DRAFT - DO NOT CITE OR QUOTE                    EPA/635/R-11/004A
                                                www.epa.gov/iris
&EPA
           TOXICOLOGICAL REVIEW

                               OF

        VANADIUM PENTOXIDE (V2O5)

                        (CAS No. 1314-62-1)
            In Support of Summary Information on the
            Integrated Risk Information System (IRIS)
                         September 2011
                             NOTICE

This document is an External Review 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.
           This document is a draft for review purposes only and does not constitute Agency policy.
                                        ii           DRAFT - DO NOT CITE OR QUOTE

-------
      CONTENTS - TOXICOLOGICAL REVIEW OF VANADIUM PENTOXIDE
                              (CAS No. 1314-62-1)

LIST OF TABLES	vi
LIST OF FIGURES	viii
LIST OF ACRONYMS AND ABBREVIATIONS	ix
FOREWORD	xii
AUTHORS, CONTRIBUTORS, AND REVIEWERS	xiii

1. INTRODUCTION	1

2. CHEMICAL AND PHYSICAL INFORMATION	3

3. TOXICOKINETICS	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	10
  3.3. METABOLISM	10
  3.4. ELIMINATION AND EXCRETION	10
      3.4.1. Inhalation Exposure	10
      3.4.2. Oral Exposure	11
      3.4.3. Dermal Exposure	11
      3.4.4. Other Routes of Exposure	11
  3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK)
      MODELS	11

4. HAZARD IDENTIFICATION	12
  4.1. STUDIES IN HUMANS-EPIDEMIOLOGY, CASE REPORTS,
      CLINICAL CONTROLS	12
      4.1.1. Oral Exposure	12
      4.1.2. Inhalation Exposure	12
  4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER
      BIOASSAYS IN ANIMALS-ORAL AND INHALATION	36
      4.2.1. Oral Exposure	36
      4.2.2. Inhalation Exposure	38
  4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL,
      INHALATION, INTRAPERITONEAL AND INJECTION	55
      4.3.1. Oral Studies	55
      4.3.2. Inhalation Studies	56
      4.3.3. Intraperitoneal and Injection Studies	57
  4.4. OTHER DURATION - OR ENDPOINT—SPECIFIC STUDIES	59
         This document is a draft for review purposes only and does not constitute Agency policy.
                                   iii          DRAFT - DO NOT CITE OR QUOTE

-------
                             CONTENTS (continued)
  4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE
      MODE OF ACTION FOR PULMONARY FIBROSIS AND CANCER	59
  4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS	60
      4.6.1. Oral	60
      4.6.2. Inhalation	62
      4.6.3. Mode of Action Information	66
  4.7. EVALUATION OF CARdNOGENICITY	68
      4.7.1. Summary of Overall Weight of Evidence	68
      4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence	69
      4.7.3. Mode of Action Information	70
  4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES	70
      4.8.1. Possible Childhood Susceptibility	70
      4.8.2. Possible Gender Differences	70
      4.8.3. Other Susceptible Populations	71

5. DOSE-RESPONSE ANALYSIS	73
  5. LORAL REFERENCE DOSE (RFD)	73
      5.1.1. Choice of Principal Study and Critical Effect with Rationale and
           Justification	73
      5.1.2. Methods of Analysis—Including Models (e.g., PBPK and BMD)	75
      5.1.3. RfD Derivation—Including Application of Uncertainty Factors	77
      5.1.4. Previous RfD Assessment	78
  5.2. INHALATION REFERENCE CONCENTRATION (RFC)	78
      5.2.1. Choice of Principal Study and Critical Effect with Rationale and
           Justification	78
      5.2.2. Methods of Analysis	81
      5.2.3. RfC Derivation—Including Application of Uncertainty Factors	86
      5.2.4. Previous RfC Assessment	87
  5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE (RFD) AND
      THE INHALATION REFERENCE CONCENTRATION (RFC)	87
  5.4. CANCER ASSESSMENT	89
      5.4.1. Choice of Study/Data—with Rationale and Justification	89
      5.4.2. Dose-Response Data	89
      5.4.3. Dose Adjustments and Extrapolation Methods	89
      5.4.4. Inhalation Unit Risk	92
      5.4.5. Oral Cancer Slope Factor	92
      5.4.6. Uncertainties in Cancer Risk Values	93

6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD
  AND DOSE RESPONSE	96
  6.1. HUMAN HAZARD POTENTIAL	96
  6.2. DOSE RESPONSE	100
      6.2.1.Noncancer	100
      6.2.2. Cancer	101

7. REFERENCES	103

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

-------
                         CONTENTS (continued)
APPENDICES:
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
   COMMENTS AND DISPOSITION	A-l

APPENDIX B: BENCHMARK CONCENTRATION MODELING OF
   INHALATION STUDIES IN RATS FROM NTP (2002)	B-l
   B. 1 TREND TESTS IN THE INHALATION DATASETS FOR BMDS
   MODELING	B-l
   B.2 DOSE CONVERSION	B-3
   B.3 BMDS MODELING FOR INHALATION DATASETS	B-5
   B.4 REFERENCES	B-31

APPENDIX C: BENCHMARK CONCENTRATION MODELING OF
   COMBINED LUNG ALVEOLAR AND BRONCHIOLAR TUMOR
   DATASETS FROM NTP STUDIES (2002)	C-l
   C.I BMR	C-l
   C.2 TREND TESTS BEFORE MODELING	C-2
   C.3 DOSE CONVERSION	C-2
   C.4 BMR CALCULATION	C-5
   C.5 BMDS MODELING FOR INHALATION CARCINOGENESIS
   DATASETS	C-5
   C.6 EXTRAPOLATION METHOD AND INHALATION CANCER
   SLOPE FACTOR	C-13
   C.7 MSW TIME-TO-TUMOR MODELING OF THE INHALATION
   CARCINOGENESIS DATASETS	C-13
   C.8 REFERENCES	C-22

APPENDIX D: SUPPLEMENTAL INFORMATION	D-l
   D.I EXPOSURE TO PM2.5 VANADIUM IN AMBIENT AIR	D-l
   D.2 LABORATORY ANIMAL AND IN VITRO STUDIES	D-4
      D.2.1 OTHER ENDPOINT AND DURATION STUDIES	D-4
      D.2.2 MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT
      OF THE MODE OF ACTION FOR PULMONARY FIBROSIS AND
      CANCER	D-13
   D.3 REFERENCES	D-30
        This document is a draft for review purposes only and does not constitute Agency policy.
                               v        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. Studies of Vanadium Pentoxide Exposure: Occupational Case
     Studies, Epidemiology Studies, and Controlled Human Exposure	22
Table 4-2. Hematological results of oral vanadium pentoxide exposure in Wistar
     rats	37
Table 4-3. Body weight gain and lung weights in rats (F344/N) exposed to
     vanadium pentoxide by inhalation for 3 months (values are means ± standard
     error)	39
Table 4-4. Selected hematology parameters in rats (F344/N) exposed to
     vanadium pentoxide by inhalation for 3 months	40
Table 4-5. Incidences of selected nonneoplastic lesions of the lung and nose in
     rats (F344/N) exposed to vanadium pentoxide by inhalation for 3 months	42
Table 4-6. Body weight gain and lung weights in mice (B6C3Fi) exposed to
     vanadium pentoxide by inhalation for 3 months (values are means ± standard
     error)	45
Table 4-7. Incidences of selected nonneoplastic lesions of the lung in mice
     (B6C3Fi) exposed to vanadium pentoxide by inhalation for 3 months	45
Table 4-8. Selected nonneoplastic lesions of the respiratory system in F344/N
     rats exposed to paniculate aerosols of vanadium pentoxide for 2 years	48
Table 4-9. Incidences of respiratory tumors in rats exposed to vanadium
     pentoxide in a 2-year inhalation study	50
Table 4-10. Selected nonneoplastic lesions of the respiratory system in B6C3Fi
     mice exposed to vanadium pentoxide in a 2-year inhalation study	51
Table 4-11. Incidences of respiratory tumors in B6C3Fi mice exposed to
     vanadium pentoxide in the 2-year inhalation study	53
Table 4-12. Lung tumor multiplicity in MCA-treated mice exposed to vanadium
     pentoxide by pharyngeal aspiration  (Rondini et al., 2010)	54
Table 4-13. Summary  of noncancer results of repeat-dose studies for oral
     exposure of experimental animals to vanadium pentoxide	61
Table 4-14. Summary  of the noncancer results for the subchronic and chronic
     inhalation exposure studies of vanadium pentoxide toxicity in experimental
     animals	64
Table 5-1. Hematological results of oral vanadium pentoxide exposure in rats	75
Table 5-2. Human equivalent dose conversion by BW34 for RfD derivation	76
Table 5-3. Selected nonneoplastic lesions of the respiratory system in F344/N
     rats exposed to paniculate aerosols of vanadium pentoxide for 2 years	82
Table 5-4. Human equivalent concentrations of vanadium pentoxide in the 2-year
     inhalation studies in F344/N rats	83
Table 5-5. Candidate PODs for vanadium pentoxide derived from NTP studies
     (2002) through BMDS modeling	84

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

-------
                             LIST OF TABLES (continued)
Table 5-6. Benchmark modeling results for incidence of larynx chronic
     inflammation in female rats	85
Table 5-7. Benchmark modeling results for incidence of larynx respiratory
     epithelium, epiglottis, hyperplasia in female rats	86
Table 5-8. Incidences of respiratory tumors in B6C3Fi mice exposed to
     vanadium pentoxide in the 2-year inhalation study	90
Table 5-9. Human equivalent concentrations of vanadium pentoxide in the 2-year
     inhalation studies	90
Table 5-10. BMDS modeling results for combined incidence of
     alveolar/bronchiolar adenoma and carcinoma in male mice	92
Table 5-11. Summary of uncertainty in the vanadium  pentoxide cancer risk
     assessment	95
Table B-l.  Candidate PODs for vanadium pentoxide derived from NTP studies
     (2002) through BMDS modeling	B-2
Table B-2.  Trend tests on the selected datasets from the 2-year inhalation studies
     in rats	B-2
Table B-3.  Average lifetime animal body weight of rats in the 2-year inhalation
     studies of vanadium pentoxide	B-4
Table B-4.  RDDRs for various concentration/sex groups in the 2-year inhalation
     studies of vanadium pentoxide	B-4
Table B-5.  Human equivalent concentrations of vanadium pentoxide in the 2-year
     inhalation studies	B-5
Table B-6.  Benchmark modeling results for incidence of lung alveolar epithelium
     hyperplasia in male rats	B-6
Table B-7.  Benchmark modeling results for incidence of lung chronic active
     inflammation in male rats	B-8
Table B-8.  Benchmark modeling results for incidence of larynx chronic
     inflammation in male rats	B-ll
Table B-9.  Benchmark modeling results for incidence of larynx respiratory
     epithelium, epiglottis, hyperplasia in male rats	B-13
Table B-10. Benchmark modeling results for incidence of nose goblet cell,
     respiratory epithelium, hyperplasia in male rats	B-16
Table B-ll. Benchmark modeling results for incidence of lung alveolar
     epithelium hyperplasia in female rats	B-l9
Table B-12. Benchmark modeling results for incidence of lung chronic active
     inflammation in female rats	B-21
Table B-13. Benchmark modeling results for incidence of larynx chronic
     inflammation in female rats	B-24
Table B-14. Benchmark modeling results for incidence of larynx respiratory
     epithelium, epiglottis, hyperplasia in female rats	B-26
           This document is a draft for review purposes only and does not constitute Agency policy.
                                       vii           DRAFT - DO NOT CITE OR QUOTE

-------
                            LIST OF TABLES (continued)
Table B-15.  Benchmark modeling results for incidence of nose goblet cell,
     respiratory epithelium, hyperplasia in female rats	B-29
Table C-l.  Summary of candidate PODs and cancer slope factors	C-2
Table C-2.  Trend tests on the inhalation carcinogenesis datasets in mice from the
     2-year inhalation studies	C-2
Table C-3.  Average lifetime animal body weight of mice for 2-year inhalation
     studies of vanadium pentoxide	C-3
Table C-4.  RDDRs for different concentration/sex group of mice in the 2-year
     inhalation studies of vanadium pentoxide	C-4
Table C-5.  Human equivalent concentrations of vanadium pentoxide in the 2-year
     inhalation studies	C-4
Table C-6.  BMR estimation for male and female mice datasets	C-5
Table C-7.  BMDS modeling results for combined incidence of
     alveolar/bronchiolar adenoma and carcinoma in male mice	C-6
Table C-8.  BMDS modeling results for combined incidence of
     alveolar/bronchiolar adenoma and carcinoma in male mice after dropping the
     highest concentration	C-8
Table C-9.  BMDS modeling results for combined incidence of
     alveolar/bronchiolar adenomas and carcinomas with all four concentrations
     in female mice	C-ll
Table C-10.  BMDS modeling results for combined incidence of
     alveolar/bronchiolar adenoma and carcinoma in female mice after dropping
     the highest concentration	C-ll
Table C-ll.  Grouped data for MSW time-to-tumor modeling; B6C3Fi male mice
     exposed to vanadium pentoxide by inhalation for 2 years	C-14
Table C-12.  Grouped data for MSW time-to-tumor modeling; B6C3Fi female
     mice exposed to vanadium pentoxide by inhalation for 2 years	C-18
Table D-1.  Genotoxicity data following exposure to vanadium pentoxide	D-20
                                  LIST OF FIGURES
Figure 2-1. Vanadium pentoxide structure	4
Figure D-l. Genomics of "V^Os-induced bronchitis	D-26
          This document is a draft for review purposes only and does not constitute Agency policy.
                                      viii          DRAFT - DO NOT CITE OR QUOTE

-------
                   LIST OF ACRONYMS AND ABBREVIATIONS
AIC
ATSDR
BAL
BALT
BMC
BMCL
BMCL[HEc]

BMD
BMDL
BMDS
BMR
bw
CASRN
cc
CD
CERCLA
CBMN Cyt
CNS
cu.m
DAF
EOF
ERK
FEF
FIFRA
FVC
g
GI
GSD
HED
HEC
Hgb
i.m.
i.p.
IRIS
IUR
i.v.
kg
L
LOAEL
LOAEL[ADJ]
LOAEL[HEC]
LOD
Akaike Information Criterion
Agency for Toxic Substances and Disease Registry
bronchiolar lavage
bronchus-associated lymphoid tissue
benchmark concentration
benchmark concentration (lower limit)
benchmark concentration (lower limit) adjusted for dosimetric differences
across species to humans
benchmark dose
benchmark dose (lower limit)
benchmark dose software
benchmark response
body weight
Chemical Abstracts Service Registry Number
cubic centimeters
caesarean delivered
Comprehensive Environmental Response, Compensation and
Liability Act of 1980
dynamic lung compliance
cytokinesis-block micronucleus cytome
central nervous system
cubic meter
dosimetric adjustment factor
epidermal growth factor
extracellular  signal-regulated protein kinase
forced expiratory flow
Federal Insecticide, Fungicide,  and Rodenticide Act
forced vital capacity
grams
gastrointestinal
geometric standard deviation
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
limit of detection
          This document is a draft for review purposes only and does not constitute Agency policy.
                                      ix          DRAFT - DO NOT CITE OR QUOTE

-------
LOH
m
MAP
MCA
MCL
MF
mg
mg/kg
mg/L
MMAD
MMP
MOA
MPO
MRL
MSW
MTD
NAAQS
NMMAPS
NOAEL
NOAEL[ADj]
NOAEL[HEc]
NOEL
NTP
OSF
OSHA
PBPK
PDGF
PFT
PHA
PM
PMN
POD
ppb
ppm
PPRTV
PWM
RBC
RCRA
RDDR
RfC
RfD
RGDR
RL
ROFA
RSV
RV
SCE
SD
SDNN
loss of heterozygosity
meter
mitogen-activated protein
3-methylcholanthrene
maximum contaminant level
modifying factor
milligram
milligrams per kilogram
milligrams per liter
median aerodynamic diameter
matrix metalloproteinases
mode of action
myeloperioxidase
minimal risk level
Multistage Weibull
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
National Toxicology Program
oral slope factor
Occupational Safety and Health Administration
physiologically based pharmacokinetic
platelet-derived growth factor
pulmonary function test
phytohemagglutinin responsiveness
particulate matter
polymorphonuclear leukocyte
point of departure
parts per billion
parts per million
Provisional Peer Reviewed Toxicity Value
pokeweed mitogen responsiveness
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)
total pulmonary resistance
residual oil fly ash
respiratory syncytial virus
residual volume
sister chromatid exchange
standard deviation
standard deviation of the normal-to-normal intervals
           This document is a draft for review purposes only and does not constitute Agency policy.
                                       x           DRAFT - DO NOT CITE OR QUOTE

-------
SDWA             Safe Drinking Water Act
SE                 standard error
sq.cm.              square centimeters
TLV                threshold limit value
TWA               time-weighted average
TSCA              Toxic Substances Control Act
UF                 uncertainty factor
ug                  microgram
umol                micromoles
VOC                volatile organic compound
           This document is a draft for review purposes only and does not constitute Agency policy.
                                       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).
           This document is a draft for review purposes only and does not constitute Agency policy.
                                        xii           DRAFT - DO NOT CITE OR QUOTE

-------
                  AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR

Maureen R. Gwinn, Ph.D., DABT
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, DC 20460

IRIS CO-LEAD/AUTHOR

Kathleen Newhouse, M.S.
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, DC 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, DC 20460

Sanjivani Diwan, Ph.D.*
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, DC 20460

Barbara Glenn, Ph.D.
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, DC 20460

Kathleen Raffaele, Ph.D.
Office of Solid Waste and Emergency Response
US Environmental Protection Agency
Washington, DC 20460

Suryanarayana V. Vulimiri, D.V.M., Ph.D., DABT
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, DC 20460

*Former EPA employee

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

-------
CONTRIBUTORS

James C. Ball, Ph.D.
National Center for Environmental Assessment
US Environmental Protection Agency
Washington, DC 20460
           This document is a draft for review purposes only and does not constitute Agency policy.
                                       xiv          DRAFT - DO NOT CITE OR QUOTE

-------
 1                                     1. INTRODUCTION

 2          This document presents background information and justification for the Integrated Risk
 3    Information System (IRIS) Summary of the hazard and dose-response assessment of vanadium
 4    pentoxide. IRIS Summaries may include oral reference dose (RfD) and inhalation reference
 5    concentration (RfC) values for chronic and other exposure durations, and a carcinogenicity
 6    assessment.
 7          The RfD and RfC, if derived, provide quantitative information for use in risk assessments
 8    for health effects known or assumed to be produced through a nonlinear (presumed threshold)
 9    mode of action (MO A). The RfD (expressed in units of mg/kg-day) is defined as  an estimate
10    (with uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
11    population (including sensitive subgroups) that is likely to be without an appreciable risk of
12    deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
13    analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
14    inhalation RfC considers toxic effects for both the respiratory system (portal-of-entry) and for
15    effects peripheral to the respiratory system (extrarespiratory or systemic effects).  Reference
16    values are generally derived for chronic exposures (up to a lifetime), but may also be derived for
17    acute (<24 hours),  short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
18    lifetime) exposure durations, all of which are derived based on an assumption of continuous
19    exposure throughout the duration specified. Unless specified otherwise, the RfD and RfC are
20    derived for chronic exposure duration.
21          The carcinogenicity assessment provides information on the carcinogenic hazard
22    potential of the substance in question, and quantitative estimates of risk from oral  and inhalation
23    exposures may be derived.  The information includes a weight-of-evidence judgment of the
24    likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
25    effects may be expressed.  Quantitative risk estimates may be derived from the application of a
26    low-dose extrapolation procedure.  If derived, the oral slope factor is a plausible upper bound on
27    the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit risk  is a
28    plausible upper bound on the estimate of risk per ug/m3 air breathed.
29          Development of these hazard identification and dose-response assessments for vanadium
30    pentoxide has followed the general guidelines for risk assessment as set forth by the National
31    Research Council (1983).  EPA Guidelines and Risk Assessment Forum technical panel reports
32    that may have been used in the development of this assessment include the following:
33    Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S. EPA, 1986b), Guidelines
3 4   for Mutagenicity Risk Assessment (U.S. EPA, 1986a), Recommendations for and Documentation
35    of Biological Values for Use in Risk Assessment (U.S. EPA, 1988), Guidelines for
36    Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Interim Policy for Particle Size and
1>1    Limit Concentration Issues in Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1        DRAFT - DO NOT CITE OR QUOTE

-------
 1   Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA,
 2   1994b), Use of the Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995),
 3   Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for
 4   Neurotoxicity Risk Assessment (U.S. EPA, 1998), Science Policy Council Handbook: Risk
 5   Characterization (U.S. EPA, 2000b), Benchmark Dose Technical Guidance Document (U.S.
 6   EPA, 2000a), Supplementary Guidance for Conducting Health Risk Assessment of Chemical
 1   Mixtures (U.S. EPA, 2000c), A Review of the Reference Dose and Reference Concentration
 8   Processes (U.S. EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a).
 9   Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
10   (U.S. EPA, 2005b), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006b), A
11   Framework for Assessing Health Risks of Environmental Exposures to Children (U.S. EPA,
12   2006a), and Recommended Use of Body Weight 3/4 as the Default Method in Derivation of the
13   Oral Reference Dose (U.S. EPA, 2011).
14          The literature search strategy employed for vanadium pentoxide was based on the
15   chemical name, the Chemical Abstracts Service Registry Number (CASRN), and multiple
16   common synonyms. Any pertinent scientific information submitted by the public to the IRIS
17   Submission Desk was also considered in the development of this document. Primary, peer-
18   reviewed literature identified through August 2011 was included where that literature was
19   determined to be critical to the assessment. The relevant literature included publications on
20   vanadium pentoxide, which were identified through Toxicology Literature Online (TOXLINE),
21   PubMed, the Toxic Substance Control Act Test Submission Database (TSCATS), the Registry of
22   Toxic Effects of Chemical Substances (RTECS), the Chemical Carcinogenesis Research
23   Information System (CCRIS), the Developmental and Reproductive Toxicology/Environmental
24   Teratology Information Center (DART/ETIC), the Hazardous Substances Data Bank (HSDB),
25   the Genetic Toxicology Data Bank (GENE-TOX), Chemical abstracts, and Current Contents.
26   Other peer-reviewed information, including health assessments developed by other
27   organizations, review articles, and independent analyses of the health effects data were retrieved
28   and may be included in the assessment where appropriate.
29          Portions of this document were developed under a Memorandum of Understanding with
30   the Agency for Toxic Substances and Disease Registry (ATSDR) as part of a collaborative effort
31   in the development of human health toxicological assessments.
32
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               2        DRAFT - DO NOT CITE OR QUOTE

-------
 1                      2. CHEMICAL AND PHYSICAL INFORMATION

 2          As an element, vanadium (V) exists in several oxidation states, from -1 to +5.  Vanadium
 3    (Chemical Abstracts Service Registry Number [CASRN] 7440-62-2) is a soft silver-gray metal
 4    commonly found in ores, tars, coals, and oils and is used as an alloy in steel (WHO, 1988). The
 5    focus of this Toxicological Review is vanadium pentoxide (CASRN 13 14-62-1), but a short
 6    description of the chemistry of vanadium and related compounds is given below for clarification.
 7          The chemistry of vanadium is complex; the valence state of vanadium can shift
 8    depending on pH and other factors.  In the body, there is an interconversion of two oxidation
 9    states of vanadium, the tetravalent form, vanadyl (V4+), and the pentavalent form, vanadate
10    (V5+). Vanadate is considered more toxic than vanadyl because vanadate is reactive with several
1 1    enzymes and is a potent inhibitor of the Na+K+-ATPase of plasma membranes (Patterson et al.,
12    1986: Harris etal..  1984).
13          Generally, the V3+ and V4+ valence states predominate in body tissues while V5+
14    predominates in plasma (IPCS, 2001). Vanadium pentoxide (V^Os), sodium metavanadate
1 5    (NaVOs), sodium orthovanadate (Nas VO/t), and ammonium metavanadate (NH/iVOs) all contain
16    vanadium in the +5 oxidation state.  Of these compounds, V2Os is the only compound that is
17    covalently bonded.
18          The physicochemical properties of vanadium compounds differ, which determines their
19    solubility under different pH conditions and their accessibility and availability in biological
20    systems [reviewed in Assem and Levy (2009)1. An acidic pH favors the tetravalent state (V4+)
21    keeping it as vanadyl, while the pentavalent state (V5+) as vanadate (Crans et al., 2004) is
22    prevalent under alkaline conditions. In the case of oral ingestion, vanadium compounds are
23    exposed to a range of pH solutions in the digestive tract starting at the stomach (pH typically
24    between 1 and 3.5), followed by the small intestine (pH around 8). Bruyere et al. (1999) state
25    that at pH between 1.3 and 3.3, the predominant form of vanadium is VO2+ and at higher pH, the
26    form is VO(OH)3. When the pH is high, V5+ (e.g., VO/^OtTh) predominates and polymerized
27    vanadium is predominant (HnVioO29(n  6)) (Bruyere et al., 1999). At physiological pH, vanadium
28    compounds have been shown to exist in monomeric tetravalent [VO(OH)3] and dimeric
29    [(VO)2(OH)s]  forms, as well as pentavalent (H2VO4 ) forms [reviewed in Assem and Levy
30    (2009)].  Thus, the valence of a vanadium compound will depend  on the pH.
3 1          The solubility of different vanadium compounds in water between 20 and 25°C differs
32    among valences as shown in Table 2-1 (HSDB, 2008: WHO. 2001). Elemental vanadium (V°) is
33    insoluble in water.  The tetravalent (V4+) compound vanadyl sulfate (VOSO4) is highly soluble
34    534.64 g/L (Rahman and Skyllas-Kazacos, 1998), while the pentavalent (V5+) vanadium
35    compounds such as vanadium pentoxide CV^Os) is less soluble (8 g/L). Other vanadium
36    compounds such as sodium metavanadate (NaVOs), sodium orthovanadate (NasVO4), and
37    ammonium metavanadate (NtLjVOs) have solubilities of 21 1 g/L, 100 g/L, and 58 g/L,
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               3        DRAFT - DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
respectively (WHO, 2001). Furthermore, the rate of dissolution (which is distinct from solubility
—an equilibrium or thermodynamic parameter) of various vanadium compounds can vary,
resulting in different concentrations of specific forms of vanadium.  The various forms of
vanadium may be absorbed differently, which could result in different physiological effects. For
example, V5+ compounds can mimic phosphate and inhibit phosphatases (Assem and Levy,
2009).
           Table 2-1. Valence states and water solubility of various vanadium
           compounds
Vanadium compound
Vanadium
Vanadium pentoxide
Sodium m-vanadate
Sodium o-vanadate
Ammonium m-vanadate
Vanadium oxytrichloride
Vanadyl sulfate
Vanadium tetrachloride
Vanadyl oxydichloride
Vanadium trioxide
Formula
V
V2O5
NaVO3
Na3VO4
NH4VO3
VOC13
VOS04
VCL,
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
Valency
0
+5
+5
+5
+5
+5
+4
+4
+3
+3
Solubility (g/L) at 20-25°C
(HSDB, 2008; WHO, 2001)
Insoluble
8
211
100
58 (WHO. 2001): 5.2 at 15°C (HSDB.
2008)
Soluble, decomposes in presence of
moisture into vanadic acid and HC1.
535at20°C
(Rahman and Sky lias -Kazacos. 1998)
Decomposes
Decomposes
Slightly soluble
     Source: Adapted from Assem and Levy (2009).
7          This Toxicological Review focuses exclusively on vanadium pentoxide (V^Os, CASRN
8    13 14-62-1) (Figure 2-1), the most common form of vanadium used commercially. Vanadium
9    pentoxide exists in the pentavalent state as a yellow-red powder (Table 2-2) (OSHA, 2007).
                           Figure 2-1.  Vanadium pentoxide structure.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               4        DRAFT - DO NOT CITE OR QUOTE

-------
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
V2O5
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)
10 g/L (20°C)
Soluble in alkalis, concentrated acids, insoluble in alcohol
0 mmHg
3.4g/cm3
Not applicable, noncombustible
1 ppm =7.44 mg/m3
1 mg/m3 = 0.134 ppm
Reference

OSHA (2007)
CAS
CAS

OSHA (2007):
O'Neil (2001): 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
    This document is a draft for review purposes only and does not constitute Agency policy.
                                    5        DRAFT - DO NOT CITE OR QUOTE

-------
 1                                     3. TOXICOKINETICS

 2           The toxicokinetics of vanadium pentoxide has been investigated in the few studies
 3    described below and as reviewed by Cooper (2007) The toxicokinetics of several other vanadium
 4    compounds has been evaluated in animal models and is reviewed elsewhere (Mukherjee et al.,
 5    2004; Barceloux, 1999; Sabbioni etal., 1996).  Vanadium pentoxide is rapidly absorbed
 6    following inhalation and oral exposures through the lungs and the gastrointestinal tract,
 7    respectively, although the amount absorbed through the gastrointestinal tract is low. Laboratory
 8    animal  studies show that vanadium pentoxide is distributed primarily to the bone, lungs, liver,
 9    and kidney following inhalation  and oral exposures.  Elimination of vanadium pentoxide, which
10    has been studied only following  inhalation exposure, occurs primarily through the urine.

11    3.1.  ABSORPTION

12    3.1.1.  Inhalation Exposure
13           Several occupational studies indicate that absorption can occur in humans following
14    inhalation  exposure. An increase in urinary vanadium levels was found in workers
15    occupationally exposed to <1 ppm (<7.44 mg/m3) of vanadium compounds, including vanadium
16    pentoxide  (Orris etal.. 1983: Kiviluoto et al.. 1981a: Gylseth et al..  1979: Lewis. 1959). with the
17    majority excreted in urine within 1 day after long-term or moderate  exposure to vanadium dust
18    (Kiviluoto et al., 1981a). The vanadium concentration in serum was higher than the
19    nonoccupationally exposed controls following exposure to vanadium pentoxide dust (Kiviluoto
20    etal.. 1981b).
21           Indirect evidence for absorption of vanadium in animals is indicated in studies involving
22    inhalation  exposure or intratracheal  administration. In rats and mice exposed to 0.28-2.2 mg
23    vanadium/m3 as vanadium pentoxide1 for 16 days or 2 years (6 hours/day, 5 days/week),
24    marginal increases in blood vanadium levels were observed, suggesting that vanadium pentoxide
25    was poorly absorbed or rapidly cleared from the blood (Dill et al.. 2004: NTP, 2002). In the
26    2-year studies by NTP (2002), the increases in blood vanadium levels were concentration-
27    related, although not statistically significant. Intratracheal studies suggest that vanadium
28    pentoxide  is readily absorbed through the lungs.  The greatest absorption of a radioactive dose,
      AR
29     V, was found to occur 5 minutes after administration in albino rats (gender not specified)
30    (Roshchin et al.. 1980). Most of the vanadium, that is, 80 and 85%  of the tetravalent (V4+) and
31    pentavalent (V5+) forms of vanadium, respectively, cleared from the lungs 3 hours after
32    intratracheal  exposure in male albino rats (Edel and Sabbioni, 1988).  After 3 days, 90% of
      1Many 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.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 6        DRAFT - DO NOT CITE OR QUOTE

-------
 1    vanadium pentoxide was eliminated from the lungs of female rats following intratracheal
 2    instillation (Conklin et al., 1982). In an intratracheal instillation study in female Fischer rats,
 3    50% was cleared in 18 minutes, and the rest within a few days (Rhoads and Sanders, 1985).
 4          Wallenborn et al. (2007) analyzed the metal components of a complex particulate matter
 5    (PM) mixture and tracked the absorption of the various metals in different tissues following  a
 6    single intratracheal dose in rats. Healthy male Wistar rats were instilled with a single
 7    intratracheal dose of combustion-derived PM containing a moderate amount of transition metals,
 8    including vanadium. The composition of vanadium in the PM was 62.95  ug/mg, with
 9    7.18 ug/mg in the water soluble fraction, 26.50 ug/mg in the acid-soluble fraction, and 29.27
10    Ug/mg in the insoluble fraction. According to calculations, of the 196.63  ug/rat of vanadium
11    instilled (theoretical), 110.32 ug/rat was measured in the lung 4 hours postinstillation and
12    62.76  ug/rat was measured in the lung 24 hours postinstillation. In the plasma and lung,
13    vanadium was significantly elevated 4 hours postinstillation (130,000 ng V/g lung tissue, and
14    350 ng V/g plasma) compared to 24 hours (60,000 ng V/g lung tissue and 110 ng V/g plasma,
15    respectively) suggesting rapid uptake of water-soluble vanadium.  The vanadium component of
16    PM was tracked to the lung, plasma, heart, and liver and was significantly increased compared to
17    controls at both 4 and 24 hours postinstillation compared to saline controls.  This study enabled
18    changes in component metals of a complex mixture to be detected in various organs and provides
19    evidence that metals can dissociate from PM and translocate to various target organs, depending
20    on solubility (Wallenborn et al.. 2007).

21    3.1.2.  Oral Exposure
22          No studies were available in the published literature regarding the rate and  extent of
23    absorption in humans after oral exposure to vanadium pentoxide.
24          Data suggest that vanadium is rapidly absorbed but the level of absorption  through the
25    gastrointestinal tract of animals is low. Vanadium was reported in tissues and urine of male
26    albino rats within hours after a single oral dose (Edel and Sabbioni, 1988), suggesting that it is
27    rapidly absorbed. Young rats that consumed vanadium in drinking water and feed were found to
28    have higher vanadium levels in tissue samples 21 days after birth than they did 115 days after
29    birth (Edel etal., 1984). The data suggest a higher absorption of vanadium in young animals due
30    to a greater nonselective permeability of the undeveloped intestinal barrier.  Thus,  age of the
31    rodents appears to be important in determining the absorption of vanadium in the gastrointestinal
32    tract.  Less than 0.1% of an intragastric dose was detectable in the blood of albino  rats at
33    15 minutes postexposure, and less than 1% at 1 hour (Roshchin et al., 1980). Similarly, only
34    2.6% of an orally administered radiolabeled dose of vanadium pentoxide was absorbed 3 days
35    after exposure in female Fischer rats (Conklin etal., 1982).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1        DRAFT - DO NOT CITE OR QUOTE

-------
 1    3.1.3.  Dermal Exposure
 2          No specific studies were available in the published literature regarding absorption in
 3    humans or animals after dermal exposure to vanadium pentoxide. However, absorption by this
 4    route, however, is generally considered to be very low (WHO, 1988). Because vanadium is a
 5    metal with low solubility, absorption through the skin is thought to be minimal.

 6    3.1.4.  Other Routes of Exposure
 7          No studies were available in the published literature regarding the nature and extent of
 8    absorption in humans through other routes of exposure to vanadium pentoxide.

 9    3.2. DISTRIBUTION
10          Distribution has been determined from autopsy cases with unknown routes of exposure.
11    Vanadium has been detected in the lung (in 52% of the cases) and intestine (in 16% of the cases)
12    of humans with no known occupational exposure using autopsy data,  as reviewed in Schroeder et
13    al. (1963). In the gastrointestinal tract, vanadium was found primarily in the ileum (37%),
14    cecum (45.1%), sigmoid colon (15.9%), and rectum (26.2%). Most positive samples had 0.01 jig
15    or less per g of tissue. The heart, aorta, brain, kidney, muscle, ovary,  and testes were found to
16    have no detectable vanadium concentrations.

17    3.2.1.  Inhalation Exposure
18          There are limited data on the distribution of vanadium in workers; serum vanadium levels
19    in workers were highest within a day after inhalation exposure followed by a rapid decline  in
20    levels upon cessation of exposure (Kiviluoto et al.,  1981b: Gvlseth etal., 1979). Analytical
21    studies have shown low levels of vanadium in human kidneys and liver, with even less in the
22    brain and heart, and in milk.  Higher levels were detected in hair, bone, and teeth (Byrne and
23    Kosta, 1978).  Inhalation exposure and intratracheal administration studies  have examined the
24    distribution of vanadium in rodents. In F344/N rats chronically exposed to 0.56 or 1.1 mg
25    vanadium/m3 as vanadium pentoxide (6 hours/day, 5 days/week), vanadium lung burdens peaked
26    after 173 days of exposure and declined until 542 days; lung levels never reached steady state
27    (NTP,  2002).  In contrast, lung burdens appeared to reach steady state by exposure day 173 in
28    rats exposed to 0.28 mg vanadium/m3 (NTP, 2002).  Similarly, lung burdens did not reach steady
29    state in B6C3Fi mice exposed to 1.1 or 2.2 mg vanadium/m3 as vanadium pentoxide, 6 hours/day
30    and 5 days/week for 542 days (NTP, 2002). Rather, lung burdens peaked near day 54 and
31    declined through day 535. Steady state was achieved in mice exposed to 0.56 mg vanadium/m3
32    during the first 26 days of exposure.
33          Vanadium is found to have a two-phase lung clearance after a single acute exposure in
34    both male Wistar rats and female Fischer rats (Rhoads and Sanders, 1985; Oberg et al., 1978).
35    The initial phase is rapid with a large percentage of the absorbed dose distributed to most organs
36    and blood 24 hours postexposure, followed by a slower clearance phase.  Vanadium is
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 8        DRAFT - DO NOT CITE OR QUOTE

-------
 1    transported mainly in the plasma. It is found in appreciable amounts in the blood initially and
 2    only at trace levels 2 days after exposure (Roshchin et al., 1980). The pentavalent and
 3    tetravalent forms of vanadium compounds were found to have similar distribution patterns in
 4    male albino Sprague-Dawley rats (Edel and Sabbioni, 1988).  Three hours after exposure to the
 5    pentavalent or tetravalent form, 15-17% of the absorbed dose was found in the lung, 2.8% in the
 6    liver, and 2% in the kidney (Edel and Sabbioni, 1988).  After intratracheal instillation of
 7    pentavalent vanadium, retention  of vanadium was observed in the lungs, liver, kidneys, bone,
 8    testes, and spleen with clearance at different time points postexposure with little to no retention
 9    observed in the stomach, intestines, heart, or trachea (Edel and Sabbioni, 1988).  This
10    distribution is similar to the distribution observed following inhalation and oral exposures.

11    3.2.2. Oral Exposure
12          No studies were available in the published literature regarding distribution in humans
13    after oral exposure to vanadium pentoxide.
14          Acute studies with rats showed the highest vanadium concentration in the bones.  Male
                                                   ,10
15    rats had approximately 0.05% of the administered   V in bones, 0.01% in the liver, and <0.01%
16    in the kidney, blood, testis, or spleen after 24 hours (Edel and Sabbioni, 1988). Other authors
17    who found that the bone had the  greatest concentration of radiolabeled vanadium, followed by
18    the kidney (Roshchin et al., 1980), noted similar findings. Conklin et al. (1982) reported that
19    after 3 days, 25% of the absorbed vanadium pentoxide was detectable in the bones and blood of
20    female Fischer rats.
21          Oral exposure for an intermediate duration produced the highest accumulation of
22    vanadium in the kidney.  In young male Sprague-Dawley rats at 3 weeks of age, the kidneys,
23    heart, and lungs had  the highest levels immediately following exposure (Edel et al., 1984).
24    Vanadium in the kidney, liver, and lung decreased significantly at 115 days of age.  An
25    accumulation in muscle and fat was related to the growing mass of the tissues  with age.  The
26    higher levels of vanadium in the  young rat tissues might be due to the higher retention capacity
27    of the undeveloped tissues, or to  greater permeability of the intestinal wall. Adult rats exposed to
28    5 or 50 ppm vanadium in drinking water for 3 months had the highest vanadium levels in the
29    kidney, followed by  bone, liver, and muscle (Parker and Sharma, 1978).  The retention in bone
30    could have been due to phosphate displacement. All tissue levels plateaued during the third
31    week of exposure. A possible explanation for the initially higher levels in the  kidney during
32    intermediate-duration exposure is the daily excretion of vanadium in the urine. When the
33    treatment is stopped, levels decrease in  the kidney.  At the cessation of treatment, vanadium
34    mobilized rapidly from the liver and slowly from the bones.  Other tissue levels decreased
35    rapidly after oral exposure was discontinued. Thus, vanadium was retained much longer in the
36    bones (Edeletal., 1984: Parker and Sharma. 1978).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 9        DRAFT - DO NOT CITE OR QUOTE

-------
 1          Radike et al. (2002) assessed the distribution of various metals, including vanadium, in
 2    female B6C3Fi mice.  Mice ingested either (1) a metal mixture containing chromium (Cr),
 3    cadmium (Cd), arsenic (As), nickel (Ni) and vanadium (V) in drinking water; or (2) a metal
 4    mixture containing Cr, Cd, As, Ni, and V in NIH-31 feed. In water and feed, the calculated
 5    vanadium concentration in the mixture was 45 ppm and 1.105 ppm, respectively.  Measured
 6    vanadium levels in the small intestine were 10 ppm at 5 weeks and 14 ppm at 8 weeks, and were
 7    significantly higher compared to controls than any other metal constituent in the small intestine.
 8    Vanadium pentoxide is distributed primarily to bone (10-25% of administered oral dose), liver
 9    (~5%), and kidney (-4%). In  addition, vanadium levels in the kidneys and the femur were
10    significantly greater than in controls at 4, 8, 12,  16, and 24 weeks following oral dosing.
11    Vanadium levels in small intestine and kidneys were lower in mice given vanadium as part of a
12    heterogeneous metal mixture in feed versus water (Radike et al.,  2002).

13    3.2.3.  Dermal Exposure
14          No studies were available in the published literature regarding distribution in humans or
15    animals after dermal exposure to vanadium.

16    3.2.4.  Other Routes of Exposure
17          After intraperitoneal administration to rats, vanadium is distributed to all organs. After
18    24 hours, the highest concentrations were found in the bones and kidney, although initial levels
19    were highest in the kidney (Roshchin et al., 1980; Sharmaetal.,  1980).

20    3.3. METABOLISM
21          Vanadium is an element, and as such, is not metabolized. In the oxygenated blood, it
22    circulates as a polyvanadate (isopolyanions containing pentavalent vanadium) but in tissues, it is
23    retained mainly as the vanadyl cation (cationic form of tetravalent vanadium). Depending on the
24    availability of reducing equivalents (such as reduced glutathione [GSH], NADPH, and NADH)
25    and oxygen, vanadium can be reduced, reoxidized, or undergo redox cycling (Byczkowski and
26    Kulkarnl 1992).

27    3.4. ELIMINATION AND EXCRETION

28    3.4.1.  Inhalation Exposure
29          Occupational studies showed that urinary vanadium levels significantly increased in
30    vanadium pentoxide-exposed workers (Orris et al., 1983; Kiviluoto et al., 198la: Gylseth et al.,
31    1979; Zenz et al., 1962: Lewis, 1959).  Male and female workers exposed to 0.1-0.19 mg/m3
32    vanadium in a manufacturing company had significantly higher urinary levels (20.6 ug/L) than
33    the nonoccupationally exposed control subjects (2.7 ug/L) (Orris et al., 1983). The correlation
34    between ambient vanadium levels and urinary levels of vanadium is difficult to determine from
35    these epidemiological  studies  (Kiviluoto et al., 1981b).  In most instances, no other excretion
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                10        DRAFT - DO NOT CITE OR QUOTE

-------
 1    routes were monitored. Analytical studies have shown very low levels in human milk (Byrne
 2    and Kosta, 1978). Evidence from animal studies supports the occupational findings.  Vanadium
 3    administered intratracheally to rats was reported to be excreted predominantly in the urine
 4    (Oberget al., 1978) at levels twice those found in the feces  (Rhoads and Sanders, 1985). Three
                                                                                       4R
 5    days after intratracheal exposure to radiolabeled vanadium pentoxide, 40% of the recovered  V
 6    dose was cleared in the urine, while 30% remained in the bones and 2-7% remained in the lungs,
 7    liver, kidneys, or blood (Conklin et al., 1982).
 8          In female rats exposed to 0.56 or 1.1 mg vanadium/m3 as vanadium pentoxide for 16 days
 9    (6 hours/day, 5 days/week), lung clearance half-times during an 8-day recovery period were
10    4.42 and 4.96 days, respectively (NTP, 2002).  In mice similarly exposed to 1.1 or 2.2 mg
11    vanadium/m3 as vanadium pentoxide, lung clearance half-times were 2.55 and 2.40 days,
12    respectively (NTP, 2002).  In contrast to the 16-day exposure data, the lung clearance half-times
13    in female rats exposed to 0.28, 0.56,  or 1.1 mg vanadium/m3 for 2 years (6 hours/day,
14    5 days/week) were 37.3, 58.6, and 61.4 days, respectively (NTP, 2002). In mice, the half-times
15    were 6.26, 10.7, and 13.9 days at 0.56, 1.1, and 2.2 mg vanadium/m3 exposure levels (NTP,
16    2002).
17          After intratracheal instillation of pentavalent vanadium, clearance from lungs was
18    initially rapid (3 hours) but with some vanadium (2% original dose) remaining at 12 days
19    postexposure. All other tissues eliminated 98-99% of the original dose by 3 hours postexposure
20    (Edel and Sabbioni, 1988).  Epidemiological studies and animal studies suggest that elimination
21    of vanadium following inhalation exposure occurs primarily through urination.

22    3.4.2.  Oral Exposure
23          No studies were available in the published literature regarding excretion in humans  or
24    laboratory animals after oral exposure to vanadium pentoxide.

25    3.4.3.  Dermal Exposure
26          No studies were available in the published literature regarding excretion in humans  or
27    laboratory animals after dermal exposure to vanadium pentoxide.

28    3.4.4.  Other Routes of Exposure
29          No studies were available in the published literature regarding excretion in humans  or
30    laboratory animals after other routes  of exposure to vanadium pentoxide.

31    3.5.  PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODELS
32          No PBPK models for vanadium pentoxide are available.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                11       DRAFT - DO NOT CITE OR QUOTE

-------
 1                                4. HAZARD IDENTIFICATION

 2    4.1.  STUDIES IN HUMANS-EPIDEMIOLOGY, CASE REPORTS, CLINICAL
 3         CONTROLS

 4    4.1.1.  Oral Exposure
 5           Few relevant studies investigating the effects of acute, subchronic, or chronic oral
 6    exposure to vanadium pentoxide in humans were identified in the peer-reviewed literature.  One
 7    study, Kucera et al. (1992), measured vanadium in the hair and blood of children exposed to
 8    vanadium by accidental drinking of contaminated water near a vanadium pentoxide plant.
 9    Vanadium levels in the hair did not differ significantly for the control and exposed groups.  An
10    increase in vanadium concentrations was found in the blood of exposed children (median:
11    0.078 ug/L) compared  to control children (median:  0.042 ug/L). No exposure-response
12    relationship could be quantified. These results suggest that vanadium levels in blood, but not
13    hair,  are a sensitive or suitable indicator of environmental exposure.

14    4.1.2.  Inhalation Exposure
15           Health effects of inhalation exposure to vanadium pentoxide include respiratory tract
16    irritation, bronchitis (often called boilermakers' bronchitis),  airway obstruction, chest pain,
17    rhinitis, pharyngitis,  laryngitis, and conjunctivitis in workers exposed to vanadium-containing
18    dust during vanadium pentoxide processing (Irsigler et al., 1999; Musk and Tees, 1982;
19    Kiviluoto et al.. 198la: Kiviluoto. 1980: Kiviluoto et al.. 1979: Zenzetal.. 1962: Sjoberg. 1956.
20    1951) or to fuel-oil ash containing vanadium during cleaning and maintenance of oil-burning
21    boilers (Kim et al.. 2004: Hauser et al.. 2001: Woodin et al.. 2000: Woodin et al.. 1999: Woodin
22    etal.. 1998: Hauser etal.. 1995a. b; Levy etal.. 1984: Ross.  1983: Lees. 1980: Sjoberg. 1955:
23    Williams, 1952). The literature includes one report of a controlled human exposure study
24    involving inhalation  of pure vanadium pentoxide dust (Zenz and Berg, 1967).  Descriptions of
25    these studies are found in Table 4-1. While exposures in the vandadium pentoxide production
26    facilities were to vanadium pentoxide in dust,  exposures to residual oil fly ash (ROFA) involved
27    a mixture of pollutants including elemental vanadium, vanadium oxides, vanadium pentoxide,
28    vanadium sulfates, paniculate  matter (PM), and other metal constituents (Hauser et al., 1995b).
29    Vanadium pentoxide is a major constituent of ROFA (2.5-40% by weight, as reported in
30    Navarro et al. , 2007), and the  health effects models evaluated in the studies of boilermakers
31    focused on vanadium content of respirable PM.  Therefore, these papers are integral to assessing
32    the nature and scope of the health response to vanadium pentoxide.
33           More recently,  epidemiology studies on the  metals content of ambient PM found that
34    communities in the United States with higher vanadium content in PM have higher PM-related
35    risk of mortality or hospitalizations for cardiovascular or respiratory disease (Bell et al., 2009:

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

-------
 1   Patel et al.. 2009: Dominici et al.. 2007: Lagorio et al.. 2006: Lippmann et al.. 2006) (described
 2   in Appendix D). Because exposures to vanadium, not vanadium pentoxide, were evaluated in
 3   the studies of ROFA or PM air pollution, information to characterize the exposure-response
 4   relationship between inhaled vanadium pentoxide alone and adverse health effects in humans is
 5   limited. Moreover, both ROFA and PM are mixtures, containing several other components that
 6   also could contribute to observed health effects.

 7   4.1.2.1. Controlled Human Exposure Study
 8          A  report of a controlled human exposure study indicated that an 8-hour exposure to pure
 9   vanadium pentoxide dust concentrations between 0.1 andl mg/m3 resulted in a productive cough
10   and wheeze that persisted for several days (Zenz and Berg,  1967). Volunteers (gender not
11   reported) were exposed to 0.1 mg/m3 (n = 2), 0.25 mg/m3 (n = 5) or 1  mg/m3 (n = 2) vanadium
12   pentoxide in an environmental chamber for 8 hours; no filtered air exposure  was included in this
13   study. Particle-size analysis revealed that 98% of particles had a diameter <0.5  jim.
14   Postexposure assessments of chest x-ray, blood, urine, nasal smear samples,  and pulmonary
15   function were similar to baseline values determined for each subject prior to exposure (data not
16   reported). No treatment-related symptoms or clinical findings were reported for any subject
17   during the 11 - to 19-month posttreatment period.

18   4.1.2.2.  Occupational Exposure during Vanadium Pentoxide Mining and Processing
19          Respiratory and other symptoms have been documented in case reports and two cross-
20   sectional studies among workers employed at facilities producing and processing vanadium
21   pentoxide (Irsigler et al.. 1999: Musk and Tees. 1982: Kiviluoto et al.. 1981b: Kiviluoto. 1980:
22   Kiviluoto et al.. 1979:  Zenz et al.. 1962: Sjoberg. 1956. 1951). Sjoberg (1951) reviewed earlier
23   reports of symptoms among workers exposed to vanadium pentoxide and described symptoms
24   among 36 employees (foremen, workers, builders/repairers) at a vanadium pentoxide production
25   factory in Falun, Sweden that began operation in  1946. The employees experienced symptoms
26   on one or multiple  occasions while under his medical surveillance between 1947 and 1950.
27   Symptoms generally persisted an average of 13 days.  Air samples in different parts of the
28   vanadium pentoxide facility were found to contain dust concentrations of 0.6-86.9  mg/m3 during
29   pulverization of iron ore slag with a vanadium content of 4.8-7.5%.  A significant proportion of
30   the dust consisted of small, respirable particles (22% < 8 jim,  39% < 12 jim). Although
31   vanadium pentoxide concentrations in air were not reported, vanadium was detected in the blood
32   and urine  of 23  and 27 individuals, respectively.
33          Sjoberg (1951) reported the main findings of his medical surveillance, including those of
34   a thorough examination of the cohort in October 1948, comparing them to an external referent
35   population.  The referent population comprised 703 workers from mines and sawmills in
36   northern Sweden that were followed during the same period and were examined using the same
37   methods.  The authors assumed that the comparison population was exposed only to inert dust
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                13       DRAFT - DO NOT CITE OR QUOTE

-------
 1   but no sampling data were reported. The main symptoms of the upper respiratory tract were
 2   nasal irritation or nasal catarrh (inflammation of mucus membranes; 42% versus 20% among
 3   unexposed) and throat dryness and pain (86% versus 8% among unexposed). These symptoms
 4   were reported to be more prevalent at the final examination in 1948 and included acute and
 5   chronic pathological changes in the nose and pharynx. Lower respiratory tract symptoms
 6   included cough with no sputum (61%), cough with sputum (39%), wheezing (86%), and
 7   shortness of breath (75%). In the comparison group, prevalence of coughing and shortness of
 8   breath was 4% and 24%, respectively. Bronchoscopy, measured in five individuals, revealed no
 9   severe changes in four cases and mild chronic bronchitis in one person. Spirometric measures
10   among the exposed and referent groups were similar. A larger decrease in hemoglobin levels
11   (7.8 ± 2.36% compared to 3.0 ± 2.07%) and red blood cell (RBC) count was observed among the
12   exposed, particularly among workers who were permanently or frequently employed at the plant
13   during the observation period, although levels  remained within normal limits.  The authors did
14   not observe leucopenia or eosinophilia in blood samples.  Two additional published case
15   summaries described symptoms similar to those reported by Sjoberg (1955, 1951) among
16   workers after the start of new plant operations  producing vanadium pentoxide pellets (n =  18)
17   (Zenz et al., 1962) or refining vanadium pentoxide (n = 4) (Musk and Tees, 1982). Both
18   exposures were to dry vanadium pentoxide powder at high concentrations (>0.5 mg/m3).
19          Kiviluoto et al. (1981a: 1981b: 1980; 1979) published a series of reports regarding an
20   investigation in 1975 of respiratory symptoms and clinical findings among 63 employees
21   (process workers, repairmen, foremen, and a laboratory worker) at a factory making vanadium
22   pentoxide from magnetite ore. A referent group of 63 men living in the same area was selected
23   from workers at the magnetite ore mine (concentrating plant, the mine, the repair shop, and the
24   office) and matched for age (within 2 years) and smoking habit (within five cigarettes per day).
25   Vanadium concentrations in the breathing zone during March-May 1976 were 0.028 mg/m3 (on
26   a time-weighted average [TWA] basis) with a range of 0.002-0.42 mg/m3. Air monitoring
27   results were not reported for the referent population.  Urinary vanadium concentrations among
28   the exposed averaged 0.26 ± 0.17 umol/L (18-hour excretion), while concentrations among the
29   referent group did not exceed the limit of detection (< 0.04 umol/L).
30          Clinical assessments were conducted by health personnel with no knowledge of exposure
31   status. Workers who reported wheezing were twice as likely to work in the vanadium factory
32   (p< 0.05) (Kiviluoto, 1980).  Prevalence of nasal  catarrh, cough, phlegm, or other respiratory
33   symptoms did not differ between the exposed and referent groups. Spirometric measurements
34   (FVC, FEVi, adjusted  for height), obtained at the  end of workers' summer holidays, did not
35   differ between the exposed and referent groups (p > 0.01) and were not related to duration of
36   exposure to vanadium  dust (p > 0.1).  Inflammation also was observed in nasal smears; these
37   results are discussed in the section on immunological endpoints (Appendix D). Nonfasting
38   serum chemistry parameters, analyzed in May-June, 1975, when vanadium concentrations were
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               14        DRAFT - DO NOT CITE OR QUOTE

-------
 1    higher (0.2-0.5 mg/m3) were compared between 16 exposed and 16 referent subjects. Among
 2    the several serum chemistry parameters tested, the difference between the exposed and referent
 3    groups for serum albumin, chloride, urea, bilirubin, and conjugated bilirubin were statistically
 4    significant, although no values were outside the range of reference values.  Hematological results
 5    (nonfasting) for 63 exposed and 16 referents, analyzed in March-May, 1976, when vanadium
 6    concentrations were lower (0.01-0.04 mg/m3), did not vary by exposure group (p > 0.05). In
 7    addition, no differences were noted for serum cholesterol, serum triglyceride, or leukocyte
 8    differential. In conclusion, this cross-sectional study found a higher prevalence of wheeze and
 9    some markers of inflammation in nasal smears among an occupational population producing
10    vanadium pentoxide with TWA concentrations between 0.002 and 0.42 mg/m3. An effect on
11    nonfasting serum chemistry and hematological markers was indicated at higher concentrations
12    between 0.2 and 0.5 mg/m3.
13          Irsigler et al.  (1999) evaluated the clinical histories of 40 men, employed at a vanadium
14    pentoxide production plant in South Africa, who were referred by the plant's medical staff for
15    more detailed medical assessment because of persistent respiratory symptoms (cough, breathing
16    difficulty, wheezing) between October 1995 and October 1997. Twelve men, aged 19-60 years
17    with bronchial hyperresponsiveness to inhaled histamine or exercise challenge (out of 40 men
18    referred) were selected for analysis along with 12 men, aged 24-54 years, who were referred and
19    did not have bronchial hyperresponsiveness, matched by age and smoking. All were free of
20    current symptoms and had no previous history of asthma when they began employment at the
21    plant.  Vanadium pentoxide concentrations in air from  area samples were <0.15 mg/m3 in the
22    mills, kiln, leaching, and pollution control areas; 1.53 mg/m3 in the fusion precipitation area; and
23    0.057 mg/m3 in the ferrovanadium area.  Vanadium pentoxide in spot urine samples was detected
24    in 10 of the 12 workers with bronchial reactivity (5.2-180 |ig/g creatinine) and was above a level
25    considered to be toxic in 3 individuals (>50 |ig/g creatinine). Levels in the 12 referent men
26    ranged between 12.0 and 55 |ig/g creatinine, with one person having >50 |ig/g creatinine.
27    Concentrations of sulfur dioxide and ammonia were above their recommended occupational
28    limits in the kiln area. Among nine  subjects who returned for a follow-up examination after 5 to
29    23 months with no vanadium pentoxide exposure, eight still exhibited bronchial reactivity.
30    Work tasks did not appear to differ between the two groups. This small study did not inform
31    understanding of what aspects of vanadium pentoxide exposure influence the prevalence of
32    bronchial reactivity.  Although the group with bronchial reactivity had two subjects with high
33    urine vanadium levels, the data reported  by the authors did not indicate that urine vanadium
34    levels or job sites varied by bronchial reactivity status and were not analyzed statistically. Also,
35    the sample size was not adequate to  determine an association of vanadium pentoxide exposure
36    with bronchial reactivity.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 15        DRAFT - DO NOT CITE OR QUOTE

-------
 1    4.1.2.3.  Occupational Exposure during Cleaning and Maintenance of Oil-Fired Boilers
 2          Vanadium pentoxide is present in significant amounts along with other vanadium oxides,
 3    vanadium sulfate, and metals in ash that accumulates in oil- and coal-fired boilers, and in other
 4    fuel types used in boilers.  Several reports of case histories and epidemiologic studies of
 5    boilermakers involved in the construction, cleaning,  and maintenance of oil-fired boilers have
 6    described upper and lower respiratory symptoms similar to those reported among workers
 7    processing vanadium pentoxide (Kim et al., 2004; Hauser et al., 2001; Woodin et al., 2000;
 8    Woodin et al.. 1999: Woodin et al.. 1998: Hauser etal.. 1995a, b; Lewetal.. 1984: Ross. 1983:
 9    Lees, 1980: Sjoberg,  1955: Williams, 1952). Additional health parameters have been
10    investigated, including pulmonary function (Woodin et al., 1999: Hauser etal.,  1995a: Levy et
11    al., 1984: Lees, 1980): biomarkers of inflammation in nasal fluid (Woodin et al., 1998: Hauser et
12    al., 1995b): and autonomic cardiac function (Magari et al., 2002).  Studies have investigated
13    acute effects among workers that cleaned or overhauled boilers over a period of a few days to
14    several weeks and chronic conditions among boilermakers who had worked in that occupation
15    for several years.
16          Several case summaries described the health  response of workers cleaning oil-fired
17    boilers in Great Britain (Ross. 1983: Williams. 1952). Sweden (Sioberg, 1955). and Canada
18    (Lees, 1980).  Two case series reports that reported measurements of vanadium pentoxide in the
19    air within the boilers  described respiratory symptoms, including rhinorrhoea, sneezing, eye
20    irritation, sore throat, and chest pain that began within  1 to 12 hours, with subsequent onset of a
21    dry cough, becoming paroxysmal and productive in some workers, wheezing, dyspnea upon
22    exertion, and fatigue  after 6-24 hours (Sjoberg,  1955: Williams, 1952). Bronchial irritation,
23    bronchitis, and the development of rales in regions of the lung of some workers also was
24    reported.  Symptoms persisted for 3 days to 1 week after exposure was ended. Concentration of
25    vanadium pentoxide particles (10-20 um in diameter) in the air inside the boilers during cleaning
26    was 17-85 mg/m3.
27          In 1981, the Occupational Safety and Health  Administration conducted an investigation
28    of work-related bronchitis  among 100 boilermakers exposed to vanadium pentoxide during an
29    oil-to-coal conversion of a utility company power plant in western Massachusetts (Levy et al.,
30    1984). The conversion occurred over approximately 6 weeks, October 15-November 30, with
31    most of the men working 10-hour days, 6 days per week. Air samples obtained in the boiler at
32    approximately 4 weeks during the conversion indicated vanadium pentoxide fume concentrations
33    of 0.05-5.3 mg/m3. Concentrations of chromium, nickel, and copper and iron oxide fumes were
34    stated to be within acceptable limits. Nitrogen dioxide and hydrogen sulfide were not detected.
35    Low concentrations of carbon dioxide (<5 ppm) and ozone (<0.1 ppm) were measured.  Sulfur
36    dioxide (<1 ppm) was measured in the boiler during  welding operations and outside the boiler
37    (1-35 ppm) when expansion joints were cut with a torch.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                16       DRAFT - DO NOT CITE OR QUOTE

-------
 1          In early December, investigators distributed a questionnaire to all 100 workers through
 2    the union president and responses were received from 55 men over the next 2 months.  All
 3    respondents, aged 23-60 years, reported symptoms, includingcough with sputum (85%), sore
 4    throat (76%), dyspnea on exertion (71%), chest pain or discomfort (65%), headache (56%),
 5    runny nose or sneezing (56%), wheezing (55%) and tiredness (51%).  The median time to onset
 6    was 7 days with clustering at 0-4 days and 6-8 days.  When the questionnaires were completed,
 7    symptoms had resolved or were improving in 41 of the 55 respondents.  Although three-fourths
 8    of the respondents stated that they had used a respirator over half the time when in the boiler,
 9    more than half stated that the respirator used was a paper mask. Respondents had been working
10    as boilermakers a median of 10 years.  A marked deficit in pulmonary function, particularly in
11    FEF, was observed in some workers.  Pulmonary function was assessed, however, in only 60%
12    of respondents by several different health providers.  The symptoms and effects on lung function
13    are consistent with previous reports about health effects among boilermakers (Lees, 1980;
14    Sjoberg, 1955; Williams, 1952).  However, this study is limited by alack of baseline information
15    on health status or comparison to a comparable occupational group, a relatively low response
16    among the exposed workers, and data collection days to weeks after exposure was discontinued.
17          A subsequent study of lung function among boilermakers overhauling an oil-fired boiler,
18    which measured baseline lung function and used a more detailed exposure assessment, did not
19    observe an association with respirable vanadium dust concentrations  (Hauser et al., 1995a). A
20    total of 36 of 80 eligible workers with an average  of 16.9 years on the job completed a baseline
21    test, and 26  completed a postexposure test.  Investigators developed daily exposure estimates for
22    each subject based on work diaries detailing tasks and locations and personal sampling. Time-
23    weighted average (1-10 hours) sampling was available for 15% of the total  number of study days
24    and these data were applied to the task/location information  to assign exposure levels for PMio
25    and vanadium dust for each worker.  Change in lung function between the baseline and
26    postexposure tests 4 weeks later,  adjusted for the average of the pre-  and postexposure values,
27    was inversely associated with peak PMio in multiple linear regression models adjusting for age
28    and current  smoking status. However, mean, peak, and day-of respirable vanadium dust
29    concentrations, were not associated with any spirometric indices (the data were not presented).
30    PMio and vanadium dust exposure also were  not related to bronchial reactivity as measured with
31    methacholine challenge tests before and after the overhaul. The authors noted that the
32    concentrations of vanadium dust were low, averaging between 2.2  and 31.3  ug/m3, and the
33    variation in  the range of concentrations might not have been wide enough to detect a relationship
34    with lung function in this small sample. Alternatively, a different constituent in PMio might
35    have caused the deficits in lung function.
36          Woodin et al. (2000; 1999; 1998) described in a series of reports a prospective clinical
37    study that evaluated health measures among 18 boilermakers and compared them to 11 utility
38    workers  involved in the overhaul of a large, oil-fired boiler over a 6-week period from  mid-May
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 17       DRAFT  - DO NOT CITE OR QUOTE

-------
 1    1995 to late June 1995. The men had volunteered for the study and had no allergic symptoms 2
 2    weeks prior to or during the overhaul. Vanadium and PMio exposures were calculated for each
 3    subject for each workday using information on task duration and location and use of personal
 4    protective equipment from job diaries, and vanadium and PMio concentrations from personal
 5    exposure monitors in the breathing zone. Utility workers did not enter the boiler during the
 6    overhaul. Vanadium levels before the boiler work were comparable for boilermakers and utility
 7    workers  (geometric mean (SD - standard deviation) ug/m3:  1.2 (1.4),  and 1.1 (1.2),
 8    respectively].  During the boiler work, vanadium levels rose to a geometric mean (SD) of 8.9
 9    (2.3) ug/m3 inside the boiler but did not change appreciably outside the boiler [geometric mean
10    (SD) ug/m3:  1.4 (1.6) (p < 0.001)].  Exposure estimates were adjusted for the type of protective
11    gear worn and its duration to calculate individual daily dose. The daily dose to the upper and
12    lower airway was estimated using values for minute volume, penetration and deposition rates,
13    and particle size.
14          Incidence of upper airway symptoms (nasal congestion/irritation, throat irritation) was
15    67% (12/18) among boilermakers and 36% (4/11) among utility workers.  The incidence of
16    lower airway symptoms (chest tightness, wheeze, cough, and sputum production) was 72%
17    (13/18) among boilers and 27% (3/11) among utility workers.  The workers recorded symptoms
18    five times per day in a log and scored them for severity with a numerical score from 0 to 3.  The
19    highest severity score for each day was used in the analysis.  Robust regression models of lower
20    airway maximum severity scores and average symptom frequency in relation to quartiles of lung
21    vanadium dose indicated a dose-related increase. Maximum lower airway severity scores were
22    increased by 0.47 (p = 0.01), 0.86 (p < 0.01), and 0.24 (0.10) in quartiles of lung vanadium doses
23    2, 3, and 4 compared to 1.  Average lower airway symptom frequency was increased by 0.19
24    (p = 0.02), 0.39 (p < 0.01), and 0.14 (0.07) in quartiles of lung vanadium doses 2, 3, and 4
25    compared to 1. The regression models were adjusted for current smoking.
26          Lung function was assessed  on three occasions: before the overhaul, during the overhaul
27    (before the shift on the last day), and 2 weeks after the work ended (Woodin et al., 1999).  No
28    changes  in the four airflow measures, FEF2s, FEF50, FEF75, and MMEF, were observed over the
29    course of the study. Change in lung function from the beginning to the end of the overhaul was
30    not associated with upper or lower airway dose levels of either vanadium or PMio when assessed
31    in linear regression models adjusting for smoking and age.  In addition, mean dose estimates
32    were not different between individuals who experienced a loss of either FEVi or FVC > 100 mL
33    or those who experienced no change or an increase (two-sample t-test).
34          Boilermakers were exposed  to higher PMio concentrations compared to utility workers
35    both before (geometric mean [SD]:  0.40 (1.60) versus 0.10 [2.70],p < 0.05) and during the
36    overhaul (geometric mean {SD]: 0.47 (1.90) versus 0.13 (4.00),/? < 0.001). In contrast to the
37    elevations in vanadium concentrations measured during the overhaul, PMio concentrations did
38    not increase appreciably (p > 0.05).  During the boiler overhaul, ozone concentrations increased
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                18       DRAFT - DO NOT CITE OR QUOTE

-------
 1    somewhat outside the boiler, but did not change inside the boiler.  The authors reported that
 2    levels of other metals (cadmium, chromium, manganese, lead, arsenic, and nickel) were low. All
 3    samples were 1-3 orders of magnitude below the 1996 TLV, and no significant changes were
 4    observed during the overhaul.
 5          Reductions in lung function over 2 years were investigated by Hauser et al. (2001) in a
 6    longitudinal study of boilermaker construction workers exposed to combustion particles from
 7    multiple sources, including power plants (oil, coal, natural gas), trash incinerators, paper mill
 8    incinerators, and other industrial sources with boiler, vessels, and tanks requiring maintenance
 9    and repair. A total of 118 boilermakers from Local 29 of the International Brotherhood of
10    Boilermakers,  Iron Shipbuilders, Blacksmiths, Forgers and Helpers (81% of those contacted)
11    were followed between 1997 and 2000. Participants completed spirometry after some days off
12    work, a modified American Thoracic Society questionnaire on respiratory symptoms, and a work
13    history questionnaire at baseline, with two annual follow-up visits. Baseline FEVi and FVC
14    were 90% and 94% of predicted.  The nine participants who were lost to follow-up had a lower
15    mean baseline FEVi compared to those who remained in the study (84% compared to 91.2%
16    predicted). The number of years worked as a boilermaker was a statistically significant predictor
17    of annual FEVi (-33.5 mL/years worked [95% CI:  -45.9,-21.1]). In generalized estimating
18    equation models adjusting for age, baseline FEVi, and cigarette smoking status, the number of
19    hours worked at gas-fired power plants in the previous year was inversely related to annual FEVi
20    (-9.8 mL/100 hours worked [95% CI: -16.0, -3.5). Adjusted models analyzing the number of
21    hours worked at oil- and coal-fired power plants also showed FEVi reductions but the
22    associations were not statistically significant.  This study provides evidence of long-term
23    declines in lung function among boilermakers exposed to combustion particles from several fuel
24    types. The study did not estimate exposure to individual substances, however, and no
25    conclusions can be drawn regarding a role for vanadium compounds.
26          The effect of occupational PM2.5 concentrations and metals components on cardiac
27    autonomic function during a work shift was studied among a panel of 39 boilermaker
28    construction workers with an average of 13 years (0-40 years) in that occupation (Magari, 2002,
29    34813}. Metals concentrations over an 8- to 10-hour work shift were determined from particle
30    samples (<2.5  um) collected using personal monitors. Vanadium concentrations were skewed
31    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
32    forty-eight personal samples were above the LOD for vanadium (0.00859 ug/m3). Average
33    PM2.5 concentrations were 1.16 ± 1.61 ug/m3 with a median of 0.56 ± 1.96 ug/m3 (range 0.09-
34    7.76). Vanadium concentrations during the shift were associated with a 3.98-msec (95% CI:
35    1.64, 6.32) per ug/m3 increase in heart rate variability estimated from the mean of the 5- minute
36    average SD of the normal-to-normal intervals (SDNN) in mixed effects regression models
37    adjusted for smoking status, age, and mean heart rate. Lead also was associated with an increase
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                19       DRAFT - DO NOT CITE OR QUOTE

-------
 1    in SDNN index (11.3 msec/ug/m3, 95% CI: 2.88, 19.73).2 Vanadium concentrations were not
 2    correlated with either lead or PM2.5, indicating that these pollutants were not likely to be
 3    confounders for the association of heart rate variability with vanadium. No associations with
 4    heart rate variability were observed for the other analyzed metals, including nickel, chromium,
 5    manganese, or copper. The increase in heart rate variability associated with vanadium and lead
 6    concentrations could have resulted from the temporal framework chosen for the analysis (e.g.,
 7    averages over the work shift).  Indeed, a subsequent study examining changes in a different
 8    index of autonomic cardiac function, rMSSD (square root of the mean squared differences of
 9    successive intervals), averaged over the nighttime hours (0:00-7:00), found an inverse
10    association with average work-shift concentrations of the PM2.5 metal, manganese (Cavallari et
11    al., 2008). The biological significance of the association of vanadium with increases in heart rate
12    variability is unclear. All-cause mortality has been associated with decreased heart rate
13    variability measured at baseline in longitudinal studies (Dekker et al.,  1997). The effect on heart
14    rate variability, however, indicates that vanadium exposure might alter autonomic function.
15    Alternatively, the observed association could have been due to chance.
16          In summary, symptoms of the upper and lower respiratory tract, including headache,
17    runny nose or sneezing, sore throat, cough with  sputum, dyspnea on exertion, chest pain or
18    discomfort, wheezing, and tiredness, were reported in case-series, cross-sectional, and panel
19    studies of workers employed at vanadium pentoxide production facilities (Ross, 1983; Musk and
20    Tees. 1982: Kiviluotq 1980: Kiviluoto et al.. 1979: ZenzetaL 1962:  Sioberg. 1951) and  of
21    boilermakers exposed to vanadium compounds in residual oil fuel ash over a few days to several
22    weeks (Woodin et al.. 2000: LevvetaL 1984: Ross. 1983: Musk and Tees. 1982: Kiviluotq
23    1980: Lees. 1980: Kiviluoto et al.. 1979: ZenzetaL. 1962: Sioberg. 1956.  1955: Williams. 1952:
24    Sjoberg, 1951).  Some symptoms began after a few hours, while the onset  of other symptoms
25    occurred after 1 to several days.  Vanadium pentoxide concentrations of particle size less than 5
26    microns in production facilities where the studies were conducted varied by work location,
27    ranging from 0.06 mg/m3 in the ferrovanadium area to 1.53 mg/m3 in the fusion precipitation
28    area at one facility (Irsigler et al., 1999).  Concentrations of vanadium pentoxide inside the boiler
29    during a cleaning operation were 85 mg/m3 (Sjoberg, 1955) and ranged from 0.05-5.3 mg/m3
30    during conversion of a power plant from coal to oil (Levy et al.,  1984).
31          Although a few of the case reports of workers exposed to ROFA during the cleaning or
32    conversion of boilers at electrical generating facilities reported measurements of vanadium
33    pentoxide, the case-control and panel studies evaluated symptoms and other health effects in
34    relation to ambient concentrations or biomarker concentrations of vanadium.  Therefore, relying
35    on these studies to draw conclusions regarding the hazard of vanadium pentoxide is not possible
      These effect estimates for vanadium and lead were reported as such in the abstract and in Table 4, but were
      transposed in the text of the results.

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

-------
 1    to. However, the findings of these studies are consistent with studies of workers in vanadium
 2    pentoxide production. The estimated dose of vanadium in the nose was associated with upper
 3    airway symptom severity scores (nasal congestion/irritation, throat irritation) during a boiler
 4    overhaul with geometric mean respirable (<10 |im) vanadium concentrations of 8.9 ± 2.3 |ig/m3
 5    (0.009 mg/m3) (Woodin et al., 2000). The estimated dose of vanadium in the lung was
 6    associated with increases in lower airway symptom frequency and severity (chest tightness,
 7    wheeze, cough, and sputum production).
 8          Vanadium pentoxide concentrations in vanadium pentoxide production facilities were not
 9    associated with pulmonary function deficits in the one epidemiological investigation that
10    included a referent group (Kiviluoto,  1980; Kiviluoto et al., 1979)1981. Pulmonary function was
11    measured after a 2- to 4- week summer holiday and changes across the shift or during the work
12    week were not assessed.  Among boilermakers, pulmonary function declines (FEVi, FVC,
13    FEF25%-75%) over the course of a boiler overhaul were observed, but no associations with
14    vanadium measures were found among the studies with more systematic and detailed reporting
15    of methods and results (Woodin et al.. 1999: Hauseretal.. 1995c: Levy etal.. 1984: Lees. 1980).
16    The authors reported that concentrations of respirable vanadium dust were relatively low and
17    insufficient variation in exposure might have precluded detection of an association with the small
18    changes in pulmonary function that occurred during the overhaul.  Uncertainties in the exposure
19    estimates, particularly inside the boiler, also might have prevented detection of associations
20    between vanadium exposure and pulmonary function.  The cohort had worked  in this occupation
21    an average of 20 years and could represent a healthy, less susceptible population.  In addition to
22    vanadium, boilermakers are exposed to increased levels of other metals during boiler overhauls
23    including nickel, chromium, and manganese (Liu et al., 2005).  The observed pulmonary
24    function declines among boilermakers might be explained by exposure to other ROFA
25    constituents.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                21        DRAFT - DO NOT CITE OR QUOTE

-------
          Table 4-1. Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
     Exposure assessment and exposure levels
                     Outcome definition and primary results
                                                                                                                                                                      Reference
Controlled human exposure;
9 healthy volunteers, aged 27^14 yr
No previous history of respiratory disease
or known exposure to pulmonary irritants

2 participants smoked cigarettes, <1
pack/day
Pure vanadium pentoxide dust, 98% < 5 microns in
diameter
0.1 (N = 2), 0.25 (N = 5) and 1 mg/m3 (N = 2) for 8 hr

Blood vanadium not detected
Urine vanadium after 3 d: max 0.013 mg/100 mL, after
7 d: not detected
Fecal vanadium: max 0.003 mg/gm, after 2 wk not
detected

No filtered air exposure
Two subjects exposed (inadvertently) to 1 mg/m3 developed coughing after 5 hr, coughing
persisted for 8 d, no other signs of irritation

Lung function tests (FVC, FEVi, MEF, MMF, MMET, FIVC) at baseline (three
occasions), after 8-h exposure, and once/week for 3 wk— no difference noted (data not
presented)

No alteration of white blood cell counts, differential cell patterns, or urinalyses (data not
presented)

No eosinophilia in nasal smears at 24 hr, 72 hr, or 1 wk later

Accidental re-exposure to dust cloud for 5 min caused coughing with sputum within 16 hr,
rales and wheezes within 2 d,  with normal lung function

0.25mg/m3for8hr(N = 5)

Productive cough by next morning with no other symptoms, resolved after 1 wk to 10 d.

Subjects continued normal activities, differential white blood counts were normal.

Pulmonary function tests were not changed (data not presented)

0.1mg/m3for8hr(N = 2)
Mucus within 24 hr with coughing, subsiding after 72 hr and resolved after 4 d
Zenz and Berg
(1967)
                                                                                         22
                                                                                            DRAFT - DO NOT CITE OR QUOTE

-------
         Table 4-1.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
     Exposure assessment and exposure levels
                    Outcome definition and primary results
                                                                                                                                                                    Reference
Case series, 1947-1950

36 employees (foremen, workers,
builders/repairers) at a V2O5 production
factory in Falun, Sweden that began
operation in 1946; compared to 703
workers from mines and sawmills in
northern Sweden

Referent group followed during same
period using same methods
Concentration in dust during pulverization of iron ore
slag:
0.6-86.9 mg/m3 with vanadium content of 4.8-7.5%;
22% < Smicrons, 39% < 12 microns

Pattern of exposure not described

Vanadium detected in blood (N = 29) and urine (N =
27) of exposed, not detected among unexposed
Main symptoms: All in series had symptoms at least once; Symptoms lasted an average of
13d

*Severe irritation of the upper respiratory tract: nasal irritation; inflammation/catarrh
(exposed 42%, unexposed 20%); throat dryness and pain (exposed 86%, unexposed 8%),
cough with no sputum (22/36, 4% among unexposed)
*Lower respiratory tract symptoms: cough with sputum (14/36), wheezing (31/36),
shortness of breath (27/36, 24% among unexposed)
*Other common symptoms:  palpitation of heart on exertion, weakness, and fatigue.
Other signs/symptoms:
*Bronchopneumonia in four cases, pneumonia in one case;
*Skin disease (N = 6):  papular eruption on the face, hands, dorsa of feet. Positive reaction
to patch test in one case with eczematous lesions.
*Tremor for a short period (N = 1).
*Greenish black discoloration of the tongue (N = 1)
*Decrease in hemoglobin levels (7.8 ± 2.36% compared to 3.0 ± 2.07%) and red blood cell
count
*No signs of chronic changes in the lung reported

Spirometry:
Sjoberg (1956.
1951)
                                                                                     VC
                                                                                     RC
                                                                                     TC
                                                        Exposed

                                                        4.84 ±0.13
                                                        1.70 ±0.08
                                                        6.54±0.16
                            Unexposed

                            4.49 ±0.07
                            1.56 ±0.04
                            6.05 ±0.09
                                                                                       23
                                                                                           DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Case series, 2 weeks during 1961

18 men, aged 21-55 yr, assigned tasks in
the pelletizing of vanadium pentoxide in a
pilot-plant operation
Exposure to pure vanadium pentoxide powder during
processing; highest exposure during unloading,
mulling, and pelletizing

Possible exposure to vanadium pentoxide fumes during
high temperature pelletizing

Concentrations of vanadium pentoxide in air calculated
as >0.5 mg/m3, mean particle size <5 microns

Urinary vanadium  detected in  12 men
Most highly exposed experienced symptoms by end of first work day (N = 3)

Burning eyes, sore throat, and dry cough.

Symptoms by end of day 3:  Slight conjunctivitis, inflamed throat, intense persistent
coughing.

After 3 d of no exposure, symptoms recurred with greater severity after work resumed with
respirator use. Cough accompanied by acute diffuse rales.

Clinical symptoms in all workers (persisting for up  to 2 wk):

Conjunctivitis, nasopharyngitis, hacking cough, fine rales, and wheezing (N = 16)

Upper respiratory irritation (N = 2)

White blood cell counts elevated (N = 9 of 17)—persisted up to 2 wk in some
"Skin itch" complaints (N = 4)
No spirometric changes attributed to exposure
Chest x-rays were normal in all 18 men
Zenz et al.
(1962)
Case Reports
4 workers at new vanadium pentoxide
refinery in Western Australia, aged 22, 24,
41, and 52 yr, all smokers
Examined  1 (N = 3) and 6 (N = 1) wk after
exposure
Exposure to dust from dry ammonium vanadate
powder

24-hr urinary vanadium (N = 1) 24 hr after
maintenance work discontinued: 0.003 mg
Respiratory symptoms reported:
Within hours: headache, green discoloration of the tongue, stuffy nose, lethargy, cough
with sputum
Within 3 d: wheezing, papular and cough with sputum

Within 1 d to 2 wk:  sore throat, hoarse voice, exertional dyspnea, wheezing

Symptoms resolved within 1-4 wk after exposure ceased

Pulmonary function at 1 wk (N = 3): mild to moderately severe airflow obstruction,
Bronchial hyperreactivity to histamine (N = 2)

One case with family history of asthma and positive skin prick tests, exhibited prolonged
asthma symptoms that persisted until 8 wk  after exposure ceased.
Musk and Tees
(1982)
Case reports: 8 maintenance workers at
large oil-fired boiler working 12 hr/d,
7d/wk
One of 2 random samples from deadspace contained
0.6% vanadium pentoxide by weight; 3 samples inside
boiler contained 14.2, 10.0, and 5.8% by weight

No protective equipment worn

Urinary vanadium concentration 0.05-11.0 mg/L
Symptoms reported in seven of eight workers: sore throat, running nose, cough with
sputum, breathlessness and wheeze, sore and watering eyes, and blepharitis

Persistent upper respiratory irritation in two workers, and eye irritation and blepharitis in
Ross (1983)
                                                                                          24
                                                                                              DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                      Study design, time period; Participants
 Study design,
  time period;
  Participants
Case series; October 1995-October 1997

12 men, aged 19-60 yr, employed at a
vanadium pentoxide production plant in
South Africa who visited the plant's
medical station for treatment of respiratory
symptoms, were referred for more detailed
assessment because of persistence of
symptoms, and who were found to have
bronchial hyperresponsiveness to inhaled
histamine or exercise challenge (out of 40
men referred); compared to 12 of the 40,
aged 24-54, who were referred and did not
have bronchial hyperresponsiveness,
matched by age and smoking
V2O5 concentration in air by plant site:

Mills, kiln, leaching, and pollution control: <0.15
mg/m3, fusion precipitation: 1.53 mg/m3,
ferrovanadium: 0.057 mg/m3

Urine V2O5 (ug/g creatinine):

5.2-180 in 10 of 12 reactive subjects; 180 ug/g (N = 1)
and 86.6 ug/g (N = 1) creatinine, nondetectable (N = 2)

12-55 ug/g creatinine in nonreactive subjects, detected
in all subjects
Histamine PC20 FEVi - Dose resulting in a decrease of 20% or greater; PC20 > 8 mg/mL
defined as normal

Exercise challenge: free treadmill running indoors for 6 min. Positive test defined as a
decrease of 200 mL or 15% or greater fall in FEVi

Bronchial reactivity persisted in eight of nine workers who returned for testing after 5-23
mo without exposure

Six reactive workers worked in the fusion precipitation area and three others worked in all
areas

Five non reactive workers worked in the fusion precipitation area

Differences in biomarker levels or job site were not analyzed statistically in relation to
bronchial reactivity status

The authors concluded that alternate explanations were not the reason for the bronchial
reactivity observed, including atopy, viral upper respiratory tract infection, or smoking, as
similar prevalence was noted in reactive and nonreactive groups.
Irsigler et al.
(1999)
                                                                                           25
                                                                                              DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Cross-sectional, prevalence, 1975

63 of 79 employees (process workers,
repairmen, foremen, and a laboratory
worker) at a factory making vanadium
pentoxide from magnetite ore. All
employed at factory >4 mo and not on
holiday or sick leave, aged 19-52 yr

Referent group: 63 men living in same area
selected from workers at magnetite ore
mine (concentrating plant, the mine, the
repair shop, and the office), matched for age
(within 2 yr), smoking habit (within 5
cigarettes/day), aged 20-52 yr
Vanadium dust concentration at various sites measured
on 8 d over 2 shifts in March-May 1976 (LOD: 0.002
mg/m3, CV: 6% at 0.05 mg/m3

Respirable fraction (<5 um)-20%
TWA (mg/m3)
Breathing zone (112 samples, 58 workers) whole shift:
0.028 (0.002-0.42)
Breathing zone grinding (2 samples, 1 worker, 1-hr):
1.7(0.25-4.7)
Stationary samples whole shift (80 samples, 15
workers): 0.012 (0.002-0.043)
Packing smelt (7 samples, 1 worker, 1-hr):
0.13 (0.02-0.37)
Grinding room  (1 sample, 1 worker): 2.3 (1-2.3)

Serum vanadium
(0.22±0.14umol/L)
Urinary vanadium
Exposed: 0.26 ± 0.17 umol/L, 18 hour excretion
Referent <0.04  umol/L LOD

Average duration of exposure: 10.8 yr
Previous concentrations in air averaged 0.2-0.5 mg/m3
1970-1975

Analyses were conducted "blind" to exposure status
Prevalence of symptoms:
Nasal catarrh cough, phlegm, or other: no difference, McNemar x^est
Wheezing more prevalent in exposed (p < 0.05)
Pathological changes in upper respiratory tract:  cytological and histological samples
Nasal smears: number of neutrophils increased (N = 45 pairs)
Exposed 16 (29%)
Referent 1 (2%),p < 0.001 McNemar's paired test
Biopsies of nasal mucosa
Number of plasma cells increased (N = 57 pairs)
Exposed 15 (26%)
Referent 4 (7%), p < 0.05)
Number of round cells increased (N = 56 pairs)
Exposed 21 (38%)
Referent 10 (18%),p < 0.05
No changes in number of eosinophils in cytological and histological samples in exposed
during 7-11 mo of exposure
Histological analysis: in exposed, subepithelial lymphocyte and plasma cell infiltration,
frequent papillarity in the mucous membrane surface and hyperemia
Spirometry (FVC,  FEVi) expressed at BTPS. Used highest value of five attempts.
Measured at the end of workers'  summer holidays (2-4 wk)
   Paired t-test of FEVi/Height2: no difference between exposed and referent for FEVi or
   FVC (p< 0.1)
Nonfasting chemical analysis, 16 exposed to high levels (0.2-0.5 mg/m3) compared to 16
referents during May and June 1975:  statistically significant differences between exposed
and referent for serum albumin, chloride, urea, bilirubin, and conjugated bilirubin.
Differences were slight and authors concluded not clinically significant

 Hematological analysis, 63 exposed, 22 referent, during March-May 1976 during low
exposure (0.01-0.04 mg/m3): no differencs

No differences for  serum cholesterol or serum triglyceride or leukocyte differential
Kiviluoto et al.
(1979):
Kiviluoto
(1980):
Kiviluoto et al.
1981a
                                                                                          26
                                                                                              DRAFT - DO NOT CITE OR QUOTE

-------
         Table4-l. Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                                                         Study design, time period; Participants
                                                                                                         Study design,
                                                                                                         time period;
                                                                                                         Participants
Studies of ROFA Exposure: Occupational Case Studies and Epidemiology Studies
Case reports, eight male workers engaged in
cleaning oil-fired boilers in Great Britain,
1946-1949

Ages 26^13 yr

Examinations before and after cleaning jobs
of approx. 1 week
Concentration of dust particles by diameter (mg/m ), 3
samples in superheater during cleaning
urn
           mg/m
0.15-1.0
1.0-5.5
5.5-11
0.36
4.09
7.86
93.6
6.14
0.26
                                      Vanadium in air (mg/m3)
                                      Superheater chamber 40.2
                                      Superheater chamber 17.2
                                      Combustion chamber 58.6

                                      Urinary vanadium (24-h collection)
                                      N= 1: 0.4 and 0.07 ug/mL
                                      N = 2:    trace
                                      N = 5:    
-------
         Table4-l. Studies of Vanadium Pentoxide Exposure: Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
 time period;
 Participants
Case studies; clinical evaluation of 17 men
occupationally exposed cleaning bottom ash
from the boiler of an oil-fired electricity
generating station, ages not stated,
Year not stated

Demographic information was not provided

No previous history of symptoms and no
positive patch skin tests at baseline
Vanadium content of bottom ash and crusted deposits
was 15.3% and 24.2-35%, respectively.
Personal sampling at breathing zone (N = 4): mean
TWA dust <10 microns: 523 ug/m3; total dust
concentration: >26 ug/m3

Men wore cartridge filter type respirators (peak
leakages 9%)

Urinary vanadium concentration < LOD (40 ug/L) (N
= 16)
280 ug/L in one worker with no respirator
Symptoms recorded on the day after completion of the cleaning operation: (N = 17)
Cough with sputum: 77%
Respiratory wheeze: 53%
Sore throat:  41%
Rhinitis: 29%
Headache: 18%
Dry cough:  18%
Eye irritation:  12%
Tongue discoloration: 6%
Epistaxis: 6%
Itch or rash:  0

Coarse rhonchi audible in those with wheeze, widespread rales found in two men

Symptoms resolved in 4-6 d

Reductions in FVC or FEVi as percent of baseline did not fall below 87%

Pumonary function: difference between mean baseline and lowest mean values during
days 1-4 and day 8, paired t-test

FVC: 5.3 ± 0.74 vs. 4.7 ±0.38 (p< 0.05)
FEV: 4.5 ± 0.38 vs. 3.9 ±0.4 (p< 0.001)
FMF (1/s) reduction: 1.16 ± 0.82 (23%), range 0.33 2.4 (9-31% decrease)

Lowest values occurred on day 2 but had not recovered on day 8
Lees(1980)
                                                                                       28
                                                                                           DRAFT - DO NOT CITE OR QUOTE

-------
         Table4-l. Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Case studies: 8 maintenance workers at
large oil-fired boiler working 12 hours/day,
7 days/week
One of two random samples from deadspace contained
0.6% vanadium pentoxide by weight; three samples
inside boiler contained 14.2, 10.0, and 5.8% by weight.

No protective equipment worn

Urinary vanadium concentration 0.05-11.0 mg/L
Symptoms reported in seven of eight workers:  sore throat, running nose, cough with
sputum, breathlessness and wheeze, sore and watering eyes, and blepharitis

Persistent upper respiratory irritation in two workers, and eye irritation and blepharitis in
Ross (1983)
OSHA investigation of 100 boilermakers
exposed during a coal to oil conversion of a
utility company power plant in western
MA, aged 23-60 yr.
Oct ober 15-November 30, 1981

Questionnaires and clinical assessment in
December 1981; 55/100 returned
questionnaire; clinical data from
examinations of 70 workers analyzed

Investigation conducted days after exposure
had ceased
Concentration of vanadium pentoxide measured on
November 17:
In boiler: V2O5 0.05-5.3 mg/m3 (N = 8)

Chromium, nickel, and fumes of copper and iron oxide
within "acceptable" limits, No CO2 or hydrogen sulfide
detected, CO2 < 5  ppm, ozone < 0.1 ppm, SO2 < 0.1
ppm

Urinary vanadium (N = 3) by AAS (LOD 10 ug/L):
Not detected
100% reported symptoms (N = 55)
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%)
Tiredness (51%)

Time to onset of symptoms: Median 7 d, clustering at 0-4 d and 6-8 d

At questionnaire, symptoms were resolved or improving in 41 workers

Pulmonary function tests (measured by study and private physicians) (N = 33):
FVC median 87% of predicted (5 of 27 < 80%)
FEVi median 93% predicted (8/27 < 80%)
FEVi/FVC median 79% predicted
FEF25_75o/0 median 57% predicted (N = 24) (4/24 > 80%)

Median FEF25-75o/0 was not correlated with history of cigarette smoking, although smoking
history was not obtained for 31%.
Levy et al.
(1984)
                                                                                        29
                                                                                           DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l. Studies of Vanadium Pentoxide Exposure: Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Cross-sectional, clinical

50 male boilermakers and utility workers at
a local electrical company; 37 with baseline
and postexposure lavage data and no cold or
flu symptoms during previous 2 weeks.
Exposure determined for each of 3 d between baseline
nasal lavage and after exposure using:
Personal sampling (37% of exposure days), self-
administered work diary for each day listing tasks
(welding, cutting, and grinding), and specific location
(boiler, airheater, and condenser). Data from personal
air samples (1- to 10-h TWA) were used to assign
exposure levels to each of 29 task and location
categories.

   io (1- to 10-hr TWA): 50^,510 ug/m3.
Respirable vanadium dust 0.10-139.2 ug/m3.
Differential cell count for PMNs, eosinophils, and epithelial cells/mL of recovered lavage
fluid.  Percent of total cells.
Adjusted by dividing the change in PMN or epithelial cell counts by the mean of baseline
and postexposure cell counts (expressed as percent).

Adjusted mean change in PMNs/mL (SD): 40% (100%), range -98% to 200%,p  < 0.05.
Hauser et al.
(1995a)
APMNatlj (%)
A%PMNs(%)
AEpitheliaUj (%)
A%Epithelial (%)
                                                                                                         Smokers            Non-smokers
                                                                                                         -0.1 (86.6)50.3 (102)
                                                                                                         -0.98(15.5)         4.56(28.5)
                                                                                                         53.0 (78.7)20.6 (82.1)
                                                                                                         14.0 (27.3)-3.9 (29.1)
                                                                                      Adjusted change in PMN, change in % PMNs, change in epithelial cell counts, or change
                                                                                      in % epithelial cells regressed on each of five exposure indices for PMio or vanadium dust
                                                                                      were not significantly associated in regression models stratified by smoking status.

                                                                                      Large variation in change in cell count parameters
                                                                                         30
                                                                                            DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
                                            Study design,
                                            time period;
                                            Participants
Prospective study, clinical
18 boilermakers (workers inside boiler) and
11 utility workers (lower exposure, outside
boiler) involved in overhaul of a large, oil-
fired boiler; 6 weeks, mid-May  1995-late
June 1995.

White men, mean age (range):
boilermakers: 37 (26-61 yr); utility
workers: 35 (30-55 yr)
% smokers:
boilermakers: 39%
utility workers: 17%

No colds or flu symptoms during the study
period
Vanadium concentration in nasal lavage (NL) fluid
(median [N, minimum, maximum]) Boilermakers: 0.5
(15, 0.5, 9.7), 4.7 (15, 0.5, 102.9), 0.5 (6, 0.5, 2.7) at
baseline, during, and 2 wk after the overhaul,
respectively, p < 0.05 (NL2 vs. NL1 or NL3).

Utility workers: 0.5 (8, 0.5, 3.8), 0.5 (8, 0.5, 8.0),
 0.05.
Nasal lavage was conducted on three occasions prior to the work shift, approx. 12 hr after
the previous day's exposure

Cytokine and enzyme levels in nasal lavage fluid
                                           Woodin et al.
                                           (1998)
Interleukin-8 (pg/mL)
               Boilermakers
Before         93.7(22.6-235.0)
During         4.7 (0.5-102.9)*
After          0.5 (0.5-2.7)

Myeloperoxidase (ng/mL)
               Boilermakers
Before         22.7 (2.0-72.8)
During         33.9(2.0-103.0)*
After          24.2(3.9-58.1)
Utility workers
69.2 (24.6-104.5)
0.5 (0.5-8.0)

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure: Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Prospective study, clinical
18 boilermakers (workers inside boiler) and
11 utility workers (lower exposure, outside
boiler) involved in overhaul of a large, oil-
fired boiler; 6 weeks, mid-May  1995-late
June 1995

White men, age range: boilermakers: 26-61
yr; utility workers: 30-55 yr
Mean length of employment as
boilermaker: 20.5 yr (3-39)

No colds or flu symptoms during the study
period

Workers recorded symptoms in log 5 times
per day, graded for severity (0-3). 50% of
participants returned for spirometry 2 weeks
after overhaul

Data collection was over 90% complete
Vanadium and PMio exposure calculated for each
subject on each work day using job diaries and data
from personal exposure monitors with adjustments for
type of protective gear worn and duration. Samplers
were located in the breathing zone, 10- to 12-hr
samples.

Daily dose to upper and lower airway estimated using
values for minute volume, penetration rates, deposition
rates, and particle size.

Exposure was estimated for subjects who did not wear
personal monitors from experience of people who did
wear a monitor and worked in same area.

Levels of cadmium, chromium, manganese, lead,
arsenic, and nickel were low, all samples 1-3 orders of
magnitude below 1996 TLV, no significant changes
during the overhaul.
Highest severity score for each day used in analysis.

Upper airway symptoms (nasal congestion/irritation, throat irritation)
                 Boilermakers        Utility workers
Incidence         12/18(67%)        4/11(36%)
Severity Score (Mean)
  Before          0.40*              0.23
  During         0.43*              0.18
Lower airway symptoms: (chest tightness, wheeze, cough, and sputum production)
                 Boilermakers        Utility workers
Incidence         13/18(72%)        3/11 (27%) Severity Score (Mean)
  Before          0.65**             0.11
  During         0.85***            0.01
*Difference between groups, p < 0.05
"Difference between groups, p < 0.01
***Difference between groups, p < 0.0001
Symptom severity score and average symptom frequency associated with quartiles of lung
vanadium dose and nasal PMio dose in a dose-related manner in regression models
adjusted for current smoking. Generally the response in the highest quartile was lower
than that in the third quartile.
Robust linear regression with cluster option to account for multiple symptom reports by
individuals
Pulmonary function:  highest of three acceptable curves used in analysis, repeated
measures analysis of variance across 3  test days.
No change over time for airflow measures (FEF25, FEF50, FEF75, and MMEF)
mean (SD)        FEVi (1)            FVC (1)
Before            3.73 (0.61)          5.01 (0.67)
During           3.76(0.54)          4.94(0.61)
After             3.65(0.42)          4.92(0.55)

% predicted FEVi: 7 subjects showed decreases > 100 mL during overhaul,  11 showed no
change or increases.
% predicted FVC: 8 subjects showed decreases > 100 mL during overhaul, 10 showed no
change or an increase.

No association of having a FEVi or FVC decrement > 100 mL with PMio or vanadium
exposure; Estimated dose of PMio or vanadium in nose or lung was not associated with
change in FEVi or FVC during overhaul in multiple regression models adjusting for
smoking status or age.
Woodin et al.
(2000; 1999);
same study as
Woodin et al.
(1998)
                                                                                          32
                                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Panel study among boilermakers
overhauling an oil-powered boiler;
spirometry before and after 4 weeks of
work; 36 of 80 eligible completed baseline
test, 26 completed postexposure spirometry,
24 completed a postexposure challenge test

The 26 subjects represent a self-selected
subgroup with higher average FEVi and
FVC % predicted and higher PC12-FEVi
(less reactive airways) compared to the 10
without postexposure spirometry—a
healthier subset of the initial study
population

Healthy worker effect—workers had a long
duration of employment
Mean age (range): 42.5  yr (27-60)

Average of 16.9 yr as boilermaker (6 mo-
35 yr)
Exposure estimated for each subject for each day based
on work diary (task/location) and personal sampling (1-
to 10-hr TWA available for 15% of total number of
study days).

Three Exposure indices:
Average, peak (1- to 10-hr TWA), and concentration
on the day of postexposure test

Vanadium concentrations in PMio
Average 12.2 ± 9.1 ug/m3 (2.2-31.3)
Peak 20.2 ±11.4 ug/m3 (2.2-32.2)
Day-of 12.1 ± 10.9(1.6-31.1)

Authors noted concentrations of vanadium low,
variation in exposure small, and sample size was low.
Spirometry:
Baseline (at least 1 d prior to start of overhaul):
For analysis, used highest of three acceptable and reproducible lung function values.
Methacholine challenge tests:  concentration of methacholine that caused a 15% fall in
FEVi.

AFEViadj:  AFEVi divided by average of pre- and postexposure FEVi

Spirometric indices showed  statistically significant decreases for AFEVi (mean of-140 ±
160 mL; range: -390 to 420), AFEV25%, AFEV50%, AFEV25-75% (mean of-270 ± 450 mL/s;
range:  -1,170 to 870), and AFVC (mean of-140 ± 200 mL; range:  -580 to 320), but not
AFEV75o/0 after exposure

Multiple regression analysis (adjusting for age and smoking status): peak PMio inversely
associated with adjusted AFEVi (p = 0.03), AFEV25%, (p = 0.07), AFEV50%, (p = 0.01),
AFVC (p = 0.01), but not AFEV25-75%, (p = 0.23) and AFEV75o/0 (p = 0.43)

Peak, average, and day-of respirable vanadium dust not associated with adjusted
spirometric indices.

The 10 workers without postexposure spirometry had worked fewer years at their present
job (12.5  ± 8.3) compared to those with postexposure spirometry (18.7 ± 8.0), p < 0.05,
and had a lower % predicted preexposure FEVi and FVC.
Hauser et al.
(1995a)
                                                                                         33
                                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure: Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                      Study design, time period; Participants
 Study design,
  time period;
  Participants
2-yr longitudinal study

118 boilermakers exposed to combustion
particles and metals through building,
maintaining, and repairing oil-, coal-, and
gas-fired boilers at power plants, boilers
and other vessels at nuclear  plants,
incinerators at trash dumps and paper mills

81% of 145 invited

Recruited through Local 29  of the
International Brotherhood of Boilermakers,
Iron Shipbuilders, Blacksmiths, Forgers and
helpers between December  1997 and March
1998

Male, Caucasian (97%)
Mean age 42.6 yr (10.5-56.5), 37 (31.4%)
current smokers
Detailed work history ascertained through self-
administered questionnaire completed 6 mo after
enrollment, and at 1- and 2-yr follow-ups. Recall
assisted through providing union records of work dates,
hours, and locations.

Hours worked per year at each plant type (oil, coal,
natural gas, trash, tree bark/sap and nuclear).

Approximately 70% worked at > types of plants.

Respirator use was 38.4 and 31.8% for years 1 and 2 at
gas-fired plants, 48.7 and 63.7% at oil-fired plants, and
59 and 66.9% at coal-fired plants.

Exposure to vanadium pentoxide was not estimated.
Annual decline in lung function:
Spirometry measurements at boilermakers union hall (Quincy, MA) or an apprentice
training site (Portland, ME).
Testing 9 am to 2 pm, between December and March to control seasonal variation.
Largest value of three acceptable curves

Data analysis using generalized estimating equation regression; Maximum FEVi at each
visit in relation to hours worked  during the previous year at each plant type.

Baseline FEVi  90% of predicted

Annual FEVi (mL/yr) adjusting for age, smoking, and baseline FEVi:
Duration of employment as boilermaker (years worked):
-33.5 (95% CI:  -45.9, -21.1)
Gas-fired plant work hours:
-9.8(95%CI:  -16.0, -3.5)
Oil-fired plant work hours:
-12.1 (95% CI:  -24.7, 0.6)
Coal-fired plant work hours:
-11.9(95%CI: -30.4,6.7)

Annual FEVi mL/Ever worked at specific fuel type
Gas-fired plant: -99.7 (95% CI: -154.8,-44.5)
Oil-fired plant: -77.6 (95% CI:  -151.1,-4.1)
Coal-fired plant: -73.4 (95% CI: -128.8,-18.0)

Association in model of ever working at a gas-fired plant remained statistically significant
with adjustment for percentage of time wearing a respirator.
Hauser et al.
(2001)
Panel study; 39 apprentice and journeyman
boilermaker construction workers; average
age 38 yr (18-59 yr), average duration of
employment 13 yr (0-40 yr)

Particulate matter concentrations and heart
rate monitored during 8- to 10-hr workshift
Metals concentrations over an 8- to 10-hr work shift
were determined from particle samples (<2.5 urn)
collected using personal monitors.
Vanadium concentrations (corrected for blank filter
metal content) were skewed, mean 0.76 ± 1.96 ug/m3,
median 0.13 ± 1.96 ug/m3 (range 0-11.62).  15/48
personal samples above LOD for vanadium (0.00859
ug/m3).  Average PM2.5 concentrations 1.16 ± 1.61
ug/m3, median 0.56 ± 1.96 ug/m3  (range 0.09-7.76).

Vanadium concentrations were not correlated with
either lead or PM2 •,
Heart rate was monitored using a five-lead Holter monitor. Heart rate variability was
estimated as the mean of 5-min average SD of the normal-to-normal intervals (SDNN).

Mixed effects regression models included random effect for each study subject and fixed
covariates for smoking status, age, and mean heart rate.

SDNN increase (index per ug/m3)
Vanadium:  3.98-msec (95% CI: 1.64,6.32)
Lead: 11.3  msec/ug/m3 (95% CI:  2.88,19.73).
PM2.5: -0.77 msec/ug/m3 (95% CI: -2.36,2.81)

No associations with heart rate variability were observed for the other analyzed metals,
including nickel, chromium, manganese, or copper.
Magari et al.
(2002)
                                                                                           34
                                                                                               DRAFT - DO NOT CITE OR QUOTE

-------
          Table4-l.  Studies of Vanadium Pentoxide Exposure:  Occupational Case Studies, Epidemiology Studies, and Controlled Human Exposure
 Study design, time period; Participants
      Study design, time period; Participants
                     Study design, time period; Participants
 Study design,
  time period;
  Participants
Prospective 5-day study of 20 boilermakers
during overhaul of power plant boilers
(74% participation), 18-59 yr of age, with
mean 21.7 ± 12.9 yr of boilermaking
experience
June 1999

10/20 were current cigarette or cigar
smokers

Demographic characteristics did not vary
significantly between participants and
nonparticipants (p > 0.25)
Personal samplers (PEMs) worn during workshift, 8-hr
TWA PM2.5
Total metal concentration including metal oxides on
particulate filter. Analyzed for vanadium, chromium,
manganese, nickel, copper, and lead.

Metal concentration and metal content (metal
mass/total PM2.5 mass).

 ;-hr TWA concentration, ug/m3 (Qjs-Q/s) (n = 39
samples)
Total PM2.5 mass:
440 (290-760)
Vanadium:
1.23(0.47-3.53)
Chromium:
0.11(0.05-0.27)
Manganese:
 0.81 (0.12-2.67)
Nickel: 0.96(0.31-1.88)
Copper: 0.98(0.25-2.53)
Lead: 0.17(0.04-0.53)

Correlations between metals: 0.52 < r < 0.92,p <
0.001
Urinary ug 8-OHdG/grams creatinine (controls for variation in urine dilution)

Preshift urine samples collected after average of 2 d away from work

Mean preshift 8-OHdG concentration (SE):

13.26 (1.04) ug/g creatinine

Mean cross-shift change:  1.88 (0.74) ug/g creatinine, p = 0.02

No difference by smoking status

Coefficient for linear mixed regression model for 8-OHdG levels and metal, adjusted for
urinary cotinine levels, chronic bronchitis status, and age:

ug/g creatinine per 1 mg/m3 increase in metal 8-hr TWA:

PM2.5:  1.67(95%CI: 0.21,3.14)
Vanadium: 0.23 (95% CI:  0.04.0.42)
Chromium:  3.08 (95% CI: -0.50. 6.67)
Manganese: 0.47 (95% CI: 0.05,0.89)
Nickel: 0.33 (95% CI: 0.01,0.64)
Copper: 0.20 (95% CI:  -0.07,0.46)
Lead: 1.67 (95% CI: 0.02,3.26)

Vanadium, Manganese, and Nickel also were significantly related to 8-OHdG levels when
expressed as proportion of total PM25 mass.

Regression models adjusted for tobacco smoking, chronic bronchitis status (yes/no), and
age.

4/20 had chronic bronchitis
Kim et al. (2004)
                                                                                         35
                                                                                             DRAFT - DO NOT CITE OR QUOTE

-------
 1    4.2.  SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
 2         ANIMALS - ORAL AND INHALATION

 3    4.2.1.  Oral Exposure

 4    4.2.1.1. Subchronic Studies
 5           No animal studies that have comprehensively examined histopathological, biochemical,
 6    and clinical endpoints of subchronic oral exposure were identified from the available literature.
 7    Mountain et al. (1953) evaluated the effects of subchronic exposure of rats to dietary vanadium
 8    pentoxide on body weight gain, erythrocyte count, hemoglobin, and cystine content of hair.
 9    Groups of five male Wistar rats were fed diets containing 0, 25, 50, 500, or 1,000 ppm of
10    vanadium (in vanadium pentoxide) for 103 days (25- and 50-ppm groups; "low-exposure"
11    groups) or 75 days (500- and 1,000-ppm groups; "high-exposure" groups).  After 35 days of
12    treatment, dietary vanadium levels of the low-exposure groups were increased to 100 and
13    150 ppm, resulting in average daily doses of 0, 74.5, 116, 500, and 1,000 ppm (0, 5.9, 9.2, 39, or
14    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
15    vanadium pentoxide.4 At the end of treatment, body weight gain, liver weight, and cystine
16    content of hair were measured in all groups; erythrocyte count and  hemoglobin level were
17    measured in control and low-exposure groups; and relative liver weight was measured in control
18    and 69.6 mg/kg-day groups. Average body weights of animals at the conclusion of the study
19    were not reported, although average weight gain in grams per rat was reported.  Compared to
20    control, average body weight gain was increased in the 10.5 mg/kg-day and 16.4 mg/kg-day
21    groups (54 and 45% increase, respectively) and decreased in the 69.6 mg/kg-day group (66%)
22    and 141.0 mg/kg-day group (no gain  in body weight over the study period).  The increase in
23    body weight gain at the low exposure levels was not explained and statistical significance or SDs
24    were not reported for any result.  Relative liver weight in the 69.6 mg/kg-day group was
25    statistically significantly increased compared to control, reported as a ratio of liver weight/body
26    weight (3.86 compared to 3.51,p < 0.05, F ratio in analysis of variance). Data on relative liver
27    weight were not reported for other dose groups.  A dose-dependent decrease in erythrocyte count
28    (12.8 and 21.3%) was observed over the duration of the study, in rats exposed to 10.5 mg/kg-day
29    and 16.4 mg/kg-day vanadium pentoxide, respectively, compared to controls (3.8% decrease)
30    (Table 4-2).  A 20-30% decrease in erythrocyte count is considered biologically significant; no
31    statistical analysis was reported by the study authors, however, and no measure of variance
32    (standard error or SD) was given for the means.  Data were not reported for high dose groups.
      Calculation: mg/kg-day = ppm(mg of compound per kg food) x mg food consumed per 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 et al.. 1953)1.
      4Conversion from mg/kg-day of vanadium to amount vanadium pentoxide: [(mg/kg-day vanadium)(molecular
      weight of vanadium pentoxide, 181.9)]/(2 x molecular weight of vanadium, 101.9).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  3 6      DRAFT - DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
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, although the researchers
speculated that vanadium might 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. No
statistical analysis was performed for the decrease in erythrocytes and body weight gain,
however, and no degree of variance was reported (precluding statistical analysis for this review).
Compared to hematological data for Wistar rats (Charles River Laboratories, 2008; Wright et al.,
1983), however, 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.  Based on decreased red blood cell count, the NOAEL and LOAEL values identified
following oral exposure to vanadium pentoxide are 10.5 and 16.4 mg/kg-day, respectively.

       Table 4-2. Hematological results of oral vanadium pentoxide exposure in
       Wistar rats

Control
10.5 mg/kg-d
16.4 mg/kg-d
Red Cell Count (M/mm3)a
Start
Finish (103d)
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 (103d)
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
16
17
18
19
20
21
aM/mm = million cells per millimeter.
Source: Mountain et al. (1953).
4.2.1.2.  Chronic Studies
       A 2.5-year dietary study on vanadium in rats (strain not described) was previously used
as the basis of the chronic reference dose (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 was 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
      5Convertedto ppm vanadium pentoxide by [(ppm vanadium)(MW vanadium pentoxide, 181.9)]/(2 x MW
      vanadium, 101.9). No information was given on average rat weights so values cannot be converted directly to
      mg/kg-day. Conversions are based on data from a subchronic study from the same group (Mountain et al.. 1953).
      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)1.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   3 7      DRAFT - DO NOT CITE OR QUOTE

-------
 1    evaluated were limited to growth rate, survival, and hair cystine content. The study did not
 2    assess comprehensive toxicity endpoints. Hair cystine content was significantly decreased in
 3    exposed animals, compared to controls but no values were given.  This study reports the oral
 4    NOAEL upon which an RfD can be based as 17.9 ppm (1.41 mg/kg-day) vanadium pentoxide.
 5    The biological significance of decreased hair cystine is unclear, however, and is not specific to
 6    vanadium pentoxide exposure. No additional oral chronic exposure studies in animals were
 7    identified in the published literature.

 8    4.2.2.  Inhalation Exposure

 9    4.2.2.1.  Subchronic Studies
10           NTP conducted 3-month exposure studies in F344/N rats to evaluate the cumulative toxic
11    effects of subchronic inhalation exposure to vanadium pentoxide (NTP, 2002).  Chemical
12    identity and purity of vanadium pentoxide was evaluated prior to the beginning and following the
13    conclusion  of all assays. Particle size given in mass median aerodynamic diameter
14    (MMAD) ± geometric standard deviation (GSD) for each dose group was as follows:
15    1 mg/m3  =  1.2 ± 2.8; 2 mg/m3 = 1.1 ± 2.8; 4 mg/m3 = 1.2 ± 2.8;  8 mg/m3 = 1.0 ± 2.9;
16    16 mg/m3 = 1.2 ± 2.8. Groups of 10 male and  10 female rats were exposed (whole-body
17    exposure) to aerosols of vanadium pentoxide at concentrations of 0, 1, 2, 4, 8 or 16 mg/m3,
18    6 hours per day,  5 days per week for 3 months. Additional groups of 10 male and 10 female rats
19    were exposed to 4, 8, or 16 mg/m3 for 12 (females) or 13 (males) weeks to  investigate effects of
20    exposure on cardiovascular function, pulmonary  function,  and pulmonary inflammation.
21    Clinical findings were recorded weekly and animals were weighed weekly  and at the end of the
22    study.  Blood and urine were collected from core study rats at study termination.  Blood was also
23    collected from cardiopulmonary physiology study rats on days 4 and 23 for hematology and
24    clinical chemistry determinations. Necropsy and histopathological evaluations (light microscopy
25    of comprehensive tissues)6 were performed on all main study rats exposed  to 0, 2 (male rats
26    only), 4, 8,  or 16 (female rats only) mg/m3 at the completion of the study.   Sperm motility and
27    vaginal cytology evaluations were analyzed from all core study rats.
28    Seven male rats and three female rats exposed to 16 mg/m3 vanadium pentoxide died during the
29    study (NTP, 2002). Abnormal breathing, emaciation, lethargy, abnormal posture, and ruffled fur
30    were observed in male and female rats exposed to concentrations of 8 mg/m3 and higher.
      6Complete histopathology was performed on 0, 8 (rats only), and 16 mg/m3 rats and mice. In addition to gross
      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 male 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.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   3 8       DRAFT - DO NOT CITE OR QUOTE

-------
 1   Diarrhea and nasal/eye discharge were also observed in some rats exposed to 16 mg/m3. Weight
 2   gain and absolute and relative lung weights are summarized in Table 4-3. Weight gain over the
 3   3-month treatment period was significantly decreased compared to control in males exposed to 4
 4   (6% decrease), 8 (10% decrease), and 16 (60% decrease) mg/m3 and in females exposed to
 5   16 mg/m3 (30% decrease). Absolute lung weights were significantly increased in males exposed
 6   to concentrations of 2 mg/m3 and greater and in females exposed to 4 mg/m3 and greater.
 7   Relative lung weights were significantly greater than control in males exposed to 2 (16%
 8   increase), 4 (30% increase), 8 (51% increase), or 16 (145% increase) mg/m3 and in females
 9   exposed to 4 (19% increase), 8 (76% increase) or 16 (117% increase) mg/m3.  Other organ
10   weight differences were considered to be related to body weight decreases.
11

             Table 4-3. Body weight gain and lung weights in rats (F344/N) exposed to
             vanadium pentoxide by inhalation for 3 months (values are
             means ± standard error)
Parameter
Exposure
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Rats
Weight gain during
3 -mo nth 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±0.19a
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 -mo nth 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
      "Significantly 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
      Source: NTP (2002).

12    Results of hematology assessments following 3 months of inhalation exposure are presented in
13    Table 4-4.  Erythrocyte count was  significantly increased in the 8 and 16 mg/m3 groups and
14    hematocrit was significantly increased in the 16 mg/m3 group in male and female rats.
15    Hemoglobin was increased significantly only in females exposed to 16 mg/m3. Microscopic
16    evaluation of the RBC morphology detected increased polychromasia and hypochromia in rats in
17    the 16 mg/m3 groups (data not presented). Significantly decreased mean cell hemoglobin
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 3 9      DRAFT - DO NOT CITE OR QUOTE

-------
 1    concentrations were observed in males exposed to 8 and 16 mg/m3 and in females exposed to 4,
 2    8, and 16 mg/m3. Reticulocyte count was significantly increased in males and females exposed
 3    to 16 mg/m3.  Mean cell volume was significantly decreased, indicative of microcytosis, in male
 4    rats at concentrations of 2 mg/m3 and above and in female rats at concentrations of 4 mg/m3 and
 5    above.  The observed hematological changes, including erythrocytosis, are consistent with
 6    pulmonary lesions that reduce pulmonary oxygen transfer, resulting in tissue hypoxia and
 7    stimulation of erythropoiesis by increased renal production of erythropoietin. Erythrocyte
 8    microcytosis is consistent with ineffective erythropoiesis, suggestive of altered iron metabolism
 9    and heme/hemoglobin production.
10

             Table 4-4.  Selected hematology parameters in rats (F344/N) exposed to
             vanadium pentoxide by inhalation for 3 months"
Parameter
Exposure
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Rats
Number
Erythrocytes (106/uL)
Reticulocytes
(106/uL)
Hematocrit (%)
Hemoglobin (g/dL)
Mean cell volume (fL)
Mean cell hemoglobin
(Pg)
9
9.2 ±0.1
0.2 ±0.02
48.5 ±0.6
15. 8 ±0.1
52.9 ±0.2
17.3 ±0.2
9
9.0 ±0.1
0.22 ±0.03
47.7 ±0.5
15.5 ±0.1
52.9 ±0.1
17.2 ±0.1
10
9.1±0.1
0.19 ±0.02
47.6 ±0.6
15.5 ±0.2
52.3±0.1b
17.1 ±0.1
9
9.3 ±0.2
0.23 ±0.03
48.7 ±0.9
15.9 ±0.2
52.2 ± 0.2b
17.1 ±0.02
10
9.7±0.2b
0.25 ± 0.02
49.9 ±0.7
16.1 ±0.2
51.3±0.2C
16.5±0.2C
3
15.1±0.3C
0.8±0.08b
71.2±2.8b
20.4 ±0.8
46.8±1.0C
13.4±0.4C
Female Rats
Number
Erythrocytes (106/uL)
Reticulocytes
(106/uL)
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)
      Source: NTP (2002).
11           Sporadic alterations in clinical chemistry and urinalysis variables were observed at
12   various time points in exposed males and females; however, no dose- or duration-related pattern
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   40      DRAFT - DO NOT CITE OR QUOTE

-------
 1    of effect was observed.  Occasional changes in serum liver enzyme activities were not consistent
 2    with hepatocellular injury.
 3          Vanadium pentoxide exposure did not affect reproductive endpoints in males (sperm
 4    count, spermatid heads,  sperm motility), but it did increase estrous cycle length by 10% in
 5    females exposed to 8 mg/m3, but not to 16 mg/m3, and reduced the number of cycling females in
 6    surviving rats in the 16 mg/m3 group (percent reduction not reported) (NTP, 2002).
 7          Complete histopathological assessments were performed on rats exposed to 0, 8, and
 8    16 mg/m3 for 3 months; only nonneoplastic lesions of the lung and nose were related to treatment
 9    (NTP, 2002). Results of histopathological evaluations of lung and nasal tissue from male and
10    female rats exposed to 1, 2, 4, 8, and 16 mg/m3 for 3  months are summarized in Table 4-5.
11    Significant increases in the incidences of epithelial hyperplasia of the lung were observed in
12    male and female rats exposed to concentrations of 2 mg/m3 or greater, compared to controls.
13    Epithelial hyperplasia occurred  in the distal airways and associated alveolar ducts and alveoli.
14    Inflammation and fibrosis were significantly increased in males (2 mg/m3 or greater) and females
15    (4 mg/m3 or greater).  In the nasal compartment, incidences of hyperplasia and metaplasia of the
16    respiratory epithelium were significantly increased in males exposed to 8 or  16 mg/m3 and in
17    females exposed to 4 mg/m3 or greater.  Nasal hyperplasia and metaplasia was localized to
18    respiratory epithelium on the ventral portion of the nasal septum, the vomeronasal organ, and, to
19    a lesser extent, the ventral lateral walls of the anterior portion of the nasal cavity. Nasal
20    inflammation was significantly increased in males and females exposed to 16 mg/m3.
21          Cardiopulmonary assessments were conducted in groups of 4-10 male and female rats
22    exposed to 0, 4, 8, and 16 mg/m3 for 3 months (NTP, 2002). No treatment-related changes in
23    cardiovascular function, as assessed by blood pressure (systolic, diastolic, and mean), heart rate,
24    and electrocardiogram, were observed in rats exposed to 4 or 8 mg/m3. Decreased heart rate and
25    systolic, diastolic, and mean blood pressure observed in male and female rats exposed to
26    16 mg/m3 were considered to be a reflection of the poor condition of the animals, and
27    complicated by anesthesia. Significant exposure-related decreases in pulmonary function (as
28    assessed by respiratory rate, tidal and minute volume, expiratory resistance, vital and total
29    capacity, diffusing capacity, and dynamic and peak compliance) were observed at all
30    concentrations of vanadium pentoxide-exposed male and female rats. Observed changes in
31    impaired capacity to diffuse carbon monoxide and reduced static and dynamic lung volumes at
32    exposure concentrations of 4 mg/m3 and greater suggest a restrictive lesion.  Changes in forced
33    expiratory maneuvers in rats exposed to 16 mg/m3 suggest the presence of an obstructive disease.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  41      DRAFT - DO NOT CITE OR QUOTE

-------
            Table 4-5. Incidences of selected nonneoplastic lesions of the lung and nose
            in rats (F344/N) exposed to vanadium pentoxide by inhalation for 3 months
Endpoint
Numbers of Animals with Lesions (Avg. Severity Score)
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Rats"
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 Rats"
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:  1 = minimal, 2 = mild, 3 = moderate, 4 = marked
     bSignificantly different from control by Fisher exact test (p < 0.01)
     Source: NTP (2002).

1           Pulmonary function results might indicate obstructive disease or could reflect the
2    deteriorating condition of the 16 mg/m3 rats, given that histopathological finding in lungs of rats
3    exposed to 8 and 16 mg/m3 were similar.  Taken together, results of pulmonary function tests
4    (PFTs) indicate that a presence of restrictive injury in male and female rats exposed to
5    concentrations of 4 mg/m3 or greater, while an obstructive lung injury could have been present in
6    rats exposed to 16 mg/m3.
7           Bronchoalveolar lavage (BAL) fluid was analyzed for markers of pulmonary
8    inflammation in rats exposed to 0, 4, 8 and 16 mg/m3 for 3 months (NTP, 2002).
9    Concentration-related increases were observed in the total numbers of cells, lymphocytes,
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  42      DRAFT - DO NOT CITE OR QUOTE

-------
 1    neutrophils, and protein recovered in BAL fluid from rats exposed to vanadium pentoxide at
 2    concentrations of 4 and 8 mg/m3, demonstrating a pulmonary inflammatory response in male and
 3    female rats. These endpoints also were increased in the 16 mg/m3 group, but to a lesser extent,
 4    and vanadium pentoxide was overtly toxic at this dose.
 5          Results of this study show that inhalation exposure of male and female rats to vanadium
 6    pentoxide aerosol for 3  months produced adverse effects on the hematological system and the
 7    lung (NTP, 2002).  Microcytic erythrocytosis, which was possibly secondary to impaired
 8    pulmonary function, was observed at concentrations of 2 mg/m3 and greater in males and
 9    4 mg/m3 and greater in  females. Absolute and relative lung weights were significantly increased
10    compared to controls at concentrations of 4 mg/m3 and greater in females and 2 mg/m3 and
11    greater and 4 mg/m3 and greater, respectively, in males. The incidence of nonneoplastic lesions
12    of the nose was increased in male and female rats at concentrations of 8 mg/m3 and greater and
13    4 mg/m3 and greater, respectively, and the incidence of nonneoplastic lesions of the lung was
14    increased in male and female rats at 2 mg/m3 and greater. Results of PFTs consistent with
15    restrictive lung disease  were observed at concentrations of 4 mg/m3 and greater. Based on
16    decreased erythrocyte size in male rats and nonneoplastic lung lesions and increased lung weight
17    in male (epithelial hyperplasia and inflammation of the lung) and female (epithelial hyperplasia)
18    rats, the NOAEL and LOAEL values identified for 3-month inhalation exposure to vanadium
19    pentoxide aerosols were 1 and 2 mg/m3, respectively.
20          NTP conducted a 3-month exposure study in B6C3Fi mice to evaluate the toxicity of
21    subchronic inhalation exposure to vanadium pentoxide (NTP, 2002). Groups of 10 male and
22    10 female mice were exposed (whole-body exposure) to vanadium pentoxide aerosols at
23    concentrations of 0, 1, 2, 4, 8, or 16 mg/m3, 6 hours per day, 5 days per week for 3 months.
24    Particle size given in mass median aerodynamic diameter (MMAD) ± geometric standard
25    deviation (GSD) for each dose group was as follows:  1 mg/m3 = 1.2 ± 2.8; 2 mg/m3 = 1.1 ± 2.8;
26    4 mg/m3 = 1.2 ± 2.8; 8 mg/m3 = 1.0 ± 2.9; 16 mg/m3 = 1.2 ± 2.8. Clinical findings were
27    recorded weekly. Animals were weighed weekly and at the end of the study. All study animals
28    were necropsied. Histopathological examinations of lungs were performed in all mice in the 0,
29    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
30    end of the 3-month exposure period, samples for sperm motility and vaginal cytology evaluations
31    were collected from mice exposed to 0, 4, 8,  or 16 mg/m3. Complete histopathological
32    examination was performed in mice in the control and 16 mg/m3 groups. Assessments of
33    cardiopulmonary function, pulmonary inflammation (analysis of pulmonary lavage), and
34    hematological parameters were not conducted in mice.
35          One male mouse in the 16 mg/m3 group died before the end of the study. Other than
36    appearing thin, no other signs of toxicity were reported (NTP, 2002). No other treatment-related
37    clinical findings were observed in any other mice in any treatment group. Weight gain and
38    absolute and relative lung weights are summarized in Table 4-6. Weight gain over the 3-month
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 43      DRAFT - DO NOT CITE OR QUOTE

-------
 1   treatment period was significantly decreased compared to control in males exposed to 8 (6%
 2   decrease) and 16 (10% decrease) mg/m3 and in females exposed to 4 (11% decrease), 8 (10%
 3   decrease) and 16 (12% decrease) mg/m3. Absolute lung weights were significantly increased
 4   compared to control at concentrations of 2 mg/m3 and higher in males and 4 mg/m3 and higher in
 5   females. Relative lung weights were significantly greater than the control in males exposed to
 6   4 (33% increase), 8 (43% increase), or 16 (82% increase) mg/m3 and in females exposed to
 7   4 (62% increase), 8 (63% increase), or!6(101% increase) mg/m3. Other organ weight
 8   differences were considered to be related to decreases in body weight by the researchers. The
 9   epididymal spermatozoal motility of males exposed to 8 or 16 mg/m3 was significantly decreased
10   by 13 and 5%, respectively. No treatment-related effects were observed for assessments of
11   estrous cycle (estrous cycle length and number of cycling females).
12          Results of histopathological  evaluations of lung tissue from male and female mice
13   exposed to 0, 1, 2, 4, 8, and 16 mg/m3 for 3 months are summarized in Table 4-7 flSTTP. 2002).
14   Epithelial hyperplasia was observed in male and female mice exposed to concentrations of
15   2 mg/m3 and above; lesion severity  increased with increasing exposure concentration.
16   Hyperplasia involved alveolar and, to a lesser extent, bronchiolar epithelium.  Inflammation was
17   characterized by multiple foci of a mixed cellular infiltrate oriented around blood vessels and
18   bronchioles and was observed in male mice exposed to 4 mg/m3 and above  and in female mice
19   exposed to 2 mg/m3 and  above. Infiltrate was composed primarily of macrophages with
20   abundant cytoplasm and  fewer lymphocytes and neutrophils. Histopathological evaluations
21   ofthe thymus of male and female mice exposed to 0, 8, and 16 mg/m3 for 3 months showed
22   lymphoid depletion in mice exposed to 16 mg/m3 (males: control, 0/9; 8 mg/m3, 0/8; 16 mg/m3,
23   2/7;  females: 0/9, 0/9, 1/10).The lung was identified as the most sensitive target organ in the
24   3-month inhalation study in mice (NTP, 2002). Based on increases in absolute lung weights at
25   concentrations of 2 mg/m3 and greater (males) and hyperplasia ofthe respiratory epithelium at
26   concentrations of 2 mg/m3 and greater (males and females), NOAEL and LOAEL values were
27   identified as 1 and 2 mg/m3, respectively.
28          In a study using cynomolgus monkeys, weekly provocation challenges (single 6-hour
29   exposures to 0.5  or 3.0 mg/m3) with inhaled vanadium pentoxide aerosol for 6 weeks produced
30   statistically significant pulmonary responses, prior to a subchronic exposure (6 hours per day, 5
31   days per week for 26 weeks) (Knecht et al., 1992). The subchronic exposure was divided into
32   three groups; one group (n = 8) was exposed to filtered, conditioned air and two exposed
33   groups(n = 8  each) received nominally equal weekly vanadium pentoxide exposures
34   (concentration x  time) with different exposure profiles. The peak exposure group received an
35   actual  concentration of 0.16 ± 0.01 mg/m3 (0.1  mg/m3 nominal) vanadium pentoxide on
36   Mondays, Wednesdays, and Fridays and 1.38 ± 0.07 mg/m3 (1.1 mg/m3 nominal) vanadium
37   pentoxide on Tuesdays and Thursdays.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 44      DRAFT - DO NOT CITE OR QUOTE

-------
            Table 4-6. Body weight gain and lung weights in mice (B6C3Fi) exposed to
            vanadium pentoxide by inhalation for 3 months (values are
            means ± standard error)
Parameter
Exposure
Control
1 mg/m3
2 mg/m3
4 mg/m3
8 mg/m3
16 mg/m3
Male Mice
Weight gain during
3 -mo nth 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 -mo nth 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)

     Source: NTP (2002).
            Table 4-7. Incidences of selected nonneoplastic lesions of the lung in mice
            (B6C3Fi) exposed to vanadium pentoxide by inhalation for 3 months
Lesion Type
Numbers of Animals with Lesions"
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)
1
2
"Numbers in parentheses indicate average severity grade of lesions in affected animals:  1 = 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

Source: NTP (2002).

       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

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

-------
 1    exposure of 0.10 mg/m3 after adjusting for exposure protocol (0.57 mg/m3 x 6/24 x 5/7). The
 2    peak exposure regimen averaged to a slightly higher continuous exposure of 0.12 mg/m3 after
 3    adjusting for exposure protocol. Vanadium pentoxide particle size was determined weekly
 4    during challenges and biweekly during exposures. Average particle size for the subchronic
 5    constant exposure group was 3.15 jim (MMAD), with a GSD of 3.25 jim. Particle sizes
 6    (MMAD ± GSD) for the peak exposure group were 3.17 ± 2.48  and 3.10 ± 2.45 for the
 7    0.1 mg/m3 and 1.1 mg/m3 exposures, respectively. PFTs, cytological and immunological
 8    analyses of blood and bronchiolar lavage fluid, and skin sensitivity tests were conducted before
 9    the pre- and postexposure provocation challenges. Immunological analyses are described in
10    Appendix D.  PFTs and bronchi olar lavage fluid analyses were also performed 1 day after the
11    provocation challenges. Cytological endpoints included complete and  differential blood cell
12    counts and leukotriene  €4 levels. Pulmonary function endpoints included total pulmonary
13    resistance (RL), forced expiratory flow (FEF), FVC, residual volume (RV), and dynamic lung
14    compliance (CLdyn).  Respiratory distress, characterized by  audible wheezing and coughing,
15    occurred in three of eight monkeys from the peak exposure group  on peak exposure days during
16    the first few weeks of the 26-week exposure; the responses  developed within 3 or 4 hours of
17    exposure and occasionally required early removal of the affected monkeys from the exposure
18    chamber.  Impaired pulmonary function accompanied preexposure provocation challenges with
19    vanadium pentoxide at 3.0 mg/m3 and was characterized by a 14% increase in RL and 13%
20    decrease in FEVso/FVC accompanied by a 14% increase in  RV and 3% decrease in FVC.
21    Pulmonary function and other study endpoints were not significantly different between the three
22    exposure groups (control, peak, and constant) at either challenge concentration when the
23    monkeys were rechallenged following subchronic exposure. The authors suggested that the
24    absence of increased pulmonary reactivity to vanadium pentoxide  following subchronic
25    inhalation might be attributed to the development of tolerance.  The study establishes a
26    subchronic NOAELjADJ] of 0.10 mg/m3 (continuous exposure) for  pulmonary function.  No
27    subchronic LOAEL was established. An apparent acute, but reversible, LOAEL of 1.38 mg/m3,
28    however, is established based on the respiratory distress observed  at 1.38 mg/m3 at an early time
29    point.
30          Hematological effects of vanadium pentoxide were  assessed in  male CD-I mice that were
31    exposed by whole-body inhalation for 1 hour per day, 2 days per week for up to 12 weeks
32    (Gonzalez-Villalva et al., 2006). A 0.02 M aqueous solution of vanadium pentoxide was
33    aerosolized, generating a reported average vanadium of 1,436 |ig/m3 (1.44 mg/m3 vanadium), as
34    measured by filters following the 12-week exposure. This study did not provide reliable
35    exposure information, thus exposure concentrations in mg/m3 could not be determined. Groups
36    of eight exposed mice and eight vehicle control mice (inhaling deionized water droplets) were
37    evaluated after 24 hours and weekly for 12 weeks. Evaluations  consisted of a complete blood
38    count and morphological examination of platelets. Platelet count was significantly increased in
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 46      DRAFT - DO NOT CITE OR QUOTE

-------
 1    the exposed mice on weeks 312; counts increased from week 3 to a maximum at week 9 and
 2    subsequently declined, but remained above controls (quantitative data inadequately reported).
 3    The morphology examinations showed the presence of giant platelets at unspecified longer
 4    exposure times.  The study establishes an apparent LOAEL for increased platelet count and
 5    altered platelet morphology from short-term intermittent exposure to vanadium pentoxide at
 6    2.56 mg/m3.  A continuous exposure equivalent concentration cannot be estimated with any
 7    confidence, as the intermittency of the exposure protocol is extreme.

 8    4.2.2.2.  Chronic Studies
 9           The toxicity of chronic inhalation exposure to particulate aerosols of vanadium pentoxide
10    was assessed in groups of 50 male and  50 female F344/N rats exposed (whole body exposure) at
11    concentrations of 0, 0.5, 1, or 2 mg/m3  for 6 hours per day, 5 days per week for 104 weeks (Ress
12    etal.. 2003: NTP. 2002). Particle MMAD ± GSD for  each dose group was reported as follows:
13    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
14    findings were recorded throughout the exposure period. Necropsy and comprehensive
15    histopathological evaluation7 were performed on all animals. No clinical findings related to
16    vanadium pentoxide exposure were  observed.  Mean body weights of females exposed to
17    2 mg/m3 were marginally less (3-6%; statistical significance not reported) than that of controls
18    throughout the 2-year study; mean body weights of exposed male rats were similar to controls
19    throughout the study.  The percent survival  of male and female rats for the entire 104-week
20    exposure period ranged from  52 to 58% in male rats and 30 to 40% in female rats. The percent
21    survival of controls was 40% for male rats and 28% for female rats (Table 4-8).  No infection
22    was reported in the study that might account for these  survival rates but the rates are comparable
23    to historical rates of survival for male and female  F344/N rats.
       In addition to gross lesions and tissue masses, the following tissues were examined in the 2-year 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.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    47       DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-8.  Selected nonneoplastic lesions of the respiratory system in F344/N
rats exposed to particulate aerosols of vanadium pentoxide for 2 years
Endpoint"
Exposure Group
Control
0.5 mg/m3
1 mg/m3
2 mg/m3
Male Rats
Percent survival"1
40
58
52
54
Lung
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)
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)
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"1
28
40
34
30
Lung
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)
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)
     This document is a draft for review purposes only and does not constitute Agency policy.
                                       48      DRAFT - DO NOT CITE OR QUOTE

-------
           Table 4-8.  Selected nonneoplastic lesions of the respiratory system in F344/N
           rats exposed to particulate aerosols of vanadium pentoxide for 2 years
Endpoint"
Exposure Group
Control
0.5 mg/m3
1 mg/m3
2 mg/m3
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:
    1 = 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 studies.
    Source: NTP (2002).

 1           The incidences of nonneoplastic lesions of the respiratory tract in male and female rats
 2    are summarized in Table 4-8 (Ress et al., 2003; NTP, 2002).  In male rats, the incidences of
 3    nonneoplastic lesions of the lungs (alveolar and bronchiolar epithelial hyperplasia and alveolar
 4    histiocyte infiltration), larynx (inflammation and epiglottis degeneration, hyperplasia and
 5    squamous metaplasia), and nose (goblet cell hyperplasia) were significantly increased compared
 6    to  controls in all vanadium pentoxide exposure groups.  In female rats, the incidences of
 7    nonneoplastic lesions of the lungs (interstitial fibrosis and alveolar histiocyte infiltration) and
 8    larynx (inflammation and epiglottis degeneration and hyperplasia) were significantly increased
 9    compared to the control in all vanadium pentoxide exposure groups.  In general, the incidences
10    and severity ratings of respiratory lesions increased with exposure level.  No treatment-related
11    histopathological findings were observed in other tissues.  A LOAEL of 0.5 mg/m3 was
12    established for nonneoplastic lesions of the respiratory tract in male and female rats; a NOAEL
13    was not identified.
14           The NTP (2002) and Ress et al. (2003) studies also conducted analysis of neoplasms in
15    rats exposed to vanadium pentoxide by inhalation for 2 years, using the protocol described
16    above. Compared to concurrent controls, the incidences of alveolar/bronchi olar adenoma,
17    alveolar/bronchi olar carcinoma or combined alveolar/bronchi olar adenoma, or carcinoma were
18    not significantly different (Poly-3 test) for male or female rats Table 4-9. Compared to historical
              O                                               Q
19    controls,  alveolar/bronchiolar adenoma in 0.5 and 2 mg/m  males (n = 8 and 6, respectively)
20    and 0.5 mg/m3 females (n = 3), alveolar/bronchiolar carcinoma in 0.5 and 2 mg/m3 males (n = 3
21    and 3, respectively), and combined alveolar/bronchiolar carcinoma in 0.5, 1, and 2 mg/m3 males
       The historical control databases were different for male and female F344/N rats. Historical control male F344/N
      rats were fed the NTP-2000 diet, while the historical control female F344/N rats were fed the NIH-07 diet and
      included a larger database.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    49      DRAFT - DO NOT CITE OR QUOTE

-------
 1    (n = 10, 6 and 9, respectively) and in 0.5 mg/m3 females (n = 3) were statistically significantly
 2    different (see Table 4-9 footnotes).
             Table 4-9.  Incidences of respiratory tumors in rats exposed to vanadium
             pentoxide in a 2-year inhalation study3
Tumor Type
Exposure Group
Historical
Control
Control
0.5 mg/m3
1 mg/m3
2 mg/m3
Male Rats
Number of animals examined
Alveolar/bronchiolar adenomab
Alveolar/bronchiolar carcinoma"1
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
Alveolar/bronchiolar 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%)
 4
 5
 6
 7
 8
 9
10
11
"Numbers 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
NTP-2000 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 NTP-2000 diet.
eHistorical incidence of combined alveolar/bronchiolar adenoma or carcinoma male F344/N rats fed in inhalation
chamber controls given NTP-2000 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.

Source: NTP (2002).


       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 because the tumors were observed only at

the lowest dose and no dose response was evident.

       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

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

-------
 1   4 mg/m3, 6 hours per day, 5 days per week, for 104 weeks (Ress et al., 2003; NTP, 2002).
 2   Particle MMAD ± GSD for each dose group was reported as follows:  1 mg/m3 = 1.3 ± 2.9;
 3   2 mg/m3 = 1.2 ± 2.9; 4 mg/m3 = 1.2 ± 2.9. Body weights and clinical findings were recorded
 4   throughout the exposure period. Necropsy and comprehensive histopathological evaluation were
 5   performed on all animals and analysis of both nonneoplastic and neoplastic lesions was
 6   performed. Many mice exposed to vanadium pentoxide were thin and exhibited abnormal
 7   breathing, particularly those exposed to 2 or 4 mg/m3 vanadium pentoxide (specific incidence
 8   data not reported). Mean body weights were generally less than control in males exposed to
 9   4 mg/m3 (decreases of 5-15%) and in females for all exposure groups (1 mg/m3, decreases  of
10   4-10%; 2 mg/m3, decreases of 14-20%; and 4 mg/m3, decreases of 4-19%) (Statistical
11   significance was not reported). The number of mice surviving for the entire 104-week exposure
12   period was similar to control (78% for male mice and 76% for female mice) for all exposure
13   groups for female mice and for males in the  1 and 2 mg/m3 groups, but survival was significantly
14   decreased in male mice exposed to 4 mg/m3  (50% survival rate) (Table 4-10).

            Table 4-10.  Selected nonneoplastic lesions of the respiratory system in
            B6C3Fi mice exposed to vanadium pentoxide in a 2-year inhalation study
Endpoint"
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 infiltration
Interstitial, fibrosis
50
3 (3.0)
0
6(1.5)
10 (2.4)
1 (1.0)
50
41C(2.2)
15C(1.0)
42C(1.5)
36C (2.4)
6 (1.7)
50
49C(3.3)
37C(1.1)
45C(1.6)
45C (2.6)
9C (1.2)
50
50C (3.9)
46C (1.7)
47C (2.0)
49C (3.0)
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)
o o
38
7 (2.4)
36
12(2.1)
40
13 (2.2)
Female Mice
Percent survival
76
64
60
64
Lung
Number of animals examined
Alveolar epithelium, hyperplasia
Bronchiole epithelium, hyperplasia
Inflammation, chronic
Alveolus, histiocyte infiltration
50
0
0
0
4(1.0)
50
31C(1.6)
12C(1.0)
37C(1.3)
34C (2.4)
50
38C (2.0)
34C(1.0)
39C(1.8)
35C (2.4)
50
50C(3.3)
48C(1.5)
49C (2.0)
45C (2.7)
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                51      DRAFT - DO NOT CITE OR QUOTE

-------
             Table 4-10. Selected nonneoplastic lesions of the respiratory system in
             B6C3Fi mice exposed to vanadium pentoxide in a 2-year inhalation study
Endpoint3
Interstitial, fibrosis
Exposure Group
Control
0
1 mg/m3
1 (2.0)
2 mg/m3
4b (2.5)
4 mg/m3
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(1.3)
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: 1 = 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)
      Source:  NTP (2002).
 1
 2           The incidences of nonneoplastic lesions of the respiratory tract in male and female mice
 3    are summarized in Table 4-10 (Ress et al., 2003; NTP, 2002).  In male mice, the incidences of
 4    nonneoplastic lesions of the lungs (hyperplasia of the alveolar and bronchiole epithelium,
 5    inflammation, and alveolus histiocyte infiltration), larynx (squamous metaplasia of the
 6    epiglottis), and nose (olfactory and respiratory epithelium degeneration) were significantly
 7    increased compared to control in all vanadium pentoxide exposure groups.  In female mice, the
 8    incidences of nonneoplastic lesions of the lungs (hyperplasia of the alveolar and bronchiole
 9    epithelium, inflammation, and alveolus histiocyte infiltration), larynx (squamous metaplasia of
10    the epiglottis), and nose (olfactory epithelium atrophy and degeneration and hyperplasia) were
11    significantly increased compared to control in all vanadium pentoxide exposure groups.
12    Incidences of interstitial fibrosis were significantly increased in male and female mice exposed
13    to 2 or 4 mg/m3.  In general, the incidences and severity ratings of lesions increased with
14    exposure level and matched the types of lesions observed in rats. No treatment-related
15    histopathological findings were observed in other tissues. The LOAEL of 1 mg/m3 was
16    established for nonneoplastic lesions of the respiratory tract in male and female  mice; a NOAEL
17    was not identified.
18           The incidences of tumors of the respiratory tract in male and female mice exposed to
19    vanadium pentoxide for 2 years are summarized in Table 4-11 (Ress  et al., 2003; NTP, 2002).
20    The incidences of alveolar/bronchiolar adenoma, alveolar/bronchiolar carcinoma, and combined
21    alveolar/bronchi olar adenoma or carcinoma were significantly increased in all groups of exposed
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   52      DRAFT - DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
female mice. In male mice, the incidences of alveolar/bronchiolar 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.

       Table 4-11. Incidences of respiratory tumors in B6C3Fi mice exposed to
       vanadium pentoxide in the 2-year inhalation study
Tumor Type"
Exposure Group
Historical
Control
Control
1 mg/m3
2 mg/m3
4 mg/m3
Male Mice
Number of animals examined
Alveolar/bronchiolar adenomab
Alveolar/bronchiolar carcinoma
Alveolar/bronchiolar 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
Alveolar/bronchiolar adenoma
Alveolar/bronchiolar carcinoma
Alveolar/bronchiolar 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%)
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
 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 = 1.2 ± 2.9.
 'Historical incidence of alveolar/bronchiolar adenoma male B6C3Fi mice fed in inhalation chamber controls
 given NIH-07 diet.
 Significantly different from control by the Poly-3 test (p < 0.01).
 Source: NTP (2002).

       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
(10 ug/g bw in corn oil; intraperitoneal injection) in week 1, followed by 5 weekly aspirations of
either vanadium pentoxide (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 4-12).  Differences  were
also observed between strains, with A/J mice showing increased tumorigenicity in response to
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  53       DRAFT - DO NOT CITE OR QUOTE

-------
 1    vanadium pentoxide. In the absence of MCA, vanadium pentoxide was not sufficient to initiate
 2    tumorigenesis in this study. C57BL/6J had no tumors following exposure (data not shown).  To
 3    evaluate the strain differences in inflammation, mice were aspirated with vanadium pentoxide
 4    (4 mg/kg bw) four times weekly, with BALF  collected at 6 hours, 1  days, 3 days, 6 days, and
 5    21 days following the last aspiration.  Cellular infiltrates and protein content were compared, and
 6    lungs were snap-frozen for histopathology. Increased pulmonary inflammation and
 7    hyperpermeability was increased in all exposed strains in a pattern similar to that of tumor
 8    induction, with greater increases observed in A/J mice than in BALB/C.  C57BL/6J showed the
 9    smallest increases in all parameters compared to all mouse strains. All results returned to
10    baseline at 21 days postexposure.  The results of PMN increases were confirmed by
11    histopathology in A/J and C57BL/6J mice. Further analysis was performed to measure
12    inflammatory chemokine production (KC, MIP-2, MCP-1) and transcription factor activity
13    (NFkB, c-Fos) and signaling pathway activation (MAPK).  Like the increased inflammatory
14    markers above, increased levels of KC and MCP-1 were observed in A/J and BALB/C mice as
15    compared to the C57BL/6J mice.   Similar strain differences were observed for the transcription
16    activity of NFkB  and c-Fos and the MAPK signaling  activity in A/J mice as compared to
17    C57BL/6J (BALB/C were not analyzed).

            Table 4-12. Lung tumor  multiplicity in MCA-treated mice exposed to
            vanadium pentoxide 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)
Vanadium Pentoxide
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 parentheses.
      Source: Rondini et al. (2010).
18
19          In summary, the identified noncancer health effects following occupational exposure via
20    inhalation to vanadium pentoxide in humans include respiratory irritation, cough, and bronchitis;
21    inhalation exposure in animals results in multiple health effects, including pulmonary
22    inflammation, lung and nasal hyperplasia, and pulmonary fibrosis. Vanadium pentoxide
23    inhalation exposure for 2 years was associated with a wide  spectrum of nonneoplastic pulmonary
24    lesions in both ratsand mice, ranging from hyperplasia to inflammation, fibrosis, and metaplasia.
25    Lesions were detected in the lung, larynx, and nose in both rats and mice exposed to vanadium
26    pentoxide.  Bronchial lymph node changes were detected in mice exposed to vanadium
27    pentoxide.  No human epidemiology studies that examined carcinogenesis following exposure to
28    vanadium pentoxide are available. The evidence of carcinogenicity in male and female mice
29    exposed to vanadium pentoxide is clear, however, and some evidence indicates carcinogenicity

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

-------
 1    in male rats, based on observations of alveolar and bronchiolar neoplasms that exceeded
 2    historical controls in groups exposed to vanadium pentoxide (NTP, 2002). A more recent study
 3    has also shown lung tumor promotion in sensitive mouse strains (Rondini et al., 2010).

 4    4.3.  REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL, INHALATION,
 5         INTRAPERITONEAL AND INJECTION

 6    4.3.1.  Oral Studies
 7          Vanadium pentoxide delivered orally to weanling rats (10-200 |imol/kg) for 3 days
 8    produced a significant increase in alkaline phosphatase activity and DNA content in the
 9    diaphysis of femoral bones, suggesting that vanadium pentoxide might be linked to bone
10    formation in the developing rat (Yamaguchi et al.,  1989). Yamaguchi et  al. (1989) examined the
11    potential effects of vanadium pentoxide on bone metabolism.  Weanling male Wistar rats (n = 5
12    per group) were exposed to 1.8-36.4 mg/kg (10-200 |imol/kg)9 vanadium pentoxide orally 1
13    hour following an oral injection of 15.3 jimol Zn/lOOg) three times at 24-hour intervals. The
14    highest dose vanadium pentoxide tested (200 jimol/kg) led to death in four of nine rats (cause of
15    death not described).  Authors state that administration of zinc prevented death in co-exposed
16    animal.  All rats were killed 24 hours following the last vanadium administration, and blood was
17    immediately removed by cardiac puncture. Statistically significant increases in calcium (27.3-
18    36.4 mg/kg; p < 0.05) and decreases in phosphorus (36.4 mg/kg; p < 0.01) were observed in
19    serum from high-dose rats. Administration of zinc completely prevented these serum changes.
20    Femurs also were immediately removed. The diaphysis and epiphysis were used for alkaline
21    phosphatase measures (right femur) and DNA content analyses (left femur).  Alkaline
22    phosphatase activity was significantly increased at the lowest dose tested (1.8 mg/kg; p < 0.05)
23    and peaked at 3.6 mg/kg. Further increasing doses led to decreases in alkaline phosphatase
24    activity. A similar pattern was observed for bone DNA content, with statistically significant
25    increases at 1.8-18.1 mg/kg (p < 0.05), but decreased at higher doses. Although the authors state
26    the interaction of vanadium and zinc led to an increase in bone calcium, based on the data
27    presented in the figures in this publication, vanadium administration did not lead to alterations of
28    bone calcium. Overall, this study shows an increase in both DNA content and alkaline
29    phosphatase activity in weanling male rats following oral administration of vanadium pentoxide,
30    suggesting a possible role of vanadium in increased bone growth.
31          Mravcova et al. (1993)assessed the extent of vanadium pentoxide accumulation in the
32    bones of rats following 6-month exposure (full study description is presented in Appendix D).
33    Vanadium accumulated in the epiphyseal cartilage of the tibia in rats. Significantly higher
34    concentrations of vanadium accumulated in the tibia and incisors of weanling rats compared to
35    adults. No dose response data for these endpoints were reported.
      9Conversion factor: umol = 0.1819 mg
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  5 5      DRAFT - DO NOT CITE OR QUOTE

-------
 1    4.3.2.  Inhalation Studies
 2          Investigations of the reproductive and developmental toxicity of subchronic or chronic
 3    inhalation exposure to vanadium pentoxide are limited to two studies. Mussali-Galante et al.
 4    (2005), used immunohistochemistry to assess the amount of gamma tubulin accumulating within
 5    somatic and testicular germ cells. Mussali-Galante et al. (2005) exposed 60 male CD-I mice to
 6    inhaled vanadium pentoxide (0.02 M, apparently aqueous solution containing aerosolized
 7    vanadium pentoxide) for 1 hour twice a week for 12 weeks. Avila-Costa et al. (2004),
 8    investigators from the same laboratory, used the same protocol and reported that droplets of the
 9    vanadium pentoxide mixture had average diameters of 0.5-5 |im.  Thirty-six control animals
10    inhaled only vehicle (deionized water).  Groups of three exposed animals and three control
11    animals were sacrificed each week for 12 weeks. Results indicated accumulation of vanadium
12    pentoxide in testes (Mussali-Galante et al., 2005). Gamma tubulin was significantly decreased in
13    testicular samples exposed to vanadium pentoxide compared to control,  starting after the first
14    week of exposure. Changes in gamma tubulin suggest changes in microtubule-involved
15    function, such as cell division, which can affect spermatogenesis.  Responses were duration
16    dependent, with the lowest percentages of immunoreactive cells occurring at the end of the
17    exposure period. Values at week 12 ranged from 1.2% for germ cells and 1.5% for Sertoli cells
18    to 10.1% for Leydig cells (compared to  87-88% in controls) (Mussali-Galante et al., 2005).
19          Fortoul et al. (2007) analyzed testes for ultrastructural changes, testosterone
20    concentration, and vanadium tissue concentration, using the same protocol as that reported above
21    (Mussali-Galante et al., 2005). No overt toxicity or changes in body or testicular weight were
22    observed. Histopathological analysis revealed necrotic cell death in spermatocytes (25%) and
23    Sertoli cells (15%) at weeks 5 and 6 in vanadium-exposed  animals. Spermatocytes exhibited
24    cytoplasmic vacuolation, nuclear distortion, and intercellular edema in response to vanadium
25    pentoxide exposure.  Spermatogonia (40% necrosis during weeks 6 and  7) were the most
26    susceptible cell type, followed by spermatocytes and Sertoli cells.  Moreover, vanadium
27    pentoxide concentrations increased dramatically after 1 week of exposure and remained
28    consistently elevated (average concentration 0.05 |ig/g dry tissue in controls, 1.63 |ig/g  dry tissue
29    in exposed animals, 33 times higher). Concentrations of testosterone were highly variable and
30    not statistically significant.
31          Vanadium pentoxide exposure did not affect reproductive endpoints in male rats (sperm
32    count, spermatid heads, and sperm motility). It did, however, increase estrous cycle length by
33    10% in female rats exposed to  8 mg/m3, but not to 16 mg/m3, and reduced the number of cycling
34    females in surviving rats in the 16 mg/m3 group (percent reduction not reported). The epididymal
35    spermatozoal motility of male mice exposed to 8 or 16 mg/m3 was significantly decreased by 13
36    and 5%, respectively. No treatment-related effects were observed for assessments of estrous
37    cycle in female mice (estrous cycle length and number of cycling females) (NTP, 2002).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  56      DRAFT - DO NOT CITE OR QUOTE

-------
 1    4.3.3.  Intraperitoneal and Injection Studies
 2           Male and female reproductive endpoints were evaluated in young rats following
 3    intraperitoneal (i.p.) administration of vanadium pentoxide (Altamirano et al., 1991). Newborn
 4    male and female rats were injected with 0 or 12.5 mg/kg vanadium pentoxide in saline on every
 5    other day from birth to age 21 days; groups sizes were five (treated males) or nine (male and
 6    female controls  and treated females). Males were sacrificed at 55 days of age and females were
 7    sacrificed on the day of first vaginal estrus. Other groups of females were injected with 0 or
 8    12.5 mg/kg-day vanadium pentoxide (n = 10 and 6, respectively) from age 21  days to the day of
 9    first vaginal estrus, at which time they were sacrificed. Reported endpoints in the males
10    consisted of absolute weights of testes, prostate, seminal vesicles, adrenals,  pituitary, thymus,
11    liver, kidneys, and submandibular glands.  The only effects in treated males were statistically
12    significant increases in seminal vesicle, thymus, and submandibular gland weights (20.1, 29.5,
13    and 19.2% higher than controls, respectively).  Endpoints evaluated in the females included body
14    weight, absolute organ weights (ovaries, uterus, adrenals, pituitary, thymus, liver, kidneys, and
15    submandibular glands), age at vaginal opening, number of ova in oviducts, and ovulation rate.
16    The only effects in treated females occurred in the group treated from 21 days of age; the effects
17    consisted of statistically significant increases in body weight (14.3% higher than controls) and
18    increased weights of thymus,  submandibular gland, and liver (31.1,  15.8 and 28.4% above
19    control weights, respectively).
20           Fertility  and sperm assessments were also performed in CD-I mice following i.p.
21    administration of vanadium pentoxide (Altamirano-Lozano et al.,  1996). In the fertility
22    assessment, groups of 20 and 15 male mice were injected with 0 and 8.5 mg/kg vanadium
23    pentoxide in saline, respectively, every three days for 60 days and were mated 24 hours after the
24    last injection. Statistically significant effects in the treated group included reduced fertility rate
25    in males (33% compared to 85% in controls), reduced numbers of implantation sites
26    (average = 5.8 compared to 10.88 in controls), reduced numbers of live fetuses (3.4 compared to
27    10.53 in controls), and increased number of resorptions per dam (2.00 compared to 0.24 average
28    resorptions in controls). In the  sperm assessment, 20 males were injected with 8.5 mg/kg
29    vanadium pentoxide every 3 days for up to 60 days with groups of five evaluated after 10, 20,
30    30, 40, 50, or 60 days of treatment.  Statistically significant effects included reduced sperm
31    motility on or after day 10, reduced sperm count and increased percentage of abnormal sperm
32    after day 20, decreased absolute testicular weight after day  50 (relative weight not reported), and
33    decreased body  weight after day 60.
34           Developmental toxicity was evaluated in groups of 13 or 15  female CD-I mice that were
35    administered 0 or 8.5  mg/kg vanadium pentoxide in distilled water, respectively, by i.p. injection
36    on days 6-15 of gestation (Altamirano-Lozano et al.,  1993). No maternal toxicity was reported
37    (endpoints not specified).  Developmental toxicity endpoints were assessed  on gestation day 18.
38    Endpoints  included number of implants, resorptions, and live fetuses.  For all fetuses, weight,
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  57      DRAFT - DO NOT CITE OR QUOTE

-------
 1    sex, and external malformations were noted. For two-thirds of the fetuses, skeletal abnormalities
 2    were also recorded. Internal soft-tissue examinations do not appear to have been conducted.
 3    The treated group had statistically significant increases in the number of litters with abnormal
 4    fetuses (9/15 compared to 3/13 in controls), number of abnormal fetuses (15/149 compared to
 5    3/124 in controls), and number of fetuses with short limbs (8/149 compared to 0/124 in controls).
 6    Additionally, the numbers of ossification centers in forelimbs and hindlimbs were significantly
 7    reduced in the treated fetuses.
 8           Zhang et al. (1991) evaluated the developmental toxicity in NIH mice  following i.p.
 9    injection of 5 mg/kg-day vanadium pentoxide on days 1-5, 6-15, 7, 8, 9,  10, 11, or 14-17 of
10    gestation.  There were no adverse effects on preimplantation or implantation,  developmental
11    toxicity, or premature births.  Increased frequencies of resorption or fetal  death were observed
12    for gestation days 6-15, 7, and 14-17. Delayed ossification (sites not specified) was observed
13    for gestation days 6-15, 8, 10, and 14-17. In a second study, Zhang et al. (1993b) evaluated
14    developmental toxicity in Wistar rats following i.p. injection of 0.33, 1.0, or 3.0 mg/kg-day on
15    days 6-15 of gestation.  Decreased placental weight and increases in embryo-fetus mortality and
16    external or skeletal malformations (unspecified) occurred at 1.0 and 3.0 mg/kg-day.  Maternal
17    toxic symptoms (unspecified), decreased maternal weight gain during treatment, and fetal growth
18    retardation were observed at 3.0 mg/kg-day. In the third study, Zhang et  al. (1993a) evaluated
19    developmental toxicity in Wistar rats following i.p. injection of vanadium pentoxide in doses of
20    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.
21    Effects in rats exposed on gestation days  6-15 and 9-12 included decreased maternal weight
22    gain, increased fetal mortality, decreased fetal weight and crown-rump  length, delayed
23    ossification of unspecified bones, and increased incidences of subcutaneous hemorrhage, wavy
24    ribs, and dilation of lateral ventricles and renal pelvis. Effects in rats exposed on a single day of
25    gestation included subcutaneous hemorrhage and unspecified visceral anomalies following
26    exposure on days 9, 10, and 11, and increased fetal mortality and delayed ossification of
27    unspecified bones following exposure on day 10.  Additional study details were not available.
28    Overall, the three studies identified an i.p. developmental toxicity LOAEL for vanadium
29    pentoxide of 1 mg/kg-day (Zhang et al.,  1991a, 1993a,b).
30           One study examined the effects of vanadium pentoxide (1.1 mg/kg in distilled water)
31    following tail vein injection in a total of 20  pregnant NMRI mice (data  pooled from three studies
32    with 6-10 mice each) (Wide, 1984).  Injections were performed before  implantation or on
33    gestation day 3 or 8, and animals were euthanized on gestation day 17 (2  days before parturition.
34    Preimplantation exposure had no effect on the fetuses relative to number per litter, weight, or
35    external and internal morphology. Exposure to vanadium pentoxide on gestation day 3 or 8 did
36    not lead to significant changes in resorption frequencies, fetal weights,  or frequencies of fetal
37    hemorrhages as compared to controls. The  number of fetuses defined as having less mature
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  5 8      DRAFT - DO NOT CITE OR QUOTE

-------
 1    skeletons10 by the authors, however, was significantly greater in mice given vanadium pentoxide
                                                                    r\
 2    on gestation day 8, but not gestation on gestation day 3 (p < 0.001, % test).
 3    Although the i.p. and tail-vein injection studies show the potential of vanadium pentoxide to
 4    cause reproductive and developmental effects in rodents, the studies are not useful for
 5    quantification of vanadium pentoxide toxicity, as equivalent oral or inhalation exposures cannot
 6    be established.

 7    4.4.  OTHER DURATION - OR ENDPOINT—SPECIFIC STUDIES
 8         Several studies determined LDso values for vanadium pentoxide following acute and short-
 9    term exposure. Among rats and mice, oral LD50 values ranged from 10 to 137 mg/kg body
10    weight (Yao et al., 1986 as cited in WHO, 2001; IARC, 2006;  Lewis et al., 2000). Acute
11    inhalation studies determined an LC67 of 1,440 mg/m3 in rats  (U.S. EPA, 1992) as well as
12    clinical signs of respiratory toxicity at 5 mg/m3 in cynomolgus monkeys.  Additionally,  acute and
13    short-term studies have evaluated immune system and nervous system toxicities. These studies
14    are described in detail in Appendix D.

15    4.5.  MECHANISTIC DATA AND OTHER STUDIES IN  SUPPORT OF THE MODE OF
16         ACTION FOR PULMONARY FIBROSIS AND CANCER
17          Noncancer effects in the lung range from histiocytic infiltration and alveolar
18    inflammation to hyperplasia of alveolar epithelium and pulmonary fibrosis.  These endpoints
19    exist in a plausible biological response continuum—from inflammation to reparative hyperplasia
20    to fibrosis. These effects also display a temporal and dose-response continuum, ranging from
21    inflammatory and hyperplastic responses that occur at earlier time points (16 days) and at lower
22    doses (2 mg/m3) to fibrosis that occurs at later time points (3 months) and at higher doses (4
23    mg/m3). Inflammation and hyperplasia are biologically relevant as precursor events to
24    pulmonary fibrosis.  Several investigators have systematically  investigated the molecular
25    mechanisms underlying vanadium pentoxide-induced pulmonary inflammation and fibrosis.
26          The evidence for mutagenicity in humans is limited (Ivancsits et al., 2002; Ehrlich et al.,
27    2008; Kleinsasser et al., 2003).  Few studies have examined genotoxicity in humans in vivo, with
28    equivocal results. Studies have demonstrated a genotoxic effect of vanadium pentoxide on
29    human cells in vitro (Ramirez et al., 1997;  Rojas et al., 1996; Roldan and  Altamirano, 1990).
30    Based on these studies, vanadium pentoxide-induced mutagenicity could occur at doses higher
31    than those measured in these occupational exposures, could be tissue specific, and could be
32    associated with oxidative stress rather than direct DNA damage. In vitro tests in bacterial and
33    yeast systems provide mixed evidence of vanadium pentoxide-induced mutagenicity.  In general,
34    classic gene mutation assays were negative, as were tests that assessed sister chromatid exchange
      10No ossification of three of four elements examined (supraoccipital bone, sternum, metatarsalia, and all caudal
      vertebrae).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  59      DRAFT - DO NOT CITE OR QUOTE

-------
 1    (SCE) and other chromosomal aberrations. DNA strand breaks (Ivancsits et al., 2002; Rojas et
 2    al., 1996) and micronucleus formation (Zhong et al., 1994) were indicated in some studies in
 3    cultured cells but depended on cell type. Fibroblasts appear to be more sensitive to vanadium
 4    exposure in vitro than are blood cells.  Similarly, experimental data from animal studies is
 5    equivocal. NTP (2002) reported that the frequency of micrenucleated normochromatic
 6    erythrocytes in peripheral blood was not increased in exposed mice compared to control mice.
 7    Several studies by Altamirano-Lozano et al. (1999; 1996; 1993), however, have noted DNA
 8    damage in specific target tissues in vanadium pentoxide-treated mice, although these studies
 9    used intraperitoneal injection as the route of exposure. These studies are described in detail  in
10    Appendix D.

11    4.6.  SYNTHESIS OF MAJOR NONCANCER EFFECTS

12    4.6.1.  Oral
13          Only one study was identified in humans exposed to vanadium pentoxide by the oral
14    route of exposure (Kucera et al., 1992). In this study, vanadium levels in the hair and blood of
15    children exposed to vanadium, by drinking contaminated water near a vanadium pentoxide plant,
16    were measured. However, no exposure-response relationship could be quantified. Limited
17    animal studies are available examining toxicity following oral exposure to vanadium pentoxide.
18    Table 4-13 presents a summary of the noncancer results for the subchronic and chronic oral
19    studies of vanadium pentoxide toxicity in experimental animals.
20          The only acute  studies available report an oral LD50 value that ranges from
21    -10-137 mg/kg body weight in rats and an oral LDso value of 64 mg/kg body weight in mice,
22    depending on the source (Yao et al., 1986 as cited in WHO, 2001).  Clinical signs of toxicity
23    included lethargy, excessive tearing (lacrimation),  and diarrhea. Histopathological analysis
24    revealed liver necrosis  and swelling of renal tubules. No acute dose-response studies are
25    available for any animal species.
26          Data on toxicity following subchronic oral  exposure to vanadium pentoxide are limited to
27    one study (Mountain et al., 1953). The primary noncancer health effects of subchronic oral
28    exposure in animals include changes in relative liver weight, decreased erythrocyte and
29    hemoglobin counts, and decreased hair cystine content in rats (Mountain et al., 1953). Mountain
30    et al., (1953) observed  decreased erythrocyte count and hemoglobin counts at > 16.4 mg/kg-day.
31    These effects were correlated, so observed decreases could not be considered as separate effects.
32    The study authors also  reported increased relative liver weights, but this endpoint was only
33    measured in the control and the high dose group (69.6 mg/kg-day) and thus no dose-response
34    could be established. Changes in hair cystine levels were observed at doses> 1.41 mg/kg-day.
35    However, the biological significance of this effect  is unknown, as changes in hair cystine content
36    can be considered a potential biomarker of exposure, rather than an adverse effect.
37
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                  60      DRAFT - DO NOT CITE OR QUOTE

-------
 1
 2
            Table 4-13. Summary of noncancer results of repeat-dose studies for oral
            exposure of experimental animals to vanadium pentoxide


Species
Wistar rat
(n = 5 per
group)

















Rat (strain
and number
not reported)








Gender
Male



















Male









Avg. Daily Dose
(mg/kg-d)
0, 10.5, 16.4,
69.6,
14 1.0 mg/kg-d
(corresponds to
0, 74.5,a116.1,b
500, and
1,000 ppm)

0, 10.5, 16.4,
69.6,
14 1.0 mg/kg-d
(corresponds to
0, 74.5,a116.1,b
500, and
1,000 ppm)
0, 10.5, 16.4,
69.6, 141.0
mg/kg-d
(corresponds to
0, 74.5,a116.1,b
500 and 1000
ppm)
1.41 and
14.1 mg/kg-d
(corresponds to
17.9 and
179 ppm)




Exposure
Duration
and Route
103 d in
food






103 d in
food





103 d in
food





2.5yrin
food








Response
at LOAEL
Decreased
erythrocyte
count





Decreased
relative
liver
weight



Decreased
hair cystine





Decreased
hair cystine








NOAEL
(mg/kg-d)



10.5






~






10.5







1.41





LOAEL
(mg/kg-d)



16.4






69.6






16.4







14.1






Comments
Dietary
exposure was
increased at
day 35 of
study (from
25 to 100 and
from 50 to
150 ppm) (See
notes below)












Study
published in
Patty 's
Industrial
Hygiene and
Toxicology 3r
ed. (1981):
Original data
not available.


Reference
Mountain
etal.
(1953)

















Stokinger
etal.
(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.
 3
 4          Chronic animal studies evaluating effects of oral exposure to vanadium pentoxide are
 5   limited.  Stokinger et al.(1953) (as cited in Clayton and Clayton, 1981) reported significantly
 6   decreased hair cystine levels (no details were provided regarding the percent decrease or dose
 7   response) in exposed rats. Comprehensive toxicity endpoints were not evaluated, and the
 8   number and strain of rats used were not reported in the study.  The biological relevance of
 9   decreased hair cystine levels is unknown. The study authors suggested that decreased hair
10   cystine could be related to poor nutritional status, as a result of an aversion to vanadium
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  61      DRAFT - DO NOT CITE OR QUOTE

-------
 1    pentoxide in the feed, rather than directly to vanadium exposure. These decreases may also
 2    reflect changes to enzymatic pathways.

 3    4.6.2. Inhalation
 4          Table 4-14 presents a summary of the noncancer effects observed following inhalation
 5    exposure to vanadium pentoxide in animals. Details of the available human studies are available
 6    in Section 4.1 and Table 4-1. The identified noncancer health effects following occupational
 7    exposure to vanadium pentoxide in humans include respiratory irritation, airway obstruction,
 8    chest pain, bronchitis, and similar effects. Inhalation exposure in animals results in multiple
 9    health effects, including pulmonary inflammation, lung and nasal hyplerplasia, pulmonary
10    fibrosis, changes in nervous system structure and function, and reproductive/developmental
11    effects.  Animal studies have identified the lung as the most sensitive organ to vanadium
12    pentoxide-induced toxicity. Other animal studies have also reported effects on the nervous
13    system, reproductive system and the developing fetus, and immune system following inhalation
14    exposure to vanadium pentoxide.
15          The primary noncancer health effects identified following acute inhalation exposure in
16    humans  are respiratory irritation, cough,  and mucus formation.  A human controlled exposure
17    study (n = 9) performed by Zenz and Berg (1967) reported respiratory irritation, cough, and
18    mucus formation in humans exposed to vanadium pentoxide for 8 hours. Levy et al.(1984)
19    described respiratory irritation in 100 workers who were reportedly exposed to 0.05-5.3 mg/m3
20    vanadium for 10 hours per day,  6 days per week, for 4 weeks. This study  did not include a
21    referent  group, and exposure to  vanadium pentoxide could not be directly  correlated to effects.
22    Respiratory and other symptoms have been documented among workers employed at facilities
23    producing and processing vanadium pentoxide (Kiviluoto et al., 1981a; (Irsigler et al., 1999;
24    Musk and Tees. 1982: Kiviluoto. 1980: Kiviluoto et al..  1979: Zenzetal., 1962: Sioberg,  1956.
25    1951).  Similar effects were observed in  occupational studies of boilermakers involved in the
26    construction, cleaning, and maintenance  of oil-fired boilers (Kim et al., 2004: Hauser et al.,
27    2001: Woodin et al.. 2000:  1999: 1998: Hauser etal.. 1995b, a;  Levyetal.. 1984: Ross. 1983:
28    Lees. 1980: Sioberg, 1955: Williams. 1952).
29           Observed respiratory effects in humans are supported by the available literature
30    evaluating toxicity in  animals.   Specifically, male monkeys exposed to vanadium pentoxide at
31    0.5 or 5.0 mg/m3 for 6 hours (Knecht et al.,  1985) were shown to have restriction of air flow at
32    high dose.   WHO (2001) also reported that a 1-hour inhalation  exposure to vanadium pentoxide
33    dust in rats led to an LCe? of 1.44 mg/L (1,440 mg/m3).  Clinical signs of toxicity included
34    respiratory difficulty,  increased  respiratory tract mucus production, and irritation of the eyes,
35    nose, and throat.  Short-term inhalation exposure in animals exposed to vanadium pentoxide for
36    3 months also increased pulmonary inflammation and relative lung weight in rodents (NTP,
37    2002).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  62      DRAFT - DO NOT CITE OR QUOTE

-------
 1          NTP (2002) evaluated pulmonary and nasal endpoints in both male and female rats and
 2    mice after a 3-month exposure to vanadium pentoxide at 0, 1, 2, 4, 8, and 16 mg/m3.  Lung
 3    inflammation, lung hyperplasia, and increased relative lung weight were observed at the low
 4    doses in rats and mice. In addition, rats exhibited bronchiolar exudates, microcytic
 5    erythrocytosis, lung fibrosis, and nasal lesions.  Body weight loss and increased absolute lung
 6    weight were reported in mice.  Rats appear to be more sensitive to inhalation exposure to
 7    vanadium pentoxide than mice following subchronic exposure to vanadium pentoxide, based on
 8    the occurrence of a wider variety of nonneoplastic lesions throughout the respiratory tract.
 9          Knecht et al. (1992) also observed that inhaled vanadium pentoxide leads to decreased
10    pulmonary function in male monkeys exposed to vanadium pentoxide at 0.1, 0.5, or 1.1 mg/m3
11    for 26 weeks (6 hours per day, 5 days per week).  However, effects on pulmonary function
12    parameters were reversible and were not replicated following subsequent challenge.
13          NTP (2002) also reported nonneoplastic pulmonary lesions in male and female rats and
14    mice exposed to vanadium pentoxide via inhalation for 2 years.  Specifically, increased alveolar
15    and bronchiolar epithelial hyperplasia in male rats; alveolar histiocyte infiltration, laryngeal
16    inflammation, and epiglottis epithelial degeneration, hyperplasia, and squamous metaplasia in
17    male and  female rats;  and goblet cell hyperplasia in nasal compartments in male rats was
18    observed. Based on these findings, the study LOAEL was 0.5 mg/m3 based on nonneoplastic
19    lesions of the lungs, larynx, and nose.  Increased incidences of pulmonary inflammation and
20    hyperplasia, nasal olfactory or respiratory epithelium degeneration, and epiglottis metaplasia
21    were reported in both  male and female mice.  The lung fibrosis and nasal inflammation were
22    noted at >2 mg/m3 in both sexes. In general, the incidence and severity of pulmonary lesions
23    increased with increasing dose. No treatment-related histopathological lesions were observed in
24    other evaluated tissues. Interstitial fibrosis was significantly elevated in exposed male and
25    female mice compared to controls at 2 and 4 mg/m3.  A LOAEL of 1.0 mg/m3 was identified
26    based on these nonneoplastic lesions of the lungs, larynx, and nose.  Increased pulmonary
27    irritation was also observed following pharyngeal aspiration in three mouse strains  (Rondini et
28    al., 2010).
29           Other effects observed following inhalation exposure to vanadium pentoxide included
30    CNS, hematological, and male reproductive effects. Neurotoxicity also has been observed in
31    mice following a single inhalation exposure.  Avila-Costa et al. (2006) noted significant changes
32    in brain morphology and impairment of memory in male CD-I mice compared to controls
33    following a single 1-hour exposure to 0.02 M (2.5 mg/m3 vanadium) inhaled vanadium
34    pentoxide. Selected hematologic parameters (number of erythrocytes and reticulocytes and
35    percent hematocrit) were significantly altered in male and female rats at the highest dose level
36    tested following subchronic exposure to vanadium pentoxide (NTP, 2002). Other subchronic
37    studies have also reported increased platelet counts (Gonzalez-Villalva et al., 2006).
38    Reproductive effects due to vanadium pentoxide inhalation exposure included morphological
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  63      DRAFT -  DO NOT CITE OR QUOTE

-------
1
2
3
4
5
changes to spermatogonia, spermatocytes, and Sertoli cells and testicular malformation (Fortoul
et al., 2007; Mussali-Galante et al., 2005; NTP, 2000).  Estrous cycle length was increased in
female rats but not in mice following inhalation exposure at high doses, while decreased
spermatozoal motility was observed in male mice but not in male rats at high doses (NTP, 2002).
     Table 4-14.  Summary of the noncancer results for the subchronic and chronic inhalation
    exposure studies of vanadium pentoxide toxicity in experimental animals

Species (sex)
Male CD-I
mice (n = 48)







Male CD-I
mice
(n=8)


Male CD-I
mice (n = 60)



Male
Cynomolgus
Monkey
(n = 8/group)















Exposure
(mg/m3)
2.56 mg/m3
1 hour, 2
times/week
for 8 weeks





2.56 mg/m3
1 hour/day,
2 days/week
for 12
weeks
2.56 mg/m3
1 hour/day,
2 days/week
for 12
weeks
0.1,0.5, or
1.1 mg/m3
6 hours/day,
5 days/week
for 26
weeks














Response at LOAEL
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


Impaired pulmonary
function

















NOAEL
(mg/m3)
-








-




-




0.10
(continuous
exposure)
















LOAEL
(mg/m3)
2.56








2.56




2.56




not
established


















Comments
No clinical signs of
toxicity or other
toxicologic
endpoints were
reported.




None




None




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.

Reference
Avila-Costa
et al. (2006;
2005: 2004)






Gonzalez-
Villalva et
al. (2006)


Mussali-
Galante et
al. (2005):
Fortoul et
al. (2007)
Knecht et
al. (1992)

















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

-------
 Table 4-14. Summary of the noncancer results for the subchronic and chronic inhalation
exposure studies of vanadium pentoxide toxicity in experimental animals

Species (sex)
F344/N rat
(n = 10/sex/
group)





B6C3FJ mice
(n =
10/sex/group)



F344/Nrat (n
=
50/sex/group)















B6C3FJ mice
(n =
50/sex/group)












Exposure
(mg/m3)
0, 1, 2, 4, 8,
16 mg/m3
6 hours/day,
5 days/week
for 3 month



0, 1, 2, 4, 8,
16 mg/m3
6 hours/day,
5 days/week
for 3
months
0,0.5, 1,
2 mg/m3
6 hours per
day, 5 days
per week
for 2 years












0, 1, 2,
4 mg/m3
6 hours per
day, 5 days
per week
for 2 years










Response at LOAEL
Lung epithelium hyperplasia
and inflammation, increased
relative lung weight, altered
levels of hematological
parameters in males. Lung
epithelium hyperplasia and
increased absolute lung
weight in females.
Increased absolute lung
weight and lung epithelium
hyperplasia in males. Lung
epithelium hyperplasia and
inflammation in females.

Effects were observed in 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) in
males. Effects were
observed in the lungs
(interstitial fibrosis and
alveolar histiocyte
infiltration) and larynx
(inflammation and epiglottis
degeneration and
hyperplasia) in females.
Effects were observed in the
lungs (alveolar and
bronchiolar epithelial
hyperplasia, inflammation,
alveolus histiocyte
infiltration in both males and
females), larynx (squamous
metaplasia of the epiglottis
in both males and females),
and nose (olfactory and
respiratory epithelium
degeneration in males and
olfactory epithelial
degeneration and atrophy in
females)
NOAEL
(mg/m3)
1.0







1.0





—

















—














LOAEL
(mg/m3)
2.0







2.0





0.5

















1.0















Comments
None







None





No other clinical
findings or altered
survival.















Mice reported as
thin, survival
significantly
decreased in male
mice at 4 mg/m3.











Reference





NTPQ002)























NTP OOO?^
IN _L IT \Zj\j\J Zj }
















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

-------
 1    4.6.3.  Mode of Action Information
 2          Currently, evidence is insufficient to establish the mode of action (MOA) for vanadium
 3    pentoxide noncancer toxicity. The limited data to inform the mechanisms of various noncancer
 4    health effects (pulmonary toxicity, neurotoxicity, and reproductive toxicity) following inhalation
 5    exposure to vanadium pentoxide are summarized below.

 6    4.6.3.1.  Pulmonary Toxicity
 1          Evidence suggests that a potential mechanism of action underlying the formation of
 8    pulmonary fibroproliferative lesions may involve inflammation, leading to a regenerative
 9    hyperplastic response (i.e., presence of mitogens and observed smooth muscle thickening).
10    Oxidative stress also has been implicated. Oxidative stress induced directly or indirectly by
11    vanadium pentoxide may work in  combination with vanadium pentoxide-induced inflammatory
12    responses to activate signaling molecules such as ERK1/2 and p38 kinase that lead to induction
13    of growth factors and generation of fibrotic lesions. These potential modes of action are
14    supported by a limited number of mechanistic analyses of inflammation and  fibrosis following
15    exposure to vanadium pentoxide.
16          Inflammation. Pierce et al.  (1996) identified proinflammatory cytokines associated with
17    vanadium pentoxide exposure. Evidence suggests that oxidative stress may be involved in
18    vanadium pentoxide-induced pulmonary injury. All species of vanadium can participate in redox
19    cycling and generate reactive oxygen species (Carter et al., 1997).  Intracellular and extracellular
20    hydrogen peroxide production increases significantly after 18-hour exposure to vanadium
21    pentoxide compared to controls (Antao-Menezes et al., 2008). Vanadium pentoxide also induced
22    significant IFN-|3 expression after 18 and 24 hours, however this expression was  inhibited by
23    catalase, an inhibitor of hydrogen peroxide.  Oxidative stress was also a suggested mechanism
24    underlying STAT-1 activation. Addition of either an NAPDH inhibitor or a xanthine oxidase
25    inhibitor ablated STAT-1 activation in cells exposed to vanadium pentoxide  (Antao-Menezes et
26    al., 2008).  Moreover, spontaneous hydrogen peroxide generation by fibroblasts was depleted
27    within minutes by addition of vanadium pentoxide (Ingram et al., 2003). In addition, vanadium
28    can contribute to inhibition of protein tyrosine phosphatases through the generation of reactive
29    oxygen species (Zhang et al., 2001). Hydrogen peroxide and vanadium pentoxide could have
30    reacted to form peroxovanadium intermediates or ROS, or both, that led to the production of HB-
31    EGF in fibroblasts.  Oxidative stress could also be inducing ERK and p38 kinase leading to the
32    production of HB-EGF (Ingram et al., 2003) or by ROS-mediated competitive inhibition of
33    protein tyrosine phosphatases (Zhang et al., 2001)(Samet et al., 1999).  Gene array analysis and
34    subsequent confirmation by PCR revealed that various oxidative stress genes such as superoxide
35    dismutase (SOD2), pipecolic acid  oxidase (PIPOX), and oxidative stress response (OXR1) were
36    altered by vanadium pentoxide exposure (Ingram et al., 2007). Thus vanadium pentoxide-
37    mediated production of ROS might lead to oxidative stress and induce downstream signaling

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

-------
 1    events that result in activation of mitogens and proinflammatory cytokines that contribute to the
 2    formation of fibroproliferative lesions in the lung.
 3          Fibrosis. Bonner et al. (1998) has published numerous studies describing the mechanism
 4    underlying the formation of fibroproliferative lesions in response to vanadium pentoxide
 5    exposure. Specifically, smooth muscle thickening and increased collagen deposition were
 6    observed beneath ciliated epithelial cells in vanadium pentoxide-exposed male Sprague-Dawley
 7    rats (Bonner et al., 1998). Additionally, proliferating myofibroblasts were the principal cell type
 8    that contributed to the observed fibrosis (Bonner et al., 2000). Using both in vitro and in vivo
 9    models, Rice et al. (1999) showed that inhibition of autophosphorylation of tyrosine kinases
10    reduced vanadium pentoxide-induced pulmonary fibrosis, thus implicating the tyrosine kinases
11    as key signaling mediators underlying the mechanism of PDGF release and ultimately,
12    fibrinogenesis.
13          Zhang et al. (2001) and Ingram et al. (2003) identified a second mitogen growth factor
14    similar to heparin-binding epidermal growth factor (HB-EGF) as an important mediator of
15    vanadium pentoxide-induced injury in vitro in normal human bronchial epithelial cells
16    (NHBECs). Further, two signaling molecules (ERK and the p38 subunit of MAP kinase) were
17    activated in response to vanadium pentoxide (Ingram et al., 2003). Gene array analysis
18    confirmed the importance of Fffi-EGF and IL-8 in vanadium pentoxide-induced lung injury and
19    identified several new candidate genes, including growth factors (VEGF, CTGF), chemokines
20    (CXCL9,  CXCL10),  oxidative response genes (SOD2, PIPOX, OXRI), and DNA-binding
21    proteins (GAS1, STAT1) (Ingram et al.. 2007).
22          Bonner et al. (2002) also has reported that fibroproliferative lesions are resolved and
23    repair is initiated in response to vanadium pentoxide. Their work illustrates that mice deficient
24    in prostaglandins (PG), such as the enzyme cyclooxygenase (COX)-2, are protected from
25    vanadium pentoxide-induced fibroproliferative lesions and indicates the potentially important
26    role of cyclooxygenases in mitigating vanadium pentoxide induced injury.  Antao-Menezes et al.
27    (2008) characterized  the role of STAT-1 in vanadium pentoxide pulmonary fibrosis. Their work
28    identified Interferon-beta (IFN-P) as a mediator of vanadium pentoxide-induced STAT-1
29    activation in normal human lung fibroblasts, and linked STAT-1 activation with STAT-1-
30    dependent production of a chemokine, CXCL10, that was identified in the Ingram et al. (2007)
31    gene array analysis. Thus, fibroblasts appear to synthesize IFN-P that activates STAT-1.  STAT-
32    1 activation simultaneously causes growth arrest and increases levels of CXCL10, which then
33    diminish fibrinogenesis, in a negative feedback loop. In summary, vanadium pentoxide-
34    stimulated production of IFN-P activates  signaling pathways that lead to the resolution of
35    fibrosis after vanadium pentoxide-induced injury.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 67      DRAFT - DO NOT CITE OR QUOTE

-------
 1    4.6.3.2.  Neurotoxicity
 2          The mechanism(s) underlying nervous system toxicity in response to vanadium pentoxide
 3    is not well characterized. A duration-dependent decrease in the number of immunoreactive TH+
 4    neurons and morphological changes to the blood-brain barrier were observed in response to
 5    vanadium pentoxide. Blood-brain barrier disruption has been suggested to be related to brain
 6    region-specific changes in metalloproteinases (MMP-2 and MMP-9) that have been observed
 7    following vanadium pentoxide exposure (Colin-Barenque et al., 2008): more work, however, is
 8    needed to characterize these findings fully.

 9    4.6.3.3.  Reproductive Toxicity
10          Limited studies of vanadium pentoxide have examined reproductive and developmental
11    toxicity.  Short-term oral exposures in weanling rats demonstrated a potential effect on bone
12    growth as measured by serum calcium concentrations and bone alkaline phosphatase activity
13    (Yamaguchi etal., 1989). Excess vanadium pentoxide was found to accumulate in the testes
14    following inhalation exposure (Mussali-Galante et al., 2005). Moreover, significant decreases in
15    gamma globulin were observed in testicular samples exposed to vanadium pentoxide (Mussali-
16    Galante et al., 2005). Decreased gamma-globulin levels could lead to changes in microtubule
17    formation that would impact spermatogenesis. Injection studies described by Wide (1984)
18    demonstrated decreased ossification in fetuses exposed to vanadium pentoxide in utero.
19    Determination of a specific MO A for reproductive toxicity is not possible due to the limited
20    studies examining these effects.

21    4.7. EVALUATION OF CARCINOGENICITY

22    4.7.1.  Summary of Overall Weight of Evidence
23          Under the U.S. EPA Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a),
24    vanadium pentoxide is "likely to be carcinogenic to humans" by the inhalation route of exposure
25    based on alveolar/bronchiolar tumors in male and female mice and rats following inhalation
26    exposure to vanadium pentoxide for 2 years (Ress et al.. 2003: NTP, 2002). NTP (2002)
27    reported statistically significant and dose-related increased incidence of alveolar/bronchiolar
28    tumors in male and female mice. Although the incidence of bronchiolar tumors in vanadium
29    pentoxide-treated rats was not significantly increased compared to control, tumor incidence was
30    elevated relative to historical control in most treatment groups in male rats and  some treatment
31    groups in female rats (see Table 4-9). No other tumor type was significantly increased in either
32    rats or mice in this study. These results are supported by a recent study by Rondini et al. (2010),
33    which also showed increased lung tumors in male mice (A/J, BALB/C) following exposure to
34    vanadium pentoxide along with an initiator (MCA).
35          U.S. EPA's Guidelines for Carcinogenic Risk Assessment (2005a) indicates that, for
36    tumors occurring at a site other than the initial point of contact, the weight of evidence for
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 68     DRAFT - DO NOT CITE OR QUOTE

-------
 1    carcinogenic potential could apply to all routes of exposure that have not been adequately tested
 2    at sufficient doses. An exception occurs when there is convincing information (e.g.,
 3    toxicokinetic data) that absorption does not occur by other routes. Information available on the
 4    carcinogenic effects of vanadium pentoxide via the inhalation route is limited to examination of
 5    the respiratory tumors. Information on the carcinogenic effects of vanadium pentoxide via the
 6    oral and dermal routes in humans or animals is absent.  Based on the observance of only
 7    portal-of-entry tumors (respiratory tumors) following inhalation exposure, and in the absence of
 8    information to establish an MO A, this  cancer descriptor applies only to the inhalation route of
 9    exposure. Therefore, vanadium pentoxide is "likely to be carcinogenic to humans" by the
10    inhalation route of exposure, and the database has "inadequate information to assess
11    carcinogenic potential" via the oral or dermal route.

12    4.7.2.   Synthesis of Human, Animal, and Other Supporting Evidence
13           Few studies are available that assess the carcinogenic potential of vanadium pentoxide.
14    Although both epidemiology and laboratory animal studies show similar respiratory tract
15    toxicity, currently no epidemiology studies are available in the published literature examining
16    carcinogenicity.
17           Only two published laboratory  animal studies provide evidence for the carcinogenic
18    potential of vanadium pentoxide (Rondini et al., 2010; NTP, 2002).  Vanadium pentoxide has
19    been shown to induce pulmonary tumors following inhalation exposure NTP (2002).  F344/N
20    rats (50 per sex per group) and B6C3Fi mice (50 per sex per group) were exposed to vanadium
21    pentoxide particles for 6 hours per day, 5 days per week, for 2 years. Rats were exposed to  0.5,
22    1.0, or 2.0 mg/m3 of vanadium pentoxide, with mice exposed to 1.0, 2.0, or 4.0 mg/m3 of
23    vanadium pentoxide.  Lung tumors were observed in male and female rats, but incidence
24    exceeded historical controls only in male rats (Table 4-9).  Both male and female mice showed
25    statistically significant increases in lung tumors as compared to controls (p < 0.01). These
26    increases were observed in both sexes  at all doses, with 50%  of the male mice in the highest
27    exposure group dying before the end of the study. Survival rates in all other exposed groups for
28    both rats and mice were not significantly different from controls. Both rats and mice  showed
29    other lesions of the respiratory tract, including inflammation, fibrosis, and hyperplasia (Table 4-8
30    and Table 4-10). Decreased  body weight gain was observed as early as 3 months postexposure
31    in the high dose groups of all exposed  animals.
32           Along with the NTP study (2002), a recent study by Rondini et al. (2010) examined
33    tumor promotion of vanadium pentoxide in three different mice strains.  Lung tumors were
34    observed in two of three mouse strains 20 weeks after MCA tumor initiation,  followed by
35    exposure (only males exposed) to vanadium pentoxide (4 mg/kg, 5 times weekly) (Table 4-12).
36    A/J and BALB/C male mice showed increases in lung tumors following exposure to both MCA
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  69      DRAFT - DO NOT CITE OR QUOTE

-------
 1    (initiator) and vanadium pentoxide, but not vanadium pentoxide alone, suggesting vanadium
 2    pentoxide may act as a tumor promoter.

 3    4.7.3.  Mode of Action Information
 4          Alveolar/bronchiolar tumors were observed in both mice and rats following inhalation
 5    exposure to vanadium pentoxide.  In vitro tests in bacterial and yeast systems provide mixed
 6    evidence of vanadium pentoxide-induced mutagenicity.  In general, gene mutation assays were
 7    negative, as were tests that assessed SCE and other chromosomal aberrations. DNA strand
 8    breaks (Ivancsits  et al., 2002; Rojas et al., 1996) and micronucleus formation (Zhong et al.,
 9    1994) were reported in some studies in cultured cells but depended on cell type. Information is
10    insufficient to establish a MOA for bronchiolar tumors observed in animals following inhalation
11    exposure to vanadium pentoxide.

12    4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES

13    4.8.1.  Possible Childhood Susceptibility
14          There are  no reports of childhood susceptibility due to exposure to vanadium pentoxide.
15    There are also no reports indicating increased susceptibility in the developing fetus but data on
16    developmental effects of vanadium pentoxide are very limited (see discussion of data gaps
17    below).  Young rats that consumed vanadium in the drinking water and feed were  found to have
18    higher tissue vanadium levels 21  days after birth than they did 115 days after birth (Edel et al.,
19    1984). The data suggest higher absorption of vanadium  in young animals  due to a greater
20    nonselective permeability of the undeveloped intestinal barrier. Thus, age of the rodents appears
21    to play an important role in the absorption of vanadium in the GI tract. Mravcova et al. (1993)
22    assessed the extent of vanadium pentoxide accumulation in the bones of rats following a 6-
23    month exposure.  Vanadium accumulated in the epiphyseal cartilage of the tibia in rats with
24    significantly higher concentrations of vanadium in the tibia and incisors of weanling rats
25    compared to adults. No dose-response data for these endpoints were reported.

26    4.8.2.  Possible Gender Differences
27          Reports of gender differences are limited to the carcinogenicity data from NTP (2002)
28    where clear evidence of carcinogenicity was reported in  male and female mice exposed to
29    vanadium pentoxide. Some evidence of carcinogenicity  was reported in male rats, based on
30    observations of alveolar and bronchiolar neoplasms that  exceeded historical controls in groups
31    exposed to vanadium pentoxide.  The number of neoplasms in female rats  was not higher than
32    that observed in historical controls and, thus, a relationship between neoplasms and vanadium
33    pentoxide could not be established in female rats (NTP, 2002)(Ress, 2003).  Therefore,  increased
34    tumor incidence in rats is equivocal overall but the lack of any increase in  females suggests a
35    gender-related increase in susceptibility in males. Whether this observed difference is applicable

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

-------
 1    to humans is unknown.  There are no reported differences for response to vanadium pentoxide
 2    between genders for either animals or humans for any noncancer endpoint.

 3    4.8.3.  Other Susceptible Populations
 4          No data exist on the role of genetic polymorphisms in differentially susceptible human
 5    populations in response to vanadium pentoxide exposure. In mice, research suggests that all
 6    strains have an inflammatory response to vanadium pentoxide exposure, but the severity of
 7    inflammation varies greatly from strain to strain.  Such variability in response suggests that a
 8    genetic component contributes to the severity of vanadium pentoxide-induced pulmonary
 9    inflammation and tumorigenicity in mice (Rondini et al., 2010). The NTP study (2002) used
10    B6C3Fi  mice in all exposure protocols for 16-day, 3-month, and 2-year exposure studies.
11    Pulmonary fibrosis was observed in this hybrid strain.
12          Kyono et al. (1999) used a rat model of acute bronchiolitis (Br) to investigate whether
13    animals with preexisting lung conditions would be differentially susceptible to inhaled vanadium
14    pentoxide.  Compared to exposed normal rats, Br rats exhibited delayed recovery from
15    preexisting lesions and exacerbated lung inflammation. Sensitive rats also showed reductions in
16    the deposition and clearance rates of inhaled particles.
17          Rondini et al. (2010) examined the effect of exposure to vanadium pentoxide in three
18    mouse strains of varying susceptibility to lung cancer (A/J, BALB/C, and C57BL/6J) in an
19    initiation/promotion model (see full study description in Appendix D).  Three mouse strains were
20    used to further understand potential susceptibility to these effects.  These particular mouse  strains
21    were selected because of their known differential susceptibility to chronic  pulmonary
22    inflammation and carcinogenesis: A/J mice are sensitive, BALB/C are intermediate, and
23    C57BL/6J are resistant. Statistically significant lung tumor increases were observed in A/J and
24    BALB/C mice as compared to the MCA-treated control (p < 0.05; Table 4-12).  Differences were
25    also observed between strains, with A/J mice showing increased tumorigenicity in response to
26    vanadium pentoxide.  In the  absence of MCA, vanadium pentoxide was not sufficient to initiate
27    tumorigenesis in this study.  C57BL/6J had no tumors following exposure (data not shown).
28          Overall, the differential inflammatory responses observed in the three strains of mice
29    appear to positively correlate with increased levels of chemokines, such as keratinocyte-derived
30    chemokine (KC) and monocyte chemotactic protein-1  (MCP-1), increased binding of
31    transcriptional factors NFicB and AP-1 (c-Fos), sustained activation of MAP kinases (MAPKs),
32    and extracellular signal-regulated kinases  1 and 2 (ERK 1/2), suggesting inflammation as a major
33    response in mice. Turpin et  al. (2010) examined pulmonary inflammation and fibrosis following
34    intranasal aspiration exposure to vanadium pentoxide with and without RSV exposure (see full
35    study description in Appendix D).  In this study, vanadium pentoxide exposure also caused a
36    significant increase in cell proliferation in the airways and lung parenchyma, lung mRNAs for
37    TGF-p-1, CTGF, PDGF-C, CollA2, and mRNAs for IFN-a and -p, and IFN-inducible

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

-------
1   chemokines CXCL9 and CXCL10 compared to controls. Pre- or posttreatment with RSV caused
2   a significant reduction in all mRNAs. Together, results from this study showed that vanadium
3   pentoxide induces inflammatory and fibrogenic response in mouse lung, and these effects were
4   suppressed by RSV infection. Yu et al. (2011) demonstrated increased mucin production from
5   the airway epithelium in female AKR mice following laryngeal aspiration of vanadium
6   pentoxide (4 mg/kg bw).  These results were supported by in vitro studies showing increased
7   gene expression for genes related to mucin production following exposure to vanadium
8   pentoxide.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                72      DRAFT - DO NOT CITE OR QUOTE

-------
 1                               5. DOSE-RESPONSE ANALYSIS

 2    5.1.  ORAL REFERENCE DOSE (RfD)

 3    5.1.1. Choice of Principal Study and Critical Effect with Rationale and Justification
 4          Two human studies are available which measured cystine levels in hair and fingernails
 5    and vanadium levels in urine (Kucera et al., 1994) or blood (Kucera et al., 1992) following oral
 6    exposure to vanadium pentoxide.  Kucera et al. (1994) detected vanadium in urine of workers
 7    from a Czechoslovakian vanadium pentoxide production plant; these urinary levels of vanadium,
 8    however, are suspected of having resulted from multiple exposure routes.  Kucera et al. (1992)
 9    measured vanadium in the hair and blood of children and in the blood of adults potentially
10    exposed through ingestion of vanadium-contaminated drinking water (concentration range:
11    0.001-0.1 mg/L). Vanadium concentrations in water supply wells exceeded the maximum
12    permissible limit in drinking water (0.01 mg/L) with the contamination continuing over 2 years.
13    Significantly increased vanadium concentrations were found in blood of exposed children
14    compared to unexposed children and adults, whereas vanadium levels in hair of exposed children
15    (adults not measured) were no different from the control group. No exposure-response
16    relationship could be determined for either endpoint. Changes in hair cystine levels have not
17    been associated with adverse health effects. Additional studies of workers occupationally
18    exposed to vanadium pentoxide exist, although these workers  are presumed to have been
19    exposed by multiple routes, and inhalation was likely the primary route of exposure (Kiviluoto et
20    al., 1981b)(Irsigleretal., 1999: Musk and Tees. 1982: Kiviluoto. 1980: Kiviluoto et al.. 1979:
21    Zenzetal.. 1962: Sioberg, 1956. 1951).
22          The results of a 2.5-year dietary study on vanadium in rats (strain not described)
23    (Stokinger et al., 1953) were summarized in Patty's Industrial Hygiene and Toxicology, 3r  ed.
24    (1981). This study  did not assess comprehensive toxicity endpoints, and details of the study are
25    limited in the available publication.  Although hair cystine content was stated to be significantly
26    decreased in exposed animals compared to controls, no quantitative data were presented.  The
27    biological significance of decreased hair cystine is unclear, however, and this effect is known to
28    be caused by other chemical exposure.  Due to these limitations, this study was not selected for
29    the principal study.  No additional oral chronic exposure studies in animals were identified in the
30    published literature.
31          Mountain et al. (1953) evaluated the effects of subchronic oral vanadium pentoxide
32    exposure in laboratory animals. Male Wistar rats were exposed to vanadium pentoxide for 103
33    days using average  daily  doses of 0,  10.5, 16.4, 69.6, 141.0 mg/kg-day vanadium pentoxide in
34    feed. Changes were observed in body weight gain, erythrocyte count, hemoglobin, and cystine
35    content of hair. Other endpoints of toxicity were  not reported.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                73        DRAFT - DO NOT CITE OR QUOTE

-------
 1          The study authors reported increased body weight gain in the low exposure groups
 2    (0-16.4 mg/kg-day) and decreased body weight gain in the highest exposure group
 3    (141.0 mg/kg-day). No explanation, statistical analysis, or measures of variability (standard
 4    error [SE] or standard deviation [SD]), however, accompanied these data. Cystine content of
 5    hair was significantly decreased compared to control at doses>16.4 mg/kg-day vanadium
 6    pentoxide. Alterations in hair cystine levels can be associated with dietary changes or altered
 7    health status (Kleinfeld et al., 1961). The authors speculated that vanadium may have inhibited
 8    enzymes, such as sulfur transferases, that decreased the availability of cystine for hair growth.
 9    Data to support a relationship between decreased hair cystine levels and adverse health
10    outcomes, however, are not available in the published literature. Thus, the biological
11    significance of decreased hair cystine content is unclear.
12          Mountain et al. (1953) also reported relative liver weight increases and decreases in
13    erythrocytes and hemoglobin levels. Mean relative liver weights, reported as a ratio of liver
14    weight to body weight, were statistically significantly elevated above controls at 69.6 mg/kg-day
15    (i.e., 3.51 ± 0.06 versus 3.86 ± 0.07). Liver weight data were not reported for other doses.
16    Apparent dose-related decreases in red blood cell (RBC) count (21.3%) and hemoglobin
17    concentration (10.5%) were observed in the 10.5 and 16.4 mg/kg-day vanadium pentoxide dose
18    groups compared to controls (Table 5-1). Study authors performed no statistical analysis on
19    these data and reported no measure of variance, precluding independent statistical analysis. The
20    effects on RBC count and hemoglobin concentration observed in this oral study are consistent
21    with the hematological effects observed in a 3-month inhalation study of vanadium  pentoxide in
22    rats (NTP, 2002).  Decreases in mean cell volume (MCV) and mean cell hemoglobin (MCH)
23    accompanied by erythrocyte microcytosis suggested altered iron metabolism and
24    heme/hemoglobin production following vanadium pentoxide inhalation exposure. Based on the
25    dose-related decreases in RBC count in the oral study supported by the hematological effects
26    observed in the inhalation study, decreased RBC count was considered adverse and selected as
27    the critical effect.
28          Although a dose-related decrease in RBC count was observed at 10.5 and 16.4 mg/kg-day
29    vanadium pentoxide (12.8% and 21.3%,  respectively) in the Mountain et al. (1953)  study, the
30    magnitude of this change was considered biologically significant at 16.4 mg/kg-day, which EPA
31    identified as the lowest-observed-adverse-effect level (LOAEL). The lowest dose of 10.5
32    mg/kg-day was identified by EPA as a no-observed-adverse-effect level (NOAEL).
33          Mountain et al., (1953) was selected as the principal study for the derivation of the RfD.
34    This study utilized four dose groups and an untreated control group with five animals per group
35    and evaluated several hematological endpoints. Limitations of the study include lack of
36    evaluation of a comprehensive number of toxicological endpoints.  The study authors reported
37    dose dependent, statistically significant decreases in RBC counts (Table 5-1) following
38    inhalation exposure to vanadium pentoxide which was selected as the critical effect for the
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 74        DRAFT - DO NOT CITE OR QUOTE

-------
 1    derivation of the RfD.  Specifically, the percent change at the low and high dose was 12.8 and
 2    21.3%, respectively compared to 3.8% in the controls. The percent change in RBC count was
 3    considered adverse and thus was selected as the critical effect for the derivation of the RfD.

            Table 5-1. Hematological results of oral vanadium pentoxide exposure in
            rats

Control
10.5 mg/kg-d
16.4 mg/kg-d
Red Blood Cell Count (M/mm3)
Start
Finish (103d)
Percent change between start and finish of
experiment (%)
8.0
7.7
3.8
7.8
6.8
12.8
8.0
6.3
21.3
Hemoglobin, %
Start
Finish (103d)
Percent change between start and finish of
experiment (%)
15.6
15.0
3.9
15.2
14.5
4.6
15.3
13.7
10.5
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
 Source: Mountain et al. (1953).

5.1.2.  Methods of Analysis—Including Models (e.g., PBPK and BMD)
       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.
       In EPA's guidance document Recommended Use of Body Weight3'4 as the Default
Method in Derivation of the Oral Reference Dose (U.S. EPA, 2011), the Agency endorses a
hierarchy of approaches for converting doses administered orally to laboratory animal species to
human equivalent oral exposures in deriving the RfD, with the preferred approach being
physiologically-based toxicokinetic modeling.  An alternate approach includes using chemical-
specific information in the absence of a complete physiologically-based toxicokinetic model. In
lieu of a toxicokinetic model or chemical-specific data  to inform the generation of human
equivalent oral exposures, EPA endorses body weight scaling to the % power (i.e., BW3 4) as a
default to extrapolate lexicologically equivalent doses of orally administered agents from
laboratory animals to humans for the purpose of deriving an RfD. When BW3 4 scaling is used in
deriving the RfD, EPA also advocates a reduction in the interspecies uncertainty factor from 10
to 3, as BW3/4 scaling addresses predominantly toxicokinetic (and some toxicodynamic) aspects
of the UF A.
       Statements in the guidance raise some important uncertainties in applying allometric
scaling, and more specifically BW3 4 scaling, when trying to extrapolate across different sized
           This document is a draft for review purposes only and does not constitute Agency policy.
                                           75       DRAFT - DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

15
17
18
19
20
21
22
23
24
25
26
27
28
individuals within a species (e.g., between neonates and adults) or across individuals in different
lifestages between species (e.g., between fetal rats and adult humans). Furthermore, the data on
which to base a default allometric scaling factor for converting the administered dose in a
laboratory animal in a different lifestage to a comparable dose in an adult human are sparse, and
thus more uncertain.  For these reasons, a BW3 4 scaling factor would not be applied as a default
approach (in combination with a reduced default UF for interspecies extrapolation) when
extrapolating from developmental effects in laboratory animals to adult humans when deriving
theRfD.
       No physiologically based  toxicokinetic modeling information exists for vanadium
pentoxide.  The selected critical effect (decreased RBC 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 (FED) through application of a
dosimetric adjustment factor (DAF) derived as follows (Table 5-2):
                                             1/4
                                                    l/4x
16    where:
       DAF
       BWa
                                DAF= (BWa1/
dosimetric adjustment factor
animal body weight
human body weight
       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 an FED of 2.56 mg/kg-day as
follows:

       FED   = 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 equivalent dose conversion by BW3/4 for RfD derivation3
Species

Rat
(0.25kg)
BWa1/4 / BWh1/4 = DAF

0.25kg1/4-70kg1/4 = 0.244
RED

10.5 mg/kg-d x
0.244 = 2.56 mg/kg-d
UF
Total = 3000
UFA = 3
UFH = 10
UFS = 10
UFD = 10
RfD(mg/kg-d)

2.56 - 3000
= 8.5 x 10~4
     a Using the BWa of 0.25 kg for rats and BWh of 70 kg for humans, and multiplying it by the NOAEL of 10.5 mg/kg-d
     from Mountain et al. (19531.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                76        DRAFT - DO NOT CITE OR QUOTE

-------
 1    5.1.3.  RfD Derivation—Including Application of Uncertainty Factors
 2          The NOAEL of 10.5 mg/kg-day for decreased RBC count in male rats (Mountain et al.,
 3    1953) was used as the POD to derive an RfD.  The uncertainty factors, selected based on EPA's
 4    A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA, 2002;
 5    Section 4.4.5), address five areas of uncertainty resulting in a composite UF of 3000. This
 6    composite uncertainty factor was applied to the selected POD to derive an RfD.
 7          An interspecies uncertainty factor, UFA, of 3 (10°'5 = 3.16, rounded to 3) was applied to
 8    account for uncertainty in extrapolating from laboratory animals to humans in the absence of
 9    information to characterize the toxicokinetic or toxicodynamic differences between rats and
10    humans after oral vanadium pentoxide exposure. For vanadium pentoxide,  toxicokinetic
11    uncertainty was accounted for by calculating an FED through application of a  DAF as outlined
12    in the U.S. EPA guidance on the use of BW3/4 scaling in the derivation of the oral RfD (U.S.
13    EPA, 2011). As the toxicokinetic differences are thus accounted for, only the toxicodynamic
14    uncertainty remains, and a UFA of 3 is retained to account for this residual uncertainty.
15          An intraspecies uncertainty factor, UFn, of 10 was applied to account for potentially
16    susceptible individuals in the  absence of data evaluating the variability of response to oral
17    vanadium pentoxide exposure in the human population.
18          A subchronic-to-chronic uncertainty factor, UFs, of 10 was applied for  extrapolation from
19    a subchronic to chronic study because a subchronic study was used for the POD.
20          A LOAEL-to-NOAEL uncertainty factor, UFL, of 1 was applied because a NOAEL was
21    identified from the principal study and used as the POD (U.S. EPA. 2002).
22          A database uncertainty factor, UFo, of  10 was applied to account for database
23    deficiencies due to the lack of a developmental toxicity study and a multigenerational
24    reproductive toxicity study for vanadium pentoxide by the oral route.  Studies using alternate
25    routes of exposure (intraperitoneal) have indicated adverse reproductive and developmental
26    effects in response to vanadium pentoxide, including statistically significant increases in seminal
27    vesicle, thymus, and submandibular gland weights in male mice and body weight, thymus,
28    submandibular gland, and liver weights in female mice (Altamirano et al., 1991) as well as
29    reduced fertility (Altamirano-Lozano et al., 1996).  No physiologically based toxicokinetic
30    models are available for conducting a route-to-route extrapolation.

31          Therefore, the RfD for vanadium pentoxide is calculated as follows:
32                               RfD = NOAELHED - UF
33                                    = 2.56 mg/kg-day - 3,000
34                                    = 0.000085 mg/kg-day or 9 x 10~4 mg/kg-day

35    Note:  Because vanadium exists in several valence states, all of which are not equivalent
36    lexicologically (WHO, 2001), the values generated here apply to vanadium pentoxide and should
37    not be applied to other vanadium compounds.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                77        DRAFT - DO NOT  CITE OR QUOTE

-------
 1    5.1.4. Previous RfD Assessment
 2          U.S. EPA previously derived an RfD of 9 x 10 3 mg/kg-day for vanadium pentoxide in
 3    1987 based on a 2.5-year dietary NOAEL of 0.89 mg/kg-day vanadium pentoxide for decreased
 4    hair cystine content [U.S. EPA (1987): Stokinger et al. (1953) reported in Patty's Industrial
 5    Hygiene and Toxicology, 3r  Ed. (1981)1. Limited details were provided, and this study is not
 6    available for analysis.  The rats (number and species unspecified) were exposed to dietary levels
 7    of vanadium  pentoxide (0.89 or 8.9 mg/kg-day for 2.5 years) and assessed for growth rate,
 8    survival, and hair cystine content. Of the endpoints reported in the book chapter, decreased hair
 9    cystine was selected as the critical effect with  a NOAEL of 0.89 mg/kg-day. Upon further
10    analysis  as described in Section 5.1.1, this study was determined to be inadequate for use in
11    deriving an RfD.

12    5.2.  INHALATION  REFERENCE CONCENTRATION (RfC)

13    5.2.1. Choice of Principal Study and Critical Effect with Rationale and Justification
14          The available human and animal data identify the respiratory tract as the primary target
15    for chronic inhalation exposure to vanadium pentoxide.  Irritation of the upper and lower
16    respiratory tract has been reported in several acute, subchronic, and chronic occupational and
17    case studies of workers exposed to vanadium pentoxide  in fuel-oil ash and vanadium dust
18    (Woodin et al.. 2000: Irsigler et al.. 1999: Woodin et al.. 1999: Hauser et al.. 1995b: Levy et al..
19    1984: Musk and Tees.  1982: Kiviluoto. 1980:  Lees. 1980: Kiviluoto et al..  1979: Zenz et al..
20    1962: Lewis. 1959: Sioberg. 1955: Vintinner et al.. 1955: Williams. 1952). These results are
21    supported by effects observed in rodents, including inflammation, hyperplasia, and fibrosis.
22    Although subchronic occupational exposure studies provide supportive evidence for the
23    respiratory tract as a target for inhaled vanadium pentoxide, studies often failed to quantify
24    vanadium pentoxide concentration as a constituent in an inhaled mixture. Thus, the available
25    occupational exposure studies are not suitable as the basis for the RfC.
26          The toxicity database for inhalation exposure in laboratory animals includes two chronic
27    studies (NTP, 2002:  Knecht etal., 1992). This study was not selected as the principal study.
28    Although this study demonstrated statistically significant pulmonary responses in monkeys
29    following provocation challenge exposure to vanadium pentoxide, an absence in increased
30    pulmonary reactivity with subchronic exposure to vanadium pentoxide was observed, suggesting
31    that the monkeys may have developed a tolerance to treatment due  to the experimental  design of
32    this study.
33          Results of the subchronic and chronic NTP (2002) studies in F344/N rats and B6C3Fi
34    mice provide evidence of toxicity to the upper and lower respiratory tract, including increased
35    lung weight, inflammation, histological lesions, and decreased pulmonary function following 3-
36    month and 2-year inhalation exposures to vanadium pentoxide.  (see Sections 4.2.2.1 and
37    4.2.2.2). Specifically, effects observed in rats at the low dose (2 mg/m3) and greater included,
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                78       DRAFT - DO NOT CITE  OR QUOTE

-------
 1    increased lung weights (both males and females), hematological effects (males), and lung effects
 2    (epithelium hyperplasia in males and females and inflammation in males) following exposure to
 3    vanadium pentoxide for 3 months.  Based on decreased erythrocyte size in male rats and
 4    increased lung weights (males) and nonneoplastic lung lesions in males (epithelial hyperplasia
 5    and inflammation in the lung) and females (epithelial hyperplasia) rats, the NOAEL and LOAEL
 6    values identified for the 3-month inhalation exposure to vanadium pentoxide aerosols in rats
 7    were 1 and 2 mg/m3, respectively. Erythrocytosis was observed in male and female rats exposed
 8    to inhaled vanadium pentoxide for 3 months; hematological endpoints were not examined in
 9    mice (NTP, 2002).  The erythrocytosis was considered to be a secondary  effect arising from the
10    primary lung lesions and was not considered as a critical effect.  Effects observed in mice at the
11    low dose and greater included, increased lung weight (males) and lung effects (epithelium
12    hyperplasia in both males and females) following exposure to vanadium pentoxide for 3 months.
13    Based on increases in absolute lung weights (males) and hyperplasia of the respiratory
14    epithelium at concentrations of 2 mg/m3 and greater (males and females), NOAEL  and LOAEL
15    values identified for the 3-month inhalation exposure to vanadium pentoxide aerosols in mice as
16    of 1 and 2 mg/m3, respectively Other observed endpoints include increased lung bronchiolar
17    exudate, lung fibrosis, altered nasal morphology, and increased nasal inflammation. Significant
18    exposure-related decreases in pulmonary function were also observed in male and female rats at
19    doses of > 4 mg/m3 (NTP, 2002).  Abnormal breathing, emaciation, and lethargy were observed
20    in male and female rats exposed to concentrations of 8 mg/m3 or higher.  Final body weight was
21    statistically significantly decreased as compared to the respective controls at 16 mg/m3 in male
22    rats (60%) and male mice (10%); 8 mg/m3 in male rats (10%) and mice (6%); 16 mg/m3 in
23    female rats (30%) and female mice (12%); 8 mg/m3 in female mice (10%); and 4 mg/m3 in
24    female mice (11%).
25          Similarly, in the chronic 2 year study by NTP (2002), effects were observed on the lung,
26    larynx, and nasal tissues.  Specifically, in rats the incidences of nonneoplastic lesions of the
27    lungs (alveolar and bronchiolar epithelial  hyperplasia [males], interstitial fibrosis[females] and
28    alveolar histiocyte infiltration[both males and females]); larynx (inflammation and epiglottis
29    degeneration and hyperplasia [males and females] and squamous metaplasia [males]);  and nose
30    (goblet cell hyperplasia [males]) effects were significantly increased compared to controls in  all
31    vanadium pentoxide exposure groups.  In general, the incidences and severity ratings of
32    respiratory lesions increased with exposure level. No treatment-related histopathological
33    findings were observed in  other tissues. A LOAEL of 0.5 mg/m3 was established for
34    nonneoplastic lesions of the respiratory tract in male and female rats; a NOAEL was not
35    identified.  In mice, the incidences of nonneoplastic lesions of the lungs (hyperplasia of the
36    alveolar and bronchiole epithelium, inflammation, and alveolus histiocyte infiltration); larynx
37    (squamous metaplasia of the epiglottis); and nose (olfactory and respiratory epithelium
38    degeneration[males and females] and atrophy [females]) were significantly increased compared
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 79        DRAFT - DO NOT  CITE OR QUOTE

-------
 1    to control in all vanadium pentoxide exposure groups.  Incidences of interstitial fibrosis were
 2    significantly increased in male and female mice exposed to 2 or 4 mg/m3. In general, the
 3    incidences and severity ratings of lesions increased with exposure level and matched the types of
 4    lesions observed in rats. No treatment-related histopathological findings were observed in other
 5    tissues. The LOAEL of 1 mg/m3 was established for nonneoplastic lesions of the respiratory tract
 6    in male and female mice; a NOAEL was not identified.Other studies identified morphological
 7    changes in the central nervous system (CNS) of male mice exposed to vanadium pentoxide for
 8    up to 8 weeks (Avila-Costa et al.. 2005: Avila-Costa et al.. 2004). Avila-Costa et al. (2005:
 9    2004), also reported morphological changes in the substantia nigra region of the basal ganglia
10    and the blood-brain barrier in male mice exposed to 1.4 mg V/m3 for up to 8 weeks; effects on
11    CNS function or other comprehensive endpoints were not reported. Using the same dose regimen
12    (1.4 mgV/m3 twice per week for up to 1 month) Avila-Costa et al. (2006) reported a time-
13    dependent loss of dendritic spines, necrotic-like cell death, and morphological changes to the
14    hippocampal region and that these changes might be related to the associated spatial memory
15    loss.  Morphological changes in the CNS reported by Avila-Costa et al. (2006; 2005; 2004) were
16    not considered as the critical effect due to lack of information on the exposure-response
17    relationship for morphological changes to the CNS, as only one exposure level was tested.
18    Moreover, the lung appears to be the most sensitive target to inhalation exposure to vanadium
19    pentoxide.
20          Results of the NTP (2002) study in rats and mice provide evidence of toxicity to the
21    upper and lower respiratory tract, including increased lung weight, inflammation, histological
22    lesions, and decreased pulmonary function following 3-month inhalation exposure to vanadium
23    pentoxide. Although body weights and a complete necropsy and histological analysis were
24    performed, adverse effects in other target organs were not identified in the chronic exposure
25    study. The series of effects described in the NTP (2002)bioassay and supporting studies from the
26    available inhalation database reflect a dose-response with increase in severity with vanadium
27    pentoxide concentration, and a progression of respiratory effects (infiltration of macrophages,
28    inflammation, hyperplastic response, fibrosis). Based on the dose-response  and  temporal
29    relationship of effects throughout the course of the 2-year study, and the concordance with
30    effects observed in humans, the NTP (2002) study is selected as the critical  study.  Specifically,
31    the NTP (2002) chronic study is well-designed and well-reported utilizing 50 animals per dose
32    group, 3 dose groups, and includes an evaluation of numerous toxicological endpoints.  As with
33    the subchronic exposure studies, results of this 2-year inhalation study identify the upper and
34    lower respiratory tract as the target for chronic inhalation exposure to vanadium  pentoxide. The
35    nasal and laryngeal lesions reported in NTP (2002) are among the most sensitive effects
36    observed and were observed in both sexes of rats and mice (Table 5-3). In addition, irritation of
37    the upper and lower respiratory tract has been reported in several occupational and case studies
38    of workers exposed for days or weeks to vanadium pentoxide in fuel-oil ash and vanadium dust
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 80       DRAFT - DO NOT CITE OR QUOTE

-------
 1    (Irsigler et al.. 1999: Musk and Tees. 1982: Kiviluoto et al.. 198la: Kiviluoto et al.. 1981b:
 2    Kiviluoto. 1980: Kiviluoto et al.. 1979: Zenzetal.. 1962: Sjoberg. 1956. 1951). Therefore, NTP
 3    (2002) was selected as the principal study, with inflammation of the larynx in female rats
 4    considered adverse and selected as the critical effect.

 5    5.2.2.  Methods of Analysis
 6          To analyze the concentration-response effect of vanadium pentoxide, the reported
 7    concentrations of vanadium pentoxide were converted to human equivalent concentrations
 8    (HEC) prior to any modeling (Table 5-4 and Appendix B).  The Methods for Derivation of
 9    Inhalation Reference Concentrations and Application of Inhalation Dosimetry (hereafter referred
10    to as the RfC Methodology) recommends converting the adjusted POD (PODjADj]) to an HEC
11    (U.S. EPA, 1994b).  The RfC Methodology separates gases into three categories based on their
12    water solubility and reactivity with tissues in the respiratory tract, and recommends the use of
13    regional deposited dose ratios (RDDRs) for converting to HECs for particles (e.g., vanadium
14    pentoxide).  RDDRs were calculated for rats with the RDDR computer program (U.S. EPA,
15    1994b) using mean body weights for male and female rats and the average particle-size mass
16    median aerodynamic diameter (MMAD) ± geometric standard deviation (GSD) of 1.24 ± 1.89
17    for rats as reported by NTP (2002).  HEC conversions (in mg vanadium pentoxide/m3) were
18    calculated by multiplying the duration-adjusted exposure concentrations (Conc[ADJ]) by the
19    pulmonary RDDR for lesions in the lung, or the extrathoracic RDDR for lesions in the larynx
20    and nose,  and are summarized in Table 5-4. ConC[ADJ] of 0.09, 0.18, and 0.36 mg/m3,
21    corresponding to nominal exposure concentrations of 0.5, 1, and 2 mg/m3, were calculated to
22    account for continuous ambient exposure:

23                                  Conc[ADJ] = Cone x 6/24 x 5/7

24          HECs were calculated by multiplying ConcjADJ] by the  extrathoracic RDDR for lesions of
25    the larynx in female rats (Appendix B, Table B-5).
26          To determine the POD for derivation of the RfC, BMD modeling was conducted on
27    lesions observed in both male and female rats in the NTP study (2002) (chronic lung
28    inflammation, alveolar epithelium hyperplasia, chronic inflammation of the larynx, respiratory
29    epithelial  hyperplasia of the larynx and nose) with the best-fitting model selected for each
30    endpoint (Appendix B).
31          As shown in Table 5-5, the lowest benchmark concentration (BMCLjHEq) of
32    0.003 mg/m3 was observed for chronic inflammation of the larynx of female rats, indicating that
33    the larynx was the most sensitive target for chronic inhalation exposure to vanadium pentoxide.
34    A 53, 55,  and 76% increase in the incidence of inflammation in the larynx compared to 16% in
35    controls was observed at the low, mid, and  high doses, respectively and was considered adverse.
36    Thus, it was chosen as the POD for the basis of the RfC calculation.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                81        DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-3.  Selected nonneoplastic lesions of the respiratory system in F344/N
rats exposed to particulate aerosols of vanadium pentoxide for 2 years
Endpoint a
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
Interstitial, fibrosis
Alveolus, histocyte infiltration
50
7(14)
3(6)
5(10)
7(14)
22 (44)
49
24b (49)
17b (35)
8(16)
7(14)
40b (82)
48
34b(71)
31b(65)
24b (50)
16C(33)
45b (94)
50
49b (98)
48b (96)
42b (84)
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, histocyte 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)
     This document is a draft for review purposes only and does not constitute Agency policy.
                                      82        DRAFT - DO NOT CITE OR QUOTE

-------
       Table 5-3. Selected nonneoplastic lesions of the respiratory system in F344/N
       rats exposed to particulate aerosols of vanadium pentoxide for 2 years
Endpoint a
Exposure Group
Control
(% incidence)
0.5 mg/m3
(% incidence)
1 mg/m3
(% incidence)
2 mg/m3
(% incidence)
Nose
Number of animals examined
Goblet cell, hyperplasia
50
13 (26)
50
19 (38)
50
16 (32)
50
30b (60)
aNumber of animals with lesions; numbers in parentheses indicate percent incidence compared to control.
bSignificantly different from control by the Poly-3 test, p < 0.01.
Significantly different from control by the Poly-3 test, p < 0.05.
Source: NTP (2002).
       Table 5-4. Human equivalent concentrations of vanadium pentoxide in the
       2-year inhalation studies in F344/N rats
Concentration as
reported (mg/m3)
Continuous exposure
adjustment factor"
RDDRb
Extrathoracic
Pulmonary
HECc(mg/m3)
Extrathoracic
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
""Continuous exposure adjustment factor = (6/24) x (5/7); animals were exposed to vanadium pentoxide 6 hr/d, 5 d/wk.
"Refer to Appendix Table B-4.
dHuman equivalent concentration = concentration as reported x continuous exposure adjustment factor x RDDR.
Source: NTP (2002).
             This document is a draft for review purposes only and does not constitute Agency policy.
                                                83        DRAFT - DO NOT CITE OR QUOTE

-------
             Table 5-5.  Candidate PODs for vanadium pentoxide derived from NTP
             studies (2002) through BMDS modeling
Endpoint
Selected model"
BMR
(extra
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
Log-Logistic
Log-Logistic
0.1
0.1
0.017
0.008
0.012
0.006
Nose
Goblet cell, respiratory epithelium,
hyperplasia
Log-Logistic
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
Log-Logistic
Log-Logistic
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
      Selected model is the best-fitting model for the dataset based on the draft Benchmark Dose Technical Guidance
      (2QQQa).
      bHEC = human equivalent concentration.
      °BMCL = the lower bound of benchmark concentration (BMC) at 95% confidence level.

 1          Degeneration of the epiglottis epithelium was not selected for BMD modeling because
 2    the incidence of this lesion did not exhibit dose dependence with the same incidence observed in
 3    the low- and high-dose groups.  Epithelial squamous metaplasia was not selected for BMD
 4    modeling because the incidence of this lesion was not significantly different from control at the
 5    low- and mid-dose groups; thus, other lesions of the larynx were more sensitive endpoints. In all
 6    vanadium pentoxide groups, lesion severity was classified as minimal to mild.
 7          Modeling was performed using the Benchmark Dose Modeling Software (BMDS;
 8    Version 2.1.2) (U.S. EPA, 2000a). Biological and statistical considerations were taken into
 9    account in the selection of a benchmark response (BMR) level for all datasets. In the absence of
10    information indicating what magnitude of inflammatory  changes in the larynx and epiglottis are
11    considered biologically significant, the BMR of 10% increase in extra risk was used as the basis
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 84        DRAFT - DO NOT CITE OR QUOTE

-------
 1   for the benchmark concentration (BMC), with the BMCLio represented by the 95% lower
 2   confidence limit on the BMCio (U.S. EPA, 2000a).  Following the model selection steps outlined
 3   in the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000a), the best-fitting model was
 4   selected. Biological and statistical considerations were taken into account  in the selection of a
 5   BMR level for this dataset.  Statistically, a 10% level of response is intended to represent a
 6   response level near the lower range of detectable observations in typical studies conducted with
 7   50 animals per dose group (U.S. EPA. 2000a).
 8          Results of the BMDS modeling for chronic inflammation of the larynx in female rats are
 9   summarized in Table 5-6 and for epiglottis hyperplasia in Table 5-7.  As assessed by the ^
10   goodness-of-fit test, several models demonstrated adequate goodness of fit (p-value > 0.1) and
11   good visual fit (Appendix B).  In accordance with the draft Benchmark Dose Technical Guidance
12   (U.S. EPA. 2000a\ the LogLogistic model was selected as the best-fitting  model.  The BMCio
13   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
Model3
Log-Logistic
Log-Probit
Gamma
Multistage*5
(Stage 1)
Weibull
Probit
Logistic
Goodness of fit
/7-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
BMCLd
(mg/m3)
0.003
0.000
0.006
0.011
0.011
      Selected (best-fitting) model shown in first row, in boldface type.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                85       DRAFT - DO NOT CITE OR QUOTE

-------
            Table 5-7.  Benchmark modeling results for incidence of larynx respiratory
            epithelium, epiglottis, hyperplasia in female rats
Model3
LogLogistic
LogProbit
Gamma
Multistage6
(Stage 1)
Weibull
Probit
Logistic
Goodness of Fit
/7-Value
0.30
0.78
0.01
0.00
0.00
AICb
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/m3)
0.004
0.000
0.006
0.016
0.016
BMCLd
(mg/m3)
0.003
failed
0.005
0.013
0.014
     "Selected (best-fitting) model shown in first row, in boldface type.
     bAIC = Akaike Information Criterion.
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     Tor Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
     presented here.
     Source: NTP (2002).
 1    5.2.3.  RfC Derivation—Including Application of Uncertainty Factors
 2          BMD modeling of incidence data for chronic inflammation of the larynx in female rats
 3    yielded a BMCLio of 0.003 mg/m3.  The uncertainty factors, selected based on EPA's A Review
 4    of the Reference Dose and Reference Concentration Processes (U.S. EPA, 2002; Section 4.4.5),
 5    address five areas of uncertainty resulting in a composite UF of 300. This composite uncertainty
 6    factor was applied to the selected POD to derive an RfC.
 7          An interspecies uncertainty factor, UFA of 3 (101/2 = 3.16, rounded to 3) was applied to
 8    account for uncertainty in characterizing the toxicokinetic or toxicodynamic differences between
 9    rats and humans following inahalation exposure to vanadium pentoxide.  In this assessment,
10    toxicokinetic uncertainty was accounted for by the calculation of an HEC by applying a DAF as
11    outlined in the RfC methodology (U.S. EPA, 1994b). As the toxicokinetic differences are thus
12    accounted for, only the toxicodynamic uncertainties remain, and a UF of 3 is retained to account
13    for this residual uncertainty.
            An intraspecies uncertainty factor, UFn, of 10 was applied to account for potentially
      susceptible individuals in the absence of data evaluating variability of response to inhalation of
      vanadium pentoxide.  Considering the pulmonary effects of vanadium pentoxide, individuals
      with preexisting respiratory disorders might be more susceptible to inhaled vanadium pentoxide.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 86       DRAFT - DO NOT CITE OR QUOTE

-------
 1          A LOAEL-to-NOAEL uncertainty factor, UFL, of 1 was applied because the current
 2    approach is to address this factor as one of the considerations in selecting a BMR for BMD
 3    modeling (U.S. EPA. 2000a. 1994b).  In this case, a BMR of 10% increase in the incidence of
 4    chronic inflammation of the larynx and epithelial hyperplasia of the epiglottis was selected
 5    assuming it represents a minimal, biologically significant change.
 6          A subchronic-to-chronic uncertainty factor, UFs, of 1 was  applied to account for
 7    extrapolation from a subchronic to chronic exposure duration because the RfC was derived from
 8    a study of chronic duration.
 9          A database uncertainty factor, UFD; of 10 wasapplied to account for database deficiencies
10    due to the lack of a developmental toxicity study and a multigeneration reproductive study for
11    vanadium pentoxide by the inhalation route.  Studies using alternative routes of exposure
12    (intraperitoneal) have indicated adverse reproductive and developmental effects in response to
13    vanadium pentoxide, including statistically significant increases in seminal vesicle, thymus, and
14    submandibular gland weights in male mice and body weight, thymus, submandibular gland, and
15    liver weights in female mice (Altamirano et al., 1991) as well as reduced fertility (Altamirano-
16    Lozano et al.,  1996). No pharmacokinetic models are available for conducting a route-to-route
17    extrapolation.
18          The chronic RfC for vanadium pentoxide is calculated as follows:
19
20                                     RfC = BMCLio-UF
21                                          = 0.003 mg/m3 - 300
22                                          = 0.00001 mg/m3 or 1  x  10~5 mg/m3
23
24    Note:  Because vanadium exists in several valence states, all of which are not equivalent
25    lexicologically (WHO, 2001), the values generated here apply to vanadium pentoxide and should
26    not be applied to other vanadium compounds.

27    5.2.4.  Previous RfC Assessment
28          An inhalation assessment for vanadium pentoxide was not previously available in IRIS.

29    5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE  (RfD) AND THE
30        INHALATION REFERENCE CONCENTRATION (RfC)
31          The following discussion identifies uncertainties associated with the quantification of the
32    RfD and RfC for vanadium pentoxide that are not accounted for by the application of the
33    uncertainty factors. As presented earlier in this section, EPA standard practices and RfC
34    guidance (U.S. EPA. 2002. 1995. 1994a. b) were followed in applying a UF approach to a POD
35    for both the RfD and RfC (see Sections 5.1.3  and 5.2.3). Factors  accounting for uncertainties
36    associated with several steps in the analyses were adopted to account for extrapolation from an
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                87        DRAFT - DO NOT CITE OR QUOTE

-------
 1    animal study to human exposure and to a diverse human population of varying susceptibilities,
 2    for extrapolation from subchronic to chronic exposure duration, and for database deficiencies.
 3          The RfD was derived based on the critical effect (decreased red blood cells) in rats
 4    exposed to vanadium pentoxide in the diet in a subchronic study (Mountain et al., 1953).
 5    Although limited human data are available, decreased red blood cells in this study following
 6    exposure to vanadium pentoxide might suggest anemia as a potential  effect of exposure. A
 7    NOAEL-to-LOAEL approach was used for derivation of the chronic RfD, as the data were not
 8    amenable for Benchmark Dose Modeling. There is no physiologically based toxicokinetic
 9    modeling information; however, based on EPA''s Recommended Use  of Body Weight3'4 as the
10    Default Method in Derivation of the Oral Reference Dose (U. S. EPA, 2011), the POD identified
11    in rats (i.e., decreased red blood cells) was converted to an HED by applying a dosimetric
12    adjustment factor. Statements in this guidance document raise some important uncertainties in
13    applying allometric scaling, and more specifically BW3 4 scaling, when trying to extrapolate
14    across different-sized individuals within a species (e.g., between neonates and adults) or across
15    individuals in different lifestages between species (e.g., between fetal rats and adult humans).
16    Furthermore, the data on which to base a default allometric scaling factor for converting the
17    administered dose in a laboratory animal in a different lifestage to a comparable dose in an adult
18    human are sparse, and thus more uncertain.  However, given that the Mountain et al.  (1953)
19    study was performed on adult male rats and decreased red blood cells are not a portal-of-entry
20    effect, the uncertainties in applying this allometric scaling are minimal.
21          An adequate range of animal toxicology data is available for the inhalation hazard
22    assessment of vanadium pentoxide, as described in  Section 4. Included in these studies are
23    short-term, subchronic, and chronic bioassays in rats, as well as a range of supporting
24    genotocixity  studies.  Toxicity associated with inhalation exposure to vanadium pentoxide is
25    observed in reproductive organs, the CNS, and particularly in the respiratory system, including a
26    range of nasal and pulmonary nonneoplastic lesions, such as pulmonary inflammation,  tissue
27    morphology changes, and  development of pulmonary fibrosis.  Recent mechanistic studies have
28    investigated vanadium pentoxide-induced pulmonary fibrosis and have contributed to some
29    understanding of a putative mode of action (MOA)  for pulmonary fibrosis in response to
30    vanadium pentoxide.
31          For derivation of the chronic RfC for vanadium pentoxide, a 2-year inhalation study in
32    rats and mice (NTP, 2002) was selected as the principal study and inflammation of the  larynx
33    was selected  as the critical effect. Lung hyperplasia, nasal inflammation, and lung fibrosis were
34    also sensitive effects. Inflammation of the larynx was selected as the critical effect because it is
35    a sensitive indicator of vanadium pentoxide-induced respiratory toxicity and yielded  the lowest
36    POD.
37          The selection of the BMD model for quantifying the RfC does not lead to significant
38    uncertainty in estimating the POD because benchmark effect levels were within the range of
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                88        DRAFT - DO NOT CITE  OR QUOTE

-------
 1    experimental data for chronic inflammation of the larynx. Although the selected log-logistic
 2    model is the best fitting model, it is not the only model that adequately describes the data.  Other
 3    models could be selected to yield more extreme results, both higher and lower than those
 4    included in this assessment.
 5          Extrapolating from animals to humans entails further issues and uncertainties as the
 6    magnitude of the effect and the effect itself associated with the concentration at the POD in
 7    rodents are extrapolated to human response. Pharmacokinetic models are useful in examining
 8    species differences in pharmacokinetics; a PBPK model for vanadium pentoxide, however, was
 9    not available.  Therefore, toxicokinetic species  differences were addressed by determining an
10    HEC through inhalation dosimetry adjustments as described in the RfC methodology (U.S. EPA,
11    1994b).

12    5.4.  CANCER ASSESSMENT

13    5.4.1. Choice of Study/Data—with Rationale and Justification
14          The 2-year NTP (2002) inhalation cancer bioassay reported an increased incidence of
15    alveolar/bronchiolar adenomas or carcinomas in male F344/N rats.  Evidence for carcinogenic
16    activity of vanadium pentoxide in female F344/N rats at the high dose was equivocal. Male and
17    female B6C3Fi mice had greater incidences of these lesions, with a statistically significantly
18    increased incidence of alveolar/bronchi olar adenomas or carcinomas in both male and female
19    B6C3Fi mice following inhalation exposure to >1 mg/m3 of vanadium pentoxide (NTP,  2002)
20    (Table 5-8). These tumor types are considered  relevant to humans. Human data on the
21    carcinogenic potential of inhalation exposure to vanadium pentoxide are not available. No
22    human or laboratory animal data to determine the carcinogenicity of vanadium pentoxide by the
23    oral or dermal route are available.

24    5.4.2. Dose-Response Data
25          Data on the incidences of alveolar/bronchiolar adenomas or carcinomas in male and
26    female mice from the NTP (2002) study were used for cancer dose-response assessment. These
27    data are shown in Table 5-9.

28    5.4.3. Dose Adjustments and Extrapolation Methods
29          The NTP (2002) 2-year carcinogenicity study in mice was used to derive an inhalation
30    unit risk  (IUR), based on the dose-response relationship for alveolar/bronchiolar neoplasms
31    (adenoma and carcinoma).
32          Using the RDDR computer program, as specified in the RfC guidelines (U.S. EPA,
33    1994b), HECs (in mg/m3) were calculated at each exposure level for male and female mice using
34    the mean body weights for males and females and the average particle size MMAD ± GSD of
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               89        DRAFT - DO NOT CITE OR QUOTE

-------
        Table 5-8.  Incidences of respiratory tumors in B6C3Fi mice exposed to
        vanadium pentoxide in the 2-year inhalation study
Tumor Type"
Exposure Group
Historical Control
(% Historical
Control)
Control
1 mg/m3
2 mg/m3
4 mg/m3
Male mice
Number of animals examined
Alveolar/bronchiolar adenomab
Alveolar/bronchiolar carcinoma
Alveolar/bronchiolar adenoma or
carcinoma
1,071
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
Alveolar/bronchiolar adenoma
Alveolar/bronchiolar carcinoma
Alveolar/bronchiolar adenoma or
carcinoma
1,075
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 = 1.2 ± 2.9.
 bHistorical incidence of alveolar/bronchiolar adenoma male B6C3FJ mice fed in inhalation chamber controls given
 NIH-07 diet.
 Significantly different from control by the Poly-3 test (p < 0.01)
 Source: NTP (2002).
       Table 5-9. Human equivalent concentrations of vanadium pentoxide in the 2-
       year inhalation studies
Concentration as reported
(mg/m3)
Continuous exposure
adjustment factor"
RDDRb
Pulmonary
HECC (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 (B6C3FO
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
""Continuous exposure adjustment factor = (6/24) x (5/7); animals were exposed to vanadium pentoxide 6 hr/d,
5d/wk.
bSee Appendix Table C-4.
°HEC = human equivalent concentration = concentration as reported x continuous exposure adjustment
factor x RDDR.
Source: NTP (2002).
            This document is a draft for review purposes only and does not constitute Agency policy.
                                               90        DRAFT - DO NOT CITE OR QUOTE

-------
 1    1.26 ± 1.87 as reported by NTP (2002). HECs were calculated by multiplying ConC[ADJ] by the
 2    RDDR for male and female mice (Table 5-9).
 3           According to the U. S. EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA,
 4    2005a), for each tumor response, a POD from the observed data should be estimated to mark the
 5    beginning of extrapolation to lower doses.  The POD is an estimated dose near the
 6    lower end of the observed range without significant extrapolation to lower doses.u All
 7    noncontrol concentrations from the NTP carcinogenesis studies (2002) showed a plateau
 8    response, and therefore the dataset provided limited information about the concentration-
 9    response relationship because the complete range of response from background to maximum
10    must occur somewhere below the lowest dose.  A BMR based on the response at the control
11    concentration and the first noncontrol concentration was therefore calculated (Table C-6) and
12    used to estimate a POD (BMCL, extra risk of 0.71 for male mice, 0.67 for female mice).
13    Modeling was performed using BMDS (Version 2.1.2) developed by the National Center for
14    Environmental Assessment (U.S. EPA, 2000a).
15           The incidences of alveolar/bronchiolar adenomas and carcinomas in mice were combined
16    the incidence of tumor-bearing animals; males and females were modeled separately (Table
17    5-10). Models were run using the default restrictions on parameters built into the BMD
                                                    9                           	
18    software. Goodness-of-fit was evaluated using the %  statistic calculated by the BMDS program.
19    Each dataset was first fitted with the dichotomous multistage cancer model; if the goodness-of-fit
20    /7-value was < 0.05, other dichotomous models were fitted; if still no model showed adequate
21    goodness of fit (p-value > 0.05), the highest dose was dropped for further modeling.
22    The BMDS modeling results for models meeting goodness-of-fit criteria are summarized in
23    Appendix C. For incidence data in male mice, the log-logistic model was the only model that
24    met goodness-of-fit criteria when all three dose groups were included, predicting BMC?i and
25    BMCL7i values of 0.360 and 0.208 mg/m3, respectively (Appendix C). The multistage model fit
26    the incidence data for male mice when the high dose was dropped, predicting BMC?i and
27    BMCL7i values of 0.306 and 0.220 mg/m3, respectively.
28          None of the dichotomous models fit tumor incidence data for female mice when all three
29    dose groups were included. One model (log-logistic) fit the incidence data for female mice when
30    the high dose was dropped, predicting BMCe? and BMCLe? values of 0.237 and 0.161 mg/m3,
31    respectively.
32           The BMCLy/i of 0.208 mg/m3 for male mice was selected as the POD for derivation of the
33    IUR as as this was the only model fit when including all doses for analysis of lung tumor
34    formation following inhalation exposure to vanadium pentoxide.
      11 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. If the POD is far below all observed data points,
      however, model and parameter uncertainty that increase with the distance between the data and the POD can be
      introduced Using a POD at the lowest level supported by the data tends to balance these considerations.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 91        DRAFT - DO NOT CITE OR QUOTE

-------
            Table 5-10. BMDS modeling results for combined incidence of
            alveolar/bronchiolar adenoma and carcinoma in male mice
Model3
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
0.01
207.0
2.05
0.72
Extra Risk
71%
0.532
0.379
Other dichotomous models
Log-Logistic
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
Extra Risk
71%
0.360
0.146
0.532
0.609
0.654
0.208
failed
0.379
0.447
0.495
      "Selected model is the best-fitting model for the dataset based on the draft Benchmark Dose Technical Guidance
      (U.S. EPA. 2000a).
      bHEC = human equivalent concentration.
      CBMCL = the lower bound of BMC at 95% confidence level.
      Source: NTP (2002).

 1    5.4.4.  Inhalation Unit Risk
 2          The U. S. EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a)
 3    recommend that the method used to characterize and quantify cancer risk from a chemical is
 4    determined by what is known about the mode of action of the carcinogen and the shape of the
 5    cancer dose-response curve.  The linear approach is recommended if the mode of action of
 6    carcinogenicity is not understood (U.S. EPA, 2005a). In the case of vanadium pentoxide, data to
 7    support an  MOA for the carcinogenicity are insufficient, although some data suggest that either a
 8    mutagenic  or a cytotoxic and reparative proliferation MOA is operative. In the absence of such
 9    data, a linear approach was used to extrapolate from the POD to lower doses.
10          The IUR represents an upper-bound, continuous lifetime exposure risk estimate and  is
11    calculated as BMR/BMCL [0.717(0.208 mg/m3)].  The HEC BMCL7i for extra risk of
12    alveolar/bronchiolar adenomas or carcinomas in male B6C3Fi mice exposed to vanadium
13    pentoxide results in  an IUR of 3.4 (mg/m3)"1.  This value was derived by linear extrapolation to
14    the origin from the POD of 0.208 mg/m3 and represents an upper-bound estimate.

15    5.4.5.  Oral Cancer Slope Factor
16          No  human data or animal studies relevant to the  carcinogenicity of vanadium pentoxide
17    following oral exposure were located in the published literature. Therefore,  an oral cancer slope
18    factor is not derived.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                92       DRAFT - DO NOT CITE OR QUOTE

-------
 1    5.4.6.  Uncertainties in Cancer Risk Values
 2          Extrapolation of study data to estimate potential risks to human populations from
 3    exposure to vanadium pentoxide introduces some uncertainty in the results. Several types of
 4    uncertainty can be considered quantitatively, but other important uncertainties cannot be.
 5    Section 5.4.5.1 and Table 5-11 summarize principal uncertainties.
 6          Carcinogenicity due to chronic exposure of vanadium pentoxide was observed in two
 7    species (NTP, 2002), with the carcinogenicity more definitive in mice than in rats, particularly
 8    female rats. For female rats, the increased incidence of neoplasms was lower than what would
 9    be expected due to spontaneous tumor formation. Similarly, spontaneous tumors were observed
10    in male rats at control levels, although tumor incidence increased in male rats in response to
11    exposure to vanadium pentoxide.  The confidence in the database is low.  In the well-conducted
12    NTP study (2002), respiratory tract carcinogenicity was reported in two rodent species (i.e., mice
13    and rats), which was supported by a study in mice (Rondini et al., 2010).  Human variability in
14    response to vanadium pentoxide, on the other hand, is unknown, and humans occupationally
15    exposed  to vanadium pentoxide are often simultaneously exposed to other valence states of
16    vanadium and to other inhaled environmental toxicants simultaneously.  Genotoxicity is
17    equivocal and no further MOA information is available.  Overall confidence in the IUR is low.
18          Choice of low-dose extrapolation approach. The MOA is a key consideration in
19    clarifying how risks should be estimated for low-dose exposure. A linear low-dose extrapolation
20    approach was used to estimate human carcinogenic risk associated with vanadium pentoxide
21    exposure due to the unavailability of data to support any specific mode of carcinogenic action of
22    vanadium pentoxide.
23          Dose metric.  Vanadium exists in the +5 valence state in vanadium pentoxide. Other
24    valence states ranging from -1 to +5 exist.  Frequently, vanadium exposures involve a mixture of
25    vanadium compounds ranging primarily from +3 to +5 valence states. The carcinogenic
26    potential of other valence states of vanadium has not been established.  Whether vanadium
27    pentoxide dissociates to other valence states with known carcinogenic potential is unknown.
28    Similarly, whether some other valence state or some combination of+5 and other valence states
29    is responsible for the observed toxicity is not known.  If the actual  carcinogenic moiety is
30    proportional to administered exposure, then use of administered exposure as the dose metric is
31    the preferred choice.
32          Statistical uncertainty at the POD. Parameter uncertainty can be assessed by examining
33    confidence intervals. Each description of parameter uncertainty assumes that the underlying
34    model and associated assumptions are valid. For the log-logistic cancer model applied to the
35    male mice data, the degree of uncertainty at a 71% increase in tumor incidence (the POD for
36    linear low-dose extrapolation) is reasonably small.
37          Bioassay selection. The NTP (2002) study was selected for development of an IUR.
38    This study was well designed and conducted on both sexes in two species with an adequate
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                93       DRAFT  - DO NOT CITE OR QUOTE

-------
 1    number of animals per dose group.  The number of test animals allocated among the two dose
 2    levels and the untreated control group was adequate, with examination of appropriate
 3    toxicological endpoints in both sexes of rats and mice. Both genders of mice exhibited a
 4    statistically significant increased incidence of lung tumors.
 5          Choice of species/gender. The IUR for vanadium pentoxide was quantified using the
 6    tumor incidence data for male mice, which were more amenable to BMD modeling than the
 7    tumor incidence in female mice. In addition, lung neoplasm incidence was reported in male and
 8    female rats, although neither gender of rat was as sensitive as mice.  A 71% tumor incidence
 9    level was observed in female mice at the lowest exposure level (1  mg/m3), suggesting that lower
10    doses might have revealed more information about the low-dose region of the dose-response
11    curve. Male mice demonstrated a high background rate of lung tumors, with spontaneous lung
12    neoplasms observed in male mice at control levels (up to 28% of male mice). Tumor incidence
13    increased significantly (84%) at the lowest dose level tested (1 mg/m3). Although these
14    incidence response rates were higher in male mice than in females at the comparable exposure
15    level, suggesting greater sensitivity of the male mice, there is no information for the dose-
16    response relationships at lower exposure levels. In other words, the behavior of vanadium
17    pentoxide at 1.0 mg/m3 in male mice might not inform the tumor response to vanadium
18    pentoxide at lower exposures.
19          Relevance to humans. In the absence of direct human data, the most appropriate animal
20    bioassays to use in the derivation of cancer risk values are chronic (i.e., lifetime) studies in two
21    species of rodents. The IUR was derived from the combined tumor incidences of lung adenomas
22    and carcinomas in male mice. Information investigating the MOA of the lung tumors observed
23    in the chronic animal bioassay, however, is limited.  Genotoxicity studies provide inadequate
24    evidence of a genotoxic MOA, and data to support alternative MOA hypotheses are inadequate.
25          Human population variability. Interindividual variability in animals for vanadium
26    pentoxide metabolism has not been characterized. Strain differences in the response to vanadium
27    pentoxide-induced pulmonary fibrosis in rodents  suggests a genetic component to susceptibility.
28    Moreover, humans occupationally exposed to vanadium pentoxide are often simultaneously
29    exposed to other valence states of vanadium and to other inhaled environmental toxicants
30    simultaneously.  This lack of understanding about potential differences in metabolism and
31    susceptibility across exposed human populations  thus represents a source of uncertainty.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                94       DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-11. Summary of uncertainty in the vanadium pentoxide cancer risk
assessment
Consideration/
Approach
Low-dose
extrapolation
procedure
Dose metric
Cross-species
scaling
Statistical
uncertainty at POD
Bioassay
Species/gender
combination
Human relevance of
mouse tumor data
Human population
variability in
metabolism and
response/sensitive
subpopulations
Impact on inhalation
unit risk
Departure from EPA's
Guidelines for
Carcinogen Risk
Assessment (2005a)
POD paradigm, if
justified, could | or |
slope factor by an
unknown extent
Alternatives could t
or 4 slope factor by an
unknown extent
Alternative methods
could | or t inhalation
unit risk [e.g., 3. 5 -fold
4 (scaling by B W) or
t 2-fold (scaling by
BW2/3)]
4 slope factor by 1.7-
fold if BMC used
rather than the BMCL
Alternatives could t
or J, slope factor by an
unknown extent
Human risk could 4 or
t, depending on
relative sensitivity
Human relevance of
mouse tumor data
could 4 slope factor
Low-dose risk f to an
unknown extent
Decision
Log-logistic cancer
model to determine
POD, linear
low-dose
extrapolation from
POD
Used administered
exposure
RDDR
BMCL (method for
calculating
reasonable
upper-bound slope
factor)
NTP study
Male mice lung
cancer
Lung tumors in
mice are relevant
to human exposure
Considered
qualitatively
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 exist to support
carcinogenicity due to other forms of vanadium.
RDDR software was used to adjust for
toxicokinetic differences in inhalation dosimetry.
The lower bound is the 95% confidence interval
on administered exposure.
Alternative bioassays were unavailable.
The lung cancer response was concordant in male
and female mice. There are no MOA data to
inform the extrapolation approach for any choice.
Humans were assumed to be as sensitive as the
most sensitive rodent gender/species tested; true
correspondence is unknown. The carcinogenic
response occurs across genders and species.
Vanadium pentoxide could be carcinogenic
through an unknown MOA.
No data are available to support the range of
human variability/sensitivity.
    This document is a draft for review purposes only and does not constitute Agency policy.
                                    95       DRAFT - DO NOT CITE OR QUOTE

-------
 1                 6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION
 2                            OF HAZARD AND DOSE RESPONSE

 3    6.1.  HUMAN HAZARD POTENTIAL
 4          Vanadium is a metal commonly found in ores, tars, coals, and oils and is used as an alloy
 5    in steel.  The toxicity of vanadium depends on its valence state, which can range from -1 to +5,
 6    depending on pH and other factors. Vanadium pentoxide (V^Os), sodium metavanadate
 7    (NaVOs), and ammonium metavanadate (NH/iVCb) all contain vanadium in the +5 oxidation
 8    state. This Toxicological Review focuses exclusively on vanadium pentoxide (V^Os), the most
 9    common form of vanadium used commercially. In addition, V2Os is the only compound that is
10    covalently bonded. Occupational exposure to vanadium pentoxide occurs primarily via
11    inhalation of dust generated during vanadium processing and fuel-oil ash generation during
12    cleaning of oil-burning boilers and furnaces.
13          The toxicokinetics of orally administered vanadium pentoxide are not available in
14    humans. Kucera et al. (1992) attempted to measure vanadium in the hair of children who had
15    been exposed to vanadium in drinking water, but exposure could not be linked to vanadium
16    pentoxide specifically. Oral toxicokinetic studies in animals are limited. One study investigated
17    the toxicokinetics of components of various metals, including vanadium, in female B6C3Fi mice
18    (Radike et al., 2002).  In this study, vanadium was detected in the small intestine, kidney, and
19    femur.
20          Similarly, a few studies have evaluated the toxicokinetics of inhaled vanadium,  but there
21    are no data for humans.  Occupational studies of inhaled vanadium pentoxide indicate that
22    vanadium is absorbed by humans, and levels in blood and urine rapidly declined following
23    exposure (Kiviluoto et al.,  1981b). Results of toxicokinetic studies of inhaled or intratracheally
24    administered vanadium pentoxide in rats show that vanadium pentoxide is absorbed from the
25    lung, undergoes a wide distribution to the liver, kidney, bone, blood, gastrointestinal (GI) tract,
26    and ovary, and is eliminated primarily in the urine (Dill et al., 2004; NTP, 2002; Roshchin et al.,
27    1980).
28          The animal oral toxicity database reported decreased hair cystine (Mountain et al., 1953;
29    Stokinger et al., 1953): hematological effects, including a decrease in red blood cell (RBC) count
30    and hemoglobin (Mountain et al., 1953): and a variety of reproductive and developmental effects
31    following exposure to vanadium pentoxide  (Altamirano-Lozano et al., 1993)(Zhang et al.,  1993a;
32    1993b).
33          Studies of human exposure to inhaled vanadium pentoxide include primarily occupational
34    reports and one controlled human exposure study. Several case study summaries reported upper
35    and lower respiratory tract irritation and inflammation among workers with inhalation exposure
36    to vanadium pentoxide and other vanadium compounds in dust during vanadium processing or to
37    fuel-oil ash during cleaning and maintenence of oil-burning boilers (Section 4.1.2).  The
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                96       DRAFT - DO NOT CITE OR QUOTE

-------
 1    chemical composition of the fuel-oil ash and vanadium dust or exposure measurements for
 2    vanadium pentoxide, however, generally were not reported. Thus, in many cases, specific
 3    relationships between vanadium pentoxide and adverse respiratory effects cannot be definitively
 4    determined.  Sjoberg (1955) published seven case reports of vanadium-induced bronchitis in
 5    workers cleaning boilers where the concentrations of vanadium pentoxide particles (10-20 um in
 6    diameter) in the air were 2-85 mg/m3.  Self-reported respiratory symptoms included cough,
 7    rhinitis, wheeze, sore throat, and conjunctivitis.  All symptoms resolved by 2 weeks
 8    postexposure, but reappeared upon re-exposure.  Of 100 workers occupationally exposed to
 9    vanadium pentoxide fumes (0.05-5.3 mg/m3) via an oil-to-coal  power-plant conversion, 74
10    reported severe respiratory tract irritation (Levy et al.,  1984). Estimated daily nasal and lung
11    doses of vanadium were associated with incidence and severity of upper airway symptoms (nasal
12    congestion/irritation, throat irritation) and lower airway symptoms (chest tightness, wheeze,
13    cough, and sputum production) in a dose-related  manner in a prospective clinical study.  This
14    clinical study involved boilermakers and utility workers overhauling a large, oil-fired boiler over
15    a 6-week period (Woodin et al.. 2000: Woodin et al.. 1999: Woodin et al.. 1998). Geometric
16    mean concentrations of vanadium (9-10 hour shifts) measured in the breathing zone ranged from
17    1.1 to 8.9 |ig/m3.  Reductions in pulmonary function were measured among boilermakers with
18    exposure to fly ash for weeks or years but an association with vanadium has not been established
19    (Hauser et al., 2001; Woodin et al., 1999; Hauser etal., 1995b). No cases of cancer were
20    reported in workers examined over a sufficient latency period as a result of exposure to
21    vanadium pentoxide.
22          Respiratory effects in experimental animals are well documented and include lesions in
23    the nasal compartment, larynx, and lung. In addition to respiratory effects observed in animals
24    (Yu etal.. 2011: Turpin  etal.. 2010: NTP. 2002:  Knechtetal.. 1992: Knechtetal.. 1985). some
25    animal studies document spermatogenesis changes and testicular ultrastructural changes
26    (Mussali-Galante et al., 2005) and central nervous system (CNS) effects (Colin-Barenque et  al.,
27    2007: Avila-Costa et al.. 2006: Avila-Costa et al.. 2005: Avila-Costa et al.. 2004) in response to
28    inhaled vanadium pentoxide. Short-term inhalation exposure (16  days) in F344/N rats and
29    B6C3Fi mice was associated with histiocytic infiltration at 1.0 mg/m3, and short-term and
30    subchronic inhalation exposures (3 months) in F344/N rats and  B6C3Fi mice were associated
31    with increased lung epithelial hyperplasia and inflammation at 2.0 mg/m3 (NTP, 2002).
32    Long-term exposure (2 years) in F344/n rats and  B6C3Fi mice similarly resulted in increased
33    incidence of respiratory  toxicity and carcinogenicity (NTP, 2002). Lung inflammation was
34    observed at 1.0 mg/m3 and lung hyperplasia was  observed at 2.0 mg/m3 in female rats. Lung
35    inflammation and hyperplasia were observed at 1.0 mg/m3 in both male and female mice.  In
36    male and female mice, nonneoplastic lesions also occurred in the larynx (squamous metaplasia of
37    epiglottis epithelium) and nasal tissues (degeneration and atrophy of olfactory epithelium and
38    degeneration of respiratory epithelium) at 1 mg/m3; it should be noted that 1 mg/m3 was the
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                97       DRAFT - DO NOT CITE OR QUOTE

-------
 1    lowest dose in mice. In mice exposed to 1 mg/m3, lesions were observed in the nose,
 2    bronchioles, and lung of male mice and the nose, larynx, bronchioles, and lung of female mice.
 3    Chronic, active inflammation and interstitial fibrosis was observed in male rats at a
 4    lowest-observed-adverse-effect level (LOAEL) of 1 mg/m3 (no-observed-adverse-effect level
 5    [NOAEL] was 0.5 mg/m3). The nasal and laryngeal lesions appear to be among the most
 6    sensitive effects observed at the lowest dose tested.  Moreover, multiple lesion types were
 7    observed in both sexes of rats and, thus, the LOAEL of 0.5 mg/m3 for nasal and laryngeal lesions
 8    was selected as the critical effect.
 9          Reproductive and developmental studies reveal various altered endpoints in response to
10    vanadium pentoxide via oral and inhalation routes. Vanadium pentoxide delivered orally to
11    weanling rats for 3 days caused a significant increase in alkaline phosphatase activity and DNA
12    content in the diaphysis of femoral bones, suggesting that vanadium pentoxide might be linked to
13    bone formation in the developing rat (Yamaguchi et al., 1989). Mussali-Galante et al., (2005)
14    identified accumulation of vanadium pentoxide in the testes after 1 week of inhalation exposure
15    in male CD-I mice (0.02 M or 1.4 mg V/m3).  Gamma tubulin was significantly decreased in
16    testes exposed to vanadium pentoxide and might suggest  changes in microtubule function that
17    could impact spermatogenesis. In a follow-up study, Fortoul et al. (2007) reported necrotic cell
18    death in spermatogonia and increased nuclear distortion in spermatocytes in male CD-I mice
19    exposed by inhalation to vanadium pentoxide (0.02 M or 1.4 mg V/m3).
20          Several intraperitoneal (i.p.) studies have evaluated reproductive and developmental
21    endpoints. Common reproductive effects include increased reproductive organ weights in male
22    rats (Altamirano et al., 1991), increased incidence of apoptotic spermatogonia in male mice
23    (Aragon et al., 2005) and reduced sperm motility, reduced sperm count, and increased numbers
24    of abnormal sperm in male CD-I mice (Altamirano-Lozano et al.,  1996). In developmental
25    studies, reduced ossification in the developing forelimbs and hindlimbs of fetuses have been
26    reported in response to i.p. exposure to vanadium pentoxide during gestation days 6-15 in
27    pregnant CD-I  mice (Altamirano-Lozano et al., 1993) and Wistar rats (Zhang etal., 1993a:
28    1993b). Other reported effects range from reduced fetal weight and increased placental weight.
29          The genotoxicity database for vanadium pentoxide is limited.  The evidence for
30    genotoxicity following exposure to vanadium pentoxide in humans is limited. A few studies
31    have examined genotoxicity in humans in vivo, with equivocal results. Ivancsits et al. (2002)
32    reported no differences in DNA strand breaks, oxidative damage, or SCE frequency in
33    leukocytes between control and vanadium pentoxide-exposed workers. Ehrlich et al., (2008)
34    noted changes in DNA stability and DNA repair in leukocytes of occupationally exposed
35    workers as compared to controls. Studies have demonstrated a genotoxic effect of vanadium
36    pentoxide on human cells in vitro. Ivancsits et al. (2002) demonstrated significant increases in
37    DNA damage as measured by the Comet assay in both leukocytes and fibroblasts but with
38    different dose sensitivity, while Kleinsasser et al (2003) noted DNA migration differences
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                98        DRAFT - DO NOT CITE OR QUOTE

-------
 1    occurred dose dependently in peripheral blood lymphocytes but not in nasal mucosa cells.
 2    Earlier studies in human lymphocyte cultures also demonstrated increased aneuploidy (Ramirez
 3    etal., 1997; Rojas etal., 1996) and DNA damage (Roldan and Altamirano, 1990) following
 4    exposure to vanadium pentoxide. Thus, vanadium pentoxide-induced mutagenicity may occur at
 5    doses higher than those measured in these occupational exposures, may be tissue-specific, and
 6    may be associated with oxidative stress rather than direct DNA damage. Experimental data in
 7    animals provide evidence of some types of genotoxicity following in vivo exposure to vanadium
 8    pentoxide. Vanadium pentoxide administered by inhalation to mice or rats did not increase the
 9    frequency of micronucleated normochromatic erythrocytes in peripheral blood (NTP, 2002).
10    Genotoxicity assessed in male CD-I mice following i.p. injection caused no treatment-related
11    effects in mitotic index, average generational time, or SCE (Altamirano-Lozano et  al., 1993).
12    DNA damage was detected, however, in six organs from vanadium pentoxide-treated mice via
13    i.p. injection (Altamirano-Lozano et al., 1999). Vanadium pentoxide produced gene mutations in
14    two bacterial test systems (Kadaetal., 1980; Kanematsu et al., 1980) but negative results in the
15    NTP (2002) study. Vanadium pentoxide produced DNA strand breaks, aneuploidy, and
16    micronucleus induction but did not produce chromosomal aberrations or SCE in various  cell
17    lines (Kleinsasser et al.. 2003: Ivancsits et al.. 2002: Ramirez et al.. 1997: Roiasetal.. 1996:
18    Zhong et al.. 1994: Roldan and Altamirano. 1990).
19          Most effects of vanadium pentoxide are thought to be produced by the parent compound,
20    primarily by inducing cell damage and pulmonary fibrosis (NTP, 2002: Bonner et al., 2000:
21    Bonner etal., 1998).  Pulmonary fibrosis is mediated either directly by vanadium
22    pentoxide-induced changes to cell signaling molecules or via vanadium pentoxide-induced
23    oxidative stress that induces cellular changes leading to fibrosis (Ingram et al., 2003: Rice et al.,
24    1999: Bonner etal., 1998). Pulmonary fibrosis could be a key event leading to eventual
25    tumorigenesis.
26          Vanadium pentoxide is "likely to be carcinogenic to humans" by the inhalation route of
27    exposure under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). Existing
28    cancer data come from two animal species (mice, rats) reported in one well-characterized study
29    (NTP, 2002).  Increased incidence of pulmonary adenomas and  carcinomas were observed  in
30    male and female mice exposed to vanadium pentoxide for 2 years. Carcinogenicity in male rats
31    exposed to vanadium pentoxide for 2 years is evident, as respiratory tumor incidence exceeds
32    that in historical controls. In female rats, the incidence of respiratory tumorigenesis in response
33    to vanadium pentoxide does not exceed historical controls. Tumor incidence has not been
34    documented in humans.
35          Information available on the carcinogenic effects of vanadium pentoxide via the
36    inhalation route is limited to examination of respiratory tumors. Information on the carcinogenic
37    effects of vanadium pentoxide via the oral and dermal routes in  humans or animals is absent.
38    Based on the observation of only respiratory tumors following inhalation exposure, and in the
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                99       DRAFT - DO NOT CITE OR QUOTE

-------
 1    absence of information to establish a mode of action (MOA), this cancer descriptor applies only
 2    to the inhalation route of exposure. Therefore, the database has "inadequate information to assess
 3    carcinogenic potential" of vanadium pentoxidevia the oral or dermal route.
 4           The MOA underlying tumorigenicity in rats and mice has not been established.  The
 5    genotoxicity database for vanadium pentoxide is equivocal, including studies that report both
 6    positive and negative results for mutation, DNA damage, and chromosomal aberrations. A
 7    nongenotoxic MOA hypothesis involving hyperplasia and development of fibrotic pulmonary
 8    lesions is supported by the presence of hyperplastic lesions and pulmonary fibrosis at earlier time
 9    points and at lower doses. The dose-response relationship is not robust, however, and a clear
10    relationship linking these effects to the tumor response has not been established. Whether
11    cytotoxicity is a required precursor event for vanadium pentoxide-induced cell proliferation is
12    unknown.  Sufficient data regarding a plausible dose response and temporal progression from
13    cytotoxicity to hyperplasia to fibrosis to tumorigenesis are not available.

14    6.2.  DOSE RESPONSE

15    6.2.1.  Noncancer
16           Limited studies are available examining the toxicity of vanadium pentoxide following
17    oral exposure in humans or laboratory animals. Decreased RBC count and hemoglobin were
18    observed following subchronic oral exposure to vanadium pentoxide in rats (Mountain et al.,
19    1953).  The authors reported the decrease of RBC as a mean with no measure of variability
20    among animals and the continuous data therefore were not amenable to benchmark dose (BMD)
21    modeling.  Derivation of an oral reference dose (RfD) was based on a NOAEL of 10.5 mg/kg-d
22    for the critical effect, decreased RBC counts, and divided by a total uncertainty factor (UF) of
23    3,000:  3 to represent interspecies toxicodynamic uncertainties, 10 for interhuman variability in
24    the absence of quantitative information on the variability of response in humans, 10 for
25    extrapolation from a subchronic to a chronic study, and 10 for database deficiencies. The result
26    was the chronic RfD of 9 x 10~4 mg/kg-day.
27           Confidence in the principal study for derivation of the RfD (Mountain et al., 1953) is low.
28    Mountain et al. (1953) is a well-conducted  study with numerous doses, but is a subchronic study
29    with a small sample size (n = 5) for one species of rodents (Wistar rat), using one gender (male),
30    and limited endpoints and time points.  Confidence in the overall database, however, is  low.
31    Mountain et al. (1953) is the single, relevant, peer-reviewed study for derivation of the chronic
32    RfD. Thus, overall confidence in the chronic RfD is low.
33           Pulmonary effects have been documented in numerous species, including humans and
34    primates, in response to inhaled vanadium pentoxide.   Inflammation and histiocytic infiltrate are
35    common hallmarks of initial vanadium-induced pulmonary injury.  The database also includes
36    occupational (inhalation) and laboratory animal (inhalation and oral) studies demonstrating
37    possible immunotoxicity following exposure to vanadium pentoxide.  In addition, the database
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                100       DRAFT - DO NOT CITE OR QUOTE

-------
 1    includes studies of neurotoxicity and reproductive and developmental toxicity studies following
 2    inhalation exposure to vanadium pentoxide in rodents. Overall, the lung is the most sensitive
 3    target for noncancer toxicity in rats and mice following chronic inhalation exposure to vanadium
 4    pentoxide. Nonneoplastic lung lesions (specifically laryngeal lesions) were selected as the most
 5    sensitive endpoint from a well-conducted chronic study (NTP, 2002) (2-year rodent bioassay).
 6    The chronic inflammation of the larynx and epithelial hyperplasia of the epiglottis in rats were
 7    chosen as critical effects because they are the most sensitive effects and the most proximal to
 8    route of exposure (inhalation).  These effects were observed in both male and female rats and
 9    mice.
10          The dose-response pattern for laryngeal lesions (NTP, 2002) was amenable to BMD
11    modeling and was used for derivation of the chronic inhalation reference concentration (RfC).
12    Two laryngeal lesions were selected for modeling because they had the lowest regional deposited
13    dose ratio (RDDR) and corresponding benchmark concentration lower limit (adjusted for
14    dosimetric differences across species to humans; BMCL^Eq). In accordance with U.S. EPA
15    BMD methodology (2000a), a benchmark response (BMR) of 10% increase in extra risk was
16    selected to represent a minimally adverse level. All available dichotomous  models were fit to the
17    incidence data for chronic inflammation and for epithelial hyperplasia of the epiglottis. The
18    model with the lowest Akaike's information criterion value was considered to provide a superior
19    fit. BMD modeling of incidence data for chronic inflammation of the larynx and epithelial
20    hyperplasia of the epiglottis in female rats yielded the same BMCLio of 0.003 mg/m3. The
21    shared BMCLio of 0.003 mg/m3 for either chronic inflammation of the larynx or epithelial
22    hyperplasia of the epiglottis was divided by a total UF of 300: 3 to represent interspecies
23    toxicodynamic uncertainties, 10 for interhuman variability in the absence of quantitative
24    information on the variability of response in humans, and 10 for database deficiencies. The
25    result was the chronic RfC of 1  x 10~5 mg/m3.
26          Confidence in the principal  study for derivation of the RfC  (NTP. 2002) is high.  NTP
27    (2002) is a well-conducted study with numerous  doses, large sample sizes of two species of
28    rodents using both genders, and numerous endpoints and time points. Confidence in the overall
29    database, however, is medium.  NTP (2002) remains the single relevant study for use  in the
30    derivation of the chronic RfC. Thus, overall confidence in the chronic RfC is medium.

31    6.2.2. Cancer
32          No studies have identified cancer effects following oral exposure to vanadium pentoxide
33    in either humans or animals, or following inhalation exposure in humans.
34          Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), the database
35    for vanadium pentoxide indicates that it is "likely to be carcinogenic to humans" via the
36    inhalation route of exposure. This determination is based predominantly on the NTP (2002)
37    study, which found positive evidence of lung tumors in both sexes  of mice and male rats after

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

-------
 1    chronic vanadium pentoxide inhalation exposure.  This weight-of-evidence conclusion takes into
 2    consideration the NTP (2002) cancer bioassay, the available human studies, and other laboratory
 3    animal studies. Information available on the carcinogenic effects of vanadium pentoxide via the
 4    inhalation route is limited to examination of the respiratory tumors. Information on the
 5    carcinogenic effects of vanadium pentoxide via the oral and dermal routes in humans or animals
 6    is absent. Based on the observation of only respiratory tumors following inhalation exposure,
 7    and in the absence of information to establish an MO A, this cancer descriptor applies only to the
 8    inhalation route of exposure. Therefore, the database has "inadequate information to assess
 9    carcinogenic potential" of vanadium pentoxide via the oral or dermal route.
10          The increased incidence of lung tumors in  male mice observed in the NTP (2002)
11    104-week inhalation study was used to calculate the inhalation unit risk (IUR) for vanadium
12    pentoxide. The calculated cancer IUR for vanadium pentoxide is 3.4 ug/m3 for the development
13    of alveolar/bronchiolar adenomas or carcinomas in male B6C3Fi mice.  This value was derived
14    from BMCLio, the 95% lower bound on the dose associated with 71% extra cancer risk of
15    respiratory carcinoma in male B6C3Fi mice, by dividing the BMR (0.71) by the BMCLio and
16    represents the  upper bound, continuous lifetime exposure estimate of cancer potency. The
17    BMCLio, lower 95% bound on exposure at 71% risk, is 2.08 x 10"1 mg/m3 and  the slope of the
18    linear extrapolation from the BMCL to the origin,  which is equal to 0.71/2.08 x 10 l mg/m3, or
19    3.4 (mg/m3)"1. A linear low-dose extrapolation approach was used to estimate human
20    carcinogenic risk associated with vanadium pentoxide exposure due to a lack of data that support
21    any specific mode of carcinogenic action of vanadium pentoxide. Therefore, the derived IUR to
22    vanadium pentoxide is 3.4 per mg/m3
23          Areas of uncertainty exist for this cancer assessment. The log-logistic model was
24    selected to model lung tumor incidence in male mice; how well this model or the linear low-dose
25    extrapolation predicts low-dose risks for vanadium pentoxide, however, is unknown. The
26    selected model, although the best-fitting, is not the only model that adequately describes the data.
27    Conceivably, other models could be selected that yield different results consistent with the
28    observed data, both higher and lower than those included in this assessment. The human
29    equivalent lURs estimated from the statistically significant increase in lung tumors ranged from
30    1.4 mg/m3 in male mice to 4.2 mg/m3 in female mice. These tumors are considered to be
31    relevant to humans.  As there is no information to  inform which species or gender of animals
32    would be most applicable to humans, the most sensitive group was selected for  the basis of the
33    IUR.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                102       DRAFT - DO NOT CITE OR QUOTE

-------
                                          7. REFERENCES
Altamirano-Lozano, M; Alvarez-Barrera, L.; Roldan-Reyes, E. (1993) Cytogenetic and teratogenic effects of
        vanadium pentoxide on mice. Med SciRes21: 711-713.
Altamirano-Lozano, M.; Alvarez-Barrera, L.; Basurto-Alcantara, F.; Valverde, M.; Rojas, E. (1996) Reprotoxic and
        genotoxic studies of vanadium pentoxide in male mice. Teratog Carcinog Mutagen 16: 7-17.
Altamirano-Lozano, M.; Valverde, M.; Alvarez-Barrera, L.; Molina, B.; Rojas, E. (1999) Genotoxic studies of
        vanadium pentoxide (V(2)O(5)) in male mice. II. Effects in several mouse tissues. Teratog Carcinog
        Mutagen 19: 243-255.
Altamirano, M.; Ayala, M. E.; Flores, A.; Morales, L.; Dominguez, R. (1991) Sex differences in the effects of
        vanadium pentoxide administration to prepubertal rats. Med Sci Res 19: 825-826.
Antao-Menezes, A.; Turpin, E. A.; Bost, P. C.; Ryman-Rasmussen, J. P.; Bonner, J. C. (2008) STAT-1 signaling in
        human lung fibroblasts is induced by vanadium pentoxide through an IFN-beta autocrine loop. J Immunol
        180: 4200-4207.
Aragon, M. A.; Ayala, M. E.; Fortoul,  T. I.; Bizarro, P.; Altamirano-Lozano, M. (2005) Vanadium induced
        ultrastructural changes and apoptosis in male germ cells. Reprod Toxicol 20: 127-134.
        http://dx.doi.0rg/10.1016/i.reprotox.2004.ll.016.
Assem, F. L. and Levy, L. S. (2009) A review of current lexicological concerns on vanadium pentoxide and other
        vanadium compounds: gaps in knowledge and directions for future research. J Toxicol Environ Health B
        CritRev 12: 289-306. http://dx.doi.org/10.1080/10937400903094166.
Avila-Costa, M. R.; Montiel Flores, E.; Colin-Barenque, L.; Ordonez, J. L.; Gutierrez, A. L.; Nino-Cabrera, H. G.,..
        . Fortoul, T. I. (2004) Nigrostriatal modifications after vanadium inhalation: An immunocytochemical and
        cytological approach. Neurochem Res 29: 1365-1369.
Avila-Costa, M. R.; Colin-Barenque, L.; Zepeda-Rodriguez, A.;  Antuna, S. B.; Saldivar, O. L.; Espejel-Maya, G.,. .
        . Fortoul, T. I. (2005) Ependymal epithelium disruption after vanadium pentoxide inhalation. A mice
        experimental model. Neurosci Lett 381: 21-25. http://dx.doi.0rg/10.1016/i.neulet.2005.01.072.
Avila-Costa, M. R.; Fortoul, T. I.; Nino-Cabrera, G.; Colin-Barenque, L.; Bizarro-Nevares, P.; Gutierrez-Valdez, A.
        L.,. . . Anaya-Martinez, V. (2006) Hippocampal cell alterations induced by the inhalation of vanadium
        pentoxide (V(2)O(5)) promote memory deterioration. Neurotoxicology 27: 1007-1012.
Barceloux, D. G. (1999) Vanadium. J Toxicol Clin Toxicol 37: 265-278.
Bell, M.; Ebisu, K.; Peng, R.; Samet, J.; Dominici, F. (2009) Hospital admissions and chemical composition of fine
        particle air pollution. Am J Respir Crit Care Med 179: 1115-1120. http://dx.doi.org/10.1164/rccm.2008Q8-
        1240OC.
Bonner, J. C.; Lindroos, P. M.; Rice, A. B.; Moomaw, C. R.; Morgan, D. L. (1998) Induction of PDGF receptor-
        alpha in rat myofibroblasts during pulmonary fibrogenesis in vivo. Am J Physiol 274: L72-L80.
Bonner, J. C.; Rice, A. B.; Moomaw, C. R.; Morgan, D. L. (2000) Airway fibrosis in rats induced by vanadium
        pentoxide. Am J Physiol Lung Cell Mol Physiol 278: L209-L216.
Bonner, J. C.; Rice, A. B.; Ingram, J. L.; Moomaw, C. R.; Nyska, A.; Bradbury, A.,... Langenbach, R. (2002)
        Susceptibility of cyclooxygenase-2-deficient mice to pulmonary fibrogenesis. Am J Pathol 161: 459-470.
Bruyere, V I. E.; Morando, P. J.; Blesa, MA. (1999) The dissolution of vanadium pentoxide  in aqueous solutions of
        oxalic and mineral acids. J Colloid Interface Sci 209:207-214.
Byczkowski, J. Z. and Kulkarni, A. P.  (1992) Vanadium redox cycling, lipid peroxidation and co-oxygenation of
        benzo(a)pyrene-7,8-dihydrodiol. BiochimBiophys Acta 1125: 134-141. http://dx.doi.org/10.1016/0005-
        2760(92)90037-V
Byrne, A. R. and Kosta, L. (1978) Vanadium in foods and in human body fluids and tissues. Sci Total Environ 10:
        17-30.
Carter, J. D.; Ohio, A. J.; Samet, J. M.; Devlin, R. B. (1997) Cytokine production by human airway epithelial cells
        after exposure to an air pollution particle is metal-dependent. Toxicol Appl Pharmacol 146:  180-188.
Cavallari, J. M.; Eisen, E. A.; Fang, S. C.; Schwartz, J.; Hauser, R.; Herrick, R. F.; Christiani, D. C. (2008) PM2.5
        metal exposures and nocturnal heart rate variability: a panel study of boilermaker construction workers.
        Environ Health 7. http://dx.doi.org/3610.1186/1476-069x-7-36.
Charles River Laboratories. (Charles River Laboratories  International Inc.). (2008) Hematology for Wistar Rat
        Colonies in North America for January-December 2008. Wilmington, MA: Author.


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

-------
Clayton, G. D. and Clayton, F. E. (1981) Patty's industrial hygiene and toxicology: Toxicology, vol. 2 (3 ed. Vol. 2).
        New York, NY: John Wiley & Sons.
Colin-Barenque, L.; Martinez-Hernandez, M. G.; Baiza-Gutman, L. A.; Avila-Costa, M. R.; Ordonez-Librado, J. L.;
        Bizarro-Nevares, P.,... Fortoul, T. I. (2008) Matrix metalloproteinases 2 and 9 in central nervous system
        and their modification after vanadium inhalation. J Appl Toxicol 28: 718-723.
        http://dx.doi.org/10.1002/iat. 1326.
Colin-Barenque, L.; Martinez-Hernandez, M. G.; Baiza-Gutman, L. A.; Avila-Costa, M. R.; Ordonez-Librado, J. L.;
        Bizarro-Nevares, P.,... Fortoul, T. I. (2007) Matrix metalloproteinases 2 and 9 in central nervous system
        and their modification after vanadium inhalation. J Appl Toxicol 28: 718-723.
        http://dx.doi.org/10.1002/iat. 1326.
Conklin, A. W.; Skinner, C. S.; Felten, T. L.; Sanders, C. L. (1982) Clearance and distribution of intratracheally
        instilled 48vanadium compounds in the rat. Toxicol Lett 11: 199-203. http://dx.doi.org/10.1016/0378-
        4274(82)90128-X.
Cooper, R. (2007) Vanadium pentoxide inhalation. Indian J Occup Environ Med 11: 97-102.
        http://dx.doi.org/10.4103/0019-5278.38457.
Crans, D. C.; Smee, J. J.; Gaidamauskas, E.; Yang, L. (2004) The chemistry and biochemistry of vanadium and the
        biological activities exerted by vanadium compounds. Chem Rev 104: 849-902.
        http://dx.doi.org/10.1021/cr020607t.
Dekker, J. M.; Schouten, E. G.; Klootwijk, P.; Pool, J.; Swenne, C. A.; Kromhout, D. (1997) Heart rate variability
        from short electrocardiographic recordings predicts mortality from all causes in middle-aged and elderly
        men. The Zutphen Study. Am J Epidemiol 145: 899-908.
Devereux, T. R.; Holliday, W.; Anna, C.; Ress, N.; Roycroft, J.; Sills, R.  C. (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, J. A.; Lee, K. M.; Mellinger, K.  H.; Bates, D. J.; Burka, L. T.; Roycroft, J. H. (2004) Lung deposition and
        clearance of inhaled vanadium pentoxide in chronically exposed F344 rats andB6C3Fl mice. Toxicol Sci
        77: 6-18. http://dx.doi.org/10.1093/toxsci/kfh005.
Dominici, F.; Peng, R. D.; Ebisu, K.; Zeger, S. L.; Samet, J. M.; Bell, M. L. (2007) Does the effect of PM10 on
        mortality depend on PM nickel and vanadium content? A reanalysis of the NMMAPS data. Environ Health
        PerspectllS: 1701-1703.
Edel, J.; Pietra, R.; Sabbioni, E.; Marafante, E.; Springer, A.; Ubertalli, L. (1984) Disposition of vanadium in rat
        tissues at different age.  Chemosphere 13: 87-93. http://dx.doi.org/10.1016/0045-6535(84)90010-9.
Edel, J. and 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, V. A.; Nersesyan,  A. K.; Hoelzl, C.; Ferk, F.; Bichler,  J.; Valic, E.,. .. Knasmuller, S. (2008) Inhalative
        exposure to vanadium pentoxide causes DNA damage in workers: Results of a multiple end point study.
        Environ Health Perspect 116: 1689-1693. http://dx.doi.org/10.1289/ehp.11438.
Fortoul, T. I.; Bizarro-Nevares, P.; Acevedo-Nava, S.; Pinon-Zarate, G.; Rodriguez-Lara, V.; Colin-Barenque, L.,. .
        . Saldivar-Osorio, L. (2007) Ultrastructural findings in murine seminiferous tubules as a consequence of
        subchronic vanadium pentoxide inhalation. Reprod Toxicol 23:  588-592.
Gonzalez-Villalva, A.; Fortoul, T. I.;  Avila-Costa, M. R.; Pinon-Zarate, G.; Rodriguez-Laraa, V.; Martinez-Levy,
        G.,. . . Colin-Barenque, L. (2006) Thrombocytosis induced in mice after subacute and subchronic V2O5
        inhalation. Toxicol Ind Health 22: 113-116. http://dx.doi.org/10.1191/0748233706th250oa.
Gylseth, B.; Leira, H. L.; Steinnes, E.; Thomassen, Y. (1979) Vanadium in the blood and urine of workers in a
        ferroalloy plant. Scand J Work Environ Health 5: 188-194.
Harris, W. R.; Friedman, S. B.; Silberman, D. (1984) Behavior of vanadate and vanadyl ion in canine blood. J Inorg
        Biochem20: 157-169. http://dx.doi.org/10.1016/0162-0134(84)80015-X.
Hauser, R.; Elreedy, S.; Hoppin, J. A.; Christiani, D. C. (1995a) Upper airway response in workers exposed to fuel
        oil ash: nasal lavage analysis. Occup Environ Med 52: 353-358.
Hauser, R.; Elreedy, S.; Hoppin, J. A.; Christiani, D. C. (1995b) Airway obstruction inboilermakers exposed to fuel
        oil ash: a prospective investigation. Am J Respir Crit Care Med  152: 1478-1484.
HHauser, R.; Eisen, E. A.; Pothier, L.; Christiani, D. C. (2001) A prospective study of lung function among
        boilermaker construction workers exposed to combustion particulates. Am J Ind Med 39: 454-462.
        http://dx.doi.org/10.1002/aiim.103.
             This document is a draft for review purposes only and does not constitute Agency policy.
                                                  104        DRAFT - DO NOT CITE OR QUOTE

-------
Hauser, R.; Eisen, E. A.; Pothier, L.; Lewis, D.; Bledsoe, T.; Christian!, D. C. (2002) Spirometric abnormalities
        associated with chronic bronchitis, asthma, and airway hyperresponsiveness among boilermaker
        construction workers. Chest 121: 2052-2060.
HSDB. (Hazardous Substances Data Bank). (2008). Vanadium pentoxide. Bethesda, MD: National Library of
        Medicine.  Retrieved from http ://toxnet.nlm.nih. gov
IARC. (International Agency for Research on Cancer). (2006) Cobalt in hard metals and cobalt sulfate, gallium
        arsenide, indium phosphide and vanadium pentoxide. Lyon, France: Author.
        http://monographs.iarc.fr/ENG/Monographs/vol86/index.php.
Ingram, J. L.; Rice, A. B.;  Santos, J.; VanHouten, B.; Bonner, J. C. (2003) Vanadium-induced HB-EOF expression
        in human lung fibroblasts is oxidant dependent;and requires MAP kinases. Am J Physiol Lung Cell Mol
        Physiol 284: L774-L782.
Ingram, J. L.; Antao-Menezes, A.; Turpin, E. A.; Wallace, D. G.; Mangum, J. B.; Pluta, J. L.,. . . Bonner, J. C.
        (2007) Genomic analysis of human lung fibroblasts exposed to vanadium pentoxide to identify candidate
        genes for occupational bronchitis. Respir Res 8: 34. http://dx.doi.org/10.1186/1465-9921-8-34.
IPCS. (International Programme on Chemical Safety). (2001) Glossary of exposure assessment-related terms: A
        compilation. Geneva, Switzerland: Author.
        http://www.who.int/ipcs/publications/methods/harmonization/en/compilation nov2001.pdf.
Irsigler, G. B.;  Visser, P. J.; Spangenberg, P. A. L. (1999) Asthma and chemical bronchitis in vanadium plant
        workers. Am J  Ind Med 35: 366-374.
Ivancsits, S.; Pilger, A.;  Diem, E.; Schaffer, A.; Rtidiger, H. W. (2002) Vanadate induces DNA strand breaks in
        cultured human fibroblasts at doses relevant to occupational exposure. Mutat Res 519: 25-35.
        htto://dx.doi.org/10.1016/S1383-5718(02)00138-9.
Kada, T.; Hirano, K.; Shirasu, Y. (1980). Screening of environmental chemical mutagens by the Rec-assay system
        with Bacillus subtilis. InF. J. d. Serres & A. Hollaender (Eds.), Chemical mutagens: Principles and
        methods for their detection (Vol. 6, pp. 149-173). New York, NY: Plenum Press.
Kanematsu, N.; Hara, M.; Kada, T. (1980) REC assay and mutagenicity studies on metal compounds. Mutat Res 77:
        109-116. http://dx.doi.org/10.1016/0165-1218(80)90127-5.
Kerckaert, G. A.; LeBoeuf, R. A.; Isfort, R. J. (1996) Use of the Syrian hamster embryo cell transformation assay
        for determining the carcinogenic potential of heavy metal compounds. Fundam Appl Toxicol 34: 67-72.
        http://dx.doi.org/10.1006/faat.1996.0176.
Kim, J. Y.; Mukherjee, S.; Ngo, L. C.; Christiani, D. C.  (2004) Urinary 8-hydroxy-2'-deoxyguanosine as a
        biomarker of oxidative DNA damage in workers exposed to fine particulates. Environ Health Perspect 112:
        666-671.
Kiviluoto, M.;  Pyy, L.; Pakarinen, A. (1979) Serum and urinary vanadium of vanadium-exposed workers. Scand J
        Work Environ Health 5: 362-367.
Kiviluoto, M. (1980) Observations on the lungs of vanadium workers. Occup Environ Med 37: 363-366.
Kiviluoto, M.;  Pakarinen, A.; Pyy, L. (1981a) Clinical laboratory results of vanadium-exposed workers. Arch
        Environ Occup Health 36: 109-113.
Kiviluoto, M.;  Pyy, L.; Pakarinen, A. (1981b) Serum and urinary vanadium of workers processing vanadium
        pentoxide. Int Arch Occup Environ Health 48:  251-256.
Kleinfeld, M; Messite, J.; Kooyman, O.; Goldwater, LJ. (1961) Fingernail cystine content in chronic mercury
        exposure. Arch Environ Health 3:676-379.
Kleinsasser, N.; Dirschedl, P.; Staudenmaier, R.; Harreus, U.; Wallner, B. (2003) Genotoxic effects of vanadium
        pentoxide on human peripheral lymphocytes and mucosal cells of the upper aerodigestive tract.  Int J
        Environ Health Res 13: 373-379.
Knecht, E. A.;  Moorman, W. J.; Clark, J. C.; Lynch, D. W.; Lewis, T. R. (1985) Pulmonary effects of acute
        vanadium pentoxide inhalation in monkeys. Am Rev Respir Dis 132:  1181-1185.
Knecht, E. A.;  Moorman, W. J.; Clark, J. C.; Hull, R. D.; Biagini, R. E.; Lynch, D. W.,.  . . Simon, S.  D. (1992)
        Pulmonary reactivity to vanadium pentoxide following subchronic inhalation exposure in a non-human
        primate animal model. J Appl Toxicol 12: 427-434.
Kucera, J.; Byrne, A. R.; Mravcova, A.; Lener, J. (1992) Vanadium levels in hair and blood of normal and exposed
        persons. Sci Total Environ 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. Biol Trace ElemRes 43-45: 327-334. http://dx.doi.org/10.1007/BF02917332.
             This document is a draft for review purposes only and does not constitute Agency policy.
                                                  105        DRAFT - DO NOT CITE OR QUOTE

-------
Kyono, H.; Serita, F.; Toya, T.; Kubota, H.; Arito, H.; Takahashi, M.,. . . Kagawa, J. (1999) A new model rat with
        acute bronchiolitis and its application to research on the toxicology of inhaled paniculate matter. Ind Health
        37: 47-54.
Lagorio, S.; Forastiere, F.; Pistelli, R.; lavarone, I.; Michelozzi, P.; Fano, V.,... Ostro, B. D. (2006) Air pollution
        and lung function among susceptible adult subjects: A panel study. Environ Health 5:11.
        http://dx.doi.org/10.1186/1476-069X-5-11.
Lees, R. E. (1980) Changes in lung function after exposure to vanadium compounds in fuel oil ash. Br J Ind Med 37:
        253-256.
Levy, B. S.; Hoffman, L.; Gottsegen, S. (1984) Boilermakers bronchitis: respiratory tract irritation associated with
        vanadium pentoxide exposure during oil-to-coal conversion of a power plant. J Occup Environ Med 26:
        567-570.
Lewis, C. E. (1959) The biological effects of vanadium: II the signs and symptoms of occupational vanadium
        exposure. AMA Arch Ind Health 19: 497-503.
Lewis, R. J. (1997) Hawley's condensed chemical dictionary (13 ed.). New York, NY: John Wiley & Sons, Inc.
Lewis, R. J., Sr.  (2000) Sax's Dangerous Properties of Industrial Materials (10 ed.). New York, NY: John Wiley &
        Sons, Inc.
Lippmann, M.; Ito, K.; Hwang, J.  S.; Maciejczyk, P.; Chen, L. C. (2006) Cardiovascular effects of nickel in ambient
        air. Environ Health Perspect 114: 1662-1669.
Liu, Y. C.; Woodin, M. A.; Smith, T. J.; Herrick, R. F.; Williams, P. L.; Hauser, R.; Christiani, D. C. (2005)
        Exposure to fuel-oil ash and welding emissions during the overhaul of an oil-fired boiler. J Occup Environ
        Hyg 2: 435-443. http://dx.doi.org/10.1080/15459620591034529.
Magari, S. R.; Schwartz, J.; Williams, P. L.; Hauser, R.; Smith, T. J.; Christiani, D. C. (2002) The association of
        paniculate air metal concentrations with heart rate variability. Environ Health Perspect 110: 875-880.
Montiel-Davalos, A.; Gonzalez-Villava, A.; Rodriguez-Lara, V.; Felipe Montano, L.; Fortoul, T,I,; Lopez-Marure,
        R. (2011) Vanadium pentoxide induces activation and death of endothelial cells. J Appl Toxicol (Epub
        ahead of print).
Motolese,  A.; Truzzi, M.; Giannini, A.; Seidenari, S. (1993) Contact dermatitis and contact sensitization among
        enamellers and decorators in the ceramics industry. Contact Dermatitis 28: 59-62.
Mountain, J. T.; Delker, L. L.; Stokinger, H. E. (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.; Lener, J. (1993) Effects of orally administered vanadium on the immune system
        and bone metabolism in experimental animals. Sci Total Environ Suppl Pt 1: 663-669.
Mukherjee, B.; Patra, B.; Mahapatra, S.; Banerjee, P.; Tiwari, A.; Chatterjee, M. (2004) Vanadium~an element of
        atypical biological significance. Toxicol Lett 150: 135-143. http://dx.doi.0rg/10.1016/i.toxlet.2004.01.009.
Musk, A. W. and Tees, J. G. (1982) Asthma caused by occupational exposure to vanadium compounds. Med J Aust
        1: 183-184.
Mussali-Galante, P.; Rodriguez-Lara, V.; Hernandez-Tellez, B.; Avila-Costa, M. R.; Colin-Barenque, L.; Bizarro-
        Nevarez, P.,... Fortoul,  T. I. (2005) Inhaled vanadium pentoxide decrease gamma-tubulin of mouse testes
        at different exposure times. Toxicol Ind Health 21: 215-222. http://dx.doi.org/10.1191/0748233705th232oa.
NIOSH. (National Institute for Occupational Safety and Health). (2005) NIOSH pocket guide to chemical hazards
        (NPG)--index of chemical abstract numbers (CAS No). Cincinnati, OH: Author.
        http://www.cdc.gov/niosh/npg/npgdcas.html.
NRC. (National Research Council). (1983) Risk assessment in the federal government: Managing the process.
        Washington, DC: National Academies Press, http://www.nap.edu/openbook.php7record id=366&page=Rl.
NTP. (National Toxicology Program). (2002) NTP toxicology and carcinogensis studies of vanadium pentoxide
        (CAS No.; 1314-62-1) in F344/N rats and B6C3F1 mice (inhalation).  Washington, DC: Author.
NTP. (National Toxicology Program). (2008) Chemical information review document for oral exposure to
        tetravalent and pentavalent vanadium compounds. Research Triangle Park, NC: U.S. Department of Health
        and Human Services.
O'Neil, M. (2001). [Entries for platinum and platinum compounds] The Merck index: an encyclopedia of chemicals,
        drugs, and biologicals 13th ed (pp. 306-307, 403-404, 1350, 1369, 1373). Whitehouse Station, NJ: Merck
        &Co.
Oberg, S. G.; Parker, R. D. R.; Sharma, R. P. (1978) Distribution and elimination of an intratracheally administered
        vanadium compound in the rat. Toxicology 11: 315-323.
             This document is a draft for review purposes only and does not constitute Agency policy.
                                                  106        DRAFT - DO NOT CITE OR QUOTE

-------
Orris, P.; Cone, I; McQuilkin, S. (1983) Health hazard evaluation report HETA 80-096-1359, Eureka Company,
        Bloomington, Illinois. (PB85163574). Washington, DC: U.S. Department of Health and Human Services,
        National Institute of Occupational Safety and Health.
OSHA. (Occupational Safety & Health Administration). (2007) Occupational safety and health guideline for
        vanadium pentoxide dust.  Washington, DC: Occupational Safety and Health Administration.
        http://www.osha.gov/SLTC/healthguidelines/vanadiumpentoxidedust/recognition.html.
Parker, R. D. and Sharma, R. P. (1978) Accumulation and depletion of vanadium in selected tissues of rats treated
        withvanadyl sulfate and sodium orthovanadate. J Environ Pathol Toxicol 2: 235-245.
Patel, M. M; Hoepner, L.; Garfinkel, R.; Chillrud, S.; Reyes, A.; Quinn, J. W.,. .  . Miller, R. L. (2009) Ambient
        metals, elemental carbon, and wheeze and cough in New York City children through 24 months of age. Am
        J Respir Crit Care Med 180: 1107-1113. http://dx.doi.org/10.1164/rccm.200901-0122OC.
Patterson, B. W.; Hansard SL, I. L; Ammerman, C. B.; Henry, P. R.; Zech, L. A.; Fisher, W. R. (1986) Kinetic
        model of whole-body vanadium metabolism: studies in sheep. Am J Physiol 251: R332.
Pazynich, V. M. (1966) Maximum permissible concentration of vanadium pentoxide in the atmosphere. Gig Sanit
        317-9:6-12.
Pierce, L. M.; Alessandrini, F.; Godleski, J. J.; Paulauskis, J. D. (1996) Vanadium-induced chemokine mRNA
        expression and pulmonary inflammation. Toxicol Appl Pharmacol 138: 1-11.
Pinon-Zarate, G.; Rodriguez-Lara, V.; Rojas-Lemus, M.; Martinez-Pedraza, M.; Gonzalez-Villalva, A.; Mussali-
        Galante, P.,... Montano, L. F. (2008) Vanadium pentoxide inhalation provokes germinal center
        hyperplasia and suppressed humoral immune responses. J Immunotoxicol 5: 115-122.
        http://dx.doi.org/10.1080/15476910802085749.
Radike, M.; Warshawsky, D.; Caruso, J.; Goth-Goldstein, R.; Reilman, R.; Collins, T.,. . .  Schneider, J. (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. J
        Toxicol Environ Health A 65: 2029-2052. http://dx.doi.org/10.1080/00984100290071324.
Rahman, F. and Skyllas-Kazacos, M. (1998) Solubility of vanadyl sulfate in concentrated sulfuric acid solutions. J
        Power Sourc 72: 105-110. http://dx.doi.org/10.1016/S0378-7753(97)02692-X.
Ramirez, P.; Eastmond, D. A.; Laclette, J.  P.; Ostrosky-Wegman, P. (1997) Disruption of microtubule assembly and
        spindle formation as a mechanism for the;induction of aneuploid cells by sodium arsenite and vanadium
        pentoxide. Mutat Res 386: 291-298.  http://dx.doi.org/10.1016/S1383-5742(97)00018-5.
Ress, N. B.; Chou, B. J.; Renne, R. A.; Dill, J. A.; Miller, R. A.; Roycroft, J. H.,.  . . Bucher, J. R. (2003)
        Carcinogenicity of inhaled vanadium pentoxide in F344/N rats and B6C3F1 mice. Toxicol Sci 74: 287-296.
        http://dx.doi.org/10.1093/toxsci/kfgl36.
Rhoads, K. and Sanders, C. L. (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, A. B.; Moomaw, C. R.; Morgan, D. L.; Bonner, J. C. (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, L. A.; Altamirano-Lozano, M.; Ostrosky-Wegman, P. (1996) Genotoxicity of
        vanadium pentoxide evaluate by the  single cell gel electrophoresis assay in human lymphocytes. Mutat Res
        359: 77-84.
Roldan, R. E. and Altamirano, L. M. (1990) Chromosomal aberrations, sister-chromatid exchanges, cell-cycle
        kinetics and satellite associations in human lymphocyte cultures exposed to vanadium pentoxide. Mutat
        Res 245: 61-65.
Rondini,  E. A.; Walters, D. M.; Bauer, A. K. (2010) Vanadium pentoxide induces pulmonary inflammation and
        tumor promotion in a strain-dependent manner. Part Fibre Toxicol 7: 9. http://dx.doi.org/10.1186/1743-
        8977-7-9.
Roshchin, A. V.; Ordzhonikidze, E. K.; Shalganova, I. V. (1980) Vanadium-toxicity, metabolism, carrier state.
        Cent Eur J Public Health 24: 377-383.
Ross, D.  S. (1983) Case study: exposure to vanadium pentoxide. Occup Health (Lond) 35:  67-71.
Sabbioni, E.; Kueera, J.; Pietra, R.; Vesterberg, O. (1996) A critical review on normal concentrations of vanadium in
        human blood, serum, and urine. Sci Total Environ 188: 49-58.
Samet, J. M.; Silbajoris, R.; Wu, W.; Graves,  L. M. (1999) Tyrosine phosphatases as targets in metal-induced
        signaling in human airway epithelial cells. Am J Respir Cell Mol Biol 21: 357-364.


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

-------
Schroeder, H. A.; Balassa, J. I; Tipton, I. H. (1963) Abnormal trace metals in man-vanadium. J Chronic Dis 16:
        1047-1071. http://dx.doi.org/10.1016/0021-9681(63)90041-9.
Sharma, R. P.; Oberg, S. G.; Parker, R. D. (1980) Vanadium retention in rat tissues following acute exposures to
        different dose levels. J Toxicol Environ Health 6: 45-54. http://dx.doi.org/10.1080/15287398009529829.
Sjoberg, S. G. (1951) Health hazards in the production and handling of vanadium pentoxide. AMA Arch Ind Hyg
        OccupMed3:631-646.
Sjoberg, S. G. (1955) Vanadium bronchitis from cleaning oil-fired boilers. AMA Arch Ind Health 11: 505-512.
Sjoberg, S. G. (1956) Vanadium dust, chronic bronchitis and possible risk of emphysema: a follow-up investigation
        of workers at a vanadium factory. J Intern Med 154: 381-386.
Sorensen, M; Schins, R. P. F.; Hertel, O.; Loft, S. (2005) Transition metals in personal samples of PM25 and
        oxidative stress in human volunteers. Cancer Epidemiol Biomarkers Prev 14: 1340-1343.
Stokinger, H. E.; Wagner, W. D.; Mountain, J. T.; Stacksill, F. R.; Dobrogorski, O. J.; Keenan, R. G. (1953)
        Unpublished results. Cincinnati, OH: Division of Occupational Health.
Turpin, E. A.; Antao-Menezes, A.; Cesta, M. F.; Mangum, J. B.; Wallace, D. G.; Bermudez, E.; Bonner, J. C. (2010)
        Respiratory syncytial virus infection reduces lung inflammation and fibrosis in mice exposed to vanadium
        pentoxide. RespirRes 11: 20. http://dx.doi.org/10.1186/1465-9921-11-20.
U.S. EPA. (U.S. Environmental Protection Agency). (1986a) Guidelines for the health risk assessment of chemical
        mixtures. (EPA/630/R-98/002). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
        Forum, http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=22567.
U.S. EPA. (U.S. Environmental Protection Agency). (1986b) Guidelines for mutagenicity risk assessment.
        (EPA/630/R-98/003). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
        http://www.epa. gov/iris/backgrd.html.
U.S. EPA. (U.S. Environmental Protection Agency). (1987) Integrated Risk Information System (IRIS). IRIS
        summary of vanadium pentoxide (CASRN 1314-62-1). Washington, DC: Office of Research and
        Development, National Center for Environmental Assessment, http://www.epa.gov/iris/subst/0125.htm.
U.S. EPA. (U.S. Environmental Protection Agency). (1988) Recommendations for and documentation of biological
        values for use in risk assessment. (EPA/600/6-87/008). Cincinnati, OH: U.S.  Environmental Protection
        Agency, Environmental Criteria and Assessment Office.
        http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=34855.
U.S. EPA. (U.S. Environmental Protection Agency). (1991) Guidelines for developmental toxicity risk assessment.
        (EPA/600/FR-91/001). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
        http://www.epa. gov/iris/backgrd.html.
U.S. EPA (U.S. Environmental Protection Agency). (1992) Support: Dust inhalation toxicity studies with cover
        letter dated 08-19-92. Office of Toxic Substances, Washington, DC; Doc #89 940000275; see also Doc
        #88-920000666, initial submission.
U.S. EPA. (U.S. Environmental Protection Agency). (1994a) Methods for derivation of inhalation reference
        concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F). Research Triangle Park, NC:
        U.S. Environmental Protection Agency, Office of Research and Development, Office of Health and
        Environmental Assessment, Environmental Criteria and Assessment Office.
        http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=71993.
U.S. EPA. (U.S. Environmental Protection Agency). (1994b) Interim policy for particle size and limit concentration
        issues in inhalation toxicity studies. Washington, DC: U.S. Environmental Protection Agency, Health
        Effects Division, Office of Pesticide Products.
        http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid= 186068.
U.S. EPA. (U.S. Environmental Protection Agency). (1995) The use of the benchmark dose approach in health risk
        assessment. (EPA/630/R-94/007). Washington, DC: U.S. Environmental Protection Agency, Risk
        Assessment Forum, http://www.epa.gov/raf/publications/useof-bda-healthrisk.htm.
U.S. EPA. (U.S. Environmental Protection Agency). (1996) Guidelines for reproductive toxicity risk assessment.
        (EPA/630/R-96/009). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
        http://www.epa.gov/raf/publications/pdfs/REPRO51.PDF.
U.S. EPA. (U.S. Environmental Protection Agency). (1998) Guidelines for neurotoxicity risk assessment.
        (EPA/630/R-95/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
        http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF.
U.S. EPA. (U.S. Environmental Protection Agency). (2000a) Science Policy Council handbook: Peer review. (EPA
        100-B-OO-OOl). Washington, DC: U.S. Environmental Protection Agency, Science Policy Council.
        http://www.ntis.gov/search/product.aspx? ABBR=PB2005109156.

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

-------
U.S. EPA. (U.S. Environmental Protection Agency). (2000b) Supplementary guidance for conducting health risk
        assessment of chemical mixtures. (EPA/630/R-00/002). Washington, DC: U.S. Environmental Protection
        Agency, Risk Assessment Forum, http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=2053 3.
U.S. EPA. (U.S. Environmental Protection Agency). (2000c) Benchmark dose technical guidance document
        [external review draft]. (EPA/630/R-00/001). Washington, DC: U.S. Environmental Protection Agency,
        Risk Assessment Forum, http://www.epa.gov/raf/publications/benchmark-dose-doc-draft.htm.
U.S. EPA. (U.S. Environmental Protection Agency). (2002) A review of the reference dose and reference
        concentration processes. (EPA/630/P-02/0002F). Washington, DC: U.S. Environmental Protection Agency,
        Risk Assessment Forum, http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=51717.
U.S. EPA. (U.S. Environmental Protection Agency). (2005a) Guidelines for carcinogen risk assessment.
        (EPA/630/P-03/00IF). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
        http://www.epa.gov/cancerguidelines/.
U.S. EPA. (U.S. Environmental Protection Agency). (2005b) Guidelines for carcinogen risk assessment and
        supplemental guidance for assessing susceptibility from early-life exposure to carcinogens. (EPA/630/P-
        03/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
        http://www.epa.gov/cancerguidelines/.
U.S. EPA. (U.S. Environmental Protection Agency). (2006a) A framework for assessing health risks of
        environmental exposures to children. (EPA/600/R-05/093A). Washington, DC: U.S. Environmental
        Protection Agency, National Center for Environmental Assessment.
        http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=158363.
U.S. EPA. (U.S. Environmental Protection Agency). (2006b) Science Policy Council handbook: Peer review.
        (EPA/100/B-06/002). Washington, DC: U.S. Environmental Protection Agency, Science Policy Council.
        http://www.epa.gov/OSA/spc/2peerrev.htm.
U.S. EPA. (U.S. Environmental Protection Agency). (2011) Recommended use  of body weight 3/4 as the default
        method in derivation of the oral reference dose.  (EPA/100/R11/0001). Washington, DC: Author.
        http://www.epa.gov/raf/publications/interspecies-extrapolation.htm.
Vintinner, F. I; Vallenas, R.; Carlin, C. E.; Weiss, R.; Macher, C.; Ochoa, R. (1955) Study of the health of workers
        employed in mining and processing of vanadium ore. Arch Environ Occup Health 12: 635-642.
Wallenborn, J. G.; McKee, J. K.; Schladweiler, M. C.; Ledbetter, A. D.; Kodavanti, U. P. (2007) Systemic
        translocation of paniculate matter-associated metals following a single intratracheal instillation in rats. J
        ToxicolSci98:231-239.
WHO. (World Health Organization). (1988) Vanadium.   Copenhagen:  Author.
WHO. (World Health Organization). (2001) Concise international chemical assessment document 29: Vanadium
        pentoxide and other inorganic vanadium compounds.  Geneva, Switzerland: World Health Organization,
        International Programme on Chemical Safety.
        http://www.inchem.org/documents/cicads/cicads/cicad29.htm.
Wide, M. (1984) Effect of short-term exposure to five industrial metals on the embryonic and fetal development of
        the mouse. EnvironRes  33: 47-53.
Williams, N. (1952) Vanadium poisoning from cleaning oil-fired boilers. Br J Ind Med 9: 50-55.
Woodin, M. A.; Hauser, R.; Liu, Y.; Smith, T. J.; Siegel,  P. D.; Lewis, D. M.,. . . Christiani, D. C. (1998) Molecular
        markers of acute upper airway inflammation in workers exposed to fuel-oil ash. Am J Respir Crit Care Med
        158:  182-187.
Woodin, M. A.; Liu, Y.; Hauser, R.; Smith, T. J.; Christiani, D. C. (1999) Pulmonary function  in workers exposed to
        low levels of fuel-oil ash. J Occup EnvironMed 41: 973-980.
Woodin, M. A.; Liu, Y.; Neuberg, D.; Hauser, R.; Smith, T. J.; Christiani, D. C. (2000) Acute respiratory symptoms
        in workers exposed to vanadium-rich fuel-oil ash. Am J Ind Med 37: 353-363.
Wright,  J. R.; Yates, A. J.; Shah, N. T.; Neff, J. C.; Covey, D. W.; Thibert, P. (1983) Hematological Characteristics
        of the BB WistarRat. Vet ClinPathol 12: 9-13.  http://dx.doi.org/10.1111/i. 1939-165X. 1983.tb00828.x.
Yamaguchi, M.; Oishi, H.; Suketa, Y. (1989) Effect of vanadium on bone metabolism in weanling rats: zinc
        prevents the toxic effect of vanadium. Res Exp Med 189: 47-53.
Yu, D.; Walters, D. M.; Zhu, L.; Lee, P.; Chen, Y. (2011) Vanadium pentoxide (V2O5) induced mucin production
        by airway epithelium. Am J Physiol Lung Cell Mol Physiol 301: L31-39.
Zenz, C.; Bartlett, J. P.; Thiede, W. H. (1962) Acute vanadium pentoxide intoxication. Arch Environ Occup Health
        5: 542-546.
             This document is a draft for review purposes only and does not constitute Agency policy.
                                                  109        DRAFT - DO NOT CITE OR QUOTE

-------
Zenz, C. and Berg, B. A. (1967) Human responses to controlled vanadium pentoxide exposure. Arch Environ Occup
        Health 14: 709-712.
Zhang, L.; Rice, A. B.; Adler, K.; Sannes, P.; Martin, L.; Gladwell, W.,. . . Bonner, J. C. (2001a) Vanadium
        stimulates human bronchial epithelial cells to produce heparin-binding epidermal growth factor-like growth
        factor: a mitogen for lung fibroblasts. Am J Respir Cell Mol Biol 24: 123-131.
Zhang, Z.; Wang, Y.; Vikis, H. G.; Johnson, L.; Liu, G.; Li, I,. . . You, M. (200Ib) Wildtype Kras2 can inhibit lung
        carcinogenesis in mice. Nat Genet 29: 25-33. http://dx.doi.org/10.1038/ng721.
Zhang, T.; Gou, X.; Yang, Z. (1991) A study on developmental toxicity of vanadium pentoxide in NIH mice. Huaxi
        yike daxue xuebao 22: 191-195.
Zhang, T.; Gou, X.; Yang, Z. (1993a) A study of teratogenicity and sensitive period of vanadium pentoxide in
        Wistar rats. Huaxi yike daxue xuebao 24: 202-205.
Zhang, T.; Yang, Z.; Zeng, C.; Gou, X. (1993b) A study on developmental toxicity of vanadium pentoxide in Wistar
        rats. Huaxi yike daxue xuebao 24: 92-96.
Zhang, Z.; Wang, Y.; Vikis, H. G.; Johnson, L.; Liu, G.; Li, J.,. . . You, M. (200Ib) Wildtype Kras2 can inhibit lung
        carcinogenesis in mice. Nat Genet 29: 25-33. http://dx.doi.org/10.1038/ng721.
Zhong, B. Z.; Gu, Z. W.; Wallace, W. E.; Whong, W. Z.; Ong, T. (1994) Genotoxicity of vanadium pentoxide in
        Chinese hamster V79 cells. MutatRes 321: 35-42.
             This document is a draft for review purposes only and does not constitute Agency policy.
                                                  110        DRAFT - DO NOT CITE OR QUOTE

-------
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
                 COMMENTS AND DISPOSITION
                 This page intentionally left blank
                         A-1        DRAFT - DO NOT CITE OR QUOTE

-------
 1               APPENDIX B.  BENCHMARK CONCENTRATION MODELING OF
 2                       INHALATION STUDIES IN RATS FROM NTP, 2002

 3          To derive the reference concentration (RfC) for vanadium pentoxide, inhalation toxicity
 4    effects observed in rats (NTP, 2002) were modeled to estimate the candidate points of departure
 5    (PODs).  Five endpoints in both male and female rats (see Table B-2) were selected for two
 6    reasons:

 7          •  The study in rats was designed with lower concentrations of vanadium pentoxide.
 8          •  The results showed both biological significance and statistically significant trends.

 9          Each dataset was fit with the dichotomous models available in EPA Benchmark Dose
10    Software (BMDS; version 2.1.2). Following the model selection steps outlined in the draft
11    Benchmark Dose Technical Guidance (U.S. EPA, 2000), the best-fitting model for each dataset
12    was used to estimate the candidate POD, which was the benchmark concentration (BMC) lower
13    limit (BMCL) at the selected benchmark response (BMR) as 10% extra risk.
14          The highlights  of the benchmark dose (BMD) modeling results are:

15          •  The POD value of 0.003 mg/m3, from two endpoints in female rats, was the lowest
16             POD derived and was used for calculating the RfC. The two endpoints were larynx
17             chronic inflammation and larynx respiratory epithelium epiglottis hyperplasia.
18          •  All candidate PODs and the selected models are summarized in Table B-l.
19    B.I.  TREND  TESTS ON THE INHALATION DATASETS FOR BMDS MODELING
20          Ten inhalation  toxicology datasets in rats were selected because of the lower
21    concentrations of vanadium pentoxide administrated (as compared to mice), the statistically
22    significant biologically adverse effects observed, and the statistical significance reported by NTP
23    (2002). A Cochran-Armitage test was conducted to confirm the significance of the statistical
24    trend for the selected datasets (Table B-2) before modeling.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-l       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"
BMR
(extra
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
Log-Logistic
Log-Logistic
0.1
0.1
0.017
0.008
0.012
0.006
Nose
Goblet cell, respiratory epithelium, hyperplasia [Log-Logistic
0.1
0.044
0.026
Female F344/N Rats
Lung
Alveolar epithelium hyperplasia
Chronic active inflammation
Gamma
Multistage (Stage 3)
0.1
0.1
0.076
0.080
0.063
0.048
Larynx
Chronic inflammation
Respiratory epithelium, epiglottis, hyperplasia
Log-Logistic
Log-Logistic
0.1
0.1
0.005
0.004
0.003
0.003
Nose
Goblet cell, respiratory epithelium, hyperplasia
Multistage (Stage 2)
0.1
0.038
0.014
""Selected model is the best-fitting model for the dataset based on the draft Benchmark Dose Technical Guidance
(U.S. EPA. 20001.
bHEC = human equivalent concentration.
CBMCL = the lower bound of BMC at 95% confidence level.
      Table B-2.  Trend tests on the selected datasets from the 2-year inhalation
      studies in rats
Endpoint
Concentration as reported (mg/m3)3
0.0
0.5
1.0
2.0
Trend testb
Z-score
p-value
Incidence0 in male F344/N rats
Lung
Alveolar epithelium hyperplasiad
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

0.0001
0.0001

O.0001
0.0001

0.0037
          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-2       DRAFT - DO NOT CITE OR QUOTE

-------
             Table B-2. Trend tests on the selected datasets from the 2-year inhalation
             studies in rats
Endpoint
Concentration as reported (mg/m3)a
0.0
0.5
1.0
2.0
Trend testb
Z-score
p-value
Incidence0 in female F344/N rats
Lung
Alveolar epithelium hyperplasiad
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.0001
O.0001

0.0001
O.0001

0.0003
      Concentrations are as reported by NTP (2002).
      bOne-sided Cochran-Armitage trend test.
      Incidence = (number of animals affected)/(number of animals examined).
      "^his dataset was calculated by combining the incidences of alveolar epithelium hyperplasia and bronchiole
      epithelium hyperplasia.
      Source: NTP (2002).
 1    B.2.  DOSE CONVERSION
 2          In the toxicology and carcinogenesis studies reported by NTP (2002), rats (F344/N) were
 3    exposed to vanadium pentoxide through inhalation. To analyze the concentration-response effect
 4    of vanadium pentoxide, the reported concentrations of vanadium pentoxide were converted to
 5    human equivalent concentrations (HECs) before any modeling and extrapolation were
 6    completed.
 7          First, the average lifetime animal body weights of rats were estimated based on the mean
 8    body weight at different weeks (Table B-3).  Then, following the Methods for Derivation of
 9    Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994),
10    regional deposited dose ratios (RDDRs) were calculated with the RDDR program designed by
11    US EPA (1994).  Although six regional RDDRs were reported by the RDDR program for each
12    concentration/sex group, only two regional RDDRs (i.e., extrathoracic and pulmonary) are
13    relevant to the selected endpoints, which are summarized in Table B-4. Finally, the HEC for
14    each concentration/sex group was calculated from the reported concentration in rats by
15    multiplying the continuous exposure adjustment factor and RDDR (Table B-5).  For endpoints in
16    the lung, the pulmonary RDDR was applied; for endpoints in the larynx and nose, the
17    extrathoracic RDDR was applied.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-3       DRAFT - DO NOT CITE OR QUOTE

-------
       Table B-3. Average lifetime animal body weight of rats in the 2-year
       inhalation studies of vanadium pentoxide
Vanadium pentoxide
concentration as reported
(mg/m3)
Mean animal body weight" (g)
1-13 wk
14-52 wk
53-104 wk
Average lifetime animal
body weight1" (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
aAs reported in Table 11-12 of the NTP report (2002).
bAverage lifetime animal body weight = (mean body weight 1-13 weeks x 13 + mean body weight 14-52 weeks
39 + mean body weight 53-104 weeks * 52)7104.
Source: NTP (2002).
       Table B-4. RDDRs for various concentration/sex groups in the 2-year
       inhalation studies of vanadium pentoxide
Vanadium pentoxide
concentration as
reported (mg/m3)
Average life time
animal body weight"
(g)
Average
MMADb
Average GSDC
RDDRd
Extrathoracic
Pulmonary
Male rats (F344/N)
0
0.5
1
2
440
449
443
432
1.24
1.89
0.516
0.530
0.520
0.503
0.496
0.494
0.495
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
"All average lifetime animal body weights were calculated in Table B-3.
bMMAD = mass median aerodynamic diameter; calculated based on the MMADs reported for 2-year studies (NTP.
2002).
°GSD = geometric standard deviation; calculated based on the GSDs reported for 2-year studies (NTP. 2002).
dRDDR = regional deposited dose ratio; calculated with the RDDR program (V.2.3. US EPA).
Source: NTP (2002).
           This document is a draft for review purposes only and does not constitute Agency policy.
                                             B-4        DRAFT - DO NOT CITE OR QUOTE

-------
           Table B-5. Human equivalent concentrations of vanadium pentoxide in the
           2-year inhalation studies
Concentration as
reported" (mg/m3)
Continuous exposure
adjustment factor1"
RDDRC
Extrathoracic
Pulmonary
Human equivalent concentration"1
(mg/m3)
Extrathoracic
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
    ^Toxicology and carcinogenesis studies of vanadium pentoxide in F344/N rats and B6C3Fj mice, NTP (2002).
    bContinuous exposure adjustment factor = (6/24) x (5/7); animals were exposed to vanadium pentoxide 6 hours per day,
    5 days per week.
    °Refer to Appendix Table B-4.
    dHuman equivalent concentration = concentration as reported x continuous exposure adjustment factor x RDDR.
    Source:  NTP (2002).

 1   B.3.  BMDS MODELING FOR INHALATION DATASETS
 2          Each selected inhalation dataset was fit with all the standard dichotomous models
 3   available in EPA BMDS (version 2.1.2.).  Following the model selection steps outlined in the
 4   Benchmark Dose Technical Guidance (U.S. EPA, 2000), the best-fitting model was used to
 5   estimate the candidate POD for each endpoint, which was the BMCL at the selected BMR as
 6   10% extra risk.
 7          The selected models and candidate PODs for all endpoints are summarized in Table B-l.

 8   B.3.1.  Dataset 1: Incidence of Lung Alveolar Epithelium Hyperplasia in Male Rats, NTP
 9   (2002)
10   Summary
11          All three concentration groups (0.5, 1, and 2 mg/m3) showed statistically significant
12   differences from the control as reported. The severity of the nonneoplastic lesions increased
13   from mild to moderate as reported when the concentration of vanadium pentoxide increased.
14   The Cochran-Armitage test confirmed the statistically significant trend with a Z-score = 8.83 and
15   one-sided/>-value < 0.0001.
16          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
17   dataset. A 10% extra risk of the combined incidence was used as the BMR to determine the
18   POD.  The goodness of fit, BMC, and BMCL for each model are summarized in Table B-6.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-5       DRAFT - DO NOT CITE OR QUOTE

-------
1          Five models demonstrated adequate goodness of fit (p-value > 0.1) and good visual fit.
2   Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Probit model
3   was selected as the best-fitting model for this dataset. Using this model, the BMCLio was
4   0.013 mg/m3 and this value was regarded as a candidate for the POD.
            Table B-6. Benchmark modeling results for incidence of lung alveolar
            epithelium hyperplasia in male rats
Model3
Probit
Logistic
Multistage6
(Stage 3)
Weibull
Gamma
Log-Probit
Log-Logistic
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
     ""Selected (best-fitting) model shown in first row, in boldface type.
     bAIC = Akaike Information Criterion
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     "For the Multistage model, up to three stages tested; only the model with lowest AIC and lowest stage is reported here.
     Source: NTP (2002X
    BMDS output file
             Probit Model.  (Version:  3.2;   Date:  10/28/2009)
             Input Data  File:
     C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLungAlBrEpHyperplasia\NTP
     _2002_Lung Alveolar Bronchiole Combined Epithelium Hyperplasia_Probit_0.1.(d)
             Gnuplot  Plotting  File:
     C:\USEPA\BMDS212\Data\V205\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  )

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

-------
 intercept
     slope
intercept
        1
    -0.78
slope
-0.78
    1

  Parameter Estimates
Interval
       Variable
Limit
      intercept
0.676464
          slope
25.5971
       Model
     Full model
   Fitted model
  Reduced model

           AIC:
Estimate
-1.02981
20.0578
Std. Err.
0.180284
2.82623
                                           95.0% Wald Confidence

                                        Lower Conf. Limit   Upper  Conf.

                                               -1.38316

                                                14.5184
          Analysis of Deviance Table
    Log(likelihood)  # Param's  Deviance  Test d.f.
         -82.2383         4
         -82.9994         2       1.52225      2
         -133.424         1       102.372      3

          169.999

                    Goodness  of  Fit
                                                                     P-value
 ChiA2 = 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
                                         0.4671
                                        <.0001

Dose
0.0000
0.0441
0.0884
0.1779

Est. Prob.
0.1515
0.4423
0.7713
0.9944

Expected
7.577
21.673
37.023
49.721

Observed
7.000
24.000
35.000
50.000

Size
50
49
48
50
Scaled
Residual
-0.228
0. 669
-0. 695
0.530
           This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-7       DRAFT - DO NOT CITE OR QUOTE

-------
                         Probit Model with 0.95 Confidence Level
           0.6
           0.2
                   0.02   0.04  0.06
       11:0009/102010
                                0.08   0.1
                                  dose
                                         0.12  0.14   0.16  0.18
 1   B.3.2. Dataset 2: Incidence of Lung Chronic Active Inflammation in Male Rats, NTP
 2   (2002)
 3   Summary
 4          Two concentration groups (1 and 2 mg/m3) showed statistically significant differences
 5   from the control group as reported.  The severity increased from minimal to mild as reported
 6   when the concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed
 7   the statistically significant trend with a Z-score = 8.36 and a one-sided/7-value < 0.0001.
 8          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 9   dataset.  A 10% extra risk of the incidence was used as the BMR to determine the POD.  The
10   goodness of fit, BMC, and BMCL for each model are summarized in Table B-7.
11          All models demonstrated  adequate goodness of fit (p-value > 0.1) and good visual fit.
12   Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Logistic model
13   was selected as the best-fitting model for this dataset.  Using this model, the BMCLio was
14   0.029 mg/m3 and this value was regarded as a candidate for the POD.
            Table B-7.  Benchmark modeling results for incidence of lung chronic active
            inflammation in male rats
Model3
Logistic
Probit
Log-Probit
Log-Logistic
Goodness of fit
/7-Value
0.40
0.37
0.67
0.60
AICb
192.5
192.7
192.8
192.9
Largest scaled
residual
-0.92
-1.00
0.29
-0.34
Scaled residual of
interest
-0.92
-1.00
-0.26
-0.34
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3)
0.035
0.032
0.046
0.045
BMCLd
(mg/m3)
0.029
0.027
0.032
0.031
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-8       DRAFT - DO NOT CITE OR QUOTE

-------
           Table B-7.  Benchmark modeling results for incidence of lung chronic active
           inflammation in male rats
Model3
Gamma
Weibull
Multistage*5
(Stage 2)
Goodness of fit
/7-Value
0.40
0.27
0.24
AICb
193.3
193.9
194.0
Largest scaled
0.58
-0.75
0.88
Scaled residual of
-0.52
-0.75
-0.66
BMR
HECC
BMC
0.042
0.038
0.040
BMCLd
0.027
0.024
0.019
     Selected (best-fitting) model shown in first row, in boldface type.
     bAIC = Akaike Information Criterion.
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
     reported here.

     Source: NTP (2002).

1   BMDS output file
             Logistic Model.  (Version:  2.13; Date: 10/28/2009)
             Input Data  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLungChronicActiveInflamma
    tion\NTP_2002_Lung  Chronic Active Inflammation_Logistic_0.1.(d)
             Gnuplot Plotting  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLungChronicActiveInflainma
    tion\NTP_2002_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,
     intercept
          slope
    Interval
                      and do not appear in the correlation  matrix )
intercept
        1
    -0.82
slope
-0.82
    1

  Parameter Estimates
                                                               95. (
                                                                     Wald Confidence
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-9       DRAFT - DO NOT CITE OR QUOTE

-------
3
4
5
6
7
            Variable
     Limit
           intercept
     1.66732
               slope
     29.3268
       Model
     Full model
   Fitted model
  Reduced model
           AIC:
                         Estimate

                         -2.29582

                          23.0148
                       Std.  Err.     Lower Conf.  Limit   Upper Conf.

                        0.320666             -2.92431

                         3.22048              16.7027
      Analysis  of  Deviance Table
Log(likelihood)  # Param's  Deviance  Test d.f.    P-value
                       4
                       2         1.8908      2           0.3885
                       1        78.6961      3          <.0001
                            -93.3159
                            -94.2613
                            -132.664
                              192.523
                                        Goodness of  Fit
          Dose
                   Est. Prob.
                                  Expected
                                              Observed
                                                       Size
                                                                        Scaled
                                                                       Residual
0.0000
0.0441
0.0884
0.1779
ChiA2 = 1.84
0.0915
0.2174
0.4350
0.8578
d.f. = 2
4.574 5.000
10.654 8.000
20.880 24.000
42.892 42.000
P-value = 0.3976
50
49
48
50

0.209
-0.919
0.908
-0.361

        Benchmark Dose Computation
     Specified effect =             0.1
     Risk Type        =      Extra risk
     Confidence level =            0.95
                  BMD =      0.0345486
                 BMDL =      0.0285337
                        Logistic Model with 0.95 Confidence Level
                   Logistic
                  BMDL BMD
      11:2409/102010
               0   0.02   0.04  0.06  0.08   0.1   0.12  0.14   0.16  0.18
                                 dose
1    B.3.3.  Dataset 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
differences 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 a Z-score = 4.79 and a
one-sided/>-value < 0.0001.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-10      DRAFT - DO NOT CITE OR QUOTE

-------
1          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
2    dataset. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
3    goodness of fit, BMC, and BMCL for each model are summarized in Table B-8.
4          Only one model demonstrated adequate goodness of fit (p-value > 0.1) and good visual
5    fit. Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Log-Logistic
6    model was selected as the best-fitting model for this dataset. Using this model, the BMCLio of
7    was 0.012 mg/m3 and this value was regarded as a candidate for the POD.
            Table B-8.  Benchmark modeling results for incidence of larynx chronic
            inflammation in male rats
Model3
Log-Logistic
Log-Probit
Multistage6
(Stage 1)
Gamma
Weibull
Probit
Logistic
Goodness of Fit
/7-Value
0.12
0.09
0.05
0.01
0.01
AICb
229.1
229.8
230.7
234.8
235.2
Largest scaled
residual
1.63
-1.37
2.14
2.45
2.43
Scaled residual of
interest
-0.27
-0.01
-0.65
2.45
2.43
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3)
0.017
0.005
0.023
0.043
0.046
BMCd
(mg/m3)
0.012
0.000
0.017
0.036
0.038
     "Selected (best-fitting) model shown in first row, in boldface type.
     bAIC = Akaike Information Criterion.
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
     reported here.

     Source:  NTP (2002).
8   BMDS output file

             Logistic Model.  (Version:  2.13;  Date:  10/28/2009)
             Input Data  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLarynxChronicInflainmation
    \NTP_2002_Larynx Chronic Inflammation_LogLogistic_0.1.(d)
             Gnuplot Plotting  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLarynxChronicInflaitimation
    \NTP_2002_Larynx Chronic Inflammation_LogLogistic_0.1.pit

        The form of the  probability function is:
        P[response]  = background+(1-background)/[1+EXP(-intercept-siope*Log(dose))]

        Dependent variable = Response
        Independent variable = Dose
        Slope parameter  is restricted as slope >= 1
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-11       DRAFT - DO NOT CITE OR QUOTE

-------
   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 user,
background
 intercept
(  ***  The  model  parameter(s)   -slope
      have been  estimated at  a boundary point,  or have been specified by

      and  do  not appear in the correlation matrix )
  background
           1
       -0.53
intercept
    -0.53
        1
                                 Parameter Estimates
Interval
       Variable
Limit
     background
      intercept
          slope
             Estimate

            0.0712314
               1.9037
                    1
             Std. Err.
         95.0% Wald Confidence

      Lower Conf. Limit   Upper Conf.
    Indicates that this value is not calculated.
                        Analysis of Deviance Table
       Model
     Full model
   Fitted model
  Reduced model

           AIC:
       Log(likelihood)
            -110.451
             -112.55
            -127.371

             229.101
        # Param's
             4
             2
             1
                                              Deviance  Test d.f.
                                                                    P-value
4.19924
33.8403
 0.1225
<.0001
                                  Goodness  of  Fit

Dose
0.0000
0.0473
0.0929
0.1796

Est. Prob.
0.0712
0.2951
0.4278
0.5789

Expected
3.490
14.754
21.390
28.366

Observed
3.000
20.000
17.000
28.000

Size
49
50
50
49
Scaled
Residual
-0.272
1. 627
-1.255
-0.106
 ChiA2 =4.31
                   d.f. = 2
                                   P-value = 0.1162
   Benchmark Dose Computation
Specified effect =            0.1
Risk Type        =      Extra risk
Confidence level =           0.95
             BMD =      0.0165573
            BMDL =      0.0117279
           This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-12      DRAFT - DO NOT CITE OR QUOTE

-------
                        Log-Logistic Model with 0.95 Confidence Level
           0.5
           0.2
                       Log-Logistic
               BMDL  BMD
       13:2209/102010
                0   0.02   0.04  0.06  0.08  0.1   0.12   0.14  0.16  0.18
                                   dose
 1   B.3.4.  Dataset 4:  Incidence of Larynx Respiratory Epithelium, Epiglottis, Hyperplasia in
 2   Male Rats, NTP (2002)

 3   Summary
 4          Three concentration groups (0.5, 1, and 2 mg/m3) showed statistically significant
 5   differences from the control group as reported. The severity increased slightly when the
 6   concentration of vanadium pentoxide increased.  The Cochran-Armitage test confirmed the
 7   statistically significant trend with a Z-score = -6.56 and a one-sided/>-value < 0.0001.
 8          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 9   dataset. A 10% extra risk of the incidence was used as the BMR to determine the POD.  The
10   goodness of fit, BMC, and BMCL for each model are summarized in Table B-9.
11          Both Log-Logistic and Log-Probit models demonstrated adequate goodness of fit
12   (p-value ^0.1) and good visual fit but the BMC/BMCL ratio in Log-Probit was infinite, which is
13   not acceptable. Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the
14   Log-Logistic model was selected as the best-fitting model for this dataset. The BMCLio was
15   0.006 mg/m3 and this  value was regarded as a candidate for the POD.
             Table B-9. Benchmark modeling results for incidence of larynx respiratory
             epithelium, epiglottis, hyperplasia in male rats
Model3
Log-Logistic
Log-Probit
Goodness of Fit
/7-Value
0.27
0.14
AICb
197.3
199.3
Largest scaled
residual
1.56
1.60
Scaled residual of
interest
0.000
0.000
BMR
Extra
Risk
HECC
BMC
(mg/m3)
0.008
0.007
BMCLd
(mg/m3)
0.006
0.000
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-13      DRAFT - DO NOT CITE OR QUOTE

-------
           Table B-9.  Benchmark modeling results for incidence of larynx respiratory
           epithelium, epiglottis, hyperplasia in male rats
Model3
Gamma
Multistage*5
(Stage 1)
Weibull
Probit
Logistic
Goodness of Fit
/7-Value
0.04
0.00
0.00
AICb
201.5
226.4
227.1
Largest scaled
-2.12
3.16
3.05
Scaled residual of
0.000
1.180
1.078
BMR
10%
HECC
BMC
0.013
0.031
0.032
BMCLd
0.010
0.027
0.027
     Selected (best-fitting) model shown in first row, in boldface type.
     bAIC=Akaike Information Criterion.
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
     reported here.

     Source:  NTP (2002).


1   BMDS output file

             Logistic Model.  (Version:  2.13; Date: 10/28/2009)
             Input  Data  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLarynxEpiglottisHyperplas
    ia\NTP_2002_Larynx  Epiglottis Hyperplasia_LogLogistic_0.1.(d)
             Gnuplot Plotting File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsLarynxEpiglottisHyperplas
    ia\NTP_2002_Larynx  Epiglottis Hyperplasia_LogLogistic_0.1.pit

        The  form  of  the  probability function is:
        P[response]  = background+(1-background)/[1+EXP(-intercept-siope*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 user,
(  ***  The model parameter(s)  -background     -slope
      have been estimated at a boundary  point,  or have been specified by

      and do not appear in the correlation  matrix )

   intercept
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-14      DRAFT - DO NOT CITE OR QUOTE

-------
      intercept
     Interval
            Variable
     Limit
         background
           intercept
               slope
                                      Parameter Estimates
Estimate        Std. Err.
 2.66112
       1
         Indicates that this value is not calculated.

                             Analysis of Deviance Table
            Model      Log(likelihood)  # Param's  Deviance   Test  d.f.
          Full  model        -95.6454         4
        Fitted  model        -97.6263         1       3.96181       3
       Reduced  model        -134.962         1       78.6325       3
   95.0% Wald Confidence

Lower Conf. Limit   Upper Conf.
                                             P-value
                AIC:
                             197.253
                                       Goodness  of  Fit
            =3.93
                        d.f.  = 3
                                        P-value = 0.2694
        Benchmark Dose Computation
     Specified effect =            0.1
     Risk  Type        =      Extra risk
     Confidence level =           0.95
                  BMD =     0.00776335
                 BMDL =     0.00584847

                       Log-Logistic Model with 0.95 Confidence Level
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
13:4009/102
'
^^— - \
^^-^ \
/^ L ;
/
/ '
/ :
/ j
/ :
.Tj
BMDL 3MD :
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
dose
010
                                                 0.2656
                                                <.0001

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
1    B.3.5.  Dataset 5:  Incidence of Nose Goblet Cell, Respiratory Epithelium, Hyperplasia in
2    Male Rats, NTP (2002)
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-15      DRAFT - DO NOT CITE OR QUOTE

-------
 1   Summary
 2          Three concentration groups (0.5, 1, and 2 mg/m3) showed statistically significant
 3   differences from the control group as reported. The severity as reported increased slightly when
 4   the concentration of vanadium pentoxide increased.  The Cochran-Armitage test confirmed the
 5   statistically significant trend with a Z-score = 2.68 and a one-sided/>-value = 0.0037.
 6          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 7   dataset. A 10% extra risk of the incidence was used as the BMR to determine the POD.  The
 8   goodness of fit, BMC, and BMCL for each model are summarized in Table B-10.
 9          Five models demonstrated adequate goodness of fit (p-value > 0.1) and good visual fit
10   but Log-Probit model failed to compute a reasonable BMCL. Based on the draft Benchmark
11   Dose Technical Guidance (U.S. EPA, 2000), the Log-Logistic model was selected as the best-
12   fitting model for this dataset.  Using this model, the BMCLio was 0.026 mg/m3 and this value
13   was regarded as a candidate for the POD.
             Table B-10. Benchmark modeling results for incidence of nose goblet cell,
             respiratory epithelium, hyperplasia in male rats
Model3
Log-Logistic
Log-Probit
Gamma
Multistage6
(Stage 1)
Weibull
Probit
Logistic
Goodness of Fit
/7-Value
0.16
0.31
0.12
0.08
0.07
AICb
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
      Selected (best-fitting) model shown in first row, in boldface type.
      bAIC = Akaike Information Criterion.
      °HEC = human equivalent concentration.
      dBMCL = the lower bound of BMC at 95% confidence level.
      Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
      reported here.
      Source:  NTP (2002).

14   BMDS output file
              Logistic  Model.  (Version: 2.13; Date: 10/28/2009)
              Input  Data  File:
     C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsNoseGobletEpiHyperplasia\
     NTP_2002_Larynx Goblet Hyperplasia _LogLogistic_0.1.(d)
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-16      DRAFT - DO NOT CITE OR QUOTE

-------
         Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerMaleRats\MaleRatsNoseGobletEpiHyperplasia\
NTP_2002_Larynx Goblet Hyperplasia _LogLogistic_0.1.pit


   The form of the probability function is:
   P[response] = background+(1-background)/[1+EXP(-intercept-siope*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 =    0.0816327
                      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
                                                         95.0% Wald Confidence
Interval
       Variable         Estimate        Std. Err.     Lower Conf. Limit   Upper Conf.
Limit
     background         0.110444            *                *                  *
      intercept         0.926382            *                *                  *
          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        -102.873         4
   Fitted model        -104.611         2       3.47486      2           0.176
  Reduced model        -109.105         1       12.4644      3         0.00595

           AIC:
Goodness of Fit

Dose
0.0000
0.0473
0.0929
0.1796

Est. Prob.
0.1104
0.2054
0.2794
0.3881

Expected
5.412
10.270
13.691
18.627

Observed
4.000
15.000
12.000
17.000

Size
49
50
49
48
Scaled
Residual
-0.643
1.656
-0.539
-0.482
       = 3.68      d.f. = 2        P-value = 0.1590
           This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-17      DRAFT - DO NOT CITE OR QUOTE

-------
         Benchmark Dose Computation
      Specified effect =            0.1
      Risk Type        =      Extra risk
      Confidence level =           0.95
                   BMD =      0.0439982
                  BMDL =       0.026051
                        Log-Logistic Model with 0.95 Confidence Level
           0.5
           0.4
           0.3
           0.2
                       Log-Logistic
       14:0809/102010
                0   0.02  0.04  0.06   0.08  0.1   0.12   0.14  0.16  0.18
                                   dose
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
B.3.6.  Dataset 6: Incidence of Lung Alveolar Epithelium Hyperplasia in Female Rats,
NTP (2002)
Summary
       Two concentration groups (1 and 2 mg/m3) showed statistically significant differences
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 a Z-score = 9.53 and a one-sided
/7-value < 0.0001.
       Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
dataset. 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 fit (p-value > 0.1) but the extreme
curvature of Log-Logistic model did not reelect in the observed data.  Based on the draft
Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Gamma model was selected as the
best-fitting model for this dataset. Using this model, the BMCLio was 0.063 mg/m3 and this
value was regarded as a candidate for the POD.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-18      DRAFT - DO NOT CITE OR QUOTE

-------
       Table B-ll.  Benchmark modeling results for incidence of lung alveolar
       epithelium hyperplasia in female rats
Model a
Gamma
Log-Logistic
Weibull
Log-Probit
Multistage6
(Stage 3)
Logistic
Probit
Goodness of Fit
P-Value
0.89
0.96
0.91
0.78
0.46
0.00
0.00
AICb
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
"Selected (best-fitting) model shown in first row, in boldface type.
bAIC = Akaike Information Criterion.
°HEC = human equivalent concentration.
dBMCL = the lower bound of BMC at 95% confidence level.
Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
reported here.

Source: NTP (2002).


BMDS output file

         Gamma Model.  (Version:  2.15;   Date:  10/28/2009)
         Input Data  File:
C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLungAlBrEpHyperplasia
\NTP_2002_Lung Alveolar  Bronchiole Combined Epithelium  Hyperplasia_Gamma_0.1.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLungAlBrEpHyperplasia
\NTP_2002_Lung Alveolar  Bronchiole Combined Epithelium  Hyperplasia_Gamma_0.1.pit

   The form of the  probability function is:

   P[response]=  background+(1-background)*CumGamma[siope*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
           This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-19      DRAFT - DO NOT CITE OR QUOTE

-------
the user,
Background
     Slope
(  ***  The  model  parameter(s)   -Power
      have been  estimated at  a boundary point,  or have been specified by

      and  do  not appear in the correlation matrix )
  Background
           1
       -0.32
Slope
-0.32
    1

  Parameter Estimates
Interval
       Variable
Limit
     Background
0.221723
          Slope
188.208
          Power
             Estimate

             0.151066

              169.414

                   18
         Std. Err.

           0.03605

           9.58875

                NA
   95.0% Wald Confidence

Lower Conf. Limit   Upper Conf.

       0.080409

        150.621
NA - Indicates that this parameter has hit a bound
     implied by some ineguality constraint and thus
     has no standard error.
       Model
     Full model
   Fitted model
  Reduced model

           AIC:
             Analysis  of Deviance Table

       Log(likelihood)   # Param's  Deviance  Test d.f.    P-value
            -75.9175          4
            -76.0968          2       0.358619      2          0.8358
            -135.531          1        119.227      3         <.0001

             156.194

                       Goodness  of  Fit

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
 ChiA2 = 0.22
                   d.f. = 2
                                   P-value = 0.8945
   Benchmark Dose Computation
Specified effect =            0.1
Risk Type        =      Extra risk
Confidence level =           0.95
             BMD =      0.0756821
            BMDL =     0.0626004
           This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-20      DRAFT - DO NOT CITE OR QUOTE

-------
           0.6
       14:4009/102010
                      Gamma Multi-Hit Model with 0.95 Confidence Level
                       Gamma Multi-Hit
                          BMDL   BMD
                         0.05
                                    0.1
                                  dose
                                              0.15
 1   B.3.7.  Dataset 7: Incidence of Lung Chronic Active Inflammation in Female Rats, NTP
 2   (2002)
 3   Summary
 4          One concentration group (2 mg/m3) showed statistically significant difference from the
 5   control group as reported.  The severity did not change much when the concentration of
 6   vanadium pentoxide increased. The Cochran-Armitage test confirmed the statistically significant
 7   trend with a Z-score = 6.71 and a one-sided/>-value < 0.0001.
 8          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 9   dataset. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
10   goodness of fit, BMC, and BMCL for each model are summarized in Table B-12.
11          Five models demonstrated adequate goodness of fit (p-value > 0.1) and good visual fit.
12   Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Multistage
13   (Stage 3) model was selected as the best-fitting model for this dataset. The BMCLio was
14   0.048 mg/m3 and this value was regarded as a candidate for the POD.

            Table B-12. Benchmark modeling results for incidence of lung chronic
            active inflammation in female rats
Model3
Multistage"
(Stage 3)
Log-Probit
Gamma
Goodness of fit
/7-Value
0.82
1.00
0.99
AICb
212.9
214.5
214.5
Largest scaled
residual
-0.51
0.00
-0.01
Scaled residual
of interest
-0.51
0.00
0.00
BMR
Extra
Risk
10%
HECC
BMC
(mg/m3)
0.080
0.094
0.094
BMCLd
(mg/m3)
0.048
0.068
0.063
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-21      DRAFT - DO NOT CITE OR QUOTE

-------
           Table B-12.  Benchmark modeling results for incidence of lung chronic
           active inflammation in female rats
Model3
Log-Logistic
Weibull
Logistic
Probit
Goodness of fit
/7-Value
0.97
0.94
0.04
0.03
AICb
214.5
214.5
218.8
219.4
Largest scaled
-0.03
-0.06
1.64
-1.71
Scaled residual
0.00
0.01
-0.48
-0.62
BMR
HECC
BMC
0.094
0.094
0.040
0.037
BMCLd
0.065
0.059
0.033
0.031
     "Selected (best-fitting) model shown in first row, in boldface type.
     AIC = Akaike Information Criterion.
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
     reported here.

     Source: NTP (2002).
1   BMDS output file
             Multistage Model.  (Version:  3.2;  Date: 05/26/2010)
             Input Data File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLungChronicActiveInfl
    ammation\NTP_2002_Lung Chronic Active Inflammation_Multi3_0.1.(d)
             Gnuplot Plotting  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLungChronicActiveInfl
    ammation\NTP_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/xl-beta2*dose/x2-beta3*dose/x3) ]

       The  parameter betas are restricted to be positive
       Dependent variable = Response
       Independent  variable = Dose

     Total  number of observations = 4
     Total  number of records with missing values =  0
     Total  number of parameters in model = 4
     Total  number of specified parameters =  0
     Degree of  polynomial = 3

     Maximum number of iterations = 250
     Relative Function Convergence has been  set to: le-008
     Parameter  Convergence has been set to:  le-008

                       Default Initial Parameter Values
                          Background =      0.17928
                             Beta(l)  =            0
                             Beta(2)  =            0
                             Beta(3)  =      214.579

                Asymptotic Correlation Matrix of Parameter Estimates
    the user,
(  ***  The model parameter(s)  -Beta(l)     -Beta(2)
      have been estimated at a boundary point, or have been specified by

      and do not appear in the correlation matrix )
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-22      DRAFT - DO NOT CITE OR QUOTE

-------
             Background       Beta (3)
Background             1         -0.41
   Beta(3)        -0.41             1
Interval
       Variable         Estimate
Limit
     Background         0.186734
        Beta(l)                 0
        Beta(2)                 0
        Beta(3)           206.262
               Parameter  Estimates


                      Std.  Err.

                           *
                           *
   95.0% Wald Confidence

Lower Conf. Limit   Upper  Conf.
* - 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.
     -104.257          4
     -104.451          2       0.389305       2
     -130.861          1         53.209       3

      212.903

                Goodness  of  Fit
                                                                     P-value
                  0.8231
                 <.0001

Dose
0.0000
0.0468
0.0936
0.1871

Est. Prob.
0.1867
0.2037
0.3132
0.7896

Expected
9.150
9.983
15.659
39.478

Observed
10.000
10.000
14.000
40.000

Size
49
49
50
50
Scaled
Residual
0.312
0.006
-0.506
0.181
 ChiA2 = 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
                   Multistage Model with 0.95 Confidence Level
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
15:2809/102
	 H^ ,!.(!;,,+,,„„ 	 -


/ '-
/
/ '-
/ ;
/ '-
/ \
/ '-
/ -
/ -
y '-
/ '-
// ;
;S :
BMDL BMD ;
0 0.05 0.1 0.15
dose
010
           This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-23       DRAFT - DO NOT CITE OR QUOTE

-------
 1   B.3.8.  Dataset 8: Incidence of Larynx Chronic Inflammation in Female Rats, NTP (2002)
 2   Summary
 3          Three concentration groups (0.5, 1, and 2 mg/m3) showed statistically significant
 4   difference from the control group as reported. The severity did not change much when the
 5   concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed the
 6   statistically significant trend with a Z-score = 5.38 and a one-sided/>-value < 0.0001.
 7          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 8   dataset. A 10%  extra risk of the incidence was used as the BMR to determine the POD. The
 9   goodness of fit, BMC and BMCL for each model are summarized in Table B-13.
10          Five models demonstrated adequate goodness of fit (p-value > 0.1) and good visual fit.
11   Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Log-Logistic
12   model was selected as the best-fitting model for this dataset. Using this model, the BMCLio was
13   0.003 mg/m3 and this value was regarded as a candidate for the POD.
14

             Table B-13.  Benchmark modeling results for incidence of larynx chronic
             inflammation in female rats
Model3
Log-Logistic
Log-Probit
Gamma
Multistage*5
(Stage 1)
Weibull
Probit
Logistic
Goodness of fit
/7-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
BMCLd
(mg/m3)
0.003
0.000
0.006
0.011
0.011
      "Selected (best-fitting) model shown in first row, in boldface type.
      bAIC = Akaike Information Criterion.
      °HEC = human equivalent concentration.
      dBMCL = the lower bound of BMC at 95% confidence level.
      Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
      reported here.
      Source:  NTP (2002).
15   BMDS output file

              Logistic Model.  (Version:  2.13;  Date: 10/28/2009)
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-24      DRAFT - DO NOT CITE OR QUOTE

-------
         Input Data File:
C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxChronicInflainma
tion\NTP_2002_Larynx Chronice Inflammation_LogLogistic_0.1.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxChronicInflaitima
tion\NTP_2002_Larynx Chronice Inflammation_LogLogistic_0.1.pit

   The form of the probability function is:
   P[response] = background+(1-background)/[1+EXP(-intercept-siope*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,
background
 intercept
                 and do not appear in the correlation matrix )
background
         1
     -0.54
intercept
    -0.54
        1
                                 Parameter Estimates
Interval
       Variable
Limit
     background
      intercept
          slope
           Estimate

           0.166778
            3.19725
                  1
             Std. Err.
* - Indicates that this value is not calculated.
   95.0% Wald Confidence

Lower Conf. Limit   Upper Conf.
                        Analysis of Deviance Table
       Model
     Full model
   Fitted model
  Reduced model

           AIC:
     Log(likelihood)
          -118.217
          -118.997
          -137.233

           241.994
# Param' s
4
2
1
Deviance

1.55893
38.0314
Test d.f.

2
3
P-value

0.4587
<.0001
Goodness of Fit

Dose
0.0000
0.0231
0.0470
0.0875

Est. Prob.
0.1668
0.4678
0. 6123
0.7347

Expected
8.339
22.925
30.001
36.735

Observed
8.000
26.000
27.000
37.000

Size
50
49
49
50
Scaled
Residual
-0.129
0.881
-0.880
0.085
       = 1.57
                   d.f. = 2
                                   P-value = 0.4553
           This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-25      DRAFT - DO NOT CITE OR QUOTE

-------
        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
u.»
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
16:2909/102
i_uy-i_uyisLio — -
^__^--^ )
,~-^^ 1 j
' /-""" J
/ :
/
/ :
- / \

BMDL BMD
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.0£
dose
010
 1   B.3.9. Dataset 9: Incidence of Larynx Respiratory Epithelium, Epiglottis, Hyperplasia in
 2   Female Rats, NTP (2002)
 3   Summary
 4          Three concentration groups (0.5, 1, and 2 mg/m3) showed statistically significant
 5   differences from the control group as reported. The severity did not change much when the
 6   concentration of vanadium pentoxide increased. The Cochran-Armitage test confirmed the
 7   statistically significant trend with a Z-score = -5.98 and a one-sided/>-value < 0.0001.
 8          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 9   dataset. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
10   goodness of fit, BMC, and BMCL for each model  are summarized in Table B-14.
11          Both the Log-Logistic and Log-Probit models  demonstrated adequate goodness of fit
12   (p-value > 0.1)  and good visual fit but the Log-Probit model failed to compute a reasonable
13   BMCL. Based  on the draft Benchmark Dose Technical Guidance (U.S. EPA. 20001 the
14   Log-Logistic model was selected as the best-fitting model for the dataset.  Using this model, the
15   BMCLio was 0.003 mg/m3 and this value was regarded as a candidate for the POD.
16

            Table B-14. Benchmark modeling results for incidence of larynx respiratory
            epithelium, epiglottis, hyperplasia in female rats
         Model3
Goodness of fit
BMR
HECC
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-26      DRAFT - DO NOT CITE OR QUOTE

-------

Log-Logistic
Log-Probit
Gamma
Multistage6
(Stage 1)
Weibull
Probit
Logistic
/7-Value
0.30
0.78
0.01
0.00
0.00
AICb
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

Extra
Risk
10%
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
     "Selected (best-fitting) model shown in first row, in boldface type.
     bAIC = Akaike Information Criterion.
     °HEC = human equivalent concentration.
     dBMCL = the lower bound of BMC at 95% confidence level.
     Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
     reported here.

     Source:  NTP (2002).
1   BMDS output file

             Logistic Model.  (Version:  2.13; Date: 10/28/2009)
             Input Data  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxEpiglottisHyper
    plasia\NTP_2002_Larynx Epiglottis Hyperplasia_LogLogistic_0.1.(d)
             Gnuplot Plotting  File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsLarynxEpiglottisHyper
    plasia\NTP_2002_Larynx Epiglottis Hyperplasia_LogLogistic_0.1.pit

       The form of the  probability function is:
       P[response]  = background+(1-background)/[1+EXP(-intercept-siope*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
    Interval
intercept
        1
                                     Parameter Estimates
                                                               95.0% Wald Confidence
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-27      DRAFT - DO NOT CITE OR QUOTE

-------
            Variable          Estimate        Std. Err.
     Limit
          background                 0             *
           intercept            3.3735             *
               slope                 1             *
     *  - Indicates that this  value is not calculated.
                                                        Lower  Conf.  Limit   Upper Conf.
            Model
          Full model
        Fitted model
       Reduced model

                AIC:
                         Analysis  of  Deviance Table
                   Log(likelihood)  # Param's  Deviance  Test  d.f.
                                                                           P-value
                        -99.8781
                         -101.66
                        -134.962
3.56345
70.1671
                               205.32

                                        Goodness  of   Fit
      ChiA2 = 3.65
                        d.f.  =  3
                                         P-value =  0.3022
        Benchmark Dose Computation
     Specified effect =             0.1
     Risk Type        =       Extra risk
     Confidence level =            0.95
                  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


          0.1


            0
                      Log-Logistic
             BMDL
                 BMD
      11:1009/162010
               0   0.01  0.02  0.03   0.04  0.05   0.06  0.07   0.08  0.09
                                  dose
 0.3126
<.0001

Dose
0.0000
0.0231
0.0470
0.0875

Est. Prob.
0.0000
0.4029
0.5781
0.7186

Expected
0.000
19.743
28.329
35.929

Observed
0.000
25.000
26.000
33.000

Size
50
49
49
50
Scaled
Residual
0.000
1.531
-0. 674
-0.921
1
2
4
5
B.3.10. Dataset 10:  Incidence of Nose Goblet Cell, Respiratory Epithelium, Hyperplasia 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

           This document is a draft for review purposes only and does not constitute Agency policy.
                                           B-28       DRAFT - DO NOT CITE OR QUOTE

-------
 1   vanadium pentoxide increased.  The Cochran-Armitage test confirmed the statistically significant
 2   trend with a Z-score = 3.46 and a one-sidedp-va\ue = 0.0003.
 3          Seven dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this
 4   dataset. A 10% extra risk of the incidence was used as the BMR to determine the POD. The
 5   goodness of fit, BMC, and BMCL for each model are summarized in Table B-15.
 6          All models demonstrated adequate goodness of fit (p-value > 0.1) and good visual fit.
 7   Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the Multistage
 8   (Stage 2) model was selected as the best-fitting model for the dataset.  Using this model, the
 9   BMCLio was 0.014 mg/m3 and this value was regarded as a candidate for the POD.
             Table B-15. Benchmark modeling results for incidence of nose goblet cell,
             respiratory epithelium, hyperplasia in female rats
Model3
Multistage6
(Stage 2)
Logistic
Probit
Weibull
(Quantal Linear)
Log-Logistic
Gamma
Log-Probit
Goodness of fit
P-Value
0.43
0.33
0.32
0.22
0.28
0.28
0.28
AICb
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
      "Selected (best-fitting) model shown in first row, in boldface type.
      bAIC = Akaike Information Criterion.
      °HEC = human equivalent concentration.
      dBMCL = the lower bound of BMC at 95% confidence level.
      Tor the Multistage model, up to three stages were tested. Only the model with the lowest AIC and lowest stage is
      reported here.

      Source:  NTP (2002).


10   BMDS output file

              Multistage  Model.  (Version:  3.2;  Date:  05/26/2010)
              Input  Data  File:
     C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsNoseGobletEpiHyperpla
     sia\NTP_2002_Nose Goblet Hyperplasia_Multi2_0.1.(d)
              Gnuplot  Plotting File:
     C:\USEPA\BMDS212\Data\V205\NTP_2002NonCancerFemaleRats\FemaleRatsNoseGobletEpiHyperpla
     sia\NTP_2002_Nose Goblet Hyperplasia_Multi2_0.1.pit

         The form of the probability function  is:
         P[response]  = background +  (1-background)*[1-EXP(
                       -betal*dose/xl-beta2*dose/x2) ]

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

-------
   The parameter betas are restricted to be positive

   Dependent variable = Response
   Independent variable = Dose

 Total number of observations = 4
 Total number of records with missing values = 0
 Total number of parameters in model = 3
 Total number of specified parameters = 0
 Degree of polynomial = 2

 Maximum number of iterations = 250
 Relative Function Convergence has been set to: le-008
 Parameter Convergence has been set to: le-008

                  Default Initial Parameter Values
                     Background =     0.270822
                        Beta(l) =            0
                        Beta(2) =      75.6009

           Asymptotic Correlation Matrix of Parameter Estimates
the user,
(  ***  The  model  parameter(s)   -Beta(l)
      have been  estimated at  a boundary point,  or have been specified by

      and  do  not appear in the correlation matrix )
             Background      Beta (2)
Background            1        -0.59
   Beta(2)        -0.59            1
Interval
       Variable
Limit
     Background
        Beta(l)
        Beta(2)
                                 Parameter Estimates
             Estimate         Std.  Err.

             0.276279             *
                    0             *
              71.2588             *
   95.0% Wald Confidence

Lower Conf. Limit   Upper Conf.
    Indicates that this value is not calculated.
                        Analysis of Deviance Table
       Model      Log(likelihood)  # Param's  Deviance  Test d.f.   P-value
     Full model        -126.318         4
   Fitted model        -127.169         2       1.70254      2          0.4269
  Reduced model        -133.292         1       13.9479      3        0.002977
           AIC:
                        258.338
Goodness of Fit

Dose
0.0000
0.0231
0.0470
0.0875

Est. Prob.
0.2763
0.3033
0.3815
0.5806

Expected
13.814
15.167
19.077
29.030

Observed
13.000
18.000
16.000
30.000

Size
50
50
50
50
Scaled
Residual
-0.257
0.872
-0.896
0.278
       = 1.71
                   d.f. = 2
                                   P-value = 0.4262
   Benchmark Dose Computation
Specified effect =            0.1
Risk Type        =      Extra risk
Confidence level =           0.95
           This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-30      DRAFT - DO NOT CITE OR QUOTE

-------
                   BMD =
                  BMDL =
                  BMDU =
0.0384521
0.0135524
0.0544996
     Taken  together,  (0.0135524,
     interval  for the BMD
    0.0544996) is a 90% two-sided  confidence
                        Multistage Model with 0.95 Confidence Level
           0.5
           0.4
           0.3
           0.2
                     Multistage
                   BMDL
                                 BMD
       16:4909/102010
                                 0.04  0.05
                                  dose
                                          0.06  0.07   0.08  0.09
 1   B.4.  REFERENCES
 2   NTP. (National Toxicology Program). (2002). NTP toxicology and carcinogensis studies of
 3          vanadium pentoxide (CAS No.; 1314-62-1) in F344/N rats and B6C3F1 mice (inhalation).
 4          Washington, DC.
 5   U.S. EPA. (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
 6          reference concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F).
 7          Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
 8          and Development, Office of Health and Environmental Assessment, Environmental
 9          Criteria and Assessment Office.
10          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993.
11   U.S. EPA. (U.S. Environmental Protection Agency). (2000). Benchmark dose technical guidance
12          document [external review draft]. (EPA/630/R-00/001). Washington, DC: U.S.
13          Environmental Protection Agency, Risk Assessment Forum.
14          http://www.epa.gov/raf/publications/benchmark-dose-doc-draft.htm.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-31       DRAFT - DO NOT CITE OR QUOTE

-------
 1         APPENDIX C. BENCHMARK CONCENTRATION MODELING OF COMBINED
 2                 LUNG ALVEOLAR AND BRONCHIOLAR TUMOR DATASETS
 3                                   FROM NTP STUDIES (2002)

 4          To derive the cancer slope factor for vanadium pentoxide, combined alveolar/bronchiolar
 5    adenoma and carcinoma datasets in B6C3Fi mice (Table C-2) from NTP (2002) were selected
 6    because of their biological and statistical significance.
 7          The highlights of the benchmark concentration (BMC) modeling results are the cancer
 8    slope factor derivation and the benchmark dose (BMD) modeling results.

 9    Cancer Slope Factor
10          The cancer slope factor was estimated as 3.4 per mg/m3 after linear extrapolation. The
11    selected model for each dataset, the candidate points of departure (PODs) and the calculation of
12    cancer slope factor are summarized in Table C-l.

13    Benchmark Dose Software (BMDS) Modeling
14          Each dataset was first fit with the recommended dichotomous model, the multistage-
15    cancer model, through EPA BMDS (version 2.1.2); if the goodness of fit test showed a
16   p-value < 0.05, other dichotomous models  available in EPA BMDS (version 2.1.2) were used; if
17    still no model showed adequate goodness of fit (p-value > 0.05), the highest dose  was dropped
18    from further modeling. Following the general model selection steps outlined in the draft
19    Benchmark Dose Technical Guidance (U.S. EPA, 2000), the best-fitting model was selected for
20    each dataset to estimate the candidate POD, which was the benchmark concentration lower limit
21    (BMCL) at the predetermined benchmark response (BMR).

22    C.I. BMR
23          Because all noncontrol concentrations showed essentially a plateau response, the tumor
24    datasets provided limited information about the concentration-response relationship. So, BMR
25    for each dataset was calculated based on the response at the control and the first noncontrol
26    concentration groups.

27   Multistage Weibull (MSW) Time-to-tumor Modeling
28          Because the survival curves reported by NTP (2002) showed a high percentage of deaths
29    (22-46%) at the end of the 2-year studies across the different concentration groups, the MSW
30    time-to-tumor model also  was used with these datasets. No adequate fit was obtained, however,
31    for either the male  or the female mouse dataset.
32
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-1      DRAFT - DO NOT CITE OR QUOTE

-------
            Table C-l. Summary of candidate PODs and cancer slope factors
Animal
Male B6C3Fi mouse
Female B6C3Fi mouse
BMR (extra
risk)3
0.71
0.67
HECb
BMC (mg/m3)
0.360
0.237
BMCLC (candidate POD,
mg/m3)
0.208
0.161
Cancer slope
factord
(per mg/m3)
3.4
4.2
     aBMR was calculated based on the response at the control concentration and the first non-control concentration
     (Table C-6).
     bHEC = human equivalent concentration.
     CBMCL = the lower bound of BMC at 95% confidence level.
     dCancer Slope Factor = BMR/BMCL, as linear extrapolation was used.
     Source:  NTP (2002).
1    C.2.  TREND TESTS BEFORE MODELING
2           Two inhalation carcinogenesis datasets were selected because of the biological and the
3    statistical significance reported by NTP (2002).  The Cochran-Armitage test was conducted to
4    confirm the significance of the statistical trend for the selected datasets (Table C-2) before
5    modeling.
            Table C-2. Trend tests on the inhalation carcinogenesis datasets in mice
            from the 2-year inhalation studies
Endpoint
Concentration as reported (mg/m3)3
0.0
1.0
2.0
4.0
Trend testb
Z -Score
^j-value
Incidence0 in male B6C3Fj mice
Lung alveolar/bronchiolar
adenoma and carcinoma
22/50
42/50
43/50
43/50
4.14
0.0001
Incidence0 in female B6C3Fj mice
Lung alveolar/bronchiolar
adenoma and carcinoma
1/48
32/47
35/48
32/49
5.19
O.0001
6
7
8
     ""Concentrations are as reported by NTP (2002).
     bOne-sided Cochran-Armitage trend test.
     Incidence = (number of animals with alveolar/bronchiolar adenoma or carcinoma)/(animal sample size). Only the
     animals alive at 52 weeks or when the first tumor appeared, whichever was earlier, were counted toward the sample
     size.
     Source: NTP (2002).

     C.3.   DOSE CONVERSION
            In the carcinogenesis studies reported by NTP (2002), mice (B6C3Fi) were exposed to
     vanadium pentoxide by inhalation. To analyze the concentration-response effect, the reported
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 C-2       DRAFT - DO NOT CITE OR QUOTE

-------
1    concentrations of vanadium pentoxide were converted to human equivalent concentrations
2    (HECs) before any modeling and extrapolation were completed.
3          First, the average lifetime animal body weights of rats were estimated based on the mean
4    body weight at different weeks (Table C-3). Then, following the EP'A''s Methods for Derivation
5    of Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA,
6    1994), regional deposited dose ratios (RDDRs) were calculated with the RDDR program
7    designed by U.S. EPA (1994) and summarized in Table C-4. Finally,  the HEC for each
8    concentration/sex group was calculated from the reported concentration in mice by multiplying
9    the continuous exposure adjustment factor and the pulmonary RDDR (Table C-5).
           Table C-3. Average lifetime animal body weight of mice for 2-year
           inhalation studies of vanadium pentoxide
Vanadium pentoxide
concentration as reported
(mg/m3)
Mean animal body weight" (g)
1-13 weeks
14-52 weeks
53-104 weeks
Average lifetime
animal body weight1"
(S)
Male Mice (B6C3FO
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 (B6C3FO
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
    aAs reported in Table 20-21 of the NTP report (2002).
    bAverage lifetime animal body weight = (mean body weight 1-13 weeks x 13 + mean body weight 14-52 weeks
    x 39 + mean body weight 53-104 weeks * 52)7104
    Source: NTP (2002).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-3      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
Concentration as reported
(mg/m3)
Average lifetime animal
body weight" (g)
Average
MMADb
Average GSDC
RDDRd
Pulmonary
Male mice (B6C3FO
0
1
2
4
48.9e
48.3e
46.9e
43.6
1.26
1.87
1.168e
1.168e
1.168e
1.134
Female mice (B6C3FO
0
1
2
4
\i.r
44.2
39.7
36.3
1.26
1.87
1.168e
1.143
1.077
1.023
aAverage lifetime animal body weight of each concentration/sex group was calculated in Table C-3.
VlMAD = mass median aerodynamic diameter; calculated based on the MMADs reported for 2-year studies (NTP.
2002).
°GSD = geometric standard deviation; calculated based on the GSDs reported for 2-year studies (NTP. 2002).
dRDDR = regional deposited dose ratio; calculated with the RDDR program from USA EPA (1994).
eBecause the average lifetime animal body weight was out of the mouse weight range (17-46g) accepted by the
RDDR program (V.2.3, US EPA), 46 g was used to calculate the corresponding RDDRs.
Source: NTP (2002).
        Table C-5.  Human equivalent concentrations of vanadium pentoxide in the 2-
        year inhalation studies
Concentration as reported"
(mg/m3)
Continuous exposure
adjustment factor1"
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 (B6C3FO
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
"Toxicology and carcinogenesis studies of vanadium pentoxide in F344/N rats and B6C3F} mice, NTP (2002).
bContinuous exposure adjustment factor = (6/24) x (5/7); animals exposed to vanadium pentoxide 6 h/d, 5 d/wk.
"Refer to Appendix Table C-4.
dHEC = human equivalent concentration = concentration as reported x continuous exposure adjustment factor x
RDDR.
Source: NTP (2002).
             This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-4       DRAFT - DO NOT CITE OR QUOTE

-------
 1    C.4.  BMR CALCULATION
 2          Because all noncontrol concentrations from the NTP carcinogenesis studies (2002)
 3    showed essentially a plateau response, the dataset provided limited information about the
 4    concentration-response relationship, given that the complete range of response from background
 5    to maximum must occur somewhere below the lowest dose. Thus, the BMD could be just below
 6    the first dose or it could be orders of magnitude lower. A BMR was therefore estimated based
 7    on the response at the control concentration, and the first noncontrol concentration was
 8    calculated (Table C-6) and then used to estimate the BMCL.
             Table C-6.  BMR estimation for male and female mice datasets
Animal
Male B6C3FJ Mice
Female B6C3FJ Mice
P(Control)a
0.44
0.02
P(lst non-control
concentration)1"
0.84
0.68
Calculated BMR (extra risk)c
0.71
0.67
      aThe combined alveolar/bronchiolar adenoma and carcinoma incidence in the control group.
      bThe combined alveolar/bronchiolar adenoma and carcinoma incidence in the first noncontrol concentration group.
      "Calculated BMR = [P(lst Noncontrol Concentration) - P(Control)]/[l - P(Control)].

 9    C.5.  BMDS MODELING FOR INHALATION CARCINOGENESIS DATASETS
10          Each dataset was first fit with the dichotomous multistage-cancer model provided in EPA
11    BMDS (version 2.1.2); if the goodness of fittest showed ap-va\ue < 0.05, other dichotomous
12    models available in EPA BMDS (version 2.1.2) were fit; if still no model showed adequate
13    goodness of fit (p-value > 0.05), the highest dose was dropped for further modeling.
14          Following the general model selection steps outlined in the draft Benchmark Dose
15    Technical Guidance (U.S. EPA, 2000), the best-fitting model was selected for each dataset to
16    estimate the candidate POD, which was the BMCL at the calculated BMR for each dataset.
17          The selected models and candidate PODs for all endpoints are summarized in Table C-l.

18    C.5.1.  Lung Tumor Dataset  1:  Combined Incidence of Alveolar/Bronchiolar Adenoma
19    and Carcinoma in Male Mice, NTP (2002)
20    Summary
21          The Cochran-Armitage  test confirmed the statistically significant trend with a
22    Z-score = 4.14 and a one-sided a/?-value < 0.0001.
23          Because all the noncontrol concentrations showed a similar effect, a BMR as 71% extra
24    risk was calculated and used to estimate the POD. The goodness of fit, BMC, and BMCL for
25    each model are summarized in  Table C-7 and Table C-8.
26          Because the primary cancer model (Multistage cancer) did not show an adequate fit, six
27    other dichotomous models available in EPA BMDS (version 2.1.2.)  were fit to this dataset.  Two
28    models demonstrated adequate  goodness of fit (p-value > 0.05) and good visual fit, but
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-5      DRAFT - DO NOT CITE OR QUOTE

-------
1   Log-Probit model failed to compute a reasonable BMCL. Based on the draft Benchmark Dose
2   Technical Guidance (U.S. EPA, 2000), the Log-Logistic model was selected as the best-fitting
3   model.  Using this model, the BMCL?! was 0.208 mg/m3.
4          Further BMD modeling was performed for comparison purposes.  Because the primary
5   cancer model (Multistage cancer) did not show adequate fit with all doses, the highest
6   concentration was dropped from further modeling.  After dropping the high dose, the Multistage-
7   cancer model demonstrated adequate goodness of fit (p- 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
Model3
Goodness of fit
/7-Value
AICb
Largest scaled
residual
Scaled residual of
interest
BMR
HEC
BMC
(mg/m3)
BMCLC
(mg/m3)
Primary cancer models
Multistage cancer
Stages 1,2, and 3
0.01
207.0
2.05
0.72
Extra
risk
71%
0.532
0.379
Other dichotomous models
Log-Logistic
Log-Probit
Gamma
Weibull
Logistic
Probit
0.19
0.88
0.01
0.00
0.00
200.63
2
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
Extra
Risk
71%
0.360
0.146
0.532
0.609
0.654
0.208
failed
0.379
0.447
0.495
     aSelected model is the best-fitting model for the dataset based on the draft Benchmark Dose Technical Guidance
     (2000).
     bAIC = Akaike Information Criterion.
     CBMCL = the lower bound of BMC at 95% confidence level.

     Source: NTP (2002).
8   BMDS output file

             Logistic Model.  (Version: 2.13; Date:  10/28/2009)
             Input Data File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002Cancer\MaleMiceTumor4DosesAllModel\NTP_2002_MaleMic
    eTumor4DosesAHModels_LogLogistic_0 . 71. (d)
             Gnuplot Plotting File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002Cancer\MaleMiceTumor4DosesAllModel\NTP_2002_MaleMic
    eTumor4DosesAHModels_LogLogistic_0 . 71. pit

        The form of  the  probability function is:
        P[response]  = background+(1-background)/[1+EXP(-intercept-siope*Log(dose))]

        Dependent variable  = Effect
        Independent  variable = Dose
        Slope parameter  is  restricted as  slope  >= 1
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-6      DRAFT - DO NOT CITE OR QUOTE

-------
   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,
background
 intercept
                 and do not appear in the correlation matrix )
background
         1
     -0.58
intercept
    -0.58
        1
                                 Parameter Estimates
Interval
       Variable         Estimate        Std. Err.
Limit
     background         0.451808            *
      intercept          1.91813            *
          slope                1            *
* - Indicates that this value is not calculated.
                                            95.0% Wald Confidence

                                         Lower Conf. Limit   Upper Conf.
       Model
     Full model
   Fitted model
  Reduced model
           AIC:
           Analysis of Deviance Table
     Log(likelihood)  # Param's  Deviance  Test d.f.   P-value
          -96.7763         4
          -98.3159         2       3.07913      2          0.2145
          -112.467         1       31.3814      3         <.0001
           200.632
Goodness of Fit

Dose
0.0000
0.2100
0.4200
0.8100

Est. Prob.
0.4518
0.7744
0.8580
0.9159

Expected
22.590
38.719
42.898
45.793

Observed
22.000
42.000
43.000
43.000

Size
50
50
50
50
Scaled
Residual
-0.168
1.110
0.041
-1.423
 ChiA2 = 3.29
                   d.f. = 2
                                   P-value = 0.1934
   Benchmark Dose Computation
Specified effect =           0.71
Risk Type        =      Extra risk
Confidence level =           0.95
             BMD =       0.359607
            BMDL =       0.208416
           This document is a draft for review purposes only and does not constitute Agency policy.
                                          C-7       DRAFT - DO NOT CITE OR QUOTE

-------
                   Log-Logistic Model with 0.95 Confidence Level
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
15:3709/202
1111 	 | .... ^ y ^'.^.'^ ' ' ' i 	 ' 	 ' 	 ' 	 ' 	 ,...._
___ 	 ^—
,-''"' j
- / \
_ / j
/ :

BMDL BMD ;
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
dose
010
       Table C-8. BMDS modeling results for combined incidence of
       alveolar/bronchiolar adenoma and carcinoma in male mice after dropping
       the highest concentration
Model3
Goodness of fit
/7-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
Log-Logistic
Gamma
Weibull
Probit
Logistic
Log-Probit
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
Selected model is the best-fitting model for the dataset based on the draft Benchmark Dose Technical Guidance
(2000).
bAIC = Akaike Information Criterion.
CBMCL = the lower bound of BMC at 95% confidence level.

Source: NTP (2002).
           This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-8       DRAFT - DO NOT CITE OR QUOTE

-------
1   BMDS output file
             Multistage Cancer Model. (Version: 1.9;  Date: 05/26/2010)
             Input Data File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002Cancer\MaleMiceTumor3DosesAllModel\NTP_2002_MaleMic
    eTumor3DosesAHModels_MultiCancl_0.71. (d)
             Gnuplot Plotting File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002Cancer\MaleMiceTumor3DosesAllModel\NTP_2002_MaleMic
    eTumor3DosesAHModels_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/xl) ]
       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
    Interval
           Variable         Estimate        Std. Err.
    Limit
         Background         0.461149             *
            Beta(l)          4.04221             *
    * - Indicates that this value is not calculated.
                                       95.0% Wald Confidence

                                    Lower Conf. Limit   Upper Conf.
           Model
         Full model
       Fitted model
      Reduced model
               AIC:
      Analysis of Deviance Table
Log(likelihood)   # 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
Dose
0.0000
0.2100
0.4200
Est. Prob.
0.4611
0.7694
0.9013
Goodness of Fit
Expected Observed Size
23.057
38.471
45.067
22.000
42.000
43.000
50
50
50
Scaled
Residual
-0.300
1.185
-0.980
     ChiA2 =2.45
                       d.f. = 1
                                       P-value =  0.1172
       Benchmark Dose Computation
    Specified effect =           0.71
    Risk Type        =      Extra risk
    Confidence level =           0.95
                 BMD =       0.306237
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-9       DRAFT - DO NOT CITE OR QUOTE

-------
                  BMDL  =        0.220508
                  BMDU  =        0.471328
     Taken  together,  (0.220508,  0.471328)  is a 90
     interval  for the  BMD
     Multistage  Cancer Slope  Factor =       3.21985
two-sided confidence
                          Multistage Cancer Model with 0.95 Confidence Level
      '•s
      £
            0.9
            0.8
            0.7
            0.6
            0.5
            0.4
            0.3
        17:3402/172011
                                 Multistage Cancer
                                Linear extrapolation
                       0.05
                                                               0.4
 1   C.5.2.  Lung Tumor Dataset 2: Combined Incidence of Alveolar/Bronchiolar Adenoma
 2   and Carcinoma in Female Mice, NTP (2002)
 3   Summary
 4          The Cochran-Armitage test confirmed the statistically significant trend with a
 5   Z-score = 5.19 and a one-sided/>-value < 0.0001.
 6          Because all noncontrol concentrations showed a similar effect, a BMR as 67% extra risk
 7   was calculated and used to determine the POD. The goodness of fit, BMC, and BMCL for each
 8   model are summarized in Table C-9 and Table C-10.
 9          Because the primary cancer model (Multistage cancer) did not show an adequate fit, six
10   other dichotomous models available in EPA BMDS (version 2.1.2.) were fit to this dataset. Only
11   one model  demonstrated adequate goodness of fit (p-value > 0.05); it failed, however, to
12   compute BMC and BMCL. So, the highest concentration was dropped from further modeling.
13          After dropping the highest dose, the Multistage-cancer model still did  not show an
14   adequate fit.  Six other dichotomous models available in EPA BMDS (version 2.1.2.) were tried.
15   Only the Log-Logistc model demonstrated adequate goodness of fit (p-value > 0.05) and good
16   visual fit. Based on the draft Benchmark Dose Technical Guidance (U.S. EPA, 2000), the
17   Log-Logistic model was selected as the best-fitting model. Using this model,  the BMCLe? was
18   0.161mg/m3.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-10      DRAFT - DO NOT CITE OR QUOTE

-------
       Table C-9.  BMDS modeling results for combined incidence of
       alveolar/bronchiolar adenomas and carcinomas with all four concentrations in
       female mice
Model a
Goodness of fit
^j-Value
AICb
Largest scaled
residual
Scaled residual of
interest
BMR
HEC
BMC
(mg/m3)
BMCLC
(mg/m3)
Primary cancer models
Multistage
cancer Stages
1,2, and 3
0.00
218.8
-3.90
1.33
Extra Risk
67%
0.435
0.359
Other dichotomous models
Log-Probit
Log-Logistic
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
aSelected model is the best-fitting model for the dataset based on draft Benchmark Dose Technical Guidance (2000)
bAIC = Akaike Information Criterion.
CBMCL = the lower bound of BMC at 95% confidence level.
Source: NTP (2002).
       Table C-10. BMDS modeling results for combined incidence of
       alveolar/bronchiolar adenoma and carcinoma in female mice after dropping
       the highest concentration
Model3
Goodness of fit
/7-Value
AICb
Largest scaled
residual
Scaled residual of
interest
BMR
HEC
BMC
(mg/m3)
BMCLC
(mg/m3)
Primary cancer models
Multistage
cancer Stages 1,2
and 3
0.06
132.2
1.42
1.42
Extra risk
67%
0.264
0.213
Other dichotomous models
Log-Logistic
Gamma
Weibull
Probit
Logistic
Log-Probit
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
Selected model is the best-fitting model for the dataset based on draft Benchmark Dose Technical Guidance (2000).
bAIC = Akaike Information Criterion.
CBMCL = the lower bound of BMC at 95% confidence level.
Source: NTP (2002).
            This document is a draft for review purposes only and does not constitute Agency policy.
                                            C-ll       DRAFT - DO NOT CITE OR QUOTE

-------
1   BMDS output file
             Logistic Model. (Version: 2.13; Date: 10/28/2009)
             Input Data File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002Cancer\FemaleMiceTumor3DosesAllModels\NTP_2002_Fema
    leMiceTumor3DosesAHModels_LogLogistic_0 . 67 . (d)
             Gnuplot Plotting File:
    C:\USEPA\BMDS212\Data\V205\NTP_2002Cancer\FemaleMiceTumor3DosesAllModels\NTP_2002_Fema
    leMiceTumor3DosesAHModels_LogLogistic_0. 67 .pit

       The form of the probability function  is:

       P[response] = background+(1-background)/[1+EXP(-intercept-siope*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
                                                                     Scaled
         Dose     Est._Prob.    Expected     Observed     Size       Residual

        0.0000     0.0211         1.011      1.000          48       -0.011


               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-12      DRAFT - DO NOT CITE OR QUOTE

-------
         0.2000
         0.3800
         ^2 =  0.80
0.6391
0.7698
    d.f.  = 1
30.036    32.000
36.952    35.000
      P-value = 0.3699
                                                             47
 0.596
-0.669
        Benchmark Dose Computation
     Specified effect =           0.67
     Risk Type        =      Extra  risk
     Confidence  level =           0.95
                  BMD =         0.23715
                 BMDL =        0.160575
           0.8
           0.6
       16:2909/21 2010
                       Log-Logistic Model with 0.95 Confidence Level
                      Log-Logistic
                             BMDL
                                        BMD
               0    0.05
                              0.15   0.2   0.25   0.3
                                  dose
 1   C.6.  EXTRAPOLATION METHOD AND INHALATION CANCER SLOPE FACTOR
 2          As explained in Chapter 5, linear extrapolation was applied in this assessment and BMCL
 3   at the calculated BMR was regarded as the POD. The inhalation cancer slope factor was the
 4   upper-bound estimation of risk and was calculated as BMR/BMCL. The inhalation slope factor
 5   was used to estimate the lifetime lung tumor risk in humans for vanadium pentoxide exposure by
 6   inhalation.  Data are summarized in Table C-l.

 7   C.7.  MSW TIME-TO-TUMOR MODELING OF THE INHALATION
 8   CARCINOGENESIS  DATASETS
 9          MSW time-to-tumor modeling also was used to model the datasets because the survival
10   curves reported by NTP (2002) showed a high percentage (22-46%) of deaths at the end of
11   2-year studies across the different concentration groups.  The individual animal data were
12   obtained through the NTP website and summarized according to concentration and week of
13   death (Table C-11).
14          Based on the results, the MSW time-to-tumor model was not recommended to derive
15   cancer slope factors for these datasets for two reasons:
16              •  The visual fit near the first noncontrol concentration was not adequate.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-13      DRAFT - DO NOT CITE OR QUOTE

-------
1             •  The parameters, including c, beta_0, and beta_l, were very close to the initial
2                values, which suggested the model was nonconvergent.

3   C.7.1.  Lung Tumor Dataset 1: Combined Incidence of Alveolar/Bronchiolar Adenomas
4   and Carcinomas in Male Mice, NTP (2002)
           Table C-ll. Grouped data for MSW time-to-tumor modeling; B6C3Fi male
           mice exposed to vanadium pentoxide by inhalation for 2 years
Human equivalent
concentration (mg/m3)
0.0
0.21
0.42
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
36
39
Response category for
alveolar/bronchiolar adenoma 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
c
c
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
1
1
              This document is a draft for review purposes only and does not constitute Agency policy.
                                            C-14      DRAFT - DO NOT CITE OR QUOTE

-------
       Table C-ll.  Grouped data for MSW time-to-tumor modeling; B6C3Fi male
       mice exposed to vanadium pentoxide by inhalation for 2 years
Human equivalent
concentration (mg/m3)

0.81
Week of death
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 or
carcinoma b
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
2
1
7
17
7
1
1
1
1
1
1
1
2
1
1
1
1
2
7
3
16
9
""Concentrations were the original doses reported in the publication from NTP (2002). No adjustment was applied.
bCategories 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 or adenoma, or both, were detected
when the subject was removed from the study due to scheduled sacrifice or unscheduled death.
           This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-l 5       DRAFT - DO NOT CITE OR QUOTE

-------
           Table C-ll. Grouped data for MSW time-to-tumor modeling; B6C3Fi male
           mice exposed to vanadium pentoxide by inhalation for 2 years
Human equivalent
concentration (mg/m3)
Week of death
Response category for
alveolar/bronchiolar adenoma or
carcinoma b
Number of
animals
1   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:  V205_NTP_MaleMiceHEC_Polyl.(d)

     V205_NTP_MaleMiceHEC

       The form of the probability function is:
       P[response] = l-EXP{-(t - t_0)Ac *
                      (beta_0+beta_l*doseAl) }

       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
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-16       DRAFT - DO NOT CITE OR QUOTE

-------
          beta 0
0.0111501
          beta 1
0.00720175

 0.0252605
0.00201449

0.00794841
0.00325342

0.00968191
0.0408391
NA - Indicates that this parameter has hit a
     bound implied by some  inequality constraint
     and  thus has no standard error.

                 Loq(likelihood)    # Param
   Fitted Model        -101.116          3
                                  AIC
                             208.233
Data Summary
CONTEXT

DOSE
0
0.21
0.42
0.81
C

28
8
7
7
F

0
0
0
0
I

22
42
43
43
U

0
0
0
0
Total

50
50
50
50
Expected

25.86
35.33
41.53
46.15
Response





   Benchmark Dose Computation
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
                  Dose = 0.00
             n  i   i   i
          0   20 40 60 80 100
                  Time
                              Incidental Risk: V2O5_NTP_MaleMiceHEC_Poly1
                                             Dose = 0.21
                                     0   20  40  60  80  100
                                             Time
              20 40
                      \   r
                      80 100
                  Time
                              .fr
                                        20  40
                                                    100
                                             Time
           This document is a draft for review purposes only and does not constitute Agency policy.
                                           C-17      DRAFT - DO NOT CITE OR QUOTE

-------
1   C.7.2.  Lung Tumor Dataset 2: Combined Incidence of Alveolar/Bronchiolar Adenoma
2   and Carcinoma in Female Mice, NTP (2002)
           Table C-12. Grouped data for MSW time-to-tumor modeling; B6C3F!
           female mice exposed to vanadium pentoxide by inhalation for 2 years
Human equivalent
concentration (mg/m3)
0.00
0.20
0.38
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
36
39
40
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
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
1
1
1
              This document is a draft for review purposes only and does not constitute Agency policy.
                                           C-l 8      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
Human equivalent
concentration (mg/m3)

0.73
Week of death
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
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
2
1
7
17
7
1
1
1
1
1
1
1
2
1
1
1
1
2
7
3
16
9
""Concentration was the original one reported in the publication from NTP (2002). No adjustment was applied.
bCategories 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 or adenoma, or both, were detected when
the subject was removed from the study due to scheduled sacrifice or unscheduled death.
Source: NTP (2002).
           This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-19       DRAFT - DO NOT CITE OR QUOTE

-------
1   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: V205_NTP_FemaleMiceHEC_Polyl.(d)

     V205_NTP_FemaleMiceHEC

       The form of the probability function is:
       P[response] =  l-EXP{-(t - t_0)Ac  *
                      (beta_0+beta_l*doseAl) }

       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 0
beta 1
c
1
-0.93
-1
beta 0
-0.93
1
0.93
beta 1
-1
0.93
1
                                     Parameter Estimates
                                                              95.0% Wald  Confidence

                                                          Lower  Conf.  Limit    Upper  Conf.

                                                                 -0.384483

                                                               -0.00146662

                                                               -0.0857143


                    Log(likelihood)   # Param             AIC
       Fitted Model        -109.298         3          224.597

                        Data Summary

               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-20      DRAFT - DO NOT CITE OR QUOTE
Interval
Variable
Estimate
Std. Err.
Limit

2

0

0

.6344

c

beta 0
1.12496

0.000224043
0.770137

0.000862599
.00191471

.114712
beta 1

0.0144991

0.0511302


-------
                          CONTEXT
                        F       I
                                        U  Total   Expected  Response
DOSE
0
0.2
0.38
0.73

49
18
15
18

0
0
0
0

1
32
35
32

0
0
0
0

50
50
50
50

1.94
19.95
30.09
42.08
   Benchmark Dose Computation
Risk  Response     =
Risk  Type
Specified effect  =
Confidence level  =
Time
              BMD  =
             BMDL  =
             BMDU  =
Incidental
     Extra
       0. 67
        0.9
        104
  0.411511
  0.339589
  0.516352
                             Incidental Risk: V2O5 MTP FemaleMiceHEC Po
                Dose= 0.00
                                              Dose = 0.20
        0   20  40 60  80  100

                Time
            \   I   I   I    I   T
            0   20  40 60  80  100

                    Time
                Dose= 0.38
                                              Dose = 0.73
           i   i   i    i   r
        0   20  40 60  80  100

                Time
            \   I   I   I    I   T
            0   20  40 60  80  100

                    Time
            This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-21       DRAFT - DO NOT CITE OR QUOTE

-------
 1   C.8.  REFERENCES
 2   NTP. (National Toxicology Program). (2002). NTP toxicology and carcinogensis studies of
 3          vanadium pentoxide (CAS No.; 1314-62-1) in F344/N rats and B6C3F1 mice (inhalation).
 4          Washington, DC.
 5   U.S. EPA. (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
 6          reference concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F).
 7          Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
 8          and Development, Office of Health and Environmental Assessment, Environmental
 9          Criteria and Assessment Office.
10          http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993.
11   U.S. EPA. (U.S. Environmental Protection Agency). (2000). Benchmark dose technical guidance
12          document [external review draft]. (EPA/630/R-00/001). Washington, DC: U.S.
13          Environmental Protection Agency, Risk Assessment Forum.
14          http://www.epa.gov/raf/publications/benchmark-dose-doc-draft.htm.

15
                This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-22       DRAFT - DO NOT CITE OR QUOTE

-------
 1                        APPENDIX D.  SUPPLEMENTAL INFORMATION

 2    D.I. EXPOSURE TO PM2.5 VANADIUM IN AMBIENT AIR
 3          Vanadium is a constituent of ambient PM generated by oil combustion.  Recent studies of
 4    the short-term health effects of PM and its constituents have found higher risks of mortality and
 5    hospitalization in locations with a higher fractional content of vanadium, nickel, and elemental
 6    carbon in PM (Bell et al.. 2009: Dominici et al.. 2007).
 7          Lippmann et al. (2006) evaluated the impact of average concentrations of 16 PM2.5
 8    components across 60 U.S. communities on the association between the daily change in PMio
 9    concentration and daily all-cause mortality risk in those communities. Community-specific
10    mortality risk estimates per daily change in PMio concentrations were obtained from the National
11    Morbidity, Mortality,  and Air Pollution Study (NMMAPS) database for 1987-1994.  Annual
12    average concentrations for PM constituents were obtained for 2000-2003 from the PM2.5
13    speciation network. The authors used weighted linear regression to evaluate for each chemical
14    constituent whether annual average concentration altered the association between PMio
15    concentration on the previous day and mortality risk. The constituents, nickel and vanadium,
16    were found to increase PMio mortality risk.
17          These results were reevaluated and extended by Domini ci and colleagues (2007) using
18    NMMAPS data for 90 communities from 1987 to 2000 and data on PM2.5 composition for 187
19    U.S. counties for 2000-2005. A total of 69 U.S. communities in the NMMAPS database also
20    had data on PM2.s composition and were included in this analysis.  Using a Bayesian hierarchical
21    regression model to estimate the  association between the 1-day lag PMio mortality risk and
22    average county-level PM constituent concentrations, counties with high average concentrations
23    of nickel and vanadium had higher PMio mortality risk with a 1-day lag.  When the three
24    counties that comprise the NMMAPS New York community were excluded from the analysis,
25    the effect of nickel and vanadium was much weaker and lost statistical significance. The authors
26    stated  that the three New York counties had nickel and vanadium concentrations that were 8.9
27    and 3.4 times higher than the other counties in the analysis.
28          Bell et al. (2009) used a Bayesian hierarchical regression model to evaluate the effect of
29    PM2.5  chemical constituents as percent of PM2.5 total mass on PM2.s-associated cardiovascular
30    and respiratory hospital admissions by county and season for 106 U.S. counties during 1999-
31    2005.  Counties with a population of 200,000 or greater for which data on PM and constituent
32    concentrations were available were selected.  Models were adjusted for day of the week,
33    seasonality, long-term trends using a smoothing function, daily temperature and dew point
34    temperature,  and temperature and dew point temperature for the previous three days.  County-
35    and season-specific PM2.5 relative risk for cardiovascular and respiratory hospital admissions
36    were higher in counties and seasons with a nickel, vanadium, or elemental carbon fraction of
37    total PM2.5 in the 75*  percentile compared to the 25  percentile. The effect of these PM
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-1       DRAFT - DO NOT CITE OR QUOTE

-------
 1    constituents was statistically significant. The average concentration of vanadium across the
 2    counties was 0.003 ug/m3 (range: 0.001-0.01) with an interquartile range of 0.001 ug/m3.  The
 3    interquartile range as percent of PIVb.s total mass was 0.01%. Each interquartile range increase
 4    in the fraction of PM2.5 total mass for vanadium was associated with a 27.5% (95% posterior
 5    interval:  10.6-44.4) increase in PM2.5-associated cardiovascular hospitalizations and a 392%
 6    (95% posterior interval: 46.3-738) increase in PM2.5-associated respiratory hospitalizations.
 7    Associations also were observed for elemental carbon and nickel, and effect estimates were not
 8    always stable in multipollutant models.
 9           The finding that communities with a higher fractional content of vanadium, nickel, and
10    elemental carbon in ambient PM have a higher risk of mortality and hospital admissions related
11    to daily change in PM2.5 concentration is intriguing:  It indicates the need for further research on
12    the contribution of fuel-oil combustion to regional and local air pollution, and the contribution of
13    specific metals, including possibly vanadium pentoxide, to elevated health risks. The time series
14    study design used in these investigations evaluates exposure-disease associations at the county
15    level and therefore, individual -level assessments of exposure and the impact of possible
16    confounders is not possible. Bell et al. (2009), however, investigated whether county-level
17    indicators of socioeconomic status, racial composition, and degree of urbanization could be
18    alternative explanations for the  observed effect modification by PM2.5 constituents and
1 9    concluded that this was not the  case.
20           Ambient concentrations of PM2.5 and PM2.5 fractions of nickel, vanadium, zinc, and
21    elemental carbon were evaluated in relation to respiratory symptoms among young Dominican
22    and African American children, aged 3-24 months, followed as part of a birth cohort study in
23    Northern Manhattan and the South Bronx in New York City between 1998 and 2007 (Patel et al..
24    2009).  Among 653 24-month-old children with questionnaire data (90% of total enrolled),
25    3 -month average ambient vanadium concentrations were associated with an increase in the
26    presence of wheeze during the cold and flu season (September 1-March 3 1). After adjusting for
27    elemental carbon, nitrogen dioxide, copper, and iron, an interquartile range increase
28    (0.003  ug/m3) in 3-month average vanadium concentrations was associated with a 3 1%
29    increased probability of wheeze during the cold season (p < 0.0003).  When not stratified by
30    season, an IQR increase in vanadium concentration was associated with a 14% increased
3 1    probability  of wheeze (p = 0.08).  When the highest 5% of vanadium concentrations were
32    excluded, however, the  association of wheeze with vanadium lost significance in the
33    multipollutant model. Vanadium concentrations were not associated with cough. Twenty-four-
34    hour average ambient concentrations of PM2.5 and PM2.5 fractions of nickel, vanadium, zinc, and
35    elemental carbon, measured every third day at two stationary sites in the Bronx, were  obtained
36    from the New York State Department of Environmental Conservation. Exposure levels were
37    assigned to  each subject by calculating 3 -month moving average concentrations of each pollutant
38    based on each follow-up questionnaire date and the previous 3 months. Exposures were assigned
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                D-2       DRAFT - DO NOT CITE OR QUOTE

-------
 1    to each subject's address using inverse-distance-weighted concentrations from the two stationary
 2    monitors. Associations were evaluated using generalized additive mixed effects models and a
 3    first-order autoregressive correlation structure to account for correlation in the up to 8 repeated
 4    observations for each individual. The models also adjusted for sex, ethnicity, postnatal ETS
 5    exposure, and a smoothed term for calendar time using natural cubic splines. Other pollutants
 6    also were associated with increased probability of wheeze (nickel) or cough (elemental carbon,
 7    nitrogen dioxide) during the cold/flu season or wheeze in other months (nitrogen dioxide). PM2.5
 8    mass concentrations were not related to an increase in probability of symptoms.
 9           The interpretation of the multipollutant models for vanadium is complicated because
10    other PM constituents are correlated with vanadium concentrations, resulting in less stable risk
11    estimates.  Nickel was not evaluated in the same model with vanadium for this reason. The
12    association with vanadium was independent of the association with nitrogen dioxide, a marker
13    for traffic emissions, and the authors suggested that oil combustion for space heating might
14    contribute to the observed respiratory symptoms in the very young children in this study.
15    Although vanadium cannot be singled out as the responsible agent for the probability of wheeze
16    observed in this study, the association is consistent with the respiratory symptoms observed
17    among boilermakers with exposure to high levels of residual oil fuel ash for periods of days to
18    weeks.
19           The effect of the metal content of ambient PM2.5 on lung function also was evaluated in
20    a time-series panel study of 29 patients with chronic obstructive lung disease, asthma, or
21    ischemic heart disease in Rome, Italy in the spring and winter of 1999 (Lagorio et al., 2006).
22    Outpatients of the Pneumology and Cardiology Departments of the Catholic University Hospital
23    in Rome who met eligibility requirements for chronic obstructive pulmonary disease (COPD)
24    (n = 11), ischemic heart disease (n  = 7), or asthma (n = 11),  and who lived in census tracts less
25    than 2 kilometers from one of six air monitoring stations were selected for the study. The
26    subjects volunteered to participate in repeated clinical examinations for two 1-month periods.
27    Pulmonary function testing was conducted according to American Thoracic Society guidelines,
28    and measures were  expressed as the percentage of predicted based on subject-specific age,
29    height, and weight.  Averages of 15, 24, and 9 observations were obtained from each participant
30    in the COPD, ischemic heart disease, and asthma panels, respectively. Daily average PM2.5
31    concentrations were calculated based on measurements obtained at two fixed site monitors set up
32    for the  study. PM content of cadmium, chromium, iron, nickel,  lead, platinum, vanadium, and
33    zinc was calculated as the ratio of the metal content in each PM sample to the air volume
34    collected during the sampling. The mean 24-hour PM2.5 concentrations during the spring and
35    winter of 1999 were 18.2 ± 5.0 ug/m3 and 36.7 ± 24.1 ug/m3, respectively. The mean 24-hour
36    vanadium concentrations during the spring and winter of 1999 were 2.4 ±1.6 ng/m3 and
37    1.1 ± 0.52 ug/m3, respectively. Vanadium was not associated with daily change in percent
38    predicted pulmonary function among subjects with COPD, ischemic heart disease, or asthma in
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                D-3       DRAFT - DO NOT CITE OR QUOTE

-------
 1    generalized estimating equation models. Models were adjusted for season (all), daily mean
 2    temperature (all), relative humidity (all), day of the week (COPD, ischemic heart disease), and B-
 3    2 agonist use (asthma).  Although the repeated measures design was a strength of the study, the
 4    number of subjects in each disease panel was small, and might have precluded detection of an
 5    association, given the very low vanadium concentrations analyzed.

 6    D.2.  LABORATORY ANIMAL AND IN VITRO STUDIES

 7    D.2.1.  Other Endpoint and Duration Studies

 8    D.2.1.1. Acute and Short-Term Studies

 9    D.2.1.1.1. Acute studies

10    D.2.1.1.1.1. Oral.  According to the Concise International Chemical Assessment Document 29
11    (WHO, 2001), rat oral LDso values for vanadium pentoxide range from 86 to 137 mg/kg body
12    weight [Yao et al., 1986 as cited in WHO (2001)1.  A rat oral LD50 value of 10 mg/kg body
13    weight was reported in IARC Monographs, volume 86 (2006) in a study by Lewis et al. (2000).
14    Lewis et al. (2000) also reported a mouse oral LDso value at 23 mg/kg body weight.  Clinical
15    signs of acute toxicity included lethargy, excessive tearing (lacrimation), and diarrhea but dose-
16    response data were not reported [Yao et al.,  1986 as cited in WHO (2001)1. Histopathological
17    analysis revealed liver necrosis and swelling of renal tubules.  Other studies (WHO, 2001) have
18    identified  LDsoS of-10 mg/kg body weight in rats and 23 mg/kg body weight in mice. An oral
19    LDso at 64 mg/kg body  weight in rabbits was established (WHO, 2001). Signs of toxicity in
20    rabbits mimicked those reported for rats.

21    D.2.1.1.1.2. Inhalation. A 1-hour inhalation exposure to vanadium pentoxide dust in rats led to
22    an LC67 of 1.44 mg/L (1440 mg/m3) [U.S. EPA, 1992 as cited in WHO (2001)1. Clinical signs of
23    toxicity included respiratory difficulty, irritation of mucosa, and diarrhea (WHO, 2001).  Knecht
24    et al. (1985) reported air flow restriction, as  measured by PFTs, in 16 adult male cynomolgus
25    monkeys (Macacafasicularis) exposed to vanadium pentoxide by whole-body inhalation at 5.0
26    mg/m3 for 6 hours but not at the lower dose tested (0.5 mg/m3). From this study, a LOAEL of
27    5.0 mg/m3 and a NOAEL of 0.5 mg/m3 were established. The lung was also identified as a target
28    organ in response to acute inhalation exposure to vanadium pentoxide.  Following a baseline
29    measurement of pulmonary function, each of 16 male cynomolgus monkeys was exposed to
30    aerosols of 0.5 mg/m3 vanadium pentoxide by whole-body inhalation for 6 hours (Knecht et al.,
31    1985). One week later, the monkeys were exposed to aerosols of 5 mg/m3 vanadium pentoxide
32    by whole-body inhalation for 6 hours. Effects on airway function were evaluated in monkeys by
33    comprehensive PFTs performed 24 hours postexposure to 0.5 and 5 mg/m3 and on pulmonary
34    inflammation by analysis of BAL fluid in monkeys performed after exposure to 5 mg/m3.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                              D-4      DRAFT - DO NOT CITE OR QUOTE

-------
 1    Significant changes in pulmonary function parameters compared to baseline values were
 2    observed only after exposure to 5 mg/m3 as follows:  16% increase in pulmonary resistance; 11%
 3    decrease in peak expiratory flow rate; 5-22% decreases in FEF maneuvers; 33% increase in RV;
 4    and 24% increase in forced residual capacity. Results are consistent with air-flow limitation in
 5    both small peripheral and large central airways. An increase (approximately 87%; data presented
 6    graphically) in the total number of cells  recovered in BAL fluid was observed 1 day after
 7    exposure to 5 mg/m3 vanadium pentoxide.  The increase in BAL fluid total cell number was due
 8    primarily to a marked increase (approximately 425%; data presented graphically) in the number
 9    of polymorphonuclear leukocytes. Results  suggest that pulmonary inflammation and release of
10    bronchoconstrictive mediators from inflammatory cells  could play a role in vanadium
11    pentoxide-induced air-flow restriction. An  acute (single 6-hour exposure) LOAEL for vanadium
12    pentoxide of 5 mg/m3 for pulmonary function in monkeys was established in this study, with a
13    NOAEL of 0.5 mg/m3.
14          A study in male CD-I mice (n = 48) by Avila-Costa et al. (2006) noted significantly
15    impaired performance on memory tasks, significantly decreased dendritic spine length, and
16    significant increases in percentages of necrotic cells in the hippocampus compared to controls
17    following a  1-hour inhalation exposure to 0.02 M vanadium pentoxide (2.5 mg/m3 as vanadium)
18    (p< 0.05). The dose-response relationship  for these effects could not be evaluated because only
19    one dose was tested, and a NO AEL could not be established.

20    D.2.1.1.2. Short-term studies

21    D.2.1.1.2.1. Inhalation and aspiration. The primary noncancer health effect of short-term
22    inhalation exposure in humans is respiratory irritation where 100 workers were reportedly
23    exposed to 0.05-5.3 mg/m3 vanadium for 10 hours per day,  6 days per week, for 4 weeks (Levy
24    et al., 1984). A LOAEL of 0.05 mg/m3 was established. Dose-response was not systematically
25    measured, however, no controls were used,  and exposure due to vanadium pentoxide could not
26    be directly correlated to effects. The primary noncancer health effects of short-term inhalation
27    exposure in animals include increased pulmonary inflammation, and dose-related decreases in
28    body weight and increases in relative lung weight in rodents (16-day exposure) (NTP, 2002).
29          Results of the NTP (2002) study in rats and mice provide evidence of toxicity to the
30    upper and lower respiratory tract, including increased lung weight, inflammation, nonneoplastic
31    lesions, and decreased pulmonary function during a 16-day inhalation exposure to vanadium
32    pentoxide. A significant increase in pulmonary inflammation and histiocytic infiltrate of
33    minimal to mild severity was observed in female rats (assessments not made in male rats)
34    exposed to vanadium pentoxide for 16 days, with a LOAEL of 1 mg/m3; a NO AEL was not
35    established.  Similar results were observed for female mice (assessments not made in male mice)
36    exposed for 16 days, with a LOAEL  of 2 mg/m3 for minimal to mild epithelial hyperplasia and
37    inflammation; a NO AEL was not established.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                D-5      DRAFT - DO NOT CITE OR QUOTE

-------
 1          Male rats (22 per group) and female mice (50 per group) were also assessed for
 2   pulmonary inflammation (BAL analysis) and systemic immunotoxicity (pulmonary bacteriocidal
 3   activity) following exposure to 0, 4, 8, and 16 mg/m3 vanadium pentoxide for 16 days (NTP,
 4   2002). Observed effects included significant alterations in the percentage of recoverable
 5   bronchial lavage cells (macrophages and neutrophils) (LOAEL of 8 mg/m3 and NOAEL of
 6   4 mg/m3), and increased lung protein and lysozyme in male rats (LOAEL of 4 mg/m3).  No
 7   NOAEL was established.  In female mice exposed to vanadium pentoxide for 16 days, a
 8   localized inflammatory response in the lungs occurred based on increase in lymphocytes, protein,
 9   and lysozymes at all concentrations; the NOAEL was not established.  A significant decrease in
10   the percentage of macrophages from BAL fluid led to a LOAEL of 8 mg/m3 and a NOAEL of
11   4 mg/m3.  These responses, NOAEL, and LOAELs are commensurate with those observed in the
12   3-month study in rats and mice of both genders (NTP, 2002). Thus, the lowest concentrations at
13   which adverse effects were observed were 1 and 2 mg/m3 (LOAEL) for lung inflammation in
14   female rats and  mice, respectively, exposed to vanadium pentoxide for 16 days.
15          An additional five male and five female mice were exposed by inhalation to vanadium
16   pentoxide for 6  hours per day, 5 days a week for 16 days at concentrations of 0, 2, 4, 8, 16, or
17   32 mg/m3 (NTP, 2002). All male mice exposed to 32 mg/m3 died before study completion.
18   Body weight was significantly decreased in male and female mice at 16 and 32 mg/m3,
19   respectively.  Absolute lung weights were significantly increased in a dose-dependent manner in
20   males at >4 mg/m3, and relative lung weight was significantly increased in males at>2 mg/m3.
21   Among females, both absolute and relative lung weights increased in all exposure groups
22   establishing the LOAEL of 2 mg/m3; no NOAEL was established.
23          Additional groups of 40-60 female mice were exposed to 0, 2, 4, or 8 mg/m3 for 6 hours
24   per day, 5 days  per week for 16 days (NTP, 2002). The nonneoplastic lung lesions noted on
25   days 6 and 13 consisted of hyperplasia of the alveolar and bronchiolar epithelium at all exposure
26   levels. Increase in severity of lesions was correlated with increasing concentration and time.
27   The LOAEL for nonneoplastic lung lesions was 2 mg/m3. A NOAEL was not established.
28          A duration-dependent decrease in the number of immunoreactive TH+ neurons (Avila-
29   Costa et al., 2004) after 4 weeks and increased quantities in metalloproteinase (MMP)-2 and
30   MMP-9 in CNS after 8 weeks in male mice (Colin-Barenque et al., 2008) following twice
31   weekly, 1-hour  inhalation exposure to 5.13 mg/m3 vanadium pentoxide.  These effects are
32   suggestive of disruption of the blood-brain barrier.
33          Turpin et al.  (2010) exposed male AKR mice to vanadium pentoxide by intranasal
34   aspiration following exposure to respiratory syncytial virus (RSV) to determine if preexposure to
35   the virus exacerbated the vanadium pentoxide-induced lung inflammation and fibrosis.  Animals
36   were  exposed intranasally to RSV (6 x 105 plaque-forming units [PFU] in 100 jiL PBS) on day 1
37   and day 8, then  exposed intranasally to vanadium  pentoxide (4 mg/kg in 50 uL PBS) on day 0
38   and day 7.  One hour before euthanasia, animals were given BrdU (50 mg/kg) by i.p. for
                This  document is a draft for review purposes only and does not constitute Agency policy.
                                              D-6      DRAFT - DO NOT CITE OR QUOTE

-------
 1    analyzing cell proliferation in bronchus-associated lymphoid tissue (BALT). Lungs were
 2    lavaged with PBS and BALF collected for analyzing differential cell counts (neutrophils,
 3    macrophages, and lymphocytes). Total RNA from lung was analyzed by real time RT-PCR for
 4    mRNA coding for profibrogenic growth factors TGF-|3-1, connective tissue growth factor
 5    (CTGF), platelet-derived growth factor-C (PDGF-C) and collagen CollA2, anti-fibrogenic type I
 6    interferon-alpha (IFN-a) and -beta (TFN-P), and IFN-inducible chemokines CXCL9 and
 7    CXCL10.
 8          Lung sections stained with Masson's trichrome staining to show collagen had an
 9    inflammation score of two, representing mild fibrosis with vanadium pentoxide exposure alone.
10    Vanadium pentoxide-induced fibrotic response, however, was less severe in the lungs of mice
11    that received either pre- or post-RSV exposure, and no differences were observed in pre- or
12    post-RSV exposure alone and the negative controls.  In addition, BALF from mice exposed to
13    vanadium pentoxide  alone or with RSV-post exposure had significantly higher total cell count
14    compared to RSV preexposure, RSV preexposure plus vanadium pentoxide, or controls. In
15    particular, vanadium pentoxide alone caused a significant increase in the levels of neutrophils
16    and lymphoctyes compared to controls. Both pre- and postexposure to  RSV led to a decrease in
17    the severity of vanadium pentoxide-induced fibrosis, and gene expression analysis showed
18    decreases in several profibrinogenic genes associated with innate immunity (Turpin et al.,  2010).
19          To analyze mucin production by the airway epithelium following exposure to vanadium
20    pentoxide, Yu et al. (2011) exposed female AKR mice (n = 5) by laryngeal aspiration to
21    vanadium pentoxide  (4 mg/kg bw) at two time points (day 1 and 7) and sacrificed on day 8.
22    Lung tissue analyzed by histopathology demonstrated an increase in the level of inflammation
23    and mucin production following exposure to vanadium pentoxide. These results were supported
24    by RT-PCR analysis demonstrating a statistically significant increase in the major airway mucin
25    (MucSac) (p< 0.05).
26          The  acute and short-term studies described here support those results described
27    previously in the subchronic and chronic studies and suggest progressive lung effects from
28    vanadium pentoxide  exposure similar to those observed in respiratory disease observed in
29    occupational studies. The studies further support that the lung is the most sensitive organ to
30    vanadium pentoxide  exposure.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-7      DRAFT - DO NOT CITE OR QUOTE

-------
 1    D.2.1.2. Immunological Endpoints

 2    D.2.1.2.1. Human studies. Study methods and results are described in Table 4-1. Some early
 3    case series observed dermatitis among affected workers employed at vanadium pentoxide
 4    processing facilities (Zenz etal., 1962; Sjoberg, 1951). Motolese et al. (1993) assessed whether
 5    exposure to vanadium pentoxide in the ceramics industry was associated with contact dermatitis
 6    or contact sensitization.  Testing for contact sensitization using skin patch testing was conducted
 7    on 126 enamellers and 64 decorators from 5 ceramics factories after exposure to a variety of
 8    substances, including vanadium pentoxide. Among the 190 workers under study, 22 individuals
 9    were found to have dermatitis and 17 reported having had skin lesions in the past. One worker
10    responded with a positive skin patch test indicating sensitization to a 10% solution of vanadium
11    pentoxide.
12          Kiviluoto published a series of reports regarding an investigation in 1975  of respiratory
13    symptoms and clinical findings among employees (process workers, repairmen, foremen, and a
14    laboratory worker) at a factory making vanadium pentoxide from magnetite ore (Kiviluoto et al.,
15    1981; Kiviluoto, 1980; Kiviluoto et al., 1979). A higher proportion of the exposed group
16    (n = 63) had an elevated number of neutrophils in nasal smears compared to the referent group
17    (n = 63) (35% versus 7%,p< 0.001, n = 55 pairs).  The referent group comprised individuals
18    who were employed at the magnetite ore mine and were matched to the exposed group members
19    by age and smoking habit. In biopsies of the nasal mucosa, a higher proportion in the exposed
20    groups had elevated plasma cells and round cells (26% versus 0%, p < 0.05, n = pairs and 48%
21    versus 29%, p < 0.05,  n = 56, respectively). The prevalence of elevated numbers of eosinophils
22    did not vary by exposure, leading the authors to conclude that the cytological and histological
23    response in the vanadium-exposed group was an irritant, not an allergic response. Vanadium
24    concentrations in the breathing zone averaged 0.028 mg/m3 (TWA) with a range of
25    0.002-0.42 mg/m3. Higher concentrations were found where grinding and packing of smelt were
26    conducted (TWA [range]:  2.3 mg/m3 [one sample] and 0.13 mg/m3 [0.02-0.37]).
27          A pilot study using nasal lavage to evaluate an inflammatory response to fuel-oil ash
28    exposure did not find an association between several exposure indices of vanadium or PMio
29    exposure and counts or percentages of polymorphonuclear cells, eosinophils, and epithelial cells
30    (Hauser et al., 1995).  Thirty-six of fifty volunteers (boilermakers and utility workers) with no
31    symptoms of cold or flu provided a nasal lavage sample both at baseline after at least 36 hours
32    away from work and after 72 hours of exposure at a local electric company. Daily exposure
33    estimates for PMio (<10 um) and vanadium (adjusted for filter extraction efficiency) were
34    assigned to each individual using data from personal air sampling (1- to 10-hour TWA) and a
35    self-completed work diary completed by each participant listing tasks and job locations during
36    the day.  Environmental  PMio concentrations based on personal sampling were 50-4,510 ug/m3.
37    Concentrations of respirable vanadium dust were 0.10-139.2 ug/m3.  Although boilermakers

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

-------
 1    (N=19) exhibited a larger increase in the number of polymorphonuclear cells per mL recovered
 2    from nasal fluid (adjusted by dividing the change by the mean of the baseline and postexposure
 3    value) and the adjusted number of epithelial cells/mL compared to utility workers, regression
 4    models, controlling for age and smoking, did not indicate an association with vanadium
 5    exposure.  Models did not control for ozone levels or respirator use. The wide variation in the
 6    change in cell counts after exposure, especially among nonsmokers, and the small number of
 7    subjects could have precluded the detection of an association with vanadium or particulate
 8    exposure.  Alternatively, the authors considered the levels of vanadium dust to be low, possibly
 9    not enough to cause inflammation in these workers.
10           Woodin et al. (1998) analyzed nasal lavage fluid from 18 boilermakers before, during,
11    and after the overhaul of a large, oil-fired boiler over a 6-week period from mid-May 1995 to late
12    June  1995. Biomarkers of upper airway inflammation in nasal lavage fluid, including
13    interleukin-8 and myeloperoxidase, were more prevalent among boilermakers during the
14    overhaul compared to 11 utility workers (Woodin et al., 1998). Mean IL-6 and eosinophilic
15    cationic protein levels did not change during the overhaul work, suggesting that the
16    inflammatory response was not due to an allergy or a respiratory infection. During the boiler
17    work, vanadium levels rose to a geometric mean SD of 8.9 (2.3) ug/m3 inside the boiler but did
18    not change appreciably outside the boiler where the utility workers were located (geometric
19    mean SD ug/m3:  1.4 [1.6] \p < 0.001]).  Vanadium concentrations in nasal lavage fluid,
20    however, were not associated with levels of either IL-8 or myeloperoxidase using Spearman's
21    Rank Order Correlation Test.
22           In conclusion, case studies of occupational  exposure to vanadium pentoxide dust or
23    ROFA have reported individuals with dermatitis, positive skin patch reactions, and bronchial
24    reactivity, although this occurrence appears to be uncommon. Increases in the numbers of
25    inflammatory cells in nasal smears or nasal lavage fluid have been observed in exposed workers
26    but no associations were observed in relation to estimates of respirable vanadium dust or
27    vanadium concentrations in nasal fluid.  The authors did not report increases in eosinophils,
28    suggesting that the inflammation was due to irritation and was not an allergic response.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                D-9       DRAFT - DO NOT CITE OR QUOTE

-------
 1    D.2.1.2.2. Animal studies. Pinon-Zarate et al. (2008) exposed 112 male CD1 mice to
 2    -1.4 mg/m3 vanadium pentoxide by inhalation for 1 hour per day, 2 times per day, for 12 weeks,
 3    as measured by filters following exposure.  Because this study did not provide reliable exposure
 4    information, exposure concentrations in mg/m3 could not be more specifically determined.
 5    Spleen weight and histology were determined. B-lymphocytes in the spleen were identified by
 6    immunohistochemical staining for CD 19 (a cell surface marker that acts as a co-receptor for
 7    other CD markers). In addition, eight control and eight vanadium pentoxide-treated mice were
 8    immunized with Hepatitis B surface antigen, a well-known T-cell-dependent antigen. Avidity to
 9    the resulting antibody was measured.  Spleen weight of vanadium pentoxide-exposed animals
10    increased significantly, peaking at 9 weeks, and then decreased significantly. Splenic germinal
11    centers significantly increased in vanadium pentoxide-treated mice and contained a significantly
12    increased number of CD19+ cells compared to controls.  The authors suggest that vanadium
13    pentoxide does not act as a direct antigen and does not induce a "host humoral response."  These
14    data suggest that vanadium pentoxide might affect the avidity of antibodies - the ability of
15    antibodies to bind effectively to substrates (affinity) and to engage multiple epitopes of the
16    antigen simultaneously.
17          Mravcova et al. (1993) conducted experiments to determine effects of subchronic
18    exposure to low doses of vanadium pentoxide on the immune system.  Weanling and  adult male
19    and female Wistar rats (n = 10 per group) were given vanadium pentoxide in drinking water (0,
20    1, 100 mg/L, or 0,  0.14,  14 mg/kg-day) for 6 months. In addition, male and female ICR mice
21    (groups of 10) were given vanadium pentoxide (0 or 6 mg/kg-day) by gavage 5 times per week
22    for 6 weeks. Immunotoxicity endpoints included spleen and thymus weight, spleen cellularity,
23    number of peripheral white blood cells, phagocytosis and natural killer cell  activity, the extent of
24    plaque-forming cell conversion to T-dependent antigen, and several cell-mediated immunity
25    endpoints (Concanavalin A [ConA], pokeweed mitogen responsiveness [PWM], and
26    phytohemagglutinin responsiveness [PHA] assays).  Of these endpoints, spleen weight was
27    significantly elevated over controls at 14 mg/kg-day in rats. The ConA and PHA assays
28    illustrated significant cell-mediated immune activation at 1 mg/L in rats over control (3 and 2.5
29    times higher than control values, respectively) but statistical significance was not indicated or
30    reported. At the high dose (14 mg/kg-day vanadium), ConA and PHA assay results were close to
31    control values.  Thus, no dose-response pattern was detected. The low-dose response could be a
32    transient or compensatory response. Results of other endpoints were not reported.  No
33    significant differences in these parameters were reported in mice. The authors suggested that the
34    high Con A response of T suppressor cells indicates that vanadium pentoxide might induce
35    hypersensitivity responses at low doses.
36          Immunological endpoints also were analyzed as part of a pulmonary study using weekly
37    provocation challenges cynomolgus monkeys (single 6-hour exposures to 0.5 or 3.0 mg/m3).
38    Inhaled vanadium pentoxide aerosol for 6 weeks produced statistically significant pulmonary
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-10      DRAFT - DO NOT CITE OR QUOTE

-------
 1    responses, prior to a subchronic exposure (6 hours per day, 5 days per week, for 26 weeks)
 2    (Knecht et al., 1992): study details are presented in Section 4.2.2.1).  Immunological analyses of
 3    blood and BAL fluid, and skin sensitivity tests were conducted before the pre- and postexposure
 4    provocation challenges. BAL fluid analyses were also performed 1 day after the provocation
 5    challenges.  Cytological endpoints included complete and differential blood cell counts and
 6    leukotriene €4 levels.  Immunological endpoints included total immunoglobulin E, total
 7    immunoglobulin G, albumin, and total protein.  The skin sensitivity tests assessed immediate and
 8    delayed responses to intradermal injections of vanadium pentoxide-monkey serum albumin
 9    conjugate. BAL fluid analysis showed a significant influx of inflammatory cells
10    (polymorphonuclear leukocytes) into the lung.  Other study endpoints were not significantly
11    different between the three exposure groups (control, peak, and constant) at either challenge
12    concentration when the monkeys were rechallenged following subchronic exposure.
13           In summary, immunological effects of vanadium pentoxide exposure have not been
14    comprehensively studied. Although the studies described here have not shown a statistically
15    significant response in  the endpoints tested, some effects were observed. These results are
16    therefore inconclusive.

17    D. 2.1.3.  Neurological Endpoints

18    D.2.1.3.1. Human studies.  No studies on the neurological effects of vanadium pentoxide were
19    reported in humans.

20    D.2.1.3.2. Animal studies.  Pazynich (1966) exposed 33 male albino rats (species not specified)
21    to 0.027 mg/m3 or 0.002 mg/m3 aerosolized vanadium pentoxide "round the clock" for 70 days.
22    A third group of rats served as control. The animals were evaluated for general  condition, body
23    weight, motor chromaxy of antagonistic muscles, whole-blood cholinesterase activity,  and
24    oxyhemoglobin content. Motor chromaxy of extensor muscles decreased significantly (p < 0.01)
25    while that of flexor muscles increased (p <  0.001) in animals exposed to 0.027 mg/m3.  Blood
26    cholinesterase levels were statistically significantly reduced after exposure to 0.027 mg/m3 and
27    the reduction persisted throughout the 90-day recovery period. A reduction in venous
28    oxyhemoglobin in rats  at 0.027 mg/m3was statistically significant. Recovery was observed after
29    20 days. No difference in these parameters was reported in the 0.002 mg/m3 group compared to
30    controls.  These results suggest a LOAEL of 0.027 mg/m3 for hematological and central nervous
31    system (CNS) effects in albino rats with a NOAEL of 0.002 mg/m3.
32           Three recent studies by Avila-Costa et al. (2006; 2005; 2004) found morphological
33    changes in CNS following inhalation exposure to vanadium pentoxide. Male CD-I mice
34    (n = 48) were exposed  to vanadium pentoxide by whole-body inhalation for 1 hour per day,
35    2 days per week for up to 8 weeks (Avila-Costa et al., 2005; Avila-Costa et al., 2004).  Particle
36    size was not reported in either study.  The exposure  concentration was reported as 0.02 M
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-11       DRAFT - DO NOT CITE OR QUOTE

-------
 1    (Avila-Costa et al.. 2006: Avila-Costa et al.. 2005: Avila-Costa et al.. 2004) or "2.5 mg/m3 V"
 2    (Avila-Costa et al., 2005). The same group of investigators (Gonzalez-Villalva et al., 2006:
 3    Mussali-Galante et al., 2005) using the same exposure protocol reported that the 0.02 M solution
 4    generated an average chamber concentration of 2.5 mg/m3, as vanadium metal (MW = 50.94),
 5    corresponding to 2.57 mg/m3 vanadium pentoxide (MW =181.9). The number of
 6    immunoreactive-TH+ neurons in the substantia nigra region of the basal ganglia in the
 7    mesencephalon (Avila-Costa et al., 2004) and the morphology of the blood-brain barrier (Avila-
 8    Costa et al., 2005) were assessed at the end of each week for up to 8 weeks of exposure. No
 9    clinical signs of toxicity were reported in either study. A duration-dependent decrease in the
10    number of immunoreactive-TH+ neurons was observed from week 3 (decrease of approximately
11    30%; data presented graphically) through week 8 (decreased by approximately 63%; data
12    presented graphically) of exposure (Avila-Costa et al., 2004). Morphological  changes to the
13    blood-brain barrier (cilia loss, cell sloughing, and ependymal cell layer detachment) were also
14    observed starting at week land increasing with duration of exposure (Avila-Costa et al., 2005).
15          Using a similar protocol, Avila-Costa et al. (2006) assessed the effects of vanadium
16    pentoxide on memory and morphology of hippocampal neurons in male  CD-I mice that were
17    exposed by whole-body inhalation for 1 hour per day, 2 days per week, for up to 4 weeks.
18    Groups of six exposed mice and six vehicle control mice (inhaling deionized water droplets)
19    were evaluated after 24 hours and  weekly for 4 weeks. No clinical signs or body weight changes
20    were observed. Spatial memory was tested using a modified Morris water maze task that was
21    learned preexposure. Performance on this test, as assessed by latency (swimming time) to locate
22    a hidden platform, was significantly  impaired in the exposed mice at all time points in an
23    increasing, time-related manner. Pyramidal neurons from the hippocampus CA1  region were
24    evaluated for cytological and ultrastructural changes because impairment in spatial memory is
25    frequently observed following damage to this region of the brain. The cytological analysis
26    assessed numbers of dendritic spines in the hippocampal cells.  Results showed a significant loss
27    of dendritic spines in the exposed mice at all time points, and the loss  increased with time, in a
28    manner that correlated with the memory impairments. The ultrastructural analysis showed a
29    significantly increased percentage of necrotic hippocampal cells at all time points, with a
30    maximum of 33% after 4 weeks of exposure; other findings included hyperdense postsynaptic
31    terminals and edema in mitochrondria, dendrites, dendritic spines, and presynaptic terminals.
32    These three studies establish a LOAEL for morphological changes to CNS accompanied by
33    behavioral effects following acute and short-term intermittent exposure to vanadium pentoxide at
34    concentrations of 2.56 mg/m3 two times per week, for 1-hour duration/exposure.
35          Colin-Barenque et al. (2008) investigated whether the vanadium pentoxide-mediated
36    disruption of the blood-brain barrier was associated with the activation of matrix
37    metalloproteinases (MMPs), protein degrading enzymes that are involved in tissue remodeling.
38    Male CD-I mice (n = 20 per group) were exposed to 0.02 M (2.56 mg/m3) aerosolized vanadium
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-12      DRAFT - DO NOT CITE OR QUOTE

-------
 1    pentoxide in deionized water or deionized water alone for 1 hour, two times a week, for up to 4
 2    weeks.  Five mice were sacrificed from each group, per time point (24 hours,  1, 2, and 4 weeks).
 3    The presence of MMP was determined by gel zymography. In the olfactory bulb, MMP-2 was
 4    the same in vanadium pentoxide-treated mice and controls, regardless of time point. MMP-9
 5    was significantly elevated (300% and -420%) in vanadium pentoxide-exposed mice compared to
 6    controls at 2 and 4 weeks, respectively. Both MMP-2 and MMP-9 were detected in the
 7    prefrontal cortex; MMP-2 was not different between controls and treated animals at any time
 8    point but MMP-9 was significantly elevated over control values at 1,2, and 4 weeks of exposure
 9    (150%, -175%, and 250% of control values, respectively).  In the hippocampus, MMP-9 from
10    vanadium pentoxide-treated mice was significantly elevated over controls after 1, 2, and 4 weeks
11    of exposure (200%, 340%, and -370%), while MMP-2 in exposed animals was significantly
12    increased at 4 weeks (-150% over controls).  In the striatum, MMP-9 from exposed mice was
13    significantly elevated over that of controls after 4 weeks of exposure; MMP-2 levels from
14    exposed mice were significantly elevated over control values after 2 (160%) and 4 weeks
15    (-250%) of exposure, but were not documented at earlier time points. These findings suggest
16    that vanadium-induced increases in MMPs in different parts of CNS occur in association with
17    dendritic spine loss and with neuronal death, and could be related to blood-brain barrier
18    disruption.

19    D.2.2.  Mechanistic Data and Other Studies in Support of the Mode of Action for
20    Pulmonary Fibrosis and Cancer
21           The preceding paragraphs highlighted the main noncancer and cancer health effects that
22    result from exposure to vanadium pentoxide.  Noncancer effects in the lung range from
23    histiocytic infiltration and alveolar inflammation to hyperplasia of alveolar epithelium and
24    pulmonary fibrosis.  These endpoints exist in a plausible biological response continuum—from
25    inflammation to reparative hyperplasia to fibrosis.  These effects also display a temporal and
26    dose-response continuum, ranging from inflammatory and hyperplastic responses that occur at
27    earlier time points (16 days) and at lower doses (2 mg/m3) to fibrosis that occurs at later time
28    points (3 months) and at higher doses (4 mg/m3). Inflammation and hyperplasia are biologically
29    relevant as precursor events to pulmonary fibrosis. Several investigators have systematically
30    investigated the molecular mechanisms underlying vanadium pentoxide-induced pulmonary
31    inflammation and fibrosis. These studies are  summarized below.

32    D.2.2.1. Genotoxicity
33           The genotoxicity assays of vanadium pentoxide are summarized in Table D-l.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-13      DRAFT - DO NOT CITE OR QUOTE

-------
 1    D.2.2.1.1. Human studies. Two studies investigated mutagenic activity in humans exposed to
 2    vanadium pentoxide (Ehrlich et al., 2008; Ivancsits et al., 2002). The in vivo genotoxicity of
 3    vanadium pentoxide in lymphocytes and whole-blood leukocytes obtained from 49 male workers
 4    exposed to vanadium pentoxide at a processing plant was compared to 12 nonexposed controls
 5    (Ivancsits et al., 2002). The average exposure duration for workers was 12.4 years. Workers
 6    reported using protective masks at least occasionally. Measurements or estimates of worker
 7    exposure to vanadium pentoxide were not reported, although exposure to vanadium was
 8    confirmed through serum and urine analyses. No significant differences between vanadium-
 9    exposed and control workers were observed for DNA strand breaks (as assessed by alkaline
10    comet assay), 8-hydroxy-2'deoxyguanosine (8-OHdG), an oxidized DNA base common
11    indicative of oxidative stress, or the frequency of sister chromatid exchange (SCE) in leukocytes.
12    When normal human leukocytes or human fibroblasts were cultured in vitro and exposed to
13    vanadate (25-500 |J,g/L), both whole blood cells and isolated nonproliferating lymphocytes
14    exhibited a significant increase in DNA migration in the alkaline comet assay compared to
15    nonexposed cells only at the highest doses tested (250500 ng/L). Cultured human fibroblasts,
16    however,  exhibited a dramatic dose-dependent increase in DNA migration after vanadate
17    treatment also at lower concentrations (as low as 0.5 ng/L) and suggest that fibroblasts are more
18    sensitive to DNA damage in the presence of vanadate than blood cells when exposed in vitro
19    (Ivancsits et al.. 2002).
20          Ehrlich et al. (2008) investigated the impact of inhaled vanadium pentoxide on DNA
21    stability in vanadium production workers (n = 52) compared to nonexposed jail wardens (n = 52)
22    during October 2004-May 2005.  All subjects studied were male and were exposed for their
23    entire 8-hour shift while wearing protective masks.  The duration of exposure and concentration
24    of the inhaled vanadium, however, was not determined. The median concentration (25th75th
25    percentile) of vanadium in plasma was sevenfold higher in exposed workers compared to the
26    unexposed reference group. Leukocytes were then assayed (Comet assay) for DNA damage, and
27    endogenous levels of oxidized purines  and pyrimidines were determined.  No differences in
28    DNA migration by exposure were noted in leukocytes under standard conditions, demonstrating
29    that exposure is not associated with increases in single- and double-strand breaks. Increases
30    were observed, however, in both oxidized purine (7% increase,/? = 0.02) and pyrimidine (33%
31    increase, p = 0.002) formation in exposed individuals.  Moreover, DNA damage induced by
32    bleomycin was 25% greater in leukocytes from the exposed workers (p < 0.0001), and DNA
33    repair after bleomycin administration was less evident (p < 0.0001). The extent of micronucleus
34    formation, necrosis, and apoptosis was determined in isolated lymphocytes using the cytokinesis-
35    block micronucleus cytome (CBMN Cyt) assay.  The number of micronuclei was 2.5-fold higher
36    in 24 workers than in 23 nonexposed referents (p= 0.01). The frequency of nucleoplasmic
37    bridges and nuclear buds (which indicate evidence of misrepaired DNA breaks and gene
38    amplification,  respectively) were significantly increased (sevenfold and threefold) over controls.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-14      DRAFT - DO NOT CITE OR QUOTE

-------
 1   Numbers of necrotic and apoptotic cells were 55% and 50% higher, respectively, in exposed
 2   workers. Together, these results suggest that occupational exposure to inhaled vanadium
 3   pentoxide could affect DNA stability by increasing levels of oxidized bases and affecting DNA
 4   repair. Age, body mass index, and smoking habit (number of cigarettes per day) were similar
 5   between the two groups. Folate levels also were similar but vitamin Be and 812  levels were
 6   lower among the unexposed, indicating that the effects on DNA in the exposed group were not
 7   due to vitamin deficits.
 8          Kim and colleagues (2004) assessed the cross-shift change in urine levels of 8-OHdG, a
 9   marker for DNA repair of oxidative DNA damage, over a 5-day period in 1999 among a group of
10   20 boilermakers involved in the overhaul of oil-fired boilers at a power plant (74% of eligible).
11   Median total 8-hour TWA PM2.5 concentration, measured using personal exposure monitoring,
12   was 0.44 mg/m3 (Q25%Q?5%: 0.290.76 mg/m3).  Total vanadium 8-hour TWA concentration,
13   including vanadium oxides, was 1.23 ug/m3 (Q25%Q75%:  0.473.53 ug/m3).  The workers were
14   18-59 years old (mean ± SD:  45.5 ± 12.0) and had been employed as boilermakers for 0.0440
15   years (mean ± SD:  21.7 ± 12.9). The mean cross-shift change in  creatinine adjusted 8-OHdG
16   levels in urine was 1.88 ug/g creatinine (SD = 0.74).  Pre-shift levels, measured  an average of
17   two days away from work, were significantly different from post-shift levels (p = 0.02). In linear
18   mixed regression models, a 1 mg/m3 increase in total PM2.5 8-hour TWA concentration was
19   associated with an increase in urinary 8-OHdG concentrations of 1.67 ug/g creatinine
20   (95% CI: 0.21, 3.14), adjusting for urinary cotinine, chronic bronchitis status, and age. A
21   1  ug/m3  increase in PM2.5 vanadium  concentration was associated with an increase in urinary 8-
22   OHdG concentrations of 0.23 ug/g creatinine (95% CI: 0.04, 0.42) in a model with the same
23   covariates. PM2.5 manganese, nickel and lead concentrations also were associated with 8-OHdG
24   levels in urine when analyzed separately in similar models.  The concentrations  of the metals
25   were  correlated (0.52 < r < 0.92) and so the association with vanadium might not have been
26   independent of the associations with the other metals.  The finding of oxidative DNA injury  and
27   repair among healthy boilermakers, however, is consistent with similar reports among vanadium
28   pentoxide workers.
29          Another marker of oxidative DNA damage, 7-hydro-8-oxo-2'-deoxyguanosine
30   (8-oxodG), was assessed in relation to water-soluble transition metal content in ambient PIVb.s
31   among male and female nonsmoking students, 20-33 years of age, living in central Copenhagen
32   (Sorensen et al., 2005).  Personal samples of PM2.5 were collected over 2 days twice in 1 year,
33   once during summer and once during autumn.  Median (interquartile range) concentrations of
34   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
35   median (interquartile range) concentration of vanadium in PM2.5 was 3.0 (0.3 - 4.7) in November
36   and 3.2 (1.4 - 5.7) in August.  Of 66 participating students, 32 provided measurements for both
37   seasons. Median (interquartile range) levels of 8-OxodG in  lymphocytes (per 105 dG) were
38   0.55 (0.34 - 0.78) and 0.58 (0.47 - 0.70) in November and August, respectively.  Vanadium and
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-15      DRAFT - DO NOT CITE OR QUOTE

-------
 1    chromium concentrations in aqueous suspensions of PM2.s were associated with the 8-oxodG
 2    concentration in lymphocytes in mixed regression models with subject as a random factor and
 3    adjusting for PM2.5 mass and season. A 1-ug/L increase in either vanadium or chromium was
 4    associated with a 1.9% (95% CI: 0.6, 3.3) or 2.2% (95% CI:  0.8, 3.5) increase in 8-oxodG
 5    concentrations in lymphocytes, respectively. Platinum, nickel, copper, and iron were not
 6    associated with 8-oxodG concentration in lymphocytes, and no metals were associated with
 7    8-oxodG concentration in urine. This study suggests that metal content of ambient fine
 8    particulate matter increases oxidative DNA damage in lymphocytes and that vanadium could be
 9    one of the responsible agents along with other metal constituents in particulate air pollution at
10    levels common in urban settings.
11          Three other studies demonstrated a genotoxic effect of vanadium pentoxide on primary
12    human lymphocytes in vitro (Ramirez et al., 1997; Rojas etal., 1996; Roldan and Altamirano,
13    1990). These studies examined chromosomal aberrations and aneuploidy by fluorescence in situ
14    hybridization (FISH) and SCE assays, as well as DNA damage by the comet assay. Cells
15    exposed to vanadium pentoxide had significantly increased DNA migration indicative of DNA
16    damage at all doses tested in primary human leukocytes (p < 0.05) and in the high doses in three
17    of four donor lymphocyte cell strains (p< 0.05). Vanadium pentoxide (00. uM) led to an increase
18    in aneuploidy with some interindividual variation observed in four primary human lymphocyte
19    cell strains (Ramirez et al., 1997).  An increase in aneuploidy was observed in one primary
20    human cell strain exposed to vanadium pentoxide (06ug/mL) but no chromosomal aberrations
21    were observed (Roldan and Altamirano, 1990).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-16      DRAFT - DO NOT CITE OR QUOTE

-------
 1    D.2.2.1.2. Laboratory in vivo and in vitro studies. Vanadium pentoxide produced gene
 2    mutations in two bacterial test systems (Bacillus subtilis and Escherichia coif) (Kada et al., 1980;
 3    Kanematsu et al., 1980): although negative results were reported by NTP (2002) in a reverse
 4    mutation assay in Salmonella typhimurium (TA97,TA98,TA100, TA102,TA1535 with or
 5    without metabolic activation). Negative results were also reported in a gene mutation assay in
 6    Chinese hamster V79 fibroblast cells (Zhongetal., 1994). DNA damage or aneuploidy,
 7    however, was observed in all in vitro studies performed in primary human cells (Kleinsasser et
 8    al.. 2003: Ivancsits et al.. 2002: Ramirez et al.. 1997: Roiasetal., 1996: Roldan and Altamirano,
 9    1990). Positive results were observed for DNA strand breaks in cultured human lymphocytes
10    (Rojas et al., 1996) at high doses  of vanadate (Ivancsits et al., 2002) and in cultured human
11    fibroblasts at lower, more environmentally relevant doses of vanadate (0.5 ng/L) (Ivancsits et al.,
12    2002). Positive results have been noted for aneuploidy (Ramirez et al., 1997) and polyploidy
13    (Roldan and Altamirano, 1990). Negative results were reported for chromosomal aberrations
14    (Roldan and Altamirano, 1990). In Chinese hamster V79 lung fibroblast cells, positive results
15    were observed for micronucleus formation (Zhongetal., 1994), altered mitosis (Zhong et al.,
16    1994), and cell transformation in  Syrian hamster embryo cells (Kerckaert et al., 1996) at
17    concentrations that were not cytotoxic. Negative results were reported for SCE and gene
18    mutations in vanadium pentoxide-treated Chinese hamster V79 fibroblast cells (Zhong et al.,
19    1994).
20          One study evaluated the genotoxicity of vanadium pentoxide in primary human cell
21    cultures. Kleinsasser et al. (2003) took mucosal biopsies from inferior nasal turbinates and blood
22    samples from 17 healthy volunteers. Isolated lymphocytes and mucosal cells were cultured and
23    exposed to 0, 0.06, 0.12, 0.24, and 0.47 mM vanadium pentoxide in vitro for 1 hour. Mucosal
24    cells and lymphocytes were assessed for DNA migration by the Comet assay. Extent of
25    migration was measured qualitatively (image analysis) and quantitatively ("Olive Tail Moment"
26    method). DNA migration was not significantly different in exposed human nasal mucosal cells
27    compared to controls. DNA migration appeared to increase dose-dependently, however, in
28    exposed human lymphocytes compared to controls (p = 0.001). Cytotoxicity was limited in both
29    cell types at all doses as measured by trypan blue exclusion assay. These results suggest that
30    human lymphocytes, but not nasal mucosal cells demonstrate genotoxic damage (single strand
31    breaks or alkali-labile damage) in response to vanadium pentoxide.
32          Experimental data in animals provide evidence of some types of genotoxicity following
33    in vivo exposure to vanadium pentoxide.  Vanadium pentoxide administered for 3  months by
34    inhalation to male and female mice (1, 2, 4, 8 or 16 mg/m3) did not increase the frequency of
35    micrenucleated normochromatic erythrocytes in peripheral blood (NTP, 2002). Additional
36    details of exposure are provided in the NTP (2002) study summary (see Section 4.2.1.2).
37    Genotoxicity was evaluated in male CD-I  mice following single intraperitoneal injections of
38    5.75, 11.5 or 23 mg/kg vanadium pentoxide (Altamirano-Lozano et al., 1996: Altamirano-
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-17      DRAFT - DO NOT CITE OR QUOTE

-------
 1    Lozanoet al., 1993). Exposure caused no treatment-related effects on mitotic index, average
 2    generational time, or SCEs in bone marrow cells (Altamirano-Lozano et al., 1993), although all
 3    doses induced DNA damage in testicular germ cells (Altamirano-Lozano et al., 1996).
 4    Altamirano-Lozano et al. (1999) assessed DNA damage in male CD1 mice 24 hours following
 5    single intraperitoneal injections of 0, 23.0,  11.5 or 5.75 mg/kg vanadium pentoxide
 6    (corresponding approximately to the LDso, l/2 LDso and l/4 LDso, respectively). Comet test
 7    results show the number of cells with DNA damage (primarily single strand breaks and alkali
 8    labile damage) was increased in liver, kidney, lung, spleen, and heart, although increases did not
 9    exhibit dose-dependence.  No evidence of DNA damage was observed in bone marrow.
10          In summary, the evidence for mutagenicity in humans is limited. Few studies examine
11    genotoxicity in humans in vivo, with equivocal results. Ivancsits et al. (2002) reported no
12    differences in DNA strand breaks, oxidative damage, or SCE frequency in leukocytes between
13    control and vanadium pentoxide-exposed workers. Ehrlich et al. (2008) noted changes in DNA
14    stability and DNA repair in leukocytes of occupationally exposed workers as compared to
15    controls.  Studies have  demonstrated a genotoxic effect of vanadium pentoxide on human cells in
16    vitro. Ivancsits et al. (2002)  demonstrated  significant increases in DNA damage as measured by
17    the Comet assay in both leukocytes and fibroblasts but with different dose sensitivity, while
18    Kleinsasser et al. (2003) noted DNA migration differences occurred  dose dependently in
19    peripheral blood lymphocytes but not in nasal mucosa. Earlier studies in human lymphocyte
20    cultures also demonstrated increased aneuploidy (Ramirez et al., 1997; Rojas etal., 1996) and
21    DNA damage (Roldan  and Altamirano, 1990) following exposure to vanadium pentoxide.  Thus,
22    vanadium pentoxide-induced mutagenicity could occur at doses  higher than those measured in
23    these occupational exposures, could be tissue specific, and could be associated with oxidative
24    stress rather than direct DNA damage.
25          In vitro tests in  bacterial and yeast systems provide mixed evidence of vanadium
26    pentoxide-induced mutagenicity. In general,  classic gene mutation assays were negative, as were
27    tests that assessed SCE and other chromosomal aberrations. DNA strand breaks (Ivancsits  et al.,
28    2002; Rojas et al., 1996) and micronucleus formation (Zhong et al., 1994) were indicated in
29    some studies in cultured cells but depended on cell type. Fibroblasts  appear to be more sensitive
30    to vanadium exposure in vitro than are blood cells. Similarly, experimental data from animal
31    studies is equivocal. NTP (2002) reported  that the frequency of micrenucleated normochromatic
32    erythrocytes in peripheral  blood was not increased in exposed compared to control mice. Several
33    studies by Altamirano-Lozano et al. (1999; 1996; 1993), however, have noted DNA damage in
34    specific target tissues in vanadium pentoxide-treated mice. It should be noted that Altamirano-
35    Lozano et al. (1993) consistently used intraperitoneal injection as the route of exposure for these
36    studies.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-18       DRAFT - DO NOT CITE OR QUOTE

-------
 1    D.2.2.2. Mechanisms of Inflammation and Fibrosis
 2          Increases in markers of pulmonary inflammation have been observed in the BAL fluid of
 3    rats and susceptible mice following intratracheal instillation exposure to vanadium pentoxide
 4    (Bonner et al., 2002; Pierce et al., 1996). These markers include macrophage inflammatory
 5    protein-2 (MTP-2), keratinocyte-derived chemokine (KC), interleukin-6 (IL-6), and IL-8.
 6    Further, the increased expression of prostaglandin-generating enzymes cyclooxygenases (COX)
 7    and prostaglandin E synthases have been associated with exposure to vanadium pentoxide
 8    (Bonner et al., 2002; Pierce et al., 1996), further suggesting increased inflammation.
 9          Pierce et al. (1996) investigated the ability of several vanadium compounds to increase
10    mRNA levels of cytokines in BAL fluid. Female CD rats received 42 or 420 |j,g of vanadium
11    pentoxide or phosphate-buffered saline (PBS) by intratracheal instillation. BAL fluid was
12    recovered 1 hour to 10 days after exposure. Significant neutrophil influx was observed after 24
13    hours of exposure to vanadium pentoxide and peaked at 48  hours postexposure with 20%
14    neutrophils. MIP-2 mRNA expression levels were significantly elevated in vanadium
15    pentoxide-treated rats compared to controls at early time points (1 hour to 48 hours), suggesting
16    pulmonary inflammation.
17          Proinflammatory prostaglandins such as PGE2 produced by the enzymes COX-1 and
18    COX-2, and PGE synthase mediate tissue homeostasis or are known to be associated with
19    various inflammatory diseases. Bonner et al. (2002) assessed the role of COX-1 and COX-2
20    enzymes in vanadium pentoxide-induced pulmonary inflammatory and fibrotic responses using
21    6- to 8-month-old male and female mice (of a hybrid C57BL/6J and 129/Ola genotype) that were
22    deficient in either COX-1 (COXl"'") or COX-2 (COX2~/~) enzyme.  These COX-deficient mice
23    and genotype-matched controls (wild type) were instilled with 50  (jL of saline (n = 3 or 4) or
24    1 mg/kg vanadium pentoxide in saline (n = 5 or 6).  Lungs were lavaged at 1, 3, 6, or 15 days
25    postinstillation and BAL fluid was collected and analyzed for tumor necrosis factor-alpha
26    (TNF-a) and prostanoids (e.g., PGE2) by ELISA. Lungs were removed and preserved for
27    histopathology, hydroxyproline assay, or COX immunoblotting. Histopathology showed marked
28    inflammation and increased injury in COX2"" mice compared to wild-type and COX1"" mice
29    3 days following vanadium pentoxide instillation. Hydroxyproline content was not different in
30    wild-type or COX1" ~ mice in response to vanadium pentoxide compared to saline-instilled
31    controls. Hydroxyproline content in vanadium pentoxide-exposed COX2"" mice was increased
32    twofold compared to saline-instilled COX2~/~, mice suggesting enhancement of lung fibrosis
33    following vanadium pentoxide exposure. PGE2 levels increased from about 500 pg/mL in
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               D-19      DRAFT - DO NOT CITE OR QUOTE

-------
Table D-l. 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 nonexposure
controls
52 exposed workers; 52
nonexposed workers
(jail wardens)

—
-

+
avg 12.4 y
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 micronucleus induction, nucleoplasmic bridges, and nuclear
bud formation. Necrosis and apoptosis levels were also increased in
exposed individuals.
Comet Assay
DNA damage
(8-OHdG)
Sister chromatid
exchange
Comet Assay
micronucleus
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)
Male CD-I mice
(n=4)
+
+
-
intraperitone
al injection,
sacrificed
24 h post
injection
intraperitone
al injection,
sacrificed
24 h post
injection
intraperitone
al injection,
sacrificed
24 h post
injection
0,5.75, 11.5,23
ug/gbw
0,5.75, 11.5,23
ug/gbw
0,5.75, 11.5,23
ug/gbw
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 (p < 0.05) dose-dependent increases were
observed.
Analysis of SCE demonstrated no effect of vanadium pentoxide
exposure in this study.
Comet assay
Comet assay
Cytogenetic
assay
Altamirano
-Lozano et
al. (1999)
Altamirano
-Lozano et
al. (1996)
Altamirano
-Lozano et
al. (19931
                                           D-20
DRAFT - DO NOT CITE OR QUOTE

-------
Table D-l. Genotoxicity data following exposure to vanadium pentoxide
Test System/Species
B6C3FJ mice (M, F)
B6C3FJ mice (M, F)
Results
+
+

Primary human
lymphocytes
Primary human whole
blood leukocytes
Cultured human
fibroblasts
Primary human
lymphocytes
Primary human nasal
mucosal cells
Primary human
lymphocytes
Primary human
lymphocytes (n=4)
Primary human
lymphocytes (n = 1)
Exposure
inhalation,
2yr
inhalation,
2yr
Dose
0, 1,2, or 4
mg/m3
0, 1,2, or 4
mg/m3
Effects
DNA was isolated from lung tumors 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 an effect of vanadium
pentoxide exposure.
Endpoint
Cytogenetic
assay
Micronucleus
assay
Reference
Devereux
etal.
(2002)
NTP
(2002)

In Vitro
+
+
+
+
—
+
+
-
+






24 h
72 h

25-500 ug/L
0-47 mM

0-0.1 uM
0.3, 30, 3,000
uM
0, 2, 4, 6 ug/mL

Vanadium pentoxide exposure led to a significant increase in DNA
migration as measured by Comet assay at the highest doses tested (250-
500 ug/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 led to an increase in aneuploidy with some
interindividual variation observed between the four primary cell strains.
Disruption of spindle formation might be due to interaction with
microtubules.
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 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.

Comet assay
Comet assay

FISH
Comet assay
SCE assay
Aneuploidy
Ivancsits et
al. (2002)
Kleinsasser
et al.
(2003)
Ramirez et
al. (1997)
Rojas et al.
(1996)
Roldan and
Altamirano
(1990)

                                           D-21
DRAFT - DO NOT CITE OR QUOTE

-------
Table D-l. Genotoxicity data following exposure to vanadium pentoxide
Test System/Species
Syrian hamster embryo
cells
Chinese hamster V79
cells
Results Exposure
+
+

-
0, 24 h, 7 d
24 h


Dose
0 to 0.875
ug/mL
0, 1,3,6,9,12
ug/mL


Effects
Vanadium pentoxide-exposed cells were positive at 7 d of exposure but
not at 24 h, similar to other tumor-promotion chemicals studied by this
group.
Vanadium pentoxide exposure led to increased micronucleus induction
(p < 0.005), apparently due to damage to the spindle apparatus but no
significant increases in mutations or SCE.
Endpoint
SHE
transformation
assay
Micronucleus
induction
SCE assay
HGPRT
mutation
Reference
Kerckaert
etal.
(1996)
Zhong et
al. (1994)
Bacterial Systems
Bacillus subtilis
Escherichia coli
Salmonella
typhimurium
Salmonella
typhimurium
(TA97,TA98,TA100,
TA102,TA1535withor
without metabolic
activation)
Positive
with and
without
activation
Positive
without
activation
(not tested
with
activation)
All strains
negative
with and
without
activation
(both
hamster
and rat S9
fractions)



48 h



0333 ug/plate
Study details not available.
Study details not available.

No increase in revertant colonies was observed following exposure.
Recombination
repair
Gene mutation

Gene mutation
Kada et al.
(1980)
Kanematsu
etal.
(1980)

NTP
(2002)

                                           D-22
DRAFT - DO NOT CITE OR QUOTE

-------
 1    saline-instilled wild-type mice to about 1,000 pg/mL in vanadium pentoxide-treated wildtype
 2    mice at 24 hours but not at other time points. The PGE2 level in saline-treated COX1"" mice was
 3    about 10 pg/mL and about 225 pg/mL after 24 hrs of vanadium pentoxide. PGE2 levels were
 4    about 200 pg/mL at 24 hours in saline-treated COX2~/~ mice and did not differ significantly from
 5    vanadium pentoxide-treated COX2"" mice regardless of time point.  This study suggests that
 6    vanadium pentoxide-induced inflammation could also be at least partially mediated by
 7    prostaglandins such as PGE2 generated by COX-2.
 8          Myofibroblasts, the principal proliferating cells that produce collagen, are involved in
 9    fibrogenic response of lungs following exposure to pulmonary irritants. Bonner et al. (2000)
10    observed that proliferating myofibroblasts were the principal cell type that contributed to the
11    observed fibrosis.  Male Sprague-Dawley rats weighing about 200 g received intratracheal
12    instillation of sterile saline or  1 mg/kg vanadium pentoxide. Rats were additionally injected with
13    BrdU (50 mg/kg, i.p.) 1 hour prior to sacrifice.  Sacrifice occurred at 3, 6, and 15 days after
14    vanadium pentoxide instillation. Excised lung tissue was assessed morphometrically and by
15    immunohistochemistry for vimentin and  desmin, two biomarkers for myofibroblasts and smooth
16    muscle cells, respectively.  Trichrome staining was used to assess collagen levels, an indicator of
17    the extent of fibrosis.  Vanadium pentoxide exposure induced thickening of the desmin-positive
18    bronchiolar smooth muscle cell layer by day 6 postexposure, which was identified as
19    myofibroblasts.  A 2.3-fold increase in airway smooth muscle cell nuclear profile, suggesting
20    increase in smooth muscle cell proliferation was due to hyperplasia. Serial sections of the
21    peribronchiolar region stained positive for vimentin and desmin, and were mainly
22    myofibroblasts.  The thickness of the subepithelial tri chrome-positive layer was 3.1- to 3.9-fold
23    higher at day 15 in vanadium  pentoxide vs. control samples.  The peak appearance of
24    peribronchiloar myofibroblasts occurred  at day 6 and declined by day  15. A thickened collagen
25    ring was apparent by day 15 in vanadium pentoxide-exposed samples compared to controls.
26          Rice et al. (1999) using both in vitro and in vivo models showed that myofibroblasts do
27    proliferate in response to vanadium pentoxide, and are dependent on platelet-derived growth
28    factor (PDGF) and epidermal  growth factor (EGF).  Rat lung myofibroblasts were isolated from
29    exposed male Sprague-Dawley rats, as stated in Bonner (1998) above, and were grown to
30    confluency. Cultures  were incubated for 24 hours with increasing concentrations of one of two
31    inhibitors of the PDGF-R (AG1296) and  EGF-R (AG1478), respectively, at a concentration of
32    100 |j,mol/L. Autophosphorlyation of PDGF-R and EGF-R in vitro was specifically blocked by
33    AG1296 and AG1478, respectively.  Tritiated [3H] thymidine uptake, a measure of mitogenesis,
34    was blocked by selective inhibition of PDGF- and EGF-receptors.  An in vivo study was carried
35    out at the same time.  Male Sprague-Dawley rats were treated with AG1296  or AG1478
36    (50 mg/kg) by intraperitoneal  injection 1  hour prior to intratracheal instillation of vanadium
37    pentoxide (1 mg/kg) and again 2 days after vanadium pentoxide was administered.  Rats were
38    sacrificed 3, 6, and 15 days after instillation and lungs were preserved for bromodeoxyuridine
                This document is a draft for review purposes only and does not constitute Agency policy.
                                           D-23         DRAFT - DO NOT CITE OR QUOTE

-------
 1    (BrdU) immunohistochemistry and hydroxyproline assays to measure DNA replication and cell
 2    division, respectively.  Quantitation of BrdU-labeled cells in the nuclei of rat lung tissue was
 3    significantly reduced in vanadium pentoxide-treated animals that had received injections of
 4    AG1296 and AG1478, compared to vanadium pentoxide-treated animals that were injected with
 5    vehicle alone. Vanadium pentoxide treatment induced a fivefold increase in lung hydroxyproline
 6    content, an indicator of lung collagen and potentially fibrosis, 15  days after instillation. Prior
 7    and posttreatment with AG1296 reduced hydroxyproline content  in vanadium pentoxide-treated
 8    animals to quantities similar to saline-instilled animals. Pre- and posttreatments with AG1478
 9    reduced hydroxyproline content by about 50%, but were still significantly higher than in saline-
10    instilled controls.
11           Two recent studies examined the mechanism of inflammation in mice (Rondini et al.,
12    2010; Turpin et al., 2010). Rondini et al. (2010) examined the effect of exposure to vanadium
13    pentoxide in three mouse strains of varying susceptibility to lung cancer (A/J, BALB/C, and
14    C57BL/6J) in an initiation/promotion model (full study description in Section 4.2.2.2).
15    Significantly higher transcriptional activity was observed for NFkB (A/J mice; peaked at 1 day
16    posttreatment) and AP-1 (A/J and B6 mice; peaked at 6 hours posttreatment) compared to PBS
17    controls in vanadium pentoxide-treated mice.  Overall, the differential inflammatory responses
18    observed in the three strains of mice appear to positively correlate with increased levels of
19    chemokines, such as keratinocyte-derived chemokine (KC) and monocyte chemotactic protein-1
20    (MCP-1), and increased binding of transcriptional factors NFicB and AP-1 (c-Fos), and sustained
21    activation of MAP kinases (MAPKs) and extracellular signal-regulated kinases 1 and 2 (ERK
22    1/2),  suggesting inflammation as a major response in mice.  Turpin et al. (2010) examined
23    pulmonary inflammation and fibrosis following intranasal aspiration exposure to vanadium
24    pentoxide with and without respiratory syncytial virus (RSV) exposure (full study description in
25    Section 4.4.1.2).  In this study, vanadium pentoxide exposure also caused a significant increase
26    in cell proliferation in the airways and lung parenchyma, lung mRNAs for TGF-|3-1, CTGF,
27    PDGF-C, CollA2, and mRNAs for IFN-a and -p and IFN-inducible chemokines CXCL9 and
28    CXCL10 compared to controls. Pre- or posttreatment with RSV caused a significant reduction in
29    the all mRNAs.  Together, results from this study showed that vanadium pentoxide induces
30    inflammatory and fibrogenic response in mouse lung and these effects were suppressed by RSV
31    infection.
32           To elucidate the potential cell signaling cascades associated with these endpoints, Antao-
33    Menezes et al. (2008) investigated the role of the signaling molecule, signal transducer, and
34    activator of transcription (ST AT)-1 in vanadium pentoxide-induced pulmonary fibrosis. Their
35    work identified another inflammatory molecule,  interferon-beta (IFN-|3), as a mediator of
36    vanadium pentoxide-induced STAT-1 activation in normal human lung  fibroblasts. Briefly,
37    confluent, quiescent cultures of human lung fibroblasts were either placed in serum-free defined
                                                         r\
38    medium (SFDM) or SFDM supplemented with 10 |j,g/cm vanadium pentoxide. Neutralizing
                This document is a draft for review purposes only and does not constitute Agency policy.
                                            D-24          DRAFT - DO NOT CITE OR QUOTE

-------
 1    anti-IFN-a and anti-IFN-p antibodies were used to quench activity of IFN-a and IFN-|3 and the
 2    ratio of phosphor-STAT-1 to STAT-1 was then measured by quantitative RT-PCR or Western
 3    blot.  Vanadium pentoxide-induced STAT-1 activation could be inhibited by a broad spectrum
 4    NADPH inhibitor at 24 hour.  Vanadium pentoxide induced significant IFN-|3  expression after
 5    18 and 24 hours. This effect was nearly completely abolished by addition of the NAPDH
 6    inhibitor. Activation of STAT-1 (as measured by a ratio of phosphor-STATl to total STAT-1)
 7    was significantly decreased by addition of neutralizing IFN-|3 antibodies and also by addition of
 8    a Janus Associated Kinase (JAK) inhibitor. In summary, STAT-1 was activated in response to
 9    vanadium pentoxide exposure, and was linked to IFN-|3 as its primary mediator. This study
10    identifies a putative signaling pathway leading to expression of genes that control proliferation of
11    myofibroblasts.
12          Ingram et al. (2007) performed gene array analysis to determine a list of candidate genes
13    altered by exposure to vanadium pentoxide. Normal human lung fibroblasts were exposed to
              r\
14    10 |j,g/cm vanadium pentoxide or saline in vitro. RNA from cells was harvested 1, 4 8, 12, and
15    24 hours posttreatment. Labeled cRNA hybridized to the Affymetrix Human Genome Array
16    U133A 2.0 gene chip was used to assess gene expression at various time points up to 24 hours of
17    exposure. About 300 genes were found to be upregulated in response to vanadium pentoxide
18    including inflammatory and immunomodulatory genes.  More than 1,000 genes were
19    downregulated in response to vanadium pentoxide, including genes from the ubiquitin cycle and
20    cell cycle genes. A dozen genes were confirmed by RT-PCR and included growth factors
21    (heparin-binding EOF-like growth factor [HB-EGF], vascular endothelial cell growth factor
22    [VEGF], and connective tissue growth factor [CTGF]); chemokines (IL-8, CXCL9, CXCL10);
23    oxidative response genes  (superoxide dismutase [SODJ2, pipecolic acid oxidase [PIPOX],
24    oxidative stress response  [OXR]1);  and DNA-binding proteins (growth arrest specific [GAS]1,
25    STAT1). The gene array analysis thus confirms that several mitogens, growth factors,
26    chemokines, cytokines, oxidative response genes, and DNA-binding proteins are all  critical to
27    the formation of fibroproliferative lesions in response to vanadium pentoxide exposure in vitro
28    (Figure D-l).
29          Together, these results indicate that proliferating myofibroblasts are the primary cell type
30    associated with vanadium pentoxide-induced pulmonary fibrosis and that cellular proliferation
31    depends  on activated mitogens such as PDGF and EGF, in addition to HB-EGF. The STAT-1
32    and MAPKinase pathways might play key roles in this process. Moreover, fibroproliferative
33    lesions contain collagen.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                           D-25          DRAFT - DO NOT CITE OR QUOTE

-------
                                    V2O5 Exposure
                                           *




fSOD2
1 PIDPIY






|OXR1

H O production
reduced
protection agair
oxidative dama

oxidative

&&H




A/FK-B





t II -8



ge




SotaL

injury ]









P
I ^
inflciiTirnatiui






STAT1






fCXCL9
I CXCL10








infiltration
_J
,.:-.





| MAPH


I


frHBEGF


airway epithelial
cell growth

suppiession of
fibroblast growth


•I

I
.,,;':4'-'^; -
repair



r
fibroblast
proliferation








fcTGF



IE

--ar
1
IVEG


collagen
deposition
1






|GAS1




loss 01 growm
arrest control

endothelial cell
proliferation
I
S£ ~*

^•MH
	 *t

\ / * ,' . -,,«.
fibrosis




K •

'''i'-'l^,'



"•'•
Stress
   *
Signaling
   I
                                                                                    I
                                                                                   Cell
                                                                                Responses
                                                                                    I
                                                                                 Pathologic
                                                                               Consequence
 1                       Figure D-l. Genomics of ViOs-induced bronchitis.
 2          Note: Reprinted with permission from Ingram et al. (2007) Respir. Res. Apr 25; 8(1):34.
 3    D.2.2.3. Mechanisms ofHyperplasia and Carcinogenicity
 4          Molecular events underlying the mechanism of reparative hyperplasia and carcinogenesis
 5    have been documented.  Bonner et al. (1998) exposed male Sprague-Dawley rats to vanadium
 6    pentoxide or sterile saline by intratracheal instillation at 1 mg/kg. Animals were sacrificed at 3,
 7    6 and 15 days after instillation. Lungs were preserved and analyzed for PDGFR-a by
 8    immunohistochemical analysis and morphometry for fibroproliferative lesions.  Smooth muscle
 9    thickening was observed beneath ciliated epithelial cells, as indicated by increased desmin
10    localization, on day 6 in exposed samples.  Trichrome staining revealed increased collagen
11    deposition around bronchioles in vanadium-pentoxide-exposed samples.  The thickness of the
12    subepithelial layer increased by 3.1- to 3.9-fold at day 15 after instillation of vanadium
13    pentoxide, determined by morphometric techniques.
14          Platelet-derived growth factor (PDGF) is a mitogen and chemoattractant for fibroblasts.
15    Male Sprague-Dawley rats were intratracheally instilled with sterile saline or 2 mg/kg of
16    vanadium pentoxide (Bonner et al., 1998).  Tissues were harvested at 24, 48, and 72 hours
17    postexposure and at 6 and 15 days.  Pulmonary myofibroblasts and alveolar macrophages were
18    isolated. Isolated total lung RNA was assessed for quantitation of PDGF by Northern blot
19    analysis. PDGF-receptor alpha mRNA and protein expression were significantly elevated in
20    vanadium pentoxide-exposed animals compared to controls at 24 and 48 hours.  PDGF-receptor
21    beta was not significantly elevated over controls at any time points.  Confluent cultures of lung
                This document is a draft for review purposes only and does not constitute Agency policy.
                                            D-26          DRAFT - DO NOT CITE OR QUOTE

-------
 1    myofibroblasts were stimulated with vanadium pentoxide. Similarly, cultures of lung
 2    macrophages were stimulated with vanadium pentoxide.  Levels of PDGF in myofibroblasts
 3    were not affected by direct stimulation by vanadium pentoxide, but were affected by a factor
 4    released by vanadium pentoxide-stimulated macrophages, and were associated with interleukin
 5    (IL)-IB, an inflammatory cytokine. Together, these results suggest that hyperplasia occurs
 6    under the control of inflammatory mediatiors such as IL-1B that can recruit mitogens such as
 7    PDGF which can stimulate growth of myofibroblasts, the main cell type involved in the
 8    development of fibrotic lesions.
 9          Zhang et al. (200 la) investigated the ability of vanadium pentoxide to induce
10    heparin-binding  epidermal growth factor-like growth factor (Fffi-EGF) (another mitogen) in
11    vitro, using normal human bronchial epithelial cells (NFffiECs). Mature cultures of NHBECs
12    were incubated with vanadium pentoxide at 0, 1, 10 and 50 |j,g/cm3 for 3 hours. Total RNA was
13    then isolated and RT-PCR was used to quantitate Fffi-EGF. HB-EGF mRNA was significantly
14    and dose-dependently increased in response to vanadium pentoxide compared to controls. In a
15    second time course study (using only the high dose 50 |j,g/cm3 of vanadium pentoxide), the peak
16    of Fffi-EGF induction occurred after 3 hours of exposure and persisted until 8 hours of exposure.
17          In a follow-up study, Ingram et al. (2003) similarly showed a peak induction of HB-EGF
18    in quiescent cultured human lung fibroblasts exposed to 10 |j,g/cm2 vanadium pentoxide at
19    3 hours.  Quiescent cultured human lung fibroblasts were exposed for 3 hours to 0, 10, 30 or
               9                      	
20    100 |j,g/cm vanadium pentoxide.  HB-EGF RNA was isolated and detected by Northern blot
21    analysis. HB-EGF was significantly elevated in a dose-dependent manner in vanadium
22    pentoxide-exposed fibroblasts compared to controls. Stimulating human lung fibroblasts with
23    H2O2 (10 (jM) similarly induced HB-EGF, with a peak mRNA expression at 1  hour
24    postexposure. HB-EGF protein expression peaked at 6 hours postexposure, as measured by
25    Western blot analysis. Quiescent human lung fibroblasts stimulated with 10 |j,g/cm2 vanadium
26    pentoxide were found to initially quench spontaneous H2O2 production (at early time points) and
27    then boost H2O2 production at a peak of 12 hours postexposure.  To assess the role of
28    extracellular signal-regulated protein kinase (ERK), and the p38 subunit of mitogen-activated
29    protein (MAP) kinase in vanadium pentoxide-induced HB-EGF production, Ingram et al. (2003)
                                                                                           r\
30    exposed quiescent  confluent human lung fibroblast cells in culture to 500 (jM H2O2 or 10 |j,g/cm
31    vanadium pentoxide for 15 minutes, 30 minutes, 1  hours, 3 hours,  6 hours, or 24 hours.  Cell
32    lysates were collected and Western blot was performed for the phosphorylated form of ERK or
33    total ERK protein or for the phosphorylated p38 subunit of MAP kinase and total p38. Peak
34    phosphorylated ERK occurred at 30 minutes postexposure to H2O2. Vanadium-induced
35    increases to p-ERK were biphasic, with one peak at 30  minutes postexposure and the next at 24
36    hours postexposure. Phosphorylated p38 was similarly maximally elevated at 30 minutes
37    posttreatment with H2O2 and 24 hours posttreatment with vanadium pentoxide.  Thus, this study

                This document is a  draft for review purposes only and does not constitute Agency policy.
                                           D-27          DRAFT - DO NOT CITE OR QUOTE

-------
 1    suggests that HB-EGF expression occurs as a result of activation of the MAP kinase and ERK
 2    pathways.
 3          NTP (2002) and Ress et al. (2003) concluded that exposure to vanadium pentoxide
 4    caused alveolar and bronchiolar adenomas and carcinomas in male and female mice and some
 5    evidence suggests carcinogenicity in male rats, based on observations of alveolar and bronchiolar
 6    neoplasms in groups exposed to vanadium pentoxide that exceeded historical controls.  The body
 7    of evidence that has investigated the MOA underlying cancer effects due to exposure of
 8    vanadium pentoxide is not as well characterized as mechanisms underlying noncancer fibrotic
 9    effects.  Loss of heterozygosity (LOH) and DNA damage, however, has been documented.
10          Using mouse lung tumor tissues from the NTP (2002) chronic inhalation study from mice
11    exposed to 0, 1, 2 or 4 mg/m3 vanadium  pentoxide, Zhang et al. (200Ib), observed LOH on
12    chromosome 6 (in the region of the K-ras gene) in 17 of 19 vanadium pentoxide-induced mouse
13    tumor samples. Moreover, 29 of 40 (73%) vanadium pentoxide-induced murine
14    adenocarcinomas from the NTP (2002) study had mutations in  K-ras2.  The K-ras2 mutations
15    typically were the result of either a GA-> AT transition or a GA-^TA transversion in the second
16    base of codon 12 (Zhang et al., 2001b).  To determine the effect of the K-ras mutations and LOH
17    on activated MAP kinase, Devereux et al. (2002) used tissues from the NTP (2002) 2-year
18    carcinogenicity study in female and male B6C3F1 mice to isolate protein from 17 vanadium
19    pentoxide-induced alveolar and bronchiolar carcinomas, 1 spontaneous carcinoma, and 2 normal
20    (untreated) lung tissue samples.  Levels of total MAP kinase and activated MAP kinase were
21    assessed by probing isolated lung protein for total and phosphorylated MAP kinase with
22    anti-phospho-MAP kinase antibody in all samples.  Only qualitative analysis was reported.  Total
23    MAP kinase was not different between normal tissue and lung tumor tissue.  Activated MAP
24    kinsase was elevated in five of six tumors that had both LOH and K-ras mutations, and was
25    barely detectable in all seven tumors examined where no K-ras mutations were detected.  Four of
26    five tumors that had K-ras mutations but were not positive for LOH had elevated phosphorylated
27    MAP kinase levels. These results should be interpreted cautiously, however, as LOH was
28    difficult to detect due to interference from infiltrating lymphocytes. In  summary, mouse tumor
29    tissue excised from mice used in the NTP (2002) study showed LOH in the region of the K-ras
30    oncogene  location in  17 of 19  samples tested. The signaling events, and specifically the role of
31    the MAP kinase pathway, associated with this oncogenic mutation, however, remain unclear.
32          Pierce et al. (1996) identified proinflammatory cytokines associated with vanadium
33    pentoxide-exposure. Pierce et al. (1996) reported increased mRNA expression of macrophage
34    inflammatory protein (MIP-2) and keratinocyte-derived cytokine (KC) in bronchiolar lavage
35    (BAL) fluid in vanadium pentoxide-treated female  CD rats compared to controls at early time
36    points (1 hour to 48 hours).
37          Monti el-Davalos et al.  (In Press) examined the effect of vanadium pentoxide exposure on
38    endothelial cells. Exposure to vanadium pentoxide to human umbilical vein endothelial cells

                This document is a draft for review purposes only and does not constitute Agency policy.
                                           D-28          DRAFT - DO NOT CITE OR QUOTE

-------
 1   (HUVEC; 3.12 |ag/cm2) induced enhanced adhesion of U937 macrophage cell line to HUVECs.
 2   HUVECs exposed to vanadium pentoxide also demonstrated an increase in reactive oxygen
 3   species and nitric oxide production, as well as diminished proliferation.  Yu et al. (2011)
 4   demonstrated in a primary cell strain and the NCI-H292 cell line an increase in genes related to
 5   mucin production, but not an increase in reactive oxygen species. The results demonstrated an
 6   increase in gene expression related to mucin expression following exposure to vanadium
 7   pentoxide that seemed unrelated to ROS production.
 8          In summary, hyperplastic responses following exposure to vanadium pentoxide are
 9   associated with increased expression of various mitogens such as PDGF and HB-EGF.
10   Moreover, PDGF activation could be dependent on inflammatory mediators such as IL-1B. HB-
11   EGF expression is dependent on activation of the MAP kinase and ERK signaling pathways.
12          No information is available concerning the rat neoplasm data.  Indeed, the marginal
13   increase in lung neoplasms observed in female rats was not statistically significant. Whether
14   lung neoplasms from male rats exhibit elevated activated MAP kinase or the rat tumors have
15   increased K-ras mutation and LOH on chromosome 6 is unknown.
16
                This document is a draft for review purposes only and does not constitute Agency policy.
                                           D-29          DRAFT - DO NOT CITE OR QUOTE

-------
  1    D.3.  REFERENCES
  2    Altamirano-Lozano. M: Alvarez-Barrera. L: Roldan-Reves. E. (1993). Cytogenetic and teratogenic effects of
  3            vanadium pentoxide on mice. Med Sci Res 21:711-713.
  4    Altamirano-Lozano. M: Alvarez-Barrera. L: Basurto-Alcantara. F: Valverde. M: Rojas. E. (1996). Reprotoxic and
  5            genotoxic studies of vanadium pentoxide in male mice. Teratog Carcinog Mutagen 16: 7-17.
  6    Altamirano-Lozano. M: Valverde. M: Alvarez-Barrera. L: Molina. B: Rojas. E. (1999). Genotoxic studies of
  7            vanadium pentoxide (V(2)O(5)) in male mice. II. Effects in several mouse tissues. Teratog Carcinog
  8            Mutagen 19: 243-255.
  9    Antao-Menezes. A: Turpin. EA: Bost PC: Ryman-Rasmussen. JP: Bonner. JC. (2008). STAT-1 signaling in human
10            lung fibroblasts is induced by vanadium pentoxide through an IFN-beta autocrine loop. J Immunol  180:
11            4200-4207.
12    Avila-Costa. MR: Montiel Flores. E: Colin-Barenque. L: Ordonez. JL: Gutierrez. AL: Nino-Cabrera. HG: Mussali-
13            Galante. P: Fortoul. TI. (2004). Nigrostriatal modifications after vanadium inhalation: An
14            immunocytochemical and cytological approach. Neurochem Res 29: 1365-1369.
15    Avila-Costa. MR: Colin-Barenque. L: Zepeda-Rodriguez. A: Antuna. SB: Saldivar. OL: Espejel-Maya. G: Mussali-
16            Galante. P: del Carmen Avila-Casado. M: Reves-Olivera. A: Anava-Martinez. V: Fortoul TI. (2005).
17            Ependymal epithelium disruption after vanadium pentoxide inhalation. A mice experimental model.
18            Neurosci Lett 381: 21-25. http://dx.doi.0rg/10.1016/i.neulet.2005.01.072.
19    Avila-Costa. MR: Fortoul TI: Nino-Cabrera. G: Colin-Barenque. L: Bizarro-Nevares. P: Gutierrez-Valdez. AL:
20            Ordonez-Librado. JL: Rodriguez-Lara. V: Mussali-Galante. P: Diaz-Bech. P: Anava-Martinez. V. (2006).
21            Hippocampal cell alterations induced by the inhalation of vanadium pentoxide (V(2)O(5)) promote
22            memory deterioration. Neurotoxicology  27: 1007-1012.
23    Bell M: Ebisu. K: Peng. R: Samet J: Dominici F. (2009). Hospital admissions and chemical composition of fine
24            particle air pollution. Am J Respir Crit Care Med 179: 1115-1120. http://dx.doi.org/10.1164/rccm.2008Q8-
25            1240OC.
26    Bonner. JC: Lindroos. PM: Rice. AB: Moomaw. CR: Morgan. PL. (1998). Induction of PDGF receptor-alpha in rat
27            myofibroblasts during pulmonary fibrogenesis in vivo. Am J Physiol 274: L72-L80.
28    Bonner. JC: Rice. AB: Moomaw. CR: Morgan. PL. (2000). Airway fibrosis in rats induced by vanadium pentoxide.
29            Am J Physiol Lung Cell Mol Physiol 278: L209-L216.
30    Bonner. JC: Rice. AB: Ingram. JL: Moomaw. CR: Nyska. A: Bradbury. A: Sessoms.  AR: Chulada. PC: Morgan.
31            PL: Zeldin. DC: Langenbach. R. (2002). Susceptibility of cyclooxygenase-2-deficient mice to pulmonary
32            fibrogenesis. Am J Pathol 161: 459-470.
33    Colin-Barenque. L: Martinez-Hernandez. MG: Baiza-Gutman. LA: Avila-Costa. MR: Ordonez-Librado. JL:
34            Bizarro-Nevares. P: Rodriguez-Lara. V:  Pinon-Zarate. G: Rojas-Lemus. M: Mussali-Galante. P: Fortoul.
35            TI. (2008). Matrix metalloproteinases 2 and 9 in central nervous system and their modification after
36            vanadium inhalation. J Appl Toxicol 28: 718-723. http://dx.doi.org/10.1002/iat. 1326.
3 7    Devereux. TR: Hollidav. W: Anna. C: Ress. N: Roycroft. J: Sills. RC.  (2002). Map kinase activation correlates with
3 8            K-ras mutation and loss of heterozygosity on chromosome 6 in alveolar bronchiolar carcinomas from
39            B6C3F1 mice exposed to vanadium pentoxide for 2 years. Carcinogenesis 23: 1737-1743.
40    Dominici. F: Peng. RD: Ebisu. K: Zeger. SL: Samet. JM: Bell. ML. (2007). Does the effect of PM10 on mortality
41            depend on PM nickel and vanadium content? A reanaly sis of the NMMAPS data. Environ Health Perspect
42            115: 1701-1703.
43    Ehrlich. VA: Nersesvan. AK: Hoelzl. C: Ferk. F:  Bichler. J: Valic. E: Schaffer. A: Schulte-Hermann. R: Fenech. M:
44            Wagner. K. -H: Knasmiiller. S. (2008). Inhalative exposure to vanadium pentoxide causes DNA damage in
45            workers: Results of a multiple end point study. Environ Health Perspect 116: 1689-1693.
46            http://dx.doi.org/10.1289/ehp. 11438.
47    Gonzalez-Villalva. A: Fortoul. TI: Avila-Costa. MR: Pinon-Zarate. G: Rodriguez-Laraa. V:  Martinez-Lew.  G:
48            Rojas-Lemus. M: Bizarro-Nevarez. P: Diaz-Bech. P: Mussali-Galante. P: Colin-Barenque. L. (2006).
49            Thrombocytosis induced in mice after subacute and subchronic V2O5 inhalation. Toxicol Ind Health 22:
50            113-116. http://dx.doi.org/10.1191/0748233706th250oa.
51    Hauser. R: Elreedv. S: Hoppin. JA: Christiani. DC. (1995). Upper airway response in workers exposed to fuel oil
52            ash: nasal lavage  analysis. Occup Environ Med 52: 353-358.
53    IARC. (International Agency for Research on Cancer). (2006). Cobalt in hard metals  and cobalt sulfate, gallium
54            arsenide, indium phosphide and vanadium pentoxide. In IARC Monographs on the Evaluation of the
55            Carcinogenic Risk of Chemicals to Humans (Vol. 86). Lyon, France.
56            http://monographs.iarc.fr/ENG/Monographs/vol86/index.php.
                   This document is a draft for review purposes only and does not constitute Agency policy.
                                                  D-3 0           DRAFT - DO NOT CITE OR QUOTE

-------
  1     Ingram. JL: Rice. AB: Santos. J: VanHouten. B: Bonner. JC. (2003). Vanadium-induced HB-EGF expression in
  2            human lung fibroblasts is oxidant dependent;and requires MAP kinases. Am J Physiol Lung Cell Mol
  3            Physiol 284: L774-L782.
  4     Ingram JL: Antao-Menezes. A: Turpin. EA: Wallace. DG: Mangum. JB: Pluta. JL: Thomas. RS: Bonner. JC.
  5            (2007). Genomic analysis of human lung fibroblasts exposed to vanadium pentoxide to identify candidate
  6            genes for occupational bronchitis. Respir Res 8: 34. http://dx.doi.org/10.1186/1465-9921-8-34.
  7     Ivancsits. S: Pilger. A: Diem E: Schaffer. A: Riidiger. HW. (2002). Vanadate induces DNA strand breaks in
  8            cultured human fibroblasts at doses relevant to occupational exposure. Mutat Res 519: 25-35.
  9            http://dx.doi.org/10.1016/S1383-5718(02)00138-9.
10     Kada. T: Hirano. K: Shirasu. Y. (1980). Screening of environmental chemical mutagens by the Rec-assay system
11            with Bacillus subtilis. In FJd Serres; A Hollaender (Eds.), Chemical mutagens: Principles and methods for
12            their detection (Vol. 6, pp. 149-173). New York, NY: Plenum Press.
13     Kanematsu. N: Kara. M: Kada. T. (1980). REC assay and mutagenicity studies on metal compounds. Mutat Res 77:
14            109-116. http://dx.doi.org/10.1016/0165-1218(80)90127-5.
15     Kerckaert GA: LeBoeuf. RA: Isfort RJ. (1996). Use of the Syrian hamster embryo cell transformation assay for
16            determining the carcinogenic potential of heavy metal compounds. Fundam Appl Toxicol 34: 67-72.
17            http://dx.doi.org/10.1006/faat.1996.0176.
18     Kim JY: Mukherjee. S: Ngo. LC: Christian! DC. (2004). Urinary 8-hydroxy-2'-deoxyguanosine as a biomarker of
19            oxidative DNA damage in workers  exposed to fine particulates. Environ Health Perspect 112: 666-671.
20     Kiviluoto. M: Pyy. L: Pakarinen. A. (1979).  Serum and urinary vanadium of vanadium-exposed workers. Scand J
21            Work Environ Health 5: 362-367.
22     Kiviluoto. M. (1980). Observations on the lungs of vanadium workers. Occup Environ Med 37: 363-366.
23     Kiviluoto. M: Pyy. L: Pakarinen. A. (1981).  Serum and urinary vanadium of workers processing vanadium
24            pentoxide. Int Arch Occup Environ Health 48: 251-256.
25     Kleinsasser. N: Dirschedl P:  Staudenmaier. R: Harreus. U: Wallner. B. (2003). Genotoxic effects of vanadium
26            pentoxide on human peripheral lymphocytes and mucosal cells of the upper aerodigestive tract. Int J
27            Environ Health Res  13:373-379.
28     Knecht. EA: Moorman. WJ: Clark. JC: Lynch. DW:  Lewis. TR. (1985). Pulmonary effects of acute vanadium
29            pentoxide inhalation in monkeys. Am Rev Respir Dis  132: 1181-1185.
30     Knecht EA: Moorman. WJ: Clark. JC: Hull RD: Biagini. RE: Lynch. DW: Boyle. TJ: Simon.  SD. (1992).
31            Pulmonary reactivity to vanadium pentoxide following subchronic inhalation exposure in a non-human
32            primate animal model. J Appl Toxicol 12: 427-434.
33     Lagorio. S: Forastiere. F: Pistelli. R: lavarone.  I: Michelozzi. P: Fano. V: Marconi. A: Ziemacki. G: Ostro. BD.
34            (2006). Air pollution and lung function among susceptible adult subjects:  A panel study. Environ Health 5:
35            11. http://dx.doi.org/10.1186/1476-069X-5-11.
36     Lew. BS: Hoffman. L: Gottsegen. S. (1984). Boilermakers bronchitis: respiratory tract irritation associated with
3 7            vanadium pentoxide exposure during oil-to-coal conversion of a power plant. J Occup Environ Med 26:
38            567-570.
39     Lewis. RJ.  Sr. (2000). Sax's Dangerous Properties of Industrial Materials. In (10 ed.). New York, NY: John Wiley
40            & Sons, Inc.
41     Lippmann. M: Ito. K: Hwang. JS: Maciejczyk. P: Chen. LC.  (2006). Cardiovascular effects of nickel in ambient air.
42            Environ Health Perspect 114:  1662-1669.
43     Montiel-Davalos. A: Gonzalez-Villava. A: Rodriguez-Lara. V: Montana LF: Fortoul TI: Lopez-Marure. R. (In
44            Press) Vanadium pentoxide induces activation and death of endothelial cells. J Appl Toxicol.
45            http://dx.doi.org/10.1002/iat.1695.
46     Motolese. A: Truzzi. M: Giannini. A: Seidenari. S. (1993). Contact dermatitis and contact sensitization among
47            enamellers and decorators in the ceramics industry.  Contact Dermatitis 28: 59-62.
48     Mravcova. A: Jirova. D: Janci. H: Lener. J. (1993). Effects of orally administered vanadium on the immune system
49            and bone metabolism in experimental animals. Sci Total Environ Suppl Pt 1: 663-669.
50     Mussali-Galante. P: Rodriguez-Lara. V: Hernandez-Tellez. B: Avila-Costa. MR: Colin-Barenque. L: Bizarro-
51            Nevarez. P: Martinez-Lew. G: Rojas-Lemus. M: Pinon-Zarate. G: Saldivar-Osorio. L: Diaz-Beck. P:
52            Herrera-Enriquez. MA: Tovar-Sanchez. E: Fortoul.  TI. (2005). Inhaled vanadium pentoxide decrease
53            gamma-tubulin of mouse testes at different exposure times. Toxicol Ind Health 21: 215-222.
54            http://dx.doi.org/10.1191/0748233705th232oa.
55     NTP. (National Toxicology Program). (2002). NTP toxicology and carcinogensis studies of vanadium pentoxide
56            (CAS No.; 1314-62-1) in F344/N rats and B6C3F1 mice (inhalation). Washington, DC.
57     Patel MM: Hoepner. L: Garfinkel R: Chillrud. S: Reyes. A: Quinn. JW: Perera. F: Miller. RL. (2009).  Ambient
5 8            metals, elemental carbon, and wheeze and cough in New York City children through 24 months of age. Am
59            J Respir Crit Care Med 180: 1107-1113. http://dx.doi.org/10.1164/rccm.200901-0122OC.

                   This document is a draft for review purposes only and does not constitute Agency policy.
                                                   D-31           DRAFT - DO NOT CITE OR QUOTE

-------
  1    Pazynich. VM. (1966). Maximum permissible concentration of vanadium pentoxide in the atmosphere. Gig Sanit
  2            317-9:6-12.
  3    Pierce. LM: Alessandrini. F: Godleski. JJ: Paulauskis. JD. (1996). Vanadium-induced chemokine mRNA expression
  4            and pulmonary inflammation. Toxicol Appl Pharmacol 138: 1-11.
  5    Pinon-Zarate. G: Rodriguez-Lara. V: Rojas-Lemus. M: Martinez-Pedraza. M: Gonzalez-Villalva. A: Mussali-
  6            Galante. P: Fortoul TI: Barquet A: Masso. F: Montano. LF. (2008). Vanadium pentoxide inhalation
  7            provokes germinal center hyperplasia and suppressed humoral immune responses. J Immunotoxicol 5: 115-
  8            122. http://dx.doi.org/10.1080/15476910802085749.
  9    Ramirez. P: Eastmond. DA: Laclette. JP: Ostrosky-Wegman. P. (1997). Disruption of microtubule assembly and
10            spindle formation as a mechanism for the;induction of aneuploid cells by sodium arsenite and vanadium
11            pentoxide. Mutat Res 386: 291-298. htto://dx.doi.org/10.1016/S1383-5742(97)00018-5.
12    Ress. NB: Chou. BJ: Renne. RA: Dill JA: Miller. RA: Roycroft JH: Hailev. JR: Haseman. JK: Bucher. JR. (2003).
13            Carcinogenicity of inhaled vanadium pentoxide in F344/N rats and B6C3F1 mice. Toxicol Sci 74: 287-296.
14            http://dx.doi.org/10.1093/toxsci/kfgl36.
15    Rice. AB: Moomaw. CR: Morgan. PL: Bonner. JC. (1999). Specific inhibitors of platelet-derived growth factor or
16            epidermal growth factor receptor tyrosine kinase reduce pulmonary fibrosis in rats. Am JPathol 155: 213-
17            221.
18    Rojas. E: Valverde. M: Herrera. LA: Altamirano-Lozano. M: Ostrosky-Wegman. P. (1996). Genotoxicity of
19            vanadium pentoxide evaluate by the single cell gel electrophoresis assay in human lymphocytes. Mutat Res
20            359:77-84.
21    Roldan. RE: Altamirano. LM. (1990). Chromosomal aberrations, sister-chromatid exchanges, cell-cycle kinetics and
22            satellite associations in human lymphocyte cultures exposed to vanadium pentoxide. Mutat Res 245: 61-65.
23    Rondinj EA: Walters. DM: Bauer. AK. (2010).  Vanadium pentoxide induces pulmonary inflammation and tumor
24            promotion in a strain-dependent manner. Part Fibre Toxicol 7:9. http://dx.doi.org/10.1186/1743-8977-7-9.
25    Sjoberg. SG. (1951). Health hazards in the production and handling of vanadium pentoxide. AMA Arch Ind Hyg
26            OccupMed3:631-646.
27    Sorensen. M: Schins. RPF: Hertel. O: Loft. S. (2005). Transition metals in personal samples of PM25 and oxidative
28            stress in human volunteers. Cancer Epidemiol Biomarkers Prev 14: 1340-1343.
29    Turpin. EA: Antao-Menezes. A: Cesta. MF: Mangum JB: Wallace. DG: Bermudez. E: Bonner. JC. (2010).
3 0            Respiratory syncytial virus infection reduces lung inflammation and fibrosis in mice exposed to vanadium
31            pentoxide. Respir Res 11: 20. http://dx.doi.org/10.1186/1465-9921-11-20.
32    WHO. (World Health Organization). (2001). Concise international chemical assessment document  29: Vanadium
33            pentoxide and other inorganic vanadium compounds. Geneva, Switzerland: World Health  Organization,
34            International Programme on Chemical Safety.
35            http://www.inchem.org/documents/cicads/cicads/cicad29.htm.
36    Woodin. MA: Hauser. R: Liu. Y: Smith. TJ: Siegel PD: Lewis. DM: Tollerud. DJ: Christiani. DC.  (1998).
3 7            Molecular markers of acute upper airway inflammation in workers  exposed to fuel-oil ash. Am J Respir
38            Crit Care Med 158: 182-187.
39    Yu. D: Walters. DM: Zhu. L: Lee. PK: Chen. Y. (2011). Vanadium pentoxide (V(2)O(5)) induced mucin production
40            by airway epithelium. Am J Physiol Lung Cell Mol Physiol 301: L31-L39.
41            http://dx.doi.org/10.1152/aiplung.00301.2010.
42    Zenz. C: Bartlett JP: Thiede. WH. (1962). Acute vanadium pentoxide intoxication. Arch Environ Occup Health 5:
43            542-546.
44    Zhang. L: Rice. AB: Adler. K: Sannes. P: Martin. L: Gladwell. W: Koo. JS:  Gray. TE: Bonner. JC.  (2001a).
45            Vanadium stimulates human bronchial  epithelial cells to produce heparin-binding epidermal growth factor-
46            like growth factor: a mitogen for lung fibroblasts. Am J Respir Cell Mol Biol 24: 123-131.
47    Zhang. Z: Wang. Y: Vikis. HG: Johnson. L: Liu. G: Li. J: Anderson. MW: Sills. RC: Hong. HL: Devereux. TR:
48            Jacks. T: Guan. KL: You. M. (200Ib). Wildtype Kras2 can inhibit lung carcinogenesis in mice. Nat Genet
49            29: 25-33. http://dx.doi.org/10.1038/ng721.
50    Zhong. BZ: Gu. ZW: Wallace. WE: Whong. WZ: Ong. T. (1994). Genotoxicity of vanadium pentoxide in Chinese
51            hamster V79 cells. Mutat Res 321: 35-42.
                   This document is a draft for review purposes only and does not constitute Agency policy.
                                                  D-32           DRAFT - DO NOT CITE OR QUOTE

-------