United States
kS^laMIjk Environmental Protection
^J^iniiil m11 Agency
EPA/690/R-ll/008F
Final
4-1-2011
Provisional Peer-Reviewed Toxicity Values for
Benzenethiol
(CASRN 108-98-5)
Superfund Health Risk Technical Support Center
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER
Harlal Choudhury, DVM, Ph.D., DABT
National Center for Environmental Assessment, Cincinnati, OH
DRAFT DOCUMENT PREPARED BY
ICF International
9300 Lee Highway
Fairfax, VA 22031
PRIMARY INTERNAL REVIEWERS
Audrey Galizia, Dr. PH.
National Center for Environmental Assessment, Washington, DC
Dan D. Petersen, Ph.D., DABT
National Center for Environmental Assessment, Cincinnati, OH
This document was externally peer reviewed under contract to
Eastern Research Group, Inc.
110 Hartwell Avenue
Lexington, MA 02421-3136
Questions regarding the contents of this document may be directed to the U.S. EPA Office of
Research and Development's National Center for Environmental Assessment, Superfund Health
Risk Technical Support Center (513-569-7300)
-------
TABLE OF CONTENTS
COMMONLY USED ABBREVIATIONS i
BACKGROUND 1
HISTORY 1
DISCLAIMERS 1
QUESTIONS REGARDING PPRTVS 2
INTRODUCTION 2
REVIEW OF POTENTIALLY RELEVANT DATA (CANCER AND NONCANCER) 4
HUMAN STUDIES 7
Oral and Inhalation Exposure 7
ANIMAL STUDIES 7
Oral Exposure 7
Subchronic Studies 7
Chronic Studies 7
Developmental and Reproduction Studies 7
Developmental Toxicity Study in Rats 7
Developmental Toxicity Study in Rabbits 9
Reproduction Study 10
Inhalation Exposure 13
Subchronic Studies 13
Chronic Studies 13
Developmental and Reproduction Studies 13
Other Data (Short-Term Tests, Other Examination) 13
Acute and Subacute Inhalation Studies 13
Metabolism, Mode-of-Action and Structure-Activity Relationship Studies 15
DERIVATION 01 PROVISIONAL VALUES 19
DERIVATION 01 ORAL REFERENCE DOSE 19
Derivation of Subchronic p-RfD 19
Derivation of Chronic p-RfD 30
DERIVATION OF INHALATION REFERENCE CONCENTRATIONS 32
Derivation of Subchronic and Chronic p-RfC 32
DERIVATION OF PROVISIONAL CANCER VALUES 32
Cancer Weight-of-Evidence Descriptor 32
Derivation of p-OSF 33
Derivation of p-IUR 33
APPENDIX A. PROVISIONAL SCREENING VALUES 34
APPENDIX B. DATA TABLES 35
APPENDIX C. BMD MODELING OUTPUTS FOR BENZENETHIOL 45
APPENDIX D. REFERENCES 109
l
Benzenethiol
-------
COMMONLY USED ABBREVIATIONS
BMC
benchmark concentration
BMCL
benchmark concentration lower bound 95% confidence interval
BMD
benchmark dose
BMDL
benchmark dose lower bound 95% confidence interval
HEC
human equivalent concentration
HED
human equivalent dose
IUR
inhalation unit risk
LOAEL
lowest-observed-adverse-effect level
LOAELadj
LOAEL adjusted to continuous exposure duration
LOAELhec
LOAEL adjusted for dosimetric differences across species to a human
NOAEL
no-ob served-adverse-effect level
NOAELadj
NOAEL adjusted to continuous exposure duration
NOAELhec
NOAEL adjusted for dosimetric differences across species to a human
NOEL
no-ob served-effect level
OSF
oral slope factor
p-IUR
provisional inhalation unit risk
p-OSF
provisional oral slope factor
POD
point of departure
p-RfC
provisional reference concentration (inhalation)
p-RfD
provisional reference dose (oral)
RfC
reference concentration (inhalation)
RfD
reference dose (oral)
UF
uncertainty factor
UFa
animal-to-human uncertainty factor
UFC
composite uncertainty factor
UFd
incomplete-to-complete database uncertainty factor
UFh
interhuman uncertainty factor
UFl
LOAEL-to-NOAEL uncertainty factor
UFS
subchronic-to-chronic uncertainty factor
WOE
weight of evidence
1
Benzenethiol
-------
FINAL
4-1-2011
PROVISIONAL PEER-REVIEWED TOXICITY VALUES FOR
BENZENE THIOL (CASRN 108-98-5)
BACKGROUND
HISTORY
On December 5, 2003, the U.S. Environmental Protection Agency's (EPA) Office of
Superfund Remediation and Technology Innovation (OSRTI) revised its hierarchy of human
health toxicity values for Superfund risk assessments, establishing the following three tiers as the
new hierarchy:
1) EPA's Integrated Risk Information System (IRIS).
2) Provisional Peer-Reviewed Toxicity Values (PPRTVs) used in EPA's Superfund
Program.
3) Other (peer-reviewed) toxicity values, including
~ Minimal Risk Levels produced by the Agency for Toxic Substances and Disease
Registry (ATSDR),
~ California Environmental Protection Agency (CalEPA) values, and
~ EPA Health Effects Assessment Summary Table (HEAST) values.
A PPRTV is defined as a toxicity value derived for use in the Superfund Program when
such a value is not available in IRIS (U.S. EPA, 2010a). PPRTVs are developed according to a
Standard Operating Procedure (SOP) and are derived after a review of the relevant scientific
literature using the same methods, sources of data, and Agency guidance for value derivation
generally used by the EPA IRIS Program. All provisional toxicity values receive internal review
by a panel of six EPA scientists and external peer review by three independently selected
scientific experts. PPRTVs differ from IRIS values in that PPRTVs do not receive the
multiprogram consensus review provided for IRIS values. This is because IRIS values are
generally intended to be used in all EPA programs, while PPRTVs are developed specifically for
the Superfund Program.
Because new information becomes available and scientific methods improve over time,
PPRTVs are reviewed on a 5-year basis and updated into the active database. Once an IRIS
value for a specific chemical becomes available for Agency review, the analogous PPRTV for
that same chemical is retired. It should also be noted that some PPRTV documents conclude that
a PPRTV cannot be derived based on inadequate data.
DISCLAIMERS
Users of this document should first check to see if any IRIS values exist for the chemical
of concern before proceeding to use a PPRTV. If no IRIS value is available, staff in the regional
Superfund and Resource Conservation and Recovery Act (RCRA) program offices are advised to
carefully review the information provided in this document to ensure that the PPRTVs used are
appropriate for the types of exposures and circumstances at the Superfund site or RCRA facility
in question. PPRTVs are periodically updated; therefore, users should ensure that the values
contained in the PPRTV are current at the time of use.
1
Benzenethiol
-------
FINAL
4-1-2011
It is important to remember that a provisional value alone tells very little about the effects
of a chemical or the quality of evidence on which the value is based. Therefore, users are
strongly encouraged to read the entire PPRTV document and understand the strengths and
limitations of the derived provisional values. PPRTVs are developed by the EPA Office of
Research and Development's National Center for Environmental Assessment, Superfund Health
Risk Technical Support Center for OSRTI. Other EPA programs or external parties who may
choose of their own initiative to use these PPRTVs are advised that Superfund resources will not
generally be used to respond to challenges of PPRTVs used in a context outside of the Superfund
Program.
QUESTIONS REGARDING PPRTVS
Questions regarding the contents of the PPRTVs and their appropriate use (e.g., on
chemicals not covered, or whether chemicals have pending IRIS toxicity values) may be directed
to the EPA Office of Research and Development's National Center for Environmental
Assessment, Superfund Health Risk Technical Support Center (513-569-7300), or OSRTI.
Benzenethiol (also called thiophenol or phenylmercaptan) is used as a mosquito larvicide,
as a food additive and, as an intermediate in the manufacture of pesticides, pharmaceuticals, and
amber dyes. It is a colorless liquid with a disagreeable odor described as penetrating, repulsive,
and garlic-like. Benzenethiol is produced commercially by reducing benzenesulfonyl chloride
with zinc dust in sulfuric acid or by reacting hydrogen sulfide with chlorobenzene (U.S. EPA,
2007). The molecular formula for benzenethiol is C6H5SH (see Figure 1). A table of its
chemico-physical properties is provided below (see Table 1).
INTRODUCTION
SH
Figure 1. Benzenethiol Structure
2
Benzenethiol
-------
FINAL
4-1-2011
Table 1. Physical Properties Table for Benzenethiola
Property (unit)
Value
Boiling point (°C)
168.3
Melting point (°C)
-14.9
Density (g/cm3)
1.07
Vapor pressure at 25°C (mm Hg)
1.93
pH (unitless)
Feebly acidic
Solubility in water (mg/L at 25°C)
836
Relative vapor density (air =1)
3.8
Molecular weight (g/mol)
110.18
Flash point (°C)
50
Octanol/water partition coefficient (unitless)
331.13
aValues from http://www.epa.gov/oppt/aegl/pubs/phenyl_mercaptan_interim_nov_2007_vl.pdf and
HSDB (searched online 02-17-2010; reviewed 4-16-2009, last revised 6-23-2005).
No reference dose (RfD), reference concentration (RfC), or cancer assessment for
benzenethiol is included on the IRIS database (U.S. EPA, 2010a) or on the Drinking Water
Standards and Health Advisories List (U.S. EPA, 2006). CalEPA (2008, 2009a,b,c) has not
derived toxicity values for exposure to benzenethiol or prepared a quantitative estimate of
carcinogenic potential. Benzenethiol is not included in the 11th Report on Carcinogens (NTP,
2005). The International Agency for Research on Cancer (IARC, 2009) has not reviewed the
carcinogenic potential of benzenethiol. An interim acute exposure guideline level (AEGL)
report stated that carcinogenicity studies in humans or animals were not available (U.S. EPA,
2007). Benzenethiol was not included in the CARA list (U.S. EPA, 1994).
The HEAST reported a subchronic RfD of 1.0 x 10 4 mg/kg-day (U.S. EPA, 2010b),
derived from a 90-day oral gavage study in rats (American Biogenics Corp., 1989) with a
LOAEL of 0.1 mg/kg-day (based on centrilobular eosinophilic changes in the liver) and an
uncertainty factor (UF) of 1000. A chronic RfD of 1.0 x 10 5 mg/kg-day was estimated from the
subchronic RfD using an additional UFL of 10 for a total UFC of 10,000. An electronic search of
the online HEAST on February 19, 2010 continued to list this value as the RfD. However, a
copy of this study was not available, and the online link to the HEAST Derivation Support
Document was not available at the time of the preparation of this PPRTV document.
The American Conference of Governmental Industrial Hygienists (ACGIH, 2009)
reported a threshold limit value (TLV) of 0.1 ppm, 0.45 mg/m3 time-weighted average (TWA),
and the National Institute of Occupational Safety and Health (NIOSH, 2005) set a Recommended
Exposure Limit (REL) at 0.30 mg/m3. The Occupational Safety and Health Administration
(OSHA, 2009) has not derived a permissible exposure limit (PEL) for benzenethiol. The toxicity
of benzenethiol has not been reviewed by ATSDR (2010) to determine oral or inhalation
Minimal Risk Levels (MRL).
3
Benzenethiol
-------
FINAL
4-1-2011
The World Health Organization (WHO) has not prepared an environmental health criteria
(EHC) document on benzenethiol (WHO, 2010). A meeting of the Joint Food and Agriculture
Organization of the United Nations (FAO)/WHO Expert Committee on Food Additives (JECFA)
in 1999 evaluated certain food additives and contaminants (FAO/WHO, 1999). Benzenethiol
(No. 525) was evaluated using the procedure for safety evaluation of flavoring agents, resulting
in new specifications prepared and conclusions of "no safety concern" based on current intake.
Literature searches were conducted on sources published from 1900 through August 2010
for studies relevant to the derivation of provisional toxicity values for benzenethiol, CAS
No. 108-98-5. The EPA Health and Environmental Research Online (HERO) evergreen
database of scientific literature was used to search the following databases: AGRICOLA;
American Chemical Society; BioOne; Cochrane Library; DOE: Energy Information
Administration; DOE: Information Bridge; DOE: Energy Citations Database; EBSCO:
Academic Search Complete; GeoRef Preview; GPO: Government Printing Office;
Informaworld; IngentaConnect; J-STAGE: Japan Science & Technology; JSTOR: Mathematics
& Statistics; JSTOR: Life Sciences; NSCEP/NEPIS (EPA publications available through the
National Service Center for Environmental Publications [NSCEP] and National Environmental
Publications Internet Site [NEPIS] database); PubMed (MEDLINE and CANCERLIT
databases); SAGE; Science Direct; Scirus; Scitopia; SpringerLink; TOXNET (Toxicology Data
Network: ANEUPL; CCRIS; ChemlDplus; CIS; CRISP; DART; EMIC; EPIDEM; ETICBACK;
FEDRIP; GENE-TOX; HAPAB; HEEP; HMTC; HSDB; IRIS; ITER; LactMed; Multidatabase
Search; NIOSH; NTIS; PESTAB; PPBIB; RISKLINE; TRI; and TSCATS); Virtual Health
Library; Web of Science (searches Current Content database among others); WHO; and
Worldwide Science. The following databases outside of HERO were searched for risk
assessment values: ACGM; AT SDR; CalEPA; EPA IRIS; EPA HEAST; EPA HEEP; EPA OW;
EPA TSCATS/TSCATS2; NIOSH; NTP; OSHA; and RTECS.
REVIEW OF POTENTIALLY RELEVANT DATA
(CANCER AND NONCANCER)
Table 2 provides information for all of the potentially relevant studies. Entries for the
principal studies (PS) are bolded. In this document, "statistically significant" denotes ap-walue
of <0.05.
4
Benzenethiol
-------
FINAL
4-1-2011
Table 2. Summary of Potentially Relevant Data for Benzenethiol (CASRN 108-98-5)
Notes3
Category
Number of Male/Female
Species, Study Type, and
Duration
Dosimetry13
Critical Effects
NOAELb
BMDL/
BMCLb
LOAELbc
Reference
(Comments)
Human
1. Oral (mg/kg-day)b
Subchronic
None
Chronic
None
Developmental
None
Reproductive
None
Carcinogenic
None
2. Inhalation (mg/m3)b
Subchronic
None
Chronic
None
Developmental
None
Reproductive
None
Carcinogenic
None
Animal
1. Oral (mg/kg-day)b
Subchronic
Albino Rat
Daily gavage for 90 days
Study is not available for review
Centrilobular eosinophilic
changes in the liver
Not
available
0.1
American
Biogenics
Corp. (1989)d
Chronic
None
PR
Developmental
5-D Rat
25 females/dose group
Daily gavage in corn oil
from Gestation Days (GDs)
6-15. Cesarean section
performed on GD 20
0, 20, 35, or
50 mg/kg-day
Maternal: increased relative and
absolute liver weights and
decreasd gravid uterine weight
Developmental: decreased fetal
body weights (females)
35
20
Not run
50
35
NTP (1994a)
5
Benzenethiol
-------
FINAL
4-1-2011
Table 2. Summary of Potentially Relevant Data for Benzenethiol (CASRN 108-98-5)
Notes3
Category
Number of Male/Female
Species, Study Type, and
Duration
Dosimetry13
Critical Effects
NOAELb
BMDL/
BMCLb
LOAELbc
Reference
(Comments)
PR
Developmental
New Zealand White Rabbit
26 does in 1-, 10-, and
30-mg/kg-day groups; 15
does in 40-mg/kg-day
group
Daily gavage in corn oil
from GDs 6-19. Cesarean
section performed on
GD 30
0, 10, 30, or
40 mg/kg-day;
50 mg/kg-day
excluded from final
assessment because
excessive maternal
toxicity (mortality
and morbidity)
Maternal: decreased body
weight gain and food
consumption. Body weight loss
for overall study, when
corrected for gravid uterine
weight.
No developmental toxicity at
40 mg/kg-day
30
40
Not run
40
Not
observed
NTP (1994b)
PS
PR
Reproductive
S-D Rat
20 breeding pairs/group
Daily gavage in corn oil
continuously for
two generations
0,9,18, or
35 mg/kg-day
Parental: increased liver and
kidney weights;
hepatocellular hypertrophy;
and renal tubule degeneration
Offspring: decreased F2 pup
body weights
Reproductive: inhibited
spermiation in F1 males
Not
observed
9
Not
observed
2.91
9
18
9
NTP (1996)
Carcinogenic
None
2. Inhalation (mg/m3)b
Subchronic
None
Chronic
None
Developmental
None
Reproductive
None
Carcinogenic
None
aPS = Principal study; PR = Peer-reviewed.
bDosimetry, NOAEL, BMDL/BMCL and LOAEL values are converted to Human Equivalent Dose (HED in mg/kg-day) or Human Equivalent Concentration (HEC
in mg/m3) units. Noncancer oral data are only adjusted for continuous exposure.
°Not reported by the study author but determined from data.
dThis study, cited in HEAST (U.S. EPA, 2010b), derived a subchronic RfD of 1 x 10 4 mg/kg-day using UF of 1000. Chronic RfD estimated at 1 x 10 " mg/kg-day
from subchronic study (UF = 10,000).
6
Benzenethiol
-------
FINAL
4-1-2011
HUMAN STUDIES
Oral and Inhalation Exposure
No studies investigating the effects of subchronic or chronic oral exposure to
benzenethiol in humans have been identified. No quantitative data were located regarding the
toxicity of benzenethiol to humans following chronic or subchronic inhalation exposure. Data
concerning human exposure to benzenethiol are limited to odor threshold data. An online search
of Haz-map (2010) reported that benzenethiol causes tearing of the eyes and is a skin and
respiratory irritant. Volunteers tolerated 8 ppm for 10 seconds (eye irritation), and a single
breath of 35 ppm (nasal irritation).
ANIMAL STUDIES
Oral Exposure
Subchronic Studies
In a subchronic oral study conducted by American Biogenics Corp. (1989), albino rats
were exposed to benzenethiol by daily gavage for 90 days. This study was cited in the HEAST
(U.S. EPA, 2010b) and reported a LOAEL of 0.1 mg/kg-day based on centrilobular eosinophilic
changes in the liver. An electronic search revealed the online HEAST on February 19, 2010
continued to list this value as the RfD. However, a copy of this study was not available, and the
online link to the HEAST Derivation Support Document was not available at the time of the
preparation of this PPRTV document. Additionally, eosinophilic changes in the hepatocytes,
without corroborating evidence of liver toxicity that may or may not be available in the original
manuscript (e.g., increased alanine aminotransferase, liver weights, or incidences of other
microscopic findings in the liver), could be considered an adaptive response to the test material.
No other studies could be located regarding the effects of subchronic oral exposure of
animals to benzenethiol.
Chronic Studies
No studies could be located regarding the effects of chronic oral exposure of animals to
benzenethiol.
Developmental and Reproduction Studies
The effects of oral exposure of benzenethiol to animals have been evaluated in a
developmental toxicity study in rats (NTP, 1994a), a range-finding toxicity study in rabbits
(NTP, 1992), a subsequent full developmental toxicity study in rabbits (NTP, 1994b), and a
reproductive toxicity study in rats (NTP, 1996).
Developmental Toxicity Study in Rats
In the developmental toxicity study in rats (NTP, 1994a), benzenethiol (>99% pure) was
administered via gavage in corn oil to time-mated Sprague-Dawley (S-D) rats (Crl:CD®BR)
(25/dose group) at dose levels of 0, 20, 35, or 50 mg/kg-day from Gestation Days (GDs) 6-15.
Animals were observed daily for clinical signs of toxicity. Body weights were recorded on
GDs 0, 3, 6 through 15, 18, and 20. Food and water consumption were recorded for the animals
in each group on GDs 1, 3, 6, 9, 12, 15, 18, and 20. The dams were euthanized on GD 20 and
subjected to a gross necropsy. The liver, right kidney, and gravid uterus were weighed. The
numbers of corpora lutea in each ovary were counted. The number of implantation sites in the
uterus was counted, and any uterus with no visible implantation sites was stained with 10%
7
Benzenethiol
-------
FINAL
4-1-2011
ammonium sulfide to detect early resorptions. Live fetuses were euthanized, weighed, examined
for external abnormalities, and dissected for visceral examination. Half of the fetuses in each
litter were decapitated prior to dissection; the heads were fixed in Bouin's solution and then
examined by a free-hand sectioning technique. All fetal carcasses were stained with Alcian
Blue/Alizarin Red S and examined for skeletal malformations.
In the NTP (1994a) study, four rats in the high-dose group died or were sacrificed in
extremis. Among the animals surviving to scheduled necropsy on GD 20, pregnancy rates were
100, 100, 96, and 100% in the 0-, 20-, 35-, and 50-mg/kg-day groups, respectively. Clinical
signs consisted primarily of rooting behavior after dose administration. The incidence of rooting
behavior increased with both dose and time, with this behavior first noted on GDs 11,8, and 6 in
the 20-, 35-, and 50-mg/kg-day groups, respectively, reaching a maximum incidence of 0%
(control), 28% (low dose), 92% (mid-dose) and 100% (high dose) by GD 15. This behavior was
not observed after the treatment period was concluded. The study authors concluded that the
rooting behavior was indicative of an aversion to the dosing formulation. Absolute and relative
(to body weight) maternal food consumption (g/kg-day) was statistically decreased by 9-28%
(p < 0.05) in all treated groups for the first 3 days of dosing (GDs 6-9). Only the high dose,
50 mg/kg-day, maintained a statistical (p < 0.05) reduction in relative food consumption for
GDs 9-12. For the last 3 days of dosing, GDs 12-15, the relative food consumption for all
treatment groups was statistically the same as the control group. Overall, absolute and relative
maternal food consumption was decreased for the high-dose (50 mg/kg-day) group by 14—18%
(p < 0.05) for the entire dosing period (GDs 6-15). Conversely, relative maternal food
consumption was increased by 11% (p < 0.05) in the 50-mg/kg-day dams during the
posttreatment period (GDs 15-20), and absolute and relative maternal water consumption at
50 mg/kg-day was 20-33% higher than the control group throughout the treatment and
posttreatment intervals. Maternal body-weight change was statistically dose-dependently
decreased by 31—102% (p < 0.05) in all treatment groups for GDs 6-9 (see Table B. 1). Maternal
body weight was significantly decreased by 6—8% at 50 mg/kg-day compared to controls from
GD 9 to termination on GD 20. Additionally in the 50-mg/kg-day group, body-weight gain was
decreased by 25% (p < 0.05) for GDs 9-12, by 33% (p < 0.05) for the overall treatment period
(GDs 6-15), by 20% (p < 0.05) for the overall gestation period (GDs 0-20), and by 17%
(p < 0.05) when corrected for gravid uterine weight. The effects on maternal food and water
consumption likely resulted in the changes in maternal body-weight gain throughout the dosing
period. Relative (percent body weight) and adjusted (for maternal body weight) liver weights
were increased by 10—18% (p < 0.05) over controls in the 50-mg/kg-day group. Gravid uterine
weight was decreased by 22% (p < 0.05) at the high dose. The maternal LOAEL is
50 mg/kg-day based on increases in relative and adjusted maternal liver weights and decreases in
gravid uterine weight. The maternal NOAEL is 35 mg/kg-day.
Table B.2 presents cesarean section and fetal examination data from the NTP (1994a)
study. Postimplantation loss was increased at 50 mg/kg-day, as evidenced by increases in the
percent of resorptions/litter (15.5% vs. 1.5% controls) and the number of litters with resorptions
(52 vs. 24 controls). The number of live fetuses per litter was decreased by 20% at
50 mg/kg-day compared to controls. Male and female fetal body weights were 10% lower than
controls at this dose. At 35 mg/kg-day, female fetal body weights were significantly decreased
by 5%> (p < 0.05) compared to controls. The incidence of external malformations (including
anophthalmia, an open eye, cleft lip and/or palate, anasarca, gastroschisis, micromelia, and
8
Benzenethiol
-------
FINAL
4-1-2011
syndactyly of the hind and forepaw) was increased in the high-dose group (1.9% fetuses;
19.0% litters) compared to concurrent controls (0.3%; 4.0% litters). This incidence also
exceeded the provided historical control incidence of 0.2% fetuses in 1.2% litters. The historical
control data comprised 1222 fetuses from 82 litters from studies conducted in 1988 by
NIEHS/NTP Contract No. N01-ES-55080 (RTI Project No. 311U-2717) and NIEHS/NTP
Contract No. N01-ES-95255 (RTI Project No. 311U-4349). Combined, these data result in a
NOAEL for developmental toxicity of 20 mg/kg-day and a LOAEL of 35 mg/kg-day based on
reduced female fetal body weight. This developmental toxicity study in rats is considered
acceptable because a maternal LOAEL was observed, and comprehensive fetal examinations
were conducted to determine external, visceral, and skeletal malformations and variations.
Developmental Toxicity Study in Rabbits
The dose levels for the definitive developmental toxicity study were based on data from
preliminary rabbit range-finding studies (NTP, 1992; summaries reported in NTP [1994b]). In
those studies, pregnant New Zealand White rabbits were dosed with corn oil (vehicle) or
benzenethiol (>99% pure) at 0.5, 1, 2, 5, or 10 mg/kg-day on GDs 6-19, and nonpregnant female
New Zealand White rabbits were dosed with 20, 40, or 50 mg/kg-day corn oil for 14 consecutive
days. Animals were weighed and observed for clinical signs of toxicity. No clinical signs were
observed in the adult animals at doses up to 40 mg/kg-day. At 50 mg/kg-day, one or
two nonpregnant females were described as slightly sedated postdosing on 2 days during the
dosing period. One of these animals died on Day 10. The study authors noted no effects on
body weight in either the pregnant or nonpregnant animals. No significant dose-related
developmental toxicity was noted. Therefore, the highest doses for the definitive developmental
toxicity study were selected to be 40 and 50 mg/kg-day in an effort to induce some maternal
toxicity without significant maternal lethality. The low exposure of 10 mg/kg-day was expected
to produce no maternal or developmental toxicity, based on the reported lack of treatment-related
effects in this preliminary study.
In the definitive developmental toxicity study in rabbits (NTP, 1994b), benzenethiol
(>99%) pure) was administered via gavage in corn oil to artificially inseminated New Zealand
White rabbits at dose levels of 0, 10, 30, 40, or 50 mg/kg-day from GDs 6-19. Twenty-four to
26 animals were assigned to each dose group, with the exception of the 40-mg/kg-day group, to
which 15 does were assigned. The authors stated that a slightly higher dose of 50 mg/kg-day
was excluded from the final assessment due to excessive maternal toxicity, with 6/13 does dying
during the first week of treatment; the remaining animals in the 50-mg/kg-day group were then
euthanized by GD 14. Maternal body weights were determined on GDs 0, 3, 6 through 19, 25,
and 30. Animals were observed for clinical signs of toxicity at least once daily before, during,
and after the treatment period. Maternal food consumption was recorded every 3 days from
GDs 0 through 18, and also on GDs 19, 22, 25, 28, and 30. All surviving does were euthanized
on GD 30. The does were subjected to postmortem examination, cesarean section, organ-weight
analyses, and gross necropsy. The liver, right kidney, and gravid uterus were weighed. The
numbers of corpora lutea in each ovary were counted. The number of implantation sites in the
uterus was counted, and any uterus with no visible implantation sites was stained with
10%) ammonium sulfide to detect early resorptions. Live fetuses were euthanized, weighed,
examined for external abnormalities, and dissected for visceral examination. Half of the fetuses
in each litter were decapitated prior to dissection; the heads were fixed in Bouin's solution and
9
Benzenethiol
-------
FINAL
4-1-2011
then examined by a free-hand sectioning technique. All fetal carcasses were stained with Alcian
Blue/Alizarin Red S and examined for skeletal malformations.
NTP (1994b) reported two deaths during the study; one doe in the 10-mg/kg-day group
died following dosing on GD 13, and one doe in the 30-mg/kg-day group died after dosing on
GD 6. Maternal relative (to body weight) food consumption was marginally affected by
treatment. In the treated animals, relative food consumption was comparable to controls before
the initiation of dosing; however, during the dosing period, relative food consumption showed a
statistically significant decreased linear trend, with decreases of 15 and 19% compared to
controls at 30 and 40 mg/kg-day, respectively (see Table B.3). Despite the decreased linear
trend, no individual exposure group demonstrated statistically significant decreases in food
consumption compared to the control group during treatment. After the dosing period, the trend
toward decreased food consumption was no longer evident.
Statistically significant pair-wise reductions in maternal body weight gain from the NTP
(1994b) study were observed in the 30- and 40-mg/kg-day groups only for GDs 12-15
(decreased 77-92%), the same period in which the largest reductions in food consumption were
seen in those two groups. A statistically significant (p < 0.05) decreased trend in maternal
weight gain was observed for the overall dosing period (GDs 6-19), with decreases of 14, 30,
and 61%) compared to controls in the 10-, 30-, and 40-mg/kg-day groups, respectively. When
corrected for gravid uterine weight, the maternal animals at 40 mg/kg-day experienced a body
weight loss of-51.0 g compared to a body-weight gain of 61.9 g in the controls. Necropsy of
maternal animals on GD 30 revealed no effects on maternal absolute or relative liver or right
kidney weight. Gravid uterine weight was also unaffected by treatment. There were no effects
of treatment on the numbers of resorptions (early, late, or complete litter), or fetal body weights
or sex ratio. There were no treatment-related external, visceral, or skeletal variations or
malformations in the fetuses. The investigators reported: a maternal NOAEL of 30 mg/kg-day;
minor and transient decreases in body weight gains and food consumption at 30- and
40 mg/kg-day; and excessive toxicity at 50 mg/kg-day based on maternal mortality and
morbidity in this study and in the dose-finding studies (as cited in NTP, 1994b). However, the
data support a maternal LOAEL of 40 mg/kg-day, based on the decreased body weight gain and
food consumption, along with the body weight loss when corrected for gravid uterine weight.
Although body weight gain and food consumption were also decreased at 30 mg/kg-day, this
dose level is considered the NOAEL because the decreases were of a smaller magnitude and did
not affect the corrected body weight gain. No effects of benzenethiol treatment on fetal
development or pregnancy were observed, indicating a developmental NOAEL of 40 mg/kg-day.
Assessment of potential developmental toxicity at 50 mg/kg-day was precluded by excessive
maternal toxicity. This developmental toxicity study in rabbits is considered acceptable because
a maternal LOAEL was observed, and comprehensive fetal examinations were conducted to
determine external, visceral, and skeletal malformations and variations.
Reproduction Study
The study by NTP (1996) is selected as the principal study for deriving the subchronic
and chronic p-RfD values. In a multigeneration reproduction toxicity study (NTP, 1996), male
and female F0 generation S-D (Crl:CD®BR) rats (Charles River Laboratories, Portage, MI) were
administered benzenethiol (101%> pure by HPLC) by daily gavage in corn oil at doses of 9, 18, or
35 mg/kg-day (20/sex/dose) and allowed to cohabitate for 16 weeks. Except when paired
10
Benzenethiol
-------
FINAL
4-1-2011
together during mating, the animals were housed individually. During cohabitation, any litters
born to the F0 animals were euthanized on Postnatal Day (PND) 1. Litters born after 17 weeks
(Fl) were raised until PND 21, when selected weanlings were administered benzenethiol at the
same doses as their parents. On PND 81, the Fl animals were allowed to cohabitate for 1 week
and were euthanized following delivery of their litters (F2).
During the continuous breeding phase of the NTP (1996) study, all litters were evaluated
on PND 1 and then euthanized. The total number of pups born, number of live and dead pups,
number of male and female pups, and total pup weight of each sex were obtained. Parental male
and female weights were obtained following delivery, and the dam was also weighed on PNDs 4,
7, 14 and 21. Feed consumption measurements for lactating dams were obtained on PNDs 1, 4,
7, 11, 14, 18 and 21. All animals were observed twice daily for mortality and signs of toxicity.
Upon sacrifice of animals, the following organs were weighed: liver, kidneys, right cauda
epididymis, right epididymis, prostate, seminal vesicles with coagulating glands, right testis, and
ovaries. Liver and kidneys were microscopically examined. However, clinical chemistry
parameters were not evalutated. Spermatid head count was determined from the right testis.
Sperm density, morphology, and motion analyses (computer-assisted) were evaluated from the
right cauda epididymis. Sperm parameters included: motility; velocity (|im/sec); linearity; ALH
max (|im); ALH mean (|im); beat/cross frequency (Hz/sec); average radius (|im); circular cells;
circular over motile cells (%); circular over all cells (%); epididymal sperm density
(1000 sperm/mg caudal tissue) and morphology (% abnormal); spermatids/mg testis; and total
spermatids/testis.
Table B.4 shows selected male body-weight results from the NTP (1996) study.
Throughout the study, the body weights of the 35-mg/kg-day F0 males were 7-15% lower than
controls. F0 female body weights were not affected by treatment. Body weights of the
35-mg/kg-day Fl parental males were 11-13% less than controls on Weeks 2 and 4,
respectively, at delivery of the Fl dams' litters, and at necropsy.
In the F0 generation (NTP, 1996), relative (to body weight) liver weights were increased
by 20, 35, and 50% (males) and by 11, 18, and 36% (females) in the 9-, 18-, and 35-mg/kg-day
groups, respectively (see Table B.5). Absolute liver weights were increased by 24, 34, and 29%
(males) and by 5, 13, and 25% (females). In the Fl generation, at 9, 18, and 35 mg/kg-day,
absolute liver weights were increased over controls by 24, 30, and 34% in the males, and by 9,
15, and 34% in the females. Relative liver weights were increased by 18, 37, and 62% in the
Fl males, and by 13, 17, and 42% in the Fl females. Centrilobular hepatocellular hypertrophy
was observed in the parents as follows (see Table B.6): in the F0 males (90-100% vs.
0% controls) at 18 and 35 mg/kg-day; in the F0 females (90-100%) vs. 0% controls) at 9, 18, and
35 mg/kg-day; in the Fl males (100% vs. 0% controls) at 9, 18, and 35 mg/kg-day; and in the
Fl females at 9 (30%), 18 (100%), and 35 (100%) mg/kg-day. The hepatocellular hypertrophy
showed a dose-dependent increase in severity in both sexes from both generations. Aside from
the hepatocellular hypertrophy, there were no other gross or microscopic changes in the liver
indicative of liver toxicity.
In the NTP (1996) study, F0 relative kidney weights of the 9-, 18-, and 35-mg/kg-day
animals were increased by 30, 53, and 104% (males) and by 8, 5, and 20% (females),
respectively (see Table B.7). Absolute kidney weights were increased by 35, 53, and 76% in the
11
Benzenethiol
-------
FINAL
4-1-2011
9-, 18-, and 35-mg/kg-day males; whereas in the females, absolute kidney weights were only
increased at 35 mg/kg-day (12% over controls). In the F1 generation, absolute kidney weights
were increased by 62, 61, and 118% over controls in the 9, 18, and 35 mg/kg-day males,
respectively, and by 17% in the 35-mg/kg-day females. Relative kidney weights were increased
by 52, 67, and 163% in the F1 males, and by 12, 6, and 26% in the F1 females. Table B.8
depicts the incidences of gross findings in the kidneys at necropsy. In the F0 males, at
35 mg/kg-day, there was a treatment-related increase in the incidence of enlarged kidneys
(2/10 vs. 0/10 controls) and pitted kidneys (4/10 vs. 1/10 controls). In the F1 males, at necropsy,
there was a treatment-related increase in the incidence of enlarged kidneys at 9, 18, and
35 mg/kg-day (20, 10, and 40%, respectively, vs. 0% controls), pale kidneys at 9, 18, and
35 mg/kg-day (90, 100, and 90% respectively, vs. 0% controls), and soft kidneys at
35 mg/kg-day (20% vs. 0% controls). Increased incidences of renal tubule degeneration (see
Table B.9) were observed in the F0 males (100% vs. 50% controls) and F1 males (100% vs.
0% controls) at 9, 18, and 35 mg/kg-day, in the F0 females at 9, 18, and 35 mg/kg-day (20-40%
vs. 10%) controls), and in the F1 females at 35 mg/kg-day (40% treated vs. 0% controls). Renal
tubule degeneration also showed a dose-dependent increase in severity in both sexes from both
generations.
The investigators reported a LOAEL for parental toxicity at 9 mg/kg-day in the NTP
(1996) study, based on liver and kidney toxicity (increases in both absolute and relative liver and
kidney weights, as well as centrilobular hepatocellular hypertrophy and renal tubule
degeneration) in both F0 and F1 generations. A parental NOAEL was not established.
Reproductive evaluations were performed on sperm and reproductive organs of both the
F0 and F1 generations in the NTP (1996) study. Sperm motility was decreased by 6% compared
to controls at 18 mg/kg-day and by 5% at 35 mg/kg-day compared to controls (see Table B.10).
Inhibited spermiation of Stage VIII-X tubules was observed in the F1 males at 9 mg/kg-day
(60%), 18 mg/kg-day (60%), and 35 mg/kg-day (90%) compared to controls (0%). The mean
percent of tubules affected was 10, 9.5, and 7.7% of the "vulnerable" tubules in the 9, 18, and
35 mg/kg-day groups, respectively. Spermatid and spermatocyte cellular morphology appeared
normal in all animals. Neither epithelial disorganization nor cell sloughing was observed in any
testes examined. No microscopic lesions were observed in the epididymis or ovaries of the
F1 animals. All of the other above-mentioned parameters regarding sperm count and
computer-assisted motion analyses in the treated groups were comparable to controls. Other
reproductive endpoints at necropsy were comparable among dose groups. The LOAEL for
reproductive toxicity is 9 mg/kg-day based on inhibited spermiation in the F1 males. A
reproductive NOAEL was not established.
The offspring (F1 and F2) of exposed parents in the NTP (1996) study were examined for
number of live pups, the number of male and female pups, and body-weight changes, and the
total pup weight of each sex was recorded on PNDs 1, 4, 7, 14, and 21. Selected pup
body-weight data are included in Table B. 11. Five F1 litters were born during the 16 weeks of
cohabitation of the F0 generation. The F1 live pup weight (adjusted for litter size) was decreased
by 4 and 6% in the 9- and 35-mg/kg-day dose groups, respectively. In the offspring from the
final F1 litter, the pup weights at 35-mg/kg-day were significantly decreased by 14—16% in the
males on PNDs 4 and 7 and in the females on PND 7; however, no differences were observed on
PNDs 1, 14, or 21. There was no treatment-related increase in preweaning mortality of the
12
Benzenethiol
-------
FINAL
4-1-2011
F1 animals. In the F1 mating trial, live F2 pup weight for the combined sexes was decreased by
9 and 12% in the 18-and 35-mg/kg-day dose groups, respectively, when compared to controls.
Other endpoints were unchanged. The investigators reported a LOAEL for offspring toxicity of
35 mg/kg-day based on decreased pup body weights. However, the data indicate a LOAEL of
18 mg/kg-day based on decreased body weights in the F2 pups. Although the decrease at this
dose is relatively minor, it is dose-dependent, statistically significant, and (for pups) often
biologically adverse. The NOAEL is 9 mg/kg-day.
Using the 35-mg/kg-day dose, a crossover mating trial in the NTP (1996) study revealed
the females as the affected sex. When naive males were mated with control or 35-mg/kg-day
females, the mean live pup weight and adjusted live pup weight were reduced in the
35-mg/kg-day group by 8-9%. No other treatment-related effects were seen (total number of
pups born, number of live and dead pups, number of males and female pups, and total pup
weight by sex were obtained). When naive females were mated with control or 35-mg/kg-day
males, reproductive parameters were comparable between dose groups. No differences were
observed in the pregnancy index, cumulative days to litter, mean average litters per pair,
proportion of pups born alive, or sex ratio of pups. This study meets the criteria for an
acceptable reproductive toxicity study, in that it was conducted for two generations under
continuous exposure and examined a comprehensive suite of parameters to determine effects on
parents, offspring, and reproduction.
Inhalation Exposure
The only inhalation studies found were short-term (acute and subacute) lethality studies
in rats and mice. Summaries of these studies are included below in the following section on
"Other Data (Short-Term Tests, Other Examination)".
Subchronic Studies
No studies could be located regarding the effects of subchronic inhalation exposure of
animals to benzenethiol.
Chronic Studies
No studies could be located regarding the effects of chronic inhalation exposure of
animals to benzenethiol.
Developmental and Reproduction Studies
No studies could be located regarding the effects of inhaled benzenethiol on reproduction
or fetal development.
Other Data (Short-Term Tests, Other Examination)
Acute and Subacute Inhalation Studies
Fairchild and Stokinger (1958) exposed groups of 5-10 Swiss-derived male mice (body
weight 25-28 g) to 20-, 31-, 41-, 52-, or 79-ppm benzenethiol and 5-10 Wistar-derived male rats
(body weight 180-220 g) to 20-, 31-, 41-, 52-, 79-, or 132-ppm benzenethiol for 4 hours,
followed by a 15-day observation period. Clinical signs included increased respiration and
restlessness (hyperactivity), uncoordinated movement, staggering gait, muscular weakness,
partial skeletal muscle paralysis beginning in the hind limbs, light to severe cyanosis, tolerance
of a prone position, and mild-to-heavy sedation. Animals exposed to "maximal lethal
13
Benzenethiol
-------
FINAL
4-1-2011
concentrations" typically died from respiratory arrest during exposure or shortly after removal
from the chamber. Animals exposed to "minimal lethal concentrations" typically died while in a
semiconscious condition of "long duration." Surviving animals often remained in a
semi-conscious state of sedation and lethargy 4 to 6 hours post exposure before showing signs of
recovery. For mice, an LC50 value of 28 ppm was calculated by the study authors. A BMC01 of
26.5 ppm and BMCL05 of 18.5 ppm were also calculated by the study authors. LC05 and LC01
values could not be calculated by the method of Litchfield and Wilcoxon because there were not
at least two concentrations showing between 0 and 100% mortality. In rats, an LC50 value of
33 ppm was calculated by the study authors. A BMC01 of 17.7 ppm and BMCL05 of 13.4 ppm
were also calculated by the study authors. An LC05 value of 15.5 ppm and LC01 value of
10.3 ppm were calculated by the method of Litchfield and Wilcoxon.
In an acute inhalation toxicity study conducted by Stauffer Chemical Company (1969),
groups of five albino rats/sex/dose were exposed to 244-, 346-, or 595-ppm benzenethiol for
1 hour, followed by a 14-day observation period. Clinical signs included ocular edema and
erythema, and slight nasal discharge in all test groups. "Acute depression" was reported in the
244-ppm group, and dyspnea, gagging, fasiculation, and cyanosis were reported in the 346- and
595-ppm groups while the animals were in the exposure chamber. There were no
treatment-related deaths in the 244-ppm group, and all animals in this dose group appeared
grossly normal at necropsy after terminal sacrifice. Treatment-related mortality was noted in
3/10 animals at 346 ppm and 10/10 animals at 595 ppm. Decedents exhibited areas of
hemorrhage in the lungs, while survivors in the 346-ppm group appeared grossly normal. The
authors calculated an LC50 of 422 ppm. No further experimental details were available.
In a subacute inhalation toxicity study conducted by Hazleton Laboratories (1951),
7 adult male albino rats and 12 adult male albino mice were exposed to an "atmosphere saturated
with benzenethiol" for 6 hours on the first exposure day and for 8 hours on each of the
3 succeeding days. The mice exhibited excitement, preening, and slight salivation during the
first 6-hour exposure period. The following morning, 7 mice were found dead, but the surviving
5 mice appeared normal (Group A). A second group of 13 adult albino male mice was added to
the experiment (Group B). All mice were then exposed 8 hours/day, for 3 consecutive days. Of
the five remaining mice from Group A, two died on Day 2 of exposure, two died on Day 4, and
the one died 3 days after the final exposure. Hemorrhagic lungs, irritation of the intestines, and
spotted livers and kidneys were noted at necropsy. Group B mice also exhibited preening,
lacrimation, and salivation immediately upon exposure and, subsequently, were lethargic and
appeared unkempt. Mortality was observed in 11/13 mice from Group B; deaths occurred from
Day 1 through 3 days after the final exposure. Hemorrhagic lungs, irritated intestines, and spotty
livers and kidneys were noted in both decedents and animals terminated 3 days after the final
exposure. Rats showed preening, lacrimation, and marked salivation during exposure, followed
by unkempt appearance and lethargy. One rat died overnight after the final exposure, and
another died 3 days after the final exposure. Hemorrhagic lungs, intestinal irritation, and mottled
livers and kidneys were noted in the decedents. Surviving rats terminated 3 days after the final
exposure showed gas-filled and irritated stomachs and intestines, pale brown kidneys, small
spleens, mottled livers, and irritated eyes.
14
Benzenethiol
-------
FINAL
4-1-2011
Metabolism, Mode-of-Action and Structure-Activity Relationship Studies
A discussion of the metabolism, mechanism of toxicity, and structure-activity
relationships with related chemicals and their relative toxicity can be found in a report deriving
the interim Acute Exposure Guideline Levels (AEGLs) for benzenethiol (U.S. EPA, 2007).
Table 3 summarizes the studies on short-term inhalation, mechanism of toxicity,
structure-activity relationships, genotoxicity, and metabolism.
15
Benzenethiol
-------
FINAL
4-1-2011
Table 3. Other Studies
Tests
Materials and Methods
Results
Conclusions
References
Acute inhalation
mouse
20-, 31-, 41-, 52-, or 79-ppm vapor for
4 hours, followed by a 15-day observation
period.
Clinical signs indicative of central
nervous system depression and
respiratory distress, including increased
respiration, restlessness (hyperactivity),
uncoordinated movement, staggering
gait, muscular weakness, skeletal muscle
paralysis, light to severe cyanosis, and
coma.
The only inhalation studies
found were short-term lethality
studies in rats and mice.
Therefore, these acute studies
are included.
LC50 = 28 ppm
BMCoi = 26.5 ppm
BMCL05 =18.5 ppm
Fairchild and
Stokinger (1958)
Acute inhalation
rat
20-, 31-, 41-, 52-, 79-, or 132-ppm vapor for
4 hours, followed by a 15-day observation
period.
Clinical signs indicative of central
nervous system depression and
respiratory distress, including increased
respiration, restlessness (hyperactivity),
uncoordinated movement, staggering
gait, muscular weakness, skeletal muscle
paralysis, light to severe cyanosis, and
coma.
LC50 = 33 ppm
LC05 = 15.5 ppm
LCoi = 10.3 ppm
BMCoi = 17.7 ppm
BMCL05 =13.4 ppm
Fairchild and
Stokinger (1958)
Acute inhalation
rat
20-, 31-, 41-, 52-, 79-, or 132-ppm vapor for
4 hours, followed by a 15-day observation
period for benzenethiol and ethyl mercaptan.
LC50s (4-hours):
Benzenethiol: 33 ppm
Ethyl mercaptan: 4740 ppm
Methyl mercaptan: 675 ppm
Hydrogen sulfide: 444 ppm
Relative toxicity compared to
similar chemicals (Quantitative
structure-activity relationship
[QSAR]) indicates that the
toxicity of benzenethiol is
greater than ethyl mercaptan
(approximately 140-fold) and
methyl mercaptan
(approximately 20-fold).
Fairchild and
Stokinger (1958)
Acute inhalation
rat
5/sex/dose group exposed to 244-, 346-, or
595-ppm vapor for 1 hour followed by a
14-day observation period.
Clinical signs included: ocular edema
and erythema and slight nasal discharge
in all groups; "acute depression" at 244
ppm; and dyspnea, gagging,
fasciculation, and cyanosis at >346 ppm.
Lung hemorrhage observed in decedents.
LC50 = 422 ppm
Stauffer Chemical
Company (1969)
16
Benzenethiol
-------
FINAL
4-1-2011
Table 3. Other Studies
Tests
Materials and Methods
Results
Conclusions
References
Subacute
inhalation rat and
mouse
"Saturated" atmosphere for 6 hours on the
first day followed by 8 hours on each of
3 succeeding days.
Similar clinical signs as acute studies
listed above.
At necropsy, hemorrhagic lungs,
intestinal irritation, and mottled liver and
kidneys noted.
No NOAF.I. LOAEL, orLC50
reported because animals were
only exposed to one
concentration, which was not
measured (i.e., only referred to
as "saturated").
Hazleton
Laboratories (1951)
Mechanistic
human RBC in
vitro
Adult human blood samples were suspended
with or without 5-mM glucose and various
concentrations of benzenethiol,
4-aminothiophenol, or corresponding
disulfides. The percentage of
oxyhemoglobin, methemoglobin, and
nonintact hemoglobin was determined.
Intracellular levels of NADH, NADPH, and
reduced glutathione were measured. Flux
through the hexose monophosphate shunt
was measured by following 14C02 formation
from the labeled glucose. Flux through
glycolysis was determined by measurements
of pyruvate and lactate in the medium and
red blood cell compartment.
Auto-oxidation of benzenethiol resulted
in production of a reactive oxygen
species, causing the conversion of
oxyhemoglobin to methemoglobin.
Reduction of the disulfide by
intracellular glutathione caused cyclic
reduction/oxidation reactions, resulting
in oxidative flux. Glycolysis and the
hexose monophosphate shunt were
inhibited at the intermediate (0.5-mM
benzenethiol) and high levels of
oxidative stress.
Benzenethiol at 0.25-mM
concentration indicated a level
of oxidative stress to which the
cell is capable of an adaptive
response.
Amrolia et al.
(1989)
Mechanistic
Benzenethiol and other mercaptans induce toxicity by interrupting electron transport via inhibition of cytochrome
oxidase. As a result of the electron transfer blockage, oxidative phosphorylation and aerobic metabolism may be
affected, peripheral tissue P02 increases, and the uploading gradient for oxyhemoglobin decreases. High oxygen
concentrations are found in the venous return, resulting in a flushed appearance of the skin and mucous membranes. An
increased demand is placed on glycolysis, resulting in lactic acidemia. Repeated-dose studies indicate that kidney
effects may be due to the phenol moiety.
EPA (2007);
NIOSH (1978)
Genotoxicity
Tested for reverse mutation in Salmonella
typhimurium (Ames assay) with and without
metabolic activation.
Negative in strain TA100 and TA98 with
or without S9 activation.
These results indicate that
benzenethiol is not mutagenic in
the Ames assay.
Lavoie et al. (1979)
17
Benzenethiol
-------
FINAL
4-1-2011
Table 3. Other Studies
Tests
Materials and Methods
Results
Conclusions
References
Metabolism rat
Oral administration of ".S'-mcthvlphcnyl
sulfone. One hour after administration,
excreted urine was extracted with benzene,
and the aqueous layer was acidified with
sulfuric acid and extracted with ether. The
benzene-soluble and water-soluble products
were analyzed using thin layer
chromatography and gas-liquid
chromatography.
".S'-mcthylphcnyl sulfone was the only
benzene-soluble metabolite identified.
Trace amounts of methylphenyl
sulfoxide were also identified.
Benzenethiol readily undergoes
.S'-mcthvlation. followed by
oxidation of phenylsulfide to
methylphenyl sulfone.
McBain and Menn
(1969)
18
Benzenethiol
-------
FINAL
4-1-2011
DERIVATION OF PROVISIONAL VALUES
Table 4 below presents a summary of noncancer oral reference values. No cancer
values could be derived (see Table 5). Because there were no subchronic or chronic inhalation
studies, the toxicity values were not converted to HEC units. For the oral noncancer studies by
gavage, the only conversion was to provide an adjusted daily dose.
DERIVATION OF ORAL REFERENCE DOSE
Derivation of Subchronic p-RfD
The multigenerational study by NTP (1996) is selected as the principal study for
derivation of a subchronic p-RfD. The critical endpoints are increased absolute and relative
kidney weights and incidences of renal tubule degeneration in the male rats.
19
Benzenethiol
-------
FINAL
4-1-2011
Table 4. Summary of Reference Values for Benzenethiol (CASRN 108-98-5)
Toxicity Type (units)
Species/Sex
Critical Effect
/;-Reference
Value
POD
Method
POD
UFC
Principal Study
Subchronic p-RfD
(mg/kg-day)
S-D Rat/M
Increased kidney weights3
1 x 1(T2
BMDS
2.91
300
NTP (1996)
Chronic p-RfD
(mg/kg-day)
S-D Rat/M
Increased kidney weights3
1 x 1(T3
BMDS
2.91
3000
NTP (1996)
Subchronic p-RfC (mg/m3)
None
None
None
None
None
None
None
Chronic p-RfC (mg/m3)
None
None
None
None
None
None
None
a Renal tubule degeneration was observed in 100% of the treated F0 males compared to 50% controls and was dose-dependently increased in severity.
Table 5. Summary of Cancer Values for Benzenethiol (CASRN 108-98-5)
Toxicity Type
Species/Sex
Tumor Type
Cancer Value
Principal Study
p-OSF
None
None
None
None
p-IUR
None
None
None
None
20
Benzenethiol
-------
FINAL
4-1-2011
Table 6 summarizes the studies available for use in deriving provisional oral toxicity
values for benzenethiol; these studies include a developmental gavage study in rats
(NTP, 1994a), a developmental gavage study in rabbits (NTP, 1994b), and a two-generation
reproductive study in rats (NTP, 1996). The developmental toxicity studies in rats and rabbits
(NTP, 1994a,b) reported developmental effects in rats only at high dose (50 mg/kg-day; external
malformations) and no developmental effects in rabbits. The body weight changes reported in
these studies were associated with decreased food intake at the lowest dose (20 mg/kg-day). In
contrast, the renal and hepatic effects (increased renal and hepatice weight, renal tubular
degeneration, and hepatocellular hypertrophy) in the reproductive study (NTP, 1996) provide a
LOAEL (9 mg/kg-day) and is therefore selected for derivation of the subchronic p-RfD. The
selection of this study is justified because the F0 parents were dosed for 16 weeks, during which
time, the dams had five litters. The pups of the first four litters were terminated on PND 1, and
parents for the F1 generation were selected from the final litter. These parental animals were
allowed to cohabitate on PND 81 for 1 week, and were euthanized following delivery of their
litters (F2). Therefore, overall post-natal dosing of these F1 parental rats was comparable to F0
generation. Renal tubular degeneration was observed in male rats from both the F0 and F1
generation at 9 mg/kg-day. Hepatocellular hypertrophy was also observed in male rats at
9 mg/kg-day in the F1 generation and 18 mg/kg-day in the F0 generation. Increased relative
kidney and liver weights were also observed at all doses in both F0 and F1 generation rats, with
the male rats demonstrating more sensitivity. Since the systemic effects were observed at a
lower dose (9 mg/kg-day) in contrast to the developmental effects observed at a higher dose
(50 mg/kg-day), it appears the parental animals are more sensitive to benezenethiol; the selection
of POD of 9.0 mg/kg-day may be protective of both parental and developmental effects.
Table 6. Summary of Oral Systemic Toxicity Studies for Benzenethiol
References
#/Sex
(M/F)
Exposure
(mg/kg-day)
Frequency/
Duration
NOAELW
(mg/kg-day)
LOAELU).|b
(mg/kg-day)
Critical Endpoint
NTP (1994a)
25 F rats
0, 20, 35, 50
7 d/wk for
GDs 6-15
gavage
C
20
Decreased body-
weight gain and
food consumption
NTP (1994b)
15-26 F
rabbits
0, 10, 30, 40,
50
7 d/wk for
GDs 6-19
gavage
10
30
Decreased body-
weight gain and
food consumption
NTP (1996)
20/20 rats
0, 9, 18, 35
7 d/wk for
16 wks
(males)/19 wks
(females)
c
9
Increased kidney
weights, renal
tubule
degeneration
NTP (1996)
20/20 rats
0, 9, 18, 35
7 d/wk for
16 wks
(males)/19 wks
(females)
c
9
Increased liver
weights,
hepatocellular
hypertrophy
aNOAELADj = NOAEL x (gavage schedule).
YOAELadj = LOAEL x (gavage schedule).
°No NOAEL was identified. A NOAEL is considered equal to a LOAEL/10 for screening purposes.
21
Benzenethiol
-------
FINAL
4-1-2011
Data from the F0 male and female rats from the reproduction study (NTP, 1996) study
indicating treatment-related findings in the liver and kidney are considered in order to select a
point of departure (POD) for the derivation of the subchronic p-RfD. The treatment-related
effects in the liver included increased liver weight and hepatocellular hypertrophy. Evidence of
toxic effects on the kidney was characterized by increased kidney weights and increased
incidence and severity of renal tubule degeneration. Given the similarity in responses between
both generations, the endpoints from which the subchronic oral RfD is derived are restricted to
the F0 generation which are more amenable to benchmark dose modeling.
Data depicting treatment-related effects on the liver in the F0 rats are considered for
BMD modeling. The histological data from the F0 males are not suitable for BMD modeling, as
the incidence of hepatocellular hypertrophy was 0, 0, 100, and 90% in the F0 males at 0, 9, 18,
and 35 mg/kg-day, respectively. An attempt at BMD modeling of dichotomous data for
hepatocellular hypertrophy in the F0 males results in all seven dichotomous models failing the
goodness-of-fit test (p-value < 0.1). In the F0 females, the incidence of hepatocellular
hypertrophy was 0, 90, 100, and 100% of the rats in the 0, 9, 18, and 35 mg/kg-day groups,
respectively. The gamma, log logistic, and log probit models all result in a BMD/BMDL ratio
> 20. The remaining dichotomous models yield BMDL values ranging from 0.2-1.3 mg/kg-day.
The low BMDL values are due to the steep dose-response curve; however, it is for this very
reason that their precision is questionable. In order to adequately describe the lower part of the
dose-response curve, an intermediate dose level between the control group (with 0% incidence)
and the 9 mg/kg-day group (with 90% incidence) is necessary. Furthermore, although 90% of
the F0 females exhibited hepatocellular hypertrophy at the low dose, the severity at this dose was
only minimal to mild.
Absolute and relative liver weight data, presented in Tables 7 through 10, were
considered for BMD modeling. The data on increased absolute and relative liver weights in
F0 male rats exposed to benzenethiol via gavage (NTP, 1996) were modeled using the
continuous-variable models in the EPA Benchmark Dose Software (BMDS), version 2.1
(U.S. EPA, 1999). Per EPA policy, in the absence of a biologically relevant benchmark response
level (BMR), a default BMR of 1 standard deviation (SD) above the control mean is used for
modeling.
22
Benzenethiol
-------
FINAL
4-1-2011
Table 7. Absolute Liver Weights (g) in the F0 Male Rats Following
16-Week Exposure to Benzenethiol to be Used for BMD Analysis"
DOSE
(mg/kg-day)
DOSEadj
(mg/kg-day)
Number of Subjects
Responseb
0
0
20
27.5 ±4.20
9
9
10
34.2 ±5.69*
18
18
10
36.9 ±6.01*
35
35
10
35.4 ±5.38*
aNTP (1996).
bMeans ± SD. Standard deviation was calculated from standard error / Vn
Statistically significantly different from control (p < 0.05) by pair-wise comparison.
Table 8. Relative Liver Weights (mg/g Body Weight) in the F0 Male Rats Following
16-Week Exposure to Benzenethiol to be Used for BMD Analysis"
DOSE
(mg/kg-day)
DOSEadj
(mg/kg-day)
Number of Subjects
Responseb
0
0
20
35.3 ±4.07
9
9
10
42.2 ±3.16*
18
18
10
47.7 ±6.96*
35
35
10
53.0 ±5.69*
aNTP (1996).
bMeans ± SD. Standard deviation was calculated from standard error / Vn
Statistically significantly different from control (p < 0.05) by pair-wise comparison.
Table 9. Absolute Liver Weights (g) in the F0 Female Rats Following
19-Week Exposure to Benzenethiol to be Used for BMD Analysis"
DOSE
(mg/kg-day)
DOSEadj
(mg/kg-day)
Number of Subjects
Responseb
0
0
20
16.3 ± 1.70
9
9
10
17.1 ± 1.14
18
18
10
18.5 ± 1.77*
35
35
10
20.3 ± 1.96*
aNTP (1996).
bMeans ± SD. Standard deviation was calculated from standard error / Vn
Statistically significantly different from control (p < 0.05) by pair-wise comparison.
23
Benzenethiol
-------
FINAL
4-1-2011
Table 10. Relative Liver Weights (mg/g Body Weight) in the F0 Female Rats Following
19-Week Exposure to Benzenethiol to be Used for BMD Analysis"
DOSE
(mg/kg-day)
DOSEadj
(mg/kg-day)
Number of Subjects
Responseb
0
0
20
35.0 ± 2.41
9
9
10
38.9 ±2.88*
18
18
10
41.3 ± 4.11*
35
35
10
47.6 ±5.06*
aNTP (1996).
bMeans ± SD. Standard deviation was calculated from standard error / Vn
Statistically significantly different from control (p < 0.05) by pair-wise comparison.
When data for absolute liver weights in the F0 males are modeled using constant
variance, the linear, polynomial, and power models fail the goodness-of-fit test (p-value
test 4 < 0.1). Although the Hill model with constant variance provides the lowest AIC and
lowest BMDL, the BMD/BMDL ratio is >20, indicating unacceptable uncertainty at the lower
end of the dose-response curve. When these data are modeled using nonconstant variance, all
four continuous variable models result in the wrong variance model (/> value test 2 < 0.1).
When relative liver weight data for the F0 males are modeled using constant variance, all
four continuous variable models indicate a poor variance model (p-value test 3 is < 0.1) and
wrong variance models. Results from subsequent modeling of the relative liver weight data
using nonconstant variance are included in Table 11. The Hill and polynomial models fail the
goodness-of-fit test. The linear and power models produce identical results, indicating that the
simpler linear model is more appropriate, with a BMDL of 5.20 mg/kg-day.
24
Benzenethiol
-------
FINAL
4-1-2011
Table 11. Model Predictions for Increases in Relative Liver Weights in the Male F0 Rats
Exposed Orally to Benzenethiol for 16 weeks"
Model Name
Homogeneity
Variance
p-Value
Goodness of
Fit
/;-Valueh
AIC for
Fitted Model
BMDisd
(mg/kg-day)
BMDLisd
(mg/kg-day)
Hill
0.060
<0.1
214.71
5.96
3.56
Linear
0.060
0.122
214.91
7.17
5.20
Polynomial0
0.060
<0.0001
10.00
-999.00
-999.00
Power
0.060
0.122
214.91
7.17
5.20
aNTP (1996).
bValues <0.10 fail to meet conventional goodness-of-fit criteria.
Invalid BMD and BMDL
AIC = Akaike's Information Criteria; BMD = benchmark dose; BMDL = lower confidence limit (95%) on the
benchmark dose.
Results from modeling liver weight data from the F0 females (Table 12) indicate that
constant variance models are appropriate for absolute liver weight data and nonconstant variance
models are appropriate for relative liver weight data. For absolute liver weights, the linear model
with constant variance results in the lowest AIC and lowest BMDL (10.70 mg/kg-day). For
relative liver weights, the linear nonconstant variance model is the simplest model that fits the
data, yielding a BMDL of 4.91 mg/kg-day.
25
Benzenethiol
-------
FINAL
4-1-2011
Table 12. Model Predictions for Increases in Absolute and Relative Liver Weights in the
Female F0 Rats Exposed Orally to Benzenethiol for 19 weeks"
Model Name
Homogeneity
Variance
p-Value
Goodness of
Fit
/;-Valueh
AIC for
Fitted Model
BMDisd
(mg/kg-day)
BMDLisd
(mg/kg-day)
Absolute Liver Weights (constant variance)
Hill
0.389
<0.1
107.53
14.15
8.15
Linear
0.389
0.877
103.79
13.89
10.70
Polynomial
0.389
0.608
105.79
13.95
10.70
Power
0.389
0.615
105.78
14.29
10.71
Model Name
Variance Model
p-Value
Goodness of
Fit
/;-Valueh
AIC for
Fitted Model
BMDisd
(mg/kg-day)
BMDLisd
(mg/kg-day)
Relative Liver Weights (nonconstant variance)
Hill
0.036
0.663
173.54
5.96
-999.00
Linear
0.036
0.850
171.68
6.56
4.91
Polynomial
0.036
0.850
171.68
6.56
4.91
Power
0.036
0.850
171.68
6.56
4.91
aNTP (1996).
bValues <0.10 fail to meet conventional goodness-of-fit criteria.
AIC = Akaike's Information Criteria; BMD = benchmark dose; BMDL = lower confidence limit (95%) on the
benchmark dose.
In summary, treatment-related effects of benzenethiol on the liver are limited to increased
liver weights and hepatocellular hypertrophy. There are no other histopathology findings in the
liver. Overall, the lowest BMDL value is for relative kidney weight in the F0 males: therefore,
the effects on kidneys are considered to be the most appropriate critical effect for the POD for
benzenethiol.
The BMDL modeling results for absolute and relative kidney weights of F0 male rats are
shown in Tables 13 and 14, respectively. The data on increased absolute and relative kidney
weights in male rats exposed to benzenethiol via gavage (NTP, 1996) are modeled using the
continuous-variable models in the EPA BMDS (version 2.1). Per EPA policy, in the absence of
a biologically relevant benchmark response level (BMR), a default BMR of 1 SD above the
control mean is used.
26
Benzenethiol
-------
FINAL
4-1-2011
Table 13. Absolute Kidney Weights (mg) in the F0 Male Rats Following
16-Week Oral Exposure to Benzenethiol to be Used for BMD Analysis"
DOSE
(mg/kg-day)
DOSEadj
(mg/kg-day)
Number of Subjects
Responseb
0
0
20
4390.1 ±581
9
9
10
5920.2 ± 683*
18
18
10
6719.7± 816*
35
35
9
7717.6 ± 1758*
aNTP (1996).
bMeans ± SD. Standard deviation was calculated from standard error x Vn
Statistically significantly different from control (p < 0.05) by pair-wise comparison
Table 14. Relative Kidney Weights (mg/g Body Weight) in the F0 Male Rats Following
16-Week Oral Exposure to Benzenethiol to be Used for BMD Analysis"
DOSE
(mg/kg-day)
DOSEadj
(mg/kg-day)
Number of Subjects
Responseb
0
0
20
5.7 ±0.63
9
9
10
7.4 ± 1.26*
18
18
10
8.7 ± 1.17*
35
35
9
11.6 ±2.22*
aNTP (1996).
bMeans ± SD. Standard deviation was calculated from standard error / Vn
Statistically significantly different from control (p < 0.05) by pair-wise comparison
When data for both absolute and relative kidney weights in the F0 males are modeled
using constant variance, all four continuous variable models indicate poor variance and wrong
variance models. The results from modeling of absolute and relative kidney weight data for the
F0 males using nonconstant variance are presented in Table 15.
27
Benzenethiol
-------
FINAL
4-1-2011
Table 15. Model Predictions for Increases in Absolute and Relative Kidney Weights in the
Male F0 Rats Exposed Orally to Benzenethiol for 16 weeks"
Model Name
Homogeneity
Variance
p-Value
Goodness-of-Fit
/;-Valueh
AIC for
Fitted Model
BMDisd
(mg/kg-day)
BMDLisd
(mg/kg-day)
Absolute Kidney Weights
Hill0
<0.0001
0.660
712.48
3.00
-999.00
Linear
<0.0001
0.136
714.28
4.83
3.52
Polynomial13
<0.0001
<0.0001
906.57
-999.00
15.68
Power
<0.0001
0.136
714.28
4.83
3.52
Model Name
Variance Model
p-Y alueb
Goodness-of-Fit
p-Y alueb
AIC for
Fitted Model
BMDisd
(mg/kg-day)
BMDLisd
(mg/kg-day)
Relative Kidney Weights
Hill6
<0.0001
0.408
60.19
3.46
2.22
Linear
<0.0001
0.634
58.41
3.84
2.91
Polynomial
<0.0001
0.634
58.41
3.84
2.91
Power
<0.0001
0.634
58.41
3.84
2.91
aNTP (1996).
bValues <0.10 fail to meet conventional goodness-of-fit criteria.
Invalid BMDL; hit bound (n = 1)
dInvalid BMD; p-value 4 < 0.1 (i.e., fails p-value criteria for goodness of fit)
"Lowest BMDL; hit bound (n = 1)
AIC = Akaike's Information Criteria; BMD = benchmark dose; BMDL = lower confidence limit (95%) on the
benchmark dose.
For absolute kidney weight data in the F0 males, the Hill model with nonconstant
variance results in an invalid BMDL, and the polynomial model fails thep-walue criteria for
goodness of fit. The linear and power models both adequately fit the data with identical results,
indicating that the BMDL of 3.52 mg/kg-day associated with the simpler linear model best
describes the data. In fact, the modeling output and graph indicate that the power model reverts
to a linear function.
For relative kidney weight data in the F0 males, nonconstant variance models for the
linear, polynomial, and power models all result in the same values, again with the linear model
best describing the data with BMDL of 2.91 mg/kg-day. Visual inspection of each graph reveals
linear outputs for the linear, polynomial, and power models. The scaled residuals for all of the
nonconstant variance models for relative kidney in the F0 males are < 2.0. Although the Hill
model results in the lowest BMDL at 2.22 mg/kg-day, the model parameter (n = 1) hits a bound
implied by some inequality constraint and thus has no standard error. Furthermore, because the
range of the BMDL values is < 3-fold, the estimated BMDL is considered sufficiently close, and
the BMDL with the lowest AIC value is selected for the POD. Therefore, the BMDL
28
Benzenethiol
-------
FINAL
4-1-2011
(2.91 mg/kg-day) from the linear nonconstant variance model of relative kidney weight data in
the F0 males is used as the POD in deriving a subchronic p-RfD.
The subchronic p-RfD for benzenethiol, based on the BMDLisd of 2.91 mg/kg-day for
increased relative kidney weight in the F0 male rat (NTP, 1996), is derived as follows:
Subchronic p-RfD = BMDLisd ^ UF
= 2.91 mg/kg-day -^300
= 0.0097 mg/kg-day or 1 x 10~2 mg/kg-day
Tables 16 and 17, respectively, summarize the UFs and the confidence descriptor for the
subchronic p-RfD for benzenethiol.
Table 16. Uncertainty Factors for Subchronic p-RfD of Benzenethiol"
UF
Value
Justification
ufa
10
A UFa of 10 is applied for interspecies extrapolation to account for potential toxicokinetic
and toxicodynamic differences between rats and humans. There are no data to determine
whether humans are more or less sensitive than rats to nephrotoxicity of benzenethiol.
ufd
3
A UFd of 3 is selected because, although the database includes by this route one
acceptable two-generation reproduction study in rats (NTP, 1996), one acceptable
developmental study in rats (NTP, 1994a), and one acceptable developmental study in
rabbits (NTP, 1994b), it is lacking a comprehensive general toxicity study.
UFh
10
A UFh of 10 for is applied for intraspecies differences to account for potentially
susceptible individuals in the absence of information on the variability of response in
humans.
ufl
1
A UFl of 1 is applied because the POD was developed using a BMDL.
UFS
1
A UFS of 1 is applied because the principal study (NTP, 1996) is a reproduction study in
which the F0 parents were dosed for 16 weeks (comparable to atypical 13-week
subchronic study). Furthermore, the endpoints utilized for the p-RfD are based on
findings in the F0 generation.
UFC < 3000
300
"Source: NTP (1996).
29
Benzenethiol
-------
FINAL
4-1-2011
Table 17. Confidence Descriptor for Subchronic p-RfD for Benzenethiol
Confidence
Categories
Designation"
Discussion
Confidence in study
M
Confidence in the key study is medium. NTP (1996)
assessed comprehensive endpoints in an appropriate
number of animals for a two-generation reproduction
study, and the duration of exposure for the parental
generation was appropriate for assessing subchronic
toxicity. However, a full complement of organs was not
examined microscopically, and clinical chemistry and
hematology measurements were not conducted. The
study included multiple effect levels, but a NOAEL is not
identified. The key study is supported by high quality
developmental toxicity studies in rats and rabbits
conducted by NTP (1994a,b). The critical effect and
subchronic p-RfD is supported by the presence of a
dose-response relationship.
Confidence in
database
M
The database includes acceptable developmental toxicity
studies in two species (rabbits and rats) and an acceptable
two-generation reproduction study in rats.
Confidence in
subchronic p-RfDb
M
The overall confidence in the subchronic p-RfD is
medium.
aL = Low, M = Medium, H = High.
bThe overall confidence cannot be greater than lowest entry in table.
Derivation of Chronic p-RfD
No chronic studies are available for the derivation of a chronic p-RfD. The available oral
studies are a developmental toxicity study in rats (NTP, 1994a), a developmental toxicity study
in rabbits (NTP, 1994b), and a two-generation reproduction study in rats (NTP, 1996). The same
study that was used for the derivation for the subchronic p-RfD (NTP, 1996) is used to derive the
chronic p-RfD.
Therefore, the chronic p-RfD for benzenethiol, based on the BMDLisd of 2.91 mg/kg-day
for increased kidney weights in the F0 male rat (NTP, 1996), is derived as follows:
Chronic p-RfD = BMDLisd UF
= 2.91 mg/kg-day ^ 3000
= 0.00097 mg/kg-day or 1 x 10~3 mg/kg-day
Tables 18 and 19, respectively, summarize the UFs and the confidence descriptor for the chronic
p-RfD for benzenethiol.
30
Benzenethiol
-------
FINAL
4-1-2011
Table 18. Uncertainty Factors for Chronic p-RfD of Benzenethiol"
UF
Value
Justification
ufa
10
A UFa of 10 is applied for interspecies extrapolation to account for potential
toxicokinetic and toxicodynamic differences between rats and humans. There are no
data to determine whether humans are more or less sensitive than rats to nephrotoxicity
of benzenethiol.
ufd
3
A UFd of 3 is applied because although the database includes one two-generation
reproduction study in rats (NTP, 1996), one developmental study in rats (NTP, 1994a),
and one developmental study in rabbits (NTP, 1994b), it is lacking a comprehensive
general toxicity study.
UFh
10
A UFh of 10 for is applied for intraspecies differences to account for potentially
susceptible individuals in the absence of information on the variability of response in
humans.
ufl
1
A UFl of 1 is applied because the POD was developed using a BMDL.
UFS
10
A UFS of 10 is applied for extrapolation of subchronic data to chronic data. The
principal study (NTP, 1996) is a reproduction study in which the F0 parents were dosed
for 16 weeks (comparable to a typical 13-week subchronic study). It is the only longer
duration study available, and the study did not evaluate all of the typical toxicity
endpoints.
UFC
<3000
3000
"Source: NTP (1996).
31
Benzenethiol
-------
FINAL
4-1-2011
Table 19. Confidence Descriptor for Chronic p-RfD for Benzenethiol
Confidence
Categories
Designation"
Discussion
Confidence in study
M
Confidence in the key study is medium. NTP (1996) assessed
comprehensive endpoints in an appropriate number of animals
for a two-generation reproduction study, and the duration of
exposure for the parental generation was appropriate for
assessing subchronic toxicity. However, a full complement of
organs was not examined microscopically, and clinical
chemistry and hematology measurements were not conducted.
The study included multiple effect levels, but a NOAEL is not
identified. The key study is supported by high quality
developmental toxicity studies in rats and rabbits conducted by
NTP (1994a,b). The critical effect and the chronic p-RfD is
supported further by the presence of a dose-response
relationship.
Confidence in database
M
The database includes developmental toxicity studies in
two species (rabbits and rats) and a two-generation
reproduction study. The database does lack a true chronic
study.
Confidence in chronic
p-RfDb
M
The overall confidence in the chronic p-RfD is medium.
aL = Low, M = Medium, H = High.
bThe overall confidence cannot be greater than lowest entry in table.
DERIVATION OF INHALATION REFERENCE CONCENTRATIONS
Derivation of Subchronic and Chronic p-RfC
No studies are available for the derivation of a subchronic or chronic p-RfC for
benzenethiol. The only inhalation studies found are short-term lethality studies in rats and mice.
Route-to-route extrapolation from oral to inhalation was not considered because suitable
physiologically-based pharmacokinetic models are not available.
DERIVATION OF PROVISIONAL CANCER VALUES
Cancer Weight-of-Evidence Descriptor
Table 20 provides a cancer weight-of-evidence descriptor of "inadequate information to
assess carcinogenic potential" for benzenethiol due to the lack of chronic toxicity or
carcinogenicity data.
32
Benzenethiol
-------
FINAL
4-1-2011
Table 20. Cancer WOE Descriptor for Benzenethiol
Possible WOE
Descriptor
Designation3
Route of Entry
(Oral, Inhalation,
or Both)
Comments
"Carcinogenic to
Humans "
N/A
N/A
No human cancer studies are available.
"Likely to be
Carcinogenic to
Humans "
N/A
N/A
There is no adequate evidence of plausible
association between human exposure and
cancer. Positive tumors were observed in
rats and mice, indicating suggestive
evidence.
"Suggestive of Evidence
of Carcinogenic
Potential"
N/A
N/A
There is no evidence from human and
animal studies that is suggestive of
carcinogenicity.
"Inadequate
Information to Assess
Carcinogenic
Potential"
X
Both
Under the 2005 Guidelines for
Carcinogenic Risk Assessment (U.S. EPA,
2005), the available evidence from
exposure to benzenethiol is inadequate to
assess carcinogenic potential.
"Not Likely to be
Carcinogenic to
Humans "
N/A
N/A
No strong evidence of noncarcinogenicity in
humans is available.
aThe designation N/A means not available, and X indicates the assigned cancer WOE descriptor.
Derivation of p-OSF
No human or animal studies examining the carcinogenicity of benzenethiol following
oral exposure have been located. Therefore, derivation of a p-OSF is precluded.
Derivation of p-IUR
No human or animal studies examining the carcinogenicity of benzenethiol following
inhalation exposure have been located. Therefore, derivation of a p-IUR is precluded.
33
Benzenethiol
-------
FINAL
4-1-2011
APPENDIX A. PROVISIONAL SCREENING VALUES
No screening values are presented.
34
Benzenethiol
-------
FINAL
4-1-2011
APPENDIX B. DATA TABLES
Table B.l. Maternal Body-Weight Changes, Food Consumption, Water Consumption, and
Liver Weights of Rats Exposed to Benzenethiol via Gavage from GDs 6-15a
Exposure Group (mg/kg-day)
Parameter/Interval
0
20
35
50
Body-weight change (g)
Treatment GDs 6-9
14.4 ± 1.0b
9.9 ± 0.9* (J.31)
7.1 ± 1.1* (J.88)
-0.3 ± 1.8* (4.102)
GDs 9-12
16.3 ± 1.3
19.1 ±0.9
16.4 ± 1.3
12.3 ± 1.7*0.25)
GDs12-15
21.1± 1.1
23.7 ± 1.2
22.2 ± 1.7
22.8 ± 1.5 (|8)
Overall treatment (GDs 6-15)
52.0 ±2.1
52.7 ± 1.9
45.7 ± 1.9
34.8 ±3.2* (|33)
Gestation (GDs 0-20)
153.7 ±4.3
153.1 ±3.9
148.5 ±3.6
122.6 ±5.9*(|20)
Gravid uterine weight (g)
92.7 ±2.1
86.9 ±3.4
89.4 ±2.1
72.0 ±5.5* (|22)
Corrected weight gain0
61.0 ±3.6
66.3 ±2.2
59.1 ±2.5
50.7 ±4.3* (|17)
Food consumption
Treatment (GDs 6-15)
Absolute (g/day)
22.2 ±0.5
21.3 ±0.4
21.1 ± 0.4
18.2 ±0.7* (|18)
Relative (g/kg-day)
71.6 ±0.9
69.3 ±0.8
68.9 ±1.1
61.9 ±2.0* (|14)
Posttreatment (GDs 15-20)
Absolute (g/day)
27.4 ±0.6
27.3 ±0.6
28.2 ±0.6
28.5 ±0.6
Relative (g/kg-day)
73.3 ± 1.3
73.1 ± 1.1
75.9 ± 1.3
81.5 ± 1.1*011)
Water consumption
Treatment (GDs 6-15)
Absolute (g/day)
37.6 ± 1.2
37.9 ±1.1
39.9 ± 1.2
45.1 ±3.0*020)
Relative (g/kg-day)
122.2 ±3.4
122.7 ±3.2
129.9 ±3.9
152.1 ±9.1* 024)
Posttreatment (GDs 15-20)
Absolute (g/day)
47.5 ± 1.5
47.3 ± 1.0
50.2 ± 1.4
59.0 ±3.0* 024)
Relative (g/kg-day)
127.0 ±3.9
126.2 ±2.4
135.4 ±4.0
168.6 ±8.2* 033)
Liver weights
Absolute (g)
17.3 ±0.3
18.3 ±0.3
18.5 ±0.3
19.2 ±0.4
Relative (% body weight)
4.4 ±0.1
4.6 ±0.1
4.6 ±0.1
5.2 ±0.1* 018)
(% adjusted weight)
5.8 ±0.1
5.8 ±0.1
6.0 ±0.1
6.4 ±0.1* (|10)
aNTP (1994a). Data were obtained from Table 3 on page 20 and Table A1-3 on page 37 of the study report.
'Means ± SE, () = percent change compared to control.
cWeight change during gestation minus gravid uterine weight.
*Significantly different from control (p < 0.05), Williams', and/or Dunnett's test.
35
Benzenethiol
-------
FINAL
4-1-2011
Table B.2. Cesarean Section and Fetal Examination Data from Time-mated Female Rats
Exposed to Benzenethiol via Gavage from GDs 6-15a
Parameter/Interval
Exposure Group (mg/kg-day)
0
20
35
50
All litters
Resorptions/litter (%)
1.5 ± 0.6b
1.0 ±0.6
2.3 ±0.9
15.5 ±5.5*
Litters with resorptions (%)
24
12
29
52*
Live litters
Number of live fetuses
15.8 ±0.4
15.1 ±0.6
15.7 ±0.4
12.6 ± 1.0* (420)
Fetal body weights
Males
3.84 ±0.06
3.84 ±0.10
3.73 ±0.06
345 ±0.11* (|10)
Females
3.70 ±0.05
3.57 ±0.06
3.51 ± 0.05* (|5)
3.34 ±0.09* (410)
Fetal external malformations
No. fetuses (litters) examined
394
378
376
265
No. fetuses (litters) with
external malformations
1(1)
0(0)
0(0)
5(4)
Percent externally malformed
fetuses (litters)
0.3 (4.0)
0.0 (0.0)
0.0 (0.0)
1.9(19.0)
Number fetuses (litters) with
Anophthalmia, right
—
—
—
1(1)
Open eye, left
—
—
—
1(1)
Anasarca
—
—
—
4(3)
Gastroschisis
—
—
—
1(1)
Micromelia
—
—
—
2(2)
Syndactyly, hindpaw
—
—
—
1(1)
Syndactyly, forepaw
—
—
—
1(1)
aNTP (1994a). Data were obtained from Tables 4 and 5 on pages 21 and 22 of the study report.
bMeans ± SE, () = percent change compared to control.
*Significantly different from control (p < 0.05), Williams', and/or Dunnett's test.
36
Benzenethiol
-------
FINAL
4-1-2011
Table B.3. Maternal Food Consumption and Body-Weight Changes
in New Zealand White Rabbits Exposed to Benzenethiol via Gavage from GDs 6-19a
Parameter/Interval
Exposure Group (mg/kg-day)
0
10
30
40
Food consumption1"
Pretreatment (GDs 0-5)
Absolute (g/day)
185 ±6.7
193 ±7.0
197 ±6.5
197 ± 10.0
Relative (g/kg-day)
52.5 ± 1.7
54.0 ±2.2
55.1 ±2.0
51.2 ±2.0
Treatment (GDs 6-19)
Absolute (g/day)
163 ±7.5
157 ±7.5
139 ±9.4
139 ±8.7
Relative (g/kg-day)
43.1 ± 1.7
41.3 ±2.0
36.6 ±2.3 (|15)
34.8 ±2.1 (419)°
Posttreatment (GDs 20-30)
Absolute (g/day)
143 ±8.8
154 ±9.3
140 ± 9.7
156 ± 12.8
Relative (g/kg-day)
36.7 ±2.1
38.6 ±2.2
35.1 ±2.2
37.6 ±2.3
Body-weight change (g)
Pretreatment (GDs 0-6)
189 ±27
210 ±33
220 ± 22
183 ±37
Treatment (GDs 6-19)
216 ±32
185 ±31 (|14)
152 ± 37 (430)
84 ± 58 (461)c
Treatment (GDs 12-15)
96 ± 14
47 ±21
22 ± 22* (477)
8 ± 24* (492)°
Posttreatment (GDs 19-30)
136 ±28
208 ± 27
114 ±35
223 ± 26
Overall (GDs 0-30)
541 ±49
603 ±61
486 ± 56
491 ±56
Gravid uterine weight (g)
476.84 ± 29.25
465.05 ±49.86
469.58 ±43.00
542.14 ±25.78
Corrected weight change
61.9 ±60.6
137.7 ±87.7
15.9 ±66.8
-51.0 ±78.7
aNTP (1994b). Data were obtained from Table 4 on page 21 and Table Al-3 on page 38 of the study report.
'Means ± SE, () = percent change compared to control.
Significant linear trend.
*Significantly different from control (p < 0.05), Williams', and/or Dunnett's test.
37
Benzenethiol
-------
FINAL
4-1-2011
Table B.4. Parental Male Body Weights in S-D Rats Exposed
to Benzenethiol via Gavagea
Time
Point/Interval
Exposure Group (mg/kg-day)
0
9
18
35
F0 Generation body weights at delivery of each litter (g)
Litter 1
512.83 ±9.61b
494.96 ±7.30
486.75 ± 9.38* (|5)
476.01 ± 10.79* (|7)c
Litter 2
584.48 ± 12.36
570.38 ± 10.50
559.32 ±9.99
532.97 ±9.77* (|9)c
Litter 3
644.46 ± 14.29
618.33 ± 12.49
612.23 ± 13.02
571.48 ± 13.50* (ill)0
Litter 4
696.85 ± 18.36
659.40 ± 18.72
647.99 ± 15.20
589.78 ± 13.67* (|15)c
Litter 5
695.42 ±25.80
660.65 ± 17.39
681.23 ± 17.67
613.90 ± 15.94* (|12)°
F0 Body weights (g)
Week 1
398.7 ±3.5
382.6 ±7.4
381.4 ±6.3*(|4)
368.7 ± 6.8* (|8)c
Week 6
528.8 ±9.0
512.0 ±6.5
501.0 ±9.1*(|5)
487.9 ± 8.3* (|8)c
Week 12
627.4 ± 12.9
619.2 ± 10.0
613.2 ± 10.8
563.1 ± 10.3* (|10)°
Week 18
690.9 ± 15.1
686.3 ± 14.5
666.1 ± 13.4
603.2 ± 11.7* (|13)°
F1 Generation body weights at delivery of litter (g)
Litter 1
595.0 ± 18.7
593.8 ± 13.8
570.1 ± 12.5
511.8 ± 16.2* (|14)°
F1 Body weights (g)
Week 2
523.0 ± 13.0
523.7 ± 11.7
511.1 ± 7.9
462.9 ± 9.6* (ill)0
Week 4
570.3 ± 13.4
572.4 ± 12.5
548.5 ±7.8
498.5 ± 8.9* (il3)°
aNTP (1996). Data were obtained from Tables 2-4,2-5,4-2, and 4-3 on pages 50, 51, 71, and 72 of the study report.
'Means ± SE, () = percent change compared to control.
Significant linear trend.
*Significantly different from control (p < 0.05) by pair-wise comparison.
38
Benzenethiol
-------
FINAL
4-1-2011
Table B.5. Parental Liver Weights in
S-D Rats Exposed
to Benzenethiol via Gavagea
Exposure Group (mg/kg-day)
Parameter
0
9
18
35
F0 Males
Terminal body weights (g)
777.3 ± 16.7b
808.4 ±32.7
777.3 ±33.9
666.0 ± 16.8* (|14)c
Liver
Absolute (g)
27.5 ±0.94
34.2 ± 1.8* (|24)
36.9 ± 1.9* (|34)
35.4 ± 1.7* (|29)0
Relative (mg/g bw)
35.3 ±0.91
42.2 ± 1.0* (|20)
47.7 ±2.2* (|35)
53.0 ± 1.8* (t50)°
F0 Females
Terminal body weights (g)
464.8 ± 11.2
441.8 ± 12.0
451.8 ± 16.1
428.8 ± 12.4
Liver
Absolute (g)
16.3 ±0.38
17.1 ±0.36 (|5)
18.5 ± 0.56* (T13)
20.3 ± 0.62* (t25)°
Relative (mg/g bw)
35.0 ±0.54
38.9 ± 0.91* (fll)
41.3 ± 1.3* (T18)
47.6 ± 1.6* (|36)°
F1 Males
Terminal body weights (g)
639.0 ± 17.1
670.4 ± 14.9
608.6 ±7.7
528.8 ± 13.2* (J. 17)°
Liver
Absolute (g)
23.8 ±0.84
29.4 ± 1.2* (|24)
31.0 ±0.78* (|30)
31.8 ± 1.0* (|34)°
Relative (mg/g bw)
37.2 ±0.86
43.9 ± 1.5* (f 18)
50.9 ±0.88* (|37)
60.2 ± 1.5* (|62)c
F1 Females
Terminal body weights (g)
371.2 ±8.2
358.0 ± 16.7
363.1 ±8.7
350.9 ± 15.9
Liver
Absolute (g)
14.3 ±0.41
15.6 ±0.83 (|9)
16.4 ± 0.78* (|15)
19.1 ± 0.66* (|34)°
Relative (mg/g bw)
38.6 ±0.79
43.7 ± 1.5* (f 13)
45.1 ± 2.1* (t 17)
54.9 ± 1.4* (|42)c
aNTP (1996). Data were obtained from Tables 3-3, 3-5, 4-5, and 4-7 on pages 59, 61, 75, and 77 and text table on page 34 of the
study report.
bMeans ± SE, () = percent change compared to control.
Significant linear trend.
*Significantly different from control (p <0.05) by pair-wise comparison.
39
Benzenethiol
-------
FINAL
4-1-2011
Table B.6. Incidences of Hepatocellular Hypertrophy in Sprague-Dawley Rats Exposed to
Benzenethiol via Oral Gavage for 2 Generations51
Generation/Sex
Exposure Group (mg/kg-day)
0
9
18
35
F0 Males Total
0
0
10
9
Minimal
—
—
4
—
Mild
—
—
5
5
Moderate
—
—
1
4
F0 Females Total
0
9
10
10
Minimal
—
6
2
...
Mild
—
3
5
3
Moderate
—
...
3
4
Marked
—
...
...
3
F1 Males Total
0
10
10
10
Minimal
—
4
2
...
Mild
—
5
3
—
Moderate
—
1
5
10
F1 Females Total
0
3
10
10
Minimal
—
3
6
—
Mild
—
—
4
3
Moderate
—
—
—
6
Marked
...
...
...
1
aNTP(1996). Number affected/10 examined. Data were obtained from Text Tables on pages 37 and 41 of the study report.
40
Benzenethiol
-------
FINAL
4-1-2011
Table B.7. Parental Kidney Weights in S-D Rats Exposed
to Benzenethiol via Gavagea
Parameter
Exposure Group (mg/kg-day)
0
9
18
35
F0 Males
Terminal body weights (g)
777.3 ± 16.7b
808.4 ±32.7
777.3 ±33.9
666.0 ± 16.8* (|14)°
Kidney
Absolute (mg)
4390.1 ± 130
5920.2 ±216* (|35)
6719.7 ±258* (|53)
7717.6 ±586* (|76)c
Relative (mg/g bw)
5.7 ±0.14
7.4 ± 0.40* (|30)
8.7 ± 0.37* (|53)
11.6 ± 0.74* (tl04)c
F0 Females
Terminal body weights (g)
464.8 ± 11.2
441.8 ± 12.0
451.8 ± 16.1
428.8 ± 12.4
Kidney
Absolute (mg)
2715.2 ±50.4
2803.6 ±62.4
2784.9 ±77.1
3033.0 ± 103* (|12)°
Relative (mg/g bw)
5.9 ± 0.11
6.4 ± 0.12* (|8)
6.2 ±0.26 (|5)
7.1 ± 0.31* (|20)c
F1 Males
Terminal body weights (g)
639.0 ± 17.1
670.4 ± 14.9
608.6 ±7.7
528.8 ± 13.2* (|17)°
Kidney
Absolute (mg)
3991.7 ± 110
6482.6 ± 588* (|62)
6407.6 ± 193* (|61)
8703.8 ± 1093* (tll8)c
Relative (mg/g bw)
6.3 ±0.07
9.6 ± 0.80* (|52)
10.5 ± 0.33* (|67)
16.6 ± 2.3* (tl63)°
F1 Females
Terminal body weights (g)
371.2 ±8.2
358.0 ± 16.7
363.1 ±8.7
350.9 ± 15.9
Kidney
Absolute (mg)
2446.7 ±71.1
2626.7 ±77.1
2550.9 ±83.0
2855.2 ±95.3* (|17)c
Relative (mg/g bw)
6.6 ±0.12
7.4 ± 0.28* (|12)
7.0 ± 0.21* (|6)
8.3 ± 0.42* (|26)c
aNTP (1996). Data were obtained from Tables 3-3, 3-5, 4-5, and 4-7 on pages 59, 61, 75, and 77 and text table on page 34 of the
study report.
bMeans ± SE, () = percent change compared to control.
Significant linear trend.
* Significantly different from control (p < 0.05) by pair-wise comparison.
41
Benzenethiol
-------
FINAL
4-1-2011
Table B.8. Incidences of Gross Findings in the Kidneys of S-D Rats Exposed to
Benzenethiol via Gavagea
Exposure Group (mg/kg-day)
Macroscopic finding
0
9
18
35
F0 Males
Number examined
20
10
10
10
Kidneys, enlarged
0
0
0
2
Pitted
1
0
1
4
F1 Males
Number examined
20
10
10
10
Kidneys, enlarged
0
2
1
4
Pale
0
9
10
9
Soft
0
0
0
2
aNTP (1996). Number affected. Data were obtained from Table 3-7 on page 63 and Table 4-9 on page 79 of the study report.
Table B.9. Incidences of Renal Tubule Degeneration in the Kidneys of S-D Rats Exposed to
Benzenethiol via Oral Gavage for 2 Generations51
Generation/Sex
Exposure Group (mg/kg-day)
0
9
18
35
F0 Males Total
5
10
10
10
Minimal
2
—
—
—
Mild
3
8
2
...
Moderate
—
2
8
6
Marked
—
...
...
4
F0 Females Total
1
2
3
4
Minimal
1
2
2
...
Mild
—
—
1
4
F1 Males Total
0
10
10
10
Minimal
—
—
1
—
Mild
—
6
5
1
Moderate
—
3
4
3
Marked
—
1
—
6
F1 Females Total
0
0
1
4
Minimal
—
...
1
...
Mild
—
...
...
3
Moderate
...
...
...
1
aNTP(1996). Number affected/10 examined. Data were obtained from Text Tables on pages 37 and 41 of the study report.
42
Benzenethiol
-------
FINAL
4-1-2011
Table B.10. Computer-Assisted Sperm Analysis of the Epididymis and Microscopic
Examination of Testis Sperm in S-D Rats Exposed to Benzenethiol via Gavagea
Parameter
Exposure Group (mg/kg-day)
0
9
18
35
F0 Males
Sperm motility (% motile)
89.2 ± 1.3b
88.1 ± 1.5
83.9 ± 2.5* (|6)
84.8 ± 1.9* (|5)°
Velocity (|im/scc)
198.9 ±4.0
187.4 ±4.6
183.2 ±7.2
186.5.5.0°
F1 Males
Sperm motility (% motile)
89.8 ±0.88
86.4 ±2.7
88.3 ± 0.74
85.8 ±2.2
Velocity (|im/scc)
210.1 ±6.1
206.8 ±6.6
210.0 ±6.1
196.1 ±7.8
Inhibited spermiation of the Stage
VIII-X tubules (# affected/10)
0
6
6
9
aNTP (1996). Data were obtained from text table on page 41, Table 3-4 on page 60, and Table 4-6 on page 76 of the study report.
Spermatid head count was determined from the right testis. Sperm density, morphology, and motion analyses (computer-
assisted) were evaluated from the right cauda epididymis. Sperm parameters included: motility, velocity (jim/sec); linearity;
ALH max (|im); ALH mean (|im); beat/cross frequency (Hz/sec); average radius (|im); circular cells; circular over motile cells
(%); circular over all cells (%); epididymal sperm density (1000 sperm/mg caudal tissue) and morphology (% abnormal);
spermatids/mg testis; and total spermatids/testis. Aside from sperm motility in the F0 gernation and inhibited spermiation in the
F1 males, none of these parameters were affected by treatment.
bMeans ± SE, () = percent change compared to control.
Significant (p < 0.05) linear trend.
*Significantly different from control (p < 0.05) by pair-wise comparison.
43
Benzenethiol
-------
FINAL
4-1-2011
Table B.ll. Selected Live Pup Body Weights of S-D Rats Exposed
to Benzenethiol via Gavagea
Exposure Group (mg/kg-day)
Time Point/Interval
0
9
18
35
F1 Pups
Absolute body weight (g)
Males
6.67 ± 0.08b
6.38 ±0.07* (4,4)
6.50 ±0.09
6.35 ±0.07* (|5)c
Females
6.36 ±0.09
6.06 ±0.09* (|5)
6.16 ±0.08
6.05 ±0.08* (|5)c
Combined
6.51 ±0.08
6.22 ±0.08* (|4)
6.33 ±0.09
6.20 ±0.07* (|5)c
Adjustedd
Males
6.68 ±0.08
6.40 ±0.08* (|4)
6.50 ±0.08
6.31 ±0.08* (|6)c
Females
6.37 ±0.08
6.09 ±0.08* (|4)
6.16 ±0.08
6.01 ±0.08* (|6)c
Combined
6.52 ±0.08
6.24 ±0.08* (|4)
6.33 ±0.08
6.16 ±0.08* (|6)c
F2 Pups
Absolute body weight (g)
Males
7.18 ±0.24
6.71 ±0.32
6.50 ±0.19
6.30 ± 0.20* (|12)°
Females
6.86 ±0.21
6.38 ±0.30
6.23 ± 0.16* (|9)
5.99 ± 0.15* (|13)c
Combined
6.99 ±0.21
6.53 ±0.30
6.35 ± 0.17* (|9)
6.16 ± 0.17* (|12)c
Adjustedd
Males
7.26 ±0.25
6.56 ±0.21
6.44 ±0.24
6.49 ±0.26
Females
6.92 ±0.24
6.28 ±0.20
6.20 ±0.22
6.12 ±0.24
Combined
7.06 ±0.23
6.41 ±0.19
6.31 ±0.22
6.32 ±0.24
aNTP (1996). Data were obtained from Table 2-2 on page 48 and Table 4-2 on page 71 of the study report.
bMeans ± SE, () = percent change compared to control.
Significant linear trend.
dLeast squares estimate of mean pup weight adjusted for average litter size ± SE (number of fertile pairs producing live pups).
*Significantly different from control (p < 0.05) by pair-wise comparison.
44
Benzenethiol
-------
FINAL
4-1-2011
APPENDIX C. BMD MODELING OUTPUTS FOR BENZENE THIOL
Hill Model
dose
15:42 01/28 2011
Figure C-l. Hill BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Hill BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Hill Model. (Version: 2.15; Date: 10/28/2009)
Input Data File:
C:/US EPA/BMDS 212/Allran2/NTP_199 6_F 0_M_Ab s_Ki dney_Wt_Hi11_1. (d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Hill_l.pit
Fri Jan 28 15:42:37 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = intercept + v*dose^n/(k^n + dose^n)
45
Benzenethiol
-------
FINAL
4-1-2011
Dependent variable = Mean
Independent variable = Dose
Power parameter restricted to be greater than 1
The variance is to be modeled as Var(i) = exp(lalpha + rho * ln(mean(i)))
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 13.7306
rho = 0
intercept = 4390.1
v = 3327.5
n = 0.0773128
k = 25.4955
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -n
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
lalpha rho intercept v k
lalpha 1 -1 -0.2 0.022 -0.037
rho -1 1 0.19 -0.023 0.037
intercept -0.2 0.19 1 0.12 0.23
v 0.022 -0.023 0.12 1 0.98
k -0.037 0.037 0.23 0.98 1
Parameter Estimates
Interval
Variable
Limit
lalpha
0.274303
rho
4.86448
intercept
4641.36
v
13691.6
n
k
92.0888
Estimate
-14.4795
3.21979
4408.97
7140.96
1
37.5881
Std. Err.
7.24771
0.839143
118.57
3342.24
NA
27.807
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-28.6848
1.5751
4176.57
590.281
-16.9126
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
46
Benzenethiol
-------
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-351.141284
-351.238144
-385.485941
# Param's
5
8
6
5
2
AIC
727.626149
714.534141
714.282567
712.476288
774.971882
FINAL
4-1-2011
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 4.39e+003
9 10 5.92e+003
18 10 6.72e+003
35 9 7.72e+003
4.41e+003
5 .79e+003
6.72e+003
7.85e+003
581
683
816
1. 76e+003
528
819
1. 04e+003
1.34e+003
-0.16
0.509
-0.00481
-0.302
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (A1 vs A2)
Test 3: Are variances adeguately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
Test 1
Test 2
Test 3
Test 4
-2*log(Likelihood Ratio) Test df
72.4377
19.092
3.74843
0.193721
p-value
<.0001
0.0002617
0.1535
0.6598
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
47
Benzenethiol
-------
FINAL
4-1-2011
The p-value for Test 2 is less than . 1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect
1
Risk Type
Estimated standard deviations from the control mean
Confidence level
0. 95
BMD
3.00245
BMDL computation failed.
48
Benzenethiol
-------
FINAL
4-1-2011
Linear Model with 0.95 Confidence Level
a)
CO
c
o
Q.
CO
a)
Q1
c
ro
a)
9000
8000
7000
6000
5000
4000
0 5
15:42 01/28 2011
Figure C-2. Linear BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Linear BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Linear_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Linear_l.pit
Fri Jan 28 15:42:44 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*dose/s2 +
Dependent variable = Mean
Independent variable = Dose
Signs of the polynomial coefficients are not restricted
49
Benzenethiol
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 13.7306
rho = 0
beta_0 = 4779.05
beta 1 = 90.829
Asymptotic Correlation Matrix of Parameter Estimates
lalpha rho beta_0 beta_l
lalpha 1 -1 0.24 -0.39
rho -1 1 -0.24 0.39
beta_0 0.24 -0.24 1 -0.53
beta 1 -0.39 0.39 -0.53 1
Parameter Estimates
Interval
Variable
Limit
lalpha
rho
beta_0
beta 1
1.16134
5.06819
4718.17
140.163
Estimate
-15.7807
3.37767
4481.6
113.816
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
7.45899 -30.4
0.862526 1.68715
120.702 4245.03
13.4425 87.4689
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 4.39e+003
9 10 5.92e+003
18 10 6.72e+003
35 9 7.72e+003
4.48e+003
5.51e+003
6.5 3e+0 03
8.47e+003
581
683
816
1. 76e+003
549
778
1. 04e+003
1.61e+003
-0.745
1.68
0.577
-1.39
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
50
Benzenethiol
-------
FINAL
4-1-2011
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-351.141284
-353.138785
-385.485941
# Param's
5
8
6
4
2
AIC
727.626149
714.534141
714.282567
714.277570
774.971882
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
3.74843
3. 995
<.0001
0.0002617
0.1535
0.1357
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
51
Benzenethiol
-------
FINAL
4-1-2011
BMD = 4.82763
BMDL = 3.52106
52
Benzenethiol
-------
FINAL
4-1-2011
Polynomial Model with 0.95 Confidence Level
8000
6000
4000
2000
°BI7IDL
Polynomial
BMl
0 500 1000 1500 2000 2500 3000 3500
dose
15:42 01/28 2011
Figure C-3. Poly3 BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Poly3 BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Poly3_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Poly3_l.pit
Fri Jan 28 15:42:44 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
The polynomial coefficients are restricted to be positive
53
Benzenethiol
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 13.7306
rho =
beta_0 =
beta_l =
beta_2 =
beta 3 =
0
4390.1
226.11
0
0. 0957401
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_2
have been estimated at a boundary point, or have been specified by
the user,
lalpha
rho
beta_0
beta_l
beta 3
and do not appear in the correlation matrix )
lalpha
1
0. 99
NA
NA
0.0065
rho
0.99
1
NA
NA
0.0064
beta_0
NA
NA
NA
NA
NA
beta_l
NA
NA
NA
NA
NA
beta_3
0.0065
0.0064
NA
NA
1
Interval
Variable
Limit
lalpha
rho
beta_0
beta 1
NA
NA
NA
NA
NA
beta_2
beta 3
Estimate
20.4374
0.431539
0. 000213826
6. 927 66e-005
6. 49404e-030
1.12497e-008
Parameter Estimates
Std. Err.
NA
NA
NA
NA
NA
NA
At least some variance estimates are negative.
THIS USUALLY MEANS THE MODEL HAS NOT CONVERGED!
Try again from another starting point.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
NA
NA
NA
NA
NA
54
Benzenethiol
-------
FINAL
4-1-2011
Table of Data and Estimated Values of Interest
Dose
Obs Mean
Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 4.39e+003 0.000214 581 4.43e+003
9 10 5.92e+003 0.000846 683 5.96e+003
18 10 6.72e+003 0.00153 816 6.76e+003
35 9 7.72e+003 0.00312 1.76e+003 7.89e+003
4.44
3.14
3.14
2.93
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-351.141284
-448.286301
-385.485941
# Param's
5
8
6
5
2
AIC
727.626149
714.534141
714.282567
906.572601
774.971882
Test 1:
Test
Test
Test
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
3.74843
194.29
<.0001
0.0002617
0.1535
<.0001
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
model appears to be appropriate
A non-homogeneous variance
55
Benzenethiol
-------
FINAL
4-1-2011
The p-value for Test 3 is greater than . 1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is less than .1. You may want to try a different
model
BMD computation failed for BMR = 4426.54
Setting BMD = 100*(maximum dose)
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = -9999
BMDL = 15.68 02
56
Benzenethiol
-------
FINAL
4-1-2011
Power Model with 0.95 Confidence Level
9000
8000
CD
(f)
% 7000
Q_
in
CD
a:
c
CO
¦i 6000
5000
4000
0 5 10 15 20 25 30 35
dose
15:42 01/28 2011
Figure C-4. Power BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Power BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Power Model. (Version: 2.16; Date: 10/28/2009)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Power_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_Power_l.pit
Fri Jan 28 15:42:45 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = control + slope * dose^power
Dependent variable = Mean
Independent variable = Dose
The power is restricted to be greater than or egual to 1
57
Benzenethiol
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 13.7306
rho = 0
control = 4390.1
slope = 435
power = 0.572274
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -power
have been estimated at a boundary point, or have been specified by
the user,
lalpha
rho
control
slope
and do not appear in the correlation matrix )
lalpha
1
-1
-0.062
0.039
rho
-1
1
0. 056
-0.045
control
-0.062
0.056
1
-0.51
slope
0. 039
-0. 045
-0.51
1
Interval
Variable
Limit
lalpha
2 .3158
rho
4.93453
control
4716.15
slope
139.198
power
Estimate
-15.7807
3.37767
4481.6
113. 816
1
Parameter Estimates
Std. Err.
6.86996
0.794332
119.67
12.9506
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-29.2456
1.82081
4247.05
88.433
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
58
Benzenethiol
-------
FINAL
4-1-2011
0
20
4.39e+003
9
10
5.92e+003
18
10
6.72e+003
35
9
7.72e+003
.48e+003 581
.51e+003 683
.53e+003 816
.47e+003 1.76e+003
549 -0.745
778 1.68
1.04e+003 0.577
1.61e+003 -1.39
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/S2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-351.141284
-353.138785
-385.485941
# Param's
5
8
6
4
2
AIC
727.626149
714.534141
714.282567
714.277570
774.971882
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
3.74843
3. 995
<.0001
0.0002617
0.1535
0.1357
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
59
Benzenethiol
-------
FINAL
4-1-2011
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 4.82763
BMDL = 3.5210 6
60
Benzenethiol
-------
0 5 10 15 20 25 30 35
dose
15:49 01/28 2011
Figure C-5. Hill BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Hill BMD Model for Relative Kidney Weight Data (NTP, 1996)
FINAL
4-1-2011
Hill Model with 0.95 Confidence Level
Hill Model. (Version: 2.15; Date: 10/28/2009)
Input Data File:
C:/US EPA/BMDS 212/Allran2/NTP_199 6_F 0_M_Rel_Ki dney_Wt_Hi11_1. (d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_Hill_l.pit
Fri Jan 28 15:49:30 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = intercept + v*dose^n/(k^n + dose^n)
Dependent variable = Mean
Independent variable = Dose
Power parameter restricted to be greater than 1
61
Benzenethiol
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + rho * In(mean(i)))
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 0.491931
rho = 0
intercept = 5.7
v = 5.9
n = 0.634978
k = 18.3462
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -n
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
lalpha rho intercept v k
lalpha 1 -0.99 -0.16 0.24 0.22
rho -0.99 1 0.14 -0.24 -0.23
intercept -0.16 0.14 1 0.22 0.24
v 0.24 -0.24 0.22 1 1
k 0.22 -0.23 0.24 1 1
Parameter Estimates
Interval
Variable
Limit
lalpha
3.33313
rho
5.03282
intercept
5.98033
v
232.798
n
k
1335.83
Estimate
-6.71853
3.35383
5.70044
45.8435
1
243.146
Std. Err.
1.72728
0.856645
0.142801
95.3867
NA
557.504
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-10.1039
1.67483
5.42055
-141.111
-849.542
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
62
Benzenethiol
-------
FINAL
4-1-2011
Table of Data and Estimated Values of Interest
Dose
Obs Mean
Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
9
18
35
20
10
10
9
5.7
7.4
8.7
11. 6
5.7
7.34
8.86
11.5
0. 62 6
1.26
1.17
2.22
0.644
0.983
1.35
2.08
-0.00306
0.203
-0.376
0.189
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-24.749829
-25.092897
-68.998702
# Param's
5
8
6
5
2
AIC
78.931884
64.272980
61. 499659
60.185795
141.997404
Test 1:
Test
Test
Test
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
89.7244
20.6589
1.22668
0.686136
<.0001
0.0001239
0.5415
0.4075
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
model appears to be appropriate
A non-homogeneous variance
63
Benzenethiol
-------
FINAL
4-1-2011
The p-value for Test 3 is greater than . 1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 3.4 62 68
BMDL = 2.21795
64
Benzenethiol
-------
0 5 10 15 20 25 30 35
dose
15:49 01/28 2011
Figure C-6. Linear BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Linear BMD Model for Relative Kidney Weight Data (NTP, 1996)
FINAL
4-1-2011
Linear Model with 0.95 Confidence Level
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_Linear_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_Linear_l.pit
Fri Jan 28 15:49:31 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
Signs of the polynomial coefficients are not restricted
65
Benzenethiol
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 0.491931
rho = 0
beta_0 = 5.76667
beta 1 = 0.166667
Asymptotic Correlation Matrix of Parameter Estimates
lalpha rho beta_0 beta_l
lalpha 1 -0.99 0.011 -0.031
rho -0.99 1 -0.012 0.032
beta_0 0.011 -0.012 1 -0.46
beta 1 -0.031 0.032 -0.46 1
Parameter Estimates
Interval
Variable
Limit
lalpha
3.4225
rho
4.81955
beta_0
5.99222
beta_l
0.201156
Estimate
-6.55592
3.27401
5.71843
0.170536
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
1.59871 -9.68933
0.788556 1.72847
0.139691 5.44464
0.0156231 0.139915
Table of Data and Estimated Values of Interest
Dose
Obs Mean
Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
9
18
35
20
10
10
9
5.7
7.4
8.7
11.6
5.72
7.25
8.79
11.7
0. 62 6
1.26
1.17
2.22
0.655
0.966
1.32
2.11
-0.126
0.48
-0.211
-0.124
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
66
Benzenethiol
-------
FINAL
4-1-2011
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-24.749829
-25.204867
-68.998702
# Param's
5
8
6
4
2
AIC
78.931884
64.272980
61. 499659
58.409735
141. 997404
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
89.7244
20.6589
1.22668
0.910076
<.0001
0.0001239
0.5415
0.6344
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
67
Benzenethiol
-------
FINAL
4-1-2011
BMD = 3.8393
BMDL = 2.90725
68
Benzenethiol
-------
0 5 10 15 20 25 30 35
dose
15:49 01/28 2011
Figure C-7. Poly3 BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Poly3 BMD Model for Relative Kidney Weight Data (NTP, 1996)
FINAL
4-1-2011
Polynomial Model with 0.95 Confidence Level
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_Poly3_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_Poly3_l.pit
Fri Jan 28 15:49:31 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
The polynomial coefficients are restricted to be positive
69
Benzenethiol
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 0.491931
rho =
beta_0 =
beta_l =
beta 2 =
0
5.7
0.227194
0
beta 3 = 9.92762e-005
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_2 -beta_3
have been estimated at a boundary point, or have been specified by
the user,
lalpha
rho
beta_0
beta 1
and do not appear in the correlation matrix )
lalpha
1
-0. 99
0.011
-0.031
rho
-0.99
1
-0.012
0. 032
beta_0
0.011
-0.012
1
-0.46
beta_l
-0.031
0. 032
-0.46
1
Interval
Variable
Limit
lalpha
3.4225
rho
4.81955
beta_0
5.99222
beta_l
0.201156
beta_2
beta 3
Estimate
-6.55592
3.27401
5.71843
0.170536
0
0
Parameter Estimates
Std. Err.
1.59871
0.788556
0.139691
0.0156231
NA
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-9.68933
1.72847
5.44464
0.139915
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
70
Benzenethiol
-------
FINAL
4-1-2011
o
9
18
35
20
10
10
9
5.7
7.4
8.7
11. 6
5.72
7.25
8.79
11.7
0. 62 6
1.26
1.17
2.22
0.655
0.966
1.32
2.11
-0.126
0.48
-0.211
-0.124
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-24.749829
-25.204867
-68.998702
# Param's
5
8
6
4
2
AIC
78.931884
64.272980
61. 499659
58.409735
141. 997404
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
89.7244
20.6589
1.22668
0.910076
<.0001
0.0001239
0.5415
0.6344
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
71
Benzenethiol
-------
FINAL
4-1-2011
The p-value for Test 4 is greater than . 1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 3.8393
BMDL = 2.90725
72
Benzenethiol
-------
FINAL
4-1-2011
Power Model with 0.95 Confidence Level
-------
FINAL
4-1-2011
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 0.491931
rho = 0
control = 5.7
slope = 0.227624
power = 0.915525
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -power
have been estimated at a boundary point, or have been specified by
the user,
lalpha
rho
control
slope
and do not appear in the correlation matrix )
lalpha
1
-0. 99
-0.18
0.22
rho
-0.99
1
0.16
-0.24
control
-0.18
0.16
1
-0. 47
slope
0.22
-0.24
-0.47
1
Interval
Variable
Limit
lalpha
3.3317
rho
4.86818
control
5.99308
slope
0.201264
power
Estimate
-6.55592
3.27401
5.71843
0.170536
1
Parameter Estimates
Std. Err.
1.64504
0.81337
0.140126
0.0156782
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-9.78013
1.67983
5.44379
0.139807
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
74
Benzenethiol
-------
FINAL
4-1-2011
0
20
5.7
9
10
7.4
18
10
8.7
35
9
11.6
5.72 0.626
7.25 1.26
8.79 1.17
11.7 2.22
0.655 -0.126
0.966 0.48
1.32 -0.211
2.11 -0.124
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-24.749829
-25.204867
-68.998702
# Param's
5
8
6
4
2
AIC
78.931884
64.272980
61. 499659
58.409735
141. 997404
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
89.7244
20.6589
1.22668
0.910076
<.0001
0.0001239
0.5415
0.6344
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
75
Benzenethiol
-------
FINAL
4-1-2011
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 3.8393
BMDL = 2.90725
76
Benzenethiol
-------
FINAL
4-1-2011
Hill Model with 0.95 Confidence Level
a)
CO
c
o
Q.
CO
a)
Q1
c
ro
a)
9000
8000
7000
6000
5000
4000
Hill
BMD
10 15 20 25 30 35
dose
0 5
15:42 01/28 2011
Figure C-9. Hill CV BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Hill CV BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Hill Model. (Version: 2.15; Date: 10/28/2009)
Input Data File:
C:/US E PA/BMDS212/A11r an2/NT P_19 9 6_F 0_M_Ab s_Ki dne y_Wt_Hi11CV_1.(d)
Gnuplot Plotting File:
C:/US E PA/BMDS 212/A11r an2/NT P_19 9 6_F 0_M_Ab s_Ki dne y_Wt_Hi11CV_1.pit
Fri Jan 28 15:42:45 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = intercept + v*dose^n/(k^n + dose^n)
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
77
Benzenethiol
-------
FINAL
4-1-2011
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 918585
rho = 0 Specified
intercept = 4390.1
v = 3327.5
n = 0.0773128
k = 25.4955
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -n
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha intercept v k
alpha 1 6.3e-008 -1.2e-007 -7.3e-008
intercept 6.3e-008 1 0.097 0.33
v -1.2e-007 0.097 1 0.95
k -7.3e-008 0.33 0.95 1
the user,
Parameter Estimates
Interval
Variable
Limit
alpha
1.17852e+006
intercept
4795.2
v
9359.91
n
k
58.0422
Estimate
844225
4393.52
5713.26
1
25.4422
Std. Err.
170559
204.94
1860.57
NA
16.633
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
509935
3991.85
2066.6
-7.15783
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
78
Benzenethiol
-------
FINAL
4-1-2011
0
20
4.39e+003
4.39e+003
581
919
-0.0167
9
10
5.92e+003
5.89e+003
683
919
0.116
18
10
6.72e+003
6.7 6e+003
816
919
-0.141
35
9
7.72e+003
7.7e+003
1. 76e+003
919
0.0514
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-358.813074
-358.831281
-385.485941
# Param's
5
8
5
4
2
AIC
727.626149
714.534141
727.626149
725.662562
774.971882
Test 1:
Test
Test
Test
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
19.092
0.0364132
<.0001
0.0002617
0.0002617
0. 8487
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
non-homogeneous variance model
Consider running a
The p-value for Test 3 is less than .1.
different variance model
You may want to consider a
79
Benzenethiol
-------
FINAL
4-1-2011
The p-value for Test 4 is greater than . 1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 4.8758
BMDL = 2.7 62 4
80
Benzenethiol
-------
FINAL
4-1-2011
Linear Model with 0.95 Confidence Level
dose
15:42 01/28 2011
Figure C-10. LinearCV BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Linear CV BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_LinearCV_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_LinearCV_l.pit
Fri Jan 28 15:42:46 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
81
Benzenethiol
-------
FINAL
4-1-2011
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1
rho = 0 Specified
beta_0 = 4779.05
beta 1 = 90.829
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
alpha beta_0 beta_l
alpha 1 8.6e-007 -2.5e-006
beta_0 8.6e-007 1 -0.68
beta 1 -2.5e-006 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
1.12089e+006
beta_0
4976.05
beta_l
117.104
Estimate
829883
4627.88
97.2536
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
148474 538880
177.641 4279.71
10.1277 77.4036
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 4.39e+003 4.63e+003 581 911 -1.17
9 10 5.92e+003 5.5e+003 683 911 1.45
18 10 6.72e+003 6.38e+003 816 911 1.18
35 9 7.72e+003 8.03e+003 1.76e+003 911 -1.03
Model Descriptions for likelihoods calculated
82
Benzenethiol
-------
FINAL
4-1-2011
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-358.813074
-361.782335
-385.485941
# Param's
5
8
5
3
2
AIC
727.626149
714.534141
727.626149
729.564669
774.971882
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
19.092
5.93852
<.0001
0.0002617
0.0002617
0.05134
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than
non-homogeneous variance model
The p-value for Test 3 is less than
different variance model
The p-value for Test 4 is less than
model
.1. Consider running a
.1. You may want to consider a
.1. You may want to try a different
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
83
Benzenethiol
-------
FINAL
4-1-2011
Confidence level = 0.95
BMD = 9.3 6705
BMDL = 7.879
84
Benzenethiol
-------
FINAL
4-1-2011
Polynomial Model with 0.95 Confidence Level
9000
8000
a)
CO
§ 7000
a.
CO
a)
a:
c
ro
6000
5000
4000
0 5 10 15 20 25 30 35
dose
15:42 01/28 2011
Figure C-ll. Poly3_CV BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Text Output for Poly3_CV BMD Model for Absolute Kidney Weight Data (NTP, 1996)
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_PolyCV3_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Abs_Kidney_Wt_PolyCV3_l.pit
Fri Jan 28 15:42:46 2011
F0_M_Abs_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
85
Benzenethiol
-------
FINAL
4-1-2011
The polynomial coefficients are restricted to be positive
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1
rho = 0 Specified
beta_0 = 4390.1
beta_l = 226.11
beta_2 = 0
beta 3 = 0.0957401
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -beta_2 -beta_3
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
alpha beta_0 beta_l
alpha 1 5.le-005 1.5e-005
beta_0 5.le-005 1 -0.68
beta 1 1. 5e-005 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
1.31807e+006
beta_0
4999.29
beta_l
118.427
beta_2
beta 3
Estimate
944168
4627.92
97.2545
0
0
Std. Err.
190772
189.478
10. 8026
NA
NA
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
570262
4256.55
76.0818
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 4.39e+003 4.63e+003 581 972 -1.09
86
Benzenethiol
-------
FINAL
4-1-2011
9
10
5.92e+003
5.5e+003
683
972
1
18
10
6.72e+003
6.38e+003
816
972
1
35
9
7.72e+003
8.03e+003
1. 76e+003
972
-0
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-358.813074
-361.569737
-385.485941
# Param's
5
8
5
3
2
AIC
727.626149
714.534141
727.626149
729.139475
774.971882
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
1:
Test
2 :
Test
3:
Test
4 :
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
19.092
5.51333
<.0001
0.0002617
0.0002617
0.0635
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than
non-homogeneous variance model
The p-value for Test 3 is less than
different variance model
The p-value for Test 4 is less than
model
.1. Consider running a
.1. You may want to consider a
.1. You may want to try a different
87
Benzenethiol
-------
FINAL
4-1-2011
Benchmark Dose Computation
Specified effect
Risk Type
Confidence level
BMD
Estimated standard deviations from the control mean
0.95
9.99114
BMDL
7.99895
88
Benzenethiol
-------
FINAL
4-1-2011
Power Model with 0.95 Confidence Level
8000
7000
CD
-------
FINAL
4-1-2011
The power is restricted to be greater than or equal to 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 918585
rho = 0 Specified
control = 4390.1
slope = 435
power = 0.572274
the user,
alpha
control
slope
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -power
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha control slope
1 2.7e-009 -9.7e-010
2.7e-009 1 -0.68
-9.7e-010 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
1.31789e+006
control
4999.23
slope
118.425
power
Estimate
944063
4627.88
97.2537
1
Std. Err.
190730
189.468
10.802
NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
570240
4256.53
76.0822
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 4.39e+003 4.63e+003 581 972 -1.09
9 10 5.92e+003 5.5e+003 683 972 1.36
18 10 6.72e+003 6.38e+003 816 972 1.11
90
Benzenethiol
-------
FINAL
4-1-2011
35 9 7.72e+003 8.03e+003 1.76e+003 972 -0.97
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-358.813074
-349.267070
-358.813074
-361.569737
-385.485941
# Param's
5
8
5
3
2
AIC
727.626149
714.534141
727.626149
729.139474
774.971882
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
72.4377
19.092
19.092
5.51333
<.0001
0.0002617
0.0002617
0.0635
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than
non-homogeneous variance model
The p-value for Test 3 is less than
different variance model
The p-value for Test 4 is less than
model
.1. Consider running a
.1. You may want to consider a
.1. You may want to try a different
91
Benzenethiol
-------
FINAL
4-1-2011
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 9.99066
BMDL = 7.99895
92
Benzenethiol
-------
0 5 10 15 20 25 30 35
dose
15:49 01/28 2011
Figure C-13. Hill CV BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Hill CV BMD Model for Relative Kidney Weight Data (NTP, 1996)
FINAL
4-1-2011
Hill Model with 0.95 Confidence Level
Hill Model. (Version: 2.15; Date: 10/28/2009)
Input Data File:
C:/US E PA/BMDS212/A11r an2/NT P_19 9 6_F 0_M_Rel_Ki dne y_Wt_Hi11CV_1.(d)
Gnuplot Plotting File:
C:/US E PA/BMDS 212/A11r an2/NT P_19 9 6_F 0_M_Rel_Ki dne y_Wt_Hi11CV_1.pit
Fri Jan 28 15:49:32 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = intercept + v*dose^n/(k^n + dose^n)
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
93
Benzenethiol
-------
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
FINAL
4-1-2011
Default Initial Parameter Values
alpha =
rho =
intercept =
v =
n =
k =
1.63547
0
5.7
5.9
0.634978
18 .3462
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -n
have been estimated at a boundary point, or have been specified by
the user,
alpha
intercept
v
k
and do not appear in the correlation matrix )
alpha
1
3.4e-006
6.8e-006
6.9e-006
intercept
3 . 4e-006
1
0.45
0.46
6. 8e-006
0. 45
1
1
6. 9e-006
0.46
1
1
Interval
Variable
Limit
alpha
2.1041
intercept
6.25201
v
1947.07
n
k
11577.5
Parameter Estimates
Estimate Std. Err.
1.50727 0.304514
5.72246 0.270183
141.882 921.029
1 NA
812.759 5492.34
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.91043
5.19292
-1663.3
-9952.03
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
94
Benzenethiol
-------
FINAL
4-1-2011
o
9
18
35
20
10
10
9
5.7
7.4
8.7
11. 6
5.72
7.28
8.8
11.6
0. 62 6
1.26
1.17
2.22
1.23
1.23
1.23
1.23
-0.0818
0.318
-0.249
0.0486
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-34.465942
-34.552287
-68.998702
# Param's
5
8
5
4
2
AIC
78.931884
64.272980
78.931884
77.104573
141. 997404
Test 1:
Test
Test
Test
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
89.7244
20.6589
20.6589
0.172689
<.0001
0.0001239
0.0001239
0.6777
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
non-homogeneous variance model
Consider running a
The p-value for Test 3 is less than .1.
different variance model
You may want to consider a
95
Benzenethiol
-------
FINAL
4-1-2011
The p-value for Test 4 is greater than . 1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 7.09421
BMDL = 4.79931
96
Benzenethiol
-------
0 5 10 15 20 25 30 35
dose
15:49 01/28 2011
Figure C-14. LinearCV BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Linear CV BMD Model for Relative Kidney Weight Data (NTP, 1996)
FINAL
4-1-2011
Linear Model with 0.95 Confidence Level
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_LinearCV_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_LinearCV_l.pit
Fri Jan 28 15:49:32 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
97
Benzenethiol
-------
FINAL
4-1-2011
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1.63547
rho = 0 Specified
beta_0 = 5.76667
beta 1 = 0.166667
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
alpha beta_0 beta_l
alpha 1 -1.8e-008 -9.9e-009
beta_0 -1.8e-008 1 -0.68
beta 1 -9.9e-009 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
2.10508
beta_0
6.21065
beta_l
0.194408
Estimate
1.50797
5.74132
0.16765
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
0.304655 0.910854
0.239459 5.27199
0.0136521 0.140893
Table of Data and Estimated Values of Interest
Dose
Obs Mean
Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
9
18
35
20
10
10
9
5.7
7.4
8.7
11.6
5.74
7.25
8.76
11.6
0. 62 6
1.26
1.17
2.22
1.23
1.23
1.23
1.23
-0.15
0.386
-0.152
-0.0222
Model Descriptions for likelihoods calculated
98
Benzenethiol
-------
FINAL
4-1-2011
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-34.465942
-34.563689
-68.998702
# Param's
5
8
5
3
2
AIC
78.931884
64.272980
78.931884
75.127378
141. 997404
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (A1 vs A2)
Test 3: Are variances adeguately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 89.7244 6 <.0001
Test 2 20.6589 3 0.0001239
Test 3 20.6589 3 0.0001239
Test 4 0.195494 2 0.9069
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. Consider running a
non-homogeneous variance model
The p-value for Test 3 is less than .1. You may want to consider a
different variance model
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
99
Benzenethiol
-------
FINAL
4-1-2011
Confidence level =
BMD =
BMDL =
0.95
7.32473
6.02296
100
Benzenethiol
-------
0 5 10 15 20 25 30 35
dose
15:49 01/28 2011
Figure C-15. Poly3_CV BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Poly3_CV BMD Model for Relative Kidney Weight Data (NTP, 1996)
FINAL
4-1-2011
Polynomial Model with 0.95 Confidence Level
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_PolyCV3_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_PolyCV3_l.pit
Fri Jan 28 15:49:32 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose/s2 + ...
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
101
Benzenethiol
-------
FINAL
4-1-2011
The polynomial coefficients are restricted to be positive
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1.63547
rho = 0 Specified
beta_0 = 5.7
beta_l = 0.227194
beta_2 = 0
beta 3 = 9.92762e-005
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -beta_2 -beta_3
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
alpha beta_0 beta_l
alpha 1 -4e-009 -3.2e-009
beta_0 -4e-009 1 -0.68
beta 1 -3.2e-009 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
2.10508
beta_0
6.21065
beta_l
0.194408
beta_2
beta 3
Estimate
1.50797
5.74132
0.16765
0
0
Std. Err.
0.304655
0.239459
0.0136521
NA
NA
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.910854
5.27199
0.140893
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 5.7 5.74 0.626 1.23 -0.15
102 Benzenethiol
-------
FINAL
4-1-2011
9
10
7.4
18
10
8.7
35
9
11.6
7.25 1.26
8.76 1.17
11.6 2.22
1.23 0.386
1.23 -0.152
1.23 -0.0222
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/S2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/S2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-34.465942
-34.563689
-68.998702
# Param's
5
8
5
3
2
AIC
78.931884
64.272980
78.931884
75.127378
141. 997404
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
1:
Test
2 :
Test
3:
Test
4 :
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 89.7244 6 <.0001
Test 2 20.6589 3 0.0001239
Test 3 20.6589 3 0.0001239
Test 4 0.195494 2 0.9069
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. Consider running a
non-homogeneous variance model
The p-value for Test 3 is less than .1. You may want to consider a
different variance model
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
103
Benzenethiol
-------
FINAL
4-1-2011
Benchmark Dose Computation
Specified effect
Risk Type
Confidence level
BMD
Estimated standard deviations from the control mean
0.95
7.32473
BMDL
6.02296
104
Benzenethiol
-------
FINAL
4-1-2011
Power Model with 0.95 Confidence Level
dose
15:49 01/28 2011
Figure C-16. PowerCV BMD Model for Relative Kidney Weight Data (NTP, 1996)
Text Output for Power CV BMD Model for Relative Kidney Weight Data (NTP, 1996)
Power Model. (Version: 2.16; Date: 10/28/2009)
Input Data File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_PowerCV_l.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Allran2/NTP_1996_F0_M_Rel_Kidney_Wt_PowerCV_l.pit
Fri Jan 28 15:49:33 2011
F0_M_Rel_Kidney_Wt
The form of the response function is:
Y[dose] = control + slope * dose^power
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
105
Benzenethiol
-------
FINAL
4-1-2011
The power is restricted to be greater than or equal to 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1.63547
rho = 0 Specified
control = 5.7
slope = 0.227624
power = 0.915525
the user,
alpha
control
slope
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -power
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha control slope
1 4.le-009 -1.5e-009
4.le-009 1 -0.68
-1.5e-009 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
2.10508
control
6.21065
slope
0.194408
power
Estimate
1.50797
5.74132
0.16765
1
Std. Err.
0.304655
0.239459
0.0136521
NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.910854
5.27199
0.140893
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 20 5.7 5.74 0.626 1.23 -0.15
9 10 7.4 7.25 1.26 1.23 0.386
18 10 8.7 8.76 1.17 1.23 -0.152
106
Benzenethiol
-------
FINAL
4-1-2011
35 9 11.6 11.6 2.22 1.23 -0.0222
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = SigmaA2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-34.465942
-24.136490
-34.465942
-34.563689
-68.998702
# Param's
5
8
5
3
2
AIC
78.931884
64.272980
78.931884
75.127378
141. 997404
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adeguately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 89.7244 6 <.0001
Test 2 20.6589 3 0.0001239
Test 3 20.6589 3 0.0001239
Test 4 0.195494 2 0.9069
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. Consider running a
non-homogeneous variance model
The p-value for Test 3 is less than .1. You may want to consider a
different variance model
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
107
Benzenethiol
-------
FINAL
4-1-2011
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 7.32473
BMDL = 6.02296
108
Benzenethiol
-------
FINAL
4-1-2011
APPENDIX D. REFERENCES
ACGIH (American Conference of Governmental Industrial Hygienists). (2008) Threshold limit
values for chemical substances and physical agents and biological exposure indices. Cincinnati,
OH: ACGIH. As cited in HSDB, 2009.
American Biogenics Corp. (1989) Ninety-day gavage study in albino rats using thiophenol.
Prepared by Dynamac Corporation, Rockville, MD for the U.S. EPA Office of Solid Waste (not
available; only cited in HEAST).
Amrolia, P; Sullivan, SG; Stern, A; Munday, R. (1989) Toxicity of aromatic thiols in the human
red blood cell. JAppl Toxicol 9(2): 113-118.
ATSDR (Agency for Toxic Substances and Disease Registry). (2010) Toxicological profile
information sheet. Atlanta, GA: U.S. Department of Health and Human Services, Public Health
Service. Available online at http://www.atsdr.cdc.gov/toxprofiles/index.asp. Accessed on
1/2/2010.
CalEPA (California Environmental Protection Agency). (2008) All OEHHA acute, 8-hour and
chronic reference exposure levels (chRELs) as of December 18, 2008. Sacramento: Office of
Environmental Health Hazard Assessment. Available online at
http://www.oehha.ca.gov/air/allrels.html. Accessed on 1/2/2010.
CalEPA (California Environmental Protection Agency). (2009a) OEHHA/ARB approved
chronic reference exposure levels and target organs. Sacramento: Office of Environmental
Health Hazard Assessment. Available online at
http://www.arb.ca.gov/toxics/healthval/chronic.pdf. Accessed on 1/2/2010.
CalEPA (California Environmental Protection Agency). (2009b) Hot spots unit risk and cancer
potency values. Sacramento, CA: Office of Environmental Health Hazard Assessment.
Available online at http://www.oehha.ca.gov/air/hot_spots/pdf/TSDlookup2002.pdf Accessed
on 1/2/2010.
CalEPA (California Environmental Protection Agency). (2009c) Technical support document
for describing available cancer potency factors. Appendix I. Sacramento, CA: Office of
Environmental Health Hazard Assessment. Available online at
http://www.oehha.ca.gov/air/hot_spots/pdf/Appendix%20I2002.pdf. Accessed on 1/2/2010.
Fairchild, EJ; Stokinger, HE. (1958) Toxicologic studies on organic sulfur compounds. I.
Acute toxicity of some aliphatic and aromatic thiols (Mercaptans). Am IndHyg Assoc J
19(3): 171—189.
Hazelton Laboratories. (1951) Acute oral, acute and chronic dermal and vapor toxicity study of
thiophenol, with cover letter dated 05/06/94. OTS0557378.
109
Benzenethiol
-------
FINAL
4-1-2011
Haz-Map. (2010). Occupational exposure to hazardous substances. Bethesda, MD: Specialized
Information Services, U.S. National Library of Medicine. Available online at
http://hazmap.nlm.nih.gov/cgi-bin/hazmap_generic?tbl=TblAgents&id=183. Accessed on
3/4/2010.
HSDB (Hazardous Substances Data Bank). (2009) Thiophenol, CASRN: 108-98-5. Bethesda,
MD: National Library of Medicine, National Toxicology Program. Available online at
http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7HSDB. Accessed on February 17, 2010.
FAO/WHO (Food and Agriculture Organization/World Health Organization). (1999)
Evaluation of certain food additives and contaminants. A meeting of the FAO/WHO Joint
Expert Committee on Food Additives (JECFA). Fifty-third meeting, Rome, 1-10 June 1999.
Geneva, Switzerland: WHO.
IARC (International Agency for Research on Cancer). (2009) IARC Monographs on the
evaluation of carcinogenic risks to humans. Lyon, France: IARC. Available online at
http://monographs.iarc.fr/ENG/Monographs/PDFs/index.php. Accessed on 1/2/2010.
Lavoie, E; Tulley, L; Fow, E; Hoffmann, D. (1979) Mutagenicity of aminophenol and
nitrophenol ethers, sulfides, and disulfides. Mutat Res 67:123-131.
McBain, JB; Menn, J. (1969) S-methylation, oxidation, hydroxylation, and conjugation of
thiophenol in the rat. Biochem Pharmacol 18(9):2282-2285.
NIOSH (National Institute for Occupational Safety and Health). (1978) Criteria for a
recommendation standard: occupational exposure to //-alkane monothiols, cyclohexanethiol, and
benzenethiol. DHEW (NIOSH) Publication No. 78-213. Atlanta, Ga: Center for Disease
Control and Prevention, U.S. Department of Health, Education and Welfare. Available online at
http://www.cdc.gov/niosh/78-213.html
NIOSH (National Institute for Occupational Safety and Health). (2005) NIOSH pocket guide to
chemical hazards. Index of Chemical Abstracts Service Registry Numbers (CAS No.). Atlanta,
Ga: Center for Disease Control and Prevention, U.S. Department of Health, Education and
Welfare. Available online at http://www.cdc.gov/niosh/npg/npgdcas.html. Accessed on
1/2/2010.
NTP (National Toxicology Program). (1992) Range-finding studies: Developmental toxicity.
Thiophenol when administered by gavage in New Zealand White Rabbits. Study Nos.
NTP-92-RF/DT-036 and NTP-92-RF/DT-043; Contract No. N01-ES-95249.
NTP (National Toxicology Program). (1994a) Final report on the developmental toxicity of
thiophenol (CAS no. 108-98-5) in Sprague-Dawley rats on gestational days 6 through 15. U.S.
Department of Health and Human Services, Public Health Service, National Institutes of Health,
Research Triangle Park, NC. NTP TER-92133, NTIS Technical Report NTIS/PB94-155009.
NTP (National Toxicology Program). (1994b) Final report on the developmental toxicity of
thiophenol (CAS no. 108-98-5) in New Zealand White (NZW) rabbits. U.S. Department of
Health and Human Services, Public Health Service, National Institutes of Health, Research
Triangle Park, NC. NTP TER-92134, NTIS Technical Report NTIS/PB94-201183.
110
Benzenethiol
-------
FINAL
4-1-2011
NTP (National Toxicology Program). (1996) Final report on the reproductive toxicity of
thiophenol (CAS no. 108-98-5) administered by gavage to Sprague-Dawley rats. U.S.
Department of Health and Human Services, Public Health Service, National Institutes of Health,
Research Triangle Park, NC. NTP RACB94001, NTIS Technical Report NTIS/PB96-211735.
NTP (National Toxicology Program). (2005) 11th Report on Carcinogens. U.S. Department of
Health and Human Services, Public Health Service, National Institutes of Health, Research
Triangle Park, NC. Available online at http://ntp-server.mehs.nih.gov/index.cfm?objectid=
32BA9724-F1F6-975E-7FCE50709CB4C932. Accessed on 1/2/2010.
OSHA (Occupational Safety and Health Administration). (2009) Air contaminants:
occupational safety and health standards for shipyard employment, subpart Z, toxic and
hazardous substances. U.S. Department of Labor, Washington, DC. OSHA Standard
1915.1000. Available online at
http://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=STANDARDS&p_id=102
86. Accessed on 1/2/2010.
Stauffer Chemical Company. (1969) Acute inhalation LC50 study of thiophenol in male and
female rats with cover letter dated 05/06/94. OTS0557380.
U.S. EPA (Environmental Protection Agency). (1994) Chemical assessments and related
activities (CARA). Office of Health and Environmental Assessment, Washington, DC.
EPA/600/R-94/904. Available online at nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=6000lG8L.txt
U.S. EPA (Environmental Protection Agency). (1999) Benchmark dose software (BMDS)
version 2.1.1 and 2.1.2 (last modified June 16, 2010). Available online at
http://www.epa.gov/ncea/bmds/.
U.S. EPA (Environmental Protection Agency). (2006) 2006 edition of the drinking water
standards and health advisories. Office of Water, Washington, DC. EPA/822/R-06/013.
Washington, DC. Available online at
http://water.epa.gOv/action/advisories/drinking/upload/2009_04_27_criteria_drinking_dwstandar
ds.pdf. Accessed on 1/2/2010.
U.S. EPA (Environmental Protection Agency). (2007) Interim acute exposure guideline levels
(AEGLs) for phenyl mercaptan (C6H5SH) (CAS Reg. No. 108-98-5). Interim 1: 11/2007.
Available online at http://www.epa.gov/oppt/aegl/pubs/phenyl_mercaptan_interim_nov_
2007_vl.pdf. Accessed on 1/2/2010.
U.S. EPA (Environmental Protection Agency). (2010a) Integrated Risk Information System
(IRIS). Office of Research and Development, National Center for Environmental Assessment,
Washington, DC. Online, http://www.epa.gov/iris. Accessed on 1/2/2010.
U.S. EPA (Environmental Protection Agency). (2010b) Health effects assessment summary
tables (HEAST). Office of Emergency and Remedial Response, Washington, DC. Available
online at http://epa-heast.ornl.gov/. Accessed on 2/19/2010.
Ill
Benzenethiol
-------
FINAL
4-1-2011
WHO (World Health Organization). (2010) Online catalogs for the Environmental Health
Criteria Series. Online, http://www.who.int/topics/environmental_health/en/. Accessed on
1/2/2010.
112
Benzenethiol
------- |