EPA/635/R-09/008F
www.epa.gov/iris
f/EPA
TOXICOLOGICAL REVIEW
OF
2-HEXANONE
(CAS No. 591-78-6)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
September 2009
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
CONTENTS—TOXICOLOGICAL REVIEW OF 2-HEXANONE (CAS No. 591-78-6)
LIST OF TABLES vi
LIST OF FIGURES vii
LIST OF ACRONYMS AND ABBREVIATIONS viii
FOREWORD ix
AUTHORS, CONTRIBUTORS, AND REVIEWERS x
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION 3
3. TOXICOKINETICS 4
3.1. ABSORPTION 4
3.1.1. Pulmonary Absorption Studies 4
3.1.2. Gastrointestinal Tract Absorption Studies 4
3.1.3. Dermal Absorption Studies 5
3.2. DISTRIBUTION 6
3.3. METABOLISM 8
3.4. ELIMINATION 16
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS 17
4. HAZARD IDENTIFICATION 18
4.1. STUDIES IN HUMANS—CASE REPORTS, EPIDEMIOLOGIC STUDIES,
AND OBSERVATIONAL STUDIES 18
4.2. ACUTE, SUBCHRONIC, AND CHRONIC STUDIES IN ANIMALS 22
4.2.1. Oral Exposure 22
4.2.1.1. Acute and Short-term Oral Exposure 22
4.2.1.2. Subchronic Toxicity Studies 22
4.2.1.3. Chronic Toxicity Study 23
4.2.2. Inhalation Exposures 25
4.2.2.1. Acute and Short-term Toxicity Studies 25
4.2.2.2. Subchronic Toxicity Study 25
4.2.2.3. Chronic Toxicity Studies 26
4.2.3. Dermal Exposure 27
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND
INHALATION 27
4.3.1. Oral Exposure 27
4.3.2. Inhalation Exposure 27
4.4. OTHER ENDPOINT-SPECIFIC STUDIES 30
4.4.1. Neurotoxicity Studies 30
4.4.1.1. Oral Exposures 30
4.4.1.2. Inhalation Exposures 35
4.4.1.3. Other Routes of Exposure 41
4.4.2. Immunotoxicity Studies 42
4.5. OTHER STUDIES 42
4.5.1. Mechanistic Studies 42
4.5.1.1. 2-Hexanone and Enzyme Induction 42
iii
-------
4.5.1.2. 2-Hexanone as a Sulfhydryl Reagent 47
4.5.1.3. Studies Exploring the Development of Neuropathy 47
4.5.2. Genotoxicity Studies 48
4.5.3. Structure-Activity Relationships 48
4.5.4. Potentiation and Other Interaction Studies 49
4.5.4.1. Methyl Ethyl Ketone 49
4.5.4.2. Chloroform 50
4.5.4.3. O-Ethyl O-4-Nitrophenyl Phenylphosphonothioate 50
4.6. SYNTHESIS AND EVALUATION OF MAJOR NONCANCER EFFECTS 50
4.6.1. Oral 50
4.6.2. Inhalation 53
4.6.3. Mode-of-Action Information 55
4.7. WEIGHT-OF-EVIDENCE EVALUATION AND CANCER
CHARACTERIZATION 57
4.7.1. Summary of Overall Weight of Evidence 57
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES 57
4.8.1. Possible Childhood Susceptibility 57
4.8.2. Possible Gender Differences 58
5. DOSE-RESPONSE ASSESSMENTS 59
5.1. ORAL REFERENCE DOSE (RfD) 59
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and
Justification 59
5.1.2. Method of Analysis: Benchmark Dose Modeling 60
5.1.3. Derivation of Human Equivalent Doses 62
5.1.4. Calculation of the RfD—Application of Uncertainty Factors 62
5.1.5. RfD Comparison Information 63
5.1.6. Previous Oral Assessment 65
5.2. INHALATION REFERENCE CONCENTRATION 65
5.2.1. Choice of Principal Study and Critical Effect—with Rationale and
Justification 65
5.2.2. Methods of Analysis: Benchmark Concentration Modeling 67
5.2.3. Exposure Duration Adjustments and Conversion to Human Equivalent
Concentrations 71
5.2.4. Calculation of the RfC: Application of Uncertainty Factors 74
5.2.5. RfC Comparison Information 75
5.2.6. Previous Inhalation Assessment 76
5.3. CANCER ASSESSMENT 76
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 77
6.1. HUMAN HAZARD POTENTIAL 77
6.2. DOSE RESPONSE 78
6.2.1. Noncancer/Oral 78
6.2.2. Noncancer/Inhalation 79
6.2.3. Cancer 80
7. REFERENCES 81
IV
-------
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS
AND DISPOSITION A-l
APPENDIX B-l. DOSE-RESPONSE MODELING FOR DERIVATION OF AN RfD
FOR 2-HEXANONE B-l
APPENDIX B-2. EXPOSURE-RESPONSE MODELING FOR DERIVATION OF AN RfC
FOR 2-HEXANONE B-9
-------
LIST OF TABLES
Table 3-1. Concentrations of 2-hexanone in plasma, liver, and lung of male rats
following oral exposure for 3 days
Table 3-2. Concentrations of 2-hexanone in plasma, liver, and lung of male rats following
inhalation exposure for 3 days 7
Table 3-3. 2-Hexanol and 2,5-hexanedione in the plasma, liver, and lung of male rats
after oral or inhalation exposure to 2-hexanone 11
Table 3-4. 2-Hexanone, 2-hexanol, and 2,5-hexanedione in the blood of male rats after
repeated administration of 400 mg/kg-day 12
Table 3-5. Serum half-lives and clearance times of 2-hexanone and its metabolites in
guinea pigs 14
Table 3-6. Tissue levels of 2-hexanone and 2,5-hexanedione in male F344 rats following
inhalation exposure to n-hexane for 5 days 15
Table 4-1. Prevalence of peripheral neuropathy among employees of a coated fabrics plant 20
Table 4-2. Results of area atmospheric sampling for MEK and 2-hexanone in a coated
fabrics plant 21
Table 4-3. Gross observations in rats exposed to 2-hexanone in drinking water for 120
days 23
Table 4-4. Pathological changes in rats exposed for 13 months to 2-hexanone 24
Table 4-5. Summary of pathological lesions in 40-week-old offspring of dams exposed to
2-hexanone during gestation 29
Table 4-6. Time course of neuropathy scores following exposure of rats to 2-hexanone in
drinking water 31
Table 4-7. Effect of 2-hexanone on guinea pig pupillary response of both eyes 32
Table 4-8. Summary of neuropathological findings in male rats administered 2-hexanone
in drinking water for 13 months 34
Table 4-9. Effects of 2-hexanone, 2,5-hexanedione, and phenobarbital on microsomal
protein and CYP450 43
Table 4-10. Effect of enzyme inducers on the activities of CYP450-related enzymes in rats
exposed to 2-hexanone or 2,5-hexanedione 43
Table 4-11. Catalytic activities of CYP450 enzyme activities in rat liver following
induction by 2-hexanone 44
VI
-------
Table 4-12. Changes in CYP450 levels following treatment with 2-hexanone 45
Table 4-13. Clinical and pathological observations with time of exposure to 2-hexanone
in rats 48
Table 4-14. Summary of the toxicities of n-hexane, MiBK, andMEK 49
Table 4-15. Synopsis of oral toxicity studies with 2-hexanone 52
Table 4-16. Synopsis of animal inhalation toxicity studies with 2-hexanone 54
Table 5-1. Summary of neuropathological findings in male rats administered 2-hexanone
in drinking water for 13 months 61
Table 5.2. Best-fit BMD modeling results for data on axonal swelling of the peripheral
nerve 62
Table 5-3. Effect of 2-hexanone inhalation exposure on the MCV of the sciatic-tibial and
ulnar nerves in monkeys (n = 8/group) 68
Table 5-4. Effect of 2-hexanone inhalation exposure on the MCV of the sciatic-tibial and
ulnar nerves in rats 69
Table 5-5. Summary of BMCLs and HECs for 2-hexanone 73
Table B-l. BMD modeling results for animals with axonal swelling of the peripheral
nerve B-l
Table B-2.1. Summary of BMDS modeling results for 2-hexanone B-9
LIST OF FIGURES
Figure 2-1. Chemical structure of 2-hexanone 3
Figure 3-1. Proposed metabolic pathway for 2-hexanone 9
Figure 4-1. Proposed mechanism for 2,5-hexanedione-induced axonopathy 57
Figure 5-1. Potential PODs for endpoints from O'Donoghue et al. (1978), with
corresponding applied UFs and derived sample oral reference values 64
Figure 5-2. Potential PODs for endpoints from Johnson et al. (1977), with corresponding
applied UFs and derived sample inhalation reference values 76
vn
-------
LIST OF ACRONYMS AND ABBREVIATIONS
AIC
ANOVA
ATSDR
AVEP
BMC
BMCL
BMD
BMDL
BMDS
BMR
CASRN
CNS
CYP
CYP450
EEC
EPA
EPN
EROD
HEC
i.p.
IRIS
LD50
LOAEL
MAI
MAP
MCV
MEK
MiBK
NLM
NOAEL
PBTK
PNS
POD
PROD
RfC
RfD
SEM
TLV
TSO
UF
w/w
Akaike Information Criterion
analysis of variance
Agency for Toxic Substances and Disease Registry
average visual evoked potential
benchmark concentration
benchmark concentration, lower 95% confidence limit
benchmark dose
benchmark dose, lower 95% confidence limit
benchmark dose software
benchmark response
Chemical Abstracts Service Registry Number
central nervous system
cytochrome
cytochrome P450
el ectroencephal ogram
Environmental Protection Agency
O-ethyl O-4-nitrophenyl phenylphosphonothioate
ethoxyresorufm O-deethylase
human equivalent concentration
intraperitoneal
Integrated Risk Information System
median lethal dose
lowest-observed-adverse-effect level
Microbiological Associates, Inc.
muscle action potential
motor (nerve) conduction velocity
methyl ethyl ketone
methyl isobutyl ketone
National Library of Medicine
no-observed-adverse-effect level
physiologically based toxicokinetic
peripheral nervous system
point of departure
pentoxyresorufin O-depentylase
reference concentration
reference dose
standard error of the mean
threshold limit value
toluene side-chain oxidase
uncertainty factor
weight/weight
Vlll
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to
2-hexanone. It is not intended to be a comprehensive treatise on the chemical or toxicological
nature of 2-hexanone.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose
Response, is to present the major conclusions reached in the derivation of the reference dose,
reference concentration and cancer assessment, where applicable, and to characterize the overall
confidence in the quantitative and qualitative aspects of hazard and dose response by addressing
the quality of data and related uncertainties. The discussion is intended to convey the limitations
of the assessment and to aid and guide the risk assessor in the ensuing steps of the risk
assessment process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
IX
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR
Amanda S. Persad, Ph.D., DABT
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
COAUTHORS
TedO. Berner, M.S.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
Glinda S. Cooper, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
Allan Marcus, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Todd Stedeford, Ph.D., J.D., DABT
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
REVIEWERS
This document has been reviewed by EPA scientists, interagency reviewers from other
federal agencies and White House offices, and the public, and peer reviewed by independent
scientists external to EPA. A summary and EPA's disposition of the comments received from
the independent external peer reviewers and from the public is included in Appendix A.
-------
INTERNAL EPA REVIEWERS
Philip J. Bushnell, Ph.D.
National Health and Environmental Effects Research Laboratory
Office of Research and Development
Lynn Flowers, Ph.D., DABT
National Center for Environmental Assessment
Office of Research and Development
Karen Hammerstrom
National Center for Environmental Assessment
Office of Research and Development
Susan Rieth, M.S.
National Center for Environmental Assessment
Office of Research and Development
Reeder L. Sams II, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
John Vandenberg, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
EXTERNAL PEER REVIEWERS
Bahie Abou-Donia, Ph.D.
Duke University Medical Center
Durham, NC 27710
A. John Bailer, Ph.D.
Miami University
Oxford, OH 45056
Frederick J. Miller, Ph.D., Fellow ATS
Fred J. Miller & Associates LLC
Cary,NC27511
Mohammad I. Sabri, Ph.D.
Oregon Health Science University
Portland, OR 97239
Alan H. Stern, D.P.H., DABT
University of Medicine and Dentistry of New Jersey
Piscataway, NJ 08854
XI
-------
Bernard Weiss, Ph.D.
University of Rochester School of Medicine and Dentistry
Rochester, NY 14642
xn
-------
1. INTRODUCTION
This document presents background information and justification for the Integrated Risk
Information System (IRIS) Summary of the hazard and dose-response assessment of 2-hexanone.
IRIS Summaries may include oral reference dose (RfD) and inhalation reference concentration
(RfC) values for chronic and other exposure durations, and a carcinogenicity assessment.
The RfD and RfC, if derived, provide quantitative information for use in risk assessments
for health effects known or assumed to be produced through a nonlinear (presumed threshold)
mode of action. The RfD (expressed in units of mg/kg-day) is defined as an estimate (with
uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without an appreciable risk of
deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
inhalation RfC considers toxic effects for both the respiratory system (portal of entry) and for
effects peripheral to the respiratory system (extrarespiratory or systemic effects). Reference
values are generally derived for chronic exposures (up to a lifetime), but may also be derived for
acute (<24 hours), short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
lifetime) exposure durations, all of which are derived based on an assumption of continuous
exposure throughout the duration specified. Unless specified otherwise, the RfD and RfC are
derived for chronic exposure duration.
The carcinogenicity assessment provides information on the carcinogenic hazard
potential of the substance in question and quantitative estimates of risk from oral and inhalation
exposure may be derived. The information includes a weight-of-evidence judgment of the
likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
effects may be expressed. Quantitative risk estimates may be derived from the application of a
low-dose extrapolation procedure. If derived, the oral slope factor is a plausible upper bound on
the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit risk is a
plausible upper bound on the estimate of risk per |ig/m3 air breathed.
Development of these hazard identification and dose-response assessments for
2-hexanone has followed the general guidelines for risk assessment as set forth by the National
Research Council (1983). EPA Guidelines and Risk Assessment Forum Technical Panel Reports
that may have been used in the development of this assessment include the following: Guidelines
for the Health Risk Assessment of Chemical Mixtures (U.S. EPA, 1986a), Guidelines for
Mutagenicity Risk Assessment (U.S. EPA, 1986b), Recommendations for and Documentation of
Biological Values for Use in Risk Assessment (U.S. EPA, 1988), Guidelines for Developmental
Toxicity Risk Assessment (U.S. EPA, 1991), Interim Policy for Particle Size and Limit
Concentration Issues in Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of
1
-------
Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA
1994b), Use of the Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995),
Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for
Neurotoxicity Risk Assessment (U.S. EPA, 1998), Science Policy Council Handbook: Risk
Characterization (U.S. EPA, 2000a), Benchmark Dose Technical Guidance Document (U.S.
EPA, 2000b), Supplementary Guidance for Conducting Health Risk Assessment of Chemical
Mixtures (U.S. EPA, 2000c), A Review of the Reference Dose and Reference Concentration
Processes (U.S. EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
(U.S. EPA, 2005b), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006a), and A
Framework for Assessing Health Risks of Environmental Exposures to Children (U.S. EPA,
2006b).
The literature search strategy employed for this compound was based on the Chemical
Abstracts Service Registry Number (CASRN) and at least one common name. Any pertinent
scientific information submitted by the public to the IRIS Submission Desk was also considered
in the development of this document. The relevant literature was reviewed through
January 2009.
-------
2. CHEMICAL AND PHYSICAL INFORMATION
Structurally, 2-hexanone consists of a keto group flanked by a methyl group and an
n-butyl group (Figure 2-1). The compound is a colorless liquid with a characteristic acetone-like
odor but more pungent (National Library of Medicine [NLM], 2005). Synonyms for 2-hexanone
include the following: methyl butyl ketone, methyl n-butyl ketone, butyl methyl ketone, MnBK,
and propylacetone.
Figure 2-1. Chemical structure of 2-hexanone.
Pertinent physical and chemical properties of 2-hexanone are listed below (NLM, 2005).
Chemical formula
Molecular weight 100.16
Melting point -55.5°C
Boiling point 127. 6°C
Flash point 23 °C
Density 0.8113at20°C
Water solubility 1 .64 x 104 mg/L at 20°C
LogKow 1.38
Vapor pressure 11.6 mm Hg at 25°C
Conversion factors 1 ppm = 4.1 mg/m3; 1 mg/m3 = 0.244 ppm
2-Hexanone is produced commercially by the catalyzed reaction of acetic acid and
ethylene under pressure followed by distillation to purify the material (NLM, 2005). The
compound has been used as a solvent for lacquers, ink thinners, nitrocellulose, resins, oils, fats,
and waxes. It is a medium evaporating solvent for nitrocellulose acrylates, vinyl, and alkyd
coatings (polyester coating derived from an alcohol and an acid or acid anhydride).
In 1977, the combined production and import of 2-hexanone in the U.S. was between 453
and 4,500 metric tons (NLM, 2005); no breakdown of these figures was provided. The only U.S.
producer of 2-hexanone, the Tennessee Eastman Company division of Eastman Kodak,
discontinued production of 2-hexanone in 1979 and sold its remaining reserves by 1981 (NLM,
2005). 2-Hexanone is not produced or used in the U.S., and no information on importation is
available (Agency for Toxic Substances and Disease Registry [ATSDR], 1992). However,
2-hexanone is still found at Superfund sites.
-------
3. TOXICOKINETICS
3.1. ABSORPTION
3.1.1. Pulmonary Absorption Studies
The available data indicate that 2-hexanone is well absorbed after administration via the
inhalation route. DiVincenzo et al. (1978) exposed three healthy male volunteers (ages 22 to
53 years) to 2-hexanone (>97% pure, containing methyl isobutyl ketone [MiBK] and traces of
2-hexanol) at 10 or 50 ppm for 7.5 hours or 100 ppm for 4 hours. The 7.5-hour exposures were
interrupted after 4 hours for a 0.5-hour lunch period. The volunteers were sedentary during the
exposure. Expired air and venous blood samples were collected before, during, and after
exposure. Exposures to 10 and 50 ppm for 7.5 hours produced mean 2-hexanone breath
concentrations of 1.4 and 9.3 ppm, respectively. Fifteen minutes after exposure to 10 or 50 ppm,
the expired air concentrations of 2-hexanone were 0.1 and 0.5 ppm, respectively. Exposure to
100 ppm for 4 hours produced an average 2-hexanone breath concentration of 22 ppm. These
results indicated that between 75 and 92% of the inhaled 2-hexanone vapor was absorbed by the
lungs and respiratory tract (DiVincenzo et al., 1978). 2-Hexanone was not detected in the
expired air 3 hours after cessation of exposure to 50 or 100 ppm 2-hexanone.
DiVincenzo et al. (1978) exposed four young male beagles to 2-hexanone (>97% pure,
containing MiBK and traces of 2-hexanol) for 6 hours at concentrations of 50 or 100 ppm. Over
the first 4 hours of the exposure period, the hexanone in exhaled air had time-weighted average
concentrations of 16 and 35 ppm for the low- and high-exposure groups, respectively. Thirty
minutes after cessation of exposure to 50 ppm 2-hexanone, the breath concentration of
2-hexanone decreased to 0.7 ppm. 2-Hexanone was below the limit of detection by 3 to 5 hours
after the exposure. It was determined that about 65-68% of the inhaled vapor was absorbed by
the lungs.
3.1.2. Gastrointestinal Tract Absorption Studies
2-Hexanone appears to be well absorbed after oral administration. DiVincenzo et al.
(1978) administered 2 jiCi of l-[14C]-hexanone dissolved in corn oil via a gelatin capsule to
human volunteers; the total dose was 0.1 mg/kg. Most of the 2-hexanone-derived radioactivity
was exhaled as 14CO2, reaching a peak within 4 hours of dosing and then decreasing slowly over
the next 3 to 5 days. The major portion of radioactivity excretion in urine occurred during the
first 48 hours but continued at measurable levels until 8 days after dosing. The cumulative 8-day
elimination of radioactivity in breath and urine averaged 39.5 and 26.3%, respectively. The
overall recovery of 14C was 65.8%. The authors presumed that the remainder of the radioactivity
was retained in tissue or fat deposits.
-------
Administration of l-[14C]-2-hexanone at 20 or 200 mg/kg by gavage to rats resulted in
excretion of about 1.1% of the administered radioactivity in the feces, about 44% in the breath,
and 38% in urine, with about 15% remaining in the carcass after 48 hours and 8% remaining
after 6 days (DiVincenzo et al., 1977). The results were similar at either dose level. These
findings suggest that about 98% of the administered dose was absorbed via the gastrointestinal
tract.
3.1.3. Dermal Absorption Studies
2-Hexanone is also absorbed after dermal application. DiVincenzo et al. (1978) exposed
six human volunteers (ages 30-53 years) to radiolabeled l-[14C]-2-hexanone (>97% purity,
contaminants not stated). The labeled compound was applied to the ventral surface of the
forearm, which had been shaved 24 hours prior to testing. l-[14C]-2-hexanone was held in
contact with the skin for 60 minutes, and precautions were taken to ensure that inhalation
exposure did not occur. The surface area of the skin subjected to the solvent was 55.6 cm2.
Calculated skin absorption rates in two volunteers were 4.8 and 8.0 |ig/cm2-minute. The
quantities of 2-hexanone absorbed systemically were 15.96 and 26.81 mg, respectively. The
major respiratory excretion metabolite of l-[14C]-hexanone was 14CC>2. A substantial portion of
the dose was also excreted in urine; however, the chemical nature of urinary radioactivity was
not characterized further.
In a similar set of experiments, DiVincenzo et al. (1978) applied l-[14C]-2-hexanone
(>97% purity, impurities not stated) to the clipped thorax (55.6 cm2) of beagles. Exposures were
carried out for 5 minutes to 1 hour. By 5 minutes, 11 mg of 2-hexanone had penetrated the skin,
and there was no apparent change in the absorption of 2-hexanone during the next 15 minutes.
However, after 20 minutes the absorption increased markedly so that, by 60 minutes, 77 mg of
2-hexanone had penetrated the skin. The 8-hour cumulative excretion of radioactivity in two
dogs dosed with l-[14C]-2-hexanone was 0.5% of the dose as unchanged 2-hexanone and 9.7%
as 14CC>2 in the breath; urinary radioactivity amounted to 6.5% of the dose. The 8-hour excretion
of radioactivity averaged 16.8% of the dose. The fraction of the applied 2-hexanone dose that
was absorbed was not calculated.
O'Donoghue and Krasavage (1981) exposed two male beagles (one of which was
pretreated with 2-hexanone) to 2-hexanone by tail dipping. Both dogs were exposed to
2-hexanone on an area of 22 cm2 on the first day of exposure, and then the exposure area was
doubled on the second day (44.1 cm2). It was found that, by 8-12 minutes, both dogs had
comparable serum levels of 2-hexanone. Doubling the exposed area increased serum levels of
2-hexanone 6 to 20 times. None of the blood samples contained detectable levels of the
2-hexanone metabolites 5-hydroxy-2-hexanone, 2,5-hexanedione, or 2,5-hexanediol. Similar
exposures were repeated with three different dogs for 16 minutes, followed by two postexposure
samples 9 and 19 minutes later (25- and 35-minute samples, respectively). One animal had
-------
detectable levels of 2-hexanone in blood within 4 minutes, but the time to detectable levels was
highly variable among the animals. The highest level observed was 3.2 jig/mL. Nineteen
minutes postexposure serum levels of 2-hexanone were still detectable. Twenty-four hours later,
no 2-hexanone was detected (O'Donoghue and Krasavage, 1981).
To examine the effects of multiple exposures, O'Donoghue and Krasavage (1981)
exposed three male dogs as above to 2-hexanone on two occasions 4 hours apart. Samples
obtained after the second treatment were not significantly different from the morning samples,
indicating the absence of accumulation of detectable 2-hexanone and 2,5-hexanedione levels in
the serum.
O'Donoghue and Krasavage (1981) performed comparison studies on percutaneous
absorption of 2-hexanone between dog and rabbit skin. Significantly more 2-hexanone was
absorbed through rabbit skin compared with dog skin, and as a probable consequence the
metabolite 5-hydroxy-2-hexanone was detected in the serum of rabbits. Overall, the skin studies
indicated that 2-hexanone was readily absorbed through the skin; detectable serum levels were
present after approximately 10 minutes of exposure to less than 1% of body skin surface;
detectable serum levels persisted for approximately 20 minutes postexposure; and, in rabbits, a
metabolite (5-hydroxy-2-hexanone) was rapidly formed and detectable in the serum.
3.2. DISTRIBUTION
Duguay and Plaa (1995) treated male Sprague-Dawley rats by gavage with 2-hexanone
(>99%, spectrophotometric grade) at 0.5, 1, or 2 mmol/kg (50, 100, or 200 mg/kg) in corn oil
(dose volume 10 mL/kg) once daily for 3 days. The animals were sacrificed 1 hour after the last
gavage. Dose-dependent increases in plasma and lung 2-hexanone levels were observed,
whereas the concentration in the liver increased only with the highest dose (Table 3-1).
Calculations for statistically significant differences among dose groups were not performed
(Duguay and Plaa, 1995).
Table 3-1. Concentrations of 2-hexanone in plasma, liver, and lung of male
rats following oral exposure for 3 days
Tissue concentration
Plasma (ug/mL)
Liver (ug/g)
Lung (ug/g)
Dose
0.5 mmol/kg
2.4 ±1.2
1.7 ±0.5
1.1 ±0.7
1 mmol/kg
4.7 ±1.1
1.6 ±0.3
4.9± 1.1
2 mmol/kg
8.5 ±2.0
3. 8 ±1.2
13.9 ±4.9
Source: Duguay and Plaa (1995).
In a parallel series of experiments from the same study, Duguay and Plaa (1995) exposed
male Sprague-Dawley rats to a total body exposure of 2-hexanone at concentrations of 75, 150,
-------
or 300 ppm (307.5, 615, or 1,230 mg/m ). Animals were exposed on 3 consecutive days for
4 hours per day. Animals were sacrificed immediately after the last exposure on the third day.
The concentration of 2-hexanone in plasma, liver, and lung increased in a dose-dependent
manner (Table 3-2). It should be noted, however, that because whole body exposures were
performed, some oral and dermal absorption may have taken place.
Table 3-2. Concentrations of 2-hexanone in plasma, liver, and lung of male
rats following inhalation exposure for 3 days
Tissue concentration
Plasma (ug/mL)
Liver (ug/g)
Lung (ug/g)
Dose
75 ppm
1.2 ±0.3
0.7 ±0.5
0.7 ±0.2
150 ppm
2.6 ±0.7
1.2 ±0.8
2.8 ±0.5
300 ppm
9.7 ±0.7
2.2 ±0.4
9.3 ±1.2
Source: Duguay and Plaa (1995).
In male CD/COBS rats administered a single gavage dose of [ C]-2-hexanone at
200 mg/kg, 2-hexanone was eliminated from the serum within 6 hours; the 2-hexanone
metabolites 5-hydroxy-2-hexanone and 2,5-hexanedione were eliminated from serum within 12
and 16 hours, respectively (DiVincenzo et al., 1977). Peak concentrations of 2-hexanone and
5-hydroxy-2-hexanone were reached at 2 hours, whereas the peak concentration of 2,5-hexanone
was reached at 6 hours. Radioactivity was detected in most tissues with highest counts in liver >
kidney > whole brain. The peak concentration of radiolabel in each of these tissues was
observed at 4 hours and was reduced to less than 50% by 24 hours. No quantitative data were
given on tissue distribution. An analysis of the subcellular distribution of the 14C-label in liver,
brain, and kidney tissue homogenates indicated the highest counts were associated with the
protein fraction, with some recovery from DNA and little or none from RNA.
Eben et al. (1979) treated male SPF-Wistar rats with 400 mg/kg 2-hexanone (98% pure,
impurities not stated) daily by stomach tube for 40 weeks. The concentrations of 2-hexanone
and metabolites in the blood were determined at intervals of 4 or 5 weeks. In the case of
2-hexanone, the maximum concentration was reached 1 hour after administration throughout the
study; thereafter, the concentration decreased rapidly. After 7 hours, only trace amounts could
be detected. During the first few weeks of the study, 2-hexanone could not be found in the urine.
Only during the third week were very small concentrations of the free compound detected in
urine, suggesting that the metabolic pathways for 2-hexanone were becoming saturated. A
maximum (approximately 20 jig) was reached in the 17th week (Eben et al., 1979).
Granvil et al. (1994) studied the distribution and disappearance of 2-hexanone (purity not
stated) from the blood and brain. Male CD-I mice were treated with 5 mmol/kg (500 mg/kg)
2-hexanone dissolved in corn oil by intraperitoneal (i.p.) injection at a volume of 10 mL/kg.
-------
Animals were killed by decapitation, and blood and brain samples were collected at 15, 30, 60,
and 90 minutes after treatment. Blood and brain concentrations at 15 minutes were -182 |ig/mL
and -126 |ig/g, respectively. By 90 minutes, the values had dropped in a uniform manner to a
blood concentration of-28 |ig/mL and a brain concentration of-25 |ig/mg. The authors noted
that the rapid decrease in the concentration of 2-hexanone was due to its active metabolism in
these tissues.
3.3. METABOLISM
2-Hexanone is hydroxylated to 5-hydroxy-2-hexanone, which is then either oxidized to
2,5-hexanedione or reduced to 2,5-hexanediol and, to a small extent, may be converted to
2,5-dimethyl-2,3-dihydrofuran (Figure 3-1). The predominant metabolite of 2-hexanone found
in blood is 2,5-hexanedione. This can be reduced to 5-hydroxy-2-hexanone and further but, to a
lesser extent, to 2,5-hexanediol. The formation of 2,5-hexanedione is favored over that of
5-hydroxy-2-hexanone. 5-Hydroxy-2-hexanone can be metabolized into 4,5-dihydroxy-
2-hexanone (not shown in Figure 3-1) before being further converted to 2,5-dimethyl-
2,3-dihydrofuran. Additionally, 4,5-dihydroxy-2-hexanone formation may be a result from
2,5-hexanedione metabolism (U.S. EPA, 2005c). Other mechanisms, such as shunting into
intermediary metabolism, may accelerate metabolic clearance of 2,5-hexanedione. Reductive
metabolism of 2-hexanone results in the formation of 2-hexanol, establishing an equilibrium
between the two compounds. 2-Hexanol can be further metabolized to 2,5-hexanediol,
5-hydroxy-2-hexanone, and 2,5-hexanedione. The findings of Ab del-Rahman et al. (1976) that
rats, guinea pigs, and rabbits exposed to 2-hexanone vapor excreted glucuronides of 2-hexanol
and 2,5-hexanediol in the urine are consistent with the results by DiVincenzo et al. (1976),
discussed later in this section. Although the proportions of metabolites may differ among
species, co-1-oxidation and carbonyl reduction appear to be the initial steps in the metabolism of
2-hexanone in all species tested so far (e.g., rat, cat, dog, guinea pig, and human). The metabolic
pathway for 2-hexanone, as proposed by DiVincenzo et al. (1977, 1976), based on 2-hexanone
metabolites identified in blood of guinea pigs, mice, and rats, is presented in Figure 3-1.
-------
o o
1! 11
OH — C— C — CH,~ CH2— CH2—CH3
2-Keto-hexanoic acid
o
II
H,C — C — CH,—
2-liexanone
OH
1
— C —CH2 —CH2—CH2—CH3
2-hexanol
OH— C —CH2 —CH2—CH2—CH3
Pentanoic acid
OH * OH
I !
H3C — CH — CH2 — — CH3
2,5-ItexitiitJcliol
Figure 3-1. Proposed metabolic pathway for 2-hexanone.
Adapted from DiVincenzo et al. (1977, 1976).
O
O
11
Ht~* m*mm f""* •:
-|\«,' \»
5
OH
O
H,C^
2.5- dime!
"X)/
.h>|.2.^-
-------
As discussed in Section 3.2, Duguay and Plaa (1995) conducted studies using male
Sprague-Dawley rats exposed to 2-hexanone by gavage (0.5, 1, or 2 mmol/kg) or by inhalation
(75, 150, or 300 ppm) and quantified the metabolites in the plasma, liver, and lung. The authors
reported that the concentrations of metabolites, such as 2-hexanol, 5-hydroxy-2-hexanone, and
2,5-hexanedione, were readily detectable in serum. 2-Hexanone concentrations in plasma
increased dose dependently regardless of the route of administration. The appearance of the
2-hexanone metabolite 2,5-hexanedione in plasma or lung, but not in liver, depended on the
route of administration. The highest oral dose and the highest inhalation concentration of
2-hexanone, 2 mmol/kg and 300 ppm, respectively, produced similar plasma 2-hexanone
concentrations, 8.5 and 9.7 |ig/mL, respectively, but the corresponding plasma 2,5-hexanedione
concentrations were 7.7 |ig/mL after oral and 25 |ig/mL after inhalation administration.
2,5-Hexanedione was not detectable in lungs when 2-hexanone was administered orally, but
significant dose-dependent amounts were found following inhalation exposure. The authors
concluded that pulmonary 2-hexanone metabolism might contribute to plasma metabolite levels.
2,5-Hexanedione concentrations in liver were dose dependent but independent of the route of
administration. Another metabolite, 2-hexanol, was found in low concentrations in plasma and
liver (0.5-1.3 |ig/mL and 0.3-1.6 |ig/g, respectively) after 2-hexanone gavage or inhalation. In
the lung, 2-hexanol concentrations were higher, ranging from 2.1 to 5.1 |ig/g. The authors noted
that, for both routes of administration, 2-hexanol concentrations did not appear to be dose
dependent. A summary of the metabolite levels found in the plasma, liver, and lung following
oral and inhalation exposures is presented in Table 3-3.
10
-------
Table 3-3. 2-Hexanol and 2,5-hexanedione in the plasma, liver, and lung of
male rats after oral or inhalation exposure to 2-hexanone
Dose"
0.5 mmol/kg
(gavage)
75 ppm
(inhalation)
1 mmol/kg
(gavage)
150 ppm
(inhalation)
2 mmol/kg
(gavage)
300 ppm
(inhalation)
Plasma concentration (fig/mL)
2-HOLb
2,5-HDc
0.6 ±0.2
5.7 ±.0.5
0.5 ±0.1
6.7 ±0.8
1.2 ±0.3
5.8 ±0.5
0.5 ±0.2
13. 5 ±2.1
1.3 ±0.4
7.7 ±0.4
0.8 ±0.3
25 ±3.1
Liver concentration (ug/g)
2-HOL
2,5-HD
0.4 ±0.1
3.1 ±0.2
0.4 ±0.2
3.1±0.1
1.6 ±0.5
4.6 ±0.5
0.3 ±0.1
4.3 ±0.6
1.2 ±0.5
5. 3 ±0.5
0.3 ±0.3
7.3 ±0.3
Lung concentration (ug/g)
2-HOL
2,5-HD
2.4 ±0.5
NDd
2.1 ±0.6
0.9 ±0.2
3.0 ±1.0
ND
2.9 ±0.5
4.0 ±0.1
5.1 ±1.5
ND
3.7 ±1.1
4.8 ±0.7
aRats were sacrificed 1 hour after the last oral treatment but immediately after the last inhalation
exposure. Values are mean ± standard error of the mean (SEM) from six animals.
b2-HOL = 2-hexanol.
C2,5-HD = 2,5-hexanedione.
dND = not detectable (O.25 ug/g).
Source: Duguay and Plaa (1995).
Eben et al. (1979) administered daily oral doses of 2-hexanone (400 mg/kg) over a
40-week period to male SPF-Wistar rats. The concentrations of 2-hexanone, 2-hexanol, and
2,5-hexanedione were determined in the blood at several intervals every 4 or 5 weeks.
2-Hexanone concentrations in blood peaked at 1 hour after administration then decreased
rapidly, and after 7 hours only traces could be detected. The metabolite 2-hexanol was
measurable in very small quantities up to 3 hours after administration (<2 |ig/mL blood). In
contrast, 2,5-hexanedione concentrations were relatively high as early as 1 hour after
administration, and maximum values were recorded after 5 or 7 hours. 2,5-Hexanediol was not
detectable in the blood at any time. A summary of the blood concentrations of 2-hexanone,
2-hexanol, and 2,5-hexanedione is presented in Table 3-4.
11
-------
Table 3-4. 2-Hexanone, 2-hexanol, and 2,5-hexanedione in the blood of male
rats after repeated administration of 400 mg/kg-day
Week
2
6
10
14
19
23
27
32
36
40
2-Hexanone
Oig/mL)a
1 hour
26.5
30.4
20.2
31.8
32.2
35.1
37.8
24.8
50.1
33.4
3 hours
15.2
21.4
7.5
25.7
22.5
19.8
21.3
12.2
13.4
18.9
5 hours
5.8
7.3
4.7
6.3
3.4
6.6
2.9
2.9
7.1
3.6
7 hours
b
1.4
3.8
2.2
0.1
0.3
0.7
0.3
1
0.4
2-Hexanol
Oig/mL)a
1 hour
—
0.6
0.7
1.7
-
-
1.3
0.6
1.5
1.2
3 hours
—
0.8
-
1.2
-
-
traces
0.1
0.1
0.9
5 hours
—
-
-
traces
-
-
-
-
-
-
7 hours
—
-
-
-
-
-
-
-
-
-
2,5-Hexanedione
Oig/mL)a
1 hour
19.8
10.9
16.7
10
16.7
10.4
8
8.4
14.6
12.5
3 hours
53.3
46.5
39.2
35.1
50.3
46.7
38.8
31.8
47.4
36.8
5 hours
65.7
59.7
60.7
55.2
62.1
68.9
49.9
41.1
55.6
51.9
7 hours
53.8
59.8
64
59.1
55
63.7
49.2
34.8
56.1
66.2
"Values represent the averages of three animals.
bA dash (-) indicates that the compound was below the limit of detection.
Source: Ebenetal. (1979).
Granvil et al. (1994) demonstrated the rapid removal of 2-hexanone from blood and brain
of male CD-I mice following a single i.p. injection of the compound at a concentration of
5 mmol/kg. The authors observed that 2-hexanol concentrations found in whole brain at several
time intervals (15, 30, and 60 minutes after dosing) were about twice as high as those found in
the blood at the same time intervals and interpreted this finding as suggesting that 2-hexanol
might be formed in the brain. Furthermore, the authors reported that the appearance of the
reduced metabolite 2-hexanol seemed to be considerably faster than the appearance of the
oxidized metabolite 2,5-hexanedione.
DiVincenzo et al. (1976) administered a single dose of 2-hexanone (450 mg/kg i.p. in
corn oil) to male guinea pigs (strain not stated). Blood was collected by heart puncture from four
animals at 1, 2, 4, 6, 8, 12, and 16 hours after dosing. In addition to 2-hexanone, three major
metabolites were identified by gas chromatography: 5-hydroxy-2-hexanone, 2,5-hexanedione,
and 2-hexanol. 2,5-Dimethyl-2,3-dihydrofuran was also detected, but additional experiments
revealed that this was an artifact because 5-hydroxy-2-hexanone underwent dehydration and
cyclization in the gas chromatograph. The authors noted that 5-hydroxy-2-hexanone may be
transformed in vivo to 2,5-dimethyl-2,3-dihydrofuran, but the equilibrium favors the formation
of 5-hydroxy-2-hexanone.
DiVincenzo et al. (1976) also conducted follow-up studies to determine the metabolic
fate of 2-hexanone metabolites in guinea pigs. Each of the principal metabolites identified in the
above study (5-hydroxy-2-hexanone, 2,5-hexanedione, 2,5-hexanediol, and 2-hexanol) was
12
-------
administered individually (450 mg/kg i.p.). 5-Hydroxy-2-hexanone was further metabolized to
2,5-hexanedione and 2,5-hexanediol. The half-life of 5-hydroxy-2-hexanone in serum was
156 minutes. The major metabolite 2,5-hexanedione was formed rapidly, and its concentration
in serum was equivalent to or greater than that of the parent compound (5-hydroxy-2-hexanone)
in all samples measured. Serum concentrations of 2,5-hexanediol were markedly lower than
those of 2,5-hexanedione. 5-Hydroxy-2-hexanone was the only metabolite detected in the serum
of guinea pigs after an i.p. injection of 2,5-hexanedione. The half-life of 2,5-hexanedione was
100 minutes. Both 5-hydroxy-2-hexanone and 2,5-hexanedione were no longer detectable in
serum by 16 hours. The principal metabolites in serum after i.p. injection with 2,5-hexanediol
were 5-hydroxy-2-hexanone and 2,5-hexanedione. 2,5-Hexanediol was cleared within 8 hours
and had a half-life of 84 minutes in serum. The following metabolites were identified after the
administration of 2-hexanol: 2-hexanone, 5-hydroxy-2-hexanone, 2,5-hexanedione, and
2,5-hexanediol. The half-life and clearance time of 2-hexanol were 72 minutes and 6 hours,
respectively.
The authors noted that the 2-hexanol was rapidly metabolized to 2-hexanone, which, in
turn, was converted to the same metabolites identified above for animals treated with
2-hexanone. They determined that the conversion of 2-hexanol to 2,5-hexanediol seemed to be a
minor pathway. The metabolites 2,5-hexanediol and 2,5-hexanedione were cleared in 8 and
16 hours, respectively. A summary of the half-life and clearance time of 2-hexanone and
metabolites is presented in Table 3-5.
13
-------
Table 3-5. Serum half-lives and clearance times of 2-hexanone and its
metabolites in guinea pigs
Compound administered
2-Hexanone
2-Hexanol
5-Hydroxy -2-hexanone
2,5-Hexanedione
2,5-Hexanediol
Half-life (minutes)3
78
72
156
100
84
Clearance time (hours)
6
6
8b
16
8
aHalf-lives were determined from the linear portion of the plasma concentration curve and
extrapolated to zero time.
Estimated value.
Source: DiVincenzo etal. (1976).
Bus et al. (1981) presented metabolism data for n-hexane that provide some insight into
the metabolism of 2-hexanone. In the study, the authors exposed male F344 rats for 1 or 5 days,
6 hours/day, to 1,000 ppm n-hexane. Animals were sacrificed immediately after exposure or at
increasing time intervals for up to 24 hours after the end of exposure, and concentrations of the
parent compound and two of its metabolites, 2-hexanone and 2,5-hexanedione, were measured in
blood, liver, kidney, brain, and sciatic nerve. Kinetics of all three compounds were similar after
1 and 5 days of exposure, with tissue levels of the metabolites frequently exceeding those of the
parent compound even immediately after the end of exposure. Tissue levels of n-hexane and
2-hexanone were always lower after 5 days of repeated exposures, compared with levels after a
single exposure, consistent with self-induction of some metabolizing enzymes. On the other
hand, tissue levels of 2,5-hexanedione were always slightly higher after 5 days of exposure,
compared with single exposure. A compilation of the data for 2-hexanone and 2,5-hexanedione
after 5 days of exposure to n-hexane is given in Table 3-6 (sciatic nerve data not included).
This experiment, although conducted with n-hexane as the parent compound, provides
some insight into the metabolism of 2-hexanone. The data shown in Table 3-6 indicate that the
metabolism of 2-hexanone to 2,5-hexanedione (intermediates not considered) proceeds rapidly,
while the further metabolism of 2,5-hexanedione and its elimination appear to proceed much
more slowly. Both the resurgence of 2-hexanone levels in kidney between 8 and 12 hours and
the precipitous drop of 2,5-hexanedione levels in kidney between 8 and 12 hours occurred in the
same fashion with single exposure, suggesting rather complex compartmentalization and
toxicokinetics that, to some extent, may be governed by the lipophilic characteristics of the
compounds. Bus et al. (1981) suggested that the high levels observed in kidney for both
metabolites, but not the parent compound, reflect the fact that the metabolites of n-hexane, and
thus 2-hexanone, are mostly eliminated via urine.
14
-------
Table 3-6. Tissue levels of 2-hexanone and 2,5-hexanedione in male F344
rats following inhalation exposure to n-hexane for 5 days
Time
(hours)
0
1
2
4
8
12
24
Blooda'b
2-Hxd
0.46 ± 0.07
0.23 ± 0.02
0.06 ±0.03
f
-
-
-
2,5-HDe
1.97 ±0.38
6.02 ±0.56
3.99 ±0.37
2.12 ±0.26
0.54 ±0.19
-
-
Livera'c
2-Hx
0.12 ±0.01
0.20 ± 0.04
-
-
-
-
-
2,5-HD
0.56 ±0.10
0.64 ±0.13
0.69 ±0.01
0.15 ±0.02
0.03 ±0.03
-
-
Braina'c
2-Hx
0.78 ±0.04
0.18 ±0.01
0.08 ±0.01
-
-
-
-
2,5-HD
5.66 ±0.17
7.41 ±0.31
7.17 ±0.74
2.75 ±0.34
-
-
-
Kidneya'c
2-Hx
22.9 ±3. 81
9.73 ±1.14
4.80 ±0.39
0.63 ±0.23
0.67 ±0.15
0.78 ±0.21
0.28 ±0.00
2,5-HD
11. 8 ±1.03
23. 5 ±1.85
26.4 ±1.61
16.8 ±3.67
9.08 ±2.45
0.86 ± 0.27
0.01 ±0.01
aMean±SEM, n = 3.
bValues in ug/mL plasma.
°Values in ug/g wet weight.
d2-Hx = 2-hexanone.
e2,5-HD = 2,5-hexanedione.
fDash (-) = below detection limit.
Source: Bus etal. (1981).
Cytochrome P450 (CYP450) enzymes catalyze the initial steps (either detoxification or
bioactivation) of 2-hexanone, but their identities have not been investigated in any great detail.
Oral administration of l-[14C]-2-hexanone to humans or rats resulted in the appearance of 14CC>2
in the exhaled breath, indicating removal of the a-carbon (DiVincenzo et al., 1978, 1977).
Administration of SKF525A (a mixed function oxidase inhibitor) to rats before oral
administration of 2-hexanone resulted in marked decrease in the respiratory excretion of 14CC>2
for the first 4 hours after administration, followed by a marked increase at 4-8 and 12-24 hours.
This suggests that this oxidative step is mediated by a microsomal mixed-function oxidase
system (DiVincenzo et al., 1977).
Because inhalation exposure of humans to l-[14C]-2-hexanone resulted in the appearance
of labeled carbon dioxide in expired air and 2,5-hexanedione in serum, DiVincenzo et al. (1978)
hypothesized that the metabolic pathway for 2-hexanone is similar in humans and experimental
animals. Metabolically, aliphatic ketones generally are in equilibrium with the corresponding
secondary alcohols, which explains the presence of 2-hexanol. An alternate pathway is oxidation
of the 5-methylene group to the corresponding alcohol, 5-hydroxy-2-hexanone. Another
possibility in the metabolism of 2-hexanone is the cyclization of 5-hydroxy-2-hexanone to the
corresponding dihydrofuran and oxidation to 2,5-dimethylfuran (DiVincenzo et al., 1977).
However, the formation of these furan moieties may be an artifact resulting from thermal
dehydration and cyclization during gas chromatography (DiVincenzo et al., 1977). In addition,
the y-valerolactone found in the urine was hypothesized to result from a-oxidation of 5-hydroxy-
2-hexanone to 2-keto-5-hydroxyhexanoic acid, decarboxylation and oxidation to
15
-------
4-hydroxypentanoic acid, and lactonization to y-valerolactone (not shown in Figure 3-1)
(DiVincenzo et al., 1977).
Although the specific isoforms of CYP450 that catalyze the metabolism of 2-hexanone
have not been fully characterized, Nakajima et al. (1991) provided some insight into the effects
of 2-hexanone on CYP450 induction. The authors treated male Wistar rats with 2-hexanone at
5 mmol/kg (500 mg/kg) i.p. for 4 days and demonstrated that various CYP450 isozyme activities
were induced. 2-Hexanone was effective in inducing several CYP450 isoforms as indicated by
the increase in activities of benzene aromatic hydroxylase (CYP2E1) and toluene side-chain
oxidation (ISO) (CYP2C6/11) two- to threefold and pentoxyresorufm O-depentylase (PROD)
(CYP2B1/2) about 30-fold but barely induced ethoxyresorufm O-deethylase activity (EROD)
(CYP1 Al/2) (Nakajima et al., 1991). Imaoka and Funae (1991) also showed that 2-hexanone
induced the immunologically measured levels of several CYP450 isozymes, foremost
cytochromes (CYPs) 2B1, 2B2, 2C6, and 2E1. Minimal or equivocal induction was observed for
CYPs 1A1, 1A2, 2C7, and 4A3. The levels of CYPs 2C11 and 2C13 were slightly reduced
(Imaoka and Funae, 1991). However, it is not evident to what extent 2-hexanone might affect its
own metabolism via enzyme induction. Similarly, the enzymes that synthesize the glucuronides
of 2-hexanone metabolites, which were identified by Abdel-Rahman et al. (1976) in the urine of
2-hexanone-exposed rats, guinea pigs, and rabbits, have not been characterized further.
3.4. ELIMINATION
In humans exposed to 2-hexanone via inhalation at 10 or 50 ppm for 7.5 hours or to
100 ppm for 4 hours, unchanged 2-hexanone (but none of its metabolites) was found in expired
air, and neither 2-hexanone nor any of its metabolites was found in urine during or after exposure
(DiVincenzo et al., 1978). 2-Hexanone was no longer detected in expired air 3 hours after
exposure to 50 or 100 ppm. In two humans who received a single oral dose of 1-[14C]-
2-hexanone, breath excretion of 14CC>2 reached a peak within 4 hours then decreased slowly over
the next 3 to 5 days. Average overall recovery of the 14C-label in 8 days was 40% in breath and
26% in urine. Feces were not analyzed (DiVincenzo et al., 1978). These results suggest slow
clearance and possibly retention of 2-hexanone in humans exposed by this route.
In beagles exposed to 2-hexanone via inhalation at 50 or 100 ppm for 6 hours, 32 and
35%, respectively, of the inhaled vapor was excreted in the expired breath (DiVincenzo et al.,
1978). By 3 to 5 hours after exposure, 2-hexanone was no longer detected in expired air.
Excretion via other routes was not addressed.
In rats administered a single oral dose of l-14C-2-hexanone, DiVincenzo et al. (1977)
observed similar results as the above findings in humans. Radioactivity in breath accounted for
about 45% of the administered dose (5% was in unchanged 2-hexanone; 40% was in 14CC>2),
38% was found in urine, 1.1% was recovered in the feces, and about 15% remained in the
carcass. In male rats that received daily gavage doses of 2-hexanone at 400 mg/kg-day for
16
-------
40 weeks, very low concentrations of free 2-hexanone were detected in the urine from the third
week on. The highest level of 2-hexanone excreted in a 24-hour period was 20 jig, observed in
the 17th week (Eben et al., 1979). Similarly, free 2,5-hexanediol was found in the urine after
3 weeks and peaked in the 17th week. Free and conjugated 2,5-hexanedione were present in the
7th week, whereas excretion levels of the free form were consistent throughout the study. A
strong correlation was observed in this study between the onset of neuropathy and the urinary
concentration of 2,5-hexanedione, when 2-hexanone, 2,5-hexanedione, or 2,5-hexanediol was
administered orally to rats at 400 mg/kg-day.
Radiolabeled 14C from 1-14C-2-hexanone applied to the forearms of two human
volunteers was found in the breath and urine (DiVincenzo et al., 1978). In one subject,
eliminated amounts in urine and breath were similar, while, in the other subject, the levels in
breath were about three times higher than in urine. Fecal elimination was not measured.
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS
2-Hexanone was considered as a metabolite in two physiologically based toxicokinetic
(PBTK) models for n-hexane that focused on its neurotoxic metabolite, 2,5-hexanedione
(Hamelin et al., 2005; Perbellini et al., 1990). PBTK models that deal specifically with
2-hexanone were not identified. A blood/air partition coefficient of 127 for 2-hexanone
measured by using preserved human blood has been reported (Sato and Nakajima, 1979).
17
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS—CASE REPORTS, EPIDEMIOLOGIC STUDIES, AND
OBSERVATIONAL STUDIES
In humans, 2-hexanone vapor caused irritation of the eyes and respiratory tract during
acute exposure to relatively high concentrations. Men exposed to 0.23, 0.65, or 2% 2-hexanone
in air (9,422, 26,600, or 81,900 mg/m3) for 1 minute or less reported strong eye and nasal
irritation (Schrenk et al., 1936). Moderate eye and nasal irritation was reported after a brief
exposure to 0.1% (4,100 mg/m3). Peripheral neuropathy was reported in printers, furniture
finishers, and spray painters occupationally exposed to 2-hexanone (Davenport et al., 1976;
Mallov, 1976; Allen et al., 1974; Billmaier et al., 1974). Several studies described the
occurrence of neurological effects after the introduction of 2-hexanone into products used in the
occupational setting.
Davenport et al. (1976) reported the occurrence of symmetrical polyneuropathy in a
35-year-old male who was occupationally exposed to 2-hexanone among other compounds. The
patient had worked as a furniture finisher for several years, most recently spraying lacquer
compounds, sometimes without using a face mask. Initially, according to the manufacturer,
MiBK was present at a concentration of 20% in the finish, 12% in the thinner, and 7% in the
sealer. Toluene, xylene, n-butyl alcohol, and acetone were also present in various proportions.
After repeated inquiries, the manufacturer disclosed that, for the 6-month period before the onset
of the man's illness, 2-hexanone had been substituted for MiBK on a volume-for-volume basis in
the formulations of lacquers and solvents because of MiBK supply limitations. The patient first
noticed tingling in the soles of his feet and mild clumsiness of gait. Weakness progressed
rapidly to the upper extremities, resulting in a wheelchair-bound condition. Three months after
the onset of the first symptoms, routine hematology, blood chemistry, urinalysis, spinal fluid, and
analysis for heavy metals and porphyrins were normal. A biopsy of the sural nerve1 at the level
of the lateral malleolus revealed diffuse fibrosis and loss of nerve fibers. Several enlarged axons,
with and without myelin sheaths, with neurofibrillary tangles were evident. A clinical evaluation
3 months later indicated improved strength and ability to walk unassisted, though with some
residual unsteadiness of gait. Tendon reflexes distal to the elbows and knees were still absent.
The case report noted that a similar progressive distal extremity weakness developed in a
19-year-old coworker of the patient. This condition also improved following removal from
contact with lacquer products.
One probable and two definite cases of 2-hexanone-induced peripheral neuropathy were
found during an investigation of 26 painters who worked at Cannelton Dam or nearby Newburgh
1 The sural nerve is a sensory nerve that innervates the skin of the back of the leg and skin and joints on the lateral
side of the heel and foot.
18
-------
Dam on the Ohio River (Mallov, 1976). Two formulations of paint were used. The older
formulation contained 22% (weight/weight [w/w]) MiBK and 22% (w/w) methyl isoamyl
ketone. In the newer otherwise identical formulation, these solvents were replaced by 44%
(w/w) 2-hexanone. While both paint formulations were reported to contain 3.1% (w/w) of the
known neurotoxicant triorthocresyl phosphate, this substance was not found in two bulk samples
of the 2-hexanone paint formulation. One definite case of peripheral neuropathy was that of a
42-year-old man, a painter for 10 years, who had been painting Cannelton Dam from September
1972 until August 1973. His initial signs, including weight loss, numbness and tingling of feet,
and progressive weakness in both lower extremities that progressed to his upper extremities as
well, began in July 1973. Weakness progressed until he could no longer stand without
assistance. Lower extremity reflexes became absent and an electromyogram was abnormal.
Blood and urine lead analysis indicated slightly elevated levels but not sufficient to cause effects.
The second case was that of a 35-year-old man who had been painting since he was 14 years old.
He painted at Cannelton Dam from April to October 1973. He felt well until about 4 weeks prior
to the termination of painting at Cannelton but eventually became unable to rise from a sitting
position without help. Urine lead levels were in the lower normal range. The third painter had
worked at either Cannelton Dam or Newburgh Dam from September 1970 until November 1973.
He also felt well until about 4 weeks prior to termination of painting. While he experienced
weakness in his extremities, he remained able to walk but reported above-normal episodes of
falling and dropping things. He was not examined by a physician until 3.5 months after the onset
of symptoms, at which time absent ankle reflex, foot weakness, and diminished sensation were
noted. None of the three patients had a history of alcoholism or family history of neurological
disease or took medications.
A cross-sectional study of peripheral neuropathy among employees at a coated fabrics
plant in Ohio was started when it was noted that six workers from the print department had
developed severe peripheral neuropathy (five hospitalized, one seen as outpatient) between April
and August 1973 (Allen et al., 1974; Billmaier et al., 1974). The plant produced plastic-coated
printed fabrics that were used mainly for wall coverings and automobile interiors. Processing
steps included mixing, calendering, laminating, coating, printing, embossing, inspecting, and
shipping. Starting in September 1973, all 1,157 employees of the plant (including the original
six cases) were screened by using electromyography and nerve conduction testing. A total of
192 employees were referred for detailed neurological evaluation. On the basis of these
examinations, it was concluded that 68 employees had definite signs, symptoms, and
electrodiagnostic findings of peripheral neuropathy. Severity ranged from mild
(electrodiagnostic findings but no physical symptoms) to moderate (distal sensory loss) to severe
(distal muscle weakness and sensory loss). There were a total of nine severe cases, including the
original six cases. Cases with possible causes other than a toxic chemical (e.g., diabetes) were
19
-------
not included in the analysis but were identified in the presentation of results. The distribution of
cases within the plant is shown in Table 4-1.
Table 4-1. Prevalence of peripheral neuropathy among employees of a
coated fabrics plant
Workplace
Non-print departments
Print department (total)
Operators
Helpers
Foreman
Service helper
Not known
Total
Number of cases
30a
38b
27
10
0
1
0
68
Number of employees
examined
984
173
69
59
21
16
8
1,157
Prevalence (%)
3
22C
39C
1?c
0
6
0
6
"Includes 18 persons with diabetes or other conditions that can cause or contribute to neuropathy.
blncludes one person with diabetes and one person on isoniazid therapy.
Significantly elevated compared with non-print departments (p < 0.001) by using the chi-square test.
Source: Billmaieretal. (1974).
The prevalence of peripheral neuropathy was significantly higher among print department
employees than among employees from other departments (22 vs. 3%,/> < 0.001). All nine
severe cases were print department workers. Within this department, prevalence was highest
among printer operators (39%, p < 0.001 compared with non-print-department employees), who
spent almost 100% of their time near the printing machines. Prevalence among helpers (17%)
who spent roughly 50% of their time near the printing machines was also significantly elevated
compared with non-print-department employees (p < 0.001). There was a 6% prevalence among
service helpers who were in and out of the premises (one case among service helpers was a pan
washer who used the solvent for cleaning). Among manufacturing departments other than the
print department, the prevalence of neuropathy ranged from 0 to 6.7%. No cases of peripheral
neuropathy were observed in supervisory personnel who remained at a distance from the
machines or in office personnel.
In addition to job category, incidence of neuropathy was also associated with working
overtime (print operators with definite neuropathy averaged 47.2 hours/week versus
42.0 hours/week for those without neuropathy [p < 0.01]) and with eating on the job (data not
shown). Each employee generally worked on the same machine all the time. No differences in
neuropathy incidence were found based on the type of printing machine or the area in which the
machine was located; data were insufficient to correlate illness with individual machines.
Among print-department employees, there were no significant differences in neuropathy
incidence related to age or tenure in the department; 90% of cases had presented within the
20
-------
previous year, and only 5% of the cases were known to have medical conditions that could cause
or contribute to neuropathy. Among non-print-department employees, cases were clustered in
older (40+) employees (p < 0.001); only 53% had onset within the previous year, and 60% of
these cases were known to have diabetes or other medical conditions that could cause or
contribute to neuropathy unrelated to compound exposure.
In the search for the etiologic agent, other chemicals known to cause peripheral
neuropathy were ruled out, either by clinical tests on workers or because they were not used in
the plant. Based on an investigation into the relationship between the cases of peripheral
neuropathy and the distribution of the roughly 275 chemicals used in the plant, the most likely
agent appeared to be contained in the solvents used as ink thinners and cleaners. These had
previously consisted of methyl ethyl ketone (MEK) and MiBK, but, starting in August 1972, the
latter was phased out and gradually replaced by 2-hexanone, which reached maximal usage in
December 1972. Substitution of 2-hexanone for MiBK was the only major change in the
production process in the previous 7 years. In September 1973, the print department was closed
for a month and 2-hexanone was removed from production materials. Thus, there was a
13-month window of exposure to 2-hexanone.
In addition to exposure to 2-hexanone, affected workers were also exposed to high
concentrations of MEK that sometimes exceeded threshold limit values (TLVs). MEK by itself
does not produce this type of neuropathy in animal studies but can potentiate the effects
produced by 2-hexanone (Saida et al., 1976). Thus, the presence of MEK in the coated fabrics
plant study could contribute to an overestimation of the risk associated with exposure to
2-hexanone itself. Workplace levels for 2-hexanone and MEK from this study are presented in
Table 4-2.
Table 4-2. Results of area atmospheric sampling for MEK and 2-hexanone
in a coated fabrics plant
Median
Front of print
machine"
MEKb
104
109
124
162
220
453
565
570
670
2-Hexanonec
2.3
2.6
4.1
5.1
5.8
9.7
11.5
19.8
21.7
Back of print machine"
MEKb
85
265
401
440
603
608
725
750
763
2-Hexanonec
2.5
3.0
9.0
9.8
21.7
23.9
48.6
49.9
156.0
Wind-up area"
MEKb
39
44
47
49
127
143
250
289
338
2-Hexanonec
1.0
2.0
2.0
2.6
5.9
6.0
7.9
9.8
17.5
aValues are in ppm, listed from lowest to highest result obtained for each solvent at each work location.
bTLV = 200 ppm.
CTLV = 100 ppm.
Source: Billmaier et al. (1974).
21
-------
Another confounding factor for this study is that exposure may not have been limited to
the inhalation route. Poor work practices documented at the plant included washing hands in
solvent, using solvent-soaked rags to clean equipment, and eating in work areas. Dermal and
even oral exposure is likely to have occurred. The significance of exposure by these routes is
suggested by the observations that eating on the job was associated with the development of
neuropathy and that a worker whose job involved washing pans with the solvent was the only
afflicted print-department worker other than the print operators and their helpers. As discussed
in Sections 3.1.2 and 3.1.3, 2-hexanone is absorbed readily through the skin and gut and can
produce neuropathy by both routes in animals.
The researchers reported that patients removed from 2-hexanone exposure showed
significant and consistent improvements. They also performed a study of workers at a
comparable coated fabrics plant in California that produced the same products as the one in Ohio
but without the use of 2-hexanone. Electrodiagnostic studies were conducted on 21 solvent-
exposed workers at the California plant, but no peripheral neuropathy was found.
4.2. ACUTE, SUBCHRONIC, AND CHRONIC STUDIES IN ANIMALS
4.2.1. Oral Exposure
4.2.1.1. Acute and Short-term Oral Exposure
Range-finding toxicity data by Smyth et al. (1954) list an oral median lethal dose (LD50)
of 2.59 g/kg of 2-hexanone for rats, while Tanii et al. (1986) provide an oral LDso of 2.43 g/kg
for mice. Details for either study (Tanii et al., 1986; Smyth et al., 1954) are limited.
4.2.1.2. Subchronic Toxicity Studies
90-Day study: hens
Abou-Donia et al. (1982) exposed adult leghorn laying hens (Gallus gallus domesticus),
three per group, to a dose of either 0 or 100 mg/kg technical grade 2-hexanone containing 70%
2-hexanone and 30% MiBK for 90 days. Hens were observed for 30 days after final dose. Body
weights were monitored weekly, and hens were examined daily. Decreased body weight was
observed in the treated hens. Hens were 89 ± 4% of their initial weight at termination compared
with controls, which were 115 ± 5% of their initial weight. Mild ataxia was observed at 12 ± 1
days with progression to severe ataxia by 50 ± 1 days. Other neurotoxicity outcomes among the
treated animals are outlined in Section 4.4.1.1.
16-Week study: female Wistarrats
Homan et al. (1977) conducted a 120-day drinking water study with female Wistar rats.
2-Hexanone (purity not stated) was administered in drinking water to rats (five/group) at 0, 0.65,
or 1.3% (0, 480, or 1,010 mg/kg-day). A dose-dependent decrease in food consumption was
observed in exposed animals versus controls. Water consumption in exposed animals was
22
-------
reduced to about half that of controls. Animals exposed to 0.65 or 1.3% 2-hexanone experienced
a 45.5 or 68.8% reduction in body weight gain, respectively. A dose-dependent decrease in
absolute liver weight was observed in exposed animals. Absolute kidney weights were
increased, and there was a dose-dependent increase in relative kidney weights. A summary of
the data for diet and water consumption, body weight gain, and organ weights is provided in
Table 4-3. Neurotoxicity outcomes among the treated animals are outlined in Section 4.4.1.
Table 4-3. Gross observations in rats exposed to 2-hexanone in drinking
water for 120 days
Dose
(mg/kg-day)
0
480
1,010
Food intake"
(g/day)
17.99
16.90
12.90
Water intake"
(mL/day)
32.29
17.98
17.33
Body weight
gain3 (g)
110.2
60.0b
34.3b
Liver weight"
Absolute (g)
10.30
9.01
7.80b
Relative
3.10
3.35
3.38
Kidney weight"
Absolute (g)
1.97
2.21
2.10
Relative
0.60
0.82b
0.92b
"Values are means from five animals/group
bSignificantly different from controls, p < 0.01.
Source: Homanetal. (1977).
40-Week study: male Wistar rats
Eben et al. (1979) treated male SPF-Wistar rats daily with 400 mg/kg 2-hexanone (98%
pure) by gavage for 40 weeks. Body weight gain in treated animals was less than in controls; a
decrease in body weights was observed from the 17th to the 25th weeks, followed by a slight
increase until study completion. There were also symptoms of neurotoxicity in the treated
animals (see Section 4.4.1).
4.2.1.3. Chronic Toxicity Study
13-Month study: CD/COBS(SD) rats
O'Donoghue et al. (1978) conducted a 13-month study in male CD/COBS(SD) rats. This
study is an unpublished study; accordingly, it was externally peer reviewed by EPA in December
2007. The animals' drinking water contained 0, 0.25, 0.5, or 1.0% (0, 143, 266, or 560 mg/kg-
day) 2-hexanone (96% pure, containing 3.2% MiBK and 0.7% unknown contaminants).
2-Hexanone produced a dose-dependent reduction in body weight at all doses tested. The effect
was present by the second week in the two highest dose levels and by the third week in the low-
dose group. A statistically significant increase in liver weight was found in the highest dose
group compared with all groups except the 0.5% group. The 0.5 and 0.25% groups showed
dose-dependent increases in relative liver weights compared with controls. A statistically
significant increase in relative kidney weights was present between the 1.0% 2-hexanone group
and all other groups and between the 0.5% group and all other groups. Similarly, a statistically
significant increase in relative testes weight was found between the 1.0% group and all other
23
-------
groups. A summary of the body weight and organ weight data is present in Table 4-4
(O'Donoghue et al., 1978).
Table 4-4. Pathological changes in rats exposed for 13 months to 2-hexanone
Control
2-Hexanone
(0.25% or 143 mg/kg-day)
2-Hexanone
(0.5% or 266 mg/kg-day)
2-Hexanone
(1.0% or 560 mg/kg-day)
Body
weight"
710
685
612
448
Liverb
Absolute
26.71 ±2.02
24.99 ±4.33
25.06 ± 2.04
20.73 ± 2.95
Relative
3.64 ±0.41
3.97 ±0.43
4.22 ±0.43
4.62±0.32C
Kidneyb
Absolute
4.66 ±0.53
4.58 ±0.69
5.33 ±0.31
4.86 ±0.38
Relative
0.63 ± 0.87
0.73 ± 0.05
0.90±0.12C
1.10±0.23C
Testesb
Absolute
2.99 ±0.81
3.24 ±0.38
3. 16 ±1.04
3.29 ±0.26
Relative
0.40 ±0.11
0.52 ±0.08
0.54 ±0.19
0.75±0.17C
"Values are means of 10 animals. No statistical test results on body weight were reported by the authors.
bValues are mean ± SEM based on four or five animals per group.
Statistically different from controls, p < 0.05.
Source: O'Donoghue etal. (1978).
Clinical neurological deficits were noted in animals exposed to either 0.5 or 1.0%
2-hexanone. Severe deficits including decreased extension of the hind limb, hind-limb
weakness, and muscular atrophy of the hind-limb musculature were noted among animals treated
with 1% 2-hexanone. Deficits among animals exposed to 0.5% 2-hexanone were slight and did
not result in clinical progression. Evidence of axonal swelling was noted at all dosing levels of
2-hexanone. Neurological effects are discussed in further detail in Section 4.4.1.1.
To determine whether the concentration of MiBK, a contaminant in the 2-hexanone
formulation used by O'Donoghue et al. (1978) and a CYP450 inducer, may have altered the
observed toxicity of 2-hexanone, other studies were evaluated that used MiBK as the test article.
In a 13-week gavage study, 30 male and female Sprague-Dawley rats were treated daily with 0,
50, 250, or 1,000 mg/kg MiBK (Microbiological Associates, Inc. [MAI], 1986). At the middle
and high doses, adverse effects were observed in the liver and kidney, which progressed in
severity in the high dose animals. No treatment-related effects of any kind were observed at
50 mg/kg-day. It should be noted that the dose of 50 mg/kg-day is 13 times higher than the
concentration of MiBk in the 2-hexanone studies, where MiBK is listed as a contaminant except
in the study of Abou-Donia et al. (1982), where the concentration of MiBK in the test material is
30%, or 30 mg/kg-day. The Carnegie-Mellon Institute of Research (1977) conducted a 120-day
drinking water study with 1.3% MiBK and using female HLA Wistar rats. The authors
estimated the dosage to be 1,040 mg/kg-day. The only statistically significant findings were
increased mean absolute and relative kidney weights in treated rats compared with controls.
Histopathological examination revealed renal tubular cell hyperplasia in only one of five of the
treated rats. No exposure-related histopathological changes were found in other organs. Based
24
-------
on the foregoing, it can be concluded that the dosage of MiBK received as an impurity in the
study by O'Donoghue et al. (1978) did not contribute to the observed 2-hexanone-related effects.
Other than increased relative organ weights, O'Donoghue et al. (1978) did not observe adverse
effects in the kidney or liver of treated animals, despite these organs being the target organs of
toxicity in experimental studies with MiBK from both the oral and inhalation routes (U.S. EPA,
2003a).
4.2.2. Inhalation Exposures
4.2.2.1. Acute and Short-term Toxicity Studies
No acute inhalation toxicity studies of 2-hexanone were identified. The NLM (2005)
Hazardous Substances Data Bank states that a 4-hour exposure of rats to 4,000 ppm 2-hexanone
did not kill all animals, while exposure to 8,000 ppm for 4 hours was an LDioo. Abdo et al.
(1982) reported the death of one out of five hens exposed continuously to 200 ppm 2-hexanone
(70% purity). No deaths were reported in hens exposed to 100 ppm or lower (Abdo et al., 1982).
4.2.2.2. Subchronic Toxicity Study
11-Week study: male rats
Groups of five male rats (Crl:COBS/CD[SD]BR) were exposed to 0 or 700 ppm (0 or
2,870 mg/m3) 2-hexanone (purity 96.1%) 72 hours/week for 11 weeks (Katz et al., 1980). The
exposure schedule was as follows: two 20-hour periods and two 16-hour periods, Monday
through Friday, separated by 8-hour nonexposure periods. Total white blood cell counts of
treated animals were significantly lower than those of controls (approximately -40%; p < 0.05);
no other differences were noted in clinical chemistry or hematological values. Gross
examination of treated animals revealed marked atrophy of the hind-limb musculature, depletion
of adipose tissue, and significantly decreased absolute and relative testicular weight (p < 0.05).
Histopathological examination was performed on selected tissues, including lung and trachea
(but not nasal cavities), eye, digestive tract, pancreas, thyroid, parathyroid, testes, epididymides,
spleen, bone marrow, mesenteric lymph nodes, thymus, and nervous system. Atrophy of
testicular germinal epithelium and grossly enlarged axons in the brain stem and cerebellum were
observed in treated animals. No damage to bone marrow was evident despite the low white
blood cell count. Although no discussion of findings in the lung or trachea was presented, the
implication is that there were no treatment-related changes in these tissues. The treatment group
developed signs of neurotoxicity (weakened hind- and forelimb grasp) by the second week of
exposure, progressing to severe hind-limb weakness by 71 days, and showed decreased weight
gain.
25
-------
4.2.2.3. Chronic Toxicity Studies
72-Week study: male Spr ague-Daw ley rats
Krasavage and O'Donoghue (1977) exposed groups of male Sprague-Dawley rats
(18/group) to 0, 100, or 330 ppm (0, 410, or 1,353 mg/m3) 2-hexanone (purity not specified)
6 hours/day, 5 days/week for 72 weeks. Clinical signs (observed daily and examined weekly),
body weight (recorded weekly), and water consumption (at 15, 22, 32, and 44 weeks of
exposure) were monitored. Beginning at 4 weeks and continuing at approximately 6-week
intervals for the first 52 weeks, unspecified numbers of animals were killed for microscopic
examination of an extensive list of tissues, including the trachea and lung. Body weights and
weight gain were comparable among groups until the 20th week. Thereafter, body weights of the
high-concentration animals fell behind those of controls (data presented graphically without
statistical analysis); a visual estimate of the graphic presentation suggested that the body weights
of high-concentration animals were at least 10% less than those of controls. After 36 weeks of
exposure, body weight gain in the low-concentration group also began to lag behind that of
controls. Water intake was comparable among groups.
Gross postmortem findings revealed no compound-related changes. Low-concentration
animals did not develop clinical signs attributed to 2-hexanone exposure or morphologic lesions
of neuropathy. Histopathological evidence for neuropathy in high-concentration rats was
equivocal. Neurological effects are discussed in further detail in Section 4.4.1.2. Spontaneous
lesions were present in the urogenital, cardiovascular, and endocrine systems of both treated and
control animals and were therefore not attributed to 2-hexanone exposure by the study authors
(Krasavage and O'Donoghue, 1977).
2-Year study: cats
Groups of four domestic shorthair cats were exposed by inhalation to 0, 100, or 330 ppm
(0, 410, or 1,353 mg/m3) 2-hexanone (purity not specified) for 6 hours/day, 5 days/week for
2 years (O'Donoghue and Krasavage, 1979). Clinical signs and body weights were monitored.
Serum was sampled after 30, 90, and 128 exposures to determine the levels of 2-hexanone and
two of its metabolites, 5-hydroxy-2-hexanone and 2,5-hexanedione. Each sample set involved
collection of serum on a Monday prior to daily exposure, the following Tuesday prior to daily
exposure, the following Friday prior to daily exposure, immediately after daily exposure, and one
and three quarter hours after daily exposure. Sera from high-dose and control animals were also
analyzed for sodium, potassium, chloride, and calcium levels. Cats were sacrificed at the end of
the treatment and were subjected to necropsy and histopathological examinations.
No clinical neurological effects attributed to exposure to 2-hexanone were identified
except that cats anesthetized with sodium pentobarbital following a 6-hour exposure had
prolonged sleeping times (O'Donoghue and Krasavage, 1979). No compound-related changes of
body weight or serum electrolyte values were found. Serum levels of 2-hexanone and the two
26
-------
metabolites 5-hydroxy-2-hexanone and 2,5-hexanedione were below the detection limit on
Monday morning following a 2-day non-exposure period. With the exception of
2,5-hexanedione in the 330 ppm group (1,353 mg/m3), serum levels on Tuesday morning
following a 6-hour exposure after 30 days of exposure remained below the detection level. Of
the three substances measured, 2-hexanone cleared the serum more quickly than 5-hydroxy-
2-hexanone, which cleared more quickly than 2,5-hexanedione. Biopsy examinations through
the first 9 months of exposure were unremarkable and did not serve as an early detection method
for neuropathy. Gross postmortem findings revealed no compound-related changes. General
histopathological examinations showed no compound-related changes other than in the nervous
system and musculature. Neurological effects are discussed in further detail in Section 4.4.1.2.
4.2.3. Dermal Exposure
90-Day study: hens
Abou-Donia et al. (1985a) treated leghorn laying hens (n = 5) with 2-hexanone (99%
pure; topical application, 1 mmol/kg). The chemical was applied daily with a micropipette over
an area of 10 cm2 on the unprotected back of the neck for 90 days. All hens developed gross
ataxia. At sacrifice, no changes were observed in treated versus control animals when compared
for size, shape, weight, or color. Equivocal histopathological changes were present in the spinal
cord of two hens. These histopathological changes were characterized by swollen axons without
obvious fragmentation of the axon or myelin sheath. No precautions against licking were
mentioned in the study, so ingestion of 2-hexanone may have taken place.
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION
4.3.1. Oral Exposure
No standard two-generation studies or other studies of reproductive and developmental
effects following oral administration of 2-hexanone were identified.
4.3.2. Inhalation Exposure
In a developmental study, Peters et al. (1981) exposed groups of 25 pregnant F344 rats to
0, 500, 1,000, or 2,000 ppm (0, 2,048, 4,096, or 8,193 mg/m3) 2-hexanone (purity not stated),
6 hours/day on gestational days 1-21. A separate control group was maintained for each
exposure group and the high-concentration controls were pair fed. Respective controls were
exposed to ambient air in similar chambers to those of their exposed counterparts. Sexually
mature female rats were impregnated and placed in exposure chambers 6 hours/day throughout
gestation. Four weeks postdelivery, the dam was separated from the pups. The maternal
500 ppm group along with its control was terminated before 3 weeks because of an apparent
lapse in care during which offspring were "unable to reach food and water," resulting in reduced
weight gain in this group. The pups in the control, 1,000, and 2,000 ppm groups were observed
27
-------
over a lifetime. At 4 (weaning), 8 (puberty), and 14 weeks (adult) and at 18-20 months of age
(geriatric), five males and five females were taken, one per litter, for gross and histopathological
studies and for measurement of organ/body-weight ratios. At different periods of development
(weaning, puberty, and adult), offspring underwent behavioral testing. Pentobarbital sleeping
time was also measured in pubescent and geriatric animals in the high-dose and control groups.
Survival in the 2,000 and 1,000 ppm dams was not affected by treatment. High-dose
dams appeared sluggish after exposure but seemed to have recovered by the next exposure. Hair
loss, lack of muscular coordination, and weakness were observed in "several" dams at the
highest concentration after 20 days of exposure. Abnormal sniffing in the air was reported for
dams in the 1,000 ppm group. Maternal gestational body weight gain was decreased by 14 and
10% in the dams exposed to 2,000 and 1,000 ppm, respectively. Rats in the 2,000 ppm exposure
group were observed to eat less than the controls. No unusual behavior or change in maternal
gestational growth was reported for the 500 ppm dams. Histopathology and neurotoxicity
evaluations were not performed in the dams.
2-Hexanone exposure was found to result in statistically significant decreases (p value
not reported) in litter size and pup weight observed at the highest exposure level (Peters et al.,
1981). However, maternal toxicity, manifested as decreased maternal body weight during
gestation, was also evident in high-dose dams, suggesting that maternal toxicity might have
affected fetal growth. There was a significant decrease in the number and weight of live
offspring of dams in the 2,000 ppm exposure group. A lifelong, statistically significant,
concentration-related reduction in growth was observed in male offspring. Only a slight
treatment-related effect on body weight was seen in female offspring. Organ weights in
weanling, pubescent, and geriatric offspring were unaffected by treatment, but brain weight in
adult 1,000 ppm offspring was significantly increased compared to that of control. Organ
weights were not measured in high-dose adult offspring. The authors did not report any gross
skeletal alterations. Beginning at 40 weeks of age, offspring of dams treated with 1,000 or
2,000 ppm showed a 3-5% decrease in survival relative to controls. The incidence of
pathological lesions and the types of lesions contributory to death were not significantly different
in treated and control groups (Table 4-5).
28
-------
Table 4-5. Summary of pathological lesions in 40-week-old offspring of dams
exposed to 2-hexanone during gestation
Number of
animals dead or
sacrificed3
Pituitary tumor
Pituitary
hemorrhage
Diaphragmatic
hernia
Ovarian cysts
Mottled testes
Control
Male
57
1
2
1
0
26
Female
57
3
0
1
2
0
Total
114
4
2
2
2
26
%
-
3.5
2
2
2
23
1,000 ]
Male
37
1
0
0
0
16
Female
34
1
0
1
7
0
ppm
Total
71
2
0
1
7
16
%
-
o
6
0
1.4
10
23
2,000 ppm
Male
24
1
1
1
0
1
Female
22
0
2
2
8
0
Total
46
1
3
3
8
1
%
-
2
6.5
6.5
18
2
aAnimals include those dying subsequent to weaning in addition to those sacrificed at 78 ± 2 weeks of age.
Source: Peters etal. (1981).
Standard hematological tests (hemoglobin, red blood cell count, white blood cell count,
lymphocytes, mean corpuscular hemoglobin, packed cell volume) showed no significant
treatment effect on the processes involved in blood cell formation and function (Peters et al.,
1981). Clinical chemistry findings were limited to a concentration-related decrease in creatinine
phosphokinase activity in pubescent offspring, with values in the 1,000 and 2,000 ppm groups
significantly lower (p < 0.05) than in controls. In geriatric offspring, there were significant
increases (p < 0.05) in serum alanine aminotransferase activity in the 1,000 and 2,000 ppm
groups and sodium in the 2,000 ppm group. The only lesions showing a significant
concentration-response relationship (p < 0.05, Fisher's exact test conducted for this assessment)
at the time of geriatric sacrifice were ovarian cysts that had 4% (2/57), 21% (7/34), and 36%
(8/22) incidences in the control, 1,000 ppm, and 2,000 ppm females, respectively.
In pubescent high-dose male offspring, pentobarbital sleep time was significantly
increased (p < 0.05) compared with controls. No significant changes in pentobarbital sleep time
were noted in pubescent females or geriatric offspring of either sex. Behavioral alterations were
reported in the offspring of pregnant rats exposed to 1,000 or 2,000 ppm 2-hexanone. These
effects consisted of reduced activity in the open field, increased activity in the running wheel,
and deficits in avoidance conditioning. Offspring of treated dams (both dose levels) clung to an
inclined screen longer than offspring of controls at all ages (newborn, weanling, puberty, and
adult) except geriatric in which results were similar to those of controls. For offspring in the
puberty and adult categories, pronounced sex differences were noted; females in all exposure
categories (including controls) were clinging from 24-100% longer than males. However, the
biological significance of this observation is unknown. There was a decreased rate of avoidance
learning in puberty-aged females of treated dams and increased random movement in both
puberty-aged and adult offspring of treated dams. Behavioral tests in most cases indicated that
29
-------
maternal exposure to 2-hexanone was associated with hyperactivity in the young and decreased
activity in the geriatric stage, which the authors (Peters et al., 1981) speculated to be due to
premature aging resulting from the earlier hyperactivity. It is not clear whether these effects are
the result of transplacental exposure to 2-hexanone or of postnatal exposure to 2-hexanone
and/or its metabolites via the milk of the exposed dams.
4.4. OTHER ENDPOINT-SPECIFIC STUDIES
4.4.1. Neurotoxicity Studies
4.4.1.1. Oral Exposures
90-Day study: hens
Abou-Donia et al. (1982) treated hens with either a single gavage dose of 2,000 mg/kg or
a 90-day daily dose of 100 mg/kg of 2-hexanone (technical grade: 70% 2-hexanone,
30% MiBK). Clinical assessment of neurotoxicity was graded by classifying the degree of ataxia
before paralysis as follows: TI, mild ataxia, characterized by diminished leg movement and
reluctance to walk, with hens tending to slide on the floor or fly; T2, gross ataxia, characterized
by a change in gait and disturbance of leg movement; T3, severe ataxia, with severe leg weakness
manifested by unsteadiness and occasional falling on the floor; 14, ataxia, with near paralysis,
marked by inability to walk. For the acute exposure, mild weakness was observed on the day of
administration, followed by apparent recovery in 4-5 days. In the subchronic (90-day) phase of
the same study, all treated hens (n = 3) developed severe ataxia. All three hens improved to a
stage of gross ataxia after cessation of 2-hexanone administration. There was also evidence of
histopathological changes, including swelling or degeneration of thoracic and lumbar regions of
the spinal cord. No neurotoxicity information was provided about the control group for either
the single gavage or subchronic experiments.
90-Day study: rats
Krasavage et al. (1980) administered 660 mg/kg 2-hexanone (96% pure) by gavage to
male CD/COBS(SD) rats for up to 90 days. The authors considered severe hind-limb weakness
or paralysis, as exhibited by "dragging" of at least one hind foot, to be clear indication of
neuropathy. When this endpoint was reached, the treatment was terminated and the animal was
processed for histological examination. There was a time- and dose-dependent depression in
body weight gain and feed consumption. Treated animals consumed an average of 21 grams/day
versus 28 grams/day for controls. The body weights of experimental and control animals at
study completion were approximately 400 and 600 grams, respectively. Histological
examination of nerve tissue collected at termination revealed morphologic changes indicative of
giant axonal neuropathy, which included multifocal axonal swellings, myelin infoldings, and
paranodal myelin retraction. In this study, atrophy of the germinal epithelium of the testes was
30
-------
also observed, but the statistical significance of this observation was not addressed (Krasavage et
al., 1980).
120-Day study: female Wistar rats
Homan et al. (1977) conducted a 120-day drinking water study with female Wistar rats
(five/group). 2-Hexanone (purity not stated) was administered in drinking water at 0, 0.65, or
1.3% (0, 480, or 1,010 mg/kg-day) (for further experimental details see Section 4.2.1.2).
Neurological evaluations were conducted to assess balance, strength, coordination, and behavior.
Performance was scored for each of the following 10 criteria: posture, gait, palpebral reflex,
startle reflex, flexor reflex, extensor reflex, placing reflex, hopping reaction, righting reflex, and
clinging reaction. Score ranged from 0 to 2, with 0 indicating normal and 2 being clearly
deficient. The net score for each rat was calculated as the sum of the individual test scores.
Scores were tabulated, ranked, and analyzed by using the Kruskal-Wallis ranks sum test. The
rank values (statistics generated from the Kruskal-Wallis test) for each treatment group for a
given day of analysis were then averaged to generate a mean rank and standard deviation. A
summary of the mean rank (mean of the values generated from the Kruskal-Wallis test) and
standard deviation is provided in Table 4-6. Gross pathological evaluation revealed mild atrophy
affecting skeletal muscles of the hind limbs in two of five animals in the 0.65% group and slight
to severe atrophy of skeletal muscles (most pronounced in muscles of the hind limbs) affecting
four of five animals in the 1.3% group (Homan et al., 1977).
Table 4-6. Time course of neuropathy scores following exposure of rats to
2-hexanone in drinking water
Treatment
Control
0.65% 2-Hexanone
1.3% 2-Hexanone
Analysis after number of treatment days
46
57
80
110
Mean rank value
26.1 ± 9.1
32.0 ± 14.9
37.5 ± 12.6
15.0 ± 0.0
30.6 ± 12.7
41.0 ± 5.7
22.1 ± 12.8
30.9 ± 9.2
40.0 ± 13.7
17.5 ± 0.0
21.5 ± 8.0
47.2 ± 2.8a
""Statistically significant versus controls within that particular number of treatment days, p < 0.05.
Source: Homan etal. (1977).
24-Week study: guinea pigs
Abdel-Rahman et al. (1978) administered 2-hexanone (purity not stated) in drinking
water to guinea pigs (five/group, sex not stated) at 0, 0.1, or 0.25% (approximately 0, 97, or
243 mg/kg-day) for 24 weeks. Bibs were used to prevent dermal absorption by inadvertent
contact of the animals' bodies with the solvent. The body weight of the guinea pigs was
monitored each week up to the eighth week of the study. At the end of the seventh week, animals
31
-------
exposed to 0.25% 2-hexanone weighed an average of 600 grams versus 440 grams in controls.
Similarly, animals exposed to 0.1% 2-hexanone weighed 618 grams by the eighth week
compared with 490 grams in controls. Decreased locomotor activity may have contributed to
increased body weights. The average motor activity counts in animals exposed to 0.25%
2-hexanone were 714 ± 130 compared to 1,173 ± 201 in controls. Locomotor activity for the
0.1% exposure group was not reported. Pupillary response to light (measured by change in
pupillary diameter in response to an intense 2-second light stimulus) was abnormal in high-dose
animals for the first 5 weeks of treatment as shown in Table 4-7 (data not provided for 0.1%
2-hexanone). The authors (Abdel-Rahman et al., 1978) reported that, by the 24th week of the
study, a greatly impaired pupillary response was observed for all treatment groups (data not
provided in the report).
Table 4-7. Effect of 2-hexanone on guinea pig pupillary response of both
eyes
Treatment
Control
0.25% 2-
Hexanone
Week
1
Right3
1.83 ±0.00
1.33 ±0.19
Left3
1.66 ±0.17
1.33±0.01b
2
Right
1.6±0.1
1.05 ±0.15
Left
1.67 ±0.19
1.17±0.01b
3
Right
1.6 ±0.06
0.67 ±0.17
Left
1.7 ±0.05
0.92±0.08b
5
Right
1.5 ±0.05
0.59±0.14b
Left
1.5±0.1
0.71±0.04b
aValues represent the mean ± SEM of the change in pupillary diameter.
bStatistically significant from controls (p < 0.001) as calculated by study authors.
Source: Abdel-Rahman etal. (1978).
40-Week study: rats
Eben et al. (1979) treated male SPF-Wistar rats daily by gavage with 400 mg/kg
2-hexanone (98% pure) for 40 weeks. The authors stated that this treatment did not cause
neuropathic symptoms; however, from the 17th week the authors noted that the animals exhibited
weakness of the hind limbs, which continued until the 28th week. Thereafter, an improvement
was observed. No further details were provided.
13-Month study: rats
As previously mentioned in Section 4.2.1.3, O'Donoghue et al. (1978) conducted a
13-month study in male CD/COBS(SD) rats. Each group of 10 rats was exposed to drinking
water containing 0, 0.25, 0.5, or 1.0% (0, 143, 266, or 560 mg/kg-day) 2-hexanone (96% pure,
containing 3.2% MiBK and 0.7% unknown contaminants). Body weight and neurological
examinations were performed weekly. At the end of the study, a dose-dependent reduction in
body weight was noted among all dose groups. All but one animal were found to have some
32
-------
evidence of neurotoxicity. Other than neural effects and body weight changes, no compound-
related clinical signs were found.
Clinical neurological deficits were found only in animals receiving 0.5 or 1.0%
2-hexanone. Deficits were recorded as slight if there was incomplete extension of the hind limb
and just detectable widening of the hind-limb stance; moderate if there was obvious weakness,
incomplete extension of the hind limbs, and waddling; and severe if there was dragging of at
least one hind paw. In the 1.0% group, all the animals exhibited severe deficits. Gross
pathological examination revealed observable muscle atrophy of hind-limb and lumbar muscles
at this high-dose level. Progression of the clinical findings to a more severe state did not occur
with time in the 0.5% group. In addition to the aforementioned changes, animals receiving 1%
hexanone in their drinking water displayed loss of tone with grossly observable atrophy of the
hind-limb musculature and axial muscles of the lumbar area. Weakness of the forelimbs with
some muscle atrophy was observable in three of nine rats at the end of the study. Pain sensation,
as judged by toe pinch, remained intact, but motor response, such as flexor response, was easily
overcome. It was noted that tactile placing in the hind limbs could be elicited even in rats with
severe weakness. Bowel and bladder functions remained normal. The clinical course was highly
variable, with improvements in the clinical symptoms being very common; thus, while all
animals showed slight deficits on at least two of the weekly examinations, they showed
improvements during other weeks.
Evidence of neuropathy was most common in the giant axons of animals of each dose
level. In peripheral nerves from the 1.0% group, swelling of giant and other axons was common.
Myelin infoldings into the axoplasm were more common than in controls. Myelin ovoids were
frequently found along with degenerating axons. The second most common site of neural
degeneration was in the spinal cord, particularly in the ventral and ventromedial funiculi of the
thoracolumbar segments. The changes were similar to those found in peripheral nerves. In
plastic embedded sections, an additional early change was noted, which consisted of clumping of
axonal organelles in otherwise normal peripheral or central axons. Examination of the dorsal
root ganglia did not reveal any effect on cell bodies, but in three animals single swollen axons
were found in adjacent roots, indicating a very minimal effect. Axonal swelling was also very
rare in the brain. No neuropathological effects were found rostral to the pons. Small numbers of
swollen axons were located in the ventromedial medulla. Rare single swollen axons were
located in the ventral spino-cerebellar tracts, cerebellar peduncles, and deep cerebellar white
matter.
Neurogenic skeletal muscle atrophy occurred in both proximal and distal hind-limb
musculature. Myofibrillar atrophy was multifocal, with foci overlapping in severe cases to
produce large diffuse areas of atrophy with fatty replacement. Intramuscular nerves frequently
showed an obvious loss of axons and rarely a swollen axon. No difference in the severity or
frequency of atrophic foci was seen between proximal and distal muscles.
33
-------
In the 0.5% group, peripheral nerve changes were identical in morphology and in the
number of animals affected compared with the higher-dose animals but were reduced in severity.
Swollen axons were generally few in number but were found in all animals. Myelin ovoids and
frankly degenerating axons were also reduced in number. In some nerve bundles, there was
obvious loss of axons. Spinal effects were reduced to a few swollen axons and rare degenerating
axons in the ventromedial fasciculi of the thoracolumbar cord. Effects on the brain stem and
cerebellum were minimal, consisting of only single or small numbers of swollen axons and
single degenerating axons in half of the animals examined. Neurogenic skeletal muscle atrophy
consisted of infrequent multifocal areas of myofibrillar atrophy that were generally regarded as
minor. Two animals without myofibrillar atrophy were considered normal. Three samples from
the quadriceps and two from the calf muscles, while not demonstrating myofibrillar atrophy, did
have early myopathic effects consisting of foci of increased numbers of angular myofibers and
increased numbers of myofibers with central or internal nuclei. In one of these animals,
intramuscular axonal swelling was found.
At the 0.25% level, peripheral nerve changes were less severe than at higher doses, and
axonal swelling was found in 8 of 10 animals examined. In these eight rats, the number of
swollen axons was very low, but additional changes, such as myelin infoldings into axons,
myelin ovoids, and degenerating axons, were more common. In one animal, while no axonal
swelling was observed, numerous degenerating axons were found. Another rat was
indistinguishable from controls. Spinal lesions were minimal, consisting of a single or very few
swollen axons. A few instances of axonal swelling were found in the medullae of two rats.
Neurogenic myofibrillar atrophy was also minimal, occurring as a single or very few foci in two
animals. Foci of angular myofibers were found in four additional animals but were of minimal
severity. In control animals, the peripheral and central nervous system (CNS) contained a few
degenerating axons and myelin ovoids, but these were minimal. A summary of animals found to
have axonal swelling and the areas in which these axons or myopathic changes were found is
presented in Table 4-8.
Table 4-8. Summary of neuropathological findings in male rats administered
2-hexanone in drinking water for 13 months
Treatment
Control
0.25%
2-Hexanone
0.5% 2-Hexanone
1.0% 2-Hexanone
Axonal swelling
Myofibrillar atrophy
Incidence per number of animals exposed
Brain
0/10
2/10
4/10
8/10
Spinal cord
0/5
7/10
5/5
5/5
Dorsal root
ganglia
0/5
0/7
0/5
3/5
Peripheral
nerve
0/10
8/10
10/10
10/10
Quadriceps
muscle
0/10
1/10
5/10
10/10
Calf muscle
0/10
2/10
6/10
10/10
Source: O'Donoghue etal. (1978).
34
-------
4.4.1.2. Inhalation Exposures
6-Week study: rats
In a short communication, Duckett et al. (1974) reported the results of a study in which
groups of nine rats (strain and sex not reported) were exposed to 200 ppm (819 mg/m3)
2-hexanone (purity unspecified) 8 hours/day, 5 days/week for 6 weeks. Four rats served as
controls. Animals presented with muscular weakness of all limbs that persisted for a few hours
after exposure termination each day. Only the sciatic nerve was examined histologically.
Axonal hypertrophy, beading, and degeneration associated with perinodal and segmental
breakdown of myelin were observed in the sciatic nerve of all treated rats.
13-Week study: rats
Duckett et al. (1979) exposed groups of 20 Wistar rats of unspecified sex to 2-hexanone
for 8 hours/day, 5 days/week at 40 ppm (164 mg/m3) for 22-88 days or at 50 ppm (205 mg/m3)
for 13 weeks. Similar numbers of control rats were sham exposed. No overt signs or
"pathological manifestations" of peripheral or central neuropathy were seen in exposed rats,
except for demyelination of the sciatic nerve in 3 of the 20 rats exposed to 50 ppm for 13 weeks.
Additional details were not provided. The results at 50 ppm for 13 weeks, when compared with
the results at 50 ppm for 6 months (described later in this section), indicate that the incidence of
neuropathy increases with increasing duration of exposure.
12-Week study: cats, rats, chickens
Mendell et al. (1974) continuously exposed groups of animals of unspecified sex (four
Sprague-Dawley rats, four domestic shorthair cats, and five domestic chickens) to 2-hexanone
(purity not stated) for 24 hours/day, 7 days/week for up to 12 weeks. Concentrations of
2-hexanone were initially 200 ppm (820 mg/m3) for chickens and 600 ppm (2,460 mg/m3) for
cats and rats but were adjusted at an unspecified time to 100 and 400 ppm (410 and
1,640 mg/m3), respectively, to minimize complications from inanition and weight loss. Pair-fed
controls were sacrificed when the exposed animals were sacrificed. After 5-8 weeks of
exposure, the cats developed hind-limb and forelimb weakness. Focal swelling of the axon along
the sciatic nerve, often associated with loss of neurotubules and denudation of myelin beginning
at the nodes of Ranvier, and areas of demyelination were observed. Abnormal electromyograms
were also observed in the cats exposed for 9-10 weeks; electromyograms were not measured in
chickens or rats (Mendell et al., 1974).
90-Day study: hens
Abdo et al. (1982) exposed adult leghorn laying hens (G. gallus domesticus), five per
group, to varying concentrations of 2-hexanone (10, 50, 100, 200, and 400 ppm; technical grade
35
-------
containing 70% 2-hexanone and 30% MiBK) for 90 days. Body weights were monitored
weekly, and hens were examined every other day for neurological signs of 2-hexanone
neurotoxicity. A 30-day observation period followed the final exposure. Clinical assessment of
neurotoxicity was graded by classifying the degree of ataxia before paralysis as follows: TI,
mild ataxia, characterized by diminished leg movement and reluctance to walk, with hens
tending to slide on the floor or fly; T2, gross ataxia, characterized by a change in gait and
disturbance of leg movement; T3, severe ataxia, with severe leg weakness manifested by
unsteadiness and occasional falling on the floor; T4, ataxia, with near paralysis, marked by
inability to walk (Abdo et al., 1982).
The spinal cord and the sciatic, peroneal, and tibial nerves were excised from hens that
died during the experiment or were killed by heart puncture and exsanguinations. Severity of
lesions was defined by the following criteria: (1) rare swollen axons without fragmentation,
phagocytosis, or loss of myelin staining were designated as equivocal histological changes; (2)
occasional degenerative changes of axons and myelin in peripheral nerve or within the spinal
cord, which may contain nests of phagocytic cells, were termed mild to moderate degeneration;
and (3) lesions were considered severe when there was almost complete destruction of axons
and myelin in a given tract, such as the anterior columns or within extensive areas of peripheral
nerve.
Only hens exposed to one of the highest two concentrations of 2-hexanone, 400 or
200 ppm, lost significant weight at the onset of ataxia; weight loss for these two groups
continued, and the hens exposed to 400 and 200 ppm 2-hexanone weighed 48.0 ± 7.4% and
63.1 ± 5.5% (mean ± SEM) of the initial weights, respectively, at the onset of paralysis.
Although the group exposed to 100 ppm 2-hexanone gained some weight at the onset of ataxia,
they lost 24.4% of their initial weight after 69 days of exposure. This weight loss coincided with
the development of severe ataxia. This treatment group, however, regained all lost weight by the
end of the 30-day observation period. No appreciable change in weight was observed in hens
exposed to 50 or 10 ppm 2-hexanone.
None of the hens continually exposed to 2-hexanone vapor showed any signs of acute
toxicity that could be attributed to the narcotic effects of 2-hexanone on the CNS. All hens
continually exposed to 50-400 ppm 2-hexanone developed ataxia after a latent period of 6-
30 days, depending on 2-hexanone concentrations. Those exposed to 400 ppm progressed to
paralysis, and two died 27 days after the beginning of exposure. The remaining three chickens
were in a distressed condition and were sacrificed at 31 days. The number of days of exposure to
2-hexanone vapor before the onset of ataxia was dependent on and inversely proportional to the
concentration of 2-hexanone.
All hens exposed to 200 ppm 2-hexanone developed paralysis 64-72 days after the
beginning of the exposure; one of these hens died at day 72, and the other four were sacrificed on
day 73. Four of the hens inhaling 100 ppm 2-hexanone developed severe ataxia (T3), while the
36
-------
fifth bird progressed to ataxia with near paralysis (T/t). Three hens of the group exposed to
50 ppm 2-hexanone showed severe ataxia (T3), while the other two developed only gross ataxia
(T2). The clinical condition of all hens in this group was gross ataxia (T2) at termination. All
hens exposed to 10 ppm 2-hexanone remained normal.
Histopathological lesions in the spinal cord were dependent on concentration, duration
of exposure, and duration of intoxication. Two of the hens exposed to 400 ppm did not exhibit
any histopathological alterations, while another two showed equivocal changes. Hens exposed to
100 ppm 2-hexanone exhibited clinical signs of neurotoxicity for 99 ± 2 days, and all hens
showed unequivocal changes in the spinal cord. Although hens exposed to 50 ppm 2-hexanone
were exposed for a mean of 97 days, only four of these hens had unequivocal changes in the
spinal cord. Similarly, the presence of histopathological lesions in peripheral nerves was a
function of both the level of 2-hexanone inhaled and, particularly, the total dose inhaled.
Although all five hens exposed to 100 ppm for 90 days survived until termination on day 120,
they showed gross to severe ataxia and each had unequivocal lesions in peripheral nerves. Hens
given high doses became paralyzed and thus could not be kept alive as long as those given
100 ppm 2-hexanone.
4-Month study: rats
Groups of six young adult rats (strain and sex not specified) were exposed to 1,300 ppm
(5,325 mg/m3) of 2-hexanone 6 hours/day, 5 days/week for up to 4 months (Spencer et al., 1975).
Three rats were exposed to air only. Animals were observed for neurological signs, and
histopathological examinations of several peripheral nerves, regions of the spinal cord, medulla,
and cerebellum were completed. In the exposed rats, narcosis, loss of coordination, weight loss
(data not presented), foot drop, and proximal hind-limb and forelimb weakness were observed.
Pathological alterations included nerve fiber degeneration in the peripheral nerves, spinal cord,
medulla, and cerebellum; axonal dilatation with localized fiber swelling; and secondary
paranodal myelin retraction.
6-Month study: male rats
Duckett et al. (1979) exposed groups of Wistar rats (sex not specified) to 0 ppm (n = 20)
or 50 ppm (n = 40) 2-hexanone (0 or 205 mg/m3) 8 hours/day, 5 days/week for 6 months. No
overt signs of toxicity were observed during the study. Electrophysiological evaluations were
performed on 5 treated and 10 control rats at the end of the experiment. The mean sciatic motor
conduction velocity (MCV) in the exposed group was significantly lower (p = 0.005) than in the
controls. No effect on the amplitude of the evoked muscle action potential (MAP) was observed.
Widespread demyelination of the sciatic nerve was reported in 32 rats from the exposed group;
two of the rats also had axonal hypertrophy and beading. No abnormalities were seen in the
37
-------
sciatic nerves of control rats. The study authors (Duckett et al., 1979) reported that the
histopathology of the CNS, liver, and kidney of all rats was normal (details were not provided).
72-Week study: male rats
Krasavage and O'Donoghue (1977) exposed male Sprague-Dawley rats (18/group) to 0,
100, or 330 ppm (0, 410, or 1,353 mg/m3) 2-hexanone (purity not specified) 6 hours/day,
5 days/week for 72 weeks (for further experimental detail, see Section 4.2.2.3). Exposure to
100 ppm did not cause clinical or pathological evidence of neurological damage. One rat
exposed to the high concentration developed progressive hind-limb weakness; another three
high-concentration animals showed slight weakness that was not progressive. One animal in the
high-concentration group developed a severe polyradiculoneuritis of the nerve roots in the
lumbar and sacral spinal nerves and in the sciatic and tibial nerves. The authors concluded that
chronic exposure to 100 ppm 2-hexanone was not neurotoxic, while findings at 330 ppm were
equivocal (Krasavage and O'Donoghue, 1977).
6-Month study: male rats
Male Sprague-Dawley rats (six/group) were exposed to 0 or 100 ppm (0 or 410 mg/m3)
2-hexanone (purity 96.66%, 2.9% MiBK) 22 hours/day, 7 days/week for 6 months (Egan et al.,
1980). Two animals from each group underwent microscopic examination for neuropathological
changes following 2, 4, and 6 months of exposure. No treated or control animals displayed
clinical signs of neurotoxicity during the exposure period. After four months of exposure, a
typical pattern of 2-hexanone-induced neuropathology began to appear in the CNS and
peripheral nervous system (PNS). At this time, PNS specimens revealed giant axonal swellings
and secondary demyelination in a few large diameter fibers in the tibial nerve branches to the
calf muscles. In the CNS, isolated giant axonal swellings were found in the medulla oblongata
and cerebellum. By 6 months, more advanced degeneration was presented in teased fibers in calf
muscle branches and giant axonal swelling had ascended to the level of the sciatic notch. The
spinal cord revealed scattered fiber degeneration in the ventral portion of the gracile tract and the
caudal portion of descending fiber tracts in the lumbar region.
10-Month study: male rats, male monkeys
Johnson et al. (1977) exposed male Sprague-Dawley rats (10/per group) and male
monkeys (Macacafascicularis) (8/group) to 0, 100, or 1,000 ppm (0, 410, or 4,100 mg/m3)
commercial grade 2-hexanone (purity not stated) for 6 hours per day, 5 days per week for up to
10 months. Rats in the 1,000 ppm exposure group exhibited progressive body weight loss
beginning at 16 weeks and reaching statistical significance at 20 weeks (p < 0.01). Monkeys in
the 1,000 ppm group progressively lost body weight beginning at 8 weeks. No significant effect
of 2-hexanone on body weight of rats or monkeys was found in the low-dose exposure groups.
38
-------
Four neurological tests were conducted on both rats and monkeys: MCV of right sciatic-
tibial nerves, MCV of the right ulnar nerve, absolute refractory period of these nerves, and MAP
recorded in response to both sciatic and ulnar stimulation. In addition, electroencephalograms
(EEGs) and visual evoked potentials were recorded from monkeys. All animals were
administered an anesthetic prior to neurological testing: rats received an i.p. injection of 35
mg/kg of sodium pentobarbital, and monkeys were given 15 mg/kg of ketamine hydrochloride
intramuscularly.
After 25 weeks, all rats and monkeys in the high-dose exposure group were removed
from further exposure because neuropathy (hind-limb drag) apparently had developed. All eight
monkeys in the 100 ppm group were exposed for a total of 41 weeks. Rats in the low-dose group
were removed from 2-hexanone exposures after 29 weeks. Beginning at 3 months of exposure,
monkeys in the 1,000 ppm group showed a progressive decrease in the MCV of the sciatic-tibial
nerves. After 6 months, the mean MCV of this group was 63% of the mean of control animals.
Commencing at 9 months, the MCV for the sciatic-tibial nerves in monkeys in the 100 ppm
group was significantly different from control values (p = 0.05). At the termination of the study,
the MCV of monkeys from the 100 ppm group was 12% less than that in the corresponding
controls (p < 0.05).
A similar pattern of sciatic-tibial neuropathy developed in rats exposed to the higher
concentration of 2-hexanone. A significant decrease in MCV was observed at approximately
3 months (13 weeks) of exposure (p = 0.05). A significant difference at 8 weeks between MCVs
of control and 1,000 ppm rats was considered spurious. In the 100 ppm group, a significant
difference in MCVs between controls and treated rats occurred at 29 weeks (p < 0.001).
A neuropathy similar to that observed for the sciatic-tibial nerves was noted in the ulnar
nerve of both the monkeys and rats. When compared with controls, commencing at 4 months,
monkeys showed a progressive decrease in the MCV of the ulnar nerve. At the end of 6 months'
exposure to 1,000 ppm, monkeys showed a significant decrease in ulnar MCV with values
approximately 64% of those of controls (p < 0.01). Ulnar MCVs in the 100 ppm group showed
a similar decreasing trend at about 6 months; however, these values were not statistically
different from controls. In rats, ulnar MCVs were significantly decreased compared with control
values (p < 0.05), beginning at about 17 weeks in both exposed groups.
Both monkeys and rats exposed to 1,000 ppm 2-hexanone showed a continuous decrease
in MAP amplitude in response to sciatic stimulation that became statistically significant in
monkeys at 6 months (p < 0.01). This effect was not noted in the low-dose group of monkeys.
Rats in the 100 ppm group had reduced MAP amplitudes for sciatic stimulation, beginning at
12 weeks. No effects of 2-hexanone on scalp-recorded EEGs of monkeys were observed.
Amplitude measures of the EEG were not affected at either exposure concentration. Visual
examination of the EEG records did not reveal any abnormal patterns (e.g., spikes or abnormal
waves).
39
-------
Evidence of 2-hexanone-induced effects on average visual evoked potential (AVEP) was
obtained in monkeys exposed to 1,000 ppm. Specifically, latencies of certain AVEP components
were increased beginning at 4 months. No effects on these latencies occurred as a result of the
low-dose 2-hexanone exposure. The refractory time (i.e., the time that must elapse between two
consecutive stimuli of a nerve in order for the second stimulus to also excite the nerve) was not
affected by 2-hexanone at either level of exposure.
Only rats were examined for effects of 2-hexanone on operant behavior at 10 and
19 weeks of exposure to 100 and 1,000 ppm, respectively. For operant behavior, animals were
trained on a multiple fixed ratio of 5, fixed interval 3-minute (multi-FR5FI3) schedule for 20-
40 days after shaping the bar press response. Once behavior was stable, animals were placed in
exposure chambers and tested after exposure. A reduction in response rate in the 1,000 ppm
group developed by the second week of exposure; however, no effects of 2-hexanone on operant
behavior were found with the 100 ppm group (Johnson et al., 1977).
2-Year study: cats
Groups of four domestic shorthair cats were exposed by inhalation to 0, 100, or 330 ppm
(0, 410, or 1,353 mg/m3) 2-hexanone (purity not specified) for 6 hours/day, 5 days/week for
2 years (O'Donoghue and Krasavage, 1979). Clinical signs and body weights were monitored
(for details, see Section 4.2.2.3). To follow the onset of neuropathy, biopsy specimens were
collected from two randomly selected cats in each group at six intervals for the first 9 months of
the exposure period. All specimens were taken from alternate hind paws and included 5-6
Pacinian corpuscles and plantar interosseous muscles. Cats were sacrificed at the end of the
treatment and underwent gross and histopathological examinations, and the nervous system was
examined microscopically in detail.
No clinical neurological effects attributed to exposure to 2-hexanone were identified.
Neuropathological examination results for the control and low-dose groups were comparable.
All cats in the high-dose group showed evidence of neuropathological changes in the CNS and
the PNS at and below the level of the cerebellum and pons (O'Donoghue and Krasavage, 1979).
In the PNS, the highest incidences of change occurred in the tibial motor nerve branches to the
musculature of the lower leg and then in the tibial nerve itself. In the branches, endoneural space
was enlarged with clear fluid. Swelling of giant axons with myelin retraction was evident, and
degenerating axons were found infrequently. No changes were found in the dorsal root ganglion
cells. In the distal portion of the PNS in the high-dose animals, unusually large preterminal
axonal processes were evident, a condition not seen in controls. Examinations of tibial nerve
fibers indicated comparable percentages of the four fiber pathology categories (i.e.,
demyelination, remyelination, swelling, and degenerative fibers) in the control and low-dose
groups, but the high-dose group had notable changes in each fiber pathology category except
degenerative fibers. Demyelination, remyelination, swelling, and degeneration occurred in 12.3,
40
-------
3.4, 6.3, and 0.4% of high-dose axons examined, respectively (average number of high-dose
axons examined = 158), compared with 0, 0.3, 0, and 0.6% of control axons (average number of
control axons examined = 84). In the CNS, swollen terminals were found in the posterior
cerebellar peduncles, folial white matter, nucleus gracilis, fasciculis gracilis, spino-cerebellar
tracts, medullary reticular formation, and all levels of the spinal cord.
4.4.1.3. Other Routes of Exposure
11-Month study: dogs (subcutaneous injections)
O'Donoghue and Krasavage (1981) administered 2-hexanone (>97% pure, with 2.9%
MiBK and trace quantities of 2-hexanol) by daily subcutaneous injection to purebred male
beagles (n = 4) for 11 months. At first, each dog received 300 mg/kg of the test compound or
saline once daily and later (time not stated) divided into two equal doses 6 hours apart. All
animals developed signs of neurotoxicity to varying degrees. The patellar reflex was lost
unilaterally in two of the four dogs receiving 133 grams of 2-hexanone over a period of 96 days.
One month later, the patellar reflex was lost bilaterally in both dogs, and clinical signs of
neurotoxicity progressed with observations of muscle weakness and difficulty walking. The
condition of both dogs gradually reversed during the course of the study, following an
unspecified cessation of exposure. In the remaining two dogs, the clinical signs of neurotoxicity
appeared later in the study or were apparent at study completion. In one dog, the patellar reflex
could not be elicited after it had received 243 grams of 2-hexanone over a period of 156 days.
Following cessation of exposure, the dog returned to apparent normality in approximately
56 days. In the remaining dog, no clear neuropathic abnormality was produced, but, although the
patellar reflex was present, the response appeared sluggish. There was occasional evidence of
hind-limb weakness.
Mean body weights of treated animals were comparable with those of controls, but
individual animals showed weight loss or decreased weight gain. Hematology, clinical
chemistry, and cerebrospinal fluid analysis were not affected by the treatment. Repeated biopsy
examinations of distal peripheral nerves showed typical giant axonal swelling. The biopsy
findings paralleled the clinical course except during a recovery phase, where the biopsy
continued to be abnormal while the clinical course improved. Electromyographic examination of
the treated dogs showed the persistence of abnormalities in two recovering dogs, no
abnormalities in one recovering dog, and no abnormalities in the one dog that had appeared
clinically normal throughout the study (O'Donoghue and Krasavage, 1981).
90-Day study: hens (i.p. injections)
Abou-Donia et al. (1982) treated five groups of leghorn laying hens (G. gallus
domesticus, n = 3) with daily i.p. injections of 2-hexanone (70% 2-hexanone, 30% MiBK) at 100
or 200 mg/kg for 90 days. Hens given 100 mg/kg 2-hexanone daily progressed through all
41
-------
successive stages of ataxia; the clinical conditions of two of them improved after treatment was
stopped, while the third hen progressed to paralysis and died after 30 days of administration.
Daily i.p. injection of 200 mg/kg 2-hexanone produced ataxia with near paralysis (T4), which
progressed to paralysis in one hen. The clinical condition of this hen, however, reverted to grade
T4 after cessation of administration.
Spinal cords from hens given daily 100 mg/kg i.p. injections of 2-hexanone did not
exhibit any histopathological changes. One of these hens, however, showed unequivocal
histopathological changes in the peripheral nerves. The sites of axonal degeneration were
accompanied by myelin degeneration, and macrophages were observed containing debris with
the staining properties of myelin. Although none of the hens given 200 mg/kg i.p. injections of
2-hexanone showed histopathological alterations in peripheral nerves, two of these hens
developed unequivocal histopathological lesions in the spinal cord. A longitudinal section from
the ventral column of the thoracic spinal cord from one of the hens showed axons with prominent
swellings. These swellings have the morphologic configuration of the paranodal swelling that
suddenly ends at the nodes of Ranvier. A longitudinal section of the thoracic spinal cord from
the other affected hen demonstrated extensive degeneration in the ventral column and a markedly
swollen axon and nests of macrophages.
4.4.2. Immunotoxicity Studies
No studies were located regarding immunological effects in humans by any route of
exposure to 2-hexanone.
A reduction in total white blood cell counts to 60% of control values (p < 0.05), but no
changes in differential white cell counts or evidence of bone marrow damage, was found in rats
intermittently exposed to 700 ppm 2-hexanone after 8 weeks, during an 11-week study (Katz et
al., 1980). These findings, although inconclusive, suggest that immunological effects may
warrant some consideration in future assessments of the potential toxicity of exposure to
2-hexanone.
4.5. OTHER STUDIES
4.5.1. Mechanistic Studies
4.5.1.1. 2-Hexanone and Enzyme Induction
2-Hexanone and its neurotoxic metabolite 2,5-hexanedione are both effective inducers of
microsomal enzyme activities. This can affect the toxicity of other xenobiotics and also can
affect the toxicity of 2-hexanone itself (or its precursor, n-hexane) by increasing or decreasing
the formation of toxic metabolites.
Nakajima et al. (1991) characterized the CYP450 enzymes in the livers of male Wistar
rats that are induced following exposure to 2-hexanone (5 mmol/kg-day), 2,5-hexanedione
(5 mmol/kg-day), or phenobarbital (80 mg/kg-day), administered intraperitoneally for 4 days. A
42
-------
control group received an equivalent volume of corn oil vehicle (4 mL/kg). All three treatments
caused a statistically significant increase in microsomal protein content and overall CYP450
activity (Table 4-9).
Table 4-9. Effects of 2-hexanone, 2,5-hexanedione, and phenobarbital on
microsomal protein and CYP450
Treatment
Control
2-Hexanone
2,5-Hexanedione
Phenobarbital
Body weight
(g)
206 ±7
192 ±6
184 ±T
197 ±5
Liver weight
(g)
6.6 ±0.2
7.3 ±0.3a
6.4 ±0.3
7.9 ±0.4a
Liver/body
weight ratio
(%)
3.21 ±0.11
3.80 ±0.05a
3.49 ±0.07a
4.01 ±0.13a
Microsomal
protein
(mg/g liver)
21.5 ±0.8
25.1 ± 1.5a
26.2 ± 1.7a
31.5 ± 3.0a
CYP450
(nmol/mg protein)
0.92 ± 0.002
1.49 ± 0.10a
1.62 ± 0.10a
2.12 ± 0.19a
"Significantly different (p < 0.05) from control.
Source: Nakajimaetal. (1991).
The enzyme activities (i.e., benzene aromatic hydroxylase [CYP2E1], TSO
[CYP2C6/11], EROD [CYP1 Al/2], and PROD [CYP2B1/2]) were measured as indicators of
CYP450 activity. All three treatments caused a statistically significant increase in the rate of
benzene hydroxylation at low (0.2 mM) and high (6.3 mM) concentrations and TSO at low (0.2
mM) and high (5.0 mM) concentrations. EROD activity was not affected by pretreatment;
however, a statistically significant increase in PROD activity was observed with all three
treatments. A summary of the results for the CYP450 activity measured with specific substrates
is listed in Table 4-10.
Table 4-10. Effect of enzyme inducers on the activities of CYP450-related
enzymes in rats exposed to 2-hexanone or 2,5-hexanedione
Treatment
Control
2-Hexanone
2,5-Hexanedione
Phenobarbital
Enzyme activity
BAHa
0.2 mM
0.68 ±0.09
1.10±0.19b
0.98±0.16b
0.48±0.11b
6.3 mM
0.53±0.11
1.76±0.23b'c
1.57±0.15b'c
2.80±0.23b'c
TSO
0.2 mM
1.87 ±0.15
5.65±0.62b
5.05±0.46b
5.59±0.87b
5.0 mM
8.34 ±0.67
19.07 ± 1.64b'd
19.98 ± 0.78b'd
25.36 ± 6.23b'd
EROD
0.32 ± 0.06
0.41 ± 0.30
0.26 ± 0.44
0.27 ± 0.04
PROD
0.11 ±0.02
3.68±0.70b
2.92 ± 0.90b
5.22 ± 0.70b
aBAH = benzene aromatic hydroxylase.
bSignificantly different (p < 0.05) from control.
Significant difference (p < 0.05) between 0.2 and 6.3 mM of the corresponding group.
dSignificant difference (p < 0.05) between 0.2 and 5.0 mM of the corresponding group.
Source: Nakajimaetal. (1991).
43
-------
Using immunoblotting and immunodetection assays, Nakajima et al. (1991) did not
detect CYP1A1/2 in microsomes from treated and control animals. CYP2B1/2 was induced by
treatment with phenobarbital > 2-hexanone = 2,5-hexanedione. Only trace amounts of CYP2E1
were detected in phenobarbital-treated rats, whereas 2-hexanone and 2,5-hexanedione both
induced this isoform efficiently.
In order to explore the effects of 2-hexanone, 2,5-hexanedione, and phenobarbital on
CYP2B1/2, CYP2E1, and CYP2C6/11, Nakajima et al. (1991) performed immunoinhibition
analyses of ISO activity by using monoclonal antibodies directed against each of these CYP450
isoforms. Anti-CYP2El inhibited TSO activity in induced microsomes as follows (values are
percent of activity in the absence of anti-CYP2El): phenobarbital, 97 ± 2%; 2,5-hexanedione, 79
± 3%; 2-hexanone, 75 ± 11%; and controls, 65 ± 2%. Anti-CYP2Bl/2 inhibited TSO activity in
induced microsomes differently: phenobarbital, 31 ± 4%; 2-hexanone, 65 ± 3%;
2,5-hexanedione, 69 ± 5%; and controls, 99 ± 2%. Anti-CYP2C6/l 1 inhibited toluene
metabolism in induced microsomes as follows: phenobarbital, 75 ± 5%; 2-hexanone, 69 ± 5%;
2,5-hexanedione, 70 ± 3%; and controls, 23 ± 4%.
Similar studies were performed by Imaoka and Funae (1991). The authors treated male
Sprague-Dawley rats (number of rats not provided) with 2-hexanone (purity not stated;
5 mmol/kg, i.p.; dissolved in corn oil) daily for 4 days. This dose was considered a maximum
tolerated dose. Control rats were given corn oil only. Hepatic microsomes were isolated, and
the activities of CYP450 enzymes were determined against specific substrates (Table 4-11).
Table 4-11. Catalytic activities of CYP450 enzyme activities in rat liver
following induction by 2-hexanone
Substrate
Aminopyrine
Aniline
7-Ethoxycoumarin
Testosterone-2a
Testosterone-2p
Testosterone-6p
Testosterone-7a
Testosterone- 1 5a
Testosterone-16a
Testosterone-16p
Enzyme activity (nmol/min-mg protein)3
Uninduced control
2.40 ±0.50
0.283 ±0.044
3.62 ±0.13
0.684 ±0.114
0.140 ±0.039
0.959 ±0.176
0.056 ±0.006
0.040 ±0.007
1.09 ±0.203
0.058 ±0.006
2-hexanone-treated
4.37±0.82b
0.421±0.070b
6.01±1.24b
0.431 ±0.158b
0.240 ±0.056b
1.45±0.341b
0.062 ±0.013
0.056 ±0.017
1.07 ±0.347
0.250 ±0.106b
"Mean ± standard deviation, number of rats not provided.
bSignificantly different from control, p < 0.05.
Source: Imaoka and Funae (1991).
44
-------
The content of total CYP450 measured photometrically did not change much with
treatment. However, the activities of aminopyrine N-demethylase, aniline hydroxylase, and
7-ethoxycoumarin O-dealkylase were increased by pretreatment with 2-hexanone. Testosterone
2|3-, 6|3-, and 16|3-hydroxylase activities were significantly increased, whereas the
2a-hydroxylase activity was decreased by treatment with 2-hexanone. Imaoka and Funae (1991)
also measured changes in the levels of 11 forms of CYP450 in hepatic microsomes caused by
treatment with 2-hexanone (Table 4-12).
Table 4-12. Changes in CYP450 levels following treatment with 2-hexanone
CYP450 isoform
2A1
2A2
2B1
2B2
2C6
2C7
2C11
2C13
2E1
4A3
CYP450 content (pmol/mg protein)3
Uninduced control
7.0 ±1.3
10.4 ±2.3
<0.5
3. 8 ±1.2
52.1 ±17.7
21.9±3.3
457.0 ±52.6
171.4 ±35.8
49.8 ±9.6
17.6 ±3.2
2-Hexanone-treated
7.9 ±1.5
11.7 ±2.8
44.3±9.4c
29.3±6.2c
93.4±16.9b
24.8 ±5.8
343.8±46.3C
159.7 ±24.5
102.6 ± 14.8b
16.7 ±2.8
"Mean ± standard deviation; number of rats not provided.
bSignificantly different from control, p < 0.01.
Significantly different from control,/) < 0.05.
Source: Imaoka and Funae (1991).
The level of CYP2C11, a male-specific form, was decreased by treatment with
2-hexanone in parallel with a decrease in testosterone 2a-hydroxylase activity, which is
catalyzed by this isozyme (Kamataki et al., 1983) (Table 4-12). CYP2A2 is a constitutive
testosterone 6|3-hydroxylase; the increase in the level of this isoform explained the increase in
testosterone 6|3-hydroxylase activity, shown in Table 4-11. CYP2B1 and 2B2 are typical
phenobarbital-inducible forms. The level of CYP2B1 in the hepatic microsomes of control rats
was very low, and CYP2B2 was detected at a slightly higher level. Both forms were strongly
induced in 2-hexanone-treated rats. These results reflected the increases in testosterone
16|3-hydroxylase and aminopyrine N-demethylase activities of hepatic microsomes (cf
Table 4-11) and suggest that 2-hexanone is a phenobarbital-type inducer.
Imaoka and Funae (1991) determined that the inducibility of CYP2B1 and 2B2 was
strongly correlated with the hydrophobicity (as estimated by the octanol/water partition
coefficients, log Kow) of several 2-hexanone homologues: 2-hexanone (1.38) > methyl n-propyl
ketone (0.91) > MEK (0.29) > acetone (-0.24). In contrast, the inducibility of CYP2E1 was not
45
-------
dependent on hydrophobicity. Each of the aforementioned chemicals, at equimolar
concentrations, induced CYP2E1 to a similar extent, approximately twofold, while acetone, a
prototypical inducer of CYP2E1, induced this isoform approximately threefold.
Based on studies of 2-hexanone and the pesticide O-ethyl O-4-nitrophenyl
phenylphosphonothioate (EPN) in hens, Abou-Donia et al. (1991, 1985b) speculated that the
potentiation of the neurotoxic effects of 2-hexanone by EPN may be due to induction of hepatic
microsomal CYP450 by EPN with increased production of 2,5-hexanedione. Similarly, MEK
may also potentiate the toxicity of 2-hexanone through induction of CYP450 as MEK but not
2-hexanone and has been shown to decrease hexobarbital sleep time in rats (Couri et al., 1977).
While MEK has been shown to potentiate the toxicity of 2-hexanone in rats (Saida et al., 1976),
Shibata et al. (2002) have demonstrated that MEK depresses the metabolism of n-hexane in
human volunteer subjects. If the metabolic pathways of 2-hexanone, as detailed in Section 3.3
and Figure 3-1, are common in humans and animals and MEK depresses the metabolism of
n-hexane but increases the metabolism of 2-hexanone, then the step in 2-hexanone metabolism
that MEK likely affects is the co-1-oxidation to 5-hydroxy-2-hexanone. While no specific
CYP450 isoenzymes have been implicated and the mechanisms are not fully elucidated, it
appears that 2-hexanone has the ability to influence its own metabolism via effects on CYP450
enzymes that need more research to be fully understood.
It should be noted that, like 2-hexanone, MiBK (a common contaminant in the
formulation of the 2-hexanone) has the potential to act as a CYP450 inducer. However, the 3.2%
concentration of MiBK in 96% pure formulations of 2-hexanone, as reported by O'Donoghue et
al. (1978), may not have a significant impact on the toxicity of 2-hexanone. To determine
whether the concentration of MiBK as a contaminant may have altered the observed toxicity of
2-hexanone, other studies were evaluated that used MiBK as a test article. In a 13-week gavage
study, 30 male and female Sprague-Dawley rats were treated with 0, 50, 250, or 1,000 mg/kg-
day MiBK (MAI, 1986). At the middle and high doses, adverse effects were observed in the
liver and kidney, which progressed in severity in the high-dose animals. No treatment-related
effects of any kind were observed at 50 mg/kg-day. The Carnegie-Mellon Institute of Research
(1977) conducted a 120-day drinking water study with 1.3% MiBK, using female HLA Wistar
rats. The authors estimated the dosage to be 1,040 mg/kg-day. The only statistically significant
finding was increased mean absolute and relative kidney weights in treated rats compared with
controls. Histopathological examination revealed renal tubular cell hyperplasia in only one of
five of the treated rats. No exposure-related histopathological changes were found in other
organs. Based on the foregoing, it can be concluded that the dosage of MiBK received as an
impurity in the study by O'Donoghue et al. (1978) did not contribute to the observed
2-hexanone-related effects. O'Donoghue et al. (1978) did not observe adverse effects in the
kidney or liver of treated animals, despite these organs being the target organs of toxicity in
experimental studies with MiBK from both the oral and inhalation routes (U.S. EPA, 2003a).
46
-------
4.5.1.2. 2-Hexanone as a Sulfhydryl Reagent
Both 2-hexanone and its metabolite 2,5-hexanedione can inhibit sulfhydryl-containing
enzymes such as fructose-6-phosphate kinase and glyceraldehyde-3-phosphate dehydrogenase
(enzymes in the pentose phosphate pathway [oxidative phase] and glycolytic pathway
[nonoxidative phase], respectively) (Sabri, 1984; Sabri et al., 1979). Both of these chemicals
inhibited fructose-6-phosphate kinase from rabbit muscle or rat brain homogenates; in each case,
2,5-hexanedione was the far more potent inhibitor (Sabri et al., 1979). Preincubation with
dithiothreitol protected this enzyme from inhibition, which suggests that these compounds
interfere with the sulfhydryl groups required for fructose-6-phosphate kinase activity. However,
dithiothreitol could not restore enzyme activity after these compounds had been added. In
addition, fructose-6-phosphate kinase activity was also reduced in brain homogenates of rats that
had received 2,5-hexanedione at 0.5% in their drinking water for 10-12 weeks (Sabri et al.,
1979). Glyceraldehyde-3-phosphate dehydrogenase from rabbit muscle (purified to crystalline
state) was also inhibited in vitro by both compounds; in this case, 2-hexanone was the more
potent inhibitor (Sabri, 1984). Levels of adenosine triphosphate were reduced in cat sciatic
nerves treated with 2,5-hexanedione (Sabri, 1984), possibly an outcome of glyceraldehyde-3-
phosphate dehydrogenase inhibition. 2-Hexanone was found to irreversibly inhibit rat brain and
rabbit muscle creatine kinase and mouse brain adenylate kinase (Lapin et al., 1982).
4.5.1.3. Studies Exploring the Development of Neuropathy
Groups of 12 Sprague-Dawley rats (sex unspecified) were continuously exposed
(24 hours/day) via inhalation to 0, 225, or 400 ppm (0, 922.5, or 1,640 mg/m3) 2-hexanone
(purity not stated) for 16-66 days (Saida et al., 1976). Rats exposed to 400 ppm were sacrificed
at 16, 28, and 42 days, and those exposed to 225 ppm were sacrificed at 16, 25, 35, 55, and
66 days to study the sequence of morphologic changes. Paralysis was observed after 66 and
42 days at the low and high concentrations, respectively. Neuropathological changes preceded
paralysis and were observed at the initial sacrifice after 16 days of exposure. Two distinct
changes occurred quite early and close to the same time: the first to appear was an increase in the
number of neurofilaments and the other was an in-pouching of the myelin sheath. In animals
exposed to 400 ppm, the first observable change at 16 days was, in larger diameter nerve fibers, a
two- to threefold increase in the number of neurofilaments. As the duration of exposure
lengthened and the number of neurofilaments increased, several interrelated morphologic
observations were made. In teased nerve fiber preparations, swelling of the axons could be seen
frequently in the paranodal area and less often at focal sites along the internodal segment. High
numbers of nerve fibers with in-pouching of the myelin sheath were found per mm2 of nerve
fascicle, increasing with time after administration of the high concentration. A summary of the
comparative sequential clinical and pathological observations is presented in Table 4-13.
47
-------
Table 4-13. Clinical and pathological observations with time of exposure to
2-hexanone in rats
Days exposed
Clinical findings
In-pouchings (no./mm2)
Denuded fibers (no./mm2)
Swollen axons >1 1 um (no. 7300 fibers)
2-Hexanone exposure
400 ppm
16
Na
6
0
0
28
N
142
4
1
42
pb
499
11
3
225 ppm
16
N
23
0
0
25
N
46
0
0
35
N
92
1
0
55
N
86
2
0
aN = normal.
bP = paralyzed.
Source: Saidaetal. (1976).
The anterior horn cells, nerve roots, nerve trunks, intramuscular nerves, and motor end
plates were studied sequentially to determine the site with the earliest pathological involvement.
In animals exposed for 16 days to 225 ppm, no abnormalities were found in the motor end plates
or intramuscular nerves of the intrinsic foot muscles. Only after prolonged exposure, 66 days,
did the authors find typical signs of denervation in the motor end plates. These end plates
showed atrophic axon terminals with Schwann cell processes interposed between the nerve
terminal and postsynaptic membrane. There was also a loss of secondary synaptic clefts.
Anterior horn cells and dorsal root ganglion cells were also examined at various intervals
of exposure. No changes were observed in these cell bodies, even after typical changes were
seen in the main trunk of the sciatic nerve. Specifically, no abnormalities were seen that would
suggest an increase in neurofilaments in these cell bodies, and no cells were observed
undergoing chromatolysis.
4.5.2. Genotoxicity Studies
Mayer and Goin (1994) tested the ability of 2-hexanone to induce chromosome loss in
strain D61.M of Saccharomyces cerevisiae. 2-Hexanone, alone or in combination with acetone
and MEK, induced only a marginally positive chromosome loss (Mayer and Goin, 1994).
No data were identified for the mutagenicity of 2-hexanone with in vitro cytogenetic tests
or in vivo tests.
4.5.3. Structure-Activity Relationships
A large body of toxicological information is available on n-hexane, a compound that is
metabolized to 2-hexanone, on MiBK (a branched-chain homologue of 2-hexanone), and on
MEK. These compounds have been reviewed in previous IRIS assessments, and a summary of
the reference values derived for each is presented in Table 4-14. n-Hexane is the only compound
48
-------
of the above three that is also capable of producing the peripheral neuropathy similar to that
observed in humans or animals exposed to 2-hexanone. Neither MiBK nor MEK can give rise to
the neurotoxic metabolite 2,5-hexanedione.
Table 4-14. Summary of the toxicities of n-hexane, MiBK, and MEK
Chemical
n-Hexane
(CASRN 110-54-3)
MiBK (CASRN 108-10-1)
MEK (CASRN 78-93-3)
Experimental
dose
NOAEL":
1,762 mg/m3
NOAEL:
1,229 mg/m3
LECb:
5,202 mg/m3
NOAEL:
594 mg/kg-day
(0.3%2-butanol)
Critical effect
Peripheral neuropathy (decreased
MCV at 12 weeks)
Reduced fetal body weight,
increased fetal death, and skeletal
variations in mice and rats
Developmental toxicity (skeletal
variations)
Decreased pup body weight
Reference
value
RfC:
7 x 10"1 mg/m3
RfC:
3 mg/m3
RfC:
5 mg/m3
RfD:
0.6 mg/kg-day
Reference
U.S. EPA
(2005c)
U.S. EPA
(2003a)
U.S. EPA
(2003b)
aNOAEL = no-observed-adverse-effect level.
bLEC = lowest effective concentration.
4.5.4. Potentiation and Other Interaction Studies
4.5.4.1. Methyl Ethyl Ketone
In a study of chemical interaction, Saida et al. (1976) exposed rats of unspecified sex
(12/group) continuously, 24 hours/day, to 225 ppm (922 mg/m3) 2-hexanone, 1,125 ppm
(3,318 mg/m3) MEK, or a combined exposure of 225 ppm (922 mg/m3) 2-hexanone and
1,125 ppm MEK for up to 66 days. No signs of neurotoxicity were observed in the MEK-
exposed rats. Paralysis occurred earlier in the rats exposed to the mixture compared with rats
exposed to 225 ppm 2-hexanone alone. In addition, an elevated severity of neuropathy in the
form of increased swollen axons, denuded fibers, and in-pouching of myelin sheaths was
observed histologically in the rats coexposed to MEK and 2-hexanone. Thus, MEK appeared to
potentiate the toxicity of 2-hexanone. Yu et al. (2002) showed that the potentiating effect of
MEK on n-hexane-induced neurotoxicity was due to an inhibitory effect of MEK on phase II
biotransformation of 2,5-hexanedione. Since n-hexane is a precursor to 2-hexanone and both
compounds form the highly toxic 2,5-hexanedione, it is likely that the results of Yu et al. (2002)
are applicable to co-exposure studies with MEK and 2-hexanone.
As a test of in vivo enzyme induction, groups of five male Wistar rats were continuously
exposed via inhalation to 225 ppm 2-hexanone, 750 ppm MEK, or the combination of 225 ppm
2-hexanone and 750 ppm MEK for 7 days (Couri et al., 1977). Subsequently, the animals were
given sodium hexobarbital (100 mg/kg, i.p.), a substrate for phenobarbital-inducible CYP450
isoenzymes (Adedoyin et al., 1994; Knodell et al., 1988), and sleep time was measured. The
49
-------
average hexobarbital-induced sleep time of 2-hexanone-treated rats was comparable to that of
controls (24.8 versus 26.0 minutes); however, the sleep times in MEK and 2-hexanone/MEK-
exposed rats were significantly (p < 0.05) less than in controls, 13.0 and 16.0 minutes,
respectively. In a study by O'Donoghue and Krasavage (1979), sodium pentobarbital-induced
sleep time was increased in 2-hexanone-treated cats.
4.5.4.2. Chloroform
Oral administration of 2-hexanone, followed by i.p. administration of chloroform to rats,
resulted in a variety of hepatic and renal effects, including decreased hepatic glutathione levels,
increased plasma levels of glutamic pyruvic transaminase and blood urea nitrogen, and
degeneration and necrosis of hepatic and renal tissue (Hewitt et al., 1990, 1987; Brown and
Hewitt, 1984; Branchflower and Pohl, 1981). Similarly, oral administration of both 2-hexanone
and chloroform to rats resulted in altered permeability of the biliary tree (Hewitt et al., 1986). In
these studies, some or no effect on the endpoints of interest was observed after administration of
2-hexanone or chloroform alone; administration of both substances resulted in statistically
significant and dramatic changes in these effects. The authors speculated that 2-hexanone
potentiated the hepatic toxicity of chloroform by decreasing glutathione levels and by increasing
the metabolism of chloroform to the potent hepatotoxicant phosgene.
4.5.4.3. O-Ethyl O-4-Nitrophenyl Phenylphosphonothioate
2-Hexanone has been shown to potentiate the neurotoxic effects of EPN. In hens, dermal
or inhalation exposure to 2-hexanone in combination with dermal application of the
organophosphate pesticide EPN has resulted in earlier onset and far more severe clinical and
histological manifestations of neurotoxic effects than with either chemical exposure alone
(Abou-Donia et al., 1991, 1985b). The authors speculated that this potentiation effect may have
been due to induction of hepatic microsomal CYP450 by EPN, leading to increased metabolism
of 2-hexanone to its neurotoxic metabolite 2,5-hexanedione. An alternate explanation is that
local trauma to the nervous tissue produced by 2-hexanone and EPN might increase vascular
permeability and thus increase the entry of these compounds and their metabolites from
circulation.
4.6. SYNTHESIS AND EVALUATION OF MAJOR NONCANCER EFFECTS
4.6.1. Oral
No studies of the possible association between oral exposure to 2-hexanone and
noncancer health effects in humans are available. There are six oral toxicity studies of
2-hexanone in experimental animals with exposures ranging from 3 to 13 months. These include
a 90-day gavage study in hens, 90-day and 40-week gavage studies in rats, 120-day and 13-
month drinking water studies in rats, and a 24-week drinking water study in guinea pigs. These
50
-------
studies demonstrate that the nervous system is the target organ for 2-hexanone toxicity following
oral exposure. For example, O'Donoghue et al. (1978), a 13-month drinking water study using
COBS CD(SD)BR rats, described the characteristic neuropathological evidence of giant axonal
neuropathy in 80% of animals at the lowest dose tested (143 mg/kg-day).
Available data suggest that the principal metabolite of 2-hexanone, 2,5-hexanedione, is
responsible for the neurotoxicity associated with oral exposure to 2-hexanone. For example,
Krasavage et al. (1980) compared the neurotoxicity of 2-hexanone with that of n-hexane,
5-hydroxy-2-hexanone, 2,5-hexanediol, and 2-hexanol by administering equimolar doses of each
chemical by gavage to five male COBS CD(SD)BR rats/group, 5 days/week for 90 days. Judged
by the time required for the rats to develop hind-limb paralysis, 2,5-hexanedione had a higher
neurotoxic potency than 2-hexanone.
In summary, the chronic and subchronic studies conducted with rats, hens, and guinea
pigs provide evidence that the nervous system is the target of toxicity following oral exposure to
2-hexanone. A summary of the oral studies with 2-hexanone is provided in Table 4-15.
51
-------
Table 4-15. Synopsis of oral toxicity studies with 2-hexanone
Species,
strain
Adult leghorn
laying hens
(G. gallus
domesticus)
Rat,
COBS/
CD(SD)BR
Rat,
Wistar
Guinea pig,
English
shorthair
Rat,
Wistar
Rat,
COBS/
CD(SD)BR
Group
size (sex)
3/group
(female)
6/group
(male)
5/group
(female)
5/group
(sex not
stated)
6/group
(male)
10/group
(male)
Dosage;
duration; purity
100 mg/kg,
gavage; 7
days/week for 90
days; technical
grade containing
70% 2-hexanone
and 30% MiBK
660 mg/kg,
gavage;
5 days/week for 90
days; 2-hexanone
containing 3. 2%
MiBK and 0.7%
unknown
contaminants
0,0.65, or 1.3%
(0, 480, or 1,010
mg/kg-day) in
drinking water;
120 days; purity
not stated
0,0.1, or 0.25%
(0,97, or 243
mg/kg-day) in
drinking water;
24 weeks; purity
not stated
400 mg/kg-day,
gavage; 40 weeks;
2-hexanone 98%
pure, contaminants
not characterized
0,0.25, 0.5, or
1.0% (0, 143,266,
or 560 mg/kg-day)
in drinking water;
13 months;
2-hexanone
containing 3. 2%
MiBK and 0.7%
unknown
contaminants
Effects at LOAEL
Mild ataxia at 12 ± 1
days with progression
to severe ataxia by
50 ± 1 days
Clinical and
histological findings of
neuropathy at 55.8 ±
4.3 days
Mild atrophy affecting
skeletal muscles of the
hind limbs in 2 of 5
animals examined
Decreased pupillary
response to light
stimulus
Hind-limb weakness
from the H^S*
week, with
improvement thereafter
Clinical neurological
deficits
Neuropathological
evidence of
myofibrillar atrophy of
the calf muscle in 1/10
animals
Neuropathological
evidence of
myofibrillar atrophy of
the quadriceps muscle
in 2/10 animals
Neuropathological
evidence of giant
axonal neuropathy in
8/10 animals
NOAEL3
(mg/kg-day)
Not
identified
Not
identified
Not
identified
Not
identified
Not
identified
143
143
143
Not
identified
LOAEL3
(mg/kg-day)
100
660
480
97
400
266
266
266
143
Reference
Abou-Donia
etal. (1982)
Krasavage et
al. (1980)
Homan et al.
(1977)
Abdel-
Rahman et
al. (1978)
Eben et al.
(1979)
O'Donoghue
etal. (1978)
aNo-observed-adverse-effect levels (NOAELs) and lowest-observed-adverse-effect levels (LOAELs) determined by
2-hexanone assessment authors.
52
-------
4.6.2. Inhalation
Several studies have established associations between inhalation exposure to 2-hexanone
and human health effects. Specifically, occupational studies and case reports suggest that
inhalation exposure to 2-hexanone in humans may be associated with neurotoxicity. For
example, a cross-sectional study of employees at a coated fabrics plant was conducted when it
was noted that six workers from the print department had developed severe peripheral
neuropathy soon after the plant began phasing in the use of 2-hexanone (Allen et al., 1974;
Billmaier et al., 1974). Definite signs, symptoms, and electrodiagnostic findings of peripheral
neuropathy were confirmed in 68 out of 192 employees. The prevalence of peripheral
neuropathy was clearly increased in jobs with evident exposure to 2-hexanone vapors and with
time spent at work sites with 2-hexanone exposure.
Mallov (1976) reported one probable and two definite cases of 2-hexanone-induced
peripheral neuropathy that were identified during an investigation of 26 painters. Similar to the
studies reported above (Allen et al., 1974; Billmaier et al., 1974), neuropathy was observed in the
painters when the formulation of paint solvents was changed from MEK and methyl isoamyl
ketone, both of which are considered not to be neurotoxic, to 2-hexanone (Mallov, 1976). In
another case of occupational exposure to 2-hexanone, symmetrical polyneuropathy was reported
in a furniture finisher (Davenport et al., 1976). Six months prior to the onset of the worker's
illness, 2-hexanone had been substituted for MiBK. A similar progressive distal extremity
weakness developed in a coworker of the patient, which also improved following the coworker's
removal from contact with lacquer products.
The toxicity of 2-hexanone via inhalation was studied extensively in experimental
animals. As with oral exposures, the target organ for toxicity following inhalation exposure to
2-hexanone was the nervous system, and the most sensitive measures of intoxication were
histopathological and clinical findings of peripheral neuropathy. Numerous subchronic and
chronic studies are available in different test species, including monkeys, rats, and cats. A
summary of the available inhalation studies with 2-hexanone is provided in Table 4-16.
53
-------
Table 4-16. Synopsis of animal inhalation toxicity studies with 2-hexanone
Species,
strain
Number
(sex)
Concentration;
duration; purity
Develo^
Rat,
pregnant
F-344
25/group
(female)
0, 1,000, or 2,000 ppm (0,
4,100, or 8,200 mg/m3);
day 0 of gestation through
day21,6h/day, 7 d/wk;
purity not stated
Effects at LOAEL
NOAEL"
(mg/m3)
LOAEL"
(mg/m3)
Reference
vmental study
Hyperactivity in
behavioral testing
Not
identified
4,100
Peters et al.
(1981)
Subchronic exposure studies
Rat,
strain not
stated
Rat,
Wistar
Rat,
Sprague-
Dawley
Rat,
COBS/
CD(SD) BR
Rat,
Sprague-
Dawley
Monkey, M.
fascicularis
Adult
leghorn
laying hens
(G. gallus
domesticus)
Domestic
chicken
Cat,
domestic,
strain not
stated
9/group
(sex not
stated)
20/group
(sex not
stated)
12/group
(sex not
stated)
5/group
(male)
4/group
(sex not
stated)
8/group
(male)
5/group
5/group
(sex not
stated)
4/group
(sex not
stated)
0 or 200 ppm (0 or 819
mg/m3); 6 weeks, 8 h/d,
5 d/wk; purity not stated
0, 40 ppm (164 mg/m3) for
22-88 days, or 50 ppm
(205 mg/m3) for 13 weeks,
8h/d, 5d/wk; purity not
stated
0, 225, or 400 ppm (0,
922.5, or 1,640 mg/m3);
42-66 days, 24 h/d,
7 d/wk; purity not stated
0 or 700 ppm (0 or 2,870
mg/m3); 81 days, 72 h/wk;
96.1% pure with 3. 2%
MiBK and 0.7%
unidentified contaminants
0 or 400 ppm (0 or 1,640
g/m3); 12 weeks, 24 h/d,
7 d/wk; purity not stated
0, 100, or 1,000 ppm (0,
410, or 4, 100 mg/m3);
10 months, 6 h/d, 5d/wk;
commercial grade,
impurities not stated
0, 10, 50, 100, 200, or 400
ppm (0,41, 205, 410, 820,
or 1,640 mg/m3); 90 days
(continuous exposure);
technical grade (70%
2-hexanone, 30% MiBK)
0 or 100 ppm (4 10 mg/m3)
(time not stated);
12 weeks, 24 h/d, 7 d/wk;
purity not stated
0 or 400 ppm (1,640
mg/m3); 12 weeks, 24 h/d,
7 d/wk; purity not stated
Axonal
hypertrophy,
beading, and
degeneration of
sciatic nerve
Peripheral
neuropathy
(demyelination of
sciatic nerve) in
3/20 animals in
50 ppm group
Increased number
of fibers with in-
pouchings per mm2
of nerve fascicle
Severe neuropathy
consisting of
difficulty extending
hind limbs and a
flat-footed gait with
feet splayed in 5/5
at 71 ±9 days
Dragging of hind
limbs at 11-12
weeks
Decreased MCV at
9 months (right
sciatic-tibia! nerve,
right ulnar nerve)
Mild ataxia (27 ± 2
days) progressing to
severe ataxia/near
paralysis
(89 ± 1 days)
Inability to stand on
legs at 4-5 weeks
Dragging of hind
limbs and forelimb
weakness at 5-8
weeks
Not
identified
164
Not
identified
Not
identified
Not
identified
Not
identified
41
Not
identified
Not
identified
819
205
922.5
2,870
1,640
410
205
410
1,640
Duckett et al.
(1974)
Duckett et al.
(1979)
Saida et al.
(1976)
Katz et al.
(1980)
Mendell et al.
(1974)
Johnson et al.
(1977)
Abdo et al.
(1982)
Mendell et al.
(1974)
Mendell et al.
(1974)
54
-------
Table 4-16. Synopsis of animal inhalation toxicity studies with 2-hexanone
Species,
strain
Number
(sex)
Concentration;
duration; purity
Effects at LOAEL
NOAEL3
(mg/m3)
LOAEL3
(mg/m3)
Reference
Chronic exposure studies
Rat,
strain not
stated
Rat,
Wistar
Rat,
Sprague-
Dawley
Rat,
Sprague-
Dawley
Rat,
Sprague-
Dawley
Cat,
domestic
shorthair
6/group
(sex not
stated)
40/group
(sex not
stated)
6/group
(male)
10/group
(male)
18/group
(male)
4/group
(female)
0 or 1,300 ppm (5,325
mg/m3); 4 months, 6 h/d,
5d/w; purity not stated
0 or 50 ppm (205 mg/m3);
6 months, 8h/d, 5 d/wk;
purity not stated
0 or 100 ppm (0 or 4 10
mg/m3); 6 months, 22 h/d,
7 d/wk; 96.66% pure,
impurities not
characterized
0, 100, or 1,000 ppm (0,
410, or 4, 100 mg/m3);
10 months, 6 h/d, 5d/wk;
commercial grade,
impurities not stated;
LOAEL based on 6 months
of exposure
0, 100, or 330 ppm (0,410,
or 1,353 mg/m3);
72 weeks, 6h/d, 5d/wk;
purity not stated
0, 100, or 330 ppm (0,410,
or 1,353 mg/m3); 2 years,
6 h/d, 5d/wk; purity not
stated
Nerve fiber
degeneration in the
peripheral nerves,
spinal cord,
medulla, and
cerebellum
Widespread
demyelination of
the sciatic nerve in
32/40
Giant axonal
swelling of
peripheral nerves
after 4 months
Decreased MCV
between treated and
control animals,
beginning at
29 weeks
Severe
polyradiculoneuritis
in the lumbar and
sacral spinal nerves
and roots and the
sciatic and tibial
nerves in one rat
Giant axonal
neuropathy of the
spinal cord and
peripheral nerve in
4/4
Not
identified
Not
identified
Not
identified
Not
identified
410
410
5,325
205
410
410
1,353
1,353
Spencer et al.
(1975)
Duckett et al.
(1979)
Egan et al.
(1980)
Johnson et al.
(1977)
Krasavage
and
O'Donoghue
(1977)
O'Donoghue
and
Krasavage
(1977)
aNo-observed-adverse-effect levels (NOAELs) and lowest-observed-adverse-effect levels (LOAELs) determined by
2-hexanone assessment authors.
4.6.3. Mode-of-Action Information
Exposure to 2-hexanone in humans and experimental animals demonstrates that the
nervous system is the target organ of toxicity, regardless of the route of exposure. The toxicity is
attributed to the neurotoxic metabolite 2,5-hexanedione. A strong relationship has been noted
between the concentration of 2,5-hexanedione in the urine and the onset of neuropathic
symptoms (Eben et al., 1979). Similarly, 2,5-hexanedione has been described as eliciting severe
neurotoxic symptoms following oral, dermal, or i.p. administration to hens and oral
administration to rats (Abou-Donia et al., 1985a; Abdo et al., 1982; Krasavage et al., 1980).
55
-------
Current research supports a mode of action for y-diketones, such as the 2-hexanone
metabolite 2,5-hexanedione, that involves the covalent cross-linking of neuronal macromolecules
with proteins as the primary target. The result is axonal swelling, specifically of giant axons,
that ultimately ends in retrograde degeneration of the axon. 2,5-Hexanedione is an electrophilic
species that reacts with nucleophilic sites of proteins via a substitution or addition reaction, with
the subsequent formation of a covalent bond (LoPachin and DeCaprio, 2005). Although
2,5-hexanedione has been shown to react with sulfhydryl groups of enzymes (Section 4.5.1.2),
the compound causes distal axonopathy by covalent reaction with nucleophilic lysine s-amino
groups to form 2,5-dimethylpyrrole adducts with neurofilaments and other proteins (LoPachin et
al., 2005, 2004). Oxidation of the pyrrole moiety with molecular oxygen can generate a cation
intermediate that can undergo further reactions with amino or sulfhydryl groups. This results in
the development of neurofilament aggregates in the distal, subterminal axon that, as they grow
larger, form massive swellings, displacing paranodal myelin sheaths (Spencer and Schaumburg,
1977). These axonal swellings often occur just proximal to the nodes of Ranvier (Graham,
1999).
One of the major hypotheses related to the mechanism of neurotoxicity of
2,5-hexanedione is covalent binding with axonal components of nerve tissue. In vitro studies in
which 2,5-hexanedione was incubated with proteins demonstrated that this compound binds to
the lysine s-amino group, resulting in the formation of the substituted pyrrole adduct s-N-(2,5-
dimethylpyrrole)norleucine (DeCaprio et al., 1982). Covalent binding of 2,5-hexanedione with
axonal components leading to pyrrole formation and protein cross-linking was hypothesized as a
possible initiation step leading to axonal degeneration and thus may account for the neurotoxic
effects observed with exposure to y-diketones in general (DeCaprio et al., 1988, 1982). In vivo
pyrrole formation was confirmed by the demonstration of s-N-(2,5-dimethylpyrrole)norleucine
in the hydrolyzed serum of a hen that had received 2,5-hexanedione at 200 mg/kg-day for
2 weeks (DeCaprio et al., 1982). The proposed mechanism for 2,5-hexanedione in the
development of progressive sensorimotor distal axonopathy is presented in Figure 4-1.
56
-------
o
o
H3C — C — CH2— CH2—C — CH3
2,5-hexan.edione
Protein
Protein
Protein
N
Protein
NH,
Figure 4-1. Proposed mechanism for 2,5-hexanedione-induced axonopathy.
Note: y-Diketones, such as 2,5-hexanedione, react with amino groups in all tissues to form
pyrroles. The pyrrole moiety can undergo further oxidation reactions with amino or sulfhydryl
groups. This results in the development of neurofilament aggregates (in the distal, subterminal
axon), which, as they grow larger, form massive swellings of the axon.
Source: Adapted from DeCaprio etal. (1988, 1982).
4.7. WEIGHT-OF-EVIDENCE EVALUATION AND CANCER CHARACTERIZATION
4.7.1. Summary of Overall Weight of Evidence
Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), there is
"inadequate information to assess the carcinogenic potential" of 2-hexanone. Specifically, there
are no animal carcinogenicity studies available that examine exposure to 2-hexanone, and there
are no studies available that assert a mutagenic potential of 2-hexanone. The available
occupational studies do not present evidence for carcinogenic action of 2-hexanone, although
these are limited by frequent co-exposure to other chemicals (e.g., MEK).
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
4.8.1. Possible Childhood Susceptibility
The susceptibility of the developing brain is based on the timing of neuronal
development, the rapid growth that occurs in the third trimester and early infancy, and the lack of
a protective barrier early in life (Costa et al., 2004). In the cerebellum, Purkinje cells develop
early, weeks 5-7 in humans, whereas granule cells are generated much later, gestational
57
-------
weeks 24-40 in humans. The developing brain is distinguished by the absence of a blood-brain
barrier. The development of this barrier is a gradual process, beginning in utero and complete at
approximately 6 months of age. Because the blood-brain barrier limits the passage of substances
from blood to brain, in its absence, toxic agents can freely enter the developing brain. Since
Purkinje-cell degeneration has been observed with adult rats exposed to high levels of
2,5-hexanedione, infants may be at an increased risk for this type of damage at lower levels of
exposures, due to the incomplete maturation of the blood-brain barrier (Hernandez-Viadel et al.,
2002). However, this would depend on the capacity of infants and small children to bioactivate
2-hexanone to 2,5-hexanedione.
Metabolism of 2-hexanone may vary between children and adults due to differences in
the development and maturity of phase I and phase II enzymes (Johnsrud et al., 2003). Studies
indicate that the mode of action of 2-hexanone toxicity involves the metabolism to a more toxic
metabolite, namely, 2,5-hexanedione. Several enzymes, such as CYP2E1, CYP2B1/2, and
CYP2C6/11, are inducible following administration of 2-hexanone in animal models (Imaoka
and Funae, 1991; Nakajima et al., 1991); however, the individual isoforms involved in
2-hexanone metabolism have not been fully elucidated. Toftgard et al. (1986) found that the
formation of 2,5-hexanediol from 2-hexanol was catalyzed by a CYP450 isozyme different from
CYP2B and present in liver but not in lung microsomes. The authors concluded that 2-hexanol
must be transported to the liver before the neurotoxic metabolite 2,5-hexanedione can be formed.
Because of this, changes in CYP450 protein levels and phase II enzymes during development
may have an impact on susceptibility to 2-hexanone. As mentioned above, the possible
susceptibility of 2-hexanone may be influenced by life stage, but there are few studies to confirm
the impact and severity of such exposure. The available information suggests that young animals
and children could more susceptible to 2-hexanone; however, the evidence of possible childhood
susceptibility is inconclusive.
4.8.2. Possible Gender Differences
Evaluations of human occupational exposures have not provided evidence that
2-hexanone acts in a gender-specific way. Most animal studies also have not brought forth
strong evidence for a sex-specific action of 2-hexanone. However, it should be mentioned that in
a few rat studies 2-hexanone appeared to affect the male reproductive system (Katz et al., 1980;
Krasavage et al., 1980; O'Donoghue et al., 1978).
58
-------
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
The RfD is an estimate (with uncertainty spanning perhaps an order of magnitude) of a
daily oral exposure to the human population (including sensitive subgroups) that is likely to be
without an appreciable risk of deleterious effects during a lifetime. It can be derived from a no-
observed-adverse-effect level (NOAEL), lowest-observed-adverse-effect level (LOAEL), or
benchmark dose (BMD), with uncertainty factors (UFs) generally applied to reflect limitations of
the data used.
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
There are no human studies that have examined the possible association between oral
exposure to 2-hexanone and noncancer health effects.
The 13-month drinking water study (10 animals/dose/sex) conducted by O'Donoghue et
al. (1978) was selected as the principal study for deriving an RfD for 2-hexanone. Five other
available subchronic studies are considered to be supporting studies. Of these five studies,
Krasavage et al. (1980) and Eben et al. (1979) both observed neurotoxicity after administration
of single doses of 2-hexanone via gavage. These two studies were not considered as principal
studies because only single, relatively high doses were administered. Abdo et al. (1982)
observed mild ataxia, which progressed to severe ataxia, in hens treated daily by gavage with
100 mg/kg 2-hexanone. Although the hen is a sensitive model for some neurotoxic effects, this
study was not chosen as the principal study because doses contained high levels of MiBK (30%).
Two subchronic drinking water studies, one in the rat and a second in the guinea pig, that utilized
multiple doses of 2-hexanone and identified neurotoxicological outcomes were considered as
candidate principal studies. The rat study by Homan et al. (1977) utilized doses that were higher
than those used by O'Donoghue et al. (1978), and the purity of 2-hexanone was not stated. The
study in the guinea pig by Abdel-Rahman et al. (1978) utilized doses of 97 and 243 mg/kg-day;
however, only data from the first 4 weeks of the study were presented. Although the 97 mg/kg-
day dose used by Abdel-Rahman et al. (1978) is lower than the lowest dose in the 13-month
study by O'Donoghue et al. (1978), the data from the 97 mg/kg-day group were not reported.
Further, the purity of the compound used was not stated.
O'Donoghue et al. (1978) administered 2-hexanone (96% pure, containing 3.2% MiBK
and 0.7% unknown contaminants) to male COBS/CD(SD) rats in drinking water at
concentrations of 0, 0.25, 0.5, or 1.0% (0, 143, 266, or 560 mg/kg-day) for 13 months. Because
rats were exposed for more than half of their life span, the study duration was considered to be
chronic. In this study, 2-hexanone produced a dose-dependent reduction in body weight at all
doses tested with decreases of 4, 14, and 37% at dose levels of 143, 266, and 560 mg/kg-day,
59
-------
respectively. The authors did not report statistical variances for the body weight data or indicate
whether the reductions were statistically significant. Dose-related increases in relative but not
absolute liver, kidney, and testes weights were statistically significant at the highest dose. The
relative kidney weight increase was also statistically significant at the 266 mg/kg-day dose.
Clinical neurological deficits were observed at the two highest doses.
Neuropathological evidence of axonal neuropathy was present in animals of each dose
level. Neuropathological evidence of myofibrillar atrophy of the calf muscle and the quadriceps
muscle was present in animals at the two highest doses. Although degenerative changes were
observed in controls, both the incidence and severity of neuropathological changes were dose
dependent, providing evidence that they were induced by 2-hexanone. The critical endpoint
selected from the O'Donoghue et al. (1978) study was the incidence of swollen axons in
peripheral nerves of male rats. This endpoint was chosen because axonal neuropathy of the
peripheral nerve is consistently identified in occupationally exposed humans and experimental
animals following low-level exposures to 2-hexanone. Axonal swelling was observed with high
incidence in the peripheral nerve at the lowest dose tested and is the most sensitive endpoint
observed in this study. Some studies (Lopachin et al., 2004, 2003; Lehning et al., 2000, 1995)
have suggested that axonal swelling may occur without progression to nerve dysfunction;
however, the neurological findings in O'Donoghue et al. (1978) provide evidence of progression.
Myofibrillar atrophy, an effect observed subsequent to axonal swelling, was present in all dose
groups that showed axonal swelling although at a lower incidence in the low- and mid-dose
groups than axonal swelling.
5.1.2. Method of Analysis: Benchmark Dose Modeling
The animal data evaluated for derivation of an RfD for 2-hexanone are displayed in
Table 5-1.
60
-------
Table 5-1. Summary of neuropathological findings in male rats administered
2-hexanone in drinking water for 13 months
Treatment
Control
0.25%
2-Hexanone
(143 mg/kg-day)
0.5% 2-Hexanone
(266 mg/kg-day)
1.0% 2-Hexanone
(560 mg/kg-day)
Axonal swelling
Myofibrillar atrophy
Incidence per number of animals exposed
Brain
0/10
2/10
4/10
8/10
Spinal cord
0/5
7/10
5/5
5/5
Dorsal root
ganglia
0/5
0/7
0/5
3/5
Peripheral
nerve"
0/10
8/10
10/10
10/10
Quadriceps
muscle
0/10
1/10
5/10
10/10
Calf muscle
0/10
2/10
6/10
10/10
"Data in bold were further evaluated for RfD derivation.
Source: O'Donoghue etal. (1978).
These data are from a 13-month toxicity study in rats in which 10 animals per dose group
were administered 2-hexanone in drinking water at four different concentrations (i.e., 0, 0.25,
0.5, and 1.0% corresponding to doses of 0, 143, 266, and 560 mg/kg-day) for 13 months
(O'Donoghue et al., 1978). The critical endpoint selected from this study was the incidence of
swollen axons in peripheral nerves of male rats. As stated above, this endpoint was selected
from the other neuropathological endpoints in Table 5-1 because peripheral neuropathy is the
most consistent and relevant effect identified in occupationally exposed humans and
experimental animals that occurs following low-level exposures to 2-hexanone. The LOAEL in
this study occurred at the lowest concentration of 2-hexanone administered (0.25%, 143 mg/kg-
day), which yielded an 80% incidence of giant or swollen axons in the peripheral nerves of
exposed animals.
U.S. EPA's BMD software (BMDS), version 1.4.1c (U.S. EPA, 1999), was used to
estimate a point of departure (POD) for deriving an RfD for 2-hexanone from data on axonal
swelling of the peripheral nerve. The POD was defined as the 95% lower confidence limit on the
BMD (BMDL) associated with a benchmark response (BMR) of 10% extra risk of axonal
swelling of the peripheral nerve. A BMR of 10% is generally used in the absence of information
regarding what level of change is considered biologically significant, and also to facilitate a
consistent basis of comparison across assessments (U.S. EPA, 2000b). All of the available
dichotomous models in BMDS were fit to the axonal swelling incidence data; Table 5-2 presents
the best-fit model result. Details of the probability function and the BMD modeling results are
contained in Appendix B-l.
61
-------
Table 5.2. Best-fit BMD modeling results for data on axonal swelling of the
peripheral nerve
Endpoint
Axonal swelling of the peripheral nerve
Model
Multistage
AICa
12.0784
p Value
0.9981
BMD10
(mg/kg-day)
36.1
BMDL10
(mg/kg-day)
5.1
aAIC = Akaike Information Criterion.
5.1.3. Derivation of Human Equivalent Doses
For 2-hexanone, no PBTK model is currently available. Therefore, the first step required
for RfD derivation is to determine whether intermittent doses were employed in the animal study
and, if so, to adjust these doses to reflect continuous exposures based on the assumption that the
product of dose and time is constant (U.S. EPA, 2002). In the principal study (O'Donoghue et
al., 1978), animals were administered 2-hexanone in drinking water 24 hours/day, 7 days/week
for 13 months. Therefore, in this case, a duration adjustment is not required (i.e., the POD
[adjusted BMDL or BMDLADj] for 2-hexanone equals the study BMDL) as follows:
BMDLAoj = BMDL x (number of hours per day exposed/24 hours) x (number of days
per week exposed/7 days)
BMDLAoj = 5 mg/kg-day x (24 hours/24 hours) x (7 days/7 days) = 5 mg/kg-day
The BMDLAoj is used as the POD to which UFs were applied.
5.1.4. Calculation of the RfD — Application of Uncertainty Factors
The RfD for axonal swelling of the peripheral nerve as the critical effect is calculated
from the BMDLio (ADJ) by application of UFs as follows:
~3
RfD = 5 mg/kg-day + 1,000 = 0.005 mg/kg-day = 5 x 10~ mg/kg-day
The composite UF of 1,000 was derived as follows:
• A default intraspecies UF (UFH) of 10 was applied to adjust for potentially sensitive
human subpopulations. A default value is warranted because insufficient information is
currently available to assess human-to-human variability in 2-hexanone toxicokinetics or
toxicodynamics.
• A default interspecies UF (UFA) of 10 was applied for extrapolation from animals to
humans. No data on the toxicity of 2-hexanone to humans exposed by the oral route were
62
-------
identified. Insufficient information is currently available to assess rat-to-human
differences in 2-hexanone toxicokinetics or toxicodynamics.
• An UF of 10 was applied to account for database deficiencies (UFo). The database
includes subchronic animal studies in rats and hens and a 13-month study in rats but does
not include a multigenerational reproductive study or developmental studies.
Additionally, there are inhalation studies that suggest the possibility of reproductive and
immunological toxicity following exposure to 2-hexanone.
• An UF for LOAEL-to-NOAEL extrapolation (UFO was not used because the current
approach is to address this factor as one of the considerations in selecting a BMR for
BMD modeling. In this case, a BMR of 10% extra risk of axonal swelling of the
peripheral nerve was selected under an assumption that it represents a minimal
biologically significant change.
• A subchronic-to-chronic UF (UFs) was not applied. Although the principal study
(O'Donoghue et al., 1978) was not a standard 2-year bioassay, rats were exposed for 13
months, or more than half of their life span. Therefore, the exposure period used in the
principal study was considered to be of chronic duration.
5.1.5. RfD Comparison Information
Figure 5-1 presents potential PODs, applied UFs, and derived sample RfDs for the
endpoints considered for 2-hexanone. As stated previously, of the available chronic and
subchronic studies, the 13-month drinking water study by O'Donoghue et al. (1978) was selected
as the principal study to derive the RfD. Axonal swelling in the peripheral nerve was selected as
the critical effect because peripheral neuropathy was deemed the most sensitive and relevant
effect. Other endpoints, such as myofibrillar atrophy of the quadriceps and calf muscles, are also
noted in O'Donoghue et al. (1978) and are thus also illustrated in Figure 5-1 for comparison
purposes. BMD modeling outputs for the endpoints in Figure 5-1 are presented in Appendix
B-l. The supporting studies outlined in Table 4-15 were deemed less relevant to human
exposure because they either involved single, relatively high doses via gavage in rodents
(Krasavage et al., 1980; Eben et al., 1979), used a test substance with high levels of MiBK (Abdo
et al., 1982), were subchronic in design with higher doses administered than the 13-month study
by O'Donoghue et al. (1978) (Homan et al., 1977), or limited data were provided (Abdel-
Rahmanetal., 1978).
63
-------
5"
•a
"Sfc
o
a
1000H
100 -
10 -
1 -
0.1 -
0.01-
0.001
Point of Departure
LTF, animal to human
LTF, human variability
LTF, database
LTF, subchronic to chronic
LTF, LOAEL to NOAEL
RiD
Axonal swelling:
periperal nerve
Myofibrillar atrophy:
quadriceps muscle
Myofibrillar atrophy:
calf muscle
Figure 5-1. Potential PODs for endpoints from O'Donoghue et al. (1978),
with corresponding applied UFs and derived sample oral reference values.
64
-------
5.1.6. Previous Oral Assessment
An RfD assessment for 2-hexanone was not previously available on IRIS.
5.2. INHALATION REFERENCE CONCENTRATION
The inhalation RfC is an estimate (with uncertainty spanning perhaps an order of
magnitude) of a continuous inhalation exposure to the human general population (including
sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects over a
lifetime. It can be derived from a NOAEL, a LOAEL, or a benchmark concentration (BMC),
with UFs generally applied to reflect uncertainties and/or limitations in the data used.
5.2.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
An inhalation study that exposed monkeys and rats to 0, 100, or 1,000 ppm (0, 410, or
4,100 mg/m3) commercial grade 2-hexanone for 6 hours/day, 5 days/week for up to 10 months
was used as the principal study in the derivation of the RfC (Johnson et al., 1977). MCV of the
sciatic-tibial nerve in monkeys was selected as the critical effect for the derivation of the RfC.
As discussed in Section 4, human and animal data indicate that neurological effects are a
characteristic and sensitive endpoint of inhalation exposure to 2-hexanone. Neuropathy has been
observed in humans following inadvertent occupational exposure (Allen et al., 1975; Billmaier et
al., 1974; Gilchrist et al., 1974) and has been demonstrated repeatedly in laboratory animals
(Egan et al., 1980; Katz et al., 1980; O'Donoghue and Krasavage, 1979; Johnson et al., 1979,
1977; Duckett et al., 1979, 1974; O'Donoghue et al., 1978; Krasavage and O'Donoghue, 1977;
Spencer et al., 1975; Mendell et al., 1974).
Several studies of workers in a coated fabrics plant (Allen et al., 1975; Billmaier et al.,
1974; Gilchrist et al., 1974) provide evidence in humans of a concentration-dependent neurotoxic
response to 2-hexanone exposure. Although personal air samples were not collected in these
studies, the available measures of exposure were sufficient to produce quantitative estimates of
2-hexanone inhalation exposure for two groups of workers (i.e., print operators and print
helpers), both of whom exhibited peripheral neuropathy. In these workers, exposure to
2-hexanone also occurred via oral and dermal routes, since the study authors noted that
individuals frequently ate at the work site and were accustomed to washing their hands with
2-hexanone. Workers were coexposed to MEK, which can potentiate the toxicity of 2-hexanone.
Because the magnitude of exposure to 2-hexanone from oral and dermal exposure routes was not
quantified by the study authors and because of co-exposure to MEK, this study was not
considered for use in RfC derivation.
Of the available animal studies of 2-hexanone, the subchronic studies by Abdo et al.
(1982), Duckett et al. (1979, 1974), Krasavage and O'Donoghue (1977), Saida et al. (1976), and
Mendell et al. (1974) and the chronic study by O'Donoghue and Krasavage (1979) were not
selected for use in deriving the RfC. Duckett et al. (1979, 1974) did not report the sex of the
65
-------
animals or the purity of 2-hexanone used. Furthermore, the authors used only one exposure
concentration per series of experiments. Krasavage and O'Donoghue (1977) utilized two
exposure concentrations (100 and 330 ppm); however, the purity of 2-hexanone was not stated
and limited data were provided. As mentioned previously, MiBK, a potential inducer of
CYP450, is a common contaminant in the formulations of 2-hexanone. Without more
information on the purity of the 2-hexanone administered, it is difficult to ascertain if MiBK
impacted the toxicity of 2-hexanone in Krasavage and O'Donoghue (1977) and Abdo et al.
(1982). Saida et al. (1976) used two exposure concentrations (225 and 400 ppm) but did not
indicate the sex of the animals or the purity of 2-hexanone used. Additionally, the purity of
2-hexanone in the study by Mendell et al. (1974) was not stated, and limited data were provided.
The study by Johnson et al. (1977) was performed in monkeys and rats with 8 and 10
animals per dose group, respectively. Two concentrations of commercial grade 2-hexanone were
employed (100 and 1,000 ppm in air) with exposures occurring 6 hours/day, 5 days/week for 10
months. Concurrent control groups were used in both species. As part of this study, Johnson et
al. (1977) conducted four neurological tests in each species (usually once per month) to identify
effects in treated versus control animals. These four tests were MCV of the right sciatic-tibial
nerve, MCV of the right ulnar nerve, absolute refractory period of these two nerves, and MAPs
in response to both sciatic and ulnar nerve stimulation.
The animal studies by Katz et al. (1980) and Egan et al. (1980) used exposure to
2-hexanone (purity >96%) at a single concentration (2,870 and 410 mg/m3, respectively) for a
period of 6 months or less, using only one strain and sex of rats. Both Katz et al. (1980) and
Egan et al. (1980) utilized clinical chemistry and histopathological changes to identify treatment-
related effects of 2-hexanone. Both studies, although limited in duration and study design,
reported neurological effects, including neuropathy consisting of difficulty extending hind limbs
and axonal swelling of the peripheral nerve, and support the findings from Johnson et al. (1977)
Despite the use of commercial grade 2-hexanone, the study by Johnson et al. (1977) was
chosen as the principal study on which to base the RfC because the authors used two different
animal species, including nonhuman primates, and two 2-hexanone exposure concentrations,
while also employing larger treatment groups and longer exposure durations than either Katz et
al. (1980) or Egan et al. (1980). Although duration of the unpublished study by Krasavage and
O'Donoghue (1977) was longer than the study by Johnson et al. (1977), the latter utilized
monkeys, a biologically more relevant species than rats, when assessing inhalation exposure.
As previously discussed, the effects seen in humans and experimental animals following
exposure to 2-hexanone via inhalation provide evidence that the nervous system is the primary
target of 2-hexanone toxicity. Data from Johnson et al. (1977) on both sciatic-tibial and ulnar
nerve MCVs in 2-hexanone-exposed monkeys and rats were considered for use in deriving the
RfC. Studies in humans have provided insight into the relationship between decreased MCV and
functional effects in humans. Sobue et al. (1978) observed a reduction in MCV among workers
66
-------
with severe polyneuropathy in a cross-sectional study of 1,662 shoe workers that were exposed
to n-hexane, a parent compound of 2-hexanone. Passero et al. (1983) also noted an association
between slowing MCV and disease severity among 98 polyneuropathy cases in a cohort of
workers exposed to n-hexane.
Johnson et al. (1977) reported neuropathy characterized by decrements in sciatic-tibial
nerve and ulnar nerve MCVs following administration of 2-hexanone. Both monkeys and rats
exposed to 100 ppm 2-hexanone exhibited statistically significant reductions in sciatic-tibial
nerve MCVs at 9 and 7 months of exposure, respectively. Similarly, MCVs were reduced in the
ulnar nerves of both monkeys and rats. Monkeys in the low-exposure group exhibited
statistically significant decreases in ulnar nerve MCVs relative to control values at 1 and
3 months. Although ulnar nerve MCVs were reduced relative to controls throughout the
remainder of the study, these reductions were not statistically significant. Rats exhibited
statistically significant decreases in ulnar nerve MCVs at 4 and 7 months exposure to 100 ppm
2-hexanone. Because monkeys have a similar respiratory tract and breathing patterns to humans
and it is known that 2,5-hexanedione (the primary metabolite of 2-hexanone) typically affects
long axons such as the sciatic-tibial nerve prior to other nerves, the sciatic-tibial nerve MCV in
monkeys was identified as the critical effect to derive the RfC. For comparison purposes,
sciatic-tibial nerve MCV in rats and ulnar nerve MCV in both monkeys and rats were considered
potential critical effects for RfC derivation.
5.2.2. Methods of Analysis: Benchmark Concentration Modeling
Table 5-3 displays monthly mean MCV values (in m/second) for both the sciatic-tibial
and ulnar nerves of monkeys exposed to three different concentrations of 2-hexanone in air (i.e.,
0, 100, or 1,000 ppm) for durations ranging from 1 to 10 months. These data were extracted (via
digitization2) from Figure 1 (for the sciatic-tibial nerve) and Figure 3 (for the ulnar nerve) of
Johnson et al. (1977). Similarly, Table 5-4 displays monthly mean MCV values (in m/second)
for both the sciatic-tibial and ulnar nerves of rats exposed to three different concentrations of
2-hexanone in air (i.e., 0, 100, or 1,000 ppm) for durations ranging from 2 to 29 weeks. These
data were extracted (via digitization) from Figure 2 (for the sciatic-tibial nerve) and Figure 4 (for
2 Values from Johnson et al. (1977) were digitized by using the line tool on Microsoft Office Word 2003, followed
by measuring the values with the distance tool function on Adobe® Acrobat® 6.0 Professional (version 6.0.0,
5/19/2003). To accomplish this task, the figures from Johnson et al. (1977) were inserted into a Word document by
using the snapshot tool from Adobe® Acrobat® 6.0 Professional. Then, horizontal lines were applied over the data
points, the measurement markers on the y-axis, and extended through the y-axis. Lines from the data points to the
x-coordinates were not traced over, since Johnson et al. (1977) provided the absolute values in the text. Once all of
the lines were traced from the data points through the y-coordinates, a vertical line was traced over the y-axis. Then
the Word document was saved in portable document format (pdf) and opened using Adobe® Acrobat® 6.0
Professional. The y-axis was viewed at 300% magnification, and the distance tool was used to measure from the
origin to each y-coordinate for each horizontal line, including data points and measurement markers. The distance
tool allows measurements to be made down to one hundredth of a millimeter, and repeated measures placed the
reproducibility of this technique at greater than 99%.
67
-------
the ulnar nerve) of Johnson et al. (1977). Note that after approximately 6 months of exposure,
monkeys and rats in the 1,000 ppm exposure group were removed from the study because
neuropathy (characterized as hind-limb drag) had developed in these animals.
Table 5-3. Effect of 2-hexanone inhalation exposure on the MCV of the
sciatic-tibial and ulnar nerves in monkeys (n = 8/group)
Exposure
duration
(months)
1
2
3
4
5
6
7
8
9
10
2-Hexanone
concentration
(ppm in air)
0
100
1,000
0
100
1,000
0
100
1,000
0
100
1,000
0
100
1,000
0
100
1,000
0
100
0
100
0
100
0
100
Mean MCV:
sciatic-tibial nerve (m/s)a
(% change from control)
42
42 (0%)
40 (5%)
51
46 (10%)
44 (14%)
54
48(11%)
46 (15%)
56
50(11%)
41C (27%)
53
48 (9%)
36C (32%)
50
47 (6%)
33C (34%)
51
48 (6%)
50
46 (8%)
53
49C (8%)
53
48C (9%)
Mean MCV:
ulnar nerve (m/s)b
(% change from control)
54
46C (15%)
47C (13%)
61
63 (3%)
49C (20%)
53
47C(11%)
45C (15%)
63
58 (8%)
49C (22%)
61
63 (3%)
43C (30%)
58
56 (3%)
41C(29%)
65
62 (5%)
58
58 (0%)
63
60 (5%)
58
57 (2%)
aValues extracted from Figure 1 in Johnson et al. (1977).
Values extracted from Figure 3 in Johnson et al. (1977).
Statistically significantly different compared with corresponding controls (p < 0.05), as determined by
the study authors. For months 1-6, analysis of variance was used to test statistical difference across the
three groups, while for months 7-10, Student's t-test was used to compare difference between controls
and the 100 ppm group.
68
-------
Table 5-4. Effect of 2-hexanone inhalation exposure on the MCV of the
sciatic-tibial and ulnar nerves in rats
Exposure
duration
(weeks)
13
17
25
29
2-Hexanone
concentration
(ppm in air)
0
100
1,000
0
100
1,000
0
100
1,000
0
100
Mean MCV:
sciatic-tibial nerve
(m/s)a
34
37
40C
_
-
-
42
41
27C
39
25C
Mean MCV:
ulnar nerve
(m/s)b
_
-
-
42
36C
38C
40
37
31C
45
30C
"Values extracted from Figure 2 in Johnson et al. (1977).
bValues extracted from Figure 4 in Johnson et al. (1977).
Statistically significantly different compared with corresponding controls (p < 0.05), as
determined by Johnson et al. (1977).
The nerve MCV data in Tables 5-3 and 5-4 were subjected to BMD modeling, employing
the available continuous models in EPA's BMDS, version 2.0 (i.e., linear, polynomial, power,
and Hill models). As shown in Table 5-3, differences from the control in mean sciatic-tibial
nerve MCV among monkeys ranged from 6 to 11% at 100 ppm, with statistically significant
changes of 8 and 9% at 9 and 10 months, respectively. Because the study authors used different
statistical tests depending on the number of exposure groups during the course of the study,
similar reductions in nerve MCVs varied in statistical significance. For example, a decrement of
9% in sciatic-tibial nerve MCV at 10 months, an exposure interval with two treatment groups,
achieved statistical significance, whereas a decrement of 9% at 5 months, an exposure interval
with three treatment groups, was not statistically significant (see Table 5-3).
Statistically significant decreases in nerve conduction velocity are indicative of a
neurotoxic effect; however, as noted in EPA's Guidelines for Neurotoxicity Risk Assessment
(U.S. EPA, 1998), normal conduction velocity may be maintained for some time after the onset
of axonal degeneration. Therefore, EPA determined that small changes in mean sciatic-tibial
nerve MCV are biologically significant. A BMR of 5% extra risk was selected based on the
following considerations: (1) this effect level is considered to be a minimal biologically
significant change; (2) the potential for nerve fiber damage (i.e., axonal degeneration) with little
to no change in MCV; and (3) the BMDLos falls within the low end of the range of the
observable data.
Data were available for three exposure groups at 6 months compared with two exposure
groups at 10 months. The magnitude of variation in MCV between the 6-month data and the 10-
69
-------
month data is similar and is supported by human studies (Metso et al., 2008; Yap and Hirota,
1967) that have reported comparable levels of motor conduction nerve velocity to those observed
by Johnson et al. (1977). Metso et al. (2008) found that motor nerve conduction velocity among
74 adults without signs or symptoms of neuropathy ranged from 47.2 to 63.4 m/second (mean:
56.3 m/second; standard deviation = 4.13) and could be influenced by age, height, and external
body temperature. Yap and Hirota (1967) noted a range of 45.3-61.1 m/second for sciatic nerve
conduction velocity among 19 individuals. Johnson et al. (1977) also noted variation in MCVs
among the control animals. Considering that the magnitude of variation in nerve MCVs between
the 6-month data (6%) and the 10-month data (9%) was similar and more treatment groups were
available for the 6-month duration of exposure, the data at 6 months were used for BMD
modeling.
A difficulty encountered in conducting a BMD analysis on these data was that no
information was provided regarding the statistical variances or confidence limits for the mean
nerve conduction velocities shown in Figures 1 through 4 in Johnson et al. (1977), nor were any
of the raw data on which these means were based presented in the paper. In BMDS, estimates of
the standard deviation of the response in each dose group are needed to calculate BMDs and their
corresponding BMDLs. Therefore, an indirect method for estimating this missing information
on response variability was devised.
Information regarding the variability in MCV measurements in Johnson et al. (1977) can
be derived from the results of statistical tests that are reported in the paper. In this study, two
different statistical procedures were employed. Analysis of variance (ANOVA) was used to test
for statistically significant differences in mean MCVs at specific test periods (usually monthly)
whenever data across the three exposure groups (i.e., 0, 100, or 1,000 ppm) were compared.
After approximately 6 months on study, however, animals (both monkeys and rats) in the highest
exposure group (1,000 ppm) were removed from further 2-hexanone exposure. Consequently,
with termination of this 1,000 ppm exposure group, only two exposure groups remained for each
species. Thus, Student's t-test was used to test for statistically significant changes in mean
MCVs across these two groups (i.e., 0 and 100 ppm) for the remaining test periods.
In ANOVA, an F statistic is used to test for a significant difference among the means of g
groups. An F statistic is defined as F(g-\, N-g) = between-group variance/within-group
variance, where g-l represents the numerator degrees of freedom and N-g represents the
denominator degrees of freedom (g is the number of groups and TV is the sample size within each
group). In the specific case where only two group means are being compared, the F statistic
reduces to a t statistic (i.e., F(\, N-g) = t(N-g)2), where t has a Student's t distribution. In order to
fit a continuous dose-response model in BMDS, an estimate of the within-group variance or s2 is
needed from which the estimated standard deviation can be obtained simply by taking the square
root of this variance estimate.
70
-------
The estimated within-group variance can be derived by using the following procedure. If
the within-group means and the numbers of observations on which each of these means is based
are known, the between-group variance can be calculated. Once the between-group variance has
been determined and the corresponding value of the F or t statistic is known, an estimate of the
within-group variance or s2 can be derived from the following equation: s2 = (between-group
variance)/F(g - 1, N - g) or s2 = (between-group variance)/?(A/-g)2. In Johnson et al. (1977), for
monkeys, F statistics were reported for mean MCVs at both 4 and 6 months, while t statistics
were reported for mean MCVs at both 9 and 10 months. These data yielded four estimates of the
within-group variance or standard deviation. The arithmetic average of these four estimates was
then used in BMD modeling as the estimated standard deviation for MCVs in each exposure
group, assuming a constant variance across dose groups. For rats, F statistics were reported in
Johnson et al. (1977) for mean MCVs at both 13 and 17 weeks, while a t statistic was reported
for mean MCVs at 29 weeks. These data yielded three estimates of the within-group variance or
standard deviation. The arithmetic average of these three estimates was then used in BMD
modeling as the estimated standard deviation for MCVs in each exposure group, assuming a
constant variance across dose groups.
The best-fit model from BMDS was selected by examining the results of the chi-squared
goodness-of-fit test and comparing the magnitudes of the Akaike Information Criterion (AIC).
All models with chi-squaredp values >0.1 were considered to exhibit an adequate fit to the data.
Of the models exhibiting adequate fit, the model with the lowest AIC (i.e., a measure of the
deviance of the model fit that allows for comparison across models for a particular endpoint) was
selected as the best-fit model. These criteria for model selection are consistent with those
described in the Benchmark Dose Technical Guidance Document (U.S. EPA, 2000b). For the
MCV data in both monkeys and rats, the lst-degree polynomial model provided the best fit for
both sciatic-tibial and ulnar nerve MCVs.
The 95% lower confidence limits on the BMC estimates (BMCLs) derived from the best-
fit models for sciatic-tibial and ulnar nerve MCV values in monkeys and rats are presented in
Table 5-5. Detailed BMDS outputs from the BMD of the monkey and rat MCV data are
contained in Appendix B-2.
5.2.3. Exposure Duration Adjustments and Conversion to Human Equivalent
Concentrations
Because the RfC is a metric that assumes continuous human exposure for a lifetime,
adjustments need to be made to animal (or human) data obtained from intermittent and/or less-
than-lifetime exposures, as outlined in the Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994b). The first step in
this process is adjusting intermittent inhalation exposures to continuous inhalation exposures,
based on the assumption that the product of exposure concentration and time is constant (U.S.
71
-------
EPA, 2002). In Johnson et al (1977), animals were exposed to 2-hexanone for 6 hours/day,
5 days/week. Therefore, the BMCLADj, reflecting continuous inhalation exposure to 2-hexanone,
is derived as follows:
BMCLAoj = BMCL x hours exposed per day724 hours x days exposed per week/7 days
BMCLAoj = 121 x 6/24 x 5/7 = 22 ppm, based on monkey sciatic-tibial nerve MCV
= 139 x 6/24 x 5/7 = 25 ppm, based on monkey ulnar nerve MCV
= 116 x 6/24 x 5/7 = 21 ppm, based on rat sciatic-tibial nerve MCV
= 176 x 6/24 x 5/7 = 31 ppm, based on rat ulnar MCV
Furthermore, because RfCs are typically expressed in units of mg/m3, the above ppm
values need to be converted to mg/m3 by using the conversion factor specific to 2-hexanone of
1 ppm = 4.1 mg/m3. Thus, the final BMCL^j values are as follows:
BMCLADJ = 22 x 4.1 = 90.2 mg/m3, monkey sciatic-tibial nerve MCV
= 25 x 4.1 = 102.5 mg/m3, based on monkey ulnar nerve MCV
= 21 x 4.1 = 86.1 mg/m3, based on rat sciatic-tibial nerve MCV
= 31 x 4.1 = 127.1 mg/m3, based on rat ulnar nerve MCV
Finally, this BMCLAoj value must be converted to a human equivalent concentration
(HEC). The HEC that elicits decreased MCV, which is not a respiratory (or portal-of-entry)
effect but a systemic effect, is derived based on the following. For systemic effects, 2-hexanone
is classified as a category 3 gas under EPA guidelines (U.S. EPA, 1994b). According to this
guidance, in order to convert the concentration effective in animals to human equivalents, a
multiplicative factor based on the ratio of blood:gas partition coefficients is employed as follows:
BMCLHEC = BMCLADJ x [(Hb/g)A/(Hb/g)H]
where
(Hb/g)A = blood:gas partition coefficient for 2-hexanone in animals
(Hb/g)n = blood:gas partition coefficient for 2-hexanone in humans
The blood:gas partition coefficient (Hb/g)n for 2-hexanone in humans is 127 (Sato and
Nakajima, 1979); however, no value has been reported for monkeys or rats. In the absence of a
measured blood:gas partition coefficient in the test species, the ratio [(Hb/g)A/(Hb/g)H] defaults to
one, in which the BMCLnEc = BMCLAoj. These values are presented in the last column of
Table 5-5.
72
-------
Table 5-5. Summary of BMCLs and HECs for 2-hexanone
Study
reference
Johnson et al.
(1977)
Johnson et al.
(1977)
Study duration
and type
10-month
inhalation
29-week
inhalation
2-Hexanone
exposure
(ppm)
0, 100, 1,000
0, 100, 1,000
Species/sex
Male monkeys
(n = 8 per dose
group)
Male rats
(n - 10 per
dose group)
Toxicological
endpoint
Sciatic -tibial nerve
MCV
(at 6 months)
Ulnar nerve MCV
(at 6 months)
Sciatic -tibial nerve
MCV
(at 25 weeks)
Ulnar motor nerve
conduction velocity
(at 25 weeks)
BMDS
best-fit
continuous
model
1st degree
polynomial
1st degree
polynomial
1st degree
polynomial
1st degree
polynomial
BMC05
(ppm)
147
167
135
235
BMCL05
or PODa
(ppm)
121
139
116
176
Adjusted
BMCL05
(BMCL05(ADJ))b
(mg/m )
90
102
86
127
BMCLos (HEC)
(mg/m3)
90
102
86
127
"BMCLs or PODs were estimated at a BMR of 0.05 or 5% relative change from controls.
bConversion factors and assumptions: molecular weight (2-hexanone) = 100.16 and 1 ppm = 100.16/24.45 = 4.1 mg/m3 (at 25°C and 760 mm Hg). Duration
adjustment of exposure concentrations and conversion to mg/m3 was accomplished as follows: BMCL05(ADj) =121 ppm x 6 hours/24 hours x 5 days/7days =
22 ppm x 4.1 mg/m3-ppm = 90 mg/m3.
°The BMCL05 (HEC) was calculated for an extrarespiratory effect of a category 3 gas. The blood:gas partition coefficient (Hb/g) value for 2-hexanone in humans is
127 (Sato and Nakajima, 1979); however, no value has been reported for monkeys or rats. According to EPA's RfC methodology (U.S. EPA, 1994b), when the
ratio of animal to human blood:gas partition coefficients [(Hb/g)A/(Hb/g)H] is greater than one or the values are unknown, a value of one is used for the ratio by
default. Thus, BMCL05(HEC) = 90 x [(Hb/g)A/(Hb/g)H] = 90 mg/m3.
73
-------
5.2.4. Calculation of the RfC: Application of Uncertainty Factors
As mentioned previously, the primary metabolite of 2-hexanone, i.e., 2,5-hexanedione,
typically affects long axons, such as the sciatic-tibial nerve, prior to affecting other nerves; thus,
the sciatic-tibial nerve MCV is used to derive the RfC. Additionally, since monkeys have a
similar respiratory tract and breathing patterns to humans, the BMCLos (HEC) based on sciatic-
tibial nerve MCV in monkeys (Table 5-5) is used to derive the RfC. It should be noted that ulnar
nerve MCV in monkeys and sciatic-tibial nerve MCV in rats were found to have similar
BMCLos(HEC) estimates as the endpoint selected above.
The RfC for 2-hexanone based on decreased sciatic-tibial nerve MCVs in monkeys as the
critical effect is derived from the BMCLos (HEC) by application of UFs as follows:
RfC = BMCLHEC - UF
RfC = 90 ^ 3,000 = 0.03 mg/m3 = 3 x 10~2 mg/m3
This composite UF of 3,000 is composed of the following:
• A default intraspecies UF (UFn) of 10 was applied to adjust for potentially sensitive
human subpopulations (intraspecies variability). A 10-fold UF is warranted because
insufficient information is currently available to assessment human-to-human variability
in 2-hexanone toxicokinetics or toxicodynamics.
• A default subchronic-to-chronic UF (UFS) of 10 was applied to account for use of data
following 6 months of exposure to 2-hexanone for the derivation of an RfC.
• An UF of 3 was applied to account for uncertainties in extrapolating from monkeys to
humans (UFA). This value is adopted by convention where an adjustment from an
animal-specific BMCLAoj to a BMCLHEc has been incorporated. Application of an UF of
10 would depend on two areas of uncertainty (i.e., toxicokinetic and toxicodynamic). In
this assessment, the toxicokinetic component is mostly addressed by the determination of
an HEC as described in the RfC methodology (U.S. EPA, 1994b). The toxicodynamic
uncertainty is also accounted for to a certain degree by the use of the applied dosimetry
method, and a UF of 3 is retained to address this component.
• An UF of 10 was applied to account for database deficiencies (UFD). The database
includes a human occupational exposure study (with co-exposure to MEK), subchronic
animal studies in rats and hens, and a chronic study in cats. One postnatal development
and behavior study (Peters et al., 1981) on 2-hexanone in F344 rats exists, identifying a
LOAEL of 1,000 ppm (no NOAEL reported). The database does not include a
multigenerational reproductive study or developmental studies. The database also lacks
information regarding axonal degeneration at concentrations similar to those inducing
74
-------
minimal reductions in nerve MCV. Additionally, Katz et al. (1980) observed a reduction
in total white blood cell counts to 60% of control values in rats exposed to 2-hexanone in
a subchronic inhalation study, suggesting that further study of immunotoxicity may be
warranted. Because of the absence of a two-generation reproductive study and studies
evaluating the developmental toxicity and possible immunotoxicity of 2-hexanone
following exposure via inhalation, an UFo of 10 is warranted.
• An UF for LOAEL-to-NOAEL extrapolation (UPi,) was not used because the current
approach is to address this factor as one of the considerations in selecting a BMR for
BMD modeling. In this case, a BMR of a 5% change in nerve conduction velocity from
the control mean was selected under an assumption that it represents a minimal
biologically significant change.
5.2.5. RfC Comparison Information
Of the chronic and subchronic studies available on inhalation exposure to 2-hexanone,
Johnson et al. (1977) was selected to serve as the principal study to derive an RfC. The
endpoints considered from Johnson et al. (1977) include MCV for both sciatic-tibial and ulnar
nerves of both rats and monkeys. The potential PODs based on the best-fit models from BMD
models, with the corresponding applied UFs and derived sample RfDs from Table 5-4, are
presented in Figure 5-2.
75
-------
1000 i
100 -
* Point of departure
f LTF, animal to human
0 LTF, human variability
[T] LTF, database
Reference concentration
| LTF, subchronic to chronic
I LTF, LOAEL to NOAEL
o
Q
10
1 -
0.1
001
m
W
t Y
^
t
1
1
m
Sciatic-tibial Ulnar MCV in Sciatic-tibial MCV Ulnar MCV in
MCV in monkeys monkeys in rats rats
Figure 5-2. Potential PODs for endpoints from Johnson et al. (1977), with
corresponding applied UFs and derived sample inhalation reference values.
5.2.6. Previous Inhalation Assessment
An RfC assessment for 2-hexanone was not previously available on IRIS.
5.3. CANCER ASSESSMENT
As discussed in Section 4.7.1, the available database for 2-hexanone contains inadequate
information to assess the carcinogenic potential according to Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005a). A cancer assessment for 2-hexanone was not previously
available on IRIS.
76
-------
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD
AND DOSE RESPONSE
6.1. HUMAN HAZARD POTENTIAL
2-Hexanone (methyl butyl ketone, CASRN 591-78-6) has the chemical formula
and a molecular weight of 100.16. It is a clear, volatile, flammable fluid with a pungent,
acetone-like odor. 2-Hexanone is most commonly used as a paint or printing ink thinner, as a
solvent for oils, waxes, and resins, or as a cleaning agent. It is currently not produced
commercially in the U.S., and no information on importation is available (ATSDR, 1992).
2-Hexanone is currently found at Superfund sites.
2-Hexanone is well absorbed by the inhalation route and in the gastrointestinal tract.
Studies in human volunteers and in experimental animals (dog and rabbit) established that 2-
hexanone is dermally absorbed. The distribution of 2-hexanone has not been studied thoroughly.
In a rat study, it appeared in the plasma and the lung at higher concentrations than in the liver
after both oral and inhalation administration (Duguay and Plaa, 1995) but did not show an
affinity for a lipid-rich tissue such as the brain (Granvil et al., 1994). In guinea pigs, 2-hexanone
was eliminated quite rapidly, with a half-life of a little more than 1 hour for the parent compound
and values not exceeding 21/2 hours for the major metabolites (DiVincenzo et al., 1976). In rats,
on the other hand, 2-hexanone was eliminated more slowly (Bus et al., 1981). The biological
half-life of 2-hexanone in humans is not known. A PBTK model has not been published.
Metabolites of 2-hexanone include 2-hexanol, 2,5-hexanediol, 5-hydroxy-2-hexanone,
2,5-hexanedione, and some cyclic furan derivatives. The enzymes that metabolize 2-hexanone
have not been well characterized. Among the metabolites of 2-hexanone, 2,5-hexanedione is the
most important because it is a well-known neurotoxicant. It causes a neuropathy, specifically of
the peripheral giant axons, that involves neurofilament cross-linking and axonal swelling and
proceeds to retrograde axonal degeneration. 2-Hexanone-induced neuropathy has been observed
clinically in occupationally-exposed humans (Davenport et al., 1976; Mallov, 1976; Allen et al.,
1974; Billmaier et al., 1974), but the findings are frequently obscured by co-exposure to other
solvents, most frequently MEK, which is known to potentiate the toxicity of 2-hexanone.
A significant number of studies have been conducted in which laboratory animals were
exposed orally or via inhalation for up to 2 years. Oral exposure studies used rats (Krasavage et
al., 1980) and guinea pigs (Abdel-Rahman et al., 1978), with doses ranging up to 600 mg/kg-day
by gavage or up to 1.3% in drinking water (amounting to 1,010 mg/kg-day). The 13-month
study in rats by O'Donoghue et al. (1978) that gave a detailed report of neuropathy incidences
was used in the RfD assessment for 2-hexanone. Inhalation studies employed rats (Egan et al.,
1980; Katz et al., 1980; Duckett et al., 1979, 1974; Johnson et al., 1977; Krasavage and
O'Donoghue, 1977; Saida et al., 1976; Spencer et al., 1975), cats (O'Donoghue and Krasavage,
77
-------
1979), and monkeys (Johnson et al., 1977), with exposures ranging from 10-1,300 ppm (41-
5,325 mg/m3). The effect observed in all of these studies was neuropathy. The 2-hexanone-
induced neuropathy also has been characterized mechanistically in animal studies (DeCaprio et
al., 1988, 1982). Additionally, a study in beagles that received 2-hexanone via the subcutaneous
route reported neuropathy (O'Donoghue and Krasavage, 1981).
It is not known whether 2-hexanone causes any other significant illness in humans. The
available animal studies do not provide information to assess the carcinogenicity of 2-hexanone.
According to the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), there is
"inadequate information to assess the carcinogenic potential" of 2-hexanone.
6.2. DOSE RESPONSE
6.2.1. Noncancer/Oral
There are no human studies of oral exposure to 2-hexanone in the literature that provide
sufficient information to support quantitative dose-response assessment. There is no standard
2-year bioassay for 2-hexanone in any animal species. A 13-month study in rats by O'Donoghue
et al. (1978) reported critical, chemical-related effects and was selected as the principal study.
Because rats were exposed for more than half of their life span, the study duration was
considered to be chronic. The dose response for the critical effect, peripheral nerve axonopathy
(axonal swelling), was not well characterized, with incidences jumping from 0% (0/10 animals)
in controls to 80% (8/10 animals) at the lowest dose (143 mg/kg-day) and 100% (10/10 animals)
at both higher doses (266 and 560 mg/kg-day). However, the dose-dependent development of
axonopathy was confirmed in related endpoints (brain axonopathy, myofibrillar atrophy).
Peripheral nerve axonopathy was chosen as the critical effect because it displayed the most
sensitive response to 2-hexanone exposure. Spinal cord axonopathy displayed a similar dose
response, but this endpoint was examined in only half of the exposed animals.
The RfD of 5 x 10~3 mg/kg-day was derived from axonal neuropathy in male
COBS/CD(SD)BR rats following 13 months of oral exposure to 2-hexanone (O'Donoghue et al.,
1978). There is evidence from other studies in experimental animals to confirm that the nervous
system is the primary target for the toxicological effects of 2-hexanone (Abdo et al., 1982;
Krasavage et al., 1980; Eben et al., 1979; Abdel-Rahman et al., 1978; Homan et al., 1977).
Treatment-related changes observed in the study by O'Donoghue et al. (1978) included clinical
neurological deficits and histological changes indicative of peripheral neuropathy. Figure 5-1
provides a graphical comparison of derived sample RfDs based on alternative neuropathological
endpoints from O'Donoghue et al. (1978).
A composite UF of 1,000 was applied: 10 for intraspecies (interindividual) variability,
10 for interspecies variability, and 10 for database deficiencies. Information was unavailable to
quantitatively assess toxicokinetic or toxicodynamic differences between animals and humans
and the potential variability in human susceptibility; thus, the interspecies and intraspecies UFs
78
-------
of 10 were applied. A 10-fold database deficiency UF was applied to reflect that, although
chronic and subchronic information on 2-hexanone was available, there are no 2-hexanone-
specific multigenerational reproductive or developmental studies. Developmental studies on
n-hexane, a precursor of 2-hexanone and 2,5-hexanedione, have shown low risk of toxicity;
however, there remains a level of concern because available studies on 2-hexanone via inhalation
exposure have suggested the possibility of reproductive toxicity and immunotoxicity.
The overall confidence in this RfD assessment is medium. Confidence in the principal
study (O'Donoghue et al., 1978) is medium. The study used 10 animals per group and reported
clinical neurological deficits and neuropathological effects within a dose range in which a
LOAEL could be identified for the critical effect. Animal studies in two additional species
(guinea pigs and hens) lend support to the choice of neurological effects as an endpoint of
concern. Confidence in the database is low to medium. The database lacks information on
developmental, reproductive, and immune system toxicity. Reflecting medium confidence in the
principal study and low to medium confidence in the database, confidence in the RfD is medium.
6.2.2. Noncancer/Inhalation
There are no human studies of inhalation exposure to 2-hexanone in the literature that
provide sufficient information to support quantitative dose-response assessment.
Dose-dependent development of 2-hexanone-induced neuropathy was confirmed in
numerous subchronic studies in rats (Egan et al., 1980; Katz et al., 1980; Duckett et al., 1979,
1974; Johnson et al., 1977; Krasavage and O'Donoghue, 1977; Saida et al., 1976; Spencer et al.,
1975) and one chronic study in cats (O'Donoghue and Krasavage, 1979). One 10-month study
(Johnson et al., 1977), used two different species (monkeys and rats; n = 8 and n = 10 per group,
respectively) with two concentrations of 2-hexanone (commercial grade). Johnson et al. (1977)
utilized four sensitive neurological tests to identify subtle changes in treated versus control
animals. The study by Johnson et al. (1977) was chosen as the principal study for RfC
development. Both sciatic-tibial MCV and ulnar MCV in 2-hexanone-exposed monkeys and rats
were considered in deriving the RfC. A graphical comparison of derived sample RfCs from
these endpoints is illustrated in Figure 5-2. Because monkeys have similar respiratory tracts and
breathing patterns to humans and it is known that 2,5-hexanedione, the primary metabolite of
2-hexanone, typically affects long axons such as the sciatic-tibial nerve prior to other nerves,
sciatic-tibial nerve MCV in monkeys was identified as the critical effect to derive the RfC. It
should be noted that ulnar nerve MCV in monkeys and sciatic-tibial nerve MCV in rats were
found to have similar BMCL0s (HEC) estimates.
The RfC of 3 x 10~2 mg/m3 was derived from the decrease in sciatic-tibial MCV in
monkeys exposed to 2-hexanone for 6 months (Johnson et al., 1977). A composite UF of 3,000
was applied in the derivation of the RfC: 10 for intraspecies (interindividual) variability, 10 for
subchronic-to-chronic uncertainty, 10 for database uncertainty, and 3 for interspecies variability.
79
-------
Information was unavailable to predict potential variability in susceptibility among the
population; thus, the intraspecies variability UF of 10 was applied. A subchronic-to-chronic UF
of 10 was applied to account for the use of data obtained after 6 months of exposure in
calculating the RfC. An UF of 10 was applied to account for database deficiencies. The
database included a human occupational exposure study (with co-exposure to MEK), subchronic
animal studies in rats and hens, and a chronic study in cats. One postnatal development and
behavior study on 2-hexanone was identified (Peters et al., 1981) that yielded a LOAEL of 1,000
ppm (no NOAEL reported). The database does not include a multigenerational reproductive
study or developmental studies. The database also lacks information regarding axonal
degeneration at concentrations similar to those inducing minimal reductions in nerve MCV.
Additionally, Katz et al. (1980) observed a reduction in total white blood cell counts to 60% of
control values in rats exposed to 2-hexanone in a subchronic inhalation study, suggesting that
further study of immunotoxicity may be warranted. Because of the absence of a two-generation
reproductive study and studies evaluating developmental toxicity and possible immunotoxicity of
2-hexanone following exposure via inhalation, a UFD of 10 is warranted. An interspecies UF of
3 (rather than 10) was applied to address toxicodynamic uncertainty, and the toxicokinetic
uncertainty was addressed by the determination of the HEC.
The overall confidence in this RfC assessment is low. Confidence in the principal study
is medium. The study included exposures in two species via the inhalation route and sensitive
diagnostic tests for determining treatment-related neurotoxicity. In addition, animal studies in
four different species (monkeys, rats, cats, and hens) and occupational exposures lend support for
the choice of neurologic effects as an endpoint of concern. Confidence in the database is low.
The database lacks a multigenerational developmental and reproductive toxicity studies. In
addition, the observation of a reduction in total white blood cell count suggests the need for
further information on immunotoxicity.
6.2.3. Cancer
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), there is
inadequate information to assess the carcinogenic potential of 2-hexanone. As such, data are
unavailable to calculate cancer risk estimates.
80
-------
7. REFERENCES
Abdel-Rahman, MS; Hetland, LB; Couri, D. (1976) Toxicity and metabolism of methyl n-butyl ketone. Am Ind
HygAssocJ37(2):95-102.
Abdel-Rahman, MS; Saladin, JJ; Bohman, CE; et al. (1978) The effect of 2-hexanone and 2-hexanone metabolites
on pupillomotor activity and growth. Am Ind Hyg Assoc J 39(2):94-99.
Abdo, KM; Graham, DG; Timmons, PR; et al. (1982) Neurotoxicity of continuous (90 days) inhalation of
technical grade methyl butyl ketone in hens. J Toxicol Environ Health 9(2): 199-215.
Abou-Donia, MB; Makkawy, HA; Graham, DG. (1982) The relative neurotoxicities of n-hexane, methyl n-butyl
ketone, 2,5-hexanediol, and 2,5-hexanedione following oral or intraperitoneal administration in hens. Toxicol
Appl Pharmacol 62(3):369-389.
Abou-Donia, MB; Makkawy, HM; Campbell, GM. (1985a) Pattern of neurotoxicity of n-hexane, methyl n-butyl
ketone, 2,5-hexanediol, and 2,5-hexanedione alone and in combination with O-ethyl O-4-nitrophenyl
phenylphosphonothioate in hens. J Toxicol Environ Health 16(1):85-100.
Abou-Donia, MB; Lapadula, DM; Campbell, G; et al. (1985b) The joint neurotoxic action of inhaled methyl butyl
ketone vapor and dermally applied O-ethyl O-4-nitrophenyl phenylphosphonothioate in hens: potentiating effect.
Toxicol Appl Pharmacol 79(l):69-82.
Abou-Donia, MB; Hu, ZH; Lapadula, DM; et al. (1991) Mechanisms of joint neurotoxicity of n-hexane, methyl
isobutyl ketone and O-ethyl O-4-nitrophenyl phenylphosphonothioate in hens. J Pharmacol Exp Ther 257(1):282-
289.
Adedoyin, A; Prakash, C; O'Shea, D; et al. (1994) Stereoselective disposition of hexobarbital and its metabolites:
relationship to the S-mephenytoin polymorphism in Caucasian and Chinese subjects. Pharmacogenetics 4(1):27-
38.
Allen, N; Mendell, JR; Billmaier, JD; et al. (1975) Toxic polyneuropathy due to methyl n-butyl ketone. Arch
Neurol32(4):209-218.
Allen, N; Mendell, JR; Billmaier, D; et al. (1974) An outbreak of a previously undescribed toxic polyneuropathy
due to industrial solvent. Trans AmNeurol Assoc 99:74-79.
ATSDR (Agency for Toxic Substances and Disease Registry). (1992) Toxicological profile for 2-hexanone (CAS
No. 591-78-6). Public Health Service, U.S. Department of Health and Human Services, Atlanta, GA. Available
online at http://www.atsdr.cdc.gov/toxpro2.html.
Billmaier, D; Allen, N; Craft, B; et al. (1974) Peripheral neuropathy in a coated fabrics plant. J Occup Med
16(10):665-671.
Bos, PM; de Mik, G; Bragt, PC. (1991) Critical review of the toxicity of methyl n-butyl ketone: risk from
occupational exposure. Am J Ind Med 20(2): 175-194.
Branchflower, RV; Pohl, LR. (1981) Investigation of the mechanism of the potentiation of chloroform-induced
hepatotoxicity and nephrotoxicity by methyl n-butyl ketone. Toxicol Appl Pharmacol 61(3):407-413.
Brown, EM; Hewitt, WR. (1984) Dose-response relationships in ketone-induced potentiation of chloroform
hepato- and nephrotoxicity. Toxicol Appl Pharmacol 76(3):437-453.
Bus, JS; White, EL; Gillies, PJ; et al. (1981) Tissue distribution of n-hexane methyl-n-butyl ketone and
2,5-hexanedione in rats after single or repeated inhalation exposure to n-hexane. Drug Metab Dispos 9(4):386-
387.
81
-------
Carnegie-Mellon Institute of Research. (1977) Comparative toxicity to rats of methoxyacetone and five other
aliphatic ketones in their drinking water (with cover letter). Produced by the Carnegie-Mellon Institute of
Research, Pittsburgh, PA for the Union Carbide Corporation, Danbury, CT. Submitted under TSCA section 8D;
EPA Document No. 87-8212140; NTIS No. OTS0206068.
Costa, LG; Aschner, M; Vitalone, A; et al. (2004) Developmental neuropathology of environmental agents. Annu
Rev Pharmacol Toxicol 44:87-110.
Couri, D; Hetland, LB; Abdel-Rahman, MS; et al. (1977) The influence of inhaled ketone solvent vapors on
hepatic microsomal biotransformation activities. Toxicol Appl Pharmacol 41(2):285-289.
Davenport, JG; Farrell, DF; Sumi, M. (1976) "Giant axonal neuropathy" caused by industrial chemicals:
neurofilamentous axonal masses in man. Neurology 26(10):919-923.
DeCaprio, AP; Olajos, EJ; Weber, P. (1982) Covalent binding of a neurotoxic n-hexane metabolite: conversion of
primary amines to substituted pyrrole adducts by 2,5-hexanedione. Toxicol Appl Pharmacol 65(3):440-450.
DeCaprio, AP; Briggs, RG; Jackowski, SJ; et al. (1988) Comparative neurotoxicity and pyrrole-forming potential
of 2,5-hexanedione and perdeuterio-2,5-hexanedione in the rat. Toxicol Appl Pharmacol 92(l):75-85.
DiVincenzo, GD; Kaplan, CJ; Dedinas, J. (1976) Characterization of the metabolites of methyl n-butyl ketone,
methyl iso-butyl ketone, and methyl ethyl ketone in guinea pig serum and their clearance. Toxicol Appl
Pharmacol 36(3):511-522.
DiVincenzo, GD; Hamilton, ML; Kaplan, CJ; et al. (1977) Metabolic fate and disposition of 14C-labeled methyl
n-butyl ketone in the rat. Toxicol Appl Pharmacol 41(3):547-560.
DiVincenzo, GD; Hamilton, ML; Kaplan, CJ; et al. (1978) Studies on the respiratory uptake and excretion and the
skin absorption of methyl n-butyl ketone in humans and dogs. Toxicol Appl Pharmacol 44(3):593-604.
Duckett, S; Williams, N; Francis, S. (1974) Peripheral neuropathy associated with inhalation of methyl-n-butyl
ketone. Experientia 30(11): 1283-1284.
Duckett, S; Streletz, LJ; Chambers, RA; et al. (1979) 50 ppmMnBK subclinical neuropathy in rats. Experientia
35:1365-1367.
Duguay, AB; Plaa, GL. (1995) Tissue concentrations of methyl isobutyl ketone, methyl n-butyl ketone and their
metabolites after oral or inhalation exposure. Toxicol Lett 75(l-3):51-58.
Eben, A; Flucke, W; Mihail, F; et al. (1979) Toxicological and metabolic studies of methyl n-butylketone,
2,5-hexanedione, and 2,5-hexanediol in male rats. Ecotoxicol Environ Saf 3(2):204-217.
Egan, G; Spencer, P; Schaumburg, H; et al. (1980) n-Hexane-"free" hexane mixture fails to produce nervous
system damage. Neurotoxicology 1:515-524.
Gilchrist, M; Hunt, W; Allen, N; et al. (1974) Toxic peripheral neuropathy. MMWR 23:9-10.
Graham, DG. (1999) Neurotoxicants and the cytoskeleton. Curr Opin Neurol 12(6):733-737.
Granvil, CP; Sharkawi, M; Plaa, GL. (1994) Metabolic fate of methyl n-butyl ketone, methyl isobutyl ketone and
their metabolites in mice. Toxicol Lett 70(3):263-267.
Hamelin, G; Charest-Tardif, G; Truchon, G; et al. (2005) Physiologically based modeling of n-hexane kinetics in
humans following inhalation exposure at rest and under physical exertion: impact on free 2,5-hexanedione in urine
and on n-hexane in alveolar air. J Occup Environ Hyg 2(2):86-97; quiz D86-87.
82
-------
Hernandez-Viadel, ML; Rodrigo, R; Felipo, V. (2002) Selective regional alterations in the content or distribution
of neuronal and glial cytoskeletal proteins in brain of rats chronically exposed to 2,5-hexanedione. Toxicol Ind
Health 18(7):333-341.
Hewitt, LA; Ayotte, P; Plaa, GL. (1986) Modifications in rat hepatobiliary function following treatment with
acetone, 2-butanone, 2-hexanone, mirex, or chlordecone and subsequently exposed to chloroform. Toxicol Appl
Pharmacol 83(3):465-473.
Hewitt, LA; Valiquette, C; Plaa, GL. (1987) The role of biotransformation-detoxication in acetone-, 2-butanone-,
and 2-hexanone-potentiated chloroform-induced hepatotoxicity. Can J Physiol Pharmacol 65(11):2313-2318.
Hewitt, LA; Palmason, C; Masson, S; et al. (1990) Evidence for the involvement of organelles in the mechanism
of ketone-potentiated chloroform-induced hepatotoxicity. Liver 10(l):35-48.
Homan, ER; Weil, CS; Cox, ER. (1977) Comparative pathology on rats given methoxyacetone and five other
aliphatic ketones in drinking water (detone neurotoxicity). Produced by the Carnegie-Mellon Institute of
Research, Pittsburgh, PA for the Union Carbide Corporation, Danbury, CT. Submitted under TSCA Section 8D;
EPA Document No. 878212141; NTIS No. OTS0206068.
Imaoka, S; Funae, Y. (1991) Induction of cytochrome P450 isozymes in rat liver by methyl n-alkyl ketones and n-
alkylbenzenes. Effects of hydrophobicity of inducers on inducibility of cytochrome P450. Biochem Pharmacol
42(Suppl):S143-S150.
Johnson, BL; Setzer, JV; Lewis, TR; et al. (1977) Effects of methyl n-butyl ketone behavior and the nervous
system. Am Ind Hyg Assoc J 38(11):567-579.
Johnson, BL; Anger, WK; Setzer, JV; et al. (1979) Neurobehavioral effects of methyl N-butyl ketone and methyl
N-amyl ketone in rats and monkeys: a summary of NIOSH investigations. J Environ Pathol Toxicol 2(5): 113-
133.
Johnsrud, EK; Koukouritaki, SB; Divakaran, K; et al. (2003) Human hepatic CYP2E1 expression during
development. J Pharmacol Exp Ther 307(1):402^07.
Kamataki, T; Maeda, K; Yamazoe, Y; et al. (1983) Sex difference of cytochrome P-450 in the rat: purification,
characterization, and quantitation of constitutive forms of cytochrome P-450 from liver microsomes of male and
female rats. Arch Biochem Biophys 225(2):758-770.
Katz, GV; O'Donoghue, JL; DiVincenzo, GD; et al. (1980) Comparative neurotoxicity and metabolism of ethyl n-
butyl ketone and methyl n-butyl ketone in rats. Toxicol Appl Pharmacol 52(1): 153-158.
Knodell, RG; Dubey, RK; Wilkinson, GR; et al. (1988) Oxidative metabolism of hexobarbital in human liver:
relationship to polymorphic S-mephenytoin 4-hydroxylation. J Pharmacol Exp Ther 245(3):845-849.
Krasavage, WJ; O'Donoghue, JL (1977) Chronic inhalation exposure of rats to methyl n-butyl ketone (MnBK).
Eastman Kodak Company, Rochester, NY. Submitted under TSCA Section 8ECP; EPA Document No. 88-
920009282; NTIS No. OTS0571036. TSCA Docket/EPA TL-77-1.
Krasavage, WJ; O'Donoghue, JL; DiVincenzo, GD; et al. (1980) The relative neurotoxicity of methyl-n-butyl
ketone, n-hexane and their metabolites. Toxicol Appl Pharmacol 52(3):433-441.
Lapin, EP; Weissbarth, S; Maker, HS; et al. (1982) The sensitivities of creatine and adenylate kinases to the
neurotoxins acrylamide and methyl n-butyl ketone. Environ Res 28(l):21-31.
Lehning, EJ; Dyer, KS; Jortner, BS; et al. (1995) Axonal atrophy is a specific component of 2,5-hexanedione
peripheral neuropathy. Toxicol Appl Pharmacol 135(l):58-66.
Lehning, EJ; Jortner, BS; Fox, JH; et al. (2000) gamma-Diketone peripheral neuropathy. I. Quality morphometric
analyses of axonal atrophy and swelling. Toxicol Appl Pharmacol 165(2): 127-140.
83
-------
LoPachin, RM; DeCaprio, AP. (2005) Protein adduct formation as a molecular mechanism in neurotoxicity.
Toxicol Sci 86(2):214-225.
LoPachin, RM; Jortner, BS; Reid, ML; Das, S. (2003) gamma-Diketone central neuropathy: quantitative
morphometric analysis of axons in rat spinal cord white matter regions and nerve roots. Toxicol Appl Pharmacol
193(l):29-46.
Lopachin, RM; He, D; Reid, ML; et al. (2004) 2,5-Hexanedione-induced changes in the monomeric neurofilament
protein content of rat spinal cord fractions. Toxicol Appl Pharmacol 198(l):61-73.
Lopachin, RM; He, D; Reid, ML. (2005) 2,5-Hexanedione-induced changes in the neurofilament subunit pools of
rat peripheral nerve. Neurotoxicology 26(2):229-240.
MAI (Microbiological Associates, Inc.). (1986) Subchronic toxicity of methyl isobutyl ketone in Sprague-Dawley
rats. Final report. Produced by Microbiological Associates, Inc., Bethesda, MD, for the Research Triangle
Institute, Research Triangle Park, NC; Study No. 5221.04. Unpublished report dated July 15, 1986.
Mallov, JS. (1976) MBK neuropathy among spray painters. JAMA 235(14): 1455-1457.
Mayer, VW; Coin, CJ. (1994) Induction of chromosome loss in yeast by combined treatment with neurotoxic
hexacarbons and monoketones. Mutat Res 341(2):83-91.
Mendell, JR; Saida, K; Ganansia, MF; et al. (1974) Toxic polyneuropathy produced by methyl N-butyl ketone.
Science 185(153):787-789.
Metso, AJ; Palmu, K; Partanen, JV. (2008) Compound nerve conduction velocity—a reflection of proprioceptive
afferents? Clin Neurophysiol 119:29-32.
Nakajima, T; Elovaara, E; Park, SS; et al. (1991) Immunochemical detection of cytochrome P450 isozymes
induced in rat liver by n-hexane, 2-hexanone and acetonyl acetone. Arch Toxicol 65(7):542-547.
NLM (National Library of Medicine). (2005) 2-Hexanone. HSDB (Hazardous Substances Data Bank). National
Institutes of Health, U.S. Department of Health and Human Services, Bethesda, MD. Available online at
http://toxnet.nlm.nih.gov.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press. Available online at http://books.nap.edu/books/POD115/html/
index, html.
O'Donoghue, JL; Krasavage, WJ. (1979) Chronic inhalation exposure of cats to methyl n-butyl ketone (MnBK).
Eastman Kodak Company, Rochester, NY. Submitted under TSCA Section 8ECP; EPA Document No. 88-
920008233; NTIS No. OTS0555051. TSCA Docket/EPA TX-80-58.
O'Donoghue, JL; Krasavage, WJ. (1981) Neurotoxicity studies on methyl n-butyl ketone (ace. no. 901958),
methyl isobutyl ketone (ace. no. 900416), methyl ethyl ketone (ace. no. 900383) and their combinations. Eastman
Kodak Company, Rochester, NY. Submitted under TSCA Section 8ECP; EPA Document No. 88-920008233;
NTISNo. OTS0555051.
O'Donoghue, JL; Krasavage, WJ; Terhaar, CJ. (1978) A comparative chronic toxicity study of methyl n-propyl
ketone, methyl n-butyl ketone, and hexane by ingestion. Eastman Kodak Company, Rochester, NY; Report No.
104657Y. Submitted under TSCA Section 8ECP; EPA Document No. 88-920008233; NTIS No. OTS0555051.
[An external peer review was conducted by EPA in December 2007 to evaluate the accuracy of experimental
procedures, results, and interpretation and discussion of the findings presented. A report of this peer review is
available through the EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris(@,epa.gov (e-mail address) and on the IRIS website (www.epa.gov/iris).]
Passero, S; Battistini, N; Cioni, R; et al. (1983) Toxic polyneuropathy of shoe workers in Italy. A clinical,
neurophysiological and follow-up study. Ital J Neurolog Sci 4:463-472.
84
-------
Perbellini, L; Mozzo, P; Olivato, D; et al. (1990) "Dynamic" biological exposure indexes for n-hexane and
2,5-hexanedione, suggested by a physiologically based pharmacokinetic model. Am Ind Hyg Assoc J 51(7):356-
362
Peters, MA; Hudson, PM; Dixon, RL. (1981) The effect totigestational exposure to methyl n-butyl ketone has on
postnatal development and behavior. Ecotoxicol Environ Saf 5(3):291-306.
Sabri, MI. (1984) Further observations on in vitro and in vivo effects of 2,5-hexanedione on glyceraldehyde-3-
phosphate dehydrogenase. Arch Toxicol 55(3): 191-194.
Sabri, MI; Ederle, K; Holdsworth, CE; et al. (1979) Studies on the biochemical basis of distal axonopathies II.
Specific inhibition of fructose-6-phosphate kinase by 2,5-hexanedione and methyl-butyl ketone. Neurotoxicology
1:285-297.
Saida, K; Mendell, JR; Weiss, HS. (1976) Peripheral nerve changes induced by methyl n-butyl ketone and
potentiationby methyl ethyl ketone. J Neuropathol Exp Neurol 35(3):207-225.
Sato, A; Nakajima, T. (1979) Partition coefficients of some aromatic hydrocarbons and ketones in water, blood
and oil. BrJ Ind Med36(3):231-234.
Schrenk, HH; Yant, WP; Patty, FA. (1936) Acute response of guinea pigs to vapors of some new commercial
organic compounds. Public Health Rep 51:624-631.
Shibata, E; Johanson, G; Lof, A; et al. (2002) Changes in n-hexane toxicokinetics in short-term single exposure
due to co-exposure to methyl ethyl ketone in volunteers. Int Arch Occup Environ Health 75(6):399-405.
Smyth, HF, Jr.; Carpenter, CP; Weil, CS; et al. (1954) Range-finding toxicity data: list V. AMA Arch Ind Hyg
Occup Med 10(l):61-68.
Sobue, I; lida, M; Yamamura, Y; et al. (1978) n-Hexane polyneuropathy. Int J Neurol 11:317-330.
Spencer, PJ; Schaumburg, HH. (1977) Ultra structural studies in the dying-back process III. The evolution of giant
axonal degeneration. J Neuropathol Exp Neurol 36:276-299.
Spencer, PJ; Schaumburg, HH; Raleigh, RL; et al. (1975) Nervous system degeneration produced by the industrial
solvent methyl n-butyl ketone. Arch Neurol 32:219-222.
Tanii, H; Tsuji, H; Hashimoto, K. (1986) Structure-toxicity relationship of monoketones. Toxicol Lett 30(1): 13-
18.
Toftgard, R; Haaparanta, T; Eng, L; et al. (1986) Rat lung and liver microsomal cytochrome P-450 isozymes
involved in the hydroxylation of n-hexane. Biochem Pharmacol 35(21):3733-3738.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment of chemical
mixtures. Federal Register 51(185):34014-34025. Available online at http://www.epa.gov/ncea/raf/rafguid.htm.
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register 51(185):34006-34012.
Available online at http://www.epa.gov/ncea/raf/rafguid.htm.
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk assessment.
Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment, Cincinnati, OH;
EPA/600/6-87/008. Available from the National Technical Information Service, Springfield, VA, PB88-
179874/AS, and online at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855.
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register 56(234):63798-63826.
Available online at http://www.epa.gov/ncea/raf/rafguid.htm.
85
-------
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation toxicity: notice of
availability. Federal Register 59(206):53799. Available online at http://www.epa.gov/EPA-
PEST/1994/October/Day-26/pr-ll.html.
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry. Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment,
Cincinnati, OH; EPA/600/8-90/066F. Available from the National Technical Information Service, Springfield,
VA, PB2000-500023, and online at http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=71993.
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/R-94/007. Available from the National Technical Information Service, Springfield,
VA, PB95-213765, and online at
http://cfpub.epa. gov/ncea/raf/raf_pubtitles.cfm?detype=document&excCol=archive.
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register 61(212):56274-56322.
Available online at http://www.epa.gov/ncea/raf/rafguid.htm.
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-26954. Available
online at http://www.epa.gov/ncea/raf/rafguid.htm.
U.S. EPA. (1999). Benchmark dose software (BMDS) version Lib (last modified June 15, 1999). Available
online at http://www.epa.gov/ncea/bmds/.
U.S. EPA. (2000a) Science policy council handbook: risk characterization. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA/100-B-00-002. Available online at
http://www.epa.gov/OSA/spc/pdfs/prhandbk.pdf.
U.S. EPA. (2000b) Benchmark dose technical guidance document [external review draft]. Risk Assessment
Forum, Washington, DC; EPA/630/R-00/001. Available online at http://cfpub.epa.gov/ncea/cfm/
nceapublication.cfm?ActType=PublicationTopics&detype=DOCUMENT&subject=
BENCHMARK+DOSE&subjtype=TITLE&excCol=Archive.
U.S. EPA. (2000c) Supplementary guidance for conducting health risk assessment of chemical mixtures. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/002. Available online at
http://cfpub.epa. gov/ncea/raf/chem_mix. cfm.
U.S. EPA. (2002) A review of the reference dose and reference concentration processess. Risk Assessment
Forum, Washington, DC; EPA/630/P-02/002F. Available online at
http://cfpub.epa.gov/ncea/raf/raf_PUB TITLES.cfm?detype=document&excCol=archive.
U.S. EPA. (2003a) Toxicological review of methyl isobutyl ketone. Integrated Risk Information System (IRIS),
National Center for Environmental Assessment, Washington, DC; EPA/635/R-03/002. Available online at
http://www.epa.gov/iris.
U.S. EPA. (2003b) Toxicological review of methyl ethyl ketone. Integrated Risk Information System (IRIS),
National Center for Environmental Assessment, Washington, DC; EPA/635/R-03/009. Available online at
http://www.epa.gov/.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Federal Register 70(66): 17765-18717. Available
online at http://www.epa.gov/cancerguidelines.
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure to carcinogens.
Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available online at
http://www.epa.gov/cancerguidelines.
U.S. EPA. (2005c) Toxicological review of n-hexane. Integrated Risk Information System (IRIS), National Center
for Environmental Assessment, Washington, DC; EPA/635/R-03/012. Available online at
http://www.epa.gov/iris.
86
-------
U.S. EPA. (2006a) Science policy council handbook: peer review. 3rd edition. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA/100/B-06/002. Available online at
http://www.epa.gov/OSA/spc/2peerrev.htm.
U.S. EPA. (2006b) A framework for assessing health risks of environmental exposures to children. National
Center for Environmental Assessment, Washington, DC; EPA/600/R-05/093F. Available online at
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=158363.
Yap, CB; Hirota, T. (1967) Sciatic nerve motor conduction velocity study. J Neurol Neurosurg Psychiat 30:233-
239.
Yu, RC; Hattis, D; Landaw, EM; et al. (2002) Toxicokinetic interaction of 2,5-hexanedione and methyl ethyl
ketone. Arch Toxicol 75(11-12):643-652.
87
-------
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION
The "Toxicological Review of 2-Hexanone" has undergone a formal external peer review
performed by scientists in accordance with EPA guidance on peer review (U.S. EPA, 2006a;
2000a). The external peer reviewers were tasked with providing written answers to general
questions on the overall assessment and on chemical-specific questions in areas of scientific
controversy or uncertainty. A summary of significant comments made by the external reviewers
and EPA's responses to these comments arranged by charge question follow. In many cases the
comments of the individual reviewers have been synthesized and paraphrased in development of
Appendix A. EPA also received scientific comments from the public. These comments and
EPA's responses are included in a separate section of this appendix.
On April 10, 2008, EPA introduced revisions to the IRIS process for developing chemical
assessments. As part of the revised process, the disposition of peer reviewer and public
comments, as found in this appendix, and the revised IRIS toxicological review were provided to
the external peer review panel members for their comments on April 22, 2009. No additional
peer review panel comments were received as part of this second review.
EXTERNAL PEER REVIEWER COMMENTS
The reviewers made several editorial suggestions to clarify specific portions of the text.
These changes were incorporated in the document as appropriate and are not discussed further.
A. General Charge Questions:
1. Is the Toxicological Review logical, clear and concise? Has EPA accurately, clearly and
objectively represented and synthesized the scientific evidence for noncancer and cancer hazard?
Comment: Four of the reviewers stated that the toxicological review was logical, concise, and
accurately summarized. However, two of the reviewers noted specific sections where the
structure could be improved with the inclusion of an introductory paragraph. One reviewer
commented on the repetition of the study descriptions in multiple sections and implied that the
document would be clearer if information was provided in a single table.
Response: Specific sections, such as Sections 5.1 and 6.2, were restructured to improve
readability. Repetitive study details were removed where appropriate.
A-l
-------
Comment: Two of the reviewers mentioned that axonal swelling was not adequately considered
or discussed and suggested that this endpoint might be a better indicator of 2-hexanone
neurotoxicity than myofibrillar atrophy.
Response: Supporting data on axonal swelling was reviewed. As a result, axonal swelling was
selected as the endpoint in the derivation of the RfD rather than myofibrillar atrophy. See
response to first comment under charge question B.2 for further details.
2. Please identify any additional studies that should be considered in the assessment of the
noncancer and cancer health effects of 2-hexanone.
Comment: Five of the reviewers did not identify any additional studies that should be included in
the assessment. One reviewer suggested the following paper to be considered in the assessment:
Spencer, PJ; Schaumburg, HH. (1977) Ultra structural studies in the dying-back process III. The
evolution of giant axonal degeneration. J Neuropathol Exp Neurol 36:276-299.
In addition, the reviewer suggested that studies related to co-exposure of 2-hexanone or
2,5-hexanedione with MEK be considered.
Two reviewers commented on the limited database for this chemical, stating that there are areas
where additional studies would be warranted and that probably establishing neurotoxicity may
have discouraged further use of and less interest in this chemical.
Response: The additional study was reviewed and included in the revised assessment. Section
4.5.4.1 summarizes studies of the co-exposure of 2-hexanone and MEK to the extent necessary
for this assessment. The studies provided by the reviewers were in response to a question about
research needs that would likely increase confidences for future assessments of 2-hexanone. The
provided references were reviewed and relevant studies were incorporated into the revised
assessment; however, most of the studies cited were beyond the scope of this assessment and
were not included.
Comment: One reviewer noted the discrepancy between EPA's requirement to utilize adult hens
to test organophosphates for delayed neurotoxicity and the assessment's narrative that hens are
not suitable models for extrapolating experimental data to humans.
Response: The rationale for not selecting a hen study as the critical study was revised.
A-2
-------
3. Please discuss research that you think would be likely to increase confidence in the database
for future assessments of 2-hexanone.
Comment: Two reviewers proposed studies on immunotoxicity, developmental toxicity, and
pharmacokinetics. One of these reviewers stressed the need for additional studies that examine
exposure level and duration relationships on the development of neuropathies used for the RfC
derivation. This reviewer also pointed out that neurotoxicity studies should include both
functional and behavioral endpoints. A third reviewer suggested immunotoxicity testing, a two-
generation combined reproductive study and 2-year chronic bioassays as well as studies
elucidating the mechanism for neurological effects of 2-hexanone. However, this reviewer stated
that the lack of these studies should not delay the derivation of references values based on the
information currently available.
Two reviewers identified and provided a number of references to be considered for inclusion in
the toxicological review. Another reviewer commented on the lack of data to assess
carcinogenic potential but noted that its utility is unclear in this assessment since neurotoxicity is
the endpoint of interest. The reviewer also noted that the critical studies that serve as the basis
for the reference values are small and lack actual data. The reviewer stated that the use of actual
data would be preferred. One reviewer suggested motor function tests to provide a better
understanding of 2-hexanone neurotoxicity.
Response: EPA agrees that additional toxicity testing would likely improve the database for
2-hexanone. The provided references were reviewed and incorporated into the revised
assessment as necessary. No additional response is required.
4. Please comment on the identification and characterization of sources of uncertainty in
Sections 5 and 6 of the assessment document. Please comment on whether the key sources of
uncertainty have been adequately discussed. Have the choices and assumptions made in the
discussion of uncertainty been transparently and objectively described? Has the impact of the
uncertainty on the assessment been transparently and objectively described?
Comment: One reviewer recommended that the database deficiency UF for both the RfC and
RfD should be 10 due to the lack of studies on the immune system and developmental effects.
Another reviewer noted that, given the manipulations and variability assumptions made in the
derivation of the RfC, the database UF should be greater than 3. A third reviewer stated that,
A-3
-------
although the UFs were objectively described in the assessment, the UF of 3 for database
deficiencies did not appear to be enough and suggested that this UF should be 10.
Response: The database UF for the RfC was changed to 10, and the rationale in Section 5.2.4
was revised.
Comment: One reviewer stated that the assumptions surrounding the uncertainties were simply
stated but not defended. The reviewer stated an example that the assessment makes the
assumption of a constant C x T product and that this assumption is seldom valid. The reviewer
pointed out that there was no discussion in the assessment with regard to the impact of this
assumption.
Response: In the absence of chemical-specific data to indicate otherwise, the C x T assumption
was retained based on the available data for 2-hexanone and current EPA guidance (U.S. EPA,
2002; 1994b).
Comment: One reviewer commented on the dose-response modeling as an area of inadequately
addressed uncertainty. The reviewer pointed out uncertainties with regard to the dose-response
modeling, noting (1) that the fit of the model in the region of interest is highly uncertain
considering the modeling was based on the use of only two exposure levels and a control group
with a significant adverse effect observed at the lowest exposure level and (2) none of the models
have an a priori claim to greater biological plausibility and selection was based solely on
goodness-of-fit. This reviewer suggested that based on these uncertainties in the modeling the
RfC should be based on the LOAEL at 10 months in monkeys or 29 weeks in rats. The reviewer
further noted that the effect at the low dose does not reach statistical significance compared with
controls and there is no mention of the lack of data for the effect at the high dose and the inherent
uncertainty in the estimation of the variance in the RfC. Additionally, the reviewer commented
on the toxicokinetic uncertainty in the animal-to-human extrapolation not being addressed
because the HEC methodology with 2-hexanone-specific data was not applied.
Response: EPA retained the BMD modeling of these data as the modeling analysis considers the
whole dose-response curve compared to the NOAEL/LOAEL approach. As noted by the
reviewer, the focus in the derivation of the RfC is in the low-response region; therefore, the
agency reconsidered the sciatic-tibial nerve MCV data in monkeys and concluded that a BMCL
associated with a BMR of 5% was representative of a response in the region of interest. EPA
agrees with the reviewer that the model selection is not based on a claim of greater biological
plausibility; rather the BMD methods fit the mathematical models in BMDS to chemical-specific
dose-response data. The best-fit model is selected based on an evaluation of the adequacy of
A-4
-------
model fit to the data (i.e., chi-square goodness-of-fit p-values > 0.1) and deviation of the model
fit (i.e., lowest AIC values that allow for comparison across models for a particular endpoint).
The rationale and justification for the BMD modeling and the selection of the BMR were revised
in Section 5.2.2. EPA acknowledges that there is uncertainty associated with the data and the
dose-response modeling; these uncertainties were considered in the application of UFs in Section
5.2.4 and in discussion of the level of confidence in the RfC in Section 6.2.
With regard to the comment that the effect at the low concentration did not reach
statistical significance, EPA considered the decrements in MCV observed in the low exposure
group at the end of the study (when two exposure groups were available for analysis) to be
similar to those observed at the six-month interval (when three exposure groups were available
for analysis). Because the study authors used different statistical tests depending on the number
of exposure groups during the course of the study, similar reductions in nerve MCVs varied in
statistical significance. This is noted in the detailed discussion of the study statistics in Section
5.2.2.
In response to the comment regarding the HEC methodology, EPA applied the current
Agency methodology (U.S. EPA, 2002; 1994b). Data are lacking to inform the blood:gas
partition coefficient for 2-hexanone; therefore, in the absence of data the default value of 1 is
used for this coefficient to calculate an HEC. This approach is described in Section 5.2.3. The
toxicokinetic and toxicodynamic uncertainty is considered in the animal to human extrapolation
UF (see Section 5.2.4).
Comment: One reviewer questioned the inclusion of the NOAEL PODs on the uncertainty figure
(Figure 5-1) given the emphasis in the assessment on BMD modeling.
Response: PODs based on NOAEL/LOAEL endpoints were removed from the uncertainty
figures (Figure 5-1 and Figure 5-2).
B. Oral reference dose (RfD) for 2-hexanone
1. A chronic RfD for 2-hexanone has been derived from a 13-month drinking water study
(O'Donoghue et al., 1978) in male rats. Please comment on whether the selection of this study
as the principal study has been scientifically justified. Has this study been transparently and
objectively described in the document? Please identify and provide the rationale for any other
studies that should be selected as the principal study.
Comment: All the reviewers agreed with the selection of this study as the principal study and that
its selection has been reasonably justified. One reviewer commented that food and water intake
data were not reported in the study description. Another reviewer noted that although
A-5
-------
Abdel-Rahman et al. (1978) used lower doses, the data were not reported for the lowest dose
level.
Response: No response is needed.
2. Myofibrillar atrophy of the quadriceps muscle was selected as the critical effect. Please
comment on whether the rationale for the selection of myofibrillar atrophy as the critical effect
has been scientifically justified. Has this selection been transparently and objectively described
in the document? Please provide a detailed explanation. Please comment on the selection of
myofibrillar atrophy of the quadriceps muscle as the critical effect rather than other endpoints
identified in O'Donoghue et al. (1978). Please comment on the selection of myofibrillar atrophy
of the quadriceps muscle as compared to the peripheral nerve axonal swelling. Please identify
and provide the rationale for any other endpoints that should be considered in the selection of the
critical effect.
Comment: One reviewer suggested that axonal swelling and degeneration in the central and
peripheral nervous systems might be a more appropriate choice as the critical effect because
these effects are characteristic of 2-hexanone-induced neuropathy. Another reviewer also
supported axonal swelling because is it a more sensitive endpoint than muscle atrophy. A third
reviewer stated that the choice of atrophy in the quadriceps muscle versus calf muscle did not
matter since there is a small difference between the two compared to the UFs applied.
One reviewer stated that the rationale for selecting myofibrillar atrophy of the quadriceps muscle
was well defended and referred to the assessment's justification that axonal swelling does not
necessarily lead to nerve dysfunction. This reviewer expressed interest in BMD results for
axonal swelling of the brain and peripheral nerve for comparison purposes. Another reviewer
stated that, although the choice of quadriceps myofibrillar atrophy as the critical effect was well
argued, other endpoints such as axonal swelling or calf muscle myofibrillar atrophy might have
yielded similar RfDs. The reviewer also stated that the observed dose-related effect with weight
reduction and liver weight increase should be further discussed.
One reviewer stated that giant axon atrophy was clearly an appropriate effect for the
development of the RfD but that it is not the most sensitive. This reviewer questioned why body
weight change was not considered. The reviewer stated that axonal atrophy appears to specific
and unique to 2-hexanone and its metabolites but this should not preclude the consideration of
body weight change. Additionally, more information is needed to explain why axonal atrophy
rather than axonal swelling is the more appropriate endpoint.
A-6
-------
Response: EPA reevaluated the axonal data, and the critical effect for the derivation of the RfD
was changed from myofibrillar atrophy of the quadriceps muscle to axonal swelling of the
peripheral nerve. Since axonal swelling precedes axonal degeneration and considering atrophy
occurred at higher doses, axonal swelling was deemed the most sensitive endpoint. Axonal
swelling of the peripheral nerve was selected because of its specificity to 2-hexanone toxicity.
Section 5.1 was revised to reflect this change. Other endpoints such as body or organ weight
changes were not considered since O'Donoghue et al. (1978) provided limited information to
assess body weight as a critical effect. Additional discussion on organ and body weight changes
was added to Section 5.
3. Please comment on the selection of the uncertainty factors applied to the point of departure
(POD) for the derivation of the RfDs. For instance, are they scientifically justified? Are they
transparently and objectively described in the document?
Comment: All of the reviewers agreed that the selection of the UFs were clearly and objectively
described and justified. One reviewer questioned the assignment of a UF of 3 for database
deficiency, which is addressed in a subsequent charge question.
Response: No response is needed.
4. Please comment specifically on the database uncertainty factor of 3 applied in the RfD
derivation. Please comment on body of information regarding reproductive, developmental
toxicity, and immunotoxicity on 2-hexanone as well as the relevance of toxicity data on n-hexane
in the determination of the database uncertainty factor. Please comment on whether the selection
of the database uncertainty factor for the RfD has been scientifically justified. Has this selection
been transparently and objectively described in the document?
Comment: Five of the reviewers stated that the database UF should be 10 rather than 3. The
sixth reviewer did not have a strong opinion on this topic. The reviewers stated that this UF does
not adequately account for reproductive, developmental, and immunotoxic effects of
2-hexanone. The reviewers noted that the available data on n-hexane and 2,5-hexanedione are
not unequivocally negative, developmental effects are seen in rats from maternal injection of
2,5-hexanedione, and immunotoxic effects are observed via other routes of exposure to
2-hexanone. With the absence of clear information about immunotoxicity and developmental
neurotoxicity from oral exposure to 2-hexanone, the database UF should be 10.
A-7
-------
Response: The database lacks a multigenerational reproductive study and developmental studies.
Additionally, a subchronic inhalation study observed a reduction in total white blood cell counts
to 60% of control values in rats exposed to 2-hexanone, suggesting that further study of
immunotoxicity may be warranted. Because of the absence of a two-generation reproductive
study and studies evaluating the developmental toxicity and possible immunotoxicity of
2-hexanone following exposure via the inhalation route, the database UF was changed from 3 to
10.
5. Please provide any other comments on the derivation of the RfD.
Comment: One reviewer stated that the total UF should be 1,000 instead of 300. No other
information was provided.
Response: The total UF for the RfD was revised to 1,000 to take into account intraspecies
variability (UFH= 10), interspecies uncertainty (UFA= 10), and database deficiencies (UFD= 10).
Comment: One reviewer stated that it is obvious that, for the RfD derivation, a four-parameter
model would yield a good fit for data with four data points. The reviewer questioned if EPA had
any concerns with convergence and suggested reseating of the doses to refit the model if that was
the case.
Response: As noted by the reviewer, a four-parameter model would provide the best fit for a
curve with four data points. In the revised assessment, axonal swelling was selected at the
critical effect. BMD modeling was performed using this critical effect; only one model, the
multistage model, provided an adequate fit of the data. Given the data available for the revised
critical effect and that only one model yielded a good fit, the issue of comparing four-parameter
models to other models becomes moot. There were no concerns regarding the existence of
convergence.
Comment: One reviewer restated the comment made in response to Charge Question A.4
regarding the assumption that the product of concentration and time was a constant.
Response: Please see the response to the second comment under Charge Question A.4.
A-8
-------
Comment: One reviewer commented on the fact that, even though other studies were published
on the toxicity of 2-hexanone, the principal study selected (O'Donoghue et al., 1978) was
justified.
Response: No response is needed.
Comment: One reviewer commented on the significance of having quantitative endpoints and
suggested that future studies on 2-hexanone should take advantage of the available measures for
assessing motor function in rodents.
Response: No response is needed.
Comment: One reviewer redid the BMD modeling in a different version of the BMDS and
pointed out differences in the modeling outputs, such as differences in models reported in each
version, and the availability of the AIC and chi-square statistics in each version. This reviewer
evaluated the modeling output for both quadriceps and calf muscle atrophy and suggested
averaging BMDLs from models with identical fits.
Response: All BMD modeling was redone in version 1.4.1c or more recent versions as they
became available. In the revised assessment, axonal swelling was selected as the critical effect,
and there was only one best-fit model for this endpoint. As a result, BMDL averaging was not
considered.
C. Inhalation reference concentration (RfC) for 2-hexanone
1. A chronic RfC for 2-hexanone has been derived from a 10-month inhalation study (Johnson et
al., 1977) in rats and monkeys. Please comment on whether the selection of this study as the
principal study has been scientifically justified. Has this study been transparently and
objectively described in the document? Please comment on the use of a 10-month monkey study
(Johnson et al., 1977) as opposed to a 72-week rat study (Krasavage and O'Donoghue, 1977).
Please identify and provide the rationale for any other studies that should be selected as the
principal study.
Comment: All of the reviewers agreed that the selection of the inhalation study by Johnson et al.
(1977) was scientifically justified. Most reviewers additionally stated that it was more
appropriate to derive the RfC from Johnson et al. (1977) rather than Krasavage and O'Donoghue
(1977), citing the use of nonhuman primates, the functional endpoint selected, and observation of
A-9
-------
adverse effects in the low-dose group in Johnson et al. (1977) and insufficient details or data in
the Krasavage and O'Donoghue (1977) study.
Response: No response is needed.
Comment: One reviewer stated that further detail and explanation is needed elsewhere in the text
to support the claim that the Krasavage and O'Donoghue (1977) study provides limited
information to serve as the basis for an RfC.
Response: The sentence stating that Krasavage and O'Donoghue (1977) provided limited
information to serve as the basis for an RfC was removed from the revised assessment.
Comment: One reviewer stated that the description of Johnson et al. (1977) as a 10-month study
was misleading because only 6-month data were used for the derivation of the RfC.
Response: In Johnson et al. (1977), animals were exposed for 10 months. Thus, it is appropriate
to use the term "10-month study" in the study description. It should be noted that after 6 months
the experiment at the high dose was terminated. Therefore, the data obtained at 6 months were
used for BMD modeling. Language has been added to the text to clarify this.
Comment: One reviewer added that, instead of extracting data from a figure, access to original
raw data would have been preferred.
Response: EPA attempted to contact the study authors to obtain the raw data from the study but
was unsuccessful.
2. Motor conduction velocity of the sciatic-tibial nerve in monkeys was selected as the critical
toxicological effect. Please comment on whether the selection of this critical effect has been
scientifically justified. Has this selection been transparently and objectively described in the
document? Please provide a detailed explanation. Please comment on the use of motor
conduction velocity of the sciatic-tibial nerve instead of motor conduction velocity of the ulnar
nerve. Please comment on the use of monkey data instead of rat data. Please identify and
provide the rationale for any other endpoints that should be considered in the selection of the
critical effect.
Comment: Three reviewers stated that the selection of sciatic-tibial nerve MCV in monkeys was
appropriate and scientifically justified in this assessment. One reviewer deferred to other
A-10
-------
reviewers to comment on the relevance of the endpoint. This reviewer noted that Table 5-3
might be better displayed as a figure and added that myofibrillar atrophy and axonal swelling
incidences could also be plotted. One reviewer stated that the selection of the endpoint and the
rationale to use monkey data versus rat data was not well defended in Section 5 and made
suggestions to aid in better articulation of the rationale for the endpoint. Another reviewer stated
that more discussion was needed in Section 5 to defend the use of sciatic tibial nerve MCV
versus ulnar nerve MCV. One reviewer was not convinced that sciatic tibial nerve MCV was the
better choice than ulnar nerve MCV, stating that the choice for selection seemed to be based on
statistical significance rather than functional significance. This reviewer suggested that this
could be a potential area for future research.
Response: Evidence that the primary metabolite of 2-hexanone, 2,5-hexanedione, typically
affects long axons such as the sciatic-tibial nerve prior to other nerves suggests that 2-hexanone-
induced effects on the sciatic-tibial nerve may occur prior to the effects on the ulnar nerve. See
Section 5.2.1 for this rationale. Suggestions presented by the reviewers were already listed in
Section 5.2.1. Table 5-3 was not converted to a figure since these values were extracted from
figures from the study by Johnson et al. (1977).
Comment: One reviewer suggested that a brief discussion of the relationship between decreased
MCV and functional effects be included.
Response: Additional text was added to Section 5.2.1 to address the relationship between
decreased MCV and functional effects.
3. Estimates of the standard deviation of the responses in each dose group are needed to
calculate benchmark doses (BMDs) and their corresponding lower confidence limits (BMDLs).
This information was not provided in Johnson et al. (1977), the principal study. Therefore, an
indirect method for estimating this missing information on response variability was devised.
Please comment on the procedure used to determine the standard deviation. Please comment on
the use of digitization as a method to abstract data from Johnson et al. (1977) for the derivation
of the inhalation reference concentration.
Comment: Three reviewers stated that the method used to estimate the missing information was
adequately described, warranted, or statistically valid. One of these reviewers questioned if the
study authors analyzed the data in multivariate repeated measures framework. Another reviewer
commented that this approach makes a strong assumption with regard to variance and the method
of obtaining mean data is not optimal. One reviewer acknowledged the use of the indirect
A-ll
-------
methodology, while another deferred to comment but stressed assumption of uniform variance
among the dose groups.
Response: No information was available to suggest an assumption other than uniform variance.
Comment: Three reviewers commented on the digitization as a method to abstract data from
Johnson et al. (1977). One reviewer stated that the use of digitization to obtain data is
unavoidable. Two reviewers stated that the digitization procedure was reasonable.
Response: No response is needed.
4. Please comment on the selection of the uncertainty factors applied to the POD for the
derivation of the RfCs. Are they scientifically justified? Are they transparently and objectively
described in the document?
Comment: Three reviewers stated that the selection of UFs was justified, but two of these
reviewers suggested that the database UF should be 10. One reviewer agreed with the
intraspecies UF and subchronic to chronic UF of 10 but did not comment on the database UF.
Another reviewer also agreed with the interspecies and intraspecies UFs but did not comment on
the database UF. One reviewer found the narrative for the UFs confusing since the text referred
to both rats and monkeys. This reviewer suggested clarification of how the monkey data was
treated with respect to the interspecies UF.
Response: The UF narrative was revised to specify that the POD was based on monkey data and
not rat data. The database UF was changed from 3 to 10 to take into account database limitations
as described in the response to the following charge question.
5. Please comment specifically on the database uncertainty factor of 3 applied in the RfC
derivation. Please comment on body of information regarding reproductive, developmental
toxicity (including developmental neurotoxicity), and immunotoxicity on 2-hexanone, as well as
the comparability and relevance of toxicity data on n-hexane and 2,5-hexanedione in the
determination of the database uncertainty factor. Please comment on whether the selection of the
database uncertainty factor for the RfC has been scientifically justified. Has the selection of the
database uncertainty factor been transparently and objectively described in the document?
A-12
-------
Comment: Five of the reviewers stated that the database UF should be 10 or at least greater than
3. One reviewer did not comment specifically on the database UF. The reviewers stated that the
rationale provided for a database UF of 3 was not defensible and information regarding
immunotoxicity and reproductive and developmental toxicity were lacking. Two reviewers
pointed out that the conversion of n-hexane to 2-hexanone is slow. One of these reviewers
further stated that quantitative information on the dose-response relationship for the reproductive
and developmental effects of n-hexane was not necessarily useful for the same effects with
2-hexanone exposure. This reviewer added that fused axons, described in the assessment as a
minimal effect, may be on the continuum of effects that includes axonal swelling, axonal
atrophy, and reduced motor nerve conduction velocity. Given these reasons, a database UF of 10
is more appropriate. Another reviewer stated that the slow metabolism of n-hexane to
2-hexanone coupled with 6-month exposure results used to establish the RfC provides a better
argument for a database UF of 10 rather than 3. Another reviewer pointed to differences in
NOAELs for n-hexane and 2-hexanone.
Response: The database UF was changed to 10 given the uncertainty surrounding the
applicability of n-hexane reproductive and developmental study findings to that of 2-hexanone.
Other data gaps in the 2-hexanone literature, such as immunotoxicity, also contributed to the
attribution of a 10-fold database UF rather than a 3-fold factor. Because of the absence of a two-
generation reproductive study and studies evaluating the possible developmental toxicity and
immunotoxicity of 2-hexanone following exposure via inhalation, a UFo of 10 was considered
warranted.
Comment: One reviewer stated that Section 5.2.4 should have included commentary on
atmospheric levels of 2-hexanone in the Billmaier et al. (1974) study relative to the TLV and
RfC.
Response: Text regarding the TLV was not deemed appropriate for Section 5.2.4. However, the
TLV and data reported by Billmaier et al. (1974) were included in Section 4.1.
6. Please provide any other comments on the derivation of the RfC.
Comment: One reviewer stated that the combined UFs for the RfC should total 3,000 instead of
1,000 (i.e., UFH = 10, UFS = 10, UFA = 10, and UFD = 3).
A-13
-------
Response: The total UF for the RfD was revised to 3,000 to take into account intraspecies
variability (UFH= 10), interspecies uncertainty (UFA= 3), subchronic to chronic uncertainty
(UFS = 10), and database deficiencies (UFD= 10).
Comment: One reviewer stated that some of the summaries in Appendix B-2 were irrelevant.
The reviewer questions why other models were included in Table B-2 and pointed out that the
only model reported was a linear fit. The reviewer asked for justification of the BMR selected
for the selected endpoint.
Response: EPA generally provides the BMD modeling output for the best-fit model for
endpoints under consideration. For the RfC, four endpoints were of particular interest: sciatic-
tibial nerve MCV and ulnar nerve MCV for both monkeys and rats. The/? values are listed for
each model. For each endpoint, the 1st degree polynomial model yielded the best fit.
Justification for selection of the BMR is provided in Section 5.2.2.
Comment: One reviewer expanded on comments provided in response to Charge Questions A.4
and B.5 questioning the assumption that the product of concentration and time is a constant. The
reviewer stated that this rule does not hold for most biological endpoints and demonstrated this
viewpoint with pulmonary absorption data from the assessment. This reviewer stated that
whenever possible experimental data should be used to determine the exponent on the
concentration component of the equation rather than assuming that the value is equal to one.
Response: The Cn x T = k assumption, or Haber's rule, was applied with respect to adjustment of
intermittent exposure to continuous chronic exposure. As demonstrated by the reviewer, this
assumption may not hold true when dealing with acute durations. Since the RfC is derived for
chronic exposure, the assumption is made that C x T is a constant. The use of the assumption is
based on current EPA guidance (U.S. EPA, 2002; 1994b). Please see the response to the second
comment under Charge Question A.4.
Comment: One reviewer stated that the RfC estimate for 2-hexanone is unlikely to have toxic
effects over a lifetime. However, co-exposure with MEK can significantly potentiate the
neurotoxicity of 2-hexanone and thus impact the RfC.
Response: The RfC is derived for chronic exposure to 2-hexanone and serves as an estimate
(with uncertainty spanning perhaps an order of magnitude) of a continuous inhalation exposure
to the human general population (including sensitive subgroups) that is likely to be without an
appreciable risk of deleterious effects over a lifetime. Section 4.5.4 outlines the impact of co-
A-14
-------
exposures on 2-hexanone neurotoxicity. Application of the RfC in cases where co-exposure is
an issue would generally be considered on a risk management level.
Comment: One reviewer recalculated the BMCLs for the endpoints of interest by using BMDS
version 1.4. Ic and pointed out differences in output in the different versions of BMDS.
Additionally, this reviewer questioned if BMD modeling is appropriate for these data, noting that
the assumed LOAEL differed based on the time point used for BMD modeling.
Response: All BMD modeling was redone in version 1.4. Ic or version 2.0. A LOAEL at a later
time point was not considered since data were not available to adequately elucidate information
for BMD modeling.
Comment: One reviewer requested a summary of TOXNET values.
Response: EPA does not generally report toxicity values of other agencies in its IRIS
assessments. Such values are frequently revised, and maintaining their accuracy is not possible.
D. Carcinogenicity of 2-hexanone
1. Under the EPA's 2005 Guidelines for Carcinogen Risk Assessment
(www.epa.gov/iris/backgr-d.htm), the Agency concluded that there is "inadequate information to
assess the carcinogenic potential" of 2-hexanone. Please comment on the scientific justification
for the cancer weight of the evidence characterization.
Comment: All of the reviewers agreed with the carcinogenicity classification. Most reviewers
stated that EPA's conclusion was justified, although one reviewer stated that the classification of
"inadequate" may be too neutral.
Response: No response is needed.
A-15
-------
PUBLIC COMMENTS
Comment: One public commenter stated that the assessment reports that 2-hexanone affects
pregnancy and reproduction. The commenter does not believe that IRIS, in general, has set safe
standards for harmful effects.
Response: IRIS assessments are based on a series of standard agency guidelines and guidance, as
listed in the introduction of the 2-hexanone assessment. The assessment has been reviewed
within the Agency as well as by interagency and external experts. All relevant studies in the
published, peer-reviewed scientific literature of which EPA is aware were considered and are
summarized in the toxicological review. A sensitive endpoint was selected as the critical effect,
and UFs were applied to develop reference values that are likely to be without an appreciable
risk of deleterious effects over a lifetime.
A-16
-------
APPENDIX B-l. DOSE-RESPONSE MODELING FOR DERIVATION OF AN RfD
FOR 2-HEXANONE
B-l.l. METHODS
The models in EPA's BMDS version 1.4.1c were fit to data sets for axonal swelling of
the peripheral nerve in a 13-month drinking water study with exposure to 2-hexanone
(O'Donoghue et al., 1978). The dose levels used were those reported in the study. A BMR of a
10% extra risk of axonal swelling of the peripheral nerve was selected under an assumption that
it represents a minimal biologically significant change. Models were run with the default
restrictions on parameters built into the BMDS. For comparison purposes models for
myofibrillar atrophy of the quadriceps and calf muscles were also run by using a BMR of 10%
extra risk.
B-1.2. RESULTS
The BMD modeling results for animals with axonal swelling of the peripheral nerve are
summarized in Table B-l. The table and the following outputs show the BMD and the 95%
lower bound on the summary of the best-fit model for this endpoint.
Table B-l. BMD modeling results for animals with axonal swelling of the
peripheral nerve
1 Model
Multistage
AIC
12.0784
p Value
0.9981
BMD
36.0688
BMDL
5.08371
BMD/BMDL
7.0949
B-l
-------
Multistage Model
AXONAL SWELLING OF THE PERIPHERAL NERVE
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose/xl-beta2*dose/x2) ]
The parameter betas are restricted to be positive
Dependent variable = Response
Independent variable = DOSE
Total number of observations = 4
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 = 1
Beta(l) = 2.00806e+017
Beta(2) = 0
the user,
Beta(2)
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
Beta(2)
1
Parameter Estimates
Interval
Variable
Limit
Background
Beta (1)
Beta (2)
Estimate
0
0
Std. Err.
8.09867e-005 *
* - Indicates that this value is not calculated.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
B-2
-------
Full model
Fitted model
Reduced model
AIC:
-5.00402
-5.0392
-24.4346
12.0784
0.0703465
38.8611
0.9951
<.0001
Goodness of Fit
Dose
0.0000
143.0000
266.0000
560.0000
Est. Prob.
0.0000
0.8091
0.9968
1.0000
Expected
0.000
8.091
9.968
10.000
Observed
0.000
8.000
10.000
10.000
Size
10
10
10
10
Scaled
Residual
0.000
-0.073
0.180
0.000
ChiA2 =0.04
d.f.
P-value = 0.9981
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
36.0688
5.08371
48.7286
Taken together, (5.08371, 48.7286) is a 90
BMD
% two-sided confidence interval for the
B-3
-------
Multistage Model with 0.95 Confidence Level
"o
I
O
ro
0.8
0.6
0.4
0.2
100
200
300
dose
400
500
12:1308/052008
Multistage Model
MYOFIBRILLAR ATROPHY OF THE QUADRICEPS MUSCLE
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose/xl-beta2*dose/x2-beta3*dose/x3)]
The parameter betas are restricted to be positive
Dependent variable = Response
Independent variable = DOSE
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations =250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
B-4
-------
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0
Beta(3) = 5.88262e+011
the user,
Beta(3)
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l) -Beta(2)
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
Beta(3)
1
Interval
Variable
Limit
Background
Beta(l)
Beta(2)
Beta (3)
Parameter Estimates
Estimate
0
0
0
Std. Err.
3.72829e-008 *
* - Indicates that this value is not calculated.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-10.1823
-10.1976
-26.9205
# Param's
4
1
1
Deviance Test d.f.
0.0306013
33.4763
Goodness of Fit
=0.02
d.f. = 3
P-value = 0.9995
P-value
0.9986
<.0001
Dose
0.0000
143.0000
266.0000
560.0000
Est. Prob.
0.0000
0.1033
0.5043
0.9986
Expected
0.000
1.033
5.043
9.986
Observed
0.000
1.000
5.000
10.000
Size
10
10
10
10
Scaled
Residual
0.000
-0.034
-0.027
0.120
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 141.38
B-5
-------
BMDL =
BMDU =
49.9434
177.391
Taken together, (49.9434, 177.391) is a 90
interval for the BMD
two-sided confidence
Multistage Model with 0.95 Confidence Level
I
c
g
t3
ro
0.8
0.6
0.4
0.2
500
10:0708/072008
Multistage Model
MYOFIBRILLAR ATROPHY OF THE CALF MUSCLE
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose/xl-beta2*dose/x2-beta3*dose/x3)
The parameter betas are restricted to be positive
Dependent variable = Response
Independent variable = DOSE
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
B-6
-------
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(l) = 0
Beta(2) = 0
Beta(3) = 5.88262e+011
the user,
Beta(l)
Beta(3)
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(2)
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
Beta(l) Beta(3)
1 -0.87
-0.87 1
Interval
Variable
Limit
Background
Beta (1)
Beta(2)
Beta(3)
Parameter Estimates
Estimate
0
0.000741607
0
3.89096e-008
Std. Err.
Indicates that this value is not calculated.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log (likelihood)
-11.7341
-11.7421
-27.5256
# Param' s
4
2
1
Deviance
0.0158412
31.5828
Test d.f.
2
3
27.4841
P-value
0.9921
<.0001
Goodness of Fit
Dose
0.0000
143.0000
266.0000
560.0000
Est. Prob.
0.0000
0.1973
0.6053
0.9993
Expected
0.000
1.973
6.053
9.993
Observed
0.000
2.000
6.000
10.000
Size
10
10
10
10
Scaled
Residual
0.000
0.021
-0.034
0.084
=0.01
d.f. = 2
P-value = 0.9956
Benchmark Dose Computation
B-7
-------
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
95.8577
30.1238
156.135
Taken together, (30.1238, 156.135) is a 90
interval for the BMD
two-sided confidence
Multistage Model with 0.95 Confidence Level
o
I
c
o
'•8
0.8
0.6
0.4
0.2
500
12:01 08/072008
B-8
-------
APPENDIX B-2. EXPOSURE-RESPONSE MODELING FOR DERIVATION OF AN
RfC FOR 2-HEXANONE
B-2.1. METHODS
The models in EPA's BMDS version 2.0 were fit to multiple data sets presented in an
inhalation study with exposure to monkeys and rats (Johnson et al., 1977). MCV was
determined to be the most relevant endpoint in both species and was modeled for the sciatic and
tibial nerves. The exposure concentrations used were those reported in the study. The U.S. EPA
(2000c) BMD methodology suggests that, in the absence of any other idea of what level of
response to consider adverse, a change in the mean equal to one control standard deviation from
the control should be used as the BMR. A BMR of a 5% change in nerve conduction velocity
from the control mean was selected under an assumption that it represents a minimal biologically
significant change.
B-2.2. RESULTS
The BMD modeling results are summarized in Table B-2.1. This table shows the BMDs
and 95% lower bounds on doses (BMDLs) derived from each endpoint modeled in monkeys and
rats. The remainder of this section shows detailed summaries of the modeling results for monkey
sciatic and ulnar nerves for both monkeys and rats (all 1st degree polynomial), presented
sequentially.
Table B-2.1. Summary of BMDS modeling results for 2-hexanone
Animal/endpoint
Monkey sciatic-tibial nerve
(MCV at 6 months)
Monkey ulnar nerve
(MCV at 6 months)
Rat sciatic-tibial nerve
(MCV at 25 weeks)
Rat ulnar nerve
(MCV at 25 weeks)
Model
1st degree polynomial
1st degree polynomial
1st degree polynomial
1st degree polynomial
p Value
0.59
0.90
0.79
0.26
AIC
107.59
107.31
123.55
124.77
BMD
146.592
167.345
135.479
235.482
BMDL
121.631
139.235
116.052
176.137
B-9
-------
1 Degree Polynomial Model.
MONKEY SCIATIC TIBIAL
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = Mean
Independent variable = DOSE
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 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
alpha = 28.6225
rho = 0 Specified
beta_0 = 49.3606
beta 1 = -0.0168361
the user,
alpha
beta_0
beta 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha beta_0 beta_l
1 3.2e-015 -4.2e-015
3.2e-015 1 -0.63
-4.2e-015 -0.63 1
Interval
Variable
Limit
alpha
39.6944
beta_0
51.9603
beta_l
0.0122456
Parameter Estimates
95.0% Wald Confidence
Estimate Std. Err. Lower Conf. Limit Upper Conf.
25.351 7.31821 11.0076
49.3606 1.3264 46.761
-0.0168361 0.0023421 -0.0214265
B-10
-------
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
98
976
50
47
33
49.4
47.7
32.9
5.35
5.35
5.35
5.03
5.03
5.03
0.359
-0.399
0.0401
Warning: Likelihood for model Al larger than the Likelihood for model A2 .
Warning: Likelihood for model A3 larger than the Likelihood for model A2 .
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma (i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e (i) } = Sigma/x2
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log (likelihood)
-50.647941
-50.647941
-50.647941
-50.793824
-64.574352
# Param's
4
6
4
3
2
AIC
109.295882
113.295882
109.295882
107.587648
133.148704
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs . R)
2: Are Variances Homogeneous? (Al vs A2 )
3: Are variances adeguately modeled? (A2 vs
4: Does the Model for the Mean Fit? (A3 vs.
(Note: When rho=0 the results of Test 3 and Test
Test
Test
Test
. A3)
fitted)
2 will be
the same.)
Test
Test 1
Test 2
Test 3
Test 4
Tests of Interest
-2*log (Likelihood Ratio) Test df
27.8528
-1.42109e-014
-1.42109e-014
0.291767
p-value
<.0001
<.0001
<.0001
0.5891
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than
non-homogeneous variance model
Consider running a
B-ll
-------
The p-value for Test 3 is less than .1. You may want to consider a
different variance model
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Relative risk
Confidence level = 0.95
BMD = 146.592
BMDL = 121.631
Linear Model with 0.95 Confidence Level
o
Q.
CO
55
50
45
40
35
30
Linear
BMDL BMD
200
400
600
800
1000
dose
09:3502/122009
1 Degree Polynomial Model.
MONKEYS MCV ULNAR
BMDS MODEL RUN
The form of the response function is:
B-12
-------
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = Mean
Independent variable = DOSE
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 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
alpha = 28.6225
rho = 0 Specified
beta_0 = 57.8551
beta 1 = -0.0172861
the user,
alpha
beta_0
beta 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha beta_0 beta_l
1 9.3e-009 -1.4e-008
9.3e-009 1 -0.63
-1.4e-008 -0.63 1
Interval
Variable
Limit
alpha
39.2394
beta_0
60.4399
beta_l
0.0127221
Parameter Estimates
95.0% Wald Confidence
Estimate Std. Err. Lower Conf. Limit Upper Conf.
25.0604 7.23432 10.8814
57.8551 1.31877 55.2704
-0.0172861 0.00232864 -0.0218502
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
98
976
8
8
8
58
56
41
57.9
56.2
41
5.35
5.35
5.35
5.01
5.01
5.01
0.0819
-0.091
0.00914
Warning: Likelihood for model Al larger than the Likelihood for model A2.
B-13
-------
Warning: Likelihood for model A3 larger than the Likelihood for model A2 .
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e (i) } = Sigma/x2
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log(likelihood)
-50.647941
-50.647941
-50.647941
-50.655476
-64.968151
# Param's
4
6
4
3
2
AIC
109.295882
113.295882
109.295882
107.310953
133.936303
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adeguately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
Test 1
Test 2
Test 3
Test 4
-2*log(Likelihood Ratio) Test df
28.6404
-1.42109e-014
-1.42109e-014
0.0150715
p-value
<.0001
<.0001
<.0001
0.9023
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
non-homogeneous variance model
The p-value for Test 3 is less than .1.
different variance model
Consider running a
You may want to consider a
The p-value for Test 4 is greater than .1.
to adeguately describe the data
The model chosen seems
Benchmark Dose Computation
Specified effect = 0.05
B-14
-------
Risk Type
Confidence level =
BMD =
BMDL =
Relative risk
0.95
167.345
139.235
Linear Model with 0.95 Confidence Level
65
60
55
Sg 50
45
40
35
12:0402/122009
200
400
600
800
1000
dose
1s Degree Polynomial Model.
RATS MCV SCIATIC TIBIAL
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = Mean
Independent variable = DOSE
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 3
Total number of records with missing values = 0
B-15
-------
Maximum number of iterations =250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 20.5209
rho = 0 Specified
beta_0 = 42.2427
beta 1 = -0.0155901
the user,
alpha
beta_0
beta 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha beta_0 beta_l
1 8.8e-008 -1.4e-007
8.8e-008 1 -0.63
-1.4e-007 -0.63 1
Interval
Variable
Limit
alpha
27.8815
beta_0
44.2287
beta_l
0.012083
Parameter Estimates
95.0% Wald Confidence
Estimate Std. Err. Lower Conf. Limit Upper Conf.
18.5129 4.78001 9.14425
42.2427 1.01325 40.2568
-0.0155901 0.00178935 -0.0190972
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
97
976
10
10
10
42
41
27
42.2
40.7
27
4.53
4.53
4.53
4.3
4.3
4.3
-0.178
0.198
-0.0197
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A3 uses any fixed variance parameters that
B-16
-------
were specified by the user
Model R: Yi = Mu + e(i)
Var{e (i) } = SigmaA2
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log(likelihood)
-58.741250
-58.741250
-58.741250
-58.777017
-77.698129
# Param's
4
6
4
3
2
AIC
125.482501
129.482501
125.482501
123.554034
159.396257
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adeguately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
-2*log(Likelihood Ratio) Test df
p-value
37.9138
0
0
0.0715333
4
2
2
1
<.0001
1
1
0.7891
Test
Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Relative risk
Confidence level = 0.95
BMD = 135.479
BMDL = 116.052
B-17
-------
Linear Model with 0.95 Confidence Level
o
Q.
CO
45
40
35
30
25
200
400
600
800
1000
dose
12:0802/122009
1 Degree Polynomial Model.
RATS MCV ULNAR
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = Mean
Independent variable = DOSE
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 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
B-18
-------
Default Initial Parameter Values
alpha = 20.5209
rho = 0 Specified
beta_0 = 38.9587
beta 1 = -0.0082721
the user,
alpha
beta_0
beta 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha beta_0 beta_l
1 -6.4e-010 -8.2e-011
-6.4e-010 1 -0.63
-8.2e-011 -0.63 1
Interval
Variable
Limit
alpha
29.0373
beta_0
40.9853
beta_l
0.00469309
Parameter Estimates
95.0% Wald Confidence
Estimate Std. Err. Lower Conf. Limit Upper Conf.
19.2803 4.97816 9.52331
38.9587 1.03404 36.932
-0.0082721 0.00182606 -0.0118511
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
97
976
10
10
10
40
37
31
39
38.2
30.9
4.53
4.53
4.53
4.39
4.39
4.39
0.75
-0.833
0.0828
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma/x2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e (i) } = Sigma/x2
B-19
-------
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log(likelihood)
-58.741250
-58.741250
-58.741250
-59.386277
-67.204199
# Param' s
4
6
4
3
2
AIC
125.482501
129.482501
125.482501
124.772554
138.408398
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adeguately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
Test 1
Test 2
Test 3
Test 4
Tests of Interest
-2*log(Likelihood Ratio) Test df
16.9259
0
0
1.29005
p-value
0.001998
1
1
0.256
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adeguately describe the data
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Relative risk
Confidence level = 0.95
BMD = 235.482
BMDL = 176.137
B-20
-------
Linear Model with 0.95 Confidence Level
o
Q.
CO
44
42
40
38
36
34
32
30
28
Linear
BMDL BMD
200
400 600
dose
800
1000
12:11 02/122009
B-21
------- |