DRAFT - DO NOT CITE OR QUOTE EPA/635/R-11/004A
www.epa.gov/iris
xvEPA
TOXICOLOGICAL REVIEW
OF
VANADIUM PENTOXIDE
(V205)
(CAS No. 1314-62-1)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
July 2011
NOTICE
This document is an Interagency Science Consultation draft. This information is distributed
solely for the purpose of pre-dissemination peer review under applicable information quality
guidelines. It has not been formally disseminated by EPA. It does not represent and should not
be construed to represent any Agency determination or policy. It is being circulated for review
of its technical accuracy and science policy implications.
U.S. Environmental Protection Agency
Washington, DC.
-------
DISCLAIMER
This document is a preliminary draft for review purposes only. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. It has not been formally disseminated by EPA. It does not represent and
should not be construed to represent any Agency determination or policy. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
DRAFT - DO NOT CITE OR QUOTE
-------
CONTENTS - TOXICOLOGICAL REVIEW OF VANADIUM PENTOXIDE
(CAS No. 1314-62-1)
AUTHORS, CONTRIBUTORS, AND REVIEWERS xiii
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION 3
3. TOXICOKTNETICS 6
3.1 Absorption 6
3.1.1 Inhalation Exposure 6
3.1.2 Oral Exposure 7
3.1.3 Dermal Exposure 8
3.1.4 Other Routes of Exposure 8
3.2 Distribution 8
3.2.1 Inhalation Exposure 8
3.2.2 Oral Exposure 9
3.2.3 Dermal Exposure 10
3.2.4 Other Routes of Exposure 11
3.3 Metabolism 11
3.4 Elimination and Excretion 11
3.4.1 Inhalation Exposure 11
3.4.2 Oral Exposure 12
3.4.3 Dermal Exposure 12
3.4.4 Other Routes of Exposure 12
3.5 Physiologically Based Pharmacokinetic (PBPK)/Pharmacodynamic (PD) Models 12
3.5.1 Animal-to-Human Extrapolations 12
4. HAZARD IDENTIFICATION 14
4.1. STUDIES IN HUMANS - EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS 14
4.1.1 Oral Exposure 14
4.1.2 Inhalation Exposure 14
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS - ORAL AND INHALATION 31
4.2.1. Oral Exposure 31
4.2.1.1. Subchronic Studies 31
4.2.1.2 Chronic Studies 32
4.2.2. Inhalation Exposure 33
DRAFT - DO NOT CITE OR QUOTE
-------
4.2.2.1 Subchronic Studies 33
4.2.2.2 Chronic Studies 43
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES - ORAL, INHALATION,
INTRAPERITONEAL AND INJECTION 53
4.3.1. Oral Studies 53
4.3.2. Inhalation Studies 54
4.3.3. Intraperitoneal and Injection Studies 55
4.4 OTHER DURATION-OR ENDPOINT-SPECIFIC STUDIES 57
4.4.1 Acute and Short-Term Studies 57
4.4.1.1 Acute Studies 57
4.4.1.2. Short-term Studies 59
4.4.2 Immunological Endpoints 61
4.4.2.1 Human Studies 61
4.4.2.2 Animal Studies 64
4.4.3 Neurological Endpoints 65
4.4.3.1 Human Studies 66
4.4.3.2 Animal Studies 66
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION FOR PULMONARY FIBROSIS AND CANCER 68
4.5.1 Genotoxicity 68
4.5.1.1 Human Studies 68
4.5.1.2 Laboratory in vivo and in vitro studies 71
4.5.2 Mechanisms of Inflammation andFibrosis 78
4.5.3 Mechanisms of Hyperplasia and Carcinogenicity 82
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS 86
4.6.1. Oral 86
4.6.1.1 Acute 86
4.6.1.2 Subchronic 88
4.6.1.3 Chronic 88
4.6.2. Inhalation 88
4.6.2.1 Acute and Short-term 89
4.6.2.2 Subchronic 90
4.6.2.3 Chronic 91
4.6.3 Mode of Action Information 98
4.6.3.1 Pulmonary Toxicity 98
4.6.3.2Neurotoxicity 100
4.6.3.3 Reproductive Toxicity 100
4.7. EVALUATION OF CARCINOGENICITY 101
4.7.1 Summary of Overall Weight of Evidence 101
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence 101
iy DRAFT - DO NOT CITE OR QUOTE
-------
4.7.3. Mode of Action Information 102
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES 103
4.8.1. Possible Childhood Susceptibility 103
4.8.2. Possible Gender Differences 103
4.8.3. Other Susceptible Populations 103
5. DOSE-RESPONSE ANALYSIS 106
5. LORAL REFERENCE DOSE (RfD) 106
5.1.1 Methods of Analysis 108
5.1.2 RfD Derivation - Including Application of Uncertainty Factors (UF) 109
5.1.2. Previous RfD Assessment 110
5.2. INHALATION REFERENCE CONCENTRATION (RfC) Ill
5.2.1 Choice of Principal Study and Critical Effect with Rationale and Justification Ill
5.2.2 Methods of Analysis 113
5.2.3 RfC Derivation- Including Application of Uncertainty Factors (UFs) 119
5.2.4. Previous RfC Assessment 120
5.3 UNCERTAINTIES IN THE INHALATION REFERENCE CONCENTRATION (RfC) 120
5.4. CANCER ASSESSMENT 123
5.4.1 Choice of Study/Data - with Rationale and Justification 123
5.4.2 Dose-Response Data 124
5.4.3 Dose Adjustments and Extrapolation Methods 124
5.4.4 Inhalation Unit Risk 127
5.4.5 Oral Cancer Slope Factor 127
5.4.6 Uncertainties in Cancer Risk Values 127
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 132
6.1 Human Hazard Potential 132
6.2. Dose Response 136
6.2.1 Noncancer 136
6.2.2 Cancer 138
7. REFERENCES 140
DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Valence states and water solubility of various vanadium compounds 4
Table 2-2: Chemical and physical properties of Vanadium Pentoxide 5
Table 4-1. Hematological results of oral vanadium pentoxide exposure in rats (Mountain
etal. 1953) 32
Table 4-2. Body Weight Gain and Lung Weights in Rats (F344/N) Exposed to Vanadium
Pentoxide by Inhalation for 3 Months (Values are Means±Standard Error) (NTP, 2002) . 35
Table 4-3. Selected Hematology Parameters in Rats (F344/N) Exposed to Vanadium
Pentoxide by Inhalation for 3 Months (NTP, 2002) 35
Table 4-4. Incidences of Selected Nonneoplastic Lesions of the Lung and Nose in Rats
(F344/N) Exposed to Vanadium Pentoxide by Inhalation for 3 Months (NTP, 2002) 37
Table 4-5. Body Weight Gain and Lung Weights in Mice (B6C3Fi) Exposed to Vanadium
Pentoxide by Inhalation for 3 Months (Values are Means±Standard Error) (NTP, 2002) . 40
Table 4-6. Incidences of Selected Nonneoplastic Lesions of the Lung in Mice (B6C3Fi)
Exposed to Vanadium Pentoxide by Inhalation for 3 Months (NTP, 2002) 41
Table 4-7. Selected Nonneoplastic Lesions of the Respiratory System in Rats Exposed to
Particulate Aerosols of Vanadium Pentoxide for 2 Years (NTP, 2002) 44
Table 4-8. Incidences of Respiratory Tumors in Rats Exposed to Vanadium Pentoxide in a
2 Year Inhalation Study (NTP, 2002)a 47
Table 4-9. Selected Nonneoplastic Lesions of the Respiratory System in Mice Exposed to
Vanadium Pentoxide in a 2 Year Inhalation Study (NTP, 2002) 48
Table 4-10. Incidences of Respiratory Tumors in Mice Exposed to Vanadium Pentoxide in
the 2 Year Inhalation Study (NTP, 2002) 51
Table 4 -11. Lung tumor multiplicity in MCA-Treated mice exposed to ViOs by
pharyngeal aspiration (Rondini et al. 2010). a'b 52
Table 4-12. Genotoxicity Data Following Exposure to Vanadium Pentoxide 74
Table 4-13: Summary of Noncancer Results of Repeat-Dose Studies for Oral Exposure of
Experimental Animals to Vanadium Pentoxide 86
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation
Exposure of Vanadium Pentoxide 92
Table 5-1. Hematological results of oral vanadium pentoxide exposure in rats (Mountain
etal. 1953) 107
Table 5-2 Human equivalence dose conversion by BW3/4 for RfD derivation 109
Table 5-3. Selected Nonneoplastic Lesions of the Respiratory System in Rats Exposed to
Particulate Aerosols of Vanadium Pentoxide for 2 Years (NTP, 2002) 114
yi DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-4: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year
Inhalation Studies in Rats (NTP, 2002) 116
Table 5-5: Candidate PODs for Vanadium Pentoxide Derived from NTP Studies (2002)
through BMDS Modeling 116
Table 5-6: Benchmark Modeling Results for Incidence of Larynx Chronic Inflammation in
Female Rats, NTP (2002) 118
Table 5-7: Benchmark Modeling Results for Incidence of Larynx Respiratory Epithelium,
Epiglottis, Hyperplasia in Female Rats, NTP (2002) 119
Table 5-8. Incidences of Respiratory Tumors in Mice Exposed to Vanadium Pentoxide in
the 2 Year Inhalation Study (NTP, 2002) 123
Table 5-9: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year
Inhalation Studies (NTP, 2002) 125
Table 5- 10. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar
Adenoma and Carcinoma in Male Mice, NTP (2002) 126
Table 5-11. Summary of uncertainty in the vanadium pentoxide cancer risk assessment.
129
Table B-l: Candidate PODs for Vanadium Pentoxide Derived from NTP Studies (2002)
through BMDS Modeling 154
Table B-2: Trend Tests on the Selected Data Sets from the 2-Year Inhalation Studies in
Rats (NTP, 2002) 156
Table B-3: Average Life Time Animal Body Weight of Rats in the 2-Year Inhalation
Studies of Vanadium Pentoxide (NTP, 2002) 157
Table B-4: RDDRs for Different Concentration/Sex Group in the 2-Year Inhalation Studies
of Vanadium Pentoxide (NTP, 2002) 157
Table B-5: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year
Inhalation Studies (NTP, 2002) 158
Table B-6: Benchmark Modeling Results for Incidence of Lung Alveolar Epithelium
Hyperplasia in Male Rats, NTP (2002) 159
Table B-7: Benchmark Modeling Results for Incidence of Lung Chronic Active
Inflammation in Male Rats, NTP (2002) 162
Table B-8: Benchmark Modeling Results for Incidence of Larynx Chronic Inflammation in
Male Rats, NTP (2002) 164
Table B-9: Benchmark Modeling Results for Incidence of Larynx Respiratory Epithelium,
Epiglottis, Hyperplasia in Male Rats, NTP (2002) 167
Table B-10: Benchmark Modeling Results for Incidence of Nose Goblet Cell, Respiratory
Epithelium, Hyperplasia in Male Rats, NTP (2002) 169
Table B-ll: Benchmark Modeling Results for Incidence of Lung Alveolar Epithelium
Hyperplasia in Female Rats, NTP (2002) 172
yii DRAFT - DO NOT CITE OR QUOTE
-------
Table B-12: Benchmark Modeling Results for Incidence of Lung Chronic Active
Inflammation in Female Rats, NTP (2002) 174
Table B-13: Benchmark Modeling Results for Incidence of Larynx Chronic Inflammation
in Female Rats, NTP (2002) 176
Table B-14: Benchmark Modeling Results for Incidence of Larynx Respiratory
Epithelium, Epiglottis, Hyperplasia in Female Rats, NTP (2002) 179
Table B-15: Benchmark Modeling Results for Incidence of Larynx Respiratory
Epithelium, Epiglottis, Hyperplasia in Female Rats, NTP (2002) 181
Table C-1. Summary of Candidate PODs and Cancer Slope Factors 185
Table C-2: Trend Tests on the Inhalation Carcinogenesis Data Sets in Mice from the 2-
Year Inhalation Studies (NTP, 2002) 185
Table C-3: Average Life Time Animal Body Weight of Mice for 2-Year Inhalation Studies
of Vanadium Pentoxide (NTP, 2002) 186
Table C-4: RDDRs for Different Concentration/Sex Group of Mice in the 2-Year
Inhalation Studies of Vanadium Pentoxide (NTP, 2002) 187
Table C-5: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year
Inhalation Studies (NTP, 2002) 187
Table C-6: BMR Estimation for Male and Female Mice Data Sets 188
Table C- 7. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar
Adenoma and Carcinoma in Male Mice, NTP (2002) 189
Table C- 8. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar
Adenoma and Carcinoma in Male Mice after Dropping the Highest Concentration, NTP
(2002) 191
Table C- 9. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar
Adenomas and Carcinomas with All Four Concentrations in Female Mice, NTP (2002). 193
Table C- 10. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar
Adenoma and Carcinoma in Female Mice after Dropping the Highest Concentration, NTP
(2002) 194
Table C-11. Grouped Data for MSW Time-to-tumor Modeling; B6C3Fi Male Mice
Exposed to Vanadium Pentoxide by Inhalation for 2 Years (NTP, 2002) 197
Table C-12. Grouped Data for MSW Time-to-tumor Modeling; B6C3Fi Female Mice
Exposed to Vanadium Pentoxide by Inhalation for 2 Years (NTP, 2002) 201
yiii DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 2-1. Vanadium pentoxide structure 3
Figure 4-1: Genomics of ViOs-Induced Bronchitis (reprinted with permission from Ingram
et al., (2007) Respir. Res. Apr 25;8(1):34) 82
• DRAFT - DO NOT CITE OR QUOTE
IX
-------
LIST OF ACRONYMS AND ABBREVIATIONS
BMC
BMCL
BMD
BMDL
BMDS
BMR
bw
cc
CD
CERCLA
CNS
cu.m
FIFRA
g
GI
HED
HEC
Hgb
i.m.
i.p.
IRIS
IUR
i.v.
kg
L
LOAEL
LOAEL[ADJ]
LOAEL[HEC]
m
MCL
MF
mg
benchmark concentration
benchmark concentration (lower limit)
benchmark dose
benchmark dose (lower limit)
benchmark dose software
benchmark response
body weight
cubic centimeters
caesarean delivered
Comprehensive Environmental Response, Compensation and
Liability Act of 1980
central nervous system
cubic meter
Federal Insecticide, Fungicide, and Rodenticide Act
grams
gastrointestinal
human equivalent dose
human equivalent concentration
hemoglobin
intramuscular
intraperitoneal
Integrated Risk Information System
inhalation unit risk
intravenous
kilogram
liter
lowest-observed-adverse-effect level
LOAEL adjusted to continuous exposure duration
LOAEL adjusted for dosimetric differences across species to a human
meter
maximum contaminant level
modifying factor
milligram
x
DRAFT - DO NOT CITE OR QUOTE
-------
mg/kg
mg/L
MRL
MTD
NAAQS
NMMAPS
NOAEL
NOAEL[ADJ]
NOAEL[HEc]
NOEL
OSF
PBPK
ppb
ppm
PPRTV
RBC
RCRA
RDDR
RfC
RfD
RGDR
SCE
SDWA
sq.cm.
TSCA
UF
ug
umol
voc
milligrams per kilogram
milligrams per liter
minimal risk level
maximum tolerated dose
National Ambient Air Quality Standards
National Morbidity, Mortality, and Air Pollution Study
no-observed-adverse-effect level
NOAEL adjusted to continuous exposure duration
NOAEL adjusted for dosimetric differences across species to a human
no-observed-effect level
oral slope factor
physiologically based pharmacokinetic
parts per billion
parts per million
Provisional Peer Reviewed Toxicity Value
red blood cell(s)
Resource Conservation and Recovery Act
Regional deposited dose ratio (for the indicated lung region)
inhalation reference concentration
oral reference dose
Regional gas dose ratio (for the indicated lung region)
sister chromatid exchange
Safe Drinking Water Act
square centimeters
Toxic Substances Control Act
uncertainty factor
microgram
micromoles
volatile organic compound
XI
DRAFT - DO NOT CITE OR QUOTE
-------
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 vanadium
pentoxide. It is not intended to be a comprehensive treatise on the chemical or toxicological
nature of vanadium pentoxide, and does not address other vanadium compounds.
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).
- DRAFT - DO NOT CITE OR QUOTE
XI1
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR
Maureen R. Gwinn, Ph.D., DABT
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, D.C. 20460
IRIS CO-LEAD/AUTHOR
Kathleen Newhouse, M.S.
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, D.C. 20460
CO-AUTHORS
Gillian Backus, Ph.D.*
Northern Virginia Community College
Sterling, VA. 20164
Yuyang Christine Cai, M.S., PMP
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, D.C. 20460
Sanjivani Diwan, Ph.D.*
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, D.C. 20460
Barbara Glenn, Ph.D.
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, D.C. 20460
Kathleen Raffaele, Ph.D.
Xlll
DRAFT - DO NOT CITE OR QUOTE
-------
Office of Solid Waste and Emergency Response
US Environmental Protection Agency
Washington, D.C. 20460
Suryanarayana V. Vulimiri, D.V.M., Ph.D., DABT
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, D.C. 20460
*Former EPA employee
xiv DRAFT - DO NOT CITE OR QUOTE
-------
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 vanadium
pentoxide. 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
exposures 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 ug/m3 air breathed.
Development of these hazard identification and dose-response assessments for vanadium
pentoxide 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
1
DRAFT - DO NOT CITE OR QUOTE
-------
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
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 May 2011.
Portions of this document were developed under a Memorandum of Understanding with
the Agency for Toxic Substances and Disease Registry (ATSDR) as part of a collaborative effort
in the development of human health toxicological assessments.
2
DRAFT - DO NOT CITE OR QUOTE
-------
2. CHEMICAL AND PHYSICAL INFORMATION
As an element, vanadium (V) exists in several oxidation states from -1 to +5. Vanadium
(CASRN 7440-62-2) is a soft silver-grey metal commonly found in ores, tars, coals and oils and
is used as an alloy in steel (WHO, 1988). The focus of this toxicological review is on vanadium
pentoxide (CASRN 1314-62-1), but a short description of chemistry of vanadium and related
compounds is given below for clarification.
The chemistry of vanadium is complex; the valence state of vanadium can shift
depending on pH and other factors. In the body, there is an interconversion of two oxidation
states of vanadium, the tetravalent form, vanadyl (V+4), and the pentavalent form, vanadate
(V+5). Vanadate is considered more toxic than vanadyl because vanadate is reactive with a
number of enzymes and is a potent inhibitor of the Na+K+-ATPase of plasma membranes
(Harris et al. 1984; Patterson et al. 1986).
Figure 2-1. Vanadium pentoxide structure
Generally, V3+ and V4+ predominate in body tissues while V5+ predominates in plasma
(IPCS, 2001). Vanadium pentoxide (V2Os), sodium metavanadate (NaVOs), sodium
orthovanadate (NasVO/t), and ammonium metavanadate (MLiVOs) all contain vanadium in the
+5 oxidation state. Of these compounds, V2Os is the only compound that is covalently bonded.
Vanadium compounds differ in their physicochemical properties which influence their
solubility under different pH conditions and their accessibility and availability in biological
systems [reviewed in (Assem and Levy 2009)]. An acidic pH favors tetravalent state (V+4)
keeping it as vanadyl, while an alkaline pH prefers pentavalent state (V+5) as vanadate (Crans et
al . 2004). In the case of oral ingestion, vanadium compounds are exposed to a range of pH
solutions in the digestive tract starting from the stomach (pH typically between 1-3.5) followed
by the small intestine (pH around 8). Bruyere et al. (1999) state that at pH between 1.3 and 3.3
the predominate form of vanadium is VO2+ and at higher pH the form is VO(OH>3. When the pH
is high V(5+) (e.g. VO4(OH)3) and polymerized vanadium is predominant (HnVioO29(n"6))(Bruyere
et al . 1999). At physiological pH vanadium compounds have been shown to exist in monomeric
tetravalent [VO(OH)3]" and dimeric [(VO)2(OH)5]" forms, as well as pentavalent (H2VO4") forms
3
DRAFT - DO NOT CITE OR QUOTE
-------
[reviewed in (Assem and Levy 2009)]. Thus, the valence of a vanadium compound will depend
on the pH.
The solubility of different vanadium compounds in water between 20 and 25°C differs
among different valences as shown in Table 2-1 (HSDB 2009; IPCS 2001). The elemental
vanadium (V°) is insoluble in water. The tetravalent (V^) compound vanadyl sulfate (VOSO/t)
is highly soluble 534.64 g/L (Rahman and Skyllas-Kazacos 1998), while the pentavalent
vanadium compounds (V+5), such as vanadium pentoxide (V^Os) is less soluble (8 g/L). Other
vanadium compounds such as sodium metavanadate (NaVOs), sodium orthovanadate (NasVO/t)
and ammonium metavanadate (NFLVOs) have solubility of 21 1 g/L, 100 g/L, 58 g/L,
respectively (IPCS 2001). Furthermore, the rate (distinguished from the solubility - an
equilibrium or thermodynamic parameter) of dissolution of various vanadium compounds may
vary, resulting in different concentrations of specific forms of vanadium. It seems a reasonable
hypothesis that these various forms of vanadium will be absorbed differently, which may result
in different physiological effects. For example, V(5+) compounds can mimic phosphate and can
inhibit phosphatases (Assem and Levy 2009).
Table 2-1. Valence states and water solubility of various vanadium compounds.
Vanadium compound
Vanadium
Vanadium pentoxide
Sodium ffj-vanadate
Sodium o-vanadate
Ammonium /w-vanadate
Vanadium oxytrichloride
Vanadyl sulfate
Vanadium tetrachloride
Vanadyl oxydichloride
Vanadium trioxide
Formula
V
V205
NaVO3
Na3VO4
NH4VO3
VOC13
VOSO4
VC14
VOC12
V203
CASRN
7440-62-2
1314-62-1
13718-26-8
13721-39-6
7803-55-6
7727-18-6
27774-13-6
7632-51-1
10213-09-9
1314-34-7
Valenc
y
0
+5
+5
+5
+5
+5
+4
+4
+3
+3
Solubility (g/L) at 20-25°C
(HSDB 2009; IPCS 2001)
Insoluble
8
211
100
58 (IPCS 2001); 5.2 at 15°C (HSDB
2009)
Soluble, decomposes in presence of
moisture into vanadic acid and HC1.
535at20°C
(Rahman and Skyllas-Kazacos 1998)
Decomposes
Decomposes
Slightly soluble
Adapted from (Assem and Levy 2009).
This Toxicological Review focuses exclusively on vanadium pentoxide (¥265, CASRN
1314-62-1) (Figure 2-1), the most common form of vanadium used commercially. Vanadium
DRAFT - DO NOT CITE OR QUOTE
-------
pentoxide exists in the pentavalent state as a yellow-red powder (Table 2-2) (OSHA 2007).
Table 2-2: Chemical and physical properties of Vanadium Pentoxide
Characteristic
Chemical name
Synonym(s)
Chemical
formula
CASRN
Molecular weight
Color
Melting point
Boiling point
Density at 18 °C
Odor threshold:
Solubility:
Water
Organic
solvents
Vapor pressure
Specific Gravity
Flash point
Conversions:
ppm to mg/m3
mg/m3 to ppm
Information
Vanadium Pentoxide
Vanadium oxide, vanadic anhydride dust,
divanadium pentaoxide, divanadium pentoxide,
vanadium pentaoxide
V205
1314-62-1
181.9
Yellowish-red powder
Yellow to rust-brown orthorhombic powder
Yellow-orange powder or dark gray flakes
dispersed in air
690 °C
1750 °C
3.357
Odorless
8 g/L (20 °C)
10g/L(20°C)
Soluble in alkalies, concentrated acids, insoluble in
alcohol
0 mm Hg
3.4g/cm3
Not applicable, Non-combustible
1 ppm = 7.44 mg/m3
1 mg/m3 = 0.134 ppm
Reference
OSHA 2007
CAS
CAS
OSHA 2007
O'Neil2001;NIOSH
2005
OSHA 2007
OSHA 2007
ChemFinder.com, HSDB
2008, Lewis, 1997.
OSHA 2007, NIOSH
2005.
OSHA 2007
ChemFinder . com
O'Neil 2001; HSDB,
2008
HSDB 2008
NTP 2008
OSHA 2007
Calculated
Calculated
DRAFT - DO NOT CITE OR QUOTE
-------
3 TOXICOKINETICS
Toxicokinetics of vanadium pentoxide have been investigated in limited studies
described below and reviewed by Cooper (2007). Toxicokinetics of several other vanadium
compounds have been evaluated in animal models and are reviewed elsewhere (Sabbioni et al.
1996; Barceloux 1999; Mukherjee et al. 2004; ATSDR 2009). Vanadium pentoxide is rapidly
absorbed by both inhalation and oral exposures through lungs and the gastrointestinal tract,
respectively, although the amount absorbed through the gastrointestinal tract is low. Laboratory
animal studies show vanadium pentoxide is mainly distributed following inhalation and oral
exposure to the bone, lungs, liver and kidney. Elimination of vanadium pentoxide has been
studied only following inhalation exposure, and is mainly through the urine.
3.1 Absorption
3.1.1 Inhalation Exposure
Several occupational studies indicate that absorption can occur in humans following
inhalation exposure. An increase in urinary vanadium levels was found in workers
occupationally exposed to <1 ppm (<7.44mg/m3) of vanadium compounds, including vanadium
pentoxide (Gylseth et al. 1979; Kiviluoto et al. 1981a; Lewis 1959; NIOSH 1983), with the
majority excreted in urine within one day post long-term or moderate exposure to vanadium dust
(Kiviluoto et al. 1981a). The vanadium concentration in serum was higher than the
nonoccupationally exposed controls following exposure to vanadium pentoxide dust (Kiviluoto
etal. 1981b).
Indirect evidence of absorption of vanadium in animals is indicated in studies involving
inhalation exposure or intratracheal administration. In rats and mice exposed to 0.28-2.2 mg
vanadium/m3 as vanadium pentoxide1 for 14 days or 2 years (6 hours/day, 5 days/week),
marginal increases in blood vanadium levels were observed, suggesting that vanadium pentoxide
was poorly absorbed or rapidly cleared from the blood (NTP 2002; Dill et al. 2004); in the 2-year
studies by NTP (2002), the increase in blood vanadium levels were concentration-related,
although not statistically significant. Intratracheal studies suggest that vanadium pentoxide is
ny studies describe exposures in terms of concentration of vanadium, particularly when describing exposure to
mixtures. When possible, a concentration is given as amounts of vanadium pentoxide. As listed here (mg
vanadium/m3 as vanadium pentoxide) shows exposure was to vanadium pentoxide, with data shown here in
concentration of vanadium.
6
DRAFT - DO NOT CITE OR QUOTE
-------
readily absorbed through the lungs. The greatest absorption of a radioactive dose, 48V, was
found to occur 5 minutes after administration in albino rats (gender not specified) (Roshchin et
al. 1980). Most of the vanadium, i.e., 80 and 85% of the tetravalent (V4+) and pentavalent (V5+)
forms of vanadium, respectively, cleared from the lungs 3 hours after intratracheal exposure in
male abino rats (Edel and Sabbioni 1988). After 3 days, 90% of vanadium pentoxide was
eliminated from the lungs of female rats following intratracheal instillation (Conklin et al. 1982).
In an intratracheal instillation study in female Fischer rats 50% was cleared in 18 minutes, and
the rest within a few days (Rhoads and Sanders 1985).
Wallenborn et. al. (2007) analyzed the components of a complex particulate matter
mixture into its metal components and tracked the absorption of different metals in different
tissues following a single intratracheal dose in rats. Healthy male Wistar rats were instilled with
a single intratracheal dose of combustion particulate matter (PM) containing a moderate amount
of transition metals, including vanadium. The composition of vanadium in the PM was 62.95
Hg/mg, with 7.18 ng/mg in the water soluble fraction, 26.50 ng/mg in the acid soluble fraction
and 29.27 ng/mg in the insoluble fraction. According to calculations, of the 196.63 |j,g/rat of
vanadium instilled (theoretical), 110.32 |j,g/rat was measured in lung 4-hrs post-instillation and
62.76 |j,g/rat was measured in lung 24-hrs post-instillation. In the plasma and lung, vanadium
was significantly elevated 4-hr post-instillation (130,000 ng V/g lung tissue, and 350 ng V/g
plasma) compared to 24-hrs (60,000 ng V/g lung tissue, and 110 ng V/g plasma, respectively)
suggesting rapid uptake of water-soluble vanadium. The vanadium component of PM was
tracked to lung, plasma, heart and liver and was significantly increased compared to controls at
both 4- and 24-hrs post-instillation compared to saline controls. This study permitted detectable
changes in component metals of a complex mixture in various organs and provides evidence that
metals dissociate from particulate matter and translocate to various target organs, depending on
solubility (Wallenborn et al 2007).
3.1.2 Oral Exposure
No studies were available in the published literature regarding the rate and extent of
absorption in humans after oral exposure to vanadium pentoxide.
The absorption of vanadium through the gastrointestinal tract of animals is low. Less
than 0.1% of an intragastric dose was detectable in the blood of albino rats at 15 minutes post-
exposure, and less than 1% at 1 hour (Roshchin et al. 1980). Similarly, only 2.6% of an orally
administered radiolabeled dose of vanadium pentoxide was absorbed 3 days after exposure in
female Fischer rats (Conklin et al. 1982). Vanadium was reported in tissues and urine of male
7
DRAFT - DO NOT CITE OR QUOTE
-------
albino rats within hours after a single oral dose (Edel and Sabbioni 1988), suggesting that it is
rapidly absorbed. Young rats that consumed vanadium in the drinking water and feed were
found to have higher tissue vanadium levels 21 days after birth than they did 115 days after birth
(Edel et al. 1984). The data suggest that there is a higher absorption of vanadium in these young
animals due to a greater nonselective permeability of the undeveloped intestinal barrier. Thus,
age of the rodents appears to play an important role in the absorption of vanadium in the
gastrointestinal tract.
3.1.3 Dermal Exposure
No specific studies were available in the published literature regarding absorption in
humans or animals after dermal exposure to vanadium pentoxide, although absorption by this
route is generally considered to be very low (WHO 1988). Vanadium is a metal with low
solubility, therefore absorption through the skin is thought to be minimal.
3.1.4 Other Routes of Exposure
No studies were available in the published literature regarding the nature and extent of
absorption in humans after other routes of exposure to vanadium pentoxide.
3.2 Distribution
Distribution was measured from autopsy cases with unknown routes of exposure.
Vanadium has been detected in the lungs (in 52% of the cases) and intestines (in 16% of the
cases) of humans with no known occupational exposure, collected from autopsy data and
reviewed in Schroeder et al. (1963). In the gastrointestinal tract, it was primarily found in the
ileum (37%), cecum (45.1%), sigmoid colon (15.9%), and rectum (26.2%). Most positive
samples had 0.01|j,g or less per g of tissue. The heart, aorta, brain, kidney, muscle, ovary, and
testes were found to have no detectable vanadium concentrations.
3.2.1 Inhalation Exposure
There are limited data on the distribution of vanadium in workers; serum vanadium levels
in workers were highest within a day after inhalation exposure followed by a rapid decline in
levels upon cessation of exposure (Gylseth et al. 1979; Kiviluoto et al. 1981b). Analytical
studies have shown low levels of vanadium in human kidneys and liver, with even less in brain,
8
DRAFT - DO NOT CITE OR QUOTE
-------
heart, and milk. Higher levels were detected in hair, bone, and teeth (Byrne and Kosta 1978).
Inhalation exposure and intratracheal administration studies have examined the distribution of
vanadium in rodents. In F344 rats chronically exposed to 0.56 or 1.1 mg vanadium/m3 as
vanadium pentoxide (6 hours/day, 5 days/week), vanadium lung burdens peaked after 173 days
of exposure and declined until 542 days; lung levels never reached steady state (NTP 2002). In
contrast, lung burdens appeared to reach steady state by exposure day 173 in rats exposed to 0.28
mg vanadium/m3 (NTP 2002). Similarly, lung burdens did not reach steady state in B6C3F1
mice exposed to 1.1 or 2.2 mg vanadium/m3 as vanadium pentoxide, 6 hours/day, 5 days/week
for 542 days (NTP 2002). Rather, lung burdens peaked near day 54 and declined through day
535. Steady state was achieved in mice exposed to 0.56 mg vanadium/m3 during the first 26
days of exposure.
Vanadium is found to have a two-phase lung clearance after a single acute exposure in
both male Wistar rats and female Fischer rats (Oberg et al. 1978; Rhoads and Sanders 1985).
The initial phase is rapid with a large percentage of the absorbed dose distributed to most organs
and blood 24 hours postexposure, followed by a slower clearance phase. Vanadium is
transported mainly in the plasma. It is found in appreciable amounts in the blood initially and
only at trace levels 2 days after exposure (Roshchin et al. 1980). The pentavalent and tetravalent
forms of vanadium compounds were found to have similar distribution patterns in male albino
Sprague-Dawley rats (Edel and Sabbioni 1988). Three hours after exposure to the pentavalent or
tetravalent form, 15-17% of the absorbed dose was found in the lung, 2.8% in the liver, and 2%
in the kidney (Edel and Sabbioni 1988). After intratracheal instillation of pentavalent vanadium,
retention of vanadium was observed in lungs, liver, kidneys, bone, testes and spleen with
clearance at different timepoints post-exposure with little to no retention observed in stomach,
intestines, heart or trachea (Edel and Sabbioni 1988). This is similar to the distribution seen
following inhalation and oral exposure.
3.2.2 Oral Exposure
No studies were available in the published literature regarding distribution in humans
after oral exposure to vanadium pentoxide.
Acute studies with rats showed the highest vanadium concentration in the skeleton. Male
rats had approximately 0.05% of the administered 48V in bones, 0.01% in the liver, and <0.01%
in the kidney, blood, testis, or spleen after 24 hours (Edel and Sabbioni 1988). Other authors
who found that the bone had the greatest concentration of radiolabeled vanadium, followed by
the kidney (Roshchin et al. 1980), noted similar findings. Conklin et al. (1982) reported that
9
DRAFT - DO NOT CITE OR QUOTE
-------
after 3 days, 25% of the absorbed vanadium pentoxide was detectable in the skeleton and blood
of female Fischer rats.
Oral exposure for an intermediate duration produced the highest accumulation of
vanadium in the kidney. In young male rats at 3 weeks of age, the kidneys, heart, and lungs had
the highest levels immediately following exposure (Edel et al. 1984). Vanadium in the kidney,
liver, and lung decreased significantly at 115 days of age. There was an accumulation in muscle
and fat, related to the growing mass of the tissues with age. The higher levels of vanadium in the
young rat tissues may be due to the higher retention capacity of the undeveloped tissues, or a
greater permeability of the intestinal wall. Adult rats exposed to 5 or 50 ppm vanadium in the
drinking water for 3 months had the highest vanadium levels in the kidney, followed by bone,
liver, and muscle (Parker and Sharma 1978). The retention in bone may have been due to
phosphate displacement. All tissue levels plateaued at the third week of exposure. A possible
explanation for the initially higher levels in the kidney during intermediate-duration exposure is
the daily excretion of vanadium in the urine. When the treatment is stopped, levels decrease in
the kidney. At the cessation of treatment, vanadium mobilized rapidly from the liver and slowly
from the bones. Other tissue levels decreased rapidly after oral exposure was discontinued.
Thus, retention of vanadium was much longer in the bones (Edel et al. 1984; Parker and Sharma
1978).
Radike et al. (2002) assessed the distribution of various metals, including vanadium, in
female B6C3F1 mice. Mice ingested either (1) a metal mixture containing Chromium (Cr),
Cadmium (Cd), Arsenic (As), Nickel (Ni) and Vanadium (V) in drinking water or (2) a metal
mixture containing Cr, Cd, As, Ni, and V in NIH-31 feed. In water and feed, the calculated
vanadium concentration in the mixture was 45 ppm and 1.105 ppm, respectively. Measured
vanadium levels in the small intestine were 10 ppm at 5 weeks and 14 ppm at 8 weeks, and were
significantly higher compared to controls than any other metal constituent in the small intestine.
Mainly, vanadium pentoxide is distributed to bone (10-25% of administered oral dose), liver
(-5%), and kidney (-4%). In addition, vanadium levels in the kidneys and the femur were
significantly greater than in controls at 4, 8, 12, 16, and 24 week following oral dosing.
Vanadium levels in small intestine and kidneys were lower in mice given vanadium as part of a
heterogeneous metal mixture in feed vs. water (Radike et al. 2002).
3.2.3 Dermal Exposure
No studies were available in the published literature regarding distribution in humans or
animals after dermal exposure to vanadium.
10
DRAFT - DO NOT CITE OR QUOTE
-------
3.2.4 Other Routes of Exposure
After intraperitoneal administration to rats, vanadium is distributed to all organs. After
24 hours, the highest concentrations were found in the bones and kidney, although initial levels
were highest in the kidney (Roshchin et al. 1980; Sharma et al. 1980).
3.3 Metabolism
Vanadium is an element, and as such, is not metabolized. In the oxygenated blood, it
circulates as a polyvanadate (isopolyanions containing pentavalent vanadium) but in tissues, it is
retained mainly as the vanadyl cation (cationic form of tetravalent vanadium). Depending on the
availability of reducing equivalents (such as reduced glutathione GSH, NADPH, NADH) and
oxygen, vanadium may be reduced, reoxidized, and/or undergo redox cycling (Byczkowski and
Kulkarni 1992).
3.4 Elimination and Excretion
3.4.1 Inhalation Exposure
Occupational studies showed that urinary vanadium levels significantly increased in
vanadium pentoxide exposed workers (Gylseth et al. 1979; Kiviluoto et al. 1981a; Lewis 1959;
NIOSH 1983; Zenz et al. 1962). Male and female workers exposed to 0.1-0.19 mg/m3 vanadium
in a manufacturing company, had significantly higher urinary levels (20.6 ug/L) than the
nonoccupationally exposed control subjects (2.7 ug/L) (NIOSH 1983). The correlation between
ambient vanadium levels and urinary levels of vanadium is difficult to determine from these
epidemiological studies (Kiviluoto et al. 1981b). In most instances, no other excretion routes
were monitored. Analytical studies have shown very low levels in human milk (Byrne and
Kosta 1978). Evidence from animal studies supports the occupational findings. Vanadium
administered intratracheally to rats was reported to be excreted predominantly in the urine
(Oberg et al. 1978) at levels twice that found in the feces (Rhoads and Sanders 1985). Three
days after intratracheal exposure to radiolabeled vanadium pentoxide, 40% of the recovered 48V
dose was cleared in the urine while 30% remained in the skeleton, and 2-7% was in the lungs,
liver, kidneys, or blood (Conklin et al. 1982).
In female rats exposed to 0.56 or 1.1 mg vanadium/m3 as vanadium pentoxide for 16 days
11
DRAFT - DO NOT CITE OR QUOTE
-------
(6 hours/day, 5 days/week), lung clearance half-times during an 8-day recovery period were 4.42
and 4.96 days, respectively (NTP 2002). In mice similarly exposed to 1.1 or 2.2 mg
vanadium/m3 as vanadium pentoxide, lung clearance half-times were 2.55 and 2.40 days,
respectively (NTP 2002). In contrast to the 16-day exposure data, the lung clearance half-times
in female rats exposed to 0.28, 0.56, or 1.1 mg vanadium/m3 for 2 years (6 hours/day, 5
days/week) were 37.3, 58.6, and 61.4 days, respectively (NTP 2002). In mice, the half-times
were 6.26, 10.7, and 13.9 days at 0.56, 1.1, and 2.2 mg vanadium/m3 exposure levels (NTP
2002).
After intratracheal instillation of pentavalent vanadium, clearance from lungs was
initially rapid (3h) but with some vanadium (2% original dose) remaining at 12 d post-exposure.
All other tissues eliminated 98-99% of original dose by 3h post-exposure (Edel and Sabbioni
1988). Epidemiological studies and animal studies suggest that elimination of vanadium
following inhalation exposure is primarily in the urine.
3.4.2 Oral Exposure
No studies were available published literature regarding excretion in humans or
laboratory animals after oral exposure to vanadium pentoxide.
3.4.3 Dermal Exposure
No studies were available published literature regarding excretion in humans or
laboratory animals after dermal exposure to vanadium pentoxide.
3.4.4 Other Routes of Exposure
No studies were available published literature regarding excretion in humans or
laboratory animals after other routes of exposure to vanadium pentoxide.
3.5 Physiologically Based Pharmacokinetic (PBPK)/Pharmacodynamic (PD) Models
No PBPK models for vanadium pentoxide are available.
3.5.1 Animal-to-Human Extrapolations
There are no relevant data available to evaluate potential toxicokinetic differences
12
DRAFT - DO NOT CITE OR QUOTE
-------
between humans and laboratory animals. Similar effects have been reported in humans and
animals following inhalation or oral exposure to vanadium pentoxide; however, this conclusion
is based on the limited human toxicity data. In absence of data to the contrary, rats or mice
appear to be valid models for extrapolation to humans.
13
DRAFT - DO NOT CITE OR QUOTE
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS - EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS
4.1.1 Oral Exposure
Few relevant studies investigating the effects of acute, subchronic or chronic oral
exposure to vanadium pentoxide in humans were identified in the peer-reviewed literature. One
study, Kucera et al. (1992), measured vanadium in the hair and blood of children exposed to
vanadium by accidental drinking of contaminated water near a vanadium pentoxide plant.
Vanadium levels in the hair did not differ significantly for the control and exposed groups. An
increase in vanadium concentrations was found in the blood of exposed children (median: 0.078
Hg/L) compared to control children (median: 0.042 ng/L). No exposure-response relationship
could be quantified. These results suggest that the vanadium levels in blood, but not hair is a
sensitive or suitable indicator of environmental exposure.
4.1.2 Inhalation Exposure
Health effects of inhalation exposure to vanadium pentoxide and other vanadium
compounds reported by case studies and epidemiological investigations include respiratory tract
irritation, bronchitis (often called boilermakers' bronchitis), airway obstruction, chest pain,
rhinitis, pharyngitis, laryngitis and conjunctivitis in workers exposed to vanadium-containing
dust during vanadium processing (Sjoberg, 1951; Sjoberg, 1956; Zenz et al., 1962; Kiviluoto et
al., 1979; Kiviluoto, 1980; Kiviluoto et al., 1981; Musk and Tees, 1982; Irsigler et al., 1999) or
to fuel-oil ash containing vanadium during cleaning and maintenence of oil-burning boilers
(Williams, 1952; Sjoberg, 1955; Lees, 1980; Ross, 1983; Levy et al., 1984; Hauser et al., 1995a;
Hauser et al., 1995b; Woodin et al., 1998; Woodin et al., 1999; Woodin et al., 2000; Hauser et
al., 2001; Kim et al., 2004). Most of these studies did not quantify the inhalation exposure
specifically to vanadium pentoxide, although it was the primary exposure in vanadium pentoxide
production. Exposures to residual oil fly ash (ROFA) involves a mixture of pollutants including
elemental vanadium, vanadium oxides, vanadium pentoxide, vanadium sulphates, particulate
matter and other metal constituents (Hauser et al., 1995b). Vanadium pentoxide is a major
14
DRAFT - DO NOT CITE OR QUOTE
-------
constituent of ROFA and the studies of health effects among boilermakers evaluated models in
relation to vanadium content of respirable particulate matter. Therefore, these papers serve to
inform the assessment of the nature and scope of the health response to vanadium pentoxide.
More recently, epidemiology studies evaluated the metals content of ambient particulate matter
(PM) and found that communities in the United States with higher vanadium content in PM have
higher PM-related risk of mortality or hospitalizations for cardiovascular or respiratory disease
(Lippmann et al., 2006; Dominici et al., 2007; Bell et al., 2009; Patel et al., 2009; Lagorio et al.,
2006). Because exposures to vanadium, not vanadium pentoxide, were evaluated in the studies
of ROFA or PM air pollution, information to characterize the exposure-response relationship
between inhaled vanadium pentoxide alone and adverse health effects in humans is limited.
Moreover, both ROFA and PM are mixtures with several components that also may contribute to
observed health effects.
4.1.2.1 Controlled Human Exposure Study
Zenz and Berg (1967) exposed nine volunteers to vanadium pentoxide dust to evaluate
respiratory effects. Volunteers (gender not reported) were exposed to 0.1 mg/m3 (n=2),
0.25 mg/m3 (n=5) or 1 mg/m3 (n=2) vanadium pentoxide in an environmental chamber for 8 hrs;
no control group was included in this study. Particle size analysis revealed that 98% of particles
had a diameter <0.5 jim. Post-exposure assessments of chest x-ray, blood, urine, nasal smear
samples and pulmonary function were compared with baseline values determined for each
subject prior to exposure. All subjects were observed for clinical symptoms for 11-19 months
after exposure. Subjects exposed to 1 mg/m3 vanadium pentoxide developed sporadic cough
after 5 hrs of exposure, which progressed to persistent cough during the last 3 hrs of exposure
and continued for 8 days. No other signs of respiratory irritation were observed. Results of
pulmonary function tests and chest x-ray 1, 2 and 3 weeks after exposure were similar to
baseline (data not reported). Hematology and urinalysis parameters were not affected by
exposure (data not reported). Nasal smears obtained 24 hrs, 72 hrs and 1 week after exposure
were negative for eosinophilia. Three weeks after the initial exposure, two of the subjects were
accidentally exposed to a "heavy cloud" of vanadium pentoxide dust (concentration not reported)
for 5 minutes. Within 16 hrs of exposure, both subjects developed a "marked" productive cough
with rales and expiratory wheeze, which continued for 1 week. However, pulmonary function
test results were comparable to baseline (data not reported). Blood and nasal smear samples
were negative for eosinophilia. Subjects in the 0.25 mg/m3 exposure group developed a "loose"
productive cough on the day after exposure, which lasted for 7-10 days. No additional
15
DRAFT - DO NOT CITE OR QUOTE
-------
symptoms were observed and post-exposure pulmonary function and laboratory tests were
comparable to baseline results (data not reported). Subjects exposed to 0.1 mg/m3 developed
"considerable" mucus formation within 24 hrs after exposure, lasting for 4 days. No other
symptoms or positive findings for pulmonary function or laboratory tests were observed. No
treatment-related symptoms or clinical findings were reported for any subject during the
11-19 months post-treatment period.
4.1.2.2 Occupational Exposure during Vanadium Pentoxide Mining and Processing
Respiratory and other symptoms have been documented among workers employed at
facilities producing and processing vanadium pentoxide (Sjoberg, 1951; Sjoberg, 1956; Zenz et
al., 1962; Kiviluoto et al., 1979; Kiviluoto, 1980; Kiviluoto et al., 1981a; Musk and Tees, 1982;
Irsigler et al., 1999). Sjoberg (1951) reviewed earlier reports of symptoms among workers
exposed to vanadium pentoxide and described symptoms among 36 employees (foremen,
workers, builders/repairers) at a X^Os production factory in Falun, Sweden that began operation
in 1946. The employees experienced symptoms on one or multiple occasions while under his
medical surveillance between 1947 and 1950. Symptoms generally persisted an average of 13
days. Air samples in different parts of the vanadium pentoxide facility were found to contain
dust concentrations of 0.6 - 86.9 mg/m3 during pulverization of iron ore slag with a vanadium
content of 4.8 to 7.5%. A significant proportion of the dust consisted of small, respirable
particles (22% < 8|i, 39% < 12|i). Although vanadium pentoxide concentrations in air were not
reported, vanadium was detected in the blood and urine of 23 and 27 individuals, respectively.
Sjoberg (1956) reported the main findings of his medical surveillance, including those of
a thorough examination of the cohort in October, 1948, comparing them to an external referent
population, a group of 703 workers from mines and sawmills in northern Sweden followed
during the same time period and examined using the same methods. The authors assumed that
the comparison population was exposed only to inert dust, however no sampling data was
reported. The main symptoms of the upper respiratory tract were nasal irritation and/or nasal
catarrh (inflammation of mucus membranes; 42% versus 20% among unexposed) and throat
dryness and pain (86% versus 8% among unexposed). These symptoms were reported to be
more prevalent at the final examination in 1948 and included acute and chronic pathological
changes in the nose and pharynx. Lower respiratory tract symptoms included cough with no
sputum (61%), cough with sputum (39%), wheezing (86%) and shortness of breath (75%). In
the comparison group, prevalence of coughing and shortness of breath was 4% and 24%,
respectively. Bronchoscopy, measured in five individuals, revealed no severe changes in four
16
DRAFT - DO NOT CITE OR QUOTE
-------
cases and mild chronic bronchitis in one person. Acute changes in the lungs indicative of
pneumonitis were noted in five workers. Spirometric measures were reported to be higher than
those of the comparison group. A decrease in hemoglobin levels (7.8 ± 2.36% compared to 3.0 ±
2.07%) and red blood cell count was observed, particularly among workers who were
permanently or frequently employed at the plant during the observation period, although levels
remained within normal limits. The authors did not observe leucopenia or eosnophilia in blood
samples. Six cases of dermatitis were observed, and one person had a positive response to a
patch test with a sodium vanadate solution. Finally, weakness and fatigue were reported to be
common symptoms after heavy exposure to dust, and symptoms of a neurasthenic character were
noted in some cases.
Six workers with persistent symptoms were re-examined in December, 1953 to January,
1954 (Sjoberg, 1956). Reported symptoms included cough and wheezing (N=5), and dyspnea
and fatigue (N=6). Clinical measurements and blood pressure were normal including sinuses,
circulating eosinophils, erythrocyte sedimentation rate, and radiographic examination of the
lungs and heart revealed no evidence of pneumoconiosis. Lung function measures had increased
which the authors attributed to differences in technique. However, bronchoscopic examination
and biopsy of bronchial mucosa showed evidence of chronic bronchitis (N=5).
Two additional published case summaries described symptoms that appeared among
workers after the start of new plant operations producing vanadium pentoxide pellets (N=18)
(Zenz et al., 1962) or refining vanadium pentoxide (N=4) (Musk and Tees, 1982). Both
exposures were to dry vanadium pentoxide powder at high concentrations (> 0.5 mg/m3).
Symptoms similar to those reported by Sjoberg (1956) appeared by the end of the first day of
exposure including eye and throat irritation and cough with sputum, and persisted 2-4 weeks
after exposure ended. Lung function tests revealed reversible airflow obstruction in 3 of the 4
case histories reported by Musk and Tees, and bronchial reactivity to histamine was
demonstrated in two of the four. One individual with a family history of asthma and positive
skin prick tests to common allergens, continued to experience wheezing for 8 weeks. Zenz et al
(1962) did not find reductions in airflow among the 18 men examined.
Kiviluoto et al. (1979; 1980; 1981a, b) published a series of reports regarding an
investigation in 1975 of respiratory symptoms and clinical findings among employees (process
workers, repairmen, foremen, and a laboratory worker) at a factory making vanadium pentoxide
from magnetite ore. A total of 79 men employed at the vanadium factory > 4 months were
eligible and 63 men, aged 19-52 years, who were not on holiday or sick leave were enrolled
(80% participation rate). A referent group of 63 men living in the same area were selected from
workers at the magnetite ore mine (concentrating plant, the mine, the repair shop, and the office)
17
DRAFT - DO NOT CITE OR QUOTE
-------
and matched for age (within 2 years) and smoking habit (within 5 cigarettes/day). Vanadium dust
concentrations at various sites measured on 8 days over 2 shifts in March-May 1976, were 0.012
mg/m3 ranging between 0.002 (LOD) and 0.043 mg/m3. Vanadium concentrations in the
breathing zone were 0.028 mg/m3 (TWA) with a range of 0.002 - 0.42 mg/m3. Higher
concentrations were found where grinding and packing of smelt were conducted (TWA (range):
2.3 mg/m3 (one sample) and 0.13 mg/m3 (0.02 - 0.37). Air monitoring results were not reported
for the referent population. Urinary vanadium concentrations among the exposed averaged 0.26
±0.17 umol/L (18 hour excretion), while concentrations among the referent group did not
exceed the < 0.04 umol/L limit of detection (LOD).
Clinical assessments were conducted by health personnel with no knowledge of exposure
status. The occurrence of wheeze differed by exposure. Workers who reported wheezing were
twice as likely to work in the vanadium factory (p<0.05) (Kiviluoto, 1980). Prevalence of nasal
catarrh, cough, phlegm, or other respiratory symptoms did not differ between the exposed and
referent groups. Spirometric measurements (FVC, FEVi, adjusted for height), obtained at the
end of workers' summer holidays, did not differ between the exposed and referent groups
(p>0.01) and were not related to duration of exposure to vanadium dust (p>0.1). Inflammation
was observed in nasal smears and these results are discussed in the section on immunological
endpoints (Section 4.2.1). Nonfasting serum chemistry parameters, analyzed in May - June,
1975, when vanadium concentrations were higher (0.2-0.5 mg/m3) were compared between 16
exposed and 16 referent subjects. Among the several serum chemistry parameters tested, the
difference between the exposed and referent groups for serum albumin, chloride, urea, bilirubin
and conjugated bilirubin were statistically significant, although no values were outside the range
of reference values. Hematological results (nonfasting) for 63 exposed and 16 referents,
analyzed in March - May, 1976, when vanadium concentrations were lower (0.01-0.04 mg/m3),
did not vary by exposure group (p>0.05). In addition, there were no differences noted for serum
cholesterol, serum triglyceride, or leukocyte differential.
Irsigler et al. (1999) evaluated the clinical histories of 40 men, who were employed at a
vanadium pentoxide production plant in South Africa, and were referred by the plant's medical
staff for more detailed medical assessment because of persistence of respiratory symptoms
(cough, breathing difficulty, wheezing) between October 1995 - October 1997. Twelve men,
aged 19-60 years with bronchial hyperresponsiveness to inhaled histamine or exercise
challenge (out of 40 men referred) were selected for analysis along with 12 men, aged 24 - 54,
who were referred and did not have bronchial hyperresponsiveness, matched by age and
smoking. The authors concluded that the asthma symptoms and bronchial reactivity had
occurred as a result of vanadium exposure because all were free of current symptoms or a
18
DRAFT - DO NOT CITE OR QUOTE
-------
previous history of asthma when they began employment at the plant. Bronchial reactivity was
determined by the repeated inhalation of histamine at successively doubled doses until a
decrease of 20% or greater in FEVi was measured (Histamine PC20 FEVi). A PC20 > 8 mg/ml
was considered to be normal. Alternatively, an exercise challenge consisting of free tread mill
running indoors for 6 minutes was used to assess bronchial reactivity. A positive test was
defined as a decrease of 200 ml or a 15% or greater fall in FEVi. Vanadium pentoxide
concentrations in air from area samples were < 0.15 mg/m3 in the mills, kiln, leaching, and
pollution control areas, 1.53 mg/m3 in the fusion precipitation area, and 0.057 mg/m3 in the
ferrovanadium area. Concentrations of 862 and NFL? were above their recommended
occupational limits in the kiln area. Vanadium pentoxide in spot urine samples was detected in
10 of the 12 workers with bronchial reactivity (5.2 - 180 jig/g creatinine) and was above a level
considered to be toxic in 3 individuals (> 50 |ig/g creatinine). Levels in the 12 referent men
ranged between 12.0 - 55 |ig/g creatinine, with one person above 50 |ig/g creatinine. Among 9
subjects who returned for a follow-up examination after 5 to 23 months with no vanadium
pentoxide exposure, 8 still exhibited bronchial reactivity. Atopy was ruled out as a primary
cause because 5 individuals with positive skin prick allergy tests were equally distributed in the
reactive and nonreactive groups. In addition, no evidence of viral upper respiratory tract
infection was found in the study subjects, and 10 cigarette smokers were also distributed equally
in both groups. Work tasks did not appear to be different between the two groups. This small
study was not informative regarding associations between vanadium pentoxide exposure and
bronchial reactivity. Although the group with bronchial reactivity had two subjects with high
urine vanadium levels, differences in biomarker levels or job site between groups were not
analyzed statistically, and the sample size and study design were not adequate to determine an
association of vanadium pentoxide exposure with bronchial reactivity.
4.1.2.3 Occupational Exposure during Cleaning and Maintenance of Oil-Fired Boilers
Vanadium pentoxide is present in significant amounts along with other vanadium oxides,
vanadium sulphate and metals in ash that accumulate in oil- and coal-fired boilers, as well as
other fuel types used in boilers. Several reports of case histories and epidemiology studies of
boilermakers involved in the construction, cleaning and maintenance of oil-fired boilers have
described upper and lower respiratory symptoms similar to those reported among workers
processing vanadium pentoxide (Williams, 1952; Sjoberg, 1955; Lees, 1980; Ross, 1983; Levy
et al., 1984; Hauser et al., 1995a; Hauser et al., 1995b; Woodin et al., 1998; Woodin et al., 1999;
Woodin et al., 2000; Hauser et al., 2001; Kim et al., 2004). Additional health parameters have
19
DRAFT - DO NOT CITE OR QUOTE
-------
been investigated including pulmonary function (Lees, 1980; Levy et al., 1984; Hauser et al.,
1995a; Woodin et al., 1999; , biomarkers of inflammation in nasal fluid (Hauser et al., 1995b;
Woodin et al., 1998), and autonomic cardiac function (Magari et al., 2002). Studies have
investigated acute effects occurring after jobs cleaning or overhauling boilers lasting a few days
to several weeks and chronic conditions among boilermakers who had worked in that occupation
for several years.
Several case summaries described the health response of workers cleaning oil-fired
boilers in Great Britain (Williams, 1952; Ross, 1983), Sweden (Sjoberg, 1955), and Canada
(Lees, 1980). Two case series reports described exposure to dust during cleaning jobs of a few
days to one week during 1946 - 1953. The onset of symptoms, including rhinorrhoea, sneezing,
eye irritation, sore throat, and chest pain, began within 1 to 12 hours (Williams, 1952; Sjoberg,
1955). Symptoms with a delayed onset of 6-24 hours included a dry cough becoming
paroxysmal and productive in some workers, wheezing, dyspnea upon exertion, and fatigue
(Williams, 1952; Sjoberg, 1955). Bronchial irritation, bronchitis, and the development of rales
in regions of the lung of some of the workers also were reported. Some workers also developed
a greenish-black coating on the tongue. Symptoms were reported to persist for 3 days to one
week after exposure was ended. Concentrations of vanadium pentoxide particles (10-20 u^ in
diameter) in the air inside the boilers during cleaning were 17 - 85 mg/m3.
Lees (1980) reported on a clinical evaluation of 17 men occupationally exposed to
bottom ash from cleaning the boiler of an oil-fired electricity generating station. Personal
sampling in the breathing zone of four of the men indicated a mean time weighted average of
523 ug/m3 (0.52 mg/m3) for dust under lOu. Vanadium content in the bottom ash and crusted
deposits was 15.3% and 24.2 - 35%, respectively. The men wore cartridge filter type respirators
during the cleaning operation, however subsequent testing documented that they leaked up to
9%. The air concentrations of vanadium pentoxide were not specifically reported. Symptoms
and lung function after exposure were compared to health status assessed before exposure began.
A medical history and clinical examination, performed before work began and the day after,
indicated symptoms similar to those described in earlier reports, including cough with sputum,
respiratory wheeze, and sore throat reported in 77%, 53%, and 41% of the men, respectively.
Serial spirometry measurements at 24 hour intervals showed reductions with the lowest mean
FVC and FEVi values at postexposure days 3 and 2, respectively, compared to baseline (percent
of baseline: 88.6% and 86.6%, respectively; p<0.05). Reductions in forced mid-expiratory flow
of 9-31% also were recorded. These measures had returned to preexposure levels after 4 weeks.
No details were provided regarding the years the study was conducted and the characteristics of
the men under study.
20
DRAFT - DO NOT CITE OR QUOTE
-------
In 1981, the Occupational Safety and Health Administration conducted an investigation
of work-related bronchitis among 100 boilermakers exposed to vanadium pentoxide during an oil
to coal conversion of a utility company power plant in western Massachusetts (Levy et al.,
1984). The conversion occurred over the course of approximately 6 weeks, October 15 -
November 30, with most of the men working 10 hour days, six days per week. Air samples
obtained in the boiler at approximately 4 weeks during the conversion indicated vanadium
pentoxide fume concentrations of 0.05 - 5.3 mg/m3. Concentrations of chromium, nickel, and
fumes of copper and iron oxide were stated to be within acceptable limits. Nitrogen dioxide and
hydrogen sulfide were not detected. Low concentrations of carbon dioxide (<5 ppm) and ozone
(< 0.1 ppm) were measured. Sulfur dioxide (< 1 ppm) was measured in the boiler during
welding operations and outside the boiler (1-35 ppm) when expansion joints were cut with a
torch.
In early December, a questionnaire was distributed to all 100 workers through the union
president and responses were received from 55 men over the next two months. All of the
respondents, aged 23 - 60 years, reported symptoms, with over half describing cough with
sputum (85%), sore throat (76%), dyspnea on exertion (71%), chest pain or discomfort (65%),
headache (56%), runny nose or sneezing (56%), wheezing (55%) and tiredness (51%). The
median time to onset was 7 days with clustering at 0 - 4 days and 6-8 days. When the
questionnaires were completed, symptoms had resolved or were improving in 41 of the 55
respondents. Although three-fourths of the respondents stated that they had used a respirator
over half the time when in the boiler, more than half stated that the respirator used was a paper
mask. Respondents had been boilermakers a median of 10 years. Pulmonary function tests were
performed on 35 individuals after they visited a physician. Median FVC was 87% of predicted,
but was < 80% of predicted in 5 of 27 individuals. Median FEVi was 93% of predicted, but was
< 80% of predicted in 8 of 27 men. Median FEVi/FVC was 79% of predicted. Median FEF25.
75% among 24 workers tested was 57% of predicted with only 4 of 24 above 80% of predicted.
FEF25-75% was not correlated with smoking history among the 69% of workers for whom this
information was obtained.
The symptoms and effects on lung function are consistent with previous reports of
occupational exposures during the cleaning and maintenance of boilers that had burned oil
contaminated with vanadium pentoxide. A marked deficit in pulmonary function, particularly in
FEF, was observed in some of the workers. However, pulmonary function was assessed in only
60% of respondents by several different health providers. In addition, this study is limited by a
lack of baseline information on health status or comparison to a comparable occupational group,
a relatively low response among the exposed workers, and data collection days to weeks after
21
DRAFT - DO NOT CITE OR QUOTE
-------
exposure was discontinued.
A subsequent study of lung function among boilermakers before and after four weeks of
work overhauling an oil-fired boiler did not observe an association with respirable vanadium
dust concentrations (Hauser et al., 1995a). A total of 36 out of 80 eligible workers completed a
baseline test, and 26 completed a postexposure test. The men averaged 42.5 years of age (27 -
60) and 16.9 years on the job (6 months - 35 years). Daily exposure estimates were developed
for each subject based on work diaries detailing tasks and locations and personal sampling.
Between 1 and 10 hour time weighted average sampling was available for 15% of the total
number of study days and these data were applied to the task/location information to assign
exposure levels for PMio and vanadium dust for each worker. Lung function values were
analyzed in relation to three exposure indices for the exposure period; average or peak
concentrations, and concentrations on the day of the postexposure test. Average, peak and mean
day-of vanadium concentrations in particulate matter < 10 jim (range) were estimated to be 12.2
± 9.1 ug/m3 (2.2 - 31.3), 20.2 ±11.4 ug/m3 (2.2-32.2) and 12.1 ± 10.9 (1.6-31.1), respectively.
For the spirometric measures, the largest value from three acceptable curves was used in the
analysis. Reductions in several lung function measures were observed over the average of 27
±4.1 days between the baseline and postexposure tests that were statistically significant. FEVi,
FVC, and FEF25-75% decreased by an average of 140 ± 160 ml (range: -390-420), 140 ± 200 ml
(range: -580-320) and 270 ± 450 ml/s (range: -1170-870), respectively. Each lung function
index was adjusted by dividing the change by the average of the pre- and post-exposure value.
The adjusted value was analyzed for associations with exposure in multiple linear regression
models adjusting for age and current smoking status. Peak PMio was inversely associated with
adjusted AFEVi (p=0.03), AFEV25%, (p=0.07), AFEV50%, (p=0.01), AFVC (p=0.01), but not
AFEV25_75o/0, (p=0.23) and &FEV75% (p=0.43). However, mean, peak, and day-of respirable
vanadium dust concentrations were not associated with any spirometric indices (the data were
not presented). PMio and vanadium dust exposure also were not related to bronchial reactivity as
measured with methacholine challenge tests before and after the overhaul. The authors noted
that the concentrations of vanadium dust were low and the variation in the range of
concentrations may not have been wide enough to detect a relation with lung function in this
small sample. Alternatively, the deficits in lung function may have been caused by a different
constituent in PMio.
Woodin et al (1998, 1999, 2000) described in a series of reports a prospective clinical
study that evaluated health measures among 18 boilermakers and compared them to 11 utility
workers involved in the overhaul of a large, oil-fired boiler over a six week period from mid-
May, 1995 to late-June, 1995. The men had volunteered for the study and did not have allergic
22
DRAFT - DO NOT CITE OR QUOTE
-------
symptoms two weeks prior to or during the overhaul. Data from one person who presented with
cold or flu symptoms at one of the clinical assessments were excluded. All were white men aged
26 - 61 years and were employed a mean of 20.5 years (range: 3 - 39). Vanadium and PMio
exposure was calculated for each subject for each work day using information on task duration
and location and use of personal protective equipment from job diaries, and PMio or vanadium
concentrations from personal exposure monitors in the breathing zone. Personal and stationary
sampling of PMio (< 10 um) was conducted over 10-12 hour shifts in major work areas in and
around the boiler. Environmental concentrations were compared before and during the boiler
overhaul, and between the two occupational groups. Utility workers did not enter the boiler
during the overhaul. Vanadium levels before the boiler work was comparable for boilermakers
and utility workers (geometric mean (SD) ug/m3: 1.2 (1.4) and 1.1 (1.2), respectively). During
the boiler work, vanadium levels rose to a geometric mean (SD) of 8.9 (2.3) ug/m3 inside the
boiler but did not change appreciably outside the boiler (geometric mean (SD) ug/m3: 1.4 (1.6)
(p < 0.001)). Exposure estimates were adjusted for the type of protective gear worn and its
duration to calculate individual daily dose. The daily dose to the upper and lower airway was
estimated using values for minute volume, penetration and deposition rates, and particle size.
Quartiles of lung vanadium dose (ug) were < 0.90, > 0.90 - < 5.30, > 5.30 - < 22.30 and > 22.30.
Quartiles of nasal vanadium dose (ug) were < 2.50, > 2.50 - < 23.20, > 23.20 - < 68.30 and >
68.30.
The workers recorded symptoms five times per day in a log and scored them for severity
with a numerical score from 0 to 3. The highest severity score for each day was used in the
analysis. Incidence of upper airway symptoms (nasal congestion/irritation, throat irritation) was
67% (12/18) among boilermakers and 36% (4/11) among utility workers. The incidence of
lower airway symptoms (chest tightness, wheeze, cough, and sputum production) was 72%
(13/18) among boilers and 27% (3/11) among utility workers. Robust regression models of
lower airway maximum severity scores and average symptom frequency in relation to quartiles
of lung vanadium dose indicated a dose-related increase. Maximum lower airway severity
scores were increased by 0.47 (p=0.01), 0.86 (p<0.01) and 0.24 (0.10) in quartiles of lung
vanadium dose 2, 3, and 4 compared to 1. Average lower airway frequency was increased by
0.19 (p=0.02), 0.39 (p<0.01) and 0.14 (0.07) in quartiles of lung vanadium dose 2, 3, and 4
compared to 1. The regression models were adjusted for current smoking.
Lung function was assessed on three occasions: before the overhaul, during the overhaul
(before the shift on the last day) and 2 weeks after the work ended (Woodin et al., 1999). The
highest value from three acceptable curves obtained during each test was used in repeated
measures analysis of variance to test for differences over time. No changes in the four airflow
23
DRAFT - DO NOT CITE OR QUOTE
-------
measures, FEF25, FEF50, FEF75, and MMEF, were observed over the course of the study. Mean
(SD) FEVi (1) values before, during and after the overhaul were 3.73 (0.61), 3.76 (0.54) and 3.65
(0.42), respectively. Mean (SD) FVC (1) values before, during and after the overhaul were 5.01
(0.67), 4.94 (0.61) and 4.92 (0.55), respectively. Change in lung function from the beginning to
the end of the overhaul was not associated with upper or lower airway dose levels of either
vanadium or PMio when assessed in linear regression models adjusting for smoking and age. In
addition, mean dose estimates were not different between individuals who experienced a loss of
either FEVi or FVC > 100 ml or those who experienced no change or an increase (two-sample li-
test).
Boilermakers were exposed to higher PMio concentrations compared to utility workers
both before (geometric mean (SD): 0.40 (1.60) versus 0.10 (2.70), p < 0.05) and during the
overhaul (geometric mean (SD): 0.47 (1.90) versus 0.13 (4.00), p < 0.001). In contrast to the
elevations in vanadium concentrations measured during the overhaul, PMio concentrations did
not increase appreciably (p > 0.05). During the boiler overhaul, ozone concentrations increased
somewhat outside the boiler, but did not change inside the boiler. The authors considered the
levels of other metals (cadmium, chromium, manganese, lead, arsenic, and nickel) to be low. All
samples were 1-3 orders of magnitude below the 1996 TLV, and no significant changes were
observed during the overhaul. Analyses of lung function and other pollutants were not reported.
Reductions in lung function over two years were investigated by Hauser et al. (2001) in a
longitudinal study of boilermaker construction workers exposed to combustion particles from
multiple sources including powerplants (oil, coal, natural gas), trash incinerators, paper mill
incinerators and other industrial sources with boiler, vessels and tanks requiring maintenance and
repair. A total of 118 boilermakers from Local 29 of the International Brotherhood of
Boilermakers, Iron Shipbuilders, Blacksmiths, Forgers and Helpers (81% of those contacted)
were followed between 1997 and 2000. Participants completed spirometry, a modified
American Thoracic Society questionnaire on respiratory symptoms, and a work history
questionnaire at baseline and two annual follow-up visits. The male cohort averaged 42.6 years
of age (range: 20.5 - 56.5 years) and 97% were Caucasian. Technicians used standardized
techniques and the same equipment to conduct spirometry testing for all workers during the
study. Each worker, after some days off work, performed 3-7 forced vital capacity maneuvers
to obtain at least three acceptable curves between 9:00 am and 2:00 pm on testing days. Baseline
FEVi and FVC were 90% and 94% of predicted. The nine participants who were lost to follow-
up had a lower mean baseline FEVi compared to those who remained in the study (84%
compared to 91.2% predicted). The number of years worked as a boilermaker was a statistically
significant predictor of annual FEVi (-33.5 ml/years worked (95% CL-45.9 - -21.1)). In
24
DRAFT - DO NOT CITE OR QUOTE
-------
generalized estimating equations adjusting for age, baseline FEVi and cigarette smoking status,
the number of hours worked at gas-fired powerplants in the previous year was inversely related
to annual FEVi (-9.8 ml/100 hours worked (95% CI: -16.0 - -3.5). Adjusted models analyzing
the number of hours worked at oil and gas-fired powerplants also showed FEVi reductions,
however the associations were not statistically significant. Statistically significant reductions in
annual FEVi also were observed among boilermakers who had ever worked at a coal, oil or gas-
fired powerplant in models evaluating each fuel type separately. This study provides evidence of
long-term declines in lung function among boilermakers exposed to combustion particles from
several fuel types. The study did not estimate exposure to individual substances however, and
no conclusions can be drawn regarding a role for vanadium compounds.
The effect of occupational PM2.5 concentrations and metals components on cardiac
autonomic function during a work shift was studied among a panel of 39 boilermaker
construction workers (Magari et al., 2002). The group of apprentice and journeyman
boilermakers was an average of 38 years old (18 - 59 years) and had worked an average of 13
years (0 - 40 years) in that occupation. Metals concentrations over an 8 - 10 hour work shift
were determined from particle samples (< 2.5 um) collected using personal monitors, and heart
rate was monitored using a five-lead Hotter monitor during the same period. Heart rate
variability was estimated as the mean of 5 minute average SDNN (standard deviation of the
normal-to-normal intervals). Vanadium concentrations (corrected for blank filter metal content)
were skewed with a mean of 0.76 ± 1.96 ug/m3 and a median of 0.13 ± 1.96 ug/m3 (range 0 -
11.62). Fifteen of 48 personal samples were above the limit of detection for vanadium (0.00859
ug/m3). Average PM2.5 concentrations were 1.16 ± 1.61 ug/m3 with a median of 0.56 ± 1.96
ug/m3 (range 0.09 - 7.76). Vanadium concentrations during the shift were associated with a 3.98
msec (95% CI: 1.64 - 6.32) increase in SDNN index per ug/m3 in mixed effects regression
models with a random effect for each study subject and fixed covariates for smoking status, age,
and mean heart rate. Lead also was associated with an increase in SDNN index (11.3
msec/ug/m3, 95% CI: 2.88 - 19.73).2 In contrast to earlier studies in this cohort that observed a
decrease in SDNN measures, average PM2 5 concentrations were not associated with SDNN
index in this study (-0.77 msec/ug/m3, 95% CI: -2.36 - 2.81). Vanadium concentrations were
not correlated with either lead or PM2.5 indicating that these pollutants were not likely to be
confounders of the association of heart rate variability with vanadium. No associations with
heart rate variability were observed for the other analyzed metals including nickel, chromium,
These effect estimates for vanadium and lead were reported as such in the abstract and Table 4, but were transposed
in the text of the Results.
25
DRAFT - DO NOT CITE OR QUOTE
-------
manganese, or copper. It is possible that the observation of an increase in heart rate variability
with unit increases in vanadium and lead concentrations was the result of the temporal
framework chosen for the analysis (e.g., averages over the work shift) rather than other time
frames for exposure and SDNN averages. Indeed, a subsequent study examining changes in a
different index of autonomic cardiac function, rMSSD (square root of the mean squared
differences of successive intervals), averaged over the night-time hours (0:00 - 7:00), found an
inverse association with average work shift concentrations of the PM2.5 metal, manganese
(Cavallari et al., 2008). The authors did not report on the results of analyses for vanadium. The
biological significance of the association of vanadium with increases in heart rate variability is
unclear. All-cause mortality has been associated with decreased heart rate variability measured
at baseline in longitudinal studies (Dekker et al., 1997). However, the effect on heart rate
variability indicates that vanadium exposure may alter autonomic function. Alternatively, the
observed association may have been due to chance.
In summary, case series, cross-sectional and longitudinal studies of occupational
exposure to vanadium pentoxide or vanadium in residual oil fuel ash over a few days to several
weeks reported symptoms of upper and lower respiratory tract irritation including headache,
runny nose or sneezing, sore throat, cough with sputum, dyspnea on exertion, chest pain or
discomfort, wheezing and tiredness (Sjoberg, 1951; Sjoberg, 1956; Zenz et al., 1962; Kiviluoto
et al., 1979; Kiviluoto, 1980; Musk and Tees, 1982; Williams, 1952; Sjoberg, 1955; Lees, 1980;
Ross, 1983; Levy et al., 1984; Woodin et al., 2000). Some symptoms began after a few hours
while the onset of other symptoms occurred after one to several days. Vanadium pentoxide
concentrations in production facilities where the studies were conducted varied by work location
from 0.06 mg/m3 in the ferrovanadium area to 1.53 mg/m3 in the fusion precipitation area at one
facility (Irsigler et al., 1999) and > 0.5 mg/m3 at a pilot-plant operation (Zenz et al., 1962).
Concentrations of vanadium pentoxide inside the boiler during a cleaning operation were 85
mg/m3 (Sjoberg, 1955) and ranged from 0.05 - 5.3 mg/m3 during conversion of a power plant
from coal to oil (Levy et al., 1984).
The frequency and severity of symptoms were associated with vanadium among
boilermakers exposed to relatively low ambient concentrations. Upper airway symptom severity
scores (nasal congestion/irritation, throat irritation) increased across quartiles of estimated
vanadium dose in the nose and the elevation was statistically significant in the third quartile
compared to the first quartile in regression models adjusted for smoking (Woodin et al., 2000).
Increases in lower airway symptom frequency and severity (chest tightness, wheeze, cough, and
sputum production) were associated in a dose-related manner with estimates of lung vanadium
dose. During the boiler overhaul, geometric mean respirable vanadium dust concentrations (SD)
26
DRAFT - DO NOT CITE OR QUOTE
-------
were 8.9 (2.3) |ig/m3 (0.009 mg/m3).
Vanadium concentrations in respirable particles (< 10 jim) inside the boiler during an
overhaul (geometric mean (SD): 8.9 (2.3) |ig/m3) associated with increased upper and lower
respiratory symptoms among boilermakers were not related to deficits in pulmonary function.
Pulmonary function declines (FEVi, FVC, FEF25%-75%) over the course of a boiler overhaul were
reported among boilermakers (Lees, 1980; Levy et al., 1984; Hauser et al., 1995a; Woodin et al.,
1999). However, those studies with more systematic and detailed reporting of methods and
results observed no association of FEVi, FVC or flow measures with average vanadium
concentrations in PMio (Mean (SD): 12.2 (9.1) |ig/m3) or estimates of vanadium or PMio dose in
nasal passages or the lung (Hauser et al., 1995a; Woodin et al., 1999). Peak PMio exposure (the
highest concentration (1 - 10 hour TWA) reached on any day during the overhaul) (Mean (SD):
4.25 (1.58) mg/m3) was inversely associated with AFEVi, AFEVso%, and AFVC, however peak
vanadium concentrations (Mean (SD): 20.3 (11.4) |ig/m3) were not.
The authors reported that concentrations of respirable vanadium dust were relatively low
and there may not have been enough variation in exposure to allow detection of an association
with the small changes in pulmonary function that occurred during the overhaul. Uncertainties
in the exposure estimates, particularly inside the boiler, also may have prevented detection of
associations between vanadium exposure and pulmonary function. The authors reported that
they obtained fewer air samples inside the boiler where wearing the monitor was uncomfortable.
Consequently, exposure and dose estimates for these locations may be more uncertain compared
to other locations. The cohort had worked in this occupation an average of 20 years and may
represent a healthy, less susceptible population. In addition to vanadium, boilermakers are
exposed to increases in the levels of other metals during boiler overhauls including nickel,
chromium and manganese (Liu et al., 2005). The observed pulmonary function declines amng
boilermakers may be explained by exposure to other ROFA constituents.
4.1.2.4 Exposure to PM2.s Vanadium in Ambient Air
Vanadium is a constituent of ambient particulate generated by oil combustion. Recent
studies of the short-term health effects of particulate matter and its constituents have found
higher risks of mortality and hospitalization in locations with a higher fractional content of
vanadium, nickel and elemental carbon in PM (Dominici et al., 2007; Bell et al., 2009).
Lippmann et al (2006) evaluated the impact of average concentrations of 16 PM2.5
components across 60 U.S. communities on the association between the daily change in PMio
concentration on daily all-cause mortality risk in those communities. Community-specific
27
DRAFT - DO NOT CITE OR QUOTE
-------
mortality risk estimates per daily change in PMio concentrations were obtained from the
National Morbidity, Mortality, and Air Pollution Study (NMMAPS) database for the years 1987
- 1994. Annual average concentrations for PM constituents were obtained for 2000-2003 from
the PM2 5 speciation network. The authors used weighted linear regression to evaluate for each
chemical constituent whether annual average concentration altered the association between PMio
concentration on the previous day and mortality risk. The constituents, nickel and vanadium,
were found to increase PMio mortality risk.
These results were re-evaluated and extended by Dominici and colleagues (2007) using
NMMAPS data for 90 communities from 1987 - 2000 and data on PM2.5 composition for 187
U.S. counties for 2000 - 2005. A total of 69 U.S. communities in the NMMAPS database also
had data on PM2.5 composition and were included in this analysis. Using a Bayesian hierarchical
regression model to estimate the association between the 1-day lag PMio mortality risk and
average county-level PM constituent concentrations, counties with high average concentrations
of nickel and vanadium had higher PMio mortality risk with a one-day lag. When the three
counties that comprise the NMMAPS New York community were excluded from the analysis,
the effect of nickel and vanadium was much weaker and lost statistical significance. The authors
stated that the three New York counties had nickel and vanadium concentrations that were 8.9
and 3.4 times higher than the other counties in the analysis.
Bell et al (2009) used a Bayesian hierarchical regression model to evaluate the effect of
PM2 5 chemical constituents as percent of PM2 5 total mass on PM2 5 associated cardiovascular
and respiratory hospital admissions by county and season for 106 U.S. counties during 1999 -
2005. Counties with a population 200,000 or greater for which data on PM and constituent
concentrations were available were selected. Models were adjusted for day of the week,
seasonality, long-term trends using a smoothing function, daily temperature and dew point
temperature, as well as the previous three days temperature and dew point temperature. County -
and season-specific PM2.5 relative risk for cardiovascular and respiratory hospital admissions
were higher in counties and seasons with a nickel, vanadium or elemental carbon fraction of total
PM2 5 in the 75th compared to the 25th percentile. The effect of these PM constituents was
statistically significant. The average concentration of vanadium across the counties was 0.003
ug/m3 (range: 0.001-0.01) with an interquartile range of 0.001 ug/m3. The interquartile range as
percent of PM2 5 total mass was 0.01%. Each interquartile range increase in the fraction of PM2 5
total mass for vanadium was associated with a 27.5% (95% posterior interval: 10.6 - 44.4)
increase in PM2.5 associated cardiovascular hospitalizations and a 392% (95% posterior interval:
46.3 - 738) increase in PM2.5 associated respiratory hospitalizations. Associations also were
observed for elemental carbon and nickel, and effect estimates were not always stable in
28
DRAFT - DO NOT CITE OR QUOTE
-------
multipollutant models.
The finding that communities with a higher fractional content of vanadium, nickel and
elemental carbon in ambient particulate matter have a higher risk of mortality and hospital
admissions related to daily change in PM2 5 concentration is intriguing and indicates the need for
further research on the contribution of fuel oil combustion to regional and local air pollution, and
the contribution of specific metals, including possibly vanadium pentoxide to elevated health
risks. The time series study design used in these investigations evaluates exposure-disease
associations at the county level and therefore, individual-level assessments of exposure and the
impact of possible confounders is not possible. However, Bell et al (2009) investigated whether
county-level indicators of socioeconomic status, racial composition, and degree of urbanization
could be alternative explanations for the observed effect modification by PM2.5 constituents and
concluded that this was not the case.
Ambient concentrations of PM2.5 and PM2.5 fractions of nickel, vanadium, zinc and
elemental carbon were evaluated in relation to respiratory symptoms among young Dominican
and African American children, aged 3-24 months, followed as part of a birth cohort study in
Northern Manhattan and the South Bronx in New York City between 1998 and 2007 (Patel et al.,
2009). Among 653 24 month old children with questionnaire data (90% of total enrolled) 3-
month average ambient vanadium concentrations were associated with an increase in the
presence of wheeze during the cold and flu season (September 1 - March 31). After adjusting
for elemental carbon, NO2, copper, and iron, an interquartile range increase (0.003 ug/m3) in
three-month average vanadium concentrations was associated with a 31% increased probability
of wheeze during the cold season (p<0.0003). When not stratified by season, an IQR increase in
vanadium concentration was associated with a 14% increased probability of wheeze (p=0.08).
However, when the highest 5% of vanadium concentrations were excluded, the association of
wheeze with vanadium lost significance in the multipollutant model. Vanadium concentrations
were not associated with cough. Twenty-four hour average ambient concentrations of PM2.5 and
PM2.5 fractions of nickel, vanadium, zinc and elemental carbon, measured every third day at two
stationary sites in the Bronx, were obtained from the New York State Department of
Environmental Conservation. Exposure levels were assigned to each subject by calculating 3-
month moving average concentrations of each pollutant based on each follow-up questionnaire
date and the previous three months. Exposures were assigned to each subject's address using
inverse-distance weighted concentrations from the two stationary monitors. Associations were
evaluated using generalized additive mixed effects models and a first-order autoregressive
correlation structure to account for correlation between the up to 8 repeated observations for
each individual. The models also adjusted for sex, ethnicity, postnatal ETS exposure, and a
29
DRAFT - DO NOT CITE OR QUOTE
-------
smoothed term for calendar time using natural cubic splines. Other pollutants also were
associated with increased probability of wheeze (nickel) or cough (elemental carbon, NO2)
during the cold/flu season or wheeze in other months (NO2). PM2.5 mass concentrations were
not related to an increase in probability of symptoms.
The interpretation of the multi-pollutant models for vanadium is complicated because
other PM constituents are correlated with vanadium concentrations resulting in less stable risk
estimates. Nickel was not evaluated in the same model with vanadium for this reason. The
association with vanadium was independent of the association with NO2, a marker for traffic
emissions, and the authors suggested that oil combustion for space heating may contribute to the
observed respiratory symptoms in the very young children in this study. Although vanadium
cannot be singled out as the responsible agent for the probability of wheeze observed in this
study, the association is consistent with the respiratory symptoms observed among boilermakers
with exposure to high levels of residual oil fuel ash for periods of days to weeks.
The effect of the metal content of ambient PM2.5 on lung function also was evaluated in a
time-series panel study of 29 patients with chronic obstructive lung disease, asthma, or ischemic
heart disease in Rome, Italy in the spring and winter of 1999 (Lagorio et al., 2006). Outpatients
of the Pneumology and Cardiology Departments of the Catholic University Hospital in Rome
who met eligibility requirements for COPD (N=l 1), asthma (N=l 1) or ischemic heart disease
(N=7), and who lived in census tracts less than 2 kilometers from one of six air monitoring
stations were selected for the study. The subjects volunteered to conduct repeated clinical
examinations for two one-month periods. Pulmonary function testing was conducted according
American Thoracic Society guidelines, and measures were expressed as the percentage of
predicted based on subject-specific age, height and weight. An average of 15, 24 and 9
observations were obtained from each participant in the COPD, ischemic heart disease and
asthma panels, respectively. Daily average PM2.5 concentrations were calculated based on
measurements obtained at two fixed site monitors set up for the study. PM content of cadmium,
chromium, iron, nickel, lead, platinum, vanadium and zinc was calculated as the ratio of the
metal content in each PM sample to the air volume collected during the sampling. The mean 24-
hour PM2.5 concentrations during the spring and winter of 1999 were 18.2 ± 5.0 ug/m3 and 36.7
± 24.1 ug/m3, respectively. The mean 24-hour vanadium concentrations during the spring and
winter of 1999 were 2.4 ±1.6 ng/m3 and 1.1 ± 0.52 ug/m3, respectively. Vanadium was not
associated with daily change in percent predicted pulmonary function among subjects with
asthma, ischemic heart disease, or COPD in generalized estimating equations models. Models
adjusted for season (all), daily mean temperature (all), relative humidity (all), day of the week
(COPD, IHD) and B-2 agonist use (asthma). Although the repeated measures design was a
30
DRAFT - DO NOT CITE OR QUOTE
-------
strength of the study, the number of subjects in each disease panel was small, and may not have
been large enough to detect an association with the very low vanadium concentrations analyzed.
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS - ORAL AND INHALATION
4.2.1. Oral Exposure
4.2.1.1. Subchronic Studies
No animal studies that have comprehensively examined histopathological, biochemical
and clinical endpoints of subchronic oral exposure were identified from the available literature.
Mountain et al. (1953) evaluated the effects of subchronic exposure of rats to dietary vanadium
pentoxide on body weight gain, erythrocyte count, hemoglobin and cystine content of hair.
Groups of 5 male Wistar rats were fed diets containing 0, 25, 50, 500 or 1000 ppm of vanadium
incorporated in the form of pentoxide for 103 days (25 and 50 ppm groups; "low-exposure"
groups) or 75 days (500 and 1000 ppm groups; "high-exposure" groups). After 35 days of
treatment, dietary vanadium levels of the "low-exposure" groups were increased to 100 and
150 ppm, resulting in average daily doses of 0, 74.5, 116, 500, and 1000 ppm (0, 5.9, 9.2, 39, or
79 mg/kg-day)3 of vanadium resulting in a doses of 0, 10.5, 16.4, 69.6, or 141.0 mg/kg-day of
vanadium pentoxide.4 At the end of treatment, body weight gain, liver weight, and cystine
content of hair were measured in all groups, erythrocyte count and hemoglobin level were
measured in control and "low-exposure" groups, and relative liver weight was measured in
control and 69.6 mg/kg-day groups. Average body weights of animals at the conclusion of the
study were not reported, though average weight gain in grams/rat was reported. Compared to
control, average body weight gain was increased in the 10.5 mg/kg-day and 16.4 mg/kg-day
groups (54 and 45% increase) and decreased in the 69.6 mg/kg-day group (66%) and 141.0
mg/kg-day group (no gain in body weight over the study period). The increase in body weight
gain at the low-exposure levels was not explained and statistical significance or standard
deviations were not reported for any result. Relative liver weight in the 69.6 mg/kg-day group
3 Calculation: mg/kg-day = ppm(mg of compound per kg food) x mg food consumed/day x I/kg body weight (using
reference food consumption rate of 0.0217 kg/day [U.S. EPA, 1988] and average body weight of 0.275 kg for male
rats [Mountain etal, 1953]).
4
Conversion from mg/kg - d of vanadium to amount vanadium pentoxide: [(mg/kg-d vanadium)(MW vanadium
pentoxide, 181.9)]/(2xMW vanadium, 101.9).
31
DRAFT - DO NOT CITE OR QUOTE
-------
was statistically significantly increased compared to control, reported as a ratio of liver
weight/body weight (3.86 compared to 3.51, p < 0.05, F ratio in analysis of variance). Data on
relative liver weight were not reported for other dose groups. A dose-dependent decrease in
erythrocyte count (12.8 and 21.3%) was observed over the duration of the study, in rats exposed
to 10.5 mg/kg-day and 16.4 mg/kg-day vanadium pentoxide respectively, compared to controls
(3.8% decrease) (Table 4-1). A 20-30% decrease in erythrocyte count is considered biologically
significant, however, no statistical analysis was reported by the study authors, and no measure of
variance (SE or SD) was given for the means. Data were not reported for high dose groups.
Hemoglobin levels decreased 4.6% and 10.5%, respectively, in the 10.5 and 16.4 mg/kg-day
groups over the duration of the study compared to 3.9% in controls. Cystine content of hair
significantly decreased in a non-dose dependent manner in all vanadium pentoxide treatment
groups compared to controls with the exception of the lowest exposure group. The biological
significance of decreased hair cystine content is not established; though the researchers
speculated that vanadium may have inhibited enzymes, such as sulfotransferases, that decreased
the availability of cystine for hair growth. This study observed potentially dose-related changes
in erythrocytes, body weight gain, and liver weight in treated animals. However, no statistical
analysis was performed for the decrease in erythrocytes and body weight gain and no degree of
variance was reported (precluding statistical analysis for this review), therefore, a NOAEL or
LOAEL could not be determined from this study. However, compared to hematological data for
Wistar rats (Wright et al., 2009; Charles River 2008), the observed decrease in erythrocyte
counts observed in this study is outside of the reference range for the historical controls and thus
is believed to be a clinically significant finding.
Table 4-1. Hematological results of oral vanadium pentoxide exposure in rats (Mountain et al.
1953).
Control
10.5 mg/kg-day
16.4 mg/kg-day
Red Cell Count (M/mm3)
Start
Finish (103 days)
Percent change between start and
finish of expt (%)
8.0
7.7
3.8
7.8
6.8
12.8
8.0
6.3
21.3
Hemoglobin, %
Start
Finish (103 days)
Percent change between start and
finish of expt (%)
15.6
15.0
3.9
15.2
14.5
4.6
15.3
13.7
10.5
4.2.1.2 Chronic Studies
32
DRAFT - DO NOT CITE OR QUOTE
-------
A 2.5-year dietary study on vanadium in rats (strain not described) was previously used
as the basis of the chronic RfD (Stokinger et al., 1953). The results were summarized in Patty's
Industrial Hygiene and Toxicology, 3rd ed., 1981. In this chronic study, an unspecified number
of rats were exposed to dietary levels of 10 or 100 ppm vanadium (about 17.9 or 179 ppm
vanadium pentoxide; 1.41 and 14. 1 mg/kg-day )5 for 2.5 years. Endpoints evaluated were
limited to growth rate, survival, and hair cystine content. The study did not assess
comprehensive toxicity endpoints. Hair cystine content was significantly decreased in exposed
animals, compared to controls but no values were given. This study reports the oral NOAEL
upon which an RfD can be derived as 17.9 ppm (1.41 mg/kg - day) vanadium pentoxide.
However, the biological significance of decreased hair cystine is unclear and is not specific to
vanadium pentoxide exposure. No additional oral chronic exposure studies in animals were
identified in the published literature.
4.2.2. Inhalation Exposure
4.2.2.1 Subchronic Studies
The 3-month exposure studies in F344/N rats were conducted to evaluate the cumulative
toxic effects of subchronic inhalation exposure to vanadium pentoxide (NTP, 2002). Chemical
identity and purity of vanadium pentoxide was evaluated prior to the beginning of and following
the conclusion of all assays. Groups of 10 male and 10 female rats were exposed (whole-body
exposure) to aerosols of vanadium pentoxide at concentrations of 0, 1, 2, 4, 8 or 16 mg/m3, 6 hrs
per day, 5 days/week for 3 months. Additional groups of 10 male and 10 female rats were
exposed to 4, 8 or 16 mg/m3 for 12 (females) or 13 (males) weeks to investigate effects of
exposure on cardiovascular function, pulmonary function and pulmonary inflammation. Clinical
findings were recorded weekly and animals were weighed weekly and at the end of the study.
Blood and urine were collected from core study rats at study termination. Blood was also
collected from cardiopulmonary physiology study rats on days 4 and 23 for hematology and
clinical chemistry determinations. Necropsy and histopathological evaluations (light microscopy
of comprehensive tissues6) were performed on all main study rats exposed to 0, 2 (male rats
5 Converted to ppm vanadium pentoxide by [(ppm vanadium)(MW vanadium pentoxide, 181.9)]/(2xMW vanadium,
101.9). No information given on average rat weights so cannot be converted to mg/kg-day directly. Conversions
based on data from subchronic study from same group (Moutain et al., 1953). Calculation: mg/kg-day =ppm(mg of
compound per kg food) x mg food consumed/day x I/kg body weight (usingreference food consumption rate of
0.0217 kg/day [U.S. EPA, 1988] and average body weight of 0.275 kg for male rats [Mountain et al., 1953])
6 Complete histopathology was performed on 0, 8 (rats only), and 16 mg/m3 rats and mice. In addition to gross
33
DRAFT - DO NOT CITE OR QUOTE
-------
only), 4, 8 or 16 (female rats only) mg/m3 at the completion of the study. Sperm motility and
vaginal cytology evaluations were analyzed from all core study rats.
Seven male rats and three female rats exposed to 16 mg/m3 vanadium pentoxide died
during the study (NTP, 2002). Abnormal breathing, emaciation, lethargy, abnormal posture and
ruffled fur were observed in male and female rats exposed to concentrations of 8 mg/m3 and
higher. Diarrhea and nasal/eye discharge were also observed in some rats exposed to 16 mg/m3.
Weight gain and absolute and relative lung weights are summarized in Table 4-2. Weight gain
over the 3-month treatment period was significantly decreased compared to control in males
exposed to 4 (6% decrease), 8 (10% decrease) and 16 (60% decrease) mg/m3 and in females
exposed to 16 mg/m3 (30% decrease). Absolute lung weights were significantly increased in
males exposed to concentrations of 2 mg/m3 and greater and in females exposed to 4 mg/m3 and
greater. Relative lung weights were significantly greater than control in males exposed to 2
(16% increase), 4 (30% increase), 8 (51% increase) or 16 (145% increase) mg/m3 and in females
exposed to 4 (19% increase), 8 (76% increase) or 16 (117% increase) mg/m3. Other organ
weight differences were considered to be related to body weight decreases.
Results of hematology assessments following 3-months of inhalation exposure are
presented in Table 4-3. Erythrocyte count was significantly increased in the 8 and 16 mg/m3
groups and hematocrit was significantly increased in the 16 mg/m3 group in male and female
rats. Hemoglobin was increased significantly only in females exposed to 16 mg/m3.
Microscopic evaluation of the red blood cell morphology detected increased polychromasia and
hypochromia in rats in the 16 mg/m3 groups (data not presented). Significantly decreased mean
cell hemoglobin concentrations were observed in males exposed to 8 and 16 mg/m3 and in
females exposed to 4, 8, and 16 mg/m3. Reticulocyte count was significantly increased in males
and females exposed to 16 mg/m3. Mean cell volume was significantly decreased, indicative of
microcytosis, in male rats at concentrations of 2 mg/m3 and above and in female rats at
concentrations of 4 mg/m3 and above. The observed hematological changes, including
erythrocytosis, are consistent with pulmonary lesions that reduce pulmonary oxygen transfer,
resulting in tissue hypoxia and stimulation of erythropoiesis by increased renal production of
erythropoietin. Erythrocyte microcytosis is consistent with ineffective erythropoiesis, suggestive
lesions and tissue masses, the following tissues were examined in the 3 month studies: adrenal gland, bone with
marrow, brain, clitoral gland, esophagus, gallbladder (mice only), heart and aorta, large intestine (cecum, colon and
rectum), small intestine (duodenum, jejunum, and ileum), kidney, larynx, liver, lung and mainstream bronchi, lymph
nodes (mandibular, mediastinal, mesenteric, and bronchial), mammary gland (except mail mice), nose, ovary,
pancreas, parathyroid gland, pituitary gland, preputial gland, prostate gland, salivary gland, skin, spleen, stomach
(forestomach and glandular), testis (with epididymis and seminal vesicle), thymus, thyroid gland, trachea, urinary
bladder, and uterus. The lung of rats and mice and nose of rats in all remaining exposure groups and the thymus in 8
mg/m3 mice were also examined.
34
DRAFT - DO NOT CITE OR QUOTE
-------
of altered iron metabolism and heme/hemoglobin production.
Table 4-2. Body Weight Gain and Lung Weights in Rats (F344/N) Exposed to Vanadium Pentoxide
by Inhalation for 3 Months (Values are Means±Standard Error) (NTP, 2002)
Parameter
Exposure
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Rats
Weight gain
during 3-
month
exposure
period (g)
Absolute
lung weight
(g)
Relative lung
weight
197±6
2.38±0.17
6.77±0.36
202±5
2.56±0.11
7.40±0.28
180±5
2.65±0.07
7.83±0.19
a
173±8a
2.93±0.09a
8.88±0.22a
161±5a
3.26±0.13a
10.20±0.30a
l±9a
1.98±0.10b
16.60±0.33a
Female Rats
Weight gain
during 3-
month
exposure
period (g)
Absolute
lung weight
(g)
Relative lung
weight
87±3
1.65±0.11C
8.37±0.58C
88±4
1.58±0.04
7.84±0.16
96±4
1.92±0.12b
9.23±0.53
83±3
1.95±0.08b'c
10.00±0.38b'c
77±4
2.16±0.06a
11.48±0.33a
25±7a
2.16±0.12a
18.15±1.06a
aSignificantly different from control by William's or Dunnett's test (p< 0.01)
bSignificantly different from control by William's or Dunnett's test (p< 0.05)
cn=9
Table 4-3. Selected Hematology Parameters in Rats (F344/N) Exposed to Vanadium Pentoxide by
Inhalation for 3 Months (NTP, 2002)a
Parameter
Exposure
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Rats
Number
Erythrocytes
(106/|iL)
Reticulocytes
9
9.2±0.1
0.2
9
9.0±0.1
0.22±0.03
10
9.1±0.1
0.19±0.02
9
9.3±0.2
0.23±0.03
10
9.7±0.2b
0.25±0.02
O
15.1±0.3C
0.8 ±0.08 b
35
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-3. Selected Hematology Parameters in Rats (F344/N) Exposed to Vanadium Pentoxide by
Inhalation for 3 Months (NTP, 2002)a
Parameter
(106/|iL)
Hematocrit (%)
Hemoglobin (g/dL)
Mean cell volume
(fL)
Mean cell
hemoglobin (pg)
Exposure
Control
±0.02
48.5±0.6
15.8±0.1
52.9±0.2
17.3±0.2
1 mg/m3
47.7±0.5
15.5±0.1
52.9±0.1
17.2±0.1
2 mg/m3
47.6±0.6
15.5±0.2
52.3±0.1b
17.1±0.1
4 mg/m3
48.7±0.9
15.9±0.2
52.2±0.2b
17.1±0.02
8 mg/m3
49.9±0.7
16.1±0.2
51.3±0.2C
16.5±0.2C
16 mg/m3
71.2±2.8b
20.4±0.8
46.8±1.0C
13.4±0.4C
Female Rats
Number
Erythrocytes
(106/|iL)
Reticulocytes
(106/|iL)
Hematocrit (%)
Hemoglobin (g/dL)
Mean cell volume
(fL)
Mean cell
hemoglobin (pg)
10
8.0±0.1
0.15 ±
0.02
45.8±0.5
15.5±0.2
56.9±0.1
19.3±0.2
10
7.8±0.1
0.17±
0.01
44.3±0.4
15.0±0.1
56.9±0.1
19.3±0.2
9
8.2±0.2
0.17±
0.01
46.1±1.2
15.5±0.2
56.6±0.1
19.0±0.2
10
8.3±0.1
0.16±
0.02
46.4±0.4
15.6±0.1
55.8±0.1C
18.7±0.2C
10
8.6±0.1b
0.17±
0.02
47.2±0.6
15.8±0.1
55.0±0.2C
18.5±0.2C
6
12.5±0.34C
0.45 ±
0.08C
60.8±1.4C
18.2±0.3C
48.7±0.6C
14.6±0.3C
aValues are means±standard error
bSignificantly different from control (p< 0.05)
°Significantly different from control (p< 0.01)
Sporadic alterations in clinical chemistry and urinalysis variables were observed at
various time-points in exposed males and females; however, no dose- or duration-related pattern
of effect was observed. Occasional changes in serum liver enzyme activities were not consistent
with hepatocellular injury.
Vanadium pentoxide exposure did not affect reproductive endpoints in males (sperm
count, spermatid heads, sperm motility), but it did increase estrous cycle length by 10% in
females exposed to 8 mg/m3, but not to 16 mg/m3, and reduced the number of cycling females in
surviving rats in the 16 mg/m3 group (percent reduction not reported) (NTP 2002).
Complete histopathological assessments were performed on rats exposed to 0, 8 and
16 mg/m3 for 3 months; only nonneoplastic lesions of the lung and nose were related to
treatment (NTP, 2002). Results of histopathological evaluations of lung and nasal tissue from
male and female rats exposed to 1, 2, 4, 8 and 16 mg/m3 for 3 months are summarized in Table
36
DRAFT - DO NOT CITE OR QUOTE
-------
4-4. Significant increases in the incidences of epithelial hyperplasia of the lung were observed in
male and female rats exposed to concentrations of 2 mg/m3 or greater, compared to controls.
Epithelial hyperplasia occurred in the distal airways and associated alveolar ducts and alveoli.
Inflammation and fibrosis were significantly increased in males (2 mg/m3 or greater) and females
(4 mg/m3 or greater). In the nasal compartment, incidences of hyperplasia and metaplasia of the
respiratory epithelium were significantly increased in males exposed to 8 or 16 mg/m3 and in
females exposed to 4 mg/m3 or greater. Nasal hyperplasia and metaplasia was localized to
respiratory epithelium on the ventral portion of the nasal septum, the vomeronasal organ, and, to
a lesser extent, the ventral lateral walls of the anterior portion of the nasal cavity. Nasal
inflammation was significantly increased in males and females exposed to 16 mg/m3.
Cardiopulmonary assessments were conducted in groups of 4-10 male and female rats
exposed to 0, 4, 8 and 16 mg/m3 for 3 months (NTP, 2002). No treatment-related changes in
cardiovascular function, as assessed by blood pressure (systolic, diastolic and mean), heart rate
and electrocardiogram, were observed in rats exposed to 4 or 8 mg/m3. Decreased heart rate and
diastolic, systolic and mean blood pressure observed in male and female rats exposed to
16 mg/m3 were considered to be a reflection of the poor condition of the animals, and
complicated by anesthesia. Significant exposure-related decreases in pulmonary function (as
assessed by respiratory rate, tidal and minute volume, expiratory resistance, vital and total
capacity, diffusing capacity, and dynamic and peak compliance) were observed at all
concentrations of vanadium pentoxide-exposed male and female rats. Observed changes in
impaired capacity to diffuse carbon monoxide and reduced static and dynamic lung volumes at
exposure concentrations of 4 mg/m3 and greater suggest a restrictive lesion. Changes in forced
expiratory maneuvers in rats exposed to 16 mg/m3 suggest the presence of an obstructive
disease. Pulmonary function results may indicate obstructive disease or may reflect the
deteriorating condition of the 16 mg/m3 rats, since histopathological finding in lungs of rats
exposed to 8 and 16 mg/m3 were similar. Taken together, results of pulmonary function tests
indicate that a presence of restrictive injury in male and female rats exposed to concentrations of
4 mg/m3 or greater, while an obstructive lung injury may have been present in rats exposed to
16 mg/m3.
Table 4-4. Incidences of Selected Nonneoplastic Lesions of the Lung and Nose in Rats (F344/N)
Exposed to Vanadium Pentoxide by Inhalation for 3 Months (NTP, 2002)
Lesion Location
and Type
Numbers of Animals with Lesions (Avg
Control
Male Ratsa
1 mg/m3
2 mg/m3
4 mg/m3
Severity Score)
8 mg/m3
16 mg/m3
37
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-4. Incidences of Selected Nonneoplastic Lesions of the Lung and Nose in Rats (F344/N)
Exposed to Vanadium Pentoxide by Inhalation for 3 Months (NTP, 2002)
Lesion Location
and Type
Numbers of Animals with Lesions (Avg Severity Score)
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Lung
Epithelium,
hyperplasia
Inflammation
Fibrosis
Bronchiole,
exudates
0
0
0
0
0
0
0
0
10b (2.0)
9b(1.0)
2(1.0)
0
10b(3.0)
10b(1.0)
10b(1.9)
0
10b(3.6)
10b(1.6)
10b(3.2)
7b(1.0)
10b(3.3)
10b(2.1)
10b(3.1)
8b(1.4)
Nose
Epithelium,
hyperplasia
Epithelium,
squamous
metaplasia
Inflammation
0
0
0
0
0
0
0
0
0
1(1.0)
1(1.0)
0
10b(1.2)
10b(1.2)
0
10b (2.0)
10b(1.8)
7b(1.6)
Female Ratsa
Lung
Epithelium,
hyperplasia
Inflammation
Fibrosis
Bronchiole,
exudates
0
0
0
0
0
0
0
0
10b(1.3)
0
0
0
10b (2.9)
10b(1.0)
10b(1.0)
0
10b(3.5)
10b(1.9)
10b (2.9)
10b(1.0)
10b(3.2)
10b(1.2)
10b(3.2)
8b(l.l)
Nose
Epithelium,
hyperplasia
Epithelium,
squamous
metaplasia
Inflammation
0
0
0
0
0
0
0
0
0
10b(1.0)
8b(1.0)
0
10b(1.8)
10b(1.8)
1(1.0)
10b (2.7)
10b (2.8)
9b(1.6))
a!0 animals per treatment group; numbers in parentheses indicate average severity grade of lesions in affected
animals: l=minimal, 2=mild, 3=moderate, 4=marked
bSignificantly different from control by Fisher exact test (p< 0.01)
BAL fluid was analyzed for markers of pulmonary inflammation in rats exposed to 0, 4, I
and 16 mg/m3 for 3 months (NTP, 2002). Concentration-related increases were observed in the
total numbers of cells, lymphocytes, neutrophils and protein recovered in BAL fluid from rats
38
DRAFT - DO NOT CITE OR QUOTE
-------
exposed to vanadium pentoxide at concentrations of 4 and 8 mg/m3, demonstrating a pulmonary
inflammatory response in male and female rats. These endpoints also were increased in the
16 mg/m3 group, but to a lesser extent, vanadium pentoxide was overtly toxic at this dose.
Results of this study show that inhalation exposure of male and female rats to vanadium
pentoxide aerosol for 3 months produced adverse effects on the hematological system and the
lung (NTP, 2002). Microcytic erythrocytosis, which was possibly secondary to impaired
pulmonary function, was observed at concentrations of 2 mg/m3 and greater in males and
4 mg/m3 and greater in females. Absolute and relative lung weights were significantly increased
compared to controls at concentrations of 4 mg/m3 and greater in females and 2 mg/m3 and
greater and 4 mg/m3 and greater, respectively, in males. The incidence of nonneoplastic lesions
of the nose was increased in male and female rats at concentrations of 8 mg/m3 and greater and
4 mg/m3 and greater, respectively, and the incidence of nonneoplastic lesions of the lung was
increased in male and female rats 2 mg/m3 and greater. Results of pulmonary function tests
consistent with restrictive lung disease were observed at concentrations of 4 mg/m3 and greater.
Based on decreased erythrocyte size in male rats and nonneoplastic lung lesions in male and
female rats, the NOAEL and LOAEL values identified for 3-month inhalation exposure to
vanadium pentoxide aerosols were 1 and 2 mg/m3, respectively.
Three-month exposure studies in B6C3Fi mice were conducted to evaluate the toxicity of
subchronic inhalation exposure to vanadium pentoxide (NTP, 2002). Groups of 10 male and
10 female mice were exposed (whole-body exposure) to vanadium pentoxide aerosols at
concentrations of 0, 1, 2, 4, 8 or 16 mg/m3, 6 hrs per day, 5 days/week for 3 months. Particle
size given in mass median aerodynamic diameter (MMAD) ± geometric standard deviation
(GSD) for each dose groups was as follows: 1 mg/m3=1.2±2.8; 2 mg/m3=l.l±2.8;
4 mg/m3=1.2±2.8; 8 mg/m3=1.0±2.9; 16 mg/m3=1.2±2.8. Clinical findings were recorded
weekly. Animals were weighed weekly and at the end of the study. All study animals were
necropsied. Histopathological examinations of lungs were performed in all mice in the 0, 1,2, 4,
8 or 16 mg/m3 groups and of thymus in all mice in the 0, 8 or 16 mg/m3 groups. At the end of
the 3-month exposure period, samples for sperm motility and vaginal cytology evaluations were
collected from mice exposed to 0, 4, 8 or 16 mg/m3. Complete histopathological examination
was performed in mice in the control and 16 mg/m3 groups. Assessments of cardiopulmonary
function, pulmonary inflammation (analysis of pulmonary lavage), and hematological parameters
were not conducted in mice.
One male mouse in the 16 mg/m3 group died before the end of the study. Other than
appearing thin, no other signs of toxicity were reported (NTP, 2002). No other treatment-related
clinical findings were observed in any other mice in any treatment group. Weight gain and
39
DRAFT - DO NOT CITE OR QUOTE
-------
absolute and relative lung weights are summarized in Table 4-5. Weight gain over the 3-month
treatment period was significantly decreased compared to control in males exposed to 8 (6%
decrease) and 16 (10% decrease) mg/m3 and in females exposed to 4 (11% decrease), 8 (10%
decrease) and 16 (12% decrease) mg/m3. Absolute lung weights were significantly increased
compared to control at concentrations of 2 mg/m3 and higher in males and 4 mg/m3 and higher in
females. Relative lung weights were significantly greater than control in males exposed to 4
(33% increase), 8 (43% increase) or 16 (82% increase) mg/m3 and in females exposed to 4 (62%
increase), 8 (63% increase) or!6(101% increase) mg/m3. Other organ weight differences were
considered to be related to decreases in body weight by the researchers. The epididymal
spermatozoal motility of males exposed to 8 or 16 mg/m3 was significantly decreased by 13 and
5%, respectively. No treatment-related effects were observed for assessments of estrous cycle
(estrous cycle length and number of cycling females).
Table 4-5. Body Weight Gain and Lung Weights in Mice (B6C3Fi) Exposed to Vanadium
Pentoxide by Inhalation for 3 Months (Values are Means±Standard Error) (NTP, 2002)
Parameter
Exposure
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Mice
Weight gain
during 3 -month
exposure period
(g)
Absolute lung
weight (g)
Relative lung
weight
8.4±0.9
0.2±0.01
7.0±0.2
7.4±0.8
0.2±0.01
6.9±0.2
8.2±0.6
0.3±0.01b
7.8±0.3
7.7±0.5
0.3±0.01b
9.3±0.2b
6.2±0.2a
0.3±0.01b
10.0±0.3b
5.6±0.7b
0.4±0.01b
12.7±0.4b
Female Mice
Weight gain
during 3 -month
exposure period
(g)
Absolute lung
weight (g)
Relative lung
weight
9.7±1.0
0.2±0.01
8.1±0.5
10.0±1.0
0.3±0.01
8.8±0.3
8.1±0.4
0.3±0.01
9.7±0.5
5.8±0.5b
0.3±0.02b
13.2±0.9b
6.1±0.4b
0.4±0.02b
13.2±0.6b
5.4±0.3b
0.4±0.02b
16.3±0.52b
Significantly different from control by William's test (p< 0.05)
bSignificantly different from control by William's test (p< 0.01)
40
DRAFT - DO NOT CITE OR QUOTE
-------
Results of histopathological evaluations of lung tissue from male and female mice
exposed to 0, 1, 2, 4, 8 and 16 mg/m3 for 3 months are summarized in Table 4-6 (NTP, 2002).
Epithelial hyperplasia was observed in male and female mice exposed to concentrations of
2 mg/m3 and above; lesion severity increased with increasing exposure concentration.
Hyperplasia involved alveolar and, to a lesser extent, bronchiolar epithelium. Inflammation was
characterized by multiple foci of a mixed cellular infiltrate oriented around blood vessels and
bronchioles and was observed in male mice exposed to 4 mg/m3 and above and in female mice
exposed to 2 mg/m3 and above. Infiltrate was composed primarily of macrophages with
abundant cytoplasm and fewer lymphocytes and neutrophils. Histopathological evaluations of
the thymus of male and female mice exposed to 0, 8 and 16 mg/m3 for 3 months showed
lymphoid depletion in mice exposed to 16 mg/m3 (males: control, 0/9; 8 mg/m3, 0/8; 16 mg/m3,
2/7; females: 0/9, 0/9, 1/10).
The lung was identified as the most sensitive target organ in the 3-month inhalation study
in mice (NTP, 2002). Based on increases in absolute lung weights at concentrations of 2 mg/m3
and greater (males) and inflammation of the respiratory epithelium at concentrations of 2 mg/m3
and greater (males and females), NOAEL and LOAEL values were identified as 1 and 2 mg/m3,
respectively.
Table 4-6. Incidences of Selected Nonneoplastic Lesions of the Lung in Mice (B6C3Fi) Exposed to
Vanadium Pentoxide by Inhalation for 3 Months (NTP, 2002)
Lesion Type
Numbers of Animals with Lesions3
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male
Number
Inflammation
Epithelium,
hyperplasia
10
0
0
10
1(1.0)
1(1.0)
10
3(1.0)
4b(1.0)
10
4b(1.0)
5b(1.0)
10
10C (2.0)
10C(1.3)
10
10C (2.0)
10C(3.0)
Female
Number
Inflammation
Epithelium,
hyperplasia
10
0
0
9
1(1.0)
0
10
7C(1.0)
6C(1.0)
9
9C(1.9)
9C(1.5)
10
10C(1.9)
10C(1.5)
10
10C (2.5)
10C (2.5)
"Numbers in parentheses indicate average severity grade of lesions in affected animals: l=minimal, 2=mild,
3=moderate, 4=marked
bSignificantly different from control by Fisher exact test, p< 0.05
Significantly different from control by Fisher exact test, p< 0.01
In a study using cynomolgus monkeys, weekly provocation challenges (single 6-hr
41
DRAFT - DO NOT CITE OR QUOTE
-------
exposures to 0.5 or 3.0 mg/m3) with inhaled vanadium pentoxide aerosol for six weeks produced
statistically significant pulmonary responses, prior to a subchronic exposure (6 hrs/day, 5
days/week for 26 weeks) (Knecht et al., 1992). The subchronic exposure was divided into three
groups; one group (n=8) was exposed to filtered, conditioned air and two exposed groups (n=8
each) received nominally equal weekly vanadium pentoxide exposures (concentration x time)
with different exposure profile. The peak exposure group received an actual concentration of
0.16 ± 0.01) mg/m3 (0.1 mg/m3 nominal) vanadium pentoxide on Mondays, Wednesdays and
Fridays and 1.38 ± 0.07) mg/m3 (1.1 mg/m3 nominal) vanadium pentoxide on Tuesdays and
Thursdays, and the constant exposure group received a constant daily actual concentration of
0.57 ± 0.03) mg/m3 (0.5 mg/m3 nominal). The constant exposure regimen corresponded to a
continuous exposure of 0.10 mg/m3 after adjusting for exposure protocol (0.57 mg/m3 x 6/24 x
5/7). The peak exposure regimen averaged to a slightly higher continuous exposure of 0.12
mg/m3 after adjusting for exposure protocol. Vanadium pentoxide particle size was determined
weekly during challenges and biweekly during exposures. Average particle size for the
subchronic constant exposure group was 3.15 jim (MMAD), with a GSD of 3.25 jim. Particle
sizes (MMAD ± GSD) for the peak exposure group were 3.17 ± 2.48 and 3.10 ± 2.45 for the 0.1
mg/m3 and 1.1 mg/m3 exposures, respectively. Pulmonary function tests, cytological and
immunological analyses of blood and bronchiolar lavage fluid, and skin sensitivity tests were
conducted before the pre- and post-exposure provocation challenges. Immunological analyses
are described in Section 4.4.1.2. Pulmonary function tests and bronchiolar lavage fluid analyses
were also performed one day after the provocation challenges. Cytological endpoints included
complete and differential blood cell counts and leukotriene €4 levels. Pulmonary function
endpoints included total pulmonary resistance (RL), forced expiratory flow (FEF), forced vital
capacity (FVC), residual volume (RV) and dynamic lung compliance (CLdyn). Respiratory
distress, characterized by audible wheezing and coughing, occurred in 3 out of 8 monkeys from
the peak exposure group on peak exposure days during the first few weeks of the 26-week
exposure; the responses developed within 3-4 hrs of exposure and occasionally required early
removal of the affected monkeys from the exposure chamber. Impaired pulmonary function
accompanied pre-exposure provocation challenges with "V^Os at 3.0 mg/m3 and was
characterized by a 14% increase in RL and 13% decrease in FEVso/FVC accompanied by a 14%
increase in RV and 3% decrease in FVC. Pulmonary function and other study endpoints were
not significantly different between the three exposure groups (control, peak and constant) at
either challenge concentration when the monkeys were rechallenged following subchronic
exposure. The authors suggested that the absence of increased pulmonary reactivity to vanadium
pentoxide following subchronic inhalation may be attributed to the development of tolerance.
42
DRAFT - DO NOT CITE OR QUOTE
-------
The study establishes a subchronic NOAEL[ADJ] of 0.10 mg/m3 (continuous exposure) for
pulmonary function. No subchronic LOAEL was established. However, an apparent acute, but
reversible, LOAEL of 1.38 mg/m3 is established based on the respiratory distress observed at
1.38 mg/m3 at an early time point.
Hematological effects of vanadium pentoxide were assessed in male CD-I mice that were
exposed by whole-body inhalation for 1 hr/day, 2 days/week for up to 12 weeks (Gonzalez-
Villalva et al., 2006). A 0.02 M aqueous solution of vanadium pentoxide was aerosolized
generating a reported average vanadium of 1436 |ig/m3 (1.44 mg/m3 V), as measured by filters
following the 12 weeks exposure. This study did not provide reliable exposure information, thus
exposure concentrations in mg/m3 could not be determined. Groups of 8 exposed mice and 8
vehicle control mice (inhaling deionized water droplets) were evaluated after 24 hrs and weekly
for 12 weeks. Evaluations consisted of a complete blood count and morphological examination
of platelets. Platelet count was significantly increased in the exposed mice on weeks 3-12;
counts increased from week 3 to a maximum at week 9 and subsequently declined, but still
remained above controls (quantitative data inadequately reported). The morphology
examinations showed the presence of giant platelets at unspecified longer exposure times. The
study establishes an apparent LOAEL for increased platelet count and altered platelet
morphology from short-term intermittent exposure to vanadium pentoxide at 2.56 mg/m3. A
continuous exposure equivalent concentration cannot be estimated with any confidence, as the
intermittency of the exposure protocol is extreme.
4.2.2.2 Chronic Studies
The toxicity of chronic inhalation exposure to particulate aerosols of vanadium pentoxide
was assessed in groups of 50 male and 50 female F344/N rats exposed (whole-body exposure) at
concentrations of 0, 0.5, 1 or 2 mg/m3 6 hrs/day, 5 days/week for 104 weeks (Ress et al., 2003;
NTP, 2002). Body weights and clinical findings were recorded throughout the exposure period.
Necropsy and comprehensive histopathological evaluation7 were performed on all animals. No
clinical findings related to vanadium pentoxide exposure were observed. Mean body weights of
females exposed to 2 mg/m3 were marginally less (3-6%; statistical significance not reported)
In addition to gross lesions and tissue masses, the following tissues were examined in the 2-yr bioassay: adrenal
gland, bone with marrow, brain, clitoral gland, esophagus, gallbladder (mice only), heart and aorta, large intestine
(cecum, colon and rectum), small intestine (duodenum, jejunum, and ileum), kidney, larynx, liver, lung and
mainstream bronchi, lymph nodes (mandibular, mediastinal, mesenteric, and bronchial), mammary gland (except
mail mice), nose, ovary, pancreas, parathyroid gland, pituitary gland, preputial gland, prostate gland, salivary gland,
skin, spleen, stomach (forestomach and glandular), testis (with epididymis and seminal vesicle), thymus, thyroid
gland, trachea, urinary bladder, and uterus.
43
DRAFT - DO NOT CITE OR QUOTE
-------
than that of controls throughout the 2-year study; mean body weights of exposed male rats were
similar to controls throughout the study. The percent survival of male and female rats for the
entire 104-week exposure period ranged from 52-58% in male rats and 30-40% in female rats.
The percent survival of controls was 40% for male rats and 28% for female rats (Table 4-7). No
infection was reported in this study that may account for these survival rates; however, these
survival rates are comparable to historical rates of survival for male and female F344/N rats.
The incidences of nonneoplastic lesions of the respiratory tract in male and female rats
are summarized in Table 4-7 (Ress et al., 2003; NTP, 2002). In male rats, the incidences of
nonneoplastic lesions of the lungs (alveolar and bronchiolar epithelial hyperplasia and alveolar
histiocyte infiltration), larynx (inflammation and epiglottis degeneration, hyperplasia and
squamous metaplasia) and nose (goblet cell hyperplasia) were significantly increased compared
to controls in all vanadium pentoxide exposure groups. In female rats, the incidences of
nonneoplastic lesions of the lungs (interstitial fibrosis and alveolar histiocyte infiltration) and
larynx (inflammation and epiglottis degeneration and hyperplasia) were significantly increased
compared to control in all vanadium pentoxide exposure groups. In general, the incidences and
severity ratings of respiratory lesions increased with exposure level. No treatment-related
histopathological findings were observed in other tissues. A LOAEL of 0.5 mg/m3 was
established for nonneoplastic lesions of the respiratory tract in male and female rats; a NOAEL
was not identified.
Table 4-7. Selected Nonneoplastic Lesions of the Respiratory System in Rats Exposed to
Particulate Aerosols of Vanadium Pentoxide for 2 Years (NTP, 2002)
Lesion Type and Location"
Exposure Group
Control
0.5 mg/m3
1 mg/m3
2 mg/m3
Male Rats
Percent survival
40
Lun
Number of animals examined)
Alveolar epithelium, hyperplasia
Bronchiole epithelium,
hyperplasia
Alveolar epithelium, squamous
metaplasia
Bronchiole epithelium, squamous
metaplasia
Inflammation, chronic active
Interstitial, fibrosis
Alveolus, histiocyte infiltration
50
7 (2.3)
3 (2.3)
1 (1.0)
0
5(1.6)
7(1.4)
22(1.3)
58 |52 |54
J
H
49
24b (2.0)
17b (2.2)
0
0
8(1.8)
7 (2.0)
40b (2.0)
48
34b (2.0)
31b(1.8)
0
0
24b(1.3)
16C(1.6)
45b (2.3)
50
49b(3.3)
48b(3.3)
21b(3.6)
7b (3/7)
42b (2.4)
38b(2.1)
50b(2.1)
44
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-7. Selected Nonneoplastic Lesions of the Respiratory System in Rats Exposed to
Particulate Aerosols of Vanadium Pentoxide for 2 Years (NTP, 2002)
Lesion Type and Location3
Exposure Group
Control
0.5 mg/m3
1 mg/m3
2 mg/m3
Larynx
Number of animals examined
Inflammation, chronic
Epiglottis epithelium,
degeneration
Epiglottis epithelium, hyperplasia
Epiglottis epithelium, squamous
metaplasia
49
3(1.0)
0
0
0
50
20b(l.l)
22b(l.l)
18b(1.5)
9b(1.7)
50
17b(1.5)
23b(l.l)
34b(1.5)
16b(1.8)
50
28b(1.6)
33b(1.5)
32b(1.9)
19b(1.9)
Nose
Number of animals examined
Goblet cell, hyperplasia
49
4(1.8)
50
15b(1.8)
49
12C (2.0)
48
17b(2.1)
Female Rats
Percent survival
28
Lun
Number of animals examined
Alveolar epithelium, hyperplasia
Bronchiole epithelium,
hyperplasia
Alveolar epithelium, squamous
metaplasia
Bronchiole epithelium, squamous
metaplasia
Inflammation, chronic active
Interstitial, fibrosis
Alveolus, histiocyte infiltration
49
4(1.0)
6(1.5)
0
0
10(1.5)
19(1.4)
26(1.4)
40 |34
30
t
49
8(1.5)
5(1.6)
0
0
10(1.1)
7b(1.3)
35C(1.3)
50
21b(1.2)
14C(1.3)
0
0
14(1.2)
12(1.6)
44b (2.0)
50
50b(3.1)
48b(3.0)
6C(3.0)
1 (2.0)
40b(1.7)
32b(1.4)
50b(1.9)
Larynx
Number of animals examined
Inflammation, chronic
Epiglottis epithelium,
degeneration
Epiglottis epithelium, hyperplasia
Epiglottis epithelium, squamous
metaplasia
50
8(1.8)
2(1.0)
0
2 (2.0)
49
26b(1.5)
33b(1.2)
25b(1.4)
7(1.9)
49
27b(1.3)
26b(1.3)
26b(1.3)
9(1.7)
50
38b(1.4)
40b(1.5)
33b(1.5)
16b(1.4)
Nose
Number of animals examined
Goblet cell, hyperplasia
50
13 (2.0)
50
19 (2.0)
50
16(1.9)
50
30b (2.0)
aNumber of animals with lesion; numbers in parentheses indicate average severity grade of lesions in affected animals:
l=minimal, 2=mild, 3=moderate, 4=marked
bSignificantly different from control by the Poly-3 test (p< 0.01)
Significantly different from control by the Poly-3 test (p< 0.05)
dNonneoplastic lesions observed at time of sacrifice; percent survival consistent with historical controls for F344 rats in NTP
45
DRAFT - DO NOT CITE OR QUOTE
-------
studies.
The NTP (2002) and Ress et. al. (2003) studies also conducted analysis of neoplasms in
rats exposed to vanadium pentoxide by inhalation for 2-years, using the protocol described
above. Compared to concurrent controls, the incidences of alveolar/bronchiolar adenoma,
alveolar/bronchiolar carcinoma or combined alveolar/bronchiolar adenoma or carcinoma were
not significantly different (Poly-3 test) for male or female rats. Compared to historical controls,
alveolar/bronchiolar adenoma in 0.5 and 2 mg/m3 males and 0.5 mg/m3 females,
alveolar/bronchiolar carcinoma in 0.5 and 2 mg/m3 males, and combined alveolar/bronchiolar
carcinoma in 0.5, 1 and 2 mg/m3 males and in 0.5 mg/m3 females were statistically significantly
different (see Table 4-8 footnotes).
NTP (2002) and Ress et al. (2003) concluded that exposure to vanadium pentoxide
caused alveolar and bronchiolar adenomas and carcinomas in male rats because incidence
exceeded historical controls. The marginal increase in lung neoplasms observed in female rats
was statistically significant only in the 0.5 mg/m3 exposure group. This increase was not
definitively attributed to vanadium pentoxide exposure since the tumors were observed only at
the lowest dose and no dose-response was evident.
46
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Incidences of Respiratory Tumors in Rats Exposed to Vanadium Pentoxide in a 2 Year
Inhalation Study (NTP, 2002)a
Tumor Type
Exposure Group
Historical
Control
Control
0.5 mg/m3
1 mg/m3
2 mg/m3
Male Rats
Number of animals examined
Al veol ar/bronchi ol ar
adenomab
Al veol ar/bronchi ol ar
carcinomad
Alveolar/bronchiolar adenoma
or carcinoma6
1054
18(10%)
8 (4%)
26 (10%)
50
4 (8%)
0 (0%)
4 (8%)
49
8 (16%)c
3 (6%)c
10 (20%)c
48
5 (10%)
1 (2%)
6 (12%)c
50
6 (12%)c
3 (6%)c
9(18%)c
Female Rats
Number of animals examined
Al veol ar/bronchi ol ar
adenomaf
Alveolar/bronchiolar
carcinoma
Alveolar/bronchiolar adenoma
or carcinoma8
1050
12 (4%)
0 (0%)
14 (4%)
49
0 (0%)
0 (0%)
0 (0%)
49
3 (6%)c
0 (0%)
3 (6%)c
50
1 (2%)
0 (0%)
1 (2%)
50
0 (0%)
1 (2%)
1 (2%)
aNumbers in parentheses indicate percent incidence; particle size mass mean aerodynamic diameter + geometric
standard deviation (MMAD±GSD): 0.5 mg/m3=1.2±2.9; 1 mg/m3=1.2±2.9; 2 mg/m3=1.3±2.9
bHistorical incidence of alveolar/bronchiolar adenoma male F344/N rats fed in inhalation chamber controls given
NIH-07 diet.
Incidence exceeds historical control (statistical comparison between NTP (2002) data and historical data not
conducted)
Historical incidence of alveolar/bronchiolar carcinoma of male F344/N rats fed in inhalation chamber controls given
NIH-07 diet.
eHistorical incidence of combined alveolar/bronchiolar adenoma or carcinoma male F344/N rats fed in inhalation
chamber controls given NIH-07 diet.
Historical incidence of alveolar/bronchiolar adenoma female F344/N rats fed in inhalation chamber controls given
NIH-07 diet.
Historical incidence of combined alveolar/bronchiolar adenoma or carcinoma female F344/N rats fed in inhalation
chamber controls given NIH-07 diet.
NTP (2002) and Ress et. al. (2003) also reported the toxicity of chronic exposure to
vanadium pentoxide in mice. Groups of 50 male and 50 female B6C3Fi mice were exposed
(whole-body exposure) to vanadium pentoxide particulate aerosol concentrations of 0, 1, 2 or
4 mg/m3, 6 hrs per day, 5 days/week, for 104 weeks (Ress et al., 2003; NTP, 2002). Particle
MMAD±GSD for each dose group was reported as follows: 1 mg/m3=1.3±2.9;
2 mg/m3=1.2±2.9; 4 mg/m3=1.2±2.9. Body weights and clinical findings were recorded
47
DRAFT - DO NOT CITE OR QUOTE
-------
throughout the exposure period. Necropsy and comprehensive histopathological evaluation were
performed on all animals and analysis of both non neoplastic and neoplastic lesions was
performed. Many mice exposed to vanadium pentoxide were thin and exhibited abnormal
breathing, particularly those exposed to 2 or 4 mg/m3 vanadium pentoxide (specific incidence
data not reported). Mean body weights were generally less than control in males exposed to
4 mg/m3 (decreases of 5-15%) and in females for all exposure groups (1 mg/m3, decreases of
4-10%; 2 mg/m3, decreases of 14-20%; and 4 mg/m3, decreases of 4-19%) (Statistical
significance not reported). The number of mice surviving for the entire 104-week exposure
period was similar to control (78% for male mice and 76% for female mice) for all exposure
groups for female mice and for males in the 1 and 2 mg/m3 groups, but survival was significantly
decreased in male mice exposed to 4 mg/m3 (50% survival rate)(Table 4-9).
The incidences of nonneoplastic lesions of the respiratory tract in male and female mice
are summarized in Table 4-9 (Ress et al., 2003; NTP, 2002). In male mice, the incidences of
nonneoplastic lesions of the lungs (hyperplasia of the alveolar and bronchiole epithelium,
inflammation, alveolus histiocyte infiltration), larynx (squamous metaplasia of the epiglottis) and
nose (olfactory and respiratory epithelium degeneration in males and olfactory epithelial
degeneration and atrophy in females) were significantly increased compared to control in all
vanadium pentoxide exposure groups. Incidences of interstitial fibrosis were significantly
increased in male and female mice exposed to 2 or 4 mg/m3. In general, the incidences and
severity ratings of lesions increased with exposure level and matched the types of lesions
observed in rats. No treatment-related histopathological findings were observed in other tissues.
The LOAEL of 1 mg/m3 was established for nonneoplastic lesions of the respiratory tract in male
and female mice; a NOAEL was not identified.
Table 4-9. Selected Nonneoplastic Lesions of the Respiratory System in Mice Exposed to
Vanadium Pentoxide in a 2 Year Inhalation Study (NTP, 2002)
Lesion Type and Location3
Exposure Group
Control
1 mg/m3
2 mg/m3
4 mg/m3
Male Mice
Percent survival
78
66
72
50b
Lung
Number of animals examined
Alveolar epithelium,
hyperplasia
Bronchiole epithelium,
hyperplasia
Inflammation, chronic
Alveolus, histiocyte
50
3 (3.0)
0
6(1.5)
10 (2.4)
50
41C (2.2)
15C(1.0)
42C(1.5)
36C (2.4)
50
49C(3.3)
37C(1.1)
45C(1.6)
45C (2.6)
50
50C(3.9)
46C(1.7)
47C (2.0)
49C(3.0)
48
DRAFT - DO NOT CITE OR QUOTE
-------
infiltration
Interstitial, fibrosis
1(1.0)
6(1.7)
9C(1.2)
12C(1.7)
Larynx
Number of animals examined
Epiglottis epithelium,
squamous metaplasia
49
2(1.0)
50
45C(1.0)
48
41C(1.0)
50
41C(1.0)
Nose
Number of animals examined
Inflammation, suppurative
Olfactory epithelium, atrophy
Olfactory epithelium,
degeneration
Respiratory epithelium,
degeneration
50
16(1.3)
6(1.0)
1 (1.0)
8(1.1)
50
11(1.4)
7(1.6)
7b(1.0)
22C(1.0)
50
32C(1.2)
9(1.3)
23b(l.l)
38C(1.2)
50
23b(1.3)
12(1.2)
30C(1.2)
41C(1.4)
Bronchial Lymph Node
Number of animals examined
Hyperplasia
40
7(2.1)
38
7 (2.4)
36
12(2.1)
40
13 (2.2)
49
DRAFT - DO NOT CITE OR QUOTE
-------
Female Mice
Percent survival
76
64
60
64
Lung
Number of animals examined
Alveolar epithelium,
hyperplasia
Bronchiole epithelium,
hyperplasia
Inflammation, chronic
Alveolus, histiocyte
infiltration
Interstitial, fibrosis
50
0
0
4(1.0)
0
50
31C(1.6)
12C(1.0)
37C(1.3)
34C (2.4)
1 (2.0)
50
38C (2.0)
34C(1.0)
39C(1.8)
35C (2.4)
4b (2.5)
50
50C(3.3)
48C(1.5)
49C (2.0)
45C (2.7)
8C(1.5)
Larynx
Number of animals examined
Epiglottis epithelium,
squamous metaplasia
50
0
50
39C(1.0)
49
45C(1.0)
50
44C(1.1)
Nose
Number of animals examined
Inflammation, suppurative
Olfactory epithelium, atrophy
Olfactory epithelium,
degeneration
Respiratory epithelium,
degeneration
Bronchial Lymph Node
(Number of animals
examined)
Hyperplasia
50
19(1.1)
2(1.5)
11(1.2)
35(1.3)
39
3 (2.0)
50
14(1.2)
8b(L3)
23C(1.0)
39(1.5)
40
13C(1.8)
50
32C(1.2)
5(1.0)
34C(1.2)
46C(1.7)
45
14C (2.3)
50
30C(1.3)
14C(1.3)
48C(1.3)
50C(1.8)
41
20C (2.3)
aNumber of animals with lesion; numbers in parentheses indicate average severity grade of lesions in affected
animals: l=minimal, 2=mild, 3=moderate, 4=marked
bSignificantly different from control by the Poly-3 test (p< 0.05)
Significantly different from control by the Poly-3 test (p< 0.01)
The incidences of tumors of the respiratory tract in male and female mice exposed to vanadium
pentoxide for 2 years are summarized in Table 4-10 (Ress et al., 2003; NTP, 2002). The
incidences of alveolar/bronchiolar adenoma, alveolar/bronchiolar carcinoma and combined
alveolar/bronchiolar adenoma or carcinoma were significantly increased in all groups of exposed
female mice. In male mice, the incidences of alveolar/bronchi olar carcinoma and combined
alveolar/bronchiolar adenoma or carcinoma were significantly increased compared to control in
all vanadium pentoxide treatment groups and alveolar/bronchiolar adenoma was significantly
increased in the 2 mg/m3 group.
50
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-10. Incidences of Respiratory Tumors in Mice Exposed to Vanadium Pentoxide in the 2
Year Inhalation Study (NTP, 2002)
Tumor Type3
Exposure Grou
Historical
Control
Control
1 mg/m3
p
2 mg/m3
4 mg/m3
Male Mice
Number of animals examined
Al veol ar/bronchi ol ar
adenomab
Al veol ar/bronchi ol ar
carcinoma
Al veol ar/bronchi ol ar
adenoma or carcinoma
1071
201 (19%)
97(9%)
285 (26.8%)
50
13 (26%)
12 (24%)
22 (28%)
50
16 (32%)
29C (58%)
42C (84%)
50
26C
(53%)
30C
(60%)
43C
(86%)
50
15 (30%)
35C (70%)
43C (86)
Female Mice
Number of animals examined
Al veol ar/bronchi ol ar
adenoma
Al veol ar/bronchi ol ar
carcinoma
Al veol ar/bronchi ol ar
adenoma or carcinoma
1075
67 (6.3%)
43 (-3.9%)
109 (-
10.1%)
50
1 (2%)
0 (0%)
1 (2%)
50
17C (34%)
23C (46%)
32C (64%)
50
23C
(46%)
18C
(36%)
35C
(70%)
50
19C(38%)
22C (44%)
32C (64%)
aNumber of animals with tumor; numbers in parentheses indicate percent incidence; particle size mass mean aerodynamic
diameter + geometric standard deviation (MMAD±GSD): 1 mg/m3= 1.3±2.9; 2 mg/m3= 1.2±2.9; 4 mg/m3=l .2±2.9
bHistorical incidence of alveolar/bronchiolar adenoma male B6C3F] mice fed in inhalation chamber controls given NIH-07 diet.
°Significantly different from control by the Poly-3 test (p<0.01)
A recent study has examined the effect of vanadium pentoxide exposure by aspiration
(Rondini et al. 2010). Rondini et al (2010) examined the induction of pulmonary inflammation
and tumor promotion in three different mice strains (A/J, BALB/c and C57BL/6J) following
oropharyngeal aspiration exposure. This study was designed to test the hypothesis that
vanadium pentoxide acts as a tumor promoter in exposed rodents. Three mouse strains were
used to further understand potential susceptibility to these effects. These particular mouse
strains were selected because of their known differential susceptibility to chronic pulmonary
inflammation and carcinogenesis: A/J mice are sensitive, BALB/C are intermediate and
C57BL/6J are resistant. The experiment was designed to measure vanadium pentoxide tumor
promotion following tumor initiation by 3-methylcholanthrene (MCA). All experimental mice
were exposed to MCA (lOug/g bw in corn oil; intraperitoneal injection) in week 1, followed by 5
weekly aspirations of either V2O5 (4 mg/kg) or PBS. Tumor incidence was measured at 20
weeks post-MCA exposure. Statistically significant lung tumor increases were observed in A/J
and BALB/C mice as compared to the MCA-treated control (p<0.05; Table 4-11). Differences
were also observed between strains, with A/J mice showing increased tumorigenicity in response
51
DRAFT - DO NOT CITE OR QUOTE
-------
to vanadium pentoxide. In the absence of MCA, V2O5 was not sufficient to initiate
tumorigenesis in this study. C57BL/6J had no tumors following exposure (data not shown). To
evaluate the strain differences in inflammation, mice were aspirated with V2O5 (4 mg/kg bw) 4
times weekly, with BALF collected at 6hr, Id, 3d, 6d, and 21d following the last aspiration.
Cellular infiltrates and protein content was compared, and lungs were snap-frozen for
histopathology. Increased pulmonary inflammation and hyperpermeability was increased in all
exposed strains in a similar pattern as the tumor induction, with greater increases observed in A/J
mice than in B ALB/C. C57BL/6J showed the least amount of increase in all parameters
compared to all mouse strains. All results returned to baseline at 21 days post-exposure. The
results of PMN increases were confirmed by histopathology in A/J and C57BL/6J mice. Further
analysis was performed to measure inflammatory chemokine production (KC, MIP-2, MCP-1)
and transcription factor activity (NFkB, c-Fos) and signaling pathway activation (MAPK). Like
the increased inflammatory markers above, increased levels of KC and MCP-1 were observed in
A/J and BALB/C mice as compared to the C57BL/6J mice. Similar strain differences were
observed for the transcription activity of NFkB and c-Fos and the MAPK signaling activity in
A/J mice as compared to C57BL/6J (BALB/C were not analyzed).
Table 4 -11. Lung tumor multiplicity in MCA-Treated mice exposed to V2O5 by pharyngeal
aspiration (Rondini et al. 2010). a'b
A/J
BALB/C
PBS Control
3.3 ±0.75
(n=4)
0.78 ±0.28
(n=8)
V205
10± 1.4
(n=15)
2.2 ±0.36
(n=12)
aNo tumors were observed in C57BL/6J mice.
bNumber of animals for each treatment in parantheses.
In summary, the identified noncancer health effects following occupational exposure via
inhalation to vanadium pentoxide in humans include respiratory irritation, cough and bronchitis;
inhalation exposure in animals results in multiple health effects, including pulmonary
inflammation, lung and nasal hyplerplasia and pulmonary fibrosis. Vanadium pentoxide
exposure for two years was associated with a wide spectrum of nonneoplastic pulmonary lesions
in both rats and mice, ranging from hyperplasia to inflammation, fibrosis, and metaplasia.
Lesions were detected in the lung, larynx, and nose in both rats and mice exposed to vanadium
pentoxide. Bronchial lymph node changes were detected in mice exposed to vanadium
pentoxide. There are currently no human epidemiology studies that examined carcinogenesis
following exposure to vanadium pentoxide. However, there is clear evidence of carcinogenicity
52
DRAFT - DO NOT CITE OR QUOTE
-------
in male and female mice exposed to vanadium pentoxide, and some evidence of carcinogenicity
in male rats, based on observations of alveolar and bronchiolar neoplasms that exceeded
historical controls in groups exposed to vanadium pentoxide (NTP 2002). A more recent study
has also shown lung tumor promotion in sensitive mouse strains (Rondini et al. 2010).
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES - ORAL, INHALATION,
INTRAPERITONEAL AND INJECTION
4.3.1. Oral Studies
Vanadium pentoxide delivered orally to weanling rats (10 to 200 |j,mol/kg) for three days
produced a significant increase in alkaline phosphatase activity and DNA content in the
diaphysis of femoral bones, and suggests that vanadium pentoxide may be linked to bone
formation in the developing rat (Yamaguchi et. al., 1989). Yamaguchi et al. (1989) examined
the potential effects of vanadium pentoxide on bone metabolism. Weanling male Wistar rats
were exposed to 1.8 to 36.4 mg/kg (10 to 200 |j,mol/kg)8 vanadium pentoxide orally 1 hr
following an oral injection of 15.3 (imol Zn/lOOg) for three times at 24-h intervals. The highest
dose vanadium pentoxide tested (200 |j,mol/kg) led to death in four of nine rats (cause of death
not described). Authors state that administration of zinc as well blocked these deaths. All rats
were killed 24h following the last vanadium administration, and blood immediately removed by
cardiac puncture. Statistically significant increases were observed in serum from high-dose rats
for calcium (27.3 - 36.4 mg/kg; p < 0.05) and decreased for phosphorus (36.4 mg/kg; p < 0.01).
Administration of zinc completely prevented these serum changes. Femurs were also
immediately removed. The diaphysis and epiphysis were used for alkaline phosphatase
measures (right femur) and DNA content analyses (left femur). Alkaline phosphatase activity
was significantly increased at the lowest dose tested (1.8 mg/kg; p< 0.05) and peaked at 3.6
mg/kg. Further increasing doses led to decreases in alkaline phosphatase activity. A similar
pattern was observed for bone DNA content, with statistically significant increases at 1.8 - 18.1
mg/kg (p < 0.05), but decreased at higher doses. Although the authors state the interaction of
vanadium and zinc led to an increase in bone calcium, based on the data presented in the figures
in this publication, vanadium administration did not lead to alterations of bone calcium. Overall,
this study shows an increase in both DNA content and alkaline phosphatase activity in weanling
male rats following oral administration of vanadium pentoxide, suggesting a possible role of
vanadium in increased bone growth.
1 Conversion factor: umol = 0.1819 mg
53
DRAFT - DO NOT CITE OR QUOTE
-------
Mravcova et. al (1993) assessed the extent of vanadium pentoxide accumulation in the
bones of rats following 6 month exposure (full study description in 4.4.4.2). Vanadium
accumulated in the epiphyseal cartilage of the tibia in rats with significantly higher
concentrations of vanadium in the tibia and incisors of weanling rats compared to adults.
However, no dose response data for these endpoints was reported.
4.3.2. Inhalation Studies
Investigations of the reproductive and developmental toxicity of subchronic or chronic
inhalation exposure to vanadium pentoxide are limited to two studies. Mussali-Galante et. al
(2005), used immunohistochemistry to assess the amount of gamma tubulin accumulating within
somatic and testicular germ cells. Mussali-Galante et al. (2005) exposed 60 male CD-I mice to
inhaled vanadium pentoxide (0.02M, apparently aqueous solution containing aerosolized
vanadium pentoxide) for 1 hr two times a week for 12 weeks. Avila-Costa et. al. (2004),
investigators from the same laboratory, used the same protocol and reported that droplets of the
vanadium pentoxide mixture had average diameters of 0.5 - 5 |j,m (Avila-Costa et. al., 2004).
Thirty-six control animals inhaled only vehicle (deionized water). Groups of three exposed
animals and three control animals were sacrificed per week for 12 weeks. Results indicated
accumulation of vanadium pentoxide in testes (Mussali-Galante et al., 2005). Moreover, gamma
tubulin was significantly decreased in testicular samples exposed to vanadium pentoxide
compared to control, starting after the first week of exposure. Changes in gamma tubulin may
suggest changes in microtubule-involved function, such as cell division, which may impact
spermatogenesis. Responses were duration-dependent, with the lowest percentages of
immunoreactive cells occurring at the end of the exposure period; values at week 12 ranged from
1.2% for germ cells and 1.5% for Sertoli cells to 10.1% for Leydig cells (compared to 87-88% in
controls) (Mussali-Galante et. al., 2005).
Fortoul et al. (2007) analyzed testes for ultrastructural changes, testosterone
concentration, and vanadium tissue concentration, using the same protocol as that reported above
(Mussali-Galante et al 2005). No overt toxicity or changes in body or testicular weight were
observed. Histopathological analysis revealed necrotic cell death in spermatocytes (25%) and
Sertoli cells (15%) at weeks 5-6 in vanadium-exposed animals. Spermatocytes exhibited
cytoplasmic vacuolation, nuclear distortion and intercellular edema in response to vanadium
pentoxide exposure. Spermatogonia (40% necrosis during weeks 6-7) were the most susceptible
cell type, followed by spermatocytes and Sertoli cells. Moreover, vanadium pentoxide
concentrations increased dramatically after one week of exposure and remained consistently
elevated (avg concentration 0.05 |j,g/g dry tissue in controls, 1.63 |j,g/g dry tissue in exposed
54
DRAFT - DO NOT CITE OR QUOTE
-------
animals, 33 times higher). Concentrations of testosterone were highly variable and not
statistically significant.
Vanadium pentoxide exposure did not affect reproductive endpoints in male rats (sperm
count, spermatid heads, sperm motility), but it did increase estrous cycle length by 10% in
female rats exposed to 8 mg/m3, but not to 16 mg/m3, and reduced the number of cycling females
in surviving rats in the 16 mg/m3 group (percent reduction not reported). The epididymal
spermatozoal motility of male mice exposed to 8 or 16 mg/m3 was significantly decreased by 13
and 5%, respectively. No treatment-related effects were observed for assessments of estrous
cycle in female mice (estrous cycle length and number of cycling females) (NTP 2002).
4.3.3. Intraperitoneal and Injection Studies
Male and female reproductive endpoints were evaluated in young rats following
intraperitoneal administration of vanadium pentoxide (Altamirano et al., 1991). Newborn male
and female rats were injected with 0 or 12.5 mg/kg vanadium pentoxide in saline on every
second day from birth to age 21 days; groups sizes were 5 (treated males) or 9 (male and female
controls and treated females). Males were sacrificed at 55 days of age and the females were
sacrificed on the day of first vaginal estrus. Other groups of females were injected with 0 or 12.5
mg/kg-day vanadium pentoxide (n = 10 and 6, respectively) from age 21 days to the day of first
vaginal estrus, at which time they were sacrificed. Reported endpoints in the males consisted of
absolute weights of testes, prostate, seminal vesicles, adrenals, pituitary, thymus, liver, kidneys
and submandibular glands. The only effects in treated males were statistically significant
increases in seminal vesicle, thymus and submandibular gland weights (20.1, 29.5 and 19.2%
higher than controls, respectively). Endpoints evaluated in the females included body weight,
absolute organ weights (ovaries, uterus, adrenals, pituitary, thymus, liver, kidneys and
submandibular glands), age at vaginal opening, number of ova in oviducts and ovulation rate.
The only effects in treated females occurred in the group treated from 21 days of age; these
consisted of statistically significant increases in body weight (14.3% higher than controls) and
increased weights of thymus, submandibular gland and liver (31.1, 15.8 and 28.4% above control
weights, respectively).
Fertility and sperm assessments were also performed in male CD-I mice following
intraperitoneal administration of vanadium pentoxide (Altamirano-Lozano et al., 1996). In the
fertility assessment, groups of 20 and 15 male mice were injected with 0 and 8.5 mg/kg
vanadium pentoxide in saline, respectively, every 3rd day for 60 days and mated 24 hrs after the
last injection. Statistically significant effects in the treated group included reduced fertility rate
in males (33% compared to 85% in controls), reduced numbers of implantation sites (avg = 5.8
55
DRAFT - DO NOT CITE OR QUOTE
-------
compared to 10.88 in controls), reduced numbers of live fetuses (3.4 compared to 10.53 in
controls) and increased number of resorptions per dam (2.00 compared to 0.24 average
resorptions in controls). In the sperm assessment, 20 males were injected with 8.5 mg/kg
vanadium pentoxide every 3 days for up to 60 days with groups of 5 evaluated after 10, 20, 30,
40, 50 or 60 days of treatment. Statistically significant effects included reduced sperm motility
after >10 days, reduced sperm count and increased percentage of abnormal sperm after >20 days,
decreased absolute testicular weight after >50 days (relative weight not reported), and decreased
body weight after 60 days.
Developmental toxicity was evaluated in groups of 13 or 15 female CD-I mice that were
administered 0 or 8.5 mg/kg vanadium pentoxide in distilled water, respectively, by
intraperitoneal injection on days 6-15 of gestation (Altamirano-Lozano et al., 1993). No
maternal toxicity was reported (endpoints not specified). Developmental toxicity endpoints were
assessed on gestation day 18. The endpoints included the number of implants, resorptions, and
live fetuses. For all fetuses, weight, sex, and external malformations were noted. For two-thirds
of the fetuses, skeletal abnormalities were also recorded. Internal soft-tissue examinations do
not appear to have been conducted. The treated group had statistically significant increases in
the number of litters with abnormal fetuses (9/15 compared to 3/13 in controls), number of
abnormal fetuses (15/149 compared to 3/124), and number of fetuses with short limbs (8/149
compared to 0/124 in controls). Additionally, the numbers of ossification centers in forelimbs
and hindlimbs were significantly reduced in the treated fetuses.
Zhang et al. (199la) evaluated the developmental toxicity in NIH mice following
intraperitoneal injection of 5 mg/kg-day vanadium pentoxide on days 1-5, 6-15, 7, 8, 9, 10, 11 or
14-17 of gestation. There were no adverse effects on preimplantation or implantation,
developmental toxicity, or premature births. Increased frequencies of resorption or fetal death
were observed for gestation days 6-15, 7 and 14-17. Delayed ossification (sites not specified)
was observed for gestation days 6-15, 8, 10 and 14-17. In a second study, Zhang et al. (1993a)
evaluated developmental toxicity in Wistar rats following intraperitoneal injection of 0.33, 1.0 or
3.0 mg/kg-day on days 6-15 of gestation. Decreased placental weight and increases in embryo-
fetus mortality and external or skeletal malformations (unspecified) occurred at 1.0 and 3.0
mg/kg-day. Maternal toxic symptoms (unspecified), decreased maternal weight gain during
treatment and fetal growth retardation were observed at 3.0 mg/kg-day. In the third study, Zhang
et al. (1993b) evaluated developmental toxicity in Wistar rats following intraperitoneal injection
of vanadium pentoxide in doses of 3 mg/kg-day on days 6-15 of gestation or 5 mg/kg-day on
days 9, 10, 11 or 9-12 of gestation. Effects in rats exposed on gestation days 6-15 and 9-12
included decreased maternal weight gain, increased fetal mortality, decreased fetal weight and
56
DRAFT - DO NOT CITE OR QUOTE
-------
crown-rump length, delayed ossification of unspecified bones, and increased incidences of
subcutaneous hemorrhage, wavy ribs, and dilation of lateral ventricles and renal pelvis. Effects
in rats exposed on a single day of gestation included subcutaneous hemorrhage and unspecified
visceral anomalies following exposure on days 9, 10 and 11, and increased fetal mortality and
delayed ossification of unspecified bones following exposure on day 10. Additional study
details were not available. Overall, the three studies identified an intraperitoneal developmental -
toxicity LOAEL for vanadium pentoxide of 1 mg/kg-day (Zhang et al., 1991a, 1993a,b).
One study examined the effects of vanadium pentoxide (l.lmg/kg in distilled water)
following tail vein injection in a total of 20 pregnant NMRI mice (data pooled from three studies
with 6-10 mice each) (Wide 1984). Injections were performed before implantation or on
gestation day 3 or 8, and animals were euthanized on gestation day 17 (two days before
parturition. Pre-implantation exposure had no effect on the fetuses as regards number per litter,
weight, or external and internal morphology. Exposure to vanadium pentoxide on gestation day
3 or 8 did not lead to significant changes in resorption frequencies, fetal weights or frequencies
of fetal hemorrhages as compared to controls. However, the number of fetuses defined as having
less mature skeletons9 by the authors was significantly greater in mice given vanadium
pentoxide on gestation day 8, but not gestation day 3 (p < 0.001, chi square test).
Although the intraperitoneal and tail vein injection studies show the potential of
vanadium pentoxide to cause reproductive and developmental effects in rodents, the studies are
of little use in the quantitation of vanadium pentoxide toxicity, as equivalent oral or inhalation
exposures cannot be established.
4.4 OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES
4.4.1 Acute and Short-Term Studies
4.4.1.1 Acute Studies
4.4.1.1.1 Oral
According to the Concise International Chemical Assessment Document 29 (WHO-IPCS,
CICAD 29, 2001), rat oral LDso values for vanadium pentoxide range from 86-137 mg/kg body
weight (Yao et. al., 1986 as cited in WHO-IPCS, 2001). A rat oral LD50 value of 10 mg/kg body
weight was reported in IARC Monographs, volume 86 (IARC, 2006) in a study by Lewis et al.
9 No ossification of three of four elements examined (supraoccipital bone, sternum, metatarsalia, and all caudal
vertebrae).
57
DRAFT - DO NOT CITE OR QUOTE
-------
(2000). Lewis et al. (2000) also reported a mouse oral LD50 value at 23 mg/kg body weight.
Clinical signs of acute toxicity included lethargy, excessive tearing (lacrimation), and diarrhea
but dose-response data was not reported (Yao et. al., 1986 as cited in WHO-IPCS, 2001).
Histopathological analysis revealed liver necrosis and swelling of renal tubules. Other studies
(WHO-IPCS, 2001) have identified LD50s of-10 mg/kg body weight in rats and 23 mg/kg body
weight in mice. An oral LDso at 64 mg/kg body weight in rabbits was established (WHO-IPCS,
2001). Signs of toxicity in rabbits mimicked those reported for rats.
4.4.1.1.2 Inhalation
A 1-hr inhalation exposure to vanadium pentoxide dust in rats led to an LCe? of 1.44
mg/L (1440 mg/m3) (US EPA, 1992). Clinical signs of toxicity included respiratory difficulty,
irritation of mucosa, and diarrhea (WHO-IPCS, 2001). Knecht et al. (1985) reported air flow
restriction, as measured by pulmonary function tests, in sixteen adult male cynomolgus monkeys
(Macaco, fasicularis) exposed to vanadium pentoxide by whole-body inhalation at 5.0 mg/m3 for
6 hrs but not at the lower dose tested (0.5 mg/m3). From this study, a LOAEL of 5.0 mg/m3 and
a NOAEL of 0.5 mg/m3 were established. The lung was also identified as a target organ in
response to acute inhalation exposure to vanadium pentoxide. Following a baseline
measurement of pulmonary function, each of sixteen male cynomolgus monkeys were exposed
to aerosols of 0.5 mg/m3 vanadium pentoxide by whole-body inhalation for 6 hrs (Knecht et al.,
1985). One week later, these monkeys were exposed to aerosols of 5 mg/m3 vanadium pentoxide
by whole-body inhalation for 6 hrs. Effects on airway function were evaluated in monkeys by
comprehensive pulmonary function tests (PFTs) performed 24 hours post-exposure to 0.5 and
5 mg/m3 and on pulmonary inflammation by analysis of bronchiolar lavage (BAL) fluid in
monkeys performed after exposure to 5 mg/m3. Significant changes in pulmonary function
parameters compared to baseline values were observed only following exposure to 5 mg/m3 as
follows: 16% increase in pulmonary resistance; 11% decrease in peak expiratory flow rate;
5-22% decreases in forced expiratory flow maneuvers; 33% increase in residual volume; and
24% increase in forced residual capacity. Results are consistent with air-flow limitation in both
small peripheral and large central airways. An increase (approximately 87%; data presented
graphically) in the total number of cells recovered in BAL fluid was observed 1 day after
exposure to 5 mg/m3 vanadium pentoxide. The increase in BAL fluid total cell number was
primarily due to a marked increase (approximately 425%; data presented graphically) in the
number of polymorphonuclear leukocytes. Results suggest that pulmonary inflammation and
release of bronchoconstrictive mediators from inflammatory cells may play a role in vanadium
58
DRAFT - DO NOT CITE OR QUOTE
-------
pentoxide-induced air-flow restriction. An acute (single 6-hr exposure) LOAEL for vanadium
pentoxide of 5 mg/m3 for pulmonary function in monkeys was established in this study, with a
NOAELof0.5mg/m3.
A study in male CD-I mice (n = 48) by Avila-Costa et al. (2006) noted significantly
impaired performance on memory tasks, significantly decreased dendritic spine length, and
significant increases in percentages of necrotic cells in the hippocampus compared to controls
following a 1 hr inhalation exposure to 0.02 M (2.5 mg/m3 as Vanadium) vanadium pentoxide (p
< 0.05). The dose-response relationship for these effects could not be evaluated, since only one
dose was tested, and a NOAEL could not be established.
4.4.1.2. Short-term Studies
4.4.1.2.1 Inhalation and Aspiration
The primary noncancer health effect of short-term inhalation exposure in humans is
respiratory irritation where 100 workers were reportedly exposed to 0.05 to5.3 mg/m3 vanadium
for 10 hr/day, 6 days/week, for 4 weeks (Levy et al., 1984). A LOAEL of 0.05 mg/m3 was
established. However, dose-response was not systematically measured, there were no controls,
and exposure due to vanadium pentoxide could not be directly correlated to effects. The primary
noncancer health effects of short term inhalation exposure in animals include increased
pulmonary inflammation, and dose-related decreases in body weight and relative lung weight in
rodents (NTP, 2002).
Results of the NTP (2002) study in rats and mice provide evidence of toxicity to the
upper and lower respiratory tract, including increased lung weight, inflammation, nonneoplastic
lesions, and decreased pulmonary function following 13-day or 16-day inhalation exposure to
vanadium pentoxide. A significant increase in pulmonary inflammation and histiocytic infiltrate
of minimal to mild severity was observed in female rats (assessments not made in male rats)
exposed to vanadium pentoxide for 13 days, with a LOAEL of 1 mg/m3; a NOAEL was not
established. Similar results were observed for female mice (assessments not made in male mice)
exposed for 13 days, with a LOAEL of 2 mg/m3 for minimal to mild epithelial hyperplasia and
inflammation; a NOAEL was not established.
Male rats (22/group) and female mice (50/group) were also assessed for pulmonary
inflammation (bronchiolar lavage analysis) and systemic immunotoxicity (pulmonary
bacteriocidal activity) following exposure to 0, 4, 8, and 16 mg/m3 vanadium pentoxide for 16-
days (NTP, 2002). Observed effects included significant alterations in the percentage of
59
DRAFT - DO NOT CITE OR QUOTE
-------
recoverable bronchial lavage cells (macrophages and neutrophils) (LOAEL of 8 , mg/m3 and
NOAEL of 4 mg/m3.), and increased lung protein and lysozyme in male rats (LOAEL of 4
mg/m3). No NOAEL was established. In female mice exposed to vanadium pentoxide for 16
days, there was a localized inflammatory response in the lungs based on increase in
lymphocytes, protein and lysozymes at all concentrations; the NOAEL was not established. A
significant decrease in the percentage of macrophages from bronchiolar lavage fluid led to a
LOAEL of 8 mg/m3 and a NOAEL of 4 mg/m3. These responses, NOAEL, and LOAELs are
commensurate with those observed in the 3-month study in rats and mice of both genders (NTP,
2002). Thus, the lowest concentrations at which adverse effects were observed were 1 and 2
mg/m3 (LOAEL) for lung inflammation in female rats and mice, respectively, exposed to
vanadium pentoxide for 13-days.
An additional 5 male and 5 female mice were exposed by inhalation to vanadium
pentoxide for 6 hrs per day, 5 days a week for 16-days at concentrations of 0, 2, 4, 8, 16, or 32
mg/m3 (NTP, 2002). All male mice exposed to 32 mg/m3 died before study completion. Body
weight was significantly decreased in male and female mice at 16 and 32 mg/m3, respectively.
Absolute lung weights were significantly increased in a dose-dependent manner in males at > 4
mg/m3, and relative lung weight were significantly increased in males at > 2 mg/m3. Among
females both absolute and relative lung weights increased in all exposure groups establishing the
LOAEL of 2 mg/m3; no NOAEL was established.
Additional groups of 40-60 female mice were exposed to 0, 2, 4 or 8 mg/m3 for 6 hrs per
day, 5 days per week for 16 days (NTP, 2002). The nonneoplastic lung lesions noted on day 6
and 13 consisted of hyperplasia of the alvelolar and bronchiolar epithelium at all exposure levels.
Increase in severity of lesions was correlated with increasing concentration and time. The
LOAEL for nonneoplastic lung lesions was 2 mg/m3. A NOAEL was not established.
A duration-dependent decrease in the number of immunoreactive TH+ neurons (Avila-
Costa et al., 2004) after 4 weeks and increased quantities in metalloproteinase (MMP)-2 and
MMP-9 in CNS after 8 weeks in male mice (Colin-Barenque et. al. 2008) following twice
weekly, one hour inhalation exposure to 5.13 mg/m3 vanadium pentoxide were suggestive of
disruption of blood-brain barrier .
Turpin et al. (2010) exposed male AKR mice to vanadium pentoxide by intranasal
aspiration following exposure to respiratory syncytial virus (RSV) to determine if pre-exposure
to the virus exacerbated the vanadium pentoxide-induced lung inflammation and fibrosis.
Animals were exposed intranasally to RSV (6 x 105 PFU in lOOul PBS) on day -1 and day 8,
then exposed intranasally to X^Os (4mg/kg in 50ul PBS) on day 0 and day 7. One hour before
euthanasia, animals were given BrdU (50 mg/kg) by i.p. for analyzing cell proliferation in
60
DRAFT - DO NOT CITE OR QUOTE
-------
bronchus-associated lymphoid tissue (BALT). Lungs were lavaged with PBS and BALF
collected for analyzing differential cell counts (neutrophils, macrophages and lymphocytes).
Total RNA from lung was analyzed by real time RT-PCR for mRNAs coding for pro-fibrogenic
growth factors TGF-|3-1, connective tissue growth factor (CTGF), platelet-derived growth factor-
C (PDGF-C) and collagen CollA2, anti-fibrogenic type I interferons-alpha (IFN-a) and -beta
(IFN-p) and IFN-inducible chemokines CXCL9 and CXCL10.
Lung sections stained with Masson's trichrome staining to show collagen had an
inflammation score of two, representing mild fibrosis with vanadium pentoxide exposure alone.
However, vanadium pentoxide-induced fibrotic response was less severe in the lungs of mice
which received either pre- or post-RSV exposure, and no difference observed in pre- or post-
RSV alone and the negative controls. In addition, BALF from mice exposed to vanadium
pentoxide alone or with RSV-post exposure had significantly higher total cell count compared to
RSV pre-exposure, RSV pre-exposure plus vanadium pentoxide or controls. In particular,
vanadium pentoxide alone caused a significant increase in the levels of neutrophils and
lymphoctyes compared to controls. Both pre- and post-exposure to RSV led to a decrease in the
severity of V2O5-induced fibrosis, and gene expression analysis showed decreases in several
pro-fibrinogenic genes associated with innate immunity.
The acute and short-term studies described here support those results describe previously
in the subchronic and chronic studies, and suggest progressive lung effects from vanadium
pentoxide exposure. These further support that the lung is the most sensitive organ to vanadium
pentoxide exposure.
4.4.2 Immunological Endpoints
4.4.2.1 Human Studies
Some of the early case series observed dermatitis among affected workers employed at
vanadium pentoxide processing facilities (Sjoberg, 1951; Zenz et al., 1962). Zenz et al. (1962)
observed that respiratory symptoms, such as conjunctivitis, nasopharyngitis, hacking cough, fine
rales, and wheezing recurred with greater severity when work resumed after three days of no
exposure, even with the use of respirators, perhaps indicating immune system involvement.
Kiviluoto published a series of reports regarding an investigation in 1975 of respiratory
symptoms and clinical findings among employees (process workers, repairmen, foremen, and a
laboratory worker) at a factory making vanadium pentoxide from magnetite ore (Kiviluoto, 1980;
Kiviluoto et al., 1979; Kiviluoto et al., 1981a). A higher proportion of the exposed group (N =
63) had an elevated number of neutrophils in nasal smears compared to the referent group (N =
61
DRAFT - DO NOT CITE OR QUOTE
-------
63) who were employed at the magnetite ore mine and matched to the exposed by age and
smoking habit (35% versus 7%, p<0.001, N=55 pairs). In biopsies of the nasal mucosa, a higher
proportion in the exposed groups had elevated plasma cells and round cells (26% versus 0%,
p<0.05, N=57 pairs and 48% versus 29%, p<0.05, N=56, respectively). The prevalence of
elevated numbers of eosinophils did not vary by exposure, leading the authors to conclude that
the cytological and histological response in the vanadium-exposed group was an irritant, not
allergic response. Vanadium concentrations in the breathing zone averaged 0.028 mg/m3 (TWA)
with a range of 0.002 - 0.42 mg/m3. Higher concentrations were found where grinding and
packing of smelt were conducted (TWA (range): 2.3 mg/m3 (one sample) and 0.13 mg/m3 (0.02 -
0.37).
Motolese et. al. (1993) assessed whether exposure to vanadium pentoxide in the ceramics
industry was associated with contact dermatitis or contact sensitization. One hundred and twenty-
six enamellers and sixty-four decorators from five ceramics factories were tested for contact
sensitization using skin patch testing after exposure to a variety of substances, including
vanadium pentoxide. Among the 190 workers under study, twenty-two individuals were found to
have dermatitis and 17 reported having had skin lesions in the past. One worker responded with a
positive skin patch test indicating sensitization to a 10% solution of vanadium pentoxide.
A pilot study using nasal lavage to evaluate an inflammatory response to fuel oil ash
exposure, did not find an association of several exposure indices of vanadium or PMio exposure
with counts or percentages of polymorphonuclear cells, eosinophils, and epithelial cells (Hauser
et al., 1995b). Thirty-six out of 50 volunteers with no symptoms of cold or flu provided a nasal
lavage sample both at baseline after at least 36 hours away from work and after 72 hours of
exposure at a local electric company. A total of 19 boilermakers involved in the overhaul of a
large oil-fired boiler and 18 utility workers, full-time employees of the power company with
lower exposure to fuel oil ash, were studied. Daily exposure estimates for PMio (< 10 urn) and
vanadium (adjusted for filter extraction efficiency) were assigned to each individual using data
from personal air sampling (1-10 hour TWA) and a self-completed work diary completed by
each participant listing tasks and job locations during the day. A total of 29 task/location
exposure categories were identified, but only 3 or fewer samples were available to estimate
concentrations at 23 of them. Environmental PMio concentrations based on personal sampling
were 50 to 4510 ug/m3. Concentrations of respirable vanadium dust were 0.10 to 139.2 ug/m3.
Compared to baseline, the number of polymorphonuclear cells/ml recovered nasal fluid (adjusted
by dividing the change by the mean of the baseline and postexposure value) increased by 40%
(SD 100%, p < 0.05) with a range of-89% - 200%. The adjusted number of epithelial cells/ml
recovered nasal fluid increased by 26.7% (SD 81.4%, p > 0.05) with a range of-83% - 200%.
62
DRAFT - DO NOT CITE OR QUOTE
-------
Regression models controlled for age and smoking. Models did not control for ozone levels or
respirator use. The wide variation in the change in cell counts after exposure, especially among
nonsmokers, and the small number of subjects may have precluded the detection of an
association with vanadium or particulate exposure. Alternatively, the authors considered the
levels of vanadium dust to be low, possibly not enough to cause inflammation in these workers.
Woodin et. al (1998) analyzed nasal lavage fluid from 18 boilermakers before, during,
and after the overhaul of a large, oil-fired boiler over a six week period from mid-May, 1995 to
late-June, 1995. Biomarkers of upper airway inflammation in nasal lavage fluid were more
prevalent among boilermakers during the overhaul compared to 11 utility workers (Woodin et al.,
1998). Interleukin-8 (pg/ml) levels increased from a mean of 93.7 pg/ml (22.6-235.0) before the
overhaul to 140.9 pg/ml (32.4-307.0) during the overhaul (p < 0.05), and decreased to levels
comparable to those before the overhaul two weeks later (mean (range): 89.0 pg/ml (20.3-75.0)).
Interleukin-8 levels among utility workers did not change substantially during the overhaul (mean
(range): 69.2 (24.6-104.5), 58.5 (14.3-108.4) and 47.5 (12.8-74.7), respectively.
Myeloperoxidase levels also increased among boilermakers during the overhaul, but not among
utility workers. Among boilermakers, myeloperoxidase levels (ng/ml) were 22.7 (2.0-72.8), 33.9
(2.0-103.0) and 24.2 (3.9-58.1) before, during and after the overhaul, respectively (before versus
during: p < 0.05). Among utility workers, myeloperoxidase levels were 25.6 (10.1-47.6), 27.2
(4.9-66.2) and 25.6 (4.9-51.7), respectively. Mean IL-6 and eosinophilic cationic protein levels
did not change during the overhaul work suggesting that the inflammatory response was not due
to an allergy or respiratory infection. During the boiler work, vanadium levels rose to a
geometric mean (SD) of 8.9 (2.3) ug/m3 inside the boiler but did not change appreciably outside
the boiler where the utility workers were located (geometric mean (SD) ug/m3: 1.4 (1.6) (p <
0.001)). However, vanadium concentrations in nasal lavage fluid were not associated with levels
of either IL-8 or myeloperoxidase using Spearman's Rank Order Correlation Test.
Hauser et. al. (2002) reported on a prospective cohort study of 118 boilermaker
construction workers (see Section 4.1.2). The participants provided information on their work at
oil, gas and coal-fired powerplants in annual work history questionnaires, but exposure to specific
components in the combustion particles were not quantified. Spirometry was used to measure
lung function. Among the cohort, 6 workers reported having asthma diagnosed by a physician
and 18 workers reported having symptoms of chronic bronchitis. Workers with asthma or chronic
bronchitis experienced greater reductions in annualFEVi associated with the number of hours
worked at gas or coal-fired powerplants during the year compared to the other workers. The
generalized estimating equation models adjusted for age and smoking status, and interaction
terms were statistically significant. In contrast, models adjusting for age, baseline FEVi, and
63
DRAFT - DO NOT CITE OR QUOTE
-------
cigarette smoking status did not find a similar effect of airway responsiveness, measured by
methacholine challenge, on exposure-related reductions in FEVi. The annual reduction in FEVi
associated with the number of hours worked at gas, oil or coal-fired powerplants was similar
between workers with and without reactive airways.
In conclusion, case studies of occupational exposure to vanadium pentoxide dust or
ROFA have reported individuals with dermatitis, positive skin patch reactions and bronchial
reactivity, although this does not appear to be a common occurence. Although increases in the
numbers of inflammatory cells in nasal smears or nasal lavage fluid have been observed in
exposed workers, no associations were observed in relation to estimates of respirable vanadium
dust or vanadium concentrations in nasal fluid. The authors did not report increases in
eosinophils suggesting that the inflammation was due to irritation and was not an allergic
response.
4.4.2.2 Animal Studies
Pinon-Zarate et. al. (2008) exposed 112 male CD1 mice to -1.4 mg/m3 vanadium
pentoxide by inhalation for 1 hr/day, 2x a day, for 12 weeks as measured by filters following
exposure. This study did not provide reliable exposure information, thus exposure concentrations
in mg/m3 could not be more specfically determined. Spleen weight and histology were
determined. B-lymphocytes in the spleen were identified by immunohistochemical staining for
CD 19 (a cell surface marker which acts as a co-receptor for other CD markers). In addition, eight
control and eight vanadium pentoxide-treated mice were immunized with Hepatitis B Surface
antigen, a well-known T-cell dependent antigen. Avidity to the resulting antibody was measured.
Spleen weight of vanadium-pentoxide exposed animals increased significantly and peaked at 9
weeks, and then decreased significantly. Splenic germinal centers were significantly increased in
vanadium pentoxide-treated mice and contained a significantly increased number of CD 19+ cells
compared to controls. The authors suggest that vanadium pentoxide does not act as a direct
antigen and does not induce a "host humoral response". These data suggest that vanadium
pentoxide may affect the avidity of antibodies - the ability of antibodies to bind effectively to
substrates (affinity) and to engage multiple epitopes of the antigen at once.
Mravcova et al (1993) conducted experiments to determine effects of subchronic
exposure to low doses of vanadium pentoxide on the immune system. Weanling and adult male
and female Wistar rats (n= 10 per group) were given vanadium pentoxide in drinking water (0, 1,
100 mg/L or 0, 0.14, 14 mg/kg-day) for 6 months. Also, male and female ICR mice (groups of
10) were given vanadium pentoxide (0 or 6 mg/kg-day) by gavage 5x a week for 6 weeks.
64
DRAFT - DO NOT CITE OR QUOTE
-------
Immunotoxicity endpoints included spleen and thymus weight, spleen cellularity, number of
peripheral white blood cells, phagocytosis and natural killer cell activity, the extent of plaque-
forming cells being converted to T-dependent antigen, and several cell-mediated immunity
endpoints (Concanavalin A (ConA), pokeweed mitogen responsiveness (PWM), and
phytohemagglutinin responsiveness (PHA) assays). Of these endpoints, spleen weight was
significantly elevated over controls at 14 mg/kg-day in rats. The ConA and PHA assays
illustrated significant cell-mediated immune activation at 1 mg/L in rats over control (3 and 2.5
times higher than control values, respectively) but statistical significance was not indicated or
reported. At the high dose (14 mg/kg-day vanadium), ConA and PHA assay results were close
to control values. Thus, no dose-response pattern was detected. The low dose response may be
a transient or compensatory response. Results of other endpoints were not reported. No
significant differences in these parameters were reported in mice. The authors suggested that the
high Con A response of T suppressor cells indicate that vanadium pentoxide may induce
hypersensitivity responses at low doses.
Immunological endpoints were also analyzed as part of a pulmonary study using
cynomolous monkeys weekly provocation challenges (single 6-hr exposures to 0.5 or 3.0
mg/m3). Inhaled vanadium pentoxide aerosol for six weeks produced statistically significant
pulmonary responses, prior to a subchronic exposure (6 hrs/day, 5 days/week for 26 weeks)
(Knecht et al., 1992; study details in Sec 4.2.2.1). Immunological analyses of blood and
bronchiolar lavage fluid, and skin sensitivity tests were conducted before the pre- and post-
exposure provocation challenges. Bronchiolar lavage fluid analyses were also performed one
day after the provocation challenges. Cytological endpoints included complete and differential
blood cell counts and leukotriene C$ levels. Immunological endpoints included total IgE, total
IgG, albumin and total protein. The skin sensitivity tests assessed immediate and delayed
responses to intradermal injections of vanadium pentoxide-monkey serum albumin conjugate.
BAL fluid analysis showed a significant influx of inflammatory cells (polymorphonuclear
leukocytes) into the lung. Other study endpoints were not significantly different between the
three exposure groups (control, peak and constant) at either challenge concentration when the
monkeys were rechallenged following subchronic exposure.
In summary, immunological effects of vanadium pentoxide exposure have not been
comprehensively studied. The studies described here have not shown a statistically significant
response in the endpoints tested, however, some effects were observed. These results are
therefore inconclusive.
4.4.3 Neurological Endpoints
65
DRAFT - DO NOT CITE OR QUOTE
-------
4.4.3.1 Human Studies
No studies on the neurological effects of vanadium pentoxide were reported in humans.
4.4.3.2 Animal Studies
Pazynich (1966) exposed 33 male albino rats (species not specified) to 0.027 mg/m3 or
0.002 mg/m3 aerosolized vanadium pentoxide "round the clock" for 70 days. A third group of
rats served as control. The animals were evaluated for general condition, body weight, motor
chromaxy of antagonistic muscles, whole blood cholinesterase activity, and oxyhemoglobin
content. Motor chromaxy of extensor muscles decreased significantly (p<0.01) while that of
flexor muscles increased (p<0.001) in animals exposed to 0.027 mg/m3. Blood cholinesterase
levels were statistically significantly reduced after exposure to 0.027 mg/m3 and the reduction
persisted throughout the 90-day recovery period. There was a statistically significant reduction
in venous oxyhemoglobin in rats from the 0.027 mg/m3. Recovery was observed after 20 days.
No difference in these parameters was reported in the 0.002 mg/m3 group compared to controls.
These results suggest a LOAEL of 0.027 mg/m3 for hematological and CNS effects in albino rats
with a NOAEL of 0.002 mg/m3.
Three recent studies by Avila-Costa et al. (2004, 2005, 2006) found morphological
changes in the central nervous system following inhalation exposure to vanadium pentoxide.
Male CD-I mice (n=48) were exposed to vanadium pentoxide by whole-body inhalation for 1
hr/day, 2 days/week for up to 8 weeks (Avila-Costa et al., 2004, 2005). Particle size was not
reported in either study. The exposure concentration was reported as 0.02 M (Avila-Costa et al.,
2004, 2005, 2006) or "2.5 mg/m3 V" (Avila-Costa et al., 2005). The same group of investigators
(Gonzalez-Villalva et al., 2006; Mussali-Galante et al., 2005) using the same exposure protocol
reported that the 0.02 M solution generated an average chamber concentration of 2.5 mg/m3, as
vanadium metal (MW = 50.94), corresponding to 2.57 mg/m3 vanadium pentoxide (MW =
181.9). The number of immunoreactive-TH+ neurons in the substantia nigra region of the basal
ganglia in the mesencephalon (Avila-Costa et. al., 2004) and the morphology of the blood-brain
barrier (Avila-Costa et. al., 2005) were assessed at the end of each week for up to 8 weeks of
exposure. No clinical signs of toxicity were reported in either study. A duration-dependent
decrease in the number of immunoreactive-TH+ neurons was observed from week 3 (decrease of
approximately 30%; data presented graphically) through week 8 (decreased by approximately
63%; data presented graphically) of exposure (Avila-Costa et al., 2004). Morphological changes
to the blood-brain barrier (cilia loss, cell sloughing and ependymal cell layer detachment) were
66
DRAFT - DO NOT CITE OR QUOTE
-------
also observed starting at one week and increasing with duration of exposure (Avila-Costa et al.,
2005).
Using a similar protocol, Avila-Costa et al. (2006) assessed the effects of vanadium
pentoxide on memory and morphology of hippocampal neurons in male CD-I mice that were
exposed by whole-body inhalation for 1 hr/day, 2 days/week for up to 4 weeks. Groups of 6
exposed mice and 6 vehicle control mice (inhaling deionized water droplets) were evaluated
after 24 hrs and weekly for 4 weeks. No clinical signs or body weight changes were observed.
Spatial memory was tested using a modified Morris water maze task that was learned pre-
exposure. Performance on this test, as assessed by latency (swimming time) to locate a hidden
platform, was significantly impaired in the exposed mice at all time points in an increasing, time-
related manner. Pyramidal neurons from the hippocampus CA1 region were evaluated for
cytological and ultrastructural changes, because impairment in spatial memory is frequently seen
following damage to this region of the brain. The cytological analysis assessed numbers of
dendritic spines in the hippocampal cells; results showed a significant loss of dendritic spines in
the exposed mice at all time points; the loss increased with time, in a manner that correlated with
the memory impairments. The ultrastructural analysis showed a significantly increased
percentage of necrotic hippocampal cells at all time points, with a maximum of 33% after 4
weeks of exposure; other findings included hyperdense postsynaptic terminals and edema in
mitochrondria, dendrites, dendritic spines and presynaptic terminals. These three studies
establish a LOAEL for morphological changes to the central nervous system accompanied by
behavioral effects following acute and short-term intermittent exposure to vanadium pentoxide at
concentrations of 2.56 mg/m3 two times per week, for 1-hr duration/exposure.
Colin-Barenque et al. (2008) investigated whether the vanadium pentoxide-mediated
disruption of the blood-brain barrier was associated with the activation of matrix
metalloproteinases (MMPs), protein degrading enzymes that are involved in tissue remodeling.
Male CD-I mice (n= 20 per group) were exposed to 0.02 M (2.56 mg/m3) aerosolized vanadium
pentoxide in deionized water or deionized water alone for 1 hr two times a week for up to four
weeks. Five mice were sacrificed from each group, per time point (24 h, 1,2 and 4 weeks). The
presence of matrix metalloproteinase (MMP) was determined by gel zymography. In the
olfactory bulb, MMP-2 was not different between vanadium pentoxide treated mice and controls,
regardless of time point. MMP-9 was significantly elevated (300% and -420%) in vanadium
pentoxide-exposed mice compared to controls at 2 and 4 weeks, respectively. Both MMP-2 and
MMP-9 were detected in the prefrontal cortex; MMP-2 was not different between controls and
treated animals at any time point, but, MMP-9 was significantly elevated over control values at 1,
2 and 4 weeks of exposure (150%, -175%, and 250% of control values, respectively). In the
67
DRAFT - DO NOT CITE OR QUOTE
-------
hippocampus, MMP-9 from vanadium pentoxide-treated mice was significantly elevated over
controls after 1, 2 and 4 weeks of exposure (200%, 340%, and -370%) while MMP-2 in exposed
animals was significantly increased at 4 weeks (-150% over controls). In the striatum, MMP-9
from exposed mice was significantly elevated over that of controls after 4 weeks of exposure;
MMP-2 levels from exposed mice were significantly elevanted over control values after 2 (160%)
and 4 weeks (-250%) of exposure, but were not documented at earlier time points. These
findings suggest that vanadium-induced increases in MMP's in different parts of CNS occur in
association with dendritic spine loss as well as neuronal death, and could be related to blood-brain
barrier disruption.
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION FOR PULMONARY FIBROSIS AND CANCER
The preceding paragraphs have highlighted the main noncancer and cancer health effects that
result from exposure to vanadium pentoxide. Noncancer effects in the lung range from
histiocytic infiltration and alveolar inflammation, to hyperplasia of alveolar epithelium and
pulmonary fibrosis. These endpoints exist in a plausible biological response continuum—from
inflammation to reparative hyperplasia, to fibrosis. These effects also display a temporal and
dose-response continuum, ranging from inflammatory and hyperplastic responses that occur at
earlier time points (13 and 16 days) and at lower doses (2 mg/m3) to fibrosis that occurs at later
time points (3 months) and at higher doses (4 mg/m3). Inflammation and hyperplasia are
biologically relevant as precursor events to pulmonary fibrosis. Several investigators have
systematically investigated the molecular mechanisms underlying vanadium pentoxide-induced
pulmonary inflammation and fibrosis. These studies are summarized below.
4.5.1 Genotoxicity
The genotoxicity assays of vanadium pentoxide are summarized in Table 4-12.
4.5.1.1 Human Studies
Two studies investigated mutagenic activity in humans exposed to vanadium pentoxide
(Ehrlich et al 2008; Ivancsits et al 2002). The in vivo genotoxicity of vanadium pentoxide in
lymphocytes and whole blood leukocytes obtained from 49 male workers exposed to vanadium
pentoxide at a processing plant was compared to 12 non-exposed controls (Ivancsits et al., 2002).
The average exposure duration for workers was 12.4 years. Workers reported using protective
68
DRAFT - DO NOT CITE OR QUOTE
-------
masks at least occasionally. Measurements or estimates of worker exposure to vanadium
pentoxide were not reported, although exposure to vanadium was confirmed through
measurement of serum and urine vanadium. No significant differences between vanadium-
exposed and control workers were observed for DNA strand breaks (as assessed by alkaline
comet assay), 8-hydroxy-2'deoxyguanosine (8-OHdG),an oxidized DNA base common
indicative of oxidative stress or the frequency of sister chromatid exchange (SCE) in leukocytes.
When normal human leukocytes or human fibroblasts were cultured in vitro and exposed to
vanadate (25-500 ng/L), both whole blood cells and isolated non-proliferating lymphocytes
exhibited a significant increase in DNA migration in the alkaline comet assay compared to non-
exposed cells only at the highest doses tested (250-500 ng/L). Cultured human fibroblasts,
however, exhibited a dramatic dose-dependent increase in DNA migration after vanadate
treatment at lower concentrations as well (as low as 0.5 ng/L) and suggest that fibroblasts are
more sensitive to DNA damage in the presence of vanadate than are blood cells when exposed in
vitro (Ivancsitis et al. 2002).
Ehrlich et al (2008) investigated the impact of inhaled vanadium pentoxide on DNA
stability in vanadium production workers (n= 52) compared to non-exposed jail wardens (n=52)
during October 2004 - May 2005. All subjects studied were male, and were exposed for their
entire 8-hr shift while wearing protective masks. However, the duration of exposure and
concentration of the inhaled vanadium was not determined. The median concentration (25th -
75th percentile) of vanadium in plasma was 7-fold higher in exposed workers compared to the
unexposed reference group. Leukocytes were then assayed (Comet assay) for DNA damage, and
endogenous levels of oxidized purines and pyrimidines were determined. No differences in
DNA migration by exposure were noted in leukocytes under standard conditions, demonstrating
that exposure is not associated with increases in single- and double-strand breaks. However,
increases were observed in both oxidized purine (7% increase, p = 0.02) and pyrimidine (33%
increase, p = 0.002) formation in exposed individuals. Moreover, DNA damage induced by
bleomycin was 25% greater in leukocytes from the exposed workers (p < 0.0001) and DNA
repair after bleomycin administration was less evident (p < 0.0001). The extent of micronuclei
formation, necrosis and apoptosis was determined in isolated lymphocytes using the CBMN Cyt
assay. The number of micronuclei was 2.5 fold higher in 24 workers than 23 non-exposed
referents (p = 0.01). The frequency of nucleoplasmic bridges and nuclear buds (which indicate
evidence of misrepaired DNA breaks and gene amplification, respectively) were significantly
increased (7-fold and 3-fold) over controls. Numbers of necrotic and apoptotic cells were
increased by 55% and 50% respectively in exposed workers. Together, these results suggest that
occupational exposure to inhaled vanadium pentoxide may affect DNA stability by increasing
69
DRAFT - DO NOT CITE OR QUOTE
-------
levels of oxidized bases and affecting DNA repair. Age, body mass index and smoking habit
(cigarettes/day) were similar between the two groups. Folate levels also were similar but
vitamin B6 and Bi2 levels were lower among the unexposed indicating that the effects on DNA in
the exposed were not due to vitamin deficits.
Kim and colleagues (2004) assessed the cross-shift change in urine levels of 8-hydroxy-
2'-deoxyguanosine (8-OHdG), a marker of DNA repair of oxidative DNA damage, over a 5-day
period in 1999 among a group of 20 boilermakers involved in the overhaul of oil-fired boilers at
a power plant (74% of eligible). Median total PM2.5 8-hour TWA concentration, measured using
personal exposure monitoring, was 0.44 mg/m3 (Q25% - Qy5%: 0.29 - 0.76 mg/m3). Total
vanadium 8-hour TWA concentration, including vanadium oxides, was 1.23 ug/m3 (Q25% - Qy5%:
0.47 - 3.53 ug/m3). The workers were 18-59 years old (mean ± SD: 45.5 ± 12.0) and had been
employed as boilermakers for 0.04 to 40 years ((mean ± SD: 21.7 ± 12.9). The mean cross-shift
change in creatinine adjusted 8-OHdG levels in urine was 1.88 ug/g creatinine (SD 0.74). Pre-
shift levels, measured an average of two days away from work, were significantly different from
post-shift levels (p = 0.02). In linear mixed regression models, a 1 mg/m3 increase in total PM2.5
8-hour TWA concentration was associated with an increase in urinary 8-OHdG concentrations of
1.67 ug/g creatinine (95% CI: 0.21 - 3.14), adjusting for urinary cotinine, chronic bronchitis
status, and age. A 1 ug/m3 increase in PM2.5 vanadium concentration was associated with an
increase in urinary 8-OHdG concentrations of 0.23 ug/g creatinine (95% CI: 0.04 - 0.42) in a
model with the same covariates. PM2 5 manganese, nickel and lead concentrations also were
associated with 8-OHdG levels in urine when analyzed separately in similar models. The
concentrations of the metals were correlated (0.52 < r < 0.92) and so the association with
vanadium may not have been independent of the associations with the other metals. However,
the finding of oxidative DNA injury and repair among healthy boilermakers is consistent with
similar reports among vanadium pentoxide workers.
Another marker of oxidative DNA damage, 7-hydro-8-oxo-2'-deoxyguanosine (8-
oxodG), was assessed in relation to water soluble transition metal content in ambient PM2.5
among male and female nonsmoking students, 20 - 33 years of age, living in central Copenhagen
(Sorensen et al., 2005). Personal samples of PM2 5 were collected over two days twice during
one year, once during summer and once during autumn. Median (interquartile range)
concentrations of PM2.5 were 20.1 ug/m3 (13.1 - 27.7) in November and 12.6 ug/m3 (9.4 - 24.3)
in August. The median (interquartile range) concentration of vanadium in PM2.5 was 3.0 (0.3 -
4.7) in November and 3.2 (1.4 - 5.7) in August. Of 66 participating students, 32 provided
measurements for both seasons. Median (interquartile range) levels of 8-OxodG in lymphocytes
(per 105 dG) were 0.55 (0.34 - 0.78) and 0.58 (0.47 - 0.70) in November and August,
70
DRAFT - DO NOT CITE OR QUOTE
-------
respectively. Vanadium and chromium concentrations in aqueous suspensions of PM2.5 were
associated with the 8-oxodG concentration in lymphocytes in mixed regression models with
subject as a random factor and adjusting for PM2.5 mass and season. A 1 ug/L increase in either
vanadium or chromium was associated with a 1.9% (95% CI: 0.6 - 3.3) or 2.2% (95% CI: 0.8 -
3.5) increase in 8-oxodG concentrations in lymphocytes, respectively. Platinum, nickel, copper,
and iron were not associated with 8-oxodG concentration in lymphocytes and no metals were
associated with 8-oxodG concentration in urine. This study suggests that metal content of
ambient fine particulate matter increases oxidative DNA damage in lymphocytes and that
vanadium may be one of the responsible agents along with other metal constituents in particulate
air pollution at levels common in urban settings.
Three other studies demonstrated a genotoxic effect of vanadium pentoxide on primary
human lymphocytes in vitro (Rojas et al. 1996; Ramirez et al. 1997; Roldan and Altamirano
1990). These studies examined chromosomal aberrations and aneuploidy by fluorescence in situ
hybridization (FISH) and SCE assays, as well as DNA damage by the comet assay. Cells
exposed to vanadium pentoxide had significantly increased DNA migration indicative of DNA
damage at all doses tested in primary human leukocytes (p<0.05) and in the high doses in three
of four donor lymphocyte cell strains (p<0.05). Vanadium pentoxide (0 - O.luM) lead to an
increase in aneuploidy with some interindividual variation observed in four primary human
lymphocyte cell strains (Ramirez et al. 1997). An increase in aneuploidy was observed in one
primary human cell strain exposed to vanadium pentoxide (0 - 6ug/ml) but no chromosomal
aberrations were observed (Roldan and Altamirano 1990).
4.5.1.2 Laboratory in vivo and in vitro studies
Vanadium pentoxide produced gene mutations in two bacterial test systems (Bacillus
subtilus and Escherichia coll) (Kada et. al., 1980, Kanematsu et. al., 1980); although negative
results were reported by NTP (2002) in a reverse mutation assay in Salmonella typhimurium
(TA97,TA98,TA100, TA102,TA1535 with or without metabolic activation). Negative results
were also reported in a gene mutation assay in Chinese hamster V79 fibroblast cells (Zhong et
al., 1994). However, DNA damage and/or aneuploidy was observed in all in vitro studies
performed in primary human cells (Ivancsits et al. 2002; Kleinsasser et al. 2003; Ramirez et al.
1997; Rojas et al. 1996; Roldan and Altamirano 1990). Positive results were observed for DNA
strand breaks in cultured human lymphocytes (Rojas et. al, 1996) at high doses of vanadate
(Ivancsits et. al, 2002) and in cultured human fibroblasts at lower, more environmentally relevant
doses of vanadate (0.5 ug/L) (Ivancsits et. al., 2002). Positive results have been noted for
71
DRAFT - DO NOT CITE OR QUOTE
-------
aneuploidy (Ramirez et. al., 1997), and polyploidy (Roldan and Altamirano et. al., 1990).
Negative results were reported for chromosomal aberrations (Roldan and Altamirano et. al.,
1990). In Chinese hamster V79 lung fibroblast cells, positive results were observed for
micronuclei formation (Zhong et. al., 1994), altered mitosis (Zhong et. al., 1994), and cell
transformation in Syrian hamster embryo cells (Kerckaert et al., 1996) at concentrations that
were not cytotoxic. Negative results were reported for sister chromatid exchanges and gene
mutations in vanadium pentoxide-treated Chinese hamster V79 firbroblast cells (Zhong et. al.,
1994).
One study evaluated the genotoxicity of vanadium pentoxide in primary human cell
cultures. Kleinsasser et. al. (2003) took mucosal biopsies from inferior nasal turbinates and blood
samples from seventeen healthy volunteers. Isolated lymphocytes and mucosal cells were
cultured and exposed to 0, 0.06, 0.12, 0.24, and 0.47 mM vanadium pentoxide in vitro for Ihr.
Mucosal cells and lymphocytes were assessed for DNA migration by the Comet assay. Extent of
migration was measured qualitatively (image analysis) and quantitatively ("Olive Tail Moment"
method). DNA migration was not significantly different in exposed human nasal mucosal cells
compared to controls. However, DNA migration appeared to increase dose-dependently in
exposed human lymphocytes compared to controls (p = 0.001). Cytotoxicity was limited in both
cell types at all doses as measured by trypan blue exclusion assay. These results suggest that
human lymphocytes, but not nasal mucosal cells demonstrate genotoxic damage (single strand
breaks and/or alkali-labile damage) in response to vanadium pentoxide.
Experimental data in animals provide evidence of some types of genotoxicity following
in vivo exposure to vanadium pentoxide. Vanadium pentoxide administered for 3 months by
inhalation to male and female mice (1, 2, 4, 8 or 16 mg/m3) did not increase the frequency of
micronucleated normochromatic erythrocytes in peripheral blood (NTP, 2002). Additional
details of exposure are provided in the NTP (2002) study summary (see Section 4.2.1.2).
Genotoxicity was evaluated in male CD-I mice following single intraperitoneal injections of
5.75, 11.5 or 23 mg/kg vanadium pentoxide (Altamirano-Lozano et al., 1993, 1996). Exposure
caused no treatment-related effects on mitotic index, average generational time or sister
chromatid exchanges in bone marrow cells (Altamirano-Lozano et al., 1993), although all doses
induced DNA damage in testicular germ cells (Altamirano-Lozano et al., 1996).
Altamirano-Lozano et al. (1999) assessed DNA damage in male CD-I mice 24 hrs following
single intraperitoneal injections of 0, 23.0, 11.5 or 5.75 mg/kg vanadium pentoxide
(corresponding approximately to the LD50, l/i LD50 and l/4 LD50, respectively). Comet test
results show the number of cells with DNA damage (primarily single strand breaks and alkali
72
DRAFT - DO NOT CITE OR QUOTE
-------
labile damage) was increased in liver, kidney, lung, spleen and heart, although increases did not
exhibit dose-dependence. No evidence of DNA damage was observed in bone marrow.
In summary, the evidence for mutagenicity in humans is limited. There are few studies
examining genotoxicity in humans in vivo, with equivocal results. Ivancsits et. al. (2002)
reported no differences in DNA strand breaks, oxidative damage, or sister chromatid exchange
frequency in leukocytes between control and vanadium pentoxide-exposed workers. Ehrlich et.
al., (2008) noted changes in DNA stability and DNA repair in leukocytes of occupationally-
exposed workers as compared to controls. Studies have demonstrated a genotoxic effect of
vanadium pentoxide on human cells in vitro. Ivancsits et al. (2002) demonstrated significant
increases in DNA damage as measured by the Comet assay in both leukocytes and fibroblasts but
with different dose sensitivity, while Kleinsasser et. al (2003) noted DNA migration differences
occurred dose-dependently in peripheral blood lymphocytes but not in nasal mucosa. Earlier
studies in human lymphocyte cultures also demonstrated increased aneuploidy (Ramirez et al.
1997; Rojas et al. 1996) and DNA damage (Roldan and Altamirano 1990) following exposure to
vanadium pentoxide. Thus, vanadium pentoxide-induced mutagenicity may occur at doses
higher than those measured in these occupational exposures, may be tissue-specific and may be
associated with oxidative stress rather than direct DNA damage.
In vitro tests in bacterial and yeast systems provide mixed evidence of vanadium
pentoxide-induced mutagenicity. In general, classic gene mutation assays were negative, as
were tests that assessed sister chromatid exchange and other chromosomal aberrations. DNA
strand breaks (Rojas et al., 1996; Ivancsits et. al., 2002) and micronuclei formation (Zhong et.
al., 1994) were indicated in some studies in cultured cells but were dependent on cell type.
Fibroblasts appear to be more sensitive to vanadium exposure in vitro than are blood cells.
Similarly, experimental data from animal studies is equivocal. NTP (2002) reported that the
frequency of micronucleated normochromatic erythrocytes in peripheral blood was not increased
in exposed compared to control mice. However, a number of studies by Altamirano-Lozano et
al. (1993, 1996, 1999) have noted DNA damage in specific target tissues in vanadium pentoxide-
treated mice. It should be noted that Altamirano-Lozano et al (1993) consistently used
intraperitoneal injection as the route of exposure for these studies.
73
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-12. Genotoxicity Data Following Exposure to Vanadium Pentoxide
Test System/Species
Results
Exposure
Dose
Effects
Endpoint
Reference
In vivo
Human
49 exposed male
workers at vanadium
pentoxide processing
plant; 12 non-
exposure controls
52 exposed workers;
52 non-exposed
workers (jail
wardens)
—
-
-
+
avg 12. 4y
inhalation
not reported -
exposure
confirmed
through blood
and urine
measurements
of vanadium
not reported -
8hr shift,
protective
masks;
exposure
confirmed
through blood
levels of
vanadium
Genotoxicity measured in isolated lymphocytes and
whole blood leukocytes. Study also examined in
vitro exposure (below).
Genotoxicity was measured in isolated leukocytes
by Comet assay, with no increases in single- and
double-DNA strand breaks observed. However,
increases were observed in oxidized purines and
pyrimidine formation in exposed workers. CBMN
Cyt assay demonstrated increased MN induction,
nucleoplasmic bridges and nuclear bud formation.
Necrosis and apoptosis levels were also increased in
exposed individuals.
Comet Assay
DNA damage
(SOHdG)
Sister
chromatid
exchange
Comet Assay
MN
induction,
oxidative
nucleotides
Ivancsits et
al. 2002
Ehrlich et
al. 2008
Laboratory Animals
Male CD-I mice (n
= 4)
Male CD-I mice (n
= 2)
+
+
intraperitoneal
injection,
sacrificed 24h
post injection
intraperitoneal
injection,
sacrificed 24h
post injection
0,5.75, 11.5,
23 (ig/g bw
0,5.75, 11.5,
23 (ig/g bw
DNA damage was observed in all tissues examined
except for bone marrow. This included liver,
kidney, lung, spleen, and heart.
As part of a larger study on reprotoxicity, DNA
damage in sperm cells was analyzed. Significant
increases were observed in a dose-dependent
manner (p <0.05).
Comet assay
Comet assay
Altamirano-
Lozano et
al. 1999
Altamirano-
Lozano et
al. 1996
74
DRAFT - DO NOT CITE OR QUOTE
-------
Male CD-I mice (n
= 4)
B6C3F1 mice (M, F)
B6C3F1 mice (M, F)
+
+
intraperitoneal
injection,
sacrificed 24h
post injection
inhalation,
2yr
inhalation,
2yr
0,5.75, 11.5,
23 (ig/g bw
0, 1,2, or 4
mg/m3
0, 1,2, or 4
mg/m3
Analysis of sister chromatid exchange demonstrated
no effect of vanadium pentoxide exposure in this
study.
DNA was isolated from lung tumros and normal
tissue from exposed animals. Of the 20 tumors
analyzed, 13 had either K-ras mutations or LOH at
chromosome 6 or both.
Analysis of frequency of micronuclei in peripheral
blood normochromatic erythrocytes demonstrated
on effect of vanadium pentoxide exposure in this
study.
cytogenetic
assay
cytogenetic
assay
Micronuclei
assay
Altamirano-
Lozano et
al. 1993
Devereux et
al. 2002
NTP 2002
In vitro
primary human
lymphocytes
primary human
whole blood
leukocytes
cultured human
fibroblasts
primary human
lymphocytes
primary human nasal
mucosal cells
primary human
lymphocytes
+
+
+
+
+
25 - 500
Mg/L
0 - 47 mM
0-0.1 nM
V2O5 exposure led to significant increase in DNA
migration as measured by Comet assay at the
highest doses tested (250 - 500ug/L) for whole
blood lymphocytes and leukocytes, and at all doses
tested in cultured fibroblasts (p values not given).
Exposure to vanadium pentoxide led to a dose-
dependent increase in DNA migration in
lymphocytes but not in mucosal cells.
Vanadium pentoxide lead to an increase in
aneuploidy with some interindividual variation
observed between the four primary cell strains.
Disruption of spindle formation may be due to
interaction with microtubules.
Comet Assay
Comet assay
FISH
Ivancsits et
al. 2002
Kleinsasser
et al. 2003
Ramirez et
al. 1997
75
DRAFT - DO NOT CITE OR QUOTE
-------
primary human
lymphocytes (n=4)
primary human
lymphocytes (n = 1)
Syrian Hamster
Embryo cells
Chinese hamster
V79 cells
+
+
+
+
-
"
24h
72h
0, 24h, 7d
24h
0.3, 30, 3000
HM
0, 2, 4, 6
fig/ml
0 to 0.875
Hg/ml
0, 1, 3, 6, 9,
12 (ig/ml
Cells exposed to vanadium pentoxide had
significantly increased DNA migration at all doses
tested in the leukocytes (p < 0.05) and in the high
doses in lymphocytes for three of the four donors (p
< 0.05). DNA repair occurred generally within in
45 min post-exposure.
Vanadium pentoxide exposed cells had an increase
in aneuploidy and a decrease in mitotic index, with
no changes in SCE or chromosomal aberrations.
Vanadium pentoxide exposed cells were positive at
7d exposure but not at 24h, similar to other tumor-
promotion chemicals studied by this group.
Vanadium pentoxide exposure led to increased MN
induction (p < 0.005), apparently due to damage to
the spindle apparatus but no significant increases in
mutations or SCE.
Comet Assay
SCE assay
aneuploidy
SHE
transformation
assay
MN induction
SCE assay
HGPRT
mutation
Rojas et al.
1996
Roldan and
Altamirano
1990
Kerckaert
etal. 1996
Zhong et al.
1994
Bacterial systems
Bacillus subtilis
positive
with and
without
activation
Study details not available.
recombination
repair
Kada et al.
1980
76
DRAFT - DO NOT CITE OR QUOTE
-------
Escherichia coli
Salmonella
typhimurium
Salmonella
typhimurium
(TA97,TA98,TA100
, TA102,TA1535
with or without
metabolic activation)
positive
without
activation
(not tested
with
activation)
All strains
negative
with and
without
activation
(both
hamster
and rat S9
fractions)
48 h
0-333
^g/plate
Study details not available.
No increase in revertant colonies was observed
following exposure.
gene mutation
gene mutation
Kanematsu
etal. 1980
NTP 2002
77
DRAFT - DO NOT CITE OR QUOTE
-------
4.5.2 Mechanisms of Inflammation and Fibrosis
Increases in markers of pulmonary inflammation have been observed in the BAL fluid of
rats and susceptible mice following intratracheal instillation exposure to vanadium pentoxide
(Pierce et al. 1996; Bonner et al. 2002). These include macrophage inflammatory protein-2
(MIP-2), keratinocyte-derived chemokine (KC), interleukin-6 (IL-6), and IL-8. Further, the
increased expression of prostaglandin-generating enzymes cyclooxygenases (COX) and
prostaglandin E synthases have been associated with exposure to vanadium pentoxide (Bonner et
al 2002; Pierce et al 1996), further suggesting increased inflammation.
Pierce et. al. (1996) investigated the ability of several vanadium compounds to increase
mRNA levels of cytokines in bronchoalveolar lavage (BAL) fluid. Female CD rats received 42
or 420 |j,g of vanadium pentoxide or phosphate-buffered saline (PBS) by intratracheal
instillation. BAL fluid was recovered one hr to ten days after exposure. Significant neutrophil
influx was observed after 24 hrs exposure to vanadium pentoxide and peaked at 48-hr post-
exposure with 20% neutrophils. MIP-2 mRNA expression levels were significantly elevated in
vanadium pentoxide-treated rats compared to controls at early time points (1 hr to 48 hrs)
suggesting pulmonary inflammation.
Pro-inflammatory prostaglandins such as PGE2 produced by the enzymes COX-1 and
COX-2, and PGE synthase mediate tissue homeostasis and/or known to be associated with
various inflammatory diseases. Bonner et. al. (2002) assessed the role of COX-1 and COX-2
enzymes in vanadium pentoxide-induced pulmonary inflammatory and fibrotic responses using
6-8 month-old male and female mice (of a hybrid C57BL/6J and 129/Ola genotype) that were
deficient in either COX-1 (COX1"7") or COX-2 (COX2"A) enzyme. These COX-deficient mice
and genotype-matched controls (wildtype) were instilled with 50 |jL of saline (n = 3 to 4) or
Img/kg vanadium pentoxide in saline (n = 5 to 6). Lungs were lavaged at 1, 3, 6, or 15 days
post-instillation and BAL fluid was collected and analyzed for tumor necrosis factor-alpha
(TNF-a) and prostanoids (e.g., PGE2) by ELISA. Lungs were removed and preserved for
histopathology, hydroxyproline assay, or COX immunoblotting. Histopathology showed marked
inflammation and increased injury in COX2"" mice compared to wild-type and COX1"7" mice 3
days following vanadium pentoxide instillation. Hydroxyproline content was not different in
wildtype or COX1"" mice in response to vanadium pentoxide compared to saline-instilled
controls. Hydroxyproline content in vanadium pentoxide-exposed COX2"" mice was increased
twofold compared to saline-instilled COX2"7" mice suggesting enhancement of lung fibrosis
following vanadium pentoxide exposure. PGE2 levels increased from -500 pg/mL in saline-
instilled wildtype mice to -1000 pg/mL in vanadium pentoxide-treated wildtype mice at 24 hrs
78
DRAFT - DO NOT CITE OR QUOTE
-------
but not at other time points. The PGE2 level in saline-treated COX1"" mice was -10 pg/mL and
about -225 pg/mL after 24 hrs of vanadium pentoxide. PGE2 levels were -200 pg/mL at 24 hrs
in saline-treated COX2"" mice and did not differ significantly from vanadium pentoxide-treated
COX2"" mice regardless of time point. This study suggests that vanadium pentoxide-induced
inflammation may also be at least partially mediated by prostaglandins such as PGE2 generated
by COX-2.
Myofibroblasts, the principal proliferating cells that produce collagen, are involved in
fibrogenic response of lungs following exposure to pulmonary irritants. Bonner et. al. (2000)
observed that proliferating myofibroblasts were the principle cell type that contributed to the
observed fibrosis. Male Sprague-Dawley rats weighing -200 g received intratreacheal
instillation of sterile saline or 1 mg/kg vanadium pentoxide. Rats were additionally injected with
BrdU (50 mg/kg, i.p.) one hr prior to sacrifice. Sacrifice occurred at 3, 6, and 15 days after
vanadium pentoxide instillation. Excised lung tissue was assessed morphometrically and by
immunohistochemistry for vimentin and desmin, two biomarkers for myofibroblasts and smooth
muscle cells, respectively. Trichrome staining was used to assess collagen levels, an indicator of
the extent of fibrosis. Vanadium pentoxide exposure induced thickening of the desmin-positive
bronchiolar smooth muscle cell layer by day 6 post exposure and were identified as
myofibroblasts. A 2.3-fold increase in airway smooth muscle cell nuclear profile, suggesting
increase in smooth muscle cell proliferation was due to hyperplasia. Serial sections of the
peribronchiolar region stained positive for vimentin and desmin, and were mainly
myofibroblasts. The thickness of the subepithelial trichrome-positive layer was 3.1-3.9-fold
higher at day 15 in vanadium pentoxide vs. control samples. The peak appearance of
peribronchiloar myofibroblasts occurred at day 6 and declined by day 15. A thickened collagen
ring was apparent by day 15 in vanadium pentoxide-exposed samples compared to controls.
Rice et. al. (1999) using both in vitro and in vivo models showed that myofibroblasts do
proliferate in response to vanadium pentoxide, and are dependent on platelet-derived growth
factor (PDGF) and epidermal growth factor (EGF). Rat lung myofibroblasts were isolated from
exposed male Sprague-Dawley rats, as stated in Bonner (1998) above, and were grown to
confluency. Cultures were incubated for 24 hrs with increasing concentrations of one of two
inhibitors of the PDGF-R (AG1296) and EGF-R (AG1478), respectively, at a concentration of
100 |j,mol/L. Autophosphorlyation of PDGF-R and EGF-R in vitro was specifically blocked by
AG1296 and AG1478, respectively. Tritiated [3H] thymidine uptake, a measure of mitogenesis,
was blocked by selective inhibition of PDGF- and EGF- receptors. An in vivo study was carried
out at the same time. Male Sprague-Dawley rats were treated with AG1296 or AG1478 (50
mg/kg) by intraperitoneal injection 1 hr prior to intratracheal instillation of vanadium pentoxide
79
DRAFT - DO NOT CITE OR QUOTE
-------
(1 mg/kg) and again two days after vanadium pentoxide was administered. Rats were sacrificed
3, 6, and 15 days after instillation and lungs were preserved for bromodeoxyuridine (BrdU)
immunohistochemistry and hydroxyproline assays to measure DNA replication and cell division,
respectively. Quantitation of BrdU-labeled cells in the nuclei of rat lung tissue was significantly
reduced in vanadium pentoxide-treated animals that had received injections of AG1296 and
AG1478, compared to vanadium pentoxide treated animals that were injected with vehicle alone.
Vanadium pentoxide treatment induced a five-fold increase in lung hydroxyproline content, an
indicator of lung collagen and potentially fibrosis, 15 days after instillation. Prior and post-
treatment with AG1296 reduced hydroxyproline content in vanadium pentoxide-treated animals
to quantities similar to saline-instilled animals. Pre- and post-treatments with AG1478 reduced
hydroxyproline content by -50%, but were still significantly higher than in saline-instilled
controls.
Two recent studies examined the mechanism of inflammation in mice (Rondini et al.
2010; Turpin et al. 2010). Rondini et al (2010) examined the effect of exposure to vanadium
pentoxide in three mouse strains of varying susceptibility to lung cancer (A/J, BALB/C and
C57BL/6J) in an initiation/promotion model (full study description in Section 4.2.2.2).
Significantly higher transcriptional activity was observed forNFkB (A/J mice; peaked at 1 day
post-treatment) and AP-1 (A/J and B6 mice; peaked at 6 hrs post-treatment) compared to PBS-
controls, the activities in the order of A/J mice > B6 mice; in vanadium pentoxide-treated mice.
Overall, the differential inflammatory responses observed in the three strains of mice appear to
positively correlate with increased levels of chemokines, such as keratinocyte-derived
chemokine (KC) and monocyte chemotactic protein-1 (MCP-1), and increased binding of
transcriptional factors NFicB and AP-1 (c-Fos), and sustained activation of MAP kinases
(MAPKs) and extracellular signal-regulated kinases 1 and 2 (ERK 1/2) suggesting inflammation
as a major response in mice. Turpin et al. (2010) examined pulmonary inflammation and fibrosis
following intranasal aspiration exposure to vanadium pentoxide with and without respiratory
syncytial virus (RSV) exposure (full study description in Section 4.4.1.2). In this study,
vanadium pentoxide exposure also caused a significant increase in cell proliferation in the
airways and lung parenchyma, lung mRNAs for TGF-|3-1, CTGF, PDGF-C, CollA2, and
mRNAs for IFN-a and -|3 and IFN-inducible chemokines CXCL9 and CXCL10 compared to
controls. Pre- or post-treatment with RSV caused a significant reduction in the all mRNAs.
Together, results from this study showed that vanadium pentoxide induces inflammatory and
fibrogenic response in mouse lung and these effects were suppressed by RSV infection.
To elucidate the potential cell signaling cascades associated with these endpoints, Antao-
Menezes et. al. (2008) investigated the role of the signaling molecule, signal transducer and
80
DRAFT - DO NOT CITE OR QUOTE
-------
activator of transcription (STAT)-l in vanadium pentoxide-induced pulmonary fibrosis. Their
work identified another inflammatory molecule, interferon-beta (IFN-|3), as a mediator of
vanadium pentoxide-induced STAT-1 activation in normal human lung fibroblasts. Briefly,
confluent, quiescent cultures of human lung fibroblasts were either placed in serum-free defined
medium (SFDM) or SFDM supplemented with 10 |j,g/cm2 vanadium pentoxide. Neutralizing
anti-IFN-a and anti-IFN-p antibodies were used to quench activity of IFN-a and IFN-|3 and the
ratio of phosphor-STAT-1 to STAT-1 was then measured by quantitative RT-PCR or Western
blot. Vanadium pentoxide-induced STAT-1 activation could be inhibited by a broad spectrum
NADPH inhibitor at 24 hr. Vanadium pentoxide induced significant IFN-|3 expression after 18
and 24 hrs. This effect was nearly completely abolished by addition of the NAPDH inhibitor.
Activation of STAT-1 (as measured by a ratio of phosphor-STATI to total STAT-1) was
significantly decreased by addition of neutralizing IFN-|3 antibodies and also by addition of a
Janus Associated Kinase (JAK) inhibitor. In summary, STAT-1 was activated in response to
vanadium pentoxide exposure, and was linked to IFN-|3 as its primary mediator. This study
identifies a putative signaling pathway leading to expression of genes that control proliferation
of myofibroblasts.
Ingram et. al. (2007) performed gene array analysis to determine a list of candidate genes
altered by exposure to vanadium pentoxide. Normal human lung fibroblasts were exposed to 10
Hg/cm2 vanadium pentoxide or saline in vitro. RNA from cells was harvested 1, 4 8, 12, and 24
hrs post-treatment. Labeled cRNA hybridized to the Affymetrix Human Genome Array U133A
2.0 gene chip was used to assess gene expression at various time points up to 24 hrs of exposure.
About 300 genes were found to be upregulated in response to vanadium pentoxide including
inflammatory and immunomodulatory genes. Over 1,000 genes were downregulated in response
to vanadium pentoxide including genes from the ubiquitin cycle and cell cycle genes. A dozen
genes were confirmed by RT-PCR and included growth factors (heparin-binding EGF-like
growth factor [HB-EGF], vascular endothelial cell growth factor [VEGF], and connective tissue
growth factor [CTGF]), chemokines (IL-8, CXCL9, CXCL10), oxidative response genes
(superoxide dismutase [SODJ2, pipecolic acid oxidase [PIPOX], oxidative stress response
[OXRJ1) and DNA-binding proteins (growth arrest specific [GAS]1, STAT1). The gene array
analysis thus confirms that a number of mitogens, growth factors, chemokines, cytokines,
oxidative response genes, and DNA-binding proteins are all critical to the formation of
fibroproliferative lesions in response to vanadium pentoxide exposure in vitro (Figure 4-1).
81
DRAFT - DO NOT CITE OR QUOTE
-------
V205 Exposure
*
Cell
Responses
Pathologic
Consequence
Figure 4-1: Genomics of ViOs-Induced Bronchitis (reprinted with permission from Ingram
et al., (2007) Respir. Res. Apr 25;8(1):34)
Together, these results indicate that proliferating myofibroblasts are the primary cell type
associated with vanadium pentoxide-induced pulmonary fibrosis, and that cellular proliferation
depends on activated mitogens such as PDGF and EOF, in addition to HB-EGF. The STAT-1
and MAPKinase pathways may play key roles in this process. Moreover, fibroproliferative
lesions contain collagen.
4.5.3 Mechanisms of Hyperplasia and Carcinogenicity
Molecular events underlying the mechanism of reparative hyperplasia and carcinogenesis
have been documented. Bonner et. al. (1998) exposed male Sprague-Dawley rats to vanadium
pentoxide or sterile saline by intratracheal instillation at 1 mg/kg. Animals were sacrificed at 3,
6 and 15 days after instillation. Lungs were preserved and analyzed for PDGFR-a by
immunohistochemical analysis and morphometry for fibroproliferative lesions. Smooth muscle
thickening was observed beneath ciliated epithelial cells, as indicated by increased desmin
localization, on day 6 in exposed samples. Trichrome staining revealed increased collagen
deposition around bronchioles in vanadium-pentoxide exposed samples. The thickness of the
subepithelial layer increased by 3.1 to 3.9 fold at day 15 after instillation of vanadium pentoxide,
determined by morphometric techniques.
Platelet-derived growth factor (PDGF) is a mitogen and chemoattractant for fibroblasts.
82
DRAFT - DO NOT CITE OR QUOTE
-------
Male Sprague-Dawley rats were intratracheally instilled with sterile saline or 2 mg/kg of
vanadium pentoxide (Bonner et al 1998). Tissues were harvested at 24, 48, and 72 hr post
exposure as well as at 6 and 15 days. Pulmonary myofibroblasts and alveolar macrophages were
isolated. Isolated total lung RNA was assessed for quantitation of PDGF by Northern blot
analysis. PDGF-receptor alpha mRNA and protein expression were significantly elevated in
vanadium pentoxide exposed animals compared to controls at 24 and 48 hrs. PDGF-receptor
beta was not significantly elevated over controls at any time points. Confluent cultures of lung
myofibroblasts were stimulated with vanadium pentoxide. Similarly, cultures of lung
macrophages were stimulated with vanadium pentoxide. Levels of PDGF in myofibroblasts
were not affected by direct stimulation by vanadium pentoxide, but were affected by a factor
released by vanadium pentoxide stimulated macrophages, and were associated with interleukin
(TL)-IB, an inflammatory cytokine. Together, these results suggest that hyperplasia occurs
under the control of inflammatory mediatiors such as IL-1B that can recruit mitogens such as
PDGF that can stimulate growth of myofibroblasts, the main cell type involved in the
development of fibrotic lesions.
Zhang et al (200 la) investigated the ability of vanadium pentoxide to induce heparin-
binding epidermal growth factor-like growth factor (HB-EGF) (another mitogen) in vitro, using
normal human bronchial epithelial cells (NFffiECs). Mature cultures of NFffiECs were
incubated with vanadium pentoxide at 0, 1, 10 and 50 |j,g/cm3 for 3 hrs. Total RNA was then
isolated and RT-PCR was used to quantitate HB-EGF. HB-EGF mRNA was significantly and
dose-dependently increased in response to vanadium pentoxide compared to controls. In a
second time course study, (using only the high dose 50 |j,g/cm3 of vanadium pentoxide) the peak
of HB-EGF induction occurred after 3hrs of exposure and persisted until 8 hrs of exposure.
In a follow-up study, Ingram et al. (2003) similarly showed a peak induction of HB-EGF
in quiescent cultured human lung fibroblasts exposed to 10 |j,g/cm2 vanadium pentoxide at 3 hrs.
Quiescent cultured human lung fibroblasts were exposed for 3 hrs to 0, 10, 30 or 100 |j,g/cm2
vanadium pentoxide. HB-EGF RNA was isolated and detected by Northern blot. HB-EGF was
significantly elevated in a dose-dependent manner in vanadium pentoxide-exposed fibroblasts
compared to controls. Stimulating human lung fibroblasts with H2O2 (10 (jM) similarly induced
HB-EGF, with a peak mRNA expression at Ihr post exposure. HB-EGF protein expression
peaked at 6 hr post—exposure, as measured by Western blot. Quiescent human lung fibroblasts
stimulated with 10 |j,g/cm2 vanadium pentoxide were found to initially quench spontaneous H2O2
production (at early time points) and then boost H2O2 production at a peak of 12 hrs post
exposure. To assess the role of extracellular signal-regulated protein kinase (ERK), and the p38
subunit of mitogen-activated protein (MAP) kinase in vanadium pentoxide induced HB-EGF
83
DRAFT - DO NOT CITE OR QUOTE
-------
production, Ingram et al (2003) exposed quiescent confluent human lung fibroblast cells in
culture to 500 |jM H2O2 or 10 |j,g/cm2 vanadium pentoxide for 15 min, 30 min, 1 hr, 3 hr, 6 hr or
24 hr. Cell lysates were collected and Western blot was performed for the phosphorylated form
of ERK or total ERK protein or for the phosphorylated p38 subunit of MAPkinase and total p38.
Peak phosphorylated ERK occurred at 30 min post exposure to H2O2. Vanadium-induced
increases to p-ERK were biphasic, with one peak at 30 min post-exposure and the next at 24 hr
post-exposure. Phosphorylated p38 was similarly maximally elevated at 30 min post-treatment
with H2O2 and 24 hrs post treatment with vanadium pentoxide. Thus, this study suggests that
HB-EGF expression occurs as a result of activation of the MAPKinase and ERK pathways.
NTP (2002) and Ress et al. (2003) concluded that exposure to vanadium pentoxide
caused alveolar and bronchiolar adenomas and carcinomas in male and female mice and there is
some evidence of carcinogenicity in male rats, based on observations of alveolar and bronchiolar
neoplasms in groups exposed to vanadium pentoxide that exceeded historical controls. The body
of evidence that has investigated the mode of action (MOA) underlying cancer effects due to
exposure of vanadium pentoxide is not as well characterized as mechanisms underlying non-
cancer fibrotic effects. However, loss of heterozygosity (LOH) and DNA damage has been
documented.
Using mouse lung tumor tissues from the NTP (2002) chronic inhalation study from mice
exposed to 0, 1, 2 or 4 mg/m3 vanadium pentoxide, Zhang et. al. (2001b), observed LOH on
chromosome 6 (in the region of the K-ras gene) in 17 of 19 vanadium pentoxide-induced mouse
tumor samples. Moreover, 29 of 40 (73%) vanadium pentoxide-induced murine
adenocarcinomas from the 2002 NTP study had mutations in Kras2. The Kras2 mutations
typically were the result of either a GA-^AT transition or a GA-^TA transversion in the second
base of codon 12 (Zhang et. al., 2001b). To determine the effect of the K-ras mutations and
LOH on activated MAP kinase, Devereux et al (2002), used tissues from the NTP (2002) 2-year
carcinogenicity study in female and male B6C3F1 mice to isolate protein from 17 vanadium
pentoxide-induced alveolar and bronchiolar carcinomas, one spontaneous carcinoma, and two
normal (untreated) lung tissue samples. Levels of total MAP kinase and activated MAP kinase
were assessed by probing isolated lung protein for total and phosphorylated MAP kinase with
anti-phospho-MAP kinase antibody in all samples. Only qualitative analysis was reported.
Total MAP kinase was not different between normal tissue and lung tumor tissue. Activated
MAP kinsase was elevated in five of six tumors that had both LOH and K-ras mutations, and
was barely detectable in all seven tumors examined where no K-ras mutations were detected.
Four of five tumors with K-ras mutations but were not positive for LOH had elevated
phosphorylated MAP kinase levels. However, these results should be interpreted cautiously as
84
DRAFT - DO NOT CITE OR QUOTE
-------
LOH was difficult to detect due to interference from infiltrating lymphocytes. In summary,
mouse tumor tissue excised from mice used in the NTP 2002 study showed LOH in the region of
the k-ras oncogene location in 17 or 19 samples tested. However, the signaling events, and
specifically the role of the MAPkinase pathway, associated with this oncogenic mutation remain
unclear.
Pierce et al. (1996) identified proinflammatory cytokines associated with vanadium
pentoxide exposure. Pierce et al. (1996) reported increased mRNA expression of macrophage
inflammatory protein (MIP-2) and keratinocyte-derived cytokine (KC) in bronchiolar lavage
(BAL) fluid in vanadium pentoxide-treated female CD rats compared to controls at early time
points (1 hrto48hrs).
In summary, hyperplastic responses following exposure to vanadium pentoxide are
associated with increased expression of various mitogens such as PDGF and HB-EGF.
Moreover, PDGF activation may be dependent on inflammatory mediators such as IL-1B. HB-
EGF expression is dependent on activation of the MAPKinase and ERK signaling pathways.
No information is available concerning the rat neoplasm data. Indeed, the marginal increase in
lung neoplasms observed in female rats was not statistically significant. It is not known whether
lung neoplasms from male rats exhibit elevated activated MAP kinase or whether the rat tumors
have increased K-ras mutation and LOH on chromosome 6.
85
DRAFT - DO NOT CITE OR QUOTE
-------
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS
4.6.1. Oral
Limited studies have been published examining the effects following oral exposure to
vanadium pentoxide. Table 4-13 presents a summary of the noncancer results for the subchronic
and chronic oral studies of vanadium pentoxide toxicity in experimental animals.
4.6.1.1 Acute
The only acute studies available report oral LD50 values in rats that range from -10-137
mg/kg body weight and 64 mg/kg-body weight in mice depending on the source (Yao et. al.,
1986b and CICAD 29, 2001). Clinical signs of toxicity included lethargy, excessive tearing
(lacrimation), and diarrhea. Histopathological analysis revealed liver necrosis and swelling of
renal tubules. No acute dose-response studies are available for any animal species.
Table 4-13: Summary of Noncancer Results of Repeat-Dose Studies for Oral Exposure of
Experimental Animals to Vanadium Pentoxide
Species
(sex)
Wistar
Rat
(male)
Avg Daily
Dose
(mg/kg-day)
0, 10.5,
16.4, 69.6,
141.0
mg/kg-day
(correspond
stoO, 74. 5a,
116.1b, 500
and 1000
ppm)
0, 10.5,
16.4, 69.6,
141.0
mg/kg-day
(correspond
stoO, 74. 5a,
116.1b, 500
and 1000
ppm)
Exposure
Duration
and
Route
103 days
in food
103 days
in food
Response
at LOAEL
Decreased
erythrocyt
e count
Decreased
relative
liver
weight
NOAEL
(mg/kg-
day)
10.5
mg/kg-
day
—
LOAEL
(mg/kg/day)
16.4 mg/kg-
day
69.6 mg/kg-
day
Comments
Dietary
exposure
was
increased
at day 35
of study
(from 25 to
100 and
from 50 to
150 ppm)
(See notes
below)
Reference
Mountain
et al.,
1953
86
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-13: Summary of Noncancer Results of Repeat-Dose Studies for Oral Exposure of
Experimental Animals to Vanadium Pentoxide
Rat
(male)
0, 10.5,
16.4, 69.6,
141.0
mg/kg-day
(correspond
stoO, 74. 5a,
116.1b, 500
and 1000
ppm)
1.41 and
14.1 mg/kg-
day
(correspond
sto 17.9 and
179 ppm)
103 days
in food
2.5 yrs in
food
Decreased
Hair
Cystine
Decreased
Hair
Cystine
10.5
mg/kg-
day
1.41
mg/kg-
day
16.4 mg/kg-
day
14.1 mg/kg-
day
Study
published
in Patty's
Industrial
Hygiene
and
Toxicology
3rd ed.
1981
Original
data not
available.
Stokinger
et. al.,
1953
a Represents an average dose based on 25 ppm for 35 days, and 100 ppm for 68 days
b Represents an average dose based on 50 ppm for 35 days and 150 ppm for 68 days.
87
DRAFT - DO NOT CITE OR QUOTE
-------
4.6.1.2 Subchronic
Data on toxicity following subchronic oral exposure to vanadium pentoxide are
extremely limited. No comprehensive studies outlining toxic effects in response to vanadium
pentoxide in humans have been reported. The primary noncancer health effects of subchronic
oral exposure in animals include changes in relative liver weight, decreased erythrocyte count
and hemoglobin counts, and decreased hair cystine content (Mountain et. al., 1953) in rats.
Erythrocyte count and hemoglobin decreases are correlated and cannot be considered as separate
effects. The hematological parameters were measured at low dose exposure levels. Relative
liver weights were only measured for control and in the highest dose (69.6 mg/kg-day) group.
Changes in hair cystine were measured in all dose groups but the biological significance of this
effect is unknown; changes in hair cystine content may serve as a biomarker of exposure, rather
than as an indication of an adverse effect. From these data, and using standard conversions for
body weight and food consumption for Wistar rats, the NOAEL of 74.5 ppm was converted to
10.5 mg/kg-day for decreased erythrocyte count.
4.6.1.3 Chronic
No studies reporting chronic exposure to vanadium pentoxide in humans are available in
the published literature. Chronic animal studies evaluating effects of oral exposure to vanadium
pentoxide are limited. Previously, a study based on a 2.5 yr dietary exposure to vanadium
(Stokinger et. al., 1953 reported in Patty's Industrial Hygiene and Toxicology, 3rd Ed., 1981.)
utilized hair cystine as the critical effect, since hair cystine was significantly decreased (no
details provided regarding the percent decrease or dose-response) in exposed rats and this change
may reflect changes to enzymatic pathways. However, comprehensive toxicity endpoints were
not evaluated, and the number and strain of rats used was not reported in this 1953 study.
Decreased hair cystine may or may not be a biologically relevant effect. Decreased hair cystine
may not be related to vanadium exposure, but rather an effect of poor nutritional status, as a
result of an aversion to vanadium pentoxide in the feed. EPA has been unable to gain access to
the raw data. No additional oral chronic exposure studies in animals were identified in the
literature.
4.6.2. Inhalation
Table 4-20 presents a summary of the noncancer results for the acute subchronic and
chronic studies of vanadium pentoxide toxicity in humans and experimental animals. The
88
DRAFT - DO NOT CITE OR QUOTE
-------
identified noncancer health effects following occupational exposure to vanadium pentoxide in
humans include respiratory irritation, airway obstruction, chest pain, bronchitis and similar
effects; inhalation exposure in animals results in multiple health effects, including pulmonary
inflammation, lung and nasal hyplerplasia, pulmonary fibrosis, changes in nervous system
structure and function, and reproductive/developmental effects (Table 4-14). The most
comprehensive animal studies evaluated toxicity following inhalation of vanadium pentoxide in
F344/N rats and B6C3Fi mice (NTP, 2002); these studies included acute, short-term, subchronic
and chronic duration exposures and multiple toxicological endpoints. In those studies, the lung
was identified as the most sensitive organ. The reported neurotoxic effects appear after as little
as a single one hour exposure to inhaled vanadium pentoxide; although multiple effects were
identified that increased in severity with increased duration of exposure, the studies were
conducted using only one exposure level. Thus, a no-effect level for neurotoxicity could not be
determined. Major data gaps also exist for studies on developmental toxicity, reproductive
toxicity, and immunotoxicity following inhalation exposure to vanadium pentoxide, although
effects have been seen in available studies (some of which used an exposure paradigm similar to
that used for neurotoxicity studies).
4.6.2.1 Acute and Short-term
The primary identified noncancer health effect following acute inhalation exposure in
humans is respiratory irritation, cough, and mucus formation. A human controlled exposure
study (n = 100) performed by Zenz and Berg (1967) reported respiratory irritation, cough, and
mucus formation in humans exposed to vanadium pentoxide at 0, 0.1, 0.5, and 1.0 mg/m3, for 8
hrs. ANOAEL of 0.5 mg/m3 was established. These effects are confirmed in male monkeys
exposed to vanadium pentoxide at 0.5 or 5.0 mg/m3 for 6 hrs (Knecht et. al., 1985). Air flow
restriction, as measured by pulmonary function tests, was reported at 5.0 mg/m3 (LOAEL); the
NOAEL was 0.5 mg/m.3 Moreover, a WHO-IPCS (CICAD) document (2001) reports a 1-hr
inhalation exposure to vanadium pentoxide dust in rats led to an LC6y of 1.44 mg/L (1440
mg/m3). Clinical signs of toxicity included respiratory difficulty, increased respiratory tract
mucus production, and irritation of the eyes, nose, and throat (WHO-IPCS, 2001, section 11.1.1).
Neurotoxicity has also been observed in mice following a single inhalation exposure; Avila-
Costa et al. (2006) noted significant changes in brain morphology and impairment in memory in
89
DRAFT - DO NOT CITE OR QUOTE
-------
male CD-I mice compared to controls following a single 1 hr exposure to 0.02 M (2.5 mg/m3 V
) inhaled vanadium pentoxide.
Respiratory irritation was observed following short-term inhalation exposure in humans.
100 workers were reportedly exposed to 0.05 to5.3 mg/m3 vanadium for 10 hr/day, 6 days/week,
for 4 weeks (Levy et al., 1984). A LOAEL of 0.05 mg/m3 was established. However, dose-
response was not systematically measured, there were no controls, and exposure to vanadium
pentoxide could not be directly correlated to effects. The primary noncancer health effects of
short term inhalation exposure in animals include increased pulmonary inflammation, and dose-
related decreases in body weight and relative lung weight in rodents (NTP, 2002).
4.6.2.2 Subchronic
No comprehensive subchronic studies have evaluated inhalation effects in humans. The
primary noncancer health effects of subchronic inhalation exposure in animals include
nonneoplastic lung and nasal lesions (NTP, 2002), hematological parameter changes (NTP,
2002), decreased pulmonary function in male monkeys (Knecht et al., 1992), morphological
changes to the CNS (Avila-Costa et al., 2004; 2005; 2006), increased platelet counts (Gonzalez-
Villalva et al., 2006), and testicular malformation (Mussali-Galante et al., 2005; Fortoul et al.,
2007). The NTP (2002) study evaluated pulmonary and nasal endpoints in both male and female
rats and mice after 3 months exposure to vanadium pentoxide at 0, 1, 2, 4, 8 and 16 mg/m3.
Lung inflammation, lung hyperplasia and increased relative lung weight were observed at similar
low doses in rats and mice. A LOAEL of 2.0 and NOAEL of 1.0 mg/m3 were established. In
addition, rats exhibited bronchiolar exudates, microcytic erythrocytosis, lung fibrosis, and nasal
lesions. Associated LOAELs and NOAELs were established and are listed in Table 4-20. Body
weight loss and increased absolute lung weight were reported in mice and resulted in a LOAEL
of 8.0 mg/m3 and NOAEL of 4.0 mg/m3. An increased absolute and relative lung weight
LOAEL was set at 4.0 mg/m3 and 2.0 mg/m3 for these lesions observed in mice. In summary,
rats appear to be more sensitive to inhalation exposure to vanadium pentoxide than mice, based
on the occurrence of a wider variety of nonneoplastic lesions throughout the respiratory tract.
Knecht et. al. (1992) observed that inhaled vanadium pentoxide leads to adverse lung
effects by measuring decreased pulmonary function in male monkeys exposed to vanadium
pentoxide at 0.1, 0.5, or 1.1 mg/m3 for 26 weeks (6 hrs/day, 5 days/week). A LOAEL of 0.5
mgm3 and a NOAEL of 0.1 mg/m3 was established. Pulmonary function parameters were
reversible and did not repeat following subsequent challenge.
Selected hematology parameters (number of erythrocytes and reticulocytes and percent
90
DRAFT - DO NOT CITE OR QUOTE
-------
hematocrit) were significantly altered in male and female rats at the highest dose level (NTP,
2002). A LOAEL of 16.0 mg/m3 and a NOAEL of 8.0 mg/m3 were established. Other effects
such as increased platelet count following 12 weeks and altered testicular morphology following
20 weeks were tested at only one dose (5.13 mg/m3).
4.6.2.3 Chronic
Limited information is available specific to the vanadium pentoxide exposure levels in
human studies. Respiratory and other symptoms have been documented among workers
employed at facilities producing and processing vanadium pentoxide (Sjoberg, 1951; Sjoberg,
1956; Zenz et al., 1962; Kiviluoto et al., 1979; Kiviluoto, 1980; Kiviluoto et al., 1981a; Musk
and Tees, 1982; Irsigler et al., 1999). Similar effects were observed in occupational studies of
boilermakers involved in the construction, cleaning and maintenance of oil-fired boilers
(Williams, 1952; Sjoberg, 1955; Lees, 1980; Ross, 1983; Levy et al., 1984; Hauser et al., 1995a;
1995b; 2001; Woodin et al., 1998; 1999; 2000; Kim et al., 2004). Noncancer health effects of
chronic inhalation exposure in animals included nonneoplastic pulmonary lesions in male and
female rats and mice exposed to vanadium pentoxide for 2 yrs at 0, 0.5, 1.0 and 2.0 mg/m3 and 0,
1.0, 2.0, and 4.0 mg/m3, respectively (NTP, 2002). Increased incidence of alveolar and
bronchiolar epithelial hyperplasia in male rats, alveolar histiocyte infiltration, laryngeal
inflammation, and epiglottis epithelial degeneration, hyperplasia, and squamous metaplasia in
male and female rats, and goblet cell hyperplasia in nasal compartments in male rats were
reported. Based on these findings, the lowest dose for which a LOAEL could be established in
rats was 0.5 mg/m3. Additional findings observed in females included increased interstitial
fibrosis. No NOAEL was established. In both male and female mice, increased incidences of
pulmonary inflammation, and hyperplasia were reported. Increased incidences of nasal olfactory
and/or respiratory epithelium degeneration, and epiglottis metaplasia were reported 1.0 mg/m3
(LOAEL) in both male and female mice. The lung fibrosis and nasal inflammation were noted at
> 2 mg/m3 in both sexes. In general, the incidence and severity of pulmonary lesions increased
with increasing dose. No treatment-related histopathological lesions were observed in other
evaluated tissues. Interstitial fibrosis was significantly elevated in exposed male and female
mice compared to controls at 2 and 4 mg/m3. No treatment-related findings were observed in
other tissues.
91
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation Exposure of Vanadium Pentoxide
Species
(sex)
Avg
Daily
Dose
(mg/m3)
Exposure
Duration
Response at LOAEL
NOAEL
(mg/m3)
LOAEL
(mg/m3)
Comments
Reference
Acute Exposure
Monkey
(male)
Human
(not
reported)
Mouse
(male)
0.5 or 5.0
mg/m3
0,0.1,
0.5, 1.0
mg/m3
2.56
mg/m3
6 hrs (single
exposure)
8 hrs (single
exposure)
1 hr (single
exposure)
Air-flow restriction (measured
by pulmonary function tests)
Respiratory irritation, cough,
mucus formation
Impaired performance in
spatial memory, decreased
dendritic spines increased
percentage of necrotic cells in
hippocampus
0.5a
0.5
5.0
1.0
2.56
Small sample size (n = 9)
Only one dose tested
Knecht et.
al., 1985
Zenz and
Berg, 1967
Avila-Costa
et. al., 2006
Short-term Exposure
Rat
and mice
(female)
Rat
(male)
Mouse
(male)
Mouse
(female)
0, 1, 2, 4
mg/m3
0, 4, 8,
16,32
mg/m3
13 days, 6
hrs/day, 5
days/week
16 days, 6
hr/day, 5
day/week
0, 2, 4, 8, 16,
32 mg/m3
16 days, , 6
hr/day, 5
day/week
Histiocytic infiltrate and lung
inflammation (rats); alveolar
and bronchial hyperplasia
(mice)
BALFluid analysis: increased
neutrophilia and decreased
macrophage infiltration
Significant increase in total cell
count
Absolute lung weight
Relative lung weight
Decreased percentage of
macrophages in B AL Fluid
8.0
4
2.0
4.0
1.0
(rats)
2.0
(mice)
16.0
8
4.0
2.0
8.0
No study performed in
male rats or mice
Sample size = 22 rats.
Rats exposed to 32 mg/m3
were emaciated, 3 male
rats died in the 32 mg/m3
group,
Sample size = 5 of each
sex. All male mice died at
32 mg/m3
Sample size 40-60.
NTP, 2002
92
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation Exposure of Vanadium Pentoxide
Mouse
(male)
Human
(not
reported)
2.56
mg/m3
Range of
0.05 to
5.3
mg/m3
30 days (1 hr,
2 times a
week for up to
4 weeks)
4 weeks, 10
hr/day, 6
days/week
Increased protein leakage into
BAL Fluid and increased
lymphocytes
Changes in matrix
metalloproteinase levels in
multiple regions of the brain
Respiratory irritation
4.0
2.56
0.05b
Females only.
Signficant increases
occurred after 1 month of
exposure. Only one dose
level was used.
Dose and response were
not systematically
measured, vanadium
pentoxide exposure not
accurate
Colin-
Barenque et.
al., 2008
Levy et. al.,
1984
Subchronic Exposure
Mouse
(male)
Mouse
(male)
Mouse
(male)
2.56
mg/m3
2.56
mg/m3
2.56
mg/m3
8 weeks, 1 hr,
2 times/week
12 week,(l
hr/day, 2
days/weeks)
12 weeks (1
hr 2 times per
week)
Morphological changes to CNS
(cilia loss, increased cell
sloughing, ependymal cell
layer detachment, decreased
dendritic spines in
hippocampus and substantia
nigra, increased cell loss,
decreased performance in
Morris water maze)
Increased platelet counts and
altered platelet morphology
Decreased gamma globulin in
testes, Increased cell death in
spermatogonia
2.56
2.56
2.56
No clinical signs of
toxicity or other
toxicologic endpoints
were reported
Only one dose used
Avila-Costa
et. al., 2004,
2005, 2006
Gonzalez-
Villalva et.
al., 2006
Mussali-
Galante et.
al., 2005,
93
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation Exposure of Vanadium Pentoxide
Monkey
(male)
Rat (both)
and
mouse
(both)
Rat (both)
Rat
(male)
Mouse
(male)
Mouse
(female)
Rat
0.1,0.5,
or 1.1
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
0, 1, 2, 4,
26 weeks (6
hrs per day, 5
days per
week)
3 months (6
hr/day, 5
days/week)
3 months (6
hr/day, 5
days/week)
3 months (6
hr/day, 5
days/week)
3 months (6
Impaired pulmonary function
Lung hyperplasia
Lung fibrosis
Microcytic erythrocytosis, lung
inflammation, and , lung
epithelial hyperplasia,
Lung inflammation
Nasal hyperplasia and
0.1
1.0
2.0
1.0
2.0
1.0
4.0
0.5
2.0
4.0
2.0
4.0
2.0
8.0
Exposures occurred on
alternate days for two sets
of animals (some received
0.1 or 1.1 mg/m3 on
alternate days, while a
second group was
exposed to a constant
concentration (0.5
mg/m3) for all 5 days.
Pulmonary function
parameters were
reversible and did not
reappear following
subsequent challenge. N=
26 animals total, n-= 8-9
per group.
Fortoul et.
al., 2007
Knecht et.
al., 1992
NTP, 2002
NTP, 2002
94
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation Exposure of Vanadium Pentoxide
(male)
Rat
(female)
Rat
and
mouse
(male)
Rat and
Mouse
(female)
Rat and
Mouse
(male)
Rat and
Mouse
(female)
Rat (both)
8, 16
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
0, 1, 2, 4,
8, 16
mg/m3
hr/day, 5
days/week)
3 months (6
hr/day, 5
days/week)
3 months (6
hr/day, 5
days/week)
3 months (6
hr/day, 5
days/week)
3 months (6
hr/day, 5
days/week)
squamous metaplasia, body
weight gain
Body weight loss
Increased Relative lung weight
Increased absolute lung weight
Increased levels of
hematological parameters
2.0
4.0
2.0
2.0
1.0
2.0
8.0
4.0
8.0
4.0
4.0
2.0
4.0
16.0
NTP, 2002
NTP, 2002
NTP, 2002
Chronic Exposure
Rat
(male)
Rat
(female)
0,0.5,1,
2 mg/m3
2 yrs (6
hr/day, 5
days/week)
Lung hyperplasia, histiocytic
infiltration, epiglottis
degeneration, , hyperplasia and
squamous metaplasia, chronic
inflammation of the larynx, ,
goblet cell hyperplasia
Inter stititial fibrosis, and
histiocytic infiltration,
epiglottis degeneration,
hyperplasia, and squamous
—
—
0.5
0.5
No other clinical findings
or altered survival
NTP, 2002
95
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation Exposure of Vanadium Pentoxide
Rat
(female)
Rat
(female)
Mouse
(both)
Rat
(female)
Rat
(male)
Mouse
(both)
Rat
(male)
Mouse
(both)
0,0.5, 1,
2 mg/m3
0,0.5,1,
2 mg/m3
0, 1, 2, 4
mg/m3
0,0.5, 1,
2 mg/m3
0,0.5, 1,
2 mg/m3
0, 1, 2, 4
mg/m3
0,0.5,1,
2 mg/m3
0, 1, 2, 4
mg/m3
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
metaplasia, chronic
inflammation of the larynx
Lung inflammation
Lung hyperplasia
Lung inflammation, and
hyperplasia
Lung squamous metaplasia of
alveolar and bronchiolar
epithelium
Lung fibrosis
Lung fibrosis
Nasal olfactory degeneration
and respiratory degeneration
Nasal inflammation
1.0
0.5
1.0
0.5
1.0
1.0
2.0
1.0
1.0
2.0
1.0
2.0
0.5
2.0
No other clinical findings
or altered survival
No other clinical findings
or altered survival
Mice reported as thin,
survival significantly
decreased in male mice at
4 mg/m3
No other clinical findings
or altered survival
No other clinical findings
or altered survival
Mice reported as thin,
survival significantly
decreased in male mice at
4 mg/m3
No other clinical findings
or altered survival
NTP, 2002
NTP, 2002
NTP, 2002
NTP, 2002
NTP, 2002
NTP, 2002
96
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14: Summary of Noncancer Results of Repeat-Dose Studies for Inhalation Exposure of Vanadium Pentoxide
Mouse
(both)
Mouse
(both)
0, 1, 2, 4
mg/m3
0, 1, 2, 4
mg/m3
2 yrs (6
hr/day, 5
days/week)
2 yrs (6
hr/day, 5
days/week)
Nasal olfactory degeneration
and respiratory degeneration
Epiglottis metaplasia
1.0
1.0
NTP, 2002
NTP, 2002
a Single exposures not adjusted for continuous exposure
b Not adjusted for continuous exposure because of the highly intermittent exposure protocol (1 hr/day, 2 days/week)
97
DRAFT - DO NOT CITE OR QUOTE
-------
4.6.3 Mode of Action Information
There is currently insufficient evidence to establish the mode of action for vanadium
pentoxide toxicity. However, the limited data to inform the mechanisms of various non-cancer
health effects (pulmonary toxicity, neurotoxicity and reproductive toxicity) following inhalation
exposure to vanadium pentoxide are summarized below.
4.6.3.1 Pulmonary Toxicity
The mechanism of action underlying the formation of pulmonary fibroproliferative
lesions has been linked to inflammation, leading to regenerative hyperplastic responses, as
evidenced by the presence of mitogens and observed smooth muscle thickening. Oxidative
stress has also been implicated. Oxidative stress induced directly or indirectly by vanadium
pentoxide may work in combination with vanadium pentoxide-induced inflammatory responses
to activate signaling molecules such as ERK 1 or 2 and p38 kinase that lead to induction to
growth factors and generation of fibrotic lesions. These potential modes of action are supported
by a limited number of mechanistic analyses of inflammation and fibrosis following exposure to
vanadium pentoxide.
Inflammation. Pierce et. al. (1996) identified proinflammatory cytokines associated with
vanadium pentoxide exposure. Oxidative stress has been implicated in the mechanism
underlying vanadium pentoxide induced pulmonary injury. All species of vanadium may
participate in redox cycling and can generate reactive oxygen species (Carter et. al.1997).
Intracellular and extracellular H2O2 production increases significantly after 18 hr exposure to
vanadium pentoxide compared to controls (Antao-Menezes et. al. 2008). Vanadium pentoxide
induced significant IFN-|3 expression after 18 and 24 hrs but could be inhibited by catalase, an
inhibitor of H2O2. Oxidative stress was also a suggested mechanism underlying STAT-1
activation, since addition of either an NAPDH inhibitor or a xanthine oxidase inhibitor ablated
STAT-1 activation in cells exposed to vanadium pentoxide (Antao-Menezes et. al. 2008).
Moreover, spontaneous hydrogen peroxide generation by fibroblasts was depleted within
minutes by addition of vanadium pentoxide (Ingram et. al. 2003). In addition, vanadium can
contribute to inhibition of protein tyrosine phosphatases through the generation of reactive
oxygen species (Zhang et. al. 2001). It is possible that hydrogen peroxide and vanadium
pentoxide reacted to form peroxovanadium intermediates and/or ROS that led to the production
of HB-EGF in fibroblasts. Oxidative stress could also be inducing ERK and p38 kinase that lead
98
DRAFT - DO NOT CITE OR QUOTE
-------
to the production of HB-EGF (Ingram et. al. 2003) or by ROS-mediated competitive inhibition
of protein tyrosine phosphatases (Zhang et. al., 2001, Samet et. al., 1999). Gene array analysis
and subsequent confirmation by PCR revealed that various oxidative stress genes such as
superoxide dismutase (SOD2), pipecolic acid oxidase (PIPOX), and oxidative stress response
(OXR1) were altered by vanadium pentoxide exposure (Ingram et. al., 2007). Thus vanadium
pentoxide-mediated production of ROS may lead to oxidative stress and induce downstream
signaling events that result in activation of mitogens and proinflammatory cytokines that
contribute to the formation of fibroproliferative lesions in the lung.
Fibrosis. Bonner et al. has published numerous studies describing the mechanism
underlying the formation of fibroproliferative lesions in response to vanadium pentoxide
exposure. Specifically, smooth muscle thickening and increased collagen deposition was
observed beneath ciliated epithelial cells in vanadium pentoxide-exposed male Sprague-Dawley
rats (Bonner et. al. 1998). Further, proliferating myofibroblasts were the principle cell type that
contributed to the observed fibrosis (Bonner et. al. 2000). Using both in vitro and in vivo
models, Rice et. al. (1999) showed that inhibition of autophosphorylation of tyrosine kinases
reduced vanadium pentoxide-induced pulmonary fibrosis, thus implicating the tyrosine kinases
as key signaling mediators underlying the mechanism of PDGF release and ultimately,
fibrinogenesis.
Zhang et. al. (2001) and Ingram et. al. (2003) identified a second mitogen, heparin-
binding epidermal growth factor-like growth factor (HB-EGF) as an important mediator of
vanadium pentoxide-induced injury in vitro in normal human bronchial epithelial cells
(NFffiECS). Further, two signaling molecules (ERK and the p38 subunit of MAPkinase) were
activated in response to vanadium pentoxide (Ingram et. al. 2003). Gene array analysis
confirmed the importance of HB-EGF and IL-8 in vanadium pentoxide-induced lung injury and
identified several new candidate genes including growth factors (VEGF, and CTGF),
chemokines (CXCL9, CXCL10), oxidative response genes (SOD2, PIPOX, OXRI) and DNA-
binding proteins (GAS1, STAT1) (Ingram et. al., 2007).
Bonner et. al. (2003) has also reported that fibroproliferative lesions are resolved and
repair initiated in response to vanadium pentoxide. Their work illustrates that mice deficient in
prostaglandins (PG) such as the enzyme cyclooxygenase (COX)-2, are protected from vanadium
pentoxide-induced fibroproliferative lesions and indicate the potential important role of
cyclooxygenases in mitigating vanadium pentoxide induced injury. Antao-Menezes et. al.
(2008) characterized the role of STAT-1 in vanadium pentoxide pulmonary fibrosis. Their work
identified Interferon-beta (IFN-|3) as a mediator of vanadium pentoxide-induced STAT-1
activation in normal human lung fibroblasts, and linked STAT-1 activation with STAT-1
99
DRAFT - DO NOT CITE OR QUOTE
-------
dependent production of a chemokine, CXCL10, that was identified in the Ingram et. al. (2007)
gene array analysis (Figure 4-1). Thus fibroblasts appear to synthesize IFN-|3 that activates
STAT-1. STAT-1 activation simultaneously causes growth arrest and increases levels of
CXCL10 which then diminish fibrinogenesis, as a negative feedback loop. In summary,
vanadium pentoxide stimulated production of IFN-|3 activates signaling pathways that lead to the
resolution of fibrosis after vanadium pentoxide induced injury.
4.6.3.2 Neurotoxicity
The mechanism(s) underlying nervous system toxicity in response to vanadium pentoxide
is not well-characterized. A duration-dependent decrease in the number of immunoreactive TH+
neurons and morphological changes to the blood-brain barrier were observed in response to
vanadium pentoxide. It has been suggested that blood-brain barrier disruption may be related to
brain region-specific changes in metalloproteinases (MMP-2 and MMP-9) that have been seen
following vanadium pentoxide exposure (Colin-Barenque et. al. 2008), however more work is
needed to fully characterize these findings.
4.6.3.3 Reproductive Toxicity
Limited studies of vanadium pentoxide have examined reproductive and developmental
toxicity. Short-term oral exposures in weanling rats demonstrated a potential effect on bone
growth as measured by serum calcium concentrations and bone alkaline phosphatase activity
(Yamaguchi et al. 1989). Reproductive effects due to vanadium pentoxide inhalation exposure
included morphological changes to spermatogonia, spermatocytes and Sertoli cells (Fortoul et.
al., 2007). Estrous cycle length was increased in female rats but not mice following inhalation
exposure at high doses, while decreased spermatozoal motility was observed in male mice but
not male rats at high doses (NTP 2002). Excess vanadium pentoxide was found to accumulate in
the testes following inhalation exposure (Mussali-Galante et al., 2005). Moreover, significant
decreases in gamma globulin were observed in testicular samples exposed to vanadium
pentoxide (Mussali-Galante et al., 2005). Decreased gamma-globulin levels may lead to changes
in microtubule formation that would impact spermatogenesis. Injection studies described by
Wide (1984) demonstrated decreased ossification in fetuses exposed to vanadium pentoxide in
utero. Determination of a specific mode of action for reproductive toxicity is not possible due to
the limited studies examining these effects.
100
DRAFT - DO NOT CITE OR QUOTE
-------
4.7. EVALUATION OF CARCINOGENICITY
4.7.1 Summary of Overall Weight of Evidence
Under the U.S. EPA Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a)
vanadium pentoxide is "likely to be carcinogenic to humans" by the inhalation route of exposure
based on inhalation studies in male rats and male and female mice (NTP 2002). No studies
evaluating the carcinogenic potential in humans exposed to inhaled vanadium pentoxide were
identified. No studies suitable for evaluation of the oral carcinogenic potential for vanadium
pentoxide were located in the published literature. There was clear evidence of carcinogenesis in
both male and female mice based on the statistically significant and dose-related increased
incidence of alveolar/bronchiolar tumors (NTP, 2002; Ress et al., 2003). There was some
evidence of carcinogenic activity in male rats and equivocal evidence in female rats (NTP,
2002). Although the incidence of bronchi olar tumors in vanadium-pentoxide-treated rats was not
significantly increased compared to control, tumor incidence was elevated relative to historical
control in most treatment groups in male rats and some treatment groups in female rats (Table 4-
8). No other tumor type was significantly increased in either rats or mice in this study. These
results are supported by a recent study by Rondini et al (2010) also showed increased lung
tumors in male mice (A/J, BALB/C) following exposure to vanadium pentoxide along with an
initiator (MCA).
U.S. EPA's Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a) indicate for
that for tumors occurring at a site other than the initial point of contact, the weight of evidence
for carcinogenic potential may apply to all routes of exposure that have not been adequately
tested at sufficient doses. An exception occurs when there is convincing information (e.g.,
toxicokinetic data) that absorption does not occur by other routes. Information available on the
carcinogenic effects of X^Os via the inhalation route is limited to examination of the respiratory
tumors. Information on the carcinogenic effects of X^Os via the oral and dermal routes in
humans or animals is absent. Based on the observance of only portal-of-entry tumors (respiratory
tumors) following inhalation exposure, and in the absence of information to establish a mode of
action, this cancer descriptor applies only to the inhalation route of exposure. Therefore, V2O5 is
"likely to be carcinogenic to humans" by the inhalation route of exposure, and the database has
"inadequate information to assess carcinogenic potential" via the oral or dermal route.
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence
Few studies are available that assess the carcinogenic potential of vanadium pentoxide.
101
DRAFT - DO NOT CITE OR QUOTE
-------
Although both epidemiology and laboratory animal studies show similar respiratory tract
toxicity, there are currently no epidemiology studies available in the published literature
examining carcinogenicity.
Only two published laboratory animal studies provide evidence for the carcinogenic
potential of vanadium pentoxide (NTP 2002; Rondini et al., 2010). Vanadium pentoxide has
been shown to induce pulmonary tumors following inhalation exposure in a study performed by
NTP (NTP 2002). F344 rats (50/sex/group) and B6C3F1 mice (50/sex/group) were exposed to
vanadium pentoxide particles for six hours a day, five days per week for two years. Rats were
exposed to 0.5, 1.0, or 2.0 mg/m3 of vanadium pentoxide, with mice exposed to 1.0, 2.0 or 4.0
mg/m3 of vanadium pentoxide. Lung tumors were observed in male and female rats, but
incidence exceeded historical controls in male rats only (Table 4-15). Both male and female
mice showed statistically significant increases in lung tumors as compared to controls (p < 0.01).
These increases were observed in both sexes at all doses, with 50% of the male mice in the
highest exposure group dying before the end of the study. Survival rates in all other exposed
groups for both rats and mice was not significantly different from controls. Both rats and mice
showed other lesions of the respiratory tract, including inflammation, fibrosis and hyperplasia
(Tables 4-7, 4-9). Decreased body weight gain was observed as early as 3 months post-exposure
in the high dose groups of all exposed animals.
Along with the NTP study (2002), a recent study by Rondini et al (2010) examined
tumor promotion of vanadium pentoxide in three different mice strains. Lung tumors were
observed in two of three mouse strains 20 weeks after MCA tumor initiation, followed by
exposure (only males exposed) to V2O5 (4 mg/kg, 5 times weekly) (Table 4-11). A/J and
BALB/C male mice showed increases in lung tumors following exposure to both MCA (initiator)
and V2O5, but not V2O5 alone, suggesting V2O5 works as a tumor promoter.
4.7.3. Mode of Action Information
The U.S. EPA (2005a) Guidelines for Carcinogen Risk Assessment defines mode of
action (MO A) as a sequence of key events and processes, starting with the interaction of an
agent with a cell, proceeding through operational and anatomical changes, and resulting in
cancer formation. Examples of possible modes of carcinogenic action include mutagenic,
mitogenic, anti-apoptotic (inhibition of programmed cell death), and cytotoxic mechanisms with
reparative cell proliferation and immunologic suppression. There is insufficient information to
establish a carcinogenic mode of action for bronchiolar tumors observed in animals following
inhalation exposure to vanadium pentoxide.
102
DRAFT - DO NOT CITE OR QUOTE
-------
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
4.8.1. Possible Childhood Susceptibility
There are no reports of childhood susceptibility due to exposure to vanadium pentoxide.
There are also no reports indicating increased susceptibility in developings, however the data on
developmental effects of vanadium pentoxide are very limited (see discussion of data gaps
below). Young rats that consumed vanadium in the drinking water and feed were found to have
higher tissue vanadium levels 21 days after birth than they did 115 days after birth (Edel et al.
1984). The data suggest that there is a higher absorption of vanadium in these young animals
due to a greater nonselective permeability of the undeveloped intestinal barrier. Thus, age of the
rodents appears to play an important role in the absorption of vanadium in the gastrointestinal
tract. Mravcova et. al (1993) assessed the extent of vanadium pentoxide accumulation in the
bones of rats following 6 month exposure. Vanadium accumulated in the epiphyseal cartilage of
the tibia in rats with significantly higher concentrations of vanadium in the tibia and incisors of
weanling rats compared to adults. However, no dose response data for these endpoints was
reported.
4.8.2. Possible Gender Differences
Reports of gender differences are limited to the carcinogenicity data from NTP, 2002
where clear evidence of carcinogenicity was reported in male and female mice exposed to
vanadium pentoxide, and some evidence of carcinogenicity was reported in male rats, based on
observations of alveolar and bronchiolar neoplasms that exceeded historical controls in groups
exposed to vanadium pentoxide. The number of neoplasms in female rats was not higher than
that observed in historical controls and thus, a relationship between neoplasms and vanadium
pentoxide could not be established in female rats (NTP, 2002 and Ress, 2002). Thus, increased
tumor incidence in rats is equivocal overall, but the lack of any increase in females may suggest
a gender-related increase in susceptibility in males. It is unknown whether this observed
difference is applicable to humans. There are no reported differences for response to vanadium
pentoxide between genders for either animals or humans for any non-cancer endpoint.
4.8.3. Other Susceptible Populations
No data exists on the role of genetic polymorphisms in differentially susceptible human
populations in response to vanadium pentoxide exposure. In mice, research suggests that all
strains have an inflammatory response to vanadium pentoxide exposure, but the severity of
103
DRAFT - DO NOT CITE OR QUOTE
-------
inflammation varies greatly from strain to strain. Such variability in response suggests that a
genetic component may contribute to the severity of vanadium pentoxide-induced pulmonary
inflammation and tumorigenicity in mice (Rondini et al., 2010). The NTP study (2002) used
B6C3F1 mice in all of their exposure protocols for 16-day, 3 month, and 2 yr exposure studies.
Pulmonary fibrosis was observed in this hybrid strain.
Kyono et. al. (1999) used a rat model of acute bronchiolitis (Br) to investigate whether
animals with pre-existing lung conditions would be differentially susceptible to inhaled
vanadium pentoxide. Compared to exposed normal rats, Br rats exhibited delayed recovery
from pre-existing lesions and exacerbated lung inflammation. Sensitive rats also showed
reductions in the deposition and clearance rates of inhaled particles.
Rondini et al (2010) examined the effect of exposure to vanadium pentoxide in three
mouse strains of varying susceptibility to lung cancer (A/J, BALB/C and C57BL/6J) in an
initiation/promotion model (full study description in Section 4.2.2.2). Three mouse strains were
used to further understand potential susceptibility to these effects. These particular mouse
strains were selected because of their known differential susceptibility to chronic pulmonary
inflammation and carcinogenesis: A/J mice are sensitive, BALB/C are intermediate and
C57BL/6J are resistant. Statistically significant lung tumor increases were observed in A/J and
BALB/C mice as compared to the MCA-treated control (p<0.05; Table 4-11). Differences
were also observed between strains, with A/J mice showing increased tumorigenicity in response
to vanadium pentoxide. In the absence of MCA, "V^Os was not sufficient to initiate tumorigenesis
in this study. C57BL/6J had no tumors following exposure (data not shown).
Overall, the differential inflammatory responses observed in the three strains of mice
appear to positively correlate with increased levels of chemokines, such as keratinocyte-derived
chemokine (KC) and monocyte chemotactic protein-1 (MCP-1), and increased binding of
transcriptional factors NFicB and AP-1 (c-Fos), and sustained activation of MAP kinases
(MAPKs) and extracellular signal-regulated kinases 1 and 2 (ERK 1/2) suggesting inflammation
as a major response in mice. Turpin et al. (2010) examined pulmonary inflammation and fibrosis
following intranasal aspiration exposure to vanadium pentoxide with and without respiratory
syncytial virus (RSV) exposure (full study description in Section 4.4.1.2). In this study,
vanadium pentoxide exposure also caused a significant increase in cell proliferation in the
airways and lung parenchyma, lung mRNAs for TGF-|3-1, CTGF, PDGF-C, CollA2, and
mRNAs for IFN-a and -|3 and IFN-inducible chemokines CXCL9 and CXCL10 compared to
104
DRAFT - DO NOT CITE OR QUOTE
-------
controls. Pre- or post-treatment with RSV caused a significant reduction in the all mRNAs.
Together, results from this study showed that vanadium pentoxide induces inflammatory and
fibrogenic response in mouse lung and these effects were suppressed by RSV infection.
105
DRAFT - DO NOT CITE OR QUOTE
-------
5. DOSE-RESPONSE ANALYSIS
5.1. ORAL REFERENCE DOSE (RfD)
Only two studies exist on human oral exposure to vanadium pentoxide, and neither
examined health effects related to this exposure. Two studies measured cystine levels in hair and
fingernails and vanadium levels in urine (Kucera et al., 1994) or blood (Kucera et al., 1992)
following oral exposure to vanadium pentoxide. Kucera et al. (1994) detected vanadium in urine
of workers from a Czechoslovakian vanadium pentoxide production plant, however, it is
expected that these urinary levels of vanadium resulted from multiple exposure routes. Kucera et
al. (1992) measured vanadium in the hair and blood of children and the blood of adults
potentially exposed through ingestion of vanadium-contaminated drinking water (concentration
range: 0.001-0.1 mg/L). Vanadium concentrations in water supply wells exceeded the maximum
permissible limit in drinking water (0.01 mg/L) with the contamination continuing over 2 years.
Significantly increased vanadium concentrations were found in blood of exposed children
compared to unexposed children and adults, whereas vanadium levels in hair of exposed children
(adults not measured) were no different from the control group. No exposure-response
relationship could be determined for either endpoint, and changes in hair cystine levels have not
been correlated with adverse health effects. Additional studies of workers occupationally
exposed to vanadium pentoxide exist, although it is presumed that these workers were exposed
by multiple routes, inhalation was likely the primary route of exposure (Sjoberg, 1951; Sjoberg,
1956; Zenz et al., 1962; Kiviluoto et al., 1979; Kiviluoto, 1980; Kiviluoto et al., 1981b; Musk
and Tees, 1982; Irsigler et al., 1999).
Mountain et al. (1953) is the only published, peer-reviewed study that evaluated the
effects of subchronic oral vanadium pentoxide exposure in laboratory animals. Male Wistar rats
(5/group; 200 - 350 gm bw) were exposed to vanadium pentoxide for 103 days using average
daily doses of 0, 10.5, 16.4, 69.6, 141.0 mg/kg-day vanadium pentoxide in feed. Changes were
observed in body weight gain, erythrocyte count, hemoglobin, and cystine content of hair. Other
endpoints of toxicity were not reported.
The study authors reported increased body weight gain in the low exposure groups (0 -
16.4 mg/kg-day) and decreased body weight gain in the highest exposure group (141.0 mg/kg-
day). However, these data were not accompanied by any explanation, statistical analysis, or
measures of variability (SE or SD). Cystine content of hair was significantly decreased
compared to control at doses > 16.4 mg/kg-day vanadium pentoxide. Alterations in hair cystine
levels can be associated with dietary changes or altered health status (Kleinfeld et al. 1961). The
106
DRAFT - DO NOT CITE OR QUOTE
-------
authors speculated that vanadium may have inhibited enzymes, such as sulfur transferases, that
decreased the availability of cystine for hair growth. However, data to support a relationship
between decreased hair cystine levels and adverse health outcomes are not available in the
published literature. Thus, the biological significance of decreased hair cystine content is
unclear.
Table 5-1. Hematological results of oral vanadium pentoxide exposure in rats (Mountain et al.
1953).
Control
10.5 mg/kg-day
16.4 mg/kg-day
Red Cell Count (M/mm3)
Start
Finish (103 days)
Percent change between start and
finish of expt(%)
8.0
7.7
3.8
7.8
6.8
12.8
8.0
6.3
21.3
Hemoglobin, %
Start
Finish (103 days)
Percent change between start and
finish of expt (%)
15.6
15.0
3.9
15.2
14.5
4.6
15.3
13.7
10.5
Relative liver weight increases and decreases in erythrocytes and hemoglobin levels were
also reported by Mountain et al. (1953). Mean relative liver weights, reported as a ratio of liver
weight to body weight, were statistically significantly elevated above controls at 69.6 mg/kg-day
(i.e., 3.51 ± 0.06 versus 3.86 ± 0.07). Liver weight data were not reported for other doses.
Apparent dose-related decreases in RBC count (21.3%) and hemoglobin concentration (10.5%)
were observed in the 10.5 and 16.4 mg/kg-day vanadium pentoxide dose groups compared to
controls (Table 5-1). No statistical analysis was performed by the study authors on these data,
and no measure of variance was reported, precluding independent statistical analysis. The
effects on RBC count and hemoglobin concentration observed in this oral study are consistent
with the hematological effects observed in a 3-month inhalation study of vanadium pentoxide in
rats (NTP, 2002). Decreases in mean cell volume (MCV) and mean cell hemoglobin (MCH)
accompanied by erythrocyte microcytosis were suggestive of altered iron metabolism and
heme/hemoglobin production following vanadium pentoxide inhalation exposure. Based on the
dose-related decreases in RBC count in the oral study supported by the hematological effects
observed in the inhalation study, decreased RBC count was selected as the critical effect.
Although a dose-related decrease in RBC count was seen at 10.5 and 16.4 mg/kg-day
vanadium pentoxide (12.8% and 21.3%, respectively) in the Mountain et al. (1953) study, the
magnitude of this change is considered biologically significant only at 16.4 mg/kg-day, which
107
DRAFT - DO NOT CITE OR QUOTE
-------
EPA thus identified as the LOAEL. The lowest dose of 10.5 mg/kg-day was then identified by
EPA as a NOAEL.
5.1.1 Methods of Analysis
The most sensitive endpoint following oral exposure to vanadium pentoxide is decreased
RBC count with a NOAEL of 10.5 mg/kg-day vanadium pentoxide (Mountain et. al., 1953).
Because the study authors reported the decrease in RBC count as a mean with no measure of
variability (SE or SD), this continuous endpoint could not be subjected to benchmark dose
(BMD) modeling. Therefore, the NOAEL of 10.5 mg/kg-day is identified as the point of
departure (POD) for use in deriving the RfD for vanadium pentoxide.
The Agency endorses a hierarchy of approaches to derive human equivalent oral
exposures from data from laboratory animal species, with the preferred approach being
physiologically based toxicokinetic modeling (U.S. EPA, 2011). Other approaches may include
using some chemical-specific information, without a complete physiologically based
toxicokinetic model. In lieu of data to support either of these types of approaches, body weight
scaling to the % power (i.e., BW3/4) is endorsed as a general default to extrapolate lexicologically
equivalent doses of orally administered agents from all laboratory animals to humans for the
purpose of deriving an oral Reference Dose (RfD). In general, the use of BW3/4 scaling is
considered appropriate for deriving an oral RfD when the observed effects are associated with
the parent compound or a stable metabolite, not related to a portal-of-entry effect, and not related
to developmental endpoints. Use of BW3/4 scaling in combination with consideration of a
reduced interspecies uncertainty factor, UFA, is recommended as the Agency default approach.
No physiologically based toxicokinetic modeling information exists for vanadium
pentoxide. The selected critical effect (decreased red blood cell counts) is associated with the
parent compound, is not considered a portal-of-entry effect and was observed in mature rats.
Therefore, consistent with U.S. EPA guidance (U.S. EPA, 2011), the POD identified in rats (i.e.,
10.5 mg/kg-day) is converted to a human equivalent dose (HED) through application of a
dosimetric adjustment factor (DAF) derived as follows:
DAF = (BWa1/4/BWh1/4),
where
DAF = dosimetric adjustment factor
BWa= animal body weight
BWh = human body weight
108
DRAFT - DO NOT CITE OR QUOTE
-------
Using a BWa of 0.25 kg for rats and a BWh of 70 kg for humans, the resulting DAF is
0.244. Applying this DAF to the POD identified in rats yields a FLED of 2.56 mg/kg-day as
follows:
FLED = Laboratory animal dose (mg/kg-day) x DAF
= 10.5 mg/kg-day x 0.244
= 2.56 mg/kg-day
Table 5-2 Human equivalence dose conversion by BW for RfD derivation.
Species
Rat
(0.25kg)
BWa1/4 / BWh1/4 = DAF
0.25kg1/4-70kg1/4= 0.244
HED
10.5 mg/kg-
d x 0.244 =
2.56 mg/kg-
d
UFs
Total = 3000
UFA=3
UFH=10
UFS= 10
UFD=10
RfD
(mg/kg-day)
2.56-3000
= 8.5 x 10'4
a Using the BWa 0.25 kg for rats and BWh 70 kg for humans, and multiplying it by the NOAEL of 10.5 mg/kg-d
from Mountain et al. (1953).
5.1.2 RfD Derivation - Including Application of Uncertainty Factors (UF)
The NOAEL of 10.5 mg/kg-day for decreased RBC count in male rats (Mountain et. al.,
1953) was used as the POD to derive an RfD. The application of uncertainty factors to the POD,
converted to a HED, results in a composite uncertainty factor of 3,000 covering four areas of
uncertainty: 3 for interspecies extrapolation from animals to humans (UFA); 10 for human
intraspecies variability (UFH), 10 for extrapolation from a subchronic to a chronic study (UFS),
and 10 for database insufficiencies (UFo). No UF is needed for extrapolation from a LOAEL to
NOAEL because a NOAEL was used to identify the POD (US EPA, 2002).
An interspecies uncertainty factor, UFA, of 3 (10°'5 = 3.16, rounded to 3) was applied to
account for uncertainty in extrapolating from laboratory animals to humans (i.e., interspecies
variability). This interspecies uncertainty factor is comprised of two separate but equal sources
of uncertainty, toxicokinetic and toxicodynamic differences between animals and humans. For
vanadium pentoxide, toxicokinetic uncertainty was accounted for by calculation of a human
equivalent dose (FLED) through application of a dosimetric adjustment factor (DAF) as outlined
in the U.S. EPA guidance on the use of BW3/4 scaling in the derivation of the oral RfD (U.S.
109
DRAFT - DO NOT CITE OR QUOTE
-------
EPA, 2011). As the toxicokinetic differences are thus accounted for, only the toxicodynamic
uncertainty remains, and an UFA of 3 is retained to account for this residual uncertainty.
An intraspecies uncertainty factor, UFH, of 10 for intraspecies differences (human
variability) was used to account for potentially susceptible individuals in the absence of
quantitative information on the variability of response in humans.
An UFs of 10 for extrapolation from a subchronic to chronic study is applied because a
subchronic study was used for the POD.
An UFL for LOAEL to NOAEL extrapolation was not used because a NOAEL was
identified from the principal study and used as the POD.
An UFD of 10 was used for database insufficiencies due to the lack of a developmental
toxicity study and a multi-generation reproductive study for ViOs by the oral route. Studies
using alternate routes of exposure (intraperitoneal) have indicated adverse reproductive and
developmental effects in response to vanadium pentoxide, including statistically significant
increases in seminal vesicle, thymus and submandibular gland weights in male mice and body
weight, thymus, submandibular gland, and liver weights in female mice (Altamirano et al., 1991)
as well as reduced fertility (Altamirano-Lozano et al., 1996). No physiologically based
toxicokinetic models are available for conducting a route-to-route extrapolation.
Therefore, the RfD for vanadium pentoxide is calculated as follows:
RfD = NO AELHED - UF
= 2.56 mg/kg-day-3000
= 0.000085 mg/kg-day or 9E-04 mg/kg-day
Note: Because vanadium exists in several different valence states, all of which are not
equivalent lexicologically (WHO-IPCS, 2001), the values generated here apply to vanadium
pentoxide and should not be applied to other vanadium compounds.
5.1.2. Previous RfD Assessment
U.S. EPA previously derived a chronic reference dose (RfD) of 9 x 10"3 mg/kg-day for
vanadium pentoxide based on a 2.5-year dietary NOAEL of 0.89 mg/kg-day vanadium pentoxide
for decreased hair cystine content that was entered on the IRIS database in 1987 (U.S. EPA,
1987; Stokinger et. al., 1953 reported in Patty's Industrial Hygiene and Toxicology, 3rd Ed.,
1981). Limited details were provided, and this study is not available for analysis. The rats
(number and species unspecified) were exposed to dietary levels of vanadium pentoxide (0.89 or
8.9 mg/kg-day for 2.5 yrs) and assessed for growth rate, survival, and hair cystine content. Of
110
DRAFT - DO NOT CITE OR QUOTE
-------
the endpoints reported in the book chapter, decreased hair cystine was selected as the critical
effect with aNOAEL of 0.89 mg/kg-day. Upon further analysis of this study as described in
Section 5.1.1, it was determined to be inadequate for use in deriving an RfD.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
5.2.1 Choice of Principal Study and Critical Effect with Rationale and Justification
The available human and animal data identify the respiratory tract as the primary target
for chronic exposure to vanadium pentoxide. Irritation of the upper and lower respiratory tract
has been reported in several acute, subchronic and chronic occupational and case studies of
workers exposed to vanadium pentoxide in fuel-oil ash and vanadium dust (Woodin et al., 2000;
Irsigler et al., 1999; Woodin et al., 1999; Hauser et al., 1995; Levy et al., 1984; Musk and Tees,
1982; Kiviluoto, 1980; Lees, 1980; Kiviluoto et al., 1979; Zenz et al., 1962; Sjoberg, 1955;
Vintinner et al., 1955; Williams, 1952, Lewis, 1959). These results are supported by effects
observed in rodents, including inflammation, hyperplasia, and fibrosis. Although subchronic
occupational exposure studies provide supportive evidence for the respiratory tract as a target for
inhaled vanadium pentoxide, studies often failed to quantify vanadium pentoxide concentration
as a constituent in an inhaled mixture. Thus, the available occupational exposure studies are not
suitable as the basis for the RfC.
The toxicity database for inhalation exposure in laboratory animals includes two chronic
studies (Knecht et al., 1992; NTP, 2002). A 6 month cynomologus monkey study (Knecht et al.,
1992) exposed monkeys (n=8) to vanadium pentoxide aerosol (0.5 or 3.0 mg/m3) for 6 hrs/day, 5
days/week for 26 weeks. This study was not selected as the principal study due to limitations in
the number of test animals, study duration and number of doses used.
Results of the NTP (2002) study in rats and mice provide evidence of toxicity to the
upper and lower respiratory tract, including increased lung weight, inflammation, histological
lesions, and decreased pulmonary function following a 3-month inhalation exposure to vanadium
pentoxide. Pulmonary lesions were observed in F344/N rats (10 male and 10 female) and
B6C3F1 mice (10 male and 10 female) exposed for 3 months, with NOAEL and LOAEL values
of 1 and 2 mg/m3, respectively, for minimal to mild epithelial hyperplasia (Tables 4-4 and 4-6).
Other observed endpoints include increased lung bronchiolar exudate, lung fibrosis, altered nasal
morphology, and increased nasal inflammation observed in rats. Significant exposure-related
decreases in pulmonary function were also observed in male and female rats, with a LOAEL of
4 mg/m3 (pulmonary function not assessed in mice) (NTP, 2002). Other effects identified from
111
DRAFT - DO NOT CITE OR QUOTE
-------
the NTP (2002) 3 month inhalation study are the erythrocytosis , body weight and lung weight
changes. Erythrocytosis was observed in male and female rats exposed to inhaled vanadium
pentoxide for 3 months; hematological endpoints were not examined in mice (NTP, 2002). The
NOAELs for mild erythrocytosis were 1 mg/m3 in male rats and 2 mg/m3 in female rats. The
erythrocytosis is considered to be a secondary effect arising from the primary lung lesions and
was not considered as a critical effect.
Abnormal breathing, emaciation, and lethargy were observed in male and female rats
exposed to concentrations of 8 mg/m3 or higher (NOAEL 4 mg/m3, LOAEL 8 mg/m3) in
response to vanadium pentoxide after 3 months of exposure. Final body weight was statistically
significantly decreased as compared to the respective controls at 16 mg/m3 in male rats (60%)
and male mice (10%); 8 mg/m3 in male rats (10%) and mice (6%); 16 mg/m3 in female rats
(30%) and female mice (12%); 8 mg/m3 in female mice (10%); and 4 mg/m3 in female mice
(11%). Thus, a NOAEL of 4 mg/m3 and a LOAEL of 8 mg/m3 suggested for body weight loss
and associated behavioral changes in rats and male mice. Relative lung weights were
significantly increased in male and female rats (4 mg/m3) following 3-months inhalation
exposure to vanadium pentoxide. Similarly, relative lung weights were significantly increased in
male and female mice starting at 4 mg/m3.
Other studies identified morphological changes to the central nervous system (CNS) in
male mice exposed to vanadium pentoxide for up to 8 weeks (Avila-Costa et al., 2004, 2005).
Avila-Costa et al. (2004, 2005) also reported morphological changes in the substantia nigra
region of the basal ganglia and the blood-brain barrier in male mice exposed to 1.4 mgV/m3 for
up to 8 weeks; effects on central nervous system function or other comprehensive endpoints
were not reported. Using the same dose regimen (1.4 mgV/m3 two times a week for up to one
month) Avila Costa et. al. (2006) reported a time-dependent loss of dendritic spines, necrotic-
like cell death, and morphological changes to the hippocampal region and that these changes
may be related to the associated spatial memory loss. Morphological changes in the CNS
reported by Avila-Costa et al. (2004; 2005; 2006) were not considered as the critical effect due
to lack of information on the exposure-response relationship for morphological changes to the
central nervous system, as only one exposure level was tested. Moreover, the lung appears to be
the most sensitive target to inhalation exposure to vanadium pentoxide.
Results of the NTP (2002) study in rats and mice provide evidence of toxicity to the
upper and lower respiratory tract, including increased lung weight, inflammation, histological
lesions, and decreased pulmonary function following 3-month inhalation exposure to vanadium
pentoxide. Though body weights and a complete necropsy and histological analysis were
performed, adverse effects in other target organs were not identified in the chronic exposure
112
DRAFT - DO NOT CITE OR QUOTE
-------
study. The series of effects described in the NTP (2002) bioassay and supporting studies from
the available inhalation database reflect a dose-response with increase in severity with vanadium
pentoxide concentration, and a progression of respiratory effects (infiltration of macrophages,
inflammation, hyperplastic response, fibrosis). Based on the dose-response and temporal
relationship of effects throughout the course of the 2-year study, and the concordance with
effects seen in humans, the NTP (2002) study is selected as the critical study. The 2-year
exposure study in F344/N rats and B6C3F1 mice (50/sex/group) (NTP, 2002) examined the
effects of chronic inhalation (0, 0.5, 1 or 2mg/m3 in rats and 0, 1, 2, 4 mg/m3 in mice for 6hrs/d,
5d/wk, 104 weeks) of vanadium pentoxide. The NTP (2002) observed nonneoplastic lesions of
the respiratory tract in male and female rats and mice. Numerous lesions of the upper and lower
respiratory tract were observed in male and female rats and mice at the lowest exposure
concentrations tested. Specifically, epiglottis epithelial hyperplasia in male and female rats
(LOAELs of 0.5 mg/m3), and alveolar and bronchiolar hyperplasia was observed in male rats
(LOAELs of 0.5 mg/m3) and in male and female mice (1 mg/m3); NOAELs were not identified
(Tables 4-7 and 4-9). Other histological lesions observed at a LOAEL of 0.5 mg/m3 include
multiple lesions in the nose and larynx of male and female rats including epiglottis epithelial
degeneration, hyperplasia, and squamous metaplasia in both sexes and nasal goblet cell
hyperplasia in males. In male and female mice, histological lesions also occurred in the larynx
and nasal tissues at 1 mg/m3; it should be noted that 0.5 mg/m3 was not used as a dose in mice.
In mice exposed to 1 mg/m3, lesions were observed in the nose, bronchioles and lung of male
mice and the nose, larynx, bronchioles and lung of female mice. Chronic, active inflammation
and interstitial fibrosis was observed in male rats at a LOAEL of 1 mg/m3 (NOAEL 0.5 mg/m3).
The NTP (2002) chronic study is well-designed, well-controlled, and well-reported. Numerous
toxicological endpoints were assessed. As with the subchronic exposure studies, results of this
two-year inhalation study identify the upper and lower respiratory tract as the target for chronic
inhalation exposure to vanadium pentoxide. The nasal and laryngeal lesions observed by NTP
(2002) are among the most sensitive effects observed and were observed in both sexes of rats
and mice (Table 5-3). Irritation of the upper and lower respiratory tract has been reported in
several occupational and case studies of workers exposed for days or weeks to vanadium
pentoxide in fuel-oil ash and vanadium dust (Sjoberg, 1951; Sjoberg, 1956; Zenz et al., 1962;
Kiviluoto et al., 1979; Kiviluoto, 1980; Kiviluoto et al., 1981a, b; Musk and Tees, 1982; Irsigler
et al., 1999). Therefore, NTP (2002) was selected as the principal study, with effects on the
upper and lower respiratory tract selected as the critical effects.
5.2.2 Methods of Analysis
113
DRAFT - DO NOT CITE OR QUOTE
-------
To analyze the concentration response effect of vanadium pentoxide, the reported
concentrations of vanadium pentoxide were converted to human equivalent concentrations prior
to any modeling (Table 5.4, Appendix B). The Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (hereafter referred to as the RfC
Methodology) recommends converting the POD[Aoj] to a human equivalent concentration (HEC)
(U.S. EPA, 1994b). The RfC Methodology separates gases into three categories based on their
water solubility and reactivity with tissues in the respiratory tract, and recommends the use of
regional deposited dose ratios (RDDRs) for converting to HECs for particles (e.g., vanadium
pentoxide). RDDRs were calculated for rats with the RDDR computer program (U.S. EPA,
1994) using mean body weights for male and female rats and the average particle size
MMAD±GSD of 1.24±1.89 for rats as reported by NTP (2002). Human equivalent concentration
(HEC) conversions (in mg vanadium pentoxide/m3) were calculated by multiplying ConqADj] by
the pulmonary RDDR for lesions in the lung, or the extrathoracic RDDR for lesions in the larynx
and nose and are summarized in Table 5-4. Duration-adjusted exposure concentrations
(ConC[ADj]) of 0.09, 0.18 and 0.36 mg/m3, corresponding to nominal exposure concentrations of
0.5, 1 and 2 mg/m3, were calculated to account for continuous ambient exposure:
ConcADj = Cone x 6/24 x 5/7
[ADj]
HECs were calculated by multiplying
the larynx in female rats (Appendix B, Table B-5).
by the extrathoracic RDDR for lesions of
Table 5-3. Selected Nonneoplastic Lesions of the Respiratory System in Rats Exposed to Particulate
Aerosols of Vanadium Pentoxide for 2 Years (NTP, 2002)
Lesion Type and Location3
Exposure Group
Control
(%
incidence)
0.5 mg/m3
(%
incidence)
1 mg/m3
(%
incidence)
2 mg/m3
(%
incidence)
Male Rats
Percent survival (%)
40
58 |52 |54
Lung
Number of animals examined
Alveolar epithelium, hyperplasia
Bronchiole epithelium,
hyperplasia
Inflammation, chronic active
50
7(14)
3(6)
5(10)
49
24b (49)
17b (35)
8(16)
48
34b (71)
31b(65)
24b (50)
50
49b (98)
48b (96)
42b (84)
114
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-3. Selected Nonneoplastic Lesions of the Respiratory System in Rats Exposed to Particulate
Aerosols of Vanadium Pentoxide for 2 Years (NTP, 2002)
Lesion Type and Location3
Interstitial, fibrosis
Alveolus, histiocyte infiltration
Exposure Group
Control
(%
incidence)
7(14)
22 (44)
0.5 mg/m3
(%
incidence)
7(14)
40b (82)
1 mg/m3
(%
incidence)
16C(33)
45b (94)
2 mg/m3
(%
incidence)
38b (76)
50b (100)
Larynx
Number of animals examined
Inflammation, chronic
Epiglottis epithelium,
degeneration
Epiglottis epithelium, hyperplasia
Epiglottis epithelium, squamous
metaplasia
49
3(6)
0(0)
0(0)
0(0)
50
20b (40)
22b (44)
18b(36)
9b(18)
50
17b (34)
23b (46)
34b (68)
16b (32)
50
28b (56)
33b (66)
32b (64)
19b (38)
Nose
Number of animals examined
Goblet cell, hyperplasia
49
4(8)
50
15b(30)
49
12C (24)
48
17b (35)
Female Rats
Percent survival (%)
28
40 |34
30
Lung
Number of animals examined
Alveolar epithelium, hyperplasia
Bronchiole epithelium,
hyperplasia
Inflammation, chronic active
Interstitial, fibrosis
Alveolus, histiocyte infiltration
49
4(8)
6(12)
10 (20)
19 (39)
26 (53)
49
8(16)
5(10)
10 (20)
7b (14)
35C(71)
50
21b (42)
14C (28)
14 (28)
12 (24)
44b (88)
50
50b (100)
48b (96)
40b (80)
32b (64)
50b (100)
Larynx
Number of animals examined
Inflammation, chronic
Epiglottis epithelium,
degeneration
Epiglottis epithelium, hyperplasia
Epiglottis epithelium, squamous
metaplasia
50
8(16)
2(4)
0(0)
2(4)
49
26b (53)
33b (67)
25b(51)
7(14)
49
27b (55)
26b (53)
26b (53)
9(18)
50
38b (76)
40b (80)
33b (66)
16b (32)
Nose
Number of animals examined
Goblet cell, hyperplasia
50
13 (26)
50
19(38)
50
16 (32)
50
30b (60)
aNumber of animals with lesion; numbers in parantheses indicate percent incidence compared to control.
115
DRAFT - DO NOT CITE OR QUOTE
-------
bSignificantly different from control by the Poly-3 test, p< 0.01
Significantly different from control by the Poly-3 test, p< 0.05
Table 5-4: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year Inhalation
Studies in Rats (NTP, 2002)
Concentration
as Reported a
(mg/m3)
Continuous
Exposure
Adjustment
Factor b
RDDRC
Extrathoraci
c
Pulmonary
Human Equivalent
Concentration d (mg/m3)
Extrathoraci
c
Pulmonary
Male Rats (F344/N)
0
0.5
1
2
0.179
0.179
0.179
0.179
0.516
0.530
0.520
0.503
0.496
0.494
0.495
0.498
0.00
0.05
0.09
0.18
0.00
0.04
0.09
0.18
Female Rats (F344/N)
0
0.5
1
2
0.179
0.179
0.179
0.179
0.263
0.259
0.263
0.245
0.524
0.524
0.524
0.524
0.00
0.02
0.05
0.09
0.00
0.05
0.09
0.19
a "Toxicology and carcinogenesis studies of vanadium pentoxide in F344/N rats and B6C3FJ Mice", NTP, 2002.
b "Continuous Exposure Adjustment Factor" = (6/24) * (5/7); animals were exposed to vanadium pentoxide 6 hours
per day and 5 days per week.
0 Please refer to Appendix Table B-4.
d "Human Equivalent Concentration" = "Concentration as Reported" * "Continuous Exposure Adjustment Factor "
* "RDDR"
To determine the POD for derivation of the RfC, benchmark dose modeling was
conducted on lesions observed in both male and female rats in the NTP study (2002; chronic
lung inflammation, alveolar epithelium hyperplasia, chronic inflammation of the larynx,
respiratory epithelial hyperplasia of the larynx and nose) with the best-fitting model selected for
each endpoint (Appendix B).
As shown in Table 5-5, the lowest BMCL^EC] of 0.003 mg/m3 was observed for chronic
inflammation of the larynx of female rats, indicating that the larynx was the most sensitive target
for chronic inhalation exposure to vanadium pentoxide. Thus, it was chosen as the POD for the
basis of the RfC calculation.
Table 5-5: Candidate PODs for Vanadium Pentoxide Derived from NTP Studies (2002) through
BMDS Modeling.
Endpoint
Selected
BM
HEC
116
DRAFT - DO NOT CITE OR QUOTE
-------
Model a
R
(Extr
a
Risk)
BMC
(mg/m3
)
BMCLC
(mg/m3)
(Candidate
POD)
Male F344/N Rats
Lung
Alveolar Epithelium Hyperplasia
Chronic Active Inflammation
Probit
Logistic
0.1
0.1
0.016
0.035
0.013
0.029
Larynx
Chronic Inflammation
Respiratory Epithelium, Epiglottis,
Hyperplasia
LogLogistic
LogLogistic
0.1
0.1
0.017
0.008
0.012
0.006
Nose
Goblet Cell, Respiratory Epithelium,
Hyperplasia
LogLogistic
0.1
0.044
0.026
Female F344/N Rats
Lung
Alveolar Epithelium Hyperplasia
Chronic Active Inflammation
Gamma
Multistage
(StageS)
0.1
0.1
0.076
0.080
0.063
0.048
Larynx
Chronic Inflammation
Respiratory Epithelium, Epiglottis,
Hyperplasia
LogLogistic
LogLogistic
0.1
0.1
0.005
0.004
0.003
0.003
Nose
Goblet Cell, Respiratory Epithelium,
Hyperplasia
Multistage
(Stage2)
0.1
0.038
0.014
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose Technical Guidance
(2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
Degeneration of the epiglottis epithelium was not selected for BMD modeling because
the incidence of this lesion did not exhibit dose-dependence, with the same incidence observed
in the low and high dose groups. Epithelial squamous metaplasia was not selected for BMD
modeling because the incidence of this lesion was not significantly different from control at the
low- and mid-dose groups; thus, other lesions of the larynx were more sensitive endpoints. In all
vanadium pentoxide groups, lesion severity was classified as minimal to mild.
Modeling was performed using the Benchmark Dose Modeling Software (BMDS;
Version 2.1.2) (U.S. EPA, 2000b). Biological and statistical considerations were taken into
account in the selection of a benchmark response (BMR) level for all data sets. In the absence of
information indicating what magnitude of inflammatory changes in the larynx and epiglottis are
117
DRAFT - DO NOT CITE OR QUOTE
-------
considered biologically significant, the benchmark response (BMR) of 10% increase in extra risk
was used as the basis for the BMC (BMCio), with the BMCLio representing by the 95% lower
confidence limit on the BMCio (US EPA 2000b). Following the model selection steps outlined in
the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the best-fitting model was
selected. Biological and statistical considerations were taken into account in the selection of a
benchmark response (BMR) level for this data set. Statistically, a 10% level of response is
intended to represent a response level near the lower range of detectable observations in typical
studies conducted with 50 animals per dose group (U.S. EPA, 2000b).
Results of the BMDS modeling for chronic inflammation of the larynx in female rats are
summarized in Table 5-6, and for epiglottis hyperplasia in Table 5-7. As assessed by the
chi-square goodness-of-fit test, several models demonstrated adequate goodness of fit^-value >
0.1 and good visual fit (Appendix B). In accordance with the draft Benchmark Dose Technical
Guidance (US EPA, 2000b), the LogLogistic model was selected as the best fitting model. The
BMCio and BMCLio were estimated as 0.005 and 0.003 mg/m3.
Table 5-6: Benchmark Modeling Results for Incidence of Larynx Chronic Inflammation in Female
Rats, NTP (2002)
Model a
LogLogistic
LogProbit
Gamma
Multistage e
(Stage 1)
Weibull
Probit
Logistic
Goodness of Fit
P-
Value
0.46
0.27
0.19
0.03
0.03
AICb
242.0
243.6
243.7
247.5
247.6
Largest
Scaled
Residual
-0.88
-0.89
1.58
1.99
1.97
Scaled
Residual of
Interest
-0.13
-0.02
-0.57
1.99
1.97
BMR
Extra
Risk
10%
HECC
BMC
(mg/
m3)
0.005
0.003
0.007
0.013
0.014
BMC
Ld
(mg/m
3)
0.003
0.000
0.006
0.011
0.011
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
118
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-7: Benchmark Modeling Results for Incidence of Larynx Respiratory Epithelium,
Epiglottis, Hyperplasia in Female Rats, NTP (2002)
Model a
LogLogist
ic
LogProbit
Gamma
Multistage
e
(Stagel)
Weibull
Probit
Logistic
Goodness of Fit
P-
Value
0.30
0.78
0.01
0.00
0.00
AIC
b
205.3
204.2
212.3
234.6
235.3
Largest
Scaled
Residual
1.53
-0.57
2.85
3.258
-3.295
Scaled
Residual of
Interest
0.000
0.000
0.000
3.258
3.174
BMR
Extra
Risk
10%
HECC
BMC
(mg/m
3)
0.004
0.000
0.006
0.016
0.016
BMCLd
(mg/m3)
0.003
failed
0.005
0.013
0.014
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
5.2.3 RfC Derivation- Including Application of Uncertainty Factors (UFs)
Benchmark dose modeling of incidence data for chronic inflammation of the larynx in
female rats yielded the BMCLio of 0.003 mg/m3. A total UF of 300 was applied to the BMCLio
of 0.003 mg/m3: 3 for interspecies extrapolation from animals to humans (UFA); 10 for human
interspecies variability (UFH), and 10 for database insufficiencies (UFD). An uncertainty factor
(UFS) was not applied for extrapolation of subchronic to chronic study as data was used from a
chronic 2-yr study.
An UF of 3 (101/2 = 3.16, rounded to 3) was applied for interspecies extrapolation (UFA)
to account for uncertainty in extrapolating from laboratory animals to humans (i.e., interspecies
variability). This uncertainty factor is comprised of two separate and equal areas of uncertainty
to account for differences in the toxicokinetics and toxicodynamics of animals and humans. In
this assessment, toxicokinetic uncertainty was accounted for by the calculation of a human
equivalent concentration by the application of a dosimetric adjustment factor as outlined in the
RfC methodology (U.S. EPA, 1994). As the toxicokinetic differences are thus accounted for,
only the toxicodynamic uncertainties remain, and a UF of 3 is retained to account for this
residual uncertainty.
119
DRAFT - DO NOT CITE OR QUOTE
-------
An UFH of 10 for intraspecies differences (human variability) was used to account for
potentially susceptible individuals in the absence of quantitative information or information on
the variability of response in humans. Considering the pulmonary effects of ¥265, individuals
with pre-existing respiratory disorders may be more susceptible to inhaled vanadium pentoxide.
An UFL for LOAEL to NOAEL extrapolation was not used because the current approach
is to address this factor as one of the considerations in selecting a BMR for BMD modeling (US
EPA 2000b; 1994b). In this case, a BMR of a 10% increase in the incidence of chronic
inflammation of the larynx and epithelial hyperplasia of the epiglottis was selected under the
assumption that it represents a minimal, biologically significant change.
An UFD of 10 was used for database insufficiencies due to the lack of a developmental
toxicity study and a multi-generation reproductive study for ViOs by the inhalation route.
Studies using alternate routes of exposure (intraperitoneal) have indicated adverse reproductive
and developmental effects in response to vanadium pentoxide, including statistically significant
increases in seminal vesicle, thymus and submandibular gland weights in male mice and body
weight, thymus, submandibular gland, and liver weights in female mice (Altamirano et al., 1991)
as well as reduced fertility (Altamirano-Lozano et al., 1996). No pharmacokinetic models are
available for conducting a route-to-route extrapolation.
The chronic RfC for vanadium pentoxide is calculated as follows:
RfC = BMCLio-UF
= 0.003 mg/m3 - 300
= 0.00001 mg/m3 or IE-OS mg/m3
Note: Because vanadium exists in several different valence states, all of which are not
equivalent lexicologically (WHO-IPCS, 2001), the values generated here apply to vanadium
pentoxide and should not be applied to other vanadium compounds.
5.2.4. Previous RfC Assessment
An inhalation assessment for ¥265 was not previously available on IRIS.
5.3 UNCERTAINTIES IN THE INHALATION REFERENCE CONCENTRATION (RfC)
The following discussion identifies uncertainties associated with the subchronic and
chronic RfCs. As presented earlier in this section, EPA standard practices and RfC guidance
(U.S. EPA 1994a,b, 1995, 2002b) were followed in applying an uncertainty factor approach to a
120
DRAFT - DO NOT CITE OR QUOTE
-------
point of departure (POD). A BMDLio approach was used for derivation of the chronic RfC.
Factors accounting for uncertainties associated with a number of steps in the analyses were
adopted to account for extrapolating from an animal bioassay to human exposure, a diverse
human population of varying susceptibilities, and to account for database insufficiencies. These
extrapolations are carried out with standard approaches given the paucity of experimental and
human data on vanadium pentoxide to inform individual steps.
An adequate range of animal toxicology data is available for the inhalation hazard
assessment of vanadium pentoxide, as described in Section 4. Included in these studies are short-
term, subchronic, and chronic bioassays in rats, as well as a range of supporting genotoxity studies.
Toxicity associated with inhalation exposure to vanadium pentoxide is observed in reproductive
organs, the CNS, and particularly in the respiratory system, including a range of nasal and
pulmonary nonneoplastic lesions such as pulmonary inflammation, tissue morphology changes, and
development of pulmonary fibrosis. Recent mechanistic studies have investigated vanadium
pentoxide-induced pulmonary fibrosis and have contributed to some understanding of a putative
mode of action for pulmonary fibrosis in response to vanadium pentoxide.
In addition to respiratory effects, immunological and neurological effects have been seen
following vanadium pentoxide exposure in experimental animals. An inhalation study using
vanadium pentoxide (12 weeks, 1.4 mg/m3) in male mice noted altered immune status, including
a temporary increase in spleen weight, and alterations in antibody avidity following exposure to
vanadium pentoxide (Pinon-Zarate et. al. 2008). This in part supports the results of an earlier
oral study (6 months, 0-14 mg/kg-day in drinking water) in rats that also observed increased
spleen weight (Mravcova et al., 1993). Several recent studies document neurotoxic
morphological and behavioral effects (decreased dendritic spine length and increased necrosis in
neural cells) in animals following inhalation exposure to vanadium pentoxide (Avila-Costa et al.,
2004; 2005; 2006; Colin-Barenque et al., 2008). Though these immune and neurological effects
occurred at a low dose (1.4 mg/m3), in most cases only one dose was tested and precludes
formation of an adequate dose-response.
For derivation of the chronic RfC for vanadium pentoxide, a two-year inhalation study in rats
and mice (NTP, 2002) was selected as the principal study and inflammation of the larynx and
epithelial hyperplasia of the epiglottis were selected as the critical effects. Lung hyperplasia, nasal
inflammation, and lung fibrosis were also sensitive effects. Inflammation of the larynx and
epithelial hyperplasia were selected as the critical effects because they are sensitive indicators of
vanadium pentoxide-induced respiratory toxicity and yielded the lowest POD.
The selection of the benchmark dose model for the quantitation of the RfC does not lead to
significant uncertainty in estimating the POD since benchmark effect levels were within the range of
121
DRAFT - DO NOT CITE OR QUOTE
-------
experimental data for chronic inflammation of the larynx. However, the selected model, the log-
logistic model, while the best fitting model, is not the only model that adequately describes the data.
Other models could be selected to yield more extreme results, both higher and lower than those
included in this assessment.
Extrapolating from animals to humans entails further issues and uncertainties as the
magnitude of the effect and the effect itself associated with the concentration at the point of
departure in rodents are extrapolated to human response. Pharmacokinetic models are useful in
examining species differences in pharmacokinetics; however, a PBPK model for vanadium
pentoxide was not available. Therefore, toxicokinetic species differences were addressed by the
determination of a HEC through inhalation dosimetry adjustments (as described in the RfC
methodology, U.S. EPA, 1994b). A UF of 3 was applied to account for the remaining
toxicokinetic uncertainties in the extrapolation from rats and humans.
Information was unavailable for quantitative assessment of toxicokinetic or toxicodynamic
differences between animals and humans, so the three-fold uncertainty factor (UF) was used to
account for uncertainty in extrapolating from laboratory animals to humans in the derivation of the
subchronic and chronic RfC values.
Heterogeneity among humans is another uncertainty associated with the RfC. In the absence
of vanadium pentoxide-specific data on human variation, an uncertainty factor of 10 was applied to
account for uncertainty associated with variation in the human population in the derivation of the
subchronic and chronic RfCs. The range of human response to vanadium pentoxide may be larger or
smaller. Vanadium pentoxide-specific data to examine the potential magnitude of over- or
underestimation are not available.
No NOAEL was identified in the NTP, 2002 study for either the subchronic or chronic
RfCs. Effects in this study were found at the lowest exposure concentrations measured,
therefore, the LOAEL determined for this study for both subchronic and chronic effects does not
indicate where a threshold of effects would lie and the data provided in the study is not sufficient
for a dose response analysis. The data were amenable to BMD analysis and a BMCL was used
so the UF for LOAEL to NOAEL extrapolation is not applicable. The selected critical effect was
observed in longer term studies (3-months and 2-yrs) at similar doses and appeared to be a
persistent and sensitive effect.
Data gaps have been identified that are associated with uncertainties specific to the
developmental and reproductive toxicities of vanadium pentoxide following inhalation exposure.
Studies that used intraperitoneal injection to study pregnant rats exposed to vanadium pentoxide
suggest that exposure to the developing fetus may have adverse health effects. However, no
inhalation studies have assessed the risk to the developing fetus. The database also does not
122
DRAFT - DO NOT CITE OR QUOTE
-------
include a comprehensive multigenerational reproductive study to help establish the full range of
developmental endpoints and thus represents a gap in the database. To account for the lack of
inhalation developmental studies and for the lack of multigenerational reproductive toxicity
studies to establish the range of toxicities across development, a full uncertainty factor of 10 was
applied to the chronic RfC derivations.
Overall confidence in the chronic RfC is medium. Confidence in the principal study
(NTP, 2002) is high. It is a well-conducted study that used rats and mice, an adequate number of
animals, and exposure at a range of doses. Multiple toxicity endpoints were assessed. However,
confidence in the overall database is medium. The NTP study is the only chronic animal study
available. Recent mechanistic studies have investigated vanadium pentoxide-induced pulmonary
fibrosis and have contributed to a better understanding of a putative mode of action for
pulmonary fibrosis in response to vanadium pentoxide. Health effects reported among workers
exposed to vanadium pentoxide and other vanadium compounds in dust are consistent with
effects observed in experimental studies. Reflecting high confidence in the principal study and
medium confidence in the database, confidence in the RfC is medium.
5.4. CANCER ASSESSMENT
5.4.1 Choice of Study/Data - with Rationale and Justification
The 2-year NTP (2002) inhalation cancer bioassay reported an increased incidence of
alveolar/bronchiolar adenomas or carcinomas in male F344/N rats and equivocal evidence of
carcinogenic activity of vanadium pentoxide in female F344/N rats at the high dose. Male and
female B6C3F1 mice had even greater incidences of these lesions, with a statistically
significantly increased incidence of alveolar/bronchiolar adenomas or carcinomas in both male
and female B6C3F1 mice following inhalation exposure to > 1 mg/m3 of vanadium pentoxide
(NTP, 2002) (Table 5-8). These tumor types are considered relevant to humans. Human data on
the carcinogenic potential of inhalation exposure to vanadium pentoxide are not available. There
are no human or laboratory animal data to determine the carcinogenicity of vanadium pentoxide
by the oral or dermal route.
Table 5-8. Incidences of Respiratory Tumors in Mice Exposed to Vanadium Pentoxide in the 2
Year Inhalation Study (NTP, 2002)
Exposure Group
123
DRAFT - DO NOT CITE OR QUOTE
-------
Tumor Type3
Historical
Control (%
Historical
Control)
Control
1 mg/m3
2 mg/m3
4 mg/m3
Male Mice
Number of animals examined
Al veol ar/bronchi ol ar
adenomab
Al veol ar/bronchi ol ar
carcinoma
Al veol ar/bronchi ol ar
adenoma or carcinoma
1071
201 (19%)
97(9%)
285 (26.8%)
50
13 (26%)
12 (24%)
22 (28%)
50
16 (32%)
29C (58%)
42C (84%)
50
26C
(53%)
30C
(60%)
43C
(86%)
50
15 (30%)
35C (70%)
43C (86)
Female Mice
Number of animals examined
Al veol ar/bronchi ol ar
adenoma
Al veol ar/bronchi ol ar
carcinoma
Al veol ar/bronchi ol ar
adenoma or carcinoma
1075
67 (6.3%)
43 (3.9%)
109(10.1%)
50
1 (2%)
0 (0%)
1 (2%)
50
17C (34%)
23C (46%)
32C (64%)
50
23C
(46%)
18C
(36%)
35C
(70%)
50
19C(38%)
22C (44%)
32C (64%)
aNumber of animals with tumor; numbers in parentheses indicate percent incidence; particle size mass mean aerodynamic
diameter + geometric standard deviation (MMAD±GSD): 1 mg/m3= 1.3±2.9; 2 mg/m3= 1.2±2.9; 4 mg/m3=l .2±2.9
bHistorical incidence of alveolar/bronchiolar adenoma male B6C3F! mice fed in inhalation chamber controls given NIH-07 diet.
'Significantly different from control by the Poly-3 test (p<0.01)
5.4.2 Dose-Response Data
Data on the incidences of alveolar/bronchiolar adenomas or carcinomas in male and
female mice from the NTP (2002) study were used for cancer dose-response assessment. These
data are shown in Table 5-9.
5.4.3 Dose Adjustments and Extrapolation Methods
The NTP (2002) 2-year carcinogenicity study in mice was used for the derivation of an
inhalation unit risk, based on the dose-response relationship for alveolar/bronchiolar neoplasms
(adenoma and carcinoma).
Using the RDDR computer program, as specified in the RfC guidelines (U.S. EPA,
1994b), HECs (in mg/m3) were calculated at each exposure level for male and female mice using
mean body weights for males and females reported by NTP (2002) and the average particle size
MMAD±GSD of 1.26±1.87 as reported by NTP (2002). HECs were calculated by multiplying
Conc[ADj] by the RDDR for male and female mice (Table 5-9).
124
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-9: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year Inhalation
Studies (NTP, 2002)
Concentration as
Reported a (mg/m3)
Continuous Exposure
Adjustment Factor b
RDDRC
Pulmonary
HECd(mg/m3)
Pulmonary
Male Mice (B6C3FO
0
1
2
4
0.179
0.179
0.179
0.179
1.168
1.168
1.168
1.134
0.00
0.21
0.42
0.81
Female Mice (B6C3Fi)
0
1
2
4
0.179
0.179
0.179
0.179
1.168
1.143
1.077
1.023
0.00
0.20
0.38
0.73
a "Toxicology and carcinogenesis studies of vanadium pentoxide in F344/N rats and B6C3F! Mice", NTP, 2002.
b "Continuous Exposure Adjustment Factor" = (6/24) * (5/7); animals were exposed to vanadium pentoxide 6 hours
per day and 5 days per week.
0 Please refer to Appendix Table C-4.
d HEC=Human Equivalent Concentration = "Concentration as Reported" * "Continuous Exposure Adjustment
Factor " * "RDDR".
According to the U.S. EPA Guidelines for Carcinogen Risk Assessment (US EPA,
2005a), for each tumor response, a POD from the observed data should be estimated to mark the
beginning of extrapolation to lower doses. The POD is an estimated dose near the lower end of
the observed range without significant extrapolation to lower doses.10 Since all non-control
concentrations from the NTP carcinogenesis studies (2002) showed a plateau response, the data
set provided limited information about the concentration-response relationship because the
complete range of response from background to maximum must occur somewhere below the
lowest dose. Therefore, a BMR based on the response at the control concentration and the first
non-control concentration was calculated (Table C-6), and then used for estimation of POD
(BMCL, extra risk of 0.71 for male mice, 0.67 for female mice).
Modeling was performed using the Benchmark Dose Modeling Software (BMDS;
Version 2.1.2) developed by the National Center for Environmental Assessment (U.S. EPA,
2000b). The incidence of alveolar/bronchiolar adenomas and carcinomas in mice were
combined; males and females were modeled separately (Table 5-10).Models were run using the
10 If the POD is above some data points, it can fail to reflect the shape of the concentration-response curve at the
lowest doses and can introduce bias into subsequent extrapolations. However, if the POD is far below all observed
data points, it can introduce model uncertainty and parameter uncertainty that increase with the distance between the
data and the POD. Use of a POD at the lowest level supported by the data seeks to balance these considerations.
125
DRAFT - DO NOT CITE OR QUOTE
-------
default restrictions on parameters built into the BMD software. Goodness-of-fit was evaluated
using the Chi-square statistic calculated by the BMDS program. Acceptable global goodness of
fit was a p-value greater than or equal to 0.1. Each data set was first fitted with the dichotomous
multi-stage cancer model; if the goodness-of-fit p-value was < 0.05, other dichotomous models
were fitted; if still no model showed adequate goodness of fit p-value > 0.05, the highest dose
was dropped for further modeling.
Table 5-10. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar Adenoma
and Carcinoma in Male Mice, NTP (2002).
Model a
Goodness of Fit
P-
Valu
e
AIC
b
Largest
Scaled
Residual
Scaled
Residual of
Interest
BM
R
HEC
BMC
(mg/m3)
BMCLC
(mg/m3)
Primary Cancer Models
Multistage-
Cancer Stage
1,2 and 3
0.01
207.
0
2.05
0.72
Extr
a
Risk
71%
0.532
0.379
Other Dichotomous Models
LogLogistic
LogProbit
Gamma
Weibull
Logistic
Probit
0.19
0.88
0.01
0.00
0.00
200.
6
199.
6
207.
0
209.
4
210.
4
-1.42
0.13
2.05
2.15
2.19
0.04
-0.06
0.72
0.96
-1.54
Extr
a
Risk
71%
0.360
0.146
0.532
0.609
0.654
0.208
failed
0.379
0.447
0.495
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose Technical Guidance
(2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
The BMDS modeling results for models meeting goodness-of-fit criteria are summarized
in Appendix C. For incidence data in male mice, the log-logistic model was the only model that
met goodness-of-fit criteria when all three dose groups were included, predicting BMC?i and
BMCLyi values of 0.360 and 0.208 mg/m3, respectively (Appendix C). The multi-stage model fit
the incidence data for male mice when the high dose was dropped, predicting BMC?i and
BMCLyi values of 0.306 and 0.220 mg/m3, respectively.
126
DRAFT - DO NOT CITE OR QUOTE
-------
None of the dichotomous models fit tumor incidence data for female mice when all three
dose groups were included. One model (log-logistic) fit the incidence data for female mice when
the high dose was dropped, predicting BMCey and BMCL6y values of 0.237 and 0.161 mg/m3,
respectively.
The BMCLyi of 0.208 mg/m3 for male mice was selected as the point of departure (POD)
for derivation of the inhalation unit risk as as this was the only model fit when including all
doses for analysis of lung tumor formation following inhalation exposure to vanadium pentoxide.
5.4.4 Inhalation Unit Risk
Data to support a mode of action for the carcinogen!city of vanadium pentoxide are
insufficient, although some data suggest that either a mutagenic or a cytotoxic and reparative
proliferation mode of action is operative. In the absence of such data, extrapolation from the
point of departure to lower doses was conducted by using a linear approach.
The inhalation unit risk represents an upper bound, continuous lifetime exposure risk
estimate and is calculated as BMR/BMCL [0.717(0.208 mg/m3)]. The HEC BMCLyi for extra
risk of alveolar/bronchiolar adenomas or carcinomas in male B6C3F1 mice exposed to vanadium
pentoxide results in an inhalation unit risk of 3.4 (mg/m3)"1. This value was derived by linear
extrapolation to the origin from the point of departure of 0.208 mg/m3 and represents an upper-
bound estimate.
5.4.5 Oral Cancer Slope Factor
No human data or animal studies relevant to the carcinogen!city of vanadium pentoxide
following oral exposure were located in the published literature. Therefore, an oral cancer slope
factor is not derived.
5.4.6 Uncertainties in Cancer Risk Values
Extrapolation of study data to estimate potential risks to human populations from
exposure to vanadium pentoxide has engendered some uncertainty in the results. Several types
of uncertainty may be considered quantitatively, but other important uncertainties cannot be
considered quantitatively. Section 5.4.5.1 and Table 5-11 summarize principal uncertainties.
Carcinogenicity due to chronic exposure of vanadium pentoxide was observed in two
species (NTP, 2002), with the carcinogenicity more definitive in mice than in rats, particularly
female rats, where the increased incidence of neoplasms did not exceed what would be expected
due to spontaneous tumor formation. Similarly, spontaneous tumors were observed in male rats
127
DRAFT - DO NOT CITE OR QUOTE
-------
at control levels, though tumor incidence increased in male rats in response to exposure to
vanadium pentoxide. The confidence in the database is low. Respiratory tract carcinogenicity
was reported in two rodent species (i.e., mice and rats) in awell-conducted NTP study (2002) and
supported by a study in mice (Rondini et al., 2010). Human variability in response to vanadium
pentoxide is unknown and, humans occupationally exposed to vanadium pentoxide are often
simultaneously exposed to other valence states of vanadium and to other inhaled environmental
toxicants simultaneously. Genotoxicity is equivocal and no further mode of action information
is available. Overall confidence in the inhalation unit risk is low.
Choice of low-dose extrapolation approach. The Mode of Action (MO A) is a key
consideration in clarifying how risks should be estimated for low-dose exposure. A linear low-
dose extrapolation approach was used to estimate human carcinogenic risk associated with
vanadium pentoxide exposure due to the unavailability of data that supports any specific mode of
carcinogenic action of vanadium pentoxide.
Dose metric. Vanadium exists in the +5 valence state in vanadium pentoxide. Other
valence states ranging from -1 to +5 exist. Frequently, vanadium exposures involve a mixture of
vanadium compounds ranging mostly from +3 to +5 valence state. The carcinogenic potential of
other valence states of vanadium has not been established. It is not known whether vanadium
pentoxide dissociates to other valence states with known carcinogenic potential or whether some
other valence state or some combination of+5 and other valence states is responsible for the
observed toxicity. If the actual carcinogenic moiety is proportional to administered exposure,
then use of administered exposure as the dose metric is the preferred choice.
Statistical uncertainty at the point of departure. Parameter uncertainty can be assessed
through confidence intervals. Each description of parameter uncertainty assumes that the
underlying model and associated assumptions are valid. For the log-logistic cancer model
applied to the male mice data, there is a reasonably small degree of uncertainty at a 71% increase
in tumor incidence (the point of departure for linear low-dose extrapolation).
Bioassay selection. The study by NTP (2002) was selected for development of an
inhalation unit risk. This was a well-designed study, conducted in both sexes in two species with
an adequate number of animals per dose group. The number of test animals allocated among the
two dose levels and an untreated control group was adequate, with examination of appropriate
toxicological endpoints in both sexes of rats and mice. Both genders of mice exhibited a
statistically significant increased incidence of lung tumors.
Choice of species/gender: The inhalation unit risk for vanadium pentoxide was quantified
using the tumor incidence data for male mice which was more amenable to BMD modeling than
tumor incidence in female mice. In addition, lung neoplasm incidence was reported in male and
128
DRAFT - DO NOT CITE OR QUOTE
-------
female rats, though neither gender of rat was as sensitive as mice. A 71% tumor incidence level
was observed in female mice at the lowest exposure level (1 mg/m3), suggesting that lower doses
may have revealed more information about low dose region of the dose response curve. Male
mice demonstrated a high background rate of lung tumors, with spontaneous lung neoplasms
observed in male mice at control levels (up to 28% of male mice). Tumor incidence increased
significantly (84%) at the lowest dose level tested (1 mg/m3). While these incidence response
rates were higher in male mice than those of the females at the comparable exposure level,
suggesting greater sensitivity of the male mice, there is no information concerning the dose-
response relationships at lower exposure levels.In other words, the behavior of vanadium
pentoxide at 1.0 mg/m3 in male mice may not inform the tumor response to vanadium pentoxide
at lower exposures.
Relevance to humans. In the absence of direct human data, the most appropriate animal
bioassays to use in the derivation of cancer risk values are chronic (i.e., lifetime) studies in two
species of rodents. The inhalation unit risk was derived from the combined tumor incidence of
lung adenomas and carcinomas in male mice. The information investigating the mode of action
of the lung tumors observed in the chronic animal bioassay, however, is limited. The
genotoxicity studies provide inadequate evidence of a genotoxic mode of action, and there are
inadequate data to support alternative mode-of-action hypotheses.
Human population variability. The extent of inter-individual variability in animals for
vanadium pentoxide metabolism has not been characterized. Strain differences in the response
to vanadium pentoxide-induced pulmonary fibrosis in rodents suggests a genetic component to
susceptibility. Moreover, humans occupationally exposed to vanadium pentoxide are often
simultaneously exposed to other valence states of vanadium and to other inhaled environmental
toxicants simultaneously. This lack of understanding about potential differences in metabolism
and susceptibility across exposed human populations thus represents a source of uncertainty.
Table 5-11. Summary of uncertainty in the vanadium pentoxide cancer risk assessment.
Consideration/
Approach
Impact on
inhalation unit
risk
Decision
Justification
129
DRAFT - DO NOT CITE OR QUOTE
-------
Consideration/
Approach
Low-dose
extrapolation
procedure
Dose metric
Cross-species
scaling
Statistical
uncertainty at
POD
Bioassay
Species /gender
combination
Impact on
inhalation unit
risk
Departure from
EPA' s Guidelines
for Carcinogen
Risk Assessment
POD paradigm, if
justified, could |
or t slope factor
an unknown
extent
Alternatives
could t or [ slope
factor by an
unknown extent
Alternative
methods could |
or | inhalation
unit risk [e.g.,
3. 5-fold |
(scaling by BW)
or t 2-fold
(scaling by BW2/3
)1
j slope factor if
MLE used rather
than lower bound
on POD
Alternatives
could t or j slope
factor by an
unknown extent
Human risk could
[ or |, depending
on relative
sensitivity
Decision
Log-logistic
cancer model
to determine
POD, linear
low-dose
extrapolation
from POD
Used
administered
exposure
RDDR
LEC (method
for calculating
reasonable
upper bound
slope factor)
NTP study
Male mice
lung cancer
Justification
Available MOA data do not inform
selection of dose-response model.
Experimental evidence supports a role of
other valence states of vanadium in the
body on health effects, but no data exists
to support carcinogenicity due to other
forms of vanadium.
RDDR software was used to adjust for
toxicokinetic differences in inhalation
dosimetry.
Lower bound is 95% confidence interval
on administered exposure.
Alternative bioassays were unavailable.
There are no MOA data to inform
extrapolation approach for any choice. It
was assumed that humans are as sensitive
as the most sensitive rodent gender/species
tested; true correspondence is unknown.
The carcinogenic response occurs across
genders and species.
130
DRAFT - DO NOT CITE OR QUOTE
-------
Consideration/
Approach
Human
relevance of
mouse tumor
data
Human
population
variability in
metabolism and
response/
sensitive
subpopulations
Impact on
inhalation unit
risk
Human relevance
of mouse tumor
data could | slope
factor
Low-dose risk |
to an unknown
extent
Decision
Lung tumors in
mice are
relevant to
human
exposure
Considered
qualitatively
Justification
Vanadium pentoxide may be carcinogenic
through an unknown mode of action.
No data to support range of human
variability/sensitivity.
131
DRAFT - DO NOT CITE OR QUOTE
-------
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE
6.1 Human Hazard Potential
Vanadium is a metal commonly found in ores, tars, coals and oils and is used as an alloy in
steel. The toxicity of vanadium depends on its valence state, which can range from -1 to +5,
depending on pH and other factors. Vanadium pentoxide (V2Os), sodium metavanadate
(NaVOs), and ammonium metavanadate (NH/tVOs) all contain vanadium in the +5 oxidation
state. This Toxicological Review focuses exclusively on vanadium pentoxide (^Os), the most
common form of vanadium used commercially. In addition, V2O5 is the only compound that is
covalently bonded. Occupational exposure to vanadium pentoxide occurs primarily via
inhalation of dust generated during vanadium processing and fuel-oil ash generation during
cleaning of oil-burning boilers and furnaces.
Toxicokinetics of orally administered vanadium pentoxide is not available in humans.
Kucera et. al. (1992) attempted to measure vanadium in the hair of children who had been
exposed to vanadium in drinking water, but exposure could not be tied to vanadium pentoxide
specifically. Oral toxicokinetic studies in animals are limited. One study investigated the
toxicokinetics of components of various metals, including vanadium, in female B6C3F1 mice
(Radike et al., 2002). In this study, vanadium was detected in small intestine, kidney and the
femur. Similarly, there are a few studies that evaluated the toxicokinetics of inhaled vanadium,
but there are no data in humans. Occupational studies of inhaled vanadium pentoxide indicate
that vanadium is absorbed by humans, and levels in blood and urine rapidly declined following
exposure (Kiviluoto et al., 1981b). Results of toxicokinetics studies of inhaled or intratracheally
administered vanadium pentoxide in rats show that vanadium pentoxide is absorbed from the
lung, undergoes a wide distribution to liver, kidney, bone, blood, gastrointestinal tract, and
ovary, and is eliminated primarily in the urine (NTP 2002; Dill et al., 2004; Roshchin et al.,
1980).
The animal oral toxicity database reported decreased hair cystine (Mountain et. al., 1953,
Stokinger et. al., 1953), hematological effects, including a decrease in RBC and hemoglobin
(Mountain et. al., 1953), and a variety of reproductive and developmental effects following
exposure to vanadium pentoxide (Altamirano et. al., 1993; Zhang et. al., 1993).
Studies of human exposure to inhaled vanadium pentoxide include primarily
occupational reports and one controlled human exposure study. Several case study summaries
reported upper and lower respiratory tract irritation and inflammation among workers with
132
DRAFT - DO NOT CITE OR QUOTE
-------
inhalation exposure to vanadium pentoxide and other vanadium compounds in dust during
vanadium processing or to fuel oil ash during cleaning and maintenence of oil burning boilers
(section 4.1.2). However, the chemical composition of the fuel-oil ash and vanadium dust or
exposure measurements for vanadium pentoxide generally were not reported. Thus, in many
cases, specific relationships between vanadium pentoxide and adverse respiratory effects cannot
be definitively determined. Sjoberg (1955) published seven case reports of vanadium-induced
bronchitis in workers cleaning boilers where the concentrations of vanadium pentoxide particles
(10-20 [i in diameter) in the air were 2-85 mg/m3. Self-reported respiratory symptoms
included cough, rhinitis, wheeze, sore throat, and conjunctivitis. All symptoms resolved by two
weeks post-exposure, but would reappear upon re-exposure. Of 100 workers occupationally
exposed to vanadium pentoxide fume (0.05 - 5.3 mg/m3) via an oil-to-coal power plant
conversion, 74 reported severe respiratory tract irritation (Levy et al., 1984). Estimated daily
nasal and lung dose of vanadium was associated with incidence and severity of upper airway
symptoms (nasal congestion/irritation, throat irritation) and lower airway symptoms (chest
tightness, wheeze, cough, and sputum production) in a dose-related manner in a prospective
clinical study of boilermakers and utility workers involved in the overhaul of a large, oil-fired
boiler over a six week period (Woodin et al., 1998, 1999, 2000). Geometric mean concentrations
of vanadium (9-10 hour shifts) measured in the breathing zone ranged from 1.1-8.9 |ig/m3.
Reductions in pulmonary function were measured among boilermakers with exposure to fly ash
for weeks or years, however an association with vanadium has not been established (Woodin et
al., 1999; Hauser et al., 1995a; 2001).There are no reported cases of cancer in workers examined
over sufficient latency period as a result of exposure to vanadium pentoxide.
Respiratory effects in experimental animals are well documented and include lesions in
the nasal compartment, larynx, and lung. In addition to respiratory effects observed in animals
(NTP, 2002; Knecht et. al., 1985, 1992), there are some animal studies documenting changes to
spermatogenesis and testicular ultrastructural changes (Mussali-Galante et. al., 2005) and CNS
effects (Colin-Barenque et. al., 2007; Avila-Costa et. al., 2004, 2005, 2006) in response to
inhaled vanadium pentoxide. Following repeated inhalation to vanadium pentoxide, the
respiratory system is the most sensitive target for noncancer toxicity in rats and mice. Short term
inhalation exposure (13-days) in F344 rats and B6C3F1 mice was associated with histiocytic
infiltration at 1.0 mg/m3, and short-term and subchronic inhalation exposure (3 month) in F344
rats and B6C3F1 mice was associated with increased lung epithelial hyperplasia and
inflammation at 2.0 mg/m3 (NTP, 2002). Long term exposure (2yr) in F344 rats and B6C3F1
mice similarly resulted in increased incidence of respiratory toxicity and carcinogenicity (NTP,
2002). Lung inflammation was observed at 1.0 mg/m3 and lung hyperplasia was observed at 2.0
133
DRAFT - DO NOT CITE OR QUOTE
-------
mg/m3 in female rats. Lung inflammation and hyperplasia were observed at 1.0 mg/m3 in both
male and female mice. In male and female mice, nonneoplastic lesions also occurred in the
larynx (squamous metaplasia of epiglottis epithelium) and nasal tissues (degeneration and
atrophy of olfactory epithelium and degeneration of respiratory epithelium) at 1 mg/m3; it should
be noted that 1 mg/m3 was the lowest dose in mice. In mice exposed to 1 mg/m3, lesions were
observed in the nose, bronchioles and lung of male mice and the nose, larynx, bronchioles and
lung of female mice. Chronic, active inflammation and interstitial fibrosis was observed in male
rats at a LOAEL of 1 mg/m3 (NOAEL 0.5 mg/m3). The nasal and laryngeal lesions appear to be
among the most sensitive effects observed at the lowest dose tested. Moreover, multiple lesion
types were observed in both sexes of rats and thus, the LOAEL of 0.5 mg/m3 for nasal and
laryngeal lesions was selected as the critical effect.
Reproductive and developmental studies reveal various altered endpoints in response to
vanadium pentoxide via oral and inhalation routes. Vanadium pentoxide delivered orally to
weanling rats for three days caused a significant increase in alkaline phosphatase activity and
DNA content in the diaphysis of femoral bones suggesting that vanadium pentoxide may be
linked to bone formation in the developing rat (Yamaguchi et. al., 1989). Mussali-Galante et. al.,
(2005) identified accumulation of vanadium pentoxide in the testes after 1 week of inhalation
exposure in male CD-I mice (0.02M or 1.4 mg V/m3). Gamma tubulin was significantly
decreased in testes exposed to vanadium pentoxide and may suggest changes in microtubule
function that may impact spermatogenesis. In a follow-up study, Fortoul et. al. (2007) reported
necrotic cell death in spermatogonia and increased nuclear distortion in spermatocytes in male
CD-I mice exposed by inhalation to vanadium pentoxide (0.02M or 1.4 mg V/m3).
A number of intraperitoneal studies have evaluated reproductive and developmental
endpoints. Common reproductive effects include increased reproductive organ weights in male
rats (Altamirano et. al., 1991), increased incidence of apoptotic spermatogonia in male mice
(Aragon et. al., 2005) and reduced sperm motility, reduced sperm count, and increased numbers
of abnormal sperm in male CD1 mice (Altamirano-Lozano et. al., 1996). In developmental
studies, reduced ossification in the developing fore and hindlimbs of fetuses have been reported
in response to intraperitoneal exposure to vanadium pentoxide during gestation days 6-15 in
pregnant CD-I mice (Altamirano-Lozano et. al., 1993) and Wistar rats (Zhang et. al., 1993a, b).
Other reported effects range from reduced fetal weight and increased placenta weight.
The genotoxicity database for vanadium pentoxide is limited. The evidence for
genotoxicity in humans is limited. There are few studies examining genotoxicity in humans in
vivo, with equivocal results. Ivancsits et. al. (2002) reported no differences in DNA strand
breaks, oxidative damage, or sister chromatid exchange frequency in leukocytes between control
134
DRAFT - DO NOT CITE OR QUOTE
-------
and vanadium pentoxide-exposed workers. Ehrlich et. al., (2008) noted changes in DNA
stability and DNA repair in leukocytes of occupationally-exposed workers as compared to
controls. Studies have demonstrated a genotoxic effect of vanadium pentoxide on human cells in
vitro. Ivancsits et al. (2002) demonstrated significant increases in DNA damage as measured by
the Comet assay in both leukocytes and fibroblasts but with different dose sensitivity, while
Kleinsasser et. al (2003) noted DNA migration differences occurred dose-dependently in
peripheral blood lymphocytes but not in nasal mucosa cells. Earlier studies in human
lymphocyte cultures also demonstrated increased aneuploidy (Ramirez et al. 1997; Rojas et al.
1996) and DNA damage (Roldan and Altamirano 1990) following exposure to vanadium
pentoxide. Thus, vanadium pentoxide-induced mutagenicity may occur at doses higher than
those measured in these occupational exposures, may be tissue-specific and may be associated
with oxidative stress rather than direct DNA damage. Experimental data in animals provide
evidence of some types of genotoxicity following in vivo exposure to vanadium pentoxide.
Vanadium pentoxide administered by inhalation to mice or rats did not increase the frequency of
micronucleated normochromatic erythrocytes in peripheral blood (NTP, 2002). Genotoxicity
assessed in male CD-I mice following intraperitoneal injection caused no treatment related
effects in mitotic index, average generational time, or sister chromatid exchange (Altamirano-
Lozano et. al., 1993). However, DNA damage was detected in six organs from vanadium
pentoxide-treated mice via intraperitoneal injection (Altamirano-Lorano et. al., 1999). Vanadium
pentoxide produced gene mutations in two bacterial test systems (Kada et. al., 1980; Kanematsu
et. al., 1980) but negative results in the NTP (2002) study. Vanadium pentoxide produced DNA
strand breaks, aneuploidy, and micronuclei induction but did not produce chromosomal
aberrations or sister chromatid exchange in various cell lines (Ivancsits et al. 2002; Kleinsasser
et al. 2003; Ramirez et al. 1997; Rojas et al. 1996; Roldan and Altamirano 1990; Zhong et al.,
1994).
Most of the effects of vanadium pentoxide are thought to be produced by the parent
compound, primarily by inducing cell damage and pulmonary fibrosis (NTP, 2002, Bonner et.
al., 1998, Bonner et. al., 2000). Pulmonary fibrosis is mediated either directly by vanadium
pentoxide-induced changes to cell signaling molecules or via vanadium pentoxide-induced
oxidative stress that induces cellular changes leading to fibrosis (Bonner et. al., 1998, Rice et.
al., 1999, Ingram et. al., 2003). It is possible that pulmonary fibrosis could be a key event
leading to eventual tumorigenesis.
Vanadium pentoxide is "likely carcinogenic to humans" by the inhalation route of
exposure under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). Existing
cancer data comes from two animal species (mice, rats) reported in one well-characterized study
135
DRAFT - DO NOT CITE OR QUOTE
-------
(NTP, 2002). Increased incidence of pulmonary adenomas and carcinomas were observed in
male and female mice exposed to vanadium pentoxide for 2 years. There is evidence of
carcinogenicity in male rats exposed to vanadium pentoxide for 2 years, as respiratory tumor
incidence exceeded that in historical controls. In female rats, the incidence of respiratory
tumorigenesis in response to vanadium pentoxide did not exceed historical controls. Tumor
incidence has not been documented in humans.
Information available on the carcinogenic effects of "V^Os via the inhalation route is
limited to examination of the respiratory tumors. Information on the carcinogenic effects of
¥265 via the oral and dermal routes in humans or animals is absent. Based on the observance of
only respiratory tumors following inhalation exposure, and in the absence of information to
establish a mode of action, this cancer descriptor applies only to the inhalation route of exposure.
Therefore, the database has "inadequate information to assess carcinogenic potential" of "V^Os
via the oral or dermal route.
The mode of action underlying tumorigenicity in rats and mice has not been established.
The genotoxicity database for vanadium pentoxide is equivocal, including both positive and
negative studies for mutation, DNA damage and chromosomal aberrations. A nongenotoxic
mode of action hypothesis involving hyperplasia and development of fibrotic pulmonary lesions
is supported by the presence of hyperplastic lesions and pulmonary fibrosis at earlier time points
and at lower doses. However, the dose-response relationship is not robust and a clear
relationship linking these effects to the tumor response has not been established. It is unknown
whether cytotoxicity may be a required precursor event for vanadium pentoxide-induced cell
proliferation. Sufficient data regarding a plausible dose response and temporal progression from
cytotoxicity to hyperplasia to fibrosis to tumorigenesis are not available.
6.2. Dose Response
6.2.1 Noncancer
Limited studies are available examining the toxicity of vanadium pentoxide following
oral exposure in humans or laboratory animals. Decreased RBC count and hemoglobin was
observed following subchronic oral exposure to vanadium pentoxide in rats (Mountain et. al.
1953). The authors reported the decrease of RBC as a mean with no measure of variability
between animals, therefore the continuous data was not amenable to BMD modeling. Derivation
of an RfD was based on a NOAEL of 10.5 mg/kg-d for the critical effect decreased RBC counts
and divided by a total UF of 3000, 3 to represent interspecies toxicodynamic uncertainties, 10 for
interhuman variability in the absence of quantitative information on the variability of response in
136
DRAFT - DO NOT CITE OR QUOTE
-------
humans, 10 for extrapolation from subchronic to chronic study and 10 for database deficiencies
to arrive at the chronic RfD of 9 x 10 "4 mg/kg-day.
Confidence in the principal study for derivation of the RfD (Mountain et al., 1953) is
low. Mountain et al. (1953) is a well-conducted study with numerous doses, but is a subchronic
study with a small sample size (n = 5) for one species of rodents (Wistar rat), using one gender
(male), limited endpoints and time points. Confidence in the critical effect is medium.
Hematological effects have been documented in other rodent species in response to inhaled
vanadium pentoxide (NTP, 2002). However, confidence in the overall database is low.
Mountain et al. (1953) is the single relevant peer-reviewed study for derivation of the chronic
RfD. Thus, overall confidence in the chronic RfD is low.
Pulmonary effects have been documented in numerous species, including humans and
primates, in response to inhaled vanadium pentoxide. Inflammation and histiocytic infiltrate are
common hallmarks of initial vanadium-induced pulmonary injury. The database also includes
occupational (inhalation) and laboratory animal (inhalation and oral) studies demonstrating
possible immunotoxicity following exposure to vanadium pentoxide. In addition, the database
includes studies of neurotoxicity and reproductive and developmental toxicity studies following
inhalation exposure to vanadium pentoxide in rodents. Overall, the lung is the most sensitive
target for noncancer toxicity in rats and mice following chronic inhalation exposure to vanadium
pentoxide. Nonneoplastic lung lesions (specifically laryngeal lesions) were selected as the most
sensitive endpoint from a well-conducted chronic study (NTP, 2002; 2-year rodent bioassay).
The chronic inflammation of the larynx and epithelial hyperplasia of the epiglottis in rats were
chosen as critical effects because they are the most sensitive effects and the most proximal to
route of exposure (inhalation). These effects were observed in both male and female rats and
mice.
The dose-response pattern for laryngeal lesions (NTP, 2002) was amenable to BMD
modeling and was used for derivation of the chronic RfC. Two laryngeal lesions were selected
for modeling because they had the lowest Regional Deposited Dose Ratio (RDDR) value and
corresponding BMCL[HEC]. In accordance with U.S. EPA Benchmark Dose methodology
(2000b), a benchmark response (BMR) of 10% increase in extra risk was selected to represent a
minimally adverse level. All available dichotomous models were fit to the incidence data for
chronic inflammation and for epithelial hyperplasia of the epiglottis. The model with the lowest
Akaike's Information Criteria (AIC) value was considered to provide a superior fit. Benchmark
dose modeling of incidence data for chronic inflammation of the larynx and epithelial
hyperplasia of the epiglottis in female rats yielded the same BMCLio of 0.003 mg/m3. The
shared BMCLio of 0.003 mg/m3 for either chronic inflammation of the larynx or epithelial
137
DRAFT - DO NOT CITE OR QUOTE
-------
hyperplasia of the epiglottis was divided by a total UF of 300, 3 to represent interspecies
toxicodynamic uncertainties, 10 for interhuman variability in the absence of quantitative
information on the variability of response in humans, and 10 for database deficiencies to arrive at
the chronic RfC of 1 x 10 "5 mg/m3.
Confidence in the principal study for derivation of the RfC (NTP, 2002) is high. NTP
(2002) is a well-conducted study with numerous doses, a large sample size of two species of
rodents, using both genders, numerous endpoints and time points. Confidence in the critical
effect is high. Pulmonary effects have been documented in numerous species, including humans
and primates, in response to inhaled vanadium pentoxide. Laryngeal lesions are a relevant
proximal portal of entry target tissue for inhalation exposure and serves as an indicator of
vanadium pentoxide-induced pulmonary injury. However, confidence in the overall database is
medium. NTP (2002) remains the single relevant study for use in the derivation of the chronic
RfC. Thus, overall confidence in the chronic RfC is medium.
6.2.2 Cancer
There are no studies identifying cancer effects in following oral exposure to vanadium
pentoxide in either humans or animals, or following inhalation exposure in humans.
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), the database
for vanadium pentoxide indicates that it is "likely to be carcinogenic to humans " via the
inhalation route of exposure. This determination is based predominantly on the NTP (2002)
study, which found positive evidence of lung tumors in both sexes of mice and male rats after
chronic vanadium pentoxide inhalation exposure. This weight of evidence conclusion takes into
consideration the NTP (2002) cancer bioassay, the available human studies, and other laboratory
animal studies. Information available on the carcinogenic effects of X^Os via the inhalation route
is limited to examination of the respiratory tumors. Information on the carcinogenic effects of
"V^Os via the oral and dermal routes in humans or animals is absent. Based on the observance of
only respiratory tumors following inhalation exposure, and in the absence of information to
establish a mode of action, this cancer descriptor applies only to the inhalation route of exposure.
Therefore, the database has "inadequate information to assess carcinogenic potential" of X^Os
via the oral or dermal route.
The increased incidence of lung tumors in male mice observed in the NTP (2002) 104
week inhalation study was used to calculate the inhalation unit risk for vanadium pentoxide. The
calculated cancer inhalation unit risk for vanadium pentoxide is 3.4 ug/m3 for the development of
alveolar/bronchiolar adenomas or carcinomas in male B6C3F1 mice. This value was derived
from BMCL10, the 95% lower bound on the dose associated with 71% extra cancer risk of
138
DRAFT - DO NOT CITE OR QUOTE
-------
respiratory carcinoma in male B6C3F1 mice, by dividing the BMR (0.71) by the BMCL10, and
represents the upper bound, continuous lifetime exposure estimate of cancer potency. The
BMCL10, lower 95% bound on exposure at 71% risk, is 2.08 x 10"1 mg/m3 and the slope of the
linear extrapolation from the BMCL to the origin = 0.71/2.08 x 10"1 mg/m3 = 3.4 (mg/m3)"1. A
linear low-dose extrapolation approach was used to estimate human carcinogenic risk associated
with vanadium pentoxide exposure due to a lack of data that supports any specific mode of
carcinogenic action of vanadium pentoxide. Therefore, the derived inhalation unit risk to
vanadium pentoxide is 3.4 per mg/m3.
Areas of uncertainty exist for this cancer assessment. The log-logistic model was selected
to model lung tumor incidence in male mice; however, it is unknown how well this model or the
linear low-dose extrapolation predicts low-dose risks for vanadium pentoxide. The selected
model, while the best fitting model, is not the only model that adequately describes the data.
Other models could conceivably be selected to yield different results consistent with the
observed data, both higher and lower than those included in this assessment. The human
equivalent inhalation unit risks estimated from the statistically significant increase in lung
tumors ranged from 1.4 mg/m3 in male mice to 4.2 mg/m3 in female mice. These tumors are
considered to be relevant to humans. As there is no information to inform which species or
gender of animals would be most applicable to humans, the most sensitive group was selected
for the basis of the inhalation unit risk.
139
DRAFT - DO NOT CITE OR QUOTE
-------
7. REFERENCES
ACGIH. (1991). Vanadium pentoxide. In:eds. Cincinnati, OH: American Conference of
Governmental Industrial Hygienists, 1677-1682.
Altamirano-Lozano, M; Alvarez-Barrera, L; Basurto-Alcantara, F; et al. (1996) Reprotoxic and
genotoxic studies of vanadium pentoxide in male mice. Teratogenesis, Carcinogenesis,
and Mutagenesis 16:7-17.
Altamirano-Lozano, M; Alvarez-Barrera, L; Roldan-Reyes, E. (1993) Cytogenetic and
teratogenic effects of vanadium pentoxide on mice. Med Sci Res 21:711-713.
Altamirano-Lozano, M; Roldan, E; Bonilla, E; et al. (1998) Effect of metal compounds on boar
sperm motility in vitro. Advances in Experimental Medicine and Biology 444:105-110;
discussion 110-111.
Altamirano-Lozano, M; Valverde, M; Alvarez-Barrera, L; et al. (1999) Genotoxic studies of
vanadium pentoxide (V(2)O(5)) in male mice. II. Effects in several mouse tissues.
Teratogenesis, Carcinogenesis, and Mutagenesis 19:243-255.
Altamirano, M; Ayala, ME; Flores, A; et al. (1991) Sex differences in the effects of vanadium
pentoxide administration to prepubertal rats. Med Sci Res 19:825-826.
Antao-Menezes, A; Turpin, EA; Bost, PC; et al. (2008) STAT-1 signaling in human lung
fibroblasts is induced by vanadium pentoxide through an IFN-beta autocrine loop.
Journal of Immunology 180:4200-4207.
Aragon, MA; Ayala, ME; Fortoul, TI; et al. (2005) Vanadium induced ultrastructural changes
and apoptosis in male germ cells. Reproductive Toxicology 20:127-134.
Assem, FL; Levy, LS. (2009) A review of current toxicological concerns on vanadium pentoxide
and other vanadium compounds: gaps in knowledge and directions for future research.
Journal of Toxicology and Environmental Health, PartB: Critical Reviews 12:289-306.
ATSDR (2009) Draft toxicological profile for vanadium. Atlanta, GA. U.S. Department of
Health and Human Services Public Health Service, Agency for Toxic Substances and
Disease Registry http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=276&tid=50
Avila-Costa, MR; Colin-Barenque, L; Zepeda-Rodriguez, A; et al. (2005) Ependymal epithelium
disruption after vanadium pentoxide inhalation A mice experimental model.
Neuroscience Letters 381:21-25.
Avila-Costa, MR; Fortoul, TI; Nino-Cabrera, G; et al. (2006) Hippocampal cell alterations
induced by the inhalation of vanadium pentoxide (V(2)O(5)) promote memory
140
DRAFT - DO NOT CITE OR QUOTE
-------
deterioration. NeuroToxicology 27:1007-1012.
Avila-Costa, MR; Montiel Flores, E; Colin-Barenque, L; et al. (2004) Nigrostriatal modifications
after vanadium inhalation: an immunocytochemical and cytological approach.
Neurochemical Research 29:1365-1369.
Barceloux, DG. (1999) Vanadium. Journal of Toxicology - Clinical Toxicology 37:265-278.
Bell, M; Ebisu, K; Peng, R; et al. (2009) Hospital Admissions and Chemical Composition of
Fine Particle Air Pollution. American Journal of Respiratory and Critical Care Medicine
179:1115-1120.
Bonner, JC; Lindroos, PM; Rice, AB; et al. (1998) Induction of PDGF receptor-alpha in rat
myofibroblasts during pulmonary fibrogenesis in vivo. American Journal of Physiology
274:L72-L80.
Bonner, JC; Rice, AB; Ingram, JL; et al. (2002) Susceptibility of cyclooxygenase-2-deficient
mice to pulmonary fibrogenesis. The American Journal of Pathology 161:459-470.
Bonner, JC; Rice, AB; Moomaw, CR; et al. (2000) Airway fibrosis in rats induced by vanadium
pentoxide. American Journal of Physiology: Lung Cellular and Molecular Physiology
278:L209-L216.
Byczkowski, JZ; Kulkarni, AP. (1992) Vanadium redox cycling, lipid peroxidation and co-
oxygenation of benzo(a)pyrene-7,8-dihydrodiol. Biochim Biophys Acta 1125:134-141.
Byrne, AR; Kosta, L. (1978) Vanadium in foods and in human body fluids and tissues. Science
of the Total Environment 10:17-30.
Carter, JD; Ohio, AJ; Samet, JM; et al. (1997) Cytokine production by human airway epithelial
cells after exposure to an air pollution particle is metal-dependent. Toxicology and
Applied Pharmacology 146:180-188.
Charles River (2008) Wistar Rat Hematology Available at:
http://www.criver.com/SiteCollectionDocuments/rm_rm_r_Wistar_rat_hematology_2008
.pdf
Colin-Barenque, L; Martinez-Hernandez, MG; Baiza-Gutman, LA; et al. (2008) Matrix
metalloproteinases 2 and 9 in central nervous system and their modification after
vanadium inhalation. Journal of Applied Toxicology 28:718-723.
Colin-Barenque, L; Martinez-Hernandez, MG; Baiza-Gutman, LA; et al. (2007) Matrix
metalloproteinases 2 and 9 in central nervous system and their modification after
vanadium inhalation. Journal of Applied Toxicology 28:718-723.
Conklin, AW; Skinner, SC; Felten, TL; et al. (1982) Clearance and distribution of intratracheally
141
DRAFT - DO NOT CITE OR QUOTE
-------
instilled vanadium compounds in the rat. Toxicol Lett 11:199-203.
Cooper, RG. (2007) Vanadium pentoxide inhalation. Indian J Occup Environ Med 11:97-102.
Dekker, JM; Schouten, EG; Klootwijk, P; et al. (1997) Heart rate variability from short
electrocardiographic recordings predicts mortality from all causes in middle-aged and
elderly men. The Zutphen Study. American Journal of Epidemiology 145:899-908.
Devereux, TR; Holliday, W; Anna, C; et al. (2002) Map kinase activation correlates with K-ras
mutation and loss of heterozygosity on chromosome 6 in alveolar bronchiolar carcinomas
from B6C3F1 mice exposed to vanadium pentoxide for 2 years. Carcinogenesis 23:1737-
1743.
Dill, JA; Lee, KM; Mellinger, KH; et al. (2004) Lung deposition and clearance of inhaled
vanadium pentoxide in chronically exposed F344 rats and B6C3F1 mice. Toxicological
Sciences 77:6-18.
Dominici, F; Peng, RD; Ebisu, K; et al. (2007) Does the effect of PM10 on mortality depend on
PM nickel and vanadium content? A reanalysis of the NMMAPS data. Environmental
Health Perspectives 115:1701-1703.
Duve, RN; Chandra, JP; Singh, SB. (1981) Open digestion modification of AOAC method for
determination of mercury in fish. Journal of the Association of Official Analytical
Chemists 64:1027-1029.
Edel, J; Pietra, R; Sabbioni, E; et al. (1984) Disposition of vanadium in rat tissues at different
age. Chemosphere 13:87-93.
Edel, J; Sabbioni, E. (1988) Retention of intratracheally instilled and ingested tetravalent and
pentavalent vanadium in the rat. J Trace Elem Electrolytes Health Dis 2:23-30.
Ehrlich, VA; Nersesyan, AK; Hoelzl, C; et al. (2008) Inhalative exposure to vanadium pentoxide
causes DNA damage in workers: Results of a multiple end point study. Environmental
Health Perspectives 116:1689-1693.
Fortoul, TI; Bizarro-Nevares, P; Acevedo-Nava, S; et al. (2007) Ultrastructural findings in
murine seminiferous tubules as a consequence of subchronic vanadium pentoxide
inhalation. Reproductive Toxicology 23:588-592.
Gonzalez-Villalva, A; Fortoul, TI; Avila-Costa, MR; et al. (2006) Thrombocytosis induced in
mice after subacute and subchronic V2O5 inhalation. Toxicology and Industrial Health
22:113-116.
Gylseth, B; Leira, HL; Steinnes, E; et al. (1979) Vanadium in the blood and urine of workers in a
ferroalloy plant. Scand J Work Environ Health 5:188-194.
142
DRAFT - DO NOT CITE OR QUOTE
-------
Hauser, R; Eisen, EA; Pothier, L; et al. (2001) A prospective study of lung function among
boilermaker construction workers exposed to combustion particulates. American Journal
of Industrial Medicine 39:454-462.
Hauser, R; Eisen, EA; Pothier, L; et al. (2002) Spirometric abnormalities associated with chronic
bronchitis, asthma, and airway hyperresponsiveness among boilermaker construction
workers. Chest 121:2052-2060.
Hauser, R; Elreedy, S; Hoppin, JA; et al. (1995a) Upper airway response in workers exposed to
fuel oil ash: nasal lavage analysis. Occupational and Environmental Medicine 52:353-
358.
Hauser, R; Elreedy, S; Hoppin, JA; et al. (1995b) Urinary vanadium as a biological indicator of
exposure to vanadium. Occupational and Environmental Medicine 52:3537358.
HSDB (2008) Vanadium pentoxide., In: National Library of Medicine,. Hazardous Substances
Databank, Available at: http://toxnet.nlm.nih.gov
IARC (2006) Cobalt in hard metals and cobalt sulfate, gallium arsenide, indium phosphide and
vanadium pentoxide, In: International Agency for Research on Cancer, pp. 1-294, Lyon,
France Available at:
Ingram, JL; Antao-Menezes, A; Turpin, EA; et al. (2007) Genomic analysis of human lung
fibroblasts exposed to vanadium pentoxide to identify candidate genes for occupational
bronchitis. Respiratory Research 8:34.
Ingram, JL; Rice, AB; Santos, J; et al. (2003) Vanadium-induced HB-EGF expression in human
lung fibroblasts is oxidant dependent;and requires MAP kinases. American Journal of
Physiology: Lung Cellular and Molecular Physiology 284:L774-L782.
Irsigler, GB; Visser, PJ; Spangenberg, PAL. (1999) Asthma and chemical bronchitis in
vanadium plant workers. American Journal of Industrial Medicine 35:366-374.
Ivancsits, S; Pilger, A; Diem, E; et al. (2002) Vanadate induces DNA strand breaks in cultured
human fibroblasts at doses relevant to occupational exposure. Mutation Research 519:25-
35.
Kada, T; Hirano, K; Shirasu, Y. (1980). Screening of environmental chemical mutagens by the
rec-assay system with Bacillus subtilis. In: de Serres, FJ.Hollaender, A eds. Chemical
Mutagens: Principles and Methods for their Detection New York: Plenum, 149-173.
Kanematsu, N; Kara, M; Kada, T. (1980) Rec assay and mutagenicity studies on metal
compounds. MutatRes 77:109-116.
143
DRAFT - DO NOT CITE OR QUOTE
-------
Kerckaert, GA; LeBoeuf, RA; Isfort, RJ. (1996) Use of the Syrian hamster embryo cell
transformation assay for determining the carcinogenic potential of heavy metal
compounds. Fundamental and Applied Toxicology 34:67-72.
Kim, JY; Mukherjee, S; Ngo, LC; et al. (2004) Urinary 8-hydroxy-2'-deoxyguanosine as a
biomarker of oxidative DNA damage in workers exposed to fine particulates.
Environmental Health Perspectives 112:666-671.
Kiviluoto, M. (1980) Observations on the lungs of vanadium workers. Occupational and
Environmental Medicine 37:363-366.
Kiviluoto, M; Pakarinen, A; Pyy, L. (1981b) Clinical laboratory results of vanadium-exposed
workers. Archives of Environmental and Occupational Health 36:109-113.
Kiviluoto, M; Pyy, L; Pakarinen, A. (1979) Serum and urinary vanadium of vanadium-exposed
workers. Scandinavian Journal of Work, Environment and Health 5:362-367.
Kiviluoto, M; Pyy, L; Pakarinen, A. (198 la) Serum and urinary vanadium of workers processing
vanadium pentoxide. International Archives of Occupational and Environmental Health
48:251-256.
Kleinsasser, N; Dirschedl, P; Staudenmaier, R; et al. (2003) Genotoxic effects of vanadium
pentoxide on human peripheral lymphocytes and mucosal cells of the upper aerodigestive
tract. International Journal of Environmental Health Research 13:373-379.
Knecht, EA; Moorman, WJ; Clark, JC; et al. (1992) Pulmonary reactivity to vanadium pentoxide
following subchronic inhalation exposure in a non-human primate animal model. Journal
of Applied Toxicology 12:427-434.
Knecht, EA; Moorman, WJ; Clark, JC; et al. (1985) Pulmonary effects of acute vanadium
pentoxide inhalation in monkeys. American Review of Respiratory Disease 132:1181-
1185.
Kucera, J; Byrne, AR; Mravcova, A; et al. (1992) Vanadium levels in hair and blood of normal
and exposed persons. Science of the Total Environment 115:191-205.
Kucera, J; Lener, J; Mnukova, J. (1994) Vanadium levels in urine and cystine levels in
fingernails and hair of exposed and normal persons. Biological Trace Element Research
43-45:327-334.
Kyono, H; Serita, F; Toya, T; et al. (1999) A new model rat with acute bronchiolitis and its
application to research on the toxicology of inhaled particulate matter. Industrial Health
37:47-54.
Lagorio, S; Forastiere, F; Pistelli, R; et al. (2006) Air pollution and lung function among
susceptible adult subjects: a panel study. Environmental Health 5:11.
144
DRAFT - DO NOT CITE OR QUOTE
-------
Lees, RE. (1980) Changes in lung function after exposure to vanadium compounds in fuel oil
ash. British Journal of Industrial Medicine 37:253-256.
Levy, BS; Hoffman, L; Gottsegen, S. (1984) Boilermakers bronchitis: respiratory tract irritation
associated with vanadium pentoxide exposure during oil-to-coal conversion of a power
plant. Journal of Occupational and Environmental Medicine 26:567-570.
Lewis, CE. (1959) The biological effects of vanadium. II. The signs and symptoms of
occupational vanadium exposure. AMA Arch Ind Health 19:497-503.
Lewis, RJ, Sr ed. 1997. Hawley's Condensed Chemical Dictionary. New York, NY: John Wiley
& Sons, Inc., p. 1163.
Lippmann, M; Ito, K; Hwang, JS; et al. (2006) Cardiovascular effects of nickel in ambient air.
Environmental Health Perspectives 114:1662-1669.
Magari, SR; Schwartz, J; Williams, PL; et al. (2002) The association of paniculate air metal
concentrations with heart rate variability. Environmental Health Perspectives 110:875-
880.
Motolese, A; Truzzi, M; Giannini, A; et al. (1993) Contact dermatitis and contact sensitization
among enamellers and decorators in the ceramics industry. Contact Dermatitis 28:59-62.
Mountain, JT; Delker, LL; Stokinger, HE. (1953) Studies in vanadium toxicology; reduction in
the cystine content of rat hair. AMA Arch Ind Hyg Occup Med 8:406-411.
Mravcova, A; Jirova, D; Janci, H; et al. (1993) Effects of orally administered vanadium on the
immune system and bone metabolism in experimental animals. Science of the Total
Environment Suppl Pt 1:663-669.
Mukherjee, B; Patra, B; Mahapatra, S; et al. (2004) Vanadium—an element of atypical biological
significance. Toxicology Letters 150:135-143.
Musk, AW; Tees, JG. (1982) Asthma caused by occupational exposure to vanadium compounds.
Medical Journal of Australia 1:183-184.
Mussali-Galante, P; Rodriguez-Lara, V; Hernandez-Tellez, B; et al. (2005) Inhaled vanadium
pentoxide decrease gamma-tubulin of mouse testes at different exposure times.
Toxicology and Industrial Health 21:215-222.
NIOSH (1983) Health hazard evaluation report HETA 80-096-1359, Eureka Company,
Bloomington, IL. Washington, DC: U.S. Department of Health and Human Services,
National Institute of Occupational Safety and Health, PB85163574.
145
DRAFT - DO NOT CITE OR QUOTE
-------
NIOSH (2005) Vanadium dust, In: NIOSH Pocket Guide to Chemical Hazard (DHHS (NIOSH)
Publication No. 2005-149), National Institute for Occupational Safety and Health,
Department of Health and Human Services, pp.328, Cincinnati, OH Available at:
http://www.cdc.gov/niosh/npg/npgd0653.html
NRC (National Research Council). (1983) Risk assessment in the federal government: Managing
the process. Washington, DC. National Academies Press Available from:
http://www.nap.edu/openbook.php?isbn=0309033497
NTP (2002) NTP toxicology and carcinogensis studies of vanadium pentoxide (CAS No.; 1314-
62-1) in F344/Nrats andB6C3Fl mice (inhalation), In: National Toxicology Program.
pp. 1-343, Washington, DC Available at:
http://ntp.niehs.nih.gov/ntp/htdocs/LT rpts/tr507.pdf
NTP (2008) Chemical Information Review Document for Oral Exposure to Tetravalent and
Pentavalent Vanadium Compounds, In: National Toxicology Program, National Institute
of Environmental Health Sciences, National Institutes of Health, U.S. Department of
Health and Human Services, pp.p. 5, Research Triangle Park, NC Available at: http://ntp-
server.niehs.nih.gov/ntp/htdocs/Chem_Background/ExSumPdf/NIEHS_Vanadium_comp
ounds_508.pdf
Oberg, SG; Parker, RD; Sharma, RP. (1978) Distribution and elimination of an intratracheally
administered vanadium compound in the rat. Toxicology 11:315-323.
O'Neal, SG; Rhoads, DB; Racker, E. (1979) Vanadate inhibition of sarcoplasmic reticulum
Ca2+-ATPase and other ATPases. Biochem Biophys Res Commun 89:845-850.
OSHA (2007) Occupational Safety and Health Guideline for Vanadium Pentoxide Dust, In:
Occupational Safety and Health Administration. Washington, DC. U.S.Department of
Labor. Available at:
http://www.osha.gov/SLTC/healthguidelines/vanadiumpentoxidedust/recognition.html
Parker, RD; Sharma, RP. (1978) Accumulation and depletion of vanadium in selected tissues of
rats treated with vanadyl sulfate and sodium orthovanadate. J Environ Pathol Toxicol
2:235-245.
Patel, MM; Hoepner, L; Garfinkel, R; et al. (2009) Ambient metals, elemental carbon, and
wheeze and cough in New York City children through 24 months of age. American
Journal of Respiratory and Critical Care Medicine 180:1107-1113.
Pazynich, VM. (1966) Maximum permissible concentration of vanadium pentoxide in the
atmosphere. Gigiena i Sanitariia 317-9:6-12.
Pierce, LM; Alessandrini, F; Godleski, JJ; et al. (1996) Vanadium-induced chemokine mRNA
146
DRAFT - DO NOT CITE OR QUOTE
-------
expression and pulmonary inflammation. Toxicology and Applied Pharmacology 138:1-
11.
Pinon-Zarate, G; Rodriguez-Lara, V; Rojas-Lemus, M; et al. (2008) Vanadium pentoxide
inhalation provokes germinal center hyperplasia and suppressed humoral immune
responses. Journal of Immunotoxicology 5:115-122.
Radike, M; Warshawsky, D; Caruso, J; et al. (2002) Distribution and accumulation of a mixture
of arsenic, cadmium, chromium, nickel, and vanadium in mouse small intestine, kidneys,
pancreas, and femur following oral administration in water or feed. Journal of
Toxicology and Environmental Health, Part A: Current Issues 65:2029-2052.
Ramirez, P; Eastmond, DA; Laclette, JP; et al. (1997) Disruption of microtubule assembly and
spindle formation as a mechanism for the;induction of aneuploid cells by sodium arsenite
and vanadium pentoxide. Mutation Research 386:291-298.
Ress, NB; Chou, BJ; Renne, RA; et al. (2003) Carcinogenicity of inhaled vanadium pentoxide in
F344/N rats and B6C3F1 mice. Toxicological Sciences 74:287-296.
Rhoads, K; Sanders, CL. (1985) Lung clearance, translocation, and acute toxicity of arsenic,
beryllium, cadmium, cobalt, lead, selenium, vanadium, and ytterbium oxides following
deposition in rat lung. Environ Res 36:359-378.
Rice, AB; Moomaw, CR; Morgan, DL; et al. (1999) Specific inhibitors of platelet-derived
growth factor or epidermal growth factor receptor tyrosine kinase reduce pulmonary
fibrosis in rats. Am J Pathol 155:213-221.
Rojas, E; Valverde, M; Herrera, LA; et al. (1996) Genotoxicity of vanadium pentoxide evaluate
by the single cell gel electrophoresis assay in human lymphocytes. Mutation Research
359:77-84.
Roldan, RE; Altamirano, LM. (1990) Chromosomal aberrations, sister-chromatid exchanges,
cell-cycle kinetics and satellite associations in human lymphocyte cultures exposed to
vanadium pentoxide. Mutation Research 245:61-65.
Rondini, EA; Walters, DM; Bauer, AK. (2010) Vanadium pentoxide induces pulmonary
inflammation and tumor promotion in a strain-dependent manner. Particle and Fibre
Toxicology 7:9.
Roshchin, AV; Ordzhonikidze, EK; Shalganova, IV. (1980) Vanadium-toxicity, metabolism,
carrier state. J Hyg Epidemiol Microbiol Immunol 24:377-383.
Ross, DS. (1983) Case study: exposure to vanadium pentoxide. Occupational health; a journal
for occupational health nurses 35:67-71.
Sabbioni, E; Kueera, J; Pietra, R; et al. (1996) A critical review on normal concentrations of
147
DRAFT - DO NOT CITE OR QUOTE
-------
vanadium in human blood, serum, and urine. Science of the Total Environment 188:49-
58.
Schroeder, HA; Balassa, JJ; Tipton, IH. (1963) Abnormal Trace Metals in Man—Vanadium. J
Chronic Dis 16:1047-1071.
Sharma, RP; Oberg, SG; Parker, RD. (1980) Vanadium retention in rat tissues following acute
exposures to different dose levels. J Toxicol Environ Health 6:45-54.
Sjoberg, SG. (1951) Health hazards in the production and handling of vanadium pentoxide.
AMA Archives of Industrial Hygiene and Occupational Medicine 3:631-646.
Sjoberg, SG. (1955) Vanadium bronchitis from cleaning oil-fired boilers. AMA Archives of
Industrial Health 11:505-512.
Sjoberg, SG. (1956) Vanadium dust, chronic bronchitis and possible risk of emphysema: a
follow-up investigation of workers at a vanadium factory. Journal of Internal Medicine
154:381-386.
Sorensen, M; Schins, RPF; Hertel, O; et al. (2005) Transition metals in personal samples of
PM25 and oxidative stress in human volunteers. Cancer Epidemiology Biomarkers and
Prevention 14:1340-1343.
Stokinger, HE; Wagner, WD; Mountain, JT; et al. (1953) Unpublished results. Division of
Occupational Health, Cincinnati, OH. (Cited in Patty's Industrial Hygiene and
Toxicology, 3rd ed., 1981).
Turpin, EA; Antao-Menezes, A; Cesta, MF; et al. (2010) Respiratory syncytial virus infection
reduces lung inflammation and fibrosis in mice exposed to vanadium pentoxide.
Respiratory Research 11:20.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment
of chemical mixtures. Federal Register 51(185):34014-34025. Available from:
.
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register
51(185):34006-34012. Available from: .
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk
assessment. Prepared by the Environmental Criteria and Assessment Office, Office of
Health and Environmental Assessment, Cincinnati, OH for the Office of Solid Waste and
Emergency Response, Washington, DC; EPA 600/6-87/008. Available from:
.
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register
56(234):63798-63826. Available from: .
148
DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation
toxicity studies. Federal Register 59(206):53799. Available from:
.
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application
of inhalation dosimetry. Office of Research and Development, Washington, DC;
EPA/600/8-90/066F. Available from: .
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:
.
U.S. EPA. (1996a) Guidelines for reproductive toxicity risk assessment. Federal Register
61(212):56274-56322. Available from: .
U.S. EPA. (1987) Vanadium Pentoxide. IRIS Assessment. Available from:
http://www.epa.gov/iris/subst/0125.htm.
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-
26954. Available from: .
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 from: .
U.S. EPA. (2000b) Benchmark dose technical guidance document [external review draft]. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/001. Available from:
.
U.S. EPA. (2000c) Supplementary guidance for conducting for health risk assessment of
chemical mixtures. Risk Assessment Forum, Washington, DC; EPA/630/R-00/002.
Available from: .
U.S. EPA. (2002) A review of the reference dose and reference concentration processes. Risk
Assessment Forum, Washington, DC; EPA/630/P-02/0002F. Available from:
.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001B. Available from:
.
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 from: .
149
DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA. (2006a) Science policy council handbook: peer review. Third edition. Office of
Science Policy, Office of Research and Development, Washington, DC; EPA/100/B-
06/002. Available from: .
U.S. EPA. (2006b) A Framework for Assessing Health Risk of Environmental Exposures to
Children. National Center for Environmental Assessment, Washington, DC, EPA/600/R-
05/093F. Available from:
.
U.S. EPA. (2011) Recommended Use of Body Weight34 at the Default Method in Derivation of
the Oral Reference Dose. Risk Assessment Forum, Office of the Science Advisor,
Washington, DC, EPA/100/R11/0001. Available from:
http://www.epa.gov/raf/publications/pdfs/recommended-use-of-bw34.pdf
Vintinner, FJ; Vallenas, R; Carlin, CE; et al. (1955) Study of the health of workers employed in
mining and processing of vanadium ore. AMA Arch Ind Health 12:635-642.
Wallenborn, JG; McKee, JK; Schladweiler, MC; et al. (2007) Systemic translocation of
particulate matter-associated metals following a single intratracheal instillation in rats.
Journal of Toxicological Sciences 98:231-239.
WHO-IPCS (2001) Vanadium pentoxide and other inorganic vanadium compounds, In: Concise
International Chemical Assessment Document 29, World Health Organization, Geneva,
Switzerland Available at: http://www.inchem.org/documents/cicads/cicads/cicad29.htm
WHO (1988) Vanadium, In: WHO Regional Publ, pp.361-371, Copenhagen Available at:
Wide, M. (1984) Effect of short-term exposure to five industrial metals on the embryonic and
fetal development of the mouse. Environmental Research 33:47-53.
Williams, N. (1952) Vanadium poisoning from cleaning oil-fired boilers. British Journal of
Industrial Medicine 9:50-55.
Woodin, MA; Hauser, R; Liu, Y; et al. (1998) Molecular markers of acute upper airway
inflammation in workers exposed to fuel-oil ash. American Journal of Respiratory and
Critical Care Medicine 158:182-187.
Woodin, MA; Liu, Y; Hauser, R; et al. (1999) Pulmonary function in workers exposed to low
levels of fuel-oil ash. Journal of Occupational and Environmental Medicine 41:973-980.
Woodin, MA; Liu, Y; Neuberg, D; et al. (2000) Acute respiratory symptoms in workers exposed
to vanadium-rich fuel-oil ash. American Journal of Industrial Medicine 37:353-363.
Wright, JR; Yates, AJ; Shah, NT; Neff JC; et al. (2009) Hematological characteristics of the BB
Wistar rat. Veterinary Clinical Pathology 12:9-13.
150
DRAFT - DO NOT CITE OR QUOTE
-------
Yamaguchi, M; Oishi, H; Suketa, Y. (1989) Effect of vanadium on bone metabolism in weanling
rats: zinc prevents the toxic effect of vanadium. Research in Experimental Medicine
189:47-53.
Zenz, C; Bartlett, JP; Thiede, WH. (1962) Acute vanadium pentoxide intoxication. Archives of
Environmental and Occupational Health 5:542-546.
Zenz, C; Berg, BA. (1967) Human responses to controlled vanadium pentoxide exposure.
Archives of Environmental and Occupational Health 14:709-712.
Zhang, L; Rice, AB; Adler, K; et al. (2001a) Vanadium stimulates human bronchial epithelial
cells to produce heparin-binding epidermal growth factor-like growth factor: a mitogen
for lung fibroblasts. American Journal of Respiratory Cell and Molecular Biology
24:123-131.
Zhang, T; Gou, X; Yang, Z. (1991a) A study on developmental toxicity of vanadium pentoxide
in Nffl mice. Hua Xi Yi Ke Da Xue Xue Bao 22:192-195.
Zhang, T; Gou, X; Yang, Z. (1993b) A study of teratogenicity and sensitive period of vanadium
pentoxide in Wistar rats. Hua Xi Yi Ke Da Xue Xue Bao 24:202-205.
Zhang, T; Yang, Z; Li, S; et al. (1991b) Transplacental passage of vanadium after treatment with
vanadium pentoxide in Wistar rat. Hua Xi Yi Ke Da Xue Xue Bao 22:296-299.
Zhang, T; Yang, Z; Zeng, C; et al. (1993a) A study on developmental toxicity of vanadium
pentoxide in Wistar rats. Hua Xi Yi Ke Da Xue Xue Bao 24:92-96.
Zhang, Z; Wang, Y; Vikis, HG; et al. (2001b) Wildtype Kras2 can inhibit lung carcinogenesis in
mice. Nature Genetics 29:25-33.
Zhong, BZ; Gu, ZW; Wallace, WE; et al. (1994) Genotoxicity of vanadium pentoxide in Chinese
hamster V79 cells. Mutation Research 321:35-42.
151
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX A - SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION
This page intentionally left blank.
152
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX B - BENCHMARK CONCENTRATION MODELING OF INHALATION
STUDIES IN RATS FROM NTP, 2002
To derive the RfC for vanadium pentoxide, inhalation toxicity effects observed in rats
(NTP, 2002) were modeled to estimate the candidate PODs. Five endpoints in both male and
female rats (Table B-2) were selected because of two reasons:
• The study in rats was designed with lower concentrations of vanadium pentoxide;
• The results showed both biological significance and statistically significant
trends.
Each data set was fitted with the dichotomous models available in EPA BMDS (version
2.1.2). Following the model selection steps outlined in the draft Benchmark Dose Technical
Guidance (US EPA, 2000b), the best-fitting model for each data set was used to estimate the
candidate POD, which was the BMCL at the selected BMR as 10% extra risk.
The highlights of the benchmark dose modeling results are:
• The POD value of 0.003 mg/m3, from two endpoints in female rats, was the lowest
POD derived and used for the calculation of RfC. The two endpoints were larynx
chronic inflammation and larynx respiratory epithelium epiglottis hyperplasia.
• All candidate PODs and the selected models are summarized in Table B-l.
153
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-l: Candidate PODs for Vanadium Pentoxide Derived from NTP Studies (2002) through
BMDS Modeling.
Endpoint
Selected
Model a
BM
R
(Extr
a
Risk)
HECb
BMC
(mg/m3
)
BMCLC
(mg/m3)
(Candidate
POD)
Male F344/N Rats
Lung
Alveolar Epithelium Hyperplasia
Chronic Active Inflammation
Probit
Logistic
0.1
0.1
0.016
0.035
0.013
0.029
Larynx
Chronic Inflammation
Respiratory Epithelium, Epiglottis,
Hyperplasia
LogLogistic
LogLogistic
0.1
0.1
0.017
0.008
0.012
0.006
Nose
Goblet Cell, Respiratory Epithelium,
Hyperplasia
LogLogistic
0.1
0.044
0.026
Female F344/N Rats
Lung
Alveolar Epithelium Hyperplasia
Chronic Active Inflammation
Gamma
Multistage(Stag
e3)
0.1
0.1
0.076
0.080
0.063
0.048
Larynx
Chronic Inflammation
Respiratory Epithelium, Epiglottis,
Hyperplasia
LogLogistic
LogLogistic
0.1
0.1
0.005
0.004
0.003
0.003
Nose
Goblet Cell, Respiratory Epithelium,
Hyperplasia
Multistage(Stag
e2)
0.1
0.038
0.014
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose
Technical Guidance (2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
B.I. TREND TESTS ON THE INHALATION DATA SETS FOR BMDS MODELING
Ten inhalation toxicology data sets in rats were selected because of the lower
concentrations of vanadium pentoxide administrated (as compared to mice), the significantly
biologically adverse effects observed and the statistical significance reported by NTP (2002). A
Cochran-Armitage test was conducted to confirm the significance of the statistical trend for the
selected data sets (Table B-2) before modeling.
154
DRAFT - DO NOT CITE OR QUOTE
-------
155
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-2: Trend Tests on the Selected Data Sets from the 2-Year Inhalation Studies in Rats (NTP,
2002)
Endpoint
Concentration as Reported
(mg/mV
0.0
0.5
1.0
2.0
Trend Test b
Z -Score
/7-value
Incidence0 in Male F344/N Rats
Lung
Alveolar Epithelium Hyperplasia d
Chronic Active Inflammation
Larynx
Chronic Inflammation
Respiratory Epithelium, Epiglottis
Hyperplasia
Nose
Goblet Cell, Respiratory Epithelium,
Hyperplasia
7/50
5/50
3/49
0/49
4/49
24/49
8/49
20/50
18/50
15/50
35/48
24/48
17/50
34/50
12/49
50/50
42/50
28/49
32/49
17/48
8.83
8.36
4.79
-6.56
2.68
O.OOOl
O.OOOl
O.OOOl
O.OOOl
0.0037
Incidence c in Female F344/N Rats
Lung
Alveolar Epithelium Hyperplasia d
Chronic Active Inflammation
Larynx
Chronic Inflammation
Respiratory Epithelium, Epiglottis
Hyperplasia
Nose
Goblet Cell, Respiratory Epithelium,
Hyperplasia
7/49
10/49
8/50
0/50
13/50
8/49
10/49
26/49
25/49
19/50
21/50
14/50
27/49
26/49
16/50
50/50
40/50
37/50
33/50
30/50
9.53
6.71
5.38
-5.98
3.46
O.OOOl
O.OOOl
O.OOOl
O.OOOl
0.0003
a Concentrations are as reported by NTP (2002).
b One-sided Cochran-Armitage trend test.
c Incidence=(number of animals affected)/(number of animals examined).
d This data set was calculated by combining the incidences of alveolar epithelium hyperplasia
and bronchiole epithelium hyperplasia.
B.2. DOSE CONVERSION
In the toxicology and carcinogenesis studies reported by NTP (2002), rats (F344/N) were
exposed to vanadium pentoxide through inhalation. To analyze the concentration response effect
of vanadium pentoxide, the reported concentrations of vanadium pentoxide were converted to
human equivalent concentrations before any modeling and extrapolation.
First, the average life time animal body weights of rats were estimated based on the mean
body weight at different weeks (Table B-3). Secondly, following the Methods for Derivation of
156
DRAFT - DO NOT CITE OR QUOTE
-------
Inhalation Reference Concentrations and Application of Inhalation Dosimetry (US EPA, 1994b),
RDDRs (regional deposited dose ratios) were calculated with the RDDR program designed by
US EPA (1994). Although six regional RDDRs were reported by the RDDR program for each
concentration/sex group, only two regional RDDRs (i.e. Extrathoracic and Pulmonary) are
relevant to the selected endpoints, which are summarized in Table B-4. Then, the human
equivalent concentration for each concentration/sex group was calculated from the reported
concentration in rats by multiplying the continuous exposure adjustment factor and RDDR
(Table B-5). For endpoints in the lung, the pulmonary RDDR was applied; for endpoints in the
larynx and nose, the extrathoracic RDDR was applied.
Table B-3: Average Life Time Animal Body Weight of Rats in the 2-Year Inhalation Studies of
Vanadium Pentoxide (NTP, 2002)
Vanadium Pentoxide
Concentration as
Reported (mg/m3)
Mean Animal Body Weight a (g)
1-13
weeks
14-52
weeks
53-104
weeks
Average Life
Time Animal
Body Weight b
(g)
Male Rats (F344/N)
0
0.5
1
2
238
241
242
233
411
422
414
404
504
513
508
494
440
449
443
432
Female Rats (F344/N)
0
0.5
1
2
151
150
150
147
227
229
227
217
326
319
326
308
269
266
269
256
a As reported in Table 11-12 of the NTP report (2002).
b "Average Life Time Animal Body Weight" = ( "Mean Body Weight 1-13 Weeks" * 13 +"
Mean Body Weight 14 -52 Weeks" * 39 + "Mean Body Weight 53-104 Weeks" * 52) 7104
Table B-4: RDDRs for Different Concentration/Sex Group in the 2-Year Inhalation Studies of
Vanadium Pentoxide (NTP, 2002)
Vanadium
Pentoxide
Concentration as
Reported (mg/m3)
Average Life
Time Animal
Body Weight a
(S)
Average
MMAD
b
Average
GSDC
RDDRd
Extrathoraci
c
Pulmonary
Male Rats (F344/N)
0
0.5
1
440
449
443
1.24
1.89
0.516
0.530
0.520
0.496
0.494
0.495
157
DRAFT - DO NOT CITE OR QUOTE
-------
2
432
0.503
0.498
Female Rats (F344/N)
0
0.5
1
2
269
266
269
256
1.24
1.89
0.263
0.259
0.263
0.245
0.524
0.524
0.524
0.524
a All average life time animal body weights were calculated in Table B-3.
b MMAD =mass median aerodynamic diameter; calculated based on the MMADs reported for 2-year studies (NTP,
2002) .
0 GSD = geometric standard deviation; calculated based on the GDSs reported for 2-year studies (NTP, 2002).
d RDDR = regional deposited dose ratio; calculated with the RDDR program (V.2.3. US EPA);
Table B-5: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year Inhalation
Studies (NTP, 2002)
Concentration
as Reported a
(mg/m3)
Continuous
Exposure
Adjustment
Factor b
RDDRC
Extrathoraci
c
Pulmonary
Human Equivalent
Concentration d (mg/m3)
Extrathoraci
c
Pulmonary
Male Rats (F344/N)
0
0.5
1
2
0.179
0.179
0.179
0.179
0.516
0.530
0.520
0.503
0.496
0.494
0.495
0.498
0.00
0.05
0.09
0.18
0.00
0.04
0.09
0.18
Female Rats (F344/N)
0
0.5
1
2
0.179
0.179
0.179
0.179
0.263
0.259
0.263
0.245
0.524
0.524
0.524
0.524
0.00
0.02
0.05
0.09
0.00
0.05
0.09
0.19
a "Toxicology and carcinogenesis studies of vanadium pentoxide in F344/N rats and B6C3F! Mice", NTP, 2002.
b "Continuous Exposure Adjustment Factor" = (6/24) * (5/7); animals were exposed to vanadium pentoxide 6 hours
per day and 5 days per week.
0 Please refer to Appendix Table B-4.
d "Human Equivalent Concentration" = "Concentration as Reported" * "Continuous Exposure Adjustment Factor "
* "RDDR"
B.3. BMDS MODELING FOR INHALATION DATA SETS
Each selected inhalation data set was fitted with all the standard dichotomous models
available in EPA BMDS (version 2.1.2.). Following the model selection steps outlined in the
Benchmark Dose Technical Guidance (US EPA, 2000b), the best-fitting model was used to
estimate the candidate POD for each endpoint, which was the BMCL at the selected BMR as
10% extra risk.
158
DRAFT - DO NOT CITE OR QUOTE
-------
The selected models and candidate PODs for all endpoints are summarized in Table B-l.
Data Set 1: Incidence of Lung Alveolar Epithelium Hyperplasia in Male Rats, NTP (2002)
Summary
All three concentration groups (0.5, 1 and 2 mg/m3) showed statistically significant
differences from the control as reported. The severity of these nonneoplastic lesions increased
from mild to moderate as reported when the concentration of vanadium pentoxide increased.
The Cochran-Armitage test confirmed the statistically significant trend with Z score as 8.83 and
one-sided/>-value as <0.0001.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the combined incidence was used as the BMR to determine the
POD. The goodness of fit, BMC and BMCL for each model are summarized in Table B-6.
Five models demonstrated adequate goodness of fit/?-value ^0.1 and good visual fit.
Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000b), the Probit model
was selected as the model for POD computation. The BMCLio of this model was 0.013 mg/m3
and regarded as one of the candidate PODs.
Table B-6: Benchmark Modeling Results for Incidence of Lung Alveolar Epithelium Hyperplasia in
Male Rats, NTP (2002)
Model a
Probit
Logistic
Multistage e
(Stage3)
Weibull
Gamma
LogProbit
LogLogisti
c
Goodness of Fit
P-
Value
0.53
0.38
0.43
0.18
0.13
0.06
0.04
AICb
170.0
171.0
171.3
172.7
173.6
175.1
176.4
Largest
Scaled
Residual
-0.70
-0.87
-0.50
-0.90
-1.05
-1.31
-1.38
Scaled
Residual of
Interest
-0.23
-0.20
-0.08
-0.15
-0.10
0.67
0.67
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
)
0.016
0.016
0.011
0.019
0.021
0.026
0.025
BMCLd
(mg/m3
)
0.013
0.013
0.007
0.010
0.009
0.016
0.015
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
159
DRAFT - DO NOT CITE OR QUOTE
-------
Probit Model. (Version: 3.2; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLungAlBrEpHyperplasia\NTP_2002_Lung
Alveolar Bronchiole Combined Epithelium Hyperplasia_Probit_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLungAlBrEpHyperplasia\NTP_2002_Lung
Alveolar Bronchiole Combined Epithelium Hyperplasia_Probit_0.1.pit
The form of the probability function is:
P[response] = CumNorm(Intercept+Slope*Dose),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Response
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
background = 0 Specified
intercept = -1.04337
slope = 20.1821
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.78
slope -0.78 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
intercept -1.02981 0.180284 -1.38316 -0.676464
slope 20.0578 2.82623 14.5184 25.5971
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -82.2383 4
Fitted model -82.9994 2 1.52225 2 0.4671
Reduced model -133.424 1 102.372 3 <.0001
AIC: 169.999
Dose
0.0000
0.0441
0.0884
0.1779
Est. Prob.
0.1515
0.4423
0.7713
0.9944
Goodness of Fit
Expected Observed Size
7.577
21.673
37.023
49.721
7.000
24 .000
35.000
50.000
50
49
48
50
Scaled
Residual
-0.228
0.669
-0.695
0.530
Chi^2 = 1.26 d.f. = 2 P-value = 0.5316
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.0155483
BMDL = 0.0129496
160
DRAFT - DO NOT CITE OR QUOTE
-------
Probit Model with 0.95 Confidence Level
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
11:0009/102010
161
DRAFT - DO NOT CITE OR QUOTE
-------
Data Set 2: Incidence of Lung Chronic Active Inflammation in Male Rats, NTP (2002)
Summary
Two concentration groups (1 and 2 mg/m3) showed statistically significant difference
from the control group as reported. The severity increased from minimal to mild as reported
when the concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed
the statistically significant trend with Z score as 8.36 and one-sided/>-value as <0.0001.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-7.
All models demonstrated adequate goodness of fit/?-value ^0.1 and good visual fit.
Based on the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the Logistic model
was selected as the model for POD computation. The BMCLio of this model was 0.029 mg/m3
and regarded as one of the candidate PODs.
Table B-7: Benchmark Modeling Results for Incidence of Lung Chronic Active Inflammation in
Male Rats, NTP (2002)
Model a
Logistic
Probit
LogProbit
LogLogist
ic
Gamma
Weibull
Multistage
e (Stage2)
Goodness of Fit
P-
Value
0.40
0.37
0.67
0.60
0.40
0.27
0.24
AIC
b
192.
5
192.
7
192.
8
192.
9
193.
3
193.
9
194.
0
Largest
Scaled
Residual
-0.92
-1.00
0.29
-0.34
0.58
-0.75
0.88
Scaled
Residual of
Interest
-0.92
-1.00
-0.26
-0.34
-0.52
-0.75
-0.66
BM
R
Extr
a
Risk
10%
HECC
BMC
(mg/m3)
0.035
0.032
0.046
0.045
0.042
0.038
0.040
BMCLd
(mg/m3)
0.029
0.027
0.032
0.031
0.027
0.024
0.019
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
162
DRAFT - DO NOT CITE OR QUOTE
-------
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLungChronicActiveInflammation\NTP_20
02_Lung Chronic Active Inflammation_Logistic_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLungChronicActiveInflammation\NTP_20
02_Lung Chronic Active Inflammation_Logistic_0.1.pit
The form of the probability function is:
P[response] = I/[1+EXP(-intercept-slope*dose)]
Dependent variable = Response
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
background = 0 Specified
intercept = -2.2183
slope = 21.8588
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept
slope
intercept
1
-0.82
slope
-0.82
1
Variable
intercept
slope
Model
Full model
Fitted model
Reduced model
AIC:
Parameter Estimates
Estimate Std. Err.
-2.29582 0.320666
23.0148 3.22048
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-2.92431 -1.66732
16.7027 29.3268
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f. P-value
-93.3159 4
-94.2613 2 1.8908 2 0.3885
-132.664 1 78.6961 3 <.0001
192.523
Dose
0.0000
0.0441
0.0884
0.1779
Est. Prob.
0.0915
0.2174
0.4350
0.8578
Goodness of Fit
Expected Observed
4 .574
10.654
20.880
42 .892
5.000
8 .000
24 .000
42 .000
Size
50
49
48
50
Scaled
Residual
0.209
-0.919
0.908
-0.361
Chi 2 =1.84
d.f. = 2
P-value = 0.3976
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.0345486
BMDL = 0.0285337
163
DRAFT - DO NOT CITE OR QUOTE
-------
Logistic Model with 0.95 Confidence Level
Logistic
BMDLl BMD
0.08 0.1
dose
11:2409/102010
Data Set 3: Incidence of Larynx Chronic Inflammation in Male Rats, NTP (2002)
Summary
Three concentration groups (0.5, 1 and 2 mg/m3) showed statistically significant
difference from the control group as reported. The severity increased from minimal to about mild
as reported when the concentration of vanadium pentoxide increased. The Cochran-Armitage test
confirmed the statistically significant trend with Z score as 4.79 and one-sided/>-value as
O.OOOl.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-8.
Only one model demonstrated adequate goodness of fit^-value ^0.1 and good visual fit.
Based on the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the LogLogistic
model was selected as the model for POD computation. The BMCLio of this model was 0.012
mg/m3 and regarded as one of the candidate PODs.
Table B-8: Benchmark Modeling Results for Incidence of Larynx Chronic Inflammation in Male
Rats, NTP (2002)
Model a
LogLogistic
LogProbit
Multistage e
(Stage 1)
Goodness of Fit
P-
Value
0.12
0.09
0.05
AIC
b
229.1
229.8
230.7
Largest
Scaled
Residual
1.63
-1.37
2.14
Scaled
Residual of
Interest
-0.27
-0.01
-0.65
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
)
0.017
0.005
0.023
BMCLd
(mg/m3
)
0.012
0.000
0.017
164
DRAFT - DO NOT CITE OR QUOTE
-------
Gamma
Weibull
Probit
Logistic
0.01
0.01
234.8
235.2
2.45
2.43
2.45
2.43
0.043
0.046
0.036
0.038
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLarynxChronicInflammation\NTP_2002_L
arynx Chronic Inflammation_LogLogistic_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLarynxChronicInflammation\NTP_2002_L
arynx Chronic Inflammation_LogLogistic_0.1.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Response
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0612245
intercept = 1.9744
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background
intercept
background
1
-0.53
intercept
-0.53
1
Parameter Estimates
Variable Estimate Std. Err.
background 0.0712314 * *
intercept 1.9037 * *
slope 1 * *
- Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f.
Full model -110.451 4
Fitted model -112.55 2 4.19924 2
Reduced model -127.371 1 33.8403 3
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
P-value
0.1225
<.0001
AIC:
229.101
Goodness of Fit
Dose
0.0000
0.0473
0.0929
Est. Prob.
0.0712
0.2951
0.4278
Expected
3 .490
14 .754
21.390
Observed
3 .000
20.000
17.000
Size
49
50
50
Scaled
Residual
-0.272
1.627
-1.255
165
DRAFT - DO NOT CITE OR QUOTE
-------
0.1796
^2 =4.31
0.5789
d.f. = 2
28.366 28.000
P-value = 0.1162
49
-0.106
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.0165573
BMDL = 0.0117279
Log-Logistic Model with 0.95 Confidence Level
13:2209/102010
Data Set 4: Incidence of Larynx Respiratory Epithelium, Epiglottis, Hyperplasia in Male
Rats, NTP (2002)
Summary
Three concentration groups (0.5, 1 and 2 mg/m3) showed statistically significant
difference from the control group as reported. The severity increased slightly when the
concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed the
statistically significant trend with Z score as -6.56 and one-sided/?-value as <0.0001.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-9.
Both LogLogistic and LogProbit models demonstrated adequate goodness of flip-value ^
0.1 and good visual fit, but the BMC/BMCL ratio in LogProbit was infinite, which was not
acceptable. Based on the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the
LogLogistic model was selected as the model for POD computation. The BMCLio of this model
was 0.006 mg/m3 and regarded as one of the candidate PODs.
166
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-9: Benchmark Modeling Results for Incidence of Larynx Respiratory Epithelium,
Epiglottis, Hyperplasia in Male Rats, NTP (2002)
Model a
LogLogist
ic
LogProbit
Gamma
Multistage
e
(Stagel)
Weibull
Probit
Logistic
Goodness of Fit
P-
Value
0.27
0.14
0.04
0.00
0.00
AIC
b
197.
3
199.
3
201.
5
226.
4
227.
1
Largest
Scaled
Residual
1.56
1.60
-2.12
3.16
3.05
Scaled
Residual of
Interest
0.000
0.000
0.000
1.180
1.078
BM
R
Extr
a
Risk
10%
HECC
BMC
(mg/m3)
0.008
0.007
0.013
0.031
0.032
BMCLd
(mg/m3)
0.006
0.000
0.010
0.027
0.027
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLarynxEpiglottisHyperplasia\NTP_2002
_Larynx Epiglottis Hyperplasia_LogLogistic_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsLarynxEpiglottisHyperplasia\NTP_2002
_Larynx Epiglottis Hyperplasia_LogLogistic_0.1.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Response
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0
intercept = 2.65432
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
167
DRAFT - DO NOT CITE OR QUOTE
-------
intercept
intercept
1
Variable
background
intercept
slope
Parameter Estimates
95.0% Wald Confidence Interval
Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
0 * * *
2.66112 * * *
1 * * *
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param's Deviance
-95.6454 4
-97.6263 1 3.96181
-134.962 1 78.6325
Test d.f.
P-value
0.2656
<.0001
197.253
Goodness of Fit
Dose
0.0000
0.0473
0.0929
0.1796
Est. Prob.
0.0000
0.4038
0.5706
0.7200
Expected
0.000
20.190
28 .532
35.279
Observed
0.000
18 .000
34 .000
32 .000
Size
49
50
50
49
Scaled
Residual
0.000
-0.631
1.562
-1.043
Chi 2 =3.93
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.00776335
BMDL = 0.00584847
P-value = 0.2694
Log-Logistic Model with 0.95 Confidence Level
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
13:4009/102010
Data Set 5: Incidence of Nose Goblet Cell, Respiratory Epithelium, Hyperplasia in Male
Rats, NTP (2002)
Summary
Three concentration groups (0.5, 1 and 2 mg/m3) showed statistically significant
difference from the control group as reported. The severity increase slightly as reported when the
168
DRAFT - DO NOT CITE OR QUOTE
-------
concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed the
statistically significant trend with Z score as 2.68 and one-sided/?-value as 0.0037.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-10.
Five models demonstrated adequate goodness of flip-value ^0.1 and good visual fit, but
LogProbit model failed to compute a reasonable BMCL. Based on the draft Benchmark Dose
Technical Guidance (U.S. EPA, 2000), the LogLogistic model was selected as the model for
POD computation. The BMCLio of this model was 0.026 mg/m3 and regarded as the one of the
candidate PODs.
Table B-10: Benchmark Modeling Results for Incidence of Nose Goblet Cell, Respiratory
Epithelium, Hyperplasia in Male Rats, NTP (2002)
Model a
LogLogistic
LogProbit
Gamma
Multistage e
(Stage 1)
Weibull
Probit
Logistic
Goodness of Fit
P-
Value
0.16
0.31
0.12
0.08
0.07
AIC
b
213.2
212.8
213.7
214.9
215.0
Largest
Scaled
Residual
1.66
-0.83
1.75
1.787
1.78
Scaled
Residual of
Interest
1.656
-0.003
1.75
-0.155
-0.128
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
)
0.044
0.002
0.052
0.077
0.081
BMCLd
(mg/m3
)
0.026
failed
0.033
0.056
0.060
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsNoseGobletEpiHyperplasia\NTP_2002_La
rynx Goblet Hyperplasia _LogLogistic_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerMaleRats\MaleRatsNoseGobletEpiHyperplasia\NTP_2002_La
rynx Goblet Hyperplasia _LogLogistic_0.1.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Response
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
169
DRAFT - DO NOT CITE OR QUOTE
-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0816327
intercept = 1.21717
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background
intercept
background
1
-0.72
intercept
-0.72
1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0.110444
0.926382
1
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f. P-value
-102.873 4
-104.611 2 3.47486 2 0.176
-109.105 1 12.4644 3 0.00595
AIC:
Dose
0.
0.
0.
0.
.0000
.0473
.0929
.1796
213 .221
Est. Prob.
0.
0.
0.
0.
.1104
.2054
.2794
.3881
Goodness of Fit
Expected Observed
5.
10.
13 .
18 .
.412
.270
.691
.627
4 .
15.
12 .
17.
.000
.000
.000
.000
Size
49
50
49
48
Scaled
Residual
-0.
1 .
-0.
-0.
.643
.656
.539
.482
Chi 2 =3.68
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.0439982
BMDL = 0.026051
P-value = 0.1590
170
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.08 0.1
dose
14:0809/102010
Data Set 6: Incidence of Lung Alveolar Epithelium Hyperplasia in Female Rats, NTP
(2002)
Summary
Two concentration groups (1 and 2 mg/m3) showed statistically significant difference
from the control as reported. The severity of these nonneoplastic lesions increased from minimal
to moderate as reported when the concentration of vanadium pentoxide increased. The Cochran-
Armitage test confirmed the statistically significant trend with Z score as 9.53 and one-sidedp-
valueasO.OOOl.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the combined incidence was used as the BMR to determine the
POD. The goodness of fit, BMC and BMCL for each model are summarized in Table B-l 1.
Five models demonstrated adequate goodness of flip-value ^ 0.1, but the extreme
curvature of LogLogistic model did not reelect in the observed data. Based on the draft
Benchmark Dose Technical Guidance (US EPA, 2000b), the Gamma model was selected as the
model for POD computation. The BMCLio of this model was 0.063 mg/m3 and regarded as one
of the candidate PODs.
171
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-ll: Benchmark Modeling Results for Incidence of Lung Alveolar Epithelium Hyperplasia
in Female Rats, NTP (2002)
Model a
Gamma
LogLogisti
c
Weibull
LogProbit
Multistage
(StageS)
Logistic
Probit
Goodness of Fit
P-
Value
0.89
0.96
0.91
0.78
0.46
0.00
0.00
AIC
b
156.2
155.9
157.9
157.9
158.1
168.1
168.4
Largest
Scaled
Residual
0.37
0.20
0.08
0.20
0.87
2.34
2.32
Scaled
Residual of
Interest
-0.12
0.00
-0.01
0.00
-0.24
-0.78
-1.11
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
)
0.076
0.086
0.071
0.085
0.055
0.035
0.031
BMCLd
(mg/m3
)
0.063
0.070
0.053
0.068
0.041
0.028
0.025
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Gamma Model. (Version: 2.15; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLungAlBrEpHyperplasia\NTP_2002_L
ung Alveolar Bronchiole Combined Epithelium Hyperplasia_Gamma_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLungAlBrEpHyperplasia\NTP_2002_L
ung Alveolar Bronchiole Combined Epithelium Hyperplasia_Gamma_0.1.pit
The form of the probability function is:
P[response]= background+(1-background)*CumGamma[slope*dose,power],
where CumGamma(.) is the cummulative Gamma distribution function
Dependent variable = Response
Independent variable = Dose
Power parameter is restricted as power >=1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
Background = 0.156863
Slope = 139.011
Power = 16.8546
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Power
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background
Slope
Background
1
-0.32
Slope
-0.32
1
Parameter Estimates
172
DRAFT - DO NOT CITE OR QUOTE
-------
Variable
Background
Slope
Power
Estimate
0.151066
169.414
18
Std. Err.
0.03605
9.58875
NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
0.080409 0.221723
150.621 188.208
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-75.9175
-76.0968
-135.531
156.194
# Param's
4
2
1
Deviance Test d.f.
P-value
0.358619
119.227
Goodness of Fit
Chi 2 = 0.22
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.0756821
BMDL = 0.0626004
P-value = 0.8945
0.8358
<.0001
Dose
0.0000
0.0468
0.0936
0.1871
Est. Prob.
0.1511
0.1523
0.4286
0.9973
Expected
7.402
7.462
21.431
49.864
Observed
7.000
8 .000
21.000
50.000
Size
49
49
50
50
Scaled
Residual
-0.160
0.214
-0.123
0.369
Gamma Multi-Hit Model with 0.95 Confidence Level
14:4009/102010
Gamma Multi-Hit
0.1
dose
Data Set 7: Incidence of Lung Chronic Active Inflammation in Female Rats, NTP (2002)
Summary
One concentration group (2 mg/m3) showed statistically significant difference from the
control group as reported. The severity did not change much when the concentration of
vanadium pentoxide increased. The Cochran-Armitage test confirmed the statistically significant
trend with Z score as 6.71 and one-sided/>-value as <0.0001.
173
DRAFT - DO NOT CITE OR QUOTE
-------
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-12.
Five models demonstrated adequate goodness of flip-value ^0.1 and good visual fit.
Based on the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the Multistage
(StageS) model was selected as the model for POD computation. The BMCLio of this model
was 0.048 mg/m3 and regarded as one of the candidate PODs.
Table B-12: Benchmark Modeling Results for Incidence of Lung Chronic Active Inflammation in
Female Rats, NTP (2002)
Model a
Multistage
e
(StageS)
LogProbit
Gamma
LogLogisti
c
Weibull
Logistic
Probit
Goodness of Fit
P-
Value
0.82
1.00
0.99
0.97
0.94
0.04
0.03
AIC
b
212.9
214.5
214.5
214.5
214.5
218.8
219.4
Largest
Scaled
Residual
-0.51
0.00
-0.01
-0.03
-0.06
1.64
-1.71
Scaled
Residual of
Interest
-0.51
0.00
0.00
0.00
0.01
-0.48
-0.62
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
)
0.080
0.094
0.094
0.094
0.094
0.040
0.037
BMCLd
(mg/m3
)
0.048
0.068
0.063
0.065
0.059
0.033
0.031
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Multistage Model. (Version: 3.2; Date: 05/26/2010)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLungChronicActiveInflammation\NT
P_2002_Lung Chronic Active Inflammation_Multi3_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLungChronicActiveInflammation\NT
P_2002_Lung Chronic Active Inflammation_Multi3_0.1.pit
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3)]
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
174
DRAFT - DO NOT CITE OR QUOTE
-------
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.17928
Beta(l) = 0
Beta(2) = 0
Beta(3) = 214.579
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l) -Beta(2)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background
Beta(3)
Background
1
-0.41
Beta(3)
-0.41
1
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Estimate
0.186734
0
0
206.262
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f. P-value
-104.257 4
-104.451 2 0.389305 2 0.8231
-130.861 1 53.209 3 <.0001
AIC:
212.903
Dose
0.0000
0.0468
0.0936
0.1871
Est. Prob.
0.1867
0.2037
0.3132
0.7896
Goodness of Fit
Expected Observed Size
9.150
9.983
15.659
39.478
10.000
10.000
14 .000
40.000
49
49
50
50
Scaled
Residual
0.312
0.006
-0.506
0.181
Chi 2 = 0.39
d.f. = 2
P-value = 0.8246
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.079938
BMDL = 0.0483842
BMDU = 0.0900697
Taken together, (0.0483842, 0.0900697) is a 90% two-sided confidence
interval for the BMD
175
DRAFT - DO NOT CITE OR QUOTE
-------
Multistage Model with 0.95 Confidence Level
Multistage
0.1
dose
15:2809/102010
Data Set 8: Incidence of Larynx Chronic Inflammation in Female Rats, NTP (2002)
Summary
Three concentration groups (0.5, 1 and 2 mg/m3) showed statistically significant
difference from the control group as reported. The severity did not change much when the
concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed the
statistically significant trend with Z score as 5.38 and one-sidedp-value as <0.0001.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-13.
Five models demonstrated adequate goodness of flip-value ^0.1 and good visual fit.
Based on the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the LogLogistic
model was selected as the model for POD computation. The BMCLio of this model was 0.003
mg/m3 and regarded as one of the candidate PODs.
Table B-13: Benchmark Modeling Results for Incidence of Larynx Chronic Inflammation in
Female Rats, NTP (2002)
Model a
LogLogistic
LogProbit
Gamma
Multistage e
(Stage 1)
Goodness of Fit
P-
Value
0.46
0.27
0.19
AIC
b
242.0
243.6
243.7
Largest
Scaled
Residual
-0.88
-0.89
1.58
Scaled
Residual of
Interest
-0.13
-0.02
-0.57
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
)
0.005
0.003
0.007
BMCLd
(mg/m3
)
0.003
0.000
0.006
176
DRAFT - DO NOT CITE OR QUOTE
-------
Weibull
Probit
Logistic
0.03
0.03
247.5
247.6
1.99
1.97
1.99
1.97
0.013
0.014
0.011
0.011
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxChronicInflammation\NTP_20
02_Larynx Chronice Inflammation_LogLogistic_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxChronicInflammation\NTP_20
02_Larynx Chronice Inflammation_LogLogistic_0.1.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Response
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.16
intercept = 3.2396
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background
intercept
background
1
-0.54
intercept
-0.54
1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0.166778
3.19725
1
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-118.217
-118.997
-137.233
# Param's
4
2
1
Deviance Test d.f.
P-value
1.55893
38.0314
0.4587
<.0001
AIC:
241.994
Goodness of Fit
Dose
0.0000
0.0231
0.0470
Est. Prob.
0.1668
0.4678
0.6123
Expected
8 .339
22 .925
30.001
Observed
8 .000
26 .000
27.000
Size
50
49
49
Scaled
Residual
-0.129
0.881
-0.880
177
DRAFT - DO NOT CITE OR QUOTE
-------
0.0875 0.7347 36.735 37.000
Chi^2 = 1.57 d.f. = 2 P-value = 0.4553
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.00454158
BMDL = 0.00312534
Log-Logistic Model with 0.95 Confidence Level
50
0.085
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
16:2909/102010
Data Set 9: Incidence of Larynx Respiratory Epithelium, Epiglottis, Hyperplasia in Female
Rats, NTP (2002)
Summary
Three concentration groups (0.5, 1 and 2 mg/m3) showed statistically significant
difference from the control group as reported. The severity did not change much when the
concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed the
statistically significant trend with Z score as -5.98 and one-sided/?-value as <0.0001.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-14.
Both LogLogistic and LogProbit models demonstrated adequate goodness of flip-value ^
0.1 and good visual fit, but LogProbit model failed to compute a reasonable BMCL. Based on
the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the LogLogistic model was
selected as the model for POD computation. The BMCLio of this model was 0.003 mg/m3 and
regarded as one of the candidate PODs.
178
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-14: Benchmark Modeling Results for Incidence of Larynx Respiratory Epithelium,
Epiglottis, Hyperplasia in Female Rats, NTP (2002)
Model a
LogLogist
ic
LogProbit
Gamma
Multistage
e
(Stagel)
Weibull
Probit
Logistic
Goodness of Fit
P-
Value
0.30
0.78
0.01
0.00
0.00
AIC
b
205.3
204.2
212.3
234.6
235.3
Largest
Scaled
Residual
1.53
-0.57
2.85
3.258
-3.295
Scaled
Residual of
Interest
0.000
0.000
0.000
3.258
3.174
BM
R
Extr
a
Risk
10%
HECC
BMC
(mg/m3)
0.004
0.000
0.006
0.016
0.016
BMCLd
(mg/m3)
0.003
failed
0.005
0.013
0.014
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxEpiglottisHyperplasia\NTP_
2002_Larynx Epiglottis Hyperplasia_LogLogistic_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxEpiglottisHyperplasia\NTP_
2002_Larynx Epiglottis Hyperplasia_LogLogistic_0.1.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Response
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0
intercept = 3.38546
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept
intercept
1
Parameter Estimates
Variable
background
intercept
Estimate
0
3.3735
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
179
DRAFT - DO NOT CITE OR QUOTE
-------
slope 1 *
- Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param's Deviance
-99.8781 4
-101.66 1 3.56345
-134.962 1 70.1671
Test d.f.
P-value
0.3126
<.0001
205.32
Goodness of Fit
Scaled
Dose Est. Prob. Expected Observed Size Residual
Chi
0.0000 0.0000 0.000 0.000
0.0231 0.4029 19.743 25.000
0.0470 0.5781 28.329 26.000
0.0875 0.7186 35.929 33.000
^2 = 3.65 d.f. = 3 P-value = 0.3022
50 0.000
49 1.531
49 -0.674
50 -0.921
Benchmark Dose Computation
Specified effect = 0.1
Risk
Type = Extra risk
Confidence level = 0.95
T3
1
|
1
t
BMD = 0.00380772
BMDL = 0.0028644
Log-Logistic Model with 0.95 Confidence Level
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0
,•••';•••.• \: ' ' ' '
^^^^
^^^
/^
s/
/^
J
BMDL BMD
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
dose
-
-
•
0.09
11:1009/162010
Data Set 10: Incidence of Nose Goblet Cell, Respiratory Epithelium, Hyperplasia in Female
Rats, NTP (2002)
Summary
One concentration groups (2 mg/m3) showed statistically significant difference from the
control group as reported. The severity did not change much when the concentration of
vanadium pentoxide increased. The Cochran-Armitage test confirmed the statistically significant
trend with Z score as 3.46 and one-sided/7-value as 0.0003.
Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this
data set. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
goodness of fit, BMC and BMCL for each model are summarized in Table B-15.
All models demonstrated adequate goodness of flip-value ^0.1 and good visual fit.
180
DRAFT - DO NOT CITE OR QUOTE
-------
Based on the draft Benchmark Dose Technical Guidance (US EPA, 2000b), the Multistage
(Stage2) model was selected as the model for POD computation. The BMCLio of this model
was 0.014 mg/m3 and regarded as one of the candidate PODs.
Table B-15: Benchmark Modeling Results for Incidence of Larynx Respiratory Epithelium,
Epiglottis, Hyperplasia in Female Rats, NTP (2002)
Model a
Multistage
(Stage2)
Logistic
Probit
Weibull
( Quantal
Linear)
LogLogisti
c
Gamma
LogProbit
Goodness of Fit
P-
Value
0.43
0.33
0.32
0.22
0.28
0.28
0.28
AIC
b
258.3
258.9
258.9
259.7
259.8
259.8
259.8
Largest
Scaled
Residual
-0.90
-1.252
-1.273
-1.45
0.78
0.77
0.765
Scaled
Residual of
Interest
-0.896
0.586
0.564
0.337
-0.027
-0.005
-0.001
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3
0.038
0.024
0.023
0.019
0.064
0.063
0.062
BMCLd
(mg/m3
0.014
0.018
0.018
0.012
0.014
0.014
0.015
a Selected (best-fitting) model shown in first row, in boldface type.
b AIC=Akaike Information Criterion.
0 HEC=Human Equivalent Concentration.
d BMCL= the lower bound of BMC at 95% confidence level.
e For Multistage model, up to 3 stages were tested. Only the model with the lowest AIC and lowest stage was
reported here.
BMDS output file
Multistage Model. (Version: 3.2; Date: 05/26/2010)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsNoseGobletEpiHyperplasia\NTP_200
2_Nose Goblet Hyperplasia_Multi2_0.1.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002NonCancerFemaleRats\FemaleRatsNoseGobletEpiHyperplasia\NTP_200
2_Nose Goblet Hyperplasia_Multi2_0.1.pit
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2)]
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
181
DRAFT - DO NOT CITE OR QUOTE
-------
Default Initial Parameter Values
Background = 0.270822
Beta(l) = 0
Beta(2) = 75.6009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background
Beta(2)
Background
1
-0.59
Beta(2)
-0.59
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0.276279
0
71.2588
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f. P-value
-126.318 4
-127.169 2 1.70254 2 0.4269
-133.292 1 13.9479 3 0.002977
AIC:
258.338
Dose
0.0000
0.0231
0.0470
0.0875
Est. Prob.
0.2763
0.3033
0.3815
0.5806
Goodness of Fit
Expected Observed Size
13 .814
15.167
19.077
29.030
13 .000
18 .000
16 .000
30.000
50
50
50
50
Scaled
Residual
-0.257
0.872
-0.896
0.278
=1.71
d.f.
P-value = 0.4262
Benchmark Dose Computation
Specified effect
Risk Type
Confidence level
BMD
BMDL
BMDU
0.1
Extra risk
0.95
0.0384521
0.0135524
0.0544996
Taken together, (0.0135524, 0.0544996) is a
interval for the BMD
two-sided confidence
182
DRAFT - DO NOT CITE OR QUOTE
-------
Multistage Model with 0.95 Confidence Level
T5 0.4
Multistage
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
16:4909/102010
183
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX C - BENCHMARK CONCENTRATION MODELING OF COMBINED
LUNG ALVEOLAR AND BRONCHIOLAR TUMOR DATA SETS FROM NTP
STUDIES (2002)
To derive the cancer slope factor of vanadium pentoxide, combined alveolar/bronchiolar
adenoma and carcinoma data sets in B6C3F1 mice (Table C-2) by NTP (2002) were selected
because of their biological and statistical significance.
The highlights of the benchmark concentration modeling results are:
• Cancer slope factor:
The cancer slope factor was estimated as 3.4 per mg/m3 after linear extrapolation. The
selected model for each data set, the candidate PODs and the calculation of cancer slope factor
are summarized in Table C-l.
• BMDS modeling:
Each data set was firstly fitted with the recommended dichotomous model, multistage-
cancer model, through EPA BMDS (version 2.1.2); if the goodness of flip-value < 0.05, other
dichotomous models available in EPA BMDS (version 2.1.2) were used; if still no model
showed adequate goodness of flip-value ^ 0.05, the highest dose was dropped for further
modeling. Following the general model selection steps outlined in the draft Benchmark Dose
Technical Guidance (U.S. EPA, 2000), the best-fitting model was selected for each data set to
estimate the candidate POD, which was the BMCL at the pre-determined BMR.
• BMR:
Since all the non-control concentrations showed essentially a plateau response, the tumor
data sets provided limited information about the concentration-response relationship. So, BMR
for each data set was calculated based on the response at the control and the first non-control
concentration groups.
• Multistage Weibull (MSW) time-to-tumor modeling:
Since the survival curves reported by NTP (2002) showed a high percentage of deaths
(22-46%) at the end of 2-year studies across the different concentrations groups, MSW time-to-
tumor model was also used with these data sets. However, no adequate fit was noticed for either
male or female mice data set.
184
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-1. Summary of Candidate POPs and Cancer Slope Factors.
Animal
Male B6C3Fi
Mice
Female B6C3Fi
Mice
BMR
(Extra
Risk) a
0.71
0.67
HECb
BMC
(mg/m3
)
0.360
0.237
BMCLC
(Candidate
POD, mg/m3)
0.208
0.161
Cancer
Slope
Factor d
(per
mg/m3)
3.4
4.2
a BMR was calculated based on the response at the control concentration and the first non-control concentration
(Table C-6).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
d Cancer Slope Factor = BMR/BMCL, as linear extrapolation was used.
C.I. TREND TESTS BEFORE MODELING
Two inhalation carcinogenesis data sets were selected because of the biological and the
statistical significance reported by NTP (2002). Cochran-Armitage test was conducted to
confirm the significance of the statistical trend for the selected data sets (Table C-2) before
modeling.
Table C-2: Trend Tests on the Inhalation Carcinogenesis Data Sets in Mice from the 2-Year
Inhalation Studies (NTP, 2002)
Endpoint
Concentration as Reported (mg/m3) a
0.0
1.0
2.0
4.0
Trend Test b
Z -Score
p-value
Incidence c in Male B6C3Fi Mice
Lung
Al veol ar/Bronchi ol ar
Adenoma and
Carcinoma
22/50
42/50
43/50
43/50
4.14
O.0001
Incidence c in Female B6C3Fi Mice
Lung
Al veol ar/Bronchi ol ar
Adenoma and
Carcinoma
1/48
32/47
35/48
32/49
5.19
O.OOOl
a Concentrations are as reported by NTP (2002).
b One-sided Cochran-Armitage trend test.
0 Incidence = (number of animals with alveolar/bronchiolar adenoma and/or carcinoma)/(animal sample size). Only
the animals alive at 52 weeks or when the first tumor appeared, whichever was earlier, was counted toward the
sample size.
C.2. DOSE CONVERSION
In the carcinogenesis studies reported by NTP (2002), mice (B6C3Fi) were exposed to
185
DRAFT - DO NOT CITE OR QUOTE
-------
vanadium pentoxide through inhalation. To analyze the concentration response effect, the
reported concentrations of vanadium pentoxide were converted to human equivalent
concentrations before any modeling and extrapolation.
Firstly, the average life time animal body weights of rats were estimated based on the
mean body weight at different weeks (Table C-3). Secondly, following the US EPA inhalation
methods (1994), RDDRs (regional deposited dose ratio) were calculated with the RDDR
program designed by US EPA (1994) and summarized in Table C-4. Then, the human equivalent
concentration for each concentration/sex group was calculated from the reported concentration in
mice by multiply the continuous exposure adjustment factor and the pulmonary RDDR (Table C-
5).
Table C-3: Average Life Time Animal Body Weight of Mice for 2-Year Inhalation Studies of
Vanadium Pentoxide (NTP, 2002)
Vanadium Pentoxide
Concentration as
Reported (mg/m3)
Mean Animal Body Weight a (g)
1-13
weeks
14-52
weeks
53-104
weeks
Average Life
Time Animal
Body Weight b
(g)
Male Mice (B6C3Fi)
0
1
2
4
30.8
30.5
30.3
29.8
46.7
47.0
45.4
43.7
54.1
52.9
51.4
46.1
48.9
48.3
46.9
43.6
Female Mice (B6C3Fi)
0
1
2
4
25.5
24.9
24.8
24.5
42.7
40.9
36.8
34.2
56.1
50.7
44.8
40.1
47.7
44.2
39.7
36.3
a As reported in Table 20-2lof the NTP report (2002).
b "Average Life Time Animal Body Weight" = ( "Mean Body Weight 1-13 Weeks" * 13 +" Mean Body Weight 14 -
52 Weeks" * 39 + "Mean Body Weight 53-104 Weeks" * 52) 7104
186
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-4: RDDRs for Different Concentration/Sex Group of Mice in the 2-Year Inhalation Studies
of Vanadium Pentoxide (NTP, 2002)
Vanadium Pentoxide
Concentration as
Reported (mg/m3)
Average Life
Time Animal
Body Weight a
(g)
Average
MMAD
b
Average
GSDC
RDDRd
Pulmonary
Male Mice (B6C3Fi)
0
1
2
4
48.9 e
48. 3 e
46.9 e
43.6
1.26
1.87
1.168e
1.168e
1.168e
1.134
Female Mice (B6C3Fi)
0
1
2
4
47.7 e
44.2
39.7
36.3
1.26
1.87
1.168e
1.143
1.077
1.023
a Average life time animal body weight of each concentration/sex group was calculated in Table C-3.
b MMAD =mass median aerodynamic diameter; calculated based on the MMADs reported for 2-year studies (NTP,
2002) .
0 GSD = geometric standard deviation; calculated based on the GDSs reported for 2-year studies (NTP, 2002).
d RDDR = regional deposited dose ratio; calculated with the RDDR program from USA EPA (1994);
e Since the average life time animal body weight was out of the mice weight range (17-46g) accepted by RDDR
program (V.2.3, US EPA), 46g was used to calculate the corresponding RDDRs.
Table C-5: Human Equivalent Concentrations of Vanadium Pentoxide in the 2-Year Inhalation
Studies (NTP, 2002)
Concentration as
Reported a
(mg/m3)
Continuous
Exposure
Adjustment
Factor b
RDDRC
Pulmonary
HECd
(mg/m3)
Pulmonary
Male Mice (B6C3Fi)
0
1
2
4
0.179
0.179
0.179
0.179
1.168
1.168
1.168
1.134
0.00
0.21
0.42
0.81
Female Mice (B6C3Fi)
0
1
2
4
0.179
0.179
0.179
0.179
1.168
1.143
1.077
1.023
0.00
0.20
0.38
0.73
a "Toxicology and carcinogenesis studies of vanadium pentoxide in F344/N rats and B6C3F! Mice", NTP, 2002.
b "Continuous Exposure Adjustment Factor" = (6/24) * (5/7); animals were exposed to vanadium pentoxide 6 hours
per day and 5 days per week.
0 Please refer to Appendix Table C-4.
187
DRAFT - DO NOT CITE OR QUOTE
-------
d HEC=Human Equivalent Concentration = "Concentration as Reported" * "Continuous Exposure Adjustment
Factor " * "RDDR"
C.3. BMR CALCULATION
Since all non-control concentrations from the NTP carcinogenesis studies (2002) showed
essentially a plateau response, the data set provided limited information about the concentration-
response relationship because the complete range of response from background to maximum
must occur somewhere below the lowest dose, thus the BMD may be just below the first dose or
orders of magnitudes lower. So, a BMR estimated based on the response at the control
concentration and the first non-control concentration was calculated (Table C-6), and then used
for estimation of BMCL.
Table C-6: BMR Estimation for Male and Female Mice Data Sets.
Animal
Male B6C3Fi Mice
Female B6C3Fi
Mice
P(Control)
a
0.44
0.02
P(lst Non-
control
Concentration)
b
0.84
0.68
Calculated BMR
(Extra Risk) c
0.71
0.67
a The combined alveolar/bronchiolar adenoma and carcinoma incidence in the control group.
b The combined alveolar/bronchiolar adenoma and carcinoma incidence in the 1st non-control concentration group.
c Calculated BMR = [P(lst Non-control Concentration) - P(Control)]/[l- P(Control)].
C.4. BMDS MODELING FOR INHALATION CARCINOGENESIS DATA SETS
Each data set was firstly fitted with dichotomous multistage-cancer model provided in
EPA BMDS (version 2.1.2); if the goodness of flip-value < 0.05, other dichotomous models
available in EPA BMDS (version 2.1.2) were fitted; if still no model showed adequate goodness
of flip-value ^ 0.05, the highest dose was dropped for further modeling.
Following the general model selection steps outlined in the draft Benchmark Dose
Technical Guidance (U.S. EPA, 2000), the best-fitting model was selected for each data set to
estimate the candidate POD, which was the BMCL at the calculated BMR for each data set.
The selected models and candidate PODs for all endpoints are summarized in Table C-l.
Lung Tumor Data Set 1: Combined Incidence of Alveolar/Bronchiolar Adenoma and
Carcinoma in Male Mice, NTP (2002).
Summary
The Cochran-Armitage test confirmed the statistically significant trend with Z score as
4.14 and one-sided/?-value as <0.0001.
188
DRAFT - DO NOT CITE OR QUOTE
-------
Because all the non-control concentrations showed a similar effect, a BMR as 71% extra
risk was calculated and used to estimate the POD. The goodness of fit, BMC and BMCL for each
model are summarized in Table C- 7 and C-8.
Since the primary cancer model (Multistage-cancer) did not show an adequate fit, six
other dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this data set.
Two models demonstrated adequate goodness of flip-value ^ 0.05 and good visual fit, but
LogProbit model failed to compute a reasonable BMCL. Based on the draft Benchmark Dose
Technical Guidance (U.S. EPA, 2000), the LogLogistic model was selected and BMCL7i of this
model was 0.208 mg/m3.
Further BMD modeling was performed for comparison purposes. Since the primary
cancer model (Multistage-cancer) did not show adequate fit with all doses, the highest
concentration was dropped for further modeling. After dropping the high dose, the primary
cancer model (Multistage-cancer) demonstrated adequate goodness of flip-value ^ 0.05 and
good visual fit.
Table C- 7. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar Adenoma
and Carcinoma in Male Mice, NTP (2002).
Model a
Goodness of Fit
P-
Valu
e
AIC
b
Largest
Scaled
Residual
Scaled
Residual of
Interest
BM
R
HEC
BMC
(mg/m3)
BMCLC
(mg/m3)
Primary Cancer Models
Multistage-
Cancer Stage
1,2 and 3
0.01
207.
0
2.05
0.72
Extr
a
Risk
71%
0.532
0.379
Other Dichotomous Models
LogLogistic
LogProbit
Gamma
Weibull
Logistic
Probit
0.19
0.88
0.01
0.00
0.00
200.
632
199.
6
207.
0
209.
4
210.
4
-1.42
0.13
2.05
2.15
2.19
0.04
-0.06
0.72
0.96
-1.54
Extr
a
Risk
71%
0.360
0.146
0.532
0.609
0.654
0.208
failed
0.379
0.447
0.495
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose Technical Guidance
(2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
189
DRAFT - DO NOT CITE OR QUOTE
-------
BMDS output file
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002Cancer\MaleMiceTumor4DosesAllModel\NTP_2002_MaleMiceTum
or4DosesAHModels_LogLogistic_0 . 71. (d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002Cancer\MaleMiceTumor4DosesAllModel\NTP_2002_MaleMiceTum
or4DosesAHModels_LogLogistic_0.71.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.44
intercept = 1.9384
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.58
intercept -0.58 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
background 0.451808 * * *
intercept 1.91813 * * *
slope 1 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -96.7763 4
Fitted model -98.3159 2 3.07913 2 0.2145
Reduced model -112.467 1 31.3814 3 <.0001
AIC: 200.632
Goodness of Fit
Dose
0.0000
0.2100
0.4200
Est. Prob.
0.4518
0.7744
0.8580
Expected
22 .590
38 .719
42 .898
Observed
22 .000
42 .000
43 .000
Size
50
50
50
Scaled
Residual
-0.168
1.110
0.041
190
DRAFT - DO NOT CITE OR QUOTE
-------
0.8100 0.9159 45.793 43.000
^2 = 3.29 d.f. = 2 P-value = 0.1934
50
-1.423
Benchmark Dose Computation
Specified effect = 0.71
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.359607
BMDL = 0.208416
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.4
dose
15:3709/202010
Table C- 8. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar Adenoma
and Carcinoma in Male Mice after Dropping the Highest Concentration, NTP (2002).
Model a
Goodness of Fit
P-
Value
AICb
Largest
Scaled
Residual
Scaled
Residual
of
Interest
BMR
HEC
BMC
(mg/m3
)
BMCLC
(mg/m3
)
Primary Cancer Models
Multistage-
Cancer
Stage 1,2
and 3
0.12
159.5
1.19
Extra
Risk
71%
0.310
0..220
Other Dichotomous Models
LogLogistic
Gamma
Weibull
Probit
Logistic
LogProbit
0.43
0.12
0.04
0.05
NA
157.7
159.5
161.5
160.1
159.1
0.52
1.19
-1.03
-1.09
0.00
Extra
Risk
71%
0.260
0.310
0.340
0.330
0.190
0.140
0.220
0.270
0.250
failed
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose Technical Guidance
(2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
191
DRAFT - DO NOT CITE OR QUOTE
-------
Multistage Cancer Model. (Version: 1.9; Date: 05/26/2010)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002Cancer\MaleMiceTumor3DosesAllModel\NTP_2002_MaleMiceTumor3Dose
sAHModels_MultiCancl_0.71. (d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002Cancer\MaleMiceTumor3DosesAllModel\NTP_2002_MaleMiceTumor3Dose
sAHModels_MultiCancl_0 . 71 .pit
Thu Feb 17 17:24:36 2011
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l)]
The parameter betas are restricted to be positive
Dependent variable = Effect
Independent variable = Dose
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.535297
Beta(l) = 3.3007
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background
Beta(l)
1
-0.6
-0.6
1
Parameter Estimates
Variable
Background
Beta(l)
Estimate
0.461149
4.04221
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f. P-value
-76.5282 3
-77.7667 2 2.47711 1 0.1155
-89.871 1 26.6857 2 <.0001
159.533
Goodness of Fit
Dose
0.0000
0.2100
0.4200
Chi^2 = 2.
Est. Prob.
0.4611
0.7694
0.9013
.45 d.f. =
Expected
23 .057
38 .471
45.067
1 P-
Observed
22 .000
42 .000
43 .000
value = 0.1172
Size
50
50
50
Scaled
Residual
-0.300
1.185
-0.980
Benchmark Dose Computation
Specified effect = 0.71
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.306237
BMDL = 0.220508
BMDU = 0.471328
Taken together, (0.220508, 0.471328)
interval for the BMD
Multistage Cancer Slope Factor =
is a 90
3 .21985
% two-sided confidence
192
DRAFT - DO NOT CITE OR QUOTE
-------
Multistage Cancer Model with 0.95 Confidence Level
0.9
0.8
0.7
0.6
0.5
0.4
0.3 -
17:3402/172011
Multistage Cancer
Linear extrapolation
0.05
0.1
0.15
0.2
dose
0.25
0.3
0.35
0.4
Lung Tumor Data Set 2: Combined Incidence of Alveolar/Bronchiolar Adenoma and
Carcinoma in Female Mice, NTP (2002).
Summary
The Cochran-Armitage test confirmed the statistically significant trend with Z score as
5.19 and one-sidedp-value as <0.0001.
Because all the non-control concentrations showed a similar effect, a BMR as 67% extra
risk was calculated and used to determine the POD. The goodness of fit, BMC and BMCL for
each model are summarized in Table C-9 and C-10.
Since the primary cancer model (Multistage-cancer) did not show an adequate fit, six
other dichotomous models available in EPA BMDS (version 2.1.2.) were fitted to this data set.
Only one model demonstrated adequate goodness of flip-value ^ 0.05, however, it failed to
compute BMC and BMCL. So, the highest concentration was dropped for further modeling.
After dropping the highest dose, the primary cancer model (Multistage-cancer) still did
not show an adequate fit. Then six other dichotomous models available in EPA BMDS (version
2.1.2.) were tried. Only the LogLogistc model demonstrated adequate goodness of flip-value ^
0.05 and good visual fit. Based on the draft Benchmark Dose Technical Guidance (U.S. EPA,
2000), the LogLogistic model was selected and BMCL6y of this model was 0.161 mg/m3.
Table C- 9. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar Adenomas
and Carcinomas with All Four Concentrations in Female Mice, NTP (2002).
I Modela
Goodness of Fit
BMR
HEC
193
DRAFT - DO NOT CITE OR QUOTE
-------
P-
Value
AICb
Largest
Scaled
Residual
Scaled
Residual
of
Interest
BMC
(mg/m3
)
BMCLC
(mg/m3
)
Primary Cancer Models
Multistage-
Cancer Stage
1,2 and 3
0.00
218.8
-3.90
1.33
Extra
Risk
67%
0.435
0.359
Other Dichotomous Models
LogProbit
LogLogistic
Gamma
Weibull
Logistic
Probit
0.72
0.00
0.00
0.00
0.00
192.6
202.7
218.8
239.4
239.6
0.62
-2.87
-3.90
-4.00
-3.97
-999.00
0.52
1.33
-2.62
-2.48
Extra
Risk
67%
Failed d
0.351
0.435
0.657
0.670
Failed d
0.255
0.359
0.532
0.553
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose Technical Guidance
(2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
Table C-10. BMDS Modeling Results for Combined Incidence of Alveolar/Bronchiolar Adenoma
and Carcinoma in Female Mice after Dropping the Highest Concentration, NTP (2002).
Model a
Goodness of Fit
P-
Value
AICb
Largest
Scaled
Residual
Scaled
Residual
of
Interest
BMR
HEC
BMC
(mg/m3
)
BMCLC
(mg/m3
)
Primary Cancer Models
Multistage-
Cancer Stage
1,2 and 3
0.06
132.2
1.42
1.42
Extra
Risk
67%
0.264
0.213
Other Dichotomous Models
LogLogistic
Gamma
Weibull
Probit
Logistic
LogProbit
0.37
0.06
0.00
0.00
NA
129.453
132.2
145.9
146.5
130.7
-0.67
1.42
3.00
2.84
0.00
0.60
1.42
-1.85
-1.99
0.00
Extra
Risk
67%
0.237
0.264
0.309
0.303
0.190
0.161
0.213
0.272
0.263
failed
a Selected model is the best-fitting model for the data set based on the draft Benchmark Dose Technical Guidance
(2000).
b HEC=Human Equivalent Concentration.
c BMCL= the lower bound of BMC at 95% confidence level.
BMDS output file
194
DRAFT - DO NOT CITE OR QUOTE
-------
Logistic Model. (Version: 2.13; Date: 10/28/2009)
Input Data File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002Cancer\FemaleMiceTumor3DosesAllModels\NTP_2002_FemaleMiceTumor
3DosesAllModels_LogLogistic_0.67.(d)
Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V2O5\NTP_2002Cancer\FemaleMiceTumor3DosesAllModels\NTP_2002_FemaleMiceTumor
3DosesAllModels_LogLogistic_0.67.pit
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0208333
intercept = 2.36858
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.14
intercept -0.14 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
background 0.0210689 * * *
intercept 2.14725 * * *
slope 1 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -62.3295 3
Fitted model -62.7264 2 0.79377 1 0.373
Reduced model -98.9486 1 73.2384 2 <.0001
AIC: 129.453
Goodness of Fit
Dose
0.0000
0.2000
0.3800
Chi^2 = 0.
Est. Prob.
0.0211
0.6391
0.7698
.80 d.f . =
Expected
1 . Oil
30.036
36 .952
Observed
1.000
32 .000
35.000
Size
48
47
48
Scaled
Residual
-0.011
0.596
-0.669
1 P-value = 0.3699
Benchmark Dose Computation
Specified effect = 0.67
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.23715
BMDL = 0.160575
195
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
16:2909/21 2010
C.5. EXTRAPOLATION METHOD AND INHALATION CANCER SLOPE FACTOR
As explained in Chapter 5, linear extrapolation was applied in this assessment and BMCL
at the calculated BMR was regarded as POD. The inhalation cancer slope factor was the upper-
bound estimation of risk and calculated as BMR/BMCL. It was used to estimate the lifetime lung
tumor risk in human for vanadium pentoxide exposure through inhalation. Data are summarized
in Table C-l.
C.6. MSW TIME-TO-TUMOR MODELING OF THE INHALATION
CARCINOGENESIS DATA SETS
Multistage Weibull Time-to-tumor modeling was also used to model these data sets
because the survival curves reported by NTP (2002) showed a high percentage (22-46%) of
deaths at the end of 2-year studies across the different concentrations groups. The individual
animal data was obtained through the NTP website and summarized according to concentration
and week of death (Table C-l 1).
Based on the results, MSW time-to-tumor model was not recommended to derive cancer
slope factor for these data sets because of two reasons:
• The visual fit near the 1st non-control concentration was not adequate.
• The parameters, including c, beta_0 and beta_l, were very close to the initial values,
which suggested the model was non-convergence.
Lung Tumor Data Set 1: Combined Incidence of Alveolar/Bronchiolar Adenomas and
196
DRAFT - DO NOT CITE OR QUOTE
-------
Carcinomas in Male Mice, NTP (2002).
Table C-11. Grouped Data for MSW Time-to-tumor Modeling; B6C3Fi Male Mice Exposed to
Vanadium Pentoxide by Inhalation for 2 Years (NTP, 2002).
Human
Equivalent
Concentration
(mg/m3)
0.0
0.21
Week of
Death
2
26
72
89
92
95
97
98
99
101
102
104
104
105
3
64
75
76
81
83
86
87
91
93
94
95
101
103
104
104
105
105
Response Category for
Alveolar/Bronchiolar
Adenoma and/or Carcinoma
b
C
C
C
C
C
C
C
C
C
C
C
C
I
C
C
C
I
I
I
I
C
I
I
I
I
C
I
I
I
C
C
I
Number of
Animals
1
1
1
1
1
1
1
2
1
1
1
6
1
31
3
1
1
1
1
1
1
1
1
1
2
1
1
1
8
1
11
13
197
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-11. Grouped Data for MSW Time-to-tumor Modeling; B6C3Fi Male Mice
Exposed to Vanadium Pentoxide by Inhalation for 2 Years (NTP, 2002). (Continued)
Human
Equivalent
Concentration
(mg/m3)
0.42
0.81
Week of
Death
36
39
40
45
50
59
70
72
76
82
83
91
92
95
95
99
101
101
104
105
105
36
68
77
79
79
81
81
87
91
93
98
99
101
104
104
105
105
Response Category for
Alveolar/Bronchiolar
Adenoma and/or Carcinoma
b
C
C
I
I
C
C
C
I
I
C
I
I
I
I
C
I
I
C
I
I
C
C
I
I
I
C
C
I
C
I
I
I
C
I
I
C
I
C
Number of
Animals
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
7
17
7
1
1
1
1
1
1
1
2
1
1
1
1
2
7
3
16
9
198
DRAFT - DO NOT CITE OR QUOTE
-------
a Concentration was the original doses reported in the publication from NTP(2002). No adjustment was applied.
b Categories of response: "C"=Neither carcinoma nor adenoma was detected when the subject was removed from the
study due to scheduled sacrifice or unscheduled death; "I"= carcinoma and/or adenoma were detected when the
subjected was removed from the study due to scheduled sacrifice or unscheduled death.
MSW Time-to-tumor output
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: V2O5_NTP_MaleMiceHEC_Polyl.(d)
V2O5_NTP_MaleMiceHEC
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 200
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
t_0 = 0
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 1.05882
t_0 = 0 Specified
beta_0 = 0.0054829
beta_l = 0.0193401
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -c ~t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
beta_0 beta_l
beta_0 1 -0.61
beta_l -0.61 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
c 1 NA
beta_0 0.00720175 0.00201449 0.00325342 0.0111501
beta_l 0.0252605 0.00794841 0.00968191 0.0408391
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -101.116 3 208.233
Data Summary
CONTEXT
C F I U Total Expected Response
DOSE
0 28 0 22 0 50 25.86
0.21 8 0 42 0 50 35.33
0.42 7 0 43 0 50 41.53
0.81 7 0 43 0 50 46.15
Benchmark Dose Computation
199
DRAFT - DO NOT CITE OR QUOTE
-------
Risk Response
Risk Type
Specified effect =
Confidence level =
Time
BMD =
BMDL =
BMDU =
Incidental
Extra
0.71
0.9
104
0.471196
0.390136
0.716564
.fr
2
Q_
Incidental Risk: V2O5_NTP_MaleMiceHEC_Poly1
Dose = 0.00 Dose = 0.21
0 20 40 60 80 100
Time
.fr
2
o_
CO
0 20 40
Time
80 100
.fr
2
o_
0 20 40 60 80 100
Time
.fr
2
o_
0 20 40
80 100
Time
200
DRAFT - DO NOT CITE OR QUOTE
-------
Lung Tumor Data Set 2: Combined Incidence of Alveolar/Bronchiolar Adenoma and
Carcinoma in Female Mice, NTP (2002).
Table C-12. Grouped Data for MSW Time-to-tumor Modeling; B6C3Fi Female Mice Exposed to
Vanadium Pentoxide by Inhalation for 2 Years (NTP, 2002).
Human
Equivalent
Concentration
(mg/m3)
0.00
0.20
Week of
Death
2
26
72
89
92
95
97
98
99
101
102
104
104
105
3
64
75
76
81
83
86
87
91
93
94
95
101
103
104
104
105
105
Response Category for
Alveolar/Bronchiolar
Adenoma and/or Carcinoma
b
C
C
C
C
C
C
C
C
C
C
C
C
I
C
C
C
I
I
I
I
C
I
I
I
I
C
I
I
I
C
C
I
Number of
Animals
1
1
1
1
1
1
1
2
1
1
1
6
1
31
3
1
1
1
1
1
1
1
1
1
2
1
1
1
8
1
11
13
201
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-12 . Grouped Data for MSW Time-to-tumor Modeling; B6C3Fi Female Mice
Exposed to Vanadium Pentoxide by Inhalation for 2 Years (NTP, 2002). (Continued)
Human
Equivalent
Concentration
(mg/m3)
0.38
0.73
Week of
Death
36
39
40
45
50
59
70
72
76
82
83
91
92
95
95
99
101
101
104
105
105
36
68
77
79
79
81
81
87
91
93
98
99
101
104
104
105
Response Category for
Alveolar/Bronchiolar
Adenoma and/or Carcinoma
b
C
c
I
I
c
c
c
I
I
c
I
I
I
I
c
I
I
c
I
I
c
c
I
I
I
c
c
I
c
I
I
I
c
I
I
c
I
Number of
Animals
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
7
17
7
1
1
1
1
1
1
1
2
1
1
1
1
2
7
3
16
202
DRAFT - DO NOT CITE OR QUOTE
-------
105
C
a Concentration was the original one reported in the publication from NTP(2002). No adjustment was applied.
b Categories of response: "C"=Neither carcinoma nor adenoma was detected when the subject was removed from the
study due to scheduled sacrifice or unscheduled death; "I"= carcinoma and/or adenoma were detected when the
subjected was removed from the study due to scheduled sacrifice or unscheduled death.
MSW Time-to-tumor output
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: V2O5_NTP_FemaleMiceHEC_Polyl.(d)
V2O5_NTP_FemaleMiceHEC
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 200
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
t 0 0
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 1.125
t_0 = 0 Specified
beta_0 = 0.000224032
beta_l = 0.0144984
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c
beta_
beta
1
-0.93
-1
beta_0
-0.93
1
0.93
beta_l
-1
0.93
1
Variable
c
beta_0
beta 1
Estimate
1.12496
0.000224043
0.0144991
Parameter Estimates
Std. Err.
0.770137
0.000862599
0.0511302
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.384483 2.6344
-0.00146662 0.00191471
-0.0857143 0.114712
Log(likelihood) # Param AIC
Fitted Model -109.298 3 224.597
DOSE
0
0.2
0.38
0.73
49
18
15
18
Data Summary
CONTEXT
F I
1
32
35
32
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Specified effect = 0.67
U Total Expected Response
50
50
50
50
1.94
19.95
30.09
42 .08
203
DRAFT - DO NOT CITE OR QUOTE
-------
Confidence level =
Time
BMD =
BMDL =
BMDU =
0.9
104
0 .411511
0.339589
0.516352
Incidental Risk: V2O5_mP_FemaleMiceHEC_Poly1
Dose = 0.00
Dose = 0.20
0 20 40 60 80 100
Time
q
d
\ I I I I r
0 20 40 60 80 100
Time
Dose= 0.38
Dose = 0.73
20 40 60 80 100
Time
\ I I I I T
0 20 40 60 80 100
Time
204
DRAFT - DO NOT CITE OR QUOTE
------- |