v>EPA
EPA/635/R-16/208b
External Review Draft
www.epa.gov/iris
Toxicological Review of Hexahydro-l,3,5-trinitro-l,3,5-triazine (RDX)
(CASRN 121-82-4]
Supplemental Information
September 2016
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.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
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 of 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ii	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
CONTENTS
APPENDIX A. ASSESSMENTS BY OTHER NATIONAL AND INTERNATIONAL HEALTH AGENCIES	A-l
APPENDIX B. ADDITIONAL DETAILS OF LITERATURE SEARCH STRATEGY | STUDY SELECTION
AND EVALUATION	B-l
B.l.	DEFENSE TECHNICAL INFORMATION CENTER (DTIC) LITERATURE SEARCH
AND SCREEN	B-l
APPENDIX C. INFORMATION IN SUPPORT OF HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS	C-l
C.l.	TOXICOKINETICS	C-l
C.1.1. Absorption	C-l
C.l.2. Distribution	C-6
C.l.3. Metabolism	C-7
C.1.4. Excretion	C-9
C.l.5. Physiologically Based Pharmacokinetic (PBPK) Models	C-ll
C.2. HUMAN STUDIES	C-32
C.3.	OTHER PERTINENT TOXICITY INFORMATION	C-38
C.3.1. Mortality in Animals	C-38
C.3.2. Other Noncancer Effects	C-43
C.3.3.	Genotoxicity	C-63
APPENDIX D. DOSE-RESPONSE MODELING FOR THE DERIVATION OF REFERENCE VALUES FOR
EFFECTS OTHER THAN CANCER AND THE DERIVATION OF CANCER RISK
ESTIMATES	D-l
D.l.	BENCHMARK DOSE MODELING SUMMARY FOR NONCANCER ENDPOINTS	D-l
D.l.l.	Evaluation of Model Fit and Model Selection	D-3
D.l.2. Modeling Results	D-4
D.1.3. Mortality: Dose-Response Analysis and BMD Modeling
Documentation	D-20
D.2. BENCHMARK DOSE MODELING SUMMARY FOR CANCER ENDPOINTS	D-30
D.2.1. Evaluation of Model Fit and Model Selection for Mouse Tumor
Data	D-31
D.2.2. Modeling Results for Mouse Tumor Data	D-31
This document is a draft for review purposes only and does not constitute Agency policy.
iii	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
D.2.3. Dose-response Analysis and BMD Modeling Documentation for
Other Tumor Data Sets	D-45
APPENDIX E. SUMMARY OF PUBLIC COMMENTS AND EPA's DISPOSITION	E-l
REFERENCES FOR APPENDICES	R-l
This document is a draft for review purposes only and does not constitute Agency policy.
iv	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
TABLES
Table A-l. Assessments by other national and international health agencies	A-l
Table B-l. Summary of detailed search strategies for RDX (Pubmed, Toxline, Toxcenter,
TSCATS)	B-3
Table B-2. Summary of detailed search strategies for RDX (DTIC)	B-12
Table B-3. Processes used to augment the search of core databases for RDX	B-13
Table C-l. Distribution of RDX or radiolabel from administered RDX	C-7
Table C-2. Principal urinary metabolites of RDX in miniature swine 24 hours after dosing with
RDX	C-9
Table C-3. Elimination ti/2 values for RDX or radiolabeled RDX	C-10
Table C-4. Parameter values used in the Sweeney et al. (2012a) and Sweeney et al. (2012b)
PBPK models for RDX in rats, humans, and mice as reported by authors	C-12
Table C-5. Parameters values used in the EPA application of the rat, human, and mouse
models	C-15
Table C-6. Doses, dosing formulations, and absorption rate constants in animal and human
studies	C-20
Table C-7. Sensitivity coefficients for rat and human RDX PBPK models	C-26
Table C-8. Summary of case reports of exposure to RDX	C-32
Table C-9. Occupational epidemiologic studies of RDX: summary of methodologic features ..C-36
Table C-10. Evidence pertaining to mortality in animals	C-39
Table C-ll. Evidence pertaining to other noncancer effects (hematological) in humans	C-47
Table C-12. Evidence pertaining to other noncancer effects in animals	C-49
Table C-13. Summary of in vitro studies of the genotoxicity of RDX	C-64
Table C-14. Summary of in vivo studies of the genotoxicity of RDX	C-68
Table C-15. Summary of in vitro and in vivo studies of the genotoxicity of RDX metabolites ..C-69
Table D-l. Noncancer endpoints selected for dose-response modeling for RDX	D-l
Table D-2. Convulsion or mortality endpoints from Crouse et al. (2006) selected for dose-
response modeling for RDX	D-2
Table D-3. Model predictions for convulsions in female F344 rats exposed to RDX by gavage for
90 days (Crouse et al., 2006); BMR= 1% ER	D-4
Table D-4. Model predictions for convulsions in male F344 rats exposed to RDX by gavage for
90 days (Crouse et al., 2006); BMR = 1% ER	D-6
Table D-5. Model predictions for convulsions in male and female F344 rats exposed to RDX by
gavage for 90 days (Crouse et al., 2006); BMR = 1% ER	D-9
Table D-6. Model predictions for convulsions in female F344 rats exposed to RDX by gavage on
GDs 6-19 (Cholakis et al., 1980); BMR = 1% ER	D-ll
Table D-7. Model predictions for combined incidence of convulsion and mortality in male and
female F344 rats exposed to RDX by gavage for 90 days (Crouse et al., 2006);
BMR = 1% ER	D-13
Table D-8. Model predictions for testicular degeneration in male B6C3Fi mice exposed to RDX
by diet for 24 months (Lish et al., 1984); BMR = 10% ER	D-15
Table D-9. Model predictions for prostate suppurative inflammation in male F344 rats exposed
to RDX by diet for 24 months (Levine et al., 1983); BMR = 10% ER	D-18
Table D-10. Mortality data selected for dose-response modeling for RDX	D-20
Table D-ll. BMD modeling results for combined mortality in male and female F344 rats
exposed to RDX by diet for 13 weeks (Levine et al., 1981b); BMR = 1% ER	D-23
This document is a draft for review purposes only and does not constitute Agency policy.
v	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
Table D-12. BMD modeling results for mortality (number found dead) in rats exposed to RDX in
the diet for 13 weeks (von Oettingen et al., 1949)	D-25
Table D-13. BMD modeling results for combined mortality (number found dead) in male and
female F344 rats exposed to RDX by gavage for 90 days (Crouse et al., 2006);
BMR = 1% ER	D-26
Table D-14. Model predictions for mortality in female Sprague-Dawley rats exposed to RDX by
gavage on gestation days 6-15 (Angerhofer et al., 1986); BMR = 1% ER	D-28
Table D-15. Cancer endpoints selected for dose-response modeling for RDX	D-30
Table D-16. Model predictions for combined alveolar/bronchiolar adenoma and carcinoma in
female B6C3Fi mice exposed to RDX by diet for 24 months (Lish et al., 1984);
BMR = 10% ER	D-32
Table D-17. Model predictions for combined alveolar/bronchiolar adenoma and carcinoma in
female B6C3Fi mice exposed to RDX by diet for 24 months (Lish et al., 1984);
BMR = 5% ER	D-34
Table D-18. Model predictions for combined hepatocellular adenoma and carcinoma in female
B6C3Fi mice exposed to RDX by diet for 24 months (Parker et al., 2006); BMR = 10%
ER D-36
Table D-19. Model predictions for B6C3Fi female mouse combined hepatocellular adenoma and
carcinoma in mice exposed to RDX by diet for 24 months (Parker et al., 2006); BMR
= 5% ER	D-38
Table D-20. Model predictions for combined hepatocellular adenoma and carcinoma in female
B6C3Fi mice exposed to RDX by diet for 24 months, using incidence frequencies
from Parker et al. (2006) and sample sizes from Lish et al. (1984); BMR = 10% ER .. D-
42
Table D-21. Liver carcinoma data from Levine et al. (1983)	D-45
Table D-22. Model predictions and oral slope factor for alveolar/bronchiolar carcinomas in male
B6C3Fi mice exposed to RDX by diet for 2 years (Lish et al., 1984)	D-46
Table D-23. Summary of BMD modeling results for model predictions for alveolar/bronchiolar
carcinoma in male B6C3Fi mice exposed to RDX by diet for 2 years (Lish et al., 1984);
BMR = 10% ER	D-46
Table D-24. Model predictions and oral slope factor for hepatocellular carcinomas in male F344
rats administered RDX in the diet for 2 years (Levine et al., 1983)	D-49
Table D-25. Model predictions for combined hepatocellular adenoma and carcinoma in F344
rats exposed to RDX by diet for 24 months (Levine et al., 1983); BMR = 5% ER ... D-50
This document is a draft for review purposes only and does not constitute Agency policy.
vi	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
FIGURES
Figure C-l. PBPK model structure for RDX in rats and humans	C-ll
Figure C-2. EPA rat PBPK model predictions fitted to observed RDX blood concentrations in male
and female Sprague-Dawley rats following i.v. exposure	C-17
Figure C-3. EPA rat PBPK model predictions fitted to observed RDX blood concentrations
following oral exposure to RDX dissolved in water	C-18
Figure C-4. EPA rat model predictions fitted to observed RDX blood concentrations following
oral exposure to RDX in dry capsules	C-18
Figure C-5. Effect of varying oral absorption parameters on EPA rat model predictions fitted to
observed RDX blood concentrations following oral exposure to coarse-grain RDX. ..C-
19
Figure C-6. EPA rat model predictions fitted to observed RDX brain tissue concentrations
following oral exposure to RDX	C-19
Figure C-7. EPA rat model predictions fitted to observed RDX blood concentrations following
oral exposure to fine-grain RDX in a saline slurry	C-21
Figure C-8. Comparison of EPA rat model predictions with data from Schneider et al. (1978) for
the subchronic gavage study	C-22
Figure C-9. Comparison of EPA rat model predictions with data from Schneider et al. (1978) for
the subchronic drinking water study	C-22
Figure C-10. EPA human model predictions fitted to observed RDX blood concentrations
resulting from an accidental ingestion of RDX by a 14.5-kg boy (Woody et al.,
1986)	C-24
Figure C-ll. EPA human model predictions fitted to observed RDX blood concentrations
resulting from accidental exposure to adults assumed to be 70 kg (Ozhan et al.,
2003)	C-24
Figure C-12. Comparison of EPA mouse PBPK model predictions with data from oral exposure to
RDX dissolved in water	C-27
Figure D-l. Plot of incidence rate by dose, with fitted curve for selected model, for convulsions
in female F344 rats exposed to RDX by gavage for 90 days (Crouse et al., 2006)... D-5
Figure D-2. Plot of incidence rate by dose, with fitted curve for selected model, for convulsions
in male F344 rats exposed to RDX by gavage for 90 days (Crouse et al., 2006)	D-7
Figure D-3. Plot of incidence rate by dose, with the fitted curve of the multistage 3° model, for
convulsions in male and female F344 rats exposed to RDX by gavage for 90 days
(Crouse et al., 2006). BMR = 1% ER; dose shown in mg/kg-day	D-10
Figure D-4. Plot of incidence rate by dose, with the fitted curve of the selected model, for
convulsions in female F344 rats exposed to RDX by gavage on GDs 6-19 (Cholakis et
al., 1980)	D-12
Figure D-5. Plot of incidence rate by dose with fitted curve for Multistage 3° model for model
predictions for combined incidence of convulsion and mortality in male and female
F344 rats exposed to RDX by gavage for 90 days (Crouse et al., 2006)	D-14
Figure D-6. Plot of incidence rate by dose, with fitted curve for selected model, for testicular
degeneration in male B6C3Fi mice exposed to RDX by diet for 24 months (Lish et al.,
1984)	D-16
Figure D-7. Plot of incidence rate by dose, with fitted curve for selected model, for prostate
suppurative inflammation in male F344 rats exposed to RDX by diet for 24 months
(Levine et al., 1983)	D-19
This document is a draft for review purposes only and does not constitute Agency policy.
vii	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
Figure D-8. Plot of incidence rate by dose, with the fitted curve of the multistage 2° model, for
combined mortality in male and female F344 rats exposed to RDX by diet for 13
weeks (Levine et al., 1981b); BMR = 1% ER	D-24
Figure D-9. Plot of incidence rate by dose with fitted curve for Dichotomous-Hill model for
Model predictions for mortality (number found dead) in rats exposed to RDX in the
diet for 13 weeks (von Oettingen et al., 1949); dose shown in mg/kg-day	D-26
Figure D-10. Plot of incidence rate by dose, with the fitted curve of the multistage 2° model, for
mortality in male and female F344 rats exposed to RDX by gavage for 90 days
(Crouse et al., 2006). BMR = 1% ER; dose shown in mg/kg-day	D-27
Figure D-ll. Plot of incidence rate by dose, with the fitted curve of the multistage 3° model, for
mortality in female Sprague-Dawley rats exposed to RDX by gavage on gestation
days 6-15 (Angerhofer et al., 1986); BMR = 1% ER	D-29
Figure D-12. Plot of incidence rate by dose, with the fitted curve for the selected model, for
combined alveolar/bronchiolar adenoma and carcinoma in female B6C3Fi mice
exposed to RDX by diet for 24 months (Lish et al., 1984)	D-32
Figure D-13. Plot of incidence rate by dose, with fitted curve for selected model, for combined
alveolar/bronchiolar adenoma and carcinoma in female B6C3Fi mice exposed to
RDX by diet for 24 months (Lish et al., 1984)	D-34
Figure D-14. Plot of incidence rate by dose, with fitted curve for selected model, for combined
hepatocellular adenoma and carcinoma in female B6C3Fi mice exposed to RDX by
diet for 24 months (Parker et al., 2006)	D-36
Figure D-15. Plot of incidence rate by dose, with fitted curve for selected model, for B6C3Fi
female mouse combined hepatocellular adenoma and carcinoma in mice exposed to
RDX by diet for 24 months (Parker et al., 2006)	D-38
Figure D-16. Plot of incidence rate by dose with fitted curve for Multistage-Cancer 1° model for
model predictions for combined hepatocellular adenoma and carcinoma in female
B6C3Fi mice exposed to RDX by diet for 24 months, using incidence frequencies
from Parker et al. (2006) and sample sizes from Lish et al. (1984)	D-42
Figure D-17. Plot of incidence rate by dose with fitted curve for Multistage-Cancer 1° model for
Model predictions for alveolar/bronchiolar carcinoma in male B6C3Fi mice exposed
to RDX by diet for 24 months (Lish et al., 1984)	D-47
Figure D-18. Plot of incidence rate by dose, with fitted curve for selected model, for combined
hepatocellular adenoma and carcinoma in F344 rats exposed to RDX by diet for
24 months (Levine et al., 1983)	D-50
This document is a draft for review purposes only and does not constitute Agency policy.
viii	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
ABBREVIATIONS
AAP	Army ammunition plant	FUDS
ACGIH	American Conference of Governmental	GABA
Industrial Hygienists	GD
AChE	acetylcholinesterase	GI
ADAF	age-dependent adjustment factor	GLP
AIC	Akaike's information criterion	HED
ALP	alkaline phosphatase	HERO
ALT	alanine aminotransferase
AOP	adverse outcome pathway	HGPRT
AST	aspartate aminotransferase
atm	atmosphere	HMX
ATSDR	Agency for Toxic Substances and
Disease Registry	IARC
AUC	area under the curve
BDNF	brain-derived neurotrophic factor	i.p.
BHC	beta-hexachlorocyclohexane	IPCS
BMC	benchmark concentration
BMCL	benchmark concentration lower	IRIS
confidence limit	IUR
BMD	benchmark dose	i.v.
BMDL	benchmark dose lower confidence limit	LDH
BMDS	Benchmark Dose Software	LOAEL
BMDU	benchmark dose upper bound	LOD
BMR	benchmark response	miRNA
BUN	blood urea nitrogen	MNX
BW	body weight
CAAC	Chemical Assessment Advisory	MOA
Committee	MRL
CASRN	Chemical Abstracts Service Registry	NAPDH
Number
CCL	Contaminant Candidate List	NAS
CI	confidence interval	NCE
CICAD	Concise International Chemical	NCEA
Assessment Document
CNS	central nervous system	NCI
CSF	cerebrospinal fluid	NCTR
CYP450	cytochrome P450
DAF	dosimetric adjustment factor	NHANES
DDT	dichlorodiphenyltrichloroethane
d.f.	degrees of freedom	NICNAS
DMSO	dimethylsulfoxide
DNA	deoxyribonucleic acid	NIEHS
DNX	l-nitro-3,5-dinitroso-
1,3,5-triazacyclohexane	NIOSH
DTIC	Defense Technical Information Center
EEG	electroencephalogram	NOAEL
EHC	Environmental Health Criteria	NOEL
EPA	Environmental Protection Agency	NPL
ER	extra risk	NRC
FDA	Food and Drug Administration	NSCEP
FOB	functional observational battery
Formerly Used Defense Sites
gamma-amino butyric acid
gestational day
gastrointestinal
good laboratory practices
human equivalent dose
Health and Environmental Research
Online
hypoxanthine-guanine
phosphoribosyltransferase
octahydro-l,3,5,7-tetranitro-
1,3,5,7-tetrazocine
International Agency for Research on
Cancer
intraperitoneal
International Programme on Chemical
Safety
Integrated Risk Information System
inhalation unit risk
intravenous
lactate dehydrogenase
lowest-observed-adverse-effect level
limit of detection
microRNA
hexahydro-l-nitroso-3,5-dinitro-
1,3,5-triazine
mode of action
Minimal Risk Level
nicotinamide adenine dinucleotide
phosphate
National Academy of Science
normochromatic erythrocyte
National Center for Environmental
Assessment
National Cancer Institute
National Center for Toxicological
Research
National Health and Nutrition
Examination Survey
National Industrial Chemicals
Notification and Assessment Scheme
National Institute of Environmental
Health Sciences
National Institute for Occupational
Safety and Health
no-observed-adverse-effect level
no-observed-effect level
National Priorities List
Nuclear Regulatory Commission
National Service Center for
Environmental Publications
This document is a draft for review purposes only and does not constitute Agency policy.
ix	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
NTP
National Toxicology Program
SGOT
glutamic oxaloacetic transaminase, also
NZW
New Zealand White

known as AST
OR
odds ratio
SGPT
glutamic pyruvic transaminase, also
ORD
Office of Research and Development

known as ALT
OSF
oral slope factor
SLE
systemic lupus erythematosus
OSHA
Occupational Safety and Health
SS
scheduled sacrifice

Administration
TLV
Threshold Limit Value
PBPK
physiologically based pharmacokinetic
TNT
trinitrotoluene
PCB
polychlorinated biphenyl
TNX
hexahydro-1,3,5-trinitroso-
PCE
polychromatic erythrocyte

1,3,5-triazine
PEL
Permissible Exposure Limit
TSCATS
Toxic Substances Control Act Test
PND
postnatal day

Submissions
POD
point of departure
TWA
time-weighted average
PWG
Pathology Working Group
U.S.
United States of America
RBC
red blood cell
UCM
Unregulated Contaminant Monitoring
RDX
Royal Demolition eXplosive
UF
uncertainty factor

(hexahydro-1,3,5-trinitro-
UFa
animal-to-human uncertainty factor

1,3,5-triazine)
UFd
database deficiencies uncertainty factor
REL
Recommended Exposure Limit
UFh
human variation uncertainty factor
RfC
inhalation reference concentration
UFl
LOAEL-to-NOAEL uncertain factor
RfD
oral reference dose
UFs
subchronic-to-chronic uncertainty
SDMS
spontaneous death or moribund

factor

sacrifice
WBC
white blood cell
SDWA
Safe Drinking Water Act
WHO
World Health Organization
This document is a draft for review purposes only and does not constitute Agency policy.
x	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
APPENDIX A. ASSESSMENTS BY OTHER NATIONAL
AND INTERNATIONAL HEALTH AGENCIES
1	Table A-l. Assessments by other national and international health agencies
Organization
Toxicity value
Agency for Toxic Substances and
Disease Registry (ATSDR, 2012)
Acute oral Minimal Risk Level (MRL)—0.2 mg/kg-d
Basis: tremors and convulsions in rats (Crouse et al., 2006); application
of a composite uncertainty factor (UF) of 30 (3 for extrapolation from
animals to humans with dosimetric adjustments [physiologically based
pharmacokinetic or PBPK modeling] and 10 for human variability)
Intermediate oral MRL—0.1 mg/kg-d
Basis: convulsions in rats (Crouse et al., 2006); application of a
composite UF of 30 (3 for extrapolation from animals to humans with
dosimetric adjustments [PBPK modeling] and 10 for human variability)
Chronic oral MRL—0.1 mg/kg-d
Basis: tremors and convulsions in rats (Levine et al.. 1983); application
of a composite UF of 30 (3 for extrapolation from animals to humans
with dosimetric adjustments [PBPK modeling] and 10 for human
variability)
National Institute for Occupational
Safetv and Health (NIOSH, 2012)
Recommended Exposure Limit (REL)—1.5 mg/m3 TWA for up to a 10-hr
workday during a 40-hr workweek; short-term (15-min) limit—
3 mg/m3
Basis: agreed with Occupational Safety and Health Administration
(OSHA)-proposed Permissible Exposure Limit (PEL) in 1988 PEL
Hearings
Skin designation indicates potential for dermal absorption
Basis: agreed with OSHA proposal for skin notation in 1988 PEL
Hearings
Occupational Safety and Health
Administration (OSHA, 2012a, b)
PEL—1.5 mg/m3 time-weighted average (TWA) for an 8-hr workday in a
40-hr workweek
Basis: adopted from the American Conference of Governmental
Industrial Hygienists (ACGIH) Threshold Limit Value (TLV) established
in 1969
Skin designation indicates that cutaneous exposure may contribute to
overall exposure and measures should be taken to prevent skin
absorption
Basis: adopted from ACGIH
Hazardous Substances Information
Svstem (Safe Work Australia, 2014)
Exposure standard—1.5 mg/m3 TWA for an 8-hr workday
Basis: adopted from the ACGIH TLV established in 1991
Skin absorption notice indicates that absorption through the skin may be
a significant source of exposure
Basis: adopted from ACGIH
This document is a draft for review purposes only and does not constitute Agency policy.
A-l	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
APPENDIX B. ADDITIONAL DETAILS OF LITERATURE
SEARCH STRATEGY | STUDY SELECTION AND
EVALUATION
The literature search for hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) was conducted in
five online scientific databases through May 2016. The detailed search strategy used to search four
of these databases—PubMed, Toxline, Toxcenter, and Toxic Substances Control Act Test
Submissions (TSCATS)—is provided in Table B-l. Toxcenter, a fee-based scientific database, was
searched outside of HERO. Toxcenter searches initially yield titles only; obtaining complete
citations and abstracts incurs additional costs. Thus, titles only were initially screened; for titles
identified as potentially relevant, complete citations with abstracts, when available, were
downloaded and rescreened. Of the rescreened citations, only those selected for full text review
were added to HERO and the RDX project page. The search strategy used to search the Defense
Technical Information Center (DTIC) database is described in Table B-2. The computerized
database searches were augmented by review of online regulatory sources, as well as "forward"
and "backward" Web of Science searches of two recent reviews (Table B-3). Forward searching
was used to identify articles that cited the selected studies (i.e., the two reviews identified in
Table B-3), and backward searching was used to identify articles that the selected studies cited.
B.l. DEFENSE TECHNICAL INFORMATION CENTER (DTIC) LITERATURE
SEARCH AND SCREEN
Among the RDX-related citations thatwere identified in the January 2015 search ofthe
DTIC database, 826 (722 after duplicate removal within DTIC) were classified with the distribution
"approved for public release", 239 (217 after duplicate removal) were classified as "distribution
limited to U.S. Government agencies and their contractors," and 199 (181 after duplicate removal)
were classified as "distribution limited to U.S. Government agencies only." A preliminary screen of
the 1,120 unique citations was performed; 85 citations with unlimited distribution and 10 citations
with limited distribution were selected for further review as potential sources of health effects data
or supporting information. The remaining 1,025 unlimited and limited-distribution DTIC
references not selected for further consideration were not studies of RDX or did not contain
information pertinent to the assessment of the health effects of RDX (e.g., documents were related
to environmental properties such as leaching explosive properties, fuel and propellant properties,
weapons systems, treatment of wastewater containing explosives, and disposal technologies). An
update of the DTIC search was performed in May 2016. The update search identified 21 items
This document is a draft for review purposes only and does not constitute Agency policy.
B-l	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	classified as "approved for public release," 9 classified as "distribution limited to U.S. Government
2	agencies and their contractors," and 9 classified as "distribution limited to U.S. Government
3	agencies only;" none of these was selected for further review, as none met the inclusion criteria
4	outlined in Table LS-1 of the main document (i.e., none contained health effects data or supporting
5	information).
6	The 85 unique selected citations with unlimited distribution were uploaded to the Health
7	and Environmental Research Online (HERO) website1 f http://hero.epa. gov). The 10 citations with
8	limited distribution were subject to a more in-depth screen to determine whether the references
9	provided additional primary health effects data and whether the U.S. Environmental Protection
10	Agency (EPA) should seek authorization for public distribution and upload to HERO. A review of
11	the abstract or full-text of the documents associated with the limited-distribution citations resulted
12	in the following determinations:
13	• one citation was excluded because it did not provide additional primary health effects data.
14	The citation reported data from a study that was subsequently published (Hathaway and
15	Buck. 1977) and had already been identified by the literature search strategy.
16	• one citation (dated 1944) provided human and animal inhalation data and was considered
17	pertinent, but was not brought forward for further review because flaws in the design of
18	both the human and animal studies were such that results would not be considered
19	credible. Experimental animal study design issues included lack of a control group, small
20	numbers of animals, incomplete information on dosage or exposure levels, and inadequate
21	reporting. The human study described a case series and lacked a referent group and
22	measures of RDX exposure.
23	• eight citations were regulatory documents, reviews, or risk assessments that did not
24	specifically identify RDX and did not appear to contain primary health effects data.
25	Based on these determinations, none of the 10 limited distribution citations that were
26	subject to further review were selected for further consideration or added to HERO.
27
'HERO is a database of scientific studies and other references used to develop EPA's assessments aimed at
understanding the health and environmental effects of pollutants and chemicals. It is developed and
managed in EPA's Office of Research and Development (ORD) by the National Center for Environmental
Assessment (NCEA). The database includes more than 1.6 million scientific references, including articles
from the peer-reviewed literature. New studies are added continuously to HERO.
This document is a draft for review purposes only and does not constitute Agency policy.
B-2	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table B-l. Summary of detailed search strategies for RDX (Pubmed, Toxline,
2	Toxcenter, TSCATS)
Database
Terms
Hits
PubMed
Date: 4/2012
((((121-82-4) OR (Cyclonite[tw] OR Cyclotrimethylenetrinitramine[tw] OR
"cyclotrimethylene trinitramine"[tw] OR "Hexahydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR "Hexahydro-1,3,5-trinitro-s-triazine"[tw] OR
Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-Triaza-
l,3,5-trinitrocyclohexane"[tw] OR "l,3,5-Trinitro-l,3,5-triazacyclohexane"[tw]
OR "l,3,5-Trinitrohexahydro-l,3,5-triazine"[tw] OR "1,3,5-Trinitrohexahydro-
s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-triazine"[tw] OR "Esaidro-
l,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-l,3,5-trinitro-
l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw] OR
Cyclotrimethylenenitramine[tw] OR Trimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] OR Trimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] OR Trinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) OR (rdx[tw])) NOT medline[sb]) OR
(((121-82-4) OR (Cyclonite[tw] OR Cyclotrimethylenetrinitramine[tw] OR
"cyclotrimethylene trinitramine"[tw] OR "Hexahydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR "Hexahydro-1,3,5-trinitro-s-triazine"[tw] OR
Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-Triaza-
l,3,5-trinitrocyclohexane"[tw] OR "l,3,5-Trinitro-l,3,5-triazacyclohexane"[tw]
OR "l,3,5-Trinitrohexahydro-l,3,5-triazine"[tw] OR "1,3,5-Trinitrohexahydro-
s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-triazine"[tw] OR "Esaidro-
l,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-l,3,5-trinitro-
l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw] OR
Cyclotrimethylenenitramine[tw] OR Trimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] ORTrimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] OR Trinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) OR (rdx[tw]» AND (to[sh] OR po[sh]
OR ae[sh] OR pk[sh] OR (me[sh] AND (humans[mh] OR animals[mh])) OR ci[sh]
OR bl[sh] OR cf[sh] OR ur[sh] OR ((pharmacokinetics[mh] OR metabolism[mh])
AND (humans[mh] OR mammals[mh])) OR "dose-response relationship,
drug"[mh] OR risk[mh] OR "toxicity tests"[mh] OR noxae[mh] OR cancer[sb]
OR "endocrine system"[mh] OR "endocrine disruptors"[mh] OR "Hormones,
Hormone Substitutes, and Hormone Antagonists"[mh] OR triazines/ai OR
("Inhalation Exposure"[Mesh] OR "Maternal Exposure"[Mesh] OR "Maximum
Allowable Concentration"[Mesh] OR "Occupational Exposure"[Mesh] OR
"Paternal Exposure"[Mesh] OR "Environmental Exposure"[Mesh:noexp]))))
NOT (((((121-82-4) OR (Cyclonite[tw] OR Cyclotrimethylenetrinitramine[tw] OR
"cyclotrimethylene trinitramine"[tw] OR "Hexahydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR "Hexahydro-1,3,5-trinitro-s-triazine"[tw] OR
Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-Triaza-
l,3,5-trinitrocyclohexane"[tw] OR "l,3,5-Trinitro-l,3,5-triazacyclohexane"[tw]
OR "l,3,5-Trinitrohexahydro-l,3,5-triazine"[tw] OR "1,3,5-Trinitrohexahydro-
s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-triazine"[tw] OR "Esaidro-
1,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-1,3,5-trinitro-
337
This document is a draft for review purposes only and does not constitute Agency policy.
B-3	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw] OR
Cyclotrimethylenenitramine[tw] ORTrimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] OR Trimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] ORTrinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) OR (rdx[tw])) NOT medline[sb]) OR
(((121-82-4) OR (Cyclonite[tw] OR Cyclotrimethylenetrinitramine[tw] OR
"cyclotrimethylene trinitramine"[tw] OR "Hexahydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR "Hexahydro-1,3,5-trinitro-s-triazine"[tw] OR
Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-Triaza-
l,3,5-trinitrocyclohexane"[tw] OR "l,3,5-Trinitro-l,3,5-triazacyclohexane"[tw]
OR "l,3,5-Trinitrohexahydro-l,3,5-triazine"[tw] OR "1,3,5-Trinitrohexahydro-
s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-triazine"[tw] OR "Esaidro-
l,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-l,3,5-trinitro-
l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw] OR
Cyclotrimethylenenitramine[tw] OR Trimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] OR Trimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] OR Trinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) OR (rdx[tw]» AND (to[sh] OR po[sh]
OR ae[sh] OR pk[sh] OR (me[sh] AND (humans[mh] OR animals[mh])) OR ci[sh]
OR bl[sh] OR cf[sh] OR ur[sh] OR ((pharmacokinetics[mh] OR metabolism[mh])
AND (humans[mh] OR mammals[mh])) OR "dose-response relationship,
drug"[mh] OR risk[mh] OR "toxicity tests"[mh] OR noxae[mh] OR cancer[sb]
OR "endocrine system"[mh] OR "endocrine disruptors"[mh] OR "Hormones,
Hormone Substitutes, and Hormone Antagonists"[mh] OR triazines/ai OR
("Inhalation Exposure"[Mesh] OR "Maternal Exposure"[Mesh] OR "Maximum
Allowable Concentration"[Mesh] OR "Occupational Exposure"[Mesh] OR
"Paternal Exposure"[Mesh] OR "Environmental Exposure"[Mesh:noexp]))))
AND (invertebrates OR aquatic organisms OR fish OR fishes OR amphibians OR
earthworm*))

PubMed
Date limit:
1/2012-
2/2013
(Cyclonite[tw] OR RDX[tw] OR Cyclotrimethylenetrinitramine[tw] OR
"cyclotrimethylene trinitramine"[tw] OR "Hexahydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR "Hexahydro-1,3,5-trinitro-s-triazine"[tw] OR
Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-Triaza-
l,3,5-trinitrocyclohexane"[tw] OR "l,3,5-Trinitro-l,3,5-triazacyclohexane"[tw]
OR "l,3,5-Trinitrohexahydro-l,3,5-triazine"[tw] OR "1,3,5-Trinitrohexahydro-
s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-triazine"[tw] OR "Esaidro-
l,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-l,3,5-trinitro-
l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw] OR
Cyclotrimethylenenitramine[tw] OR Trimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] OR Trimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] OR Trinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) AND (("2012/01/01"[Date - MeSH] :
"3000"[Date - MeSH]) OR ("2012/01/01"[Date - Entrez] : "3000"[Date -
Entrez]) OR ("2012/01/01"[Date - Create] : "3000"[Date - Create]))
112
This document is a draft for review purposes only and does not constitute Agency policy.
B-4	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits
PubMed
Date limit:
11/2012-
1/2014
(Cyclonite[tw] OR RDX[tw] OR Cyclotrimethylenetrinitramine[tw] OR
"cyclotrimethylene trinitramine"[tw] OR "Hexahydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR "Hexahydro-1,3,5-trinitro-s-triazine"[tw] OR
Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-Triaza-
l,3,5-trinitrocyclohexane"[tw] OR "l,3,5-Trinitro-l,3,5-triazacyclohexane"[tw]
OR "l,3,5-Trinitrohexahydro-l,3,5-triazine"[tw] OR "1,3,5-Trinitrohexahydro-
s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-triazine"[tw] OR "Esaidro-
l,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-l,3,5-trinitro-
l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw] OR
Cyclotrimethylenenitramine[tw] OR Trimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] OR Trimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] OR Trinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) AND (("2012/ll/01"[Date - MeSH] :
"3000"[Date - MeSH]) OR ("2012/ll/01"[Date - Entrez] : "3000"[Date -
Entrez]) OR ("2012/ll/01"[Date - Create] : "3000"[Date - Create]))
138
PubMed
Date limit:
11/2013-
1/2015
("cyclonite"[nm] OR Cyclonite[tw] OR RDX[tw] OR
Cyclotrimethylenetrinitramine[tw] OR "cyclotrimethylene trinitramine"[tw] OR
"Hexahydro-l,3,5-trinitro-l,3,5-triazine"[tw] OR "Hexahydro-l,3,5-trinitro-
s-triazine"[tw] OR Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR
"l,3,5-Triaza-l,3,5-trinitrocyclohexane"[tw] OR "1,3,5-Trinitro-
l,3,5-triazacyclohexane"[tw] OR " 1,3,5-Trinitrohexahydro-1,3,5-triazine"[tw]
OR "l,3,5-Trinitrohexahydro-s-triazine"[tw] OR "1,3,5-Trinitroperhydro-
l,3,5-triazine"[tw] OR "Esaidro-1,3,5-trinitro-l,3,5-triazina"[tw] OR
"Hexahydro-1,3,5-trinitro-l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-
l,3,5-triazine"[tw] OR Cyclotrimethylenenitramine[tw] OR
Trimethylenetrinitramine[tw] OR "Trimethylene trinitramine"[tw] OR
Trimethyleentrinitramine[tw] OR "Trinitrocyclotrimethylene triamine"[tw] OR
Trinitrotrimethylenetriamine[tw] OR "CX 84A"[tw] OR Cyklonit[tw] OR
Geksogen[tw] OR Heksogen[tw] OR Hexogeen[tw] OR Hexolite[tw] OR "KHP
281"[tw] OR "PBX (af) 108"[tw] OR "PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw])
AND (2013/11/01: 3000[crdat] OR 2013/11/01: 3000[edat])
76
PubMed
Date limit:
11/2014-
5/2016
("cyclonite"[nm] OR Cyclonite[tw] OR RDX[tw] OR
Cyclotrimethylenetrinitramine[tw] OR "cyclotrimethylene trinitramine"[tw] OR
"Hexahydro-l,3,5-trinitro-l,3,5-triazine"[tw] OR "Hexahydro-l,3,5-trinitro-s-
triazine"[tw] OR Hexogen[tw] OR "1,3,5-trinitro-l,3,5-triazine"[tw] OR "1,3,5-
Triaza-l,3,5-trinitrocyclohexane"[tw] OR " 1,3,5-Trinitro-1,3,5-
triazacyclohexane"[tw] OR "1,3,5-Trinitrohexahydro-1,3,5-triazine"[tw] OR
"l,3,5-Trinitrohexahydro-s-triazine"[tw] OR "l,3,5-Trinitroperhydro-l,3,5-
triazine"[tw] OR "Esaidro-1,3,5-trinitro-l,3,5-triazina"[tw] OR "Hexahydro-
l,3,5-trinitro-l,3,5-triazin"[tw] OR "Perhydro-1,3,5-trinitro-l,3,5-triazine"[tw]
OR Cyclotrimethylenenitramine[tw] OR Trimethylenetrinitramine[tw] OR
"Trimethylene trinitramine"[tw] OR Trimethyleentrinitramine[tw] OR
"Trinitrocyclotrimethylene triamine"[tw] OR Trinitrotrimethylenetriamine[tw]
OR "CX 84A"[tw] OR Cyklonit[tw] OR Geksogen[tw] OR Heksogen[tw] OR
Hexogeen[tw] OR Hexolite[tw] OR "KHP 281"[tw] OR "PBX (af) 108"[tw] OR
"PBXW 108(E)"[tw] OR "Pbx(AF) 108"[tw]) AND (2014/11/01: 3000[crdat] OR
2014/11/01: 3000[edat])
118
This document is a draft for review purposes only and does not constitute Agency policy.
B-5	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits
Toxline
Date: 4/2012
Notes: Searched CASRN or synonyms; removed invertebrates, aquatic
organisms, amphibians, earthworms.
507
Toxline
Date limit:
2011-2/2013
@OR+("Cyclonite"+"RDX"+"Cyclotrimethylenetrinitramine"+"cyclotrimethylen
e trinitramine"+"Hexahydro-l,3,5-trinitro-l,3,5-triazine"+" Hexahydro-
l,3,5-trinitro-s-triazine"+"Hexogen"+"l,3,5-trinitro-
l,3,5triazine"+"l,3,5-Triaza-l,3,5-trinitrocyclohexane"+"l,3,5-Trinitro-
l,3,5-triazacyclohexane"+"l,3,5-Trinitrohexahydro-l,3,5-triazine"+
"l,3,5-Trinitrohexahydro-s-triazine"+@term+@rn+121-82-4)+
@AND+@range+yr+2011+2013+@NOT+@org+pubmed+pubdart+crisp+tscats
5

@OR+("l,3,5-Trinitroperhydro-l,3,5-triazine"+"Esaidro-l,3,5-trinitro-
1,3,5-triazina"+" Hexahydro-1,3,5-trinitro-l,3,5-triazin"+" Perhydro-
1,3,5-trinitro-l,3,5-triazine"+"Cyclotrimethylenenitramine"+
"Trimethylenetrinitramine"+"Trimethylene+trinitramine"+
"Trimethyleentrinitramine"+"Trinitrocyclotrimethylene+triamine"+
"Trinitrotrimethylenetriamine"+"CX+84A"+"Cyklonit"+"Geksogen"+
"Heksogen"+"Hexogeen"+"Hexolite"+"KHP+281")+@AND+@range+yr+2011+
2013+@NOT+@org+pubmed+pubdart+crisp+tscats
0
Toxline
Date limit:
2012-1/2014
@OR+("Cyclonite"+"RDX"+"Cyclotrimethylenetrinitramine"+"cyclotrimethylen
e trinitramine"+"Hexahydro-1,3,5-trinitro-l,3,5-triazine"+" Hexahydro-
1,3,5-trinitro-s-triazine"+" Hexogen"+"l,3,5-trinitro-l,3,5-triazine"+
"l,3,5-Triaza-l,3,5-trinitrocyclohexane"+"l,3,5-Trinitro-
l,3,5-triazacyclohexane"+"l,3,5-Trinitrohexahydro-l,3,5-triazine"+
"l,3,5-Trinitrohexahydro-s-triazine"+@term+@rn+121-82-4)+ @AND+
@range+yr+2012+2014+@NOT+@org+pubmed+pubdart+crisp+tscats
10

@OR+("l,3,5-Trinitroperhydro-l,3,5-triazine"+"Esaidro-l,3,5-trinitro-
1,3,5-triazina"+" Hexahydro-1,3,5-trinitro-l,3,5-triazin"+" Perhydro-
1,3,5-trinitro-l,3,5-triazine"+"Cyclotrimethylenenitramine"+
"Trimethylenetrinitramine"+"Trimethylene+trinitramine"+
"Trimethyleentrinitramine"+"Trinitrocyclotrimethylene+triamine"+
"Trinitrotrimethylenetriamine"+"CX+84A"+"Cyklonit"+"Geksogen"+
"Heksogen"+"Hexogeen"+"Hexolite"+"KHP+281")+@AND+@range+yr+2012+
2014+@NOT+@org+pubmed+pubdart+crisp+tscats
0
Toxline
Date limit:
2013-1/2015
@SYN0+@OR+(RDX+"cyclotrimethylene+trinitramine"+"l,3,5-trinitro-
l,3,5-triazine"+"l,3,5-Triaza-l,3,5-trinitrocyclohexane"+"l,3,5-Trinitro-
l,3,5-triazacyclohexane"+"l,3,5-Trinitrohexahydro-l,3,5-triazine"+
"l,3,5-Trinitrohexahydro-s-triazine"+"l,3,5-Trinitroperhydro-l,3,5-triazine"+
"CX+84A")+@AND+@range+yr+2012+2015+@NOT+@org+pubmed+pubdart+"
nih+reporter"+tscats+crisp
19

@SYN0+@OR+("Cyclonite"+"Cyclotrimethylenenitramine"+"Cyclotrimethylene
trinitramine"+"Hexahydro-1,3,5-trinitro-l,3,5-triazine"+"Hexahydro-
l,3,5-trinitro-s-triazine"+"Hexogen"+"Hexolite"+"KHP+281"+"PBX+(af)+108"+
"PBXW+108(E)"+"Pbx(AF)+108"+"Perhydro-1,3,5-trinitro-l,3,5-triazine")+
@AND+@range+yr+2012+2015+@NOT+@org+pubmed+pubdart+
"nih+reporter"+tscats+crisp
9

@SYN0+@OR+("Research+Development+Explosive"+"Royal+Demolition+eXpl
osive+"Trimethylenetrinitramine"+"Trinitrocyclotrimethylene+triamine"+"Trini
trotrimethylenetriamine"+"sym-Trimethylene+trinitramine"+@term+
0
This document is a draft for review purposes only and does not constitute Agency policy.
B-6	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

@rn+121-82-4+@term+@rn+204655-61-8+@term+@rn+50579-23-
2+@term+@rn+53800-53-6+@term+@rn+57608-45-4+@term+@rn+82030-
42-0)+ @AND+@range+yr+2012+2015+@NOT+@org+pubmed+pubdart+
"nih+reporter"+tscats+crisp

Toxline
Date limit:
2014-5/2016
@SYNO+@OR+(RDX+"cyclotrimethylene+trinitramine"+"l, 3,5-trinitro-l, 3,5-
triazine"+"l,3,5-Triaza-l,3,5-trinitrocyclohexane"+"l,3,5-Trinitro-l,3,5-
triazacyclohexane"+"l,3,5-Trinitrohexahydro-l,3,5-triazine"+"l,3,5-
Trinitrohexahydro-s-triazine"+"l,3,5-Trinitroperhydro-l,3,5-
triazine"+"CX+84A")+@AND+@range+yr+2014+2016+@NOT+@org+pubmed+
pubdart+"nih+reporter"+tscats+crisp
1

@SYNO+@OR+("Cyclonite"+"Cyclotrimethylenenitramine"+"Cyclotrimethylene
trinitramine"+"Hexahydro-1,3,5-trinitro-l, 3,5-triazine"+"Hexahydro-1,3,5-
trinitro-s-
triazine"+"Hexogen"+"Hexolite"+"KHP+281"+"PBX+(af)+108"+"PBXW+108(E)"
+"Pbx(AF)+108"+"Perhydro-l,3,5-trinitro-l,3,5-
triazine")+@AND+@range+yr+2014+2016+@NOT+@org+pubmed+pubdart+"n
ih+reporter"+tscats+crisp
0

@SYNO+@OR+("Research+Development+Explosive"+"Royal+Demolition+eXpl
osive+"Trimethylenetrinitramine"+"Trinitrocyclotrimethylene+triamine"+"Trini
trotrimethylenetriamine"+"sym-
Trimethylene+trinitramine"+@term+@rn+121-82-4+@term+@rn+204655-61-
8+@term+@rn+50579-23-2+@term+@rn+53800-53-6+@term+@rn+57608-
45-4+@term+@ rn+82030-42-
0)+@AND+@range+yr+2014+2016+@NOT+@org+pubmed+pubdart+"nih+rep
orter"+tscats+crisp
0
TSCATS
Date: 2/2013
@term+@rn+121-82-4+@AND+@org+tscats
4
TSCATS 2
Date: 5/2016
121-82-4 from EPA receipt date 01/01/2000
0
TSCATS
8e/FYI
Date: 5/2016
("121-82-4" OR "1,3,5-Triazine, hexahydro-1,3,5-trinitro-") tsca (8e OR FYI)
0
This document is a draft for review purposes only and does not constitute Agency policy.
B-7	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits
Toxcenter
Date: 4/2012
((121-82-4 OR Cyclonite OR RDX OR Cyclotrimethylenetrinitramine OR
"cyclotrimethylene trinitramine" OR "Hexahydro-1,3,5-trinitro-l,3,5-triazine"
OR "Hexahydro-1,3,5-trinitro-s-triazine" OR Hexogen OR "1,3,5-trinitro-
1,3,5-triazine" OR "l,3,5-Triaza-l,3,5-trinitrocyclohexane" OR "1,3,5-Trinitro-
1,3,5-triazacyclohexane" OR "1,3,5-Trinitrohexahydro-l,3,5-triazine" OR
"1,3,5-Trinitrohexahydro-s-triazine" OR "1,3,5-Trinitroperhydro-l,3,5-triazine"
OR "Esaidro-1,3,5-trinitro-l,3,5-triazina" OR "Hexahydro-1,3,5-trinitro-
1,3,5-triazin" OR "Perhydro-1,3,5-trinitro-l,3,5-triazine" OR
Cyclotrimethylenenitramine ORTrimethylenetrinitramine OR "Trimethylene
trinitramine" ORTrimethyleentrinitramine OR "Trinitrocyclotrimethylene
triamine" OR Trinitrotrimethylenetriamine OR "CX 84A" OR Cyklonit OR
Geksogen OR Heksogen OR Hexogeen OR Hexolite OR "KHP 281" OR "PBX (af)
108" OR "PBXW 108(E)" OR "Pbx(AF) 108") NOT (patent/dt OR tscats/fs))AND
((chronic OR immunotox? OR neurotox? OR toxicokin? OR biomarker? OR
neurolog? OR pharmacokin? OR subchronic OR pbpk OR epidemiology/st,ct,it)
OR acute OR subacute OR Id50# OR Ic50# OR (toxicity OR adverse OR
poisoning)/st,ct,it OR inhal? OR pulmon? OR nasal? OR lung? OR respir? OR
occupation? OR workplace? OR worker? OR oral OR orally OR ingest? OR
gavage? OR diet OR diets OR dietary OR drinking(w)water OR (maximum and
concentration? and (allowable OR permissible)) OR (abort? OR abnormalit? OR
embryo? OR cleft? OR fetus? OR foetus? OR fetal? OR foetal? OR fertil? OR
malform? OR ovum OR ova OR ovary OR placenta? OR pregnan? OR prenatal
OR perinatal? OR postnatal? OR reproduc? OR steril? OR teratogen? OR sperm
OR spermac? OR spermag? OR spermati? OR spermas? OR spermatob? OR
spermatoc? OR spermatog? OR spermatoi? OR spermatol? OR spermator? OR
spermatox? OR spermatoz? OR spermatu? OR spermi? OR spermo? OR
neonat? OR newborn OR development OR developmental? OR zygote? OR
child OR children OR adolescen? OR infant OR wean? OR offspring OR
age(w)factor? OR dermal? OR dermis OR skin OR epiderm? OR cutaneous? OR
carcinog? OR cocarcinog? OR cancer? OR precancer? OR neoplas? OR tumor?
OR tumour? OR oncogen? OR lymphoma? OR carcinom? OR genetox? OR
genotox? OR mutagen? OR genetic(w)toxic? OR nephrotox? OR hepatotox? OR
endocrin? OR estrogen? OR androgen? OR hormon? OR rat OR rats OR mouse
OR mice OR muridae OR dog OR dogs OR rabbit? OR hamster? OR pig OR pigs
OR swine OR porcine OR goat OR goats OR sheep OR monkey? OR macaque?
OR marmoset? OR primate? OR mammal? OR ferret? OR gerbil? OR rodent?
OR lagomorpha OR baboon? OR bovine OR canine OR cat OR cats OR feline OR
pigeon? OR occupation? OR worker? OR epidem?) AND ((biosis/fs AND
py>1999) OR caplus/fs))
Notes: Duplicates were removed; Biosis subfile results were date limited to
avoid extensive overlap with Toxline.
337 titles
screened
(20 selected for
full records and
added to HERO)
Toxcenter
Date limit:
1/2012-
2/2013
(((121-82-4 OR Cyclonite OR RDX OR Cyclotrimethylenetrinitramine OR
"cyclotrimethylene trinitramine" OR "Hexahydro-1,3,5-trinitro-l,3,5-triazine"
OR "Hexahydro-1,3,5-trinitro-s-triazine" OR Hexogen OR "1,3,5-trinitro-
1,3,5-triazine" OR "l,3,5-Triaza-l,3,5-trinitrocyclohexane" OR "1,3,5-Trinitro-
1,3,5-triazacyclohexane" OR "1,3,5-Trinitrohexahydro-l,3,5-triazine" OR
"1,3,5-Trinitrohexahydro-s-triazine" OR "1,3,5-Trinitroperhydro-l,3,5-triazine"
OR "Esaidro-1,3,5-trinitro-l,3,5-triazina" OR "Hexahydro-1,3,5-trinitro-
1,3,5-triazin" OR "Perhydro-1,3,5-trinitro-l,3,5-triazine" OR
Cyclotrimethylenenitramine ORTrimethylenetrinitramine OR "Trimethylene
26 titles screened
(6 selected for
full records and
added to HERO)
This document is a draft for review purposes only and does not constitute Agency policy.
B-8	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

trinitramine" ORTrimethyleentrinitramine OR "Trinitrocyclotrimethylene
triamine" OR Trinitrotrimethylenetriamine OR "CX 84A" OR Cyklonit OR
Geksogen OR Heksogen OR Hexogeen OR Hexolite OR "KHP 281" OR "PBX (af)
108" OR "PBXW 108(E)" OR "Pbx(AF) 108") NOT (patent/dt OR tscats/fs)) AND
(py>2012 OR ed>20120101)) AND (chronic OR immunotox? OR neurotox? OR
toxicokin? OR biomarker? OR neurolog? OR pharmacokin? OR subchronic OR
pbpk OR epidemiology/st,ct, it) OR acute OR subacute OR Id50# OR Ic50# OR
(toxicity OR adverse OR poisoning)/st,ct,it OR inhal? OR pulmon? OR nasal? OR
lung? OR respir? OR occupation? OR workplace? OR worker? OR oral OR orally
OR ingest? OR gavage? OR diet OR diets OR dietary OR drinking(w)water OR
(maximum and concentration? and (allowable OR permissible)) OR (abort? OR
abnormalit? OR embryo? OR cleft? OR fetus? OR foetus? OR fetal? OR foetal?
OR fertil? OR malform? OR ovum OR ova OR ovary OR placenta? OR pregnan?
OR prenatal OR perinatal? OR postnatal? OR reproduc? OR steril? OR
teratogen? OR sperm OR spermac? OR spermag? OR spermati? OR spermas?
OR spermatob? OR spermatoc? OR spermatog? OR spermatoi? OR spermatol?
OR spermator? OR spermatox? OR spermatoz? OR spermatu? OR spermi? OR
spermo? OR neonat? OR newborn OR development OR developmental? OR
zygote? OR child OR children OR adolescen? OR infant OR wean? OR offspring
OR age(w)factor? OR dermal? OR dermis OR skin OR epiderm? OR cutaneous?
OR carcinog? OR cocarcinog? OR cancer? OR precancer? OR neoplas? OR
tumor? OR tumour? OR oncogen? OR lymphoma? OR carcinom? OR genetox?
OR genotox? OR mutagen? OR genetic(w)toxic? OR nephrotox? OR hepatotox?
OR endocrin? OR estrogen? OR androgen? OR hormon? OR rat OR rats OR
mouse OR mice OR muridae OR dog OR dogs OR rabbit? OR hamster? OR pig
OR pigs OR swine OR porcine OR goat OR goats OR sheep OR monkey? OR
macaque? OR marmoset? OR primate? OR mammal? OR ferret? OR gerbil? OR
rodent? OR lagomorpha OR baboon? OR bovine OR canine OR cat OR cats OR
feline OR pigeon? OR occupation? OR worker? OR epidem?) AND (biosis/fs OR
caplus/fs))
Notes: Duplicates were removed.

Toxcenter
Date limit:
11/2012-
1/2014
(((121-82-4 OR Cyclonite OR RDX OR Cyclotrimethylenetrinitramine OR
"cyclotrimethylene trinitramine" OR "Hexahydro-1,3,5-trinitro-l,3,5-triazine"
OR "Hexahydro-1,3,5-trinitro-s-triazine" OR Hexogen OR "1,3,5-trinitro-
1,3,5-triazine" OR "l,3,5-Triaza-l,3,5-trinitrocyclohexane" OR "1,3,5-Trinitro-
1,3,5-triazacyclohexane" OR "1,3,5-Trinitrohexahydro-l,3,5-triazine" OR
"1,3,5-Trinitrohexahydro-s-triazine" OR "1,3,5-Trinitroperhydro-l,3,5-triazine"
OR "Esaidro-1,3,5-trinitro-l,3,5-triazina" OR "Hexahydro-1,3,5-trinitro-
1,3,5-triazin" OR "Perhydro-1,3,5-trinitro-l,3,5-triazine" OR
Cyclotrimethylenenitramine ORTrimethylenetrinitramine OR "Trimethylene
trinitramine" ORTrimethyleentrinitramine OR "Trinitrocyclotrimethylene
triamine" OR Trinitrotrimethylenetriamine OR "CX 84A" OR Cyklonit OR
Geksogen OR Heksogen OR Hexogeen OR Hexolite OR "KHP 281" OR "PBX (af)
108" OR "PBXW 108(E)" OR "Pbx(AF) 108") NOT (patent/dt OR tscats/fs)) AND
(py>2012 OR ed>20121101)) AND (chronic OR immunotox? OR neurotox? OR
toxicokin? OR biomarker? OR neurolog? OR pharmacokin? OR subchronic OR
pbpk OR epidemiology/st,ct, it) OR acute OR subacute OR Id50# OR Ic50# OR
(toxicity OR adverse OR poisoning)/st,ct,it OR inhal? OR pulmon? OR nasal? OR
lung? OR respir? OR occupation? OR workplace? OR worker? OR oral OR orally
OR ingest? OR gavage? OR diet OR diets OR dietary OR drinking(w)water OR
20 titles screened
(0 selected for
full records; none
added to HERO)
This document is a draft for review purposes only and does not constitute Agency policy.
B-9	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

(maximum and concentration? and (allowable OR permissible)) OR (abort? OR
abnormalit? OR embryo? OR cleft? OR fetus? OR foetus? OR fetal? OR foetal?
OR fertil? OR malform? OR ovum OR ova OR ovary OR placenta? OR pregnan?
OR prenatal OR perinatal? OR postnatal? OR reproduc? OR steril? OR
teratogen? OR sperm OR spermac? OR spermag? OR spermati? OR spermas?
OR spermatob? OR spermatoc? OR spermatog? OR spermatoi? OR spermatol?
OR spermator? OR spermatox? OR spermatoz? OR spermatu? OR spermi? OR
spermo? OR neonat? OR newborn OR development OR developmental? OR
zygote? OR child OR children OR adolescen? OR infant OR wean? OR offspring
OR age(w)factor? OR dermal? OR dermis OR skin OR epiderm? OR cutaneous?
OR carcinog? OR cocarcinog? OR cancer? OR precancer? OR neoplas? OR
tumor? OR tumour? OR oncogen? OR lymphoma? OR carcinom? OR genetox?
OR genotox? OR mutagen? OR genetic(w)toxic? OR nephrotox? OR hepatotox?
OR endocrin? OR estrogen? OR androgen? OR hormon? OR rat OR rats OR
mouse OR mice OR muridae OR dog OR dogs OR rabbit? OR hamster? OR pig
OR pigs OR swine OR porcine OR goat OR goats OR sheep OR monkey? OR
macaque? OR marmoset? OR primate? OR mammal? OR ferret? OR gerbil? OR
rodent? OR lagomorpha OR baboon? OR bovine OR canine OR cat OR cats OR
feline OR pigeon? OR occupation? OR worker? OR epidem?) AND (biosis/fs OR
caplus/fs))
Notes: Duplicates were removed.

Toxcenter
Date limit:
11/2013-
1/2015
(((121-82-4 OR Cyclonite OR RDX OR Cyclotrimethylenetrinitramine OR
"cyclotrimethylene trinitramine" OR "Hexahydro-1,3,5-trinitro-l,3,5-triazine"
OR "Hexahydro-1,3,5-trinitro-s-triazine" OR Hexogen OR "1,3,5-trinitro-
1,3,5-triazine" OR "l,3,5-Triaza-l,3,5-trinitrocyclohexane" OR "1,3,5-Trinitro-
1,3,5-triazacyclohexane" OR "1,3,5-Trinitrohexahydro-l,3,5-triazine" OR
"1,3,5-Trinitrohexahydro-s-triazine" OR "1,3,5-Trinitroperhydro-l,3,5-triazine"
OR "Esaidro-1,3,5-trinitro-l,3,5-triazina" OR "Hexahydro-1,3,5-trinitro-
1,3,5-triazin" OR "Perhydro-1,3,5-trinitro-l,3,5-triazine" OR
Cyclotrimethylenenitramine ORTrimethylenetrinitramine OR "Trimethylene
trinitramine" ORTrimethyleentrinitramine OR "Trinitrocyclotrimethylene
triamine" OR Trinitrotrimethylenetriamine OR "CX 84A" OR Cyklonit OR
Geksogen OR Heksogen OR Hexogeen OR Hexolite OR "KHP 281" OR "PBX (af)
108" OR "PBXW 108(E)" OR "Pbx(AF) 108") NOT (patent/dt OR tscats/fs)) AND
(py >2013 OR ed>20131101)) AND (chronic OR immunotox? OR neurotox? OR
toxicokin? OR biomarker? OR neurolog? OR pharmacokin? OR subchronic OR
pbpk OR epidemiology/st,ct, it) OR acute OR subacute OR Id50# OR Ic50# OR
(toxicity OR adverse OR poisoning)/st,ct,it OR inhal? OR pulmon? OR nasal? OR
lung? OR respir? OR occupation? OR workplace? OR worker? OR oral OR orally
OR ingest? OR gavage? OR diet OR diets OR dietary OR drinking(w)water OR
(maximum and concentration? and (allowable OR permissible)) OR (abort? OR
abnormalit? OR embryo? OR cleft? OR fetus? OR foetus? OR fetal? OR foetal?
OR fertil? OR malform? OR ovum OR ova OR ovary OR placenta? OR pregnan?
OR prenatal OR perinatal? OR postnatal? OR reproduc? OR steril? OR
teratogen? OR sperm OR spermac? OR spermag? OR spermati? OR spermas?
OR spermatob? OR spermatoc? OR spermatog? OR spermatoi? OR spermatol?
OR spermator? OR spermatox? OR spermatoz? OR spermatu? OR spermi? OR
spermo? OR neonat? OR newborn OR development OR developmental? OR
zygote? OR child OR children OR adolescen? OR infant OR wean? OR offspring
OR age(w)factor? OR dermal? OR dermis OR skin OR epiderm? OR cutaneous?
80 titles screened
(3 selected for
full records and
added to HERO)
This document is a draft for review purposes only and does not constitute Agency policy.
B-10	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

OR carcinog? OR cocarcinog? OR cancer? OR precancer? OR neoplas? OR
tumor? OR tumour? OR oncogen? OR lymphoma? OR carcinom? OR genetox?
OR genotox? OR mutagen? OR genetic(w)toxic? OR nephrotox? OR hepatotox?
OR endocrin? OR estrogen? OR androgen? OR hormon? OR rat OR rats OR
mouse OR mice OR muridae OR dog OR dogs OR rabbit? OR hamster? OR pig
OR pigs OR swine OR porcine OR goat OR goats OR sheep OR monkey? OR
macaque? OR marmoset? OR primate? OR mammal? OR ferret? OR gerbil? OR
rodent? OR lagomorpha OR baboon? OR bovine OR canine OR cat OR cats OR
feline OR pigeon? OR occupation? OR worker? OR epidem?) AND (biosis/fs OR
caplus/fs))
Note: Duplicates were removed.

Toxcenter
Date limit:
11/2014-
5/2016
(((121-82-4 OR Cyclonite OR RDX OR Cyclotrimethylenetrinitramine OR
"cyclotrimethylene trinitramine" OR "Hexahydro-1,3,5-trinitro-l,3,5-triazine"
OR "Hexahydro-1,3,5-trinitro-s-triazine" OR Hexogen OR "1,3,5-trinitro-l,3,5
triazine" OR "l,3,5-Triaza-l,3,5-trinitrocyclohexane" OR "l,3,5-Trinitro-l,3,5-
triazacyclohexane" OR "1,3,5-Trinitrohexahydro-l,3,5-triazine" OR "1,3,5-
Trinitrohexahydro-s-triazine" OR "1,3,5-Trinitroperhydro-1,3,5-triazine" OR
"Esaidro-l,3,5-trinitro-l,3,5-triazina" OR "Hexahydro-1,3,5-trinitro-l,3,5
triazin" OR "Perhydro-1,3,5-trinitro-l,3,5-triazine" OR
Cyclotrimethylenenitramine ORTrimethylenetrinitramine OR "Trimethylene
trinitramine" ORTrimethyleentrinitramine OR "Trinitrocyclotrimethylene
triamine" OR Trinitrotrimethylenetriamine OR "CX 84A" OR Cyklonit OR
Geksogen OR Heksogen OR Hexogeen OR Hexolite OR "KHP 281" OR "PBX (af)
108" OR "PBXW 108(E)" OR "Pbx(AF) 108") NOT (patent/dt OR tscats/fs)) AND
(py >2013 OR ed>20131101)) AND (chronic OR immunotox? OR neurotox? OR
toxicokin? OR biomarker? OR neurolog? OR pharmacokin? OR subchronic OR
pbpk OR epidemiology/st,ct, it) OR acute OR subacute OR Id50# OR Ic50# OR
(toxicity OR adverse OR poisoning)/st,ct,it OR inhal? OR pulmon? OR nasal? OR
lung? OR respir? OR occupation? OR workplace? OR worker? OR oral OR orally
OR ingest? OR gavage? OR diet OR diets OR dietary OR drinking(w)water OR
(maximum and concentration? and (allowable OR permissible)) OR (abort? OR
abnormalit? OR embryo? OR cleft? OR fetus? OR foetus? OR fetal? OR foetal?
OR fertil? OR malform? OR ovum OR ova OR ovary OR placenta? OR pregnan?
OR prenatal OR perinatal? OR postnatal? OR reproduc? OR steril? OR
teratogen? OR sperm OR spermac? OR spermag? OR spermati? OR spermas?
OR spermatob? OR spermatoc? OR spermatog? OR spermatoi? OR spermatol?
OR spermator? OR spermatox? OR spermatoz? OR spermatu? OR spermi? OR
spermo? OR neonat? OR newborn OR development OR developmental? OR
zygote? OR child OR children OR adolescen? OR infant OR wean? OR offspring
OR age(w)factor? OR dermal? OR dermis OR skin OR epiderm? OR cutaneous?
OR carcinog? OR cocarcinog? OR cancer? OR precancer? OR neoplas? OR
tumor? OR tumour? OR oncogen? OR lymphoma? OR carcinom? OR genetox?
OR genotox? OR mutagen? OR genetic(w)toxic? OR nephrotox? OR hepatotox?
OR endocrin? OR estrogen? OR androgen? OR hormon? OR rat OR rats OR
mouse OR mice OR muridae OR dog OR dogs OR rabbit? OR hamster? OR pig
OR pigs OR swine OR porcine OR goat OR goats OR sheep OR monkey? OR
macaque? OR marmoset? OR primate? OR mammal? OR ferret? OR gerbil? OR
rodent? OR lagomorpha OR baboon? OR bovine OR canine OR cat OR cats OR
feline OR pigeon? OR occupation? OR worker? OR epidem?) AND (biosis/fs OR
caplus/fs))
33 titles screened
(0 selected for
full records and
added to HERO)
This document is a draft for review purposes only and does not constitute Agency policy.
B-ll	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

Note: Duplicates were removed.

1
2	Table B-2. Summary of detailed search strategies for RDX (DTIC)
Database
Terms
Hits
DTIC Online
Access
Controlled
Date
searched:
1/2015
Synonyms in all fields search box
("121-82-4" OR "RDX" OR "Cyclotrimethylenetrinitramine" OR "Cyclonite" OR
"cyclotrimethylene trinitramine" OR "Hexogen" OR "Hexahydro-1,3,5-
trinitro-1,3,5-triazine" OR "Hexahydro-1,3,5-trinitro-s-triazine" OR
"Trimethylene trinitramine" OR "Trimethylenetrinitramine" OR "Hexolite" OR
"Trinitrotrimethylenetriamine")
Keywords in citation box
("toxicity" OR "toxicology" OR "poisoning" OR "cancer" OR "carcinogens" OR
"carcinogen" OR "neoplasms" OR "neoplasm" OR "oncogenesis" OR
"teratogenic compounds" OR "lethality" OR "death" OR "body weight" OR
"immunology" OR "genotoxicity" OR "mutagenicity" OR "mutagens" OR
"mutations" OR "oral" OR "gavage" OR "inhalation" OR "dermal" OR
"metabolism" OR "pharmacokinetics" OR "pharmacokinetic" OR "PBPK" OR
"pharmacology" OR "organs" OR "skin" OR "tissues" OR "body fluids" OR
"toxic agents" OR "rats" OR "mice" OR "mouse" OR "rat")
Limited to
Content type: Documents


Distribution: Approved for Public Release
826 (85 selected
and added to HERO)

Distribution: U.S. Gov't and Contractors
239 (0 selected and
added to HERO)

Distribution: U.S. Gov't Only
199 (0 selected and
added to HERO)
DTIC Online
Access
Controlled
Date
searched:
5/2016
Synonyms in all fields search box
("121-82-4" OR "RDX" OR "Cyclotrimethylenetrinitramine" OR "Cyclonite" OR
"cyclotrimethylene trinitramine" OR "Hexogen" OR "Hexahydro-1,3,5-
trinitro-1,3,5-triazine" OR "Hexahydro-1,3,5-trinitro-s-triazine" OR
"Trimethylene trinitramine" OR "Trimethylenetrinitramine" OR "Hexolite" OR
"Trinitrotrimethylenetriamine")
Keywords in citation box
("toxicity" OR "toxicology" OR "poisoning" OR "cancer" OR "carcinogens" OR
"carcinogen" OR "neoplasms" OR "neoplasm" OR "oncogenesis" OR
"teratogenic compounds" OR "lethality" OR "death" OR "body weight" OR
"immunology" OR "genotoxicity" OR "mutagenicity" OR "mutagens" OR
"mutations" OR "oral" OR "gavage" OR "inhalation" OR "dermal" OR
"metabolism" OR "pharmacokinetics" OR "pharmacokinetic" OR "PBPK" OR
"pharmacology" OR "organs" OR "skin" OR "tissues" OR "body fluids" OR
"toxic agents" OR "rats" OR "mice" OR "mouse" OR "rat")
Limited to Content type: Documents

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

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Database
Terms
Hits

Distribution: Approved for Public Release
21 (0 selected and
added to HERO)
Distribution: U.S. Gov't and Contractors
9 (0 selected and
added to HERO)
Distribution: U.S. Gov't Only
9 (0 selected and
added to HERO)
1	Table B-3. Processes used to augment the search of core databases for RDX
Selected key reference(s) or sources
Date
Additional references
identified
"Forward" and "backward" Web of Science searches3
Sweenev et al. (2012a). Assessing the non-cancer risk for RDX (hexahvdro-
1,3,5-trinitro-l,3,5-triazine) using physiologically based pharmacokinetic (PBPK)
modeling. Regul Toxicol Pharmacol 62(1):107-114. (forward search)
1 search result
3/2013
0 citations added
Sweenev et al. (2012b). Cancer mode of action, weight of evidence, and
proposed cancer reference value for hexahydro-1,3,5-trinitro-l,3,5-triazine
(RDX). Regul Toxicol Pharmacol 64(2):205-224 (backwards search)
0 search results
3/2013
0 citations added
Sweenev et al. (2012a). Assessing the non-cancer risk for RDX (hexahvdro-
1,3,5-trinitro-l,3,5-triazine) using physiologically based pharmacokinetic (PBPK)
modeling. Regul Toxicol Pharmacol 62(1):107-114.
(review of 35 references cited in this paper)
3/2013
0 citations added
Sweenev et al. (2012b). Cancer mode of action, weight of evidence, and
proposed cancer reference value for hexahydro-1,3,5-trinitro-l,3,5-triazine
(RDX). Regul Toxicol Pharmacol 64(2):205-224
(review of 69 references cited in this paper)
3/2013
3 citations added
Online regulatory sources
Combination of Chemical Abstracts Registry Number (CASRN) and synonyms
searched on the following websites:
Agency for Toxic Substances and Disease Registry (ATSDR)
htto://www.atsdr.cdc.gov/substances/index.aso
(Note: the reference list for the ATSDR toxicological profile for RDX was
compared to the search results and relevant references were added)
California Environmental Protection Agency (Office of Environmental Health
Hazard Assessment) (htto://www. oehha.ca.gov/risk. html)
eChemPortal
(httD://www.echemDortal.org/echemDortal/DarticiDant/Dage.action?DagelD=9)
4/2012
1/2014
1/2015
5/2016
15 citations added
1 citation added
0 citations added
0 citations added
EPA Acute Exposure Guideline Levels
(htto://www. eoa.gov/oDDt/aegl/Dubs/chemlist. htm)
(htto://www.eoa.gov/ncea/iris/index.html) to find data
EPA National Service Center for Environmental Publications (NSCEP)
(htto://www. eoa.gov/nceoihom/)
EPA Science Inventory


This document is a draft for review purposes only and does not constitute Agency policy.
B-13	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Selected key reference(s) or sources
Date
Additional references
identified
(http://cfpub.epa.gov/si/)
Federal Docket
www.regulations.gov
Health Canada First Priority List Assessments
(httD://www.hc-sc.gc.ca/ewh-semt/Dubs/contaminants/Dsll-lsDl/index-


eng.php)
Health Canada Second Priority List Assessments
(httD://www.hc-sc.gc.ca/ewh-semt/Dubs/contaminants/Dsl2-lsD2/index-
eng.php)
International Agency for Research on Cancer (IARC)
(httD://monograohs. iarc.fr/htdig/search. html)
International Programme on Chemical Safety (IPCS) INCHEM
(htto://www. inchem.org/)
National Academy of Science (NAS) via the National Academies Press
(htto://www. nao.edu/)
National Cancer Institute (NCI)
(htto://www. cancer.gov)
National Center for Toxicological Research (NCTR)
(httD://www.fda.gov/AboutFDA/CentersOffices/OC/OfficeofScientificand Medic
alPrograms/NCTR/default.htm)
National Institute of Environmental Health Sciences (NIEHS)
(htto://www. niehs.nih.gov/)
National Institute for Occupational Safety and Health (NIOSH) NIOSHTIC 2
(htto://www2a. cdc.gov/nioshtic-2/)
National Toxicology Program (NTP)—RoC, status, results, and management
reports
(htto://ntDsearch.niehs. nih.gov/auerv. html)
World Health Organization (WHO) assessments—Concise International
Chemical Assessment Documents (CICADs), Environmental Health Criteria
(EHC)
(httD://www.who.int/iDcs/assessment/en/)
1
2	aSweenev et al. (2012a) and Sweeney et al. (2012b) were selected for forward and backward searching in the Web
3	of Science as the two more recent reviews of the health effects of RDX toxicity in the published literature.
This document is a draft for review purposes only and does not constitute Agency policy.
B-14	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
APPENDIX C. INFORMATION IN SUPPORT OF
HAZARD IDENTIFICATION AND DOSE-RESPONSE
ANALYSIS
C.l. TOXICOKINETICS
Hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) is absorbed following exposure by inhalation
and oral routes. The rate and extent of absorption are dependent upon the dosing preparation.
RDX is systemically distributed, can be transferred from mother to fetus, and can transfer in
maternal milk. Metabolism of RDX is extensive and includes denitration, ring cleavage, and
generation of C02 possibly through cytochrome P450 (CYP450). RDX metabolites are eliminated
primarily via urinary excretion and exhalation of C02.
C.l.l. Absorption
Absorption of RDX following oral exposure has been demonstrated in humans and
laboratory animals (rats, mice, swine, and voles) through measurement of radiolabeled RDX and/or
metabolites in excreta (urine and expired air) and tissues (including blood). Quantitative estimates
of oral absorption (e.g., oral bioavailability or fractional absorption) are not available in humans.
Results of animals studies indicate that oral bioavailability ranges from approximately 50 to 90%
and may vary based on the physical form of RDX and the matrix (e.g., soil, plants) in which it is
administered. Studies investigating absorption of RDX following inhalation exposure were not
identified. Results of an intratracheal administration study in rats provide limited evidence of
absorption of RDX from the respiratory tract. The only data evaluating dermal absorption of RDX is
provided by in vitro studies showing that RDX can be absorbed through excised skin of humans and
animals.
Oral Absorption
Quantitative information on blood levels following accidental exposure to RDX is limited to
two studies of accidental oral exposures fKuctikardali etal.. 2003: Woody etal.. 19861 and one
study of mixed dermal and inhalation exposure fOzhan etal.. 20031. A number of qualitative case
studies of accidental exposures with similar toxic effects provide additional support that RDX is
absorbed into the body (Hettand Fichtner. 2002: Harrell-Bruder and Hutchins. 1995: Goldberg et
al.. 1992: Ketel and Hughes. 1972: Hollander and Colbach. 1969: Stone etal.. 19691. The oral
absorption of RDX in humans was demonstrated in a case report of a 3-year-old male child who
ingested plasticized RDX material that adhered to his mother's work boots and clothing (Woody et
This document is a draft for review purposes only and does not constitute Agency policy.
C-l	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
al.. 19861. RDX was measured in serum, urine, cerebrospinal fluid, and feces. Based on a kinetic
analysis of the serum RDX concentrations, the dose was estimated to be 85 mg/kg and the first-
order absorption rate constants were estimated to be 0.34-2.20 hour-1 fWoodv et al.. 198612.
Sweeney etal. (2012a) estimated the absorption rate constant for this same subject to be
0.060 hour-1. The large range in the calculated absorption rate constants resulted from uncertainty
in the dose and time to peak serum RDX levels, and the models that were used to simulate the RDX
toxicokinetics. Ozhan etal. (2003) summarized plasma RDX levels in five military personnel who
were accidentally exposed to toxic levels of RDX. Although Ozhan etal. (2003) reported that
personnel were exposed through dermal and inhaled routes, secondary oral exposure may have
occurred. Based on physiologically based pharmacokinetic (PBPK) model fits to the plasma RDX
concentration data, Sweeney etal. (2012a) estimated a first-order absorption rate constant of
0.033 hour-1. Kiiciikardali et al. (2003) summarized plasma RDX levels in five military personnel
who ingested toxic levels of RDX (doses were not reported). RDX was detected in plasma of all
patients within 3 hours after ingestion.
Quantitative data to directly support estimates of oral bioavailability are available from
studies in rats and mice (Guo etal.. 1985: Schneider etal.. 1978.19771. Results of single and
repeated oral dose studies in adult Sprague-Dawley rats indicate that approximately 83-87% of the
administered dose is absorbed from the gastrointestinal (GI) tract Following gavage
administration of 50 mg/kg [14C]-RDX dissolved in dimethylsulfoxide (DMSO), approximately 90%
of the administered carbon-14 was recovered 4 days after dosing, with ~3% in feces, 34% in urine,
43% in expired air, and 10% in the carcass (Schneider et al.. 1977). It is unclear if the carcass
includes the GI tract, which may include unabsorbed RDX. Assuming that all of the carbon-14 in
feces represents unabsorbed RDX (rather than RDX that was absorbed and subsequently secreted
into the intestine), results of this study indicate that at least 87% of the administered dose was
absorbed from the GI tract Similar results were observed following repeated daily oral exposure of
Sprague-Dawley rats to [14C]-RDX by gavage (in DMSO) or drinking water for 1 week. Based on
recovery of carbon-14 in urine and expired air and the carbon-14 retained in carcass,
approximately 83% (drinking water) to 85 % (gavage) of the administered dose was absorbed
(Schneider etal.. 1978).
An estimate of oral bioavailability in rats can also be obtained from data on blood RDX
concentrations reported in Krishnan etal. f20091. Male Sprague-Dawley rats received a single
intravenous (i.v.) (0.77 or 1.04 mg/kg) or oral (1.53 or 2.07 mg/kg, dissolved in water) dose of RDX.
Estimates of bioavailability were obtained based on the reported blood RDX concentrations,
terminal elimination rate constants (estimated for this review by fitting the serum RDX data with a
first-order exponential function, see Table C-5 in Section C.1.4, Excretion) and the blood area under
the curve (AUC) values (calculated for this review using the trapezoid rule extrapolated to infinite
2Woodv et al. (19861 reported the absorption rate constants in units of L/hour; however, this appears to have
been a typographical error for l/hour or hour-1.
This document is a draft for review purposes only and does not constitute Agency policy.
C-2	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
time). Calculated dose-adjusted AUC values were 9.6 and 8.4 hours-kg/L for the i.v. studies and
4.7 and 6.0 hours-kg/L for the oral dosing studies. These AUC values correspond to estimated oral
bioavailability ranging from 50 to 70%. Recovery of administered radiolabel was incomplete
(~90% of the administered carbon-14) in the studies f Schneider et al.. 1978.1977): therefore, it is
possible that oral bioavailability is actually higher than 83-87%. Guo etal. Q9851 reported data on
blood tritium kinetics in mice that received i.v. (0.055 mg RDX or ~2.5 mg/kg body weight) or oral
(50 mg/kg) doses of [3H]-RDX. Based on the reported blood tritium concentrations (% of dose/g)
and terminal ti/2 values for concentrations of tritium in blood (1.1 days for i.v. and 2.2 days for
oral), the corresponding AUCs of the blood concentration versus time curves were calculated
(calculated for this review using the trapezoid rule extrapolated to infinite time) to be 30 and
16 hours-% dose/g for i.v. and oral dosing, respectively. This corresponds to an oral bioavailability
of RDX-derived tritium concentration of approximately 50% (i.e., 16/30).
In Yucatan miniature swine administered a single dose of [14C]-RDX (43-45 mg/kg as a
suspension in carboxymethylcellulose), approximately 0.8-6% of the administered carbon-14 was
eliminated in feces 24 hours after dosing (Musick etal.. 2010: Major etal.. 20071. Although results
of swine studies suggest that GI absorption of RDX was nearly complete, data cannot be used to
determine a quantitative estimate of oral bioavailability because it is unlikely that fecal excretion of
unabsorbed RDX was complete 24 hours after dosing fSnoeck et al.. 2004).
Oral bioavailability of RDX appears to vary depending upon the physical form of RDX and
the matrix (e.g., soil, vegetation) in which it is administered. Schneider et al. (1977) compared the
oral absorption of a single 100 mg/kg gavage dose of coarse granular [14C]-RDX as a slurry in
isotonic saline with a single 50 mg/kg gavage dose of a finely powdered [14C]-RDX solution in saline
in Sprague-Dawley rats. Plasma carbon-14 levels were measured for 24 hours after dosing. For
both [14C]-RDX preparations, peak plasma levels of carbon-14 were observed 24 hours after oral
administration, with a higher 24-hour plasma concentration for the 50 mg/kg dose (~4.7 [ig/mL)
compared to the 100 mg/kg dose (3.12 [ig/mL). Results of this study indicate that the oral
bioavailability of RDX may be greater for the finely powdered preparation than for the coarse
granular preparation consistent with a greater surface area available for absorption with finely
powdered RDX. However, blood levels were only measured 24 hours after dosing, and the lower
24-hour carbon-14 plasma concentration for the coarse granular preparation could be due to
slower absorption of coarse RDX granules compared with fine RDX powder, rather than lower
overall bioavailability.
Oral bioavailability of RDX is lower when administered as RDX-contaminated soil or when
RDX is in plant materials that were grown on RDX-contaminated soils. Crouse etal. (2008)
investigated the oral bioavailability of RDX in contaminated soils relative to pure RDX by
comparison of the AUC for the RDX blood concentration versus time curves. Adult male
Sprague-Dawley rats were administered oral doses (in gelatin capsules) of pure RDX (99.9% purity;
neat) or an equivalent amount of RDX in contaminated soils from the Louisiana Army Ammunition
This document is a draft for review purposes only and does not constitute Agency policy.
C-3	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Plant (AAP) or Fort Meade. Blood concentrations for rats dosed with Louisiana AAP soil
(1.24 mg/kg) and neat RDX (1.24 mg/kg) peaked at approximately 6 hours. The AUC and 6-hour
RDX blood concentration were both approximately 25% lower for Louisiana AAP soil than for neat
RDX (p < 0.003 for AUC), suggesting that oral bioavailability of RDX from Louisiana AAP soil was
25% lower than neat RDX. For Fort Meade soil (0.2 mg/kg), RDX blood concentrations peaked at
6 hours compared to 4 hours for neat RDX (0.2 mg/kg). The 4-hour blood concentration for Fort
Meade soil was approximately 15% lower than for neat RDX, although the AUC for Fort Meade soil
was only 5% lower than for neat RDX (not statistically significant). Collectively, these results
suggest that RDX in soil is absorbed following oral exposure and that it has a lower bioavailability
than neat RDX.
Fellows etal. (2006) showed that plants (alfalfa shoots and corn leaves) incorporated
[14C]-RDX grown on [14C]-RDX-amended soils. [14C]-RDX and plant metabolites of [14C]-RDX were
absorbed by voles following oral administration (Fellows etal.. 20061. In adult male prairie voles
[Microtus ochrogaster) fed diets containing RDX incorporated in plants for 5 or 7 days (average RDX
dose per animal of 2.3 mg/kg-day), 94.8 and 96.6% (respectively) of the administered carbon-14
was eliminated in excreta (combined feces, urine, and CO2) and 3-5% was retained in the carcass.
Feces, urine, and CO2 contained 74-79,13-14, and 8-12% of the total carbon-14 in excreta,
respectively. Based on carbon-14 elimination in urine and CO2 plus that retained by the carcass, the
study authors estimated the oral bioavailability of plant-derived RDX to be >20%. However, if
biliary excretion of RDX and/or RDX metabolites is a major excretory pathway in voles (as is the
case with mice), estimates of bioavailability of plant-derived RDX could be substantially higher.
In Yorkshire piglets administered single doses of 5 or 10 mg/kg in gelatin capsules, peak
plasma concentrations were proportional to the administered dose (Bannon. 2006). However, the
potential for dose-dependence has not been evaluated over a wide range of doses.
RDX appears in blood within 1 hour following oral dosing; however, the rate of absorption
may depend upon the physical form or dose of RDX fBannon et al.. 2009a: Crouse etal.. 2008:
Bannon. 2006: Guo etal.. 1985: MacPhail etal.. 1985: Schneider et al.. 1977). Oral absorption of
RDX was rapid in LACA mice following stomach perfusion with [3H]-RDX (50 mg/kg in methyl
cellulose) (Guo etal.. 1985). The tritium radiolabel was detected in blood 15 minutes following
dosing and the highest concentrations in blood were observed 30 minutes after dosing. Based on
an analysis of the blood tritium concentration kinetics, the authors estimated an absorption rate
constant of 8.7 hour"1. In Sprague-Dawley rats administered single doses (0.2-18.0 mg/kg) of RDX
in gelatin capsules, peak blood RDX concentrations were observed between 2.5 and 6 hours
(Bannon et al.. 2009a: Krishnan etal.. 2009: Crouse etal.. 2008). Peak blood concentrations
occurred at 24 hours after Sprague-Dawley rats were administered a single oral dose (100 mg/kg)
of coarse granular RDX in saline (Schneider etal.. 1977). Similarly, peak RDX blood concentrations
in swine administered single doses (5-10 mg/kg) of finally powdered (>98% pure) RDX in gelatin
capsules occurred at 3-8 hours after dosing fBannon etal.. 2009al. compared to 24 hours after a
This document is a draft for review purposes only and does not constitute Agency policy.
C-4	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
single dose (100 mg/kg) of RDX administered as a finely powdered in saline (Bannon et al.. 2009a:
Schneider et al.. 19771. Peak plasma concentrations in Yucatan miniature swine administered a
single dose of [14C]-RDX (45 mg/kg as a suspension in carboxymethylcellulose) were reached
within 6-12 hours after dosing fMusick etal.. 20101. Krishnan etal. (20091 and Sweeney et al.
f2012al estimated absorption rates in rats dosed with higher doses of coarse granular RDX to be
approximately 100 times slower than absorption rates in rats dosed with lower doses of finely
powdered neat RDX preparations or neat RDX dissolved in aqueous vehicles. For example,
Krishnan etal. (20091 estimated the absorption rate constant to be 0.75 hour"1 for rats dosed with
neat RDX dissolved in water (1.53 or 2.07 mg/kg) or neat RDX in a gelatin capsule (0.2 or
1.24 mg/kg) fCrouse etal.. 20081. compared to 0.0075 hour-1 for rats dosed with coarse granular
RDX (100 mg/kg) fSchneider et al.. 19771.
Inhalation Absorption
Studies evaluating absorption of RDX in humans following inhalation exposure were not
identified. Several case reports have documented seizures and other neurological effects in
individuals exposed to RDX either in a manufacturing setting or in the course of using RDX as a
cooking fuel fTestud etal.. 1996a: Testud etal.. 1996b: Ketel and Hughes. 1972: Hollander and
Colbach. 1969: Kaplan etal.. 1965: Barsotti and Crotti. 19491. These reports suggest that RDX may
be absorbed by the respiratory system. However, in several cases, the study authors were unable
to clearly identify the primary route of exposure. In some cases, incidental oral exposure was
suggested. Studies in laboratory animals have not investigated the absorption of RDX following
inhalation exposure.
Dermal Absorption
In vitro studies have demonstrated the dermal absorption of RDX in human and pig skin
fReddv etal.. 2008: Re i fen rath etal.. 20081. Reddvetal. (20081 reported that 5.7% of the applied
RDX dose (in acetone) was absorbed into excised human skin in 24 hours. Dermal absorption of
[14C]-RDX from both a low-carbon (1.9%) and a high-carbon (9.5%) soil was also assessed in this
system. Approximately 2.6% of the RDX applied in the low-carbon soil and 1.4% applied in the
high-carbon soil was absorbed after 24 hours. Thus, the dermal absorption of RDX from soils was
reduced when compared with absorption from acetone, and absorption was lower in the
high-carbon soil than in the low-carbon soil.
Reifenrath etal. (20081 investigated the influence of skin surface moisture conditions, soil
carbon content, and soil aging on the in vitro percutaneous penetration of [14C]-labeled RDX in
excised pig skin. Mean skin absorption of RDX was higher for low-carbon soil (1.2%), regardless of
soil age and skin surface moisture. Absorption and evaporation were <1% for RDX regardless of
soil type and age. While dermal absorption of certain munitions (e.g., 2,6-dinitrotoluene) is greatly
enhanced by hydration of the skin surface, hydration had minimal effect on RDX, mostly due to the
lack of RDX volatility (e.g., <0.5% evaporation).
This document is a draft for review purposes only and does not constitute Agency policy.
C-5	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
C.1.2. Distribution
Information on the distribution of absorbed RDX in humans is limited to a few cases of
accidental exposures to RDX that provide data on the kinetics of RDX in blood and cerebrospinal
fluid fKiiciikardali etal.. 2003: Ozhan etal.. 2003: Woody etal.. 19861. Concentrations of RDX in
serum and cerebrospinal fluid were similar (11 and 9 mg/L, respectively) in a child 24 hours after
ingesting an estimated dose of 85 mg/kg RDX (Woody etal.. 19861. More extensive information on
tissue distribution is available for animals, including mice, rats, and swine (Musick etal.. 2010:
Bannon. 2006: Reddv etal.. 1989: Guo etal.. 1985: MacPhail etal.. 1985: Schneider etal.. 19771. In
these studies, RDX or radiolabeled RDX ([14C] or [3H) was administered by the oral, intraperitoneal
(i.p.), i.v., or intratracheal route and the distribution of the RDX or radiolabel was measured. Since
metabolism of RDX can result in loss of carbon-14 or tritium from the parent compound, the
distribution of radiolabel will not necessarily reflect the distribution of RDX f Schneider etal.. 19771.
To compare tissue distributions in studies in which animals received different doses by different
routes of administration, distribution data are expressed as ratios of tissue RDX or radiolabel to
that of either whole blood or plasma, whichever was reported. RDX in blood distributes into red
blood cells (RBCs) and plasma to achieve concentration ratios that are close to unity. The
plasma:whole blood carbon-14 ratio in swine that received a single oral dose of [14C]-RDX
(45 mg/kg) was approximately 1.3 fMusick etal.. 20101. and whole rat blood incubated in vitro
with RDX had a plasma:RBC RDX ratio of approximately 1.0 (Krishnan et al.. 20091. As a result of
the similarity between plasma and whole blood concentrations, tissue distribution is approximately
equivalent when expressed as ratios of blood or plasma.
Studies conducted in rats, mice, and swine indicate that absorbed RDX distributes to many
different tissues. Schneider et al. T19771 estimated the volume of distribution of RDX to be
approximately 2.18 L/kg in rats, based on plasma RDX kinetics in rats that received a single i.p.
dose of RDX (5-6 mg/kg). Consistent with this estimate are observations of tissue:blood (or
plasma) concentration ratios that exceed 1 in various tissues, including brain (showing that RDX
can cross the blood:brain barrier), heart, kidney, and liver (Musick etal.. 2010: Bannon etal.. 2006:
MacPhail etal.. 1985: Schneider etal.. 19771. Distribution within the brain may not be uniform.
However, Bannon etal. (20061 observed tissue:blood concentrations for RDX of approximately 4 in
brain hippocampus and 3.5 in brain cortex of swine that received a single oral dose of 10 mg/kg
[14C]-RDX, although this is the only study that reported distribution for brain regions. Reported
tissue:blood (or plasma) concentration ratios of RDX 24 hours following a single dose (oral or i.p.)
were 1-9 for kidney, 1-7 for liver, and 1-3 for heart (Table C-l) (Bannon. 2006: Schneider etal..
19771. With repeated oral dosing (e.g., 30-90 days), tissue:blood ratios of RDX for these tissues are
consistently greater than unity (Schneider et al.. 19781. There is no consistent evidence that RDX
accumulates in fat, although estimates of the fat:blood partition coefficient range from 6 to 8 and
exceed that of other tissues f Sweeney etal.. 2012a: Krishnan et al.. 20091.
This document is a draft for review purposes only and does not constitute Agency policy.
C-6	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-l. Distribution of RDX or radiolabel from administered RDXa
Animal
Route
Dose (mg/kg)
Time (hrs)
Brain
Heart
Kidney
Liver
Fat
Source
Swine
Oral
45b
24
0.6°
0.7
2.4
7.3
0.4
Musick et al. (2010)
Swine
Oral
10d
3
3.5-4.0d
2
<1
<1
NAg
Bannon et al. (2006)
Swine
Oral
100d
24
1.5°
1.1
1.2-1.9
0.9
1.8
Schneider et al. (1977)
Rat
Oral
100d
24
3.4°
2.9
6.6
0.7
NA
Schneider et al. (1977)
Rat
i.p.
50d
2
3.4°
2.6
8.8
5.7
NA
Schneider et al. (1977)
Rat
i.p.
500d
<6.5
2.5°
2.1
4.8
3.3
NA
Schneider et al. (1977)
Mouse
Oral
50e
24
lc
0.8
1
1.4
0.8
Guo et al. (1985)
Mouse
i.v.
2.5e
24
0.6f
0.8
0.7
1.6
0.4
Guo et al. (1985)
2
3	aValues are tissue:blood or tissue:plasma ratios following a single dose of either RDX, [14C]-RDX, or [3H]-RDX.
4	bCarbon-14
5	cTissue:plasma
6	dRDX
7	eTritium
8	tissueiblood
9	gNot available
10
11	In rats, RDX can cross the placental:blood barrier resulting in exposure to the fetus, and can
12	also be transported into maternal milk. Hess-Ruth et al. (2007) detected RDX in the brain tissue of
13	postnatal day (PND) 1 rat pups (concentrations ranged from 0.64 to 7.6 ng/g brain tissue, with no
14	differences between males and females) after maternal exposure to 6 mg/kg RDX via gavage from
15	gestational day (GD) 6 to PND 10. RDX was also detected in maternal milk (concentrations ranged
16	from 3 to 5.7 ng/mL on PND 1 and from 0.7 to 3.1 ng/mL on PND 10).
17	C.1.3. Metabolism
18	The metabolism of RDX is not well characterized. No studies investigating the metabolism
19	of RDX in humans were identified. Studies in animals indicate that RDX undergoes extensive
20	metabolism, including denitration, ring cleavage, and generation of CO2. Predominant metabolic
21	pathways and major organs involved in RDX metabolism have not been identified, although results
22	of in vitro studies suggest a role for CYP450.
23	RDX undergoes metabolism through processes that generate CO2. In Sprague-Dawley rats
24	administered a single 50 mg/kg gavage dose of [14C]-RDX, 43% was recovered as exhaled [14CC>2]
25	after 4 days f Schneider etal.. 19771. Similarly, approximately 30-50% of the radioactivity was
26	recovered as exhaled [14CC>2] in rats administered [14C]-RDX in saturated drinking water or daily
27	gavage for up to 3 months f Schneider etal.. 1978). Metabolism of RDX to CO2 was also observed in
28	prairie voles following dietary exposure (average RDX dose per animal of 2.3 mg/kg-day) to
This document is a draft for review purposes only and does not constitute Agency policy.
C-7	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
[14C]-RDX incorporated plant materials for 5-7 days, with approximately 9% of the administered
[14C]-RDX dose eliminated as exhaled [14CC>2] fFellows etal.. 20061.
Terminal metabolites of RDX have been identified in the urine of rats and swine, with very
little urinary excretion of parent compound, indicating extensive metabolism of RDX. Following
oral administration of a single 50 mg/kg gavage dose of [14C]-RDX, 3.6% of the urinary radioactivity
was identified as unmetabolized RDX f Schneider etal.. 19771. Total urinary radiolabel accounted
for about one-third of the administered label and unmetabolized RDX contributed 3-5% of total
urinary radioactivity in rats exposed to [14C]-RDX-saturated drinking water for 1 or 13 weeks
(Schneider etal.. 19781. Similar results were observed in Yucatan swine administered a single
45 mg/kg oral dose of [14C]-RDX, with approximately 1-3.5% of the urinary radioactivity as parent
RDX (Major etal.. 20071. Urinary metabolites were not characterized in these studies f Schneider et
al.. 1978.19771. However, Schneider et al. Q9781 cited unpublished findings in their laboratory
that, in addition to carbon dioxide, other one-carbon intermediates were produced, including
bicarbonate and formic acid.
In the environment, the predominant breakdown products of RDX are methylene
dinitramine and 4-nitro-2-diazbutanal (Sweeney etal.. 2012b: Paquetetal.. 20111. RDX
metabolism in animals is less well understood. N-Nitroso RDX metabolites have been identified as
derived through anaerobic metabolism (ATSDR. 2012: Pan etal.. 2007bl. Based on characterization
of RDX metabolites in urine and plasma of miniature swine, metabolism of RDX appears to involve
loss of nitro groups and ring cleavage (Musick etal.. 2010: Major etal.. 20071. The two principal
urinary metabolites identified in miniature swine following a single oral dose of 43 or 45 mg/kg
were 4-nitro-2,4-diazabutanal and 4-nitro-2,4-diaza-butanamide (see Table C-2). Bhushan et al.
(2003) suggested that the formation of the 4-nitro-2,4-diazabutanal metabolite occurred via
denitration followed by hydroxylation and spontaneous hydrolytic decomposition resulting in ring
cleavage and aldehyde formation. In the miniature swine gavage studies, only trace amounts of the
nitrosamine RDX metabolites, hexahydro-l-nitroso-3,5-dinitro-l,3,5-triazine (MNX) and 1-nitro-
3,5-dinitroso-l,3,5-triazacyclohexane (DNX), were found in urine (Musick etal.. 2010: Major etal..
20071. In plasma, most of the radioactivity existed as RDX, with trace levels of MNX, DNX, and
l,3,5-trinitroso-l,3,5-triazacyclohexane (TNX). The study authors suggested that the trace levels of
MNX, DNX, and TNX in plasma may have been formed within the GI tract via sequential nitrogen
reduction by intestinal bacteria fMaior etal.. 20071. The low levels of these compounds in urine
and plasma were attributed to the nearly complete absorption of RDX from the GI tract, leaving
little parent compound available for bacterial metabolism within the GI tract. In a study of female
deer mice (Peromyscus maniculatus) fed diets containing RDX at concentrations of 12 and
120 mg/kg for 9 days, MNX and DNX were identified in the stomach, but TNX was not detected (Pan
etal.. 2007b). MNX and DNX were also measured in various organs of female B6C3Fi mice provided
RDX in feed at doses of 0.38-522 mg/kg; TNX was also detected in some organ compartments, but
not in the liver. The authors concluded that RDX can be metabolized into its N-nitroso compounds
This document is a draft for review purposes only and does not constitute Agency policy.
C-8	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	in mice, but did not identify a mechanism for the formation of the metabolites. Comparing RDX
2	with MNX and TNX, RDX was the most potent compound at causing overt signs of toxicity (seizures
3	and mortality) as determined through identification of the median lethal dose using the EPA up-
4	and-down procedure in deer mice of varying ages f Smith etal.. 2009: Rispin etal.. 20021.
5	Table C-2. Principal urinary metabolites of RDX in miniature swine 24 hours
6	after dosing with RDX
Sample origin
Metabolite name
Metabolite structure
Urine peak 1 Ml
4-Nitro-2,4-diazabutanal
H H
q3n" ^ ""jf
O
Urine peak 2 M2
4-Nitro-2,4-diaza-butanamide
H H
.N NHi
o2n""
6
7
8	Sources: Major et al. (2007): Musick et al. (2010)
9
10	Although the metabolic pathways and major tissues involved in RDX metabolism have not
11	been identified, there is some evidence for the involvement of the liver and CYP450 enzymes.
12	Comparison of hepatic radioactivity to liver concentrations of RDX after a single gavage dose to rats
13	suggested the presence of RDX metabolites and a possible role for hepatic metabolism of RDX
14	f Schneider etal.. 19771. In vitro data indicated that CYP450 may be involved in the metabolism of
15	RDX fBhushan et al.. 20031. Incubation of RDX with nicotinamide adenine dinucleotide phosphate
16	(NADPH) and rabbit liver CYP450 2B4 under anaerobic conditions produced nitrite, 4-nitro-
17	2,4-diazabutanal, formaldehyde, and ammonium ion fBhushan etal.. 20031. The reaction rate
18	under aerobic conditions was approximately one-third of that observed under anaerobic
19	conditions. Several CYP450 inhibitors (ellipticine, metyrapone, phenylhydrazine, 1-aminobenzo-
20	triazole, and carbon monoxide) decreased the formation of RDX metabolites (55-82% inhibition),
21	providing support for the role of CYP450 in RDX metabolism.
22	C.1.4. Excretion
23	The primary routes of elimination of absorbed RDX are excretion of RDX and metabolites in
24	urine, and exhalation of CO2 liberated from metabolism of RDX (Sweeney etal.. 2012a: Musick etal..
25	2010: Krishnan etal.. 2009: Major etal.. 2007: Schneider etal.. 19771. Tritium derived from
26	administered [3H]-RDX has been detected in mouse gall bladder contents, suggesting biliary
27	secretion in this species fGuo etal.. 19851: however, biliary secretion of RDX or metabolites has not
28	been confirmed in other animal species. Studies conducted in the rat and swine suggest that
29	metabolism is the dominant mechanism of elimination of absorbed RDX. In both species,
30	metabolites dominated the carbon-14 distribution in urine of animals that received doses of
This document is a draft for review purposes only and does not constitute Agency policy.
C-9	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	[14C]-RDX, with RDX accounting for <5% of the urinary carbon-14 fMusick etal.. 2010: Schneider et
2	al.. 19771
3	Data on kinetics of elimination of absorbed RDX from blood are available from reports of
4	accidental exposures of humans to RDX (Table C-3). Woody etal. (1986) estimated the elimination
5	ti/2 to be approximately 15 hours in a child who ingested approximately 85 mg of RDX per kg of
6	body weight The ti/2 estimate was based on measured serum concentrations of RDX made
7	between 24 and 120 hours following ingestion for RDX. Based on plasma RDX concentration data
8	from five adults exposed to RDX (measurements made between 24 and 96 hours following
9	exposure) (Ozhan et al.. 2003). a first-order elimination ti/2 of 20-30 hours was derived (calculated
10	for this review by fitting the serum RDX data to a first-order exponential function). It needs to be
11	noted that it is not possible to draw reliable inferences from these values since they are based on
12	accidental, acute exposures and, in particular, the data for the child are based on a single set of
13	measurements for one individual.
14	Table C-3. Elimination ti/2 values for RDX or radiolabeled RDX
Animal
Route
Dose (mg/kg)
Time3
ti/z (hrs)
Source
Human (child)
Oral
_Q
LO
00
24-120 hrs
15.0°
Woodvetal. (1986)
Human (adult)
Oral
NA
24-96 hrs
21-29cd
Ozhan et al. (2003)
Rat
i.v.
5-6
0.5 min-6 hrs
_Q
o
1
Schneider et al. (1977)
Rat
i.v.
0.8-1.0
30 min-10 hrs
4.6c,d
Krishnan et al. (2009)
Rat
Oral
1.53-2.07
1-10 hrs
6.9c,d
Krishnan et al. (2009)
Mouse
Oral
35, 60, 80
45 min-4 hrs
1.2d
Sweeney et al. (2012b)
15
16	Observation period following exposure on which the ti/2 values were based.
17	bReported estimate of dose based on blood kinetics.
18	cValue for blood RDX.
19	Calculated for this review based on reported blood RDX concentrations.
20
21
22	The kinetics of elimination of absorbed RDX from blood has been evaluated in rats and
23	mice. In rats elimination kinetics were biphasic fKrishnan et al.. 2009: Guo etal.. 1985: Schneider et
24	al.. 19771. As shown in Table C-3, estimated ti/2 values for the terminal elimination phase in rats
25	range from 5 to 10 hours fKrishnan etal.. 2009: Schneider et al.. 1977). Blood concentration time
26	course measurements of RDX can be used to estimate an apparent metabolism and elimination of
27	RDX from blood. The RDX blood concentrations reported in Sweeney etal. (2012 b) after gavage
28	dosing of 35, 60, and 80 mg/kgRDX found a consistent terminal elimination ti/2 of approximately
29	1.2 hours. The elimination ti/2 estimated for rats fKrishnan etal.. 2009: Schneider et al.. 1977) is as
30	much as an order of magnitude longer than mice fSweenev etal.. 2012b).
This document is a draft for review purposes only and does not constitute Agency policy.
C-10	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
C.1.5. Physiologically Based Pharmacokinetic (PBPK) Models
Overview of Available PBPK Models
A PBPK model to simulate the pharmacokinetics of RDX in rats was first developed by
Krishnan etal. f20091 and improved upon to extend the model to humans and mice fSweeney etal..
2012a: Sweeney etal.. 2012b). The Sweeney etal. (2012a) model consists of six main
compartments: blood, brain, fat, liver, and lumped compartments for rapidly perfused tissues and
slowly perfused tissues (Figure C-l). The model can simulate RDX exposures via the i.v. or oral
route. Distribution of RDX to tissues is assumed to be flow-limited. Oral absorption is represented
in this model as first-order uptake from the GI tract into the liver, with 100% of the dose absorbed.
RDX is assumed to be cleared by first-order metabolism in the liver. However, there is no
representation of the kinetics of any RDX metabolites. The acslX model code (Advanced Continuous
Simulation Language, Aegis, Inc., Huntsville, Alabama) was obtained from the authors of Sweeney et
al. f2012a).
IV dose
KQC
KQB
KQF
KQR
KQS
KQL
KAD
KfC
KfC
KAS
KAS
Metabolism
Metabolism
Stomach
Duodenum
Fat
Slowly Perfused
Liver
Blood
Liver
Richly Perfused
Brain
Oral dose	Oral dose
Exposure to RDX is by the i.v. or oral route, and clearance occurs by metabolism in the liver. See Table C-4 for
definitions of parameter abbreviations. The GI tract is represented as one compartment in Krishnan et al. (2009)
(on the left) and two compartments in Sweeney et al. (2012a) (on the right).
Figure C-l. PBPK model structure for RDX in rats and humans.
This document is a draft for review purposes only and does not constitute Agency policy.
C-ll	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
The parameter values used in the Sweeney etal. (2012a) rat model are listed in Table C-4.
The physiological model parameter values for cardiac output, tissue volumes, and blood perfusion
of tissues were obtained from the literature fT imchalk et al.. 2 0 0 2: Brown etal.. 19971. RDX
tissue:blood partition coefficients for liver (PL), brain (PB), and richly perfused tissues (PR) were
estimated with an algorithm that relates the measured n-octanol:water partition coefficient for RDX
to reported compositions of water and lipids in specific rat tissues fPoulin and Theil. 2000: Poulin
and Krishnan. 1995). Tissue:blood partition coefficients for fat (PF) and slowly perfused tissues
(PS), as well as the metabolic rate constant (KfC), were simultaneously optimized to fit rat blood
RDX concentrations following i.v. doses of 0.77 or 1.04 mg/kg RDX (Krishnan etal.. 20091
producing values of 5.57, 0.15, and 2.6 kg0 33/hour for PF, PS, and KfC, respectively. While the
optimized value for PS is much smaller than that used by Krishnan etal. (2009) (1.0 kg0 33/hour),
the optimized values for PF and KfC were fairly similar to those used by Krishnan et al. f20091
(7.55, and 2.2 kg0 33/hour). The rat model with these parameter values also had good agreement
with blood RDX concentrations after a 5-6 mg/kg i.v. exposure (Schneider etal.. 1977).
Table C-4. Parameter values used in the Sweeney etal. f2012a) and Sweeney
etal. (2012b) PBPK models for RDX in rats, humans, and mice as reported by
authors
Parameter (abbreviation; units)
Rat
Human
Mouse
Source
Body weight (BW; kg)
0.3
70
0.0206
Default values; study-specific values used
if available
Cardiac output (KQC, L/hr/kg074)
15
14
15
Timchalk et al. (2002); Brown et al. (1997)
Tissue volumes (fraction of BW)
Liver (KVL)
0.04
0.026
0.04
Timchalk et al. (2002); Brown et al. (1997)
Brain (KVB)
0.012
0.02
0.012
Timchalk et al. (2002); Brown et al. (1997)
Fat (KVF)
0.07
0.21
0.07
Timchalk et al. (2002); Brown et al. (1997)
Richly perfused tissues (KVR)
0.04
0.052
0.04
Timchalk et al. (2002); Brown et al. (1997)
Blood (KVV)
0.06
0.079
0.06
Timchalk et al. (2002); Brown et al. (1997)
Slowly perfused tissues (KVS)
0.688
0.523
0.688
0.91 - (KVL + KVB + KVF + KVR + KVV)
Blood flows (fraction of cardiac output)
Liver (KQL)
0.25
0.175
0.25
Timchalk et al. (2002); Brown et al. (1997)
Brain (KQB)
0.03
0.114
0.03
Timchalk et al. (2002); Brown et al. (1997)
Fat (KQF)
0.09
0.085
0.09
Timchalk et al. (2002); Brown et al. (1997)
Slowly perfused tissues (KQS)
0.2
0.2449
0.2
Timchalk et al. (2002); Brown et al. (1997)
Richly perfused tissues (KQR)
0.43
0.3811
0.43
1 - (KQL + KQB + KQF + KQS)
This document is a draft for review purposes only and does not constitute Agency policy.
C-12	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Parameter (abbreviation; units)
Rat
Human
Mouse
Source
Tissue:blood partition coefficients
Liver (PL)
1.2
1.3
1.3
Kr
shnan et al. (2009)a
Brain (PB)
1.4
1.6
1.6
Kr
shnan et al. (2009)a
Richly perfused tissues (PR)
1.4
1.6
1.6
Kr
shnan et al. (2009)a
Fat:blood (PF)
5.57
5.57
5.57
Sweeney et al. (2012a)b
Slowly perfused tissues (PS)
0.15
0.15
0.15
Sweeney et al. (2012a)b
Metabolism
First-order metabolic rate constant
(KfC; kg033/hr)
2.6
9.87 (child);
11.2 (adult)
102
Sweeney et al. (2012a)b c;
Sweeney et al. (2012b)d
GI absorption
Dosing via gavage




Absorption from compartment 1
(KAS, /hr)
0.83
0.033
0.51
Sweeney et al. (2012a); Sweeney et al.
(2012b)cde
Transfer from compartment 1 to
compartment 2 (KT, /hr)
1.37
0
0
Sweeney et al. (2012a)c d
Absorption from compartment 2
(KAD, /hr)
0.0258
0
0
Sweeney et al. (2012a)c d
Dosing via capsule (KAS, /hr)
0.12
NA
NA
Sweeney et al. (2012a)e
"coarse" RDX formulation (KAS, /hr)
0.005
NA
NA
Sweeney et al. (2012a)e
1
2	Predicted from n-octanol:water partition coefficient.
3	bOptimized from rat i.v. data.
4	cOptimized from human data of Ozhan et al. (2003) and Woody et al. (1986).
5	dOptimized from mouse oral data.
6	eOptimized from rat oral data of Bannon et al. (2009a), Crouse et al. (2008), Krishnan et al. (2009), and Schneider
7	et al. (1977).
8
9	Note: Parameter values used in the Sweeney et al. (2012a) and Sweeney et al. (2012b) PBPK models for RDX in
10	rats, humans, and mice.
11
12	The GI tract oral absorption rate constant (KAS) was optimized to fit the time-course
13	concentration data for rat oral dosing studies. The Krishnan etal. f20091 model used a
14	one-compartment GI tract KAS was fit to the RDX blood concentrations in Krishnan et al. (2009).
15	and the model with this parameter value had good agreement with the blood RDX concentrations
16	after 0.2 and 1.24 mg/kg oral exposures fCrouse etal.. 20081. The value of KAS was adjusted to fit
17	the RDX blood concentrations in the Schneider etal. (1977) study. Sweeney etal. (2012a) modified
18	the GI tract description by adding a second GI compartment and corresponding oral absorption
19	parameters (KAS, KAD, and KT) to fit the blood concentrations from Krishnan etal. f20091. For the
20	other oral dosing studies, the two-compartment GI model did not improve the model fit to the data,
This document is a draft for review purposes only and does not constitute Agency policy.
C-13	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
so KT was set equal to zero making the GI submodel equivalent to a one-compartment model. The
value of KAS was adjusted separately to fit the oral studies with RDX in capsules (Bannon etal..
2009a: Grouse etal.. 20081 and coarse-grain RDX in a saline slurry f Schneider etal.. 19771.
The Sweeney etal. f2012al model fits to blood and brain RDX concentrations for rats were
mostly within a factor of 1.5 of the experimentally measured values indicating a tightly calibrated
model.
Human RDX toxicokinetics were modeled with the same model structure as for rats. Values
for the human physiological parameters such as tissue volumes and blood perfusion of tissues were
obtained from the literature (Brown et al.. 19971. Human absorption and metabolic clearance rate
constants were optimized to fit observed RDX blood concentrations from a case study of ingestion
by a 3-year-old boy fWoodv et al.. 19861. and a study where five soldiers were intentionally or
accidentally exposed to RDX powder via inhalation or dermal contact fOzhan etal.. 20031. The
amounts of RDX ingested in both studies were unknown, so Sweeney etal. f2012al estimated the
dose amount by optimizing this parameter to fit the data (Table C-4). Sweeney etal. (2012a)
initially simulated each individual soldier's blood level data separately. The resulting parameter
values were similar, so data from the five soldiers were combined and the rate constants were re-
estimated using the combined data. For comparison, the rat metabolic rate constant (KfC) was
scaled to humans; the rat KfC (from fitting to in vivo data) was multiplied by the ratio of the human
to rat metabolic rate constants measured in vitro and by the ratio of human to rat microsomal
protein levels (Cao etal.. 2008: Lipscomb and Poet. 20081. The scaling from rats yielded a human in
vivo metabolic rate constant of 12.4 kg-BW0 33/hour, which is similar to the values that Sweeney et
al. (2012a) derived by fitting the combined Ozhan etal. (2003) adult data (11.2 kg-bw° 33/hour) and
the Woody etal. (1986) child data (9.87 kg-bw° 33/hour).
Mouse RDX toxicokinetics were also modeled by Sweeney etal. f 2 012 b 1 using the same
model structure as for rats. Values for the mouse physiological parameters such as tissue volumes
and blood perfusion of tissues were assumed to be the same as the body weight normalized
parameter values in the rat model. RDX tissue:blood partition coefficients for liver (PL), brain (PB),
and richly perfused tissues (PR) were estimated with an algorithm that relates the measured
n-octanol:water partition coefficient for RDX to reported compositions of water and lipids in
specific mouse tissues (Poulin and Theil. 2000: Poiilin and Krishnan. 1995). The KfC and KAS were
optimized to fit measured mouse RDX blood concentrations fSweenev et al.. 2012bl. The KfC value
estimated for the mouse (102 kg0 33/hour) is much higher than those estimated for rats and humans
(2.6 and 11.2 kgO 33/hour, respectively); however, the KAS value (0.51/hour) fit to mouse data is
similar to the value (0.83/hour) used in the RDX rat model for gavage in water. The Sweeney et al.
(2012b) model predictions of blood RDX concentrations were in good agreement with the
experimental mouse gavage data reported in the same study.
The above PBPK model was evaluated and subsequently modified by EPA for use in dose-
response modeling in this assessment This is detailed in the following section.
This document is a draft for review purposes only and does not constitute Agency policy.
C-14	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	PBPK Model Evaluation and Further Development of the Sweeney etal. (2012a) and Sweeney et
2	at (2012b) Models
3	EPA evaluated and performed a quality control check of the PBPK models for RDX in rats,
4	humans, and mice published by Sweeney and colleagues fSweenev et al.. 2012a: Sweeney etal..
5	2012b). The conclusions from these analyses are summarized below and then discussed in more
6	detail:
7	1) The model code and the parameter values matched the published reports. Minor
8	discrepancies in physiological parameters (KVR and KQS) were identified and updated in
9	the model by EPA.
10	2) The absorption of RDX from the GI tract did not use a consistent structure; for gavage doses,
11	the model used a two-compartment GI submodel and for other oral exposures (e.g., gelatin
12	capsule), the model used a one-compartment GI submodel. The model was revised to have
13	a one-compartment GI submodel to simulate all oral exposures with a consistent set of
14	absorption parameters for each dosage formulation of administered RDX.
15	3) Additional oral rat data were identified from single-dose studies fMacPhail etal.. 1985:
16	Schneider et al.. 19771 and subchronic studies fSchneider et al.. 19781 and were used for
17	model calibration as well as for independent comparison against model predictions.
18	4) In addition to the sensitivity analysis conducted by Sweeney etal. f2012bl on the mouse
19	model, a sensitivity analysis in the rat and human models was performed.
20	5) The Sweeney et al. f 2012b 1 mouse model used the same physiological parameters scaled to
21	body weight as the rat model. This mouse model was revised to use mouse-specific
22	physiological parameters.
23	The Sweeney etal. f2012al model for rats was modified by changing the oral absorption
24	rate constants (as discussed below) and the partition coefficients for the fat and slowly perfused
25	tissues (PF and PS) as shown in Table C-5. The partition coefficients for the fat and slowly perfused
26	tissues were set to the values calculated by Krishnan etal. (20091 relating the measured n-octanol:
27	water partition coefficient for RDX to reported compositions of water and lipids in those tissues.
28	The fits to RDX blood time course data after i.v. exposure (Figure C-2) are slightly worse than the
29	Sweeney etal. f2012al rat model because the Sweeney etal. f2012al rat model optimized the
30	fat:blood and slowly perfused tissue partition coefficients to fit the data.
31	Table C-5. Parameters values used in the EPA application of the rat, human,
32	and mouse models
Parameter (abbreviation; units)
Rat
Human
Mouse
Source
Body weight (BW; kg)
0.3
70
0.0206
Default values shown; study-specific
values used if available
Cardiac output (KQC; L/hr/kg0 74)
15
14
15
Timchalk et al. (2002): Brown et al. (1997)
This document is a draft for review purposes only and does not constitute Agency policy.
C-15	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Parameter (abbreviation; units)
Rat
Human
Mouse
Source
Tissue volumes (fraction of BW)
Liver (KVL)
0.04
0.026
0.055
Timchalk et al. (2002); Brown et al. (1997)
Brain (KVB)
0.012
0.02
0.017
Timchalk et al. (2002); Brown et al. (1997)
Fat (KVF)
0.07
0.21
0.07
Timchalk et al. (2002); Brown et al. (1997)
Richly perfused tissues (KVR)
0.04
0.054
0.071
Timchalk et al. (2002); Brown et al. (1997)
Blood (KVV)
0.06
0.079
0.049
Timchalk et al. (2002); Brown et al. (1997)
Slowly perfused tissues (KVS)
0.688
0.523
0.648
0.91 - (KVL + KVB + KVF + KVR + KVV)
Blood flows (fraction of cardiac output)
Liver (KQL)
0.25
0.175
0.25
Timchalk et al. (2002); Brown et al. (1997)
Brain (KQB)
0.03
0.114
0.03
Timchalk et al. (2002); Brown et al. (1997)
Fat (KQF)
0.09
0.085
0.09
Timchalk et al. (2002); Brown et al. (1997)
Slowly perfused tissues (KQS)
0.2
0.249
0.2
Timchalk et al. (2002); Brown et al. (1997)
Richly perfused tissues (KQR)
0.43
0.377
0.43
1 - (KQL + KQB + KQF + KQS)
Tissue:blood partition coefficients and metabolism
Liver (PL)
1.2
1.3
1.3
Krishnan et al. (2009)a
Brain (PB)
1.4
1.6
1.6
Krishnan et al. (2009)a
Richly perfused tissues (PR)
1.4
1.6
1.6
Krishnan et al. (2009)a
Fat:blood PC (PF)
7.55
7.55
7.55
Krishnan et al. (2009)a
Slowly perfused tissues (PS)
1.0
1.0
0.9
Krishnan et al. (2009)a
First-order metabolic rate constant
(KfC; kg033/hr)
2.6
9.87 (small
boy); 11.2
(soldiers)
77
Sweeney et al. (2012a)bc; Sweeney et al.
(2012b)d
Absorption
Absorption from Gl to liver (KAS;
/hr)
Table C-6
1.75
0.6
Fit to rat, human, and mouse oral data
Absorption from lung to blood
(Klung; /hr)

0.75

Fit to human data
1
2	Predicted from n-octanol:water partition coefficient.
3	bOptimized from rat i.v. data.
4	cOptimized from human data of Ozhan et al. (2003) and Woody et al. (1986).
5	dOptimized from mouse oral data, and differs from that obtained by Sweeney et al. (2012b).
6
This document is a draft for review purposes only and does not constitute Agency policy.
C-16	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
A) 2.5
B) 5
¦ 5.5 mg/kg data 	5.5mg/kgfit
2
¦ 1.038 mg/kg data 	1.038 mg/kg fit
A 0.767 mg/kg data	0.767 mg/kg fit
0
0
0123456789 10
time (hours)
0
1
2	3	4
time (hours)
5
6
A) data from Krishnan et al. (2009) (0.4 kg rats) and B) data from Schneider et al. (1977)
(simulation of 0.25 kg rats and 5.5 mg/kg dose for 0.2-0.25 kg rats and 5-6 mg/kg dose).
Figure C-2. EPA rat PBPK model predictions fitted to observed RDX blood
concentrations in male and female Sprague-Dawley rats following i.v.
exposure.
Absorption of RDX from the GI Tract
As discussed above in the oral absorption section under toxicokinetics (Section C.l.l), the
rate of oral absorption depends on the physical form of RDX. This was demonstrated by comparing
the Schneider etal. (1977) studies, which used gavage doses of 100 mg/kg of course, granular RDX
and 50 mg/kg finely powdered RDX, and observing that the 50 mg/kg finely powdered RDX had a
higher peak plasma level. These results are likely explained by the smaller surface area to mass
ratio of the coarse-grain RDX leading to slower dissolution and absorption.
To follow the rule of model parsimony (i.e., use no more parameters than needed for the
best fit to all of the data), oral absorption was modeled with a one-compartment GI tract submodel
for all simulations. To account for the differences in absorption due to the physical form of RDX,
separate rate constants for RDX oral absorption were optimized to fit measured blood
concentrations of RDX according to the type of dosing formulation; the model fits obtained with the
EPA's revised parameters for rats are shown in Figures C-3 to C-5. The oral dosing formulations
were grouped into four categories: RDX dissolved in water, RDX in capsules, fine-grain RDX, and
coarse-grain RDX. The absorption rate constant for RDX dissolved in water was optimized to the
data in the Krishnan et al. (2009) study (Figure C-3). The absorption rate constant for RDX in
capsules was optimized to the data in the Crouse etal. (2008) and Bannonetal. (2009a) studies
(Figure C-4). The absorption rate constant for fine-grain RDX was optimized to the data described
below (Additional RDX Time-Course Data) in the MacPhail etal. T19851 and Schneider et al. T19771
studies (Figure C-7). The Schneider et al. (1977) study was used to estimate the absorption rate
constants for coarse-grain RDX (Figure C-5; as represented by the fit to the data obtained by the
solid curve at 100% bioavailability). Overall, the fits of the EPA revised model to the blood 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
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
course data of these studies are similar to the fits of the Sweeney etal. (2012a) rat model. The fits
to RDX brain time course data after oral exposure to RDX in capsules (Figure C-6A) are similar to
the fits of the Sweeney etal. f2012al rat model. The absorption rate constants for each dosing
formulation are listed in Table C-6.
2.25
¦ 1.53 mg/kg data 	l.SBmg/kgfit
2
A 2.07 mg/kg data 	2.07 mg/kg fit
1.75
1.5
1.25
1
0.75
0.5
0.25
0
0
1
2
3
4
5
6
7
8
9 10
time (hours)
Male and female Sprague-Dawley rats (0.4 kg) were dosed by gavage (Krishnan et al., 2009).
Figure C-3. EPA rat PBPK model predictions fitted to observed RDX blood
concentrations following oral exposure to RDX dissolved in water.
	18 mg/kg fit
¦ 18 mg/kg data
	3 mg/kg fit
A 3 mg/kg data
A)
0.55
0.5
0.45
5 0.4
£0.35
"g 0.3
I 0.25
I °'2
g 0.15
0.1
0.05
0
5 6 7
time (hours)
20 24 28
time (hours)
The ingested RDX doses were: A) 0.2 and 1.24 mg/kg RDX in male Sprague-Dawley rats (0.4 kg,
data from Crouse et al. (2008)) and B) 3 and 18 mg/kg RDX in male and female Sprague-Dawley
rats (0.35 kg, data from Bannon et al. (2009a)) for KAS = 0.35/hour.
Figure C-4. EPA rat model predictions fitted to observed RDX blood
concentrations following oral exposure to RDX in dry capsules.
	1.24 mg/kg fit ¦ 1.24 mg/kg data
This document is a draft for review purposes only and does not constitute Agency policy.
C-18	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
3.75
3.5
3.25
3
3-2.75
"Si 2.5
-§-2.25
2
£1.75
1.5
gl.25
: 1
0.75
0.5
0.25
0
0 2 4 6 8 10 12 14 16 IS 20 22 24
time (hours)
Symbols denote observed RDX blood concentrations measured in male Sprague-Dawley rats
(0.225 kg) resulting from oral doses of 100 mg/kg RDX (Schneider et al., 1977). The KAS fit to
these data assuming 100% bioavailability resulted in the same estimate (KAS = 0.00497/hour) as
obtained by Sweeney et al. (2012a). Alternatively, for KAS fixed at the value fit to fine-grain RDX
in a saline slurry (KAS = 0.019/hour fit to data from Schneider et al. (1977) and MacPhail et al.
(1985); Figure C-7), the estimate of oral bioavailability fit to the RDX blood concentrations was
35%. A bioavailability of 40% and KAS = 0.019/hour is also shown for comparison.
Figure C-5. Effect of varying oral absorption parameters on EPA rat model
predictions fitted to observed RDX blood concentrations following oral
exposure to coarse-grain RDX.
	18 mg/kg fit
	3 mg/kg fit
¦ 18 mg/kg data
A 3 mg/kg data
B) 10
9

A
	50 mg/kg fit
¦ female data


8 -


A male data
16 20 24 28 32
time (hours)
16 20 24 28 32 36 40 44 48
time (hours)
A) 3 and 18 mg/kg RDX in dry capsules (0.35 kg male and female rat data from Bannon et al.
(2009a); best fit KAS = 0.35/hour. B) 50 mg/kg fine-grain RDX in a saline slurry (0.25 kg male and
female rats data from MacPhail et al. (1985); best fit KAS = 0.019/hour.
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
C
Figure C-6. EPA rat model predictions fitted to observed RDX brain tissue
concentrations following oral exposure to RDX.
¦ 100 mg/kg data
	35%bioavail KAS=0.019/hrfit
	40%bioavail KAS=0.019/hrfit
	100%bioavail KAS=0.00497/hrfit
This document is a draft for review purposes only and does not constitute Agency policy.
C-19	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-6. Doses, dosing formulations, and absorption rate constants in
2	animal and human studies
Formulation
Study
Dose
Estimated KA (/hr)
RDX dissolved in water
Krishnan et al. (2009)
1.53, 2.07 mg/kg, single gavage
1.75
Schneider et al. (1978)
~5-8 mg/kg-d, drinking water
90 d
Dry RDX in capsules3
Crouse et al. (2008)
0.2, 1.24 mg/kg, single dose
0.35
Bannon et al. (2009a)
3,18 mg/kg, single dose
Fine-grain RDX in saline slurry
Schneider et al. (1977)
50 mg/kg, single gavage
0.19
MacPhail etal. (1985)b
50 mg/kg, single gavage
Coarse-grain RDX in saline slurry
Schneider et al. (1977)
100 mg/kg, single gavage
0.00497
3
4	aCapsules were filled with dry RDX from stock solution of acetone, and acetone was evaporated off.
5	bRDX particle size was <66 urn in diameter suspended in a 2% solution of carboxymethylcellulose.
6
7	An alternative to varying the KAS for each RDX formulation would be to vary the oral
8	bioavailability, in effect modifying the administered exposure concentration. Therefore, the
9	sensitivity of the model fitto variations in oral bioavailability was examined in Figure C-5 and an
10	analysis of model sensitivity to oral bioavailability was conducted as discussed further in the
11	section, Sensitivity Analysis of the Rat PBPK Model.
12	Additional RDX Time-Course Data
13	The EPA revised models were simultaneously fitted against additional RDX time-course
14	data (not used in the original Sweeney etal. (2012a) model calibration). These data came from
15	(1) two studies in which animals received oral doses of fine-grain RDX (MacPhail etal.. 1985:
16	Schneider et al.. 19771 (Figure C-7) and (2) RDX brain time-course data from a study in which
17	animals received oral doses of fine-grain RDX fMacPhail etal.. 19851 (Figure C-6B). Overall, the
18	calibrated EPA rat model predictions are within a factor of 1.5 of the measured values from
19	different data sets, and are therefore likely to provide a more robust estimated parameter.
20
This document is a draft for review purposes only and does not constitute Agency policy.
C-20	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
A) 5
	 B) 4.5
	SOmg/kgfit
	50 mg/kg fit
4.5
4
¦ 50 mg/kg data
¦ female data
0
-¦
0 24 48 72 96 120 144 168 192 216 240 264
time (hours)
0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96
time (hours)
Oral doses of 50 mg/kg RDX were administered to: A) male Sprague-Dawley rats (0.225 kg)
(Schneider et al., 1977) and B) male and female Sprague-Dawley rats (0.25 kg) data (MacPhail et
al„ 1985). Best fit KAS = 0.019/hour.
Figure C-7. EPA rat model predictions fitted to observed RDX blood
concentrations following oral exposure to fine-grain RDX in a saline slurry.
Following calibration, the EPA model was further tested by comparison with results from
two other subchronic oral studies in male and female rats (Schneider etal.. 19781. These were a
gavage study where 20 mg/kg RDX was administered in saline slurry and a drinking water study
where rats were provided with RDX-saturated drinking water (50-70 |ig/mL] ad libitum for which
the study authors estimated a daily dose between 5 and 8 mg RDX/kg body weight It is striking
that the observed RDX blood concentrations in the gavage study (Figure C-8, symbols) were
virtually the same, or only slightly elevated, as compared to the blood concentrations reported in
the drinking water exposures, with an approximately threefold lower daily administered dose in
the drinking water study (Figure C-9, symbols). This is counter to the expectation that higher doses
cause higher blood levels and is discussed further below.
This document is a draft for review purposes only and does not constitute Agency policy.
C-21	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
20 mg/kg
model
20 mg/kg
data
4.5
™ 3.5
¦a 3
0
1	2.5
c
X
Q
OS 1.5
0.5
10 20 30 40 50 60 70 80 90 100
0
time (days)
Model fits and mean observed RDX blood concentrations resulting from daily gavage doses of
20 mg/kg RDX for 90 days to male and female Sprague-Dawley rats (0.225 kg). The RDX in saline
slurry was assumed to be coarse-grained with an oral absorption rate constant
KAS = 0.00497/hour,
Figure C-8. Comparison of EPA rat model predictions with data from
Schneider et al. Q978) for the subchronic gavage study.
model
0 10 20 30 40 50 60 70 80 90 100
time (days)
Model fits and mean observed RDX blood concentrations resulting from a daily estimated dose of
6.5 mg RDX/kg-day for 90 days to male and female Sprague-Dawley rats (0.225 kg). The large
peak to trough change in the simulation results from model representation of the daily oral
ingestion of drinking water primarily during the waking state. The oral absorption rate constant
for RDX dissolved in water was used (KAS = 1.75/hour).
Figure C-9. Comparison of EPA rat model predictions with data from
Schneider et al. (1978) for the subchronic drinking water study.
This document is a draft for review purposes only and does not constitute Agency policy,
C-22	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
EPA's modified PBPK model was set up to simulate drinking water exposures with a
noncontinuous sipping pattern based on Spiteri (1982). which assumed 80% of the consumption to
occur episodically at night when the rats were awake3. The model predicts blood concentrations to
increase in proportion to the total dose; for the gavage study, the model predictions yielded an RDX
blood concentration approximately threefold higher than the reported mean blood concentrations
(Figure C-8), while for the subchronic drinking water study, the model fit the data reasonably well
(Figure C-9).
It is possible that multiple mechanisms such as elimination of unabsorbed RDX or metabolic
induction may explain why the observed RDX blood concentrations did not increase in proportion
to the higher administered dose in the gavage studies compared to the drinking water study.
Elimination of unmetabolized RDX may be an insignificant factor in the single-dose studies used for
calibration of the absorption constant for the RDX in saline slurry, but for repeated, higher doses
this elimination route could be significant. Schneider et al. (1978) found similar RDX
concentrations in the feces of rats in the gavage and drinking water studies (3.1 ± 2.0 and
2.7 ± 1.3 [ig RDX per g dry weight feces, respectively). The total recovery of radioactivity in feces
was also similar in the gavage study (4.8 ± 0.8%, week 1 only) and drinking water study
(4.4 ± 0.6%, measured over the course of the study). Thus, the difference in fecal elimination for
the two routes does not appear significant
It is also possible that metabolic induction occurred during the repeated dosing of RDX in
the gavage study leading to the lower observed RDX blood concentrations. The reasonably good fits
of the model to the drinking water data set demonstrated achievement of regular periodic levels,
and indicate a lack or much lower extent of metabolic induction over time from those repeated
doses, possibly because the dose rate was lower: 5-8 versus 20 mg/kg-day in the gavage study.
Overall, the reasonable agreement of the modified EPA RDX rat model with the subchronic drinking
water data support the use of the model in estimating and extrapolating blood levels following
chronic exposure at or below this exposure range (5-8 mg/kg-day), particularly in drinking water.
Simulating Exposures in Humans
The Sweeney etal. f 2012a 1 model for humans was modified in the same ways as the rats, by
changing the partition coefficients for the fat and slowly perfused tissues (PF and PS) as shown in
Table C-5 and fitting the rate constants for oral absorption and metabolism to RDX blood
concentration data. In the studies of humans with measured RDX blood concentrations by Woody
etal. (1986) and Ozhan etal. (2003). the RDX doses were unknown and the doses were therefore
also optimized to fit the data. The model predictions for the Woody etal. fl9861 data using the best
fit values of dose = 58.9 mg/kg, KAS = 1.75/hour, and KfC = 9.87 kg0 33/hour are shown in
3A constant drinking water ingestion rate interspaced between episodes of no ingestion was assumed. Each
12-hour awake period consisted of eight cycles that alternated between 1.5-hour cycles of frequent sipping
(continuous ingestion) and zero ingestion for 45 minutes each. Each 12-hour sleeping period consisted of
four cycles with regular sipping periods of 30 minutes followed by 2.5 hours of no ingestion.
This document is a draft for review purposes only and does not constitute Agency policy.
C-23	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Figure C-10. The model predictions for the Ozhan etal. (2003) data using the best fit values of an
2	oral dose = 3.5 mg/kg, KAS = 1.75/hour, and KfC = 9.87 kg0 33/hour are shown in Figure C-ll.
3
40
	58.9 mg/kg fit
~ 58.9 mg/kg data
35
-a
§ 20
.a
.£ 15
x
~
^ 10
0
12
24
36
48
60
72
84
96 108 120
time (hours)
4
5	The best fit values were KAS = 1.75/hour, dose = 58.9 mg/kg, and KfC = 9.87 kg0 33/hour.
6	Figure C-10. EPA human model predictions fitted to observed RDX blood
7	concentrations resulting from an accidental ingestion of RDX by a 14.5-kg boy
8	(Woody et al.. 1986).
¦ data from accidental exposure
	inhalation fit assuming3.5 mg/
	oral fit assuming 3.5 mg/kg
2.5
E
0.5
0
12
24
36
48
60
72
84
96
time (hours)
10	For an assumed oral exposure, the best fit values were KAS = 1.75/hour, dose = 3.5 mg/kg, and
11	KfC = 9.87 kg0 33/hour. For the same 3.5 mg/kg dose and metabolism rate constant, an inhalation
12	exposure found a best fit value for Klung of 0.75/hour.
13	Figure C-ll. EPA human model predictions fitted to observed RDX blood
14	concentrations resulting from accidental exposure to adults assumed to be
15	70 kg (Ozhan et al.. 20031.
This document is a draft for review purposes only and does not constitute Agency policy.
C-24	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
EPA's calibration of the model differed in another important respect from that carried out
by Sweeney etal. (2012a). As previously mentioned, Sweeney etal. (2012a) simulated the soldiers'
exposure from the Ozhan etal. (2003) study as an oral exposure, although the study report states
that the exposure was via inhalation and dermal routes. An inhalation or dermal exposure could
change the amount of RDX reaching the blood compared with an oral exposure due to first-pass
metabolism in the liver after oral absorption. Dermal absorption was not considered by EPA to be a
significant route of RDX exposure and was therefore not modeled. This decision is supported by a
study that used excised human skin and reported that only 5.7% of the applied dose was absorbed
into the skin by 24 hours post dosing (Reddv et al.. 2008). The model was modified to simulate an
inhalation exposure and compared with the data from Ozhan etal. (2003). There are insufficient
data on blood:air partitioning to modify the Sweeney etal. (2012a) model with a lung
compartment; therefore, inhalation exposure was modeled in an approximate manner as a direct
input to the blood with an optimized absorption rate to represent absorption from air containing
RDX into the blood. The soldiers' inhalation exposure was simulated as a continuous 8-hour
exposure (i.e., assuming that the soldiers were exposed occupationally during an 8-hour workday),
and for the same dose of 3.5 mg RDX/kg that was estimated by Sweeney etal. (2012a). The model
assumed that 100% of the inhaled dose was absorbed and that the absorption rate constant was
optimized to fit the measured blood concentrations of RDX. The model predictions were in good
agreement with the RDX blood concentrations reported by Ozhan etal. (2003) as shown in
Figure C-ll.
Sensitivity Analysis of the Rat and Human PBPK Models
A sensitivity analysis was performed to see how each model parameter affects the model
output A sensitivity coefficient, defined as the change in a specified dose metric due to a 1%
increase in the value of a parameter, was calculated for each parameter in the rat and human
models. This analysis was carried out for both short-term (24 hours following a single oral dose of
1.5 mg/kg RDX) and longer-term (90 days of repeated oral dosing with 1.5 mg/kg RDX) exposures
for the dose-metric of blood AUC. Parameters with sensitivity coefficients >0.1 in absolute value
(i.e., considered sensitive) are presented in Table C-7. For the blood AUC dose-metric, the only
sensitive RDX-specific parameter is the KfC. This sensitivity is likely because bioavailability was
assumed to be 100% and metabolism is the only route of elimination in the model. These
assumptions mean that all administered RDX will be absorbed and metabolized; in other words, the
blood AUC is proportional to the dose and inversely proportional to the metabolic clearance rate
constant. For the parameter values in this model, the rate of metabolism is relatively slow
compared to the transport of RDX between other tissues and the site of metabolism in the liver, so
that the blood AUC is not sensitive to parameters that impact transport such as KQC and KQL.
Because the metabolic clearance rate constant is scaled to body weight and by liver volume, the
blood AUC is also sensitive to these parameters. The sensitivity analysis by Sweeney etal. (2012b)
This document is a draft for review purposes only and does not constitute Agency policy.
C-25	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	for the AUC of RDX in the liver found the model was sensitive to the liver:blood partition coefficient
2	(PL) in addition to the same parameters (KfC, KVL, and BW) found for the blood AUC.
3	Table C-7. Sensitivity coefficients for rat and human RDX PBPK models
Parameter
Rat sensitivity coefficient
Human sensitivity coefficient
Fractional liver volume (KVL)
-1
-1
Body weight (BW)
0.3
0.3
Metabolic rate constant (KfC)
-1
-1
4
5	Parameters with sensitivity coefficients <0.1 in absolute value are considered not sensitive, and are listed below:
6	• cardiac output (KQC);
7	• fractional blood flow to all tissues (liver, KQL; fat, KQF; slowly perfused tissues, KQS; brain, KQB)
8	• fractional tissue volume of fat (KVF), brain (KVB), and blood volume (KVV)
9	• blood partition coefficients to all tissues (liver, PL; fat, PF; rapidly perfused, PR; slowly perfused, PS; brain, PB)
10	• absorption rates from Gl (KAS, KT, KAD)
11
12	The model is also very sensitive to oral bioavailability, with a sensitivity coefficient of 0.8 in
13	the case of the rat model. As discussed above in the oral absorption section of toxicokinetics
14	(Section C.l.l), estimates of the bioavailability of RDX range from 50 to 87% or greater and may
15	depend upon the physical form of RDX fKrishnan etal.. 2009: Schneider et al.. 1978.19771.
16	However, as seen in Figure C-5, it was not possible to identify the bioavailability and the absorption
17	rate (KAS) as separate parameters by fitting to the available RDX blood concentration time course.
18	Introducing oral bioavailability as an additional unknown parameter and recalibrating the model
19	did not appear to provide an advantage. Therefore, 100% bioavailability was assumed in the model
20	and was acknowledged as an uncertainty.
21	Sim ulating Exposures in Mice
22	Physiological parameters specific to mice were obtained from the literature f Brown etal..
23	19971 and are shown in Table C-5. The partition coefficients calculated for mice by Sweeney et al.
24	(2 012b) were used, and include the liver, brain, and richly perfused tissues. The partition
25	coefficients for the fat and slowly perfused tissues from the Sweeney etal. (2012b) mouse model
26	were not used because they were estimated via optimization of fits to rat i.v. data. Instead, the
27	partition coefficient for fat tissues was set equal to the value calculated by Krishnan etal. f20091 for
28	rat fat tissue, 7.55. The partition coefficient for slowly perfused tissues (0.9) was calculated for
29	mouse tissues using the same methodology as Krishnan etal. f20091. The rate constants for oral
30	absorption and metabolism were optimized to fit the data from Sweeney etal. f2012bl for mouse
31	blood RDX concentrations. The model predictions were in good agreement with the RDX blood
32	concentrations reported by Sweeney etal. (2012b). as shown in Figure C-12.
This document is a draft for review purposes only and does not constitute Agency policy.
C-26	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
~ 80mg/kgdata 	80mg/kgfit
time (hours)
0.5
60mg/kgdata 	60mg/kgfit
1.5 2 2.5
time (hours)
3.5
u
c
o
u
"D
O
_o
-Q
X
o
en
< 2
Ofl
E
— 1
0.5
~ 35mg/kgdata 	35mg/kgfit
1.5 2 2.5 3 3.5
time (hours)
Model fits and mean and standard deviation of observed RDX blood concentrations in female
B6C3Fi mice (0.0205 kg) for doses of 35, 60, and 80 mg/kg with KAS = 0.6/hour and
KfC = 77 kg0 33/hour. Experimental data from Sweeney et al. (2012b).
Figure C-12. Comparison of EPA mouse PBPK model predictions with data
from oral exposure to RDX dissolved in water.
The mouse RDX blood concentrations reported by Sweeney etal. f2012bl. as shown in
Figure C-12 were evaluated with a non-compartmental analysis and compared with the rat data.
The estimate of the area under the curve for blood concentration versus time from the time of
dosing to the time RDX is completely eliminated (AUCtotai) was calculated with a linear trapezoidal
sum plus an extrapolation of the blood concentration at the last time point divided by the terminal
elimination rate constant as shown in the following equation:
This document is a draft for review purposes only and does not constitute Agency policy.
C-27	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
AUCtotal = 2(Ablood concentrations) At/2 + blood concentration at last time point/Kei
where Ablood concentrations are the successive blood concentrations, At is the time
between measured concentrations, and Kei is the terminal elimination rate constant
(calculated from the slope of the linear regression line to the log of blood
concentrations)
For the mouse data from Sweeney etal. (2012b) for the doses of 35, 60, and 80 mg/kg, the
results of the AUCtotai calculation are 3.35, 3.70, and 4.75 mg/L hour; normalized to the
administered dose, these are 0.096, 0.062, and 0.059 mg/L hour per mg/kg. For the blood
concentrations measured in rats in the Krishnan et al. (2009) study (Figure C-3), the animals
received a single oral (gavage) dose of RDX dissolved in water similar to the Sweeney etal. (2012b)
study. The Krishnan et al. (2009) study used doses of 1.53 and 2.07 mg/kg and the results of the
AUCtotai calculation are 6.1 and 11.9 mg/L hour. Including the extrapolation ofthe blood
concentration from the last time point with the terminal elimination rate constant, Kei had a major
contribution to the AUCtotai (approximately one-third), which adds uncertainty to the result, so the
AUCtotai was also calculated without this term and the results are 4.1 and 7.5 mg/L hour. The
AUCtotai values normalized to the administered doses are 4.0 and 5.8 mg/L hour per mg/kg
(including the extrapolation from the last time point) or 2.7 and 3.6 mg/L hour per mg/kg
(excluding the extrapolation from the last time point). Overall, the AUCtotai normalized to the
administered doses for the rat are of the order 10-100 times greater than for the mouse. This non-
compartmental analysis of the data is independent of the PBPK modeling and shows the extent of
the toxicokinetic differences for RDX between the mouse and rat.
The only additional information on RDX metabolism in the mouse comes from a study by
Pan etal. (2013). Pan etal. (2013) measured nitrosamine RDX metabolites of RDX (MNX, DNX, and
TNX, the latter representing a minor metabolic pathway) in mice at the end of a 28-day exposure to
RDX in feed (ad libitum). These measurements were a single time point without controlling the
time between the last RDX ingestion and measurement, and were therefore judged not to be
sufficient for use in parameterizing a PBPK model of the nitrosamine metabolites.
Rat to Human Extrapolations
The rat and human PBPK models as described above were applied to derive human
equivalent doses (HEDs) for candidate points of departure (PODs) for endpoints selected from rat
bioassays. The rat and human PBPK models were used to estimate two dose metrics—the AUC and
the peak concentration (Cmax) for RDX concentration in arterial blood. The relationships between
administered dose and both internal metrics (AUC and Cmax) were evaluated with the rat PBPK
model over the range of 1 ng/kg-day to 100 mg/kg-day and with the human PBPK model over the
range of 0.05 ng/kg-day to 200 mg/kg-day, ranges that encompass the PODs. The times to reach
steady state for the dose metrics were shorter than the duration of the toxicity studies, so the
This document is a draft for review purposes only and does not constitute Agency policy.
C-28	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
steady state values were considered representative of the study and were used. To calculate
steady-state values for daily exposure, the simulations were run until the daily average had a <1%
change between consecutive days. For both the rat and human PBPK models, both dose metrics
correlated linearly with the administered dose. For rats dosed via gavage, the slope of
administered dose versus AUC was 6.800 mg/L-day / mg/kg-day and that for Cmax was
0.4718 mg/L / mg/kg-day. For a continuous dose, the slope of dose versus AUC was the same
(6.800 mg/L-day / mg/kg-day) and for Cmax was 0.3951 mg/L / mg/kg-day. For humans, assuming
a drinking water dose sipping pattern, the slope of administered dose versus AUC was
13.95 mg/L-day / mg/kg-day and that for Cmax was 0.7316 mg/L / mg/kg-day. Given this linearity
in internal metrics and assuming that equal internal metrics in rats and humans are associated with
the same degree of response, the HEDs could then be directly determined by multiplying the lower
bound on the benchmark dose (BMDL) in rats by the ratio of these slopes. For a gavage dose in rats
converted to a human drinking water dose, the ratio for AUC was 6.800 / 13.95 = 0.487 and Cmax
was 0.4718 / 0.7316 = 0.645. For a continuous dose in rats converted to a human drinking water
dose, the ratio for AUC was 6.800 / 13.95 = 0.487 and for Cmax was 0.3951 / 0.7316 = 0.540. These
ratios were applied in Table 2-2 to calculate the PODhed from the rat benchmark dose lower
confidence limits (BMDLs) and no-observed-adverse-effect levels (NOAELs) for each endpoint
Mouse to Human Extrapolations
The mouse and human PBPK models as described above were applied to derive HEDs for
candidate PODs for endpoints selected from mouse bioassays. The mouse and human PBPK models
were used to estimate two dose metrics—the area under the curve (AUC) and peak concentration
(Cmax) for RDX concentration in arterial blood. The relationships between administered dose and
both internal metrics (AUC and Cmax) were evaluated with the mouse PBPK model over the range
10 |ig/kg-day to 100 mg/kg-day and with the human PBPK model over the range 0.05 ng/kg-day to
200 mg/kg-day, ranges that encompass the PODs. The times to reach steady state for the dose
metrics were shorter than the duration of the toxicity studies, so the steady state values were
considered representative of the study and were used. To calculate steady state values for daily
exposure, the simulations were run until the daily average had a <1% change between consecutive
days. For both the mouse and human PBPK models, both dose metrics correlated linearly with the
administered dose. For mouse dosed via gavage, the slope of administered dose versus AUC was
0.0656 mg/L-day / mg/kg-day and that for Cmax was 0.0273 mg/L / mg/kg-day. For a continuous
dose, the slope of dose versus AUC was the same 0.0656 mg/L-day / mg/kg-day and for Cmax was
0.0081 mg/L / mg/kg-day. For humans, assuming a drinking water dose sipping pattern, the slope
of administered dose versus AUC was 13.95 mg/L-day / mg/kg-day and that for Cmax was
0.7316 mg/L / mg/kg-day. Given this linearity in internal metrics and assuming that equal internal
metrics in mice and humans are associated with the same degree of response, the HEDs could then
be directly determined by multiplying the BMDL in mice by the ratio of these slopes. For a gavage
dose in mice converted to a human drinking water dose, the ratio for AUC was 0.0656 / 13.95 =
This document is a draft for review purposes only and does not constitute Agency policy.
C-29	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
0.0047 and Cmax was 0.0273 / 0.7316 = 0.373, respectively. For a continuous dose in mice
converted to a human drinking water dose, the ratio for AUC was 0.0656 / 13.95 = 0.0047 and for
Cmax was 0.0081 / 0.7316 = 0.011. These ratios were applied in Table 2-2 to calculate the PODhed
from the mouse BMDLs and NOAELs for each endpoint
Summary of Confidence in PBPK Models for RDX
Overall, good fits to the rat, mouse, and human time-course data for RDX internal
concentrations were obtained. For the rat and human models, calibration was based generally on
fitting to more than one data set obtained from different studies originating in different
laboratories or accidental exposure settings. Predictions from the rat model compared well with
data from a subchronic drinking water study that was not used in model calibration.
The metabolic rate constant used in the human model was fit to limited data from
accidentally exposed humans; however, the value of the metabolic rate constant has additional
support from in vitro experimental data. The rat metabolic rate constant, fit to multiple
experimental data sets, was scaled to humans using the ratio of human to rat rate constants
measured with in vitro methods. This scaled value of the human metabolic rate constant was very
similar to the rate constant estimated by fitting the model to the human data. The congruence in
values increases the confidence in using the EPA-modified PBPK model for predicting human blood
RDX concentrations.
There are several uncertainties in these models (listed below), the most significant of which
pertain to the mouse PBPK model. The mouse model was based on a single data set, which used
high RDX doses to obtain detectable RDX blood concentrations, and the types of additional data that
increased the confidence in the rat and human models are not available for mice. The additional
data not available for mice are in vitro measurements of RDX metabolism by mouse cells and
quantification of potential routes of RDX elimination in mice. Overall, these uncertainties result in
lower confidence in the mouse model than in the rat and human models.
1)	RDX is readily metabolized in several species, yet there are no data on the toxicokinetics of
RDX metabolites in the rat and human. Some data are available for the n-nitrosoamine
metabolites (a minor metabolic pathway) in mice, but the data are too sparse to help better
parameterize a PBPK model. Consequently, the PBPK models used in this assessment do
not incorporate the kinetics of RDX metabolites.
2)	The available toxicokinetic data are not sufficient to uniquely identify a parameter value for
RDX oral bioavailability. Consequently, the model assumes 100% bioavailability even
though some studies in rats suggest that a lower bioavailability is likely.
3)	The human model is based on single accidental exposures, and the exposure concentrations
are not known.
4)	The only route of clearance of RDX used in the models is that of total metabolism, which
appears reasonable for the rat for which only roughly 5% of the RDX was detected
This document is a draft for review purposes only and does not constitute Agency policy.
C-30	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
unmetabolized in urine and feces. However, no data on the excretion of RDX are available
for the mouse. This inability to properly characterize the fraction of RDX that is
metabolized in the mouse is problematic considering some evidence to indicate that the role
of metabolism in RDX toxicity may be different across species. This uncertainty decreases
the confidence in the mouse PBPK model.
5)	The PBPK model for the mouse is based on a single data set This single data set is used to
fit both the absorption and metabolic rate constants. There are no in vitro data to
independently estimate the metabolic rate constant for the mouse. Consequently, the
confidence in the mouse model parameter values is low.
6)	The analytical detection limit in the mouse pharmacokinetic study is too high to enable
detection at the lower doses. The lowest dose that resulted in a detectable level of RDX in
blood was 35 mg/kg; this dose was high enough to manifest some toxicity in the chronic
mouse bioassay. The measured blood concentration at the final 4-hour timepoint at the
35 mg/kg dose was based on the level measured from one animal only (in the other five
animals exposed at this dose, three were non-detects, one was excluded as an outlier, and
one animal died). Data from a single animal decreases the confidence in the calibration of
the mouse PBPK model.
7) The metabolic rate constant as estimated by the PBPK model for mice was 30-fold higher
than the rat (after accounting for body weight differences), suggesting that the
toxicokinetics of RDX could be significantly different in the mouse than in the rat. Mice may
have more efficient or higher expression of the CYP450 enzymes. Alternatively, mice may
have other unknown metabolic pathways responsible for metabolizing RDX. Identifying the
specific CYP450 enzymes and measuring expression levels and in vitro metabolic rate
constants in mice would allow for in vitro scaling from rats to mice, which could be used to
independently evaluate the mouse metabolic rate constant. Given the high sensitivity of the
model to the metabolic rate constant, this uncertainty in the mouse toxicokinetics
significantly decreases the confidence in using the mouse PBPK model for predicting mouse
blood RDX concentrations.
Model Code for RDX PBPK Model Used in the Assessment
The PBPK acslX model code is made available electronically through the Health and
Environmental Research Online (HERO) database. All model files may be downloaded in a zipped
workspace from HERO (U.S. EPA. 2014).
This document is a draft for review purposes only and does not constitute Agency policy.
C-31	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	C.2. HUMAN STUDIES
2	Table C-8 presents a summary of case reports of humans acutely exposed to RDX. Table C-9
3	provides a chronological summary of the methodologic features of the available epidemiology
4	studies of RDX.
5	Table C-8. Summary of case reports of exposure to RDX
Reference,
number of cases,
exposure setting
Exposure
Effects observed
Comments
Barsotti and Crotti
(1949)
17 males among
20 male Italian
workers
(1939-1942)
Inhalation of RDX powder
during the drying, cooling,
sieving, and packing
processes of its manufacture
Generalized convulsions of a
tonic-clonic (epileptic) type
followed by postictal coma; loss
of consciousness without
convulsions; vertigo; vomiting and
confusion; transient arterial
hypertension
Tobacco and alcohol use
were considered by the
study authors to be
aggravating factors
Manufacturing

Symptoms occurred either
without prodromal symptoms or
were preceded by several days of
insomnia, restlessness, irritability,
or anxiety

Kaplan et al. (1965)
5 males among
26 workers (April-
July 1962)
Manufacturing
Inhalation, ingestion, and
possible skin absorption as a
result of the release of RDX
dust into the workroom air
during the dumping of dried
RDX powder, screening and
blending, and cleanup of
spilled material without
adequate ventilation
Sudden convulsions or loss of
consciousness without
convulsions; few or no
premonitory symptoms (e.g.,
headache, dizziness, nausea,
vomiting); stupor, disorientation,
nausea, vomiting, and weakness;
no changes in complete blood
counts or urinalysis
Mild cases of RDX
intoxication may have
been masked by viral
illness with nonspecific
symptoms (e.g.,
headache, weakness,
upset Gl tract); no
method was available
for determining RDX
concentrations in air;
recovery was complete
without sequelae
Merrill (1968)
2 males
Wartime, Vietnam
Ingestion of unknown
quantity of C-4 with
moderate amounts of alcohol
Coma, vomiting, hyperirritability,
muscle twitching, convulsions,
mental confusion, and amnesia;
kidney damage (oliguria, gross
hematuria, proteinuria, elevated
BUN); liver or muscle damage
(high AST); leukocytosis
Confounding factors
included ingestion of C-4
while intoxicated with
ethanol (vodka), which
may have caused Gl
symptoms, and smoking
(1-1.5 packs of
cigarettes per day)
This document is a draft for review purposes only and does not constitute Agency policy.
C-32	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference,
number of cases,
exposure setting
Exposure
Effects observed
Comments
Stone et al. (1969)
4 males (March-
December 1968)
Wartime, Vietnam
Ingestion of 180 g (patient 1),
or 25 g (patients 2, 3) of C-4
(91% RDX)
Generalized seizures, lethargy,
nausea, vomiting, fever, muscle
soreness, headaches, twitching,
(semi)comatose, headaches,
hematuria, abnormal laboratory
findings, muscle injury, elevated
AST; no kidney damage
For the patient who ingested the
highest dose, anemia and loss of
recent memory present after 30 d
Troops ingested small
quantities of RDX to get
a feeling of inebriation
similar to that induced
by ethanol
Hollander and
Colbach (1969)
5 males (June
1968-January 1969)
Wartime, Vietnam
Inhalation (all five cases) and
ingestion of unknown
quantity of C-4 (two cases)
Tonic-clonic seizures; nausea and
vomiting occurred before and
after admission; hyperirritability,
muscle twitching, convulsions,
mental confusion, and amnesia;
kidney damage (oliguria, gross
hematuria, proteinuria, elevated
BUN); liver or muscle damage
(high AST); leukocytosis;
symptoms cleared by the next day
except for amnesia (in case 2),
oliguria (lasted for 4 d), and gross
hematuria (decreased by
9th hospital day)

Kneoshield and
Stone (1972)
6 males
Wartime, Vietnam
Ingestion of C-4, range
25-180 g, average 77 g
Generalized seizures, coma,
lethargy, severe neuromuscular
irritability with twitching and
hyperactive reflexes, myalgia,
headache, nausea, vomiting,
oliguria, gross hematuria, low-
grade fever; abnormal laboratory
findings (neutrophilic
leukocytosis, azotemia, elevated
AST)
Includes data on
two patients from
Merrill (1968)
Ketel and Hughes
(1972)
18 males
(December
1968-December
1969)
Wartime, Vietnam
Inhalation while cooking with
C-4 and possible ingestion
CNS signs (confusion, marked
hyperirritability, involuntary
twitching of the extremities,
severe prolonged generalized
seizures, prolonged postictal
mental confusion, amnesia); renal
effects (oliguria and proteinuria,
one case of acute renal failure
requiring hemodialysis); Gl
toxicity (nausea, vomiting)
C-4 was cut with the
same knife used to
stir/prepare food
This document is a draft for review purposes only and does not constitute Agency policy.
C-33	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference,
number of cases,
exposure setting
Exposure
Effects observed
Comments
Woodv etal. (1986)
1 male child (August
1984)
Manufacturing
Ingestion of plasticized RDX
from mother's clothing
and/or boots; estimated
ingested dose of 1.23 g RDX
was normalized to the
patient's body weight
(84.82 mg/kg)
Seizures before and after
admission; EEG revealed
prominent diffuse slowing that
was greatest in the occipital
regions with no evidence of
epileptiform activity; elevated
AST on admission and after
24 hrs; within 24 hrs, the child
was extubated and intensive care
withdrawn; normal mental status
and normal neurological
examination at discharge
Mother worked at an
explosive plant in which
RDX was manufactured
in a plasticized form
Goldberg et al.
(1992)
1 male
Nonwartime
Ingestion after chewing a
piece (unknown size) of
"Semtex" plastic explosive
4 hrs before first seizure
Frontal headache and two tonic-
clonic seizures; progressively
disseminating petechial rash
suggestive of meningococcal
infection apyrexial; normotensive;
no photophobia; no neurological
abnormalities; florid petechial
rash over the face and trunk;
lacerated tongue
Initial results included leukocyte
count of 10.8 x 109/dL (87%
neutrophils); hemoglobin, platelet
count, coagulation screen, serum
and CSF biochemistry all within
normal limits; CSF and blood
bacteriologically unremarkable
Shortly following admission,
headache and rash disappeared;
no further seizures

Harrell-Bruder and
Hutchins (1995)
1 male
Ingestion of C-4 (chewing on
a piece of undetermined size)
Tonic-clonic seizures; postictal
state; EEGs were normal; brisk
deep tendon reflexes

Nonwartime



Testud et al.
(1996a)
1 male
Manufacturing
Inhalation and possible
dermal exposure during the
RDX manufacturing process
Malaise with dizziness, headache,
and nausea progressing to
unconsciousness and generalized
seizures without involuntary
urination or biting of the tongue;
blood chemistries were in the
normal range and blood alcohol
content was null

This document is a draft for review purposes only and does not constitute Agency policy.
C-34	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference,
number of cases,
exposure setting
Exposure
Effects observed
Comments
Testud et al.
(1996b)
2 males
Manufacturing
Inhalation and possible
dermal exposure during the
RDX manufacturing process
Sudden loss of consciousness and
generalized seizures; blood serum
level of 2 mg/L RDX measured
Smoker and alcohol
drinker
Hett and Fichtner
(2002)
1 male
Nonwartime
Ingestion of a cube (1 cm
across) of C-4
Nausea and vomiting; tonic-clonic
seizure lasting 2 min, followed by
two seizures of about 30 sec each;
myoclonic jerks in all limbs;
petechial hemorrhages around
face and trunk after seizures

Kucukardali et al.
(2003)
5 males
Nonwartime
Ingestion (accidental) of
37-250 mg/kg body weight
RDX during military training
via food contaminated with
RDX
Abdominal pain, nausea,
vomiting, myalgia, headache,
generalized weakness, repetitive
tonic-clonic convulsions, lethargic
or comatose between seizures,
hyperactive deep tendon reflexes,
sinusal tachycardia; elevated
serum levels of AST and ALT;
kidney damage; plasma RDX
levels 3 hrs after ingestion ranged
from 268 to 969 pg/mL

Davies et al. (2007)
17 males
Nonwartime
Ingestion of unknown
quantity C-4 under unclear
circumstances, but unrelated
to recreational abuse
Seizures, headache, nausea, and
vomiting; hypokalemia and
elevated creatine kinase, lactate
dehydrogenase, and phosphate
noted in all but two patients;
metabolic acidosis only occurred
in two patients directly following
seizures
Patient histories may
have been affected by
the fact that the
incident was the focus
of a military police
investigation
Kasuske et al.
(2009)
2 males
Nonwartime
Ingestion of C-4 after
handling explosive ordnance
Seizures, postictal state,
confusion, drowsiness, headache,
nausea, and vomiting; blood work
revealed high WBC count and
elevated creatine phosphokinase;
proteinuria and gross hematuria
observed

1
2	ALT = alanine aminotransferase; AST = aspartate transaminase; BUN = blood urea nitrogen; CNS = central nervous
3	system; CSF = cerebrospinal fluid; EEG = electroencephalogram; WBC = white blood cell
This document is a draft for review purposes only and does not constitute Agency policy.
C-35	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-9. Occupational epidemiologic studies of RDX: summary of methodologic features
Reference,
Participants,
Consideration


Consideration


setting and
selection,
of likely
Exposure measure and
Outcome
of likely
Analysis and

design
follow-up
selection bias
range
measure
confounding
results details
Sample size
Ma and Li
Details of industrial
Sparse reporting
Details of exposure
Neurobehavioral
No adjustment
Comparisons of
60 exposed;
(1993)
process and subject
of details on
monitoring not reported.
battery
for other risk
mean scaled score
Group A
(China)3
selection framework
subject
Two groups of exposed
administered by
factors (e.g.,
on memory
(n = 30;
Industrial
not reported;
recruitment and
subjects:
trained personnel:
alcohol
retention, letter
26 males,
workers
referents chosen
participation
Group A, intensity,
memory
consumption);
cancellation, or
4 females);
(translated
from same plant;

0.407 (0.332) mg/m3
retention, simple
no consideration
block design test;
Group B
article)
age, employment

[mean (standard
reaction time,
or attempt to
mean time on
(n = 30);

duration, and

deviation)], daily
choice reaction
distinguish TNT
reaction tests;
32 referents

education similar

cumulative,
time, letter

and total
(27 males,

across groups;

2.66 (1.89) mg/m3.
cancellation, and

behavioral score;
5 females)

participation rate

Group B, intensity,
block design

variance (F test),


not reported

0.672 (0.556) mg/m3; daily


linear and




cumulative,


multiple




2.53 (8.40) mg/m3.


regression, and







correlation







analysis

Hathawav
2,022 workers
Potential healthy
Atmospheric samples of all
Liver function,
Workers with
Comparison of
69 RDX
and Buck
(1,017 exposed to
worker effect
operations with potential
renal function,
TNT exposure
mean value;
exposed
(1977)
open explosives

exposure to open
and hematology
excluded from
prevalence of
(43 males,
(United
(TNT, RDX, HMX),

explosives taken in 1975.
tests [blood]
exposed groups
elevated value on
26 females),
States)
1,005 referents) at

Range: not detected to


an individual test
24 RDX/HMX
Ammunition
five U.S. Army

1.57 mg/m3. Seventy



exposed (all
workers
ammunition plants

exposed workers with RDX



males),

(Iowa, Illinois,

at >0.01 mg/m3 [the LOD];



338 referents

Tennessee);

mean: 0.28 mg/m3



(237 males,

participation rate:

[standard deviation not



101 females)

76% exposed, 71%

presented]. Job title used





referents

to initially identify exposed







or unexposed status and







reassigned to one of two




This document is a draft for review purposes only and does not constitute Agency policy.
C-36	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference,
Participants,
Consideration


Consideration


setting and
selection,
of likely
Exposure measure and
Outcome
of likely
Analysis and

design
follow-up
selection bias
range
measure
confounding
results details
Sample size



exposed groups







(nondetected,







>0.01 mg/m3) based on







subject's industrial hygiene







monitoring results.




West and
378 of 404 subjects
Former
Semiquantitative
Abnormal
Cases and
Unadjusted
32 cases
Stafford
(excluded 3 deaths
employees who
assessment; source of
hematology value
controls were
prevalence odds
(29 males,
(1997)
and 23 subjects with
were unable to
industrial hygiene data not
in 1991 survey
not matched and
ratios and 95%
3 females) and
(United
unknown addresses)
work due to
reported. Interviews with
indicating possible
statistical
CIs; analyses
322 controls
Kingdom)
previously studied in
adverse health
current and past
myelodysplasia:
analyses were
limited to males
(282 males,
Ammunition
1991, 32 cases with
outcome were
employees and job title
neutropenia
not adjusted for
because of low
12 females)
workers
abnormal
not included in
analysis were conducted to
(2.0 x 109/L), low
other risk factor
frequency of

(case-
hematology test and
the 1991
identify potential exposure
platelet count
or occupational
exposure to

control
322 controls with
prevalence study
to 100 agents, including
(<150 x 109/L), or
exposures; no
females

study)
normal hematology
test; participation
rate among eligible
subjects: 97% cases,
93% controls

RDX. Exposure surrogate
was >50 hrs in duration
and intensity was low
(1-10 ppm), moderate
(10-100 ppm), or high
(100-1,000 ppm). RDX
exposure prevalence
(males) was 83%.
macrocytosis
(mean
corpuscular
volume = 99 fL or
>6% macrocytes)
consideration or
attempt to
distinguish TNT


1
2	aMa and Li (1993) describe symptoms reported by subjects during a physical examination, but these are not included in the evidence table because responses
3	for individual symptoms were not identified.
4
5	CI = confidence interval; HMX = octahydro-l,3,5,7-tetranitro-l,3,5,7-tetrazocine; LOD = limit of detection; TNT = trinitrotoluene
This document is a draft for review purposes only and does not constitute Agency policy.
C-37	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
C.3. OTHER PERTINENT TOXICITY INFORMATION
C.3.1. Mortality in Animals
Evaluations of the evidence for specific health effects associated with RDX exposure are
provided in Sections 1.2.1 to 1.2.6. In addition to these specific organ/system health effects,
reduced survival associated with RDX exposure has been observed in experimental animals across
multiple studies of varying exposure duration and study design (Table C-10). Evidence pertaining
to mortality in experimental animals exposed to RDX is summarized in Table C-10; studies are
ordered in the evidence table by duration of exposure and then species.
Following chronic dietary exposure, an increased rate of mortality in F344 rats, and in
particular male rats, at 40 mg/kg-day was largely attributed to RDX-related effects on the kidneys
(Levine etal.. 1983)4: see further discussion in Section 1.2.2. In a comparable chronic study, mice
were less sensitive than rats with respect to mortality following RDX exposure. After the high dose
was reduced from 175 to 100 mg/kg-day at week 11 in a 2-year dietary study in B6C3Fi mice
because of high mortality, the mortality curve at 100 mg/kg-day in surviving mice was not
significantly different from the control group for the duration of the 2-year study (Lish etal.. 19841.
The investigators did not identify the probable cause of death at 175 mg/kg-day.
Increased rates of mortality were also observed in experimental animals that ingested RDX
for durations up to 6 months (Lish etal.. 1984: Levine etal.. 1983: Levine etal.. 1981a: Cholakis et
al.. 1980: von Oettingen etal.. 19491. The most detailed data on RDX-related mortality come from a
90-day gavage study in F344 rats by Grouse etal. (20061. Across groups of rats exposed to
8-15 mg/kg-day RDX, pre-term deaths occurred in male rats as early as day 26-78 and in female
rats as early as day 8-16 flohnson. 2015: Grouse etal.. 20061. Treatment-related mortality was
also observed in the dams of rats exposed gestationally by gavage at doses ranging from 20 to
120 mg/kg-day fAngerhofer et al.. 1986: Cholakis etal.. 19801. Deaths were additionally reported
in one of 40 pregnant dams in both 2 and 6 mg/kg-day groups in the rat developmental toxicity
study by Angerhofer et al. (19861
In general, the evidence suggests that mortality occurs at lower doses in rats than in mice
(e.g., comparison of rates from the 2-year dietary studies in mice by Lish etal. f 19841 and in rats by
Levine etal. (198311. and at lower doses following gavage administration than dietary
administration (e.g., comparison of rates from the 13-week rat studies using gavage f Grouse etal..
20061 and dietary f Levine etal.. 198 lal administration). An RDX formulation with a larger particle
size (e.g., ~200 [im) (Cholakis etal.. 19801. which could potentially reduce the ability of RDX to
4 Deaths in high-dose (40 mg/kg-day] male rats were reported as early as week 4 (estimated from Volume 1,
Figure 4 in Levine etal. (198311: the cause of death in rats that died prior to 6 months on study was generally
not determined (Levine et al.. 19831. Survival rates in both male and female rats at doses <8 mg/kg-day RDX
were similar to the control.
This document is a draft for review purposes only and does not constitute Agency policy.
C-38	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
enter the bloodstream, appears to produce less mortality than formulations with finer particle sizes
(e.g., median particle diameter of 20 |im) (Levine etal.. 1981al. There is evidence that mortality
may be associated with nervous system effects; several investigators reported that unscheduled
deaths were frequently preceded by convulsions or seizures (Grouse etal.. 2006: Levine etal.. 1983:
Cholakis etal.. 19801. In a number of studies, treatment-related mortality was observed at doses as
low as doses associated with nervous system effects fCrouse etal.. 2006: Angerhofer etal.. 1986:
Levine etal.. 1983: Levine etal.. 1981a: Cholakis etal.. 1980: von Oettingenetal.. 19491. The
evidence for an association between nervous system effects and mortality is discussed in more
detail in Section 1.2.1, Nervous System Effects.
In humans, there were no reports of mortality in case reports involving workers exposed to
RDX during manufacture or in individuals exposed acutely as a result of accidental or intentional
ingestion (see Appendix C, Section C.2).
Table C-10. Evidence pertaining to mortality in animals3
Reference and study design
Results
Lish etal. (1984)
Doses
LO
O
7.0
35
175/100
Mice, B6C3Fi, 85/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
Mortality at 11 wks (incidence)
M
1/85 0/85
0/85
0/85
30/85
contaminant; 83-89% of particles <66 urn
0,1.5, 7.0, 35, or 175/100 mg/kg-d (high
F
0/85 0/85
0/85
0/85
36/85
dose reduced to 100 mg/kg-d in wk 11 due
Mortality at 6 mo (incidence)
to excessive mortality)
Diet
M
1/85 2/85
3/85
3/85
34/85
2 yrs (mortality incidence also provided for
F
0/85 1/85
0/85
0/85
36/85
mice through week 11 when the high dose
was dropped because of high mortality at
that dose, and from the report of the 6-
Mortality at 2 yrs (incidence)
M
20/65 23/65
25/65
29/65
41/65
month interim sacrifice)
F
16/65 21/65
14/65
21/65
42/65

After the high dose was reduced to 100 mg/kg-d, survival was

similar to controls.



Levine et al. (1983)
Doses
m
o
o
1.5
8.0
40
Rats, F344, 75/sex/group; interim sacrifices
(10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
Mortality at 13 wks (incidence)
M
0/75 0/75
0/75
0/75
0/75
contaminant: 83-89% of particles <66 nm
0, 0.3,1.5, 8.0, or 40 mg/kg-d
F
0/75 0/75
0/75
0/75
0/75
Diet
Mortality in 6-mo interim sacrifice animals (incidencef

2 yrs (mortality incidence also provided for
"includes spontaneous death and moribund sacrifice

mice through week 13, and from the report
of the 6-month scheduled sacrifice)
M
0/75 0/75
0/75
0/75
5/75

F
0/75 0/75
0/75
0/75
0/75

Mortality at 2 yrs (incidence)
This document is a draft for review purposes only and does not constitute Agency policy.
C-39	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results

M
17/55
19/55 30/55* 26/55

51/55*

F
12/55
10/55 13/55 14/55

27/55*
Cholakis et al. (1980)
Doses
0

79.6
147.8


256.7
Mice, B6C3Fi, 10-12/sex/group
88.6% pure, with 9% HMX and 2.2% water
as contaminants; ~200 urn particle size
Mortality (incidence)
M
0/10

0/10
0/10


4/10*
0, 40, 60, or 80 mg/kg-d for 2 wks followed
by 0, 320,160, or 80 mg/kg-d (TWA doses
F
0/11

0/12
0/10


2/12
of 0, 79.6,147.8, or 256.7 mg/kg-d for








males and 0, 82.4,136.3, or 276.4 mg/kg-d








for females)b








Diet








13 wks








Cholakis et al. (1980)
Doses
0
10
14
20
28
40
Rats, F344,10/sex/group
88.6% pure, with 9% HMX and 2.2% water
as contaminants; ~200 nm particle size
Mortality (incidence)
M
0/10
0/10
0/10
0/10
0/10
0/10
0,10,14, 20, 28, or 40 mg/kg-d
Diet
13 wks
F
1/10
0/10
0/10
0/10
0/10
0/10

(accidental
death)






Cholakis et al. (1980)
Doses
0

5
16


50
Rats, CD, two-generation study;
F0: 22/sex/group; Fl: 26/sex/group;
Mortality in F0 adults (incidence)c








F2: 10/sex/group
M (F0)
0/22

0/22
0/22


2/22
88.6% pure, with 9% HMX and 2.2% water
as contaminants; ~200 nm particle size
F (F0)
0/22

0/22
0/22


6/22
F0 and Fl parental animals: 0, 5,16, or
M + F
0/44

0/44
0/44


8/44*
50 mg/kg-d
(F0)







Diet








F0 exposure: 13 wks pre-mating, and during








mating, gestation, and lactation of Fl;








Fl exposure: 13 wks after weaning, and








during mating, gestation, and lactation of








F2; F2 exposure: until weaning








Crouse et al. (2006)
Doses
0 4
8
10
12
15
Rats, F344,10/sex/group
99.99% pure
0, 4, 8,10,12, or 15 mg/kg-d
Mortality (incidence)
M
0/10 0/10
1/10
3/10
2/10
3/10
Gavage
13 wks
F
0/10 0/10
1/10
2/10
5/10
4/10
Levine et al. (1990): Levine et al. (1981a):
Doses
0 10
30
100
300
600
Levine et al. (1981b)d
Rats, F344,10/sex/group; 30/sex for
control
Mortality (incidence)e
M
0/30 0/10
0/10
8/10
10/10
10/10
This document is a draft for review purposes only and does not constitute Agency policy.
C-40	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
84.7 ± 4.7% purity, ~10% HMX, median
particle diameter 20 urn, ~90% of particles
<66 urn
0,10, 30,100, 300, or 600 mg/kg-d
Diet
13 wks
F
0/30 1/10 0/10 5/10 10/10 10/10
von Oettingen et al. (1949)
Rats, sex/strain not specified, 20/group
90-97% pure, with 3-10% HMX; particle
size not specified
0,15, 25, or 50 mg/kg-d
Diet
13 wks
Doses
0 15 25 50
Mortality (incidence)

0/20 1/20 8/20 8/20
(probably
not related
to RDX)
Hart (1974)
Dogs, Beagle, 3/sex/group
Pre-mix with ground dog chow containing
20 mg RDX/g-chow, 60 g dog food; purity
and particle size not specified
0, 0.1,1, or 10 mg/kg-d
Diet
13 wks
Doses
o
1
1
1
o
o
Mortality (incidence)
M
F
0/3 0/3 1/3 0/3
(not related to
RDX)
0/3 0/3 0/3 0/3
Martin and Hart (1974)
Monkeys, Cynomolgus or Rhesusf,
3/sex/group
Purity and particle size not specified
0, 0.1,1, or 10 mg/kg-d
Gavage
13 wks
Doses
o
1
1
1
o
o
Mortality (incidence)
M
F
0/3 0/3 0/3 0/3
0/3 0/3 0/3 1/3
(animal exhibited
neurological
effects; euthanized)
von Oettingen et al. (1949)
Dogs, breed not specified, 5 females/group
(control); 7 females/group (exposed)
90-97% pure, with 3-10% HMX; particle
not specified
0 or 50 mg/kg-d
Diet
6 d/wk for 6 wks
Doses
0 50
Mortality (incidence)
F
0/5 1/7
MacPhail et al. (1985)
Rats, Sprague-Dawley derived CD,
8-10 males or females/group
Purity 84 ± 4.7%; <66 nm particle size
0,1, 3, or 10 mg/kg-d
Gavage
30 d
No mortality was reported (incidence data were not provided).
This document is a draft for review purposes only and does not constitute Agency policy.
C-41	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Cholakis et al. (1980)
Doses
0
0.2
2.0
20
Rabbits, New Zealand White (NZW),
11-12 pregnant females/group
88.6% pure, with 9% HMX and 2.2% water
Mortality (incidence)
F
0/11
0/11
0/11
0/12
as contaminants; ~200 nm particle size





0, 0.2, 2.0, or 20 mg/kg-d





Diet





GDs 7-29





Cholakis et al. (1980)
Doses
0
0.2
2.0
20
Rats, F344, 24-25 females/group
88.6% pure, with 9% HMX and 2.2% water
as contaminants
Mortality (incidence)
F
0/24
0/24
0/24
5/24
0, 0.2, 2.0, or 20 mg/kg-d



(1 rat accidentally
Gavage



killed; removed from
GDs 6-19




analysis)
Angerhofer et al. (1986) (range-finding
Doses
0 10
20
40
80 120
study)
Rats, Sprague-Dawley, 6 pregnant
females/group
Mortality (incidence)
F
0/6 0/6
0/6
6/6
6/6 6/6
Purity 90%; 10% HMX and 0.3% acetic acid





occurred as contaminants





0,10, 20, 40, 80, or 120 mg/kg-d





Gavage





GDs 6-15





Angerhofer et al. (1986)
Doses
0
2
6
20
Rats, Sprague-Dawley, 39-51 mated
females/group
Purity 90%; 10% HMX and 0.3% acetic acid
Mortality (incidence)
F
0/39
1/40
1/40
16/51
occurred as contaminants
0, 2, 6, or 20 mg/kg-d
Gavage
GDs 6-15


Not reported whether
deaths at 2 and 6 mg/kg-d



related or likely related to
RDX exposure

^Statistically significant (p < 0.05) based on analysis by the study authors.
aThe 2-year rat study by Hart (1976) was not included in this evidence table because a malfunctioning heating
system incident resulted in the premature deaths of 59 animals on study days 75-76 across groups, thereby
confounding mortality findings.
bDoses were calculated by the study authors.
cData for male and female rats were combined for statistical analysis.
dLevine et al. (1981a) is a laboratory report of a 13-week study of RDX in F344 rats; two subsequently published
papers (Levine et al., 1990; Levine et al., 1981b) present subsets of the data provided in the full laboratory report.
eAnimals receiving 300 mg/kg-day died by week 3 of the study; animals receiving 600 mg/kg-day died by week 1 of
the study.
the species of monkey used in this study was inconsistently reported in the study as either Cynomolgus (in the
methods section) or Rhesus (in the summary).
TWA = time-weighted average
This document is a draft for review purposes only and does not constitute Agency policy.
C-42	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
C.3.2. Other Noncancer Effects
There are isolated reports of RDX inducing systemic effects in several organs/systems,
including the eyes and the musculoskeletal, cardiovascular, immune, and GI systems. However,
there is less evidence for these effects compared to organ systems described in Section 1.2. Overall,
at the present time, the evidence does not support identifying these other systemic effects as
human hazards of RDX exposure. Summaries of the evidence for other systemic effects in humans
and animals are shown in Tables C-ll and C-12, respectively. Experimental animal studies are
ordered in the evidence table by type of effect, and then by duration of exposure and by species.
Ocular Effects
There are no reports of ocular effects in human case reports or epidemiological studies. In
experimental animals, evidence of ocular effects comes from cataract findings in one 2-year
bioassay. Specifically, the incidence of cataracts was 73% in female F344 rats exposed to
40 mg/kg-day RDX in the diet for 2 years, compared with 32% in the control group (Levine etal..
19831. After 76 weeks of exposure, the incidence of cataracts in female rats at 40 mg/kg-day (23%)
was also elevated compared to controls (6%). The incidence of cataracts was not increased in RDX-
exposed male rats in the same study (Levine etal.. 19831. and other studies have not observed
ocular effects associated with RDX exposure. Only two rats (dose groups not reported) were
observed to have mild cataracts in a 90-day study of male and female F344 rats exposed to RDX at
doses up to 15 mg/kg-day by gavage; however, the authors noted that these observations are
common in F344 rats at 4 months of age and should not be attributed to treatment (Grouse etal..
20061. Furthermore, cataracts were not observed in male or female F344 rats exposed to
40 mg/kg-day RDX by diet for 90 days fCholakis etal.. 19801 or in male or female B6C3Fi mice
exposed to RDX in the diet for 2 years at doses up to approximately 100 mg/kg-day (Lish etal..
19841. A statistically significant increase in the incidence of cataracts in male mice was initially
noted by Lish etal. (19841. but was not confirmed when mice used for orbital bleedings were
excluded from the analysis, suggesting that the effect was not treatment related. No ocular effects
were observed in monkeys exposed by gavage for 90 days at doses up to 10 mg/kg-day (Martin and
Hart. 19741.
In summary, the incidence of cataracts was statistically significantly increased in high-dose
female rats in one chronic oral study; however, this finding was not reproduced in other subchronic
and chronic studies in rats or mice.
Cardiovascular Effects
Human evidence for cardiovascular effects is limited to case reports that include
observations of transient arterial hypertension in male workers following inhalation of RDX during
manufacturing (Barsotti and Crotti. 19491. sinus tachycardia, and in one instance, premature
ventricular beats in five men following accidental ingestion of RDX at 37-250 mg/kg body weight
fKiiciikardali etal.. 20031 (see Appendix C, Section C.2).
This document is a draft for review purposes only and does not constitute Agency policy.
C-43	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Inconsistent observations of cardiovascular effects have been reported in animal studies.
An increase in the relative heart-to-body weight ratio was observed at the highest dose tested in
B6C3Fi mice (male: 13%; female: 17%) and F344 rats (male: 22%; female: 15%) following chronic
dietary administration of RDX (Lish etal.. 1984: Levine etal.. 1983): however, these doses also
resulted in reductions in body weight in both males and females. Dose-related decreases in
absolute heart weight in rats were reported in some subchronic (dietary) studies f Levine etal..
1990: Levine etal.. 1981a. b; Cholakis etal.. 1980). whereas little change or modest increases in
absolute heart weight were observed in other subchronic studies in rats or mice (Grouse etal..
2006: Cholakis etal.. 1980). A subchronic study in male dogs reported a 31% increase in absolute
heart weight at the highest dose tested (10 mg/kg-day) fHart. 19741.
Evidence for changes in histopathology associated with RDX exposure is limited to findings
of an increased incidence of focal myocardial degeneration in female rats (6/10 versus 2/10,
respectively) and male mice (5/10 versus 0/10, respectively) compared with controls following
exposure to RDX in the diet for 90 days (Cholakis etal.. 19801. With the exception of one male
mouse, the severity of the lesion was characterized as minimal. In each study, the finding of
myocardial degeneration was limited to one sex and to the high-dose group only; the high dose in
the male mouse study caused 40% mortality. Other studies in monkeys f Martin and Hart. 19741
and rats (von Oettingen etal.. 1949) reported no observable cardiovascular effects.
In summary, evidence for cardiovascular effects associated with RDX exposure consists of
two case reports of cardiovascular effects following acute exposure, inconsistent findings of
changes in heart weight in experimental animals, and one report of minimal histopathological
changes in a 90-day rat study that was not confirmed in other toxicity studies.
Musculoskeletal Effects
Evidence of musculoskeletal effects in humans consists of case reports that include
observations of muscle twitching, myalgia/muscle soreness, and muscle injury as indicated by
elevated levels of aspartate aminotransferase (AST) or myoglobinuria (Kiiciikardali etal.. 2003:
Hett and Fichtner. 2002: Hollander and Col bach. 1969: Stone etal.. 1969: Merrill. 1968) (see
Appendix C, Section C.2). Histological evaluations of musculature or skeletal tissue did not reveal
any alterations in mice (Lish etal.. 1984) or rats (Levine etal.. 1983: Hart. 1976) following chronic
oral exposure to RDX, in mice and rats following subchronic exposure fCholakis etal.. 1980). or in
dogs following a 90-day dietary exposure fHart. 19741.
Immune System Effects
RDX is structurally similar to various drugs known to induce the autoimmune disorder
systemic lupus erythematosus (SLE). Based on the initial identification of three cases of SLE at
one U.S. Army munitions plant, further investigation was conducted on a population of
69 employees at five U.S. Army munitions plants with potential exposure to RDX fHathawav and
Buck. 19771: no additional cases of SLE were identified. Increased WBC counts have been reported
This document is a draft for review purposes only and does not constitute Agency policy.
C-44	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
in some case reports of individuals (troops during the Vietnam war) who ingested or inhaled RDX
or C-4 (91% RDX) (Knepshield and Stone. 1972: Hollander and Colbach. 1969: Stone et al.. 1969:
Merrill. 19681.
In animal studies, increased WBC count in female rats following subchronic dietary
exposure to RDX was the only dose-related immune effect reported fLevine et al.. 1990: Levine et
al.. 1981a. b); WBC counts in male rats were unaffected. Conversely, decreased WBC counts were
reported in male and female rats in a 2-year study (Hart. 19761. Changes in spleen weights were
observed across studies, but the responses were not consistent and did not appear to be dose-
related. For example, in 90-day studies, Cholakis etal. (19801 identified a statistically significant
decrease in absolute spleen weight in female F344 rats at 40 mg/kg-day, while Crouse etal. f2 0061
observed a statistically significant increase in spleen weight at >10 mg/kg-day. Across studies,
there was no significant or dose-dependent pattern of response to suggest that the WBC changes
reflect RDX-induced immunotoxicity. No dose-related immune effects from oral exposure to RDX
were observed in other animal studies, including a 90-day study in F344 rats specifically designed
to evaluate immunotoxicity (parameters included evaluation of red blood cell [RBC] and WBC
populations, proportion of cell surface markers, cellularity in proportion to organ weight, B and
T cells in the spleen, and CD4/CD8 antigens of maturing lymphocytes in the thymus) fCrouse etal..
20061. Routine clinical and histopathology evaluations of immune-related organs in a two-
generation study in rats fCholakis etal.. 19801 and chronic studies in rats fLevine etal.. 19831 and
mice (Lish etal.. 19841 provide no evidence of immunotoxicity associated with oral (dietary)
exposure to RDX.
In summary, evidence for immunotoxicity associated with RDX exposure is limited to
findings from one study of increased WBC counts in female rats (Levine etal.. 1981a. b). Evidence
that RDX is not immunotoxic comes from several animal studies, including other repeat-dose oral
studies in mice and rats fCrouse etal.. 2006: Lish etal.. 1984: Levine etal.. 1983: Cholakis etal..
19801.
Gastrointestinal Effects
Clinical signs of nausea and/or vomiting have been frequently identified in case reports of
accidental or intentional RDX poisonings, and have generally been concurrent with severe
neurotoxicity fKasuske etal.. 2009: Davies etal.. 2007: Kiiciikardali et al.. 2003: Hett and Fichtner.
2002: Ketel and Hughes. 1972: Knepshield and Stone. 1972: Hollander and Colbach. 1969: Stone et
al.. 1969: Merrill. 1968: Kaplan et al.. 1965: Barsotti and Crotti. 19491 (see Appendix C, Section C.2).
In animal studies, vomiting has also been observed following oral exposure in swine (single-dose
study) (Musick et al.. 20101. dogs (Hart. 19741. and monkeys f Martin and Hart. 19741. One
subchronic oral (diet) rat study from the early literature reported congestion of the GI tract at
doses also associated with elevated mortality (von Oettingen etal.. 19491: however, none of the
subsequent subchronic or chronic animal studies reported histological changes of the GI tract
This document is a draft for review purposes only and does not constitute Agency policy.
C-45	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
related to RDX administered via gavage or the diet (Grouse etal.. 2006: Lish etal.. 1984: Levine et
al.. 1983: Hart. 1974: Martin and Hart. 19741.
In summary, evidence for GI tract effects associated with RDX exposure consists largely of
reports of nausea and vomiting in humans acutely exposed to RDX and similar reports of vomiting
in swine, dogs, and monkeys. In general, histopathological changes have not be reported in
experimental animals exposed to RDX in the diet.
Hematological Effects
Elevated prevalence odds ratios (ORs) for hematological abnormalities (i.e., neutropenia,
low platelet count, or macrocytosis; see Table C-ll for criteria used to define abnormal) were
observed in a case-control study of men (24 cases, 199 controls) exposed to RDX in ordnance
factories fWest and Stafford. 19971 (see Table C-ll). The prevalence OR for an association between
RDX exposure and hematological abnormalities was 1.7 (95% confidence interval [CI] 0.7-4.2) for
men with >50 hours of low-intensity exposure (based on 22 cases), while the prevalence OR was
1.2 (95% CI 0.3-5.3) for men with >50 hours of high-intensity exposure (based on 2 cases). The
ORs from this study must be interpreted with caution given the small sample size, wide CIs, and
lack of identification of co-exposures. No changes in hematological parameters (including
hemoglobin, hematocrit, and reticulocyte count) were observed in a cross-sectional epidemiologic
study of 69 workers exposed to RDX by inhalation (RDX exposure range: undetectable
[<0.01 mg/m3] to 1.6 mg/m3) (Hathaway and Buck. 19771. Humans who ingested or inhaled
unknown amounts of RDX or C-4 (~91% RDX) for an acute duration displayed temporary
hematological alterations, including anemia, decreased hematocrit, hematuria, and
methemoglobinemia (Kasuske etal.. 2009: Kiiciikardali etal.. 2003: Knepshield and Stone. 1972:
Hollander and Col bach. 1969: Stone etal.. 1969: Merrill. 19681. In other case reports, normal blood
counts were observed in accidentally exposed individuals fTestud et al.. 1996a: Goldberg etal..
1992: Woody etal.. 1986: Ketel and Hughes. 1972: Kaplan etal.. 19651 (see Appendix C,
Section C.2).
In animals, hematological alterations were observed following oral exposure in chronic and
subchronic studies in both sexes of rats (F344 or Sprague-Dawley) and B6C3Fi mice (see
Table C-12). Increases in platelet count were observed in male and female mice and rats in some
subchronic and chronic studies at doses ranging from 0.3 to 320 mg/kg-day (Lish etal.. 1984:
Levine etal.. 1983: Cholakis etal.. 19801: however, changes were generally inconsistent across
studies and were not generally dose-dependent. Similarly, decreased hemoglobin levels/anemia
were observed in some chronic and subchronic studies (Levine etal.. 1983: Cholakis etal.. 1980:
von Oettingen etal.. 19491. particularly at doses >15 mg/kg-day, but trends in hemoglobin levels
across studies did not show a consistent relationship with dose. Other hematological parameters,
including WBC counts, reticulocyte counts, and hematocrit, showed conflicting results between
studies, marginal responses, or inconsistent changes with increasing dose. Other subchronic
This document is a draft for review purposes only and does not constitute Agency policy.
C-46	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	studies in rats and dogs (Grouse etal.. 2006: Hart. 1974: von Oettingen et al.. 19491 did not identify
2	any changes in hematological parameters.
3	In summary, evidence for hematological effects associated with RDX exposure in humans
4	comes from several case reports that found transient fluctuations in hematological endpoints after
5	acute exposures. Hematological findings from the case-control study and the cross-sectional study
6	were not consistent The small number of cases or exposed individuals, respectively, from the case-
7	control and cross-sectional study may contribute to the difficulty in interpreting the results across
8	studies (Table C-ll). In general, animal studies of chronic and subchronic durations showed no
9	consistent, dose-related pattern of increase or decrease in hematological parameters.
10	Table C-ll. Evidence pertaining to other noncancer effects (hematological) in
11	humans
Reference and study design
Results
Hematological effects
West and Stafford (1997) (United Kingdom)
Case-control study of ordnance factory
workers, 32 cases with abnormal and
322 controls with normal hematology test
drawn from 1991 study of 404 workers at
ammunitions plant; participation rate 97% of
cases, 93% of controls. Analysis limited to men
(29 cases, 282 controls). Analysis specific to
RDX: 22 low- and 2 high-intensity cases;
182 low- and 17 high-intensity controls.
Exposure measures: Exposure determination
based on employee interviews and job title
analysis; data included frequency (hrs/d, d/yr),
duration (yrs), and intensity (low [1-10 ppm],
moderate [10-100 ppm], and high
[100-1,000 ppm], based on ventilation
considerations).
Effect measures: Hematology tests; blood
disorder defined as neutropenia (2.0 x 109/L),
low platelet count (<150 x 109/L), or
macrocytosis (mean corpuscular volume = 99 fl
or >6% macrocytes).
Analysis: Unadjusted OR.
Hematological abnormality (neutropenia, low platelet count, or
macrocytosis) (OR; 95% CI [number of exposed cases])
Low intensity, 50-hr-duration 1.7; 0.7,4.2 [22]
Medium intensity, 50-hr duration 1.6; not reported [not
reported]
High intensity, 50-hr duration 1.2; 0.3, 5.3 [2]
This document is a draft for review purposes only and does not constitute Agency policy.
C-47	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Hathawav and Buck (1977) (United States)
Hematology tests in men (mean; standard deviation not
reported)
Cross-sectional study, 2,022 workers,
1,491 participated (74% response rate).
Analysis limited to whites; 69 exposed to RDX
alone and 24 exposed to RDX and HMX; 338 not
exposed to RDX, HMX, or TNT.
Exposure measures: Exposure determination
based on job title and industrial hygiene
evaluation. Exposed subjects assigned to two
groups: 0.01 mg/m3 (mean for
employees with exposures >LOD: 0.28 mg/m3).
Effect measures: Hematology tests.
Analysis: Types of statistical tests were not
reported (assumed to be t-tests for comparison
of means and x2 tests for comparison of
proportions).
Test
RDX exposed*
Undetected
Referent (0.01 mg/m3
(n = 237) (n = 22) (n = 45)
Hemoglobin
Hematocrit
Reticulocyte
count
15.2 14.7 15.2
42 45.6 47
0.7 0.9 0.7
includes both workers exposed to RDX alone and RDX and
HMX.
No differences were statistically significant. Similar results in
women.
Hematology tests in men (prevalence of abnormal values)
Test
(abnormal
range)
RDX exposed*
Undetected
Referent (0.01 mg/m3
Hemoglobin
(<14)
Hematocrit
(<40)
Reticulocyte
count (>1.5)
15/237 3/22 4/45
1/237 1/22 1/45
18/237 3/22 2/45
*lncludes both workers exposed to RDX alone and RDX and
HMX.
No differences were statistically significant. Similar results in
women.
This document is a draft for review purposes only and does not constitute Agency policy.
C-48	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-12. Evidence pertaining to other noncancer effects in animals3
Reference and study design
Results
Ocular effects
Lish etal. (1984)
Mice, B6C3Fi, 85/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 urn
0,1.5, 7.0, 35, or 175/100 mg/kg-d (high
dose reduced to 100 mg/kg-d in wk 11
due to excessive mortality)
Diet
2 yrs
Doses
0 1.5 7.0 35 175/100
Cataracts; 103 wks (incidence)b
M
F
2/47 2/41 0/41 2/37 2/16
2/50 1/37 6/52 0/46 1/26
Levine et al. (1983)
Rats, F344, 75/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 nm
0, 0.3,1.5, 8.0, or 40 mg/kg-d
Diet
2 yrs
Doses
o
o
00
LO
m
o
o
Cataracts; 103 wks (incidence)
M
F
8/40 6/39 6/31 8/35 2/6
14/44 4/48 11/44 8/43 22/30*
Cholakis et al. (1980)
Rats, F344,10/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 nm
particle size
0,10,14, 20, 28, or 40 mg/kg-d
Diet
13 wks
No ocular effects were observed (gross examination of eye was
performed in all animals, and microscopic examination was performed
in control and 40 mg/kg-d animals).
Crouse et al. (2006)
Rats, F344,10/sex/group
99.99% pure
0, 4, 8,10,12, or 15 mg/kg-d
Gavage
13 wks
No ocular effects were observed (ophthalmic examinations were
performed in all animals within 1 wk of sacrifice, and microscopic
examination of the eye was performed in control and
15 mg/kg-d animals).
Martin and Hart (1974)
Monkeys, Cynomolgus or Rhesus0,
3/sex/group
Purity of test material not specified
0, 0.1,1, or 10 mg/kg-d
Gavage
13 wks
No ocular effects were observed (ophthalmoscopic examination was
performed at the end of exposure).
This document is a draft for review purposes only and does not constitute Agency policy.
C-49	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Cardiovascular effects
Lish etal. (1984)
Mice, B6C3Fi, 85/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 urn
0,1.5, 7.0, 35, or 175/100 mg/kg-d (high
dose reduced to 100 mg/kg-d in wk 11
due to excessive mortality)
Diet
2 yrs
Doses
0 1.5 7.0 35 175/100
Absolute heart weight; 104 wks (percent change compared to control)
M
F
0% 4% 4% 5% 7%
0% 1% 5% 2% -5%
Relative heart-to-body weight; 104 wks (percent change compared to
control)
M
F
0% 7% 5% 5% 13%*
0% 0% 6% 4% 17%*
Body weight was significantly lower at termination in males and
females exposed to 175/100 mg/kg-d (-5 and -19%, respectively).
Hart (1976)
Rats, Sprague-Dawley, 100/sex/group
Purity and particle size not specified
0,1.0, 3.1, or 10 mg/kg-d
Diet
2 yrs
Doses
o
1
1
rn
o
O
Myocardial fibrosis (percent incidence; number not reported)
M
F
20% - - 5%
5% - - 1%
Endocardial disease (percent incidence; number not reported)
M
F
1% - - 3%
0% - - 0%
Absolute heart weight; 104 wks (percent change compared to control)
M
F
0% -6% -2% -5%
0% 13% 3% 15%
Relative heart-to-body weight; 104 wks (percent change compared to
control)
M
F
0% -2% 4% 1%
0% 23% 13% 27%
Levine et al. (1983)
Rats, F344, 75/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 nm
0, 0.3,1.5, 8.0, or 40 mg/kg-d
Diet
2 yrs
Doses
o
o
00
LO
m
o
o
Absolute heart weight; 104 wks (percent change compared to control)
M
F
0% 3% -2% -2% 1%
0% -1% 0% -4% -3%
Relative heart-to-body weight; 104 wks (percent change compared to
control)
M
F
0% 2% 6% 0% 22%
0% -2% 3% -1% 15%
This document is a draft for review purposes only and does not constitute Agency policy.
C-50	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Cholakis et al. (1980)
Mice, B6C3Fi, 10-12/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 urn
particle size
Experiment 1: 0,10,14, 20, 28, or
40 mg/kg-d
Diet
13 wks
Experiment 2: 0, 40, 60, or 80 mg/kg-d
for 2 wks followed by 0, 320,160, or
80 mg/kg-d (TWA doses of 0, 79.6,
147.8, or 256.7 mg/kg-d for males and 0,
82.4,136.3, or 276.4 mg/kg-d for
females)d
Diet
13 wks
Doses
0 10 14 20 28 40
Absolute heart weight (percent change compared to control)
M
F
0% - 7% 7%
0% - - - 0% 0%
Relative heart weight (percent change compared to control)
M
F
0% - - - 6% 0%
0% - - - -4% 0%
Doses
0 80 160 320
Focal myocardial degeneration (incidence)
M#
F+
0/10 - - 5/10*-*
0/11 - - 2/11
Absolute heart weight (percent change compared to control)
M
F
0% 0% 0% 8%
0% 0% 0% 8%
Relative heart-to-body weight (percent change compared to control)
M
F
0% 0% -2% 6%
0% 0% -2% 2%
"Includes one affected and three unaffected animals that died
prematurely.
includes one unaffected animal that died prematurely.
*Minimal severity in four rats, moderate in one.
Cholakis et al. (1980)
Rats, F344,10/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 nm
particle size
0,10,14, 20, 28, or 40 mg/kg-d
Diet
13 wks
Doses
0 10 14 20 28 40
Focal myocardial degeneration, minimal severity (incidence)
M
F
3/10 - 1/10
2/10 - 6/10
Absolute heart weight (percent change compared to control)
M
F
0% - - - 0% -8%*
0% - -6% -11%*
Relative heart-to-body weight (percent change compared to control)
M
F
0% - - - 3% 0%
0% - - - -3% -8%
Relative heart-to-brain weight (percent change compared to control)
M
F
0% - -4% -10%*
0% - -5% -11%*
This document is a draft for review purposes only and does not constitute Agency policy.
C-51	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Cholakis et al. (1980)
Rats, CD, two-generation study; FO:
22/sex/group; Fl: 26/sex/group; F2:
10/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 urn
particle size
FO and Fl parental animals: 0, 5,16, or
50 mg/kg-d
Diet
FO exposure: 13 wks pre-mating, and
during mating, gestation, and lactation
of Fl; Fl exposure: 13 wks after
weaning, and during mating, gestation,
and lactation of F2; F2 exposure: until
weaning
No cardiac effects were observed (microscopic examination of heart
was performed in randomly selected F2 animals).
Doses
0 5 16 50
Absolute heart weight (percent change compared to control)
F2 M
F2 F
0% 3.2% -6.5%
0% 15% -3.7%
Crouse et al. (2006)
Rats, F344,10/sex/group
99.99% pure
0, 4, 8,10,12, or 15 mg/kg-d
Gavage
13 wks
Doses
0 4 8 10 12 15
Cardiomyopathy (incidence)
M
F
2/10 - - - - 3/8
0/10 - - - - 1/6
Absolute heart weight (percent change compared to control)
M
F
0% -2% -7% -1% 1% 11%
0% -2% 0% 8% 7% 6%
Relative heart-to-body weight (percent change compared to control)
M
F
0% 4% 2% 1% -1% 8%
0% -2% -7% -6% -9% -16%*
Levine et al. (1990): Levine et al.
(1981a): Levine et al. (1981b)e
Rats, F344,10/sex/group; 30/sex for
control
84.7 ± 4.7% purity, ~10% HMX, median
particle diameter 20 nm, ~90% of
particles <66 nm
0,10, 30,100, 300, or 600 mg/kg-d
Diet
13 wks
All animals in the 300 and 600 mg/kg-d groups died prior to study
termination.
Doses
0 10 30 100 300 600
Chronic focal myocarditis (incidence)
M
F
8/30 8/10 6/10 1/10 1/10 0/10
8/30 3/10 1/10 1/10 1/10 1/9
Absolute heart weight (percent change compared to control)
M
F
0% -2% -10% -15%
0% -3% 0% -5%
Relative heart-to-body weight (percent change compared to control)
M
F
0% 2% -4% 3%
0% -2% 0% -3%
This document is a draft for review purposes only and does not constitute Agency policy.
C-52	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
von Oettingen et al. (1949)
Rats (sex/strain not specified); 20/group
Purity and particle size not specified
0,15, 25, or 50 mg/kg-d
Diet
13 wks
The study authors reported that there were no cardiac effects
(microscopic examination of the heart was performed in all rats; data
were not shown).
Hart (1974)
Dogs, Beagle, 3/sex/group
Pre-mix with ground dog chow
containing 20 mg RDX/g-chow, 60 g dog
food; purity and particle size not
specified
0, 0.1,1, or 10 mg/kg-d
Diet
13 wks
Doses
o
1
1
1
o
o
Focal hyalinization of the heart (incidence)
M
F
0/3 - - 0/3
0/3 - - 1/3
Absolute heart weight (percent change compared to control)
M
F
0% - - 31%
0% - - 5.7%
Martin and Hart (1974)
Monkeys, Cynomolgus or Rhesus0,
3/sex/group
Purity of test material not specified
0, 0.1,1, or 10 mg/kg-d
Gavage
13 wks
Doses
o
1
1
1
o
o
Myocarditis (percent change compared to control)
M
F
1/3 - - 1/3
0/3 - - 0/3
Absolute heart weight (percent change compared to control)
M
F
0% 7% -1% 5%
0% 10% 12% -12%
Immune effects
Lish et al. (1984)
Mice, B6C3Fi, 85/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 nm
0,1.5, 7.0, 35, or 175/100 mg/kg-d (high
dose reduced to 100 mg/kg-d in wk 11
due to excessive mortality)
Diet
2 yrs
No immune effects were observed with routine hematology, clinical
chemistry, or histopathology evaluations.
Doses
0 1.5 7.0 35 175/100
WBC count; 105 wks (percent change compared to control)
M
F
0% -13% -8% -16% -30%
0% 12% 39%* 28% 0%
Absolute spleen weight; 105 wks (percent change compared to
control)
M
F
0% 24% 31% -10% -28%
0% 4% 15% -17% 16%
Relative spleen weight; 105 wks (percent change compared to control)
M
F
0% 26% 32% -11% -21%
0% 4% 15% -17% 44%
This document is a draft for review purposes only and does not constitute Agency policy.
C-53	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Hart (1976)
Rats, Sprague-Dawley, 100/sex/group
Purity and particle size not specified
0,1.0, 3.1, or 10 mg/kg-d
Diet
2 yrs
Doses
o
l->
o
w
h->
o
WBC count; 104 wks (percent change compared to control)
M
F
0% -13% -22%* -34%*
0% 5% -32%* -12%
Absolute spleen weight; 104 wks (percent change compared to
control)
M
F
0% -11% -16% -4%
0% 58% 8% 37%
Relative spleen weight; 104 wks (percent change compared to control)
M
F
0% -11% -14% 1%
0% 77% 19% 55%
Levine et al. (1983)
Rats, F344, 75/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 urn
0, 0.3,1.5, 8.0, or 40 mg/kg-d
Diet
2 yrs
No immune effects were observed with routine hematology, clinical
chemistry, or histopathology evaluations.
Doses
o
o
00
LO
m
o
o
WBC count; 105 wks (percent change compared to control)
M
F
0% -11% 103%f 184%f 15%
0% 7% 12% 354%f 251%f
Absolute spleen weight; 105 wks (percent change compared to
control)
M
F
0% 5% -10% -32% -49%
0% -28% -44% -35% 17%
Relative spleen weight; 105 wks (percent change compared to control)
M
F
0% 9% 4% -29% -38%
0% -34% -45% -36% 9%
Cholakis et al. (1980)
Mice, B6C3Fi, 10-12/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 nm
particle size
Experiment 1: 0,10,14, 20, 28, or
40 mg/kg-d
Diet
13 wks
Doses
0 10 14 20 28 40
Absolute spleen weight (percent change compared to control)
M
F
0% - 18% 13%
0% - - - -2% -8%
Relative spleen weight (percent change compared to control)
M
F
0% - 24% 14%
0% - - - -3% -5%
This document is a draft for review purposes only and does not constitute Agency policy.
C-54	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Experiment 2: 0, 40, 60, 80 mg/kg-d for
2 wks followed by 0, 320,160, or
80 mg/kg-d (TWA doses of 0, 79.6,
147.8, or 256.7 mg/kg-d for males and 0,
82.4,136.3, or 276.4 mg/kg-d for
females)d
Diet
13 wks
Doses
0 80 160 320
WBC count (percent change compared to control)
M
F
0% -27% -12% 30%
0% -17% 3% -3%
Absolute spleen weight (percent change compared to control)
M
F
0% 17% 0% -17%
0% -22% 0% 0%
Relative spleen weight (percent change compared to control)
M
F
0% 25% 5% 0%
0% -12% 0% -3%
Cholakis et al. (1980)
Rats, F344,10/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 nm
particle size
0,10,14, 20, 28, or 40 mg/kg-d
Diet
13 wks
Doses
0 10 14 20 28 40
WBC count (percent change compared to control)
M
F
0% - -12% 7%
0% - 17% 30%
Absolute spleen weight (percent change compared to control)
M
F
0% - 2% -4%
0% - -10% -12%*
Relative spleen weight (percent change compared to control)
M
F
0% - - - 5% 5%
0% - - - -8% -8%
Cholakis et al. (1980)
Rats, CD, two-generation study; F0:
22/sex/group; Fl: 26/sex/group;
F2: 10/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 nm
particle size
F0 and Fl parental animals: 0, 5,16, or
50 mg/kg-d
Diet
F0 exposure: 13 wks pre-mating, and
during mating, gestation, and lactation
of Fl; Fl exposure: 13 wks after
weaning, and during mating, gestation,
and lactation of F2; F2 exposure: until
weaning
No immune effects were observed upon routine histopathology
evaluation.
This document is a draft for review purposes only and does not constitute Agency policy.
C-55	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Crouse et al. (2006)
Rats, F344,10/sex/group
99.99% pure
0, 4, 8,10,12, or 15 mg/kg-d
Gavage
13 wks
No effects were observed on thymus or spleen histology, RBC or WBC
populations, or lymphocyte populations.
Doses
0 4 8 10 12 15
WBC count (percent change compared to control)
M
F
0% -5% -12% -7% 1% -3%
0% 22% 45% 12% 52% 29%
Absolute spleen weight (percent change compared to control)
M
F
0% -3% -6% 3% 1% 5%
0% 1% 8% 23%* 17%* 24%*
Relative spleen weight (percent change compared to control)
M
F
0% 3% 4% 7% -1% 2%
0% 1% 0% 6% -1% -2%
Absolute thymus weight (percent change compared to control)
M
F
0% -1% 3% -10% -12% -25%
0% -7% 12% 19% 32% 19%
Relative thymus weight (percent change compared to control)
M
F
0% -1% 3% -10% -12% -25%
0% -7% 4% 4% 12% -6%
Levine et al. (1990); Levine et al.
(1981a); Levine et al. (1981b)e
Rats, F344,10/sex/group; 30/sex for
control
84.7 ± 4.7% purity, ~10% HMX, median
particle diameter 20 urn, ~90% of
particles <66 urn
0,10, 30,100, 300, or 600 mg/kg-d
Diet
13 wks
Data were not reported for rats in the 300 or 600 mg/kg-d dose groups
because all of the rats died before the 13-wk necropsy.
Doses
0 10 30 100 300 600
WBC count (percent change compared to control)
M
F
0% 4% 7% 15%
0% 23%* 24%* 62%*
Absolute spleen weight (percent change compared to control)
M
F
0% -11% -16% -34%
0% 2% 12% 0%
Relative spleen weight (percent change compared to control)
M
F
0% -9% -12% -21%
0% 2% 12% 3%
This document is a draft for review purposes only and does not constitute Agency policy.
C-56	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
von Oettingen et al. (1949)
Rats, sex/strain not specified, 20/group
90-97% pure, with 3-10% HMX; particle
size not specified
0,15, 25, or 50 mg/kg-d
Diet
13 wks
Doses
0 15 25 50
WBC count (percent change compared to control)
M
0% -30% 7% -6%
Hart (1974)
Dogs, Beagle, 3/sex/group
Pre-mix with ground dog chow
containing 20 mg RDX/g-chow, 60 g dog
food; purity and particle size not
specified
0, 0.1,1, or 10 mg/kg-d
Diet
13 wks
Doses
o
1
1
1
o
o
WBC count (percent change compared to control)
M
F
0% 5% 2% -19%
0% -2% 24% 6%
Absolute spleen weight (percent change compared to control)
M
F
0% - - 123%
0% - - -11%
Martin and Hart (1974)
Monkeys, Cynomolgus or Rhesus0,
3/sex/group
Purity of test material not specified
0, 0.1,1, or 10 mg/kg-d
Gavage
13 wks
Doses
o
1
1
1
o
o
WBC count (percent change compared to control)
M
F
0% -32% 0% -3%
0% -38% -1% -41%
Gastrointestinal effects
Lish et al. (1984)
Mice, B6C3Fi, 85/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 nm
0,1.5, 7.0, 35, or 175/100 mg/kg-d (high
dose reduced to 100 mg/kg-d in wk 11
due to excessive mortality)
Diet
2 yrs
No Gl tract effects were observed as clinical signs or on gross pathology
or histopathology examination.
Levine et al. (1983)
Rats, F344, 75/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 nm
0, 0.3,1.5, 8.0, or 40 mg/kg-d
Diet
2 yrs
No Gl tract effects were observed as clinical signs or on gross pathology
or histopathology examination.
This document is a draft for review purposes only and does not constitute Agency policy.
C-57	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results
Crouse et al. (2006)
Rats, F344,10/sex/group
99.99% pure
0, 4, 8,10,12, or 15 mg/kg-d
Gavage
13 wks
No Gl tract effects were observed on gross pathology or histopathology
examination. Increased salivation and blood stains around the mouth
were noted (affected doses and incidences were not reported); it is not
clear whether these effects occurred in animals also experiencing
convulsions.
von Oettingen et al. (1949)
Rats (sex/strain not specified); 20/group
90-97% pure, with 3-10% HMX; particle
size not specified
0,15, 25, or 50 mg/kg-d
Diet
13 wks
Congestion of the Gl tract was observed in 50 and 100 mg/kg-d rats
that also exhibited mortality (40%) and severe neurotoxicity.
Martin and Hart (1974)
Monkeys, Cynomolgus or Rhesus0,
3/sex/group
Purity of test material not specified
0, 0.1,1, or 10 mg/kg-d
Gavage
13 wks
Vomiting was observed more frequently in the 1 and 10 mg/kg-d
groups compared to the control or 0.1 mg/kg-d groups, although some
episodes occurred during the intubation procedure.
Hart (1974)
Dogs, Beagle, 3/sex/group
Pre-mix with ground dog chow
containing 20 mg RDX/g-chow, 60 g dog
food; purity and particle size not
specified
0, 0.1,1, or 10 mg/kg-d
Diet
13 wks
Some nausea and vomiting were reported (incidences and affected
dose groups were not reported).
Hematological effects
Lish et al. (1984)
Mice, B6C3Fi, 85/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 nm
0,1.5, 7.0, 35, or 175/100 mg/kg-d (high
dose reduced to 100 mg/kg-d in wk 11
due to excessive mortality)
Diet
2 yrs
Doses
0 1.5 7.0 35 175/100
RBC count; 105 wks (percent change compared to control)
M
F
0% -4% 3% -3% 14%
0% 4% -7% 5% 3%
Hemoglobin; 105 wks (percent change compared to control)
M
F
0% -6% 3% -5% 9%
0% 2% -7% 3% 1%
Hematocrit; 105 wks (percent change compared to control)
M
F
0% -4% 3% -4% 9%
0% 3% -6% 3% 1%
Platelets; 105 wks (percent change compared to control)
M
0% 33% 9% 21% 27%
This document is a draft for review purposes only and does not constitute Agency policy.
C-58	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results

F
0% -14% -7% 1% 5%
Hart (1976)
Rats, Sprague-Dawley, 100/sex/group
Purity and particle size not specified
0,1.0, 3.1, or 10 mg/kg-d
Diet
2 yrs
Doses
o
1
1
rn
o
O
RBC count; 104 wks (percent change compared to control)
M
F
0% 3% 7% -2%
0% -14% 7% 2%
Reticulocyte count; 104 wks (percent change compared to control)
M
F
0% 250% 500%* 850%*
0% 180%* -40% 20%
Hemoglobin; 104 wks (percent change compared to control)
M
F
0% 3% 4% 0%
0% -1% 1% -2%
Levine et al. (1983)
Rats, F344, 75/sex/group; interim
sacrifices (10/sex/group) at 6 and 12 mo
89.2-98.7% pure, with 3-10% HMX as
contaminant; 83-89% of particles
<66 urn
0, 0.3,1.5, 8.0, or 40 mg/kg-d
Diet
2 yrs
Doses
o
o
00
LO
m
o
o
Hemoglobin levels; 105 wks (percent change compared to control)
M
F
0% 6% 6% 3% -13%
0% -5% 1% -9% -14%
RBC count; 105 wks (percent change compared to control)
M
F
0% 5% 2% -1% -9%
0% -2% 2% -9% -13%
Platelet count; 105 wks (percent change compared to control)
M
F
0% 6% -4% -10% -7%
0% 14% -4% 5% 22%
Hematocrit; 105 wks (percent change compared to control)
M
F
0% 5% 5% 2% -7%
0% -5% 0% -8% -12%
Cholakis et al. (1980)
Mice, B6C3Fi, 10-12/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 nm
particle size
0, 80, 60, or 40 mg/kg-d for 2 wks
followed by 0, 80,160, or 320 mg/kg-d
(TWA doses of 0, 79.6, 147.8, or
256.7 mg/kg-d for males and 0, 82.4,
136.3, or 276.4 mg/kg-d for females)d
Diet
13 wks
Doses
0 80 160 320
RBC count (percent change compared to control)
M
F
0% -5% -12%* -2%
0% -10% -1% 1%
Reticulocytes (percent change compared to control)
M
F
0% -36% -13% 15%
0% 21% 25% -19%
Hematocrit (percent change compared to control)
M
0s-
O
0s-
ID
1
0s-
1
1
0s-
O
This document is a draft for review purposes only and does not constitute Agency policy.
C-59	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results

F
0% -8% 2%

1%

Hemoglobin (percent change compared to control)

M
0% -2% -7%*

-3%

F
0% -5% 4%

1%

Platelets (percent change compared to control)

M
0% 33% 28%

22%

F
0% 3% 9%

39%
Cholakis et al. (1980)
Doses
0 10 14 20
28
40
Rats, F344,10/sex/group
88.6% pure, with 9% HMX and 2.2%
water as contaminants; ~200 urn
RBC count (percent change compared to control)
M
0% -
3%
-1%
particle size
0,10,14, 20, 28, or 40 mg/kg-d
F
0% -
-1%
-7%
Diet
Hemoglobin (percent change compared to control)
13 wks
M
0% -
2%
-1%

F
0% -
-1%
-1%

Platelet (percent change compared to control)

M
0% -
11%
16%*

F
0% -
-23%
-13%

Reticulocytes (percent change compared to control)

M
0% -
26%
76%*

F
0% -
-2%
17%

Hematocrit (percent change compared to control)

M
0% -
3%
0%

F
0% -
0%
-2%
Crouse et al. (2006)
Doses
0 4 8 10
12
15
Rats, F344,10/sex/group
99.99% pure
0, 4, 8,10,12, or 15 mg/kg-d
RBC count (percent change compared to control)
M
0% 1% -7% -2%
-4%
-5%
Gavage
13 wks
F
0s-
1
1
0s-
m
0s-
m
0s-
O
2%
-2%

Hemoglobin (percent change compared to control)

M
0s-
O
0s-
LO
1
0s-
1
1
0s-
O
-1%
-6%

F
0% 2% 4% -1
4%
-4%

Platelet count (percent change compared to control)

M
0% 21% 11% 13%
-8%
34%

F
0% 6% 40% 47%
34%
-36%
This document is a draft for review purposes only and does not constitute Agency policy.
C-60	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results

Hematocrit (percent change compared to control)
M
F
0% 2% -5% 0% -1% -4%
0% 3% 4% 0% 4% -2%
Levine et al. (1990); Levine et al.
(1981a); Levine et al. (1981b)e
Rats, F344,10/sex/group; 30/sexfor
control
84.7 ± 4.7% purity, ~10% HMX, median
particle diameter 20 urn, ~90% of
particles <66 urn
0,10, 30,100, 300, or 600 mg/kg-d
Diet
13 wks
Data were not reported for rats in the 300 or 600 mg/kg-d dose groups
because all of the rats died before the 13-wk necropsy.
Doses
0 10 30 100 300 600
Hematocrit (percent change compared to control)
M
F
0% -2% -1% -5%
0% 0% -4% -7%
Hemoglobin (percent change compared to control)
M
F
0% -3% -1% -6%
0% 0% -4% -8%*
RBC count (percent change compared to control)
M
F
0% -2% -2% -5%
0% -1% -4% -5%
Reticulocytes (percent change compared to control)
M
F
0% -4% 10% 28%
0% 9% 73% 71%
von Oettingen et al. (1949)
Rats, sex/strain not specified, 20/group
90-97% pure, with 3-10% HMX; particle
size not specified
0,15, 25, or 50 mg/kg-d
Diet
13 wks
Doses
0 15 25 50
RBC count (percent change compared to control)
M + F
0% -23% -12% -14%
Hemoglobin (percent change compared to control)
M + F
0% -25% -7% -11%
Hart (1974)
Dogs, Beagle, 3/sex/group
Pre-mix with ground dog chow
containing 20 mg RDX/g-chow, 60 g dog
food; purity and particle size not
specified
0, 0.1,1, or 10 mg/kg-d
Diet
13 wks
Doses
0 0.1 1 10
RBC count (percent change compared to control)
M
F
0% -3% 3% 2%
0% 13% 7% 11%
Reticulocyte count (percent change compared to control)
M
F
0% -66% 0% -50%
0% -17% -50% 0%
Hematocrit (percent change compared to control)
M
F
0% -4% 2% 0%
0% 6% 1% 7%
This document is a draft for review purposes only and does not constitute Agency policy.
C-61	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Reference and study design
Results

Hemoglobin (percent change compared to control)
M
F
0% 5% -2% 0%
0% 8% -2% 8%
Martin and Hart (1974)
Monkeys, Cynomolgus or Rhesus0,
3/sex/group
Purity of test material not specified
0, 0.1,1, or 10 mg/kg-d
Gavage
13 wks
Histopathological examination revealed increased numbers of
degenerate or necrotic megakaryocytes in all bone marrow sections.
Doses
o
1
1
1
o
o
RBC count (percent change compared to control)
M
F
0% -3% 2% -3%
0% 0% -1% 2%
Reticulocyte count (percent change compared to control)
M
F
0% -33% -50% -50%
0% -18% -36% 45%
Hematocrit (percent change compared to control)
M
F
0% -7% -4% -1%
0% 10% 7% 3%
Hemoglobin (percent change compared to control)
M
F
0% -10% -8% -6%
0% 6% 6% 3%
*Statistically significantly different compared to the control, as determined by study authors (p < 0.05).
aNo musculoskeletal evidence is presented in this table as no animal study reported effects on the musculoskeletal
system and all human effects were in case reports (see summary in Appendix C, Section C.2).
incidence counts exclude individuals from which blood was obtained via the orbital sinus.
cThe species of monkey used in this study was inconsistently reported in the study as either Cynomolgus (in the
Methods section) or Rhesus (in the Summary).
dDoses were calculated by the study authors.
eLevine et al. (1981a) is a laboratory report of a 13-week study of RDX in F344 rats; two subsequently published
papers (Levine et al., 1990; Levine et al., 1981b) present subsets of the data provided in the full laboratory report.
'Standard deviations accompanying the mean response in a given dose group were high, suggesting uncertainty in
the accuracy of the reported percent change compared to control.
Note: A dash ("-") indicates that the study authors did not measure or report a value for that dose group.
This document is a draft for review purposes only and does not constitute Agency policy.
C-62	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
C.3.3. Genotoxicity
RDX
RDX has tested negative in a variety of in vitro tests for genotoxicity, including mutation
assays in multiple strains of Salmonella typhimurium (with or without metabolic activation),
recombination in Saccharomyces cerevisiae strain D3, and forward mutations in both V79 Chinese
hamster lung cells and mouse lymphoma L5178Y cells. However, in genotoxicity assays designed to
be more sensitive, RDX did show some positive results. For example, when the concentration of S9
was doubled, the mutagenicity of RDX was about twice that of background. RDX also showed
positive mutagenic results with metabolic activation in a chemiluminescent assay (Mutatox assay).
In some cases, the interpretation of testing data for RDX was complicated by the tendency of the
compound to precipitate outof DMSO solution (the usual vehicle) at concentrations >250 |J.g/mL
(Reddv etal.. 2005). As with other studies of RDX, the purity of the test compound was unknown in
several (particularly older) studies. A summary of the results of in vitro genotoxicity studies of RDX
is presented in Table C-13.
RDX has produced negative results in all reverse mutation assays in S. typhimurium that use
standard levels of the metabolic activation system (S9). No evidence of reverse mutation was
observed inS. typhimurium (strains TA98, TA100, TA1535, TA1537, andTA1538), either with or
without the addition of S9 metabolic activating mixture fNeuwoehner etal.. 2007: George etal..
2001: Lachance etal.. 1999: Tan etal.. 1992: Cholakis etal.. 1980: Whongetal.. 1980: Cotruvo etal..
1977: Simmon et al.. 1977). One exception is a finding of weak mutagenic activity of RDX to
S. typhimurium strainTA97a (mutagenicity index = 1.5-2.0) (Pan etal.. 2007al. However, this assay
used a high percentage of S9 fraction (9% instead of 4%), indicating that extensive metabolic
activation is needed to elicit a mutagenic response.
RDX did not cause gene recombination in S. cerevisiae strain D3 at concentrations up to
23 [.ig/mL, with or without metabolic activation (Cotruvo etal.. 1977: Simmon etal.. 1977).
Simmon etal. (1977) noted that the negative findings should be considered in the context of the low
concentrations tested. RDX was negative in assays with S. choleraesius and E. coli with and without
metabolic activation fNeuwoehner et al.. 2007). Similarly, RDX did not induce forward mutations
(HGPRT locus) in V79 Chinese hamster lung cells, with or without metabolic activation, although
minimal cytotoxicity was observed at 180 |iM f Lachance etal.. 19991. However, RDX produced
revertants in two of three trials in the Mutatox assay with the bacterium Vibrio fisheri when tested
at doses up to 2.5 ng/tube, with and without S9 (Arfsten et al.. 19941. In the presence of S9, a dose-
response was observed; in the absence of S9, no dose-response relationship was detected (Arfsten
etal.. 1994). RDX did not induce forward mutations in mouse lymphoma L5178Y cells with or
without metabolic activation (Reddv etal.. 2005). During an accompanying range-finding study,
precipitates occurred at doses >250 [ig/mL, suggesting that concentrations of RDX in DMSO
reported beyond 250 |J.g/mL may not be accurate.
This document is a draft for review purposes only and does not constitute Agency policy.
C-63	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-13. Summary of in vitro studies of the genotoxicity of RDX
Endpoint
Test system
Dose/
concentration3
Results'5
Comments
Reference
Without
activation
With
activation
Genotoxicity studies in prokaryotic organisms
Reverse
mutation
Salmonella typhimurium
TA1535, TA1537, TA1538, TA98,
TA100
1,000 ng/plate
"
"
Metabolic activation with S9
Cholakis et al.
(1980)
Reverse
mutation
5. typhimurium TA1535,
TA1537, TA1538 TA100, TA98
14 ng/plate


Effect of disinfection treatments on
mutagenicity tested: RDX was not mutagenic
in any strain before or after disinfection
treatment with chlorine or ozone
Simmon et al.
(1977)
Reverse
mutation
S. typhimurium TA98, TA100
250 ng/plate
-
-
Study authors noted that results were
consistent with literature
George et al.
(2001)
Reverse
mutation
S. typhimurium TA98, TA100
1 mg/plate
-
-
Metabolic activation with S9
Tan et al.
(1992)
Reverse
mutation
S. typhimurium TA98, TA100
1,090 ng/plate
-
-
High S9 activation (9%) used
Pan et al.
(2007a)
Reverse
mutation
5. typhimurium TA97a
32.7 ng/plate

+
High S9 activation (9%) used; study authors
concluded that RDX "required intensive
metabolic activation" to exhibit mutagenicity
in this strain
Pan et al.
(2007a)
Reverse
mutation
5. typhimurium TA1535,
TA1537, TA1538 TA100, TA98
Up to
2.5 mg/plate


Results were reported qualitatively only;
quantitative results were not presented. Not
clear if assay was also performed without S9
Whong et al.
(1980)
Reverse
mutation
Vibrio fischeri
0.004 ng/tube
+
+
Mutatox assay with metabolic activation (S9)
Arfsten et al.
(1994)
This document is a draft for review purposes only and does not constitute Agency policy.
C-64	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Endpoint
Test system
Dose/
concentration3
Results'5
Comments
Reference
Without
activation
With
activation
Reverse
mutation (umu
test)
Salmonella choleraesius subsp.
chol. (prior S. typhimurium)
TA1535/pSK1002
20.6 Hg/mL


No observed effect concentration; tested at
highest concentration where the induction
rate was below 1.5 for the first time and the
growth factor was below 0.5
Neuwoehner
et al. (2007)
Reverse
mutation
(NM2009 test)
5. choleraesius subsp. chol.
NM2009,
TA1535/pS K1002/pN M12
20.6 Hg/mL


No observed effect concentration; tested at
highest concentration where the induction
rate was below 1.5 for the first time and the
growth factor was below 0.5
Neuwoehner
et al. (2007)
Induction of
the sfiA gene
(SOS
chromotest)
Escherichia coli PQ37
20.6 Hg/mL


No observed effect concentration; tested at
highest concentration where the induction
rate was below 1.5 for the first time and the
growth factor was below 0.5
Neuwoehner
et al. (2007)
Reverse
mutation
S. typhimurium, TA98, TA100
24.8 Hg/mL
-
-
No observed effect concentration; metabolic
activation with S9
Neuwoehner
et al. (2007)
Reverse
mutation
S. typhimurium TA98, TA100
2.6 ng/mL


No observed effect concentration; metabolic
activation with S9; minimal cytotoxicity was
observed at 180 nM
Lachance et al.
(1999)
Reverse
mutation
5. typhimurium TA1535,
TA1536, TA1537, TA1538
TA100, TA98
30.8 Hg/mL


Metabolic activation with S9
Cotruvo et al.
(1977)
Genotoxicity studies in nonmammalian eukaryotic organisms
Recombination
induction
Saccharomyces cerevisiae D3
23 ng/mL


Study authors concluded that this
microorganism did not appear to be useful
for detecting mutagenicity in several
compounds tested
Simmon et al.
(1977)
Recombination
induction
5. cerevisiae D3
30.8 Hg/mL
-
-
Metabolic activation with S9
Cotruvo et al.
(1977)
This document is a draft for review purposes only and does not constitute Agency policy.
C-65	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Endpoint
Test system
Dose/
concentration3
Results'5
Comments
Reference
Without
activation
With
activation
Genotoxicity studies in mammalian cells
Forward
mutation
Chinese hamster lung
fibroblasts V79 cells
40 Hg/mL
-
-
Minimal cytotoxicity observed at 40 ng/mL
(limit of solubility)
Lachance et al.
(1999)
Mutation
L5178Y mouse lymphoma cells
500 Hg/mL


No or low cytotoxicity seen at these
concentrations; however, precipitate was
observed at >250 ng/mL
Reddv et al.
(2005)
Unscheduled
DNA synthesis;
DNA repair
WI-38 cells, human diploid
fibroblasts
4,000 |Jg/mL


Precipitates were observed at concentrations
of RDX >40 ng/mL
Dillev et al.
(1979)
1
2	aLowest effective dose for positive results; highest dose tested for negative results.
3	b+ = positive; ± = equivocal or weakly positive; - = negative.
This document is a draft for review purposes only and does not constitute Agency policy.
C-66	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
RDX did not induce unscheduled DNA synthesis, with or without metabolic activation, using
human diploid fibroblasts (WI-38 cells) when tested in DMSO at concentrations up to 4,000 [ig/mg;
however, precipitation of RDX at high concentrations in cell culture media makes interpretation of
these results difficult (Dillev etal.. 19791. Only two in vivo studies are available for the genotoxicity
of RDX; these are summarized in Table C-14. RDX did not decrease the ratio of polychromatic
erythrocytes (PCEs) to normochromatic erythrocytes (NCEs), nor did it induce micronucleated
PCEs in an in vivo mouse bone marrow micronucleus assay in young adult male CD-I mice (Reddv
etal.. 20051. RDX was considered negative for the induction of dominant lethal mutations in male
CD rats fed RDX at nominal doses from 0 to 50 mg/kg-day for 15 weeks prior to mating with
untreated virgin females fCholakis etal.. 19801. Females sacrificed at midgestation showed no
statistically significant effects on number of corpora lutea, implants, or live or dead embryos
fCholakis etal.. 19801.
Metabolites of RDX
Several metabolites of RDX, N-nitroso derivatives of the parent compound (mononitroso,
dinitroso, and trinitroso compounds, abbreviated MNX, DNX, andTNX, respectively) fMusicketal..
2010: Major etal.. 20071 have been tested directly for genotoxicity (Pan etal.. 2007a: George etal..
2001: Snodgrass. 19841. Miniature pigs were used to detect these trace metabolites because the
swine model of the GI tract more closely resembles that of humans fMaior etal.. 20071: an
identification and quantification of RDX metabolites in humans has not been conducted. A
summary of the results of in vitro and in vivo genotoxicity studies of metabolites of RDX is
presented in Table C-15.
Pan etal. f2007al studied the mutagenicity of two metabolites, MNX and TNX. These
metabolites were not mutagenic in S. typhimurium strain TA97a at normal levels of S9, but were
clearly mutagenic at enhanced concentrations of S9 (4% versus 9% S9). The observation that these
metabolites were positive in S. typhimurium strain TA97a is likely due to this strain's higher
sensitivity for frameshift mutations that occur at a cluster of cytosine residues in the mutated gene
for histidine synthesis in this strain (Pan etal.. 2007a). These metabolites were also weakly
mutagenic in S. typhimurium strain TA102, again with high levels of S9. Strain TA102 was
developed with an A:T base pair at the site of mutation and its sensitivity was increased by the
addition of some 30 copies of a plasmid containing the mutant gene that is available for back
mutation. This strain is sensitive to many oxidative mutagenic compounds fLevin etal.. 19821.
Other metabolites with potential human relevance identified in the urine of miniature pigs have not
been assessed for their genotoxicity (Major etal.. 2007). In assays with S. typhimurium strains
TA98 and TA100, TNX was positive in strain TA100 with and without S9, but not in strain TA98;
MNX and DNX were not mutagenic in either strain (George etal.. 2001).
This document is a draft for review purposes only and does not constitute Agency policy.
C-67	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-14. Summary of in vivo studies of the genotoxicity of RDX
Endpoint
Test system
Dose/
concentration
Results
Comments
Reference
In vivo genotoxicity studies in mammalian systems
Micronucleus
formation
CD-I mouse bone marrow
Single dose of 62.5,
125, or 250 mg/kg
No significant decrease in
PCE:NCE ratios; no
induction of
micronucleated PCE at any
dose
250 mg/kg was maximum
tolerated dose determined in
dose range-finding study
Reddy et al. (2005)
Dominant
lethal
mutations
Male CD rats dosed and mated with
untreated female rats
0, 5,16, or
50 mg/kg-d for
15 wk
No statistically or
biologically significant
effects on fertility;
determined to be negative
for the induction of lethal
mutations
Males in the high-dose group
experienced lower food
consumption and weight gain
compared with all other
groups
Cholakis et al.
(1980)
This document is a draft for review purposes only and does not constitute Agency policy.
C-68	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table C-15. Summary of in vitro and in vivo studies of the genotoxicity of RDX metabolites
Endpoint
Test system
Dose/
concentration3
Results'5
Comments
Reference
Without
activation
With
activation
Genotoxicity studies in prokaryotic organisms
Reverse
mutation
Salmonella typhimurium TA97a,
TA102
22 ng/plate
-
+
Mono and trinitroso metabolites (MNX
and TNX); high S9 activation (9%) used
Pan et al. (2007a)
Reverse
mutation
S. typhimurium TA98, TA100
500 ng/plate
+
+
Positive in TA100 (but not in TA98) only
for TNX; MNX and DNXwere negative
George et al.
(2001)
Reverse
mutation
5. typhimurium TA1535,
TA1537, TA1538, TA98, TA100
NR
-
-
Mononitroso metabolite, MNX;
metabolic activation with S9
Snodgrass (1984)
Genotoxicity studies in mammalian cells—in vitro
Forward
mutation
Mouse lymphoma thymidine
kinase
NR
+
+
Mononitroso metabolite, MNX;
metabolic activation with S9
Snodgrass (1984)
Chromosomal
aberrations
Chinese hamster ovary cells
NR
-
+
Mononitroso metabolite, MNX;
metabolic activation with S9
Snodgrass (1984)
Unscheduled
DNA synthesis;
DNA repair
Primary rat hepatocytes
NR
+
ND
Mononitroso metabolite, MNX;
additional metabolic activation not
required with S9
Snodgrass (1984)
In vivo genotoxicity studies in mammalian systems
Dominant
lethal
mutations
Male mice dosed and mated
with untreated female mice
NR

ND
Mononitroso metabolite, MNX;
additional metabolic activation not
required with S9
Snodgrass (1984)
2
3	aLowest effective dose for positive results; highest dose tested for negative results; NR = not reported.
4	b+ = positive; ± = equivocal or weakly positive; - = negative; ND = not determined.
This document is a draft for review purposes only and does not constitute Agency policy.
C-69	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
The genotoxicity of MNX was positive in three out of five assays conducted for the U.S. Army
(Snodgrass. 19841. MNX was positive with or without metabolic activation in the mouse lymphoma
forward mutation assay at the thymidine kinase locus, for chromosomal aberrations in Chinese
hamster ovary cells, and in the primary rat hepatocyte unscheduled DNA synthesis assay. MNX was
not considered positive inS. typhimurium (strains TA98, TA100, TA1535, TA1537, andTA1538),
either with or without the addition of S9 metabolic activating mixture or in an in vivo dominant
lethal mutation assay in mice. However, this study is of limited use due to a significant lack of
details including information on dosing, raw data, and statistical reporting.
In summary, RDX is not mutagenic or genotoxic in vitro or in vivo in typical assays used to
detect genotoxicity. In two in vitro studies using more sensitive assays and conditions for detecting
mutagenicity, RDX was found to be positive. Several studies showed that the N-nitroso metabolites
are genotoxic, but the formation and quantification of these metabolites in humans is not known.
This document is a draft for review purposes only and does not constitute Agency policy.
C-70	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
APPENDIX D. DOSE-RESPONSE MODELING FOR
THE DERIVATION OF REFERENCE VALUES FOR
EFFECTS OTHER THAN CANCER AND THE
DERIVATION OF CANCER RISK ESTIMATES
This appendix provides technical detail on dose-response evaluation and determination of
points of departure (POD) for relevant toxicological endpoints. The endpoints were modeled using
the U.S. Environmental Protection Agency (EPA) Benchmark Dose Software (BMDS, Versions 2.4).
Sections D.l (noncancer) and D.2 (cancer) describe the common practices used in evaluating the
model fit and selecting the appropriate model for determining the POD, as outlined in the
Benchmark Dose Technical Guidance Document (U.S. EPA. 2012b). In some cases, it may be
appropriate to use alternative methods, based on statistical judgement; exceptions are noted as
necessary in the summaiy of the modeling results.
D.l. BENCHMARK DOSE MODELING SUMMARY FOR NONCANCER
ENDPOINTS
The noncancer endpoints that were selected for dose-response modeling are presented in
Table D-l. For each endpoint, the doses and response data used for the modeling are presented.
Table D-l. Noncancer endpoints selected for dose-response modeling for RDX
Endpoint and
reference
Species/sex
Dose
(mg/kg-d)
Incidence/total (%)
Convulsions
Female F344 rat
0
0/10 (0%)
Crouse et al. (2006)a

4
0/10 (0%)


8
2/10 (20%)


10
3/10 (30%)


12
5/10 (50%)


15
5/10 (50%)

Male F344 rat
0
0/10 (0%)


4
0/10 (0%)


8
1/10 (10%)


10
3/10 (30%)


12
8/10(80%)


15
7/10 (70%)
This document is a draft for review purposes only and does not constitute Agency policy.
D-l	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Endpoint and

Dose

reference
Species/sex
(mg/kg-d)
Incidence/total (%)

Male and female
0
0/20 (0%)

F344 rat, combined
4
0/20 (0%)


8
3/20(15%)


10
6/20 (30%)


12
13/20 (65%)


15
12/20 (60%)
Convulsions
Female F344 rat
0
0/24 (0%)
Cholakis et al. (1980)
(gestational
0.2
0/24 (0%)

exposure)
2
1/24 (4%)


20
18/24 (75%)
Testicular
Male B6C3Fi mouse
0
0/63 (0%)
degeneration

1.5
2/60 (3%)
Lish etal. (1984)

7
2/62 (3%)


35
6/59 (10%)


107
3/27 (11%)
Prostate suppurative
Male F344 rat
0
2/54 (4%)
inflammation

0.3
4/55 (7%)
Levine et al. (1983)

1.5
9/52 (17%)


8
12/55 (22%)


40
19/31 (61%)
1
2	aFor convulsions in Crouse et al. (2006), the incidence rates across doses were determined to be not statistically
3	significantly different between the males and females using an exact Cochran-Mantel-Haenszel test (p > 0.10).
4	The data were combined across sex for this endpoint prior to modeling.
5
6	In addition to the endpoints presented in Table D-l, the combined incidence of seizure and
7	mortality was modeled for Crouse etal. (2006) to determine the effect of possible underestimation
8	of seizures, as discussed in Section 2.1.6. Table D-2 presents the data on this combined incidence.
9	Table D-2. Convulsion or mortality endpoints from Crouse et al. f2006)
10	selected for dose-response modeling for RDX
Endpoint and

Dose

reference
Species/sex
(mg/kg-d)
Incidence/total (%)
Convulsion or
Female F344 rat
0
0/10 (0%)
mortality

4
0/10 (0%)
Johnson (2015)

8
3/10 (30%)


10
5/10 (50%)


12
9/10 (90%)


15
8/10(80%)
This document is a draft for review purposes only and does not constitute Agency policy.
D-2	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Endpoint and
reference
Species/sex
Dose
(mg/kg-d)
Incidence/total (%)

Male F344 rat
0
0/10 (0%)


4
0/10 (0%)


8
2/10 (20%)


10
4/10 (40%)


12
8/10(80%)


15
7/10 (70%)

Male and female
0
0/20 (0%)

F344 rat, combined
4
0/20 (0%)


8
5/20 (25%)


10
9/20 (45%)


12
17/20 (85%)


15
15/20 (75%)
incidence was defined for each animal as the presence of convulsion or mortality. The incidence rates across
doses for this endpoint were determined to be not statistically significantly different between the males and
females using an exact Cochran-Mantel-Haenszel test (p > 0.10). The data were combined across sex for this
endpoint prior to modeling.
D.l.l. Evaluation of Model Fit and Model Selection
For each dichotomous endpoint, BMDS dichotomous models5 were fitted to the data using
the maximum likelihood method. Each model was tested for goodness-of-fit using a chi-square
goodness-of-fit test (x2 p-value < 0.10 indicates lack of fit). Other factors were also used to assess
model fit, such as scaled residuals, visual fit, and adequacy of fit in the low-dose region and in the
vicinity of the benchmark response (BMR).
From among the models exhibiting adequate fit, the best-fit model was selected for
estimation of the BMD. This model selection was conducted in two stage, first from among only the
multistage models to determine a representative multistage model, and second from among the
representative multistage model and the non-multistage models. In each stage, the BMDL estimates
(95% lower confidence limit on the BMD, as estimated by the profile likelihood method) and AIC
values of the models considered in that stage were used to make the selection, as follows. If the
BMDL estimates were "sufficiently close," that is, differed by threefold or less, the model selected
was the one that yielded the lowest AIC value. If the BMDL estimates were not sufficiently close, the
model with the lowest BMDL was selected. The model selected in the second stage was considered
the best-fit model.
The BMDL estimate (95% lower confidence limit on the benchmark dose [BMD], as
estimated by the profile likelihood method) and Akaike's information criterion (AIC) value were
used to select a best-fit model from among the models exhibiting adequate fit If the BMDL
5Unless otherwise specified, all available BMDS dichotomous models besides the alternative and nested
dichotomous models were fitted. The following parameter restrictions were applied: for the Log-Logistic
model, restrict slope >1; for the Gamma and Weibull models, restrict power >1.
This document is a draft for review purposes only and does not constitute Agency policy.
D-3	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	estimates were "sufficiently close" (i.e., differed by at most threefold), the model selected was the
2	one that yielded the lowest AIC value. If the BMDL estimates were not sufficiently close, the lowest
3	BMDL was selected as the POD.
4	D.1.2. Modeling Results
5	The tables that follow summarize the modeling results for the noncancer endpoints
6	modeled.
7	Nervous System Effects
8	Tables D-3 to D-5 (and Figures D-l to D-3) presentthe BMD modeling results for incidence
9	of convulsions for female, male, and male and female F344 rats combined based on data from
10	Grouse etal. (2006). using BMRs of 10, 5, and 1% extra risk (ER). Table D-6 (and Figure D-4)
11	presentthe BMD modeling results for incidence of convulsions for female F344 rats based on data
12	from Cholakis etal. (1980). using BMRs of 10. 5. and 1% ER. Table D-7 (and Figure D-5) presents
13	the BMD modeling results for combined incidence of convulsions and mortality for male and female
14	rats combined based on data from Grouse etal. (2006).
15	Table D-3. Model predictions for convulsions in female F344 rats exposed to
16	RDX by gavage for 90 days (Grouse et al.. 2006): BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.923
55.085
3.10
0.355
The Quantal-Linear model had a
BMD more than 10 times lower
than the lowest dose, and the
residual at the lowest dose was
modereatly high (-1.3). Thus,
this model was excluded from
consideration. Of the higher
degree multistage models, the
multistage 2° model was selected
as the representative multistage
model based on lowest AIC.
From among the multistage 2°
and non-multistage models, the
multistage 5° model was selected
based on lowest BMDL (BMDLs
differed by more than threefold).
Logistic
0.733
56.607
1.60
0.681
LogLogistic
0.929
55.076
2.87
0.468
Probit
0.793
56.086
1.86
0.649
LogProbit
0.952
54.798
3.63
0.919
Weibull
0.892
55.420
2.30
0.259
Multistage 2°
0.954
53.595
1.69
0.236
Quantal-Linear
0.733
56.131
0.263
0.176
Multistage 3°
0.885
55.525
1.99
0.238
Multistage 4°
0.885
55.525
1.99
0.236
Multistage 5°
0.885
55.525
1.99
0.235
Dichotomous-Hill
0.964
56.265
4.77
0.778
17
18	aSelected model in bold; scaled residuals for selected model for doses 0, 4, 8,10,12, and 15 mg/kg-day were 0.00,
19	-0.67, 0.14, 0.11, 0.64, and -0.51, respectively. The BMDio and BMDLio values for the selected model were
20	5.46 and 2.47 mg/kg-day, respectively; the BMDos and BMDLos values for the selected model were 3.81 and
21	1.21 mg/kg-day, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
D-4	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Model, with BMR of 1% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BIV
Multistage
0.8
0.7
0.6
0.5
0.4
0.3
0.2
e:mdl
BMD
0
2
4
6
8
10
12
14
dose
14:06 02/27 2014
Figure D-l. Plot of incidence rate by dose, with fitted curve for selected model,
for convulsions in female F344 rats exposed to RDX by gavage for 90 days
(Crouse et al.. 2006).
Multistage Model (Version: 3.3; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
Benchmark Dose Computation
BMR = 1% Extra risk
BMD = 1.68508
BMDL atthe 95% confidence level = 0.236479
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Beta(l)
0
0.0172961
Beta(2)
0.00353947
0.002476
Analysis of Deviance Table
Model
Log(likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-24.9756
6



Fitted model
-25.7976
1
1.64388
5
0.8959
Reduced model
-33.7401
1
17.529
5
0.003598
This document is a draft for review purposes only and does not constitute Agency policy.
D-5	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	AIC: = 53.5951
2
3	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
10
0
4
0.0551
0.551
0
10
-0.763
8
0.2027
2.027
2
10
-0.021
10
0.2981
2.981
3
10
0.013
12
0.3993
3.993
5
10
0.65
15
0.549
5.49
5
10
-0.312
4
5	ChiA2 = 1.1 d.f. = 5 p-value = 0.9538
6
7	ChiA2 = 1.16 d.f
8	Table D-4. Model predictions for convulsions in male F344 rats exposed to
9	RDX by gavage for 90 days fGrouse et al.. 20061: BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.482
48.534
4.96
2.32
Of the multistage models that
provided an adequate fit, the
multistage 2° model was selected
based on lowest AIC. From
among the multistage 2° and
non-multistage models, the
multistage 2° model was selected
based on lowest BMDL (BMDLs
differed by more than threefold).
Logistic
0.335
49.692
2.86
0.975
LogLogistic
0.522
48.248
4.79
2.38
Probit
0.363
49.460
3.60
1.01
LogProbit
0.530
48.224
5.41
3.00
Weibull
0.376
49.496
3.52
1.43
Multistage 2°
0.307
50.335
1.40
0.363
Quantal-Linear
0.0553
56.530
0.189
0.131
Multistage 5°b
0.361
49.607
3.42
0.392
Multistage 4°c
0.361
49.607
3.42
0.392
Multistage 3°
0.515
47.803
2.82
0.457
Dichotomous-Hill
0.701
48.408
6.64
3.47
10
11	aSelected model in bold; scaled residuals for selected model for doses 0, 4, 8,10,12, and 15 mg/kg-day were 0.00,
12	-0.92, -1.26, -0.65,1.76, and 0.11, respectively. The BMDio and BMDLio values for the selected model were
13	4.54 and 2.95 mg/kg-day, respectively; the BMDos and BMDLos values for the selected model were 3.17 and
14	1.63 mg/kg-day, respectively.
15	bThe Multistage 5° model may appear equivalent to the Multistage 4° model; however, differences exist in digits
16	not displayed in the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-6	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Model, with BMR of 1 % Extra Risk for the BMD and 0.95 Lower Confidence Limit for the Bh
Multistage
0.8
0.6
0.4
0.2
I3MD
BMD
0
2
4
6
8
10
12
14
dose
14:09 02/27 2014
BMR = 1% ER; dose shown in mg/kg-day.
Figure D-2. Plot of incidence rate by dose, with fitted curve for selected model,
for convulsions in male F344 rats exposed to RDX by gavage for 90 days
(Crouse et al.. 2006).
Multistage Model (Version: 3.3; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
Benchmark Dose Computation
BMR = 1% ER
BMD = 1.40125
BMDL atthe 95% confidence level = 0.363499
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Beta(l)
0
0
Beta(2)
0.00511858
0.00691555
This document is a draft for review purposes only and does not constitute Agency policy.
D-7	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1 Analysis of Deviance Table	
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-20.4721
6



Fitted model
-24.1672
1
7.39017
5
0.1932
Reduced model
-37.4599
1
33.9755
5
<0.0001
2
3	AIC:= 50.3345
4
5	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
10
0
4
0.0786
0.786
0
10
-0.924
8
0.2793
2.793
1
10
-1.264
10
0.4006
4.006
3
10
-0.649
12
0.5215
5.215
8
10
1.763
15
0.6839
6.839
7
10
0.11
6
7	ChiA2 = 5.99 d.f. = 5 p-value = 0.3069
8
This document is a draft for review purposes only and does not constitute Agency policy.
D-8	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table D-5. Model predictions for convulsions in male and female F344 rats
2	exposed to RDX by gavage for 90 days (Grouse et al.. 2006): BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.484
101.79
4.02
2.03
The Quantal-Linear model did not
fit the data adequately
(goodness-of-fit p-value <0.10),
so it was excluded from
consideration. Of the higher
degree multistage models, the
Multistage 3° model was selected
based on lowest AIC. From
among the multistage 3° and
non-multistage models, the
multistage 3° model was selected
based on lowest BMDL (BMDLs
differed by more than threefold).
Logistic
0.231
104.55
2.04
0.987
LogLogistic
0.512
101.66
3.79
2.00
Probit
0.291
103.61
2.57
1.03
LogProbit
0.557
101.25
4.50
2.69
Weibull
0.369
102.91
2.94
1.35
Multistage 2°
0.364
103.03
1.53
0.544
Quantal-Linear
0.0369
111.56
0.222
0.169
Multistage 5°b
Multistage 4°
0.502
100.91
3.02
0.549
Multistage 3°
0.502
100.91
3.02
0.569
Dichotomous-Hill
0.696
101.64
5.62
2.90
3
4	aSelected model in bold; scaled residuals for selected model for doses 0, 4, 8,10,12, and 15 mg/kg-day were 0.00,
5	-0.69, -0.25, -0.06,1.62, and -1.08, respectively. The BMDio and BMDLio values for the selected model were
6	6.60 and 4.59 mg/kg-day, respectively; the BMDos and BMDLos values for the selected model were 5.19 and
7	2.66 mg/kg-day, respectively.
8	bFor the Multistage 5° model, the beta coefficient estimates were 0 (boundary of parameters space). The models
9	in this row reduced to the Multistage 4° model.
This document is a draft for review purposes only and does not constitute Agency policy.
D-9	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Model, with BMR of 1 % Extra Risk for the BMD and 0.95 Lower Confidence Limit for the Bh
0.6
0.4
BMDL
14:03 02/27 2014
Multistage
BMD
6	8
dose
10
12
14
Figure D-3. Plot of incidence rate by dose, with the fitted curve of the
multistage 3° model, for convulsions in male and female F344 rats exposed to
RDX by gavage for 90 days (Crouse etal.. 20061. BMR = 1% ER; dose shown in
mg/kg-day.
Multistage Model (Version: 3.3; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
Benchmark Dose Computation
BMR =1% Extra risk
BMD = 3.01676
BMDL atthe 95% confidence level = 0.569284
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Beta(l)
0
0.00163806
Beta(2)
0
0.00485933
Beta(3)
0.000366065
0
This document is a draft for review purposes only and does not constitute Agency policy.
D-10	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1 Analysis of Deviance Table	
Model
Log(likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-47.0806
6



Fitted model
-49.4567
1
4.75213
5
0.4469
Reduced model
-71.5289
1
48.8965
5
<0.0001
2
3	AIC: = 100.913
4
5	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
20
0
4
0.0232
0.463
0
20
-0.689
8
0.1709
3.418
3
20
-0.248
10
0.3065
6131
6
20
-0.063
12
0.4688
9.375
13
20
1.624
15
0.7093
14.186
12
20
-1.076
6
7	ChiA2 = 4.34 d.f. = 5 p-value = 0.5021
8	Table D-6. Model predictions for convulsions in female F344 rats exposed to
9	RDX by gavage on GDs 6-19 fCholakiset al.. 19801: BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.989
42.003
0.866
0.149
Of the multistage models, the
quantal-linear model was
selected based on lowest AIC.
From among the quantal-linear
and non-multistage models, the
quantal-linear model is selected
based on lowest BMDL (BMDLs
differed by more than threefold).
Logistic
0.526
43.556
2.46
1.05
LogLogistic
0.991
41.996
0.902
0.201
Probit
0.577
43.348
1.96
0.871
LogProbit
1.000
41.963
1.11
0.335
Weibull
0.983
42.026
0.798
0.148
Multistage 3°b
0.960
42.113
0.638
0.146
Multistage 2°c
0.960
42.113
0.638
0.146
Quantal-Linear
0.669
42.077
0.179
0.123
10
11	aSelected model in bold; scaled residuals for selected model for doses 0, 0.2, 2, and 20 mg/kg-day were 0.00,
12	-0.52, -1.03, and 0.49, respectively. The BMDio and BMDLio values for the selected model were 1.88 and
13	1.29 mg/kg-day, respectively; the BMDos and BMDLos values for the selected model were 0.915 and
14	0.628 mg/kg-day, respectively.
15	bThe Multistage 3° model may appear equivalent to the Multistage 2° model; however, differences exist in digits
16	not displayed in the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-ll	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Quantal Linear Model, with BMR of 1% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the I
Quantal Linear
j	
5
_i	
10
_i	
15
	i	
20
dose
14:19 07/28 2014
BMR = 1% ER; dose shown in mg/kg-day.
Figure D-4. Plot of incidence rate by dose, with the fitted curve of the selected
model, for convulsions in female F344 rats exposed to RDX by gavage on
GDs 6-19 fCholakisetal.. 19801.
Quantal Linear Model using Weibull Model (Version: 2.16; Date: 2/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-slope*dose)]
Benchmark Dose Computation
BMR = 1% ER
BMD = 0.179224
BMDL atthe 95% confidence level = 0.122966
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0.0384615
Slope
0.056077
0.0588587
Power
Not applicable
1
This document is a draft for review purposes only and does not constitute Agency policy.
D-12	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1 Analysis of Deviance Table	
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-18.9808
4



Fitted model
-20.0384
1
2.11537
3
0.5488
Reduced model
-47.9793
1
57.9972
3
<0.0001
2
3	AIC:= 42.0769
4
5	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
24
0
0.2
0.0112
0.268
0
24
-0.52
2
0.1061
2.546
1
24
-1.025
20
0.6742
16.856
18
25
0.488
6
7	ChiA2 = 1.56 d.f. = 3 p-value = 0.6686
8	Table D-7. Model predictions for combined incidence of convulsion and
9	mortality in male and female F344 rats exposed to RDX by gavage for 90 days
10	(Grouse et al.. 2006): BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.245
99.260
3.73
2.10
The log-logistic and quantal-linear
models did not achieve an
adequate fit (goodness-of-fit
p-value <0.10). The multistage 2°
model was excluded from model
selection because the residual in
the lowest dose group, near the
BMD, was above 1.5 in absolute
value. Of the remaining
multistage models, the
multistage 3° model was selected
based on lowest AIC. From
among the multistage 3° and
non-multistage models, the
multistage 3° model was selected
based on lowest BMDL (BMDLs
differed by more than threefold).
Dichotomous-Hill
0.436
98.317
5.22
3.04
Logistic
0.0859
102.17
1.81
0.846
LogLogistic
0.305
98.593
3.70
2.20
Probit
0.101
101.85
2.16
0.853
LogProbit
0.316
98.465
4.22
2.75
Weibull
0.152
101.16
2.45
1.24
Multistage 4°b
0.229
99.182
2.56
0.486
Multistage 30c
0.229
99.182
2.56
0.486
Multistage 2°
0.165
102.01
1.22
0.470
Quantal-Linear
0.0052
113.90
0.144
0.113
11
12	aSelected model in bold; scaled residuals for selected model for doses 0, 4, 8,10,12, and 15 mg/kg-day were 0,
13	-0.88, -0.14, -0.01,1.92, and -1.55, respectively. The BMDio and BMDLio values for the selected model were
14	5.60 and 3.85 mg/kg-day, respectively; the BMDos and BMDLos values for the selected model were 4.41 and
15	2.25 mg/kg-day, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
D-13	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	bThe Multistage 4° model may appear equivalent to the Multistage 3° model; however, differences exist in digits
2	not displayed in the table.
3	cThe Multistage 3° model may appear equivalent to the Multistage 4° model; however, differences exist in digits
4	not displayed in the table.
5
Multistage Model, with BMR of 1% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
Multistage
1
0.8
0.6
0.4
0.2
0
2
4
6
8
10
12
14
dose
g	13:21 01/06 2016
7	BMR = 1% ER; dose shown in mg/kg-day.
8	Figure D-5. Plot of incidence rate by dose with fitted curve for Multistage 3°
9	model for model predictions for combined incidence of convulsion and
10	mortality in male and female F344 rats exposed to RDX by gavage for 90 days
11	(Crouse et al.. 2006).
12
13	Multistage Model (Version: 3.4; Date: 05/02/2014)
14	The form of the probability function is: P [response] = background +
15	(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
16
17	Benchmark Dose Computation
18	BMR = 1% Extra risk
19	BMD = 2.56012
20	BMDL at the 95% confidence level = 0.486284
21
22	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Beta(l)
0
0.0272036
Beta(2)
0
0.00626035
Beta(3)
0.000598962
0
This document is a draft for review purposes only and does not constitute Agency policy.
D-14	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1
2	Analysis of Deviance Table	
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-44.71
6



Fitted model
-48.59
1
7.76102
5
0.17
Reduced model
-9.88
1
70.3406
5
<0.0001
3
4	AIC:= 99.1817
5
6	Goodness-of-Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
20
0
4
0.0376
0.752
0
20
-0.88
8
0.2641
5.282
5
20
-0.14
10
0.4506
9.012
9
20
-0.01
12
0.6448
12.896
17
20
1.92
15
0.8675
17.351
15
20
-1.55
7
8	ChiA2 = 6.88 d.f. = 5 p-value = 0.2294
9
10	Male Reproductive Effects
11	Table D-8 (and Figure D-6) presents the BMD modeling results for incidence of testicular
12	degeneration for male B6C3Fi mice based on data from Lish etal. (1984). using a BMR of 10% ER.
13	Table D-8. Model predictions for testicular degeneration in male B6C3Fi mice
14	exposed to RDX by diet for 24 months fLish etal.. 19841: BMR = 10% ER
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMDLioPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gammab
Weibull
Quantal-Linear
0.357
101.10
66.5
35.4
The Log-Probit model was
selected based on lowest BMDL
(BMDLs differed by more than
threefold. The multistage models
had the same AIC values and
BMDLs, so selection of a
Logistic
0.159
103.40
97.1
66.1
LogLogistic
0.377
100.91
63.6
32.3
Probit
0.178
103.12
93.1
61.4
This document is a draft for review purposes only and does not constitute Agency policy.
D-15	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMD LioPct
(mg/kg-d)
Basis for model selection
representative multistage model
was unnecessary.)
p-value
AIC
LogProbit
0.876
97.564
56.0
16.3
Multistage 2°c
Multistage 3°
Multistage 4°
0.357
101.10
66.5
35.4
aSelected model in bold; scaled residuals for selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were 0.00,
0.32, -0.61, 0.43, and -0.17, respectively. The BMDos and BMDLos values for the selected model were 7.42 and
0.0477 mg/kg-day, respectively; the BMDoi and BMDLoi values for the selected model were 0.168 and
2.83 x 10"13 mg/kg-day, respectively.
bFor the Gamma and Weibull models, the power parameter estimates were 1 (boundary of parameter space). The
models in this row are equivalent to the Quantal-Linear model.
The Multistage 3° and 4° model had b3 and b4 coefficient estimates of 0 (boundary of parameters space). The
models in this row reduced to the Multistage 2° model. The models in this row may appear equivalent to the
Gamma model; however, differences exist in digits not displayed in the table.
LogProbit Model, with BMR of 10% Extra Risk for the BMD arid 0.95 Lower Confidence Limit for the B
Log Pro bit
0.3
0.25
0.2
0.15
0.1
0.05
BMDL
BMD
O
20
40
60
80
100
dose
1 3:42 02/1 4 201 4
Figure D-6. Plot of incidence rate by dose, with fitted curve for selected model,
for testicular degeneration in male B6C3Fi mice exposed to RDX by diet for
24 months (Lish et al.. 1984).
Probit Model (Version: 3.3; Date: 2/28/2013)
The form of the probability function is: P [response] = Background + (1-Background) *
CumNorm(Intercept+Slope*Log(Dose)), where CumNorm(.) is the cumulative normal
distribution function
Slope parameter is not restricted
This document is a draft for review purposes only and does not constitute Agency policy.
D-16	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Benchmark Dose Computation
2	BMR = 10% ER
3	BMD = 55.9784
4	BMDL at the 95% confidence level = 16.2787
5
6	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Intercept
-2.0054E+00
-1.9976E+00
Slope
0.179828
0.172286
7
8	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-46.4212
5



Fitted model
-46.7817
2
0.721088
3
0.8682
Reduced model
-52.1663
1
11.4902
4
0.02157
9
10	AIC:= 97.5635
11
12	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
63
0
1.5
0.0267
1.599
2
60
0.321
7
0.0489
3.033
2
62
-0.608
35
0.086
5.072
6
59
0.431
107
0.122
3.294
3
27
-0.173
13
14	ChiA2 = 0.69 d.f. = 3 p-value = 0.8759
15
16	Kidney/Urogenital System Effects
17	Table D-9 (and Figure D-7) presents the BMD model results for incidence of suppurative
18	inflammation of the prostate in male F344 rats based on data from Levine etal. f 19831 using a BMR
19	of 10% ER.
This document is a draft for review purposes only and does not constitute Agency policy.
D-17	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Table D-9. Model predictions for prostate suppurative inflammation in male
F344 rats exposed to RDX by diet for 24 months (Levine et al.. 1983):
BMR = 10% ER
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMDLioPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gammab
Multistage 2°
Quantal-Linear
Multistage 3°
Multistage 4°
0.288
200.37
4.61
3.24
The Log-Probit model is selected
based on lowest BMDL (BMDLs
differ by more than threefold.
The multistage models had the
same AIC values and BMDLs, so
selection from among the
multistage models was
unnecessary.)
Logistic
0.102
203.50
10.8
8.58
LogLogistic
0.328
200.05
3.33
2.09
Probit
0.116
203.10
9.91
7.96
LogProbit
0.204
202.03
1.67
0.469
Weibull8
0.288
200.37
4.61
3.24
aSelected model in bold; scaled residuals for selected model for doses 0, 0.3,1.5, 8, and 40 mg/kg-day were
-0.289, 0.172, 0.846, -1.298, and 0.819, respectively. The BMDos and BMDLos values for the selected model were
0.702 and 0.122 mg/kg-day, respectively; the BMDoi and BMDLoi values for the selected model were 0.137 and
0.00906 mg/kg-day, respectively.
bThe Gamma model had a power parameter estimate of 1 (boundary of parameter space). The Multistage 2°, 3°,
and 4° models had b2, b3, and b4 coefficients of 0 (boundary of parameter space). The models in this row are
equivalent to the Quantal-Linear model.
cThe Weibull model may appear equivalent to the Quantal-Linear model; however, differences exist in digits not
displayed in the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-18	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
LogProbit Model, with BMR of 10% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BIN
0.8
0.7
0.6
H	05
o
.CD
^	0.4
c
o
o
£	0.3
LL.
0.2
0.1
O
13:39 02/14
Figure D-7. Plot of inddence rate by dose, with fitted curve for selected model,
for prostate suppurative inflammation in male F344 rats exposed to RDX by
diet for 24 months (Levine et al.. 1983).
Probit Model (Version: 3.3; Date: 2/28/2013)
The form of the probability function is: P [response] = Background + (1-Background) *
CumNorm(Intercept+Slope*Log(Dose)), where CumNorm(.) is the cumulative normal
distribution function
Slope parameter is not restricted
Benchmark Dose Computation
BMR =10% ER
BMD = 1.67454
BMDL at the 95% confidence level = 0.468688
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.0452202
0.037037
Intercept
-1.4970E+00
-1.3564E+00
Slope
0.417872
0.36341
LogProbit
BMD
This document is a draft for review purposes only and does not constitute Agency policy.
D-19	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1 Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-96.3905
5



Fitted model
-98.0147
3
3.24837
2
0.1971
Reduced model
-118.737
1
44.6933
4
<0.0001
2
3	AIC:= 202.029
4
5	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0.0452
2.442
2
54
-0.289
0.3
0.0669
3.682
4
55
0.172
1.5
0.1332
6.927
9
52
0.846
8
0.2982
16.402
12
55
-1.298
40
0.5396
16.726
19
31
0.819
6
7	ChiA2 = 3.18 d.f. = 2 p-value = 0.2035
8
9	D.1.3. Mortality: Dose-Response Analysis and BMD Modeling Documentation
10	This appendix also presents a quantitative dose-response analysis of mortality incidence
11	from studies identified in Section 2.1.6 (see Table D-10).
12	Table D-10. Mortality data selected for dose-response modeling for RDX



Incidence/total (%) or
Reference
Species/sex
Dose
mean ± SD (number of animals)
Lish etal. (1984)
Male B6C3Fi
0 mg/kg-d
1/85 (1%)
(mortality at 11 wks)
mouse
1.5
0/85 (0%)


7
0/85 (0%)


35
0/85 (0%)


175/100
30 / 85 (35%)
Lish etal. (1984)
Female
0 mg/kg-d
0/85 (1%)
(mortality at 11 wks)
B6C3Fi mouse
1.5
0/85 (0%)


7
0/85 (0%)


35
0/85 (0%)


175/100
36 / 85 (42%)
Levine et al. (1981b)a
Female F344
0 mg/kg-d
0/30 (0%)

rat
10
1/10(10%)


30
0/10 (0%)


100
5/10 (50%)


300
10 /10 (100%)


600
10 /10 (100%)
This document is a draft for review purposes only and does not constitute Agency policy.
D-20	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine



Incidence/total (%) or
Reference
Species/sex
Dose
mean ± SD (number of animals)

Male F344 rat
0 mg/kg-d
0/30 (0%)


10
0/10 (0%)


30
0/10 (0%)


100
8/10 (80%)


300
10 /10 (100%)


600
10 /10 (100%)

Male and
0 mg/kg-d
0/60 (0%)

female F344
10
1/20 (5%)

rat, combined
30
0/20 (0%)


100
13 / 20 (65%)


300
20 / 20 (100%)


600
20 / 20 (100%)
von Oettingen et al.
Rats,
0 mg/kg-d
0/20 (0%)
(1949)
sex/strain not
15
0/19 (0%)b

specified
25
8/20 (40%)


50
8/20 (40%)
Cholakis et al. (1980)
Female CD rat
0 mg/kg-d
0/22 (0%)
(2-generation study)

5
0/22 (0%)


16
0/22 (0%)


50
6/22 (27%)
Levine et al. (1983)
Male F344 rat
0 mg/kg-d
0/75 (0%)
(mortality at 13 wks)

0.3
0/75 (0%)


1.5
0/75 (0%)


8
0/75 (0%)


40
0/75 (0%)

Female F344
0 mg/kg-d
0/75 (0%)

rat
0.3
0/75 (0%)


1.5
0/75 (0%)


8
0/75 (0%)


40
0/75 (0%)
Cholakis et al. (1980)
Male F344 rat
0 mg/kg-d
0/10 (0%)
(13-wk study)

10
0/10 (0%)


14
0/10 (0%)


20
0/10 (0%)


28
0/10 (0%)


40
0/10 (0%)

Female F344
0 mg/kg-d
0/9 (0%)c

rat
10
0/10 (0%)


14
0/10 (0%)


20
0/10 (0%)


28
0/10 (0%)


40
0/10 (0%)
This document is a draft for review purposes only and does not constitute Agency policy.
D-21	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine



Incidence/total (%) or
Reference
Species/sex
Dose
mean ± SD (number of animals)
Crouse et al. (2006)
Female F344
0 mg/kg-d
0/10 (0%)

rat
4
0/10 (0%)


8
1/10 (20%)


10
2/10 (20%)


12
5/10 (50%)


15
4/10 (40%)

Male F344 rat
0 mg/kg-d
0/10 (0%)


4
0/10 (0%)


8
1/10(10%)


10
3/10 (30%)


12
2/10 (20%)


15
3/10 (30%)

Male and
0 mg/kg-d
0/20 (0%)

female F344
4
0/20 (0%)

rat, combined
8
2/20(10%)


10
5/20 (25%)


12
7/20 (35%)


15
7/20 (35%)
Cholakis et al. (1980)
Female F344
0 mg/kg-d
0/24 (0%)
(gestational exposure)
rats
0.2
0/24 (0%)

(gestational
2
0/24 (0%)

exposure)
20
5/24 (21%)
Angerhofer et al. (1986)
Female SD
0 mg/kg-d
0/39 (0%)

rate
2
1/40 (3%)

(gestational
6
1/40 (3%)

exposure)
20
16/51 (31%)
Cholakis et al. (1980)
Female New
0 mg/kg-d
0/11 (0%)

Zealand white
0.2
0/11 (0%)

rabbit
2
0/11 (0%)

(gestational
20
0/12 (0%)

exposure)


1
2	aFor Levine et al. (1981a) and Crouse et al. (2006), the incidence rates across doses were determined to be not
3	statistically significantly different between the males and females using an exact Cochran-Mantel-Haenszel test
4	(p > 0.10). The data were combined across sex for each of these endpoints prior to modeling.
5	bFor von Oettingen et al. (1949), one mortality was reported in the 15 mg/kg-day dose group. However, this
6	mortality was most likely not related to RDX, so the animal that died was excluded.
7	Tor Cholakis et al. (1980), one accidental death was reported in the 0 mg/kg-day dose group. The animal that died
8	was excluded.
9
10	Tables D-ll to D-14 presentthe BMD modeling results for incidence of mortality from
11	Crouse etal. (2006). von Oettingen etal. (1949). Levine etal. (1983). and Angerhofer etal. (1986).
12	The following datasets were not modeled because each had either no response or a positive
13	response only in the highest dose group: 11-week mortality from Lish etal. (1984). both male (one
14	death in control group) and female; 13-week mortality data from Levine etal. f!983I both male
This document is a draft for review purposes only and does not constitute Agency policy.
D-22	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	and female; mortality in female CD rats and male and female F344 rats from Cholakis etal. (1980):
2	and mortality in female F344 rats during gestational exposure from Cholakis et al. (1980).
3	Table D-ll. BMD modeling results for combined mortality in male and female
4	F344 rats exposed to RDX by diet for 13 weeks (Levine etal.. 1981b):
5	BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.401
41.100
49.8
12.7
The quantal-linear model was
excluded because the fit has a
residual below -2 at 4 mg/kg-d.
The multistage 4° model was
selected as the representative
multistage model based on
lowest AIC. From among the
multistage 4° and non-multistage
models, the multistage 4° model
was selected based on lowest
BMDL (BMDLs differed by more
than threefold).
Logistic
0.346
41.429
18.3
8.25
LogLogistic
0.257
43.098
73.2
16.9
Probit
0.328
41.727
15.0
6.82
LogProbit
0.257
43.098
58.6
19.5
Weibull
0.257
43.101
56.6b
6.51b
Multistage 2°
0.424
42.942
7.72
2.01
Quantal-Linear
0.139
50.257
1.12
0.818
Multistage 3°
0.503
41.520
7.71
2.08
Multistage 4°
0.535
40.935
7.85
2.15
Multistage 5°
0.371
42.928
7.86
2.15
6
7	aSelected model in bold; scaled residuals for selected model for doses 0,10, 30,100, 300, and 600 mg/kg-day were
8	0.00,1.48, -0.97, 0.05, 0.00, and 0.00, respectively. The BMDio and BMDLio estimates for the selected model
9	were 47.2 and 22.2 mg/kg-day, respectively; the BMDos and BMDLos estimates for the selected model were
10	32.4 and 11.0 mg/kg-day, respectively.
11	bThe parameter convergence parameter was increased to 2 x 10"8 to obtain convergence.
12
This document is a draft for review purposes only and does not constitute Agency policy.
D-23	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Model, with BMR of 1% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BIV
Multistage
O
1 OO
Figure D-8. Plot of inddence rate by dose, with the fitted curve of the
multistage 2° model, for combined mortality in male and female F344 rats
exposed to RDX by diet for 13 weeks (Levine etal.. 1981b): BMR = 1% ER.
Multistage Model (Version: 3.3; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
Benchmark Dose Computation
BMR = 1% Extra risk
BMD = 7.85287
BMDL atthe 95% confidence level = 2.15059
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Beta(l)
0.00127544
1.9710E+17
Beta(2)
0
0
Beta(3)
0
0
Beta(4)
9.0721E-09
0
Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-16.9192
6



Fitted model
-18.4677
2
3.09685
4
0.5418
Reduced model
-102.298
1
170.758
5
<0.0001
This document is a draft for review purposes only and does not constitute Agency policy.
D-24	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	AIC:= 40.9353
2
3	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
60
0
10
0.0128
0.255
l
20
1.484
30
0.0446
0.892
0
20
-0.966
100
0.6447
12.894
13
20
0.05
300
1
20
20
20
0
600
1
20
20
20
0
4
5	ChiA2 = 3.14 d.f. = 4 p-value = 0.5352
6
7	Table D-12. BMD modeling results for mortality (number found dead) in rats
8	exposed to RDX in the diet for 13 weeks (von Oettingen etal.. 19491
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.0341
66.088
3.14
0.648
All of the models besides
dichotomous-Hill had goodness-
of-fit p-values <0.10 and thus did
not provide an adequate fit to the
data. For the dichotomous-Hill
model, the slope parameter
acheived the BMDS internal
upper bound (18), so the results
from this model were not
reliable. No model was selected.
Dichotomous-Hill
0.984
57.888
17.2
10.9
Logistic
0.0044
70.074
3.40
2.08
LogLogistic
0.0397
65.853
3.39
0.529
Probit
0.0056
69.283
3.28
1.94
LogProbit
0.0426
65.464
5.67
0.409
Weibull
0.0349
66.233
2.30
0.641
Multistage 3°a
0.0351
66.517
1.22
0.628
Multistage 2°b
0.0351
66.517
1.22
0.628
Quantal-Linear
0.0995
64.639
0.919
0.623
9
10	aThe Multistage 3° model may appear equivalent to the Multistage 2° model; however, differences exist in digits
11	not displayed in the table.
12	bThe Multistage 2° model may appear equivalent to the Multistage 3° model; however, differences exist in digits
13	not displayed in the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-25	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Dichotomous-Hill Model, with BMR of 1 % Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
0.7
0.6
0.5
-a 0.4
.J 0.3
0.2
0.1
O
O	10	20	30	40	50
14:56 02/04 2016
Figure D-9. Plot of incidence rate by dose with fitted curve for Dichotomous-
Hill model for Model predictions for mortality (number found dead) in rats
exposed to RDX in the diet for 13 weeks (von Oettingen et al.. 1949): dose
shown in mg/kg-day.
Table D-13. BMD modeling results for combined mortality (number found
dead) in male and female F344 rats exposed to RDX by gavage for 90 days
fCrouse et al.. 20061: BMR= 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.794
93.263
3.46
0.840
The multistage 2° model was
selected as the representative
multistage model based on
lowest AIC. From among the
multistage 2° and non-multistage
models, the multistage 2° model
was selected based on lowest
BMDL (BMDLs differed by more
than threefold).
Logistic
0.474
95.709
2.11
1.11
LogLogistic
0.794
93.332
3.17
0.872
Probit
0.574
94.797
2.40
1.07
LogProbit
0.854
92.832
3.96
1.48
Weibull
0.743
93.698
2.76
0.641
Multistage 2°
0.858
91.926
2.11
0.463
Quantal-Linear
0.535
95.345
0.405
0.288
Multistage 5°b
0.731
93.851
2.42
0.433
Multistage 4°c
0.731
93.851
2.42
0.433
Multistage 3°
0.731
93.851
2.42
0.439
Dichotomous-Hill
0.998
93.343
5.96
1.95
aSelected model in bold; scaled residuals for selected model for doses 0, 4, 8,10,12, and 15 mg/kg-day were 0.00,
-0.86, -0.46, 0.53, 0.72, and 0.45, respectively. The BMDio and BMDLio values for the selected model were
6.82 and 4.41 mg/kg-d, respectively.
bThe Multistage 5° model may appear equivalent to the Multistage 4° model; however, differences exist in digits
not displayed in the table.
This document is a draft for review purposes only and does not constitute Agency policy.
D-26	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Model, with BMR of 1 % Extra Risk for the BMD arid 0.95 Lower Confidence Limit for the BPu
Multistage
0.6
0.5
0.4
0.3
0.2
0.1
O
BMDI
BMD
O
2
4
6
8
10
12
14
dose
14:12 02/27 201 4
Figure D-10. Plot of inddence rate by dose, with the fitted curve of the
multistage 2° model, for mortality in male and female F344 rats exposed to
RDX by gavage for 90 days (Crouse etal.. 20061. BMR = 1% ER; dose shown in
mg/kg-day.
Multistage Model (Version: 3.3; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
Benchmark Dose Computation.
BMR = 1% Extra risk
BMD = 2.10625
BMDL at the 95% confidence level = 0.462994
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0
Beta(l)
0
0.0134587
Beta(2)
0.00226548
0.00141278
This document is a draft for review purposes only and does not constitute Agency policy.
D-27	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1 Analysis of Deviance Table	
Model
Log (likelihood)
Number of
parameters
Deviance
Test d.f.
p-value
Full model
-43.6462
6



Fitted model
-44.963
1
2.63354
5
0.7563
Reduced model
-55.6472
1
24.0019
5
0.0002169
2
3	AIC:= 91.926
4
5	Goodness-of-Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
20
0
4
0.0356
0.712
0
20
-0.859
8
0.135
2.699
2
20
-0.458
10
0.2027
4.054
5
20
0.526
12
0.2784
5.567
7
20
0.715
15
0.3993
7.987
7
20
-0.451
6
7	ChiA2 = 1.94 d.f. = 5 p-value = 0.8576
8	Table D-14. Model predictions for mortality in female Sprague-Dawley rats
9	exposed to RDX by gavage on gestation days 6-15 (Angerhofer et al.. 19861:
10	BMR = 1% ER
Model3
Goodness of fit
BMDiPct
(mg/kg-d)
BMDLiPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Gamma
0.314
89.496
5.15
0.538
The multistage 3° model was
selected as the representative
multistage model based on
lowest AIC. From among the
multistage 3° and non-multistage
models, the multistage 3° model
was selected based on lowest
BMDL (BMDLs differed by more
than threefold).
Logistic
0.667
87.213
3.88
2.16
LogLogistic
0.312
89.473
4.88
0.560
Probit
0.643
87.196
3.37
1.87
LogProbit
0.319
89.522
5.58
0.885
Weibull
0.309
89.458
4.62
0.541
Quantal-Linear
0.450
87.502
0.652
0.452
Multistage 3°
0.655
86.906
1.68
0.588
Multistage 2°
0.554
87.291
1.78
0.555
11
12	aSelected model in bold; scaled residuals for selected model for doses 0, 2, 6, and 20 mg/kg-day were 0.00, 0.76,
13	-0.52, and 0.04, respectively. The BMDio and BMDLio values for the selected model were 10.9 and
14	6.09 mg/kg-day, respectively; the BMDos and BMDLos estimates for the selected model were 5.23 and
15	7.29 mg/kg-day, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
D-28	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Model, with BMR of 1 % Extra Risk for the BMD arid 0.95 Lower Confidence Limit for the BPu
Multistage
BMDL
BMD
O
5
10
15
20
dose
1 0:18 05/22 2014
Figure D-ll. Plot of incidence rate by dose, with the fitted curve of the
multistage 3° model, for mortality in female Sprague-Dawley rats exposed to
RDX by gavage on gestation days 6-15 fAngerhofer et al.. 19861: BMR = 1% ER.
Multistage Model (Version: 3.3; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
Benchmark Dose Computation
BMR = 1% Extra risk
BMD = 1.68097
BMDL atthe 95% confidence level = 0.587568
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0
0.00807857
Beta(l)
0.00588873
0.00216407
Beta(2)
0
0
Beta(3)
0.0000319123
0.0000406218
This document is a draft for review purposes only and does not constitute Agency policy.
D-29	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1 Analysis of Deviance Table	
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-41.0771
4



Fitted model
-41.4531
2
0.752152
2
0.6866
Reduced model
-57.4292
1
32.7043
3
<0.0001
2
3	AIC:= 86.9063
4
5	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0
0
0
39
0
2
0.012
0.478
l
40
0.759
6
0.0413
1.654
l
40
-0.519
20
0.3114
15.881
16
51
0.036
6
7	ChiA2 = 0.85 d.f. = 2 p-value = 0.6549
8
9	D.2. BENCHMARK DOSE MODELING SUMMARY FOR CANCER ENDPOINTS
10	The cancer endpoints in the mouse that were selected for dose-response modeling are
11	presented in Table D-15. For each endpoint, the doses and tumor incidence data used for the
12	modeling are presented.
13	Table D-15. Cancer endpoints selected for dose-response modeling for RDX
Endpoint and reference
Species/sex
Dose (mg/kg-d)
Incidence/total
Hepatocellular adenomas or
Female B6C3Fi
0
1/67 (1%)a
carcinomas
mouse
1.5
4/62 (6%)
Parker et al. (2006)

7
5/63 (8%)


35
10 /64 (16%)


107
4/31 (13%)
Alveolar/bronchiolar adenomas or
Female B6C3Fi
0
7/65(11%)
carcinomas
mouse
1.5
3/62 (5%)
Lish et al. (1984)

7
8/64 (13%)


35
12/64 (19%)


107
7/31 (23%)
14
15	aFor female mouse hepatocellular tumors from Lish et al. (1984), tumor incidence and totals are those reported in
16	the Pathology Working Group (PWG) reevaluation (Parker et al., 2006).
This document is a draft for review purposes only and does not constitute Agency policy.
D-30	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
D.2.1. Evaluation of Model Fit and Model Selection for Mouse Tumor Data
First, to determine whether a time-to-tumor analysis was warranted, the survival curves
were compared across dose groups for female mice in Lish etal. f19841 in the study to determine
whether time of death should be incorporated in the dose-response analysis of tumors. A log-rank
test on the Kaplan-Meier survival curves per dose was used to do the comparison, excluding
animals that died prior to week 11 when the dose was reduced in the high-dose group to
100 mg/kg-day. The test yielded a nonsignificant result (p = 0.51), so a time-to-tumor analysis was
not necessary for this study.
Therefore, non-time-dependent dose-response analyses were conducted using standard
BMDS models. For each tumor type, BMDS multistage-cancer models6 were fitted to the data using
the maximum likelihood method. Each model was tested for goodness-of-fit using a chi-square
goodness-of-fit test (x2 p-value <0.057 indicates lack of fit). Other factors were used to assess model
fit, including scaled residuals, visual fit, and adequacy of fit in the low-dose region and in the
vicinity of the BMR. The BMDL estimate and AIC value were used to select a best-fit model from
among the models exhibiting adequate fit If the BMDL estimates were "sufficiently close"
(i.e., differed by threefold or less), then the model selected was the one that yielded the lowest AIC
value. If the BMDL estimates were not sufficiently close, then the lowest BMDL was selected as the
POD.
After selecting models for the two endpoints, the results were combined using MS-COMBO
in BMDS. This procedure analyzes the incidence of a tumor (adenoma or carcinoma) defined as
present if either the hepatocellular or alveolar/bronchiolar tumor (or both) was present, and not
present otherwise. The two endpoints were assumed to be independent.
D.2.2. Modeling Results for Mouse Tumor Data
The BMD modeling results for mouse tumor data sets are provided in Tables D-16 to D-20
(and Figures D-12 to D-16).
6The coefficients of the multistage-cancer models were restricted to be nonnegative (beta values >0).
7 A significance level of 0.05 from U.S. EPA (2012b) is used for selecting cancer models because the model
family (multistage) is selected a priori.
This document is a draft for review purposes only and does not constitute Agency policy.
D-31	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Mouse Tumor Data—BMD Modeling Results
Table D-16. Model predictions for combined alveolar/bronchiolar adenoma
and carcinoma in female B6C3Fi mice exposed to RDX by diet for 24 months
fLish etal.. 19841: BMR = 10% ER
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMDLiopct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage lob
Multistage 2°
Multistage 3°
Multistage 4°
0.417
218.68
52.8
27.7
All of the models reduced to the
Multistage 1° model, so this model
was selected.
aSelected model in bold. Scaled residuals for the selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were
0.40, -1.27, 0.50, 0.73, and -0.52, respectively.
bFor the Multistage 2°, 3°, and 4° models, the b2, b3, and b4 coefficient estimates were 0 (boundary of parameter
space). The models in this row reduced to the Multistage 1° model.
Multistage Cancer Model, with BMR of 10% Extra Risk for the BMD arid 0.95 Lower Confidence Limit fort
q 45 p	Multistage Cancer	1
Linear extrapolation 	
0.4 i-	-i
0.35
0.3
0.25
0.2
0.15
0.1
0.05
O
BMD
BMDI
O
20
40
60
80
100
dose
16:09 02/14 2014
Figure D-12. Plot of incidence rate by dose, with the fitted curve for the
selected model, for combined alveolar/bronchiolar adenoma and carcinoma
in female B6C3Fi mice exposed to RDX by diet for 24 months fLish etal..
1984)-
Multistage Cancer Model (Version: 1.10; Date: 02/28/2013)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
The parameter betas are restricted to be positive
This document is a draft for review purposes only and does not constitute Agency policy.
D-32	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Benchmark Dose Computation
2	BMR = 10% ER
3	BMD = 52.8078
4	BMDL at the 95% confidence level = 27.748
5	Benchmark dose upper bound (BMDU) at the 95% confidence level = 194.806
6	Taken together, (27.748,194.806) is a 90% two-sided confidence interval for the BMD
7	Multistage Cancer Slope Factor = 0.00360387
8
9	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.093168
0.0998927
Beta(l)
0.00199517
0.00155773
10
11	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-105.777
5



Fitted model
-107.341
2
3.12764
3
0.3724
Reduced model
-110.164
1
8.77367
4
0.06701
12
13	AIC:= 218.682
14
15	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0.0932
6.056
7
65
0.403
1.5
0.0959
5.944
3
62
-1.27
7
0.1057
6.768
8
64
0.501
35
0.1543
9.877
12
64
0.734
107
0.2675
8.292
7
31
-0.524
16
17	ChiA2 = 2.84 d.f. = 3 p-value = 0.4168
18
This document is a draft for review purposes only and does not constitute Agency policy.
D-33	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Table D-17. Model predictions for combined alveolar/bronchiolar adenoma
and carcinoma in female B6C3Fi mice exposed to RDX by diet for 24 months
fLish etal.. 19841: BMR = 5% ER
Model3
Goodness of fit
BMDspct
(mg/kg-d)
BMDLspct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage l°b
Multistage 2°
Multistage 3°
Multistage 4°
0.417
218.68
25.7
13.5
All of the models reduced to the
Multistage 1° model, so this model
was selected.
aSelected model in bold. Scaled residuals for the selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were
0.40, -0.40, -1.27, 0.50, 0.73, and -0.52, respectively. The BMDio and BMDLio values for the selected model were
52.8 and 27.7 mg/kg-day, respectively.
Multistage Cancer Model, with BMR of 5% Extra Risk for the BMD and 0.95 Lower Confidence Limit forth
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
16:11 02/14 2014
Multistage Cancer
Linear extrapolation
BMDL
BMD
20
40
60
dose
80
100
Figure D-13. Plot of incidence rate by dose, with fitted curve for selected
model, for combined alveolar/bronchiolar adenoma and carcinoma in female
B6C3Fi mice exposed to RDX by diet for 24 months fLish et al.. 19841.
This document is a draft for review purposes only and does not constitute Agency policy.
D-34	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Multistage Cancer Model (Version: 1.10; Date: 02/28/2013)
2	The form of the probability function is: P [response] = background +
3	(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
4	The parameter betas are restricted to be positive
5
6	Benchmark Dose Computation
7	BMR = 5% ER
8	BMD = 25.7088
9	BMDL at the 95% confidence level = 13.5087
10	BMDU at the 95% confidence level = 94.8384
11	Taken together, (13.5087, 94.8384) is a 90% two-sided confidence interval for the BMD
12	Multistage Cancer Slope Factor = 0.00370131
13
14	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.093168
0.0998927
Beta(l)
0.00199517
0.00155773
15
16	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-105.777
5



Fitted model
-107.341
2
3.12764
3
0.3724
Reduced model
-110.164
1
8.77367
4
0.06701
17
18	AIC:= 218.682
19
20	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0.0932
6.056
7
65
0.403
1.5
0.0959
5.944
3
62
-1.27
7
0.1057
6.768
8
64
0.501
35
0.1543
9.877
12
64
0.734
107
0.2675
8.292
7
31
-0.524
21
22	ChiA2 = 2.84 d.f. = 3 p-value = 0.4168
23
This document is a draft for review purposes only and does not constitute Agency policy.
D-35	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Table D-18. Model predictions for combined hepatocellular adenoma and
carcinoma in female B6C3Fi mice exposed to RDX by diet for 24 months
fParker etal.. 20061: BMR = 10% ER
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMD LioPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage l°b
Multistage 2°
Multistage 3°
Multistage 4°
0.160
164.06
64.2
32.6
All of the models reduced to the
Multistage 1° model, so this model was
selected.
aSelected model in bold. Scaled residuals for the selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were
-1.37, 0.35, 0.54,1.34, and -1.05, respectively.
bFor the Multistage 2°, 3°, and 4° models, the b2, b3, and b4 coefficient estimates were 0 (boundary of parameter
space). The models in this row reduced to the Multistage 1° model.
Multistage Cancer Model, with BMR of 10% Extra Risk for the BMD arid 0.95 Lower Confidence Limit fort
Multistage Cancer
Linear extrapolation
0.3
0.25
0.2
0.15
0.1
0.05
O
BMD
BMD
O
20
40
60
80
100
dose
16:21 02/14 2014
Figure D-14. Plot of incidence rate by dose, with fitted curve for selected
model, for combined hepatocellular adenoma and carcinoma in female B6C3Fi
mice exposed to RDX by diet for 24 months fParker et al.. 20061.
This document is a draft for review purposes only and does not constitute Agency policy.
D-36	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Multistage Cancer Model (Version: 1.10; Date: 02/28/2013)
2	The form of the probability function is: P [response] = background +
3	(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
4	The parameter betas are restricted to be positive
5
6	Benchmark Dose Computation
7	BMR = 10% ER
8	BMD = 64.203
9	BMDL atthe 95% confidence level = 32.6282
10	BMDU atthe 95% confidence level = 281.385
11	Taken together, (32.6282, 281.385) is a 90% two-sided confidence interval for the BMD
12	Multistage Cancer Slope Factor = 0.00306483
13
14	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.0520755
0.0658334
Beta(l)
0.00164105
0.000876864
15
16	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-77.1516
5



Fitted model
-80.0315
2
5.75967
3
0.1239
Reduced model
-82.5216
1
10.74
4
0.02965
17
18	AIC: = 164.063
19
20	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0.0521
3.489
l
67
-1.369
1.5
0.0544
3.373
4
62
0.351
7
0.0629
3.963
5
63
0.538
35
0.105
6.719
10
64
1.338
107
0.2047
6.347
4
31
-1.045
21
22	ChiA2 = 5.17 d.f. = 3 p-value = 0.16
23
This document is a draft for review purposes only and does not constitute Agency policy.
D-37	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Table D-19. Model predictions for B6C3Fi female mouse combined
hepatocellular adenoma and carcinoma in mice exposed to RDX by diet for
24 months (Parker et al.. 2006): BMR = 5% ER
Model3
Goodness of fit
BMDspct
(mg/kg-d)
BMDLspct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage l°b
Multistage 2°
Multistage 3°
Multistage 4°
0.160
164.06
31.3
15.9
All of the models reduced to the
Multistage 1° model, so this model was
selected.
aSelected model in bold; scaled residuals for selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were -1.37,
0.35, 0.54,1.34, and -1.05, respectively. The BMDio and BMDLio values for the selected model were 64.2 and
32.6 mg/kg-day, respectively.
bFor the Multistage 2°, 3°, and 4° models, the b2, b3, and b4 coefficient estimates were 0 (boundary of parameter
space). The models in this row reduced to the Multistage 1° model.
Multistage Cancer Model, with BMR of 5% Extra Risk for the BMD arid 0.95 Lower Confidence Limit forth
Multistage Cancer
Linear extrapolation
0.3
0.25
0.2
0.15
0.1
0.05
O
BMDI
BMD
O
20
40
60
80
100
dose
16:21 02/14 2014
Figure D-15. Plot of incidence rate by dose, with fitted curve for selected
model, for B6C3Fi female mouse combined hepatocellular adenoma and
carcinoma in mice exposed to RDX by diet for 24 months (Parker et al.. 20061.
This document is a draft for review purposes only and does not constitute Agency policy.
D-38	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Multistage Cancer Model (Version: 1.10; Date: 02/28/2013)
2	The form of the probability function is: P [response] = background +
3	(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
4	The parameter betas are restricted to be positive
5
6	Benchmark Dose Computation
7	BMR = 5% ER
8	BMD = 31.2563
9	BMDL at the 95% confidence level = 15.8846
10	BMDU at the 95% confidence level = 136.989
11	Taken together, (15.8846,136.989) is a 90% two-sided confidence interval for the BMD
12	Multistage Cancer Slope Factor = 0.0031477
13
14	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.0520755
0.0658334
Beta(l)
0.00164105
0.000876864
15
16	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-77.1516
5



Fitted model
-80.0315
2
5.75967
3
0.1239
Reduced model
-82.5216
1
10.74
4
0.02965
17
18	AIC: = 164.063
19
20	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0.0521
3.489
l
67
-1.369
1.5
0.0544
3.373
4
62
0.351
7
0.0629
3.963
5
63
0.538
35
0.105
6.719
10
64
1.338
107
0.2047
6.347
4
31
-1.045
21
22	ChiA2 = 5.17 d.f. = 3 p-value = 0.16
23
This document is a draft for review purposes only and does not constitute Agency policy.
D-39	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Combined results for presence of hepatocellular or alveolar/bronchiolar adenoma or
carcinoma in B6C3Fi female mice exposed to RDX by diet for 24 months; BMR = 10% ER
BMD = 29.0 mg/kg-day; BMDL = 17.7 mg/kg-day
MSCOMBO results
BMR of 10% Extra Risk
**** Start of combined BMD and BMDL Calculations.****
Combined Log-Likelihood	-187.37235968 92213
Combined Log-likelihood Constant	166.0173 7 62 6058 8 41
Benchmark Dose Computation
Specified effect =	0.1
Risk Type = Extra risk
Confidence level =	0.95
BMD = 28.9753
BMDL = 17.6574
Multistage Cancer Slope Factor = 0.00566334
This document is a draft for review purposes only and does not constitute Agency policy.
D-40	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Supplemental Information—Hexahydro-1,3,5-trinitro-l,3,5-triazine
Combined results for presence of hepatocellular or alveolar/bronchiolar adenoma or
carcinoma in B6C3Fi female mice exposed to RDX by diet for 24 months; BMR = 5% ER
BMD = 29.0 mg/kg-day; BMDL = 17.7 mg/kg-day
MSCOMBO results
BMR of 5% Extra Risk
**** Start of combined BMD and BMDL Calculations.****
Combined Log-Likelihood	-187.37235968 92213
Combined Log-likelihood Constant	166.0173 7 62 6058 8 41
Benchmark Dose Computation
Specified effect =	0.05
Risk Type = Extra risk
Confidence level =	0.95
BMD = 14.1062
BMDL = 8.59627
Multistage Cancer Slope Factor = 0.00581647
This document is a draft for review purposes only and does not constitute Agency policy.
D-41	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Table D-20. Model predictions for combined hepatocellular adenoma and
carcinoma in female B6C3Fi mice exposed to RDX by diet for 24 months, using
incidence frequencies from Parker et al. (2006) and sample sizes from Lish et
al. fl9R41: BMR = 10% ER
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMD LioPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage lob
Multistage 2°
Multistage 3°
Multistage 4°
0.171
163.98
64.8
32.8
All of the models reduced to the
Multistage 1° model, so this model was
selected.
aSelected model in bold. Scaled residuals for the selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were
-1.34, 0.34, 0.49,1.34, and -1.03, respectively.
bFor the Multistage 2°, 3°, and 4° models, the b2, b3, and b4 coefficient estimates were 0 (boundary of parameter
space). The models in this row reduced to the Multistage 1° model.
Multistage Cancer Model, with BMR of 10% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
Multistage Cancer
Linear extrapolation
0.3
0.25
0.2
0.15
0.1
0.05
O
O
20
40
60
80
100
Figure D-16. Plot of incidence rate by dose with fitted curve for Multistage-
Cancer 1° model for model predictions for combined hepatocellular adenoma
and carcinoma in female B6C3Fi mice exposed to RDX by diet for 24 months,
using incidence frequencies from Parker et al. (2006) and sample sizes from
Lish et al. (1984).
This document is a draft for review purposes only and does not constitute Agency policy.
D-42	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Multistage Model (Version: 3.4; Date: 05/02/2014)
2	The form of the probability function is: P [response] = background +
3	(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
4	The parameter betas are restricted to be positive
5
6	Benchmark Dose Computation
7	BMR = 10% Extra risk
8	BMD = 64.7853
9	BMDL at the 95% confidence level = 32.7981
10	BMDU at the 95% confidence level = 291.495
11	Taken together, (32.7981, 291.495) is a 90% two-sided confidence interval for the BMD
12	Multistage Cancer Slope Factor = 0.00304896
13
14	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.0525105
0.0656105
Beta(l)
0.0016263
0.000878945
15
16	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-77.2
5



Fitted model
-79.99
2
5.57217
3
0.13
Reduced model
-82.43
1
10.462
4
0.03
17
18	AIC: = 163.978
19
20	Goodness-of-Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled residuals
0
0.0525
3.413
l
65
-1.34
1.5
0.0548
3.399
4
62
0.34
7
0.0632
4.047
5
64
0.49
35
0.1049
6.716
10
64
1.34
107
0.2038
6.319
4
31
-1.03
21
22	ChiA2 = 5.02 d.f. = 3 p-value = 0.1706
23
This document is a draft for review purposes only and does not constitute Agency policy.
D-43	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Combined results for presence of hepatocellular or alveolar/bronchiolar adenoma or
carcinoma in B6C3Fi female mice exposed to RDX by diet for 24 months; for hepatocellular
adenoma or carcinoma, the incidence frequencies from Parker et al. (2006) and the sample
sizes from Lish et al. f19841 were used; BMR = 10% ER
BMD = 29.0 mg/kg-day; BMDL = 17.7 mg/kg-day
MSCOMBO results
BMR of 10% Extra Risk
**** Start of combined BMD and BMDL Calculations.****
Combined Log-Likelihood	-187.33008597565913
Combined Log-likelihood Constant	166.0 68 4165505 47
Benchmark Dose Computation
Specified effect =	0.1
Risk Type	=	Extra risk
Confidence level =	0.95
BMD
29.0933
BMDL
17.7048
Multistage Cancer Slope Factor
0.00564818
This document is a draft for review purposes only and does not constitute Agency policy.
D-44	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	D.2.3. Dose-response Analysis and BMD Modeling Documentation for Other Tumor Data
2	Sets
3	This appendix also presents a quantitative dose-response analysis of incidence of liver
4	carcinomas in male F344 rats fLevine etal.. 19831 and incidence of lung carcinomas in male B3C6Fi
5	mice (Table D-21). The resulting candidate oral slope factors (OSFs) are presented for comparison
6	with OSF estimates provided in Section 2.3.3 of the Toxicological Review.
7	Table D-21. Liver carcinoma data from Levine et al. f1983)
Endpoint and reference
Species/sex
Dose (mg/kg-d)
Incidence/total
Alveolar/bronchiolar carcinomas
Male B6C3Fi
0
3/63 (5%)
Lish etal. (1984)
mouse
1.5
6/60(10%)


7
3/62 (5%)


35
7/59 (12%)


107
5/27(19%)
Hepatocellular carcinomas
Male F344 rat
0
1/55 (2%)
Levine et al. (1983)

0.3
0/55 (0%)


1.5
0/52 (0%)


8
2/55 (4%)


40
2/31a (6%)
8
9	aThe denominators listed in the table represent the number of animals that were alive 1 year after dosing began.
10
11	For male mice in Lish etal. (1984). a log-rank test on the Kaplan-Meier survival curves,
12	stratified by dose, yielded a nonsignificant result (p-value >0.10), indicating that the survival curves
13	were similar across dose groups. Therefore, a time-to-tumor analysis was not necessary for
14	hepatocellular carcinomas in Lish etal. (1984). A non-time-dependent dose-response analysis was
15	conducted using BMDS multistage-cancer models, and the model selection procedures described in
16	Section D.2.1 were used to select the appropriate models. Subsequently, the administered dose was
17	converted to a human equivalent dose (HED) on the basis of (body weight)3/4 fU.S. EPA. 19921. as
18	described in Section 2.3.2. The POD estimate for male mouse carcinomas and OSF calculated from
19	this POD are provided in Table D-22; detailed BMD modeling results are provided in Table D-23
20	(and Figure D-17).
This document is a draft for review purposes only and does not constitute Agency policy.
D-45	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table D-22. Model predictions and oral slope factor for alveolar/bronchiolar
2	carcinomas in male B6C3Fi mice exposed to RDX by diet for 2 years (Lish etal..
3	19841
Tumor type
Selected
model
BMR
BMD,
mg/kg-d
BMDL,
mg/kg-d
POD =
BMDLio-hed3
mg/kg-d
Candidate OSFb
(mg/kg-d)1
Alveolar/bronchiolar
carcinomas
Multistage 1°
10% ER
76.1
36.2
5.41
0.018
4
5	aBased on allometric scaling of administered RDX dose; BMDLio-hed = BMDLio x (BWa1/4/BWh1/4), BWa = 0.035 kg,
6	and BWh = 70 kg.
7	"Slope factor = BMR/BMDLio-hed, where BMR = 0.10 (10% ER).
8	Table D-23. Summary of BMD modeling results for model predictions for
9	alveolar/bronchiolar carcinoma in male B6C3Fi mice exposed to RDX by diet
10	for 2 years (Lish et al.. 1984): BMR = 10% ER
Model3
Goodness of fit
BMDioPct
(mg/kg-d)
BMDLioPct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage lob
Multistage 2°
Multistage 3°
Multistage 4°
0.561
162.00
76.1
36.2
All of the models reduced to the
multistage 1° model, so this model was
selected.
11
12	aSelected model in bold; scaled residuals for selected model for doses 0,1.5, 7, 35, and 107 mg/kg-day were -0.52,
13	1.08, -0.74, 0.27, and -0.1, respectively.
14	bFor the Multistage 2°, 3°, and 4° models, the b2, b3, and b4 coefficient estimates were 0 (boundary of parameter
15	space). The models in this row reduced to the Multistage 1° model.
16
This document is a draft for review purposes only and does not constitute Agency policy.
D-46	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Multistage Cancer Model, with BMR of 10% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
Multistage Cancer
Linear extrapolation
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0
20
40
60
100
Figure D-17. Plot of incidence rate by dose with fitted curve for Multistage-
Cancer 1° model for Model predictions for alveolar/bronchiolar carcinoma in
male B6C3Fi mice exposed to RDX by diet for 24 months fLish et al.. 19841.
Multistage Model (Version: 3.4; Date: 05/02/2014)
The form of the probability function is: P [response] = background +
(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
The parameter betas are restricted to be positive
Benchmark Dose Computation
BMR = 10% Extra risk
BMD = 76.1197
BMDL at the 95% confidence level = 36.2316
BMDU at the 95% confidence level = 27443100
Taken together, (36.2316, 27443100) is a 90% two-sided confidence interval for the BMD
Multistage Cancer Slope Factor = 0.00276003
Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.0635642
0.0652915
Beta(l)
0.00138414
0.00131052
This document is a draft for review purposes only and does not constitute Agency policy.
D-47	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-78
5



Fitted model
-79
2
1.99294
3
0.57
Reduced model
-81.08
1
6.15622
4
0.19
AIC: = 162.001
Goodness-of-Fit Table
Dose
Est. Prob.
Expected
Observed
Size
Scaled residuals
0
0.0636
4.005
3
63
-0.52
1.5
0.0655
3.93
6
60
1.08
7
0.0726
4.501
3
62
-0.74
35
0.1078
6.363
7
59
0.27
107
0.1925
5.197
5
27
-0.1
ChiA2 = 2.06 d.f. = 3 p-value = 0.561
For male rats in Levine etal. (1983). the high-dose group had a markedly lower survival
curve than the other dose groups, indicating a substantial number of early deaths in the high-dose
group. A log-rank test on the Kaplan-Meier survival curves, stratified by dose, yielded a significant
result (p-value <0.001), in which case a time-to-tumor analysis is generally preferred. However,
such an analysis was not possible because the data were insufficient to allow this analysis.
Although tumor incidence was listed for each animal in the source article, the pathology report
used a different animal numbering system than the experimental report where the times of death
were listed, and the relationship between the two systems was not documented. This prohibited
the matching of the times of death and the tumor incidence of the animals, thus prohibiting the use
of a time-to-tumor analysis.
Therefore, a non-time-dependent dose-response analysis was conducted using BMDS
multistage-cancer models. The model selection procedures described in Section D.2.1 were used to
select the appropriate models. To account for the difference in the survival curves across the
groups for rats, the number of animals alive at 12 months was used as the denominator in the
analysis (denominators listed in Table D-21). Because the maximum liver tumor response in the
male rat was 6.4%, a BMR of 5% was used to model male rat liver tumor data in order to obtain a
BMD and BMDL in the range of the experimental data, as recommended in Section 3.2 of Guidelines
for Carcinogen Risk Assessment fU.S. EPA. 2005al.
To estimate the HED at the BMDL, HEDs based on both administered dose scaled by BW3/4
and physiologically based pharmacokinetic (PBPK) modeling were considered. Confidence in the
revised rat PBPK model is relatively high (see Appendix C, Section C.1.5); however, the choice of an
This document is a draft for review purposes only and does not constitute Agency policy.
D-48	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	internal dose is not straightforward. First, evidence regarding the involvement of metabolites has
2	been discussed in the literature only in the context of the mouse, and the rate of metabolism
3	(allometrically adjusted) appears to be qualitatively slower for the rat Second, metabolism in the
4	model represents the total of all pathways, whereas it is only the minor N-nitroso metabolic route,
5	and not the oxidative route that has been proposed as a factor in RDX-induced mouse
6	carcinogenicity. Third, while blood concentration of RDX as an internal dose would be more
7	proximally relevant to the tissue than administered dose, there are no data to indicate that the
8	parent RDX is directly related to its carcinogenicity. Therefore, given the uncertainties, HEDs based
9	on both administered dose scaled by BW3/4 and area under the curve (AUC) of RDX arterial blood
10	concentration (calculated using the PBPK model) are presented. Extrapolation based on the
11	internal dose of the parent compound is accomplished by assuming toxicological equivalence when
12	dose is expressed in terms of the AUC of the RDX blood concentration.
13	The POD estimates for rat liver carcinomas and the OSFs calculated from these PODs are
14	provided in Table D-24; detailed BMD modeling results are provided in Table D-25 (and
15	Figure D-18). Results based on two dose-metrics are presented: administered dose of RDX scaled
16	by BW3/4 (when dose is expressed in terms of mg/kg-day, this entails scaling by BW"1/4) and AUC of
17	RDX arterial blood concentration (using PBPK modeling).
18	Table D-24. Model predictions and oral slope factor for hepatocellular
19	carcinomas in male F344 rats administered RDX in the diet for 2 years (Levine
20	et al.. 19831
Tumor type
Selected
model
BMR
BMD,
mg/kg-d
BMDL,
mg/kg-d
POD =
BMDLos-hed,
mg/kg-d
Candidate OSFa
(mg/kg-d)"1
Hepatocellular carcinomas
Multistage 1°
5% ER
28.5
11.8
2.88b, 5.75°
0.017b, 0.009°
21
22	aSlope factor = BMR/BMDLos-hed, where BMR = 0.05 (5% ER).
23	bBased on allometric scaling of administered RDX dose; BMDLos-hed = BMDLos x (BWa1/4/BWh1/4), BWa = 0.25 kg, and
24	BWh = 70 kg.
25	cBased on toxicological equivalence of PBPK model derived AUC of RDX blood concentration.
This document is a draft for review purposes only and does not constitute Agency policy.
D-49	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Table D-25. Model predictions for combined hepatocellular adenoma and
2	carcinoma in F344 rats exposed to RDX by diet for 24 months (Levine et al..
3	1983); BMR = 5% ER
Model3
Goodness of fit
BMDspct
(mg/kg-d)
BMDLspct
(mg/kg-d)
Basis for model selection
p-value
AIC
Multistage l°b
Multistage 2°
Multistage 3°
Multistage 4°
0.493
49.095
28.5
11.8
All of the models reduced to the
Multistage 1° model, so this model
was selected.
4
5	aSelected model in bold. Scaled residuals for the selected model for doses 0, 0.3,1.5, 8, and 40 mg/kg-day were
6	0.89, -0.67, -0.74, 0.74, and -0.26, respectively.
7	bFor the Multistage 2°, 3°, and 4° models, the b2, b3, and b4 coefficient estimates were 0 (boundary of parameter
8	space). The models in this row reduced to the Multistage 1° model.
9
Multistage Cancer Model, with BMR of 5% Extra Risk for the BMD arid 0.95 Lower Confidence Limit for tl*
0.25
Multistage Cancer
Linear extrapolation
0.2
0.15
0.1
0.05
o
BMD
BMDI
O
5
10
15
20
25
30
35
40
O	5	10	15	20	25	30	35	40
dose
10	17:39 07/29 2014
11	Figure D-18. Plot of incidence rate by dose, with fitted curve for selected
12	model, for combined hepatocellular adenoma and carcinoma in F344 rats
13	exposed to RDX by diet for 24 months (Levine etal.. 19831.
14	Multistage Model (Version: 3.4; Date: 05/02/2014)
15	The form of the probability function is: P [response] = background +
16	(l-background)*[l-EXP(-betal*doseAl-beta2*doseA2...)]
17	The parameter betas are restricted to be positive
18
This document is a draft for review purposes only and does not constitute Agency policy.
D-50	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	Benchmark Dose Computation
2	BMR = 5% ER
3	BMD = 28.4525
4	BMDL atthe 95% confidence level = 11.8487
5	BMDU atthe 95% confidence level = 235.886
6	Taken together, (11.8487, 235.886) is a 90% two-sided confidence interval for the BMD
7	Multistage Cancer Slope Factor = 0.00421987
8
9	Parameter Estimates
Variable
Estimate
Default initial parameter values
Background
0.00766363
0.00949438
Beta(l)
0.00180277
0.00149364
10
11	Analysis of Deviance Table
Model
Log (likelihood)
Number of parameters
Deviance
Test d.f.
p-value
Full model
-21.0055
5



Fitted model
-22.5473
2
3.08372
3
0.3789
Reduced model
-24.4692
1
6.92747
4
0.1398
12
13	AIC:= 49.0947
14
15	Goodness-of-Fit Table
Dose
Est. prob.
Expected
Observed
Size
Scaled residuals
0
0.0077
0.421
l
55
0.894
0.3
0.0082
0.451
0
55
-0.674
1.5
0.0103
0.538
0
52
-0.737
8
0.0219
1.203
2
55
0.735
40
0.0767
2.378
2
31
-0.255
16
17	ChiA2 = 2.4 d.f. = 3 p-value = 0.493
18
19
This document is a draft for review purposes only and does not constitute Agency policy.
D-51	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
APPENDIX E. SUMMARY OF PUBLIC COMMENTS
AND EPA's DISPOSITION
The Toxicological Review of Hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) was released for
a 60-day public comment period on March 10, 2016. Public comments on the assessment were
submitted to the U.S. Environmental Protection Agency (EPA) by the U.S. Army Public Health
Command and Uniformed Services University of the Health Sciences (posted May 5, 2016), Johns
Hopkins Bloomberg School of Public Health, Special Studies in Risk Assessment class (posted May
11, 2016), Ronald Melnick (posted May 19, 2016), and an anonymous member of the public (posted
May 5, 2016). The anonymous public comment consisted of the word "good," and is not further
discussed in this appendix.
A summary of major public comments provided in these submissions and EPA's response to
these comments are provided in the sections that follow. The comments have been synthesized and
paraphrased, and organized by topic and commenter. Editorial changes and factual corrections
offered by public commenters were incorporated in the document as appropriate and are not
discussed further. All public comments provided were taken into consideration in revising the
draft assessment prior to release for external peer review. The complete set of public comments is
available on the docket at http://www.regulations.gov (Docket ID No. EPA-HQ-ORD-2013-0430).
A public science meeting was held on May 10, 2016 to provide the public an opportunity to
engage in early discussions on the draft Integrated Risk Information System (IRIS) toxicological
review and the draft charge to the peer review panel prior to release for external peer review. The
following three sets of slides were presented at the May 2016 public meeting on RDX and
subsequently submitted to the RDX docket.
•	Andy Nong (Health Canada) provided an overview of the physiologically based
pharmacokinetic (PBPK) modeling of RDX that provided a framework for the discussion of
modeling during the public science meeting. The slides did not provide specific comments
on the IRIS assessment.
•	Nancy Beck (American Chemistry Council) raised several broader programmatic topics
concerning the Preamble and the use of quantitative analyses for chemicals with suggestive
evidence of carcinogenicity. Beck also raised several issues specific to RDX, including
further justification for the choice of a benchmark response (BMR) of 1% for convulsions
and for selection of a gavage study (Crouse etal.. 20061 over a dietary study as the basis for
the reference dose (RfD). Questions related to both issues had been included in the charge
to external peer reviewers.
This document is a draft for review purposes only and does not constitute Agency policy.
E-l	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
• Larry Williams (formerly with U.S. Army Public Health Command) presented a slide titled
"Realities of Human RDX Dose," suggesting that if an individual drank water containing RDX
(at the chemical's water solubility), the toxicity of water would be greater than the toxicity
of RDX. Williams identified the no-observed-adverse-effect level (NOAEL) in rats from the
90-day Grouse etal. f20061 study as 8 mg/kg-day, and estimated the corresponding
equivalent dose in humans as 560 mg/kg-day (or approximately 0.5 g). Using a water
solubility of RDX of 60 mg/L, Williams estimated that a 70-kg human would need to drink
9 L of RDX-saturated water to ingest 0.5 g RDX; he compared this value to a reported LD50
for water of 6 L.
EPA identified the following issues with this analysis:
o The dose of 8 mg/kg-day in the Grouse etal. (2006) study was not, in fact, a NOAEL, but
rather a dose that caused convulsions in 15% of exposed rats and death in 10% of
exposed rats.
o Williams' analysis also failed to consider the statistical uncertainty around a NOAEL. In
a study using 10 animals of each sex per dose group as in the Grouse etal. (2006) study,
the 95% upper confidence limit on an observed response rate of 0% is 31%.
Increasing the sample size to 20 animals per dose group by combining male and female
rats would result in an upper confidence limit of 17%. Therefore, in this case, a
response greater than 0% at the NOAEL is possible, even in the absence of statistical
significance at that dose.
o Williams' estimation of a human equivalent dose (HED) failed to take into consideration
(1) possible differences in susceptibility between rodents and humans, including
allometric scaling between rats and humans to account for species differences in
toxicokinetics, and (2) potentially greater variation in sensitivity to RDX in the human
population than in an in-bred strain of rat.
In light of these omissions and misidentification of the NOAEL for RDX, the conclusion
reached by Williams is not supported.
Comments Related to the Mechanisms by which RDX Induces Seizures
Comment: On behalf of the U.S. Army Public Health Command and Uniformed Services University
of the Health Sciences, Williams and colleagues submitted slides that summarized their research on
the mechanism by which RDX induces seizures, including the measurement of in vivo acetylcholine
and RDX concentrations in blood and brain samples following gavage exposure of rats, receptor
binding assays, and in vitro extracellular and whole cell patch-clamp recordings. The authors
concluded that (1) binding of RDX to the GABAa receptor convulsant site is the primary mechanism
of seizure induction by RDX, (2) reduction of GABAergic inhibitory transmission in the rat
basolateral amygdala is involved in RDX-induced seizures, and (3) the mechanism for RDX-induced
seizures in rats is probably similar to humans. The submission provided no comments on the IRIS
assessment of RDX.
This document is a draft for review purposes only and does not constitute Agency policy.
E-2	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
EPA Response: The submitted slide set summarizes the findings presented in the paper by
Williams etal. (2011) published in Environmental Health Perspectives titled "RDX binds to the
GABA(A) receptor-convulsant site and blocks GABA(A) receptor-mediated currents in the
amygdala: A mechanism for RDX-induced seizures." This paper is cited extensively by EPA in the
discussion of mechanistic evidence for nervous system effects associated with RDX exposure
(Section 1.2.1), and was considered an important primary source of information on the mechanism
by which RDX induces seizures. The discussion of the potential mechanism of RDX-induced
seizures in the RDX assessment is consistent with the Williams etal. f 20 111 paper and with the
slides submitted to the docket
Comments Related to the Cancer Descriptor
Comment: Ronald Melnick commented that the cancer descriptor of suggestive evidence of
carcinogenic potential was not supported, and that RDX clearly met the criteria for likely to be
carcinogenic to humans according to the U.S. EPA (2005a) Guidelines for Carcinogen Risk Assessment
(Cancer Guidelines) because RDX induced dose-related increases in tumors in two species (mouse
and rat), in both sexes, and at two sites (liver and lung).
EPA Response: As noted in the Cancer Guidelines fU.S. EPA. 2005al "[c]hoosing a descriptor is a
matter of judgment and cannot be reduced to a formula. Each descriptor may be applicable to a
wide variety of potential data sets and weights of evidence... Descriptors represent points along a
continuum of evidence; consequently, there are gradations and borderline cases that are clarified
by the full narrative" (p. 2-51).
Interpretation of the evidence of carcinogenicity for RDX is not straight forward, and
arguments for selecting more than one descriptor can be made. Section 1.3.2 of the public comment
draft of the Toxicological Review had already presented the argument supported by Melnick for
likely to be carcinogenic to humans; based on tumor findings in two species, both sexes, and two
sites, as one of two plausible cancer descriptors, along with the argument for suggestive evidence of
carcinogenic potential. The scientific support for the selection of the cancer descriptor for RDX had
already been posed as a charge question to the Science Advisory Board's Chemical Assessment
Advisory Committee (CAAC).
Melnick's assertion that EPA assigns cancer descriptors based on a set of criteria does not
accurately characterize the selection of descriptors as discussed in the Cancer Guidelines. As noted
in Section 2.5 of the Cancer Guidelines (p. 2-53), the bullets included under each cancer descriptor
are examples that are illustrative of the combinations of evidence consistent with each of the five
descriptors. As the Cancer Guidelines note, "[t]he examples are neither a checklist nor a limitation
for the descriptor."
This document is a draft for review purposes only and does not constitute Agency policy.
E-3	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Comment: Melnick supported his criticism of the selection of the suggestive evidence of
carcinogenic potential descriptor with the observations that although the Toxicological Review
identified statistically significant positive trends for hepatocellular adenomas or carcinomas
(combined) in female mice and alveolar/bronchiolar carcinomas in male mice, it was incomplete by
failing to note statistically significant increases in tumor incidence in individual dose groups based
on pair-wise comparisons to the control.
EPA Response: Melnick mischaracterized the role of statistical testing in evaluating evidence of an
association between exposure and tumor response. In general, trend tests are preferred for
evaluating response patterns across dose groups or exposure levels because they are more
powerful in detecting overall dose-response trends than multiple pairwise comparisons to control.
The presence of statistically significant pairwise comparisons to control does not provide
additional information for assessing cancer hazard in this case.
Since trend tests often are not presented in study reports or journal articles (as is the case
for RDX bioassays), EPA calculates trend tests where necessary. In Tables 1-13 and 1-14, the
results of statistical analysis as reported by the authors are provided; EPA did not conduct
statistical analyses using pairwise comparisons where the study authors did not, but did conduct
the more informative trend tests as needed. In Section 1.3.2, EPA's evaluation of the carcinogenicity
evidence for RDX intentionally relied on trend tests over pairwise comparisons. Thus,
consideration of statistical analysis in Section 1.3.2 is not incomplete, but rather provides results of
the more informative statistical tests.
Comment: Melnick also offered the following comments on Section 1.3.2:
•	The incidence of hepatocellular carcinomas in male rats, identified in the Toxicological
Review as showing a statistically significant positive trend, should also have been compared
to historical controls because it is a rare tumor in the F344 rat
•	It is inappropriate to emphasize the number of tumors in male rats in the mid- and high-
dose groups in Levine et al. f 19831 without adjusting for differences in the denominators in
these groups.
•	It is misleading to discuss the lack of tumor findings in the Hart Q9761 study in Sprague-
Dawley rats without discussing the limitations of that study.
EPA Response: EPA agrees with these observations and revised Section 1.3.2 for greater
transparency as follows:
•	Text identifying hepatocellular carcinomas in male F344 rats as rare tumors was added to
Section 1.3.2. Statistical comparison of the incidence of hepatocellular carcinomas in male
rats in the RDX bioassay with historical control data from the National Toxicology Program
had already been presented in Section 1.2.5 of the Toxicological Review.
This document is a draft for review purposes only and does not constitute Agency policy.
E-4	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
•	Text on the number of male rats with liver tumors was revised by providing the total
number of animals examined. EPA agrees with including the denominator so that the tumor
rate can be adjusted for the number of animals examined histopathologically. The incidence
had already been provided in Table 1-13.
•	Limitations of the Hart (1976) study had already been discussed in Section 1.2.5 and
summarized in Section 1.3.2. Discussion of the limitations of the study in Section 1.3.2 was
expanded to include that fact that examination of pathology in treated rats was limited to
the high-dose group.
Comments submitted by the Johns Hopkins Bloomberg School of Public Health, Special
Studies in Risk Assessment class
Comments were developed by the Special Studies in Risk Assessment class at the Johns
Hopkins Bloomberg School of Public Health and submitted by Dr. Mary A. Fox, Assistant Professor
and Acting Director of the Risk Sciences and Public Policy Institute. Selected major comments are
summarized below.
Comment: Planning and scoping for this review were not sufficiently explained. The assessment
should include discussion of why RDX was selected for further study at this time, discussion of the
specific public health concerns related to RDX (e.g., groundwater contamination, ingesting
contaminated food), and whether there is concern that even small exposures to RDX could lead to
illness. The assessment did not include information on past and present production quantities of
RDX, geographic regional areas of use and distribution, summaries of potentially impacted
populations, or demographic information (e.g., socio-economic status) that would assist in the
evaluation of susceptible populations, support cumulative risk assessment activities, and help
determine public health improvement metrics through proposed modification of the RfD.
EPA Response: A planning and scoping step (introduced as part of the July 2013 enhancements to
the IRIS process) was implemented well after the RDX assessment was initiated. Therefore, a
formal planning and scoping step was not conducted for RDX; however, a brief discussion of
environmental releases and occurrence, exposure potential, and regulatory interest that
contributed to the selection of RDX for assessment development were provided in the Preface of the
Toxicological Review. EPA notes that the mission of the IRIS Program is to identify and characterize
the health hazards of chemicals, and that the scope of an IRIS assessment covers the first two steps
of the risk assessment process: hazard identification and dose-response assessment While the
Preface includes discussion of uses and environmental occurrence, in general, exposure
information falls outside the scope of an IRIS assessment.
This document is a draft for review purposes only and does not constitute Agency policy.
E-5	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Comment: More detailed explanation of criteria used for deciding which research to include or
exclude in both hazard identification and selection of studies for dose-response modeling is needed.
Specifically, the comment was offered that three rodent studies fMacPhail etal.. 1985: Cholakis et
al.. 1980: Hart. 19761 were excluded in Section 1.2.1 because they showed no evidence of RDX-
associated neurotoxicity, and explanation was sought for the aspects of these studies that made
them unacceptable.
The Special Studies in Risk Assessment class also noted that three studies were selected as
the basis for calculating candidate reference values for nervous system effects in Section 2.1.4, but
the overall RfD for nervous system effects was based on a value derived from Grouse etal. (2006)
rather than from Cholakis etal. (19801 that resulted in the lowest of the three values. They
questioned that if the results of the Cholakis etal. (1980) study were going to be dismissed, why
this study would have been selected as one of the key studies for RfD derivation purposes. It would
be more appropriate that a study would either be dismissed during the principal study selection
phase, with all of the appropriate evidence and justification, than after going through the motions of
modeling and calculating benchmark doses (BMDs) and RfDs to ultimately dismiss the results
anyway.
EPA Response: The studies by MacPhail etal. (19851. Cholakis etal. f 19801. and Hart fl9761 were
not excluded or determined to be unacceptable studies. Study evaluation was documented in the
Literature Search Strategies | Study Selection and Evaluation section; studies determined to be
uninformative were identified in that step and were not brought forward into hazard identification.
MacPhail etal. (1985). Cholakis etal. (1980). and Hart (1976) were all included in Section 1.2
(Hazard Identification) as informative studies, and were considered in weighing the total
evidence—positive and negative—for nervous system effects as a human hazard of RDX exposure
in Section 1.2.1 (Integration of Nervous System Effects). Unlike the majority of toxicity studies of
RDX, these three studies found no evidence of RDX-associated neurotoxicity. Section 1.2.1 provides
a discussion of possible reasons that could account for a lack of nervous system response in these
studies, but does not dismiss or exclude those findings.
In considering the comment concerned with bringing forward multiple studies for deriving
candidate reference values, EPA points to Section 2.1.1, and particularly Table 2-1, that identifies
factors considered in moving one or more studies forward for dose-response analysis. Among
these are measurement of a representative outcome, reporting of incidence data, multiple dose
groups and observation of a dose-related increase in the outcome, observation of an effect at a
relatively low dose, and consideration of route of administration (e.g., diet or gavage). It is
generally the case that studies considered for dose-response analysis have different strengths and
limitations for estimating dose-response relationships; infrequently can a single "best" study be
identified as the basis for a reference value. In bringing forward multiple studies for dose-response
analysis, study strengths and limitations can be weighed in the context of the candidate values
This document is a draft for review purposes only and does not constitute Agency policy.
E-6	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
derived from each study, and the influence of differences in study design (e.g., dose spacing, route of
administration) and study quality can be examined. Similarities or differences in multiple
candidate values from different studies can provide information on the confidence or uncertainty in
the final reference value. Thus, performing dose-response analysis on datasets from multiple
studies is more than "going through the motions of modeling"; rather, it informs selection of the
overall reference value.
In the case of RDX, three studies of varying study design—Grouse etal. (2006). Cholakis et
al. (1980). and Levine et al. (1983)—were selected for dose-response analysis of nervous system
effects with the above considerations in mind. The rationale for selecting the candidate value based
on Grouse etal. f20061 over candidate values from the other studies is provided in Section 2.1.4.
Comment: The public comments stated that justification of the hazard descriptor of suggestive
evidence of carcinogenic potential was considered clear and sufficient. In another section of the
comment document, however, the Special Studies in Risk Assessment class noted that it was not
immediately clear how EPA applied the Cancer Guidelines, particularly the weight-of-evidence
evaluation. Specifically, more detail was requested on specific factors that would increase or
decrease weight of evidence, including the number of independent studies with consistent results,
multiple observations across species/strain/site, route of administration (including the fact that the
most common route of exposure in humans is inhalation but animal studies used oral
administration; differences between species), and severity and progression.
EPA Response: Several of the factors identified as contributing to the weight-of-evidence
evaluation for carcinogenicity were already considered in Section 1.3.2 (Carcinogenicity). Other
considerations, such as route of administration (only bioassays involving dietary exposure were
available) did not influence the weight of evidence. As discussed in Section 1.3.2, the descriptor
suggestive evidence of carcinogenic potential applies to all routes of human exposure, even where
there is inadequate testing by an exposure route (i.e., inhalation exposure in the case of RDX), in the
absence of convincing evidence to prove otherwise (see Cancer Guidelines, U.S. EPA. 2005a).
As noted in response to comments from Melnick, the charge to external peer reviewers
already included a question as to whether the conclusion regarding weight of evidence for
carcinogenicity was supported.
Comment: The limitations of gavage dosing studies, as well as why those limitations are of less
concern than the limitations from the other studies, should be more clearly stated in Section 2.1.1.
EPA Response: EPA recognizes the influence of method of dosing on the response to RDX in animal
toxicity studies. Considerations related to the selection of a gavage study as the basis for the RfD
over a dietary study were addressed in Sections 1.2.1 and 2.1.7, and this was identified as a key
This document is a draft for review purposes only and does not constitute Agency policy.
E-7	DRAFT-DO NOT CITE OR QUOTE

-------
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
1	issue in the Executive Summary. In addition, a question regarding the appropriateness of selecting
2	Grouse etal. (2006). which used gavage administration, as the basis for the RfD is included in the
3	charge to external peer reviewers.
4
This document is a draft for review purposes only and does not constitute Agency policy.
E-8	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
REFERENCES FOR APPENDICES
Angerhofer. R: Davis. G: Balezewski. L. (1986). Teratological assessment of Trinitro - RDX in rats.
(75-51-0573-86). Aberdeen Proving Ground, MD: U.S. Army Environmental Hygiene
Agency.
Arfsten. DP: Davenport. R: Schaeffer. DT. (1994). Reversion of bioluminescent bacteria (Mutatox
TM) to their luminescent state upon exposure to organic compounds, munitions, and metal
salts. Biomed Environ Sci 7: 144-149.
ATSDR (Agency for Toxic Substances and Disease Registry). (2012). Toxicological profile for RDX
(Update) [ATSDR Tox Profile], (NTIS/10590213). Atlanta, GA: U.S. Department of Health
and Human Services, Public Health Service.
Bannon. PI. (2006). Toxicology study no. 87-XE-059F-06: Biomarkers of RDX exposure in breath of
swine, August - September 2006. Aberdeen Proving Ground, MD: U.S. Army Center for
Health Promotion and Preventive Medicine.
Bannon. PI: Dillman. IF: Hable. MA: Phillips. CS: Perkins. El. (2009a). Global gene expression in rat
brain and liver after oral exposure to the explosive hexahydro-1,3,5-trinitro-l,3,5-triazine
(RPX). Chem Res Toxicol 22: 620-625. http://dx.doi.org/10.1021/tx800444k
Bannon. PI: Pillman. IF: Perkins. El: Bao. W: Wolfinger. RP: Chu. T: Phillips. CS. (2006). Acute RPX
exposure and gene expression in the rat brain [Abstract], Oxford, United Kingdom: Oxford
University Press. http: //www.toxic0l0gy.0rg/pubs/d0cs/Tox/2006Tox.pdf
Bannon, PI; Johnson, M; Williams, L; Adams, V; Perkins, E; Gust, K; Gong, P. (2009b). RPX and
miRNA expression in B6C3F1 mice [Letter], Environ Health Perspect 117: A98.
http://dx.doi.org/10.1289/ehp.0800276
Barsotti. M: Crotti. G. (1949). [Attacchi epileptici come manifestazione di intossicazione
professionale da trimetilen-trinitroamina (T4)]. Med Lav 40: 107-112.
Bhushan. B: Trott. S: Spain. TC: Halasz. A: Paquet. L: Hawari. I. (2003). Biotransformation of
hexahydro-1,3,5-trinitro-l,3,5-triazine (RPX) by a rabbit liver cytochrome P450: Insight
into the mechanism of RPX biodegradation by Rhodococcus sp. strain PN22. Appl Environ
Microbiol 69: 1347-1351. http://dx.doi.Org/10.1128/AEM.69.3.1347-1351.2003
Brown. RP: Pelp. MP: Lindstedt. SL: Rhomberg. LR: Beliles. RP. (1997). Physiological parameter
values for physiologically based pharmacokinetic models [Review], Toxicol Ind Health 13:
407-484. http://dx.doi.Org/10.1177/074823379701300401
Cao. CI: Reddv. G: Bannon. PI: Tohnson. MS. (2008). In vitro study of hexahydro-1,3,5-trinitro-l,3,5-
triazine (RPX) metabolism in human liver. Aberdeen Proving Ground, MP: U.S. Army Center
for Health Promotion and Preventive Medicine.
Cholakis. TM: Wong. LCK: Van Goethem. PL: Minor. I: Short. R: Sprinz. H: Ellis. HV. III. (1980).
Mammalian toxicological evaluation of RPX. (PAMP17-78-C-8027). Kansas City, MO:
Midwest Research Institute.
Cotruvo. TA: Simmon. VF: Spanggord. RT. (1977). Investigation of mutagenic effects of products of
ozonation reactions in water. Ann N Y Acad Sci 298: 124-140.
http://dx.doi.Org/10.llll/i.1749-6632.1977.tbl9259.x
Crouse. LCB: Michie. MW: Major. M: Tohnson. MS: Lee. RB: Paulus. HI. (2006). Subchronic oral
toxicity of RPX in rats. (Toxicology Study No. 85-XC-5131-03). Aberdeen Proving Ground,
MP: U.S. Army Center for Health Promotion and Preventive Medicine.
This document is a draft for review purposes only and does not constitute Agency policy.
R-l	PRAFT-PO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Grouse. LCB: Michie. MW: Major. MA: Leach. GT: Reddv. G. (2008). Oral bioavailability of
cyclotrimethylenetrinitramine (RPX) from contaminated site soils in rats. Int J Toxicol 27:
317-322. http://dx.doi.Org/10.1080/10915810802366885
Davies. TOT: Roberts. DM: Hittarage. A: Buckley. NA. (2007). Oral C-4 plastic explosive in humans - A
case series. Clin Toxicol 45: 454-457. http://dx.doi.org/10.1080/15563650601118044
Dillev. IV: Tyson. GA: Newell. GW. (1979). Mammalian toxicological evaluation of TNT wastewaters.
Volume II. Acute and subacute mammalian toxicity of TNT and the LAP mixture.
(ADA080957. DAMD17-76-C-6050.). Menlo Park, CA: SRI International.
http://oai.dtic.mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADA080957
Fellows. RT: Driver. CR: Cataldo. DA: Harvey. SD. (2006). Bioavailability of hexahydro-1,3,5-trinitro-
1,3,5-triazine (RDX) to the prairie vole (Microtus ochrogaster). Environ Toxicol Chem 25:
1881-1886. http://dx.doi.Org/lQ.1897/Q5-446Rl.l
George. SE: Huggins-Clark. G: Brooks. LR. (2001). Use of a Salmonella microsuspension bioassay to
detect the mutagenicity of munitions compounds at low concentrations. MutatRes 490: 45-
56. http://dx.doi.org/lQ.1016/S1383-5718rQ01Q015Q-9
Goldberg. PI: Green. ST: Nathwani. D: McMenamin. T: Hamlet. N: Kennedy. PH. (1992). RPX
intoxication causing seizures and a widespread petechial rash mimicking
meningococcaemia. J R Soc Med 85: 181.
Guo. L: Xu. H: Chen. Y: Chang. Y. (1985). Pistribution and metabolism of tritium-labeled hexogen in
white mice. 3:335-339.
Harrell-Bruder. B: Hutchins. KL. (1995). Seizures caused by ingestion of composition C-4. Ann
EmergMed 26: 746-748. http://dx.doi.org/10.1016/S0196-0644f95170Q49-8
Hart. ER. (1974). Subacute toxicity of RPX and TNT in dogs. Final report. (A717530). Kensington,
MP: Litton Bionetics, Inc.
https://ntrl.ntis.gov/NTRL/dashboard/searchResul ts.xhtml?searchQuery=APA03 5717
Hart. ER. (1976). Two-year chronic toxicity study in rats. (N00014-73-C-0162). Kensington, MP:
Litton Bionetics, Inc.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchQuery=APA040161
Hathawav. IA: Buck. CR. (1977). Absence of health hazards associated with RPX manufacture and
use. J Occup Med 19: 269-272.
Hess-Ruth. A: Crouse. L: Roszell. L. (2007). RPX pilot development neurotoxicity test in rats.
(Toxicology Study No. 85-XC-064Y-07). Aberdeen Proving Ground: U.S. Army Center for
Health Promotion and Preventive Medicine.
Hett. PA: Fichtner. K. (2002). A plastic explosive by mouth. J R Soc Med 95: 251-252.
http://dx.doi.Org/10.1258/irsm.95.5.251
Hollander. AT: Colbach. EM. (1969). Composition C-4 induced seizures: A report of five cases. Mil
Med 134: 1529-1530.
lohnson. MS. (2015). Memorandum for National Center for Environmental Assessment (8101R/Pr.
Louis P'Amico), U.S. EPA- Office ofResearch and Pevelopment, 1200 Pennsylvania Ave.,
NW, Washington, PC 20460. Subject: Additional data from the oral subchronic toxicity of
RPX in rats, 2006. (Toxicology study no. 85-XC-5131-03), U.S. Army Center for Health
Promotion and Preventative Medicine, Aberdeen Proving Ground, Maryland; Incidence of
seizure relative to mortality events. (MCHB-IP-T). Aberdeen Proving Ground, MP:
Pepartment of the Army, US Army Institute of Public Health.
Kaplan. AS: Berghout. CF: Peczenik. A. (1965). Human intoxication from RPX. Arch Environ Health
10: 877-883.
Kasuske. L: Schofer. IM: Hasegawa. K. (2009). Two marines with generalized seizure activity. J
Emerg Nurs 35: 542-543. http://dx.doi.Org/10.1016/i.ien.2008.05.001
Ketel. WB: Hughes. IR. (1972). Toxic encephalopathy with seizures secondary to ingestion of
composition C-4. A clinical and electroencephalographic study. Neurology 22: 871-876.
This document is a draft for review purposes only and does not constitute Agency policy.
R-2	PRAFT-PO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Knepshield. IH: Stone. WI. (1972). Toxic effects following ingestion of C-4 plastic explosive. In W
Keup (Ed.), Drug Abuse: Current Concepts and Research (pp. 296-300). Springfield, IL:
Charles C. Thomas.
Krishnan. K: Crouse. LCB: Bazar. MA: Major. MA: Reddv. G. (2009). Physiologically based
pharmacokinetic modeling of cyclotrimethylenetrinitramine in male rats. J Appl Toxicol 29:
629-637. http://dx.doi.org/10.1002/iatl455
Kiiciikardali. Y: Acar. HV: Ozkan. S: Nalbant. S: Yazgan. Y: Atasovu. EM: Keskin. 0: Naz. A: Akvatan. N:
Gokben. M: Danaci. M. (2003). Accidental oral poisoning caused by RDX (cyclonite): A report
of 5 cases. J Intensive Care Med 18: 42-46. http://dx.doi.org/10.1177/0885066602239123
Lachance. B: Robidoux. PY: Hawari. I: Ampleman. G: Thiboutot. S: Sunahara. GI. (1999). Cytotoxic
and genotoxic effects of energetic compounds on bacterial and mammalian cells in vitro.
MutatRes 444: 25-39. http://dx.doi.org/10.1016/S1383-5718f99100073-X
Levin. DE: Hollstein. M: Christman. MF: Schwiers. EA: Ames. BN. (1982). A new Salmonella tester
strain (TA102) with AT base pairs at the site of mutation detects oxidative mutagens. Proc
Natl Acad Sci USA 79: 7445-7449. http://dx.doi.org/10.lQ73/pnas.79.23.7445
Levine. BS: Furedi. EM: Gordon. DE: Barklev. II: Lish. PM. (1990). Toxic interactions of the munitions
compounds TNT and RDX in F344 rats. Fundam Appl Toxicol 15: 373-380.
http:/ /dx.doi.org/10.1016/0272-0590f90190062-Q
Levine. BS: Furedi. EM: Gordon. DE: Burns. IM: Lish. PM. (1981a). Thirteen week oral (diet) toxicity
study of trinitrotoluene (TNT), hexahydro-1, 3, 5-trinitro-l, 3, 5-triazine (RDX) and
TNT/RDX mixtures in the Fischer 344 rat. Final report (ADA108447. DAMD17-79-C-9120.
DAMD17-79-C-9161). Chicago, IL: IIT Research Institute.
http://oai.dtic. mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADAl 08447
Levine. BS: Furedi. EM: Gordon. DE: Burns. IM: Lish. PM. (1981b). Thirteen week toxicity study of
hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) in Fischer 344 rats. Toxicol Lett 8: 241-245.
http://dx.doi.Org/10.1016/0378-4274f81190108-9
Levine. BS: Lish. PM: Furedi. EM: Rac. VS: Sagartz. IM. (1983). Determination of the chronic
mammalian toxicological effects of RDX (twenty-four month chronic
toxicity/carcinogenicity study of hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) in the
Fischer 344 rat): Final report-phase V. Chicago, IL: IIT Research Institute.
Lipscomb. IC: Poet. TS. (2008). In vitro measurements of metabolism for application in
pharmacokinetic modeling [Review], Pharmacol Ther 118: 82-103.
http://dx.doi.Org/10.1016/i.pharmthera.2008.01.006
Lish. PM: Levine. BS: Furedi. EM: Sagartz. IM: Rac. VS. (1984). Determination of the chronic
mammalian toxicological effects of RDX: Twenty-four month chronic
toxicity/carcinogenicity study of hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) in the
B6C3F1 hybrid mouse (Volumesl-3). (ADA181766. DAMD17-79-C-9161). FortDetrick,
Frederick, MD: U.S. Army Medical Research and Development Command,
http://oai.dtic. mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADAl 81766
Ma. B: Li. H. (1993). Neurobehavioral effects of hexogen on exposed workers. Gongye Weishengyu
Zhiyebin 19: 20-23.
MacPhail. RC: Walker. OP: Cook. L. (1985). Neurotoxicology of cyclotrimethy 1 enetrinitramine
(RDX). Final report (ADA168266). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Health Effects Research Laboratory, Neurotoxicology Division.
https://ntrl.ntis.gov/NTRL/dashboard/searchResul ts.xhtml?searchQuery=ADAl 68266
Major. MA: Reddv. G: Berge. MA: Patzer. SS: Li. AC: Gohdes. M. (2007). Metabolite profiling of
[14C]hexahydro-l,3,5-trinitro-l,3,5-triazine (RDX) in Yucatan miniature pigs. J Toxicol
Environ Health A 70: 1191-1202. http://dx.doi.org/10.1080/15287390701252717
This document is a draft for review purposes only and does not constitute Agency policy.
R-3	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Martin. D: Hart. E. (19741. Subacute toxicity of RDX and TNT in monkeys (pp. 1-216). (ADA044650).
Kensington, MD: Litton Bionetics, Inc.
https://ntrl.ntis.gov/NTRL/dashboard/searchResul ts.xhtml?searchQuery=ADA044650
Merrill. SL. (1968). Ingestion of an explosive material, composition C-4: A report of two cases.
USARVMed Bull 40: 5-11.
Musick. TI: Berge. MA: Patzer. SS: Tilch. KR. (2010). Absorption, distribution, metabolism, and
excretion of 14C-RDX following oral administration to minipigs (DAAD05-02-P-2319).
(ADA526472). Madison, WI: Covance Laboratories Inc. http://www.dtic.mil/cgi-
bin/GetTRDoc?AD=ADA526472
Neuwoehner. I: Schofer. A: Erlenkaemper. B: Steinbach. K: Hund-Rinke. K: Eisentraeger. A. (2007).
Toxicological characterization of 2,4,6-trinitrotoluene, its transformation products, and two
nitramine explosives. Environ Toxicol Chem 26: 1090-1099. http://dx.doi.org/10.1897/06-
471R.1
NIOSH (National Institute for Occupational Safety and Health). (2012). NIOSH pocket guide to
chemical hazards: Cyclonite, CAS 121-82-4. Cincinnati, OH: U.S. Department of Health and
Human Services, Centers for Disease Control and Prevention, National Institute for
Occupational Safety and Health, http: / /www,cdc.gov/niosh/npg/npgdOl69.html
OSHA (Occupational Safety & Health Administration). (2012a). Air contaminants. Table Z-
Shipyards. CFR 29 CFR1915, Subpt. P, App A.
OSHA (Occupational Safety & Health Administration). (2012b). Gases, vapors, fumes, dusts, and
mists. Appendix A to 1926.55-1970 American Conference of Governmental Industrial
Hygienists' threshold limit values of airborne contaminants. CFR 29 CFR 1926.55.
Ozhan. G: Topuz. S: Alpertunga. B. (2003). Determination of cyclonite (RDX) in human plasma by
high-performance liquid chromatography. Farmaco 58: 445-448.
http://dx.doi.Org/10.1016/S0014-827Xf03100069-7
Pan. X: Ochoa. KM: San Francisco. MI: Cox. SB: Dixon. K: Anderson. TA: Cobb. GP. (2013).
Absorption, distribution, and biotransformation of hexahydro-1,3,5-trinitro-l,3,5-triazine
(RDX) in B6C3F1 mice (Mus musculus). Environ Toxicol Chem 32: 1295-1303.
http://dx.doi.org/10.1002/etc.2188
Pan. X: San Francisco. MI: Lee. C: Ochoa. KM: Xu. X: Liu. I: Zhang. B: Cox. SB: Cobb. GP. (2007a).
Examination of the mutagenicity of RDX and its N-nitroso metabolites using the Salmonella
reverse mutation assay. MutatRes Genet Toxicol Environ Mutagen 629: 64-69.
http://dx.doi.Org/10.1016/i.mrgentox.2007.01.006
Pan. X: Zhang. B: Smith. IN: San Francisco. M: Anderson. TA: Cobb. GP. (2007b). N-Nitroso
compounds produced in deer mouse (Peromyscus maniculatus) GI tracts following
hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) exposure. Chemosphere 67: 1164-1170.
http://dx.doi.Org/10.1016/i.chemosphere.2006.10.077
Paquet. L: Monteil-Rivera. F: Hatzinger. PB: Fuller. ME: Hawari. I. (2011). Analysis of the key
intermediates of RDX (hexahydro-1,3,5-trinitro-l,3,5-triazine) in groundwater: Occurrence,
stability and preservation. J Environ Monit 13: 2304-2311.
h ttp: / /dx. do i. o r g /10.10 3 9 /c 1 e m 10 3 2 9 f
Parker. GA: Reddv. G: Major. MA. (2006). Reevaluation of a twenty- four- month chronic
toxicity/carcinogenicity study of hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX) in the
B6C3F1 hybrid mouse. Int J Toxicol 25: 373-378.
http:/ /dx.doi.org/10.1080/10915810600846245
Pou 1 in. P: Krishnan. K. (1995). An algorithm for predicting tissue : blood partition coefficients of
organic chemicals from n-octanol: water partition coefficient data. J Toxicol Environ Health
46: 117-129. http://dx.doi.org/10.1080/15287399509532021
Poulin. P: Theil. F. -P. (2000). A priori prediction of tissue : plasma partition coefficients of drugs to
facilitate the use of physiologically-based pharmacokinetic models in drug discovery. J
This document is a draft for review purposes only and does not constitute Agency policy.
R-4	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Pharm Sci 89: 16-35. http://dx.doi.org/10.1002/fSICI11520-6017f200001189:l<:16::AID-
IPS3>:3.0.CQ:2-E
Reddv. G: Allen. NA: Major. MA. (2008). Absorption of 14C-cyclotrimethylenetrinitramine (RDX)
from soils through excised human skin. Toxicol Mech Meth 18: 575-579.
http://dx.doi.Org/10.1080/15376510701703466
Reddv. G: Eisenhut. K: Morrison. SA: Kelly. LA. (1989). Toxicokinetics of 14C RDX
(cyclotrimethylenetrinitramine) in rats after intratracheal administration [Abstract],
Reddv. G: Erexson. GL: Cifone. MA: Major. MA: Leach. GT. (2005). Genotoxicity assessment of
hexahydro-1,3,5-trinitro-l,3,5-triazine (RDX). Int J Toxicol 24: 427-434.
http://dx.doi. org/10.1080/10915810500366922
Reifenrath. WG: Kammen. HO: Reddv. G: Maior. MA: Leach. GL (2008). Interaction of hydration,
aging, and carbon content of soil on the evaporation and skin bioavailability of munition
contaminants. J Toxicol Environ Health A 71: 486-494.
http://dx.doi.Org/10.1080/15287390801906956
Rispin. A: Farrar. D: Margosches. E: Gupta. K: Stitzel. K: Carr. G: Greene. M: Meyer. W: McCall. D.
(2002). Alternative methods for the median lethal dose (LD(50)) test: The up-and-down
procedure for acute oral toxicity. ILAR J 43: 233-243.
http://dx.doi.Org/10.1093/ilar.43.4.233
Safe Work Australia. (2014). Hazardous substances information system (HSIS). Exposure standard
documentation: Cyclonite [Database], Retrieved from
http://www.hsis.safeworkaiistralia.gov.aii/ExposureStandards/Document7exposiireStanda
rdlD=l 74
Schneider. NR: Bradley. SL: Andersen. ME. (1977). Toxicology of cyclotrimethylenetrinitramine:
Distribution and metabolism in the rat and the miniature swine. Toxicol Appl Pharmacol 39:
531-541. http://dx.doi.Org/10.1016/0041-008Xf77190144-2
Schneider. NR: Bradley. SL: Andersen. ME. (1978). The distribution and metabolism of
cyclotrimethylenetrinitramine (RDX) in the rat after subchronic administration. Toxicol
Appl Pharmacol 46: 163-171. http://dx.doi.Org/10.1016/0041-008Xf78190147-3
Simmon. VF: Spanggord. RL Eckford. SL: McClurg. V. (1977). Mutagenicity of some munition
wastewater chemicals and chlorine test kit reagents (pp. 1-205). (ADA057680. DAMD17-
76-C-6013). Menlo Park, CA: SRI International.
http://oai.dtic. mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADA05 7680
Smith. IN: Espino. MA: Liu. I: Romero. NA: Cox. SB: Cobb. GP. (2009). Multigenerational effects in
deer mice (Peromyscus maniculatus) exposed to hexahydro-1,3,5-trinitroso-l,3,5-triazine
(TNX). Chemosphere 75: 910-914. http://dx.doi.Org/10.1016/i.chemosphere.2009.01.010
Snodgrass. HL. Ir. (1984). Preliminary assessment of relative toxicity and mutagenicity potential of
l-nitroso-3,5-dinitro-l,3,5-triazacylcohexane (mononitroso-RDX) (Final Reported.).
(ADA149351. USAEHA-75-51-0345-85). Aberdeen Proving Ground, MD: U.S. Army
Environmental Hygiene Agency.
http://oai.dtic. mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=ADA149351
Snoeck. V: Huvghebaert. N: Cox. E: Vermeire. A: Saunders. I: Remon. IP: Verschooten. F: Goddeeris.
BM. (2004). Gastrointestinal transit time of nondisintegrating radio-opaque pellets in
suckling and recently weaned piglets. J Control Release 94: 143-153.
http://dx.doi.Org/10.1016/i.iconrel.2003.09.015
Spiteri. NI. (1982). Circadian patterning of feeding, drinking and activity during diurnal food access
in rats. Physiol Behav 28: 139-147. http://dx.doi.org/10.1016/0031-9384f82190115-9
Stone. WI: Paletta. TL: Heiman. EM: Bruce. II: Knepshield. IH. (1969). Toxic effects following
ingestion of C-4 plastic explosive. Arch Intern Med 124: 726-730.
http://dx.doi.Org/10.1001/archinte.1969.00300220078015
This document is a draft for review purposes only and does not constitute Agency policy.
R-5	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Sweeney. LM: Gut. CP. Ir: Gargas. ML: Reddv. G: Williams. LR: lohnson. MS. (2012a). Assessing the
non-cancer risk for RDX (hexahydro-1,3,5-trinitro-l,3,5-triazine) using physiologically
based pharmacokinetic (PBPK) modeling [Review], Regul Toxicol Pharmacol 62: 107-114.
http://dx.doi.Org/10.1016/i.vrtph.20ri.12.007
Sweeney. LM: Okolica. MR: Gut. CP. Ir: Gargas. ML. (2012b). Cancer mode of action, weight of
evidence, and proposed cancer reference value for hexahydro-1,3,5-trinitro-l,3,5-triazine
(RDX). Regul Toxicol Pharmacol 64: 205-224.
http://dx.doi.Org/10.1016/i.vrtph.2012.07.005
Tan. EL: Ho. CH: Griest. WH: Tvndall. RL. (1992). Mutagenicity of trinitrotoluene and its metabolites
formed during composting. J Toxicol Environ Health A 36: 165-175.
http://dx.doi.org/10.1080/15287399209531632
Testud. F: Glanclaude. 1. -M: Descotes. 1. (1996a). Acute hexogen poisoning after occupational
exposure. J Toxicol Clin Toxicol 34: 109-111.
http://dx.doi.org/10.3109/15563659609020244
Testud. F: Glanclaude. 1: Imperatori. 1: Le Meur. B: Descotes. 1. (1996b). [Acute hexogen poisoning
after occupational exposure, report of 2 cases]. Arch Mai Prof Med Trav Secur Soc 57: 342-
346.
Timchalk. C: Nolan. RI: Mendrala. AL: Dittenber. DA: Brzak. KA: Mattsson. IL. (2002). A
physiologically based pharmacokinetic and pharmacodynamic (PBPK/PD) model for the
organophosphate insecticide chlorpyrifos in rats and humans. Toxicol Sci 66: 34-53.
http://dx.doi.Org/10.1093/toxsci/66.l.34
U.S. EPA (U.S. Environmental Protection Agency). (1992). A cross-species scaling factor for
carcinogen risk assessment based on equivalence of mg/kg3/4/day [EPA Report] (pp.
24152-24161). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines for carcinogen risk
assessment [EPAReport] (pp. 1-166). (EPA/630/P-03/001F). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
http://www2.epa.gov/osa/guidelines-carcinogen-risk-assessment
U.S. EPA (U.S. Environmental Protection Agency). (2005b). Supplemental guidance for assessing
susceptibility from early-life exposure to carcinogens (pp. 1-125). (EPA/630/R-03/003F).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
U.S. EPA (U.S. Environmental Protection Agency). (2012a). Advances in inhalation gas dosimetry for
derivation of a reference concentration (RfC) and use in risk assessment (pp. 1-140).
(EPA/600/R-12/044). Washington, DC.
https://cfpub.epa. gov/ncea/risk/recordisplay.cfm?deid=244650&CFID=50524762&CFTOK
EN=17139189
U.S. EPA (U.S. Environmental Protection Agency). (2012b). Benchmark dose technical guidance (pp.
1-99). (EPA/100/R-12/001). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum, https://www.epa.gov/risk/benchmark-dose-technical-guidance
U.S. EPA (U.S. Environmental Protection Agency). (2012c). EPA Federal facilities forum issue paper:
Site characterization for munitions constituents [EPAReport] (pp. 1-170). (EPA-505-S-11-
001). Washington, D.C. http://www2.epa.gov/fedfac/epa-federal-facilities-forum-issue-
paper-site-characterization-munitions-constituents
U.S. EPA (U.S. Environmental Protection Agency). (2014). RDX PBPK files in acsIX format [EPA
Report],
von Oettingen. WF: Donahue. DP: Yagoda. H: Monaco. AR: Harris. MR. (1949). Toxicity and potential
dangers of cyclotrimethylenetrinitramine (RDX). J Ind Hyg Toxicol 31: 21-31.
West. RR: Stafford. DA. (1997). Occupational exposures and haematological abnormalities among
ordnance factory workers: A case control study. Leuk Res 21: 675-680.
http://dx.doi.Org/10.1016/S0145-2126f97100042-8
This document is a draft for review purposes only and does not constitute Agency policy.
R-6	DRAFT-DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
Supplemental Information— Hexahydro-1,3,5-trinitro-l,3,5-triazine
Whong. W. -Z: Speciner. ND: Edwards. GS. (1980). Mutagenic activity of tetryl, a nitroaromatic
explosive, in three microbial test systems. Toxicol Lett 5: 11-17.
http://dx.doi.Org/10.1016/0378-4274r80190142-3
Williams. LR: Aroniadou-Anderiaska. V: Oashu. F: Finne. H: Pidoplichko. V: Bannon. PI: Braga. MFM.
(2011). RDXbinds to the GABA(A) receptor-convulsantsite and blocks GABA(A) receptor-
mediated currents in the amygdala: A mechanism for RDX-induced seizures. Environ Health
Perspect 119: 357-363. http://dx.doi.org/10.1289/ehp.1002588
Woody. RC: Kearns. GL: Brewster. MA: Turlev. CP: Sharp. GB: Lake. RS. (1986). The neurotoxicity of
cyclotrimethylenetrinitramine (RDX) in a child: A clinical and pharmacokinetic evaluation.
Clin Toxicol 24: 305-319. http://dx.doi.org/10.3109/15563658608992595
This document is a draft for review purposes only and does not constitute Agency policy.
R-7	DRAFT-DO NOT CITE OR QUOTE

-------