United States
Environmental Protection
1=1 m m Agency
EPA/690/R-09/027F
Final
9-08-2009
Provisional Peer-Reviewed Toxicity Values for
Hydroquinone
(CASRN 123-31-9)
Superfund Health Risk Technical Support Center
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
COMMONLY USED ABBREVIATIONS
BMD
Benchmark Dose
IRIS
Integrated Risk Information System
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
p-RfC
provisional inhalation reference concentration
p-RfD
provisional oral reference dose
RfC
inhalation reference concentration
RfD
oral reference dose
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
1
-------
FINAL
9-8-2009
PROVISIONAL PEER-REVIEWED TOXICITY VALUES FOR
HYDROQUINONE (CASRN 123-31-9)
Background
On December 5, 2003, the U.S. Environmental Protection Agency's (U.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. U.S. EPA's Integrated Risk Information System (IRIS).
2. Provisional Peer-Reviewed Toxicity Values (PPRTVs) used in U.S. 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 U.S. EPA's IRIS. 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 U.S. EPA IRIS Program. All provisional toxicity values receive internal
review by two U.S. 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 U.S. 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.
It is important to remember that a provisional value alone tells very little about the
adverse 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 U.S. EPA
1
-------
FINAL
9-8-2009
Office of Research and Development's National Center for Environmental Assessment,
Superfund Health Risk Technical Support Center for OSRTI. Other U.S. 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 U.S. EPA Office of Research and Development's National Center for Environmental
Assessment, Superfund Health Risk Technical Support Center (513-569-7300), or OSRTI.
INTRODUCTION
There is no RfD for hydroquinone on IRIS (U.S. EPA, 2007) or in the Drinking Water
Standards and Health Advisories list (U.S. EPA, 2006). A subchronic RfD of 0.4 mg/kg-day and
a chronic RfD of 0.04 mg/kg-day are listed in the HEAST (U.S. EPA, 1997) for hydroquinone.
These RfD values were derived using a NOAEL of 4.29 mg/kg-day for hematological effects
from a 3-5 month study in humans (Carlson and Brewer, 1953) and an uncertainty factor of
10 (subchronic RfD) or 100 (chronic RfD). The source document for these derivations is a 1987
Health and Environmental Effects Document (HEED) (U.S. EPA, 1987). This HEED and a
Reportable Quantity document (U.S. EPA, 1988a) are the only U.S. EPA reports on
hydroquinone in the Chemical Assessments and Related Activities (CARA) list (U.S. EPA,
1991a, 1994). ATSDR (2006) has not produced a Toxicological Profile for hydroquinone. An
Environmental Health Criteria Document (WHO, 1994) is available, but it does not derive any
risk assessment values.
There are no RfC values for hydroquinone on IRIS or in the HEAST, and both include
messages stating that an RfC for hydroquinone is not verifiable (U.S. EPA, 2007, 1997).
Occupational exposure limits are available for hydroquinone that include a threshold limit
value-time weighted average (TLV-TWA) of 2 mg/m , recommended by the American
Conference of Governmental Industrial Hygienists (ACGIH, 2005), and a permissible exposure
limit-time weighted average (PEL-TWA) of 2 mg/m3 promulgated by the Occupational Safety
and Health Administration (OSHA, 2007). The National Institute of Occupational Safety and
Health (NIOSH, 1978, 2005) has established a recommended exposure limit (REL) of 2 mg/m3
as a 15-minute ceiling for hydroquinone. These limits are intended to protect against eye injury
(irritation or corneal changes), dermatitis, and CNS effects potentially associated with
occupational exposure.
No carcinogenicity assessment is available on IRIS (U.S. EPA, 2008), and hydroquinone
is not listed in the HEAST cancer table (U.S. EPA, 1997) or indicated in the Drinking Water
Standards and Health Advisories list (U.S. EPA, 2006). The Office of Pesticide Programs (OPP)
does not list a cancer classification for hydroquinone, but it does report an oral slope factor
(OSF) of 5.6 x 10"2 per mg/kg-day for this chemical (basis not reported). The carcinogenicity of
hydroquinone was tested by the National Toxicology Program (NTP, 1989) and evaluated by the
International Agency for Research on Cancer (IARC, 1999), who categorized hydroquinone in
Group 3 (Not classifiable as to its carcinogenicity to humans) based on inadequate evidence in
2
-------
FINAL
9-8-2009
humans and limited evidence in experimental animals. The Environmental Health Criteria
Document (WHO, 1994) on hydroquinone concluded that insufficient data are available for a
thorough assessment of the carcinogenic potential for humans due to limited evidence in animals
and the lack of adequate epidemiological studies.
Literature searches for studies relevant to the derivation of provisional toxicity values for
hydroquinone (CASRN 123-31-9) were conducted in MEDLINE, TOXLINE special, and
DART/ETIC (1960's-January 2007); BIOSIS (2000-January 2007); TSCATS/TSCATS2,
RTECS, CCRIS, HSDB, and GENETOX (not date limited); and Current Contents (previous
6 months). A final search in PubMed was conducted (January 2007-July 2008).
REVIEW OF PERTINENT DATA
Human Studies
The oral toxicity of hydroquinone was assessed in two men who ingested 500 mg/day for
5 months and 17 men and women (numbers/sex not reported) who ingested 300 mg/day
for3-5 months (Carlson and Brewer, 1953). Total daily chemical intake was consumed with
meals in three divided doses. Assuming an average human body weight of 70 kg (U.S. EPA,
1987), the estimated daily doses on a per-kg basis were 7.1 and 4.3 mg/kg-day. Hematology
indices (red blood cell [RBC] count, hematocrit, percent hemoglobin, differential white blood
cell count, sedimentation rate, platelet count, coagulation time, and iteric index) and urine
indices (albumin, reducing sugars, white and red cell counts, casts, and urobilinogen) were
evaluated during a control period for 1 month prior to exposure and again while the experiment
was in progress, enabling each subject to serve as his/her own control. Results of the blood
analyses and urinalyses revealed no abnormal findings. No additional information was reported
on the design or results of this study. Because the high dose was administered to only two
subjects, the low dose is used to identify a NOAEL of 4.3 mg/kg-day for hematological and renal
effects in humans.
Corneal alterations with loss of visual acuity have been reported in workers with years of
prolonged exposure to hydroquinone dust and quinone vapor (Anderson, 1947; Anderson and
Oglesby, 1958; Miller, 1954; Naumann, 1966; Sterner et al., 1947). The eye lesions generally
increased in severity with the length of exposure and progressed in some individuals even after
exposure had ended. All studies describing eye lesions involved exposure to both hydroquinone
dust and quinone vapor. The relative contributions of the two chemicals, as well as the relative
roles of direct contact and systemic toxicity in causing corneal injury, are uncertain (DeCaprio,
1999; U.S. EPA, 1987).
Mortality from cancer and noncancer causes was not increased in a cohort of 879 workers
(858 men and 21 women) who had worked for at least 6 months between January 1930 and
December 1990 in a Tennessee plant in which hydroquinone was manufactured and used
(Pifer et al., 1995). Average hydroquinone dust levels ranged from 0.1-6.0 mg/m2 from
1949-1990. Mean employment duration was 13.7 years and mean follow-up from first exposure
was 6.8 years; follow-up was essentially complete. Standardized mortality ratios (SMRs) were
determined by comparing observed deaths with those expected in the general population of
Tennessee, as well as in an occupational cohort from an out-of-state plant not exposed to
hydroquinone. The SMRs for all causes of death combined (n = 168), all cancers combined
3
-------
FINAL
9-8-2009
(n = 33), and most site-specific cancers and noncancer diseases were significantly below 1.0 (less
than expected) within both comparison populations. No SMRs were significantly greater than
1.0 within either population of referents. Only two sites, the colon (5 cases) and the lungs
(14 cases), had more than three observed cancer cases; this indicates that the power of the study
to detect site-specific effects was weak (IARC, 1999). There were no significant mortality
excesses from the endpoints suggested by the animal studies of hydroquinone, including kidney
cancer, liver cancer, leukemia, and nephrotoxicity. Dose-response analyses of deaths from
selected cancers and other diseases did not show any significant heterogeneities or linear trends
according to estimated career exposure (mg/m3-years) or latency (time from first exposure).
Animal Studies
Oral Exposure
Subchronic Studies—The systemic effects of oral exposure to hydroquinone have been
investigated in numerous animal studies. Studies have been performed using drinking water,
dietary, and gavage exposure.
Drinking water studies were conducted by Christian et al. (1980, as cited by U.S. EPA,
1987). In one study, groups of 6 male and 6 female Carworth Farm adult rats were exposed to 0,
2500, 5000, or 10,000 ppm of hydroquinone (purity not reported) in the drinking water for
8 weeks (Christian et al., 1980). The estimated doses based on measured water consumption
were 0, 230, 390, or 700 mg/kg-day for males and 0, 270, 470, or 810 mg/kg-day for females.
The effects included dose-related decreased water intake throughout the study in both males and
females, decreased body weight gain in the females at >470 mg/kg-day and males at
700 mg/kg-day, and increased relative liver and kidney weights in males at >390 and females at
>470 mg/kg-day. Blood analysis (hematocrit, hemoglobin content, and leukocyte count after
5 weeks of exposure) and comprehensive histological examinations (23 tissues at the end of the
study) revealed no compound-related abnormalities. Additional information on the experimental
design and results were not provided in the available U.S. EPA (1987) summary of this study.
The low dose (230-270 mg/kg-day) is a NOAEL and the mid dose (390-470 mg/kg-day) a
minimal LOAEL for effects on body and organ weights in this study.
Weanling rats were used in a second study by Christian et al. (1980, as summarized by
U.S. EPA, 1987). Groups of 15 male and 15 female Carworth Farm weanling rats were exposed
to 0, 1000, 2000, or 4000 ppm of hydroquinone (purity not reported) in the drinking water for
15 weeks. Estimated doses based on measured water consumption were 0, 110, 200, or
360 mg/kg-day for males and 0, 140, 240, or 430 mg/kg-day for females. Effects included
dose-related decreased water intake throughout the study in both sexes, slightly decreased body
weight gain in males at 360 mg/kg-day, highly significant increases in relative liver weight in
males at >110 mg/kg-day and females at >140 mg/kg-day, and significant increases in relative
kidney weight in males at >110 mg/kg-day and females at >240 mg/kg-day. Hemoglobin levels
were slightly reduced in males at >200 mg/kg-day after 5 and 10 weeks of exposure, but they
were normal after 14 weeks of exposure. No changes in total or differential leukocyte counts in
the males or any hematological effects in the females were reported. Measurements of
spontaneous locomotor activity and comprehensive histological examinations (23 tissues)
revealed no compound-related abnormalities. Additional information on the experimental design
and results was not provided in the available U.S. EPA (1987) summary of this study. The low
dose of 110 mg/kg-day in males is identified as a minimal LOAEL for increased liver and kidney
weights.
4
-------
FINAL
9-8-2009
Feeding studies with hydroquinone were conducted in rats, hamsters, and dogs.
Sprague-Dawley rats (groups of 14 adults of unspecified sex) were exposed to diets containing
0 or 5% (0 or 50,000 ppm) hydroquinone (purity not reported) for 9 weeks (Carlson and Brewer,
1953). Using reference values for food consumption and body weight in adult Sprague-Dawley
rats (U.S. EPA, 1988b), the 50,000 ppm diet provided an estimated dose of 3333 mg/kg-day of
hydroquinone in males and 4000 mg/kg-day in females. The exposed rats experienced a 46%
loss in weight and developed aplastic anemia (incidence rate not reported). Examination of the
bone marrow showed a 66% average decrease in cellularity relative to controls, with marked
atrophy of the hematopoietic elements. Other effects included atrophy of the liver cord cells,
splenic lymphoid tissue, adipose tissue and striated muscle, as well as superficial ulceration and
hemorrhage of the stomach mucosa. The authors indicated that the effects were partly due to
reduced food intake. The observed reduction in food consumption and weight loss introduces
uncertainty into the dose estimate. No additional information on the design and results of this
study were reported. The severity of the body weight and the hematological and
histopathological changes indicate that the 5% dietary level (estimated as 3333 mg/kg-day in
males and 4000 mg/kg-day in females) is a FEL for subchronic exposure in rats.
Hamsters (groups of 15 male 6-week-old Syrian strain) were exposed to 0 or 0.5%
(5000 ppm) hydroquinone (>99% pure) in the diet for 20 weeks (Hirose et al., 1986). Using
reference values for food consumption and body weight in male Syrian hamsters with subchronic
exposure (U.S. EPA, 1988b), the 5000 ppm diet provided an estimated dose of 474 mg/kg-day.
Evaluation of body, liver, and kidney weights and histology of seven tissues (cheek pouch,
esophagus, stomach, liver, kidneys, bladder, pancreas, and lung) showed no effects other than an
equivocal increase in mild hyperplasia of the forestomach. The lack of any clear
exposure-related changes indicates that 474 mg/kg-day is a NOAEL for subchronic
histopathology in hamsters.
For 80 weeks, 4-month old mongrel dogs from two litters were fed diets containing
hydroquinone (purity not reported) in tablets (Carlson and Brewer, 1953). A single pup was
exposed to 16 mg/kg-day, two pups were exposed to 1.6 mg/kg-day for 31 weeks followed by
40 mg/kg-day for 49 weeks (TWA dose 25.1 mg/kg-day), and two pups were maintained as
controls. The sex of the treated and control dogs was not reported. In a second experiment,
5 adult male dogs were fed 100 mg/kg-day in tablets for 26 weeks. Routine analyses of the
blood and urine were performed during the study (indices not specified) and limited histological
examinations (liver, kidney, spleen, bone marrow, and 8 other tissues) were performed at
necropsy. No further evaluations were reported. No exposure-related effects were observed in
any exposed dogs, indicating that 100 mg/kg-day is a subchronic NOAEL for systemic toxicity
in dogs.
Gavage studies reported high mortality in rats exposed to hydroquinone at doses of
500 mg/kg-day and above. High mortality was observed in groups of 20-48 rats (age not stated)
that were treated with gavage doses ranging from 500 to 1750 mg/kg for up to 9 times in 12 days
(Carlson and Brewer, 1953). Almost 75% of the deaths in this study occurred within 24 hours of
the first dosing. More than 50% of the rats died within the first 2 months of exposure in a group
of 16 Sprague-Dawley rats given hydroquinone by gavage (vehicle not specified; sex not
specified) at a dose of 500 mg/kg for up to 101 times in 151 days (5 months) (Carlson and
Brewer, 1953). Gross and histological examinations on the survivors at the end of the study
showed no remarkable effects. No additional information was provided on the results and no
5
-------
FINAL
9-8-2009
other endpoints were evaluated. The mortality data indicate that 500 mg/kg-day is a FEL for
subchronic gavage exposure.
In a 40-day study, groups of 25 rats were dosed 6 days/week with hydroquinone by
gavage in water at doses of 7.5 or 15 mg/kg-day (Delcambre et al., 1962, as summarized by
U.S. EPA, 1987). A group of 20 rats dosed with 5% glucose served as controls. Hematological
examinations were performed on 2 rats per dose on Exposure Days 8, 15, 26, and 36 and on
7 rats per dose on Day 40. No changes in total red, total white or differential white blood cell
counts were observed on Days 8 and 15. On Day 26, one rat in the 7.5 mg/kg-day dosing group
developed anisocytosis (considerable variation in the size of erythrocytes) with
polychromatophilia (young or degenerating erythrocytes with unusual staining characteristics);
these effects were not observed at 15 mg/kg-day on Day 26. No abnormalities were observed at
7.5 mg/kg-day on Day 36, although one rat in the 15 mg/kg-day dosing group developed slight
anemia with decreased neutrophils, anisocytosis, severe polychromatophilia, and numerous
erythroblasts in peripheral blood. On Day 40, one control rat had anisocytosis and
erythroblastosis, one rat in the 7.5 mg/kg-day dosing group had erythroblastosis and several rats
in the 15 mg/kg-day rats had anisocytosis (4 rats), definite polychromatophilia (5 rats), and
erythroblasts in peripheral blood (4 rats). No additional information on the experimental design
or results was reported in the U.S. EPA (1987) summary of this study. A NOAEL of
7.5 mg/kg-day and a LOAEL of 15 mg/kg-day are identified on the basis of the hematological
effects. The LOAEL is considered by the U.S. EPA (1987) to be of minimal toxicological
significance because the anisocytosis, polychromatophilia, and erythroblastosis were not
accompanied by reduced numbers of circulating erythrocytes.
In a longer-term study by the same researchers, groups of 15 rats were treated with
hydroquinone by gavage in water at doses of 5 or 10 mg/kg-day for 6 days/week for 4 months
(Delcambre et al., 1962, as summarized by U.S. EPA, 1987). A group of 15 rats dosed with
5% glucose was used as controls. During the study, 1,1, and 7 rats in the 0, 5, and 10 mg/kg-day
groups died; causes of death were not reported, although 3 of the 7 high-dose rats died during a
scabies epidemic. A decrease in body weight gain was observed at 10 mg/kg-day during the
3rd month, when mortality was greatest. No exposure-related changes in blood parameters were
reported. Additional information on the experimental design and results was not provided in the
U.S. EPA (1987) summary of this study. Identification of a reliable effect level is precluded by
the insufficient information on the cause of death and by the lack of mortality in rats that were
exposed to a higher dose of 15 mg/kg-day for 40 days by the same investigators, as summarized
above.
A 13-week gavage study was performed in rats by Topping et al. (2007). Groups of
Sprague-Dawley rats (10/sex/group) were given hydroquinone (99% purity) in degassed distilled
water via gavage doses of 0, 20, 64, or 200 mg/kg-day, 5 days/week, for 13 weeks. Clinical
observations were made daily and measurements of body weight and food consumption were
made weekly. To assess nervous system impairment, a functional observational battery (FOB)
was administered 3 days prior to first dose, after dosing on Day 1 (1 and 6 hours after dosing)
and prior to dosing on Days 2, 7, 14, 30, 60 and 91. At necropsy, the brain and kidney were
weighed and subjected to histopathological examination.
Increased incidences of depression (reduced locomotor and home cage activity) and
tremors were statistically significant in the 64- and 200-mg/kg-day dose groups. Both male and
female rats were affected at 1 and 6 hours after the first dosing (Topping et al., 2007). These
6
-------
FINAL
9-8-2009
neurological effects appear to be acute, as recovery occurred prior to subsequent FOB
observations. Body weights of 200 mg/kg-day males were decreased by 7% (p < 0.05)
compared with controls at necropsy. Food consumption in the 200 mg/kg-day males was
significantly lower than in controls for the first 5 days of the study, but it was increased to
controls levels thereafter. No significant pathological findings were observed. Based on acute
neurological effects, this study identified a LOAEL of 64 mg/kg-day and a NOAEL of
20 mg/kg-day in rats.
The NTP (1989) conducted 13-week toxicity studies in F344/N rats and B6C3F1 mice to
determine the doses to be used in subsequent two-year studies (summarized below). Groups of
10 males and 10 females of each species were administered 0, 25, 50, 100, 200, or
400 mg/kg-day doses of hydroquinone (>99% pure) by gavage in corn oil on a 5 days/week
regimen for 13 weeks. Clinical signs and body weight were evaluated throughout the study.
Necropsies and liver weight measurements were performed on all animals. Comprehensive
histological examinations were performed on all vehicle controls, animals receiving 200 or
400 mg/kg-day, and animals dying before the end of the study. Tissues examined in the
100 mg/kg-day dose groups were limited to the liver, kidneys, and stomach of male rats and
kidneys of female rats.
All rats receiving 400 mg/kg-day and 3/10 female rats receiving 200 mg/kg-day died
before the end of the study. Tremors and convulsions occurred after dosing in most males and
females at 400 mg/kg-day and in some females at 200 mg/kg-day. There were also signs of
lethargy in both sexes at >200 mg/kg-day. Other effects in the 400 mg/kg-day rats included
red-to-brown perioral staining in 4/10 males and 5/10 females, reddened stomach mucosa in
1/10 males and 2/10 females, and meningial hemorrhage in 1/10 males. The 200 mg/kg-day
males exhibited intra-abdominal bleeding in 2/10 rats, while 1/10 females exhibited blood in the
stomach and 2/10 had perioral bleeding. Also in the 200 mg/kg-day group, inflammation and/or
epithelial hyperplasia (acanthosis) of the forestomach (4/10 males and 1/10 females) and toxic
nephropathy (tubular cell degeneration in the renal cortex) (7/10 males and 6/10 females) were
seen. Tubular cell degeneration in the renal cortex was observed in 1/10 females at
100 mg/kg-day. Significant 8-9% reductions in body weight, compared to controls, were seen in
males at >100 mg/kg-day (with a smaller, but still statistically significant 5% decrease at
50 mg/kg-day). Increased absolute and relative liver weights were observed in females at
>50 mg/kg-day. Decreases in absolute and relative liver weight in males were probably
secondary to the reduced body weight in males. The kidney lesions in males were judged by
NTP (1989) to be of moderate-to-marked severity and consisted of tubular cell degeneration and
regeneration in the renal cortex. Kidney lesions in the females were similar to those in males but
of lesser (minimal-to-mild) severity. The changes in liver weight (without corresponding
histopathology at any dose) and the small decrease in body weight at 50 mg/kg-day are not
considered adverse. Therefore, a NOAEL of 50 mg/kg-day and a LOAEL of 100 mg/kg-day are
identified based on kidney lesions in females and 8—9% decreased body weight in males. Overt
toxic effects and death were observed at 200 mg/kg-day.
The mice exposed to hydroquinone for 13 weeks experienced the following effects:
mortality (8/10 males and 8/10 females at 400 mg/kg-day and 2/10 males at 200 mg/kg-day with
one death in this group attributed by the study authors to gavage error); forestomach
histopathology with ulceration, inflammation or epithelial hyperplasia (3/10 males and
2/10 females at 400 mg/kg-day and 1/10 females at 200 mg/kg-day); tremors after dosing
(females at 400 mg/kg-day and males at >200 mg/kg-day); lethargy (females at >100 mg/kg-day
7
-------
FINAL
9-8-2009
and males at >25 mg/kg-day); and increased absolute and relative liver weights (males at
>25 mg/kg-day with no clear relationship to dose). There were no consistent dose-related
changes in body weight and no hepatic histopathology (NTP, 1989). Although lethargy was the
most common clinical sign and was observed in all dosed males, NTP (1989) concluded that
doses of 100 mg/kg-day and below resulted in no discernible indices of toxicity that would
preclude long-term growth and survival in mice. Additionally, the significance of the lethargy is
questionable because there was no lethargy or any other signs of toxicity in mice exposed to 50
or 100 mg/kg-day for up to 103 weeks in the chronic experiment, as summarized below. The
dose of 200 mg/kg-day is a FEL for marked gastric histopathology, tremors, and death. The
NOAEL is 100 mg/kg-day.
Chronic Studies—Chronic toxicology/carcinogenesis studies were performed in which
groups of 65 F344/N rats of each sex were treated with 0, 25, or 50 mg/kg-day doses of
hydroquinone (>99% pure) by gavage in deionized water for 5 days/week for up to 103 weeks
(NTP, 1989). Groups of 64 or 65 B6C3F1 mice of each sex were similarly exposed to doses of
0, 50, or 100 mg/kg-day. Clinical signs and body weight were evaluated throughout the studies.
Hematology exams (total red and white blood cell counts, differential white cell counts,
hematocrit, hemoglobin concentration, and reticulocyte counts) and clinical chemistry exams
(6 indices including blood urea nitrogen) were performed on 10 animals from each group after
65 weeks of exposure. Necropsies, organ weight measurements (liver, kidney, brain), and
histological examinations were performed on all rats after 65 or 103 weeks of exposure. The
histological exams were comprehensive in all rats (except that preputial gland and thyroid were
not examined in low-dose males) and vehicle control and high-dose mice; tissues examined in
low-dose mice were limited to gross lesions, liver, spleen, thyroid, and adrenal glands in males,
and gross lesions, liver, lungs, ovaries, salivary glands, and thyroid in females.
No compound-related clinical signs were observed in the rats (NTP, 1989). Survival in
treated rats was similar to controls. Males treated with 50 mg/kg-day had reduced body weight
(up to 13% lower than controls) throughout the second year of the study. Body weight was also
reduced in males treated with 25 mg/kg-day, but the difference from controls was small (less
than 10%) and the effect was only seen late in the study (Weeks 89-103). There was no effect
on body weight in female rats. Increases in relative brain, kidney, and liver weights in the
50 mg/kg-day males appeared to be secondary to decreased body weight in this group. The
relative weights of these organs were not different from controls in the female rats. Spontaneous
nephropathy occurred in nearly all male and most female rats of all dosed groups and vehicle
controls. The nephropathic changes occurred at both 15-month and 2-year sacrifices, were
consistent with age-related advanced renal disease and were more severe in the 50 mg/kg-day
males than in controls. No hyaline droplet formation was seen in the kidneys. Other effects in
the rats included significantly (p < 0.05) decreased RBC count, hematocrit percent, and
hemoglobin concentration in females at 50 mg/kg-day (evaluated at 15 months). Based on the
hematological changes in females and the increased severity of toxic nephropathy and the
reduced body weight in males, 50 mg/kg-day is a LOAEL and 25 mg/kg-day a NOAEL for
nonneoplastic effects in rats in this study.
The NTP (1989) concluded that there was some evidence of carcinogenic activity of
hydroquinone in male and female rats based on increases in renal tubular adenomas in male rats
and mononuclear cell leukemia in females. The incidences of renal tubule cell adenomas in male
rats are shown in Table 1. There was a statistically significant trend for increased renal tumors
with dose and the incidence in the high dose-group was statistically increased in pairwise
8
-------
FINAL
9-8-2009
comparison to concurrent controls. The incidence in both dose groups exceeded the highest
historical incidence of this tumor in either untreated (3/50 = 6%) or water gavage (1/50 = 2%)
controls, and it is markedly higher than the overall historical incidence of less than 0.5% in both
types of controls. A reanalysis of the histology data from this study found that the adenomas
were located in areas of severe chronic progressive nephropathy (Hard et al., 1997). The
incidences of mononuclear cell leukemia are also given in Table 1. There was a statistically
significant trend for increased mononuclear cell leukemia with dose and the incidence in the
high-dose group was statistically increased in pair-wise comparison to concurrent controls. The
historical incidence of mononuclear cell leukemia for water gavage vehicle control female
F344/N rats was 25 ± 15% (n = 299), while that for untreated controls was 19% ± 7% (// = 1983).
The incidence of leukemia in the high-dose females was just within the historical control range.
The researchers graded the severity of the observed leukemia as three stages. Features of
Stage 1 include limited distortion of splenic architecture, no infiltration of other organs that are
not likely to cause death. Stage 2 effects include an effacement of splenic architecture, limited
infiltration of the liver, and possibly other organs that may contributed to mortality. Stage 3
effects include a marked effacement of splenic architecture and advanced infiltration of the liver
and other organs that were the most probable cause of death in affected animals. The severity of
the observed leukemia was increased in the high-dose group relative to controls. Of the
leukemias observed in each group, 5/9 (56%), 8/15 (53%), and 14/22 (64%) were classified as
Stage 3 in the control, low-, and high-dose groups, respectively.
Table 1. Incidences of Neoplastic Lesions in Male and Female F344/N Rats Given Gavage
Doses of Hydroquinone for 103 Weeks3
Tumor Type
0 mg/kg-d
25 mg/kg-d
50 mg/kg-d
Males
Renal tubule cell adenoma
0/5 5b (0%)
4/55 (7%)
8/55° (14%)
Females
Mononuclear cell leukemia
9/5 5b (16%)
15/55 (27%)
22/55d (40%)
aNTP, 1989
bp < 0.005 by logistic regression trend test
cp < 0.005 by logistic regression pairwise test
dp < 0.01 by logistic regression pairwise test
No compound-related clinical signs or effects on survival were observed in the mice
(NTP, 1989). Body weight was slightly reduced (less than 10% reduction from controls) in
100 mg/kg-day males over the last 10 weeks of the study. Body weight was reduced 10—15%
throughout the second year of the study in 100 mg/kg-day females. Small increases in relative
liver weight in the 100 mg/kg-day males and females are consistent with the reduced body
weight in these groups. An increased incidence of nonneoplastic hepatic lesions was observed
after 103 weeks in males at 100 mg/kg-day (anisokaryosis, syncytial alterations, basophilic foci).
Follicular cell hyperplasia of the thyroid gland was significantly increased in male and
particularly female mice at both dose levels. Incidences of thyroid follicular cell hyperplasia in
the control, low-, and high-dose groups were 5/55, 15/53, and 19/54 in the males and 13/55,
47/55, and 45/55 in the females, indicating that the low dose of 50 mg/kg-day is a LOAEL in this
mouse study.
9
-------
FINAL
9-8-2009
The NTP (1989) concluded that there was some evidence of carcinogenic activity of
hydroquinone in the female mice, as shown by increases in liver hepatocellular neoplasms,
mainly adenomas (Table 2). Incidences of hepatocellular adenomas and combined incidences of
hepatocellular adenomas or carcinomas in the female mice were significantly elevated in both
treated groups. The incidence of combined liver tumors in both dose groups exceeded the
historical range for untreated (9.1 ± 4.7%, max = 20%) or water gavage (8.3 ± 5.0%,
max = 14%>) controls. The incidence of hepatocellular adenomas was also increased in treated
male mice, but this increase was offset by a decrease in hepatocellular carcinoma in the treated
males (Table 2) so that the combined incidence of hepatocellular adenoma or carcinoma was not
increased in the treated males. Historical incidences for combined liver tumors in untreated and
water gavage controls averaged 30%>, but ranged as high as 58%>, for male mice. The NTP
concluded that there was no evidence of carcinogenicity in exposed male mice.
Table 2. Incidences of Neoplastic Lesions in Male and Female B6C3F1 Mice Given Gavage
Doses of Hydroquinone for 103 Weeks3
Tumor Type
0 mg/kg-d
50 mg/kg-d
100 mg/kg-d
Males
Hepatocellular adenoma
9/5 5 b
21/54°
20/55d
Hepatocellular carcinoma
13/55
11/54
7/55
Hepatocellular adenoma or carcinoma
20/55 (36%)
29/54 (54%)
25/55 (45%)
Females
Hepatocellular adenoma
2/5 5e
15/55f
12/5 5f
Hepatocellular carcinoma
1/55
2/55
2/55
Hepatocellular adenoma or carcinoma
3/55d (5%)
16/55e (29%)
13/55e (24%)
aNTP, 1989
hp < 0.05 by logistic regression trend test
°p < 0.01 by logistic regression pairwise test
dp < 0.05 by logistic regression pairwise test
ep < 0.01 by logistic regression trend test
fp < 0.005 by logistic regression pairwise test
A chronic feeding study in rats was conducted by Carlson and Brewer (1953). Groups of
10 male and 10 female weanling Sprague-Dawley rats were exposed to 0, 0.1, 0.5, or 1.0% (0,
1000, 5000, or 10,000 ppm) hydroquinone (purity not reported) in the diet for 103 weeks
(Carlson and Brewer, 1953). Using chronic reference values for food consumption and body
weight in Sprague-Dawley rats (U.S. EPA, 1988b), the diets provided estimated doses of 0, 69,
344, or 688 mg/kg-day in males and 0, 80, 399, or 799 mg/kg-day in females. Endpoints
included body weight, limited hematology (RBC counts, percent hemoglobin, and differential
white blood cell counts; other indices not reported) at unspecified times during the study, and
limited histology (liver, kidney, spleen, bone marrow, and 8 other tissues) at the end of the study.
Terminal body weights in the treated rats were not significantly different from the controls,
although rats in the mid- and high-dose groups reportedly gained weight more slowly than
controls during the first month of the study (e.g., 10 g per week in high-dose rats versus 22-27 g
per week in controls). No hematological or histopathological changes were found. Some of the
exposed males and females were mated with each other after 6 months of exposure and two
successive litters were produced. The average number of offspring in control and treated groups
10
-------
FINAL
9-8-2009
were almost identical (data not presented). The offspring (16-23/sex) were fed hydroquinone
that had been heated with lard at 190°C for 30 minutes before being incorporated into the diet in
concentrations of 0, 0.1, 0.25, or 0.5% for 103 weeks. The offspring showed normal growth
rates; no other details were reported. The transient effect on body weight gain in the mid- and
high-dose groups is not considered adverse; the high-dose (688 mg/kg-day in males and
799 mg/kg-day in females) is a NOAEL in this study.
Effects from chronic dietary exposure were studied in rats and mice by Shibata et al.
(1991). Groups of F344 rats (30/sex/group) and B6C3F1 mice (30/sex/group) were exposed to 0
or 0.8% hydroquinone in the diet for 104 or 96 weeks, respectively. Reported average
hydroquinone intakes were 351 and 368 mg/kg-day in the male and female rats and 1046 and
1486 mg/kg-day in the male and female mice, respectively. Clinical signs, food and water
intake, and body weight were assessed throughout the study. Evaluations performed at the end
of the exposure period included gross necropsy, liver and kidney weights, and comprehensive
histopathology. Effects in the exposed rats included reduced body weight gain in females (final
weight 7.5%) less than controls,/? < 0.05), increased absolute and relative liver (49%> absolute,
55%o relative,/? < 0.01) and kidney (69%> absolute, 74%> relative,/? < 0.01) weights in males,
increased relative kidney weights in females (10%>,/? < 0.01), and increased severity of
age-related chronic nephropathy in males (13/30 treated rats with moderate or severe lesions vs.
0 controls,/? < 0.01) and females (8/30 treated rats with slight lesions vs. 1/30 controls,
/? < 0.05). The prevalence and severity of the nephropathy was more severe in the males than
females; the nephropathy was not of the alpha 2-u-globulin type, as determined by the study
authors. Other renal changes in the male rats included increased incidences of renal hyperplasia
(100%o compared to 3%> in controls) and adenomas. The treatment level in rats of 351 mg/kg-day
is a LOAEL based on increased severity of age-related nephropathy and increased liver and
kidney weights. A 47%> increased incidence of renal adenomas in the male rats (14/30 vs. 0/30
in controls,/? < 0.01) was observed.
Effects in the exposed mice included reduced body weight gain in females (final weight
27.6%o less than controls,/? < 0.01) and increased relative liver (46%>,/? < 0.01) and kidney
(41%,/? < 0.01) weights in females, hepatic centrilobular hypertrophy (26/30 treated mice vs.
0/30 controls,/? < 0.01), and renal tubular hyperplasia (9/30 treated mice vs. 0/30 controls,
/? < 0.01) in the males, and forestomach hyperplasia in both males (11/30 treated mice vs.
1/30 controls,/? < 0.01) and females (14/30 treated mice vs. 3/30 controls,/? < 0.01) (Shibata et
al., 1991). The treatment level in mice of 1046 mg/kg-day is a LOAEL based on pathology and
increased liver and kidney weights. No statistically significant neoplastic changes in the kidneys
were found in mice of either sex, although the incidences of hepatocellular adenomas were
increased 25%> in the exposed males (14/30 vs. 6/28 in controls,/? < 0.05). No statistically
significant increased incidences of tumors of any type were observed in the female rats or mice.
A long-term study in dogs was performed by Woodard (1951, as summarized by
U.S. EPA, 1987). There were 4 mongrel dogs (2 males and 2 females) that were treated with a
single 100 mg/kg dose of hydroquinone by stomach tube. Treatment was interrupted because
this initial dose caused swelling of the eyes. The dogs were divided into two groups of one male
and one female 10 days after the initial dose and administered daily capsules containing 25 or
50 mg/kg on a 6 days/week regimen for 809 days; 2 untreated dogs served as controls. Body
weight measurements and blood counts were performed during the study and gross and
histological examinations were performed at the end of the experiment. Treatment was
suspended on Days 20-73 in one 25 mg/kg-day dog due to weight loss and on Days 238-309 for
11
-------
FINAL
9-8-2009
all dogs to assess body weight gain effects. No significant changes in weight gain or blood
counts were observed overall. Bone marrow hyperplasia and excessive pigment deposits in the
spleen were observed in all exposed dogs, but it is not clear if these lesions were also observed in
the controls (U.S. EPA, 1987), precluding reliable identification of a LOAEL.
Reproductive and Developmental Toxicity Studies—The effect of hydroquinone on rat
teratogenicity was reported by Eastman Kodak Co. (1984). Groups of 10 pregnant COBS-CD
BR rats were given gavage doses of hydroquinone (purity unspecified) of 0, 50, 100, or
200 mg/kg-day in water on Gestational Days 6 through 15. Clinical observations were made
twice daily, while body weights and food consumption were measured on Days 1, 6, 9, 12, and
15. On Day 16, the rats were anesthetized, exsanguinated, and subjected to necropsy and gross
pathology examinations. Maternal blood samples underwent hematological (hematocrit,
hemoglobin concentration, total white and red blood cell, nucleated RBC and platelet counts,
mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin
concentration, and RBC morphology) and clinical chemistry (serum levels of alanine
aminotransferase, alkaline phosphatase, urea nitrogen, glucose, creatinine, and sorbitol
dehydrogenase) measurements. Liver and kidney weights were recorded. Implantation sites and
corpora lutea were counted. Individual fetuses were weighed.
Brownish, dose-related urine discoloration was noted (Eastman Kodak Co., 1984). Body
weights were not significantly affected by treatment, but food consumption was significantly
lower on Gestational Days 6-9. Hematological and clinical chemistry parameters were not
affected by treatment. Although sorbitol dehydrogenase levels in the 50 mg/kg-day dose group
were significantly higher than controls, this finding was not considered treatment-related by the
U.S. EPA, because the higher dose groups were not similarly affected. Gross and histological
pathology were unremarkable. Relative kidney weights in the 50 and 200 mg/kg-day groups
were slightly, but significantly, increased (6-7%) over controls. Since the increase was small
and not dose-related, the toxicological significance of this finding is unknown. No
treatment-related effects were seen on pregnancy rate, number of corpora lutea, implantations
per dam, viable fetuses, fetal resorptions per litter, or fetal weight. This study identified a
NOAEL of 200 mg/kg-day for maternal and fetal toxicity.
Groups of 30 pregnant COBS-CD BR rats were administered hydroquinone (>99% pure)
in aqueous solution by gavage in doses of 0, 30, 100, or 300 mg/kg-day on Days 6-15 of
gestation (Krasavage et al., 1992). The maternal rats were euthanized on Gestation Day 20.
Maternal endpoints included food consumption, body weight, clinical signs, gross appearance of
the thoracic and abdominal viscera, and liver and kidney weights and histology. Developmental
endpoints included implantation and resorption indices, corpora lutea per dam, viable and dead
fetuses per litter, pre- and postimplantation losses, gravid uterine and fetal weights, sex ratio,
external abnormalities, internal soft tissue abnormalities (approximately half the fetuses in each
litter examined), and skeletal malformations and variations (remaining half of the fetuses). No
compound-related effects on reproductive or teratogenicity indices were observed. The
300 mg/kg-day dose level produced a slight reduction in maternal body weight gain and feed
intake with a concomitant slight reduction in mean fetal body weight (6.3% less than controls,
p < 0.05). The total number of fetuses with vertebral variations was slightly, but significantly,
increased at 300 mg/kg-day (6.7% greater than controls, p < 0.05), but the study authors did not
consider this finding to be toxicologically significant. The incidence of vertebral variations was
not statistically increased on a litter basis. There is no evidence indicating that hydroquinone
was selectively toxic to the developing rats; based on this study, 300 mg/kg-day is a minimal
12
-------
FINAL
9-8-2009
LOAEL and 100 mg/kg-day a NOAEL for both maternal and developmental toxicity,
respectively (slight reductions in maternal and fetal body weight).
A range-finding developmental toxicity study was conducted in which groups of 5 mated
New Zealand White rabbits were administered 0, 50, 100, 200, 300, 400, or 500 mg/kg-day by
gavage on Gestation Days 6-18 (Murphy et al., 1992). The range-finding study found
slight-to-moderate decreases in the mean maternal weight gain and feed consumption at 50 and
100 mg/kg-day and effects clearly indicative of maternal toxicity at the higher doses, including
weight loss at >200 mg/kg-day, tremors at >300 mg/kg-day and mortality at 500 mg/kg-day.
There was no developmental toxicity at 50 or 100 mg/kg-day, although results suggestive (but
not statistically significant) of fetotoxicity (reduced fetal body weight and slightly increased
resorptions with no effect on litter size) were observed at 200 mg/kg-day.
In the definitive study (Murphy et al., 1992), groups of 18 mated New Zealand White
rabbits were administered hydroquinone (100% pure) in aqueous solution by gavage at doses of
0, 25, 75, or 150 mg/kg-day on Days 6-18 of gestation (Murphy et al., 1992). The maternal
rabbits were euthanized on Gestation Day 30. Maternal endpoints included food consumption,
body weight, clinical signs, gross postmortem condition, and liver and kidney weights.
Developmental endpoints included implantation and resorption indices, corpora lutea per dam,
viable and dead fetuses per litter, pre- and postimplantation losses, gravid uterine and fetal
weights, and sex ratio; all fetuses were examined for external abnormalities, visceral
malformations/variations, and skeletal malformations/variations. No adverse maternal or
developmental effects were observed at 25 mg/kg-day. Significantly reduced feed consumption
on Gestation Days 12 and 13 (26 and 36% less than controls,/? < 0.05), but no effects on body
weight of the dams or developmental toxicity, occurred at 75 mg/kg-day. Effects at
150 mg/kg-day included significantly reduced maternal body weight gain and food consumption
and slight, nonsignificant increases in the incidences of ocular and minor skeletal malformations
(microophthalmia, vertebral/rib defects, angulated hyoid arch) on a per-fetus and a per-litter
basis. The dose of 150 mg/kg-day is identified as a maternal and developmental LOAEL for
reduced maternal body weight gain and slight increases in fetal ocular and skeletal
malformations, respectively. The dose of 75 mg/kg-day is a NOAEL for both maternal and
developmental toxicity.
A two-generation study was performed in which groups of 30 Sprague-Dawley F0 and Fi
rats of each sex were treated with 0, 15, 50, or 150 mg/kg-day of hydroquinone by gavage in
water on 7 days/week for at least 70 days prior to cohabitation and subsequently throughout
mating, gestation, and lactation until scheduled termination (3-4 weeks after the mating period
in males and following weaning in females) (Blacker et al., 1993). Gavage exposure of the Fi
generation was initiated at 25 days of age. Evaluations included clinical signs, body weight and
feed consumption in parental animals throughout the study, as well as fertility, number of live
pups/litter, sex ratio, pup weights, and gross external abnormalities in pups during the lactation
period. All parental animals were necropsied, and histological examinations were performed on
the reproductive tissues and the pituitary gland in all Fo and Fi control and high-dose animals, as
well as in all F0 and Fi parents in the low- and mid-dose groups that failed to produce a litter.
Gross external and internal examinations were performed on all pups culled on Day 4 of
lactation, Fi pups not selected as parental animals, and F2 pups and intact pups found dead at
birth or during lactation. No adverse effects on feed consumption, survival, or reproduction
parameters were found in the F0 or Fi parental animals. Mild transient tremors occurred shortly,
but infrequently, after dosing at 50 mg/kg-day in 1 Fo male and at 150 mg/kg-day in a number of
13
-------
FINAL
9-8-2009
F0 rats (11 males, 13 females) and Fi rats (5 males, 16 females). The tremors were considered to
reflect an acute effect on the nervous system. Body weight gain was reduced slightly in Fi males
at 50 and 150 mg/kg-day (approximately 6% less than controls in both dose groups,/? < 0.05)
exposures, but body weight was unaffected in Fo males or females of either generation. This
study identified a LOAEL of 150 mg/kg-day and a NOAEL of 50 mg/kg-day based on acute
neurological effects (mild transient tremors in a number of parental animals in both the Fo and Fi
generations). There was no evidence for reproductive toxicity of hydroquinone at doses up to
150 mg/kg-day.
Inhalation Exposure
No information was located regarding effects of inhaled hydroquinone in animals.
Other Studies
Toxicokinetics
The metabolism of hydroquinone appears to be very similar in humans and rodents
(IARC, 1999). The compound is metabolized mainly to sulfate and glucuronide conjugates, with
glucuronidation in human liver microsomes being somewhat less than in mouse but greater than
in rat microsomes (IARC, 1999). A small percentage can be converted to 1,4-benzoquinone by
several cellular enzymes, particularly macrophage peroxidases. In rats receiving gavage doses of
up to 350 mg/kg, the majority of hydroquinone was recovered as glucuronides (45-53%) and
O-sulfate conjugates (19-33%) in the urine, with a small fraction metabolized to
1,4-benzoquinone and then to the hydroquinone mercapturate (<5%) (English and Deisinger,
2005). 1,4-Benzoquinone is a very reactive metabolite that can be conjugated with glutathione
or form DNA adducts (IARC, 1999). Such adducts have been identified in promyelocytic HL-60
cell cultures. Similar macrophage peroxidase-mediated formation of 1,4-benzoquinone seems to
be important in the myelotoxicity of benzene (IARC, 1999).
Genotoxicity
The genotoxicity of hydroquinone has been extensively tested. As evaluated by IARC
(1999) and summarized below, hydroquinone is genotoxic in many in vitro systems using a
variety of endpoints. Hydroquinone caused gene mutations in Salmonella typhimurium strains
TA104 and TA102 (strains sensitive to oxidative mutagens) and induced gene conversion and
mutations in Saccharomyces cerevisiae, although it did not induce sex-linked recessive lethal
mutations in Drosophila melanogaster. In cultured rodent and human cells, hydroquinone
induced DNA strand breaks, gene mutations, chromosomal aberrations, sister chromatid
exchanges, and micronuclei formation; and hydroquinone inhibits intercellular communication.
Hydroquinone also caused micronuclei and chromosomal aberrations in mouse bone marrow
cells and chromosomal aberrations and hyperploidy in mouse spermatocytes, after intraperitoneal
injection (IARC, 1999). Hydroquinone-derived DNA adducts were not observed in F344 rat
kidney cells (English et al., 1994) following 6 weeks of oral (gavage) dosing at nephrotoxic
levels.
More recent genotoxicity literature corroborates the genotoxic effects summarized by
IARC (1999). Genotoxic effects were detected in human (supF forward mutation via DNA
adduction or gene deletion in embryonic adenovirus-transformed kidney cells) and animal
(micronucleus formation in Chinese hamster V79 cells) in vitro assay systems
(von der Hude et al., 2000; Silva et al., 2003; Nakayama et al., 2004; Gaskell et al., 2004,
2005a,b). The effects on meiotic segregation, exhibited as meiotic nondisjunction, were
observed in treated oocytes of Drosophila melanogaster (Munoz and Barnett, 2000). Roza et al.
14
-------
FINAL
9-8-2009
(2003) reported that hydroquinone treatment of cultured human lymphocytes did not result in
clastogenic effects, while Silva et al. (2004) identified the expression of the polymorphism
GSTM1 as protective against hydroquinone-induced micronuclei in the same cell type.
Hydroquinone treatment of human whole blood cultures resulted in a particular form of
aneuploidy (loss of chromosomes 5 and 7) associated with development of acute myeloid
leukemia (Zhang et al., 2005).
Initiation/Promotion
A number of studies were performed in which oral administration of hydroquinone was
predominantly inactive in promoting neoplasms initiated by other chemicals (IARC, 1999).
Administration of 0.2% hydroquinone in the diet for 22 weeks caused no increase in bladder
lesions in rats when given alone or following initiation by 0.05%
A'-nitrosobutyl-A'-(4-hydroxybutyl)amine in the drinking water for 2 weeks (Miyata et al., 1985).
Likewise, administration of 0.8% hydroquinone in the diet for 36 weeks did not induce bladder
tumors in rats when given alone, nor did it increase the incidence or multiplicity of bladder
tumors initiated by 0.05% A'-nitrosobutyl-A'-(4-hydroxybutyl)amine in the drinking water for
4 weeks (Kurata et al., 1990).
Administration of 0.8% hydroquinone in the diet for 49 weeks caused no increase in the
incidence of upper digestive tract tumors when given alone or following initiation by
6 intraperitoneal injections of 25 mg/kg of A'-nitrosornethyl-//-arnylamine, although the
multiplicity of esophageal carcinomas was increased in rats given the initiator (Yamaguchi et al.,
1989). Administration of 0.8% hydroquinone in the diet for 51 weeks caused no increase in
forestomach or glandular stomach lesions in rats when given alone or following initiation by a
single 150 mg/kg gavage dose of jV-methyl-jV-nitro-jV-nitrosoguanidine (Hirose et al., 1989).
Administration of hydroquinone in dietary doses of 100 or 200 mg/kg for 6 weeks
following initiation by partial hepatectomy and intraperitoneal injection of 300 mg/kg of
A'-nitrosodi ethyl amine caused an increase in the multiplicity of liver enzyme-altered
(y-glutamyltranspeptidase) foci in rats, but the response was not dose-related (Stenius et al.,
1989). Rats that underwent the same hepatectomy/A'-nitrosodi ethyl amine regimen to initiate
liver carcinogenesis, but were promoted with 1 mg/kg of hydroquinone by oral gavage on a
5 days/week regimen for 7 weeks, had no increase in the multiplicity of enzyme-altered foci,
although the area and volume of the foci were increased (Stenius et al., 1989). Dietary
administration of 0.8% hydroquinone for 36 weeks did not induce preneoplastic or neoplastic
liver or kidney lesions in rats when given alone, although this exposure did increase the
multiplicity of renal cell tumors and microadenomas that were initiated by 0.1%
A'-nitrosoethyl-A'-hydroxyethylamine in their drinking water for 3 weeks (Kurata et al., 1990).
Dietary administration of 1.5% hydroquinone for 16 weeks did not induce neoplastic lesions in
the pancreas or liver of hamsters when administered alone or following initiation by two
70 mg/kg subcutaneous injections of /V-nitrosobis(2-oxopropyl)amine, although administration
of hydroquinone after the initiator reduced the multiplicity of pancreatic lesions (Maruyama et
al., 1991). Finally, dietary administration of 0.8% hydroquinone for 30 weeks did not induce
tumors in rats when given alone, nor did it promote thyroid, lung, kidney, or bladder tumors that
were initiated by 0.1% A'-nitroso-bis(2-hydroxypropyl)amine in their drinking water for 2 weeks
(Hasegawa et al., 1990).
Dermal application and bladder implantation studies of hydroquinone are reviewed by
IARC (1977), NTP (1989), and U.S. EPA (1987). Hydroquinone was inactive as a complete
15
-------
FINAL
9-8-2009
dermal carcinogen, skin cocarcinogen, or initiator of skin carcinogenesis in dermal application
studies in mice (Roe and Salaman, 1955; Van Duuren and Goldschmidt, 1976). Bladder
implantation of hydroquinone in cholesterol pellets increased the incidence of bladder
carcinomas in mice (Boyland et al., 1964).
DERIVATION OF PROVISIONAL SUBCHRONIC AND CHRONIC
ORAL RfD VALUES FOR HYDROQUINONE
The systemic toxicity of repeated oral exposures to hydroquinone has been evaluated in
one subchronic human study (Carlson and Brewer, 1953); several subchronic and chronic studies
in rats (Carlson and Brewer, 1953; Delcambre et al., 1962; Christian et al., 1980; NTP, 1989;
Shibata et al., 1991); subchronic and chronic studies in mice (NTP, 1989; Shibata et al., 1991),
hamsters (Hirose et al., 1986), and dogs (Carlson and Brewer, 1953); and reproductive and
developmental toxicity studies in rats and rabbits (Eastman Kodak Co., 1984; Krasavage et al.,
1992; Murphy et al., 1992; Blacker et al., 1993) (Table 3). The human study identifies a
NOAEL of 4.3 mg/kg-day for lack of hematological and renal effects in 17 subjects who
ingested hydroquinone in 3 divided doses with meals for 3-5 months. The human NOAEL
reflects normal results in blood analyses (RBC count, hematocrit, percent hemoglobin,
differential white blood cell count, sedimentation rate, platelet count, coagulation time, and iteric
index) and urinalyses (albumin, reducing sugars, white and red cell counts, casts, and
urobilinogen), as determined by comparison of each subject with preexposure values.
The erythrocyte, kidney, and thyroid gland are the most sensitive targets of hydroquinone
toxicity in the animal studies. Studies showing hematologic effects, including RBC changes in
rats, identified a NOAEL of 7.5 mg/kg-day and minimal LOAEL of 15 mg/kg-day (for
anisocytosis, polychromatophilia, and erythroblastosis) following gavage exposure on a
6 days/week regimen for 40 days (Delcambre et al., 1962). A NOAEL of 25 mg/kg-day and
LOAEL of 50 mg/kg-day (for decreased hematocrit value, hemoglobin concentration, and
erythrocyte count) was also identified following gavage exposure on a 5 days/week regimen for
15 months (NTP, 1989). These findings are consistent with the known toxicity of hydroquinone
toward bone marrow, as illustrated by the development of aplastic anemia in rats exposed to
3333 mg/kg-day in the diet for 9 weeks (Carlson and Brewer, 1953), as well as mononuclear cell
leukemia in rats exposed to 50 mg/kg-day by gavage on a 5 days/week regimen for 103 weeks
(NTP, 1989). Hydroquinone is a metabolite of benzene and is suspected of playing a role in the
myelodepressive and leukemogenic activity of benzene (IARC, 1999). Effects of hydroquinone
on bone marrow have been demonstrated in mechanistic studies performed in vitro or by acute
parenteral administration (IARC, 1999).
16
-------
FINAL
9-8-2009
Table 3. Noncancer Effects and Effect Levels Identified from Studies of Oral (drinking
water, dietary, and gavage dosing) Hydroquinone Exposure to Humans and Animals
Source
Species
Exposure duration /
sub-route
NOAEL
mg/kg-day
LOAEL
mg/kg-day
Effect
Subchronic Studies
Carlson and
Brewer, 1953
Human
3-5 months,
diet
4.3
ND
No dose-related hematological
or renal effects
Christian et al.,
1980
Rat
8 weeks,
drinking water
230
390
Decreased body weight gain,
increased relative liver and
kidney weight
Christian et al.,
1980
Rat
15 weeks,
drinking water
ND
110
Increased relative liver and
kidney weight
Carlson and
Brewer, 1953
Rat
9 weeks,
diet
ND
3333
FEL: decreased body weight,
mutlitissue atrophy, aplastic
anemia, stomach mucosa
hemorrhage
Hirose et al., 1986
Hamster
20 weeks,
diet
474
ND
No dose-related histopathology
Carlson and
Brewer, 1953
Dog
26 weeks,
diet
100
ND
No dose-related effects
Carlson and
Brewer, 1953
Rat
12 days,
gavage
ND
500
FEL: Death
Carlson and
Brewer, 1953
Rat
22 weeks,
gavage
ND
500
FEL: Death
Delcambre et al.,
1962
Rat
40 days,
gavage
7.5
15
Anisocytosis,
polychromatophilia,
erythroblastosis
Topping et al.,
2007
Rat
13 weeks,
gavage
20
64
Acute neurological effects
(tremors, reduced activity)
NTP, 1989
Rat
13 weeks,
gavage
50
100
Decreased body weight (8-9%),
renal tubule degeneration
NTP, 1989
Mouse
13 weeks,
gavage
100
200
FEL: tremors, forestomach
hyperplasia and ulceration,
death
Chronic Studies
NTP, 1989
Rat
103 weeks,
gavage
25
50
Hematological changes,
increased severity of age-related
nephropathy, decreased body
weight
NTP, 1989
Mouse
103 weeks,
gavage
ND
50
Thyroid follicular cell
hyperplasia
Carlson and
Brewer, 1953
Rat
103 weeks,
diet
688
ND
No dose-related effects
17
-------
FINAL
9-8-2009
Table 3. Noncancer Effects and Effect Levels Identified from Studies of Oral (drinking
water, dietary, and gavage dosing) Hydroquinone Exposure to Humans and Animals
Source
Species
Exposure duration /
sub-route
NOAEL
mg/kg-day
LOAEL
mg/kg-day
Effect
Shibata et al., 1991
Rat
104 weeks,
diet
ND
351
Increased relative liver and
kidney weights, increased
severity of age-related
nephropathy
Shibata et al., 1991
Mouse
96 weeks,
diet
ND
1046
Increased relative liver and
kidney weight, forestomach
hyperplasia, renal tubular
hyperplasia
Reproductive and developmental effects
Blacker et al., 1993
Rat
Two generations,
gavage
50
150
Acute neurological effects
(tremors). No
repro/developmental effects
observed at any dose
Eastman Kodak
Co., 1984
Rat
Gestational Days
6-15 (terminated at
Day 15), gavage
200
ND
No repro/developmental effects
observed
Krasavage et al.,
1992
Rat
Gestational Days
6-15 (terminated at
Day 20), gavage
100
300
Decreased fetal weight
Murphy et al., 1992
Rabbit
Gestational Days
6-18 (terminated at
Day 30), gavage
75
150
Increased fetal ocular and
skeletal malformations
ND = Not determined
Kidney lesions included tubular cell degeneration in rats with subchronic gavage
exposure at a LOAEL of 100 mg/kg-day (NTP, 1989), an increased severity of age-related toxic
nephropathy in rats with chronic exposure at a LOAEL of 50 mg/kg-day for gavage exposure
(NTP, 1989), and a LOAEL of 351 mg/kg-day for dietary exposure (Shibata et al., 1991). The
nephropathic changes were more severe in male rats than female rats, but they were not
associated with hyaline droplet formation. Increases in renal tubular cell hyperplasia and
adenomas were also observed (NTP, 1989; Shibata et al., 1991).
Follicular cell hyperplasia of the thyroid gland was increased in male and female mice
that were exposed to hydroquinone by gavage on a 5 days/week regimen for 103 weeks (NTP,
1989); the LOAEL is 50 mg/kg-day and no NOAEL was identified. The toxicological
significance of this effect is unclear because thyroid and pituitary hormone levels were not
assessed, thyroid hyperplasia was not observed in the only other chronic study of hydroquinone
(Shibata et al., 1991), and incidences of follicular cell neoplasia were not increased (NTP, 1989;
Shibata et al., 1991).
A two-generation study found no effects on fertility and reproduction in rats exposed to
hydroquinone at doses <150 mg/kg-day (Blacker et al., 1993). Developmental toxicity studies
found no evidence indicating that hydroquinone was selectively toxic to the developing
organism. In rats treated during gestation, 300 mg/kg-day was a LOAEL of minimal
18
-------
FINAL
9-8-2009
toxicological significance and 100 mg/kg-day aNOAEL for both maternal and developmental
toxicity (slight reductions in maternal and fetal body weight) (Krasavage et al., 1992). In rabbits,
150 mg/kg-day was a LOAEL and 75 mg/kg-day a NOAEL for reduced maternal body weight
gain and slightly increased fetal ocular and skeletal malformations (Murphy et al., 1992). These
data indicate that reproductive and developmental toxicity are not critical effects of concern for
hydroquinone.
Studies, summarized above, yielded several potential points of departure (POD) for
comparison. The human study identified a NOAEL of 4.3 mg/kg-day for lack of hematological
and renal effects in subjects who ingested hydroquinone for 3-5 months (Carlson and Brewer,
1953). A LOAEL of minimal toxicological significance (15 mg/kg-day) for RBC changes
(Delcambre et al., 1962) and aNOAEL and LOAEL (50 and 100 mg/kg-day, respectively) for
kidney toxicity were identified for subchronic exposure to hydroquinone in rats (NTP, 1989).
The reliability of the 15 mg/kg-day subchronic hematological LOAEL is questionable due to
limited information in the available summary of the study and a lack of corroborating effects on
RBCs in rats chronically exposed to a higher (25 mg/kg-day) dose level of hydroquinone (NTP,
1989). The chronic LOAEL for hematological effects in rats, kidney toxicity in rats, and thyroid
follicular cell hyperplasia in mice is 50 mg/kg-day (NTP, 1989; Shibata et al., 1991).
The Carlson and Brewer (1953) study was selected as the basis for the POD because it
reflects a human assessment of the two main targets of hydroquinone, as observed in animals,
and is supported by the 50 mg/kg-day subchronic NOAEL for kidney effects (NTP, 1989) and
the 25 mg/kg-day chronic NOAEL for hematologic effects in animals (NTP, 1989; Shibata et al.,
1991). The subchronic NOAEL for hematological and renal effects in humans (4.3 mg/kg-day)
is also the highest NOAEL below all identified LOAELs (Carlson and Brewer, 1953). Because
of a lack of sufficient data for benchmark dose (BMD) modeling, the human NOAEL is the most
appropriate basis for subchronic and chronic p-RfD derivation.
A subchronic p-RfD is derived by applying a composite Uncertainty Factor (UF) of 10
to the POD, the subchronic human NOAEL of 4.3 mg/kg-day. UFs are applied to the POD for
low-dose extrapolation when specific data are lacking or insufficient. The composite UF of 10 is
composed of an UF of 10 that is applied to account for variation in human sensitivity. An UF for
extrapolation from animals to humans is not applied because a human study is available; an UF
for extrapolating from a LOAEL to a NOAEL is not applied because a NOAEL is available; and
an UF for database deficiencies is not applied because numerous well designed subchronic and
chronic studies in animals are available, including developmental studies in multiple species and
a mutligeneration reproduction study.
Subchronic p-RfD = NOAEL ^ composite UF
= 4.3 mg/kg-day -M0
= 0.4 mg/kg-day or 4 x 10"1 mg/kg-day
A chronic p-RfD is similarly derived by applying a composite UF of 100 to the
subchronic human NOAEL. The composite UF is composed of the following two component
factors: An UF of 10 is applied to account for variation in human sensitivity; and an UF of 10 is
applied for extrapolation from subchronic to chronic exposure. An UF for animal-to-human
extrapolation is not applied because a human study is available; an UF for extrapolation from a
LOAEL is not applied because a NOAEL is available; and an UF for database deficiencies is not
applied due to the robust database of available studies.
19
-------
FINAL
9-8-2009
Chronic p-RfD = NOAEL composite UF
= 4.3 mg/kg-day -MOO
= 0.04 mg/kg-day or 4 x 10"2 mg/kg-day
Confidence in the key study (Carlson and Brewer, 1953) is low. Although blood analyses
and urinalyses were performed and showed no indications of adverse hematological and renal
effects in humans, the study is poorly reported with minimal details, only a limited number of
relevant indices were tested, the number of subjects is marginal, and only one dose level was
adequately tested. Confidence in the database is high because subchronic, chronic, and
developmental toxicity have been adequately tested in two species and reproductive toxicity has
been evaluated in a mutligeneration study. Medium confidence in the subchronic and chronic
p-RfD values follows.
FEASIBILITY OF DERIVING PROVISIONAL SUBCHRONIC AND CHRONIC
INHALATION RfC VALUES FOR HYDROQUINONE
The derivation of p-RfC values for hydroquinone is precluded by inadequate information
on inhalation toxicity. Human data were limited to case reports of corneal lesions in workers
with unquantified exposure to mixtures of hydroquinone dust and quinone vapor and a cohort
mortality study that showed no increase in mortality from noncancer causes in a cohort of
workers exposed to hydroquinone for a mean duration of 13.7 years (Pifer et al., 1995). No
subchronic or chronic duration inhalation studies of hydroquinone in animals were located.
PROVISIONAL CARCINOGENICITY ASSESSMENT
FOR HYDROQUINONE
Weight-of-Evidence Descriptor
Inadequate information is available on the carcinogenicity of hydroquinone in humans
from two occupational studies. Standardized mortality ratios for total cancer and site-specific
cancers were not elevated in a cohort of 879 workers who were exposed in a plant in which
hydroquinone was manufactured and used (Pifer et al., 1995). This study is limited by a weak
power to detect effects, due to the relatively small cohort size and small numbers of deaths from
site-specific cancers. An increased number of malignant melanoma cases was observed in a
cohort of 836 lithographers—about 200 of whom had worked regularly by hand with
photographic chemicals and were exposed to hydroquinone (Nielsen et al., 1996). This study is
limited by a small number of cases (the excess of malignant melanoma was based on 5 cases,
only 2 of which had reported exposure to hydroquinone), as well as mixed chemical exposures to
various pigments, dyes, and organic solvents used in lithography printing processes.
Chronic oral carcinogenicity tests of hydroquinone were conducted in two-dose-level
studies in rats and mice (NTP, 1989) and in single-dose-level studies in rats and mice
(Shibata et al., 1991). Hydroquinone induced significantly increased incidences of renal tubule
adenomas in male rats in both studies, hepatocellular adenomas in female mice in the NTP
(1989) study, and hepatocellular adenomas in male mice in the Shibata et al. (1991) study.
20
-------
FINAL
9-8-2009
Incidences of mononuclear cell leukemia were significantly greater than controls in female rats
in the NTP (1989) study, although the incidences in the exposed groups were just within the
historical control range.
Orally-administered hydroquinone was predominantly inactive in promoting neoplasms
initiated by other chemicals (Hasegawa et al., 1990; Hirose et al., 1989; Kurata et al., 1990;
Miyata et al., 1985; Maruyama et al., 1991; Stenius et al., 1989; Yamaguchi et al., 1989).
Hydroquinone was inactive as a complete dermal carcinogen, skin cocarcinogen, or initiator of
skin carcinogenesis in dermal application studies in mice (Roe and Salaman, 1955; Van Duuren
and Goldschmidt, 1976), although bladder implantation of hydroquinone in cholesterol pellets
increased the incidence of bladder carcinomas in mice (Boyland et al., 1964).
Hydroquinone was genotoxic in many in vitro systems using a variety of endpoints
(IARC, 1999). Hydroquinone induced gene mutations in some strains of Salmonella
typhimurium, gene conversion and mutations in Saccharomyces cerevisiae; and DNA strand
breaks, gene mutations, chromosomal aberrations, sister chromatid exchanges, and micronuclei
in cultured rodent and human cells. In vivo effects induced by hydroquinone included
micronuclei and chromosomal aberrations in mouse bone marrow cells and spermatocytes. In
addition, structurally similar DNA adducts have been observed in bone marrow following either
in vitro hydroquinone or in vivo benzene exposure in mice (Pathak et al., 1995). No
hydroquinone-related DNA adducts were detected in the kidneys of rats gavage-dosed with up to
50 mg/kg-day for up to six weeks (English et al., 1994).
Although the positive findings of carcinogenicity in animals were not unequivocally
treatment-related (leukemia) or indicated development of benign tumors (renal tubular and
hepatocellular adenomas), evidence of hydroquinone-induced cancer was identified in male and
female mice and rats in two separate studies. Further, mutagenicity was clearly exhibited in
in vitro studies of bacterial and mammalian cell cultures, including human kidney cells. Under
the U.S. EPA (2005) cancer guidelines, these data provide evidence that hydroquinone is "Likely
to be Carcinogenic in Humans"
Mode-of-Action Discussion
There are no data for hydroquinone carcinogenicity in humans. Available evidence in
animals shows that oral exposure to hydroquinone is carcinogenic, producing significant
increases in the incidence of kidney tumors in male rats, liver tumors in male and female mice,
and mononuclear cell leukemia in female rats. The MOA for these tumor types in animals is not
known. However, hydroquinone has consistently given positive results in standard prokaryotic
and eukaryotic tests of mutagenicity and other types of genotoxicity in animals and human cell
cultures. Consequently, a mutagenic MOA is plausible for both renal cancer and mononuclear
cell leukemia in rats and liver cancer in mice but data are inadequate to make such a
determination.
Renal Cancer
Based on the available data, the MOA for hydroquinone-induced renal adenomas in male
rats cannot currently be determined. While a single study failed to find hydroquinone-DNA
adducts in the kidneys of rats (English et al., 1994), the genotoxicity database for this chemical is
consistently positive for mutagenicity (IARC, 1999). Therefore, a mutagenic MOA is possible
but a determination cannot be made given the available data.
21
-------
FINAL
9-8-2009
Key Events—The available cancer bioassay data show that high doses of hydroquinone
produce statistically significant increases in renal adenomas in male rats, but not in females or in
mice (NTP, 1989; Shibata et al., 1991). This sex and species specificity suggests the possibility
of alpha 2u-globulin-induced cancer as the MOA. As summarized in the Risk Assessment
Forum Review of alpha 2u-globulin renal toxicity and neoplasia (U.S. EPA, 1991b), the first step
in a sequence of events leading to the alpha 2u-globulin-mediated renal tumor formation in male
rats is excessive accumulation of hyaline droplets containing alpha 2u-globulin in renal proximal
tubules. Both the NTP (1989) gavage study and the Shibata et al. (1991) feeding study reported
the lack of hyaline droplet accumulation in the renal tubules. Thus, renal adenomas in male rats
resulting from different doses (50 and 351 mg/kg-day) and sub-routes of exposure (gavage and
diet) do not appear to be alpha 2u-globulin-mediated. No other data identifying key events for
renal tumor development are available.
Strength, Consistency, Specificity of Association—Information to support a genotoxic
MOA for hydroquinone includes hydroquinone-induced mutagenicity in several in vitro assays
using bacterial strains and animal and human cell cultures.
Dose-Response Concordance—Although dose-response concordance for tumor
development was observed in mice and rats (NTP, 1989), there are no dose-response data for
precursor events, precluding an assessment of dose-response concordance for a mutagenic MOA.
In the studies of NTP (1989) and Shibata et al. (1991), the extent of age-related nephropathy is
exacerbated by hydroquinone treatment. A reanalysis of the histology data from the NTP (1989)
study found that renal tubular adenomas were located in areas of severe chronic progressive
nephropathy (Hard et al., 1997). However, the increase in nephropathy-severity across the dose
groups (dose-related incidence of moderate or marked severity: 69%, 71%, and 85%) did not
show high concordance with renal adenoma incidence (0%, 7%, and 15%).
Biological Plausibility and Coherence—The plausibility and coherence of a mutagenic
MOA for hydroquinone-induced renal adenomas in rats is provided by several positive in vitro
genotoxicity assays in bacterial, animal, and human culture systems. The basis for the observed
sex-specific difference in tumorigenic response is not currently known. Based on the positive
mutagenic response to cell culture systems across species, the human relevance of
hydroquinone-induced renal adenomas in male rats is assumed.
Conclusions—The available data on the increased incidences of renal adenomas in
hydroquinone-exposed male rats are considered suitable for quantitative cancer assessment. The
available data do not support a definitive MOA, but the largely positive results from the
mutagenicity assays suggest that a mutagenic MOA is plausible.
Liver Cancer
The MOA for hydroquinone-induced liver adenomas and/or carcinomas in mice is not
known. However, the genotoxicity database for this chemical is consistently positive for
mutagenicity (IARC, 1999). Therefore, a mutagenic MOA is possible but data are not available
to make such a determination.
Key Events—No key events have been identified leading to the development of
hydroquinone-induced hepatic tumors in mice. The nonneoplastic lesions observed in mice by
NTP (1989) and Shibata et al. (1991) are indicative of nuclear alterations, but not cytotoxicity.
Shibata et al. (1991) reported significant increases in the incidence of hepatocyte hypertrophy
22
-------
FINAL
9-8-2009
and foci of cellular alterations occurring with hepatic adenoma and carcinoma. Nonneoplastic
lesions reported by NTP (1989) in male mice included anisokaryosis and syncitial alteration
(more than 5 nuclei per cell). Basophilic foci, lesions that are considered to be a precursor to
adenomas, have been observed in male and female mice. The observed nuclear alterations and
preneoplastic lesions and the lack of observable cytotoxicity support a mutagenic MOA for liver
carcinogenicity in mice.
Strength, Consistency, Specificity of Association—As with renal tumor development in
rats, information to support a genotoxic MOA for hydroquinone includes hydroquinone-induced
mutagenicity in several in vitro assays using bacterial strains and animal and human cell cultures.
Dose-Response Concordance—There are currently no dose-response data for key events
for hydroquinone-induced liver tumors in mice, precluding an assessment of dose-response
concordance. The available tumor data did not show an increase in incidence at the high-dose
group in comparison to the low-dose group; response rates were similar in both groups (NTP,
1989).
Biological Plausibility and Coherence—The plausibility and coherence of a mutagenic
MOA for liver adenomas or carcinomas in hydroquinone-exposed mice is based on several
positive in vitro genotoxicity assays and lack of observable hepatic cytotoxicity. The basis for
the observed species-specific difference in tumor development is not currently known. Based on
the positive mutagenic response to cell culture systems across species, the human relevance of
hydroquinone-induced liver cancer in mice is assumed.
Conclusions—The available data for increased incidences of liver adenomas or
carcinomas in hydroquinone-exposed female mice (NTP, 1989, based on positive effects at two
dose levels) are considered suitable for quantitative cancer assessment. The available data do not
support a definitive MOA, but the consistently positive results from mutagenicity assays suggest
that a mutagenic MOA is possible.
Mononuclear Cell Leukemia
The NTP (1989) reported that chronic gavage exposure to hydroquinone significantly
increased the incidence of mononuclear cell leukemia in female F344 rats relative to concurrent
(but not historical) controls. Development of mononuclear cell leukemia was not observed in
other species exposed to hydroquinone. Mononuclear cell leukemia (also called "large
granular lymphocytic [LGL] leukemia" or "Ty leukemia") is a spontaneous, rapidly fatal
neoplasm that is common in aged, untreated F344 rats (Haseman et al., 1998). The MOA for
induction of mononuclear cell leukemia in rats is unknown. Hence, the MOA analysis in this
case is applied in an attempt to address the question of human relevance of rat tumor responses.
Key Events—The MOA of development of mononuclear cell leukemia in rats is
currently unknown, precluding review and discussion of key events.
Strength, Consistency, Specificity of Association—Mononuclear cell leukemia is a
spontaneous, rapidly fatal neoplasm that is common (average incidence of 28.1%) in aged,
untreated F344 rats in 2-year carcinogenicity studies conducted by NTP (Haseman et al., 1998);
it is, uncommon in other strains of laboratory rats and unknown in mice. For this reason, some
pathologists regard it as a unique cancer that is not relevant to humans. In the case of phthalates,
Caldwell (1999) argued that the increased incidences of mononuclear cell leukemia in F344 rats
23
-------
FINAL
9-8-2009
treated with phthalates are not relevant to humans because 1) the equivalent cell type (based on
morphological criteria) does not exist in humans or other animals, 2) the course of disease differs
between species, and 3) phthalates are not genotoxic and induce leukemia only at high dose
levels. However, the species-dependent arguments are weakened by a reliance on relatively old
literature (largely predating 1990) and an overly strict definition of the apparently related LGL
leukemia in humans. Hydroquinone differs from the phthalates in that the results of
mutagenicity and other genotoxicity assays have largely been positive.
Dose-Response Concordance—Dose-response data for key events in development of
mononuclear cell leukemia in hydroquinone-exposed rats do not currently exist, precluding an
assessment of dose-response concordance. However, the severity of observed leukemia
(stages 1, 2, or 3) increased with increasing dose thus making the data suitable for modeling.
Biological Plausibility and Coherence—In a brief discussion of mononuclear cell
leukemia, CHAPDP (2001) noted that the human correlate to rat mononuclear cell leukemia is
chronic Ty lymphoproliferative disease, which is characterized by abnormal expansion of large
granular lymphocytes (LGL). Patients with this disease are predominately older males, who
exhibit disease-related changes in lymphocytes; bone marrow and the spleen that are reported to
be morphologically, functionally, and clinically similar to mononuclear cell leukemia in rats.
Other reviews of human LGL leukemia have concluded that the disease has a diverse origin
(NK- or T-cells) and a wide spectrum of acute or chronic clinical presentations (Lamy and
Loughran, 1998). Canine LGL leukemias also may present as acute or chronic diseases—either
of which may be caused by NK- or T-cells (Vernau and Moore, 1999). The mode(s) of action
for induction of mononuclear cell leukemia in rats and LGL leukemia in humans and canines are
currently unknown, preventing an assessment of relevance to human cancer on this basis.
Conclusions—Based on the clinical and pathological similarities between chronic Ty
lymphoproliferative disease in humans and mononuclear cell leukemia in rats, increases in the
incidence of mononuclear cell leukemia in treated rats were considered relevant to human health
and suitable for quantitative cancer assessment of hydroquinone.
Quantitative Estimates of Carcinogenic Risk
Oral Exposure
Data for hydroquinone are sufficient to perform benchmark dose (BMD) modeling
(U.S. EPA, 2000). Modeling was performed based on the incidences of renal tubule adenomas
in male rats, mononuclear cell leukemia in female rats, and hepatocellular tumors (adenomas and
carcinomas combined) in female mice in the NTP (1989) study. The Shibata et al. (1991) data
are not amenable to dose-response modeling because the study included only a single dose level
in each species. Dose-response modeling of the NTP (1989) data (Table 4) was performed using
the methodologies in the U.S. EPA (2005) Guidelines for Carcinogen Risk Assessment. A linear
extrapolation is appropriate for all three tumor types as indicated above as the MOA(s) for these
tumors is generally unknown. In accordance with the 2005 cancer guidelines, the BMDLio
(lower bound on dose estimated to produce a 10% increase in tumor incidence over background)
for each tumor site was estimated using the U.S. EPA (2000) benchmark dose methodology.
The incidence data were analyzed using all available models for dichotomous data in the
benchmark dose software (BMDS) program (version 1.3.2) developed by U.S. EPA. Risk was
calculated as extra risk. Confidence bounds were automatically calculated by the BMDS using a
maximum likelihood profile method.
24
-------
FINAL
9-8-2009
Output from the BMDS program was evaluated using the criteria described in U.S. EPA
(2000). Goodness-of-fit was evaluated using the chi-square statistic calculated by the BMDS
program. Acceptable global goodness-of-fit is indicated by a/;-value greater than or equal to
0.1. Models that did not meet these criteria were eliminated from consideration. Local fit is
evaluated visually on the graphic output by comparing the observed and estimated results at each
data point. BMDLio estimates that are within a factor of three are considered to show no model
dependence and are ranked using the AIC reported by the BMDS program. The model with the
lowest AIC is considered to provide a superior fit.
Table 4. Data Selected for BMD Modeling of Cancer Incidence in Rats and Mice Given
Gavage Doses of Hydroquinone for 103 Weeksa
Tumor Type
Species
Sex
0
mg/kg-d
18
mg/kg-d
36
mg/kg-d
Renal tubule adenoma
rat
male
0/55
4/55
8/55
Mononuclear cell leukemia
rat
female
9/55
15/55
22/55
0
mg/kg-d
36
mg/kg-d
71
mg/kg-d
Hepatocellular adenoma or carcinoma
mouse
female
3/55
16/55
13/55
Doses listed are daily average doses (gavage dose x 5days/7 days)
aNTP, 1989
Modeling results are shown in Table 5 and Figures A-l and A-2. For mononuclear cell
leukemia, fits of the gamma, log logistic, and Weibull models to the data could not be evaluated
due to the availability of only three dose groups—these models each have three estimable
parameters and the inability of the BMD software to provide an initial specification for one of
the three model parameters. Thus, too few degrees of freedom were available for calculation of a
p-walue with which to evaluate model fit. This did not occur for the renal or liver tumor data.
None of the available models resulted in an adequate fit to the hepatocellular tumor data in mice
(Table 5).
Table 5. Multistage Benchmark Dose Modeling Results for Rats and Mice
Exposed to Hydroquinone by Gavage for 2 Yearsa'b
Incidence
Species
Strain
Sex
AIC
/>-Valuce
BMD10
(mg/kg-day)
BMDL10
(mg/kg-day)
Renal Tubule Adenomas'1
Rats
F344/N
M
76.2961
0.9979
24.4596
15.7456
Mononuclear Cell Leukemiad
Rats
F344/N
F
191.571
0.8016
11.7719
7.3221
Hepatocellular Adenoma or
Carcinoma (Combined)6
Mice
B6C3FJ
F
157.99
0.0358
25.3683
16.4603
aNTP, 1989
bBest fitting model(s) in bold text
°Values <0.1 fail to meet conventional goodness-of-fit criteria
dBetas restricted to >0; polydegree = 1 (lowest degree polynomial with adequate fit)
"Betas restricted to >0; polydegree = 1 (no adequate fit at any polydegree; higher polydegrees default to 1 degree)
Abbreviations: AIC = Akaike Information Criterion; BMDi0 = maximum likelihood estimate of the dose producing
a 10% extra risk of effect; BMDLio = 95% lower confidence limit on the BMD10
25
-------
FINAL
9-8-2009
Human equivalent doses (BMDLio hed) were calculated for each animal BMDLio using
U.S. EPA's cross-species scaling factor of body weight raised to the 3/4 power (U.S. EPA,
2005). Adjustment from animal-to-human administered dose is performed by multiplying the
animal BMDLio by the ratio of animal-to-human body weight raised to the 1/4 power. The
BMDLio hed represents the chronic daily dose (mg/kg-d) expected to result in 10% extra risk for
tumor development extrapolated from the animal bioassay data. The BMDLio hed values for
renal tumors in male rats and mononuclear cell leukemia in female rats are shown in Table 6.
Comparison of the BMDLio hed values in Table 6 shows that mononuclear cell leukemia
in female rats is a more sensitive endpoint than renal tubule adenoma in male rats because it
occurs at a lower dose.
Table 6. BMDLi0s and Human Equivalent Doses from Models Adequately Fit to
Incidence Data for Tumors in Animals Chronically Treated with Hydroquinone
Test Group
Tumor Location
& Type
/j-valuc
BMDL10
(mg/kg-day)
BMDL10 hed3
(mg/kg-day)
Male Rats
renal tubule adenoma
0.9979
15.8
4.4
Female Rats
mononuclear cell leukemia
0.8016
7.3
1.8
Female Mice
Hepatocellular adenoma or
carcinoma (combined)
0.035 (failed)
16.46
2.5
aHuman cancer equivalent dose of the BMDL10 calculated as animal BMDL10 x (W:imm(l|/W|lllm:m) where Whuman =
70 kg (human reference body weight) and W:iMlnl(l = 0.416 kg for male rats and 0.273 kg for female rats (time
weighted average body weights in the study).
In order to linearly extrapolate cancer risks from the BMDLio hed to the origin, a cancer
oral slope factor (OSF) was calculated as the ratio 0.1/BMDLio hed- Taking the BMDLio hed of
1.8 mg/kg-day for mononuclear cell leukemia in female rats as the POD, a provisional OSF of
0.06 (mg/kg-day)"1 is calculated as follows:
p-OSF = 0.1 BMDLio hed
= 0.1 ^ 1.8 mg/kg-day
= 0.06 (mg/kg-day)"1
The OSF for hydroquinone should not be used with exposures exceeding the POD
(BMDLio hed =1.8 mg/kg-day) because above this level the fitted dose-response model better
characterizes what is known about the carcinogenicity of hydroquinone.
Inhalation Exposure
No data are currently available for the quantitative estimate of cancer risk following
inhalation exposure to hydroquinone.
26
-------
FINAL
9-8-2009
REFERENCES
ACGIH (American Conference of Governmental Industrial Hygienists). 2005. Threshold Limit
Values for Chemical Substances and Physical Agents and Biological Exposure Indices. ACGIH,
Cincinnati, OH.
Anderson, B. 1947. Corneal and conjunctival pigmentation among workers engaged in
manufacture of hydroquinone. AMA Arch. Ophthalmol. 38:812-826. (Cited in U.S. EPA,
1987).
Anderson, B. and F. Oglesby. 1958. Corneal changes from quinonehydroquinone exposure.
AMA Arch. Ophthalmol. 59:495-501. (Cited in U.S. EPA, 1987).
AT SDR (Agency for Toxicol ogical Substances Disease Registry). 2006. Internet HazDat
Toxicological Profile Query. U.S. Department of Health and Human Services, Public Health
Service. Atlanta, GA. Online, http://www.atsdr.cdc.gov/gsql/toxprof.script.
Blacker, A.M., R.E. Schroeder, J.C. English et al. 1993. A two-generation reproductive study
with hydroquinone in rats. Fund. Appl. Toxicol. 21(4):420-424.
Boyland, E., E.R. Busby, C.E. Dukes et al. 1964. Further experiments on implantation of
materials into the urinary bladder of mice. Br. J. Cancer. 18:575-581. (Cited in IARC, 1977).
Caldwell, D.J. 1999. Review of mononuclear cell leukemia (MNCL) in F-344 rat bioassays and
its significance to human cancer risk: A case study using alkyl phthalates. Reg. Toxicol.
Pharmacol. 30:45-53.
Carlson, A.J. and N.R. Brewer. 1953. Toxicity studies on hydroquinone. Proc. Soc. Exp. Biol.
Med. 84:684-688.
CHAPDP (Consumer Hazard Advisory Panel on Diisononyl Phthalate). 2001. Report to the
U.S. Consumer Product Commission. Bethesda: U.S. Product Safety Commission Directorate
for Health Sciences. Online, http://www.cpsc.gov/librarv/foia/foia01/os/dinp.pdf.
Christian, R.T., C.S. Clark, T.E. Cody et al. 1980. The development of a test for the potability
of water treated by a direct reuse system. Washington, DC: U.S. Army Medical Research and
Development Command. 20314. Contract No. DADA-17-73-C-3013. University of Cincinnati,
Cincinnati, OH. (Cited in U.S. EPA, 1987).
DeCaprio, A.P. 1999. The toxicity of hydroquinone - relevance to occupational and
environmental exposure. Crit. Rev. Toxicol. 29(3):283-330.
Delcambre, J.P., B. Weber and C. Baron. 1962. Toxicite'de l'hydroquinone. Rev.
Agressologic III. 2:311-315. [Article in French] (Cited in NIOSH, 1978; U.S. EPA, 1987).
Eastman Kodak Co. 1984. Hydroquinone: Teratology probe study in rats. TSCA Section 8D
submission. OTS0206392.
27
-------
FINAL
9-8-2009
English, J.C., T. Hill, J.L. O'Donoughe et al. 1994. Measurement of nuclear DNA modification
32
by P-postlabeling in the kidneys of male and female Fisher 344 rats after multiple gavage doses
of hydroquinone. Fund. Appl. Toxicol. 23:391-396.
English, J.C. and P.J. Deisinger. 2005. Metabolism and disposition of hydroquinone in Fischer
344 rats after oral or dermal administration. Food Chem. Toxicol. 43:483-493.
Gaskell, M., K.I.E. McLuckie and P.B. Farmer. 2004. Comparison of the mutagenic activity of
the benzene metabolites, hydroquinone and para-benzoquinone in the supF forward mutation
assay: a role for minor DNA adducts formed from hydroquinone in benzene mutagenicity. Mut.
Res. 554:387-398.
Gaskell, M., K.I.E. McLuckie and P.B. Farmer. 2005a. Comparison of the repair of DNA
damage induced by the benzene metabolites hydroquinone and p-benzoquinone: a role for
hydroquinone in benzene genotoxicity. Carcinogenesis 26:673-680.
Gaskell, M., K.I.E. McLuckie and P.B. Farmer. 2005b. Genotoxicity of the benzene metabolites
para-benzoquinone and hydroquinone. Chem. Biochem. Interact. 153-154:267-270.
Hard, G.C., J. Whysner, J.C. English et al. 1997. Relationship of hydroquinone-associated rat
renal tumors with spontaneous chronic progressive nephropathy. Toxicol. Pathol. 25:132-143.
Hasegawa, R., R. Furukawa, K. Toyoda et al. 1990. Inhibitory effects of antioxidants on
N-bis(2-hydroxypropyl)nitrosamine-induced lung carcinogenesis in rats. Jpn. J. Cancer Res.
81:871-877.
Haseman, J.K., J.R. Hailey and R.W. Morris. 1998. Spontaneous neoplasm incidences in
Fischer 344 rats and B6C3F1 mice in two-year carcinogenicity studies: a National Toxicology
Program update. Toxicol. Pathol. 26(3):428-441.
Hirose, M., T. Inoue, M. Asamoto et al. 1986. Comparison of the effects of 13 phenolic
compounds in induction of proliferative lesions of the forestomach and increase in the labeling
indices of the glandular stomach and urinary bladder epithelium of Syrian Golden hamsters.
Carcinogenesis. 7(8): 1285-1289.
Hirose, M., M. Yamaguchi, S. Fukushima et al. 1989. Promotion by dihydroxybenzene
derivative of N-methyl-N-nitro-N-nitrosoguanidine-induced F344 rat forestomach and glandular
stomach carcinogenesis. Cancer Res. 49:5143-5147.
IARC (International Agency for Research on Cancer). 1977. Dihydroxybenzenes. Some
fumigants, the herbicides 2,4-D and 2,4,5-T, chlorinated dibenzodioxins and miscellaneous
industrial chemicals. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans.
IARC (International Agency for Research on Cancer). 1999. Hydroquinone. Re-evaluation of
some organic chemicals, hydrazine and hydrogen peroxide. IARC Monographs on the
Evaluation of Carcinogenic Risks to Humans.
Krasavage, W.J., A.M. Blacker, J.C. English et al. 1992. Hydroquinone: A developmental
toxicity study in rats. Fund. Appl. Toxicol. 18(3):370-375.
28
-------
FINAL
9-8-2009
Kurata, Y., S. Fukishima, R. Hasegawa et al. 1990. Structure-activity relations in promotion of
rat urinary bladder carcinogenesis by phenolic antioxidants. Jpn. J. Cancer Res. 81(8):754-759.
Lamy, T. and T.P. Loughran, Jr. 1998. Large granular lymphocyte leukemia. Online.
http://www.moffitt.usf.edu/pubs/cci/v5nl/article3.html.
Maruyama, H., T. Amanuma, D. Nakae et al. 1991. Effects of catechol and its analogs on
pancreatic carcinogenesis initiated by N-nitrosobis(2-oxopropyl)amine in Syrian hamsters.
Carcinogenesis. 12:1331-1334. (Cited in IARC, 1999).
Miller, S.J.H. 1954. Ocular ochronosis. Trans. Ophthalmol. Soc. UK. 74:349-366. (Cited in
U.S. EPA, 1987).
Miyata, Y., S.S. Fukushima, M. Hirose et al. 1985. Short-term screening of promoters of
bladder carcinogenesis in N-butyl-N-(4-hydroxybutyl)nitrosamine-initiated, unilaterally
ureter-ligated rats. Jpn. J. Cancer Res. (Gann) 76:828-834.
Munoz, E.R. and B.M. Barnett. 2000. Effect of hydroquinone on meiotic segregation in
Drosophilia malongaster females. Mut. Res. 469:215-221.
Murphy, S.J., R.E. Schroeder, A.M. Blacker et al. 1992. A study of developmental toxicity of
hydroquinone in the rabbit. Fund. Appl. Toxicol. 19(2):214-221.
Naumann, G. 1966. Corneal damage in hydroquinone workers. A clinicopathologic study.
Arch. Ophthalmol. 76(2): 189-194. (Cited in U.S. EPA, 1987).
Nakayama, A., Y. Noguchi, T. Mori et al. 2004. Comparison of mutagenic potentials and
mutation spectra of benzene metabolites using supF shuttle vectors in human cells. Mutagen.
19:91-97.
NIOSH (National Institute for Occupational Safety and Health). 1978. Criteria for a
Recommended Standard Occupational Exposure to Hydroquinone. NIOSH, Cincinnati, OH.
NTIS PB 81-226508. (Cited in U.S. EPA, 1987).
NIOSH (National Institute for Occupational Safety and Health). 2005. NIOSH Pocket Guide to
Chemical Hazards. Online, http://www.cdc.gov/niosh/npg/npe.html.
Nielsen, H., L. Henriksen and J.H. Olsen. 1996. Malignant melanoma among lithographers.
Scand. J. Work Environ. Health. 22:108-111.
NTP (National Toxicology Program). 1989. Toxicology and carcinogenesis studies of
hydroquinone (CAS No. 123-31-9) in F344/N rats and B6C3F1 mice (gavage studies). NTP TR
366. NM Publication No. 90-2821.
OSHA (Occupational Safety and Health Administration). 2007. Regulations for air
contaminants (Standards 29 CFR 1910.1000 and 1915.1000).
29
-------
FINAL
9-8-2009
Pathak, D.N., G. LeVay and W.J. Bodell. 1995. DNA adduct formation in the bone marrow of
B6C3F1 mice treated with benzene. Carcinogenesis. 16:1803-1808. (Cited in DeCaprio,
1999).
Pifer, J.W., F.T. Hearne, F.A. Swanson et al. 1995. Mortality study of employees engaged in
the manufacture and use of hydroquinone. Int. Arch. Occup. Environ. Health. 67(4):267-280.
Roe, F.J.C. and M.H. Salaman. 1955. Further studies on incomplete carcinogenesis: triethylene
melamine (T.E.M.), 1,2-benzanthracene and P-propiolactone as initiators of skin tumour
formation in the mouse. Br. J. Cancer. 9:177-203. (Cited in IARC, 1977).
Roza, L., N. de Vogel, J.H.M. van Delft. 2003. Lack of clastogenic effects in cultured human
lymphocytes treated with hydroquinone. Food Chem. Toxicol. 41:1299-1305.
Shibata, M.-A., M. Hirose, H. Tanaka et al. 1991. Induction of renal cell tumors in rats and
mice, and enhancement of hepatocellular tumor development in mice after long-term
hydroquinone treatment. Jpn. J. Cancer Res. 82(11):1211—1219.
Silva, M.C., J. Gaspar, I.D. Silva et al. 2003. Mechanisms of induction of chromosomal
aberrations by hydroquinone in V79 cells. Mutagen. 18:491-496.
Silva, M.C., J. Gaspar, I.D. Silva et al. 2004. GSTM1, GSTT1, and GSTP1 genotypes and the
genotoxicity of hydroquinone in human lymphocytes. Environ. Mol. Mutagen. 43:258-264.
Stenius, U., M. Warholm, A. Rannug et al. 1989. The role of GSH depletion and toxicity in
hydroquinone-induced development of enzyme-altered foci. Carcinogenesis. 10(3):593-599.
Sterner, J.H., F.L. Oglesby and B. Anderson. 1947. Quinone vapors and their harmful effects.
I. Corneal and conjunctival injury. J. Ind. Hyg. Toxicol. 29:60-73. (Cited in U.S. EPA, 1987).
Topping, D.C., L.G. Bernard, J.L. O'Donoghue et al. 2007. Hydroquinone: acute and
subchronic toxicity studies with emphasis on neurobehavioral and nephrotoxic effects. Food
Chem. Toxicol. 45:70-78.
U.S. EPA. 1987. Health and Environmental Effects Document for p-Hydroquinone. Prepared
by the Office of Health and Environmental Assessment, Environmental Criteria and Assessment
Office, Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington, DC
EC AO-CIN-GO15.
U.S. EPA. 1988a. Reportable Quantity Document for Hydroquinone. Prepared by the Office of
Health and Environmental Assessment, Environmental Criteria and Assessment Office,
Cincinnati, OH, for the Office of Solid Waste and Emergency Response, Washington, DC.
U.S. EPA. 1988b. Recommendations for and Documentation of Biological Values for Use in
Risk Assessment. Prepared by the Office of Health and Environmental Assessment,
Environmental Criteria and Assessment Office, Cincinnati, OH for the Office of Solid Waste and
Emergency Response, Washington, DC. EPA/600/6-87/008. NTIS PB 88-179874.
U.S. EPA. 1991a. Chemical Assessments and Related Activities (CARA). Office of Health and
Environmental Assessment, Washington, DC. April.
30
-------
FINAL
9-8-2009
U.S. EPA. 1991b. Alpha 2u-globulin: Association with Chemically Induced Renal Toxicity and
Neoplasia in the Male Rat. Risk Assessment Forum, Washington, DC. EPA/625/3-91/019F.
U.S. EPA. 1994. Chemical Assessments and Related Activities (CARA). Office of Health and
Environmental Assessment, Washington, DC. December.
U.S. EPA. 1997. Health Effects Assessment Summary Tables (HEAST). FY-1997 Update.
Prepared by the Office of Research and Development, National Center for Environmental
Assessment, Cincinnati, OH for the Office of Emergency and Remedial Response, Washington,
DC. July. EPA/540/R-97/036. NTIS PB 97-921199.
U.S. EPA. 2000. Benchmark Dose Technical Guidance Document. Risk Assessment Forum,
National Center for Environmental Assessment, Washington, DC. External Review Draft.
October 2000. EPA/630/R-00/001.
U.S. EPA. 2005. Guidelines for carcinogen risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001F. Federal Register 70(66): 17765-17817. Online.
http://www.epa.eov/raf.
U.S. EPA. 2006. Drinking Water Standards and Health Advisories. Summer 2006. Office of
Water, Washington, DC. Online, http://www.epa.gov/waterscience/criteria/drinkine/
dwstandards.pdf.
U.S. EPA. 2007. Integrated Risk Information System (IRIS). Online. Office of Research and
Development, National Center for Environmental Assessment, Washington, DC.
http ://www. epa. gov/iri s/.
Van Duuren, B.L. and B.M. Goldschmidt. 1976. Cocarcinogenic and tumor-promoting agents
in tobacco carcinogenesis. J. Natl. Cancer Inst. 56:1237-1242. (Cited in NTP, 1989).
Vernau, W. and P.F. Moore. 1999. An immunophenotypic study of canine leukemias and
preliminary assessment of clonality by polymerase chain reaction. Vet. Immunol. Immunopath.
69:145-164.
von der Hude, W., S. Kalweit, G. Englehardt et al. 2000. In vitro micronucleus assay with
Chinese hamster V79 cells - results of a collaborative study with in situ exposure to 26 chemical
substances. Mut. Res. 468:137-163.
WHO (World Health Organization). 1994. Environmental Health Criteria 157. Hydroquinone.
Online, http://www.inchem.org/documents/ehc/ehc/ehc 157.htm.
Woodard, G.D.L. 1951. The toxicity, mechanism of action and metabolism of hydroquinone.
Dissertation, George Washington Univ., Washington, DC. (Cited in U.S. EPA, 1987).
Yamaguchi, S., M. Hirose, S. Fukushima et al. 1989. Modification by catechol and resorcinol of
upper digestive tract carcinogenesis in rats treated with methyl-n-amyl-nitrosamine. Cancer Res.
49:6015-6018.
31
-------
FINAL
9-8-2009
Zhang, L., W. Wang, A.E. Hubbard et al. 2005. Nonrandom aneuploidy of chromosomes 1, 5,
6, 7, 8, 9, 11, 12, and 21 induced by benzene metabolites hydroquinone and benzenetriol.
Environ. Mol. Mutagen. 45:388-396.
32
-------
FINAL
9-8-2009
APPENDIX A. DETAILS OF BENCHMARK DOSE MODELING
FOR HYDROQUIN ONE
Quantal Linear Model with 0.95 Confidence Level
Quantal Linear
BMD Lower Bound
0.3
0.25
0.2
CD
<
C
o
0.15
t5
ro
LL.
0.05
BMDL
BMD
0
10
20
30
40
50
Dose
16:42 02/19 2007
Figure A-l. Observed and Predicted Incidences of Renal Tubule Adenomas in Male Rats
Gavaged with Hydroquinone for 2 Years by NTP (1989)
33
-------
FINAL
9-8-2009
Multistage Model with 0.95 Confidence Level
0.55
Multistage
BMD Lower Bound
0.5
0.45
0.4
0.35
.2
<
c
o
0.3
t5
ro
LL.
0.25
0.2
0.15
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
11:41 02/22 2007
Figure A-2. Observed and Predicted Incidences of Mononuclear Cell Leukemia in Female
Rats Gavaged with Hydroquinone for 2 Years by NTP (1989)
34
-------
FINAL
9-8-2009
BMD Modeling of Rat and Mouse Tumor Data from 2-year Oral Hydroquinone Exposure (NTP
1989)
Part I. Male Rats: Renal Tubule Adenoma
Gamma model
$Revision: 2.2 $ $Date: 2001/03/14 01:17:00 $
Input Data File: C:\BMDS\HYDROQUINONE\KENAl_ADENOMA_GAMMA.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\RENAL_ADENOMA_GAMMA.plt
Mon Feb 19 16:24:04 2007
HMDS MODEL RUN
The form of the probability function is:
P[response]= backgrounds- (1-background)*CumGamma[slope*dose,power],
where CumGamma(.) is the cummulative Gamma distribution function
Dependent variable = renal_adenoma
Independent variable = Dose
Power parameter is restricted as power >=1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.00892857
Slope = 0.00618629
Power = 1.18252
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
Slope Power
Slope 1 0.99
Power 0.99 1
Parameter Estimates
Variable
Background
Slope
Power
Estimate
0
0.00512205
1.06717
Std. Err.
NA
0.0130769
1.03886
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
35
-------
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-37.1459
-37.1459
-43.0051
78.2918
Deviance Test DF
2.2952e-010
11.7185
1
2
P-value
0.002853
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.0000 0.000 0 55
18.0000 0.0727 4.000 4 55
36.0000 0.1455 8.000 8 55
Chi-square = 0.00 DF = 1 P-value = 1.0000
-5.573e-006
-1.409e-005
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
24.6551
15.7507
36
-------
FINAL
9-8-2009
Gamma Multi-Hit Model with 0.95 Confidence Level
Gamma Multi-Hit
BMD Lower Bound
0.3
0.25
0.2
0.15
0.05
0
BMDL
BMD
0
10
20
30
40
Dose
16:24 02/19 2007
37
-------
FINAL
9-8-2009
Log Logistic model
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\EMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 14:14:51 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose)) ]
Dependent variable = renal_M_resp
Independent variable = rat_dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0
intercept = -5.77649
slope = 1.11784
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -1
slope -1 1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
-5.77649
1.11784
Std. Err.
*
*
*
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
* - Indicates that this value is not calculated.
Warning: Likelihood for the fitted model larger than the Likelihood for the full
model.
38
-------
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-37.1459
-37.1459
-43.0051
78.2918
# Param's Deviance Test d.f. P-value
3
2 -1.42109e-014 1
1 11.7185 2 0.002853
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
18.0000
36.0000
ChiA2
0.0000
0.0727
0.1455
0.000
4.000
8.000
0.00
d.f.
P-value
1.0000
55
55
55
0.000
-0.000
-0.000
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 24.5807
BMDL = 15.0507
Log-Logistic Model with 0.95 Confidence Level
0.3
Log-Logistic
BMD Lower Bound
£
<
0.25
0.2
0.15
0.1
0.05
BMDL
BMD
14:19 04/18 2007
10
15
20
Dose
25
30
35
40
39
-------
FINAL
9-8-2009
Logistic model
Logistic Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:20 $
Input Data File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_LOGISTIC.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\RENAL_ADENOMA_LOGISTIC.plt
Mon Feb 19 16:31:37 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = 1/[1+EXP(-intercept-slope*dose)]
Dependent variable = renal_adenoma
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
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
background = 0 Specified
intercept = -4.45033
slope = 0.0830241
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.94
slope -0.94 1
Parameter Estimates
Variable
intercept
slope
Estimate
-4.34388
0.0742432
Std. Err
0.889063
0.0283803
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-37.1459
-38.238
-43.0051
Deviance Test DF
2.18422
11.7185
1
2
P-value
0.1394
0.002853
AIC:
80.476
40
-------
Dose
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.0128
0.0471
0.1583
1.59
0.705
2.590
8.705
DF
0
4
8
P-value
55
55
55
0.2076
-0.8451
0.8976
-0.2605
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
30.5583
25.0151
FINAL
9-8-2009
41
-------
FINAL
9-8-2009
Logistic Model with 0.95 Confidence Level
Logistic
BMD Lower Bound
0.3
0.25
0.2
0.15
0.05
BMDL
BMD
0
5
10
15
20
25
30
35
40
Dose
16:31 02/19 2007
42
-------
FINAL
9-8-2009
Multistage model: 2 degree polynomial
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_MS.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_MS.plt
Mon Feb 19 16:35:48 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2) ]
The parameter betas are restricted to be positive
Dependent variable = renal_adenoma
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 5.55112e-017
Beta(1) = 0.00402346
Beta(2) = 9.52234e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
Beta(l) Beta(2)
Beta(1) 1 -0.97
Beta(2) -0.97 1
Parameter Estimates
Variable
Background
Beta(1)
Beta(2)
Estimate
0
0.00402346
9.52238e-006
Std. Err.
NA
0.0160776
0.000487293
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
43
-------
FINAL
9-8-2009
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-37.1459
-37.1459
-43.0051
78.2918
Deviance Test DF
6.39488e-013
11.7185
1
2
P-value
0.002853
Dose
Goodness of Fit
Est._Prob. Expected Observed
Size
ChiA2 Res.
i: 1
0.0000
i: 2
18.0000
i: 3
36.0000
Chi-square
0.0000
0.0727
0.1455
0.00
0.000
4.000
8.000
DF = 1
55
55
55
0.000
-0.000
-0.000
P-value = 1.0000
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
24.7382
15.7507
44
-------
Multistage Model with 0.95 Confidence Level
FINAL
9-8-2009
Multistage
BMD Lower Bound
BMDL
16:35 02/19 2007
45
-------
FINAL
9-8-2009
Multistage model: 1 degree polynomial
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 14:24:26 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
The parameter betas are restricted to be positive
Dependent variable = renal_M_resp
Independent variable = rat_dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(1) = 0.00436627
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
Beta(1)
Beta(1) 1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Limit
Background 0 * * *
Beta(1) 0.00430754 * * *
* - Indicates that this value is not calculated.
46
-------
FINAL
9-8-2009
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-37.1459
-37.148
-43.0051
76.2961
# Param's Deviance Test d.f. P-value
3
1 0.00429126 2 0.9979
1 11.7185 2 0.002853
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
18.0000
36.0000
ChiA2 =0.00
0.0000
0.0746
0.1436
d.f. = 2
0.000 0 55
4.103 4 55
7.901 8 55
P-value = 0.9979
0.000
-0.053
0.038
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
BMDU =
0.1
Extra risk
0. 95
24.4596
15.7456
41.3552
Taken together, (15.7456, 41.3552) is a 90
interval for the EMD
two-sided confidence
47
-------
FINAL
9-8-2009
Multistage Model with 0.95 Confidence Level
Multistage
BMD Lower Bound
0.3
0.25
0.2
0.15
0.05
0
BMDL
BMD
0
10
20
30
40
50
Dose
14:24 04/18 2007
48
-------
FINAL
9-8-2009
Log Probit model
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: C:\BMDS\HYDROQUINONE\KENAl_ADENOMA_LOGPROBIT.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\RENAL_ADENOMA_LOGPROBIT.plt
Mon Feb 19 16:41:07 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+Slope*Log(Dose)),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = renal_adenoma
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial (and Specified) Parameter Values
background = 0
intercept = -4.34615
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept
intercept 1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0
-4.52904
1
Std. Err.
NA
0.162761
NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
49
-------
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-37.1459
-37.5373
-43.0051
77.0747
Deviance Test DF
0.782885
11.7185
P-value
0.6761
0.002853
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.0000
0.0506
0.1722
0.83
0.000
2.785
9.471
DF
0 55 0
4 55 0.747
8 55 -0.5253
P-value = 0.6590
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
25.7257
19.8319
FINAL
9-8-2009
50
-------
FINAL
9-8-2009
Probit Model with 0.95 Confidence Level
Probit
BMD Lower Bound
0.3
0.25
0.2
0.15
0.05
BMDL
BMD
0
5
10
15
20
25
30
35
40
45
Dose
16:41 02/19 2007
51
-------
FINAL
9-8-2009
Probit model
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_PROBIT.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\RENAL_ADENOMA_PROBIT.plt
Mon Feb 19 16:39:37 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = CumNorm(Intercept+Slope*Dose),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = renal_adenoma
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
background = 0 Specified
intercept = -2.61578
slope = 0.047858
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.92
slope -0.92 1
Parameter Estimates
Variable Estimate Std. Err.
intercept -2.33003 0.398082
slope 0.0371221 0.0134772
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-37.1459
-38.0507
-43.0051
Deviance Test DF
I.80952
II.7185
1
2
P-value
0.1786
0.002853
52
-------
FINAL
9-8-2009
AIC:
80.1013
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.0099 0.545 0 55 -0.7417
18.0000 0.0483 2.655 4 55 0.8461
36.0000 0.1602 8.811 8 55 -0.2981
Chi-square = 1.35 DF = 1 P-value = 0.2444
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
29.5699
23.8061
53
-------
FINAL
9-8-2009
Probit Model with 0.95 Confidence Level
Probit
BMD Lower Bound
0.3
0.25
0.2
0.15
0.05
BMDL
BMD
0
5
10
15
20
25
30
35
40
Dose
16:39 02/19 2007
54
-------
FINAL
9-8-2009
Quantal Linear model
Quantal Linear Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_QL.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_QL.plt
Mon Feb 19 16:42:30 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*dose)]
Dependent variable = renal_adenoma
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.00892857
Slope = 0.0043237
Power = 1 Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Power
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
Slope
Slope 1
Parameter Estimates
Variable Estimate Std. Err.
Background 0 NA
Slope 0.00430754 0.00124441
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-37.1459
-37.148
-43.0051
Deviance Test DF
0.00429126
11.7185
P-value
0.9979
0.002853
AIC:
76.2961
55
-------
FINAL
9-8-2009
Dose
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.0000
0.0746
0.1436
0.00
0.000
4.103
7.901
DF
0 55 0
4 55 -0.05303
8 55 0.03824
P-value = 0.9979
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
24.4596
15.7456
56
-------
FINAL
9-8-2009
-------
FINAL
9-8-2009
Quantal Quadratic model
Quantal Quadratic Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_QQ.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_QQ.plt
Mon Feb 19 16:43:51 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*doseA2)]
Dependent variable = renal_adenoma
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.00892857
Slope = 0.000120103
Power = 2 Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Power
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
Slope
Slope 1
Parameter Estimates
Variable Estimate Std. Err.
Background 0 NA
Slope 0.000144402 4.17271e-005
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-37.1459
-37.6677
-43.0051
Deviance Test DF
I.04354
II.7185
P-value
0.5935
0.002853
AIC:
77.3353
58
-------
FINAL
9-8-2009
Dose
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.0000
0.0457
0.1707
1.17
0.000
2.514
9.387
DF
0 55 0
4 55 0.9594
8 55 -0.4972
P-value = 0.5578
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
27.0117
21.6709
59
-------
FINAL
9-8-2009
Quantal Quadratic Model with 0.95 Confidence Level
0.3
0.25
0.2
"a
| 0.15
o
o
ro
0.1
0.05
0
Quantal Quadratic
BMDL
BMD
10 15 20 25 30 35 40
Dose
16:43 02/19 2007
60
-------
FINAL
9-8-2009
Weibull model
Weibull Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_WEIBULL.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\RENAL_ADENOMA_WEIBULL.plt
Mon Feb 19 16:46:28 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*doseApower)]
Dependent variable = renal_adenoma
Independent variable = Dose
Power parameter is restricted as power >=1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.00892857
Slope = 0.00192321
Power = 1.22607
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
Slope Power
Slope 1 -1
Power -1 1
Parameter Estimates
Variable Estimate Std. Err.
Background 0 NA
Slope 0.00354973 0.0105672
Power 1.05777 0.88385
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model Log(likelihood) Deviance Test DF P-value
Full model -37.1459
61
-------
Fitted model
Reduced model
AIC:
-37.1459
-43.0051
78.2918
2.31637e-011
11.7185
1
2
0.002853
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.0000
0.0727
0.1455
0.00
0.000
4.000
8.000
DF
0 55
4 55
8 55
P-value = 1.0000
1.236e-006
4.652e-006
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
24.6637
15.7507
FINAL
9-8-2009
62
-------
FINAL
9-8-2009
Weibull Model with 0.95 Confidence Level
Weibull
BMD Lower Bound
0.3
0.25
0.2
0.15
0.05
BMDL
BMD
0
5
10
15
20
25
30
35
40
45
Dose
16:46 02/19 2007
63
-------
FINAL
9-8-2009
Part II. Female Rats: Mononuclear Cell Leukemia
Gamma model
$Revision: 2.2 $ $Date: 2001/03/14 01:17:00 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_GAMMA.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\MNCL_GAMMA.plt
Mon Feb 19 16:51:11 2007
HMDS MODEL RUN
The form of the probability function is:
P[response]= backgrounds- (1-background)*CumGamma[slope*dose,power],
where CumGamma(.) is the cummulative Gamma distribution function
Dependent variable = MNCL
Independent variable = Dose
Power parameter is restricted as power >=1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.169643
Slope = 0.0163268
Power = 1.39303
Asymptotic Correlation Matrix of Parameter Estimates
Background Slope Power
Background 1 0.24 0.33
Slope 0.24 1 0.99
Power 0.33 0.99 1
Parameter Estimates
Variable
Background
Slope
Power
Estimate
0.163636
0.015406
1.33159
Std. Err
0.0498785
0.0300769
1.46377
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood) Deviance Test DF
-93.7538
-93.7538 4.80355e-010 0
-97.6478 7.78794 2
P-value
NA
0.02036
AIC:
193.508
64
-------
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.1636
0.2727
0.4000
0.00
9.000
15.000
22.000
DF = 0
9
15
22
P-value
55 2 .448e-006
55 1.149e-006
55 -2.175e-005
NA
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 14.4093
BMDL = 7.35611
Gamma Multi-Hit Model with 0.95 Confidence Level
0.5
0.4
0.3
0.2
0.1
Gamma Multi-Hit
BMD Lower Bound
BMDL
10
15 20
Dose
25
30
35
16:51 02/19 2007
65
-------
Log Logistic model
Logistic Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:20 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_LOGLOGISTIC.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\MNCL_LOGLOGISTIC.plt
Mon Feb 19 16:57:16 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose
Dependent variable = MNCL
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.163636
intercept = -5.92344
slope = 1.39301
Asymptotic Correlation Matrix of Parameter Estimates
background intercept slope
background 1 -0.38 0.31
intercept -0.38 1 -0.99
slope 0.31 -0.99 1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0.163636
-5.92344
1.39301
Std. Err
0.0498825
3.94306
1.12951
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-93.7538
-93.7538
-97.6478
Deviance Test DF
0
7.78794
P-value
NA
0.02036
AIC:
193.508
66
-------
FINAL
9-8-2009
Dose
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.1636
0.2727
0.4000
0.00
9.000
15.000
22.000
DF = 0
9
15
22
P-value
55
55
55
NA
-1.721e-014
1.076e-014
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
14.5114
6.26573
67
-------
FINAL
9-8-2009
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
BMD Lower Bound
0.5
0.4
0.3
0.2
0.1
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
16:57 02/19 2007
68
-------
FINAL
9-8-2009
Logistic model
Logistic Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:20 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_LOGISTIC.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\MNCL_LOGISTIC.plt
Mon Feb 19 16:55:35 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = 1/[1+EXP(-intercept-slope*dose)]
Dependent variable = MNCL
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
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
background = 0 Specified
intercept = -1.57728
slope = 0.0330592
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.83
slope -0.83 1
Parameter Estimates
Variable
intercept
slope
Estimate
-1.61418
0.0338498
Std. Err
0.318545
0.012471
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-93.7538
-93.7587
-97.6478
Deviance Test DF
0.00982294
7.78794
1
2
P-value
0.9211
0.02036
AIC:
191.517
69
-------
Dose
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square
0.1660
0.2680
0.4024
0.01
9.130
14.739
22.130
DF
9
15
22
P-value
55
55
55
0.9210
-0.04729
0.07946
-0.03588
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
15.1377
11.3236
FINAL
9-8-2009
70
-------
FINAL
9-8-2009
Logistic Model with 0.95 Confidence Level
Logistic
BMD Lower Bound
0.5
0.4
0.3
0.2
0.1
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
16:55 02/19 2007
71
-------
FINAL
9-8-2009
Multistage model: 2 degree polynomial
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:10:46 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2) ]
The parameter betas are restricted to be positive
Dependent variable = mcl_F_resp
Independent variable = rat_dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.163636
Beta(1) = 0.00630316
Beta(2) = 8.11882e-005
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l) Beta(2)
Background 1 -0.56 0.38
Beta(1) -0.56 1 -0.96
Beta(2) 0.38 -0.96 1
Parameter Estimates
Interval
Variable
Limit
Background
Beta(1)
Beta(2)
Estimate
0.163636
0.00630316
8.11882e-005
Std. Err.
*
*
*
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
* - Indicates that this value is not calculated.
Error in computing chi-square; returning 2
72
-------
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-93.7538
-93.7538
-97.6478
193.508
# Param's
3
3
1
Deviance Test d.f.
0
7.78794
P-value
NA
0.02036
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
18.0000
36.0000
0.1636
0.2727
0.4000
9.000
15.000
22.000
9
15
22
ChiA2
0.00
d.f.
P-value
NA
55
55
55
-0.000
0.000
0.000
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
BMDU =
0.1
Extra risk
0. 95
14.1401
7.35611
31.2005
Taken together, (7.35611, 31.2005) is a 90
interval for the EMD
two-sided confidence
73
-------
FINAL
9-8-2009
Multistage Model with 0.95 Confidence Level
Multistage
BMD Lower Bound
0.5
0.4
0.3
0.2
0.1
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
15:12 04/18 2007
74
-------
FINAL
9-8-2009
Multistage model: 1 degree polynomial
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_MSC.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\MNCL_MSC.plt
Thu Feb 22 11:41:43 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl)]
The parameter betas are restricted to be positive
Dependent variable = MNCL
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.156271
Beta(1) = 0.00922594
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.74
Beta(1) -0.74 1
Parameter Estimates
Variable
Background
Beta(1)
Estimate
0.160156
0.00895016
Std. Err
0.113667
0.00621176
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-93.7538
-93.7855
-97.6478
Deviance Test DF
0.0634144
7.78794
1
2
P-value
0.8012
0.02036
AIC:
191.571
75
-------
Dose
Goodness of Fit
Est. Prob. Expected Observed
Size
ChiA2 Res.
i: 1
0.0000
i: 2
18.0000
i: 3
36.0000
Chi-square =
0.1602
0.2851
0.3915
0.06
8.809
15.682
21.532
DF = 1
9
15
22
55
55
55
0.026
-0.061
0.036
P-value = 0.8016
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
0.1
Extra risk
0. 95
11.7719
FINAL
9-8-2009
BMDL = 7.32213
76
-------
FINAL
9-8-2009
Multistage Model with 0.95 Confidence Level
BMDL
BMD
15 20
Dose
11:41 02/22 2007
Multistage
BMD Lower Bound
77
-------
FINAL
9-8-2009
Log Probit model
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_LOGPROBIT.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\MNCL_LOGPROBIT.plt
Mon Feb 19 17:00:30 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+Slope*Log(Dose)),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = MNCL
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial (and Specified) Parameter Values
background = 0.163636
intercept = -4.01471
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.64
intercept -0.64 1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0.169336
-4.13687
1
Std. Err.
0.0481103
0.26196
NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
78
-------
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-93.7538
-93.8046
-97.6478
191.609
Deviance Test DF
0.101464
7.78794
1
2
P-value
0.7501
0.02036
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.1693 9.313 9 55 -0.1127
18.0000 0.2576 14.170 15 55 0.2561
36.0000 0.4102 22.563 22 55 -0.1543
Chi-square = 0.10 DF = 1 P-value = 0.7493
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 17.3801
BMDL = 12.0988
Probit Model with 0.95 Confidence Level
Probit
BMD Lower Bound
BMDL
10
15 20
Dose
25
30
35
17:00 02/19 2007
79
-------
FINAL
9-8-2009
Probit model
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_PROBIT.(d)
Gnuplot Plotting File: C:\EMDS\HYDROQUINONE\MNCL_PROBIT.plt
Mon Feb 19 16:59:41 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = CumNorm(Intercept+Slope*Dose),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = MNCL
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
background = 0 Specified
intercept = -0.968445
slope = 0.0200245
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.81
slope -0.81 1
Parameter Estimates
Variable Estimate Std. Err.
intercept -0.97479 0.180177
slope 0.0201363 0.00731154
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-93.7538
-93.7552
-97.6478
Deviance Test DF
0.0028213
7.78794
1
2
P-value
0.9576
0.02036
80
-------
FINAL
9-8-2009
AIC:
191.51
Goodness of Fit
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.1648 9.066 9 55 -0.0239
18.0000 0.2702 14.859 15 55 0.04292
36.0000 0.4013 22.074 22 55 -0.02026
Chi-square = 0.00 DF = 1 P-value = 0.9576
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
14.6549
10.8081
81
-------
FINAL
9-8-2009
Probit Model with 0.95 Confidence Level
Probit
BMD Lower Bound
0.5
0.4
0.3
0.2
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
16:59 02/19 2007
82
-------
FINAL
9-8-2009
Quantal Linear model
Quantal Linear Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_QL.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\MNCL_QL.plt
Mon Feb 19 17:01:36 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*dose)]
Dependent variable = MNCL
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.169643
Slope = 0.00910852
Power = 1 Specified
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,
and do not appear in the correlation matrix )
Background Slope
Background 1 -0.62
Slope -0.62 1
Parameter Estimates
Variable
Background
Slope
Estimate
0.160154
0.00895032
Std. Err
0.0470496
0.00321617
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-93.7538
-93.7855
-97.6478
Deviance Test DF
0.0634144
7.78794
1
2
P-value
0.8012
0.02036
AIC:
191.571
83
-------
Dose
FINAL
9-8-2009
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.1602 8.808 9 55 0.07043
18.0000 0.2851 15.682 15 55 -0.2036
36.0000 0.3915 21.532 22 55 0.1293
Chi-square = 0.06 DF = 1 P-value = 0.8016
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
11.7717
7.32213
84
-------
FINAL
9-8-2009
Quantal Linear Model with 0.95 Confidence Level
Quantal Linear
BMD Lower Bound
0.5
0.4
0.3
0.2
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
17:01 02/19 2007
85
-------
FINAL
9-8-2009
Quantal Quadratic model
Quantal Quadratic Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_QQ.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\MNCL_QQ.plt
Mon Feb 19 17:02:50 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*doseA2)]
Dependent variable = MNCL
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.169643
Slope = 0.000253015
Power = 2 Specified
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,
and do not appear in the correlation matrix )
Background Slope
Background 1 -0.59
Slope -0.59 1
Parameter Estimates
Variable
Background
Slope
Estimate
0.179025
0.000257272
Std. Err
0.0457014
0.000101905
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-93.7538
-93.9292
-97.6478
Deviance Test DF
0.350778
7.78794
1
2
P-value
0.5537
0.02036
AIC:
191.858
86
-------
Dose
FINAL
9-8-2009
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.1790 9.846 9 55 -0.2977
18.0000 0.2447 13.458 15 55 0.4838
36.0000 0.4118 22.649 22 55 -0.1778
Chi-square = 0.35 DF = 1 P-value = 0.5517
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
20.2368
15.5633
87
-------
FINAL
9-8-2009
Quantal Quadratic Model with 0.95 Confidence Level
Quantal Quadratic
BMD Lower Bound
0.5
0.4
0.3
0.2
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
17:02 02/19 2007
88
-------
FINAL
9-8-2009
Weibull model
Weibull Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\MNCL_WEIBULL.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\MNCL_WEIBULL.plt
Mon Feb 19 17:03:52 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*doseApower)]
Dependent variable = MNCL
Independent variable = Dose
Power parameter is restricted as power >=1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.169643
Slope = 0.0030436
Power = 1.30589
Asymptotic Correlation Matrix of Parameter Estimates
Background Slope Power
Background 1 -0.4 0.34
Slope -0.4 1 -0.99
Power 0.34 -0.99 1
Variable
Background
Slope
Power
Parameter
Estimate
0.163636
0.00378274
1.2488
Estimates
Std. Err
0.0498834
0.0138287
1.04037
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood) Deviance Test DF
-93.7538
-93.7538 1.55467e-011 0
-97.6478 7.78794 2
P-value
NA
0.02036
AIC:
193.508
Goodness of Fit
89
-------
FINAL
9-8-2009
Dose
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
18.0000
36.0000
Chi-square =
0.1636
0.2727
0.4000
0.00
9.000
15.000
22.000
DF = 0
9
15
22
P-value =
55
55
55
NA
-1.558e-006
3.215e-006
-1.667e-006
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
14.3552
7.35611
90
-------
FINAL
9-8-2009
Weibull Model with 0.95 Confidence Level
Weibull
BMD Lower Bound
0.5
0.4
0.3
0.2
BMDL
BMD
0
5
10
15
20
25
30
35
Dose
17:03 02/19 2007
91
-------
FINAL
9-8-2009
Part III. Female Mice: Hepatocellular Adenoma or Carcinoma
Gamma model
Gamma Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:54:48 2007
HMDS MODEL RUN
The form of the probability function is:
P[response]= background+(1-background)*CumGamma[slope*dose,power],
where CumGamma(.) is the cummulative Gamma distribution function
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Power parameter is restricted as power >=1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.0625
Slope = 0.0136565
Power = 1.3
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,
and do not appear in the correlation matrix )
Background Slope
Background 1 -0.62
Slope -0.62 1
Interval
Variable
Limit
Background
0.144596
Slope
0.00679
Power
Estimate
0.0704472
0.00415324
1
Parameter Estimates
Std. Err.
0.0378316
0.00134531
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-0.00370129
0.00151648
92
-------
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-74.8827
-76.995
-81.161
157.99
# Param's
3
2
1
Deviance Test d.f.
4.22464
12.5567
P-value
0.03984
0.001877
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
0.0704
0.1995
0.3078
3.875
10.975
16.931
3
16
13
ChiA2
4.41
d.f.
P-value
0.0358
55
55
55
-0.461
1. 696
-1.148
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
25.3683
16.4603
93
-------
FINAL
9-8-2009
Gamma Multi-Hit Model with 0.95 Confidence Level
Gamma Multi-Hit
BMD Lower Bound
0.4
0.3
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
15:54 04/18 2007
94
-------
FINAL
9-8-2009
Log Logistic model
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\EMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:55:49 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose)) ]
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0545455
intercept = -4.66122
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.6
intercept -0.6 1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
0.0656374
-5.30636
1
Std. Err.
*
*
*
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
Indicates that this value is not calculated.
95
-------
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-74.8827
-76.7324
-81.161
157.465
# Param's
3
2
1
Deviance Test d.f.
3.6995
12.5567
P-value
0.05443
0.001877
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
0.0656
0.2072
0.3090
3.610
11.396
16.994
3
16
13
ChiA2
3.82
d.f.
P-value
0.0508
55
55
55
-0.332
1.532
-1.166
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 22.4017
BMDL = 13.7873
Log-Logistic Model with 0.95 Confidence Level
o
it
<
Log-Logistic
BMD Lower Bound
BMDL
15:55 04/18 2007
96
-------
FINAL
9-8-2009
Logistic model
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\EMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:56:29 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = 1/[1+EXP(-intercept-slope*dose)]
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Slope parameter is not restricted
Total number of observations = 3
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
background = 0 Specified
intercept = -2.36514
slope = 0.0221272
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.85
slope -0.85 1
Interval
Variable
Limit
intercept
-1.37287
slope
0.0314579
Estimate
-2.11535
0.0172474
Parameter Estimates
Std. Err.
0.378823
0.00725038
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-2.85783
0.00303693
Analysis of Deviance Table
97
-------
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-74.8827
-78.1435
-81.161
160.287
# Param's
3
2
1
Deviance Test d.f.
6.52164
12.5567
P-value
0.01066
0.001877
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
ChiA2
0.1076
0.1833
0.2909
5.919
10.079
16.002
3
16
13
6. 67
d.f.
P-value
0.0098
55
55
55
-1.270
2.064
-0.891
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 41.123
BMDL = 30.2749
o
it
<
0.4
0.3
0.2
0.1
Logistic Model with 0.95 Confidence Level
Logistic
BMD Lower Bound
10
20
BMDL
30
BMD
40
50
Dose
60
70
15:56 04/18 2007
98
-------
FINAL
9-8-2009
Multistage model: 2 degree polynomial
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:57:20 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2)]
The parameter betas are restricted to be positive
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.10867
Beta(1) = 0.00303181
Beta(2) = 0
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,
and do not appear in the correlation matrix )
Background Beta(l)
Background 1 -0.79
Beta(1) -0.79 1
Parameter Estimates
Interval
Conf. Limit
Variable
Background
Beta(1)
Estimate
0.0704472
0.00415324
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper
99
-------
Beta(2)
* - Indicates that this value is not calculated.
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-74.8827
-76.995
-81.161
157.99
# Param's
3
2
1
Deviance Test d.f.
4.22464
12.5567
P-value
0.03984
0.001877
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
ChiA2 =4.41
0.0704
0.1995
0.3078
d.f. = 1
3.875 3 55
10.975 16 55
16.931 13 55
P-value = 0.0358
-0.461
1. 696
-1.148
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
EMD =
BMDL =
BMDU =
0.1
Extra risk
0. 95
25.3683
16.4603
56.2399
Taken together, (16.4603, 56.2399) is a 90
interval for the EMD
two-sided confidence
100
-------
FINAL
9-8-2009
Multistage Model with 0.95 Confidence Level
Multistage
BMD Lower Bound
0.4
0.3
0.2
0.1
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
15:57 04/18 2007
101
-------
FINAL
9-8-2009
Multistage model: 1 degree polynomial
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:58:11 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
The parameter betas are restricted to be positive
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.10867
Beta(1) = 0.00303181
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.79
Beta(1) -0.79 1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Limit
Background 0.0704472 * * *
Beta(1) 0.00415324 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
102
-------
FINAL
9-8-2009
Full model
Fitted model
Reduced model
AIC:
-74.8827
-76.995
-81.161
157.99
3
2
1
4.22464
12.5567
0.03984
0.001877
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
0.0704
0.1995
0.3078
3.875
10.975
16.931
3
16
13
ChiA2
4.41
d.f.
P-value
0.0358
55
55
55
-0.461
1. 696
-1.148
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
BMDU =
0.1
Extra risk
0. 95
25.3683
16.4603
56.2399
Taken together, (16.4603, 56.2399) is a 90
interval for the EMD
two-sided confidence
103
-------
FINAL
9-8-2009
Multistage Model with 0.95 Confidence Level
Multistage
BMD Lower Bound
0.4
0.3
0.2
0.1
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
15:58 04/18 2007
104
-------
FINAL
9-8-2009
Log Probit model
Probit Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\EMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:59:00 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+Slope*Log(Dose)),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial (and Specified) Parameter Values
background = 0.0545455
intercept = -4.24001
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.75
intercept -0.75 1
Parameter Estimates
Interval
Variable
Limit
background
0.180869
intercept
-4.3374
slope
Estimate
0.0888211
-4.89853
1
Std. Err.
0.046964
0.2863
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-0.00322662
-5.45967
105
-------
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-74.8827
-78.6887
-81.161
161.377
# Param's
3
2
1
Deviance Test d.f.
7.61211
12.5567
P-value
0.005798
0.001877
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
ChiA2 =8.04
0.0888
0.1747
0.3279
d.f. = 1
4.885 3 55
9.609 16 55
18.037 13 55
P-value = 0.0046
-0.894
2.270
-1.447
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
37.225
25.8879
106
-------
FINAL
9-8-2009
Probit Model with 0.95 Confidence Level
Probit
BMD Lower Bound
0.4
0.3
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
15:59 04/18 2007
107
-------
FINAL
9-8-2009
Probit model
Probit Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\EMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 15:59:36 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = CumNorm(Intercept+Slope*Dose) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Slope parameter is not restricted
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
background = 0 Specified
intercept = -1.39179
slope = 0.0124078
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.83
slope -0.83 1
Parameter Estimates
Interval
Variable
Limit
intercept
-0.860662
slope
0.0183628
Estimate
-1.26742
0.0102247
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
0.207531 -1.67417
0.00415216 0.00208665
Analysis of Deviance Table
108
-------
FINAL
9-8-2009
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-74.8827
-78.0008
-81.161
160.002
# Param's
3
2
1
Deviance Test d.f.
6.23629
12.5567
P-value
0.01252
0.001877
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
ChiA2
0.1025
0.1842
0.2941
5.638
10.133
16.175
3
16
13
6.42
d.f.
P-value
0.0113
55
55
55
-1.173
2.041
-0.940
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
38.9049
28.4317
Probit Model with 0.95 Confidence Level
Probit
BMD Lower Bound
0.4
0.3
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
15:59 04/18 2007
109
-------
FINAL
9-8-2009
Quantal Linear model
Quantal Linear Model using Weibull Model (Version: 2.7; Date: 2/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 16:00:28 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*dose)]
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.0625
Slope = 0.00297618
Power = 1 Specified
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,
and do not appear in the correlation matrix )
Background Slope
Background 1 -0.62
Slope -0.62 1
Parameter Estimates
Interval
Variable
Limit
Background
0.144598
Slope
0.00679008
Estimate
0.0704471
0.00415323
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
0.0378326 -0.00370342
0.00134535 0.00151639
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -74.8827 3
Fitted model -76.995 2 4.22464 1 0.03984
110
-------
Reduced model
AIC:
-81.161
157.99
12.5567
FINAL
9-8-2009
0.001877
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
ChiA2 =4.41
0.0704
0.1995
0.3078
d.f. = 1
3.875 3 55
10.975 16 55
16.931 13 55
P-value = 0.0358
-0.461
1. 696
-1.148
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 25.3683
BMDL = 16.4603
Quantal Linear Model with 0.95 Confidence Level
Quantal Linear
BMD Lower Bound
0.4
0.3
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
16:00 04/18 2007
111
-------
FINAL
9-8-2009
Quantal Quadratic model
Quantal Quadratic Model $Revision: 2.2 $ $Date: 2000/03/17 22:27:16 $
Input Data File: C:\BMDS\HYDROQUINONE\LIVER_TUMOR_QQ.(d)
Gnuplot Plotting File: C:\BMDS\HYDROQUINONE\LIVER_TUMOR_QQ.plt
Wed Apr 18 16:33:32 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*doseA2)]
Dependent variable = COLUMN3
Independent variable = COLUMN1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.0625
Slope = 5.64397e-005
Power = 2 Specified
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,
and do not appear in the correlation matrix )
Background Slope
Background 1 -0.62
Slope -0.62 1
Parameter Estimates
Variable
Background
Slope
Estimate
0.0866145
6.25592e-005
Std. Err
0.0388928
2.26923e-005
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-74.8827
-76.6626
-81.161
Deviance Test DF
3.55985
12.5567
1
2
P-value
0.05919
0.001877
AIC:
157.325
112
-------
Dose
FINAL
9-8-2009
Goodness of Fit
Est. Prob.
Expected
Observed
Size
Scaled
Residual
0.0000 0.0866 4.764 3 55 -0.8456
36.0000 0.1577 8.676 13 55 1.6
71.0000 0.3337 18.351 16 55 -0.6724
Chi-square = 3.73 DF = 1 P-value = 0.0536
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
HMD =
BMDL =
0.1
Extra risk
0. 95
41.0387
32.1823
113
-------
FINAL
9-8-2009
Quantal Quadratic Model with 0.95 Confidence Level
Quantal Quadratic
BMD Lower Bound
0.4
0.3
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
16:33 04/18 2007
114
-------
FINAL
9-8-2009
Weibull model
Weibull Model using Weibull Model (Version: 2.7; Date: 2/20/2007)
Input Data File: C:\BMDS\DATA\HYDROQUINONE_CARC.(d)
Gnuplot Plotting File: C:\BMDS\DATA\HYDROQUINONE_CARC.plt
Wed Apr 18 16:01:36 2007
HMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-slope*doseApower)]
Dependent variable = liver_F_resp
Independent variable = mouse_dose
Power parameter is restricted as power >=1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.0625
Slope = 0.00297618
Power = 1
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,
and do not appear in the correlation matrix )
Background Slope
Background 1 -0.62
Slope -0.62 1
Parameter Estimates
Interval
Variable
Limit
Background
0.144598
Slope
0.00679008
Power
Estimate
0.0704471
0.00415323
1
Std. Err.
0.0378326
0.00134535
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.
-0.00370342
0.00151639
115
-------
FINAL
9-8-2009
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log (likelihood)
-74.8827
-76.995
-81.161
157.99
# Param's Deviance Test d.f. P-value
3
2 4.22464 1 0.03984
1 12.5567 2 0.001877
Dose
Goodness of Fit
Est. Prob. Expected Observed Size
Scaled
Residual
0.0000
36.0000
71.0000
ChiA2
0.0704
0.1995
0.3078
3.875
10.975
16.931
3
16
13
4.41
d.f.
P-value
0.0358
55
55
55
-0.461
1. 696
-1.148
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 25.3683
BMDL = 16.4603
116
-------
FINAL
9-8-2009
Weibull Model with 0.95 Confidence Level
Wei bull
BMD Lower Bound
0.4
0.3
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
Dose
16:01 04/18 2007
117
------- |