47/59* (2.2)


               This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-39           DRAFT—DO NOT CITE OR QUOTE

-------
                                                                  Toxicological Review of tert-ButyI Alcohol
Reference and study design
F344/N rat; 60/sex/treatment
(10/sex/treatment evaluated at 15
months)
Drinking water (0, 1.25, 2.5, 5, or 10
mg/mL)
M: 0, 90, 200, or 420a mg/kg-d
F: 0, 180, 330, or 650a mg/kg-d
2 years






NTP (1995)
B6C3Fi mouse; 60/sex/treatment
Drinking water (0, 5, 10, or 20 mg/mL)
M: 0, 540, 1,040, or 2,070a mg/kg-d
F: 0, 510, 1,020, or 2,110 mg/kg-d
2 years



A
^





Results

Dose (mg/kg-d'
Male

0
90

200
420 a
Female
0
180
330
650 a
Incidence

Follicular cell
adenoma


2/50
0/49

0/50
0/50

1/50
0/50
1/50
0/50


Dose Follicular cell Follicular cell
(mg/kg-d)
Male
0
540
1,040
2,070 a
Female
0
510
1,020
2,110
adenoma carcinoma

1/60 0/60
0/59 0/59
4/59 0/59
1/57 17^7

2/58 0/58
3/60 0/60
2/59 0/59
9/59* 0/59
Follicular cell
carcinoma


2/50
0/49

0/50
0/50

1/50
0/50
1/50
0/50

Follicular cell
adenoma or
carcinoma
(mortality
adjusted rates)-^

1/60 (3.6%)
0/59 (0.0%)
4/59 (10.1%)
2/57 (8.7%)

2/58 (5.6%)
3/60 (8.6%)
2/59 (4.9%)
9/59* (19.6%)
















Animals
surviving to
study
termination

27/60
36/60
34/60
17/60

36/60
35/60
41/60
42/60
1    aThere was a significant decrease in survival in the high-dose group.
2    bResults do not include the animals sacrificed at 15 months.
3    cMortality-adjusted rates were not calculated by study authors for follicular cell carcinoma. The mortality-adjusted rates for the
4     incidence of adenomas are the same as the combined rates, with the exception of the male high-dose group, where the rate
5     for adenomas alone was 5.9%.
6    dCochran-Armitage trend test was applied to mortality-adjusted thyroid tumor incidences, by applying the NTP adjusted rates
7     to the observed numbers of tumors to estimate the effective number at risk in each group. For male mice, p = 0.041; for
8     female mice, p = 0.028.* Statistically significant p < 0.05 as determined by the study authors.

9    Note: Conversions from drinking water concentrations to mg/kg-d performed by study authors.
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                           1-40             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
2
3
4
                • = exposures at which the endpoint was reported statistically significant by study authors
                D = exposures at which the endpoint was reported not statistically significant by study authors
                   Hyperplasia; M mouse -
                    Hyperplasia; F mouse -
         NONCANCER
                      Hyperplasia; M ral
                       Hyperplasia; F ral -
                     Adenoma; M mouse
                      Adenoma; F mouse
         CANCER
                        Adenoma; M rat
                         Adenoma; F rat
                                      10
            Source: NTP (1995)
                                            D
-B	B
                                                 Q	B	B
                                                          B	\1
                                                                     B	B
                                                             -B	B
                                                             B	B	B
                                            100              1,000
                                             Dose (mg/kg-day)
Figure 1-10. Exposure-response array of thyroid follicular cell effects
following chronic oral exposure to tert-butanol. (Note: Only one carcinoma
was observed in male mice at the high-dose group.)
                             10,000
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-41           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   Mode of Action Analysis—Thyroid Effects
 2          The MOA responsible for tert-butanol-induced thyroid effects has not been the subject of
 3   much study. One hypothesis is that tert-butanol increases liver metabolism of thyroid hormones,
 4   triggering a compensatory increase in pituitary thyroid-stimulating hormone (TSH) production.
 5   Such sustained increases in TSH could induce elevated thyroid follicular cell proliferation and
 6   hyperplasia and lead to follicular cell adenoma and carcinoma, that, an antithyroid MOA, as
 7   identified in U.S. EPA's guidance on the assessment of thyroid follicular cell tumors [U.S. EPA.
 8   1998a].
 9          To determine if the thyroid follicular cell tumors result from a chemically induced
10   antithyroid MOA, U.S. EPA[1998a] requires that the available database demonstrate: (1) increases
11   in thyroid cell growth, (2) thyroid  and pituitary hormone changes consistent with the antithyroid
12   MOA, (3) site(s) of the antithyroid action, (4) dose correlation among the various effects, and (5)
13   reversibility of effects in the early  stages of disruption. The available evidence pertaining to each of
14   these aspects of antithyroid activity following tert-butanol exposure is discussed below.

15   1] Increases in cell growth [required]
16          U.S. EPA[1998a] considers increased absolute or relative thyroid weights, histological
17   indicators of cellular hypertrophy  and hyperplasia, DNA labeling, and other measurements (e.g., Ki-
18   67 or proliferating cell nuclear antigen expression) to be indicators of increased cell growth. Only a
19   few studies [NTP. 1997.1995] have evaluated the thyroid by routine histological examination
20   following tert-butanol exposure, and none investigated specific molecular endpoints. None of the
21   available long-term studies measured thyroid weight in mice, likely due to the technical limitations
22   involved, and no thyroid effects were attributed to tert-butanol exposure in rats treated up to 2
23   years [NTP. 1997.1995]. Although the  short-term female mouse study by Blancketal. [2010]
24   stated that thyroids were weighed, no results were reported.
25          An increase in thyroid follicular cell hyperplasia was observed in both female and male mice
26   after a 2-year drinking water exposure to tertbutanol- [NTP. 1995]. The increase was dose
27   dependent in female mice with a slight increase in severity in the highest dose, while male mice
28   experienced a similar magnitude of hyperplasia induction at all doses, with increased severity at
29   the highest dose [NTP. 1995]. Thyroid follicular cell hyperplasia was not observed in any mouse
30   study with less than 2 years of exposure: no treatment-related histological alterations in the thyroid
31   of tert-butanol-treated (2 or 20 mg/mL] female mice after 3 or 14 days of drinking water exposure
32   [Blancketal.. 2010] were reported, in male or female mice after 13 weeks of drinking water
33   exposure [NTP. 1995] or in male or female mice following 18-day or 13-week inhalation studies
34   [NTP. 1997]. The observation of increased hyperplasia in male and female mice  after 2 years of
35   exposure is sufficient evidence to support increased thyroid cell  growth.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-42          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   2] Changes in thyroid and relevant pituitary hormones [required]
 2          Evidence of hormonal changes, including decreases in thyroxine (14) and triiodothyronine
 3   [Ts] and increases in TSH, are required to demonstrate a disruption in the thyroid-pituitary
 4   signaling axis [U.S. EPA, 1998a]. Blancketal. [2010] evaluated serum thyroid hormones in mice
 5   after 3 or 14 days of exposure to tert-butanol. No tert-butanol-related effects were observed in Ts,
 6   T4, or TSH levels after 3 days, and although both T3 and T4 levels were significantly decreased
 7   approximately 10-20% after 14 days of treatment with tert-butanol, TSH levels remained
 8   unaffected. Similar results were reported with the positive control [phenobarbital]. The limited
 9   evidence available from this single study suggests that although T3 and T4 levels were decreased
10   after 14 days, this perturbation likely was not in excess of the range  of homeostatic regulation in
11   female B6C3Fi mice and thus not likely to induce compensatory thyroid follicular cell proliferation.
12   Multiple lines of evidence support this observation: [1] TSH levels were unaffected, indicating that
13   the decrease in Ts and T4 levels was not severe enough to stimulate  increased TSH secretion by  the
14   pituitary; [2] thyroid hyperplasia was not induced in this study, or any others exposing mice for
15   2.5-13 weeks, suggesting that thyroid proliferation was either not induced by the hormone
16   fluctuations or that any follicular cell proliferation during this period was too slight to be detected
17   by routine histopathological examination; [3] the maximal decrease in Ts or T4 hormone levels
18   induced by tert-butanol exposure after 14 days (i.e., ~20%] was well within the range of fluctuation
19   in Ts and T4 hormone levels reported to occur between the 3- and 14-day control groups [15-40%;
20   [Blancketal.. 2010]]. Although the lower Ts and T4 levels following tert-butanol were later
21   attributed by the study authors to an increase in liver metabolism (see next section], they could  in
22   fact be due to a decrease in thyroid hormone production, resulting from some, as of yet,
23   uninvestigated molecular interactions of tert-butanol in the thyroid, pituitary, or hypothalamus.
24          The absence of information regarding thyroid hormone levels in male mice and lack of
25   molecular studies evaluating exposures >2 weeks in female mice are significant deficiencies in the
26   available database. Together, although small decreases in some thyroid hormone levels have been
27   reported in female mice, the available evidence is inadequate to determine if tert-butanol
28   negatively affects the pituitary-thyroid signaling axis in female mice; furthermore, no evidence was
29   available to evaluate this effect in male mice.

30   3] Site[s] of antithyroid action [required]
31          The thyroid and liver are two of several potential sites of antithyroid action, with the liver
32   the most common site of action, where increased microsomal enzyme activity could enhance
33   thyroid hormone metabolism and removal (U.S. EPA. 1998a]. Rats are thought to be more sensitive
34   than mice to this aspect of antithyroid activity (Rogues etal.. 2013: Qatanani et al.. 2 005: U.S. EPA.
35   1998a]: however, rats exposed to tert-butanol for 2 years did not exhibit treatment-related thyroid
36   effects, while mice did. Typically, chronic induction of liver microsomal enzyme activity resulting
37   from repeated chemical exposure would manifest some manner of liver histopathology, such as
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-43           DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   hepatocellular hypertrophy or hyperplasia [U.S. EPA. 1998a: NTP. 1995]. In a 14-day mechanistic
 2   investigation, tert-butanol had no effect on liver weight when compared to the control group, but
 3   centrilobular hepatocellular hypertrophy was reported in 2/5 livers from high-dose mice (versus
 4   0/6 in control and 0/5 in low-dose mice [Blancketal., 2010]. Relative liver weights increased in
 5   male and female mice after 13 weeks of oral exposure [NTP. 1995] to higher doses than those
 6   evaluated by Blancketal. [2010]. although absolute liver weight measurements in treated animals
 7   showed little change from controls suggesting that the relative measures could have been related to
 8   decreases in body weight rather than specific liver effects. Relative (and absolute] liver weights
 9   were increased in female mice (only] after 13 weeks of inhalation exposure at the two highest
10   concentrations (NTP. 1997]: liver weight was not reported in mice orally exposed for 2 years (NTP.
11   1995]. No increase in mouse hepatocellular hypertrophic or hyperplastic histopathology was
12   reported following 2.5 weeks to 2 years of exposure (NTP. 1997.1995]. In fact, the only liver
13   pathology associated with tert-butanol exposure in these studies was an increase in fatty liver in
14   male mice in the high-dose group after 2 years of oral exposure (NTP.  1995]. Although increased
15   fatty liver could indicate some non-specific metabolic alteration, the absence of a similar treatment-
16   related effect in livers from female  mice, which were sensitive to both thyroid follicular cell
17   hyperplasia and tumor induction, suggests that it might not be related to the thyroid tumorigenesis.
18          One study evaluated liver enzyme expression and found highly dose-responsive induction
19   of a single phase I cytochrome p450 enzyme (CYP2B10] following 14 days of tert-butanol exposure
20   in female mice, with much smaller increases in the expression of another phase I enzyme CYP2B9,
21   and the phase II thyroid hormone-metabolizing enzyme, sulfotransferase 1A1 [(SULT1A1; Blanck et
22   al. (2010]]. CYP2B enzyme induction is commonly used as an indication of constitutive androstane
23   receptor (CAR] activation; CAR can induce expression of a wide range  of hepatic enzymes, including
24   several CYPs along with thyroid hormone-metabolizing sulfotransferases (Rogues etal.. 2013]. The
25   only thyroid hormone-metabolizing enzyme induced by tert-butanol, however, was SULT1A1,
26   which has been reported to be inducible in a CAR-independent manner in mice (Oatanani etal..
27   2005]. Based on alterations in hepatic phase I and phase II enzyme activities and gene expression,
28   the above data suggest a possible role for increased thyroid hormone clearance in the liver
29   following repeated tert-butanol exposure; however, the expression changes in these few enzymes
30   are not supported by any liver histopathological effects in mice exposed for longer durations, so
31   whether this enzyme induction is transient, or simply insufficient to induce liver pathology after >2
32   weeks of exposure, is unknown. No evidence is available to evaluate the potential for intrathyroidal
33   or any other extrahepatic effects in female mice or for any of these molecular endpoints in male
34   mice;  therefore, the available evidence is inadequate to determine if major site(s] of antithyroid
35   action are affected.

36   4] Dose correlation (required]
37          Confidence in the disruption of the thyroid-pituitary function is enhanced when dose
38   correlation is present among the hormone levels producing various changes in thyroid
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-44          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   histopathology, including thyroid tumors [U.S. EPA. 1998a]. Furthermore, if thyroid hormone levels
 2   were affected by liver enzyme induction, confidence would be increased by a concordance among
 3   liver effects, thyroid hormone levels, and thyroid pathology. Thyroid hormone levels were
 4   evaluated only in female mice exposed to tert-butanol; after 2 weeks of exposure, both T4 and T3
 5   were decreased with both doses (2 and 20 mg/L), and TSH was unaffected at either dose [Blanck et
 6   al.. 2010). Liver expression of CYP2B10 was increased in a dose-responsive manner, while
 7   SULT1A1 mRNA was induced by 20-30% at both doses [Blanck etal.. 2010]. As described above,
 8   induction of liver microsomal enzyme activity would manifest some manner of liver histopathology
 9   [Maronpotetal.. 2010: U.S. EPA. 1998a:  NTP.  1995). Toxicol Pathol 38:776-795), and consistent
10   with this expected association, centrilobular hepatocellular hypertrophy was reported in 2/5 high-
11   dose mice exposed for 2 weeks [Blanck etal., 2010]. No liver histopathology, however, was
12   attributed to tert-butanol exposure in female mice exposed for 2.5 weeks to 2 years to comparable
13   tert-butanol concentrations [NTP. 1997.1995]. Although liver enzyme levels and activity were not
14   specifically evaluated following subchronic to chronic exposure, the lack of liver pathology suggests
15   a comparable lack of enzyme induction.  Conversely, no histopathological alterations were reported
16   in the thyroids of female mice after 2 weeks of oral exposure at doses that elevated some liver
17   enzyme levels [Blanck etal.. 2010].
18          Following 2 years of oral exposure, both follicular cell hyperplasia and follicular cell tumor
19   incidence was increased in mice despite a lack of treatment-related liver pathology [NTP. 1995]
20   (Table 1-6]. Any associations relating hormone changes to thyroid pathology or liver enzyme
21   induction are limited due to the inadequate database (described above]; the available  evidence
22   suggests little concordance among reports of liver, pituitary, and thyroid effects in female mice, and
23   no evidence was available to evaluate these associations in male mice.

24   5] Reversibility [required]
25          Chemicals acting via an antithyroid MOA have effects (e.g., increased TSH levels, thyroid
26   follicular cell proliferation] that are reversible after cessation of treatment [U.S. EPA. 1998a].
27   Although increased TSH levels have not  been demonstrated following tert-butanol exposure,
28   thyroid follicular cell proliferation was observed following chronic exposure. As no studies have
29   evaluated changes in thyroid hormones  or thyroid histopathology after cessation of tert-butanol
30   treatment, however, the available evidence is  inadequate to evaluate reversibility of these effects.
31          In summary, the available database sufficiently supports only [1] increases in thyroid cell
32   growth. The existing data are inadequate to evaluate [2] thyroid and pituitary hormone changes
33   consistent with the antithyroid MOA, (3] site(s]  of the antithyroid action, or (5] reversibility of
34   effects in the early stages of disruption. Although these inadequacies also limit the evaluation of [4]
35   dose correlation among the various effects, the available evidence suggests that little correlation
36   exists among reported thyroid, pituitary, and liver endpoints. Together, the database is inadequate
37   to determine if an antithyroid MOA is operating in mice. In the absence of information to indicate
38   otherwise, the thyroid tumors observed in mice are considered relevant to humans.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-45           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   Integration of thyroid effects
 2          The thyroid endpoints reported following chronic exposure to tert-butanol include
 3   increases in follicular cell hyperplasia and tumors in male and female mice. As discussed above, due
 4   to inadequacies in four of the five required areas [U.S. EPA, 1998a], the evidence is inadequate to
 5   determine if an antithyroid MOA is operating in mice; therefore, the MOA(s) for thyroid
 6   tumorigenesis has not been identified. EPA considers the thyroid follicular cell hyperplasia to be an
 7   early event in the neoplastic progression of thyroid follicular cell tumors, and no other noncancer
 8   effects on the thyroid were observed. Thus, the hazard and dose-response conclusions regarding
 9   the thyroid follicular cell hyperplasia and tumors associated with tert-butanol exposure are
10   discussed as partof the overall weight of evidence for carcinogenicity in Section 1.3.2.

11   1.2.3.  Developmental Effects

12   Synthesis of effects related to development
13          Four studies evaluated developmental effects [three oral or inhalation developmental
14   studies [Faulkner etal.. 1989: Nelson etal.. 1989: Daniel and Evans. 1982]  and a one-generation,
15   oral reproductive study [Lyondell Chemical Co., 2004]] in animals exposed to tert-butanol via liquid
16   diet (i.e., maltose/dextrin], oral gavage, or inhalation. No developmental epidemiology studies are
17   available for tert-butanol. The animal studies are arranged in the evidence tables by species, strain,
18   and route of exposure. The design, conduct, and reporting of each study were reviewed, and each
19   study was considered adequate to provide information pertinent to this assessment. One study was
20   considered less informative, Faulkner et al. [1989]. because it did not provide sufficient information
21   on the dams to determine if fetal effects occurred due to maternal toxicity.
22          Developmental effects of tert-butanol observed after oral exposure (liquid diets or gavage]
23   in several mouse strains and one rat strain include measures of fetal loss or viability (e.g., increased
24   number of resorptions, decreased numbers of neonates per litter] and decreased fetal body weight
25   (Lyondell Chemical Co.. 2004: Faulkner etal.. 1989: Daniel and Evans. 1982]. Daniel and Evans
26   (1982] also observed decreases in body weight gain during post-natal days (PNDs] 2-10; data
27   suggest, however, that this effect might be due to altered maternal  behavior or nutritional status. In
28   addition, a  single dose study reported a small increase in the incidence  of variations of the skull or
29   sternebrae in two mouse strains (Faulkner etal., 1989]. Although variations in skeletal
30   development were noted in the study, no malformations were reported. Similar developmental
31   effects were observed after whole-body inhalation exposure in Sprague-Dawley rats for 7
32   hours/day  on gestation days (CDs] 1-19  (Nelson etal., 1989]. Fetal effects included dose-related
33   reductions  in body weight in male and female fetuses and higher incidence of skeletal variations
34   when analyzed based on individual fetuses (but not on a per litter basis].
35          In these studies, fetal effects are generally observed at doses that cause toxicity in the dams
36   as measured by clinical signs (e.g., decreased body weight gain, food consumption]  (Table 1-7;
37   Figure 1-11; Figure 1-12]. As stated in the Guidelines for Developmental Toxicity Risk Assessment

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-46           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   [U.S. EPA. 1991b]. "an integrated evaluation must be performed considering all maternal and
 2   developmental endpoints." "[W]hen adverse developmental effects are produced only at doses that
 3   cause minimal maternal toxicity; in these cases, the developmental effects are still considered to
 4   represent developmental toxicity and should not be discounted." Although, at doses of "excessive
 5   maternal toxicity...information on developmental effects may be difficult to interpret and of limited
 6   value." In considering the fetal and maternal toxicity data following tert-butanol exposure, the
 7   severity of the maternal effects were minimal and therefore the developmental effects in the fetuses
 8   should not be discounted [U.S. EPA, 1991b]. The observed fetal effects occurred, however, at doses
 9   resulting in maternal toxicity across all available studies. Therefore, whether the fetal effects are
10   directly related to tert-butanol treatment or are secondary to maternal toxicity remains unclear.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-47           DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review of tert-ButyI Alcohol
        Table 1-7. Evidence pertaining to developmental effects in animals following
        exposure to tert-butanol
    Reference and study design
                              Results
Lyondell Chemical Co. (2004)

Sprague-Dawley rat; 12/sex/treatment
Gavage 0, 64,160, 400, or 1,000 mg/kg-d
FO males: 9 weeks beginning 4 weeks prior
to mating
FO females: 4 weeks prior to mating
through PND21
Fl Males and Females: 7 weeks
(throughout gestation and lactation; 1 male
and 1 female from each litter was dosed
directly from PND 21-28)
Response relative to control

Dose
(mg/kg-d)       0            64          160        400        1000

Maternal effects

Body weight gain GD 0-20

               0            -3-40         -16*

Food consumption GD 0-20

               0            00+40

Body weight gain PND 1-21

               0            +3          -10        +3        +100*


Food consumption LD1-14

               0            -2-60         -16

Live pups/litter response relative to control

              0             -9          -11        -7        -33*

Dams dosed with 400 or 1000 mg/kg-d showed CNS effects (e.g.. ataxia. lethargy)
which were undetectable by 4-weeks of exposure in animals exposed to 400
mg/kg-d but not those in the higher dose group.

Fl effects


Viability index (pup survival to PND4)

             96.4%        98.7%       98.2%      99.4%      74.1%*

Lactation index (pup survival to PND21)

              100%        100%       100%      99.2%      98.8%

Sex ratio (% males)

              54.4          52.3         50.9        53.4        52.1

Pup weight/litter PND 1 relative to control (%)

               0            +6          +4        +7         -10

Pup weight PND 28 relative to control (%)

      M:       0            +200         -12*

       F:       0            0-4        -2         -8
            This document is a draft for review purposes only and does not constitute Agency policy,
                                                   1-48            DRAFT—DO NOT CITE OR QUOTE

-------
                                         Toxicological Review of tert-ButyI Alcohol
Reference and study design
Daniel and Evans (1982)
Swiss Webster (Cox) mouse; 15 pregnant
dams/treatment
Liquid diet (0, 0.5, 0.75, 1.0%, w/v)
0 (isocaloric amounts of maltose/dextrin)
3,324, 4,879, 6,677 mg/kg-d
GD6-20


Faulkner etal. (1989)
CBA/J mouse; 7 pregnant females in
control, 12 pregnant females in treated
Gavage (10.5 mmoles/kg twice a day);
0 (tap water), 1,556 mg/kg-d
GD6-18
Results
No statistical analysis was conducted on any of these data
Maternal
Percent change compared to control:
Number of litters
Dose Food consumption Body weight (% pregnant
(mg/kg-d) (mean g/animal/day) gain dams)
000 11(77%)
3,324 +2 -3 12(80%)
4,879 -3 -19 8(53%)
6,677 -4 -20 7(47%)
^H
Authors note that lower food consumption in higher tert-butanol dose groups
reflects problems with pair feeding and maternal sedation.
Fetal
Percent change compared to control:
Dose Number of Fetal body weight
(mg/kg-d) neonates/litter on PND 2
000
3,324 -1 -7
4,879 -29 -19
6,677 -49 -38
Number of stillborn also increased with dose (3, 6, 14, and 20, respectively), but
the number of stillborn per litter was not provided. The high dose also caused a
delay in eye opening and a lag in weight gain during PND 2-10 (information was
only provided in text or figures)
Maternal results not reported.
Fetal
Percent change compared to control: Incidence:
Live
Dose fetuses/ Fetal Sternebral Skull
(mg/kg-d) Resorptions/litter litter weight variations variations
0 000 4/28 1/28
1,556 +118* -41* -4 7/30 3/30
Sternal variations: misaligned or unossified sternebrae
Skull variations: moderate reduction in ossification of supraoccipital bone
Number of total resorptions (10 resorptions/66 implants in controls, 37/94
implants in treated) increased (p < 0.05)
This document is a draft for review purposes only and does not constitute Agency policy,
                                  1-49           DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review of tert-ButyI Alcohol
Reference and study design
Faulkner etal. (1989)
C57BL/6J mouse; 5 pregnant females in
controls, 9 pregnant females treated
Gavage (10.5 mmoles/kg twice a day)
0 (tap water), 1,556 mg/kg-d
GD6-18
Nelson etal. (1989)
Sprague-Dawley rat; 15 pregnant
dams/treatment
Inhalation analytical concentration: 0,
2,200, 3,510, 5,030 ppm (0, 6,669, 10,640,
15,248 mg/m3), (dynamic whole body
chamber)
7hr/d
GD 1-19


Results
Maternal results not reported.
Fetal
Percent change compared to control: Incidence:
Live
Dose fetuses/ Fetal Sternebral Skull
(mg/kg-d) Resorptions/litter litter weight variations variations
0 000 5/21 1/21
1,556 +428* -58* -4 9/16 7/16
Sternal variations: misaligned or unossified sternebrae
Skull variations: moderate reduction in ossification of supraoccipital bone
Number of total resorptions (4 resorptions/44 implants in controls, 38/68
implants in treated) increased (p < 0.05)
Maternal: Unsteady gait (no statistical tests reported), dose-dependent •]/ in
body weight gain (results presented in figure only), dose-dependent •]/ in food
consumption ranging from 7-36% depending on dose and time
Fetal
Percent change compared to control (mean ± standard error):
Dose Number of live Resorptions
(mg/m3-) fetuses/litter per litter
0 0(13±2) 0(1.1+1.2)
6,669 0(13±4) +9(1.2±1.1)
10,640 +15 (15±2) -18 (0.9±1.0)
15,248 +8(14±2) 0(1.1±0.9)
Percent change compared to
control: Incidence:
Skeletal Skeletal
Dose Fetal weight Fetal weight variation variation
(mg/m-) (males) (females) by litter by fetus
000 10/15 18/96
6,669 -9* -9* 14/17 35/104
10,640 -12* -13* 14/14 53/103*
15,248 -32* -31* 12/12 76/83*
Skeletal variation by litter refers to the number of variations observed in the
number of litters examined. Skeletal variation by fetus refers to the number of
variations observed in the total number of fetuses examined. Fetuses are not
categorized by litter.
1
2
3
4
5
* Statistically significant p < 0.05 as determined by study authors. Conversions from diet concentrations to mg/kg-d
 performed by study authors. Conversion from ppm to mg/m3 is 1 ppm = 3.031 mg/m3.

Note: Percentage change compared to control = (treated value - control value) -f control value x 100.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-50           DRAFT—DO NOT CITE OR QUOTE

-------
                                                                 Toxicological Review of tert-ButyI Alcohol
                   • = exposures at which the endpoint was reported statistically significant by study authors
                   D = exposures at which the endpoint was reported not statistically significant by study authors
                  DEVELOPMENTAL
                         i Maternal body weight gain
                            (CD 0-20]; F rat (C)

                         TMaternal body weight gain
                            (LD 1-21); F rat (C)

                  •iNumberof live pups per litter; M+F
                             rat(C)


                        -IViability index; M+F rat(C)
                        Lactation index; M+F rat (C)
                             Sex ratio; M+F rat (C)


                             -I Pup weight per litter
                             [PNDl);M+Frat(C)

                             JPup weight per litter
                              (PND28);Mrat(C)

                             IPup weight per litter
                              (PND28);Frat(C)

                      -1 Maternal body weight gain; F
                              mouse (A) *

                     -INumberof neonates/Iitter, fetal
                      body weight; M+F mouse (A)*

                    TNumberof resorptions per litter;
                           M+F mouse (B)

                    INumberof live fetuses per litter;
                           M+F mouse (B)


                       -IFetal weight; M+F mouse (B)
                   Skeletal variations; M+F mouse (B)





























































































































































































































































































E





















,








,




Q-B-B



m-m-m



•









D
                                              10
                                                              100            1,000

                                                              Dose (mg/kg-day)
                                                                               10,000
3
4


5
6
* Study authors did not conduct statistical analysis on these endpoints, but results are determined by EPA
to be biologically significant.
Sources: (A) Daniel and Evans (1982); (B) Faulkner et al. (1989); Lyondell Chemical Co. (2004)

Figure 1-11. Exposure-response array of developmental effects following oral
exposure to tert-butanol.
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                         1-51             DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review of tert-Butyl Alcohol
• = exposures at which the endpoint was reported statistically significant by study authors
D = exposures at which the endpoint was reported not statistically significant by study authors
 DEVELOPMENTAL
        Number of live fetuses per litter; M+F rat
                 (Nelson etal., 1989)
        Number of resorpn'ons per litter; M+F rat
                 (Nelson etal., 1989)
                         ^Fetal weight; M rat
                         (Nelson etal., 1989)
                         iFetal weight; F rat
                         (Nelson etal., 1989)
             Skeletal variation by litter; M+F rat
                   (Nelson etal., 1989)
             Skeletal variation by fetus; M+F rat
                   (Nelson etal., 1989)






D	B—B
D	B—B
                                        1,000              10,000
                                          Exposure Concentration (rng/m3)
                    100,000
 Figure 1-12. Exposure-response array of developmental effects following
 inhalation exposure to tert-butanol.
    This document is a draft for review purposes only and does not constitute Agency policy,
                                         1-52            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol

 1   Integration of developmental effects
 2          There is suggestive evidence of developmental effects associated with tert-butanol
 3   exposure. Exposure to tert-butanol during gestation resulted in increased fetal loss, decreased fetal
 4   body weight, and increases in skeletal variations in exposed offspring. Dams had body weight losses
 5   or gains (or both), decreased food consumption, and clinical signs of intoxication at the same doses
 6   of tert-butanol causing fetal effects. Therefore, determining whether tert-butanol exposure results
 7   in specific developmental toxicity or whether the fetal effects are due to maternal toxicity is
 8   difficult The observed maternal effects are minimal, however, and thus, the developmental effects
 9   observed in the fetuses are not discounted as being secondary to maternal toxicity [U.S. EPA.
10   1991b) and the evidence is considered suggestive of developmental toxicity.

11   1.2.4.  Neurodevelopmental Effects

12   Synthesis of effects related to neurodevelopment
13          Three studies evaluated neurodevelopmental effects [Nelson etal.. 1991: Daniel and Evans.
14   1982.)[one in male rats; one in female rats] following tert-butanol exposure via liquid diet (i.e.,
15   maltose/dextrin) or inhalation. No epidemiology studies on neurodevelopment are available. The
16   animal studies evaluating neurodevelopmental effects of tert-butanol contain study design
17   limitations. Daniel and Evans (1982) had a small number of animals per treatment group, lacked
18   comparison of treatment-related effects  to controls for all endpoints investigated, and did not use
19   long-term neurodevelopmental testing. The two studies by Nelson etal. (1991) evaluated
20   neurodevelopmental effects after either  paternal or maternal exposure but did not run the
21   exposures concurrently or provide exposure methods to indicate the studies were conducted
22   similarly. The studies are arranged in the evidence tables by species and sex.
23          Various neurodevelopmental effects have been observed in the available studies. These
24   include changes in rotarod performance following oral or inhalation exposures and decreases in
25   open field behavior and cliff avoidance following oral exposure, and reduced time hanging on wire
26   after inhalation exposure during gestation (Table 1-8).

27   Rotarod performance
28          Inconsistent results were observed across studies. Although Daniel and Evans (1982) found
29   decreased rotarod performance in mouse pups of dams orally exposed during gestation, Nelson et
30   al. (1991) observed an increase in rotarod performance  in rat pups of dams exposed via inhalation
31   during gestation.

32   Neurochemical measurements
33          Biochemical or physiological changes in the brain of offspring exposed during gestation or
34   early in the postnatal period were examined in one study. In this study, Nelson etal. (1991)
35   reported statistically significant changes in neurochemical measurements in the brain in offspring

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-53           DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review of tert-ButyI Alcohol


 1    of both dams exposed via inhalation during gestation and treated adult males mated with untreated
 2    dams. The strength of these results is compromised, however, because the two concentrations
 3    tested (in both experiments) were not run concurrently, and only data on statistically significant
 4    effects were reported. Therefore, comparison across doses or trend analysis for the effects is not
 5    feasible.
 6    Physiological and psychomotor development
 7           Daniel and Evans [1982] cross-fostered half the mouse pups born to treated mothers with
 8    untreated surrogate females to test the effects of maternal nutrition and behavioral factors on the
 9    pups' physiological and psychomotor development Results indicated that pups fostered to control
10    dams performed significantly better than those maintained with treated dams (Table l-8][Daniel
11    and Evans. 1982). Data suggest that neurodevelopmental effects were not solely due to in utero
12    exposure to tert-butanol [Daniel and Evans, 1982]. Interpretation of these results is limited,
13    however, as the neurodevelopmental data were presented only in figures and could not be
14    compared with controls.
15
16
Table 1-8. Evidence pertaining to neurodevelopmental effects in animals
following exposure to tert-butanol
            Reference and study design
                                                         Results
      Daniel and Evans (1982)
      Swiss Webster (Cox) mouse; 15 pregnant
      dams/treatment
      Liquid diet (0, 0.5, 0.75, or 1.0%, w/v); GD6-20;
      after birth half the pups were nursed with their
      treated dams and the other half were fostered
      by untreated dams who recently gave birth
      0 (isocaloric amounts of maltose/dextrin),
      3,324, 4,879, or 6,677 mg/kg-d
                                   •   a dose-dependent increase in righting reflex time, with more time
                                       needed in animals maintained with maternal dams

                                   •   a dose-dependent decrease in open field behavior, with less
                                       activity in pups maintained with maternal dams

                                   •   a dose-dependent decrease in rotarod performance with the
                                       pups from maternal dams having lower performances

                                   •   a dose-dependent decrease in the amount of time the pups were
                                       able to avoid a cliff, with animals maintained with their maternal
                                       dams having less avoidance time
      Nelson etal.(1991)
      Sprague-Dawley rat; 15 pregnant
      dams/treatment
      Inhalation analytical concentration: 0, 6,000, or
      12,000 mg/m3; (dynamic whole body chamber)
      7hr/d
      GD 1-19
                                Data were not presented specifically by dose nor were any tables or figures
                                of the data provided

                                Maternal toxicity was noted by decreased food consumption and body
                                weight gains

                                Results in offspring

                                    •   increase in rotarod performance in high-dose group (16 versus 26
                                       revolutions/min for controls and 12,000 mg/m3 animals,
                                       respectively)

                                    •   decreased time held on wire in the performance ascent test in the
                                       low-dose group (16 sec versus 10 sec for controls and 1,750
                                       mg/m3 animals, respectively)
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                      1-54            DRAFT—DO NOT CITE OR QUOTE

-------
                                                                Toxicological Review of tert-ButyI Alcohol
            Reference and study design
                           Results
                                               The following differences in neurochemical measurements in the brain
                                               between control and treated offspring were observed,

                                                   •   53% decrease in norepinephrine in the cerebellum at
                                                       12,000 mg/m3

                                                   •   57% decrease in met-enkephalin in the cerebrum at
                                                       12,000 mg/m3 and 83% decrease at 6,000 mg/m3

                                                   •   61% decrease in P-endorphin in the cerebellum at 12,000 mg/m3

                                                   •   67% decrease in serotonin in the midbrain at 6,000 mg/m3
       Nelson et al. (1991)
       Adult male Sprague-Dawley rats (18/treatment)
       mated to untreated females
       Inhalation analytical concentration: 0, 6,000, or
       12,000 mg/m3; (dynamic whole body chamber)
       7 hr/d for 6 wk
Data were not presented specifically by dose nor were any tables or figures
of the data provided

Results (generally only specified as paternally treated versus controls) in
offspring indicate

    •    increase in rotarod performance (16 versus 20 revolutions/min
        for controls and 12,000 mg/m3 animals, respectively)

    •    decreased time in open field (less time to reach the outer circle of
        the field, 210 sec versus 115 seconds for controls and 12,000
        mg/m3 animals, respectively)

The following differences in neurochemical measurements in the brain
between control and treated offspring were observed

    •    39% decrease in norepinephrine in the cerebellum at
        12,000 mg/m3

    •    40% decrease in met-enkephalin in the cerebrum at
        12,000 mg/m3 and 75% decrease at 6,000 mg/m3

    •    71% decrease in P-endorphin in the cerebellum at 12,000 mg/m3

    •    47% decrease in serotonin in the midbrain at 6,000 mg/m3
 1    * Statistically significant p < 0.05 as determined by study authors.
 2
 3    Note: Conversions from diet concentrations to mg/kg-d performed by study authors.
 4    Percentage change compared to control = (treated value - control value) 4- control value x 100.

 5    Mechanistic Evidence
 6            No mechanistic evidence is available for reproductive, developmental, or
 7    neurodevelopmental effects.

 8    Integration of neurodevelopmental effects
 9            Neurodevelopmental effects, including decreased brain weight, changes in brain
10    biochemistry, and changes in behavioral performances, have been observed. Each study evaluating
11    neurodevelopmental effects, however, had limitations in study design, reporting, or both. In
12    addition, results were not always consistent between studies or across dose. At this time, there is
13    inadequate information to draw conclusions regarding neurodevelopmental toxicity.
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                        1-55            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   1.2.5.  Reproductive Effects
 2   Synthesis of effects related to reproduction
 3          Several studies evaluated reproductive effects [a one-generation, oral reproductive study
 4   [Lyondell Chemical Co.. 2004] and subchronic evaluations in rats and mice following oral and
 5   inhalation exposure [NTP. 1997.1995]] in animals exposed to tert-butanol via oral gavage, drinking
 6   water, or inhalation for >63 days. The studies are arranged in the evidence tables by sex, route of
 7   exposure, duration of exposure, and species. The collection of studies evaluating reproductive
 8   effects of tert-butanol is limited by the absence of two-generation reproductive oral or inhalation
 9   studies and by having no human studies on reproduction. The design, conduct, and reporting of
10   each study were reviewed, and each study was considered adequate to provide information
11   pertinent to this assessment.
12          Reproductive endpoints, such as reproductive organ weights, estrous cycle length, and
13   sperm effects were examined following either oral or inhalation exposure [Lyondell Chemical Co..
14   2004: NTP. 1997.1995] (Table l-9;Figure 1-13; Figure 1-14]. In males, the only significant effect
15   observed was a slight decrease in sperm motility for FO males treated with 1000 mg/kg-day of tert-
16   butanol [Lyondell Chemical Co.. 2004]. No significant changes in sperm motility were reported
17   following oral exposure in other rat studies or via  inhalation exposure in mice or rats. In addition,
18   the reduced motility in treated animals falls within the range of historical control data and,
19   therefore, its biological significance is uncertain. In female B6C3Fi mice, estrous cycle length was
20   increased 28% following oral exposure to 11,620 mg/kg-day [NTP. 1995]. No significant changes in
21   estrous cycle length were observed following oral exposure in rats, or inhalation exposure in mice
22   or rats.
23
24
Table 1-9. Evidence pertaining to reproductive effects in animals following
exposure to tert-butanol
Reference and study design
Results
Male reproductive effects
Lvondell Chemical Co. (2004)
Sprague-Dawley rat; 12/sex/treatment
Gavage 0, 64, 160, 400, or 1,000 mg/kg-d
FO males: 9 weeks beginning 4 weeks prior to
mating
PND21
NTP (1995)
F344/N rat; 10/sex/treatment
Drinking water (0, 2.5, 5, 10, 20, or 40 mg/mL)
M: 0, 230, 490, 840, 1,520, 3,610a mg/kg-d
13 weeks
FO reproductive effects
Sperm motility (only control and high-dose groups examined)
0: 94% 1000: 91%*
No other significant effect on weights of male reproductive organs or sperm
observed
No significant effect on weights of male reproductive organs or sperm
observed
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                 1-56           DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review of tert-ButyI Alcohol
      Reference and study design
                             Results
NTP (1995)
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, or 40 mg/mL)
 M: 0, 350, 640, 1,590, 3,940, 8,210a mg/kg-d
13 weeks
No significant effect on weights of male reproductive organs or sperm
observed
NTP (1997)
F344/N rat; 10/sex/treatment
Inhalation analytical concentration: 0,134, 272,
542, 1,080, or 2,101 ppm (0, 406, 824, 1,643,
3,273 or 6,368 mg/m3) (dynamic whole body
chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist Ultrasonic spray
nozzle nebulizer), analytical concentration and
method were reported
No significant effect on weights of male reproductive organs or sperm
observed

Evaluations were performed only for concentrations >542 ppm
(1,643 mg/m3)
NTP (1997)
B6C3Fi mouse; 10/sex/treatment
Inhalation analytical concentration: 0,134, 272,
542, 1,080, or 2,101 ppm (0, 406, 824, 1,643,
3,273 or 6,368 mg/m3) (dynamic whole body
chamber)
6 hr/d, 5 d/wk
13 weeks
Generation method (Sonimist Ultrasonic spray
nozzle nebulizer), analytical concentration and
method were reported
No significant effect on weights of male reproductive organs or sperm
observed

Evaluations were performed only for concentrations >542 ppm
(1,643 mg/m3)
Female reproductive effects
Lyondell Chemical Co. (2004)
Sprague-Dawley rat; 12/sex/treatment
Gavage 0, 64,160, 400, or 1,000 mg/kg-d
FO females: 4 weeks prior to mating through
PND21
Pregnancy index

91.7%    91.7%    100%    100%    91.7%
NTP (1995)
F344/N rat; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, or 40 mg/mL)
 F: 0, 290, 590, 850, 1,560, 3,620a mg/kg-d
13 weeks
No significant effect on female estrous cycle (0, -2, -4, 0, +8 % change
relative to control)
NTP (1995)
B6C3Fi mouse; 10/sex/treatment
Drinking water (0, 2.5, 5,10, 20, or 40 mg/mL)
F: 0, 500, 820, 1,660, 6,430, ll,620a mg/kg-d
13 weeks
T" length of estrous cycle

Response relative to control: 0, +5, +5, +5, +6, +28*%
            This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-57             DRAFT—DO NOT CITE OR QUOTE

-------
                                                                   Toxicological Review of tert-ButyI Alcohol
            Reference and study design
                            Results
      NTP (1997)
      F344/N rat; 10/sex/treatment
      Inhalation analytical concentration: 0,134, 272,
      542, 1,080, or 2,101 ppm (0, 406, 824, 1,643,
      3,273 or 6,368 mg/m3) (dynamic whole body
      chamber)
      6 hr/d, 5 d/wk
      13 weeks
      Generation method (Sonimist Ultrasonic spray
      nozzle nebulizer), analytical concentration and
      method were reported
No significant effect on female estrous cycle (0, -4, +2, +4 % change relative
to control)

Evaluations were performed only for concentrations >542 ppm
(1,643 mg/m3)
      NTP (1997)
      B6C3Fi mouse; 10/sex/treatment
      Inhalation analytical concentration: 0,134, 272,
      542, 1,080, or 2,101 ppm (0, 406, 824, 1,643,
      3,273 or 6,368 mg/m3) (dynamic whole body
      chamber)
      6 hr/d, 5 d/wk
      13 weeks
      Generation method (Sonimist Ultrasonic spray
      nozzle nebulizer), analytical concentration and
      method were reported
No significant effect on female estrous cycle (0, -3, -9, -5 % change relative
to control)

Evaluations were only performed for concentrations >542 ppm
(1,643 mg/m3)
1    * Statistically significant p < 0.05 as determined by the study authors.

2    Notes: Conversions from drinking water concentrations to mg/kg-d performed by study authors.
3    Conversion from ppm to mg/m3 is 1 ppm = 3.031 mg/m3.
4    Percentage change compared to control = (treated value - control value) 4- control value x 100
                  This document is a draft for review purposes only and does not constitute Agency policy,
                                                           1-58             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review of tert-ButyI Alcohol
            • = exposures at which the endpoint was reported statistically significant by study authors
            D = exposures at which the endpoint was reported not statistically significant by study authors
             REPRODUCTIVE EFFECTS
/[ale reproductive effects
Reproductive organs or sperm; M
rat (A]

Reproductive organs or sperm; M
rat(B)



Reproductive organs or sperm; M
mouse (B)


emale reproductive effects




Estrous cycle length; F rat (B)






































































































































































































































































































































































                                        10
                                       100         1,000       10,000

                                            Dose (mg/kg-day)
100,000
2
3
Sources: (A) Lyondell Chemical Co. (2004); (B) NTP (1995).

Figure 1-13. Exposure-response array of reproductive effects following oral
exposure to tert-butanol.

   This document is a draft for review purposes only and does not constitute Agency policy.
                                       1-59           DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-Butyl Alcohol
• = exposures at which the endpoint was reported statistically significant by study authors
D = exposures at which the endpoint was reported not statistically significant by study authors
REPRODUCTIVE EFFECTS
Male reproductive effects
Reproductive organs or sperm; M rat
(NTP, 1997]
Reproductive organs or sperm; M mouse
(NTP, 1997)
Female reproductive effects
Estrous cycle; F rat (NTP, 1997)
Estrous cycle; F mouse (NTP, 1997)






                                             1,000
                                                                       10,000
                                                    Exposure Concentration (mg/m:i)
1
2
Figure 1-14. Exposure-response array of reproductive effects following
inhalation exposure to tert-butanol.
              This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-60           DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   Integration of reproductive effects
 2          At this time, no conclusions are drawn in regard to reproductive toxicity. The database is
 3   limited to a one-generation study [Lyondell Chemical Co.. 2004: NTP. 1995). No two-generation
 4   reproductive studies are available that evaluate oral or inhalation exposure. In males, the only
 5   observed effect was a slight decrease in sperm motility for FO males in the highest dose group of
 6   rats treated with tert-butanol. This effect was not observed, however, in other studies with orally
 7   treated rats and mice or in rats exposed via inhalation. In females, NTP [1995] reported an
 8   increased length of the estrous cycle in the highest dose group of orally exposed mice. This effect
 9   was not observed in similarly treated rats or in mice and rats exposed via inhalation.

10   1.2.6.  Other Toxicological Effects
11          Effects other than those related to kidney, thyroid, reproductive, developmental, and
12   neurodevelopmental toxicity were observed in some of the available rodent studies; these include
13   liver and urinary bladder effects. Due to a lack of consistency in the liver effects and minimal to
14   mild effects with a lack of progression in urinary bladder, however, inadequate information is
15   available to draw conclusions regarding liver or urinary bladder toxicity at this time.
16          Additionally, central nervous system (CNS) effects similar to those caused by ethanol
17   (animals appearing intoxicated and having withdrawal symptoms after cessation of oral or
18   inhalation exposure) were observed. Due to study quality concerns (e.g., lack of data reporting,
19   small number of animals per treatment group), however, adequate information to assess CNS
20   toxicity is unavailable at this time. For more information on these other toxicological effects, see
21   Appendix B.3.

22   1.3.   INTEGRATION AND EVALUATION

23   1.3.1.  Effects Other Than Cancer
24          Kidney effects were identified as a potential human hazard of tert-butanol exposure based
25   on several endpoints, including suppurative inflammation in female rats, transitional epithelial
26   hyperplasia in male and female rats, severity of nephropathy in male and female rats, incidences of
27   nephropathy in female rats, mineralization in male rats, and increased kidney weights in both male
28   and female rats. These effects are similar to the kidney effects observed with ETBE exposure (e.g.,
29   CPN and urothelial hyperplasia) and MTBE (e.g., CPN and mineralization) fATSDR. 19961
30          Several effects were observed in the kidneys of rats. Based on mechanistic evidence
31   indicating that an a2u-globulin-related process is operating in male rats (Hardetal., 2011: Cirvello
32   etal., 1995: NTP, 1995:  Lindamoodetal., 1992), any kidney effects associated with a2U-globulin
33   nephropathy are not considered relevant for human hazard identification. In addition, CPN played a
34   role in the renal tubule nephropathy observed following tert-butanol exposure, and effects
35   associated with such nephropathy are not considered relevant for human hazard identification.
36   Although increases in severity (males and females) or incidence (females) of nephropathy were

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-61           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   related to tert-butanol exposure and could have arisen from chemical-specific processes
 2   independent from CPN, the association of these effects with CPN makes this measure less suitable
 3   for dose-response analysis, and therefore these effects were not considered for the derivation of
 4   reference values. Furthermore, some uncertainty exists regarding whether mineralization is also
 5   associated with CPN in male rats; due to this uncertainty, and because other kidney effects were
 6   identified as being associated with tert-butanol exposure and yet independent from CPN,
 7   mineralization in male rats was not considered for dose-response analysis. The remaining effects
 8   (suppurative inflammation, transitional epithelial hyperplasia, and increased kidney weights) are
 9   considered the result of tert-butanol exposure and relevant to human hazard characterization.
10   These effects therefore are suitable for consideration for dose-response analysis and derivation of
11   reference values, in Section 2.
12          There is suggestive evidence of developmental effects associated with tert-butanol
13   exposure. Increased fetal loss, decreased fetal body weight, and increases in skeletal variations  in
14   exposed offspring were observed following exposure to relatively high doses of tert-butanol during
15   gestation.. These effects are similar to the developmental effects observed with MTBE exposure
16   (e.g., decreased fetal body weight and increases in skeletal variations) [ATSDR. 1996).
17          No mechanistic evidence is available for developmental effects of tert-butanol. Although the
18   evidence is suggestive of developmental toxicity, due to the uncertainty as to whether fetal effects
19   were due to direct effects of tert-butanol or indirect effects of maternal toxicity and the lack of
20   consistency across some endpoints, developmental effects were not considered for dose-response
21   analysis and derivation of reference values in Section 2. Furthermore, no adverse effects were
22   reported in one- and two-generation reproductive/developmental studies on ETBE [Gaoua. 2004a.
23   b), providing further support for the lack of evidence supporting reproductive  or developmental
24   effects as possible human hazards following tert-butanol exposure.
25          At this time, there is inadequate information to draw conclusions regarding
26   neurodevelopmental effects as a human hazard of tert-butanol exposure. Although
27   neurodevelopmental effects have been observed, the studies had limitations in design or reporting,
28   or both, and results were inconsistent between studies and across dose groups. No mechanistic
29   evidence is available to inform the MOA for neurodevelopmental effects of tert-butanol. These
30   effects were not considered further for dose-response analysis and derivation of reference values.
31          At this time, no conclusions are drawn regarding reproductive effects as a human hazard of
32   tert-butanol exposure. The only reproductive effect observed due to tert-butanol exposure was
33   increased length of estrous cycle [NTP. 1995) in the  highest dose group of orally exposed mice,  and
34   this effect was not observed in orally exposed rats or in mice and rats exposed via inhalation.
35   Further, the database was limited and contained only two oral exposure studies and one subchronic
36   inhalation study. No mechanistic or MOA information is available for reproductive effects of tert-
37   butanol. These effects were not considered further for dose-response analysis and derivation of
38   reference values.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-62           DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1          At this time, there is inadequate information to draw conclusions regarding liver or urinary
 2   bladder toxicity due to lack of consistency of effects and minimal/mild effects showing a lack of
 3   progression, respectively. No mechanistic evidence is available for these effects. The liver and
 4   urinary bladder effects were not considered further for dose-response analysis and the derivation
 5   of reference values.

 6   1.3.2.  Carcinogenicity
 7   Summary of evidence
 8          In F344/N rats, administration of tert-butanol in drinking water increased the incidence of
 9   renal tubule tumors, mostly adenomas, in males; no renal tumors in females were reported [Hard et
10   al.. 2011: NTP. 1995). As discussed in Section 1.2.1, some of these tumors might be associated with
11   a2u-globulin nephropathy, an MOA considered specific to the male rat [U.S. EPA, 1991a]. Evidence
12   in support of this hypothesized MOA includes the accumulation of hyaline droplets in renal tubule
13   cells, the presence of (X2u-globulin in the hyaline droplets, and additional aspects associated with
14   (X2u-globulin nephropathy, including linear papillary mineralization and foci of tubular hyperplasia.
15   Other evidence, however, is not supportive: The accumulation of hyaline droplets was minimal;
16   concentrations of (X2u-globulin were low at doses that induced tumors; and no significant necrosis
17   or cytotoxicity was associated with compensatory regenerative proliferation or induction of
18   granular casts observed within a timeframe consistent with (X2u-globulin-mediated nephropathy.
19   Renal tumors also could be associated with chronic progressive nephropathy, but the data on CPN
20   are not coherent: Dose-response relationships for CPN, renal tubule hyperplasia, and renal tubule
21   tumors were different; in addition, CPN was nearly as severe in female rats as in male rats, yet no
22   female rats developed renal tumors. Thus, some renal tumors may be attributable to a2u-globulin
23   nephropathy and some to other, yet unspecified, processes. Taken together, and according to EPA's
24   guidance on renal tumors in male rats [U.S. EPA.  1991a]. renal tumors induced by tert-butanol are
25   relevant for human hazard identification.
26          InB6C3Fi mice, administration of tert-butanol in drinking water increased the incidence of
27   thyroid follicular cell adenomas in females, and adenomas or carcinomas (only one carcinoma
28   observed) in males [NTP, 1995], as discussed in Section 1.2.2. According to EPA's thyroid tumor
29   guidance [U.S. EPA, 1998a], chemicals that produce thyroid tumors in rodents might pose a
30   carcinogenic hazard to humans.
31          In addition, as mentioned in Section 1.1.4, tert-butanol is a primary metabolite of MTBE and
32   of ETBE, two compounds tested in rats and mice that could provide supplementary information on
33   the carcinogenicity of tert-butanol. For MTBE, the most recent cancer evaluation by a national or
34   international health agency is from IARC [1999]. IARC reported that oral gavage exposure in
35   Sprague-Dawley rats resulted in testicular tumors in males and lymphomas and leukemias
36   (combined] in females; inhalation exposure in male and female F344 rats resulted in renal tubule
37   adenomas in males; and inhalation exposure in male and female CD-I mice resulted in

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-63           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   hepatocellular adenomas in females [IARC. 1999]. For ETBE, a draft IRIS assessment under
 2   development concurrently with this assessment reports that inhalation exposure in male and
 3   female F344 rats resulted in hepatocellular tumors, mostly adenomas, in males; no significant
 4   tumor increases were reported for 2-year studies by drinking water exposure in male and female
 5   F344 rats or by oral gavage in male and female Sprague-Dawley rats.

 6   Integration of evidence
 7          This evidence leads to consideration of two hazard descriptors under EPA's cancer
 8   guidelines [U.S. EPA. 2005a]. The descriptor likely to be carcinogenic to humans is appropriate when
 9   the evidence is "adequate to demonstrate carcinogenic potential to humans" but does not support
10   the descriptor carcinogenic to humans. One example from the cancer guidelines is "an agent that has
11   tested positive in animal experiments in more than one species, sex, strain, site, or exposure route,
12   with or without evidence of carcinogenicity in humans." tert-Butanol matches the conditions of this
13   example, having increased tumor incidences in two species, in both sexes, and at two sites.
14          Alternatively, the descriptor suggestive evidence of carcinogenic potential is appropriate
15   when the evidence raises "a concern for potential carcinogenic effects in humans" but is not
16   sufficient for a stronger conclusion. The results for tert-butanol raise a concern for  cancer but none
17   of the effects is particularly strong. The kidney tumors resulted, in part, from an MOA that is specific
18   to male rats, while no kidney tumors occurred in female rats. The thyroid tumors induced in male
19   and female mice were almost entirely benign. In addition, while MTBE was also associated with
20   male rat kidney tumorigenesis, there is little coherence of results between tert-butanol and ETBE
21   associated tumorigenesis in rats. MTBE or ETBE effects following chronic oral exposure in mice
22   have not been investigated, however, so no evidence exists to evaluate the coherence of the thyroid
23   tumorigenesis observed following tert-butanol exposure  in B6C3Fi mice.
24          These considerations, interpreted in light of the cancer guidelines, support the conclusion
25   that there is suggestive evidence of carcinogenic potential for tert-butanol. Although increased tumor
26   incidences were reported for two species, two sexes, and two sites, none of the tumor responses
27   was strong or coherent with the results for ETBE, and this was decisive in selecting a hazard
28   descriptor.
29          The descriptor suggestive evidence of carcinogenic potential applies to all routes of human
30   exposure. Oral administration of tert-butanol to rats and  mice induced tumors at sites beyond the
31   point of initial contact, and inhalation exposure for 13 weeks resulted in absorption and
32   distribution of tert-butanol into the systemic circulation,  as discussed in Section 1.2.1. According to
33   the cancer guidelines, this information provides sufficient basis to apply the cancer descriptor
34   developed from oral studies to other exposure routes.

35   Biological considerations for dose-response analysis
36          Regarding hazards to bring forward to Section 2 for dose-response analysis, EPA's guidance
37   on renal tumors in male rats [U.S. EPA. 1991a] advises that unless the relative contribution of

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-64           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   (X2u-globulin nephropathy and other process can be determined, dose-response analysis should not
 2   be performed. As discussed in Section 1.2.1, the available data do not allow such determination, and
 3   so an analysis of kidney tumors does not appear in Section 2.
 4          EPA's guidance on thyroid tumors and EPA's cancer guidelines [U.S. EPA, 1998a] advises
 5   that for thyroid tumors resulting from thyroid-pituitary disruption, dose-response analysis should
 6   use nonlinear extrapolation, in the absence of MOA information to indicate otherwise. As discussed
 7   in Section 1.2.2, increases in thyroid follicular cell hyperplasia in male and female mice provide
 8   partial support for thyroid-pituitary disruption. Other necessary data on tert-butanol, however, are
 9   not adequate or are not supportive. There is little correlation among thyroid, pituitary, and liver
10   effects in female mice, and no data are available to evaluate the potential for antithyroid effects in
11   male mice. Data are not adequate to conclude that thyroid hormone changes exceed the range of
12   homeostatic regulation or to evaluate effects on extrahepatic sites involved in thyroid-pituitary
13   disruption. Also, no data are available to evaluate reversibility of effects upon cessation of exposure.
14   Thus, according to EPA's thyroid tumor guidance, concluding that the thyroid tumors result from
15   thyroid-pituitary disruption is premature, and dose-response analysis should use linear
16   extrapolation. The data are well suited to dose-response analysis, coming from an NTP study that
17   tested multiple dose levels.

18   1.3.3.  Susceptible Populations and Lifestages for Cancer and Noncancer Outcomes
19          No chemical-specific data that would allow for the identification of populations with
20   increased susceptibility to tert-butanol exposure exist In vitro studies have implicated the liver
21   microsomal mixed function oxidase (MFO) system, namely CYP450 [Cederbaum etal.. 1983:
22   Cederbaum and Cohen. 1980]. as playing a role in the metabolism of tert-butanol. No studies,
23   however, have identified the specific CYPs responsible for the biotransformation of tert-butanol.
24   Pharmacokinetic differences among the fetus, newborns, children, and the aged might alter
25   responses to chemicals compared to adults, resulting in differences in health effects. In the
26   presence of environmental chemicals, metabolic homeostasis is maintained by the liver's ability to
27   detoxify and eliminate xenobiotics.  This process is accomplished, in part, by the expression of
28   xenobiotic metabolizing enzymes and transporters (XMETs), which metabolize and transport
29   xenobiotics and determine whether exposure will result in altered responses. The expression of
30   XMETs, including various CYPs, has been found to be underexpressed in the mouse fetus and
31   neonate [Lee etal.. 2011] and decreased in older mice [Lee etal.. 2011] and rats [Lee etal.. 2008].
32   Decreased ability to detoxify and transport tert-butanol out of the body could result in increased
33   susceptibility to tert-butanol in the  young and old.
34          In regard to cancer, although children are more sensitive than adults to thyroid
35   carcinogenesis resulting from ionizing radiation, relative differences in lifestage sensitivity to
36   chemically induced thyroid carcinogenesis are unknown [U.S. EPA, 1998a]. In addition, the data on
37   tert-butanol mutagenicity are inconclusive.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-65          DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol


1           Collectively, there is little evidence on tert-butanol itself to identify susceptible populations
2    or lifestages.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-66           DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol
     2.    DOSE-RESPONSE ANALYSIS
 2   2.1.   ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER
 3          The reference dose (RfD, expressed in units of mg/kg-day) is defined as an estimate (with
 4   uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human population
 5   (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects
 6   during a lifetime. The RfD can be derived from a no-observed-adverse-effect level (NOAEL), lowest-
 7   observed-adverse-effect level (LOAEL), or the 95% lower bound on the benchmark dose (BMDL),
 8   with uncertainty factors (UFs) generally applied to reflect limitations of the data used.

 9   2.1.1.  Identification of Studies and Effects for Dose-Response Analysis
10          EPA identified kidney effects as a potential human hazard of tert-butanol exposure (see
11   Section 1.2.1). Studies within this effect category were evaluated using general study quality
12   characteristics [as discussed in Section 6 of the Preamble; see also U.S. EPA (2002)] to help inform
13   the selection of studies from which to derive toxicity values. No other hazards were identified for
14   further for consideration in the derivation of reference values.
15          Human studies are preferred over animal studies when quantitative measures of exposure
16   are reported and the reported effects are determined to be associated with exposure. No human
17   occupational or epidemiological studies of oral exposure to tert-butanol, however, are available.
18          Animal studies were evaluated to determine which studies provided: (1) the most relevant
19   routes and durations of exposure, (2) multiple exposure levels to provide information about the
20   shape of the dose-response curve, and (3) power to detect effects atlow exposure levels. The
21   database for tert-butanol includes both chronic and subchronic studies showing effects in the
22   kidney that are suitable for deriving reference values.

23   Kidney Toxicity
24          EPA identified kidney effects as a potential human hazard of tert-butanol-induced toxicity
25   based on findings in male and female rats (summarized in Section 1.3.1). Kidney toxicity was
26   observed across multiple chronic, subchronic, and short-term studies following oral and inhalation
27   exposure. Kidney effects observed after chronic exposure, such as suppurative inflammation and
28   transitional epithelial hyperplasia, may impact the ability of the kidney to filter waste. Observed
29   changes in kidney weight could also indicate toxic effects in the kidney. For the oral tert-butanol
30   database, there are several studies available that evaluated these kidney effects. Lyondell Chemical
31   Co. (2004) conducted a reproductive study in Sprague-Dawley rats that was of shorter duration,
32   and reported changes in kidney weight but did not examine changes in histopathology. NTP
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-1            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   conducted a 2-year drinking water study [NTP. 1995] in F344 rats that evaluated multiple doses in
 2   both males and females, and reported on all three endpoints highlighted above. NTP [1995] was
 3   identified as most suitable for dose-response assessment considering the study duration,
 4   comprehensive reporting of outcomes, and multiple doses tested.
 5          In the NTP [1995] 2-year drinking water study, male F344 rats were exposed to
 6   approximate doses of 0, 90, 200, or 420 mg/kg-day; female F344 rats were exposed to approximate
 7   doses of 0,180, 330, or 650 mg/kg-day. Reduced body weights and survival were observed and
 8   reflected in some of the effects. Kidney effects, including changes in organ weight, histopathology,
 9   or both, were observed in both sexes of rats after 13 weeks, 15 months, and 2 years of treatment
10   [NTP. 1995]. Specific endpoints chosen for dose-response analysis were absolute kidney weight
11   (observed in males and females], kidney suppurative inflammation (observed in females], and
12   kidney transitional epithelial hyperplasia (observed in males and females]. For absolute kidney
13   weight, data from 15 months was selected as described in Section 1.2.1; for the other endpoints,
14   data at the longest duration of 2 years were  selected.

15   2.1.2.  Methods of Analysis
16          No biologically based dose-response models are available for tert-butanol. In this situation,
17   EPA evaluates a range of dose-response models thought to be consistent with underlying biological
18   processes to determine how best to empirically model the dose-response relationship in the range
19   of the observed data. The models in EPA's Benchmark Dose  Software (BMDS] were  applied.
20   Consistent with EPA's Benchmark Dose Technical Guidance [U.S. EPA. 2012b). the BMD and the
21   BMDL are estimated using a benchmark response [BMR] to  represent a minimal, biologically
22   significant level of change. In the absence  of information regarding the level of change that is
23   considered biologically significant, a BMR of 1 standard deviation from the control mean for
24   continuous data or a BMR of 10% extra risk for dichotomous data is used to estimate the BMD and
25   BMDL, and also to facilitate a consistent basis of comparison across endpoints, studies, and
26   assessments. Endpoint-specific BMRs, where feasible, are described further below. When modeling
27   was feasible, the estimated BMDLs were used as points of departure (PODs]; the PODs are
28   summarized in Table 2-1. Further details including the modeling output and graphical results for
29   the model selected for each endpoint can be found in Appendix C of the Supplemental Information
30   to this Toxicological Review.
31          Kidney weights were  analyzed as absolute weights rather than relative to body weight In
32   general, absolute and relative kidney weight data can both be considered appropriate endpoints for
33   analysis [Bailey etal.. 2004]. In the NTP [1995] 2-year drinking water study, body weight in
34   exposed animals noticeably decreased relative to controls at the 15-month interim sacrifice (see
35   Table 1-1], but this decrease in body weight impacted the measure of relative kidney weight
36   resulting in an exaggeration of the kidney weight change. There was greater confidence in the
37   absolute kidney weight measure; thus, it was considered more appropriate for dose-response
38   analysis, and changes in relative kidney weights were not analyzed. A 10% relative  change from
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-2            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   control was used as a BMR for absolute kidney weight by analogy with a 10% change in body
 2   weight as an indicator of toxicity. A BMR of 10% extra risk was considered appropriate for the
 3   quantal data on incidences of kidney suppurative inflammation and kidney transitional epithelial
 4   hyperplasia.
 5          Human equivalent doses (HEDs) for oral exposures were derived from the PODs according
 6   to the hierarchy of approaches outlined in EPA's Recommended Use of Body Weight3/4 as the Default
 7   Method in Derivation of the Oral Reference Dose [U.S. EPA, 2011]. The preferred approach is
 8   physiologically based toxicokinetic modeling (PBPK). Other approaches include using chemical-
 9   specific information in the absence of a complete PBPK model. As discussed in Appendix B of the
10   Supplemental Information, human PBPK models for inhalation of ETBE or inhalation and dermal
11   exposure to MTBE have been published, which include tert-butanol submodels. A validated human
12   PBPK model for tert-butanol, however, is not available for extrapolating doses from animals to
13   humans. In lieu of either chemical-specific models or data to inform the derivation of human
14   equivalent oral exposures, body weight scaling to the % power (i.e., BW3/4) is applied to extrapolate
15   toxicologically equivalent doses of orally administered agents from adult laboratory animals to
16   adult humans for the purpose of deriving an oral RfD.
17          Consistent with EPA guidance (U.S. EPA. 2011). the PODs estimated based on effects in adult
18   animals were converted to HEDs employing a standard dosimetric adjustment factor (DAF) derived
19   as follows:
                                                                %
20                        DAF = (BWa1/* / BWh1/4),
21          where
22                        BWa = animal body weight
23                        BWh = human body weight

24          Using a standard BWa of 0.25 kg for rats and a BWh of 70 kg for humans (U.S. EPA. 1988).
25   the resulting DAF is 0.24 for rats. Applying this DAF to the POD identified for effects in adult rats
26   yields a PODHED as follows (see Table 2-1):

27                PODHED = Laboratory animal dose (mg/kg-day) x DAF

28          Table 2-1 summarizes all PODs and the sequence of calculations leading to the derivation of
29   a human-equivalent POD for each endpoint discussed above.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-3            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
            Table 2-1. Summary of derivations of points of departure following oral
            exposure for up to 2 years
Endpointand
reference
Species/
sex
Model3
BMR
BMD
(mg/kg-d)
BMDL
(mg/kg-d)
PODADjb(m
g/kg-d)
Kidney
Increased absolute
kidney weight at 15
months
NTP (1995)

Increased absolute
kidney weight at 15
months
NTP (1995)

Kidney inflammation
(suppurative)
NTP (1995)

Kidney transitional
epithelial
hyperplasia
NTP (1995)

Kidney transitional
epithelial
hyperplasia
NTP (1995)
Rat/M
Rat/F
Rat/F
Rat/M
Rat/F
Linear
(constant
variance)
Exponential
(M4)
(constant
variance)
Log-probit
Log-logistic
Multistage,
3-degree
10%
10%
10%
10%
10%
657
164
254
30
412
296
91
200
16
339
296
91
200
16
339
PODHEDc(mg/
kg-d)

71
22
48
3.84
81.4
 3
 4
 5
 6
 8
 9
10
11
12
13
14
15
16
17
18
aFor modeling details, see Appendix C in Supplemental Information.
bFor studies in which animals were not dosed daily, EPA would adjust administered doses to calculate the TWA
 daily doses prior to BMD modeling. However, this adjustment was not required for the NTP (1995) study.
CHED PODs were calculated using BW3/4scaling (U.S. EPA, 2011).
NA= not applicable
2.1.3.  Derivation of Candidate Values
       Consistent with EPA's A Review of the Reference Dose and Reference Concentration Processes
[[U.S. EPA. 2002]: Section 4.4.5], also described in the Preamble, five possible areas of uncertainty
and variability were considered when determining the application of UFs to the PODs presented in
Table 2-1. An explanation follows:
       An intraspecies uncertainty factor, UFn, of 10 was applied to all PODs to account for
potential differences in toxicokinetics and toxicodynamics in the absence of information on the
variability of response in the human population following oral exposure to tert-butanol [U.S. EPA.
2002).
       An interspecies uncertainty factor, UFA, of 3 (10°5 = 3.16, rounded to 3) was applied to all
PODs because BW3/4 scaling was used to extrapolate oral doses from laboratory animals to humans.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  2-4            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   Although BW3/4 scaling addresses some aspects of cross-species extrapolation of toxicokinetic and
 2   toxicodynamic processes, some residual uncertainty in the extrapolation remains. In the absence of
 3   chemical-specific data to quantify this uncertainty, EPA's BW3/4 guidance [U.S. EPA. 2011]
 4   recommends use of an uncertainty factor of 3.
 5          A subchronic to chronic uncertainty factor, UFS, of 1 was applied to all PODs because the
 6   endpoints were all observed following chronic exposure.
 7          A LOAEL to NOAEL uncertainty factor, UFL, of 1 was applied to all PODs derived because the
 8   current approach is to address this factor as one of the considerations in selecting a BMR for
 9   benchmark dose modeling. In this case, BMRs of a 10% relative change in absolute kidney weight, a
10   10% extra risk of kidney suppurative inflammation, and a 10% extra risk of transitional cell
11   hyperplasia were selected assuming they represent minimal biologically significant response levels.
12          A database uncertainty factor, UF, of 1 was applied to all PODs. The tert-butanol oral toxicity
13   database includes chronic and subchronic toxicity studies in rats and mice [Acharyaetal.. 1997:
14   Acharyaetal.. 1995: NTP. 1995] and developmental toxicity studies in rats and mice [Lyondell
15   Chemical Co.. 2004: Faulkner et al.. 1989: Daniel and Evans. 1982]. In the developmental studies, no
16   effects were observed at exposure levels below 1000 mg/kg-day, and effects observed at
17   >1000 mg/kg-day were accompanied by evidence of maternal toxicity. These exposure levels are
18   much higher than the PODs for kidney effects, suggesting developmental toxicity is not as sensitive
19   an endpoint as kidney effects. No immunotoxicity or multigenerational reproductive studies are
20   available for tert-butanol. Studies on ETBE, which is rapidly metabolized to systemically available
21   tert-butanol, are informative for consideration of the gaps in the tert-butanol oral database. The
22   database for ETBE  does not indicate immunotoxicity [Bantonetal.. 2011: Lietal.. 2011]. suggesting
23   immune system effects would not be a sensitive target for tert-butanol. No adverse effects were
24   reported in one- and two-generation reproductive/developmental studies on ETBE  [Gaoua. 2004a.
25   b], indicating that reproductive/developmental effects would not be a sensitive target for tert-
26   butanol. Additionally, a one-generation, reproductive toxicity study in rats from a Toxic Substances
27   Control Act submission [Lyondell Chemical Co.. 2004] is available  for tert-butanol. This study did
28   not observe reproductive effects. Although the oral toxicity database for tert-butanol has some
29   gaps, the available data on tert-butanol, informed by the data on ETBE, do not suggest that
30   additional studies would lead to identification of a more sensitive  endpoint or a lower POD.
31   Therefore, a database UFD of 1 was applied.
32          Figure 2-1 presents graphically the candidate values, UFs,  and PODHED values, with each bar
33   corresponding to one data set described in Tables 2-1 and 2-2.
34          Table 2-2 is a continuation of Table 2-1 and summarizes the application of UFs to each POD
35   to derive a candidate value for each data set, preliminary to the derivation of the organ/system-
36   specific RfDs. These candidate values are considered individually in the selection of a
37   representative oral reference value for a specific hazard and subsequent overall RfD for tert-
38   butanol. Figure 2-1 presents graphically the candidate values, UFs, and PODHED values, with each
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-5            DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-Butyl Alcohol
1   bar corresponding to one data set described in Tables 2-1 and 2-2.
           Table 2-2. Effects and corresponding derivation of candidate values


Endpoint and reference

PODhHED
(mg/kg-d)


POD type


UFA


UFH


UFL


UFs


UFD

Composite
UF
Candidate
value
(mg/kg-d)
Kidney
Increased absolute kidney weight;
male rat at 15 months
NTP (1995)

Increased absolute kidney weight;
female rat at 15 months
NTP (1995)

Kidney inflammation (suppurative);
female rat NTP (1995)
Kidney transitional epithelial
hyperplasia; male rat
NTP (1995)

Kidney transitional epithelial
hyperplasia; female rat
NTP (1995)
71



22



48

3.8



81
BMDLio%



BMDLio%



BMDLio%

BMDLio%



BMDLio%
^
3



3



3

3



3

10



10



10

10



10

1



1



1

1



1

1



1



1

1



1

1



1



1

1



1

30



30



30

30



30

2x10°



7 x 10'1



2x10°

1 x 10'1



3x10°

              This document is a draft for review purposes only and does not constitute Agency policy,
                                                2-6           DRAFT—DO NOT CITE OR QUOTE

-------
                                              Toxicological Review of tert-Butyl Alcohol
     T Absolute kidney
     weight; male rat
       (NTP, 1995)
     T Absolute kidney
     weight; female rat
       (NTP, 1995)
 Kidney inflammation;
 female rat (NTP, 1995]
   Kidney transitional
 epithelial hyperplasia;
  male rat (NTP, 1995)
  Kidney transitional
 epithelial hyperplasia;
 female rat (NTP, 1995)
                   0.1
                                                         10
                                                                           100
      Candidate RfD
      PODHED
      Composite UF
mg/kg-day
Figure 2-1. Candidate values with corresponding POD and composite UF. Each
bar corresponds to one data set described in Table 2-1 and Table 2-2.
   This document is a draft for review purposes only and does not constitute Agency policy,
                                       2-7             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
 1
 2
 3
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16

17
2.1.4.  Derivation of Organ/System-Specific Reference Doses
       Table 2-3 distills the candidate values from Table 2-2 into a single value for each organ or
system. Organ or system-specific RfDs are useful for subsequent cumulative risk assessments that
consider the combined effect of multiple agents acting at a common site.
Kidney Toxicity
       For tert-butanol, candidate values were for several different kidney effects in both sexes,
spanning a range from 1 x lO1 to 3 x 10° mg/kg-day, for an overall 30-fold range. To estimate an
exposure level below which kidney toxicity from tert-butanol exposure is not expected to occur, the
RfD for increased incidence of transitional epithelial hyperplasia in male rats (1 x 10'1 mg/kg-day)
was selected as  the kidney-specific reference dose for tert-butanol. Unlike kidney suppurative
inflammation, this effect was observed in both sexes, with males appearing to be more sensitive
than females. Additionally, this indicator of kidney toxicity is more specific and more sensitive than
the relatively non-specific endpoint of absolute kidney weight changes. Confidence in this kidney-
specific RfD is high. The PODs are based on benchmark dose modeling, and the candidate values are
derived from a well-conducted long-term study, involving a sufficient number of animals per group,
including both sexes, and assessing a wide range of kidney endpoints.
       Table 2-3. Organ/system-specific RfDs and overall RfD for tert-butanol
Effect
Kidney
Overall RfD
Basis
Incidence of transitional
epithelial hyperplasia (NTP
(1995)
Kidney
RfD (mg/kg-day)
1 x 10'1
1 x 1C'1
Study exposure
description
Chronic
Chronic
Confidence
High
High
18
19
20
21
22
23
24
25
26
27
28
29
2.1.5.  Selection of the Overall Reference Dose
       For tert-butanol, only kidney effects were identified as a hazard and carried forward for
dose-response analysis; thus only one organ/system-specific reference dose was derived.
Therefore, the kidney specific RfD of (1 x 1Q-1 mg/kg-day) is the overall RfD for tert-butanol. This
value is based on increased incidence of transitional epithelial hyperplasia in male rats exposed to
tert-butanol.
       The overall reference dose is derived to be protective of all types of effects for a given
duration of exposure and is intended to protect the population as a whole, including potentially
susceptible subgroups [U.S. EPA. 2002). Decisions concerning averaging exposures over time for
comparison with the RfD should consider the types of toxicological effects and specific lifestages of
concern. Fluctuations in exposure levels that result in elevated exposures during these lifestages
could lead to an appreciable risk, even if average levels over the full exposure duration were less
          This document is a draft for review purposes only and does not constitute Agency policy.
                                            2-8           DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   than or equal to the RfD. In the case of tert-butanol, there is potential for early lifestage
 2   susceptibility to tert-butanol exposure as discussed in Section 1.3.3.

 3   2.1.6.  Confidence Statement
 4          A confidence level of high, medium, or low is assigned to the study used to derive the RfD,
 5   the overall database, and the RfD, as described in Section 4.3.9.2 of EPA's Methods for Derivation of
 6   Inhalation Reference Concentrations and Application of Inhalation Dosimetry [U.S. EPA, 1994]. The
 7   overall confidence in this RfD is high. Confidence in the principal study [NTP. 1995] is high. This
 8   study was well conducted, complied with Food and Drug Administration (FDA] Good Laboratory
 9   Practice (GLP] regulations, involved a sufficient number of animals per dose group (including both
10   sexes], and assessed a wide range of tissues and endpoints. Although the toxicity database for tert-
11   butanol has some gaps, they are informed by the data on ETBE, a parent compound of tert-butanol.
12   Therefore, the confidence in the database is high. Reflecting high confidence in the principal study
13   and high confidence in the database, confidence in the RfD is high.

14   2.1.7.  Previous IRIS Assessment
15          No previous oral assessment for tert-butanol is available in IRIS.

is   2.2.   INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER
17          THAN CANCER
18          The inhalation RfC (expressed in units of mg/m3] is defined as an estimate  (with
19   uncertainty spanning perhaps an order of magnitude] of a continuous inhalation exposure to the
20   human population (including sensitive subgroups] that is likely to be without an appreciable risk of
21   deleterious effects during a lifetime. It can be derived from a NOAEL, LOAEL, or the 95% lower
22   bound on the benchmark concentration (BMCL], with UFs generally applied to reflect limitations of
23   the data used.

24   2.2.1.  Identification of Studies and Effects for Dose-Response Analysis
25          As for oral exposure, EPA identified kidney effects as a potential human hazard of tert-
26   butanol inhalation exposure (summarized in Section 1.3.1]. No chronic inhalation study for tert-
27   butanol is available; there is only one 13-week study in rats and mice (NTP. 1997]. Sufficient data
28   were available to modify and utilize a PBPK model in rats for both oral and inhalation exposure in
29   order to perform a route-to-route extrapolation, so rat studies from both routes of exposure were
30   considered for dose-response analysis.
31          The database for tert-butanol includes oral and inhalation studies and data sets that are
32   potentially suitable for use in deriving inhalation reference values. Specifically, effects associated
33   with tert-butanol exposure in animals include observations of organ weight and histological
34   changes in the kidney in chronic and subchronic studies in male and female rats.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-9            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-Butyl Alcohol

 1   Kidney Toxicity
 2          EPA identified kidney effects as a potential human hazard of tert-butanol exposure based on
 3   findings of organ weight changes and histopathology primarily in male rats. These findings were
 4   observed across multiple chronic, subchronic, and short-term studies following oral and inhalation
 5   exposure. The subchronic NTP [1997] inhalation study is the only route-specific study available,
 6   and was carried forward for further analysis. For oral studies considered for route-to-route
 7   extrapolation, see Section 2.1.1 for a summary of considerations for selecting oral studies for dose-
 8   response analysis. Overall, the NTP 2-year drinking water study NTP [1995] was identified as the
 9   study most suitable for dose-response assessment, given the study duration, comprehensive
10   reporting of outcomes, use of multiple species tested, multiple doses tested, and availability of a
11   PBPK model for route-to-route extrapolation. This study was discussed previously in Section 2.1.1
12   as part of the derivation of the oral reference dose, so is not reviewed here again. The NTP [1997]
13   subchronic inhalation study shares many strengths with the 2-year drinking water study [NTP,
14   1995], and is described in more detail below.
15          NTP [1997] was a well-designed subchronic study that evaluated the effect of tert-butanol
16   exposure on multiple species at multiple inhalation doses. Absolute kidney weights were elevated
17   [10-11%] in male rats exposed at >3,273 mg/m3; relative kidney weights were elevated [~9%] in
18   males at >3,273 mg/m3 and in females at 6,368 mg/m3. Male rats exhibited an increase in the
19   severity of chronic nephropathy (characterized as number of foci of regenerative tubules]. Few
20   endpoints were available for consideration in the subchronic inhalation study, but changes in
21   kidney weights also were observed in the oral studies, such as the NTP [1995]  2-year drinking
22   water study.

23   2.2.2.  Methods of Analysis
24          No biologically based dose-response models are available for tert-butanol. In this situation,
25   EPA evaluates a range of dose-response models considered consistent with underlying biological
26   processes to determine how best to model the dose-response relationship empirically in the range
27   of the observed data. Consistent with this approach, all models available in EPA's BMDS were
28   evaluated. Consistent with EPA's Benchmark Dose Technical Guidance [U.S. EPA. 2012b], the
29   benchmark dose or concentration [BMD/C] and the 95% lower confidence limit on the BMD/C
30   [BMD/CL] were estimated using a  BMR of 10% change from the control mean for absolute kidney
31   weight changes (as described in  Section 2.1.2]. As noted in Section 2.1.2., a BMR of 10% extra risk
32   was considered appropriate for the quantal data on incidences of kidney suppurative inflammation
33   and kidney transitional epithelial hyperplasia. The estimated BMD/CLs were used as PODs. Where
34   dose-response modeling was not feasible, NOAELs or LOAELs were identified and summarized in
35   Table 2-4. Further details, including the modeling output and graphical results for the best-fit
36   model for each endpoint, can be  found in Appendix C of the Supplemental Information.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-10           DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-Butyl Alcohol

 1   PODsfrom Inhalation Studies
 2          Because the RfC is applicable to a continuous lifetime human exposure but derived from
 3   animal studies featuring intermittent exposure, EPA guidance [U.S. EPA. 1994] provides
 4   mechanisms for (1] adjusting experimental exposure concentrations to a value reflecting
 5   continuous exposure duration (ADJ) and (2) determining a human equivalent concentration (HEC)
 6   from the animal exposure data. The former employs an inverse concentration-time relationship to
 7   derive a health-protective duration adjustment to time weight the intermittent exposures used in
 8   the studies. The modeled benchmark concentration from the inhalation study [NTP, 1997] was
 9   adjusted to reflect a continuous exposure by multiplying it by (6 hours per day] 4- (24 hours per
10   day] and (5 days per week] 4 (7 days per week] as follows:

11          BMCLADj       =      BMCL (mg/m3] x (6 4 24] x (5 4 7]
12                        =      BMCL (mg/m3] x (0.1786]

13          The RfC methodology provides a mechanism for deriving an HEC from the duration-
14   adjusted POD (BMCLADj] determined from the animal data. The approach takes into account the
15   extra-respiratory nature of the toxicological responses and accommodates species differences by
16   considering blood:air partition coefficients for tert-butanol in the laboratory animal (rat or mouse]
17   and humans. According to the RfC guidelines (U.S. EPA. 1994]. tert-butanol is a Category 3 gas
18   because extra-respiratory effects were observed. Kaneko etal. (2000] measured a blood:gas
19   partition coefficient [(Hb/g]A] of531 ± 102 for tert-butanol in the male Wistar rat, while Borghoff et
20   al. (1996] measured a value of 481 ± 29 in male F344 rats. Ablood:gas partition coefficient
21   [(Hb/g]n] of 462 was reported for tert-butanol in humans (Nihlen et al.. 1995]. The calculation
22   (Hb/g]A 4 (Hb/g]H was used to calculate a blood:gas partition coefficient ratio to apply to the
23   delivered concentration. Because F344 rats were used in the study, the blood:gas partition
24   coefficient for F344 rats was used. Thus, the calculation was 481 4 462 = 1.04. Therefore, a ratio of
25   1.04 was used to calculate the HEC. This allowed a BMCLHEc to be derived as follows:

26          BMCL.HEC      =      BMCLADj (mg/m3] x (interspecies conversion]
27                        =      BMCLADj (mg/m3] x (481 4 462]
28                        =      BMCLADj (mg/m3] x (1.04]

29          Table 2-4 summarizes the sequence of calculations leading to the derivation of a human-
30   equivalent POD for each inhalation data set discussed above.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-11          DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
            Table 2-4. Summary of derivation of PODs following inhalation exposure
Endpoint and
reference
Species/
Sex
Model"
BMR
BMCb
(mg/m3)
BMCLb
(mg/m3)
PODADjb
(mg/m3)
PODHEcc(
mg/m3)
Kidney
Increased absolute
kidney weight
NTP (1997)

Increased absolute
kidney weight
NTP (1997)
Male F344
rats
Female F344
rats
Hill
No model
selectedd
10%
10%
1931
~
1705
~
304
1137
304
1137
 2   aFor modeling details, see Appendix C in Supplemental Information.
 3   bBMCs, BMCLs, and PODs were adjusted for continuous daily exposure by multiplying by (hours exposed per day /
 4    24 hr) x (days exposed per week / 7 days).
 5   CPODHEC calculated by adjusting the PODAoj by the DAF (=1.0, rounded from 1.04) for a Category 3 gas (U.S. EPA,
 6    1994).
 7   dBMD modeling failed to calculate a BMD value successfully (see Appendix C); POD calculated from no-observed
 8    adverse effect level (NOAEL) of 6368 mg/m3.

 9   PODs from oral studies - use ofPBPK model for route-to-route extrapolation
10          A PBPK model for tert-butanol in rats has been modified, as described in Appendix B of the
11   Supplemental Information. Using this model, route-to-route extrapolation of the oral BMDLs to
12   derive inhalation PODs was performed as follows. First, the internal dose in the rat at each oral
13   BMDL (assuming continuous exposure) was estimated using the PBPK model, to derive an "internal
14   dose BMDL." Then, the inhalation air concentration (again, assuming continuous exposure) that led
15   to the same internal dose in the rat was estimated using the PBPK model. The resulting BMCL was
16   then converted to a human equivalent concentration POD (PODHEc) using the methodology
17   previously described in "PODs from inhalation studies":

18          BMCL.HEC      =      BMCLADj (mg/m3) x (interspecies conversion)
19                        =      BMCLADj (mg/m3) x (481 -H 462)
20                        =      BMCLADj (mg/m3) x (1.04)

21          A critical decision in the route-to-route extrapolation is selection of the internal dose metric
22   that establishes "equivalent" oral and inhalation exposures. For tert-butanol-induced kidney effects,
23   the two options are the concentration of tert-butanol in blood and rate of tert-butanol metabolism.
24   Note that using the kidney concentration of tert-butanol will lead to the same route-to-route
25   extrapolation relationship as tert-butanol in blood because the distribution from blood to kidney is
26   independent of route. Data are not available that suggest that metabolites of tert-butanol mediate
27   its renal toxicity. Without evidence that suggests otherwise, tert-butanol is assumed the active
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  2-12           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol

 1   toxicological agent. Therefore, the concentration of tert-butanol in blood was selected as the dose
 2   metric.
 3          Table 2-5 summarizes the sequence of calculations leading to the derivation of a human-
 4   equivalent inhalation POD from each oral data set discussed above.
            Table 2-5. Summary of derivation of inhalation points of departure derived
            from route-to-route extrapolation from oral exposures


Endpoint and reference


Species/sex


BMR

BMDL
(mg/kg-d)

Internal dose3
(mg/L)
Equivalent
PODHEcb
(mg/m3)
Kidney
Mean absolute kidney weight
at 15 months NTP (1995)

Mean absolute kidney weight
at 15 months NTP (1995)

Kidney inflammation
(suppurative) NTP (1995)

Kidney transitional epithelial
hyperplasia
NTP (1995)


Kidney transitional epithelial
hyperplasia
NTP (1995)
Rat/M


Rat/F


Rat/F


Rat/M
10%


10%


10%


10%


Rat/F


10%


296


91


200


16


339


22.4


4.76


12.6


0.745


27.9


551


155


359


26.1


638


 7   a Average blood concentration of tert-butanol under continuous oral exposure at the BMDL
 8   b Continuous inhalation human equivalent concentration that leads to the same average blood concentration of
 9    tert-butanol as continuous oral exposure at the BMDL.

10   2.2.3.   Derivation of Candidate Values
11           In EPA's A Review of the Reference Dose and Reference Concentration Processes [[U.S. EPA,
12   2002]: Section 4.4.5], also described in the Preamble, five possible areas of uncertainty and
13   variability were considered. Several PODs for the candidate inhalation values were derived using a
14   route-to-route extrapolation from the PODs estimated from the chronic oral toxicity study in rats
15   [NTP. 1995] in the derivation of the oral RfD (Section 2.1]. With the exception of the subchronic
16   inhalation [NTP. 1997] study, the uncertainty factors (UFs] selected and applied to PODs derived
17   from the chronic oral [NTP, 1995] study for route-to-route extrapolation are the same as those for
18   the RfD for tert-butanol (see Section 2.1.3). The model used to perform this route-to-route
19   extrapolation is a well-characterized model considered appropriate for the purposes of this
20   assessment One source of uncertainty regarding the route-to-route extrapolation is the assumption
21   of that 100% of inhaled tert-butanol reaches the gas-exchange region, that  is, 100% of the inhaled
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  2-13           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   tert-butanol could be absorbed and distributed to the rest of the body in rats. If not all of the
 2   compound is bioavailable for the rat, a lower blood concentration would be expected compared to
 3   the current estimate, and thus, a higher RfC would be calculated.
 4          For the PODs derived from the subchronic inhalation [NTP. 1997] study, a UFS of 10 was
 5   applied to account for extrapolation from subchronic to chronic duration.
 6          Table 2-6 is a continuation of Table 2-4 and Table 2-5, and summarizes the application of
 7   UFs to each POD to derive a candidate value for each data set. The candidate values presented in the
 8   table below are preliminary to the derivation of the organ/system-specific reference values. These
 9   candidate values are considered individually in the selection of a representative inhalation
10   reference value for a specific hazard and subsequent overall RfC for tert-butanol.
11          Figure 2-2 presents graphically the candidate values, UFs, and PODnEc values, with each bar
12   corresponding to one data set described in Tables 2-4, 2-5, and 2-6.
13
Table 2-6. Effects and corresponding derivation of candidate values

Endpoint (sex and species) and
reference

PODHEC3
(mg/m3)

POD
type


UFA


UFH


UFL


UFs


UFD

Composite
UF
Candidate
value
(mg/m3)
Kidney
Increased absolute kidney weight
at 13 weeks; male rat
NTP (1997)

Increased absolute kidney weight
at 13 weeks; female rat
NTP (1997)

Increased absolute kidney weight
at 15 months; male rat
NTP (1995)

Increased absolute kidney weight
at 15 months; female rat
NTP (1995)

Kidney inflammation
(suppurative); female rat
NTP (1995)

Kidney transitional epithelial
hyperplasia; male rat
NTP (1995)

Kidney transitional epithelial
hyperplasia; female rat
NTP (1995)
304



1137



551



155



359



26.1



638


BMCLio%



NOAEL



BMCLio%



BMCLio%



BMCLio%



BMCLio%



BMCLio%


3



3



3



3



3



3



3


10



10



10



10



10



10



10


1



1



1



1



1



1



1


10



10



1



1



1



1



1


1



1



1



1



1



1



1


300



300



30



30



30



30



30


1x10°



4x10°



2 x 101 *



5 x 10° *



1 x 101 *



9 x 10'1 *



2 x 101 *


               This document is a draft for review purposes only and does not constitute Agency policy,
                                                 2-14           DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review of tert-ButyI Alcohol
1    These candidate values are derived using route-to-route extrapolated PODs based on NTP's chronic drinking
2     water study.
              TAbsolute kidney weight;
                male rat [NTP, 1997]
              TAbsolute kidney weight;
               female rat (NTP, 1997]
              TAbsolute kidney weight;
                male rat (NTP, 1995]
              tAbsolute kidney weight;
               female rat (NTP, 1995]
                 Kidney inflammation;
                female rat (NTP, 1995]
                  Kidney transitional
                 epithelial hyperplasia;
                 male rat (NTP, 1995]
                  Kidney transitional
                epithelial hyperplasia;
                female rat (NTP, 1995]
                   Candidate RfC
                   PODHEC
                   Composite UF
0.1
                       10         100

                         mg/m3
                                             1000       10000
             Figure 2-2. Candidate RfC values with corresponding POD and composite UF.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                       2-15            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
 1
 2
 3
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21

22
2.2.4.  Derivation of Organ/System-Specific Reference Concentrations
       Table 2-7 distills the candidate values from Table 2-6 into a single value for the kidney.
Organ- or system-specific reference values can be useful for subsequent cumulative risk
assessments that consider the combined effect of multiple agents acting at a common site.
Kidney Toxicity
       For the derivation of candidate values, whether PODs from the subchronic inhalation study
of NTP [1997] would provide a better basis than the route-to-route extrapolated PODs based on the
chronic oral study of NTP [1995] must be considered. Candidate values were derived for increased
kidney weight observed in the subchronic inhalation study [NTP, 1997] and several kidney effects
observed in the chronic oral study [NTP. 1995] in both sexes of rat, spanning a range from 9 x lO1
to 2 x 101 mg/m3, for an overall 20-fold range. To estimate an exposure level below which kidney
toxicity from tert-butanol exposure  is not expected to occur, the RfC for increased incidence of
transitional epithelial hyperplasia in male rats (9 x 1Q-1 mg/m3] was selected as the kidney-specific
RfC for tert-butanol, consistent with the selection of the kidney-specific RfD (see Section 2.1.4). As
discussed in Section 2.1.4, unlike kidney suppurative inflammation, this effect was observed in both
sexes, with males appearing to be more sensitive than females. Additionally, it is based on a longer
(chronic) duration and a more specific and sensitive indicator of kidney toxicity than the relatively
non-specific endpoint of kidney weight change. Confidence in this kidney-specific RfC is high. The
PODs are based on BMD modeling, and the candidate values  are derived from a well-conducted
study, involving a sufficient number of animals per group, including both sexes, assessing a wide
range of kidney endpoints, and availability of a PBPK model for route-to-route extrapolation.
       Table 2-7. Organ/system-specific RfCs and  overall RfC for tert-butanol
Effect
Kidney
Overall RfC
Basis
Incidence of transitional
epithelial hyperplasia (NTP,
1995)

Kidney
RfC
(mg/m3)
9 x 10'1
9 x 1C'1
Study exposure
description
Chronic
Chronic
Confidence
High
High
23
24
25
26
27
28
29
2.2.5.  Selection of the Overall Reference Concentration
       For tert-butanol, kidney effects were identified as the primary hazard; thus, a single
organ-/system-specific RfC was derived. The kidney-specific RfC of 9 x 1Q-1 mg/m3 is selected as
the overall RfC, representing an estimated exposure level below which deleterious effects from
tert-butanol exposure are not expected to occur.
       The overall RfC is derived to be protective of all types of effects for a given duration of
exposure and is intended to protect the population as a whole, including potentially susceptible
          This document is a draft for review purposes only and does not constitute Agency policy.
                                             2-16          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol

 1   subgroups [U.S. EPA. 2002]. Decisions concerning averaging exposures over time for comparison
 2   with the RfC should consider the types of toxicological effects and specific lifestages of concern.
 3   Fluctuations in exposure levels that result in elevated exposures during these lifestages could
 4   potentially lead to an appreciable risk, even if average levels over the full exposure duration were
 5   less than or equal to the RfC. In the case of tert-butanol, there is potential for early lifestage
 6   susceptibility to tert-butanol exposure as discussed in Section 1.3.3.

 7   2.2.6.  Confidence Statement
 8          A confidence level of high, medium, or low is assigned to the study used to derive the RfC,
 9   the overall database, and the RfC itself, as described in Section 4.3.9.2 of EPA's Methods for
10   Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry [U.S. EPA.
11   1994). A PBPK model was utilized to perform a route-route extrapolation to determine a POD for
12   the derivation of the RfC from the NTP [1995] oral study and corresponding critical effect.
13   Confidence in the principal study [NTP.  1995] is high. This study was well-conducted, complied
14   with FDA GLP  regulations, involved a sufficient number of animals per group (including both
15   sexes], and assessed a wide range of tissues and endpoints. Although there are some gaps in the
16   toxicity database for tert-butanol, these  areas are informed by the data on ETBE, a parent
17   compound of tert-butanol. Therefore, the confidence in the database is high. Reflecting high
18   confidence in the principal study, high confidence in the database, and minimal uncertainty
19   surrounding the application of the modified PBPK model for the purposes of a route-to-route
20   extrapolation,  the overall confidence in the RfC for tert-butanol is high.

21   2.2.7.  Previous IRIS Assessment
22          No previous inhalation assessment for tert-butanol is available in IRIS.

23   2.2.8.  Uncertainties in the Derivation of the Reference Dose and Reference Concentration
24          The following discussion identifies uncertainties associated with the RfD and RfC for
25   tert-butanol. To derive the RfD, the UF approach [U.S. EPA. 2000a. 1994] was applied to a POD
26   based on kidney toxicity in rats treated chronically. UFs were applied to the POD to account for
27   extrapolating from an animal bioassay to human exposure, and the likely existence of a diverse
28   human population of varying susceptibilities. These extrapolations are carried out with default
29   approaches given the lack of data to inform individual steps. To derive the RfC, this same approach
30   was applied, but a PBPK model was used to extrapolate from oral to inhalation exposure.
31          The database for tert-butanol contains no human data on adverse health effects from
32   subchronic or  chronic exposure, and the PODs were calculated from data on the effects of tert-
33   butanol reported by studies in rats. The  database for tert-butanol exposure  includes one lifetime
34   bioassay, several reproductive/developmental studies, and several subchronic oral studies.
35          Although the database is adequate for reference value derivation, there is uncertainty
36   associated with the lack of a comprehensive multigeneration reproductive toxicity study.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  2-17            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-Butyl Alcohol

 1   Additionally, only subchronic and short-term inhalation studies have been conducted, and no
 2   chronic inhalation studies are available. Developmental studies identified significant increases in
 3   fetal loss, decreases in fetal body weight, and possible increases in skeletal variations in exposed
 4   offspring or pups. However, effects were not always consistent across exposure routes, and
 5   maternal toxicity was present whenever developmental effects were observed.
 6          The toxicokinetic and toxicodynamic differences for tert-butanol between the animal
 7   species in which the POD was derived and humans are unknown. The tert-butanol database lacks
 8   an adequate model that would inform potential interspecies differences (A limited data set exists
 9   for tert-butanol appearing as a metabolite from ETBE exposure in humans, but none for direct
10   exposure to tert-butanol.) Generally, it was found that rats appear more susceptible than mice, and
11   males appear more susceptible than females to tert-butanol toxicity. However, the underlying
12   mechanistic basis of these apparent differences is not understood. Most importantly, it is unknown
13   which animal species and/or sexes may be more comparable to humans.
14          Another uncertainty to consider relates to the MOA analysis conducted for the kidney
15   effects. The assessment concluded that tert-butanol is a weak inducer of (X2u-globulin which is
16   operative in male kidney tumors; therefore, noncancer effects related to (X2u-globulin were
17   considered not relevant for hazard identification and, therefore, not suitable for dose response
18   consideration. If this conclusion was incorrect and  the noncancer effects characterized in this
19   assessment as being related to a2u-globulin were relevant to humans, then the RfD and RfC values
20   could be underestimating toxicity. Similarly, the renal effects characterized as CPN and dismissed as
21   not being treatment related, if considered relevant, would likewise contribute to the hazard
22   potential and dose-response analysis for the kidney-specific RfD and RfC.

23   2.3.  ORAL SLOPE FACTOR FOR CANCER
24          The oral slope factor (OSF) is a plausible upper bound on the estimate of risk per mg/kg day
25   of oral exposure. The OSF can be multiplied by an estimate of lifetime exposure (in mg/kg-day) to
26   estimate the lifetime cancer risk.

27   2.3.1.  Analysis of Carcinogenicity Data
28          As noted in Section 1.3.2, that there is "suggestive evidence of carcinogenic potential" for
29   tert-butanol. The Guidelines for Carcinogen Risk Assessment [U.S. EPA. 2005a] state:

30          When there is suggestive evidence, the Agency generally would not attempt a dose-
31          response assessment, as the nature of the data generally would not support one; however
32          when the evidence includes a well-conducted study, quantitative analysis may be useful for
33          some purposes, for example, providing a sense of the magnitude and uncertainty of
34          potential risks, ranking potential hazards, or setting research priorities.
35          No human data relevant to an evaluation of the carcinogenicity of tert-butanol were
36   available. The cancer descriptor was based on the 2-year drinking water study in rats and mice by

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-18          DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol

 1   [NTP. 1995]. which reported renal tumors in male rats and thyroid tumors in both male and female
 2   mice. This study was considered suitable for dose-response analysis. It was conducted in
 3   accordance with FDA GLP regulations, and all aspects were subjected to retrospective quality
 4   assurance audits. The study included histological examinations for tumors in many different
 5   tissues, contained three exposure levels and controls, contained adequate numbers of animals per
 6   dose group (~50/sex/group], treated animals for up to 2 years, and included detailed reporting of
 7   methods and results. Additionally, the  renal tumors were reexamined by a Pathology Working
 8   Group  [Hard etal.. 2011].
 9          Based on a mode of action analysis, it was concluded that the (X2u-globulin process was at
10   least partially responsible for the male rat renal tumors, in addition to other, unknown, processes.
11   Because the relative contribution of each process to tumor formation cannot be determined [U.S.
12   EPA. 1991a]. the male rat renal tumors are not considered suitable for quantitative analysis.
13   Conversely, the mouse thyroid tumors are suitable for dose-response analysis and unit risk
14   estimation, as described in Section 1.3.2.

15   2.3.2.  Dose-Response Analysis—Adjustments and Extrapolations Methods
16          The EPA Guidelines for Carcinogen Risk Assessment [U.S. EPA, 2005a] recommend that the
17   method used to characterize and quantify cancer risk from a chemical be determined by what is
18   known about the MOA of the carcinogen and the shape of the cancer dose-response curve.  EPA
19   uses a two-step approach that distinguishes analysis of the observed dose-response data from
20   inferences about lower doses [U.S. EPA, 2005a]. Within the observed range, the preferred approach
21   is to use modeling to incorporate a wide range of data into the analysis, such as through a
22   biologically based model, if supported  by substantial data. Without a biologically based model, as  in
23   the case of tert-butanol, a standard model is used to curve-fit the data and estimate a POD. EPA uses
24   the multistage model in IRIS dose-response analyses for cancer [Gehlhaus etal.. 2011] because it
25   parallels the multistage carcinogenic process and fits a broad array of dose-response patterns.
26          The second step, extrapolation to lower exposures from the POD, considers what is known
27   about the modes of action for each effect As above, a biologically based model is preferred [U.S.
28   EPA, 2005a]. Otherwise, linear low-dose extrapolation is recommended if the MOA of
29   carcinogenicity is  mutagenic or has not been established [U.S. EPA, 2005a]. For tert-butanol, the
30   mode(s] of carcinogenic action for thyroid follicular cell tumors has not been established (see
31   Section 1.3.2]. Therefore, linear low-dose  extrapolation was used to estimate human carcinogenic
32   risk.
33          The dose-response modeling used administered dose because a PBPK model to characterize
34   internal dosimetry in mice was not available. For the analysis of male mice thyroid tumors, the
35   incidence data were adjusted to account for the increased mortality in high-dose male mice, relative
36   to the other groups, that reduced the number of mice at risk for developing tumors. The Poly-3
37   method [Bailer and Portier, 1988] was used to estimate the number at risk of developing tumors,
38   by weighting the length of time each animal was on study (details in Appendix C of the
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-19          DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-Butyl Alcohol

 1   Supplemental Information). This method was not applied to the female mice data because a
 2   difference in survival with increasing exposure was not appreciable and only one tumor, in the
 3   high-dose group, occurred before study termination.
 4          The data modeled and other details of the modeling are provided in Appendix C. The BMDs
 5   and BMDLs recommended for each data set are summarized in Table 2-8. The modeled tert-butanol
 6   PODs were scaled to HEDs according to EPA guidance (U.S. EPA. 2011. 2005al In particular, the
 7   BMDL was converted to an HED by assuming that doses in animals and humans are toxicologically
 8   equivalent when scaled by body weight raised to the 3/4 power. Standard body weights of 0.025 kg
 9   for mice and 70 kg for humans were used (U.S. EPA. 1988]. The following formula was used for the
10   conversion of oral BMDL to oral HED for mouse endpoints:

11                 HED in mg/kgday- = (BMDL in mg/kgday-) x (animal body weight/70)1/4
12                                      = (BMDL in mg/kgday-) x 0.14

13          PODs for estimating low-dose risk were identified at doses at the lower end of the observed
14   data, corresponding to 10% extra risk in female mice and 5% extra risk in male mice.

15   2.3.3.  Derivation of the Oral Slope Factor
16          The PODs estimated for each tumor data set are summarized in Table 2-8. The lifetime oral
17   cancer slope factor for humans is defined as the slope of the line from the lower 95% bound on the
18   exposure atthe POD to the control response (slope factor = BMR/BMDLBMR = 0.1/BMDLio). This
19   slope represents a plausible upper bound on the true population average risk. Using linear
20   extrapolation from the BMDLio, human equivalent oral slope factors were derived for male and
21   female mice and are listed in Table 2-8.
22          The oral slope factor based on the incidence of thyroid follicular cell adenomas in female
23   mice was 5 x 104 per mg/kg-day. Despite high mortality in high-dose male mice, estimating slope
24   factors using the poly-3 method was feasible for addressing competing risks. Whether using the full
25   data set (including the only thyroid follicular cell carcinoma observed at the highest dose) or
26   omitting the high-dose group altogether (under the assumption that mortality in this group was too
27   extensive to interpret the results), oral slope factors based on the incidence of thyroid follicular cell
28   adenomas or carcinomas in male mice were similar when rounded to one significant digit—5 x 1Q-4
29   per mg/kg-day or 6 x 1Q-4 per mg/kg-day, respectively.
30          The recommended slope factor for lifetime oral exposure to tert-butanol is
31   5 x KM per mg/kg-day, based on the thyroid follicular cell adenoma or carcinoma response in
32   male or female B6C3Fi mice. This slope factor should not be used with exposures exceeding
33   1400 mg/kg-day, the highest POD from the two data sets, because above this level the cancer risk
34   might not increase linearly with exposure. The slope of the linear extrapolation from the central
35   estimate BMDioHED derived from the female mouse data set is 0.1/[0.14 x (2002 mg/kg-day)] =
36   4 x IQ-4 per mg/kg-day.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-20           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-Butyl Alcohol
            Table 2-8. Summary of the oral slope factor derivation



Tumor
Thyroid follicular
cell adenoma
Thyroid follicular
cell adenoma or
carcinoma






Species/ sex
B6C3Fi
mouse/Female
B6C3Fi
mouse/Male






Selected
model
3° Multistage

All dose
groups: 1°
Multistage
High dose
omitted: 2°
Multistage



BMR
10%

5%c


5%c



BMD
(mg/kg-
d)
2002

1788


1028


POD=
BMDL
(mg/kg-
d)
1437

787


644




BMDLHEDa
(mg/kg-d)
201

110


90



Slope factor15
(mg/kg-day)
i
5 x 10'4

5 x 10'4


6 x 10'4


 2   aHED PODs were calculated using BW3/4scaling (U.S. EPA, 2011).
 3   bHuman equivalent slope factor = 0.1/BMDLioHEo; see Appendix C of the Supplemental Information for details of
 4    modeling results.
 5   °Because the observed responses were <10%, a BMR of 5% was used to represent the observed response range for
 6    low-dose extrapolation; human equivalent slope factor = 0.05/BMDLsHED.

 7   2.3.4. Uncertainties in the Derivation of the Oral Slope Factor
 8          There is uncertainty when extrapolating data from animals to estimate potential cancer
 9   risks to human populations from exposure to tert-butanol.
10          Table 2-9 summarizes several uncertainties that could affect the oral slope factor. There are
11   no other chronic studies to replicate these findings or that examined other animal models, no data
12   in humans to confirm a cancer response in general or the specific tumors observed in the NTP
13   [1995] bioassay, and no other data (e.g., MOA) to support alternative approaches for deriving the
14   oral slope factor.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                   2-21           DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review of tert-ButyI Alcohol
       Table 2-9. Summary of uncertainties in the derivation of the oral slope factor
       for tert-butanol
       Consideration and
   impact on cancer risk value
                                         Decision
                                       Justification
Selection of tumor type and
relevance to humans:
Mouse thyroid tumors are the basis
for estimating human cancer risk, as
the fraction of rat kidney tumors not
attributed to the male rat specific
o2u.-globulin process could not be
determined. Alternatively,
quantifying rat kidney tumors could
T* slope factor to 1 x 10'2 mg/kg-day
(see Appendix C, Supplemental
Information)
                                 Thyroid tumors in female
                                 and male mice were
                                 selected U.S. EPA(1998a),
                                 U.S. EPA(1991a).
                          MOA data suggested that mouse thyroid
                          tumors were relevant to humans.
                          Quantitation of thyroid tumors in male mice
                          was impacted only slightly by high mortality
                          in the high-dose group, and supports the
                          estimate based on female mice.
Selection of data set:
No other studies are available.
                                  NTP (1995), oral (drinking
                                 water) study, was selected
                                 to derive cancer risks for
                                 humans.
                          NTP (1995), the only chronic bioassay
                          available, was a well-conducted study.
                          Additional bioassays might add support to
                          the findings, facilitate determination of what
                          fraction of kidney tumors are not attributable
                          to a2u.-globulin process, or provide results
                          for different (possibly lower) doses, which
                          would affect (possibly increase) the oral
                          slope factor.
Selection of dose metric:
Alternatives could 4, or 1
factor
                                 Used administered dose.
                        slope
                          For mice, PBPK-estimated internal doses
                          could impact the OSF value for thyroid
                          tumors if the carcinogenic moiety is not
                          proportional to administered dose, but no
                          PBPK model was available, and no
                          information is available to suggest if any
                          metabolites elicit carcinogenic effects.
Interspecies extrapolation of
dosimetry and risk:
Alternatives could 4, or ^ slope
factor (e.g., 3.5-fold 4, [scaling by
body weight] or ^ 2-fold [scaling by
BW2/3])
                                 The default approach of
                                 body weight374 was used.
                          No data to suggest an alternative approach
                          for tert-butanol. Because the dose metric
                          was not an area under the curve, BW3/4
                          scaling was used to calculate equivalent
                          cumulative exposures for estimating
                          equivalent human risks. Although the true
                          human correspondence is unknown, this
                          overall approach is expected neither to over-
                          or underestimate human equivalent risks.
Dose-response modeling:
Alternatives could 4, or 1"
factor
                        slope
Used multistage dose-
response model to derive a
BMDand BMDL
No biologically based models for tert-butanol
were available. The multistage model has
biological support and is the model most
consistently used in EPA cancer assessments.
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                 2-22            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review of tert-ButyI Alcohol
            Consideration and
         impact on cancer risk value
       Decision
             Justification
      Low-dose extrapolation:
      4, cancer risk estimate would be
      expected with the application of
      nonlinear low-dose extrapolation
Linear extrapolation of risk
in low-dose region used
U.S. EPA(1998a).
Linear low-dose extrapolation for agents
without a known MOA is supported (U.S.
EPA, 2005a) and recommended for rodent
                       thyroid tumors arising from an unknown
                       MOA (U.S. EPA, 1998a).
      Statistical uncertainty at POD:
      4> oral slope factor 1.4-fold if BMD
      used as the POD rather than BMDL
BMDL (preferred approach
for calculating slope factor)
Limited size of bioassay results in sampling
variability; lower bound is 95% Cl on
administered exposure at 10% extra risk of
thyroid tumors.
      Sensitive subpopulations:
      1" oral slope factor to unknown
      extent
No sensitive populations
have been identified.
No chemical-specific data are available to
determine the range of human
toxicodynamic variability or sensitivity,
including the susceptibility of children.
Because determination of a mutagenic MOA
has not been made, an age-specific
adjustment factor is not applied.
 1    2.3.5.  Previous IRIS Assessment: Oral Slope Factor
 2           No previous cancer assessment for tert-butanol is available in IRIS.

 3    2.4.   INHALATION UNIT RISK FOR CANCER
 4           The carcinogenicity assessment provides information on the carcinogenic hazard potential
 5    of the substance in question, and quantitative estimates of risk from oral and inhalation exposure
 6    may be derived. Quantitative risk estimates may be derived from the application of a low-dose
 7    extrapolation procedure. If derived, the inhalation unit risk (IUR) is a plausible upper bound on the
 8    estimate of risk per [J.g/m3 air breathed.
 9           No chronic inhalation exposure studies to tert-butanol are available. Lifetime oral exposure
10    has been associated with increased renal tubule adenomas and carcinoma in male F344 rats,
11    increased thyroid follicular cell adenomas in female B6C3Fi mice,  and increased thyroid follicular
12    cell adenomas and carcinomas in male B6C3Fi mice. Because only a rat PBPK model exists,
13    however, route-to-route extrapolation cannot be performed for thyroid tumors in mice at this time.
14    The NTP [1995] drinking water study in rats and mice was the only chronic bioassay available for
15    dose-response analysis. Still, the rat PBPK model and kidney tumors from the NTP [1995] drinking
16    water study were not used for route-to-route extrapolation because enough information to
17    determine the relative contribution of (X2u-globulin nephropathy and other processes to the overall
18    renal tumor response [U.S. EPA. 1991a] is not available. Alternatively, if kidney tumors were
19    considered acceptable for quantitation, then route-to-route extrapolation could be conducted to
20    calculate an IUR (see Appendix  C in Supplemental Information].
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                   2-23           DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review of tert-ButyI Alcohol

1   2.4.1.  Previous IRIS Assessment: Inhalation Unit Risk
2          An inhalation cancer assessment for tert-butanol was not previously available on IRIS.

3   2.5. APPLICATION OF AGE-DEPENDENT ADJUSTMENT FACTORS
4   As discussed in the Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to
5   Carcinogens [U.S. EPA. 2005b]. either default or chemical-specific age-dependent adjustment
6   factors (ADAFs) are recommended to account for early-life exposure to carcinogens that act
7   through a mutagenic MOA Because chemical-specific lifestage susceptibility data for cancer are not
8   available, and because the MOA for tert-butanol carcinogenicity is not known (see Section 1.3.2),
9   application of ADAFs is not recommended.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              2-24          DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-Butyl Alcohol

     REFERENCES
 1   Acharya. S: Mehta. K: Rodrigues. S: Pereira. I: Krishnan. S: Rao. CV. [1995]. Administration of
 2          subtoxic doses of t-butyl alcohol and trichloroacetic acid to male Wistar rats to study the
 3          interactive toxicity. Toxicol Lett 80: 97-104. http://dx.doi.org/10.1016/0378-
 4          4274(95)03340-0.
 5   Acharya, S: Mehta, K: Rodriguez, S: Pereira, T: Krishnan, S: Rao, CV. (1997). Ahistopathological study
 6          of liver and kidney in male Wistar rats treated with subtoxic doses of t-butyl alcohol and
 7          trichloroacetic acid. Exp Toxicol Pathol 49: 369-373.
 8   Amberg. A: Rosner. E: Dekant. W. (1999). Biotransformation and kinetics of excretion of methyl-
 9          tert-butyl ether in rats and humans. Toxicol Sci 51: 1-8.
10   Amberg. A: Rosner. E: Dekant. W. (2000). Biotransformation and kinetics of excretion of ethyl tert-
11          butyl ether in rats and humans. Toxicol Sci 53: 194-201.
12          http://dx.doi.0rg/10.1093/toxsci/53.2.194.
13   ARCO (ARCO Chemical Company). (1983). Toxicologist's report on metabolism and
14          pharmacokinetics of radiolabeled TEA 5 34 tertiary butyl alcohol with cover letter dated
15          03/24/1994. (8EHQ86940000263). Newton Square, PA.
16   ATSDR (Agency for Toxic Substances and Disease Registry). (1996). Toxicological profile for
17          methyl-tert-butyl ether [ATSDR Tox Profile]. Atlanta, GA: U.S. Department of Health and
18          Human Services, Public Health Service. http://www.atsdr.cdc.gov/ToxProfiles/tp91.pdf.
19   Bailer. AT: Portier. CT. (1988). Effects of treatment-induced mortality and tumor-induced mortality
20          on tests for carcinogenicity in small samples. Biometrics 44: 417-431.
21   Bailey, SA: Zidell, RH: Perry, RW. (2004).  Relationships between organ weight and body/brain
22          weight in the rat: What is the best analytical endpoint? Toxicol Pathol 32: 448-466.
23          http://dx.doi.org/10.1080/01926230490465874.
24   Banton. MI: Peachee. VL: White. KL: Padgett. EL. (2011). Oral subchronic immunotoxicity study of
25          ethyl tertiary butyl ether in the rat. J Immunotoxicol 8: 298-304.
26          http://dx.doi.org/10.3109/1547691X.2011.598480.
27   Bernauer, U: Amberg, A: Scheutzow, D: Dekant, W. (1998). Biotransformation of 12C- and 2-13C-
28          labeled methyl tert-butyl ether, ethyl tert-butyl ether, and tert-butyl alcohol in rats:
29          identification of metabolites in urine by 13C nuclear magnetic resonance and gas
30          chromatography/mass spectrometry. Chem Res Toxicol 11: 651-658.
31          http://dx.doi.org/10.1021/tx970215v.
32   Blancato. IN: Evans. MV: Power. FW: Caldwell. 1C. (2007). Development and use of PBPK modeling
33          and the impact of metabolism on variability in dose metrics for the risk assessment of
34          methyl tertiary butyl ether (MTBE). J Environ Prot Sci 1: 29-51.
35   Blanck. 0: Fowles. T: Schorsch. F: Fallen. C: Espinasse-Lormeau. H: Schulte-Koerne. E: Totis. M:
36          Banton. M. (2010). Tertiary butyl alcohol in drinking water induces phase I and II liver
37          enzymes with consequent effects on thyroid hormone homeostasis in the B6C3F1 female
38          mouse. J Appl Toxicol 30: 125-132. http://dx.doi.org/10.1002/iat.1478.
39   Borghoff, S: Murphy, I: Medinsky, M. (1996). Development of physiologically based pharmacokinetic
40          model for methyl tertiary-butyl ether and tertiary-butanol in male Fisher-344 rats. Fundam
41          Appl Toxicol 30: 264-275. http://dx.doi.org/10.1006/faatl996.0064.
42   Borghoff. S: Parkinson. H:  Leavens. T. (2010). Physiologically based pharmacokinetic rat model for
43          methyl tertiary-butyl ether; comparison of selected dose metrics following various MTBE
44          exposure scenarios used for toxicity and carcinogenicity evaluation. Toxicology 275: 79-91.
45          http://dx.doi.0rg/10.1016/i.tox.2010.06.003.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-l             DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol


 1   Borghoff. ST: Prescott. TS: Tanszen. DB: Wong. BA: Everitt. TI. (2001). alpha2u-Globulin nephropathy,
 2          renal cell proliferation, and dosimetry of inhaled tert-butyl alcohol in male and female F-
 3          344rats. ToxicolSci 61: 176-186.
 4   Gal/EPA (California Environmental Protection Agency). (1999). Expedited evaluation of risk
 5          assessment for tertiary butyl alcohol in drinking water. Available online at
 6          http://www.oehha.ca.gov/water/pals/tba.html (accessed
 7   CDC (Centers for Disease Control and Prevention). (2004). The health consequences of smoking: A
 8          report of the Surgeon General. Washington, DC: U.S. Department of Health and Human
 9          Services, http://www.cdc.gov/tobacco/data statistics/sgr/2004/index.htm.
10   Cederbaum. AI: Cohen. G. (1980). Oxidative demethylation of t-butyl alcohol by rat liver
11          microsomes. Biochem Biophys Res Commun 97: 730-736.
12   Cederbaum. AI: Qureshi. A: Cohen. G. (1983). Production of formaldehyde and acetone by hydroxyl-
13          radical generating systems during the metabolism of tertiary butyl alcohol. Biochem
14          Pharmacol 32: 3517-3524. http://dx.doi.org/10.1016/0006-2952(83)90297-6.
15   Chen, M. (2005). Amended final report of the safety assessment of t-butyl alcohol as used in
16          cosmetics [Review]. Int J Toxicol 24 Suppl 2: 1-20.
17          http://dx.doi.org/10.1080/10915810590953833.
18   Cirvello. ID: Radovsky. A: Heath. IE: Farnell. PR: III. LC. (1995). Toxicity and carcinogenicity of t-
19          butyl alcohol in rats and mice following chronic exposure in drinking water. Toxicol Ind
20          Health 11: 151-165.
21   Craig. EA: Yan. Z: Zhao. 01. (2014). The relationship between chemical-induced kidney weight
22          increases and kidney histopathology in rats. J Appl Toxicol 35: 729-736.
23          http://dx.doi.org/10.1002/iat3036.
24   Daniel. MA: Evans. MA. (1982). Quantitative comparison of maternal  ethanol and maternal tertiary
25          butanol diet on postnatal development. J Pharmacol Exp Ther 222: 294-300.
26   Faulkner. TP: Wiechart. ID: Hartman. DM:  Hussain. AS. (1989). The effects of prenatal tertiary
27          butanol administration in CBA/J and C57BL/6J mice. Life Sci 45: 1989-1995.
28   FDA (U.S. Food and Drug Administration). (201 la). Indirect food additives: Adjuvants, production
29          aids, and sanitizers. Surface lubricants used in the manufacture of metallic articles. 21 CFR
30          178.3910.
31          http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch. cfm?fr=l 78.3910.
32   FDA (U.S. Food and Drug Administration). (201 Ib). Indirect food additives: Paper and paperboard
33          components. Defoaming agents used in coatings. 21 CFR 176.200.
34          http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=176.200.
35   Gaoua. W. (2004a). Ethyl tertiary butyl ether (ETBE): prenatal developmental toxicity study by the
36          oral route (gavage) in rats. (CIT Study No. 24860 RSR). unpublished study for Totalfinaelf
37          on behalf of the ETBE Producers' Consortium.
38   Gaoua. W. (2004b). Ethyl tertiary butyl ether (ETBE): Two-generation study (reproduction and
39          fertility effects) by the oral route (gavage) in rats. (CIT Study  No. 24859 RSR). unpublished
40          study for Totalfinaelf on behalf of the ETBE Producers' Consortium.
41   Gehlhaus. MW. Ill: Gift. TS: Hogan. KA: Kopylev. L: Schlosser. PM: Kadry. A. -R. f20111 Approaches to
42          cancer assessment in EPA's Integrated Risk Information System [Review]. Toxicol Appl
43          Pharmacol 254: 170-180. http://dx.doi.Org/10.1016/j.taap.2010.10.019.
44   Guyatt. GH: Oxman. AD: Kunz. R: Vist. GE: Falck-Ytter. Y: Schiinemann. HI. (2008a). What is "quality
45          of evidence" and whyis it important to clinicians? [Review]. BMJ 336: 995-998.
46          http://dx.doi.org/10.1136/bmj.39490.551019.BE.
47   Guyatt. GH: Oxman. AD: Vist. GE: Kunz. R: Falck-Ytter. Y: Alonso-Coello.  P: Schiinemann. HI. (2008b).
48          GRADE: An emerging consensus on rating quality of evidence and strength of
49          recommendations. BMJ 336: 924-926. http://dx.doi.org/10.1136/bmj.39489.470347.AD.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-2             DRAFT—DO NOT CITE OR  QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol


 1   Hard, GC. (1986). Experimental models for the sequential analysis of chemically-induced renal
 2          carcinogenesis. Toxicol Pathol 14: 112-122.
 3   Hard, GC. (2008). Some aids to histological recognition of hyaline droplet nephropathy in ninety-
 4          day toxicity studies. Toxicol Pathol 36: 1014-1017.
 5          http://dx.doi.org/10.1177/0192623308327413.
 6   Hard. GC: Banton. MI: Bretzlaff. RS: Dekant. W: Fowles. 1. R: Mallett. AK: Mcgregor. DB: Roberts. KM:
 7          Sielken. RL: Valdez-Flores. C: Cohen. SM. (2013). Consideration of ratchronic progressive
 8          nephropathy in regulatory evaluations for carcinogenicity. Toxicol Sci 132: 268-275.
 9          http://dx.doi.org/10.1093/toxsci/kfs305.
10   Hard. GC: Bruner. RH: Cohen. SM: Fletcher. TM: Regan. KS. (2011). Renal histopathology in toxicity
11          and carcinogenicity studies with tert-butyl alcohol administered in drinking water to F344
12          rats: A pathology working group review and re-evaluation.  Regul Toxicol Pharmacol 59:
13          430-436. http://dx.doi.0rg/10.1016/j.yrtph.2011.01.007.
14   Hard, GC: Johnson, KT: Cohen, SM. (2009). A comparison of ratchronic progressive nephropathy
15          with human renal disease-implications for human risk assessment [Review]. Crit Rev
16          Toxicol 39: 332-346. http://dx.doi.org/10.1080/10408440802368642.
17   Hard. GC: Khan. KN. (2004). A contemporary overview of chronic progressive nephropathy in the
18          laboratory rat, and its significance for human risk assessment [Review]. Toxicol Pathol 32:
19          171-180. http://dx.doi.org/10.1080/01926230490422574.
20   Hard, GC: Seely, 1C. (2005). Recommendations for the interpretation of renal tubule proliferative
21          lesions occurring in rat kidneys with advanced chronic progressive nephropathy (CPN).
22          Toxicol Pathol 33: 641-649. http://dx.doi.org/10.1080/01926230500299716.
23   Hard, GC: Seely, 1C. (2006). Histological investigation of diagnostically challenging tubule profiles in
24          advanced chronic progressive nephropathy (CPN) in the fischer 344 RaT. Toxicol Pathol 34:
25          941-948. http://dx.doi.org/10.1080/01926230601083381.
26   Hard, GC: Wolf, DC. (1999). Re-evaluation of the chloroform 2-year drinking water bioassy in
27          Osborne-Mendel rats indicates that sustained renal tubule injury is associated with renal
28          tumor development [Abstract]. Toxicologist 48: 30.
29   HEW (U.S. Department of Health, Education and Welfare). (1964). Smoking and health: Report of
30          the advisory committee to the surgeon general of the public health service. Washington, DC:
31          U.S. Department of Health, Education, and Welfare.
32          http://profiles.nlm.nih.gov/ps/retrieve/ResourceMetadata/NNBBMO.
33   Hill. AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58:295-
34          300.
35   HSDB (Hazardous Substances Data Bank). (2007). t-Butyl alcohol [Database]. Bethesda, MD:
36          National Library of Medicine.  Retrieved from http://toxnetnlm.nih.gov
37   IARC (International Agency for Research on Cancer). (1999). Methyl tert-butyl ether (group 3) (pp.
38          339-383). Lyon, France.
39   IARC (International Agency for Research on Cancer). (2006). IARC monographs on the evaluation of
40          carcinogenic risks to humans: Preamble. Lyon, France: World Health Organization.
41          http://monographs.iarc.fr/ENG/Preamble/.
42   IOM (Institute of Medicine). (2008). Improving the presumptive disability decision-making process
43          for veterans. In JM Samet; CC Bodurow (Eds.). Washington,  DC: National Academies Press.
44          http://dx.doi.org/10.17226/11908.
45   IPCS (International Programme on Chemical Safety). (1987a). Butanols: Four isomers: 1-butanol, 2-
46          butanol, tert-butanol, isobutanol  [WHO EHC]. In Environmental Health Criteria. Geneva,
47          Switzerland: World Health Organization.
48          http://www.inchem.org/documents/ehc/ehc/ehc65.htm.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-3            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol


 1   IPCS (International Programme on Chemical Safety). (1987b). Tert-Butanol. In Health and Safety
 2          Guide. Geneva, Switzerland: World Health Organization.
 3          http://www.inchem.org/documents/hsg/hsg/hsg007.htm.
 4   Kaneko, T: Wang, PY: Sato, A. (2000). Partition coefficients for gasoline additives and their
 5          metabolites. J Occup Health 42: 86-87. http://dx.doi.org/10.1539/joh.42.86.
 6   Leavens. T: Borghoff. S. (2009). Physiologically based pharmacokinetic model of methyl tertiary
 7          butyl ether and tertiary butyl alcohol dosimetry in male rats based on binding to alpha2u-
 8          globulin. Toxicol Sci 109:321-335. http://dx.doi.org/10.1093/toxsci/kfp049.
 9   Lee. JS: Ward. WO: Liu. J: Ren. H: Vallanat. B: Delker. D: Gorton. JC. (2011). Hepatic xenobiotic
10          metabolizing enzyme and transporter gene expression through the life stages of the mouse.
11          PLoS ONE 6: e24381. http://dx.doi.org/10.1371/iournal.pone.0024381.
12   Lee. IS: Ward. WO: Wolf. DC: Allen. TW: Mills. C: Devito. MI: Gorton. 1C. (2008). Coordinated changes
13          in xenobiotic metabolizing enzyme gene expression in aging male rats. Toxicol Sci 106: 263-
14          283. http://dx.doi.org/10.1093/toxsci/kfnl44.
15   Li, Q: Kobayashi, M: Inagaki, H: Hirata, Y: Hirata, K: Shimizu, T: Wang, RS: Suda, M: Kawamoto, T:
16          Nakajima. T: Kawada. T. (2011). Effects of subchronic inhalation exposure to ethyl tertiary
17          butyl ether on splenocytes in mice. Int J Immunopathol Pharmacol 24: 837-847.
18   Lindamood. C: Farnell. D: Giles. H: Preiean. T: Collins. T: Takahashi. K: Maronpot. R. T19921
19          Subchronic toxicity studies of t-butyl alcohol in rats and mice. Fundam Appl Toxicol 19: 91-
20          100.
21   Lyondell Chemical Co. (Lyondell Chemical Company). (2004). Reproductive and developmental
22          toxicity screening test in rats by oral gavage. (Document Control Number: 89-040000106).
23   Maronpot. RR: Yoshizawa. K: Nyska. A: Harada. T: Flake. G: Mueller. G: Singh. B: Ward. TM. (2010).
24          Hepatic enzyme induction: histopathology [Review]. Toxicol Pathol 38: 776-795.
25          http://dx.doi.org/10.1177/0192623310373778.
26   McGregor, D. (2010). Tertiary-butanol: A toxicological review [Review]. Crit Rev Toxicol 40: 697-
27          727. http://dx.doi.org/10.3109/10408444.2010.494249.
28   Melnick, R: Burns, K: Ward, I: Huff, I. (2012). Chemically Exacerbated Chronic Progressive
29          Nephropathy Not Associated with Renal Tubule Tumor Induction in Rats: An  Evaluation
30          Based on 60 Carcinogenicity Studies by the National Toxicology Program. Toxicol Sci  128:
31          346-356. http://dx.doi.org/10.1093/toxsci/kfsl56.
32   Melnick. RL: Ward. TM: Huff. I. (2013). War on carcinogens: Industry disputes human relevance of
33          chemicals causing cancer in laboratory animals based on unproven hypotheses, using
34          kidney tumors as an example [Editorial]. Int J Occup Environ Health 19: 255-260.
35          http://dx.doi.org/10.1179/1077352513Z.00000000090.
36   Nelson. BK: Brightwell. WS: Khan. A: Burg. TR: Goad. PT. (1989). Lack of selective developmental
37          toxicity of three butanol isomers administered by inhalation to rats. Fundam Appl Toxicol
38          12:469-479.
39   Nelson. BK: Brightwell. WS: Khan. A: Shaw. PB: Krieg. EF. Jr: Massari. VJ. (1991). Behavioral
40          teratology investigation of tertiary-butanol administered by inhalation to rats.
41          Pharmacopsychoecologia 4: 1-7.
42   Nihlen. A: Lof. A: Johanson. G. (1995). Liquid/air partition coefficients of methyl and ethyl t-butyl
43          ethers, t-amyl methyl ether, and t-butyl alcohol. J Expo Anal Environ Epidemiol 5: 573-582.
44   Nihlen. A: Lof. A: Tohanson. G. (1998a). Controlled ethyl tert-butyl ether (ETBE) exposure of male
45          volunteers: I Toxicokinetics. Toxicol Sci 46: 1-10.
46          http://dx.doi.org/10.1006/toxs.1998.2516.
47   Nihlen, A: Lof, A: Tohanson, G. (1998b). Experimental exposure to methyl tertiary-butyl ether: I
48          Toxicokinetics in humans. Toxicol Appl Pharmacol 148: 274-280.


               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-4             DRAFT—DO NOT CITE OR  QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol


 1   NIOSH (National Institute for Occupational Safety and Health). (2005). NIOSH pocket guide to
 2          chemical hazards: tert-Butyl alcohol. Atlanta, GA: Center for Disease Control.
 3          http://www.cdc.gov/niosh/npg/npgd0078.html.
 4   NIOSH (National Institute for Occupational Safety and Health). (2007). NIOSH pocket guide to
 5          chemical hazards. (DHHS-2005-149. CBRNIAC-CB-112149). Cincinnati, OH.
 6          http://www.cdc.gov/niosh/docs/2005-149/.
 7   NRC (National Research Council). (1983). Risk Assessment in the Federal Government: Managing
 8          the Process. Washington, DC: National Academy Press.
 9          http://dx.doi.org/10.1080/00139157.1983.9931232.
10   NRC (National Research Council). (2009). Science and decisions: Advancing risk assessment
11          Washington, DC: National Academy Press, http://www.nap.edu/catalog/12209.html.
12   NRC (National Research Council). (2011). Review of the Environmental Protection Agency's draft
13          IRIS assessment of formaldehyde (pp. 194). Washington, DC: National Academies Press.
14          http://www.nap.edu/catalog/13142.html.
15   NSF International. (2003). t-Butanol: Oral Risk Assessment Document (CAS 75-65-0). Ann Arbor,
16          MI.
17   NTP (National Toxicology Program). (1995). Toxicology and carcinogenesis studies of t-butyl
18          alcohol (CAS no 75-65-0) in F344/N rats and B6C3F1 mice (Drinking water studies) (pp. 1-
19          305). (NTPTR436). Research Triangle Park, NC.
20   NTP (National Toxicology Program). (1997). NTP technical report on toxicity studies of t-butyl
21          alcohol (CAS no 75-65-0) administered by inhalation to F344/N rats and B6C3F1 mice (pp.
22          1-56, A51-D59). Research Triangle Park, NC.
23          http://ntp.niehs.nih.gov/ntp/htdocs/ST  rpts/tox053.pdf.
24   NTP (National Toxicology Program). (2015). Handbook for conducting a literature-based health
25          assessment using OHAT approach for systematic review and evidence integration. U.S. Dept
26          of Health and Human Services, National Toxicology Program.
27          http://ntp.niehs.nih.gov/pubhealth/hat/noms/index-2.html.
28   OSHA (Occupational Safety & Health Administration). (1992). Occupational safety and health
29          guideline for tert-butyl alcohol. Cincinnati, OH: National Institute for Occupational Safety
30          and Health.
31          http://www.osha.gov/SLTC/healthguidelines/tertbutylalcohol/recognition.html.
32   OSHA (Occupational Safety & Health Administration). (2006). Table Z-l: Limits for air
33          contaminants. Occupational safety and health standards, subpart Z, toxic and hazardous
34          substances. (OSHA standard 1910.1000, 29 CFR). Washington, DC: U.S. Department of
35          Labor.
36          http://www.osha.gov/pls/oshaweb/owadisp.show document?p table=STANDARDS&p id=
37          9992.
38   Poet. TS: Valentine. JL: Borghoff. SJ. (1997). Pharmacokinetics of tertiary butyl alcohol in male and
39          female Fischer 344 rats. Toxicol Lett 92: 179-186.
40   Oatanani. M: Zhang. I: Moore. DP. (2005). Role of the constitutive androstane receptor in
41          xenobiotic-induced thyroid hormone metabolism. Endocrinology 146: 995-1002.
42          http://dx.doi.org/10.1210/en.2004-1350.
43   Rogues. BB: Leghait. I: Lacroix. MZ: Lasserre. F: Pineau. T: Viguie. C: Martin. PG. (2013). The nuclear
44          receptors pregnane X receptor and constitutive androstane receptor contribute to the
45          impact of fipronil on hepatic gene expression linked to thyroid hormone metabolism.
46          Biochem Pharmacol 86: 997-1039. http://dx.doi.Org/10.1016/j.bcp.2013.08.012.
47   Rothman, KT: Greenland, S. (1998). Modern Epidemiology (2nd ed.). Philadelphia, PA: Lippincott,
48          Williams, & Wilkins.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-5            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol


 1   Saito. A: Sasaki. T: Kasai. T: Katagiri. T: Nishizawa. T: Noguchi. T: Also. S: Nagano. K: Fukushima. S.
 2          (2013). Hepatotumorigenicity of ethyl tertiary-butyl ether with 2-year inhalation exposure
 3          in F344 rats. Arch Toxicol 87: 905-914. http://dx.doi.org/10.1007/s00204-012-0997-x.
 4   Salazar, KD: Brinkerhoff, CT: Lee, IS: Chiu, WA. (2015). Development and application of a ratPBPK
 5          model to elucidate kidney and liver effects induced by ETBE and tert-butanol. Toxicol Appl
 6          Pharmacol 288: 439-452. http://www.ncbi.nlm.nih.gov/pubmed/26341290.
 7   Scorecard. (2014). t-butanol. Available online at http://scorecard.goodguide.com/chemical-
 8          profiles/summary.tcl?edf substance id=+75-65-0 (accessed
 9   Seely. JC: Haseman. JK: Nyska. A: Wolf. DC: Everitt. JI: Hailey. JR. (2002). The effect of chronic
10          progressive nephropathy on the incidence of renal tubule cell neoplasms in control male
11          F344 rats. Toxicol Pathol 30: 681-686.
12   Short. BG: Burnett. VL: Swenberg. TA.  (1986). Histopathology and cell proliferation induced by 2,2,4-
13          trimethylpentane in the male rat kidney. Toxicol Pathol 14: 194-203.
14   Short, BG: Burnett, VL: Swenberg, TA.  (1989). Elevated proliferation of proximal tubule cells and
15          localization of accumulated "alpha"2u-globulin in F344 rats during chronic exposure to
16          unleaded gasoline or 2,2,4-trimethylpentane. Toxicol Appl Pharmacol 101: 414-431.
17   Suzuki. M: Yamazaki. K: Kano. H: Aiso. S: Nagano. K: Fukushima. S. (2012). No carcinogenicity of
18          ethyl tertiary-butyl ether by 2-year oral administration in rats. J Toxicol Sci 37: 1239-1246.
19   Swenberg, TA: Lehman-McKeeman, LD. (1999). alpha 2-Urinary globulin-associated nephropathy as
20          a mechanism of renal tubule cell carcinogenesis in male rats. In CC Capen; E Dybing; JM
21          Rice; JD Wilbourn (Eds.), IARC Scientific Publications (pp. 95-118). Lyon, France:
22          International Agency for Research on Cancer.
23          http://apps.who.int/bookorde rs/anglais/detartl.jsp?sesslan=l&codlan=l&codcol=73&cod
24          cch=147.
25   Takahashi, K: Lindamood, C: Maronpot, R. (1993). Retrospective study of possible alpha-2 mu-
26          globulin nephropathy and associated cell proliferation in male Fischer 344 rats dosed with
27          t-butyl alcohol. Environ Health Perspect 101: 281-285.
28   U.S. EPA (U.S. Environmental Protection Agency). (1986a). Guidelines for mutagenicity risk
29          assessment (EPA/630/R-98/003). Washington, DC: U.S. Environmental Protection Agency,
30          Risk Assessment Forum, http://www.epa.gov/iris/backgrd.html.
31   U.S. EPA (U.S. Environmental Protection Agency). (1986b). Guidelines for the health risk
32          assessment of chemical mixtures. (EPA/630/R-98/002). Washington, DC: U.S.
33          Environmental Protection Agency, Risk Assessment Forum.
34          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=22567.
35   U.S. EPA (U.S. Environmental Protection Agency). (1988). Recommendations for and documentation
36          of biological values for use in  risk assessment (EPA/600/6-87/008). Cincinnati, OH: U.S.
37          Environmental Protection Agency, Office of Research and Development, Office of Health and
38          Environmental Assessment http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855.
39   U.S. EPA (U.S. Environmental Protection Agency). (1991a). Alpha-2u-globulin: Association with
40          chemically induced renal toxicity and neoplasia in the male rat  (EPA/625/3-91/019F).
41          Washington, DC: U.S. Environmental Protection Agency, National Center for Environmental
42          Assessment.
43          https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchOuery=PB92143668.
44   U.S. EPA (U.S. Environmental Protection Agency). (1991b). Guidelines for developmental toxicity
45          risk assessment. (EPA/600/FR-91/001). Washington, DC: U.S. Environmental Protection
46          Agency, Risk Assessment Forum.
47          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2 3162.
48   U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods  for derivation of inhalation
49          reference concentrations (RfCs) and application of inhalation dosimetry [EPA Report].

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-6            DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review of tert-ButyI Alcohol


 1          (EPA/600/8-90/066F). Washington, DC: U.S. Environmental Protection Agency, Office of
 2          Research and Development, Office of Health and Environmental Assessment, Environmental
 3          Criteria and Assessment Office.
 4          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993.
 5   U.S. EPA (U.S. Environmental Protection Agency). (1996). Guidelines for reproductive toxicity risk
 6          assessment (EPA/630/R-96/009). Washington, DC: U.S. Environmental Protection Agency,
 7          Risk Assessment Forum, http://www.epa.gov/raf/publications/guidelines-reproductive-
 8          tox-risk-assessmenthtm.
 9   U.S. EPA (U.S. Environmental Protection Agency). (1997). Drinking water advisory: consumer
10          acceptability advice and health effects analysis on methyl tertiary-butyl ether (MTBE) [EPA
11          Report].
12   U.S. EPA (U.S. Environmental Protection Agency). (1998a). Assessment of thyroid follicular cell
13          tumors [EPA Report]. (EPA/630/R-97/002). Washington, DC.
14   U.S. EPA (U.S. Environmental Protection Agency). (1998b). Guidelines for neurotoxicity risk
15          assessment [EPA Report]. (EPA/630/R-95/001F). Washington, DC: U.S. Environmental
16          Protection Agency, Risk Assessment Forum, http://www.epa.gov/risk/guidelines-
17          neurotoxicity-risk-assessment
18   U.S. EPA (U.S. Environmental Protection Agency). (2000a). Science policy council handbook: Risk
19          characterization. (EPA/100/B-00/002). Washington, D.C.: U.S. Environmental Protection
20          Agency, Science Policy Council.
21          http://nepis.epa. gov/Exe/ZyNET.exe/40000006.TXT?ZyActionD=ZyDocument&Client=EPA
22          &Index=2000+Thru+2005&Docs=&Ouery=&Time=&EndTime=&SearchMethod=l&TocRest
23          rict=n&Toc=&TocEntry=&OField=&OFieldYear=&OFieldMonth=&OFieldDay=&IntOFieldOp
24          =0&ExtOFieldOp=0&XmlQuery=&File=D%3A%5Czyfiles%5CIndex%20Data%5COOthru05
25          %5CTxt%5C00000002%5C40000006.txt&User=ANONYMOUS&Password=anonymous&Sor
26          tMethod=h%7C-
27          &MaximumDocuments=l&FuzzyDegree=0&ImageOuality=r75g8/r75g8/xl50yl50gl6/i42
28          5&Display=p%7Cf&DefSeekPage=x&SearchBack=ZyActionL&Back=ZyActionS&BackDesc=R
29          esults%20page&MaximumPages=l&ZyEntry=l&SeekPage=x&ZyPURL.
30   U.S. EPA (U.S. Environmental Protection Agency). (2000b). Supplementary guidance for conducting
31          health risk assessment of chemical mixtures. (EPA/630/R-00/002). Washington, DC:  U.S.
32          Environmental Protection Agency, Risk Assessment Forum.
33          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=20533.
34   U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and
35          reference concentration processes. (EPA/630/P-02/002F). Washington, DC: U.S.
36          Environmental Protection Agency, Risk Assessment Forum.
37          http://www.epa.gov/osa/review-reference-dose-and-reference-concentration-processes.
38   U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines  for carcinogen risk
39          assessment [EPA Report]. (EPA/630/P-03/001F). Washington, DC: U.S. Environmental
40          Protection Agency, Risk Assessment Forum, http://www2.epa.gov/osa/guidelines-
41          carcinogen-risk-assessment
42   U.S. EPA (U.S. Environmental Protection Agency). (2005b). Supplemental guidance for assessing
43          susceptibility from early-life exposure to carcinogens. (EPA/630/R-03/003F). Washington,
44          DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
45          http://www.epa.gov/raf/publications/pdfs/childrens supplement final.pdf.
46   U.S. EPA (U.S. Environmental Protection Agency). (2006a). Approaches for the application of
47          physiologically based pharmacokinetic (PBPK) models and supporting data in risk
48          assessment (Final Report) [EPA Report]. (EPA/600/R-05/043F). Washington, DC: U.S.
49          Environmental Protection Agency, Office of Research and Development, National Center for

               This document is a draft for review purposes only and does not constitute Agency policy.
                                               R-7             DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review of tert-ButyI Alcohol


 1          Environmental assessment.
 2          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 57668.
 3   U.S. EPA (U.S. Environmental Protection Agency). (2006b). A framework for assessing health risk of
 4          environmental exposures to children. (EPA/600/R-05/093F). Washington, DC: U.S.
 5          Environmental Protection Agency, Office of Research and Development, National Center for
 6          Environmental Assessment
 7          ht±p://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 58363.
 8   U.S. EPA (U.S. Environmental Protection Agency). (2009). EPA's Integrated Risk Information
 9          System: Assessment development process [EPA Report]. Washington, DC.
10          http://epa.gov/iris/process.htm.
11   U.S. EPA (U.S. Environmental Protection Agency). (2010). Integrated science assessment for carbon
12          monoxide [EPA Report]. (EPA/600/R-09/019F). Research Triangle Park, NC: U.S.
13          Environmental Protection Agency, Office of Research and Development, National Center for
14          Environmental Assessment
15          http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=218686.
16   U.S. EPA (U.S. Environmental Protection Agency). (2011). Recommended use of body weight 3/4 as
17          the default method in derivation of the oral reference dose. (EPA/100/R11/0001).
18          Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum, Office of
19          the Science Advisor, http://www.epa.gov/raf/publications/interspecies-extrapolation.htm.
20   U.S. EPA (U.S. Environmental Protection Agency). (2012a). Advances in inhalation gas dosimetry for
21          derivation of a reference concentration (RFC) and use in risk assessment. (EPA/600/R-
22          12/044). Washington, DC. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=244650.
23   U.S. EPA (U.S. Environmental Protection Agency). (2012b). Benchmark dose technical guidance.
24          (EPA/100/R-12/001). Washington, DC: U.S. Environmental Protection Agency, Risk
25          Assessment Forum, http://www.epa.gov/raf/publications/benchmarkdose.htm.
26   U.S. EPA (U.S. Environmental Protection Agency). (2012c). Releases: Facility Report. Toxics Release
27          Inventory. Available online at http://iaspub.epa.gov/triexplorer/tri release.chemical
28          (accessed
29   U.S. EPA. (2012d). Toxicological review of tetrahydrofuran. In support of summary information on
30          the integrated risk information system (IRIS). (EPA/635/R-11/006F). Washington, DC: U.S.
31          Environmental Protection Agency
32   U.S. EPA (U.S. Environmental Protection Agency). (2014). Chemical Data Reporting 2012, reported
33          in the Chemical Data Access Tool. Available online at
34          http://www.epa.gov/oppt/cdr/index.html (accessed
35   Williams. TM: Borghoff. ST. (2001). Characterization of tert-butyl alcohol binding to "alpha"2u-
36          globulin in F-344 rats. Toxicol Sci 62: 228-235. http://dx.doi.Org/10.1093/toxsci/62.2.228.
37   Yuan. Y: Wang. HF: Sun. HF: Du. HF: Xu. LH: Liu. YF: Ding. XF: Fu. DP: Liu. KX. (2007). Adduction of
38          DNA with MTBE and TEA in mice studied by accelerator mass Spectrometry. Environ
39          Toxicol 22: 630-635. http://dx.doi.org/10.1002/tox.20295.
40
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                R-8            DRAFT—DO NOT CITE OR QUOTE

-------