jfMk ¦¦¦¦¦Ik Mk

9 PPA
\sc,tr\

EPA/635/R-24/013

IRIS Assessment Protocol

www.epa.gov/iris

Protocol

for the Uranium IRIS Assessment (Oral)
(Preliminary Assessment Materials)

CASRN 7440-61-1

February 2024

Integrated Risk Information System
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC


-------
Protocol for the Uranium IRIS Assessment (Oral)

DISCLAIMER

This document is a public comment draft for review purposes only. This information is
distributed solely for the purpose of public comment. It has not been formally disseminated by EPA.
It does not represent and should not be construed to represent any Agency determination or policy.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.

This document is a draft for review purposes only and does not constitute Agency policy.

ii	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

CONTENTS

AUTHORS | CONTRIBUTORS | REVIEWERS	x

1.	INTRODUCTION	1-1

2.	SCOPING AND INITIAL PROBLEM FORMULATION SUMMARY	2-1

2.1. BACKGROUND	2-1

2.1.1.	Physical and Chemical Properties	2-1

2.1.2.	Sources, Production, and Use	2-2

2.1.3.	Environmental Fate and Transport	2-3

2.1.4.	Potential Human Exposure (Oral)	2-3

2.1.5.	Previous Assessments of Oral Exposure to Uranium by the Environmental

Protection Agency and Other Health Agencies	2-4

2.2.SCOPING SUMMARY	2-8

2.3.	PROBLEM FORMULATION	2-9

2.4.	KEY SCIENCE ISSUE	2-10

3.	OVERALL OBJECTIVES AND SPECIFIC AIMS	3-1

3.1. SPECIFIC AIMS	3-1

4.	LITERATURE SEARCH AND SCREENING STRATEGIES	4-1

4.1.	USE OF EXISTING ASSESSMENTS	4-1

4.2.	POPULATIONS, EXPOSURES, COMPARATORS, AND OUTCOMES CRITERIA FOR THE
SYSTEMATIC EVIDENCE MAP	4-1

4.3.SUPPLEMENTAL CONTENT SCREENING CRITERIA	4-3

4.4.	LITERATURE SEARCH STRATEGIES	4-9

4.4.1.	Database Search Term Development	4-9

4.4.2.	Database Searches	4-9

4.4.3.	Searching Other Sources	4-10

4.4.4.	Non-Peer-Reviewed Data	4-11

4.5.	LITERATURE SCREENING	4-11

4.5.1.	Title and Abstract Screening	4-12

4.5.2.	Full-Text Screening	4-13

4.5.3.	Multiple Publications of the Same Data	4-13

4.5.4.	Literature Flow Diagram	4-13

This document is a draft for review purposes only and does not constitute Agency policy.

iii	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

4.6. LITERATURE INVENTORY	4-15

4.6.1.	Studies That Meet Problem Formulation PECO Criteria	4-15

4.6.2.	Organizational Approach for Supplemental Material	4-16

5.	REFINED PROBLEM FORMULATION AND ASSESSMENT APPROACH	5-1

5.1.COMPARISON	WITH ATSDR TOXICOLOGICAL PROFILE (2013)	5-1

5.2.	REFINEMENTS TO PECO CRITERIA	5-3

5.2.1. Other Exclusions Based on Full-Text Content	5-6

5.3.	UNITS OF ANALYSES FOR DEVELOPING EVIDENCE SYNTHESIS AND INTEGRATION
JUDGMENTS FOR HEALTH EFFECT CATEGORIES	5-6

5.4.	CONSIDERATIONS OF SUPPLEMENTAL MATERIAL	5-8

5.4.1.	Noncancer MOA Mechanistic Information	5-8

5.4.2.	ADME and PK/PBPK Model Information	5-8

5.4.3.	Other Supplemental Material Content	5-9

6.	STUDY EVALUATION (RISK OF BIAS AND SENSITIVITY)	6-1

6.1.STUDY EVALUATION OVERVIEW FOR HEALTH EFFECT STUDIES	6-1

6.2.	EPIDEMIOLOGY STUDY EVALUATION	6-5

6.3.	EXPERIMENTAL ANIMAL STUDY EVALUATION	6-14

6.4.	MECHANISTIC AND OTHER NON-PECO STUDY EVALUATION	6-24

6.5.	PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODEL DESCRIPTIVE SUMMARY

AND EVALUATION	6-24

7.	DATA EXTRACTION OF STUDY METHODS AND RESULTS	7-1

7.1.STANDARDIZING ADMINISTERED DOSE LEVELS/CONCENTRATIONS	7-3

8.	EVIDENCE SYNTHESIS AND INTEGRATION	8-1

8.1.	EVIDENCE SYNTHESIS	8-5

8.2.	EVIDENCE INTEGRATION	8-15

9.	DOSE-RESPONSE ASSESSMENT: STUDY SELECTION AND QUANTITATIVE ANALYSIS	9-1

9.1.OVERVIEW	9-1

9.2.SELECTING STUDIES FOR DOSE-RESPONSE ASSESSMENT	9-2

9.3.CONDUCTING	DOSE-RESPONSE ASSESSMENTS	9-5

9.3.1.	Dose-Response Analysis in the Range of Observation	9-5

9.3.2.	Extrapolation: Reference Values	9-8

REFERENCES	R-l

APPENDIX A. ELECTRONIC DATABASE SEARCH STRATEGIES	A-l

APPENDIX B. SURVEY OF EXISTING TOXICITY VALUES	B-l

This document is a draft for review purposes only and does not constitute Agency policy.

iv	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

APPENDIX C. PROCESS FOR SEARCHING AND COLLECTING EVIDENCE FROM SELECTED OTHER

RESOURCES 	C-l

APPENDIX D. COMPARISON BETWEEN ATSDR 2013 AND IRIS LITERATURE SEARCH

INVENTORY 	D-l

This document is a draft for review purposes only and does not constitute Agency policy.

v	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

TABLES

Table 2-1. Chemical identity and physiochemical properties of selected uranium compounds as

curated by EPA's CompTox Chemicals Dashboard	2-2

Table 2-2. Details on derivation of the available health effect reference values for oral exposure

to uranium3	2-6

Table 2-3. EPA Program and Regional Office interest in an assessment of uranium	2-8

Table 4-1. Problem formulation populations, exposures, comparators, and outcomes criteria

used for the systematic evidence map	4-2

Table 4-2. Categories of potentially relevant supplemental material	4-4

Table 5-1. Health effect categories from ATSDR 2013 (ATSDR, 2013) selected for hazard ID, dose

response, or no further consideration	5-3

Table 5-2. Assessment populations, exposures, comparators, and outcomes criteria for uranium	5-5

Table 5-3. Dose-response: Health effect categories and human and animal evidence unit of

analysis endpoint groupings for dose response	5-7

Table 5-4. Hazard evaluation: Health effect categories and human and animal evidence unit of

analysis endpoint groupings for hazard evaluation	5-8

Table 6-1. Questions to guide the development of criteria for each domain in epidemiology

studies	6-6

Table 6-2. Questions to guide the development of criteria for each domain in experimental

animal toxicology studies	6-15

Table 8-1. Generalized evidence profile table to show the relationship between evidence
synthesis and evidence integration to reach judgment of the evidence for

hazard	8-3

Table 8-2. Generalized evidence profile table to show the key findings and supporting rationale

from mechanistic analyses	8-4

Table 8-3. Considerations that inform evaluations and judgments of the strength of the evidence

for hazard	8-7

Table 8-4. Framework for strength of evidence judgments from studies in humans	8-12

Table 8-5. Framework for strength of evidence judgments from studies in animals	8-13

Table 8-6. Considerations that inform evidence integration judgments	8-16

Table 8-7. Framework for summary evidence integration judgments in the evidence integration

narrative	8-18

Table 9-1. Attributes used to evaluate studies for derivation of toxicity values	9-3

Table A-l. Database search strategy	A-l

Table B-l. Sources searched for existing human health reference values	B-l

Table C-l. Summary table for other sources search results	C-3

Table D-l. Studies of cardiovascular endpoints in humans identified 2011-2021 	D-3

Table D-2. Summary of animal studies reporting on uranium-induced cardiovascular effects	D-5

Table D-3. Studies of developmental endpoints in humans identified 2011-2022 	D-7

Table D-4. Summary of toxicological studies reporting on uranium-induced developmental

effects	D-9

Table D-5. Studies of endocrine endpoints in humans identified 2011-2022 	D-10

Table D-6. Studies of hematological endpoints in humans identified 2011-2022	D-13

Table D-7. Summary of toxicological studies reporting on uranium-induced hepatic effects	D-14

Table D-8. Studies of immunological endpoints in humans identified 2011-2022	D-16

This document is a draft for review purposes only and does not constitute Agency policy.

vi	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table D-9. Summary of toxicological studies reporting on uranium-induced immunological

effects	D-17

Table D-10. Studies of metabolic endpoints in humans identified 2011-2022	D-18

Table D-ll. Studies of musculoskeletal endpoints in humans identified 2011-2022	D-21

Table D-12. Summary of toxicological studies reporting on uranium-induced musculoskeletal

effects	D-21

Table D-13. Studies of neurological endpoints in humans identified 2011-2022	D-23

Table D-14. Summary of toxicological studies reporting on uranium-induced neurological effects	D-23

Table D-15. Studies of reproductive endpoints in humans identified 2011-2022	D-26

Table D-16. Summary of toxicological studies reporting on uranium-induced reproductive effects .... D-27

Table D-17. Studies of respiratory endpoints in humans identified 2011-2022	D-28

Table D-18. Studies of urinary endpoints in humans identified 2011-2022	D-31

Table D-19. Summary of toxicological studies reporting on uranium-induced urinary effects	D-32

FIGURES

Figure 1-1. Integrated Risk Information System systematic review problem formulation and

method documents	1-1

Figure 2-1. Available health effect reference values for oral exposure to uranium (current as of

November 2022)	2-5

Figure 4-1. IRIS literature search flow diagram for uranium	4-14

Figure 4-2. Visual summary of approach for tagging major categories of supplemental material.

See interactive HAWC link: Uranium Literature Tagtree	4-17

Figure 5-1. Approach and decision tree used to compare ATSDR 2013 (ATSDR, 2013) with IRIS

literature search results	5-2

Figure 6-1. Overview of Integrated Risk Information System study evaluation approach.

(a) individual evaluation domains organized by evidence type, and (b) individual

evaluation domains judgments and definitions for overall ratings (i.e., domain

and overall judgments are performed on an outcome-specific basis)	6-2

This document is a draft for review purposes only and does not constitute Agency policy.

vii	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

ABBREVIATIONS

AC50 activity concentration at 50%

AD ME absorption, distribution, metabolism,

and excretion
AIC Akaike's information criterion
ALT alanine aminotransferase
AOP adverse outcome pathway
AST aspartate aminotransferase
atm atmosphere
ATSDR Agency for Toxic Substances and

Disease Registry
BMC benchmark concentration
BMCL benchmark concentration lower

confidence limit
BMD benchmark dose

BMDL benchmark dose lower confidence limit
BMDS Benchmark Dose Software
BMR benchmark response
BUN blood urea nitrogen
BW body weight

BW3/4 body weight scaling to the 3/4 power
CA	chromosomal aberration

CAA Clean Air Act
CAS Chemical Abstracts Service
CASRN Chemical Abstracts Service registry
number

CERCLA Comprehensive Environmental

Response, Compensation, and Liability
Act

CHO	Chinese hamster ovary (cell line cells)

CI	confidence interval

CL	confidence limit

CNS	central nervous system

CO I	conflict of interest

COPD	chronic obstructive pulimary disease

CPAD	Chemical and Pollutant Assessment
Division

CPHEA Center for Public Health and
Environmental Assessment
CYP450 cytochrome P450

DAF	dosimetric adjustment factor

DMSO	dimethylsulfoxide

DNA	deoxyribonucleic acid

eGFR	estimated glomerular filtration rate

EPA	Environmental Protection Agency

ER	extra risk

FDA	Food and Drug Administration

FEVi	forced expiratory volume of 1 second

FSH	follicle-stimulating hormone

GD	gestation day

GDH	glutamate dehydrogenase

GGT	y-glutamyl transferase

GLP	Good Laboratory Practice

GSH	glutathione

GST	glutathione-^"-transferase

HAP	hazardous air pollutant

HAWC	Health Assessment Workspace

Collaborative

Hb/g-A	animal blood:gas partition coefficient

Hb/g-H	human blood:gas partition coefficient

HBCD	hexabromocyclododecane

HEC	human equivalent concentration

HED	human equivalent dose

HERO	Health and Environmental Research
Online

HPV	high production volume

i.p.	intraperitoneal

i.v.	intravenous

IAP	IRIS Assessment Plan

IARC	International Agency for Research on
Cancer

IRIS	Integrated Risk Information System

IUR	inhalation unit risk

LCso	median lethal concentration

LD50	median lethal dose

LH	luteinizing hormone

LOAEL	lowest-observed-adverse-effect level

LOEL	lowest-observed-effect level

MAC	maximum acceptable concentration

MeSH	Medical Subject Headings

MLE	maximum likelihood estimation

MN	micronuclei

MNPCE	micronucleated polychromatic

erythrocyte

MOA	mode of action

MRL	minimal risk level

MTD	maximum tolerated dose

NCI	National Cancer Institute

NMD	normalized mean difference

NOAEL	no-observed-adverse-effect level

NOEL	no-observed-effect level

NTP	National Toxicology Program

NZW	New Zealand White (rabbit breed)

OAR	Office of Air and Radiation

OECD	Organisation for Economic

Co-operation and Development

OLEM	Office of Land and Emergency

Management

ORD	Office of Research and Development

OSF	oral slope factor

OW	Office of Water

This document is a draft for review purposes only and does not constitute Agency policy.

viii	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

PBPK

physiologically based pharmacokinetic

SEM

systematic evidence map

PECO

populations, exposures, comparators,

SGOT

serum glutamic oxaloacetic



and outcomes



transaminase, also known as AST

PK

pharmacokinetic

SGPT

serum glutamic pyruvic transaminase,

PND

postnatal day



also known as ALT

POD

point of departure

TDI

tolerable daily intake

POD [AD J]

duration-adjusted POD

TIAB

title and abstract

QAPP

quality assurance project plan

TK

toxicokinetic

QSAR

quantitative structure-activity

TSCA

Toxic Substances Control Act



relationship

TSCATS

Toxic Substances Control Act Test

RD

relative deviation



Submissions

RfC

inhalation reference concentration

TWA

time-weighted average

RfD

oral reference dose

UF

uncertainty factor

RfV

reference value

UFa

animal-to-human uncertainty factor

RGDR

regional gas dose ratio

UFd

database deficiencies uncertainty factor

RNA

ribonucleic acid

UFh

human variation uncertainty factor

ROBINS I

Risk of Bias in Nonrandomized Studies

UFl

LOAEL-to-NOAEL uncertainty factor



of Interventions

UFs

subchronic-to-chronic uncertainty

SAR

structure-activity relationship



factor

SCE

sister chromatid exchange

WOS

Web of Science

SD

standard deviation





SDH

sorbitol dehydrogenase





SE

standard error





This document is a draft for review purposes only and does not constitute Agency policy.

ix	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

AUTHORS | CONTRIBUTORS | REVIEWERS

Assessment Team

Xabier Arzuaga. Ph.D. (Assessment Comanager)	EPA/ORD/CPHEA/CPAD

Martha Powers. Ph.D. (Assessment Comanager)

Thomas F. Bateson. Sc.D., M.P.H.

Bevin Blake. Ph.D.

Channa Keshava. Ph.D.

Amanda Persad. Ph.D.

Margaret Pratt. Ph.D.

Hongvu Ru. Ph.D.

Executive Direction

Wayne Cascio, M.D. (CPHEA Director)	EPA/ORD/CPHEA

V. Kay Holt, M.S. (CPHEA Deputy Director)

Samantha Jones, Ph.D. (CPHEA Associate Director)

Kristina Thayer. Ph.D. (CPAD Director)

Andrew Kraft. Ph.D. (CPAD Associate Director)

Ravi Subramaniam. Ph.D. (CPAD Senior Science Advisor)

Paul White, (CPAD Senior Science Advisor)

Elizabeth Radke. Ph.D. (Branch Chief)

lanice Lee. Ph.D. (Branch Chief)

Kathleen Newhouse. M.S. (Acting Branch Chief)

Glenn Rice, Ph.D. (Branch Chief)

Viktor Morozov, Ph.D. (Branch Chief)

Vicki Soto, B.S. (Branch Chief)

Production Team

Maureen Johnson (CPHEA Webmaster)	EPA/ORD/CPHEA

Ryan Jones (HERO Director)

Dahnish Shams (Production Team)

Jessica Soto-Hernandez (Production Team)

Samuel Thacker (HERO Team)

Garland Waleko (Production Team)

This document is a draft for review purposes only and does not constitute Agency policy.

x	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Protocol for the Uranium IRIS Assessment (Oral)

The Integrated Risk Information System (IRIS) Program is undertaking a reassessment of
the noncancer health effects of natural and/or depleted uranium via oral exposure. Enriched
uranium is not a subject of this assessment

IRIS assessments provide high quality, publicly available information on the toxicity of
chemicals to which the public might be exposed. These science assessments are not regulations and
do not constitute U.S. Environmental Protection Agency (EPA) policy. Science assessments such as
these provide a critical part of the scientific foundation for subsequent risk assessment and risk
management decisions made by EPA program and regional offices to protect public health. IRIS
assessments are also used by states and local health agencies, Tribes, other federal agencies,
international health organizations, and other external stakeholders.

This protocol document includes the IAP content, revised in response to public input and
updated EPA scoping needs, and presents the methods for conducting the systematic review and
dose-response analysis for the assessment. While the IAP described what the assessment will cover,
this protocol describes how the assessment will be conducted (see Figure 1-1).

The systematic review methods described in this protocol are based on the Office of
Research and Development (ORD) Staff Standard Operating Procedures for Developing Integrated
Risk Information System (IRIS) Assessments (Version 2.0, referred to as the "IRIS Handbook") (U.S.
EPA. 2022al.

Systematic Review: A structured and documented process for transparent
literature review using explicit, pre-specified scientific methods to identify, select,
assess, and summarize the findings of similar but separate studies.

Assessment
Initiated

Scoping/Initial
Problem
Formulation

Specify Assessment
Approach

Assessment
Developed

Figure 1-1. Integrated Risk Information System systematic review problem
formulation and method documents.

This document is a draft for review purposes only and does not constitute Agency policy.

1-1	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

Protocol for the Uranium IRIS Assessment (Oral)

2. SCOPING AND INITIAL PROBLEM FORMULATION
SUMMARY	

2.1. BACKGROUND

2.1.1. Physical and Chemical Properties

Uranium (U), the 92nd element in the periodic table, is a naturally occurring radioactive
actinide element,1 which has the highest atomic mass among naturally occurring elements. The
half-life of naturally occurring uranium ranges between 159,200 and 4.5 billion years. It is a silvery-
gray metal in the actinide series of elements, and a uranium atom has 92 protons and 92 electrons
of which 6 are valence electrons. In nature, uranium can be found in rock and ores. In the United
States it can be naturally found in greatest concentrations in western states (including Arizona,
Colorado, New Mexico, Texas, Utah, and Wyoming) (U.S. EPA. 2023a: ATSDR. 20131. Table 2-1 lists
the properties of elemental uranium and the most common uranium compounds used in
toxicological studies (uranyl nitrate, uranyl acetate, uranyl fluoride, uranium tetrachloride, and
uranyl fluoride).

In nature uranium exists as a mixture of three isotopes: 234U, 235U, and 238U, with 238U being
the most abundant By weight, natural uranium is mostly (99.27%) 238U, with 0.72% 235U and
0.006% 234U CUSEPAOGWDW. 20001. The specific activities of U-238, U-235, and U-234 in natural
uranium are about 12.4, 80, and 231,000 becquerels [Bq]/mg, respectively (Kim etal.. 2012). or
0.34, 2.2, and 6,253 pCi/kg. The specific activity of natural uranium in rock is 0.68 pCi/[ig (USEPA
OGWDW. 2000). Uranium is "enriched" by processes that remove and concentrate 235U from 0.72%
to 2-4%, with the remaining uranium being termed "depleted." Depleted uranium has a greater
concentration of 238U than natural uranium, but the toxicity of the two are believed to be essentially
identical. In its refined state uranium is malleable, dense, ductile, and slightly paramagnetic
fUNSCEAR. 2017: ATSDR. 20131.

Uranium is chemically reactive and can combine with most elements. In air, the metal easily
oxidizes and becomes coated with a layer of oxide (Bleise etal.. 2003). Uranium forms compounds
in which the valence of the element can range between +3 and +6. The most prevalent form of
uranium in the environment is the uranyl ion U022+ (the +6-oxidation state). It can form
complexes with phosphate, carbonate, and sulfur ions fSheppard etal.. 20051. In aqueous solutions,
only the +4 and +6 compounds are sufficiently stable, both thermodynamically and kinetically, to be

1 Actinide elements are 15 metallic chemical elements that are all radioactive and found in the f-block of the
periodic table.

This document is a draft for review purposes only and does not constitute Agency policy.

2-1	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	of biological importance. These are the compounds that are commonly identified in and transported

2	by ground and surface waters fNRC. 19881.

Table 2-1. Chemical identity and physiochemical properties of selected
uranium compounds as curated by EPA's CompTox Chemicals Dashboard

Name

Elemental
uranium

Uranyl nitrate

Uranium
tetrachloride

Uranyl fluoride

Uranyl acetate

CASRN

7440-61-1

10102-06-4

10026-10-5

13536-84-0

541-09-3

DTXSID3

1042522

2037136

1064906



3060243

Structure

U

0 0
Ox .0 //

| // xo^ cr

0- o

CI

1

-U- ¦ LI

Cf

0

II

F	U	F

II

0

r-i

J,

0

1

0=u=0

1

fUi.

Molecular weight
(g/mol)

238.029

394.035

379.83

308.024

388.115

Molecular formula

U

U02(N0b)2

UCI4

F2O2U

C4H606U

Selected synonyms

238U

Uranium dinitrate
dioxide, uranyl
dinitrate

Uranium chloride

Uranium difluoride
dioxide, Difluoride
[bis(oxido)]
uranium

Uranium,
bis(acetato-
.kappa.0)dioxo-,
(T-4)

Water solubility
(mol/L)b



-

-

-

-

LogKow: Octanol -
Water"



-

-

-

-

Melting point (°C)b

1.1! x 103

-

-

-

-

Boiling point (°C)b

3.82 x 103

-

-

-

-

aDTXSIDs are unique substance identifiers used for curation by EPA's Distributed Structure-Searchable Toxicity (DSSTox) project
(https://www.epa.gov/chemical-research/distributed-structure-searchable-toxicitv-dsstox-database).

Experimental average values for physiochemical properties are shown here. Median values and ranges for physiochemical
properties are also provided on EPA's Chemicals Dashboard at https://comptox.epa.gov/dashboard/ (U.S. EPA. 2023a). If no
experimental or predicted values were available on the Chemicals Dashboard, is shown.

2.1.2. Sources, Production, and Use

3	Uranium is naturally present in many soils with an average concentration in the United

4	States and worldwide of about 3 ppm; some areas, particularly in the western US, have higher

5	concentrations. Uranium is found as a component of various minerals (e.g., uraninite, pitchblende,

6	and carnotite) in its natural state, but not in its metallic state fATSDR. 20131. Commercially viable

7	phosphate ore deposits contain uranium fUlrich etal.. 2014: Sattouf et al.. 20071. The major

8	producers of uranium in the world are the US, China, Australia, Kazakhstan, Namibia, Niger, Russia,

This document is a draft for review purposes only and does not constitute Agency policy.

2-2	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

Protocol for the Uranium IRIS Assessment (Oral)

and Uzbekistan fKeith etal.. 20151. In the United States higher concentrations in rocks and ores
occur in westerns states including Arizona, Colorado, New Mexico, Texas, Wyoming, and Utah
CATSDR. 20131.

The main commercial use for uranium is to create fuel for electricity (NRC. 20121. Uranium
is mined primarily for the U235 isotope, and the process of enrichment adjusts the ratio of U234, U235,
and U238 to an increased amount of U235 (Yelamanchili and Fox. 20101. In addition to energy and
weapons production, uranium is also used in a variety of products such as X-ray targets, glass
tinting agents, gyroscope wheels, ceramic glazes, and shields for radioactive sources. Enriched
uranium2 is used in nuclear reactor fuel and in nuclear weapons.

Depleted uranium is the by-product of the uranium enrichment process. It is less
radioactive than natural uranium (approximately 60%) and it has a density higher than lead fUNEP.
2022: U.S. EPA. 2006al. Because of its physical properties depleted uranium is used for several
applications including: as a counterbalance in aircraft, for shielding against ionizing radiation, as a
gyroscope component, and both in military armor and in armor penetrating munitions fUNEP.
2022: ATSDR. 20131.

2.1.3.	Environmental Fate and Transport

Uranium is naturally mobilized from the Earth's crust by chemical and mechanical
weathering of rocks. Uranium mining, milling, and processing operations can release it into the
environment leading to elevated levels of uranium in affected soils, dusts, and surface and ground
water fU.S. EPA. 2023b: ATSDR. 20131. Uranium mining and the treatment of uranium ore creates
waste in the form of tailings which contain uranium and other radioactive elements such as radium
and plutonium (Brugge and Buchner. 2011: Yelamanchili and Fox. 20101. Depleted uranium has
also been introduced into the environment because of its use in military conflicts (WHO. 20011. and
can be found in soil, water, biota, and airborne particles (U.S. EPA. 2006al.

2.1.4.	Potential Human Exposure (Oral)

The general population is primarily exposed to uranium through intake of food and
drinking water. Higher levels of uranium are seen in water from wells in uranium-rich rock. Human
daily intake from water and food has been estimated to range from 0.9 to 1.5 |ig U/day depending
on the drinking water source and type of diet (Keith etal.. 20151. Uranium from soil is adsorbed
onto the roots of plants; root crops including potatoes, onions, and other root vegetables are a
source of uranium in the diet (ATSDR. 20131.

Environmental exposures to uranium include ingestion of soil, foods, surface water, or
ground water including ingestion of locally grown or foraged food. Such routes of exposure may be
important at a number of Superfund sites with uranium contamination that are located on or near
Indian Country fArnold. 2014: ATSDR. 2013: Middlecamp etal.. 2006: Brugge and Goble. 20021.

2Enriched uranium is not a subject of this assessment.

This document is a draft for review purposes only and does not constitute Agency policy.

2-3	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

Protocol for the Uranium IRIS Assessment (Oral)

Depending on the chemical form of uranium and circumstances of intake, about 0.1%-6% of
ingested uranium is absorbed by the gastrointestinal tract and enters the systemic circulation in
humans, with soluble uranium compounds (e.g., uranyl nitrate and uranyl acetate) being more
readily absorbed (Keith etal.. 20151. Urinary excretion is the principal elimination pathway for
absorbed uranium. Absorbed uranium is retained in many organ systems with the highest levels
found in bones, liver, and kidneys. It is estimated that 66% of the typical human body burden of
uranium is found in the skeleton. Uranium in the skeleton is retained for a longer period, with a
half-life on the order of 70-200 days; most of the uranium in other tissues leaves the body within
1-2 weeks following exposure fATSDR. 20131.

2.1.5. Previous Assessments of Oral Exposure to Uranium by the Environmental Protection

Agency and Other Health Agencies

Existing human health oral reference values for uranium from federal, state, and
international agencies were searched in October 2022 as described in Appendix B and are depicted
in Figure 2-1, and Table 2-2. IRIS published health effect assessments on uranium soluble salts in
1989, which included a reference dose (RfD) for lifetime oral exposure to uranium (U.S. EPA. 19891.
The RfD was based on a study by Mavnard and Hodge (19491 in which rabbits were administered
uranyl nitrate hexahydrate in the diet at 0%, 0.02%, 0.1%, or 0.5% (2.8,14, or 71 mg/kg-day) for
30 days. An RfD of 0.003 mg/kg-day for uranium was derived based on the Lowest Observed
Adverse Effects Level (LOAEL) of 2.8 mg/kg-day for renal histopathological damage. The RfD was
calculated by applying an uncertainty factor of 1,000 (a factor of 10 for interspecies extrapolation,
10 for intraspecies extrapolation, and 10 for use of a LOAEL).

The EPA Office of Water (OW) also developed an RfD for chronic (lifetime) exposure to
uranium (USEPAOGWDW. 20001. These values were based on renal histopathology (dilation of
tubules, apical displacement, vesiculation of tubular nuclei, and cytoplasmic vacuolation and
degranulation in kidneys of male rats exposed to uranyl nitrate) observed in a subchronic exposure
study in which Sprague-Dawley (SD) rats were exposed to uranyl nitrate at 0.06, 0.31,1.52, 7.54,
36.73 mg/kg-day for 91 days (Gilman et al.. 19981. A chronic RfD of 0.0006 mg/kg-day was derived
based on a LOAEL of 0.06 mg/kg-day and applying a UF of 100 (3 for animal to human
extrapolation, 10 for interhuman variability, 3 for LOAEL to NOAEL extrapolation, and 1 for
subchronic to chronic adjustment).

Health Canada calculated a tolerable daily intake (TDI), health-based value (HBV), and a
maximum acceptable concentration (MAC) for chronic exposure to uranium in drinking water.

Their analysis was also based on renal lesions reported in the Gilman et al. 1998 study, which
exposed male rats to uranyl nitrate for 91 days (Health Canada. 2019: Gilman etal.. 19981. This
study was selected for the Health Canada risk assessment point of departure as it reported the
lowest LOAEL for kidney effects. A total uncertainty of 100 (10 for animal to human extrapolation,
and 10 for interhuman variability) was applied to the selected LOAEL of 0.06 mg U/kg-day. The TDI

This document is a draft for review purposes only and does not constitute Agency policy.

2-4	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

Protocol for the Uranium IRIS Assessment (Oral)

of 0.0006 mg/kg-bw was used to determine an HBV for total uranium in drinking water of
0.014 mg/L and a MAC of 0.02 mg/L total natural uranium in drinking water fHealth Canada. 20191.

In 2013, the Agency for Toxic Substances and Disease Registry (ATSDR) completed its
Toxicological Profile for Uranium (ATSDR. 20131. which includes a detailed review of the available
human epidemiology and experimental toxicology data. The ATSDR Toxicological Profile examines
the substantial data available on the kidney reproductive, developmental, and other effects of
uranium and recommends an intermediate-duration oral minimal risk level (MRL) of
2 x 10"4 mg U/kg/day for soluble uranium compounds. This intermediate-duration MRL is also
based on the 91-day study in rats by Gilman et al. 1998 fGilman et al.. 19981. This MRL calculation
uses a LOAEL value of 0.06 mg U/kg-day for renal effects in rats, divided by an uncertainty factor of
300. This includes a factor of 3 because of the use of a "minimal" LOAEL, a factor of 10 for animal to
human extrapolation, and a factor of 10 for human variability.

Uranium and Compounds Oral Reference Values

o.i

0.01

>•
(U
"O

J. 0.001

a>

¦»-»
oj
cc

0)

o
G

0.0001

0.00001

Acute

Short Term

Subchronic

Chronic

1 1 1 1 1 1 1 1

24-Hours

«/)
>
re
O
o
m

7-Years

70-Years



:

ATSDR-MRL
$	^



EPA/IRIS RfD
~





* —

ATSDR-MRL

•	-X









Health CanadaTDI (Natu

alU) XttD EPA/

'OW RfD

10

100	1,000

Duration (Days)

Values apply to soluble uranium salts unless otherwise noted

10,000

100,000

January 2024

X ATSDR-MRL

~ EPA/IRIS RfD

~ EPA/OW RfD

X Health Canada TDI

Figure 2-1. Available health effect reference values for oral exposure to
uranium (current as of November 2022).

This document is a draft for review purposes only and does not constitute Agency policy.

2-5	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 2-2. Details on derivation of the available health effect reference values for oral exposure to uranium3

Reference
value
name

Duration

Uranium
form(s)

Reference

value
(mg/kg-d)

Health effect

Point of
departure

Qualifier

Source

Uncertainty
factors

Notes on
derivation

Review
status

EPA RfD
(IRIS)

Chronic

Soluble

uranium

salts

0.003

Initial BW loss
and mild
nephrotoxicity
in rabbits
exposed to
uranyl nitrate
hexahydrate
for 30 d

2.8 mg
U/kg-d

LOAEL

Mavnard and Hodge
(1949)

Total
UF= 1,000
UFa= 10
UFh = 10
UFl= 10

NA

Final

NCEA (1989)

EPA RfD
(OW)

Chronic

Soluble

uranium

salts

0.0006

Renal

histological
lesions in male
rats exposed
to uranyl
nitrate
hexahydrate
for 91 d

0.06 mg
U/kg-d

LOAEL

Gilman et al. (1998)

Total
UF= 100

UFa= 3

UFh = 10

UFl = 3

UFs= 1

NA

Final

USEPA

OGWDW

(2000)

ATSDR
MRL

Acute
(1-14 d)

Soluble

uranium

salts

0.002

Cleft palate
and other
developmental
effects in fetal
mice exposed
to uranyl
acetate
dihydrate in
utero

0.2 mg
U/kg-d

BMDLos

Domingo et al. (1989)

Total
UF= 100
UFa= 10
UFh = 10

NA

Final

ATSDR (2013)

This document is a draft for review purposes only and does not constitute Agency policy.

2-6	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference





Reference















value



Uranium

value



Point of





Uncertainty

Notes on

Review

name

Duration

form(s)

(mg/kg-d)

Health effect

departure

Qualifier

Source

factors

derivation

status



Intermediate



0.0002

Renal

0.06 mg

LOAEL

Gilman et al. (1998)

Total







(15-365 d)





histological
lesions in male
rats exposed
to uranyl
nitrate
hexahydrate
for 91 d

U/kg-d





UF = 300
UFa= 10
UFh = 10
UFl = 3





Health

Chronic

Natural

0.0006

Renal

0.06 mg

LOAEL

Gilman et al. (1998)

Total

NA

Final

Canada



uranium



histological

U/kg-d





UF= 100



Health

TDI







lesions in male
rats exposed
to uranyl
nitrate
hexahydrate
for 91 d







UFa= 10
UFh = 10



Canada
(2019)

ATSDR = Agency for Toxic Substances and Disease Registry; BMDL = benchmark dose level; BW = body weight; EPA = U.S. Environmental Protection Agency; IRIS = Integrated Risk
Information System; LOAEL = lowest-observed-adverse-effect level; MRL = minimal risk level; OGWDW = Office of Groundwater and Drinking Water; OW = Office of Water;
RfD = reference dose; TDI = tolerable daily intake; UF = uncertainty factor; UFA = animal to human variability; UFH = interhuman variability; UFL = LOAEL-to-NOAEL adjustment;
UFs = subchronic-to-chronic adjustment.
aCurrent as of January 2020; please consult citation source entities and other entities in Appendix Table B-l for current values.

This document is a draft for review purposes only and does not constitute Agency policy.

2-7	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

Protocol for the Uranium IRIS Assessment (Oral)

2.2. SCOPING SUMMARY

During scoping, the IRIS Program met with EPA program and regional offices that had
interest in an IRIS assessment for uranium to discuss specific assessment needs. Table 2-3 below
provides a summary of input from this outreach.

Table 2-3. EPA Program and Regional Office interest in an assessment of
uranium

EPA
program or
regional
office

Oral

Inhalation

Anticipated uses/interest

OW

V



Uranium is found as a natural contaminant of ground water in certain
geologic situations. OW periodically updates drinking water standards
under the Safe Drinking Water Act.

OLEM

V



Uranium is found at approximately 60 Superfund sites across the United
States. Uranium is a hazardous constituent at Resource Conservation and
Recovery Act (RCRA) sites. Uranium is also found at a number of Federal
Facility sites that are managed under CERCLA or RCRA. Sites include
uranium and phosphate mines and the Hanford Nuclear Reservation (non-
enriched uranium).

Region 10

V



Updated uranium reference values are needed to conduct regional risk
assessment-related activities at contaminated sites.

Oral exposure to uranium is of concern to several EPA Program and Regional Office,
including the Office of Water (OW), Office of Land and Emergency Management (OLEM), and Region
10. Uranium is of concern to the OLEM-administered Superfund Program (approximately 60
Superfund sites) and Federal Facility sites managed under the Comprehensive Environmental
Response and Liability Act (CERCLA) or the Resource Conservation and Recovery Act (RCRA), with
oral intake driving site exposure assessments. EPA regulated uranium as a drinking water
contaminant in 2000 based primarily on radiological exposures, but also considering kidney
toxicity. The EPA's Office of Water (OW) periodically updates drinking water regulations and has
need for an IRIS assessment of uranium that examines the more recent literature, and the EPA's
Office of Land and Emergency Management (OLEM) manages Superfund sites (see Table 2-3). The
EPA has been involved in the cleanup or of abandoned uranium mines in Utah, New Mexico, and
Arizona; and Navajo and Hopi lands (U.S. EPA. 2021).

An IRIS assessment plan (IAP) for uranium (IRIS. 2018) was presented at a public science
meeting on March 14, 2018 (https://www.epa.gov/iris/iris-public-science-meeting-mar-2018) to
seek input on the problem formulation components of the assessment plan. The 2018 IAP specifies
why uranium was selected for evaluation, specifies the objectives and specific aims of the

This document is a draft for review purposes only and does not constitute Agency policy.

2-8	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

Protocol for the Uranium IRIS Assessment (Oral)

assessment, provides draft populations, exposures, comparators, and outcomes (PECO) criteria, and
identifies key areas of scientific complexity. However, in April 2019 the uranium assessment was
suspended because of changes in how EPA identified priorities for the IRIS Program (April 2019
IRIS Program Outlook). In June 2021, the assessment work was restarted after interest was
expressed by the EPA Office of Land and Emergency Management (OLEM), Office of Water (OW),
and Region 10. This assessment may also be used to support actions in other EPA programs and
regions and can inform efforts to address uranium by tribes, states, and international health
agencies.

This reassessment focuses on noncancer effects associated with uranium exposure because
cancer risks from uranium have generally been attributed to and assessed as the result of radiation
exposures. In addition, this reassessment focuses only on oral exposure because the oral pathway
has been the primary route of exposure for environmental exposures to uranium (e.g., drinking
water, soils at contaminated sites). Studies on both natural uranium and depleted uranium will be
considered in this reassessment; studies of enriched uranium or the radiological effects of uranium
are not within the assessment scope. This reassessment will include examination of potentially
susceptible populations including women of childbearing age, pregnant women, infants, and
children.

2.3. PROBLEM FORMULATION

EPA's IRIS assessment of uranium dates from 1989 (IRIS. 2018). Much research on the
health effects of uranium has been subsequently published. Systematic review methods were used
to identify a preliminary literature inventory for uranium compounds using the literature search
and screening methods described in Section 4. The ATSDR Toxicological Profile for Uranium
(ATSDR. 2013). was selected as the starting point for the literature search. All references from the
ATSDR Toxicological Profile were retrieved and stored in the EPA's Health and Environmental
Research Online (HERO) database

(https://heronet.epa.gov/heronet/index.cfm/proiect/page/proiect id/3609).3 and a literature
search was conducted to identify studies published since the end of the period covered by the
ATSDR Toxicological Profile (see Section 4).

In this reassessment, EPA will include the literature review and scientific analysis contained
in ATSDR's Toxicological Profile. (ATSDR. 2013) identified urinary, hepatic, neurological,
reproductive, and developmental effects of uranium as being of possible concern. Data on these
effects provided the basis for the Toxicological Profile's MRL values for different durations of
exposure (ATSDR. 2013). The IRIS assessment will examine whether newly available data could be
considered for dose-response analysis for these hazards. Newly available studies and data will also

3EPA's HERO database provides access to the scientific literature behind EPA science assessments. The
database includes more than 600,000 scientific references and data from the peer-reviewed literature used
by EPA to develop its health assessment documents.

This document is a draft for review purposes only and does not constitute Agency policy.

2-9	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

Protocol for the Uranium IRIS Assessment (Oral)

be examined to determine whether there are additional health hazards related to uranium
exposure that have been reported and may provide a basis for hazard evaluation and the
development of toxicity values. As described below, the review of the new literature will be
integrated with the studies and evidence compiled in the ATSDR Toxicological Profile to develop an
updated characterization of health hazards and provide the basis for the derivation of an oral RfD
for uranium.

These methods were implemented in accordance with EPA Quality Assurance policies and
procedures [Quality Policy Procedures4 and CIO 2105.0 (formerly 5360.1 A2)5]. The results
obtained from this systematic compilation of the evidence helped inform the specific aims and key
science issues that will be the focus of the assessment (see Section 2.4 below).

2.4. KEY SCIENCE ISSUE

The preliminary literature survey identified the following key scientific issue, which
warrants evaluation in this assessment

• Earlier life stages appear to be more susceptible to uranium-induced musculoskeletal effects in
experimental studies (Arzuaga etal.. 20151. A toxicological study using SD rats suggests that
newborns are more sensitive than sexually mature animals to uranium-induced effects in the
skeletal system such as decreased cortical bone diameter and trabecular bone development in
the femur fWade-Gueve etal.. 20121. To evaluate potentially increased susceptibility in younger
individuals the available epidemiological and animal evidence will be evaluated and
synthesized according to the recommendations presented in the EPA's Framework for
Assessing Health Risk of Environmental Exposures to Children fBrown etal.. 2008: Makris etal..
2008: U.S. EPA. 2006bl

4U.S. Environmental Protection Agency Procedures for Quality Policy:
https://www.epa.gOv/sites/production/files/2015-10/documents/21060.pdf.

5Policy and Program Requirements for the Mandatory Agency-Wide Quality System:
littps://www.epa.gov/sites/production/files/2015-09/documents/epa order cio 21050.pdf.

This document is a draft for review purposes only and does not constitute Agency policy.

2-10	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

21

22

23

24

25

26

27

28

29

30

Protocol for the Uranium IRIS Assessment (Oral)

3. OVERALL OBJECTIVES AND SPECIFIC AIMS

The overall objectives of this assessment are to identify adverse health effects and
characterize oral exposure-response relationships for noncancer effects from ingestion of uranium
to support development of oral toxicity values (RfD). This assessment will use systematic review
methods to evaluate the epidemiological and toxicological literature for uranium, including
consideration of relevant supplemental material. The assessment methods described in this
protocol utilize EPA guidelines.6

3.1. SPECIFIC AIMS

•	Develop a systematic evidence map (SEM) to identify an initial literature inventory of
epidemiological studies (i.e., human), toxicological studies (i.e., experimental animal), PBPK
models, and supplemental literature pertinent to characterizing the noncancer, health effects of
oral uranium exposure, according to the methods for literature search, screening, and inventory
described in Section 4. The literature search will build on findings from the ATSDR
Toxicological Profile (ATSDR. 2013) and will focus on publications published since the ATSDR
literature search was conducted; the current search addresses publications from 2011 to 2022.

° Epidemiological studies, toxicological studies, and PBPK models are identified for inclusion
based on the predefined populations, exposure, comparators, and outcomes (PECO) criteria
(referred to as the "problem formulation PECO").

° Supplemental material content includes: mechanistic studies, including in vivo, in vitro, ex
vivo, or in silico models; pharmacokinetic and absorption, distribution, metabolism, and
excretion (ADME) studies; studies with routes of exposure other than oral; case studies;
studies that evaluate exposure and health effects associated with exposure to enriched
uranium; studies in non-PECO animal models, such as nonmammalian systems; mixture
studies; case reports or case series; records with no original data; and studies that are
abstract-only or did not have the full text available.

•	Examine whether newly available data indicate a need to update evidence conclusions and (or)
toxicity values for principal health systems from the ATSDR Toxicological Profile. Also examine
whether newly available data on other health systems support identification of additional
uranium health hazards and may plausibly support deriving a toxicity value (RfD) for uranium.

° Informed by these examinations: (1) develop "assessment PECO" criteria that define the
subset of health systems that will be the focus of the systematic review; (2) define the
unit(s) of analysis at the level of endpoint or health system for hazard characterization; and

6EPA guidance documents: http://www.epa.gov/iris/basic-information-about-integrated-risk-information-
svstem#guidance/.

This document is a draft for review purposes only and does not constitute Agency policy.

3-1	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

Protocol for the Uranium IRIS Assessment (Oral)

(3) identify priority analyses of supplemental material to address the specific aims,
uncertainties in hazard characterization, susceptibility, and dose-response analysis.

•	If important newer studies on relevant health systems are identified, these findings will be
considered along with key studies7 cited in the ATSDR Toxicological Profile for evidence
synthesis/integration and RfD derivation purposes using the methods described below.

•	Conduct study evaluations (risk of bias and sensitivity) for individual epidemiological and
toxicological studies that meet the assessment PECO criteria.

•	Conduct a scientific and technical review of available PBPK models and their use. If a PBPK or
PK model is selected for use, the most reliable dose metric will be applied based on analyses of
the available dose metrics and the outcomes to which they are being applied.

•	Conduct data extraction (summarizing study methods and results) from epidemiological and
animal toxicological studies that meet the assessment PECO criteria.

•	For the identified health effect categories with important new data, synthesize evidence across
studies (including both new and older studies cited in ATSDR Toxicological Profile) within the
human and animal evidence streams, using a structured framework to develop and describe
weight of evidence judgments across evidence streams and the supporting rationale for those
judgments ("evidence integration"). The evidence integration analysis presents inferences and
conclusions on human relevance of findings in animals, cross-evidence stream coherence,
potentially susceptible populations and lifestages, and other critical inferences supported by
mechanistic, or ADME, or PK/PBPK data (e.g., biological plausibility). For health systems
examined by ATSDR where important new studies are not identified, EPA will seek to base its
hazard conclusions on ATSDR's findings.

•	For each health effect category, summarize evidence synthesis and evidence integration
conclusions in an evidence profile table (see Section 8).

•	As supported by the currently available evidence, derive noncancer chronic and subchronic oral
reference doses (RfDs) and organ- or system-specific RfDs. Apply pharmacokinetic and
dosimetry modeling (possibly including PBPK modeling) to account for interspecies differences,
as appropriate. Characterize confidence in any toxicity values that are derived.

•	Characterize uncertainties and identify key data gaps and research needs, such as limitations of
the evidence base, limitations of the systematic review, and consideration of dose relevance and
pharmacokinetic differences when extrapolating findings from higher dose animal studies to
lower levels of human exposure.

7Key studies cited in the ATSDR Toxicological Profile document are those that appear to provide informative
data on relevant health outcomes and may plausibly support deriving noncancer toxicity values for uranium.
These will be identified through the study summaries and analysis in the ATSDR Toxicological Profile.
Considerations include studies providing data in dose ranges proximate to toxicological findings considered
in ATSDR's MRL derivation and/or used in important newly identified literature; studies of relevant
durations for toxicity value development (generally studies of subchronic or chronic duration as well as
developmental or reproductive studies using relevant shorter exposure durations); and studies that were not
determined by ATSDR to have major methodological shortcomings.

This document is a draft for review purposes only and does not constitute Agency policy.

3-2	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

Protocol for the Uranium IRIS Assessment (Oral)

4. LITERATURE SEARCH AND SCREENING
STRATEGIES

The literature search and screening processes described in this section were used to
conduct a systematic evidence map (SEM) and identify an initial literature inventory for uranium,
using problem formulation PECO criteria (see Section 4.2) and supplemental screening criteria (see
Section 4.3) to guide the inclusion of studies. The resulting initial literature inventoiy was used to
develop assessment PECO criteria (described in Section 5). The initial literature search as well as all
subsequent literature search updates are conducted using the processes described in this section,
and therefore for the purposes of this assessment the literature inventory developed as part of the
SEM will be continually updated with new studies as the assessment progresses.

4.1.	USE OF EXISTING ASSESSMENTS

The IRIS assessment of uranium will build on findings from the ATSDR Toxicological Profile
for Uranium, (ATSDR. 2013) which included an extensive search of the existing literature. The
literature search for the current uranium assessment will focus on publications since the ATSDR
literature search was conducted (i.e., publications from 2011 to 2022). The United Nations
Scientific Committee on the Effects of Atomic Radiation published a review of uranium that
included examination of toxicological and epidemiological studies fUNSCEAR. 20171. so this
reference will also be consulted to aid in identification of literature. Finally, any unique references
from the 1989 U.S. EPA IRIS summary will also be incorporated (U.S. EPA. 1989).

4.2.	POPULATIONS, EXPOSURES, COMPARATORS, AND OUTCOMES
CRITERIA FOR THE SYSTEMATIC EVIDENCE MAP

PECO (Populations, Exposures, Comparators, and Outcomes) criteria are used to focus the
research question(s), search terms, and inclusion/exclusion criteria. The PECO criteria used to
develop the SEM are referred to hereafter as the "problem formulation PECO" (see Table 4-1) and
were intentionally broad to identify the available evidence in humans and animal models. During
problem formulation, exposure to uranium from routes other than ingestion were determined to be
out of scope for this assessment

This document is a draft for review purposes only and does not constitute Agency policy.

4-1	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 4-1. Problem formulation populations, exposures, comparators, and
outcomes criteria used for the systematic evidence map

PECO element

Evidence

Population

Human: Any population and lifestage (occupational or general population, including children
and other sensitive populations). Note: Case reports and case series will be tracked during
study screening as potentially relevant supplemental material



Animal: Nonhuman mammalian animal species (whole organism) of any life stage (including
preconception, in utero, lactation, peripubertal, and adult stages).

Exposure

Exposure to natural or depleted uranium based on administered dose or concentration,
biomonitoring data (e.g., urine, blood, or other specimens), environmental or occupational-
setting measures (e.g., air, water levels), or job title or residence. Studies on natural uranium
and depleted uranium will be included, studies on enriched uranium or specific to radiation
exposure from uranium will not be included but will be tracked as potentially relevant
supplemental information.

Oral exposure will be examined. Other exposure routes, such as those that are clearly dermal,
or inhalation will be tracked during title and abstract screening as "supplemental information."

Animal studies involving exposures to mixtures will be included only if they include an arm with
exposure to uranium alone.

Comparator

Human: A comparison or reference population exposed to lower levels (or no
exposure/exposure below detection levels) of uranium or to uranium for shorter periods. Any
study with a comparison group, control group, or referent group, including:

•	A comparison group that does not have the disease or outcome of interest (such as a
case-control study); or

•	Any study comparing exposed individuals to unexposed or lower-exposed individuals
including:

•	A comparison group with no exposure to the chemical of interest or exposure below
detection limits, or

•	A comparison group exposed to lower levels of the chemical of interest; or

•	A comparison group exposed to the chemical of interest for shorter periods of time;
or

•	Any study assessing the association between a continuous measure of exposure and a
health outcome; or

For studies in which humans are intentionally exposed to the chemical of interest, an individual
can serve as their own control.



Animal:

A concurrent control group exposed to vehicle-only treatment and/or untreated control. The
control could be a baseline measurement (e.g., acute toxicity studies of mortality) or a
repeated measure design.

Outcomes

All noncancer health effect categories. In general, endpoints related to clinical diagnostic
criteria, disease outcomes, histopathological examination, or other apical/phenotypic
outcomes will be prioritized for evidence synthesis over outcomes such as biochemical
measures.

This document is a draft for review purposes only and does not constitute Agency policy.

4-2	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

Protocol for the Uranium IRIS Assessment (Oral)

4.3. SUPPLEMENTAL CONTENT SCREENING CRITERIA

During the literature screening process, studies containing information that may be
potentially relevant to the specific aims of the assessment are tagged as supplemental material by
category. Because the major health effect categories and units of analysis are not fully identified
when screening is initially conducted, the broad tagging categorization, described in Table 4-2, was
used to characterize the available evidence base and facilitate further screening and analysis of the
supplemental material after PECO refinement Some studies could emerge as being critically
important to the assessment and may need to be evaluated and summarized at the individual study
level (e.g., certain MOA or ADME studies), or might be helpful to provide context (e.g., provide
hazard evidence from routes or durations of exposure not meeting the PECO), or might not be cited
at all in the assessment (e.g., individual studies that contribute to a well-established scientific
conclusion). The categories are designed to help the assessment team prioritize citations for
consideration in the assessment based on the likelihood of impacting assessment conclusions.

This document is a draft for review purposes only and does not constitute Agency policy.

4-3	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 4-2. Categories of potentially relevant supplemental material

Category

Evidence

Typical assessment use

Mechanistic

Studies that do not meet PECO criteria but report measurements that
inform the biological or chemical events associated with phenotypic
effects related to a health outcome. Experimental design may include
in vitro, in vivo (by various routes of exposure; includes all transgenic
models), ex vivo, and in silico studies in mammalian and
nonmammalian model systems. Studies using new approach
methodologies (NAMs, e.g., high-throughput testing strategies, read-
across applications) are also categorized here. Studies where the
chemical is used as a laboratory reagent (e.g., as a chemical probe
used to measure antibody response) generally are not considered
relevant and should be excluded).

Prioritized studies of mechanistic endpoints are described
in the mechanistic synthesis sections; subsets of the most
informative studies may become part of the units of
analysis. Mechanistic evidence can provide support for
the relevance of animal effects to humans and biological
plausibility for evidence integration judgments (including
MOA analyses, e.g., using the MOA framework in the U.S.
EPA Cancer Guidelines). (U.S. EPA, 2005a)

Enriched uranium

Studies that evaluate health effects caused by the enriched fissionable
uranium isotope. Uranium is enriched by processes that concentrate
235U. Enriched uranium is used in nuclear reactor fuel and in nuclear
weapons; it is not a subject of this assessment.

Studies of non-PECO animals, exposures, or durations can
be summarized to inform evaluations of consistency
(e.g., across species, routes, or duration), coherence, or
adversity; subsets of the most informative studies may be
included in the unit of analysis. These studies may also be
used to inform evidence integration judgments of
biological plausibility and/or MOA analyses and thus may
be summarized as part of the mechanistic evidence
synthesis.

Non-PECO animal model
(i.e., nonmammalian
systems)

Studies reporting outcomes in animal models that meet the outcome
criteria but do not meet the "P" in the PECO criteria. Depending on
the endpoints measured in these studies, they can also provide
mechanistic information (in these cases studies should also be tagged
"mechanistic or MOA").

Non-PECO route of
exposure

Epidemiological or animal studies that use a non-PECO route of
exposure, (e.g., injection studies or dermal studies if the dermal route
is not part of the exposure criteria).

This categorization generally does not apply to epidemiological studies
where the exposure route is unclear; such studies are considered to
meet PECO criteria if the relevant route(s) of exposure are plausible,
with exposure being more thoroughly evaluated at later steps.

Non-PECO exposure
duration

For assessments that focus on chronic exposure, acute exposure
durations (defined as animal studies of less than 1 d in duration) are
generally considered supplemental. In rare cases and for very large

This document is a draft for review purposes only and does not constitute Agency policy.

4-4	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Category

Evidence

Typical assessment use



evidence bases, short-term (i.e., less than subchronic) exposure
durations could also be categorized as supplemental.

Some assessment teams might prefer to keep these studies as PECO
relevant and summarize them in the literature inventory rather than
track them as supplemental.



Susceptible populations

Studies that help identify potentially susceptible subgroups, including
citations investigating how intrinsic factors such as sex, lifestage,
genotype, or other factors (e.g., health status) that can influence
toxicity. These are often co-tagged with other supplemental material
categories, such as mechanistic or ADME. Studies meeting PECO
criteria that also address susceptibility should be co-tagged as
supplemental.

Susceptibility based on most extrinsic factors, such as increased
exposure due to residential proximity to exposure sources, is not
considered an indicator of susceptible populations for the purposes of
IRIS assessments.

Provides information on factors that might predispose
sensitive populations or lifestages to a higher risk of
adverse health effects following exposure to the
chemical. This information is summarized during evidence
integration for each health effect and is considered during
dose-response, where it can directly impact modeling
decisions.

Classical pharmacokinetic
(PK) or physiologically
based pharmacokinetic
(PBPK) model studies

Classical pharmacokinetic or dosimetry model studies: Classical PK or
dosimetry modeling usually divides the body into just one or two
compartments, which are not specified by physiology, wherein
movement of a chemical into, between, and out of the compartments
is quantified empirically by fitting model parameters to absorption,
distribution, metabolism, and excretion (ADME) data. This category is
for papers that provide detailed descriptions of PK models that are not
physiologically based PK (PBPK) models.

•	The data are typically the concentration time course in blood
or plasma after oral and or intravenous exposure, but other
exposure routes can be described.

•	A classical PK model might be elaborated from the basic
structure applied in standard PK software, for example to
include dermal or inhalation exposure, or growth of body
mass over time, but otherwise does not use specific tissue
volumes or blood flow rates as model parameters.

PBPK and PK model studies are included in the
assessment and evaluated for possible use in conducting
quantitative extrapolations. PBPK/PK models are
categorized as supplemental material with the
expectation that each one will be evaluated for
applicability to address assessment extrapolation needs
and technical conduct. Specialized expertise is required
for their evaluation.

Standard operating procedures for PBPK/PK model
evaluation and the identification, organization, and
evaluation of ADME studies are outlined \nAn Umbrella
Quality Assurance Project Plan (QAPP)for PBPK models
(U.S. EPA, 2018b).

This document is a draft for review purposes only and does not constitute Agency policy.

4-5	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Category

Evidence

Typical assessment use



•	Such models can be used for extrapolation similar to PBPK
models, although such use might be more limited.

•	Note: ADME studies often report classical PK parameters,
such as bioavailability (fraction of an oral dose absorbed),
volume of distribution, clearance rate, and/or half-life or half-
lives. If a paper provides such results only in tables with
minimal description of the underlying model or software (i.e.,
uses standard PK software without elaboration), including
"noncompartmental analysis," it should only be listed as a
supplemental material ADME study.

Physiologically based pharmacokinetic or mechanistic dosimetry
model studies: PBPK models represent the body as various
compartments (e.g., liver, lung, slowly perfused tissue, richly perfused
tissue) to quantify the movement of chemicals or particles into and
out of the body (compartments) by defined routes of exposure,
metabolism, and elimination, and thereby estimate concentrations in
blood or target tissues.

•	Usually specific to humans or defined animal species; often a
single model structure is calibrated for multiple species.

•	Some mechanistic dosimetry models might not be
compartmental PBPK models but predict dose to the body or
specific regions or tissues based on mechanistic data, such as
ventilation rate and airway geometry.

•	A defining characteristic is that key parameters are
determined from a substance's physicochemical parameters
(e.g., particle size and distribution, octanol-water partition
coefficient) and physiological parameters (e.g., ventilation
rate, tissue volumes); that is, data that are independent of in
vivo ADME data that are otherwise used to estimate model
parameters.



Pharmacokinetic (ADME)

Pharmacokinetic (ADME) studies are primarily controlled experiments
in which defined exposures usually occur by intravenous, oral,
inhalation, or dermal routes, and the concentration of particles, a

ADME studies are inventoried and prioritized for possible
inclusion in an ADME synthesis section on the chemical's
PK properties and for conducting quantitative
adjustments or extrapolations (e.g., animal to human).

This document is a draft for review purposes only and does not constitute Agency policy.

4-6	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Category

Evidence

Typical assessment use



chemical, or its metabolites in blood or serum, other body tissues, or
excreta are then measured.

•	These data are used to estimate the amount absorbed (A),
distributed to different organs (D), metabolized (M), and/or
excreted (E) through urine, breath, or feces.

•	The most informative studies involve measurements over
time such that the initial increase and subsequent
concentration decline is observed, preferably at multiple
exposure levels.

•	Data collected from multiple tissues or excreta at a single
time point also inform distribution.

•	ADME data can also be collected from human subjects who
have had environmental or workplace exposures that are not
quantified or fully defined. However, to be useful such data
must involve either repeated measurements over a time
period when exposure is known (e.g., is zero because
previous exposure ended) or time- and subject-matched
tissue or excreta concentrations (e.g., plasma and urine, or
maternal and cord blood).

•	ADME data, especially metabolism and tissue partition
coefficient information, can be generated using in vitro
model systems. Although in vitro data may not be as
definitive as in vivo data, these studies should also be tracked
as ADME. For large evidence bases it may be appropriate to
separately track the in vitro ADME studies.

Note: Studies describing environmental fate and transport or
metabolism in bacteria or model systems not applicable to humans or
animals should not be tagged.

Specialized expertise in PK is necessary for inventory and
prioritization.

Standard operating procedures for PBPK/PK model
evaluation and the identification, organization, and
evaluation of ADME studies is outlined in An Umbrella
Quality Assurance Project Plan (QAPP)for PBPK models
(U.S. EPA, 2018b).

Exposure and
biomonitoring (no health
outcome)

Exposure characteristic studies include data that are unrelated to
toxicological endpoints, but which provide information on exposure
sources or measurement properties of the environmental agent (e.g.,
demonstrate a biomarker of exposure).

This information may be useful for developing exposure
criteria for study evaluation or refining problem
formulation decisions.

This document is a draft for review purposes only and does not constitute Agency policy.

4-7	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Category

Evidence

Typical assessment use

Mixture studies

Mixture studies that are not considered PECO relevant because they
do not contain an exposure or treatment group assessing only the
chemical of interest. This categorization generally does not apply to
epidemiological studies in which the exposure source might be
unclear.

Mixture studies are tracked to help inform cumulative risk
analyses, which may provide useful context for risk
assessment but fall outside the scope of an IRIS
assessment.

Case reports or case
series

All study designs such as case reports, case series, and case studies
without a comparison group in any setting (e.g., occupational, general
population).

Tracking case studies can facilitate awareness of potential
human health issues missed by other types of studies
during problem formulation.

Records with no original
data

Records that do not contain original data, such as other agency
assessments, informative scientific literature reviews, editorials, or
commentaries.

Studies that are tracked for potential use in identifying
missing studies, background information, or current
scientific opinions (e.g., hypothesized MOAs).

Conference abstracts /
proceedings, abstract-
only

Records that do not contain sufficient documentation to support study
evaluation and data extraction.



ADME = absorption, distribution, metabolism, and excretion; MOA = mode of action; NAM = new approach methodology; PECO = populations, exposures, comparators, and
outcomes; PK = pharmacokinetic; PBPK = physiologically based pharmacokinetic.

This document is a draft for review purposes only and does not constitute Agency policy.

4-8	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

Protocol for the Uranium IRIS Assessment (Oral)

4.4. LITERATURE SEARCH STRATEGIES

4.4.1.	Database Search Term Development

In accordance with the Uranium IAP (IRIS. 20181. the EPA conducted an in-depth literature
search to identify relevant studies published since the completion of the ATSDR literature search.
EPA's search strategy for the literature published since 2011 was developed using key terms and
words related to the PECO criteria.

4.4.2.	Database Searches

The literature search focused on studies published after the period covered by the ATSDR
Toxicological Profile for Uranium, covering the period January 2011 to November 2022. No
language restrictions were applied. The detailed search strategies are presented in Appendix A.
Literature searches were conducted using EPA's Health and Environmental Research Online
(HERO) database.8 The following databases were searched:

•	PubMed (National Library of Medicine)

•	Web of Science (Thomson Reuters)

•	Scopus

•	Toxline9

After deduplication in HERO, records were imported into SWIFT Review software (Howard
etal.. 2016) to identify those references most likely to be applicable to a human health assessment.
In brief, SWIFT Review has preset literature search strategies ("filters") developed and applied by
information specialists to identify studies more likely to be useful for identifying human health
content from those that likely are not (e.g., analytical methods). The filters function like a typical
search strategy in which studies are tagged as belonging to a certain filter if the terms appear in
title, abstract, keyword or MeSH. The applied SWIFT Review filters focused on lines of evidence:
human, animal models for human health, and in vitro studies. The details of the search strategies
that underlie the filters are available online (Sciome. 2019). Studies not retrieved using these filters
were not considered further. Studies that included one or more of the search terms in the title,
abstract, keyword, or MeSH fields were exported as a RIS (Research Information System) file for
screening in SWIFT-Active Screener (Sciome. 2019) and then DistillerSR. as described below in
Section 4.5 (Evidence Partners. 2022).

The literature searches are updated annually throughout the assessment's development and
review process to identify newly published literature. During this period, the literature search

8Health and Environmental Research Online: https: //hero.epa.gov/hero/.

9The Toxline database was migrated to PubMed after the 2019 literature search update, thus it was not
included in subsequent literature search updates.

This document is a draft for review purposes only and does not constitute Agency policy.

4-9	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

Protocol for the Uranium IRIS Assessment (Oral)

terms do not change from those used in the initial search and studies are screened according to
both the problem formulation PECO criteria. Thus, the SEM literature inventory is updated during
the process of developing the draft assessment. The last full literature search update is conducted
several months prior to the planned release of the draft document for public comment. Studies
identified after peer review begins are only considered for inclusion if they are directly relevant to
the assessment PECO criteria and are expected to fundamentally alter the draft assessment
conclusions.

4.4.3. Searching Other Sources

The literature search strategy described above was designed to be broad, but like any
search strategy, studies can be missed [e.g., cases where the specific chemical is not mentioned in
title, abstract, or keyword content; ability to capture "gray" literature (studies not reported in the
peer-reviewed literature) that is not indexed in the databases listed above]. Thus, in addition to the
database searches, the sources below were used to identify studies that could have been missed
based on the database search. Searching of these resources occurs during preparation of the SEM
literature inventory. After preparation of the SEM literature inventory, references can be identified
during public comment periods, by technical consultants, and during peer review. Records that
appeared to meet the problem formulation PECO criteria and that had not been previously
identified in the literature search are uploaded into DistillerSR, annotated with respect to source of
the record, and screened using the methods described in Section 4.5. Appendix C describes the
specific methods and results for searching the sources below. Searching of these sources is
summarized to include the source type or name, the search string (when applicable), number of
results present within the resource, and the URL (uniform resource locator, when available and
applicable). The list of other sources consulted includes:

•	Manual review (at the title level) of the reference list from other publicly available final or draft
assessments from other non-EPA Agencies (e.g., 2016 UNSCEAR Report to the United Nations
General Assembly) or published journal review specifically focused on human health. Reviews
can be identified from the database search or from the resources listed in Appendix B.

•	European Chemicals Agency (ECHA) registration dossiers to identify data submitted by
registrants http://echa.europa.eu/information-on-chemicals/information-from-existing-
substances-regulation.

•	EPA ChemView database fU.S. EPA. 20191 to identify unpublished studies, information
submitted to EPA under Toxic Substances Control Act (TSCA) Section 4 (chemical testing
results), Section 8(d) (health and safety studies), Section 81 (substantial risk of injury to health
or the environment notices), and FYI (For Your Information, voluntary documents). Other
databases accessible via ChemView include the EPA High Production Volume (HPV) Challenge
database and the Toxic Release Inventory database.

•	The National Toxicology Program (NTP) database of study results and research projects
(https://ntp.niehs.nih.gov/results/index.html).

This document is a draft for review purposes only and does not constitute Agency policy.

4-10	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

21

22

23

24

25

26

27

28

29

30

31

32

33

34

Protocol for the Uranium IRIS Assessment (Oral)

•	The Organization for Economic Cooperation and Development (OECD) Screening Information
DataSet (SIDS) High Production Volume Chemicals

https://www.echemportal.org/echemportal/substancesearch/page.action?pageID=9

•	References identified during public comment periods by technical consultants, and during peer
review.

•	References that had been previously added to the HERO database for the uranium assessment
during the development of the IAP.

4.4.4. Non-Peer-Reviewed Data

IRIS assessments rely mainly on publicly accessible, peer-reviewed studies. However, it is
possible that unpublished data directly relevant to the PECO may be identified during assessment
development In these instances, the EPA will try to get permission to make the data publicly
available (e.g., in HERO); data that cannot be made publicly available are not used in IRIS
assessments. In addition, on rare occasions where unpublished data would be used to support key
assessment decisions (e.g., deriving a toxicity value), EPA may obtain external peer review if the
owners of the data are willing to have the study details and results made publicly accessible, or if an
unpublished report is publicly accessible (or submitted to EPA in a non-confidential manner) (U.S.
EPA. 2015). This independent, contractor driven, peer review would include an evaluation of the
study similar to that for peer review of a journal publication. The contractor would identify and
typically select three scientists knowledgeable in scientific disciplines relevant to the topic as
potential peer reviewers. Persons invited to serve as peer reviewers would be screened for conflict
of interest In most instances, the peer review would be conducted by letter review. The study and
its related information, if used in the IRIS assessment, would become publicly available. In the
assessment, EPA would acknowledge that the document underwent external peer review managed
by the EPA, and the names of the peer reviewers would be identified. In certain cases, IRIS will
assess the utility of a data analysis of accessible raw data (with descriptive methods) that has
undergone rigorous quality assurance/quality control review (e.g., ToxCast/Tox21 data, results of
NTP studies not yet published) but that have not yet undergone external peer review.

Unpublished data from personal author communication can supplement a peer-reviewed
study as long as the information is made publicly available. If such ancillary information is acquired,
it will be documented in the Health Assessment Workspace Collaborative (HAWC) or HERO project
page (depending on the nature of the information received).

4.5. LITERATURE SCREENING

The problem formulation PECO criteria described in Section 4.2 are used to determine
inclusion or exclusion of a reference as a primary source of health effects data or a published PBPK
model. In general, records identified from the literature searches are housed in the HERO system

and imported into SWIFT-Active Screener (httpsi //www.sciome.com/swift~activescreener/) for an

This document is a draft for review purposes only and does not constitute Agency policy.

4-11	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

Protocol for the Uranium IRIS Assessment (Oral)

initial title and abstract (TIAB) screen using machine learning, followed by import into DistillerSR
(Evidence Partners; https://distillercer.eom/products/distillersr-systematic-review-software/l for
manual TIAB screening and full-text screening by two independent reviewers. One batch of
literature search results corresponding to the literature search update was imported directly into
DistillerSR for title-abstract screening without the initial import into SWIFT-Active Screener (see
Figure 4-1).

In addition to the inclusion of studies that meet the problem formulation PECO criteria,
studies containing supplemental material that is potentially relevant to the specific aims are
tracked during the screening process. Although not considered to directly meet PECO criteria, these
studies are not strictly excluded unless otherwise specified. Unlike studies that meet PECO criteria,
supplemental studies may not be subject to systematic review unless specifically defined questions
are identified that focus the mechanistic (or other) analysis to inform the specific aims.

4.5.1. Title and Abstract Screening

The studies identified from the searches described above are imported into SWIFT-Active
Screener for TIAB screening. SWIFT-Active Screener is a web-based collaborative software
application that utilizes active machine learning approaches to reduce the screening effort (Howard
etal.. 2020). Following a pilot phase to calibrate screening guidance, two screeners independently
perform a TIAB screen using a structured form. Studies considered "relevant" or "unclear" based on
meeting all problem formulation PECO criteria at the TIAB level are considered for inclusion and
advanced to full-text screening. TIAB screening is conducted by two independent reviewers and
any screening conflicts are resolved by discussion between the primary screeners with consultation
by a third reviewer, if needed. For citations with no abstract, articles are initially screened based on
the following: title relevance (title should indicate clear relevance), and page length (articles two
pages in length or less are assumed to be conference reports, editorials, or letters). Eligibility status
of non-English studies is assessed using the same approach with online translation tools or
engagement with a native speaker.

The machine learning screening process is designed to prioritize references that appear to
meet the problem formulation PECO criteria or supplemental material content for manual review
(i.e., both types of references are screened as "include" for machine learning purposes). Screening
continues until SWIFT-Active Screener indicates that it was likely at least 95% of the relevant
studies are identified, a percent identification often used to evaluate the performance of machine
learning applications and considered comparable to human error rates (Bannach-Brown et al..
2018: Howard etal.. 2016: Cohen etal.. 2006). Any studies with "partially screened" status at the
time of reaching the 95% threshold are then fully screened. Studies identified as meeting the
problem formulation PECO criteria, unclear, or supplemental material by SWIFT-Active Screener
are then imported into DistillerSR software either for conflict resolution or for an additional round
of more specific TIAB tagging (i.e., to separate studies meeting PECO criteria versus supplemental
content and to tag the evidence stream or specific type of supplemental content). In DistillerSR,

This document is a draft for review purposes only and does not constitute Agency policy.

4-12	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

Protocol for the Uranium IRIS Assessment (Oral)

TIAB screening is conducted manually by two independent reviewers and any screening conflicts
resolved by discussion between the primary screeners with consultation by a third reviewer, if
needed. Conflicts between screeners in applying the supplemental tags, which primarily occur at
the TIAB level, are resolved similarly, erring on the side of over-tagging based on TIAB content.

4.5.2.	Full-Text Screening

Full-text references are sought through the EPA's HERO database for studies screened as
meeting the problem formulation PECO criteria or "unclear" based on the TIAB screening. Full-text
screening occurs in DistillerSR. Full-text copies of these records are retrieved, stored in the HERO
database, and independently assessed by two screeners using a structured form in DistillerSR to
confirm eligibility. Screening conflicts are resolved by discussion among the primary screeners with
consultation by a third reviewer or technical advisor (as needed to resolve any remaining
disagreements). Rationales for excluding studies are documented, e.g., study did not meet PECO,
full-text not available. Approaches for language translation include online translation tools or
engagement of a native speaker. Fee-based translation services for non-English studies are typically
reserved for studies that are anticipated as being useful for toxicity value derivation.

4.5.3.	Multiple Publications of the Same Data

When there are multiple publications using the same or overlapping data, all publications
are included, with one selected for use as the primary study; the others are considered as
secondary publications with annotation in HAWC and HERO indicating their relationship to the
primary record during data extraction. For epidemiology studies, the primary publication is
generally the one with the longest follow-up, the largest number of cases, or the most recent
publication date. For animal studies, the primary publication is typically the one with the longest
duration of exposure, the largest sample size, or with the outcome(s) most informative to the PECO
criteria. For both epidemiology and animal studies, the assessments include relevant data from all
publications of the study, although if the same data are reported in more than one study, the data
are only extracted once (see Section 7). For corrections, retractions, and other companion
documents to the included publications, a similar approach to annotation is taken and the most
recently published data are incorporated into the assessments.

4.5.4.	Literature Flow Diagram

The results of the screening process are posted on the project page for the assessment in
the HERO database fhttps://heronet.epa.gov/heronet/index.cfm/proiect/page/project id/2970}.
Results are also summarized in a literature flow diagram (see Figure 4-1) and interactive HAWC
literature trees (where additional sub-tagging beyond what is presented in HERO is documented
and visualized, e.g., more details on the nature of mechanistic or ADME studies).

The literature flow diagram represents the results of the original literature searches as well
as several updates. The original literature search was conducted preceding the absorption of the

This document is a draft for review purposes only and does not constitute Agency policy.

4-13	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Toxline database into PubMed. Most of the literature was initially screening in SWIFT-Active
Screener prior to being screened in DistillerSR. However, the gray literature and one of the
literature updates was directly imported into DistillerSR for screening. For large datasets, the use of
SWIFT-Active Screener before DistillerSR allowed for more efficient screening via the use of the
inherent predictive relevance component. Less than 10% of the references screened at the TIAB
screening level made it to the full-text screening phase and of those, only about half (143 out of
257) were deemed PECO relevant. In addition to identifying references that were PECO relevant,
the screening process identified nearly 1,000 references that can be categorized as supplemental
material.

Literature Searches (January 2011 - November 2022)

PubMed
(n = 1,666)

SCOPUS
(n = 8,119)

WOS
(n = 18,396)

'

Toxline
(n = 1,748)

SWIFT Review Software Applied to 19,718 Records from Literature Search

Identification of potentially relevant records based on application of SWIFT Review evidence stream
tags (n = 4774). Not included in 4774 are 2375 references from 2021 literature update were directly

TIAB Screen in

1 <«.=<

SWIFT Active

774)





788 records consider
SWIFT Activ

ed relevant based on
e screening

TIAB Screen in DistillerSR
(n=3390)

Full-Text Screening
(n = 257)

Studies Meeting PECO
(n = 143)

1 Human health effect records (n = 110)
1 Animal health effect records (n = 33)
1 PBPK models (n = 0)

1 Studies meeting PECO criteria that also
reported mechanistic(n = 38) or ADME (n
= 29) information

Excluded (n = 3986)
2140 records manually screened and
excluded

1846 records predicted as not relevant in
SWIFT Active (and not manually screened)

Additional Database Search

Grey
Literature
(" = 20)

2021 Literature Update
directly imported to
DistillerSR (n = 2375)

Excluded (n = 3140)

Not relevantto PECO and not considered
supplemental (n = 2237)

1 Tagged as supplemental material (n = 903)

Excluded (n = 60)

Not relevantto PECO (n = 54)
Unable to obtain full text (n = 6)

• Tagged as supplemental material (n = 50)

Tagged as Supplemental Material
(n = 953; 903 TIAB + 50 full text)

Mechanistic or MOA (n = 150)

Enriched Uranium (n = 7)

Non-mammalian model systems (n = 76)

Non-oral route of exposure (n = 91)

ADME (n = 68)

Exposure characteristics (n =427)

Mixture studies (n = 27)

Case report or case study (n = 17)

Review, commentary, letter, no original data (n = 219)

Conference abstracts (n = 52)

Figure 4-1. IRIS literature search flow diagram for uranium.

This document is a draft for review purposes only and does not constitute Agency policy.

4-14	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

Protocol for the Uranium IRIS Assessment (Oral)

The Toxline database was migrated to PubMed after the 2019 literature search update, thus it was not included in subsequent
literature search updates.

Tagged as Supplemental Material: these numbers represent the total number of unique citations that were identified; because
some citations are given multiple tags, the sum of the individual material tags is greater than the total number of citations.

4.6. LITERATURE INVENTORY

During TIAB or full-text level screening, studies that meet the problem formulation PECO
criteria are categorized by evidence type (human, or animal) or category of supplemental
information (e.g., mechanistic, ADME, PK/PBPK, reviews). Next, study design details for studies that
meet the problem formulation PECO criteria are summarized. The results of this categorization are
referred to as the literature inventory and is the key analysis output of the SEM. Literature
inventories for PECO-relevant studies were created to develop summary level, sortable lists that
include some basic study design information (e.g., study population, exposure information such as
doses administered or biomarkers analyzed, age/life stage of exposure, endpoints examined).

These literature inventories facilitated subsequent review of individual studies and effects for
comparison with the ATSDR Toxicological Profile.

4.6.1. Studies That Meet Problem Formulation PECO Criteria

Human and animal studies that meet the problem formulation PECO criteria after TIAB and
full-text review are briefly summarized using structured DistillerSR Hierarchical Data Extraction
forms to create literature evidence inventories, which were used to display the extent and nature of
the available evidence (see Section 4.2). The literature inventories are used to inform the
assessment PECO criteria and evaluation plan. Studies were extracted by one team member and the
extracted data were qualitatively reviewed by at least one other team member. The extraction fields
in the forms are available in Microsoft (MS) Excel format upon request. See
https: //www.epa.gov/iris/forms/contact-us-about-iris for requests. The literature inventories
were exported from Distiller SR in MS Excel format.

For experimental animal studies, which are typically studies in rodents, the following
information is captured: chemical form, study type (acute [<24 hours], short term [<7 days], short
term [7-27 days], subchronic [28-90 days], chronic [>90 days10] and developmental, which
includes multigeneration studies), duration of treatment, route, species, strain, sex, dose or
concentration levels tested, dose units, health system and specific endpoints assessed, and a
summary of the results reported in the study.

For epidemiological studies the following information was summarized: uranium
compound, population type (e.g., residential/school based, occupational, other), sex, study design
(e.g., cross-sectional, cohort, case-control, ecological, case-report, controlled trial, meta-analysis),

10EPA considers chronic exposure to be more than approximately 10% of the life span in humans. For typical
laboratory rodent species, this can lead to consideration of exposure durations of approximately 90 days to 2
years. However, studies in duration of 1-2 years are typical of what is considered representative of chronic
exposure rather than durations just over 90 days.

This document is a draft for review purposes only and does not constitute Agency policy.

4-15	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

Protocol for the Uranium IRIS Assessment (Oral)

study location, life stage (adults, children/infants), exposure measurement (air sampling,
occupational history, other), biomonitoring matrix, health system studied, endpoints assessed, and
a brief description of the observed effects. More detail on the process of summarizing studies is
presented in Sections 5 and 7.

4.6.2. Organizational Approach for Supplemental Material

The results of the supplemental material tagging conducted in DistillerSR are imported into
the literature review module in HAWC, where more granular sub-tagging within a type of
supplemental material content category is conducted. A publication can have multiple tags,
including PECO studies that also contain supplemental material. The degree of sub-tagging depends
on the extent of content for a given type of supplemental material and needs of the assessment with
respect to developing human health hazard conclusions and derivation of toxicity values. Tagging
judgments in DistillerSR and HAWC are made by one assessment member and confirmed during the
screening step by another member of the assessment team. The overall approach for supplemental
material content is presented in Figure 4-2, with details on subtagging presented in the following
sections under the specific type of supplemental content (see Table 4-2).

This document is a draft for review purposes only and does not constitute Agency policy.

4-16	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Protocol for the Uranium IRIS Assessment (Oral)

o Has additional sub-tagging
O No additional sub-tagging

,43
Included

Uranium Toxicological Review (2023): Literature Tagtree

(m)

Human Study

©
Animal Study

Uranium Toxicological Review
(2023)



,50



Mechanistic or MOA



©

Enriched uranium



©

Non-mammalian model systems



©

Non-oral route of exposure



©

953

ADME/TK

V-/
ernentary Material

©

( 2295 J

Exposure only

Excluded

©

Mixture studies

©

Case report or case study
200

Review, commentary, letter, no
original data

©

Conference abstracts

Figure 4-2. Visual summary of approach for tagging major categories of
supplemental material. See interactive HAWC link: Uranium Literature Tagtree.

Organization of Mechanistic Information

If a mechanistic analysis is considered necessary to assist with the interpretation and
integration of the epidemiological and experimental evidence of a specific hazard or health effect,
EPA will rely on previously published reviews and analyses to identify potential pathways of
toxicity and identify critical studies through forward/backward searches. To facilitate this analysis,
publications tagged as reviews or commentaries that included a mechanistic analysis were sub-
tagged according to health system/target tissue. With respect to health system/target tissue
tagging, the following organizational categories were applied: cardiovascular, dermal,
developmental, endocrine, gastrointestinal, hematologic, hepatic, immune, metabolic,
musculoskeletal/connective tissue, multi-system, nervous, ocular, reproductive, respiratory,
sensory, urinary, or whole body. The same publication could have multiple tags and studies that
address broad physiological processes were tagged as systemic.

Depending on the extent of evidence for a given health system target tissue/cellular
response category (e.g., liver, nervous system, immune), an additional level of sub-tagging
describing the biological processes presented in the studies may be utilized. This level of sub-

This document is a draft for review purposes only and does not constitute Agency policy.

4-17	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

tagging is based on the content of the available studies (e.g., specific receptor interaction,
inflammation pathway).

This document is a draft for review purposes only and does not constitute Agency policy.

4-18	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

Protocol for the Uranium IRIS Assessment (Oral)

5. REFINED PROBLEM FORMULATION AND
ASSESSMENT APPROACH

The primary purpose of this step is to provide further specification to the assessment
methods based on characterization of the extent and nature of the evidence identified from the
literature inventory. This includes refinements to PECO criteria and defining the unit(s) of analysis
for health endpoints/outcomes during evidence synthesis, and presenting analysis approaches for
mechanistic, ADME or other types of supplemental material content A unit of analysis is an
outcome or group of related outcomes within a health effect category that are considered together
during evidence synthesis (see Section 8). The systematic review will focus on the health outcome
categories that appear to have sufficient information available to support hazard identification,
based upon the availability of animal and human studies as cited in ATSDR Toxicological Profile
fATSDR. 20131. and the updated literature search conducted by EPA.

5.1. COMPARISON WITH ATSDR TOXICOLOGICAL PROFILE (2013)

In this reassessment, EPA builds on the scientific review and analysis from the ATSDR
Toxicological Profile for Uranium fATSDR. 20131. The following categories of health effects of oral
uranium exposure were identified in ATSDR 2013: urinary, hepatic, neurological, reproductive, and
developmental.11 While ATSDR 2013 did not identify the following as hazards, they also considered
uranium-induced body weight changes, mortality, metabolic alterations, and effects on the
endocrine, musculoskeletal, cardiovascular, gastrointestinal, hematological, immune, and
respiratory systems.

This protocol examines newly available literature since the publishing of ATSDR 2013. The
newly available literature as determined by the IRIS literature search (i.e., studies that met problem
formulation PECO criteria) was examined to determine whether the data warranted a revision of
ATSDR health effect categories and their hazard findings or identified additional noncancer health
effect categories for examination in the IRIS assessment The proposed approach to compare
ATSDR 2013 with the IRIS literature search results is shown in Figure 5-1:

nThese were identified by EPA based on the "Summary of Health Effects" section of the Profile (see
Section 1.2) and were confirmed by ATSDR staff in a meeting with EPA in August 2023. Furthermore, urinary,
and developmental effects of uranium were considered the bases for MRL values for intermediate and acute
duration oral exposures, respectively (ATSDR. 20131.

This document is a draft for review purposes only and does not constitute Agency policy.

5-1	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

Protocol for the Uranium IRIS Assessment (Oral)

Figure 5-1. Approach and decision tree used to compare ATSDR 2013 (ATSDR.

2013) with IRIS literature search results.

PECO-relevant studies were examined by two reviewers who compared the IRIS literature
search results with ATSDR 2013 conclusions for each health effect category. The initial examination
was done independently followed by discussion. Expert judgment from the reviewers was used to
look for associations between uranium exposure and health effects, noting potential study
limitations. Appendix D contains the review for each health effect category: summary of the ATSDR
2013 conclusion; description of the new epidemiological data; and description of the new
toxicological data.

As described in Appendix D and Table 5-1 below, health effect categories that will undergo
full evaluation by EPA according to the methods described in Sections 6, 7, 8, and 9 are:
cardiovascular, endocrine, immune, musculoskeletal, and respiratory effects. Health systems with
hazards previously identified by ATSDR 2013 that will not undergo hazard re-evaluation by EPA
but will be considered for dose-response analysis include: developmental, hepatic, neurological,
reproductive, and urinary effects.

This document is a draft for review purposes only and does not constitute Agency policy.

5-2	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Protocol for the Uranium IRIS Assessment (Oral)

Table 5-1. Health effect categories from ATSDR 2013 (ATSDR. 20131 selected
for hazard ID, dose response, or no further consideration

Hazard evaluation

Update ATSDR Toxicological Profile hazard conclusions by
performing new hazard identification for health effect categories,
using studies from both the IRIS literature search and ATSDR 2013.

•	Cardiovascular

•	Endocrine

•	Immune

•	Musculoskeletal

•	Respiratory

Dose-response

Accept ATSDR Toxicological Profile hazard conclusion3 and conduct
dose-response analysis for health effect categories using studies
from both the IRIS literature search and ATSDR 2013.

•	Developmental

•	Hepatic

•	Neurological

•	Reproductive

•	Urinary

No further consideration

Accept ATSDR Toxicological Profile conclusion with no further
consideration for health effect categories.

•	Body weight

•	Gastrointestinal

•	Hematological

•	Metabolic

aFor the purposes of this IRIS Assessment, the evidence for the health effects identified as hazards by ATSDR 2013 were
considered to support an evidence integration judgment of at least "evidence indicates [likely]," as defined in Section 8.

Because of a lack of evidence in epidemiological studies and/or lack of evidence from
experimental studies, EPA will not consider the following health effect categories effects for hazard
evaluation or dose-response (see Table 5-1): body weight, due to new animal studies, the majority
of which reported no effect, and no new epidemiological studies (see Appendix D.I.);
gastrointestinal, due to no new animal studies and two epidemiological studies that did not show a
negative effect (see Appendix D.5.); hematological, due to two animal studies reporting null
evidence and two epidemiological studies with potential limitations (see Appendix D.6.); or
metabolic, due to no new animal studies and only one new epidemiological study that observed an
association (see Appendix D.9.). EPA will continue to monitor the literature and these decisions will
be re-evaluated when the literature search is annually updated.

5.2. REFINEMENTS TO PECO CRITERIA

The problem formulation PECO criteria were refined based on the analysis of the literature
inventory and comparison with the ATSDR Toxicological Profile to develop the assessment PECO
criteria (see Table 5-2 with changes underlined! The assessment PECO criteria focused on the
health systems listed below which EPA determined to have new available data that indicated a need
to revise hazard evaluation conclusions or derive new toxicity values (see Appendix D, and

This document is a draft for review purposes only and does not constitute Agency policy.

5-3	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Protocol for the Uranium IRIS Assessment (Oral)

Table 5-1). The hazards listed from ATSDR 2013 were triaged for evaluation in the IRIS assessment
as follows:

• For the hazards previously identified by ATSDR f20131 (urinary, hepatic, neurological,
reproductive, and developmental), EPA considered the evidence to be sufficient to support
reference value derivation. For the purposes of this IRIS Assessment, the evidence for the health
effects identified as hazards by ATSDR 2013 were considered to support an evidence
integration judgment of at least "evidence indicates [likely]," as defined in Section 8. EPA will
not conduct a de novo hazard synthesize the evidence for these outcomes. EPA will perform
study evaluations (see Section 6) on the studies considered for dose response, based on the
considerations in Section 9, from both the IRIS literature search and studies cited in fATSDR.
20131 (see Table 5-1).

For other health effect categories, if the newly available evidence from PECO-relevant
toxicological and epidemiological studies suggests a need to update hazard conclusions, EPA
will perform a complete evaluation of the studies identified in the IRIS literature search plus the
studies cited in (ATSDR. 2013). In such cases, both new studies and the studies cited in ATSDR
(2013) will be summarized and evaluated jointly using the methods described in Sections 6, 7,
8, and 9 (see Table 5-1).

This document is a draft for review purposes only and does not constitute Agency policy.

5-4	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 5-2. Assessment populations, exposures, comparators, and outcomes
criteria for uranium

PECO element

Evidence

Population

Human: Any population and lifestage (occupational or general population, including children
and other sensitive populations). Note: Case reports and case series will be tracked during
study screening as potentially relevant supplemental material.



Animal: Nonhuman mammalian animal species (whole organism) of any life stage (including
preconception, in utero, lactation, peripubertal, and adult stages).

Exposure

Exposure based on administered dose or concentration, biomonitoring data (e.g., urine, blood,
or other specimens), environmental or occupational-setting measures (e.g., air, water levels),
or job title or residence. Studies on natural uranium and depleted uranium will be included,
studies on enriched uranium or specific to radiation exposure from uranium will not be
included but will be tracked as potentially relevant supplemental information.

Oral exposure will be examined. Other exposure routes, such as those that are clearly dermal,
or inhalation will be tracked during title and abstract screening as "supplemental information."
Animal studies involving exposures to mixtures will be included only if they include an arm with
exposure to uranium alone.

Comparator

Human: A comparison or reference population exposed to lower levels (or no
exposure/exposure below detection levels) of uranium or to uranium for shorter periods. Any
study with a comparison group, control group, or referent group, including:

•	A comparison group that does not have the disease or outcome of interest (such as a
case-control study); or

•	Any study comparing exposed individuals to unexposed or lower-exposed individuals
including:

•	A comparison group with no exposure to the chemical of interest or exposure below
detection limits, or

•	A comparison group exposed to lower levels of the chemical of interest; or

•	A comparison group exposed to the chemical of interest for shorter periods of time;
or

•	Any study assessing the association between a continuous measure of exposure and a
health outcome; or

•	For studies in which humans are intentionally exposed to the chemical of interest, an
individual can serve as their own control.



Animal: A concurrent control group exposed to vehicle-only treatment and/or untreated
control. The control could be a baseline measurement (e.g., acute toxicity studies of mortality)
or a repeated measure design.

Outcomes

Outcomes considered for hazard evaluation by EPA: cardiovascular, endocrine, immune,
musculoskeletal, and respiratory effects. These outcomes may also be considered for dose
response after evidence synthesis and integration (see Sections 8 and 9) Outcomes for which
EPA will rely on ATSDR's hazard conclusions but will be considered for dose-response analysis:
developmental, hepatic, neurological, reproductive, and urinary effects. In general, endpoints
related to clinical diagnostic criteria, disease outcomes, histopathological examination, or
other apical/phenotypic outcomes will be prioritized for evidence synthesis over outcomes
such as biochemical measures.

This document is a draft for review purposes only and does not constitute Agency policy.

5-5	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

Protocol for the Uranium IRIS Assessment (Oral)

5.2.1. Other Exclusions Based on Full-Text Content

In addition to failure to meet PECO criteria (described above), epidemiological and
toxicological studies may be excluded at the full-text level due to critical reporting limitations.
Reporting limitations can be identified during full-text screening but are more commonly identified
during subsequent phases of the assessment (e.g., literature inventory, data extraction, study
evaluation). Regardless of when the limitation is identified, exclusions based on full-text content are
documented at the level of full-text exclusions in literature flow diagrams with a rationale of
"critical reporting limitation." Critical reporting information for different study types are
summarized below. For each piece of information, if the information can be inferred (when not
directly stated) for an exposure/endpoint combination, the study should be included.

Epidemiology studies

Sample size

Exposure characterization and/or measurement method
Outcome ascertainment method
Study design

Animal studies

Species

Test article name

Levels and duration of exposure

Route of exposure

Quantitative or qualitative (e.g., photomicrographs; author-reported lack of an effect on the
outcome) results for at least one endpoint of interest

5.3. UNITS OF ANALYSES FOR DEVELOPING EVIDENCE SYNTHESIS AND
INTEGRATION JUDGMENTS FOR HEALTH EFFECT CATEGORIES

The planned units of analysis based on health systems identified in the assessment PECO
are summarized in Tables 5-3 and 5-4. General considerations for defining the units of analysis are
presented in the IRIS Handbook. For dose-response analysis units of analysis captured in Table 5-3
will be analyzed as described in Section 9. For hazard evaluation each unit of analysis captured in
Table 5-4 is initially synthesized and judged separately within an evidence stream (see Section 8.1).
Depending on the specific health endpoint or outcome, PK data, mechanistic information, and other
supporting evidence (e.g., from studies of non-PECO routes of exposure) may be included in a unit
of analysis.

This document is a draft for review purposes only and does not constitute Agency policy.

5-6	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	The units of analysis can also include or be framed to focus on precursor events (e.g.,

2	biomarkers). Evidence integration judgments focus on the stronger within evidence stream

3	synthesis when multiple units of analysis are synthesized. The evidence synthesis judgments are

4	used alongside other key considerations (i.e., human relevance of findings in animal evidence,

5	coherence across evidence streams, information on susceptible populations or lifestages, and other

6	critical inferences that draw on mechanistic evidence) to draw an overall evidence integration

7	judgment for each health effect category or more granular health outcome grouping (see Section

8	8.2).

Table 5-3. Dose-response: Health effect categories and human and animal
evidence unit of analysis endpoint groupings for dose response

Health effect
categories for dose
response

Units of analysis for dose-response analysis
(each bullet represents a unit of analysis)

Human evidence

Animal evidence

Developmental

•	Pregnancy
outcomes

•	Congenital
malformations

•	Fetal viability/survival or other birth parameters
(e.g., resorptions, number of pups per litter)

•	Fetal/pup growth (e.g., weight or length)

•	Note: An analysis of dam health (e.g., weight gain, food
consumption) is also conducted to support conclusions
of specificity of the effects as being developmental
(versus derivative of maternal toxicity)

Hepatic

• Liver disease

•	Organ weight

•	Clinical measures of liver function (including liver
enzymes)

•	Clinical measures of biliary function

•	Organ morphology/histopathology

Neurological

•	Cognitive
function

•	Brain
disorders

•	Learning/memory

•	Brain morphology/histopathology

•	Neurodegenerative disease

•	Neurotransmitter levels/function

•	Organ weights

Reproductive

• Semen quality

•	Organ morphology/histopathology

•	Developmental measures

•	Reproductive hormone measures

•	Functional measures

Urinary

•	Kidney disease

•	Markers of
kidney
function

•	Urinary and serum markers of renal disease/function

•	Organ weights

•	Organ morphology/histopathology

This document is a draft for review purposes only and does not constitute Agency policy.

5-7	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 5-4. Hazard evaluation: Health effect categories and human and animal
evidence unit of analysis endpoint groupings for hazard evaluation

Health effect
categories for
evidence
integration

Units of analysis for evidence synthesis that inform evidence integration
(each bullet represents a unit of analysis)

Human evidence

Animal evidence

Cardiovascular

•	Cardiovascular disease

•	Blood pressure

•	Blood and arteriole pressure, peripheral
resistance, and other measures of
cardiovascular function

•	Heart and vessel morphology and
histopathology

•	Organ weights

Endocrine

•	Thyroid hormone measures

•	Diabetes

•	Hormone measures

•	Organ morphology/histopathology

•	Organ weights

Immune

•	Autoimmune disease and
measures

•	Immunotoxicity

•	Clinical endpoints (e.g., immune cell
counts/responses)

•	Organ weights

•	Organ morphology/histopathology

•	Immune functional measures

Musculoskeletal

•	Musculoskeletal conditions

•	Muscle and bone health

•	Muscular & skeletal
morphology/histopathology

•	Clinical markers of musculoskeletal disease

•	Parameters/measures of bone
development and function

Respiratory

•	Respiratory disease

•	Pulmonary symptoms

•	Organ weights

•	Organ morphology/histopathology

•	Functional measures

5.4. CONSIDERATIONS OF SUPPLEMENTAL MATERIAL

5.4.1.	Noncancer MOA Mechanistic Information

1	For uranium, evaluating individual mechanistic studies is not anticipated to be critical for

2	this noncancer assessment given the extent of the epidemiological and experimental animal

3	evidence for included outcomes well as the availability of earlier reviews that include mechanistic

4	analyses fMa etal.. 2020: Shaki etal.. 2019: IRIS. 2018: Yue etal.. 20181. For mechanistic

5	information, this assessment will primarily rely on other published sources, such as public health

6	agency reports and expert review articles (see Section 4.6.2).

5.4.2.	ADME and PK/PBPK Model Information

7	Studies containing ADME and PK/PBPK content were screened and tagged as described in

8	Section 4.5. Oral pharmacokinetics of uranium compounds are the primary focus since the current

This document is a draft for review purposes only and does not constitute Agency policy.

5-8	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

Protocol for the Uranium IRIS Assessment (Oral)

assessment focuses on the derivation of oral toxicity values. However, pharmacokinetic studies
from alternate routes of exposure can still inform various aspects of ADME and are also considered.
The ATSDR Toxicological Profile identified two PK/PBPK models for inhalation exposure (ICRP.
1995.19931 and oral exposure (19951: (ATSDR. 20131. These models do not include dosimetric
adjustments from animals to humans, and therefore could not be used for human extrapolation.
The ATSDR Toxicological Profile did not incorporate these models into their dose-response
analysis. Furthermore, no new PK/PBPK models were identified in the date-limited IRIS literature
search. These decisions will be re-evaluated when the literature search is annually updated.

5.4.3. Other Supplemental Material Content

Structured approaches to organize evidence were not developed for the supplemental
material. Instead, the tagged material was reviewed during preparation of the draft to determine
whether the available studies addressed specific uncertainties of the health study evidence base,
inform susceptibility conclusions, and ensure completeness of identifying primary data papers
most pertinent to the assessment.

•	Titles of studies tagged as exposure-only are reviewed to see if they provided information
pertinent to establish study evaluation considerations for the exposure domain.

•	Titles of review articles are reviewed to identify those that are directly pertinent to the scope of
the assessment. The reference lists of such reviews are scanned to identify primary data studies
that might have been missed from database search queries. The reviews may also be used to
provide perspective on interpretation of foundational science cited in the assessment.

•	Other types of supplemental material did not undergo additional analysis because the
information was not considered likely to impact toxicity value development (including
application of uncertainty factors). The specific categories are case reports, enriched uranium,
nonmammalian model systems, mixtures, or conference abstracts.

This document is a draft for review purposes only and does not constitute Agency policy.

5-9	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Protocol for the Uranium IRIS Assessment (Oral)

6. STUDY EVALUATION (RISK OF BIAS AND
SENSITIVITY)

The general approach for evaluating primary health effect studies that meet assessment
PECO criteria is described in Section 6.1. Instructional and informational materials for study
evaluations are available athttps://hawcprd.epa.gov/assessment/100000039/. The approach is
conceptually the same for epidemiology, animal toxicology, and in vitro studies but the application
specifics differ; thus, they are described separately in Sections 6.2, 6.3, and 6.4, respectively. Any
PBPK models used in the assessment are evaluated using methods described in the Quality
Assurance Project Plan for PBPK models fU.S. EPA. 2018b! which is summarized below (see
Section 6.5).

6.1. STUDY EVALUATION OVERVIEW FOR HEALTH EFFECT STUDIES

The IRIS Program uses a domain-based approach to evaluate studies. Key concerns for the
review of epidemiology and animal toxicology studies are potential bias (factors that affect the
magnitude or direction of an effect in either direction) and insensitivity (factors that limit the
ability of a study to detect a true effect; low sensitivity is a bias toward the null when an effect
exists). The study evaluations are aimed at discerning the expected magnitude of any identified
limitations (focusing on limitations that could substantively change a result), considering the
expected direction of the bias. The study evaluation approach is designed to address a range of
study designs, health effects, and chemicals. The general approach for reaching an overall judgment
regarding confidence in the reliability of the results is illustrated in Figure 6-1.

This document is a draft for review purposes only and does not constitute Agency policy.

6-1	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

(a) Individual evaluation domains

Epidemiology

Animal

In vitro

» Exposure measurement

•	Outcome ascertainment

*	Participant selection

•	Confounding

~	Analysis

•	Selective reporting

*	Sensitivity

•	Allocation

•	Observational bias/blinding

•	Confounding

•	Attrition

•	Chemical administration and
characterization

•	Endpoint measurement

•	Results presentation

•	Selective reporting

•	Sensitivity

•	Observational bias/blinding

•	Variable control

•	Selective reporting

•	Chemical administration and
characterization

•	Endpoint measurement

•	Results presentation

•	Sensitivity

(b) Domain level judgements and overall study rating

Domain judgments

Judgment

Interpretation

0 Good
Adequate

Deficient

Q Critically
Deficient

Appropriate study conduct relating to the domain, and minor
deficiencies not expected to influence results.

A study that may have some limitations relating to the
domain, but they are not likely to be severe or to have a
notable impact on results.

Identified biases or deficiencies interpreted as likely to have
had a notable impact on the results or prevent reliable
interpretation of study findings.

A serious flaw identified that makes the observed effect(s)
uninterpretable. Studies with a critical deficiency are
considered "uninformative:: overall.

Overall study rating for an outcome

Rating

Interpretation

High

Medium

Low

Uninformative

No notable deficiencies or concerns identified; potential for

bias unlikely or minimal; sensitive methodology.

Possible deficiencies or concerns noted but they are unlikely

to have a significant impact on results.

Deficiencies or concerns were noted, and the potential for

substantive bias or inadequate sensitivity could have a

significant impact on the study results or their interpretation.

Serious flaw(s) makes study results uninterpretable but may
be used to highlight possible research gaps.

Figure 6-1. Overview of Integrated Risk Information System study evaluation
approach, (a) individual evaluation domains organized by evidence type, and
[b] individual evaluation domains judgments and definitions for overall ratings
(i.e., domain and overall judgments are performed on an outcome-specific basis).

1	To calibrate the assessment-specific considerations, the study evaluation process includes a

2	pilot phase to assess and refine the evaluation process. Following this pilot, at least two reviewers

This document is a draft for review purposes only and does not constitute Agency policy,

6-2	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

Protocol for the Uranium IRIS Assessment (Oral)

independently evaluate studies to identify characteristics that bear on the informativeness
(i.e., validity and sensitivity) of the results. The independent reviewers use structured web-based
forms for study evaluation housed within the EPA's version of HAWC to record separate judgments
for each domain and the overall study for each outcome and unit of analysis, to reach consensus
between reviewers, and when necessary, resolve differences by discussion between the reviewers
or consultation with additional independent reviewers. As reviewers examine a group of studies,
additional chemical-specific knowledge or methodological concerns could emerge, and a second
pass of all pertinent studies might become necessary.

In general, considerations for reviewing a study with regard to its conduct for specific
health outcomes are based on considerations presented in the IRIS Handbook fU.S. EPA. 2022a) and
use of existing guideline documents when available, including EPA guidelines for carcinogenicity,
neurotoxicity, reproductive toxicity, and developmental toxicity fU.S. EPA. 2005a. 1998.1996.

1991).

Authors might be queried to obtain critical information, particularly that involving missing
key study design or results information, or additional analyses that could address potential study
limitations. During study evaluation, the decision on whether to seek missing information focuses
on information that could result in a re-evaluation of the overall study confidence for an outcome.
Any information obtained through personal correspondence with the authors must be made public
to be used in the assessment. If this information cannot be obtained, the study will be rated
Deficient in the "Chemical administration and characterization" domain and Low confidence
overall. Outreach to study authors is documented in HAWC and considered unsuccessful if
researchers do not respond to an email or phone request within 1 month of the attempt to contact.
Only information or data that can be made publicly available (e.g., within HAWC or HERO) will be
considered.

When evaluating studies that examine more than one outcome, the evaluation process is
explicitly conducted at the individual outcome level within the study. Thus, the same study may
have different outcome domain judgments for different outcomes. These measures could still be
grouped for evidence synthesis.

During review, for each evaluation domain, reviewers reach a consensus judgment of good,
adequate¦, deficient, not reported, or critically deficient. If a consensus is not reached, a third
reviewer performs conflict resolution. It is important to emphasize that evaluations are performed
in the context of the study's utility for identifying individual hazards. Limitations specific to the
usability of the study for dose-response analysis are useful to note and applicable to selecting
studies for that purpose (see Section 9), but they do not contribute to the study confidence
classifications. These four categories are applied to each evaluation domain for each outcome
considered within a study, as follows:

This document is a draft for review purposes only and does not constitute Agency policy.

6-3	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

Protocol for the Uranium IRIS Assessment (Oral)

•	Good represents a judgment that the study was conducted appropriately in relation to the
evaluation domain, and any minor deficiencies noted are not expected to influence the study
results or interpretation of the study findings.

•	Adequate indicates a judgment that methodological limitations related to the evaluation domain
are (or are likely to be) present, but those limitations are unlikely to be severe or to notably
impact the study results or interpretation of the study findings.

•	Deficient denotes identified biases or deficiencies interpreted as likely to have had a notable
impact on the results, or that limit interpretation of the study findings.

•	Not reported indicates the information necessary to evaluate the domain question was not
available in the study. Depending on the expected impact, the domain may be interpreted as
adequate or deficient for the purposes of the study confidence rating.

•	Critically deficient reflects a judgment that the study conduct relating to the evaluation domain
introduced a serious flaw that is interpreted to be the primary driver of any observed effect(s)
or makes the study uninterpretable. Studies with critically deficient judgments in any
evaluation domain are almost always classified as overall uninformative for the relevant
outcome (s).

Once the evaluation domains are rated, the identified strengths and limitations are
considered collectively to reach a study confidence classification of high, medium, or low confidence,
or uninformative for each specific health outcome(s). This classification is based on the reviewer
judgments across the evaluation domains and considers the likely impact that the noted
deficiencies in bias and sensitivity have on the outcome-specific results. There are no pre-defined
weights for the domains, and the reviewers are responsible for applying expert judgment to make
this determination. The study confidence classifications, which reflect a consensus judgment
between reviewers, are defined as follows:

1)	High confidence: No notable deficiencies or concerns were identified; the potential for bias
is unlikely or minimal, and the study used sensitive methodology. High confidence studies
generally reflect judgments of good across all or most evaluation domains.

2)	Medium confidence: Possible deficiencies or concerns were identified, but the limitations are
unlikely to have a significant impact on the study results or their interpretation. Generally,
medium confidence studies include adequate or good judgments across most domains, with
the impact of any identified limitation not being judged as severe.

3)	Low confidence: Deficiencies or concerns are identified, and the potential for bias or
inadequate sensitivity is expected to have a significant impact on the study results or their
interpretation. Typically, low confidence studies have a deficient evaluation for one or more
domains, although some medium confidence studies might have a deficient rating in
domain(s) considered to have less influence on the magnitude or direction of effect
estimates. Low confidence results are given less weight compared to high or medium
confidence results during evidence synthesis and integration (see Sections 7 and 8) and are
generally not used as the primary sources of information for hazard identification or
derivation of toxicity values unless they are the only studies available (in which case, this
significant uncertainty would be emphasized during dose-response analysis). Studies rated

This document is a draft for review purposes only and does not constitute Agency policy.

6-4	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

Protocol for the Uranium IRIS Assessment (Oral)

low confidence only because of sensitivity concerns are asterisked or otherwise noted
because they often require additional consideration during evidence synthesis. Effects
observed in studies that are biased toward the null may increase confidence in the results,
assuming the study is otherwise well-conducted (see Section 8).

4) Uninformative: Serious flaw(s) are judged to make the study results uninterpretable for use
in the assessment Studies with critically deficient judgments in any evaluation domain are
almost always rated uninformative. Studies with multiple deficient judgments across
domains may also be considered uninformative. Given that the findings of interest are
considered uninterpretable based on the identified flaws (see above definition of critically
deficient) and do not provide information of use to assessment interpretations, these
studies have no impact on evidence synthesis or integration judgments and are not usable
for dose-response analyses but may be used to highlight research gaps.

As previously noted, study evaluation determinations reached by each reviewer and the
consensus judgment between reviewers are recorded in HAWC. Final study evaluations housed in
HAWC are made available when the draft is publicly released. The study confidence classifications
and their rationales are carried forward and considered as part of evidence synthesis (see Section
8) to help interpret the results across studies.

6.2. EPIDEMIOLOGY STUDY EVALUATION

Evaluation of epidemiology studies of health effects to assess risk of bias and study
sensitivity are conducted for the following domains: exposure measurement, outcome
ascertainment, participant selection, potential confounding, analysis, study sensitivity, and selective
reporting. Bias can result in false positives and negatives, whereas study sensitivity is typically
concerned with identifying the latter.

The principles and framework used for evaluating epidemiology studies are adapted from
the principles in the Cochrane Risk of Bias in Nonrandomized Studies of Interventions [ROBINS-I;
(Sterne etal.. 20161]. modified to address environmental and occupational exposures. Core and
prompting questions, presented in Table 6-1, are used to collect information to guide evaluation of
each domain. Core questions represent key concepts while the prompting questions help the
reviewer focus on relevant details under each key domain. Exposure- and outcome-specific criteria
to use during study evaluation are developed using the core and prompting questions and refined
during a pilot phase with engagement from topic-specific experts.

This document is a draft for review purposes only and does not constitute Agency policy.

6-5	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 6-1. Questions to guide the development of criteria for each domain in epidemiology studies

Domain and core
question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes

Exposure
measurement

Does the exposure
measure reliably
distinguish
between levels of
exposure in a time
window

considered most
relevant for a
causal effect with
respect to the
development of
the outcome?

For all:

•	Does the exposure measure capture
the variability in exposure among the
participants, considering intensity,
frequency, and duration of exposure?

•	Does the exposure measure reflect a
relevant time window? If not, can the
relationship between measures in this
time and the relevant time window be
estimated reliably?

•	Was the exposure measurement likely
to be affected by knowledge of the
outcome?

•	Was the exposure measurement likely
to be affected by the presence of the
outcome (i.e., reverse causality)?

For case-control studies of occupational
exposures:

•	Is exposure based on a comprehensive
job history describing tasks, setting,
period, and use of specific materials?

For biomarkers of exposure, general population:

•	Is a standard assay used? What are the
intra- and interassay coefficients of
variation? Is the assay likely to be
affected by contamination? Are values
less than the limit of detection dealt
with adequately?

•	What exposure period is reflected by
the biomarker? If the half-life is short,
what is the correlation between serial
measurements of exposure?

Is the degree of
exposure
misclassifi cation
likely to vary by
exposure level?

If the correlation
between exposure
measurements is
moderate, is there an
adequate statistical
approach to
ameliorate variability
in measurements?

If potential for bias is
a concern, is the
predicted direction
or distortion of the
bias on the effect
estimate (if there is
enough
information)?

Good

•	Valid exposure assessment methods used, which
represent the etiologically relevant period of interest.

•	Exposure misclassification is expected to be minimal.

Adequate

•	Valid exposure assessment methods used, which
represent the etiologically relevant period of interest.

•	Exposure misclassification could exist but is not
expected to greatly change the effect estimate.

Deficient

•	Valid exposure assessment methods used, which
represent the etiologically relevant time period of
interest. Specific knowledge about the exposure and
outcome raises concerns about reverse causality, but
whether it is influencing the effect estimate is
uncertain.

•	Exposed groups are expected to contain a notable
proportion of unexposed or minimally exposed
individuals, the method did not capture important
temporal or spatial variation, or other evidence of
exposure misclassification would be expected to
notably change the effect estimate.

Critically deficient

•	Exposure measurement does not characterize the
etiologically relevant period of exposure or is not valid.

•	Evidence exists that reverse causality is very likely to
account for the observed association.

•	Exposure measurement was not independent of
outcome status.

This document is a draft for review purposes only and does not constitute Agency policy.

6-6	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core







question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes

Outcome

For all:

Is there a concern

Good

ascertainment

• Is outcome ascertainment likely

that any outcome

• High certainty in the outcome definition (i.e., specificity

Does the outcome

affected by knowledge, or presence, of

misclassification is

and sensitivity), minimal concerns with respect to

measure reliably

exposure (e.g., consider access to

nondifferential,

misclassification.

distinguish the

healthcare, if based on self-reported

differential, or both?

• Assessment instrument was validated in a population

presence or

history of diagnosis)?

What is the

comparable to the one from which the study group was

absence (or degree

For case-control studies:

predicted direction

selected.

of severity) of the
outcome?

• Is the comparison group without the

or distortion of the
bias on the effect

Adequate



outcome (e.g., controls in a

estimate (if there is

enough

information)?

• Moderate confidence that outcome definition was



case-control study) based on objective

specific and sensitive, some uncertainty with respect to



criteria with little or no likelihood of

misclassification but not expected to greatly change



inclusion of people with the disease?

the effect estimate.



For mortality measures:



• Assessment instrument was validated but not



• How well does cause-of-death data



necessarily in a population comparable to the study



reflect occurrence of the disease in an



group.



individual? How well do mortality data



Deficient



reflect incidence of the disease?



• Outcome definition was not specific or sensitive.



For diagnosis of disease measures:



• Uncertainty regarding validity of assessment



• Is the diagnosis based on standard



instrument.



clinical criteria? If it is based on



• Critically deficient



self-report of the diagnosis, what is the



• Invalid/insensitive marker of outcome.



validity of this measure?



• Outcome ascertainment is very likely to be affected by



For laboratory-based measures (e.g., hormone



knowledge of, or presence of, exposure.



levels):



Note: Lack of blinding should not be automatically construed to





be critically deficient.



• Is a standard assay used? Does the





assay have an acceptable level of







interassay variability? Is the sensitivity







of the assay appropriate for the







outcome measure in this study







population?





This document is a draft for review purposes only and does not constitute Agency policy.

6-7	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Participant
selection

Is there evidence
that selection into
or out of the study
(or analysis
sample) was jointly
related to
exposure and to
outcome?

For longitudinal cohort:

•	Did participants volunteer for the
cohort on the basis of knowledge of
exposure or preclinical disease
symptoms? Was entry into, or
continuation in, the cohort related to
exposure and outcome?

For occupational cohort:

•	Did entry into the cohort begin with
the start of the exposure?

•	Was follow-up or outcome assessment
incomplete, and if so, was follow-up
related to both exposure and outcome
status?

•	Could exposure produce symptoms
that would result in a change in work
assignment/work status ("healthy
worker survivor effect")?

For case-control study:

•	Were controls representative of
population and periods from which
cases were drawn?

•	Are hospital controls selected from a
group whose reason for admission is
independent of exposure?

•	Could recruitment strategies, eligibility
criteria, or participation rates result in
differential participation relating to
both disease and exposure?

For population-based survey:

•	Was recruitment based on
advertisement to people with
knowledge of exposure, outcome, and
hypothesis?

Were differences in
participant
enrollment and
follow-up evaluated
to assess bias?

If potential for bias is
a concern, what is
the predicted
direction or
distortion of the bias
on the effect
estimate (if there is
enough
information)?

Were appropriate
analyses performed
to address changing
exposures over time
relative to
symptoms?

Is there a comparison
of participants and
nonparticipants to
address whether
differential selection
or study

retention/continuati
on is likely?

Good

•	Minimal concern for selection bias based on
description of recruitment process and follow-up
(e.g., selection of comparison population,
population-based random sample selection,
recruitment from sampling frame including current and
previous employees).

•	Exclusion and inclusion criteria specified and would not
induce bias.

•	Participation rate is reported at all steps of study
(e.g., initial enrollment, follow-up, selection into
analysis sample). If rate is not high, appropriate
rationale is given for why it is unlikely to be related to
exposure (e.g., comparison between participants and
nonparticipants or other available information
indicates differential selection is not likely).

Adequate

•	Enough of a description of the recruitment process to
be comfortable that there is no serious risk of bias.

•	Inclusion and exclusion criteria specified and would not
induce bias.

•	Participation rate is incompletely reported but
available information indicates participation is unlikely
to be related to exposure.

Deficient

•	Little information on recruitment process, selection
strategy, sampling framework, and participation OR
aspects of these processes raises the potential for bias
(e.g., healthy worker effect, survivor bias).

Critically deficient

•	Aspects of the processes for recruitment, selection
strategy, sampling framework, or participation result in
concern that selection bias is likely to have had a large
impact on effect estimates (e.g., convenience sample
with no information about recruitment and selection,

This document is a draft for review purposes only and does not constitute Agency policy.

6-8	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core
question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes







cases and controls are recruited from different sources
with different likelihood of exposure, recruitment
materials stated outcome of interest and potential
participants are aware of or are concerned about
specific exposures).

Confounding
Is confounding of
the effect of the
exposure likely?

Is confounding adequately addressed by
considerations in:

•	Participant selection (matching or
restriction)?

•	Accurate information on potential
confounders and statistical adjustment
procedures?

•	Lack of association between
confounder and outcome, or
confounder and exposure in the study?

•	Information from other sources?

Is the assessment of confounders based on a
thoughtful review of published literature,
potential relationships (e.g., as can be gained
through directed acyclic graphing), and
minimizing potential overcontrol (e.g., inclusion
of a variable on the pathway between exposure
and outcome)?

If potential for bias is
a concern, what is
the predicted
direction or
distortion of the bias
on the effect
estimate (if there is
enough
information)?

Good

•	Conveys strategy for identifying key confounders,
including co-exposures. This may include a priori
biological consideration, published literature, causal
diagrams, or statistical analyses, with the recognition
that not all "risk factors" are confounders.

•	Inclusion of potential confounders in statistical models
not based solely on statistical significance criteria
(e.g., p < 0.05 from stepwise regression).

•	Does not include variables in the models likely to be
influential colliders or intermediates on the causal
pathway.

•	Key confounders are evaluated appropriately and
considered unlikely sources of substantial confounding.
This often will include:

o Presenting the distribution of potential
confounders by levels of the exposure of
interest or the outcomes of interest (with
amount of missing data noted);
o Consideration that potential confounders
were rare among the study population, or
were expected to be poorly correlated with
exposure of interest;
o Consideration of the most relevant functional

forms of potential confounders;
o Examination of the potential impact of
measurement error or missing data on
confounder adjustment; or

This document is a draft for review purposes only and does not constitute Agency policy.

6-9	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core
question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes







o Presenting a progression of model results with
adjustments for different potential
confounders, if warranted.

Adequate

•	Similar to good but might not have included all key
confounders, or less detail might be available on the
evaluation of confounders (e.g., sub bullets in good).
That residual confounding could explain part of the
observed effect is possible, but concern is minimal.

Deficient

•	Does not include variables in the models shown to be
influential colliders or intermediates on the causal
pathway.

•	And any of the following:

o The potential for bias to explain some results
is high based on an inability to rule out
residual confounding, such as a lack of
demonstration that key confounders of the
exposure-outcome relationships were
considered;

o Descriptive information on key confounders
(e.g., their relationship relative to the
outcomes and exposure levels) are not
presented; or
o Strategy of evaluating confounding is unclear
or is not recommended (e.g., only based on
statistical significance criteria or stepwise
regression [forward or backward elimination]).

Critically deficient

•	Includes variables in the models that are colliders or
intermediates in the causal pathway, indicating that
substantial bias is likely from this adjustment; or

This document is a draft for review purposes only and does not constitute Agency policy.

6-10	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core
question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes







• Confounding is likely present and not accounted for,
indicating that all results were most likely due to bias.

Analysis

Does the analysis
strategy and
presentation
convey the
necessary
familiarity with the
data and
assumptions?

•	Are missing outcome, exposure, and
covariate data recognized, and if
necessary, accounted for in the
analysis?

•	Does the analysis appropriately
consider variable distributions and
modeling assumptions?

•	Does the analysis appropriately
consider subgroups or lifestages of
interest (e.g., based on variability in
exposure level or duration or
susceptibility)?

•	Is an appropriate analysis used for the
study design?

•	Is effect modification considered,
based on considerations developed a
priori?

•	Does the study include additional
analyses addressing potential biases or
limitations (i.e., sensitivity analyses)?

If potential for bias is
a concern, what is
the predicted
direction or
distortion of the bias
on the effect
estimate (if there is
enough
information)?

Good

•	Use of an optimal characterization of the outcome
variable, including presentation of subgroup- or
lifestage-specific comparisons (as appropriate for the
outcome).

•	Quantitative results presented (effect estimates and
confidence limits or variability in estimates) (i.e., not
presented only as a p-value or "significant"/"not
significant").

•	Descriptive information about outcome and exposure
provided (where applicable).

•	Amount of missing data noted and addressed
appropriately (discussion of selection issues—missing
at random vs. differential).

•	Where applicable, for exposure, includes LOD (and
percentage below the LOD), and decision to use log
transformation.

•	Includes analyses that address robustness of findings,
e.g., examination of exposure-response (explicit
consideration of nonlinear possibilities, quadratic,
spline, or threshold/ceiling effects included, when
feasible); relevant sensitivity analyses; effect
modification examined based only on a priori rationale
with sufficient numbers.

•	No deficiencies in analysis evident. Discussion of some
details might be absent (e.g., examination of outliers).

Adequate

•	Same as "Good," except:

•	Descriptive information about exposure provided
(where applicable) but might be incomplete; might not

This document is a draft for review purposes only and does not constitute Agency policy.

6-11	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core
question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes







have discussed missing data, cut points, or shape of
distribution(s).

•	Includes analyses that address robustness of findings
(examples in 'Good'), but some important analyses are
not performed.

Deficient

•	Does not conduct analysis using optimal
characterization of the outcome variable.

•	Descriptive information about exposure levels not
provided (where applicable).

•	Effect estimates and p-value presented, without
standard error or confidence interval.

•	Results presented as statistically "significant"/"not
significant."

Critically deficient

•	Analysis methods are not appropriate for design or
data of the study.

Selective reporting
Is there reason to
be concerned
about selective
reporting?

•	Were results provided for all the
primary analyses described in the
methods section?

•	Is appropriate justification given for
restricting the amount and type of
results shown?

•	Are only statistically significant results
presented?

If potential for bias is
a concern, what is
the predicted
direction or
distortion of the bias
on the effect
estimate (if there is
enough
information)?

Good

•	The results reported by study authors are consistent
with the primary and secondary analyses described in a
registered protocol or methods paper.

Adequate

•	The authors described their primary (and secondary)
analyses in the methods section and results were
reported for all primary analyses.

Deficient

•	Concerns were raised based on previous publications, a
methods paper, or a registered protocol indicating that
analyses were planned or conducted that were not
reported, or that hypotheses originally considered to

This document is a draft for review purposes only and does not constitute Agency policy.

6-12	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core
question

Prompting questions

Follow-up questions

Criteria that apply to most exposures and outcomes







be secondary were represented as primary in the
reviewed paper.

•	Only subgroup analyses were reported, suggesting that
results for the entire group were omitted.

•	Only statistically significant results were reported.

Sensitivity

Is there a concern
that sensitivity of
the study is not
adequate to detect
an effect?

•	Is the exposure contrast adequate to
detect associations and
exposure-response relationships?

•	Was the appropriate population or
lifestage included?

•	Was the length of follow-up adequate?
Is the time/age of outcome
ascertainment optimal given the
interval of exposure and the health
outcome?

•	Do other aspects related to risk of bias
or otherwise raise concerns about
sensitivity?



Good

•	There is sufficient variability/contrast in exposure to
evaluate primary hypotheses.

•	The study population was sensitive to the development
of the outcomes of interest (e.g., ages, lifestage, sex).

•	The timing of outcome ascertainment was appropriate
given expected latency for outcome development
(i.e., adequate follow-up interval).

•	The study was adequately powered to observe an
effect.

•	No other concerns raised regarding study sensitivity.

Adequate

•	Same considerations as Good, except:

•	There may be issues identified that could reduce
sensitivity, but they are considered unlikely to
substantially impact the overall findings of the study.

Deficient

•	Concerns were raised about the considerations
described for Good that are expected to notably
decrease the sensitivity of the study to detect
associations for the outcome.

Critically deficient

•	Severe concerns were raised about the sensitivity of
the study such that any observed associations are likely
to be explained by bias.

This document is a draft for review purposes only and does not constitute Agency policy.

6-13	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

6.3. EXPERIMENTAL ANIMAL STUDY EVALUATION

1	Using the principles described in Section 6.1, the identified animal studies are evaluated for

2	the following domains to assess risk of bias and sensitivity: allocation, observational bias/blinding,

3	confounding, selective reporting, attrition, chemical administration and characterization, endpoint

4	measurement and validity, results presentation and comparisons, and sensitivity (see Table 6-2).

5	The rationale for judgments is documented at the outcome level. The evaluation

6	documentation in HAWC includes the identified limitations and their expected impact on the overall

7	confidence level. To the extent possible, the rationale will reflect an interpretation of the potential

8	influence on the outcome-specific results, including the direction or magnitude of influence

9	(or both).

This document is a draft for review purposes only and does not constitute Agency policy.

6-14	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 6-2. Questions to guide the development of criteria for each domain in experimental animal toxicology
studies

Domain and core question

Prompting questions

General considerations

Allocation

Were animals assigned to
experimental groups using a
method that minimizes
selection bias?

For each study:

Did each animal or litter have an
equal chance of being assigned to any
experimental group (i.e., random
allocation)?

Is the allocation method described?
Aside from randomization, were any
steps taken to balance variables
across experimental groups during
allocation?

These considerations typically do not need to be refined by assessment teams.

A judgment and rationale for this domain should be given for each cohort or experiment

in the study.

Good

•	Experimental groups were randomized, and any specific randomization
procedure was described or inferable (e.g., computer-generated scheme. Note
that normalization is not the same as randomization [see response for
adequate]).

Adequate

•	Authors report that groups were randomized but do not describe the specific
procedure used (e.g.," animals were randomized"). Alternatively, authors used
a nonrandom method to control for important modifying factors across
experimental groups (e.g., body-weight normalization).

Not reported

•	(Interpreted as deficient): No indication of randomization of groups or other
methods (e.g., normalization) to control for important modifying factors across
experimental groups.

Critically deficient

•	Bias in the animal allocations was reported or inferable.

Observational bias/blinding

Did the study implement
measures to reduce
observational bias?

For each endpoint/outcome or
grouping of endpoints/outcomes in a
study:

Does the study report blinding or
other procedures for reducing
observational bias?

If not, did the study use a design or
approach for which such procedures
can be inferred?

These considerations typically do not need to be refined by the assessment teams.
(Note that it can be useful for teams to identify highly subjective measures of
endpoints/outcomes where observational bias may strongly influence results prior to
performing evaluations.)

A judgment and rationale for this domain should be given for each endpoint/outcome
or group of endpoints/outcomes investigated in the study.

This document is a draft for review purposes only and does not constitute Agency policy.

6-15	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations



What is the expected impact of failure
to implement (or report
implementation) of these procedures
on results?

Good

•	Measures to reduce observational bias were described (e.g., blinding to
conceal treatment groups during endpoint evaluation; consensus-based
evaluations of histopathology-lesions).a

Adequate

•	Methods for reducing observational bias (e.g., blinding) can be inferred or were
reported but described incompletely.

Not reported

•	Measures to reduce observational bias were not described.

Interpreted

•	(Interpreted as adequate) The potential concern for bias was mitigated based
on use of automated/computer driven systems, standard laboratory kits,
relatively simple, objective measures (e.g., body or tissue weight), or
screening-level evaluations of histopathology.

•	(Interpreted as deficient) The potential impact on the results is major
(e.g., outcome measures are highly subjective).

Critically deficient

•	Strong evidence for observational bias that impacted the results.

Confounding

Are variables with the
potential to confound or
modify results controlled
for and consistent across
experimental groups?

Note:

Consideration of overt
toxicity (possibly masking
more specific effects) is
addressed under endpoint
measurement reliability.

For each study:

Are there differences across the
treatment groups, considering both
differences related to the exposure
(e.g., co-exposures, vehicle, diet,
palatability) and other aspects of the
study design or animal groups (e.g.,
animal source, husbandry, or health
status), that could bias the results?
If differences are identified, to what
extent are they expected, based on a
specific scientific understanding, to
impact the results?

These considerations may need to be refined by assessment teams, as the specific
variables of concern can vary by experiment or chemical.

A judgment and rationale for this domain should be given for each cohort or experiment
in the study, noting when the potential for confounding is restricted to specific
endpoints/outcomes.

Good

•	Outside of the exposure of interest, variables that are likely to confound or
modify results appear to be controlled for and consistent across experimental
groups.

Adequate

•	Some concern that variables that were likely to confound or modify results
were uncontrolled or inconsistent across groups but are expected to have a
minimal impact on the results.

This document is a draft for review purposes only and does not constitute Agency policy.

6-16	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations





Deficient

•	Notable concern that potentially confounding variables were uncontrolled or
inconsistent across groups and are expected based on to substantially impact
the results.

Critically deficient

•	Confounding variables were presumed to be uncontrolled or inconsistent
across groups and are expected to be a primary driver of the results.

Attrition

Did the study report the
results for all tested
animals?

For each study:

Are all animals accounted for in the
results?

If there is attrition, do authors
provide an explanation (e.g., death or
unscheduled sacrifice during the
study)?

If unexplained attrition of animals for
outcome assessment is identified,
what is the expected impact on the
interpretation of the results?

These considerations typically do not need to be refined by assessment teams.

A judgment and rationale for this domain should be given for each cohort or experiment

in the study.

Good

•	Results were reported for all animals. If animal attrition is identified, the
authors provide an explanation, and these are not expected to impact the
interpretation of the results.

Adequate

•	Results are reported for most animals. Attrition is not explained but this is not
expected to significantly impact the interpretation of the results.

Deficient

•	Moderate to high level of animal attrition that is not explained and may
significantly impact the interpretation of the results.

Critically deficient

•	Extensive animal attrition that prevents comparisons of results across
treatment groups.

Chemical administration
and characterization

Did the study adequately
characterize exposure to
the chemical of interest and
the exposure administration
methods?

For each study:

Are there concerns [specific to this
chemical] regarding the source and
purity and/or composition
(e.g., identity and percent distribution
of different isomers) of the chemical?
Was independent analytical
verification of the test article (e.g.,

It is essential that these considerations are considered, and potentially refined, by
assessment teams, as the specific variables of concern can vary by chemical
(e.g., stability may be an issue for one chemical but not another).

A judgment and rationale for this domain should be given for each cohort or experiment
in the study.

This document is a draft for review purposes only and does not constitute Agency policy.

6-17	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations

Note:

Consideration of the
appropriateness of the
route of exposure (not the
administration method) is
not a risk of bias
consideration. Relevance
and utility of the routes of
exposure are considered in
the PECO criteria for study
inclusion and during
evidence synthesis.
Relatedly, consideration of
exposure level selection
(e.g., were levels sufficiently
high to elicit effects) is
addressed during evidence
synthesis and is not a risk of
bias consideration.

composition, homogeneity, and
purity) performed?

Were nominal exposure levels verified
analytically? Are there concerns about
the methods used to administer the
chemical (e.g., inhalation chamber
type, gavage volume)?

Good

•	Chemical administration and characterization are complete (i.e., source and
purity are provided or can be obtained from the supplier and test article is
analytically verified). There are no notable concerns about the composition,
stability, or purity of the administered chemical, or the specific methods of
administration. Exposure levels are verified using reliable analytical methods.

Adequate

•	Some uncertainties in the chemical administration and characterization are
identified but these are expected to have minimal impact on interpretation of
the results (e.g., purity of the test article is suboptimal but interpreted as
unlikely to have a significant impact; analytical verification of exposure levels is
not reported or verified with nonpreferred methods).

Deficient

•	Uncertainties in the exposure characterization are identified and expected to
substantially impact the results (e.g., source of the test article is not reported,
and composition is not independently verified; impurities are substantial or
concerning; administration methods are considered likely to introduce
confounders, such as use of static inhalation chambers or a gavage volume
considered too large for the species or lifestage at exposure).

Critically deficient

•	Uncertainties in the exposure characterization are identified and there is
reasonable certainty that the study results are largely attributable to factors
other than exposure to the chemical of interest (e.g., identified impurities are
expected to be a primary driver of the results).

Endpoint measurement

Are the selected
procedures, protocols, and
animal models adequately
described and appropriate
for the

endpoint(s)/outcome(s) of
interest?

Notes:

For each endpoint/outcome or
grouping of endpoints/outcomes in a
study:

Are the evaluation methods and

animal model adequately described

and appropriate?

Are there concerns regarding the

methodology selected for endpoint

evaluation?

Considerations for this domain are highly variable depending on the
endpoint(s)/outcome(s) of interest and typically must be refined by assessment teams.

A judgment and rationale for this domain should be given for each endpoint/outcome
or group of endpoints/outcomes investigated in the study.

Some considerations include the following:

Good

• Adequate description of methods and animal models.

This document is a draft for review purposes only and does not constitute Agency policy.

6-18	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations

Considerations related to
the sensitivity of the animal
model and timing of
endpoint measurement are
evaluated under Sensitivity
Considerations related to
adjustments/corrections to
endpoint measurements
(e.g., organ weight
corrected for body weight)
are addressed under results
presentation.

Are there concerns about the
specificity of the experimental
design?

Are there serious concerns regarding
the sample size or how endpoints
were sampled?

Are appropriate control groups for
the study/assay type included?

•	Use of generally accepted and reliable endpoint methods.

•	Sample sizes are generally considered adequate for the assay or protocol of
interest and there are no notable concerns about sampling in the context of
the endpoint protocol (e.g., sampling procedures for histological analysis).

•	Includes appropriate control groups and any use of nonconcurrent or historical
control data (e.g., for evaluation of rare tumors) is justified (e.g., authors or
evaluators considered the similarity between current experimental animals
and laboratory conditions to historical controls).

Ratings of Adequate, Deficient, and Critically Deficient are generally defined as follows:

Adequate

•	Issues are identified that may affect endpoint measurement but are considered
unlikely to substantially impact the overall findings or the ability to reliably
interpret those findings.

Deficient

•	Concerns are raised that are expected to notably affect endpoint measurement
and reduce the reliability of the study findings.

Critically deficient

•	Severe concerns are raised about endpoint measurement and any findings are
likely to be largely explained by these limitations.

The following specific examples of relevant concerns are typically associated with a

Deficient rating, but Adequate or Critically Deficient might be applied depending on

the expected impact of limitations on the reliability and interpretation of the results:

•	Study report lacks important details that are necessary to evaluate the
appropriateness of the study design (e.g., description of the assays or
protocols; information on the strain, sex, or lifestage of the animals).

•	Selection of protocols that are nonpreferred or lack specificity for investigating
the endpoint of interest. This includes omission of additional experimental
criteria (e.g., inclusion of a positive control or dosing up to levels causing
minimal toxicity) when required by specific testing guidelines/protocols.*

•	Over toxicity (e.g., mortality, extreme weight loss) is observed or expected
based on findings from similarly designed studies and may mask interpretation
of outcome(s) of interest.

This document is a draft for review purposes only and does not constitute Agency policy.

6-19	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations





•	Sample sizes are smaller than is generally considered adequate for the assay or
protocol of interest. Inadequate sampling can also be raised within the context
of the endpoint protocol (e.g., in a pathology study, bias that is introduced by
only sampling a single tissue depth or an inadequate number of slides per
animal).

•	Control groups are not included, considered inappropriate, or comparisons to
non-concurrent or historical controls are not adequately justified.

'These limitations typically also raise a concern for insensitivity

"Sample size alone is not a reason to conclude an individual study is critically deficient.

Results presentation

Are the results presented
and compared in a way that
is appropriate and
transparent?

For each endpoint/outcome or
grouping of endpoints/outcomes in a
study:

Does the level of detail allow for an
informed interpretation of the
results?

Are the data compared, or presented,
in a way that is inappropriate or
misleading?

Considerations for this domain are highly variable depending on the outcomes of

interest and typically must be refined by assessment teams.

A judgment and rationale for this domain should be given for each endpoint/outcome

or group of endpoints/outcomes investigated in the study.

Some considerations include the following:

Good

•	No concerns with how the data are presented.

•	Results are quantified or otherwise presented in a manner that allows for an
independent consideration of the data (assessments do not rely on author
interpretations).

•	No concerns with completeness of the results reporting.*

Ratings of Adequate, Deficient, and Critically Deficient are generally defined as follows:

Adequate

•	Concerns are identified that may affect results presentation but are considered
unlikely to substantially impact the overall findings or the ability to reliably
interpret those findings.

Deficient

•	Concerns with results presentation are identified and expected to substantially
impact results interpretation and reduce the reliability of the study findings.

Critically deficient

•	Severe concerns about results presentation were identified and study findings
are likely to be largely explained by these limitations.

This document is a draft for review purposes only and does not constitute Agency policy.

6-20	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations





The following specific examples of relevant concerns are typically associated with a
Deficient rating but Adequate or Critically Deficient might be applied depending on
expected impact of limitations on the reliability and interpretation of the results:

•	Nonpreferred presentation of data (e.g., developmental toxicity data averaged
across pups in a treatment group, when litter responses are more appropriate;
presentation of only absolute organ weight data when relative weights are
more appropriate).

•	Pooling data when responses are known or expected to differ substantially
(e.g., across sexes or ages).

•	Incomplete presentation of the data* (e.g., presentation of mean without
variance data; concurrent control data are not presented; dichotomizing or
truncating continuous data).

*Failure to describe any findings for assessed outcomes (i.e., report lacks any qualitative
or quantitative description of the results in tables, figures, or text) is addressed under
Selective Reporting.

Selective reporting

Did the study report the
results for all prespecified
outcomes?

Note:

This domain does not
consider the
appropriateness of the
analysis/results
presentation. This aspect of
study quality is evaluated in
another domain.

For each study:

Are results presented for all
endpoints/outcomes described in the
methods (see note)?

If unexplained results omissions are
identified, what is the expected
impact on the interpretation of the
results?

These considerations typically do not need to be refined by assessment teams.

A judgment and rationale for this domain should be given for each cohort or experiment

in the study.

Good

•	Quantitative or qualitative results were reported for all prespecified outcomes
(explicitly stated or inferred), exposure groups and evaluation time points.

Data not reported in the primary article is available from supplemental
material. If results omissions are identified, the authors provide an
explanation, and these are not expected to impact the interpretation of the
results.

Adequate

•	Quantitative or qualitative results are reported for most prespecified outcomes
(explicitly stated or inferred) and evaluation time points. Omissions and are not
explained but are not expected to significantly impact the interpretation of the
results.

This document is a draft for review purposes only and does not constitute Agency policy.

6-21	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations





Deficient

•	Quantitative or qualitative results are missing for many prespecified outcomes
(explicitly stated or inferred), omissions are not explained and may significantly
impact the interpretation of the results.

Critically deficient

•	Extensive results omission is identified and prevents comparisons of results
across treatment groups.

Sensitivity

Are there concerns that
sensitivity in the study is not
adequate to detect an
effect?

Note:

Consideration of exposure
level selection (e.g., were
levels sufficiently high to
elicit effects) is addressed
during evidence synthesis
and is not a study sensitivity
consideration.

Was the exposure period, timing (e.g.,
lifestage), frequency, and duration
sensitive for the outcome(s) of
interest?

Given knowledge of the health hazard
of concern, did the selection of
species, strain, and/or sex of the
animal model reduce study
sensitivity?

Are there concerns regarding the
timing (e.g., lifestage) of the outcome
evaluation?

Are there aspects related to risk of
bias domains that raise concerns
about insensitivity (e.g., selection of
protocols that are known to be
insensitive or nonspecific for the
outcome(s) of interest)

These considerations may require customization to the specific exposure and outcomes.
Some study design features that affect study sensitivity may have already been included
in the other evaluation domains; these should be noted in this domain, along with any
features that have not been addressed elsewhere. Some considerations include:

Good

•	The experimental design (considering exposure period, timing, frequency, and
duration) is appropriate and sensitive for evaluating the outcome(s) of interest.

•	The selected animal model (considering species, strain, sex, and/or lifestage) is
known or assumed to be appropriate and sensitive for evaluating the
outcome(s) of interest.

•	No significant concerns with the ability of the experimental design to detect
the specific outcome(s) of interest, (e.g., outcomes evaluated at the
appropriate lifestage; study designed to address known endpoint variability
that is unrelated to treatment, such as estrous cyclicity or time of day).

•	Timing of endpoint measurement in relation to the chemical exposure is
appropriate and sensitive (e.g., behavioral testing is not performed during a
transient period of test chemical-induced depressant or irritant effects;
endpoint testing does not occur only after a prolonged period, such as weeks
or months, of non-exposure)

•	Potential sources of bias toward the null are not a substantial concern.
Adequate

Same considerations as Good, except:

•	The duration and frequency of the exposure was appropriate, and the
exposure covered most of the critical window (if known) for the outcome(s) of
interest.

This document is a draft for review purposes only and does not constitute Agency policy.

6-22	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Domain and core question

Prompting questions

General considerations





•	Potential issues are identified that could reduce sensitivity, but they are
unlikely to impact the overall findings of the study.

Deficient

•	Concerns were raised about the considerations described for Good or
Adequate that are expected to notably decrease the sensitivity of the study to
detect a response in the exposed group(s).

Critically deficient

•	Severe concerns were raised about the sensitivity of the study and
experimental design such that any observed associations are likely to be
explained by bias. The rationale should indicate the specific concern(s).

Overall confidence

Considering the identified
strengths and limitations,
what is the overall
confidence rating for the
endpoint(s)/outcome(s) of
interest?

For each endpoint/outcome or
grouping of endpoints/outcomes in a
study:

Were concerns (i.e., limitations or
uncertainties) related to the risk of
bias or sensitivity identified?

If yes, what is their expected impact
on the overall interpretation of the
reliability and validity of the study
results, including (when possible)
interpretations of impacts on the
magnitude or direction of the
reported effects?

The overall confidence rating considers the likely impact of the noted concerns
(i.e., limitations or uncertainties) in reporting, bias, and sensitivity on the results.

Reviewers should mark studies that are rated lower than high confidence only due to
low sensitivity (i.e., bias toward the null) for additional consideration during evidence
synthesis. If the study is otherwise well conducted and an effect is observed, it may
increase the strength of evidence judgment.

A confidence rating and rationale should be given for each endpoint/outcome or group
of endpoints/outcomes investigated in the study. Confidence ratings are described
above (see Section 6.1).

aFor nontargeted or screening-level histopathological outcomes often used in guideline studies, blinding during the initial evaluation of tissues is generally not recommended, as
masked evaluation can make "the task of separating treatment-related changes from normal variation more difficult" and "there is concern that masked review during the
initial evaluation may result in missing subtle lesions." Generally, blinded evaluations are recommended for targeted secondary review of specific tissues or in instances when
there is a predefined set of outcomes that is known or predicted to occur (Crissman et al.. 2004).

This document is a draft for review purposes only and does not constitute Agency policy.

6-23	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

Protocol for the Uranium IRIS Assessment (Oral)

6.4.	MECHANISTIC AND OTHER NON-PECO STUDY EVALUATION

As described in Sections 4.4, 4.5, and 4.6, the initial literature screening identifies sets of
other potentially informative studies, including mechanistic studies, as potentially relevant
supplemental information that do not meet the assessment PECO criteria. The approach for the
prioritization and evaluation of mechanistic and other non-PECO studies is targeted to the
assessment needs, depending on the extent and nature of the human and animal evidence. An
intensive analysis may not be warranted for health outcomes or specific mechanistic events not
expected to meaningfully impact assessment approaches or conclusions or for those already well
accepted scientifically. Given the literature inventory and findings from the ATSDR assessment used
as a starting point for the IRIS assessment, evaluating individual mechanistic studies is not
anticipated to be impactful for most, if not all, health effects identified for review for this
assessment As described in Section 5.4, this assessment will primarily rely on other published
authoritative sources, such as public health agency reports and literature reviews, to summarize
the available mechanistic information (when such context aids the evidence synthesis narrative)
unless substantial scientific issues or new, impactful studies are identified during the course of
developing the assessment.

6.5.	PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODEL
DESCRIPTIVE SUMMARY AND EVALUATION

PBPK (or classical pharmacokinetic [PK]) models should be used in an assessment when an
applicable one exists and no equal or better alternative for dosimetric extrapolation is available.
Any models used should represent current scientific knowledge and accurately translate the
science into computational code in a reproducible, transparent manner. For a specific target
organ/tissue, it may be possible to employ or adapt an existing PBPK model or develop a new PBPK
model or an alternate quantitative approach. Data for PBPK models may come from studies across
various species and may be in vitro or in vivo in design.

Note that the terms "pharmacokinetic" (adjective) and "pharmacokinetics" (noun), which
are both abbreviated as "PK," are used in this document when discussing absorption, distribution,
metabolism, and excretion (ADME) of a substance by an organism or any related quantities,
experiments, or models. The terms "toxicokinetic" and "toxicokinetics," which are both abbreviated
as "TK," are frequently used as synonyms for "pharmacokinetic" and "pharmacokinetics" in the
literature, but the latter terms are used preferentially here for document-wide consistency. Also,
PBPK models are sometimes described as "physiologically based toxicokinetic models"

(abbreviated "PBTK models") or even as "physiologically based kinetic models" (abbreviated "PBK
models") in the literature, but in this document the term "PBPK model" is used preferentially for
purposes of consistency.

This document is a draft for review purposes only and does not constitute Agency policy.

6-24	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	As described in Section 5.4.2, the ATSDR Toxicological Profile identified two PK/PBPK

2	models for inhalation and oral exposures, but these models do not include a dosimetric adjustments

3	from animals to humans and were not considered further. No PBPK models for uranium have been

4	identified in the preliminary survey of the date-limited literature search.

This document is a draft for review purposes only and does not constitute Agency policy.

6-25	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

Protocol for the Uranium IRIS Assessment (Oral)

7. DATA EXTRACTION OF STUDY METHODS AND
RESULTS

The process of summarizing study methods and results is referred to as data extraction.
Studies that met initial PECO criteria after full-text review are briefly summarized in data extraction
forms available in the Distiller and serve as a literature inventory. These study summaries are
exported from DistillerSR in Excel format to create interactive literature inventory used for analysis
of the available evidence. For experimental animal studies, which are typically studies in rodents,
the following information is captured: chemical form, study type (acute [<24 hours], short term
[<7 days], short term [7-27 days], subchronic [28-90 days], chronic [>90 days] and developmental,
which includes multigeneration studies), duration of treatment, route, species, strain, sex, dose or
concentration levels tested, dose units, health system and specific endpoints assessed, and the no-
observed-effect level/low-observed-effect level (NOEL/LOEL) based on author-reported statistical
significance. Expert judgment may be used to identify NOEL/LOELs in cases where only qualitative
results are reported (e.g., "no effects on liver weight were observed at any dose level") or when the
findings indicate an apparent clear and strong effect of exposure (e.g., large magnitude of change)
but the authors did not present a statistical comparison. When findings are not analyzed by the
authors and are not readily interpretable, then NOEL/LOELs are not identified, and the extraction
field entry indicates "not reported." For human studies, the following information is summarized in
DistillerSR forms: chemical form, population type (e.g., general population-adult, occupational,
pregnant women, infants and children), study type (e.g., cross-sectional, cohort, case-control), sex,
major route of exposure (if known), description of how exposure was assessed, health system
studied, specific endpoints assessed and a quantitative summary of findings at the endpoint level
(or narrative only if the finding was qualitatively presented).

For epidemiology and animal studies that met the assessment PECO criteria, the HAWC is
used for study evaluation and for full extraction of study methods and results. Compared with the
literature inventory, full data extraction in HAWC includes summarizing more details of study
design and gathering effect size information. For animal studies, compared with the literature
inventory forms used to described studies that meet problem formulation PECO criteria, full data
extraction in HAWC includes summarizing more details of study design (e.g., diet, chemical purity)
and gathering effect size information. Instructions on how to conduct data extraction in HAWC are
available at https: //hawcproiect.org/resources /. An additional resource used to implement use of a
consistent vocabulary to summarize endpoints assessed in animal studies is available in HAWC (the
Environmental Health Vocabulary (EHV); https: //hawc.epa.gov/vocab/ehv/.

This document is a draft for review purposes only and does not constitute Agency policy.

7-1	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

Protocol for the Uranium IRIS Assessment (Oral)

In some cases, EPA may conduct their own statistical analysis of human and animal
toxicology data (assuming the data are amenable to doing so and the study is otherwise well-
conducted) during evidence synthesis.

If necessary, data extraction for mechanistic studies (including in vivo and in vitro studies)
will be conducted in Distiller SR or Microsoft Excel and presented in tabular format The extracted
evidence is available in MS Excel format upon request See
https: //www.epa.gov/iris/forms/contact-us-about-iris for requests.

All findings are considered for extraction, regardless of statistical significance. The level of
extraction for specific outcomes within a study could differ (i.e., narrative only if the finding was
qualitative). For quality control, studies were extracted by one member of the evaluation team and
independently verified by at least one other member. Discrepancies were resolved by discussion or
consultation within the evaluation team. Data extraction results are presented via figures, tables, or
interactive web-based graphics in the assessment. The information is also made available for
download in Excel format when the draft is publicly released. Download of full data extraction for
animal studies is done directly from HAWC.

For non-English studies online translation tools (e.g., Google translator) or engagement with
a native speaker can be used to summarize studies at the level of the literature inventory. Fee-based
translation services for non-English studies are typically reserved for studies considered potentially
informative for dose response, a consideration that occurs after preparation of the initial literature
inventory during draft assessment development. Digital rulers, such as WebPlotDigitizer
fhttp://arohatgi.info/WebPlotDigitizer/). are used to extract numerical information from figures,
and their use is be documented during extraction. For studies that evaluate endpoints at multiple
time points (e.g., 7 days, 3 weeks, 3 months) data are generally summarized for the longest duration
in the study report, but other durations may be summarized if they provide important contextual
information for hazard characterization (e.g., an effect was present at an interim time point but did
not appear to persist or the magnitude of the effect diminished). A free text field is available in
HAWC to describe cases when the approach for summarizing results requires explanation.

Author queries may be conducted for studies considered for hazard identification or dose-
response to facilitate study evaluation and quantitative analysis (e.g., information on variability or
availability of individual animal data). Outreach to study authors or designated contact persons is
documented and considered unsuccessful if researchers do not respond to email or phone requests
within 1 month of initial attempt(s) to contact Only information or data that can be made publicly
available (e.g., within HAWC or HERO) will be considered.

Exposures are standardized to common units when possible. For hazard characterization,
exposure levels are typically presented as reported in the study and standardized to common units.

This document is a draft for review purposes only and does not constitute Agency policy.

7-2	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

7.1. STANDARDIZING ADMINISTERED DOSE LEVELS/CONCENTRATIONS

1	Exposures are standardized to common units. Exposure levels in oral studies are expressed

2	in units of mg uranium/kg-day. When study authors provide exposure levels in concentrations in

3	the diet or drinking water, dose conversions are made using study-specific food or water

4	consumption rates and body weights when available. Otherwise, EPA defaults are used fU.S. EPA.

5	19881. addressing age and study duration as relevant for the species/strain and sex of the animal of

6	interest Exposure levels are converted to uranium equivalents. For example, doses administered as

7	uranyl nitrate are expressed as uranium using a molecular weight conversion. Unless otherwise

8	reported by study authors, the background level in experimental animal studies is assumed to be

9	0 ppm (0 mg/kg-day).

This document is a draft for review purposes only and does not constitute Agency policy.

7-3	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

Protocol for the Uranium IRIS Assessment (Oral)

8. EVIDENCE SYNTHESIS AND INTEGRATION	

As described in Sections 5.1 and 5.2 if the newly available evidence from PECO-relevant
toxicological and epidemiological studies suggests a need to update hazard conclusions, EPA will
perform a complete evaluation of the studies identified in the IRIS literature search plus the studies
cited in fATSDR. 20131.12 Within-stream evidence synthesis is conducted separately for human,
animal, and mechanistic evidence to directly inform the integration across the streams of evidence
and draw overall conclusions for each of the assessed human health effects. The phrases "evidence
synthesis" and "evidence integration" used here are analogous to the phrases "strength of evidence"
and "weight of evidence," respectively, used in some other assessment processes fEFSA. 2017: U.S.
EPA. 2017: NRC. 2014: U.S. EPA. 2005al. A structured framework approach is used to guide both
evidence synthesis and integration. This structured framework includes consideration of
mechanistic information during both evidence synthesis and integration, although the focus of the
analysis differs. Similarly other types of supplemental information (e.g., ADME, non-PECO route of
exposure) can also inform evidence synthesis and integration analyses.

•	Evidence synthesis: Judgment(s) regarding the strength of the evidence for hazard for each unit
of analysis from the available human and animal studies are made in parallel, but separately.
These judgments can incorporate PK, mechanistic, and other supplemental evidence when the
unit of analysis is defined as such (see Section 5.2). The units of analysis can also include or be
framed to focus on precursor events (e.g., biomarkers). In addition, this includes an evaluation
of coherence across units of analysis within an evidence stream. At this stage, the animal
evidence judgment(s) does not yet consider the human relevance of that evidence.

•	Evidence integration: The animal and human evidence judgments are combined to draw an
overall evidence integration judgment(s) that incorporates inferences drawn based on
information on the human relevance of the animal evidence, coherence across evidence
streams, potential susceptibility, and other critical inferences (e.g., biological plausibility)
informed by mechanistic, ADME, or other supplemental data.

Evidence synthesis and integration judgments are expressed both narratively in the
assessment and summarized in tabular format in evidence profile tables (see Table 8-1). Key
findings and analyses of mechanistic and other supplemental content are also summarized in
narrative and tabular format to inform evidence synthesis and integration judgments (see
Table 8-2). In brief, a synthesis (strength of evidence) judgment is drawn for each unit of analysis
summarized as robust, moderate, slight, indeterminate, or compelling evidence of no effect (see

12Health systems that will undergo full evaluation by EPA: cardiovascular (see Appendix D.2), endocrine (see
Appendix D.4), immune (see Appendix D.8), musculoskeletal (see Appendix D.10), and respiratory (see
Appendix D.13).

This document is a draft for review purposes only and does not constitute Agency policy.

8-1	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	Section 8.1). Next, evidence synthesis judgments are used to inform evidence integration (weight of

2	evidence) judgments summarized as evidence demonstrates¦, evidence indicates¦, evidence suggests¦,

3	evidence inadequate, or strong evidence supports no effect) (see Section 8.2). These summary

4	judgments are included as part of the evidence synthesis and integration narratives. When multiple

5	units of analysis are synthesized, the main evidence integration judgments13 typically focus on the

6	unit of analysis with the strongest evidence synthesis judgments, although exceptions may occur.

7	Structured evidence profile tables are used to summarize these analyses and foster consistency

8	within and across assessments. Instructions for using HAWC to create these tables are available at

9	the HAWC project "IRIS PPRTV SEM Template Figures and Resources" (see "Attachments," then
10	select the "Creating Evidence Profile Tables in HAWC")

13In some cases, as discussed in Section 8.2, it will be appropriate to draw multiple evidence integration
judgments within a given health effect category. This is generally dependent on data availability (i.e., more
narrowly defined categories may be possible with more evidence) and the ability to integrate the different
evidence streams at the level of these more granular categories. More granular categories will generally be
organized by pre-defined manifestations of potential toxicity. For example, within the health effect category
of immune effects, separate and different evidence integration judgments might be appropriate for
immunosuppression, immunostimulation, and sensitization and allergic response (i.e., the three types of
immunotoxicity described in the 2012 WHO Guidance for immunotoxicity risk assessment for chemicals
(WHO. 201211 Likewise, within the category of developmental effects, it may be appropriate to draw separate
judgments for potential effects on fetal death, structural abnormality, altered growth, and functional deficits
(i.e., the four manifestations of developmental toxicity described in EPA guidelines (U.S. EPA. 1991)). These
separate judgments are particularly important when the evidence supports that the different manifestations
might be based on different toxicological mechanisms. As described for the evidence synthesis judgments, the
strongest evidence integration judgment will typically be used to reflect certainty in the broader health effect
category.

This document is a draft for review purposes only and does not constitute Agency policy.

8-2	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 8-1. Generalized evidence profile table to show the relationship between evidence synthesis and evidence
integration to reach judgment of the evidence for hazard











Evidence integration



Evidence synthesis (strength of evidence) judgments



(weight of evidence)

(note that many factors and judgments require elaboration or evidence-based justification; see IRIS Handbook for details)

judgment(s)





Factors that increase





Describe overall evidence





certainty





integration judgment(s):



Summary of key

(applied to each unit of

Factors that decrease certainty

Evidence synthesis



Studies

findings

analysis)

(applied to each unit of analysis)

judgment(s)

©©© Evidence demonstrates

Evidence from human studies







©©O Evidence indicates (likely)











©OO Evidence suggests
OOO Evidence inadequate

	Strong evidence supports no

effect

Unit of analysis #1

Studies considered
and study
confidence

Description of
the primary
results

•	All/Mostly medium
or high confidence
studies

•	Consistency

•	All/Mostly low
confidence studies

•	Unexplained
inconsistency

Judgment reached for
each unit of analysis3

©0© Robust

©©O Moderate





•	Dose-response
gradient

•	Large or concerning
magnitude of effect

•	Coherence3

•	Imprecision

•	Concerns about
biological significance3

•	Indirect outcome
measures3

•	Lack of expected
coherence3



Unit of analysis #2
Studies considered
and study
confidence

Description of
the primary
results

©OO Slight
OOO Indeterminate

	Compelling

evidence of no effect

Highlight the primary supporting
evidence for each integration
judgment3

Present inferences and conclusions
on:

• Human relevance of

Evidence from animal studies

findings in animals3

Unit of analysis #1

Description of

• All/Mostly medium

• All/Mostly low

Judgment reached for

• Cross-stream coherence3

Studies considered

the primary

or high confidence

confidence studies

each unit of analysis

• Potential susceptibility3

and study

results

studies

• Unexplained

©©© Robust

• Understanding of

confidence



• Consistency

inconsistency

©©O Moderate

biological plausibility and





•	Dose-response
gradient

•	Large or concerning
magnitude of effect

•	Coherence3

•	Imprecision

•	Concerns about
biological significance3

•	Indirect outcome
measures3

•	Lack of expected
coherence3

MOA3

• Other critical inferences3

Unit of analysis #2
Studies considered
and study
confidence

Description of
the primary
results

©OO Slight
OOO Indeterminate

	Compelling

evidence of no effect

aCan be informed by key findings from the mechanistic analyses (see Table 8-2).

This document is a draft for review purposes only and does not constitute Agency policy.

8-3	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 8-2. Generalized evidence profile table to show the key findings and supporting rationale from mechanistic
analyses

Mechanistic analyses

Biological events or pathways (or
other relevant evidence grouping)

Summary of key findings and interpretation

Judgment(s) and rationale

Different analyses can be presented
separately, e.g., bv exposure route
or key uncertainty addressed.

Each analysis can include multiple
rows separated bv biological events
or other feature of the approach
used for the analysis

•	Generally, will cite
mechanistic synthesis
(e.g., for references; for
detailed analysis).

•	Does not have to be
chemical-specific
(e.g., read-across).

Can include separate summaries, for example bv studv
tvpe (e.g., new approach methods vs. in vivo
biomarkers), dose, or design.

Interpretation: Summary of expert interpretation for
the body of evidence and supporting rationale.

Key findings: Summary of findings across the body of
evidence (may focus on or emphasize highly
informative designs or findings), including key sources
of uncertainty or identified limitations of the study
designs tested (e.g., regarding the biological event or
pathway being examined).

Overall summary of expert interpretation across the assessed
set of biological events, potential mechanisms of toxicity, or
other analysis approach (e.g., AOP).

•	Includes the primary evidence supporting the
interpretation(s).

•	Describes and informs the extent to which the
evidence influences inferences across evidence
streams.

•	Characterizes the limitations of the evaluation and
highlights existing data gaps.

•	May have overlap with factors summarized for other
streams.

This document is a draft for review purposes only and does not constitute Agency policy.

8-4	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

Protocol for the Uranium IRIS Assessment (Oral)

8.1. EVIDENCE SYNTHESIS

IRIS assessments synthesize the evidence separately for each unit of analysis by focusing on
factors that increase or decrease certainty in the reported findings as evidence for hazard (see
Table 8-1). These factors are adapted from considerations for causality introduced by Austin
Bradford Hill (Hill. 19651 with some expansion and adaptation of how they are applied to facilitate
transparent application to chemical assessments that consider multiple streams of evidence.
Specifically, the factors considered are confidence in study findings (risk of bias [RoB] and
sensitivity), consistency across studies or experiments, dose/exposure-response gradient, strength
(effect magnitude) of the association, directness of outcome or endpoint measures, and coherence
[see Table 8-3; see additional discussion in (U.S. EPA. 2022a. 2005a. 1994)]. These factors are
similar to the domains considered in the GRADE (Grading of Recommendations Assessment,
Development, and Evaluation) Quality of Evidence framework (Schunemann et al.. 2013). Each of
the considered factors and the certainty of evidence judgments requires elaboration or evidence-
based justification in the synthesis narrative. Analysis of evidence synthesis considerations is
qualitative (i.e., numerical scores are not developed, summed, or subtracted).

As previously described, the units of analysis may include predefined categories of
mechanistic evidence or other supplemental information (e.g., from studies of non-PECO routes of
exposure). This may include consideration of biomarkers or precursor events. Biological
understanding (e.g., knowledge of how an effect is manifest or progresses) or mechanistic inference
(e.g., dependency on a conserved key event across outcomes) can also be used to define which
related outcomes are considered as a unit of analysis. These considerations also inform the
evaluation of coherence and adversity within a unit of analysis and coherence with other units of
analyses. Mechanistic analyses outside the context of defining and evaluating the units of analysis
during evidence synthesis are considered as part of across stream evidence integration (see
Section 8.2).

Typically, human and animal evidence synthesis sections are structured similarly across
different units of analysis, health effects, and assessments. In contrast, the presentation, and
analyses of mechanistic and other types of supplemental information often differs within and
across assessments. This is due to the diversity of supplemental data that may be available and the
complexity of conducting supplemental analyses. For example, these data may inform unit of
analysis considerations, evidence integration judgments, or both. Each of the key analyses
informing the synthesis judgments are described in the narrative and summarized in an evidence
profile table.

Five levels of certainty in the evidence for (or against) a hazard are used to summarize
evidence synthesis judgments: robust (©©©, very little uncertainty exists), moderate (©©O,
some uncertainty exists), slight (©OO, large uncertainty exists), indeterminate (OOO), or
compelling evidence of no effect (—, little to no uncertainty exists for lack of hazard) (see

This document is a draft for review purposes only and does not constitute Agency policy.

8-5	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Protocol for the Uranium IRIS Assessment (Oral)

Tables 8-3 and 8-4 for descriptions). Conceptually, before the evidence synthesis framework is
applied, certainty in the evidence is neutral (i.e., functionally equivalent to indeterminate). Next, the
level of certainty regarding the evidence for (or against) hazard is increased or decreased
depending on interpretations using the factors described in Table 8-3. Observations that increase
certainty are having consistency across high or medium confidence studies or experiments, the
presence of medium or high confidence studies with a strong dose-response gradient or observing a
large or concerning magnitude of effect, and coherent findings across medium or high confidence
studies for closely related endpoints (can include mechanistic endpoints) within the unit of analysis
within an evidence stream. Evidence from low confidence studies can further strengthen
observations from medium or high confidence studies but do not increase certainty on their own.
Observations that decrease certainty are having an evidence base of mostly low confidence studies,
unexplained inconsistency, lack of expected coherence, imprecision, unclear biological significance,
null findings with concerns for insensitivity (which decreases certainty in the lack of an effect), or
indirect measures of outcomes. Table 8-3 provides additional detail on how these factors are
considered when evaluating units of analysis.

This document is a draft for review purposes only and does not constitute Agency policy.

8-6	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 8-3. Considerations that inform evaluations and judgments of the strength of the evidence for hazard

Consideration

Increased evidence certainty
(of the human or animal evidence for hazard3)

Decreased evidence certainty
(of the human or animal evidence for hazard3)

Risk of bias and
sensitivity (across
studies)

•	An evidence base of mostly (or all) high or
medium confidence studies is interpreted as
being only minimally affected by bias and
insensitivity.

•	This factor should not be used if no other
factors would increase or decrease the
confidence for a given unit of analysis.

•	In addition, consideration of risk of bias and
sensitivity should inform how other factors are
evaluated, i.e., can inconsistency be potentially
explained by variation in confidence judgments?

•	An evidence base of mostly (or all) low confidence studies decreases
strength. An exception to this is an evidence base of studies in which
the issues resulting in low confidence are related to insensitivity. This
may increase evidence certainty in cases where an association is
identified because the expected impact of study insensitivity is
toward the null.

•	An evidence base of mostly null findings where insensitivity is a
serious concern decreases certainty that the evidence is sufficient to
support a lack of health effect or association.

•	Decisions to increase certainty for other considerations in this table
should generally not be made if there are serious concerns for risk
of bias.

Consistency

• Similarity of findings for a given outcome

(e.g., of a similar direction) across independent
studies or experiments, especially when
medium or high confidence, increases certainty.
The increase in certainty is larger when
consistency is observed across populations (e.g.,
geographical location) or exposure scenarios in
human studies, and across laboratories, species,
or exposure scenarios (e.g., route; timing) in
animal studies. When seemingly inconsistent
findings are identified, patterns should be
further analyzed to discern if the inconsistencies
can potentially be explained based on study
confidence, dose or exposure levels, population,
or experimental model differences, etc. This
factor is typically given the most attention
during evidence synthesis.

• Unexplained inconsistency [i.e., conflicting evidence; see (U.S. EPA,
2005a)l decreases certainty. Generally, certainty should not be
decreased if discrepant findings can be reasonably explained by
considerations such as study confidence conclusions (including
sensitivity); variation in population or species, sex, or lifestage
(including understanding of differences in pharmacokinetics); or
exposure patterns (e.g., intermittent versus continuous), levels (low
versus high), or duration. Similar to current recommendations in the
Cochrane Handbook [(Higgins et al., 2022), see Section 7.8.61, clear
conflicts of interest (COI) related to funding source can be considered
as a factor to explain apparent inconsistency. For small evidence
bases, it might be hard to assess consistency. An evidence base of a
single or a few studies where consistency cannot be accurately
assessed does not, alone, increase or decrease evidence certainty.
Similarly, a reasonable explanation for inconsistency does not
necessarily result in an increase in evidence certainty.

This document is a draft for review purposes only and does not constitute Agency policy.

8-7	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Consideration

Increased evidence certainty
(of the human or animal evidence for hazard3)

Decreased evidence certainty
(of the human or animal evidence for hazard3)

Effect magnitude and
imprecision

•	Evidence of a large or concerning magnitude of
effect can increase strength (generally only
when observed in medium or high confidence
studies).

•	Judgments on effect magnitude and imprecision
consider the rarity and severity of the effect.

•	Certainty could be decreased if the findings are considered not likely
to be biologically significant. Effects that are small in magnitude
might not be considered biologically significant (adverseb) based on
information such as historical responses and variability. However,
effects that appear to be of small magnitude could be meaningful at
the population level e.g., IQ shifts); in such cases, certainty would not
be decreased.

•	Certainty might also be decreased for imprecision, particularly if
there are only a few studies available to evaluate consistency in
effect magnitude across studies.

Dose-response

•	Evidence of dose-response or exposure-
response in high or medium confidence studies
increases certainty. Dose-response can be
demonstrated across studies or within studies
and it can be dose- or duration-dependent. It
could also not be a monotonic dose-response
(monotonicity should not necessarily be
expected as different outcomes might be
expected at low vs. high doses due to factors
such as activation of different mechanistic
pathways, systemic toxicity at high doses or
tolerance/acclimation). Sometimes, grouping
studies by level of exposure is helpful to identify
the dose-response pattern.

•	Decreases in a response (e.g., symptoms of
current asthma) after a documented cessation
of exposure also might increase certainty in a
relationship between exposure and outcome
(this is primarily applicable to epidemiology
studies because of their observational nature).

•	A lack of dose-response when expected on the basis of biological
understanding can decrease certainty in the evidence. If the data are
not adequate to evaluate a dose-response pattern, however,
certainty is neither increased nor decreased.

•	In some cases, duration-dependent patterns in the dose-response can
decrease evidence certainty. Such patterns are generally only
observable in experimental studies. Specifically, the magnitude of
effects at a given exposure level might decrease with longer
exposures (e.g., due to tolerance or acclimation). Or effects might
rapidly resolve under certain experimental conditions (e.g.,
reversibility after removal of exposure). As many reversible and
short-lived effects can be of high concern, decisions about whether
such patterns decrease evidence certainty depend on considering the
pharmacokinetics of the chemical and the conditions of exposure [see
U.S. EPA (1998)], endpoint severity, judgments regarding the
potential for delayed or secondary effects, the underlying
mechanism(s) involved, and the exposure context focus of the
assessment (e.g., addressing intermittent or short-term exposures).

Directness of

outcome/endpoint

measures

• Not applicable

• If the evidence base primarily includes outcomes or endpoints that
are indirect measures (e.g., biomarkers) of the unit of analysis,
certainty (for that unit of analysis) is typically decreased. Judgments
to decrease certainty based on indirectness should focus on findings

This document is a draft for review purposes only and does not constitute Agency policy.

8-8	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Consideration

Increased evidence certainty
(of the human or animal evidence for hazard3)

Decreased evidence certainty
(of the human or animal evidence for hazard3)





for measures that have an unclear linkage to an apical or clinical
(adverseb) outcome. Scenarios where the magnitude of the response
is not considered to reflect a biologically meaningful level of change
(i.e., biological significance; see "effect magnitude and imprecision"
row, above) are not considered under indirectness of outcome
measures.

• Related to indirectness, certainty in the evidence can be decreased
when the findings are determined to be nonspecific to the hazard
under evaluation. This consideration is generally only applicable to
animal evidence and the most common example is effects only with
exposures (level, duration) shown to cause excessive toxicity in that
species and lifestage (including consideration of maternal toxicity in
developmental evaluations). This does not apply when an effect is
viewed as secondary to other changes (e.g., effects on pulmonary
function because of disrupted immune responses).

Coherence

•	Biologically related findings within or across
studies, within an organ system or across
populations (e.g., sex), increase certainty
(generally only when observed in medium or
high confidence studies). Certainty is further
increased when a temporal or dose-dependent
progression of related effects is observed within
or across studies, or when related findings of
increasing severity are observed with increasing
exposure.

•	Coherence across findings within a unit of
analysis (e.g., consistent changes in disease
markers and biological precursors in exposed
humans) can increase certainty in the evidence
for an effect.

•	Coherence within or across biologically related
units of analysis can also increase certainty for a
given (or multiple) unit(s) of analysis. This
considers certainty in the biological

• An observed lack of expected coherent changes (e.g., in well-
established biological relationships) within or across biologically
related units of analysis will typically decrease evidence certainty.

This includes mechanistic changes when included in the unit of
analysis. However, as described for decisions to increase certainty,
confidence in the understanding of the biological relationships
between the endpoints being compared, and the sensitivity and
specificity of the measures used, need to be carefully examined. The
decision to decrease certainty depends on the availability of evidence
across multiple related endpoints for which changes would be
anticipated, and it considers factors (e.g., dose and duration of
exposure, strength of expected relationship) across the studies of
related changes.

This document is a draft for review purposes only and does not constitute Agency policy.

8-9	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Consideration

Increased evidence certainty
(of the human or animal evidence for hazard3)

Decreased evidence certainty
(of the human or animal evidence for hazard3)



relationships between the endpoints being
compared, and the sensitivity and specificity of
the measures used.

• Mechanistic support for, or biological

understanding of, the relatedness between
different endpoints within (or across different)
units of analysis, can inform an understanding of
coherence.



Other factors

• Unusual scenarios that cannot be addressed by
the considerations above, e.g., read-across
inferences supporting the adversity of observed
changes.

• Unusual scenarios that cannot be addressed by the considerations
above, e.g., strong evidence of publication bias.c

aAlthough the focus is on identifying potential adverse human health effects (hazards) of exposure, these factors can also be used to increase or decrease certainty in the
evidence supporting lack of an effect (e.g., leading to a judgment of compelling evidence of no effect). The latter application is not explicitly outlined here.
bWithin this framework, evidence synthesis judgments reflect an interpretation of the evidence for a hazard; thus, consideration of the adversity of the findings is an explicit
aspect of the analyses. To better define how adversity is evaluated, the consideration of adversity is broken into the two, sometimes related, considerations of the indirectness
of the outcome measures and the interpreted biological significance of the effect magnitude.

Publication bias involves the influence of the direction, magnitude, or statistical significance of the results on the likelihood of a paper being published; it can result from
decisions made, consciously or unconsciously, by study authors, journal reviewers, and journal editors (Dickersin. 1990). This could make the available evidence base
unrepresentative. However, publication bias can be difficult to evaluate (NTP. 2019) and should not be used as a factor that decreases certainty unless there is strong evidence.

This document is a draft for review purposes only and does not constitute Agency policy.

8-10	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

21

22

23

24

25

Protocol for the Uranium IRIS Assessment (Oral)

A structured framework approach is used to draw evidence synthesis judgments for human
and animal evidence. Tables 8-4 and 8-5 (for human and animal evidence, respectively) provide the
criteria that guide how to draw the strength of evidence judgments for each unit of analysis within
a health effect category and the terms used to summarize those judgments. These terms are applied
to human and animal evidence separately. The terms robust and moderate are characterizations for
judgments that the evidence (across studies) supports a conclusion that the effect(s) results from
the exposure being assessed. These two terms are differentiated by the quality and amount of
information available to rule out alternative explanations for the results. For example, repeated
observations of effects by independent studies or experiments examining various aspects of
exposure or response (e.g., different exposure settings, dose levels or patterns, populations or
species, biologically related endpoints) result in increased certainty in the evidence for hazard. The
term slight indicates situations in which there is some evidence supporting an association within
the evidence stream, but substantial uncertainties in the data exist to prevent judgments that the
effect(s) can be reliably attributed to the exposure being assessed. Indeterminate reflects judgments
for a wide variety of evidence scenarios, including when no studies are available or when the
evidence from studies of similar confidence has a high degree of unexplained inconsistency.
Compelling evidence of no effect represents a rare situation in which extensive evidence across a
range of populations and exposures has demonstrated that no effects are likely attributable to the
exposure being assessed. This category is applied at the health effect level (e.g., hepatic effects)
rather than more granular units of analysis level to avoid giving the impression of confidence in
lack of a health effect when aspects of potential toxicity have not been adequately examined.
Reaching this judgment is infrequent because it requires both a high degree of confidence in the
conduct of individual studies, including consideration of study sensitivity, as well as comprehensive
assessments of outcomes and lifestages of exposure that adequately address concern for the hazard
under evaluation.

This document is a draft for review purposes only and does not constitute Agency policy.

8-11	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 8-4. Framework for strength of evidence judgments from studies in
humans

Evidence synthesis
judgment

Description

Robust (©©©)

...evidence in
human studies

(strong signal of
effect with very
little uncertainty)

A set of high or medium confidence independent studies (e.g., in different populations)
reporting an association between the exposure and the health outcome(s), with reasonable
confidence that alternative explanations, including chance, bias, and confounding, can be
ruled out across studies. The set of studies is primarily consistent, with reasonable
explanations when results differ; the findings are considered adverse (i.e., biologically
significant and without notable concern for indirectness); and an exposure-response
gradient is demonstrated. Additional supporting evidence, such as associations with
biologically related endpoints in human studies (coherence) or large estimates of risk or
severity of the response, can increase certainty but are not required. Supplemental
evidence included in the unit of analysis (e.g., mechanistic studies in exposed humans or
human cells) could raise the certainty in the evidence to robust for a set of studies that
otherwise would be described as moderate. Such evidence not included in the unit of
analysis can also inform evaluations of the coherence of the human evidence, the directness
of the outcome measures, and the biological significance of the findings. Causality is
inferred for a human evidence base of robust.

Moderate

(©©O)

...evidence in
human studies

(signal of effect
with some
uncertainty)

A set of evidence that does not reach the degree of certainty required for robust, but which
includes at least one high or medium confidence study reporting an association and
additional information increasing certainty in the evidence. For multiple studies, there is
primarily consistent evidence of an association with reasonable support for adversity, but
there might be some uncertainty due to potential chance, bias, or confounding or because
of the indirectness of some measures. When only a single study is available in the unit of
analysis, there is a large magnitude or severity of the effect, or a dose-response gradient, or
other supporting evidence, and there are no serious residual methodological uncertainties.
Supplemental evidence included in the unit of analysis might address the above factors and
raise certainty in the evidence to moderate for a set of studies that otherwise would be
described as slight or, in exceptional cases, could support raising to moderate evidence that
would otherwise be described as indeterminate. Mechanistic evidence not included in the
unit of analysis can also inform evaluations of the coherence of the human evidence, the
directness of the outcome measures, and the biological significance of the findings.

Slight

(©OO)

...evidence in
human studies

(signal of effect
with large amount
of uncertainty)

One or more studies reporting an association between exposure and the health outcome,
but considerable uncertainty exists and supporting coherent evidence is sparse. In general,
the evidence is limited to a set of consistent low confidence studies, or higher confidence
studies with significant unexplained heterogeneity or other serious residual uncertainties. It
also applies when one medium or high confidence study is available within the unit of
analysis without additional information strengthening the likelihood of a causal association
(e.g., coherent findings within the same study or from other studies). This category serves
primarily to encourage additional study where evidence does exist that might provide some
support for an association, but for which the evidence does not reach the degree of
confidence required for moderate.

This document is a draft for review purposes only and does not constitute Agency policy.

8-12	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Evidence synthesis
judgment

Description

Indeterminate

(OOO)

...evidence in
human studies

(signal cannot be
determined for or
against an effect)

No studies available in humans or situations when the evidence is inconsistent and primarily
of low confidence. In addition, this might include situations where higher confidence studies
exist, but there are major concerns with the evidence base such as unexplained
inconsistency, a lack of expected coherence from a stronger set of studies, very small effect
magnitude (i.e., major concerns about biological significance), or uncertainties or
methodological limitations that result in an inability to discern effects from exposure. It also
applies for a single low confidence study in the absence of factors that increase certainty. A
set of largely null studies could be concluded to be indeterminate if the evidence does not
reach the level required for compelling evidence of no effect.

Compelling
evidence of no
effect
(...)

...in human studies

(strong signal for
lack of an effect
with little
uncertainty)

A set of high confidence studies examining a reasonable spectrum of endpoints showing
null results (e.g., an odds ratio of 1.0), ruling out alternative explanations including chance,
bias, and confounding with reasonable confidence. Each of the studies should have used an
optimal outcome and exposure assessment and adequate sample size (specifically for
higher exposure groups and for susceptible populations). The set as a whole should include
diverse sampling (across sexes [if applicable] and different populations) and include the full
range of levels of exposures that human beings are known to encounter, an evaluation of an
exposure-response gradient, and an examination of at-risk populations and lifestages.
Supplemental evidence can help to address the above considerations or, when included in
the unit of analysis, provide additional support for this judgment.

Table 8-5. Framework for strength of evidence judgments from studies in
animals

Evidence synthesis
judgment

Description

Robust (©©©)

...evidence in
animal studies

(strong signal of
effect with very
little uncertainty)

The set of high or medium confidence, independent experiments (i.e., across laboratories,
exposure routes, experimental designs [for example, a subchronic study and a
multigenerational study], or species) reporting effects of exposure on the health
outcome(s). The set of studies is primarily consistent, with reasonable explanations when
results differ (i.e., due to differences in study design, exposure level, animal model, or study
confidence), and the findings are considered adverse (i.e., biologically significant and
without notable concern for indirectness). At least two of the following additional factors in
the set of experiments increase certainty in the evidence: coherent effects across multiple
related endpoints (within or across biologically related units of analysis); an unusual
magnitude of effect, rarity, age at onset, or severity; a strong dose-response relationship; or
consistent observations across animal lifestages, sexes, or strains. Supplemental evidence
included in

the unit of analysis (e.g., mechanistic studies in exposed animals or animal cells) might raise
the certainty of evidence to robust for a set of studies that otherwise would be described as
moderate. Such evidence not included in the unit of analysis can also inform evaluations of
the coherence of the animal evidence, the directness of the outcome measures, and the
biological significance of the findings.

This document is a draft for review purposes only and does not constitute Agency policy.

8-13	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Evidence synthesis
judgment

Description

Moderate

(©©O)

...evidence in
animal studies

(signal of effect
with some
uncertainty)

A set of evidence that does not reach the degree of certainty required for robust, but which
includes at least one high or medium confidence study and additional information
increasing certainty in the evidence. For multiple studies or a single study, the evidence is
primarily consistent or coherent with reasonable support for adversity, but there are
notable remaining uncertainties (e.g., difficulty interpreting the findings due to concerns for
indirectness of some measures); however, these uncertainties are not sufficient to reduce
or discount the level of concern regarding the positive findings and any conflicting findings
are from a set of experiments of lower confidence. The set of experiments supporting the
effect provide additional information increasing certainty in the evidence, such as
consistent effects across laboratories or species; coherent effects across multiple related
endpoints (can include mechanistic endpoints within the unit of analysis); an unusual
magnitude of effect, rarity, age at onset, or severity; a strong dose-response relationship; or
consistent observations across exposure scenarios (e.g., route, timing, duration), sexes, or
animal strains. Supplemental evidence included in the unit of analysis could address the
above factors and raise certainty in the evidence to moderate for a set of studies that
otherwise would be described as slight or, in exceptional cases, might support raising to
moderate evidence that would otherwise be described as indeterminate. Mechanistic
evidence not included in the unit of analysis can also inform evaluations of the coherence
of the animal evidence, the directness of the outcome measures, and the biological
significance of the findings.

Slight

(©OO)

...evidence in
animal studies

(signal of effect
with large amount
of uncertainty)

One or more studies reporting an effect on an exposure on the health outcome, but
considerable uncertainty exists and supporting coherent evidence is sparse. In general, the
evidence is limited to a set of consistent low confidence studies, or higher confidence
studies with significant unexplained heterogeneity or other serious uncertainties (e.g.,
concerns about adversity) across studies. It also applies when one medium or high
confidence experiment is available within the unit of analysis without additional
information increasing certainty in the evidence (e.g., coherent findings within the same
study or from other studies). Biological evidence from mechanistic studies could also be
independently interpreted as slight. This category serves primarily to encourage additional
study where evidence does exist that might provide some support for an association, but
for which the

evidence does not reach the degree of confidence required for moderate.

Indeterminate

(OOO)

...evidence in
animal studies

(signal cannot be
determined for or
against an effect)

No studies available in animals or situations when the evidence is inconsistent and primarily
of low confidence. In addition, this might include situations where higher confidence
studies exist, but there are major concerns with the evidence base such as unexplained
inconsistency, a lack of expected coherence from a stronger set of studies, very small effect
magnitude (i.e., major concerns about biological significance), or uncertainties or
methodological limitations that result in an inability to discern effects from exposure. It also
applies for a single low confidence study in the absence of factors that increase certainty. A
set of largely null studies could be concluded to be indeterminate if the evidence does not
reach the level required for compelling evidence of no effect.

This document is a draft for review purposes only and does not constitute Agency policy.

8-14	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

Protocol for the Uranium IRIS Assessment (Oral)

Evidence synthesis
judgment

Description

Compelling
evidence of no
effect
(...)

...in animal studies

(strong signal for
lack of an effect
with little
uncertainty)

A set of high confidence experiments examining a reasonable spectrum of endpoints that
demonstrate a lack of biologically significant effects across multiple species, both sexes, and
a broad range of exposure levels. The data are compelling in that the experiments have
examined the range of scenarios across which health effects in animals could be observed,
and an alternative explanation (e.g., inadequately controlled features of the studies'
experimental designs; inadequate sample sizes) for the observed lack of effects is not
available. Each of the studies should have used an optimal endpoint and exposure
assessment and adequate sample size. The evidence base should represent both sexes and
address potentially susceptible populations and lifestages. Supplemental evidence can help
to address the above considerations or, when included in the unit of analysis, provide
additional support for this judgment.

8.2. EVIDENCE INTEGRATION

The phase of evidence integration combines animal and human evidence synthesis
judgments while also considering information on the human relevance of findings in animal
evidence, coherence across evidence streams ("cross-stream coherence"), information on
susceptible populations or lifestages, understanding of biological plausibility or MOA, and
potentially other critical inferences (e.g., read-across analyses) that generally draw on mechanistic
and other supplemental evidence (see Table 8-6). This analysis culminates in an evidence
integration judgment and narrative for each potential health effect category (i.e., each noncancer
health effect and specific type of cancer, or broader grouping of related outcomes as defined during
problem formulation). To the extent it can be characterized prior to conducting dose-response
analyses, exposure context is also provided.

Given the extent of human and animal toxicology studies, in vitro and other mechanistic
studies will not be a focus of the systematic review because noncancer toxicity values for uranium
are likely to be based directly on human and mammalian studies of uranium's apical effects. If a
mechanistic analysis is considered necessary to assist with the interpretation and integration of the
epidemiological and experimental evidence of a specific hazard or health effect, EPA will rely on
previous reviews and analyses to identify relevant pathways and key studies (see Section 4.5).

With respect to susceptibility, the assessment describes the evidence (i.e., human, animal,
mechanistic) on populations and lifestages most likely to be susceptible to the hazards identified
and, to the extent possible, the factors that increase their risk for the hazards. In addition to
assessment-specific health effects evidence, background information about biological mechanisms
or ADME, as well as biochemical and physiological differences among lifestages and sexes, may be
used. At a minimum, particular consideration is given to infants and children, pregnant women, and
women of childbearing age. Many of the foundational analyses for summarizing susceptibility in the
evidence integration narrative are undertaken during evidence synthesis as patterns across studies
are evaluated with respect to consistency, coherence, and the magnitude and direction of effect
measures. Relevant factors for exploring patterns may include intrinsic factors (e.g., age, sex,

This document is a draft for review purposes only and does not constitute Agency policy.

8-15	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	genetics, health status, behaviors) and certain extrinsic factors (e.g., socioeconomic status, access to

2	healthcare), although information on the latter is rarely available in human health studies of

3	environmental chemicals.

Table 8-6. Considerations that inform evidence integration judgments

Judgment

Description

Human relevance
of findings

Used to describe and justify the interpreted relevance of the data from experimental animals
(or other model systems) to humans. In the absence of chemical-specific evidence informing
human relevance, the evidence integration narrative will briefly describe the interpreted
underlying biological similarity across species. As noted in EPA guidelines (U.S. EPA, 2005a),
there needs to be evidence or a biological explanation to support an interpreted lack of
human relevance for findings in animals, and site concordance is neither expected nor
required. Thus, in the absence of specific evidence or cross-species understanding of the
underlying biology, it is appropriate to use a statement such as, "without evidence to the
contrary, [health effect] responses in animals are presumed relevant to humans."

Cross-stream
coherence

Used to address the concordance of biologically related findings across human, animal, and
mechanistic studies, considering features of the available evidence such as exposure timing
and cancer), it is not necessary or expected that effects manifest in humans are identical to
those observed in animals (e.g., tumors in animals can be predictive of carcinogenic potential
in humans, but not necessarily at the same site), although this typically provides stronger
evidence. Biological understanding of the manner in which the outcomes are manifest in
different species can inform cross-stream coherence. Evidence supporting a biologically
plausible mechanistic pathway across species adds coherence (see below).

Susceptible
populations and
lifestages

Used to summarize analyses relating to individual and social factors that may increase

susceptibility to exposure-related health effects in certain populations or lifestages, or to

highlight the lack of such information. These analyses are based on knowledge about the
health outcome or organ system affected and focus on the influence of intrinsic biological
factors but can also include consideration of mechanistic and ADME evidence.

Biological
plausibility and
MOA

considerations

Used to summarize the interpreted biological plausibility of an association between exposure
and the health effect, based primarily on the extent to which the available evidence comports
with the known development and characteristics of the health effect (and thus dependent on
sufficient information being available to draw such an interpretation). Importantly, because
this interpretation is dependent on canonical scientific knowledge about the health effect, the
lack of such understanding does not provide a rationale to decrease certainty in the evidence
for an effect (NTP, 2015; NRC, 2014). These analyses can be detailed (e.g., when attempting to
establish MOA understanding) and, if so, are typically conducted separately (e.g., as part of
the mechanistic evidence synthesis) and then referenced in the evidence integration
narrative.

Other critical

inferences

(optional)

Can be used to describe the consideration of other evidence or non-chemical-specific

information that informs evidence integration judgments (e.g., use of read-across analyses or
ADME understanding used to inform the other considerations described below; judgments on
other health effects expected to be linked to the health effect under evaluation).

ADME = absorption, distribution, metabolism, and excretion; MOA = mode of action.

This document is a draft for review purposes only and does not constitute Agency policy.

8-16	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

Protocol for the Uranium IRIS Assessment (Oral)

Using a structured framework approach, one of five phrases is used to summarize the
evidence integration judgment based on the integration of the evidence synthesis judgments, taking
into account the additional considerations assessed across evidence streams: evidence
demonstrates, evidence indicates (likely), evidence suggests, evidence is inadequate, or strong evidence
supports no effect (see Table 8-7). The five evidence integration judgment levels reflect the
differences in the amount and quality of the data that inform the evaluation of whether exposure is
interpreted as capable of causing the health effect(s). As it is assumed that any identified health
hazards will only be manifest given exposures of a certain type and amount (e.g., a specific route; a
minimal duration, periodicity, and level), the evidence integration narrative and summary
judgment levels include the generic phrase, "given sufficient exposure conditions." This highlights
that, for those assessment-specific health effects identified as potential hazards, the exposure
conditions associated with those health effects will be defined (as will the uncertainties in the
ability to define those conditions) during dose-response analysis (see Section 9). More than one
evidence integration judgment level can be used when the evidence base is able to support that a
chemical's effects differ by exposure level or route fU.S. EPA. 2005al The analyses and judgments
are summarized in the evidence profile table (see Table 8-1).

Similar to the description for summarizing noncancer judgments above, the cancer
descriptor and evidence integration narrative for carcinogenicity also consider the conditions of
carcinogenicity, including exposure (e.g., route; level) and susceptibility (e.g., genetics; lifestage), as
the data allow fFarland. 2005: U.S. EPA. 2005a. b). As with noncancer effects, the specific exposure
conditions necessary for carcinogenicity are further defined during dose-response analysis.

This document is a draft for review purposes only and does not constitute Agency policy.

8-17	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 8-7. Framework for summary evidence integration judgments in the evidence integration narrative

Summary evidence integration judgment3
in narrative

Evidence integration judgment level

Explanation and example scenarios'3

The currently available evidence
demonstrates that [chemical] causes
[health effect] in humans0 given sufficient
exposure conditions. This conclusion is
based on studies of [humans or animals]
that assessed [exposure or dose] levels of
[range of concentrations or specific cutoff
level concentration01].

Evidence demonstrates

A strong evidence base demonstrating that [chemical] exposure
causes [health effect] in humans.

•	This conclusion level is used if there is robust human
evidence supporting an effect.

•	This conclusion level could also be used with moderate
human evidence and robust animal evidence if there is
strong mechanistic evidence that MOAs and key precursors
identified in animals are anticipated to occur and progress in
humans.

The currently available evidence indicates
that [chemical] likely causes [health
effect] in humans given sufficient
exposure conditions. This conclusion is
based on studies of [humans or animals]
that assessed [exposure or dose] levels of
[range of concentrations or specific cutoff
level concentration].

Evidence indicates (likely0)

An evidence base that indicates that [chemical] exposure likely
causes [health effect] in humans, although there may be outstanding
questions or limitations that remain, and the evidence is insufficient
for the higher conclusion level.

•	This conclusion level js used if there is robust animal
evidence supporting an effect and slight-to-indeterminate
human evidence, or with moderate human evidence when
strong mechanistic evidence is lacking.

•	This conclusion level could also be used with moderate
human evidence supporting an effect and moderate-to-
indeterminate animal evidence, or with moderate animal
evidence supporting an effect and moderate-to-
indeterminate human evidence. In these scenarios, any
uncertainties in the moderate evidence are not sufficient to
substantially reduce confidence in the reliability of the
evidence, or mechanistic evidence in the slight or
indeterminate evidence base (e.g., precursors) exists to
increase confidence in the reliability of the moderate
evidence.

This document is a draft for review purposes only and does not constitute Agency policy.

8-18	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Summary evidence integration judgment3
in narrative

Evidence integration judgment level

Explanation and example scenarios'3

The currently available evidence suggests
that [chemical] may cause [health effect]
in humans given sufficient exposure
conditions. This conclusion is based on
studies of [humans or animals] that
assessed [exposure or dose] levels of
[range of concentrations or specific cutoff
level concentration].

Evidence suggests

An evidence base that suggests that [chemical] exposure may cause
[health effect] in humans, but there are very few studies that
contributed to the evaluation, the evidence is very weak or
conflicting, and/or the methodological conduct of the studies is poor.

•	This conclusion level is used if there is slight human evidence
and \ndeterminate-to-slight animal evidence.

•	This conclusion level is also used with slight animal evidence
and indeterminate-to-slight human evidence.

•	This conclusion level could also be used with moderate
human evidence and slight or indeterminate animal
evidence, or with moderate animal evidence and slight or
indeterminate human evidence. In these scenarios, there are
outstanding issues or uncertainties regarding the moderate
evidence (i.e., the synthesis judgment was borderline with
slight), or mechanistic evidence in the slight or
indeterminate evidence base (e.g., null results in well-
conducted evaluations of precursors) exists to decrease
confidence in the reliability of the moderate evidence.

•	Exceptionally, when there is general scientific understanding
of mechanistic events that result in a health effect, this
conclusion level could also be used if there is strong
mechanistic evidence that is sufficient to highlight potential
human toxicity'—in the absence of informative conventional
studies in humans or in animals (i.e., indeterminate evidence
in both).

This document is a draft for review purposes only and does not constitute Agency policy.

8-19	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Summary evidence integration judgment3
in narrative

Evidence integration judgment level

Explanation and example scenarios'3

The currently available evidence is
inadequate to assess whether [chemical]
may cause [health effect] in humans.

Evidence inadequate

his conveys either a lack of information or an inability to interpret the
available evidence for [health effect]. On an assessment-specific
basis, a single use of this "inadequate" conclusion level might be used
to characterize the evidence for multiple health effect categories (i.e.,
all health effects that were examined and did not support other
conclusion levels).8

•	This conclusion level is used if there is indeterminate human
and animal evidence.

•	This conclusion level is also used with slight animal evidence
and compelling evidence of no effect human evidence.

•	This conclusion level could also be used with sliaht-to-robust
animal evidence and indeterminate human evidence if
strong mechanistic information indicated that the animal
evidence is unlikely to be relevant to humans.

A conclusion of inadequate is not a determination that the agent
does not cause the indicated health effect(s). It simply indicates that
the available evidence is insufficient to reach conclusions.

This document is a draft for review purposes only and does not constitute Agency policy.

8-20	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Summary evidence integration judgment3
in narrative

Evidence integration judgment level

Explanation and example scenarios'3

Strong evidence supports no effect in
humans. This conclusion is based on
studies of [humans or animals] that
assessed [exposure or dose] levels of
[range of concentrations].

Strong evidence supports no effect in

humans. This conclusion is based on
studies of [humans or animals] that
assessed [exposure or dose] levels of
[range of concentrations].

This represents a situation in which extensive evidence across a range
of populations and exposure levels has identified no
effects/associations. This scenario requires a high degree of
confidence in the conduct of individual studies, including
consideration of study sensitivity, and comprehensive assessments of
the endpoints and lifestages of exposure relevant to the heath effect
of interest.

•	This conclusion level js used if there is compelling evidence
of no effect in human studies and compelling evidence of no
effect to indeterminate in animals.

•	This conclusion level is also used if there is indeterminate
human evidence and compelling evidence of no effect animal
evidence in models concluded to be relevant to humans.

•	This conclusion level could also be used with compelling
evidence of no effect in human studies and moderate-to-
robust animal evidence if strong mechanistic information
indicated that the animal evidence is unlikely to be relevant
to humans.

aEvidence integration judgments are typically developed at the level of the health effect when there are sufficient studies on the topic to evaluate the evidence at that level; this
should always be the case for "evidence demonstrates" and "strong evidence supports no effect," and typically for "evidence indicates (likely)." However, some databases only
allow for evaluations at the category of health effects examined; this will more frequently be the case for conclusion levels of "evidence suggests" and "evidence inadequate."
A judgment of "strong evidence supports no effect" is drawn at the health effect level.

Terminology of "is" refers to the default option; terminology of "could also be" refers to situational options dependent on mechanistic understanding.
cln some assessments, these conclusions might be based on data specific to a particular lifestage of exposure, sex, or population (or another specific group). In such cases, this
would be specified in the narrative conclusion, with additional detail provided in the narrative text. This applies to all conclusion levels.
dlf concentrations cannot be estimated, an alternative expression of exposure level such as "occupational exposure levels," are provided. This applies to all conclusion levels.
eFor some applications, such as benefit-cost analysis, categories of "evidence demonstrates" and "evidence indicates," should be interpreted as evidence that supports an

exposure-effect linkage that is likely to be causal.

'Scientific understanding of adverse outcome pathway (AOPs) and of the human implications of new toxicity testing methods (e.g., from high-throughput screening, from
short-term in vivo testing of alternative species or from new in vitro testing) will continue to increase. This may make possible the development of hazard conclusions when
there are mechanistic or other relevant data that can be interpreted with a similar level of confidence to positive animal results in the absence of conventional studies in
humans or in animals.
gSpecific narratives for each of these health effects may also be deemed unnecessary.

This document is a draft for review purposes only and does not constitute Agency policy.

8-21	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

Protocol for the Uranium IRIS Assessment (Oral)

9. DOSE-RESPONSE ASSESSMENT: STUDY
SELECTION AND QUANTITATIVE ANALYSIS

9.1. OVERVIEW

Selection of specific datasets for dose-response assessment and performance of the
dose-response assessment is conducted after hazard identification is complete and involves
database- and chemical-specific biological judgments. A number of EPA guidance and support
documents detail data requirements and other considerations for dose-response modeling,
especially EPA's Benchmark Dose Technical Guidance (U.S. EPA. 2012b). EPA's Review of the
Reference Dose and Reference Concentration Processes fU.S. EPA. 2005a. 20021. Guidelines for
Carcinogen Risk Assessment fU.S. EPA. 2005al. and Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005b). This section of the protocol
provides an overview of considerations for conducting the dose-response assessment, particularly
statistical considerations specific to dose-response analysis that support quantitative risk
assessment Importantly, these considerations do not supersede existing EPA guidance.

The focus of this assessment is to develop an oral noncancer reference dose (RfD). An RfD-is
an estimate, with uncertainty spanning perhaps an order of magnitude, of an exposure to the
human population (including susceptible populations and life stages) that is likely to be without an
appreciable risk of deleterious health effects over a lifetime (U.S. EPA. 20021. A reference
concentration (RfC) for inhalation noncancer will not be derived, nor will inhalation unit risk and
oral slop factors to characterize cancer dose response.

The derivation of noncancer toxicity values depends on the nature of the hazard conclusion.
For noncancer outcomes dose-response is conducted based on having stronger evidence of a
hazard (generally, "evidence demonstrates" and "evidence indicates [likely]." When the noncancer
outcome is considered "evidence suggests" of potential hazard to humans, EPA generally would not
conduct a dose-response assessment and derive a RfD. Cases where suggestive evidence might be
used to develop a noncancer toxicity value include when the evidence base includes a
well-conducted study (overall medium or high confidence for the outcome), quantitative analyses
may be useful for some purposes, (e.g., providing a sense of the magnitude and uncertainty of
potential risks, ranking potential hazards, or setting research priorities) (U.S. EPA. 2005a).

Dose-response assessments for noncancer hazards are typically performed following
chronic exposure14 to the chemical of interest, if supported by existing data. In addition to an RfD,

14Dose-response assessments may also be conducted for shorter durations, particularly if the evidence base
for a chemical indicates risks associated with shorter exposures to the chemical (U.S. EPA. 20021.

This document is a draft for review purposes only and does not constitute Agency policy.

9-1	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

Protocol for the Uranium IRIS Assessment (Oral)

this assessment will attempt to derive organ- or system-specific RfDs (osRfDs) when the data are
sufficiently strong (i.e., noncancer conclusions of evidence demonstrate or evidence indicates
[likely]). If the available data are appropriate for doing so, the assessments will derive a
less-than-lifetime toxicity value (a "subchronic" reference dose) for noncancer hazards. Both
less-than-lifetime and hazard-specific values may be useful to EPA risk assessors within specific
decision contexts.

9.2. SELECTING STUDIES FOR DOSE-RESPONSE ASSESSMENT

The assessment presents a summary of hazard identification conclusions to transition to
dose response considerations, highlighting the feasibility of extracting, or deriving, a dose-response
function corresponding to each identified hazard. If PODs are based on modeled internal dose
levels, there will need to be physiologically based pharmacokinetic (PBPK) modeling to convert
internal POD into human equivalent doses (POD(hed)S). If such PBPK models have not been
established, then it may not be feasible to derive P0D(hed)S. Once the feasibility of using dose-
response information to derive PODs has been established, the next step is to identify and justify
the selection of one or more benchmark response (BMR) levels for the derivation of points of
departure (PODs).

The pool of outcomes and study-specific endpoints is discussed to identify which categories
of effects and study designs are considered the strongest and most appropriate for quantitative
assessment of a given health effect, particularly among the studies that exemplify the study
attributes summarized in Table 9-1. Consideration will also be given as to whether toxicity values
can be derived to protect specific populations or life stages.

Also considered is whether there are opportunities for quantitative evidence integration.
Examples of quantitative integration, from simplest to more complex, include (1) combining results
for an outcome across sex (within a study); (2) characterizing overall toxicity, as in combining
effects that comprise a syndrome, or occur on a continuum (e.g., precursors and eventual overt
toxicity, benign tumors that progress to malignant tumors); and (3) conducting a meta-analysis or
meta-regression of all studies addressing a category of important health effects.

Some studies that are used qualitatively for hazard identification may or may not be useful
quantitatively for dose-response assessment due to such factors as the lack of quantitative
measures of exposure or lack of variability measures for response data. If the needed information
cannot be located, semiquantitative analysis may be feasible (e.g., via NOAEL/LOAEL). In the draft
and final assessments, specific endpoints considered for dose response are summarized in a tabular
format that includes rationales for decisions to proceed (or not) for POD derivation. In addition,
mechanistic evidence that influences the dose-response analyses is highlighted, for example,
evidence related to susceptibility or other uncertainty factors, or if MOA may influence the potential
shape of the dose-response curve (i.e., linear, nonlinear, or threshold model).

This document is a draft for review purposes only and does not constitute Agency policy.

9-2	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table 9-1. Attributes used to evaluate studies for derivation of toxicity values





Considerations

Study attributes

Human studies

Animal studies

Study confidence

High or medium confidence studies are highly preferred over low confidence studies. The selection of low confidence studies
should include an additional explanatory justification (e.g., only low confidence studies had adequate data for toxicity value
derivation). The available high and medium confidence studies are further differentiated on the basis of the study attributes
below, as well as a reconsideration of the specific limitations identified and their potential impact on dose-response analyses.

Rationale for choice of
species

Human data are preferred over animal data to
eliminate interspecies extrapolation uncertainties

(e.g., in pharmacodynamics, dose-response pattern in
relevant dose range, relevance of specific health
outcomes to humans).

Animal studies provide supporting evidence when adequate human
studies are available, and they are considered the studies of primary
interest when adequate human studies are not available. For some
hazards, studies of particular animal species known to respond
similarly to humans would be preferred over studies of other species.

Relevance of

exposure

paradigm

Exposure
route

Studies involving human environmental exposures
(oral, inhalation).

Studies by a route of administration relevant to human
environmental exposure are preferred. A validated pharmacokinetic
or PBPK model can also be used to extrapolate across exposure
routes.



Exposure
durations

When developing a chronic toxicity value, chronic or subchronic studies are preferred over studies of acute exposure durations.
Exceptions exist, such as when a susceptible population or life stage is more sensitive in a particular time window (e.g.,
developmental exposure).



Exposure
levels

Exposures near the range of typical environmental human exposures are preferred. Studies with a broad exposure range and
multiple exposure levels are preferred to the extent that they can provide information about the shape of the
exposure-response relationship (see the EPA Benchmark Dose Technical Guidance, §2.1.1) and facilitate extrapolation to more
relevant (generally lower) exposures.

Subject selection

Studies that provide risk estimates in the most susceptible groups are preferred.

Controls for possible
confounding3

Studies with a design (e.g., matching procedures, blocking) or analysis (e.g., covariates or other procedures for statistical
adjustment) that adequately address the relevant sources of potential critical confounding for a given outcome are preferred.

This document is a draft for review purposes only and does not constitute Agency policy.

9-3	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Study attributes

Considerations

Human studies

Animal studies

Measurement of exposure

Studies that can reliably distinguish between levels of
exposure in a time window considered most relevant
for development of a causal effect are preferred.
Exposure assessment methods that provide
measurements at the level of the individual and that
reduce measurement error are preferred.
Measurements of exposure should not be influenced by
knowledge of health outcome status.

Studies providing actual measurements of exposure (e.g., analytical
inhalation concentrations vs. target concentrations) are preferred.
Relevant internal dose measures may facilitate extrapolation to
humans, as would availability of a suitable animal PBPK model in
conjunction with an animal study reported in terms of administered
exposure.

Health outcome(s)

Studies that can reliably distinguish the presence or absence (or degree of severity) of the outcome are preferred. Outcome
ascertainment methods using generally accepted or standardized approaches are preferred.

Studies with individual data are preferred in general. For example, individual data allow you to characterize experimental
variability more realistically and to characterize overall incidence of individuals affected by related outcomes (e.g., phthalate
syndrome).

Among several relevant health outcomes, preference is generally given to those outcomes with less concern for indirectness or
with greater biological significance.

Study size and design

Preference is given to studies using designs reasonably expected to have power to detect responses of suitable magnitude.15
This does not mean that studies with substantial responses, but low power would be ignored, but that they should be
interpreted in light of a confidence interval or variance for the response. Studies that address changes in the number at risk
(through decreased survival, loss to follow-up) are preferred.

aAn exposure or other variable that is associated with both exposure and outcome but is not an intermediary between the two.

bPower is an attribute of the design and population parameters, based on a concept of repeatedly sampling a population; it cannot be inferred post hoc using data from one
experiment (Hoenie and Heisev. 2001).

This document is a draft for review purposes only and does not constitute Agency policy.

9-4	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

Protocol for the Uranium IRIS Assessment (Oral)

9.3. CONDUCTING DOSE-RESPONSE ASSESSMENTS

EPA uses a two-step approach for dose-response assessment that begins with analysis of
the dose-response data in the range of observation. However, when data are available, they often
cover only a portion of the possible range of the dose-response relationship, in which case some
extrapolation must be done in order to estimate the effects of exposures that are lower than the
range of data obtained from scientific studies fU.S. EPA. 2012b. 2005a):

1)	Step 1: Take an assessment of all data that are available from selected studies or can be
gathered through experiments. This is in order to document the dose-response
relationship (s) over the range of observed doses (i.e., the doses that are reported in the data
collected) to derive an estimated POD). See Section 9.3.1 for more details. However,
frequently this range of observation may not include sufficient data to identify a dose where
the adverse effect is not observed in the human population fU.S. EPA. 2022b. 2000).

2)	Step 2: This consists of extrapolations to estimate the risk of adverse effects beyond the
lower range of available observed data. This is in order to make inferences about the critical
region where the dose level begins to cause the adverse effect in the human population fU.S.
EPA. 2022b. 2000). See Section 9.3.2.

When sufficient and appropriate human data and laboratory animal data are both available
for the same outcome, human data are generally preferred for the dose-response assessment
because their use eliminates the need to perform interspecies extrapolations.

For noncancer analyses, IRIS assessments typically derive a candidate value from each
suitable dataset, whether for human or animal. Evaluating these candidate values grouped within a
particular organ/system yields a single organ/system-specific reference value for each
organ/system under consideration. Next, evaluation of these organ/system-specific reference
values results in the selection of a single overall reference value to cover all health outcomes across
all organs/systems. While this overall reference value is the focus of the assessment, the
organ/system-specific reference values can be useful for subsequent cumulative risk assessments
that consider the combined effect of multiple agents acting at a common organ/system.

9.3.1. Dose-Response Analysis in the Range of Observation

For conducting a dose response assessment, pharmacodynamic ("biologically based")
modeling can be used when there are sufficient data to ascertain the mode of action and
quantitatively support model parameters that represent rates and other quantities associated with
the key precursor events of the modes of action. If there is not an applicable pharmacodynamic
model available to assess health effects associated with oral exposure to uranium, empirical dose-
response modeling is used to fit the data (on the apical outcomes or a key precursor events) in the
ranges of observation. For this purpose of empirical dose-response modeling, EPA has developed a
standard set of models f http: / /www.epa.gov/ncea /bmds] that can be applied to typical
dichotomous and continuous datasets, including those that are nonlinear. In situations where there

This document is a draft for review purposes only and does not constitute Agency policy.

9-5	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

Protocol for the Uranium IRIS Assessment (Oral)

are alternative models with significant biological support, the users of the assessment can be
informed by the presentation of these alternatives along with the models' strengths and
uncertainties. The EPA has developed guidelines on modeling dose-response data, assessing model
fit, selecting suitable models, and reporting modeling results [see the EPA Benchmark Dose
Technical Guidance (U.S. EPA. 2012b)].

U.S. EPA Benchmark Dose Software (BMDS) is designed to model dose-response datasets in
accordance with EPA Benchmark Dose Technical Guidance (U.S. EPA. 2012b). For noncancer effects,
a benchmark dose lower confidence limit (BMDL) is computed from a model selected from the
BMDS suite of models using statistical and graphical criteria. Additional judgments or alternative
analyses may be used if initial modeling procedures fail to yield results in reasonable agreement
with the data. For example, modeling may be restricted to the lower doses, especially if there is
competing toxicity at higher doses. Modeling may also need to accommodate cases of nonlinear
dose-response data.

For noncancer datasets, EPA recommends (1) application of a preferred set of models that
use maximum likelihood estimation (MLE) methods (default models in BMDS) and (2) selection of a
POD from a single model based on criteria designed to limit model selection subjectivity (auto
implemented in BMDS version 3 and higher). For the linear analysis of cancer datasets, EPA
recommends (1) application of the Multistage MLE model; (2) selection of a single Multistage
degree; and (3) in cases where tumors are observed in multiple organ systems, use of a multi-tumor
model (i.e., MS-Combo) that appropriately estimates combined tumor risk (both (2) and (3) are
available in BMDS).15

Version 3.2 and higher of BMDS also provides an alternative modeling approach that uses
Bayesian model averaging for dichotomous modeling average (DMA). EPA makes DMA available as
an alternative approach but has not yet finalized guidelines for their use. DMA may be applied to
uranium as a supplemental analysis.

For each modeled dataset for an outcome, a POD from the observed data should be
estimated to mark the beginning of extrapolation to lower doses. The POD is an estimated dose
(expressed in human equivalent terms) near the lower end of the observed range without
significant extrapolation to lower doses. For linear extrapolation of cancer risk, the POD is used to
calculate an OSF, and for nonlinear extrapolation, the POD is used in calculating an RfD.

The selection of the response level at which the POD is calculated is guided by the severity
of the endpoint. Nonlinear approaches consider both statistical and biologic considerations. For
dichotomous data, a response level of 10% extra risk is generally used for minimally adverse
effects, 5% or lower for more severe effects or effects observed in studies with increased statistical
sensitivity. Lower BMRs are often supported for developmental toxicity studies. For continuous

15The Multistage degree selection process outlined in the memo is auto-implemented in the BMDS multi
tumor model, which can be run on one or more tumor datasets, but only the noncancer model selection
process is auto-implemented for individual Multistage model runs in the current version, BMDS 3.2).

This document is a draft for review purposes only and does not constitute Agency policy.

9-6	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

Protocol for the Uranium IRIS Assessment (Oral)

data, a response level is ideally based on an established definition of biologic significance. In the
absence of such definition, one control standard deviation from the control mean is often used for
minimally adverse effects, one-half standard deviation for more severe effects. As with
dichotomous endpoints, lower BMRs may also be supported for endpoints observed in studies with
greater statistical sensitivity (e.g., developmental toxicity studies). The POD is the 95% lower bound
on the dose associated with the selected response level.

EPA has developed standard approaches for determining the relevant dose to be used in the
dose-response modeling in the absence of appropriate pharmacokinetic modeling. These standard
approaches also facilitate comparison across exposure patterns and species:

•	Intermittent study exposures are standardized to a daily average over the duration of exposure.
For chronic effects, daily exposures are averaged over the lifespan. Exposures during a critical
period, however, are not averaged over a longer duration (YU.S. EPA. 2005a). see §3.1.1; fU.S.
EPA. 19911. see §3.2). Note that this will typically be done after modeling because the
conversion is linear.

•	Doses are standardized to equivalent human terms to facilitate comparison of results from
different species. Oral doses are scaled allometrically using mg/kg3/4day as the equivalent dose
metric across species. Allometric scaling pertains to equivalence across species, not across life
stages, and is not used to scale doses from adult humans or mature animals to infants or
children fU.S. EPA. 2011a. 2005a). §3.1.3. Inhalation exposures are scaled using dosimetry
models that apply species-specific physiologic and anatomic factors and consider whether the
effect occurs at the site of first contact or after systemic circulation fU.S. EPA. 2012a. 1994). §3.

•	It can be informative to convert doses across exposure routes. If this is done, the assessment
describes the underlying data, algorithms, and assumptions fU.S. EPA. 2005a). §3.1.4.

•	In the absence of study specific data on, for example, intake rates or body weight, the EPA has
developed recommended values for use in dose response analysis fU.S. EPA. 1988).

•	The preferred approach for dosimetry extrapolation from animals to humans is through PBPK
modeling. Elements of more than one published model can be combined if the effort involved is
minimal and no one model has all the features desired.

Briefly, PBPK model simulations are used to estimate internal dose metrics corresponding
to the applied doses for each experimental animal bioassay. By simulating the exposure scenario for
each toxicity study, the resulting internal metric effectively accounts for the difference between the
pattern and a nominal daily exposure. The set of internal dose metrics for each toxicity study and
endpoint can then be used in dose-response analysis to identify a BMDL or other POD for individual
animal toxicity studies. In this assessment, the internal dose metric is either the tissue-specific rate
of oxidative metabolism or a daily average blood concentration. The human version of the PBPK
model can then be used to estimate the exposure dose that would result in an internal dose at the
POD. Any remaining uncertainty factors, including the factor of 10 for human interindividual
variability (UFH) will then be applied for derivation of the HECs.

This document is a draft for review purposes only and does not constitute Agency policy.

9-7	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

Protocol for the Uranium IRIS Assessment (Oral)

9.3.2. Extrapolation: Reference Values

Reference value derivation is EPA's most frequently used type of nonlinear extrapolation
method. For each dataset selected for reference value derivation, reference values are estimated by
applying relevant adjustments to the PODs to account for the conditions of the reference value
definition—for human variation, extrapolation from animals to humans, extrapolation to chronic
exposure duration, and extrapolation to a minimal level of risk (if not observed in the dataset).
Increasingly, data-based adjustments (U.S. EPA. 20141 and Bayesian methods for characterizing
population variability fNRC. 2014] are feasible and may be distinguished from the UF
considerations outlined below. The assessment discusses the scientific bases for estimating these
data-based adjustments and UFs:

•	Animal-to-human extrapolation: If animal results are used to make inferences about humans,
the reference value derivation incorporates the potential for cross-species differences, which
may arise from differences in pharmacokinetics or toxicodynamics. If available, a biologically
based model that adjusts fully for pharmacokinetic and toxicodynamic differences across
species maybe used. Otherwise, the POD is standardized to equivalent human terms or is based
on pharmacokinetic or dosimetry modeling, that may range from detailed chemical-specific to
default approaches (U.S. EPA. 2014. 2011a). and a factor of 101/2 (rounded to 3) is applied to
account for the remaining uncertainty involving pharmacokinetic and toxicodynamic
differences.

•	Human variation: The assessment accounts for variation in susceptibility across the human
population and the possibility that the available data may not represent individuals who are
most susceptible to the effect, by using a data-based adjustment, a UF, or a combination of the
two. Where appropriate data or models for the effect or for characterizing the internal dose are
available, the potential for data-based adjustments for toxicodynamics or pharmacokinetics is
considered fU.S. EPA. 2014. 2002.).16 17 When sufficient data are available, an intraspecies UF
either less than or greater than 10-fold may be justified fU.S. EPA. 20021. This factor may be
reduced if the POD is derived from or adjusted specifically for susceptible individuals [not for a
general population that includes both susceptible and non-susceptible individuals (U.S. EPA.
20021. §4.4.5; CIJ.S. EPA. 19981. §4.2; CIJ.S. EPA. 19961. §4; CIJ.S. EPA. 19941. §4.3.9.1; CIJ.S. EPA.
19911. §3.4], When the use of such data or modeling is not supported, a UF with a default value
of 10 is considered.

•	LOAEL-to-NOAEL\ If a POD is based on a LOAEL, the assessment includes an adjustment to an
exposure level where such effects are not expected. This can be a matter of great uncertainty if
there is no evidence available at lower exposures. A factor of 10 is generally applied to
extrapolate to a lower exposure expected to be without appreciable effects. A factor other than

"Examples of adjusting the pharmacokinetic portion of interhuman variability include the IRIS boron
assessment's use of nonchemical-specific kinetic data [e.g., glomerular filtration rate in pregnant humans as a
surrogate for boron clearance (U.S. EPA. 20041] and the IRIS trichloroethylene assessment's use of population
variability in trichloroethylene metabolism, via a PBPK model, to estimate the lower 1st percentile of the dose
metric distribution for each POD (U.S. EPA. 2011bl.

17Note that when a PBPK model is available for relating human internal dose to environmental exposure,
relevant portions of this UF may be more usefully applied prior to animal-to-human extrapolation, depending
on the correspondence of any nonlinearities (e.g., saturation levels) between species.

This document is a draft for review purposes only and does not constitute Agency policy.

9-8	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Protocol for the Uranium IRIS Assessment (Oral)

10 may be used depending on the magnitude and nature of the response and the shape of the
dose-response curve fU.S. EPA. 2002.1998.1996.1994.19911.

•	Subchronic-to-chronic exposure: When using subchronic studies to make inferences about
chronic/lifetime exposure, the assessment considers whether lifetime exposure could have
effects at lower levels of exposure. A factor of up to 10 may be applied to the POD, depending on
the duration of the studies and the nature of the response fU.S. EPA. 2002.1998.19941.

•	Database deficiencies: In addition to the adjustments above, if database deficiencies raise
concern that further studies might identify a more sensitive effect, organ system, or life stage,
the assessment may apply a database UF fU.S. EPA. 2002.1998.1996.1994.19911. The size of
the factor depends on the nature of the database deficiency. For example, the EPA typically
follows the recommendation that a factor of 10 be applied if both a prenatal toxicity study and a
two-generation reproduction study are missing and a factor of 101/2 (i.e., 3) if either one or the
other is missing fU.S. EPA. 20021.

The POD for a reference value (RfV) is divided by the product of these factors. fU.S. EPA.

20021 recommends that any composite factor that exceeds 3,000 represents excessive uncertainty

and recommends against relying on the associated RfV.

This document is a draft for review purposes only and does not constitute Agency policy.

9-9	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

Protocol for the Uranium IRIS Assessment (Oral)

REFERENCES

. (1995). Age-dependent doses to members of the public from intake of radionuclides: Part 3.

Ingestion dose coefficients. A report of a Task Group of Committee 2 of the International
Commission on Radiological Protection. Ann ICRP 25: 1-74.

Adams. IB: Audhva. T: Mcdonough-Means. S: Rubin. RA: Ouig. D: Geis. E: Gehn. E: Loresto. M:

Mitchell. I: Atwood. S: Barnhouse. S: Lee. W. (2013). Toxicological status of children with
autism vs. neurotypical children and the association with autism severity. Biol Trace Elem
Res 151: 171-180. http://dx.doi.Org/l 0.1007/sl 2011 -012-9551-1.

Al-Sahlanee. MHR: Ramli. RM: Ali. MAH: Tawfiq. NF: Azman. NZN: Rahman. AA: Mustafa. IS: Razak.
N: Yahava. N: Al-Marri. HM: Avob. N: Zakaria. N. (2017). Trace of heavy metals in maternal
and umbilical cord blood samples in association with birth outcomes in Baghdad, Iraq. In
MH Koh; G Krishnan; F Mohd Noor (Eds.), EPJ Web of Conferences. CEDEX A: E D P
SCIENCES, http://dx.doi.org/10.1051 /epiconf/201715600003.

A1 Rashida. VIM: Wang. X: Myers. OB: Bovce. TW: Kocher. E: Moreno. M: Karr. R: Ass'ad. N: Cook. LS:
Sood. A. (2019). Greater odds for angina in uranium miners than nonuranium miners in
New Mexico. J Occup Environ Med 61: 1-7.
http://dx.doi.Org/10.1097/TOM.0000000000001482.

Alaani. S: Tafash. M: Busby. C: Hamdan. M: Blaurock-Busch. E. (2011). Uranium and other

contaminants in hair from the parents of children with congenital anomalies in Fallujah,
Iraq. Conflict and Health 5:15. http: //dx.doi.org/10.1186/1752-15Q5-5-15.

Arnold. C. (2014). Once upon a mine: the legacy of uranium on the Navajo Nation. Environ Health
Perspect 122: A44-A49. http://dx.doi.org/10.1289/ehp.122-A44.

Arzuaga. X: Gehlhaus. M: Strong. I. (2015). Modes of action associated with uranium induced

adverse effects in bone function and development [Review], Toxicol Lett 236: 123-130.
http://dx.doi.Org/10.1016/i.toxlet2015.05.006.

Ass'ad. NA: Shore. X: Myers. 0: Camacho. AR: Tacquez. 0: Pollard. C: Cook. LS: Leng. S: Page. K: Sood.
A: Zvchowski. KE. (2021). VCAM-1 is upregulated in uranium miners compared to other
miners. Life 11: 1223. http: / /dx. doi. o r g /10.3 3 9 0 /life 111112 2 3.

ATSDR (Agency for Toxic Substances and Disease Registry). (2013). Toxicological profile for
uranium. Atlanta (GA): Agency for Toxic Substances and Disease Registry (US).

Aung. MT: Meeker. ID: Boss. I: Bakulski. KM: Mukheriee. B: Cantonwine. DE: Mcelrath. TF: Ferguson.
KK. (2019). Title: Manganese is associated with increased plasma interleukin-ip during
pregnancy, within a mixtures analysis framework of urinary trace metals. Reprod Toxicol
93: 43-53. http: / /dx. do i. or g /10.1016 /i. r ep r otox. 2 019.12.0 04.

Bai. 1: Forma. A: Kowalska. B: Teresiriski. G: Buszewicz. G: Maierek. D: Flieger. W: Macieiewski. R:

Karakuta. K: Flieger. M: Czeczelewski. M: Kqdzierawski. P: Flieger. I. (2022). Multi-Elemental
Analysis of Human Optic Chiasm-A New Perspective to Reveal the Pathomechanism of
Nerve Fibers' Degeneration. Int J Environ Res Public Health 19.
http://dx.doi.org/10.3390/iierphl9074420.

Bannach-Brown. A: Przybvla. P: Thomas. 1: Rice. ASC: Ananiadou. S: Liao. I: Macleod. MR. (2018).
Machine learning algorithms for systematic review: reducing workload in a preclinical
review of animal studies and reducing human screening error (pp. 1-26). bioRxiv.
http://dx.doi.org/10.1101/255760.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

41

42

43

44

45

46

47

48

Protocol for the Uranium IRIS Assessment (Oral)

Bleise. A: Danesi. PR: Burkart. W. (2003). Properties, use and health effects of depleted uranium
(DU): a general overview [Review], J Environ Radioact 64: 93-112.
http://dx.doi.org/10.1016/s0265-931xr02100041-3.

Bloom. MS: Buck Louis. GM: Sundaram. R: Maisog. 1M: Steuerwald. AT: Parsons. PI. (2015). Birth

outcomes and background exposures to select elements, the Longitudinal Investigation of
Fertility and the Environment (LIFE). Environ Res 138: 118-129.
http://dx.doi.Org/10.1016/i.envres.2015.01.008.

Bolt. AM: Medina. S: Lauer. FT: Liu. KT: Burchiel. SW. (2019). Minimal uranium immunotoxicity

following a 60-day drinking water exposure to uranyl acetate in male and female C57BL/6J
mice. Toxicol Appl Pharmacol 372: 33-39. http://dx.doi.Org/10.1016/i.taap.2019.04.003.

Bouet. S: Samson. E: Tovanovic. I: Laurier. D: Laurent. 0. (2018). First mortality analysis in the

French cohort of uranium millers (F-Millers), period 1968-2013. Int Arch Occup Environ
Health 91: 23-33. http://dx.doi.Org/10.1007/s00420-017-1254-7.

Branch. FM: Perry. MI: Chen. Z: Louis. GMB. (2021). Metal(loid)s and human semen quality: The
LIFE Study. Reprod Toxicol 106: 94-102.
http://dx.doi.0rg/lO.lOl6/i.reprotox.202i.lO.OO6.

Brown. RC: Barone. S. Tr: Kimmel. CA. (2008). Children's health risk assessment: Incorporating a
lifestage approach into the risk assessment process [Review], Birth Defects Res B Dev
Reprod Toxicol 83: 511-521. http://dx.doi.org/10.1002 /bdrb.20172.

Brugge. D: Buchner. V. (2011). Health effects of uranium: New research findings [Review], Rev
Environ Health 26: 231-249. http://dx.d0i.0rg/l 0.1515/REVEH.2011.032.

Brugge. D: Goble. R. (2002). The history of uranium mining and the Navajo people. Am J Public
Health 92: 1410-1419. http://dx.doi.Org/10.2105/ATPH.92.9.1410.

Butler-Dawson. I: Krisher. L: Dally. M: Tames. KA: Tohnson. RT: Taramillo. D: Yoder. H: Tohnson. EC:

Pilloni. D: Asensio. C: Cruz. A: Newman. LS. (2021). Sugarcane workweek study: Risk factors
for daily changes in creatinine. Kidney Int 6: 2404-2414.
http://dx.doi.0rg/lO.lOl6/i.ekir.202i.O6.OO3.

Chafe. R: Aslanov. R: Sarkar. A: Gregory. P: Comeau. A: Newhook. LA. (2018). Association of type 1
diabetes and concentrations of drinking water components in Newfoundland and Labrador,
Canada. BMJ Open Diabetes Res Care 6: e000466. http://dx.doi.org/10.1136/bmidrc-2017-
000466.

Chen. L: Sun. 0: Peng. S: Tan. T: Mei. G: Chen. H: Zhao. Y: Yao. P: Tang. Y. (2022a). Associations of
blood and urinary heavy metals with rheumatoid arthritis risk among adults in NHANES,
1999-2018. Chemosphere 289: 133147.
http://dx.doi.Org/10.1016/i.chemosphere.2021.133147.

Chen. L: Zhao. Y: Liu. F: Chen. H: Tan. T: Yao. P: Tang. Y. (2022b). Biological aging mediates the
associations between urinary metals and osteoarthritis among U.S. adults. BMC Med 20:
207. http://dx.doi.org/10.1186/sl2916-022-02403-3.

Choi. HI: Ko. HI: Kim. AS: Moon. H. (2019). The association between mineral and trace element
concentrations in hair and the 10-year risk of atherosclerotic cardiovascular disease in
healthy community-dwelling elderly individuals. Nutrients 11.
http://dx.doi.org/10.3390/nullQ30637.

Christensen. KLY. (2012). Metals in blood and urine, and thyroid function among adults in the
United States 2007-2008. Int J Hyg Environ Health 216: 624-632.
http://dx.doi.0rg/lO.lOl6/i.iiheh.2Oi2.O8.OO5.

Cohen. AM: Hersh. WR: Peterson. K: Yen. PY. (2006). Reducing workload in systematic review

preparation using automated citation classification. J Am Med Inform Assoc 13: 206-219.
http://dx.doi.org/10.1197/iamia.M1929.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

39

40

41

42

43

44

45

46

47

48

49

Protocol for the Uranium IRIS Assessment (Oral)

Crissman. TW: Goodman. DG: Hildebrandt. PK: Maronpot. RR: Prater. DA: Riley. TH: Seaman. WT:

Thake. DC. (2004). Best practices guideline: Toxicologic histopathology. Toxicol Pathol 32:
126-131. http://dx.doi.Org/10.1080/01926230490268756.

De Benedetti. S: Lucchini. G: Del Bo. C: Peon. V: Marocchi. A: Penco. S: Lunetta. C: Gianazza. E:

Bonomi. F: Iametti. S. (2017). Blood trace metals in a sporadic amyotrophic lateral sclerosis
geographical cluster. Biometals 30: 355-365. http://dx.doi.org/10.1007/sl0534-017-0011-
4.

Denisova. 0: Chernogorvuk. G: Baranovskava. N: Rikhvanov. L: Shefer. N: Cherniavskava. G:

Palchikova. I: Kalacheva. T. (2020). Trace elements in the lung tissue affected by sarcoidosis.
Biol Trace Elem Res 196: 66-73. http://dx.doi.org/l 0.1007/sl 2011 -019-01915-z.

Denisova. OA: Chernogorvuk. GE: Baranovskava. NY: Rikhvanov. LP: Chernyavskava. GM. (2018).

Macroelement and microelement composition of mediastinal lymph nodes of patients with
sarcoidosis. 27: 754-759. http://dx.doi.org/10.18093/0869-0189-2017-27-6-754-759.

Devssenroth. MA: Gennings. C: Liu. SH: Peng. SN: Hao. K: Lambertini. L: Tackson. BP: Karagas. MR:
Marsit. CI: Chen. 1. (2018). Intrauterine multi-metal exposure is associated with reduced
fetal growth through modulation of the placental gene network. Environ Int 120: 373-381.
http://dx.doi.0rg/lO.lOl6/i.envint.2Ol8.O8.OlO.

Dickersin. K. (1990). The existence of publication bias and risk factors for its occurrence. JAMA 263:
1385-1389.

Dinocourt. C: Culeux. C: Legrand. M: Elie. C: Lestaevel. P. (2017). Chronic exposure to uranium from
gestation: Effects on behavior and neurogenesis in adulthood. Int J Environ Res Public
Health 14. http://dx.doi.org/10.3390/iierphl4050536.

Domingo. TL: Paternain. TL: Llobet. TM: Corbella. I. (1989). The developmental toxicity of uranium in
mice. Toxicology 55: 143-152. http://dx.doi.org/10.1016/0300-483Xr89190181-9.

Duan. W: Xu. C: Liu. 0: Xu. I: Weng. Z: Zhang. X: Basnet. TB: Dahal. M: Gu. A. (2020). Levels of a

mixture of heavy metals in blood and urine and all-cause, cardiovascular disease and cancer
mortality: A population-based cohort study. Environ Pollut 263: 114630.
http://dx.doi.Org/10.1016/i.envpol.2020.114630.

Dublineau. I: Souidi. M: Gueguen. Y: Lestaevel. P: Bertho. TM: Manens. L: Delissen. 0: Grison. S:
Paulard. A: Monin. A: Kern. Y: Rouas. C: Loven. 1: Gourmelon. P: Aigueperse. 1. (2014).
Unexpected lack of deleterious effects of uranium on physiological systems following a
chronic oral intake in adult rat. BioMed Res Int 2014: 181989.
http://dx.d0i.0rg/l 0.1155/2014/181989.

EFSA (European Food Safety Authority). (2017). Guidance on the use of the weight of evidence
approach in scientific assessments. EFSA J15: 1-69.
http://dx.doi.Org/10.2903/i.efsa.2017.4971.

Elmhiri. G: Gloaguen. C: Grison. S: Kereselidze. D: Elie. C: Tack. K: Benderitter. M: Lestaevel. P:

Legendre. A: Souidi. M. (2018). DNA methylation and potential multigenerational epigenetic
effects linked to uranium chronic low-dose exposure in gonads of males and females rats.
Toxicol Lett 282: 64-70. http://dx.doi.Org/10.1016/i.toxlet.2017.10.004.

Erdei. E: Shuev. C: Pacheco. B: Caiero. M: Lewis. 1: Rubin. RL. (2019). Elevated autoimmunity in

residents living near abandoned uranium mine sites on the Navajo Nation. J Autoimmun 99:
15-23. http://dx.doi.0rg/lO.lOl6/i.iaut.2Oi9.Ol.OO6.

Evidence Partners. (2022). DistillerSR. Retrieved from

https://www.evidencepartners.com/products/distillersr-systematic-review-software/

Farland. WH. (2005). [Memo to Science Policy council regarding implementation of the cancer
guidelines and accompanying supplemental guidance - Science Policy Council Cancer
Guidelines. Implementation Workgroup communication I: Application of the mode of action
framework in mutagenicity determinations for carcinogenicity]. Available online at

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

Protocol for the Uranium IRIS Assessment (Oral)

https://www.epa.gov/sites/production/files/2015-
01/documents/cgiwgcommuniation i.pdf (accessed

Feng. W: He. X: Chen. M: Deng. S: Oiu. G: Li. X: Liu. C: Li. T: Deng. 0: Huang. S: Wang. T: Dai. X: Yang. B:
Yuan. T: He. M: Zhang. X: Chen. W: Kan. H: Wu. T. (2014). Urinary metals and heart rate
variability: A cross-sectional study of urban adults in Wuhan, China. Environ Health
Perspectl23: 217-222. http://dx.doi.org/10.1289/ehp.1307563.

Feng. W: Huang. X: Zhang. C: Liu. C: Cui. X: Zhou. Y: Sun. H: Oiu. G: Guo. H: He. M: Zhang. X: Yuan. 1:
Chen. W: Wu. T. (2015). The dose-response association of urinary metals with altered
pulmonary function and risks of restrictive and obstructive lung diseases: A population-
based study in China. BMJ Open 5: e007643. http://dx. doi.org/10.1136/bmiopen-2015-
007643.

Fiore. M: Barone. R: Copat. C: Grasso. A: Cristaldi. A: Rizzo. R: Ferrante. M. (2020). Metal and

essential element levels in hair and association with autism severity. J Trace Elem Med Biol
57: 126409. http://dx.doi.Org/10.1016/i.itemb.2019.126409.

Gilman. AP: Villeneuve. DC: Secours. YE: Yagminas. AP: Tracy. BL: Ouinn. TM: Valli. YE: Willes. RT:

Moss. MA. (1998). Uranyl nitrate: 28-day and 91-day toxicity studies in the Sprague-Dawley
rat. Toxicol Sci 41: 117-128. http: //dx.doi.org/10.1006/toxs.1997.2367.

Greene. AD: Kendziorski. TA: Buckholz. TM: Niu. L: Xie. C: Pinnev. SM: Burns. KA. (2019). Elevated
serum chemokines are independently associated with both endometriosis and uranium
exposure. Reprod Toxicol 84: 26-31. http://dx.doi.Org/10.1016/i.reprotox.2018.12.006.

Grison. S: Elmhiri. G: Gloaguen. C: Elie. C: Kereselidze. D: Tack. K: Lestaevel. P: Legendre. A: Manens.
L: Benadiaoud. MA: Lobaccaro. TM: Souidi. M. (2018). Low dose of uranium induces
multigenerational epigenetic effects in rat kidney. IntJ RadiatBiol 94: 975-984.
http://dx.doi.Org/10.1080/09553002.2018.1493242.

Grison. S: Fave. G: Maillot. M: Manens. L: Delissen. 0: Blanchardon. E: Banzet. N: Defoort. C: Bott. R:
Dublineau. I: Aigueperse. 1: Gourmelon. P: Martin. TC: Souidi. M. (2013). Metabolomics
identifies a biological response to chronic low-dose natural uranium contamination in urine
samples. Metabolomics 9: 1168-1180. http://dx.doi.org/10.1007/sll306-013-0544-7.

Grison. S: Fave. G: Maillot. M: Manens. L: Delissen. 0: Blanchardon. E: Dublineau. I: Aigueperse. 1:
Bohand. S: Martin. TC: Souidi. M. (2016). Metabolomics reveals dose effects of low-dose
chronic exposure to uranium in rats: identification of candidate biomarkers in urine
samples. Metabolomics 12: 154. http://dx.doi.org/10.1007/sll306-Q16-1092-8.

Grison. S: Kereselidze. D: Cohen. D: Gloaguen. C: Elie. C: Lestaevel. P: Legendre. A: Manens. L:

Habchi. B: Benadiaoud. MA: Tarlet. G: Milliat. F: Martin. TC: Lobaccaro. TM: Souidi. M. (2019).
Applying a multiscale systems biology approach to study the effect of chronic low-dose
exposure to uranium in rat kidneys. IntJ RadiatBiol 95: 737-752.
http://dx.doi.Org/10.1080/09553002.2019.1577567.

Grison. S: Legendre. A: Svilar. L: Elie. C: Kereselidze. D: Gloaguen. C: Lestaevel. P: Martin. TC: Souidi.
M. (2022). Multigenerational exposure to uranium changes sperm metabolome in rats.
International Journal of Molecular Sciences 23: 8349.
http://dx.doi.org/10.3390/iims23158349.

Gueguen. Y: Grandcolas. L: Baudelin. C: Grison. S: Tissandie. E: Tourdain. TR: Paquet. F: Voisin. P:
Aigueperse. I: Gourmelon. P: Souidi. M. (2007). Effect of acetaminophen administration to
rats chronically exposed to depleted uranium. Toxicology 229: 62-72.
http://dx.doi.0rg/lO.lOl6/i.tox.2OO6.lO.OO6.

Gueguen. Y: Rouas. C: Monin. A: Manens. L: Stefani. 1: Delissen. 0: Grison. S: Dublineau. I. (2014).
Molecular, cellular, and tissue impact of depleted uranium on xenobiotic-metabolizing
enzymes. Arch Toxicol 88: 227-239. http://dx.doi.org/10.1007/sQ0204-013-1145-y.

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

Protocol for the Uranium IRIS Assessment (Oral)

Guo. Y: Liu. L: Ni. W: Pan. Y: Chen. Y: Xie. 0: Liu. Y: Tin. L: Li. Z: Ren. A: Wang. L. (2020). Uranium

concentration in umbilical cord may increase the risk for orofacial clefts. Environ Res 182:
109103. http://dx.doi.Org/10.1016/i.envres.2019.109103.

Guseva Canu. I: Zhivin. S: Garsi. TP: Caer-Lorho. S: Samson. E: Collomb. P: Acker. A: Laurier. D.

(2014). [Effects of chronic uranium internal exposure on mortality: Results of a pilot study
among French nuclear workers]. Rev Epidemiol Sante Publique 62: 339-350.
http://dx.doi.Org/10.1016/i.respe.2014.09.006.

Hao. Y: Li. R: Leng. Y: Ren. T: Liu. T: Ai. G: Xu. H: Su. Y: Cheng. T. (2012). The reproductive effects in
rats after chronic oral exposure to low-dose depleted uranium. J RadiatRes (Tokyo) 53:
377-384. http://dx.doi.org/10.1269/irr.l 1192.

Hao. Y: Ren. T: Li. R: Liu. T: Yang. Z: Su. Y. (2013a). Immunological changes associated with chronic
ingestion of depleted uranium in rats. Health Phys 105: 3-10.
http://dx.doi.Org/10.1097/HP.0b013e31828730a9.

Hao. Y: Ren. 1: Liu. 1: Yang. Z: Liu. C: Li. R: Su. Y. (2013b). Immunological changes of chronic oral
exposure to depleted uranium in mice. Toxicology 309: 81-90.
http://dx.doi.Org/10.1016/i.tox.2013.04.013.

Harchaoui. H: Azzaoui. FZ: Achouri. I: Samih. M: Aboussaleh. Y: Ahami. APT. (2020). High

concentration of toxic metals in children's scalps is likely the cause of autism. Current
Toxicol 16: 109-112.

Harmon. ME: Lewis. I: Miller. C: Hoover. 1: Ali. AS: Shuev. C: Caiero. M: Lucas. S: Pacheco. B: Erdei. E:
Ramone. S: Nez. T: Campen. Ml: Gonzales. M. (2018). Arsenic association with circulating
oxidized low-density lipoprotein in a Native American community. J Toxicol Environ Health
A 81: 535-548. http://dx.doi.org/10.1080/15287394.2018.1443860.

Health Canada. (2019). Guidelines for Canadian drinking water quality guideline technical

document- Uranium [Standard], (H144-13/12-2019E). Ottawa, Ontario: Water and air
quality bureau, healthy environments and consumer safety branch, Health Canada.
https://www.canada.ca/content/dam/hc-sc/documents /services /publications /healthy-
living/guidelines/drinking- water-quality-uranium/uranium- mav-2019-eng.pdf.

Henriquez-Hernandez. LA: Boada. LP: Carranza. C: Perez-Arellano. TL: Gonzalez-Antuna. A:

Camacho. M: Almeida-Gonzalez. M: Zumbado. M: Luzardo. OP. (2017). Blood levels of toxic
metals and rare earth elements commonly found in e-waste may exert subtle effects on
hemoglobin concentration in sub-Saharan immigrants. Environ Int 109: 20-28.
http://dx.doi.Org/10.1016/i.envint.2017.08.023.

Higgins. TPT: Thomas. I: Chandler. I: Cumpston. M: Li. T: Page. Ml: Welch. VA. (2022). Cochrane
handbook for systematic reviews of interventions version 6.3. Higgins, JPT; Thomas, J;
Chandler, J; Cumpston, M; Li, T; Page, MJ; Welch, VA.
http: / /www, training. Cochrane. o r g /handbo ok.

Hill. AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58: 295-
300. http://dx.doi.Org/10.1177/003591576505800503.

Hoenig. TM: Heisev. DM. (2001). The abuse of power: The pervasive fallacy of power calculations for
data analysis. Am Stat 55: 19-24.

Howard. BE: Phillips. 1: Miller. K: Tandon. A: Mav. D: Shah. MR: Holmgren. S: Pelch. KE: Walker. V:
Roonev. AA: Macleod. M: Shah. RR: Thayer. K. (2016). SWIFT-Review: A text-mining
workbench for systematic review. Syst Rev 5: 87. http: //dx.doi.org/10.1186/sl3643-016-
0263-z.

Howard. BE: Phillips. 1: Tandon. A: Maharana. A: Elmore. R: Mav. D: Sedvkh. A: Thayer. K: Merrick.
BA: Walker. V: Roonev. A: Shah. RR. (2020). SWIFT-Active Screener: Accelerated document
screening through active learning and integrated recall estimation. Environ Int 138:

105623. http://dx.doi.Org/10.1016/i.envint.2020.105623.

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

Protocol for the Uranium IRIS Assessment (Oral)

Howe. CG: Nozadi. SS: Garcia. E: O'Connor. TG: Starling. AP: Farzan. SF: Tackson. BP: Madan. TC:

Alshawabkeh. AN: Cordero. TF: Bastain. TM: Meeker. TP: Breton. CV: Karagas. MR. (2022).
Prenatal metal(loid) mixtures and birth weight for gestational age: A pooled analysis of
three cohorts participating in the ECHO program. Environ Int 161: 107102.
http://dx.doi.Org/10.1016/i.envint.2022.107102.

Huang. X: Xie. 1: Cui. X: Zhou. Y: Wu. X: Lu. W: Shen. Y: Yuan. 1: Chen. W. (2016). Association between
concentrations of metals in urine and adult asthma: A case-control study in Wuhan, China.
PLoS ONE 11: e0155818. http://dx.doi.org/10.1371/iournal.pone.0155818.

ICRP (International Commission on Radiological Protection). (1993). Appendix A: Age-specific
biokinetic models for the alkaline earth elements and lead. Ann ICRP 23(3-4): 95-120.
http://dx.doi.org/10.1016/0146-6453r93190031-3.

ICRP (International Commission on Radiological Protection). (1995). Age-dependent doses to

members of the public from intake of radionuclides: Part 4: Inhalation dose coefficients (pp.
1-405). (ISSN 0146-6453

EISSN1872-969X

ICRP Publication 71). Oxford, United Kingdom: Pergamon. http://dx.doi.org/10.1016/SQ146-
6453C00180008-1.

IRIS (Integrated Risk Information System). (2018). IRIS assessment plan for uranium (Oral

reference dose) (Scoping and problem formulation materials) [EPA Report], (EPA/635/R-
17/787). Washington, DC.

https://nepis.epa.gov/Exe/ZyNET.exe/P1012GZX.txt?ZyActionD=ZyDocument&Client=EPA

&Index=2016%20Thru%202020&Docs=&0uery=Assessment%20Plan%20Uranium&Time

=&EndTime=&SearchMethod=2&TocRestrict=n&Toc=&TocEntry=&0Field=&0FieldYear=&

0FieldMonth=&0FieldDay=&Use0Field=&Int0FieldOp=0&Ext0FieldQp=0&Xml0uery=&Fil

e=D%3A%5CZYFILES%5CINDEX%20DATA%5C16THRU20%5CTXT%5C00000024%5CPl

012 GZX.txt&User=AN0NYM0U S&Password=anonymous&SortMethod=h%7C-

&MaximumDocuments=15&FuzzyDegree=0&ImageOuality=r85gl6/r85gl6/xl50yl50gl6

/i500&Display=hpfr&DefSeekPage=x&SearchBack=ZyActionL&Back=ZyActionS&BackDesc

=Results%20page&MaximumPages=l&ZyEntrv=l&SeekPage=x#.

Karakis. I: Landau. D: Gat. R: Shemesh. N: Tirosh. 0: Yitshak-Sade. M: Sarov. B: Novack. L. (2021).
Maternal metal concentration during gestation and pediatric morbidity in children: an
exploratory analysis. Environ Health Prev Med 26: 40. http: //dx.doi.org/10.1186/sl2199-
021-00963-z.

Kavembe-Kitenge. T: Kabange Umba. I: Musa Obadia. P: Mbuyi-Musanzavi. S: Nkulu Banza. P:

Katoto. P: Katshiez Nawei. C: Kalenga Ilunga. G: Haufroid. V: Banza Lubaba Nkulu. C: Nawrot.
T: Nemerv. B. (2020). Respiratory Health and Urinary Trace Metals among Artisanal Stone-
Crushers: A Cross-Sectional Study in Lubumbashi, DR Congo. Int J Environ Res Public Health
17. http: //dx.doi.org/10.3390/iierph 17249384.

Keith. LS: Faroon. 0M: Fowler. BA. (2015). Uranium. In G Nordberg; M Costa (Eds.), Handbook on
the Toxicology of Metals, vol 1 (4th ed., pp. 1307-1345). Amsterdam, the Netherlands:
Academic Press. http://dx.doi.org/10.1016/B978-0-444-59453-2.00059-7.

Kim. HP: Shin. HS: Song. DY: Lee. TH: Han. BY: Ahn. SK: Park. SH. (2012). Application of safeguards-
By-Design for the pyroprocessing facilities in the ROK. 40: 24-31.

Kim. K: Argos. M: Perskv. VW: Freels. S: Sargis. RM: Turvk. ME. (2022). Associations of exposure to
metal and metal mixtures with thyroid hormones: Results from the NHANES 2007-2012.
Environ Res 212: 113413. http://dx.doi.Org/10.1016/i.envres.2022.113413.

Kim. SS: Meeker. ID: Carroll. R: Zhao. SS: Mourgas. MI: Richards. Ml: Aung. M: Cantonwine. DE:

Mcelrath. TF: Ferguson. KK. (2018). Urinary trace metals individually and in mixtures in

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

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

Protocol for the Uranium IRIS Assessment (Oral)

association with preterm birth. Environ Int 121: 582-590.
http://dx.doi.Org/10.1016/i.envint:2018.09.052.

Kim. SS: Meeker. TP: Keil. AP: Aung. MT: Bommarito. PA: Cantonwine. DE: Mcelrath. TF: Ferguson.
KK. (2019). Exposure to 17 trace metals in pregnancy and associations with urinary
oxidative stress biomarkers. Environ Res 179: 108854.
http://dx.doi.Org/10.1016/i.envres.2019.108854.

Kocher. E: Rendon. KT: Kesler. D: Bovce. TW: Myers. 0: Evans. K: Cook. LS: Sood. A. (2016). Uranium
workers demonstrate lower lobe predominant irregular pneumoconiotic opacities on chest
radiographs. J Health Care Poor Underserved 27: 116-127.
http://dx.doi.org/10.1353/hpu.2016.0193.

Kocvlowski. R: Grzesiak. M: Gai. Z: Lorenc. W: Bakinowska. E: Baratkiewicz. D: von Kaisenberg. CS:
Lamers. Y: Suliburska. 1. (2019). Associations between the Level of Trace Elements and
Minerals and Folate in Maternal Serum and Amniotic Fluid and Congenital Abnormalities.
Nutrients 11. http://dx.doi.org/10.3390/null020328.

Kudvasheva. AG: Zagorskava. NG: Shishkina. LN. (2020). [Early and remote effects of gamma

irradiation and uranyl nitrate in the liver lipids of mice], Teoreticheskaya i Prikladnaya
Ekologiya (2): 157-165. http: //dx.doi.org/10.25750/1995-4301 -2020-2-157-165.

Legendre. A: Elie. C: Ramambason. C: Manens. L: Souidi. M: Froment. P: Tack. K. (2016). Endocrine
effects of lifelong exposure to low-dose depleted uranium on testicular functions in adult
rat. Toxicology 368-369: 58-68. http://dx.doi.Org/10.1016/i.tox.2016.08.014.

Legendre. A: Elmhiri. G: Gloaguen. C: Magneron. V: Kereselidze. D: Saci. N: Elie. C: Vavsset. E:
Benadiaoud. MM: Tack. K: Grison. S: Souidi. M. (2019). Multigenerational exposure to
uranium changes morphometric parameters and global DNA methylation in rat sperm. 342:
175-185. http://dx.doi.Org/10.1016/i.crvi.2019.07.002.

Legrand. M: Elie. C: Stefani. 1: N Flores. 1: Culeux. C: Delissen. 0: Ibanez. C: Lestaevel. P: Eriksson. P:
Dinocourt. C. (2016a). Cell proliferation and cell death are disturbed during prenatal and
postnatal brain development after uranium exposure. Neurotoxicology 52: 34-45.
http://dx.doi.Org/10.1016/i.neuro.2015.10.007.

Legrand. M: Lam. S: Anselme. I: Gloaguen. C: Ibanez. C: Eriksson. P: Lestaevel. P: Dinocourt. C.
(2016b). Exposure to depleted uranium during development affects neuronal
differentiation in the hippocampal dentate gyrus and induces depressive-like behavior in
offspring. Neurotoxicology 57: 153-162. http://dx.doi.Org/10.1016/i.neuro.2016.09.006.

Lestaevel. P: Airault. F: Racine. R: Bensoussan. H: Dhieux. B: Delissen. 0: Manens. L: Aigueperse. 1:
Voisin. P: Souidi. M. (2014). Influence of environmental enrichment and depleted uranium
on behaviour, cholesterol and acetylcholine in apolipoprotein E-deficient mice. J Mol
Neurosci 53: 469-479. http://dx.doi.Org/10.1007/sl2031-013-0038-0.

Lestaevel. P: Bensoussan. H: Dhieux. B: Delissen. 0: Vacher. CM: Dublineau. I: Voisin. P: Taouis. M.
(2013). Cerebral cortex and hippocampus respond differently after post-natal exposure to
uranium. J Toxicol Sci 38: 803-811. http://dx.doi.org/10.2131/its.38.803.

Lestaevel. P: Dhieux. B: Delissen. 0: Benderitter. M: Aigueperse. 1. (2015). Uranium modifies or not
behavior and antioxidant status in the hippocampus of rats exposed since birth. J Toxicol Sci
40: 99-107. http://dx.doi.org/10.2131/its.40.99.

Lestaevel. P: Grison. S: Fave. G: Elie. C: Dhieux. B: Martin. TC: Tack. K: Souidi. M. (2016). Assessment
of the central effects of natural uranium via behavioural performances and the
cerebrospinal fluid metabolome. Neural Plast 2016: 9740353.
http: / /dx.doi.org/10.1155/2016/9740353.

Lewicka. I: Kocvlowski. R: Grzesiak. M: Gai. Z: Sainog. A: Bar alkie wicz. D: von Kaisenberg. C:

Suliburska. I. (2019). Relationship between pre-pregnancy body mass index and mineral
concentrations in serum and amniotic fluid in pregnant women during labor. J Trace Elem
Med Biol 52: 136-142. http://dx.doi.Org/10.1016/i.itemb.2018.12.007.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

39

40

41

42

43

44

45

46

47

48

49

Protocol for the Uranium IRIS Assessment (Oral)

Li. X: Fan. Y: Zhang. Y: Huang. X: Huang. Z: Yu. M: Xu. 0: Han. X: Lu. C: Wang. X. (2021). Association

between selected urinary heavy metals and asthma in adults: a retrospective cross-sectional
study of the US National Health and Nutrition Examination Survey. Environ Sci Pollut Res
Int 28: 5833-5844. http://dx.doi.org/10.1007/sll356-020-10906-w.

Lin. YK: Liang. CS: Tsai. CK: Tsai. CL: Lee. IT: Sung. YF: Chou. CH: Shang. HS: Yang. BH: Lin. GY: Su.
MW: Yang. FC. (2022). A Metallomic Approach to Assess Associations of Plasma Metal
Levels with Amnestic Mild Cognitive Impairment and Alzheimer's Disease: An Exploratory
Study. J Clin Med 11. http://dx.doi.org/10.3390/icmlll33655.

Liu. B: Feng. W: Wang. 1: Li. Y: Han. X: Hu. H: Guo. H: Zhang. X: He. M. (2016). Association of urinary
metals levels with type 2 diabetes risk in coke oven workers. Environ Pollut 210: 1-8.
http://dx.doi.Org/10.1016/i.envpol.2015.ll.046.

Long. T: Wang. R: Wang. 1: Wang. F: Xu. Y: Wei. Y: Zhou. L: Zhang. X: Yuan. 1: Yao. P: Wei. S: Guo. H:
Yang. H: Wu. T: He. M. (2019). Plasma metals and cardiovascular disease in patients with
type 2 diabetes. Environ Int 129: 497-506. http://dx.doi.Org/10.1016/i.envint.2019.05.038.

Lourenco. 1: Pereira. R: Pinto. F: Caetano. T: Silva. A: Carvalheiro. T: Guimaraes. A: Goncalves. F:
Paiva. A: Mendo. S. (2013). Biomonitoring a human population inhabiting nearby a
deactivated uranium mine. Toxicology 305: 89-98.
http://dx.doi.Org/10.1016/i.tox.2013.01.011.

Lu-Fritts. PY: Kottvan. LC: Tames. TA: Xie. C: Buckholz. TM: Pinnev. SM: Harlev. IB. (2014). Association
of systemic lupus erythematosus with uranium exposure in a community living near a
uranium-processing plant: a nested case-control study. Arthritis and Rheumatology 66:
3105-3112. http: //dx.doi.org/10.1002 /art.38786.

Lu. B: Ran. Y: Wang. S: Li. I: Zhao. Y: Ran. X: Li. R: Hao. Y. (2021). Chronic oral depleted uranium

leads to reproductive damage in male rats through the ROS-hnRNP A2/B1-COX-2 signaling
pathway. Toxicology 449: 152666. http://dx.doi.Org/10.1016/i.tox.2020.152666.

Ma. 1: Wang. B: Gao. X: Wu. H: Wang. D: Li. N: Tan. I: Wang. 1: Yan. L. (2018). A comparative study of
the typical toxic metals in serum by patients of schizophrenia and healthy controls in China.
Psychiatry Res 269: 558-564. http://dx.doi.Org/10.1016/i.psychres.2018.08.114.

Ma. M: Wang. R: Xu. L: Xu. M: Liu. S. (2020). Emerging health risks and underlying toxicological
mechanisms of uranium contamination: Lessons from the past two decades [Review],
Environ Int 145: 106107. http://dx.doi.Org/10.1016/i.envint.2020.106107.

Makris. SL: Thompson. CM: Euling. SY: Selevan. SG: Sonawane. B. (2008). A lifestage-specific
approach to hazard and dose-response characterization for children's health risk
assessment [Review], Birth Defects Res B Dev Reprod Toxicol 83: 530-546.
http://dx.doi.org/10.1002/bdrb.20176.

Malamba-Lez. D: Tshala-Katumbav. D: Bito. V: Rigo. I: Kvandabike. R: Yolola. E: Katchunga. P: Koba-
Bora. B: Ngov-Nkulu. D. (2021). Concurrent Heavy Metal Exposures and Idiopathic Dilated
Cardiomyopathy: A Case-Control Study from the Katanga Mining Area of the Democratic
Republic of Congo. Int J Environ Res Public Health 18.
http://dx.doi.org/10.3390/iierphl8094956.

Manduca. P: Nairn. A: Signoriello. S. (2014). Specific association of teratogen and toxicant metals in
hair of newborns with congenital birth defects or developmentally premature birth in a
cohort of couples with documented parental exposure to military attacks: observational
study at A1 Shifa Hospital, Gaza, Palestine. Int J Environ Res Public Health 11: 5208-5223.
http://dx.doi.Org/10.3390/iierphll0505208.

Marie. I: Gehanno. IF: Bubenheim. M: Duval-Modeste. AB: Tolv. P: Dominique. S: Bravard. P: Noel. D:
Cailleux. AF: Benichou. 1: Levesque. H: Goulle. TP. (2017). Systemic sclerosis and exposure to
heavy metals: A case control study of 100 patients and 300 controls [Review], Autoimmun
Rev 16: 223-230. http://dx.doi.org/10.1016/i.autrev.2017.01.004.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

44

45

46

47

48

49

50

Protocol for the Uranium IRIS Assessment (Oral)

Mavnard. EA: Hodge. HC. (1949). Studies of the toxicity of various uranium compounds when fed to
experimental animals. In IC Voegtlin; HC Hodge (Eds.), Pharmacology and toxicology of
uranium compounds (pp. 309-376). New York, NY: McGraw-Hill.

Mckeating. PR: Clifton. VL: Hurst. CP: Fisher. 11: Bennett. WW: Perkins. AY. (2021). Elemental
metabolomics for prediction of term gestational outcomes utilising 18-week maternal
plasma and urine samples. Biol Trace Elem Res 199: 26-40.
http://dx.doi.org/10.1007/sl 2011 -020-02127-6.

McKeating. PR: Fisher. 11: Zhang. P: Bennett. WW: Perkins. AY. (2020). Elemental metabolomics in
human cord blood: Method validation and trace element quantification. J Trace Elem Med
Biol 59: 126419. http://dx.doi.org/l 0.1016/i.itemb.2019.126419.

Medina. S: Lauer. FT: Castillo. EF: Bolt. AM: Ali. AS: Liu. KT: Burchiel. SW. (2020). Exposures to

uranium and arsenic alter intraepithelial and innate immune cells in the small intestine of
male and female mice. Toxicol Appl Pharmacol 403: 115155.
http://dx.doi.Org/10.1016/i.taap.2020.115155.

Mendv. A: Gasana. 1: Vieira. ER. (2012). Urinary heavy metals and associated medical conditions in
the US adult population. Int J Environ Health Res 22: 105-118.
http://dx.doi.org/10.1080/09603123.2011.605877.

Menke. A: Guallar. E: Cowie. CC. (2016). Metals in Urine and Piabetes in U.S. Adults. Piabetes 65:
164-171. http://dx.doi.org/10.2337/dbl5-0316.

Middlecamp. CH: Phillips. MF: Bentlev. AK: Baldwin. 0. (2006). Chemistry, society, and civic
engagement (part 2): uranium and American Indians. J Chem Educ 83: 1308.

Nanavakkara. S: Senevirathna. STM. LP: Harada. KH: Chandraiith. R: Hitomi. T: Abevsekera. T: Muso.
E: Watanabe. T: Koizumi. A. (2019). Systematic evaluation of exposure to trace elements
and minerals in patients with chronic kidney disease of uncertain etiology (CKPu) in Sri
Lanka. J Trace Elem Med Biol 54: 206-213. http://dx.doi.Org/10.1016/i.itemb.2019.04.019.

NCEA (National Center for Environmental Assessment). (1989). Uranium, soluble salts: IRIS
summary. U.S. Environmental Protection Agency.

Nozadi. SS: Li. L: Luo. L: Mackenzie. P: Erdei. E: Pu. R: Roman. CW: Hoover. 1: O'Ponald. E: Burnette.
C: Lewis. 1. (2021). Prenatal metal exposures and infants' developmental outcomes in a
Navajo population. Int J Environ Res Public Health 19: 425.
http://dx.doi.org/10.3390/iierphl9010425.

NRC (National Research Council). (1988). Health risks of radon and other internally deposited
alpha emitters: BEIRIV. Washington, PC.

http://books.nap.edu/openbook.php7record id=1026&page=Rl.

NRC (National Research Council). (2012). Uranium mining in Virginia: Scientific, technical,

environmental, human health and safety, and regulatory aspects of uranium mining and
processing in Virginia. Washington, PC: National Academies Press.
http://dx.doi.org/10.17226/13266.

NRC (National Research Council). (2014). Review of EPA's Integrated Risk Information System
(IRIS) process. Washington, PC: The National Academies Press.
http://dx.doi.org/10.17226/18764.

NTP (National Toxicology Program). (2015). Handbook for conducting a literature-based health

assessment using OHAT approach for systematic review and evidence integration. Research
Triangle Park, NC: U.S. Peptartment of Health and Human Services, National Toxicology
Program, Office of Health Assessment and Translation.
https://ntp.niehs.nih.gov/ntp/ohat/pubs/handbookian2015 508.pdf.

NTP (National Toxicology Program). (2019). Handbook for conducting a literature-based health

assessment using OHAT approach for systematic review and evidence integration. Research
Triangle, NC: National Institute of Environmental Health Sciences.
https://ntp.niehs.nih.gov/ntp/ohat/pubs/handbookmarch2019 508.pdf.

This document is a draft for review purposes only and does not constitute Agency policy.

R-9	PRAFT-P0 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

Protocol for the Uranium IRIS Assessment (Oral)

Okaneku. I: Vearrier. D: Mckeever. R: Lasala. G: Greenberg. MI. (2015). Urine uranium

concentrations and renal function in residents of the United States-2001 to 2010. Clin
Toxicol 53: 931-934. http://dx.doi.org/10.3109/15563650.2015.1094704.

Oruc. M: Mercan. S: Bakan. S: Kose. S: Ikitimur. B: Trabulus. S: Altiparmak. MR. (2022). Do trace
elements play a role in coronary artery calcification in hemodialysis patients? Int Urol
Nephrol 55: 173-182. http://dx.doi.org/l 0.1007/sl 1255-022-03303-4.

Park. RM: An. Y. (2022). Continuous NHANES survey data for environmental ambient and

occupational hazard identification—Feasibility and preliminary findings for osteoporosis
and kidney disease. J Occup Environ Hyg 19: 489-499.
http://dx.doi. org/10.1080/15459624.2022.2088769.

Pavlvushchik. 0: Afonin. V: Fatvkhava. S: Shabunva. P: Sarokina. V: Khapaliuk. A. (2017). Macro- and
Microelement Status in Animal and Human Hypertension: the Role of the ACE Gene I/D
Polymorphism. Biol Trace Elem Res 180: 110-119. http://dx.doi.org/10.1007/sl2011-017-
0990-6.

Poisson. C: Rouas. C: Manens. L: Dublineau. I: Gueguen. Y. (2014a). Antioxidant status in rat kidneys
after coexposure to uranium and gentamicin. Hum Exp Toxicol 33: 136-147.
http://dx.doi.org/10.1177/0960327113493297.

Poisson. C: Stefani. 1: Manens. L: Delissen. 0: Suhard. D: Tessier. C: Dublineau. I: Gueguen. Y.

(2014b). Chronic uranium exposure dose-dependently induces glutathione in rats without
any nephrotoxicity. Free Radic Res 48: 1218-1231.
http://dx.doi.org/10.3109/10715762.2014.945441.

Rahman. HH: Niemann. D: Munson-Mcgee. SH. (2022a). Association between environmental toxic
metals, arsenic and polycyclic aromatic hydrocarbons and chronic obstructive pulmonary
disease in the US adult population. Environ Sci Pollut Res Int 29: 54507-54517.
http://dx.doi.org/10.1007/sll356-022-19695-w.

Rahman. HH: Niemann. D: Munson-Mcgee. SH. (2022b). Urinary metals, arsenic, and polycyclic

aromatic hydrocarbon exposure and risk of chronic bronchitis in the US adult population.
Environ Sci Pollut Res Int 29: 73480-73491. http: //dx.doi.org/10.1007/sll356-022-
20982-9.

Rahman. HH: Niemann. D: Munson-Mcgee. SH. (2022c). Urinary Metals, Arsenic, and Polycyclic
Aromatic Hydrocarbon Exposure and Risk of Self-reported Emphysema in the US Adult
Population. Lung 200: 237-249. http://dx.doi.org/l 0.1007/s00408-022-00518-1.

Rahman. HH: Niemann. D: Munson-Mcgee. SH. (2022d). Urinary polycyclic aromatic hydrocarbon,
arsenic, and metal exposure and correlation with emphysema in smokers. Toxicol Appl
Pharmacol 450: 116168. http://dx.doi.Org/10.1016/i.taap.2022.116168.

Rango. T: Teuland. M: Manthrithilake. H: Mccornick. P. (2015). Nephrotoxic contaminants in

drinking water and urine, and chronic kidney disease in rural Sri Lanka. Sci Total Environ
518-519: 574-585. http://dx.doi.Org/10.1016/i.scitotenv.2015.02.097.

Rhaifal-Sahlanee. M: Ramli. R: Ali. M: Tawfiq. N: Rahman. A: Mustafa. I: Azman. N: Razak. N: Yahava.
N: Al-Marri. H: Avob. N: Zakaria. N. (2016). Analysis of uranium concentration on maternal
and umbilical cord blood samples after delivery in Iraq. In Ieee (Ed.), Proceedings of the 6th
IEEE International Conference on Control System, Computing and Engineering (ICCSCE
2016) (pp. 360-365). New York, NY: IEEE.
http://dx.doi. org/10.1109 /TCCSCE.2016.7893599.

Richardson. DB: Rage. E: Demers. PA: Do. MT: Debono. N: Fenske. N: Deffner. V: Kreuzer. M: Samet.
I: Wiggins. C: Schubauer-Berigan. MK: Kelly-Reif. K: Tomasek. L: Zablotska. LB: Laurier. D.
(2021). Mortality among uranium miners in North America and Europe: the Pooled
Uranium Miners Analysis (PUMA). Int J Epidemiol 50: 633-643.
http: / /dx. do i. or g /10.109 3 /ii e /dvaa 195.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

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

Protocol for the Uranium IRIS Assessment (Oral)

Rodrigues. G: Arruda-Neto. TP: Pereira. RM: Kleeb. SR: Geraldo. LP: Primi. MC: Takavama. L:

Rodrigues. TE: Cavalcante. GT: Genofre. GC: Semmler. R: Nogueira. GP: Fontes. EM. (2013).
Uranium deposition in bones of Wistar rats associated with skeleton development. Appl
Radiatlsot 82: 105-110. http://dx.doi.Org/10.1016/i.apradiso.2013.07.033.

Roos. PM: Vesterberg. 0: Svversen. T: Flaten. TP: Nordberg. M. (2013). Metal Concentrations in
Cerebrospinal Fluid and Blood Plasma from Patients with Amyotrophic Lateral Sclerosis.
Biol Trace Elem Res 151: 159-170. http://dx.doi.org/l 0.1007/sl 2011-012-9547-x.

Rouas. C: Stefani. 1: Grison. S: Grandcolas. L: Baudelin. C: Dublineau. I: Pallardv. M: Gueguen. Y.
(2011). Effect of nephrotoxic treatment with gentamicin on rats chronically exposed to
uranium. Toxicology 279: 27-35. http://dx.doi.Org/10.1016/i.tox.2010.09.003.

Saint-Marc. B: Elie. C: Manens. L: Tack. K: Benderitter. M: Gueguen. Y: Ibanez. C. (2016). Chronic
uranium contamination alters spinal motor neuron integrity via modulation of SMN1
expression and microglia recruitment. Toxicol Lett 254: 37-44.
http://dx.doi.Org/10.1016/i.toxlet2016.05.004.

Samson. E: Piot. I: Zhivin. S: Richardson. DB: Laroche. P: Serond. AP: Laurier. D: Laurent. 0. (2016).
Cancer and non-cancer mortality among French uranium cycle workers: the TRACY cohort
BMJ Open 6: e010316. http://dx.doi.org/10.1136/bmiopen-2015-010316.

Sankar. P: Telang. AG: Ramva. K: Viiayakaran. K: Kesavan. M: Sarkar. SN. (2014). Protective action of
curcumin and nano-curcumin against arsenic-induced genotoxicity in rats in vivo. Mol Biol
Rep 41: 7413-7422. http://dx.doi.org/10.1007/sll033-014-3629-0.

Sattouf. M: Kratz. S: Diemer. K: Rienitz. 0: Fleckenstein. 1: Schiel. D: Schnug. E. (2007). Identifying
the origin of rock phosphates and phosphorus fertilizers through high-precision
measurement of the strontium isotopes 87 Sr and 86 Sr. Landbauforschung Voelkenrode
57: 1-11.

Savabieasfahani. M: Basher Ahamadani. F: Mahdavi Damghani. A. (2020). Living near an active U.S.
military base in Iraq is associated with significantly higher hair thorium and increased
likelihood of congenital anomalies in infants and children. Environ Pollut 256: 113070.
http://dx.doi.Org/10.1016/i.envpol.2019.113070.

Scammell. MK: Sennett. C: Laws. RL: Rubin. RL: Brooks. PR: Amador. IT: Lopez-Pilarte. D: Ramirez-
Rubio. 0: Friedman. DT: Mcclean. MP: Lewis. 1: Erdei. E. (2020). Urinary metals
concentrations and biomarkers of autoimmunity among Navajo and Nicaraguan men. Int J
Environ Res Public Health 17: 1-17. http://dx.doi.org/10.3390/iierphl7155263.

Schiinemann. H: Brozek. I: Guvatt. G: Oxman. A. (2013). GRADE handbook. Available online at
https://gdt.gradepro.org/app/handbook/handbook.html (accessed April 22, 2022).

Sciome. (2019). SWIFT Review: Evidence stream search strategies. Sciome.

https://www.sciome.eom/wp-content/uploads/2019/08/SWIFT-Review-Search-
Str ate gie s- E vide nee - Str e am. docx.

Shaki. F: Zamani. E: Arimand. A: Pourahmad. 1. (2019). AReviewonToxicodynamics of Depleted
Uranium [Review], Iran J Pharm Res 18: 90-100.
http://dx.doi. org/10.22037/iipr.2020.113045.14085.

Shelley. R: Kim. NS: Parsons. PI: Lee. BK: Agnew. I: Taar. BG: Steuerwald. AT: Matanoski. G:

Fadrowski. 1: Schwartz. BS: Todd. AC: Simon. D: Weaver. VM. (2014). Uranium associations
with kidney outcomes vary by urine concentration adjustment method. J Expo Sci Environ
Epidemiol 24: 58-64. http://dx.doi.org/10.1038/ies.2013.18.

Sheppard. SC: Sheppard. MI: Gallerand. M0: Sanipelli. B. (2005). Derivation of ecotoxicity thresholds
for uranium [Review], J Environ Radioact 79: 55-83.
http://dx.doi.Org/10.1016/i.ienvrad.2004.05.015.

Shiue. I. (2013). Urinary environmental chemical concentrations and vitamin D are associated with
vision, hearing, and balance disorders in the elderly. Environ Int 53: 41-46.
http://dx.doi.0rg/lO.lOl6/i.envint.2Oi2.i2.OO6.

This document is a draft for review purposes only and does not constitute Agency policy.

R-ll	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

Protocol for the Uranium IRIS Assessment (Oral)

Shiue. I. (2014). Relationship of environmental exposures and ankylosing spondylitis and spinal
mobility: US NHAENS, 2009-2010. Int J Environ Health Res 25: 1-8.
http://dx.doi.Org/10.1080/09603123.2014.945512.

Shiue. I. (2015). Urinary arsenic, heavy metals, phthalates, pesticides, polyaromatic hydrocarbons
but notparabens, polyfluorinated compounds are associated with self-rated health: USA
NHANES, 2011-2012. Environ Sci PollutRes Int 22: 9570-9574.
http://dx.doi. org/10.1007/sl 1356-015-4604-6.

Shiue. I: Hristova. K. (2014). Higher urinary heavy metal, phthalate and arsenic concentrations
accounted for 3-19% of the population attributable risk for high blood pressure: US
NHANES, 2009-2012. Hypertens Res 37: 1075-1081.
http://dx.doi.org/10.1038/hr.2014.121.

Shumate. AM: Yeoman. K: Victoroff. T: Evans. K: Karr. R: Sanchez. T: Sood. A: Lanev. AS. (2017).

Morbidity and health risk factors among New Mexico miners: A comparison across mining
sectors. J Occup Environ Med 59: 789-794.
http://dx.doi.Org/10.1097/TOM.0000000000001078.

Souidi. M: Gueguen. Y: Linard. C: Dudoignon. N: Grison. S: Baudelin. C: Marquette. C: Gourmelon. P:
Aigueperse. 1: Dublineau. I. (2005). In vivo effects of chronic contamination with depleted
uranium on CYP3A and associated nuclear receptors PXR and CAR in the ray. Toxicology
214: 113-122. http://dx.doi.Org/10.1016/i.tox.2005.06.006.

Souidi. M: Racine. R: Grandcolas. L: Grison. S: Stefani. I: Gourmelon. P: Lestaevel. P. (2012). Influence
of depleted uranium on hepatic cholesterol metabolism in apolipoprotein E-deficient mice. J
Steroid Biochem Mol Biol 129: 201-205. http://dx.doi.Org/10.1016/i.isbmb.2011.12.007.

Souidi. M: Wade-Gueve. NM: Manens. L: Blanchardon. E: Aigueperse. I. (2018). [Uranium's effects on
bone integrity], E R S 17: 31-39. http://dx.doi.org/10.1684/ers.2017.1119.

Sterne. TAC: Hernan. MA: Reeves. BC: Savovic. I: Berkman. ND: Viswanathan. M: Henry. D: Altman.

DG: Ansari. MT: Boutron. I: Carpenter. TR: Chan. AW: Churchill. R: Peeks. IT: Hrobiartsson. A:
Kirkham. 1: Tiini. P: Loke. YK: Pigott. TP: Ramsay. CR: Regidor. D: Rothstein. HR: Sandhu. L:
Santaguida. PL: Schiinemann. HI: Shea. B: Shrier. I: Tugwell. P: Turner. L: Valentine. TC:
Waddington. H: Waters. E: Wells. GA: Whiting. PF: Higgins. TPT. (2016). ROBINS-I: A tool for
assessing risk of bias in non-randomised studies of interventions. BMJ 355: i4919.
http://dx.doi.org/10.1136/bmi.i4919.

Stoisavlievic. A: Rovcanin. B: Tagodic. 1: Radoikovic. DP: Paunovic. I: Gavrovic-lankulovic. M:
Manoilovic. P. (2020a). Significance of arsenic and lead in Hashimoto's thyroiditis
demonstrated on thyroid tissue, blood, and urine samples. Environ Res 186: 109538.
http://dx.doi.Org/10.1016/i.envres.2020.109538.

Stoisavlievic. A: Rovcanin. B: Krstic. P: Borkovic-Mitic. S: Paunovic. I: Piklic. A: Gavrovic-lankulovic.
M: Manoilovic. P. (2020b). Risk assessment of toxic and essential trace metals on the
thyroid health at the tissue level: The significance of lead and selenium for colloid goiter
disease. Exposure and Health 12: 255-264. http://dx.doi.org/10.1007/sl2403-019-003Q9-
9.

Stoisavlievic. A: Rovcanin. B: Krstic. P: Borkovic-Mitic. S: Paunovic. I: Kodranov. I: Gavrovic-
lankulovic. M: Manoilovic. P. (2019). Evaluation of trace metals in thyroid tissues:
Comparative analysis with benign and malignant thyroid diseases. Ecotoxicol Environ Saf
183: 109479. http://dx.doi.org/l0.1016/i.ecoenv.2019.109479.

Strand. LA: Martinsen. II: Borud. EK. (2014). Cancer risk and all-cause mortality among Norwegian
military United Nations peacekeepers deployed to Kosovo between 1999 and 2011. Cancer
PetectPrev 38: 364-368. http://dx.doi.Org/10.1016/i.canep.2014.04.003.

Su. M: Zhang. T: Zhao. T: Li. F: Ni. Y: Wang. X: Chen. T: Zhao. A: Oiu. Y: Bao. Y: Tia. W: Tia. W. (2012).
Human gouty arthritis is associated with a distinct serum trace elemental profile.
Metallomics 4: 244-252. http://dx.doi.org/10.1039/c2mt00178k.

This document is a draft for review purposes only and does not constitute Agency policy.

R-12	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

Protocol for the Uranium IRIS Assessment (Oral)

Sukhn. C: Awwad. I: Ghantous. A: Zaatari. G. (2018). Associations of semen quality with non-
essential heavy metals in blood and seminal fluid: Data from the Environment and Male
Infertility (EMI) study in Lebanon. J Assist Reprod Genet 35: 1691-1701.
http://dx.doi.Org/10.1007/sl0815-018-1236-z.

Suliburska. 1: Kocvlowski. R: Komorowicz. I: Grzesiak. M: Bogdariski. P: Baratkiewicz. D. (2016).

Concentrations of Mineral in Amniotic Fluid and Their Relations to Selected Maternal and
Fetal Parameters. Biol Trace Elem Res 172: 37-45. http://dx.doi.org/10.1007/sl2011-015-
0557-3.

Tissandie. E: Gueguen. Y: Lobaccaro. TM: Grandcolas. L: Voisin. P: Aigueperse. 1: Gourmelon. P:

Souidi. M. (2007). In vivo effects of chronic contamination with depleted uranium on
vitamin D3 metabolism in rat Biochim Biophys Acta 1770: 266-272.
http://dx.doi.0rg/lO.lOl6/i.bbagen.2OO6.lO.OO6.

Torrente. M: Gascon. M: Vriiheid. M: Sunver. 1: Forns. I: Domingo. 1: Nadal. M. (2013). Levels of

Metals in Hair in Childhood: Preliminary Associations with Neuropsychological Behaviors.
Toxics 2: 1-16. http://dx.doi.org/10.3390/toxics2010001.

Tret'iakov. SV: Shpagina. LA: Kuznetsova. GV: Shurkevich. AA. (2011). [Depending on exposure,

functional state of left ventricle in combined cardiovascular disease among chronic uranium
intoxication patients], Med Tr Prom Ekol (3): 37-40.

Tretvakov. SV: Khabarova. EA: Ermakova. MA. (2011). [Cognitive disorders in occupational diseases
of after contact period, associated with cardiovascular problems], Med Tr Prom Ekol (10):
27-32.

U.S. EPA (U.S. Environmental Protection Agency). (1988). Recommendations for and documentation
of biological values for use in risk assessment [EPA Report], (EPA600687008). Cincinnati,
OH. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855.

U.S. EPA (U.S. Environmental Protection Agency). (1989). Uranium, soluble salts; no CASRN.
Chemical assessment summary. Washington, DC: National Center for Environmental
Assessment, Integrated Risk Information System.

https://cfpub.epa.gov/ncea/iris/iris documents/documents/subst/0421 summarv.pdf.

U.S. EPA (U.S. Environmental Protection Agency). (1991). Guidelines for developmental toxicity risk
assessment Fed Reg 56: 63798-63826.

U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
reference concentrations and application of inhalation dosimetry [EPA Report],
(EPA600890066F). Research Triangle Park, NC.

https://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=71993&CFID=51174829&CFTOKE
N=25006317.

U.S. EPA (U.S. Environmental Protection Agency). (1996). Guidelines for reproductive toxicity risk
assessment [EPA Report], (EPA/630/R-96/009). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=30004Y0B.txt.

U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for neurotoxicity risk

assessment [EPA Report], (EPA/630/R-95/001F). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum, http: //www.epa.gov/risk/guidelines-
neurotoxicity-risk-assessment.

U.S. EPA (U.S. Environmental Protection Agency). (2000). Science policy council handbook: risk

characterization. Office of Science Policy, Office of Research and Development, Washington,
DC; EPA 100-B-00-002. http://www.epa.gov/iris/backgr-d.htm.

U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and
reference concentration processes [EPA Report], (EPA630P02002F). Washington, DC.
https://www.epa.gov/sites/production/files/2014-12/documents/rfd-final.pdf.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

39

40

41

42

43

44

45

46

47

48

49

Protocol for the Uranium IRIS Assessment (Oral)

U.S. EPA (U.S. Environmental Protection Agency). (2004). Toxicological review of boron and

compounds. In support of summary information on the Integrated Risk Information System
(IRIS) [EPA Report], (EPA/635/04/052). Washington, DC: U.S. Environmental Protection
Agency, IRIS. http://nepis.epa.gov/exe/ZyPURL.cgi?Dockey=P1006CK9.txt.

U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines for carcinogen risk
assessment [EPA Report], (EPA630P03001F). Washington, DC.
https://www.epa.gov/sites/production/files/2Q13-
09/documents/cancer guidelines final 3-25-05.pdf.

U.S. EPA (U.S. Environmental Protection Agency). (2005b). Supplemental guidance for assessing

susceptibility from early-life exposure to carcinogens [EPA Report], (EPA/630/R-03/003F).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/risk/supplemental-guidance-assessing-susceptibility-early-life-
exposure-carcinogens.

U.S. EPA (U.S. Environmental Protection Agency). (2006a). Depleted uranium: Technical brief [EPA
Report], (EPA-402-R-06-011). Washington, DC: U.S. Environmental Protection Agency,

Office of Air and Radiation. http://nepis.epa.gov/exe/ZyPURL.cgi?Dockey=P1005MCQ.txt.

U.S. EPA (U.S. Environmental Protection Agency). (2006b). A framework for assessing health risk of
environmental exposures to children (pp. 1-145). (EPA/600/R-05/093F). Washington, DC:
U.S. Environmental Protection Agency, Office of Research and Development, National Center
for Environmental Assessment.

http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 58363.

U.S. EPA (U.S. Environmental Protection Agency). (2011a). Recommended use of body weight 3/4
as the default method in derivation of the oral reference dose. (EPA100R110001).
Washington, DC. https://www.epa.gov/sites/production/files/2013-
09/documents/recommended-use-of-bw3 4.pdf.

U.S. EPA (U.S. Environmental Protection Agency). (2011b). Toxicological review of

trichloroethylene (CAS No. 79-01-6) in support of summary Information on the Integrated
Risk Information System (IRIS) [EPA Report], (EPA635R09011F). Washington, DC.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100CB6V.txt.

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
[EPA Report], (EPA100R12001). 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). (2014). Guidance for applying quantitative data to
develop data-derived extrapolation factors for interspecies and intraspecies extrapolation
[EPA Report], (EPA/100/R-14/002F). Washington, DC: Risk Assessment Forum, Office of
the Science Advisor, https://www.epa.gov/sites/production/files/2015-
01/documents/ddef-final.pdf.

U.S. EPA (U.S. Environmental Protection Agency). (2015). Peer review handbook [EPA Report] (4th
ed.). (EPA/100/B-15/001). Washington, DC: U.S. Environmental Protection Agency, Science
Policy Council, https: //www.epa.gov/osa/peer-review-handbook-4th-edition-2015.

U.S. EPA (U.S. Environmental Protection Agency). (2017). Guidance to assist interested persons in
developing and submitting draft risk evaluations under the Toxic Substances Control Act
(EPA/740/R17/001). Washington, DC: U.S Environmental Protection Agency, Office of
Chemical Safety and Pollution Prevention.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

Protocol for the Uranium IRIS Assessment (Oral)

https://www.epa.gov/sites/production/files/2017-
06/documents/tsca ra guidance final.pdf.

U.S. EPA (U.S. Environmental Protection Agency). (2018a). Chemistry Dashboard. Washington, DC.

Retrieved from https: //comptox.epa.gov/dashboard
U.S. EPA (U.S. Environmental Protection Agency). (2018b). An umbrella Quality Assurance Project
Plan (QAPP) for PBPK models [EPA Report], (ORD QAPP ID No: B-0030740-QP-1-1).
Research Triangle Park, NC.

U.S. EPA (U.S. Environmental Protection Agency). (2019). ChemView [Database], Retrieved from

https://chemview.epa.gov/chemview
U.S. EPA (U.S. Environmental Protection Agency). (2021). Annual progress update: Federal actions
to address uranium contamination on Navajo Nation. Washington, DC.
https://www.epa.gov/system/files/documents/2Q22-ll/ten-year-plan-annual-progress-
update- nnaum- 2 0 2 2 - 0 7.pdf.

U.S. EPA (U.S. Environmental Protection Agency). (2022a). ORD staff handbook for developing IRIS
assessments [EPA Report], (EPA 600/R-22/268). Washington, DC: U.S. Environmental
Protection Agency, Office of Research and Development, Center for Public Health and
Environmental Assessment

https://cfpub.epa.gov/ncea/iris drafts/recordisplay.cfm?deid=356370.

U.S. EPA (U.S. Environmental Protection Agency). (2022b). Risk assessment: Conducting a human
health risk assessment Available online at https: //www.epa.gov/risk/conducting-human-
health-risk-assessment (accessed November 7, 2023).

U.S. EPA (U.S. Environmental Protection Agency). (2023a). CompTox chemicals dashboard:

Uranium. Available online at

https://comptox.epa.gov/dashboard/chemical/details/DTXSID1042522 (accessed October
18,2023).

U.S. EPA (U.S. Environmental Protection Agency). (2023b). Radioactive waste from uranium mining
and milling. Available online at https: //www.epa.gov/radtown/radioactive-waste-uranium-
mining- and-milling (accessed December 11, 2023).

Ulrich. AE: Schnug. E: Prasser. HM: Frossard. E. (2014). Uranium endowments in phosphate rock.

Sci Total Environ 478: 226-234. http://dx.doi.Org/10.1016/i.scitotenv.2014.01.069.

UNEP (United Nations Environment Programme). (2022). The environmental impact of the conflict
in Ukraine: A preliminary review. Nairobi, Kenya: United Nations Environment Programme
:: UNEP. https://wedocs.unep.org/20.500.11822/4Q746.

UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). (2017). Annex
D: Biological effects of selected internal emitters—Uranium. In Sources, effects and risks of
ionizing radiation: Report to the General Assembly with scientific annexes (pp. 361-502).
New York, NY: United Nations.

https://www.unscear.org/unscear/uploads/documents/publications/UNSCEAR 2016 Ann
ex-D-CORRpdf.

USE PA OGWDW (USEPA Office of Ground Water and Drinking Water). (2000). Radionuclides notice
of data availability technical support document (NTIS/02926440_a). Washington, DC:
United States Environmental Protection Agency.
https://www.epa.gov/sites/production/files/2015-

09/documents/2009 04 16 radionuclides regulation radionuclides rulemaking techsuppo
rtdoc.pdf.

van Gerwen. M: Alpert. N: Lieberman-Cribbin. W: Cooke. P: Ziadkhanpour. K: Liu. B: Genden. E.

(2020). Association between Uranium Exposure and Thyroid Health: A National Health and
Nutrition Examination Survey Analysis and Ecological Study. Int J Environ Res Public Health
17. http://dx.doi.org/10.3390/iierphl7030712.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

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

Protocol for the Uranium IRIS Assessment (Oral)

Vicente-Vicente. L: Ferreira. L: Gonzalez-Buitrago. TM: Lopez-Hernandez. FT: Lopez-Novoa. TM:
Morales. AI. (2013). Increased urinary excretion of albumin, hemopexin, transferrin and
VDBP correlates with chronic sensitization to gentamicin nephrotoxicity in rats. Toxicology
304: 83-91. http://dx.doi.Org/10.1016/i.tox.2012.12.006.

Wade-Gueve. NM: Delissen. 0: Gourmelon. P: Aigueperse. 1: Dublineau. I: Souidi. M. (2012). Chronic
exposure to natural uranium via drinking water affects bone in growing rats. Biochim
Biophys Acta 1820: 1121-1127. http://dx.doi.Org/10.1016/i.bbagen.2012.04.019.

Wang. X: Karvonen-Gutierrez. CA: Herman. WH: Mukheriee. B: Harlow. SD: Park. SK. (2020).

Urinary metals and incident diabetes in midlife women: Study of Women's Health Across
the Nation (SWAN). BMJ Open Diabetes Res Care 8: e001233.
http://dx.doi.org/10.1136/bmidrc-2020-001233.

Wang. X: Xiao. P: Wang. R: Luo. C: Zhang. Z: Yu. S: Wu. 0: Li. Y: Zhang. Y: Zhang. H: Zhao. X. (2022).
Relationships between urinary metals concentrations and cognitive performance among
U.S. older people in NHANES 2011-2014. Front Public Health 10: 985127.
http://dx.doi.org/10.3389/fpubh.2022.985127.

Wang. YX: Sun. Y: Huang. Z: Wang. P: Feng. W: Li. 1: Yang. P: Wang. M: Sun. L: Chen. YT: Liu. C: Yue. 1:
Gu. LI: Zeng. 0: Lu. WO. (2016). Associations of urinary metal levels with serum hormones,
spermatozoa apoptosis and sperm DNA damage in a Chinese population. Environ Int 94:
177-188. http://dx.doi.Org/10.1016/i.envint.2016.05.022.

Wang. YX: Wang. P: Feng. W: Liu. C: Yang. P: Chen. YT: Sun. L: Sun. Y: Yue. 1: Gu. LI: Zeng. 0: Lu. W0.

(2017). Relationships between seminal plasma metals/metalloids and semen quality, sperm
apoptosis and DNA integrity. Environ Pollut 224: 224-234.
http://dx.doi.Org/10.1016/i.envpol.2017.01.083.

Weaver. VM: Vargas. GG: Silbergeld. EK: Rothenberg. ST: Fadrowski. 11: Rubio-Andrade. M: Parsons.
PI: Steuerwald. AT: Navas-Acien. A: Guallar. E. (2014). Impact of urine concentration
adjustment method on associations between urine metals and estimated glomerular
filtration rates (eGFR) in adolescents. Environ Res 132: 226-232.
http://dx.doi.Org/10.1016/i.envres.2014.04.013.

Wei. Y: Tin. L: Li. Z: Liu. I: Wang. L: Pi. X: Yin. S: Wang. C: Ren. A. (2019). Levels of uranium and

thorium in maternal scalp hair and risk of orofacial clefts in offspring. J Environ Radioact
204: 125-131. http://dx.doi.Org/10.1016/i.ienvrad.2019.04.007.

WHO (World Health Organization). (2001). Depleted uranium: Sources, exposure and health effects.
(WHO/SDE/PHE/Ol.l). Geneva, Switzerland: World Health Organization, Department of
Protection of the Human Environment. https: //www.who.int/publications /i/item/WHO-
SDE-PHE-01.1.

WHO (World Health Organization). (2012). Guidance for immunotoxicity risk assessment for
chemicals. (Harmonization Project Document No. 10). Geneva, Switzerland.
https://apps.who.int/iris/bitstream/handle/10665/330098/9789241503303-
eng.pdf?sequence=l&isAllowed=y.

Wu. H: Xu. B: Guan. Y: Wang. W: Huang. R: Zhang. T: Sun. R: Xie. K: Chen. M. (2020). A metabolomic
study on the association of exposure to heavy metals in the first trimester with primary
tooth eruption. Sci Total Environ 723: 138107.
http://dx.doi.Org/10.1016/i.scitotenv.2020.138107.

Wu. M: Shu. Y: Wang. Y. (2022). Exposure to mixture of heavy metals and muscle strength in

children and adolescents: a population-based study. Environ Sci Pollut Res Int 29: 60269-
60277. http://dx.doi.org/10.1007/sll356-022-19916-2.

Wu. W: Tiang. S: Zhao. 0: Zhang. K: Wei. X: Zhou. T: Liu. D: Zhou. H: Zeng. 0: Cheng. L: Miao. X: Lu. 0.
(2018a). Environmental exposure to metals and the risk of hypertension: A cross-sectional
study in China. Environ Pollut 233: 670-678.
http://dx.doi.Org/10.1016/i.envpol.2017.10.lll.

This document is a draft for review purposes only and does not constitute Agency policy.

R-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

30

31

32

33

34

Protocol for the Uranium IRIS Assessment (Oral)

Wu. W: Zhang. K: liang. S: Liu. D: Zhou. H: Zhong. R: Zeng. 0: Cheng. L: Miao. X: Tong. Y: Lu. 0.

(2018b). Association of co-exposure to heavy metals with renal function in a hypertensive
population. Environ Int 112: 198-206. http: / /dx.doi.org/10.1016/i.envint2017.12.023.

Xu. YL: Wei. Y: Long. TF: Wang. RX: Li. ZY: Yu. CZ: Wu. TC: He. MA. (2020). Association between
urinary metals levels and metabolic phenotypes in overweight and obese individuals.
Chemosphere 254: 126763. http://dx.doi.Org/10.1016/i.chemosphere.2020.126763.

Yang. F: Huang. Z: Yuan. H: He. M: Shen. M: Chen. X: Yi. X: Guo. 1: Xu. S: Xiao. Y: Huang. X: Duan. Y:
Luo. D: Xiao. S. (2019). Association of plasma and urine metals levels with kidney function:
A population-based cross-sectional study in China. Chemosphere 226: 321-328.
http://dx.doi.Org/10.1016/i.chemosphere.2019.03.171.

Yang. 1: Chan. K: Choi. C: Yang. A: Lo. K. (2022). Identifying Effects of Urinary Metals on Type 2
Diabetes in U.S. Adults: Cross-Sectional Analysis of National Health and Nutrition
Examination Survey 2011-2016. Nutrients 14. http://dx.doi.org/10.3390/nul4081552.

Yang. X: Li. Y: Li. 1: Bao. S: Zhou. A: Xu. S: Xia. W. (2020). Associations between exposure to metal
mixtures and birth weight Environ Pollut 263: 114537.
http://dx.doi.Org/10.1016/i.envpol.2020.114537.

Yelamanchili. SV: Fox. HS. (2010). Defining Larger Roles for "Tiny" RNA Molecules: Role of miRNAs
in Neurodegeneration Research. J Neuroimmune Pharmacol 5: 63-69.
http://dx.doi.Org/10.1007/sll481-009-9172-4.

Yin. S: Tian. T: Wang. C: Wang. D: Pi. X: Liu. M: Tin. L: Liu. 1: Wang. L: Li. Z: Ren. A: Yin. C. (2022).

Prenatal uranium exposure and risk for fetal neural tube defects: A case-control study in
women living in a rural area of northern China. J Hazard Mater 424: 127466.
http://dx.doi.Org/10.1016/i.ihazmat.2021.127466.

Yue. Z: Lin. 1: Silver. MA: Han. L: Li. X: Zhou. 1: Guo. X: Bao. H: Huang. YY: Wang. 10. (2018). Anionic
uranyl oxyfluorides as a bifunctional platform for highly selective ion-exchange and
photocatalytic degradation of organic dyes. Dalton Transactions (Online) 47: 14908-14916.
http://dx.doi.org/10.1039/c8dt02309c.

Zablotska. LB: Lane. RS: Frost. SE. (2013). Mortality (1950-1999) and cancer incidence (1969-1999)
of workers in the Port Hope cohort study exposed to a unique combination of radium,
uranium andy-ray doses. BMJ Open 3. http://dx.doi.org/10.1136/bmiopen-2012-002159.

Zhang. W: Liu. W: Bao. S: Liu. H: Zhang. Y: Zhang. B. in: Zhou. A: Chen. 1. ia: Hao. K. e: Xia. W. ei: Li. Y:
Sheng. X. ia: Xu. S. (2020). Association of adverse birth outcomes with prenatal uranium
exposure: A population-based cohort study. Environ Int 135: 105391.
http://dx.doi.Org/10.1016/i.envint.2019.105391.

This document is a draft for review purposes only and does not constitute Agency policy.

R-17	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

APPENDIX A. ELECTRONIC DATABASE SEARCH
STRATEGIES

Table A-l. Database search strategy

Database

Search string

Results3

Scopus

((TITLE-ABS-KEY-AUTH("Uranium tetrachloride*") OR TITLE-ABS-KEY-
AUTHf'Uranium chloride*") OR TITLE-ABS-KEY-AUTH("Sodium diuranate*") OR
TITLE-ABS-KEY-AUTH("Sodium uranate*") OR TITLE-ABS-KEY-AUTH("Sodium
uranium oxide*") ORTITLE-ABS-KEY-AUTH("Disodium heptaoxodiuranate*") OR
TITLE-ABS-KEY-AUTH("Ammonium uranyl tricarbonate*") ORTITLE-ABS-KEY-
AUTH("Ammonium uranium carbonate*") ORTITLE-ABS-KEY-
AUTH("Tetraammonium uranyl tricarbonate*") OR TULEf'uranium*") OR
TITLE("diuranium*") ORTITLE("triuranium*") OR TULEf'uranic") OR
TITLE("uranous") ORTITLE("uranyl") ORTITLE("uranate") OR TULEf'uranates") OR
TITLE("diuranate") ORTITLE("diuranates") ORTITLE("dioxouranium") OR
TITLE("uranyldifluoride") ORTITLE("uranyldifluorides") OR
TITLEf'diacetatodioxouranium") OR TITLE("difluorodioxouranium") OR
TITLE("dinitratodioxouranium") ORTITLEf'yellowcake") ORTITLE("234U") OR
TITLE("235U") OR TITLE("238U") OR TITLE("u-234") OR TITLE("u-235") ORTITLE("u-
238")) OR ((TITLE-ABS-KEY-AUTH("uranium*") OR TITLE-ABS-KEY-
AUTH("diuranium*") OR TITLE-ABS-KEY-AUTH("triuranium*") OR TITLE-ABS-KEY-
AUTH("uranic") OR TITLE-ABS-KEY-AUTH("uranous") OR TITLE-ABS-KEY-
AUTH("uranyl") OR TITLE-ABS-KEY-AUTH("uranate") OR TITLE-ABS-KEY-
AUTH("uranates") OR TULE-ABS-KEY-AUTH("diuranate") OR TITLE-ABS-KEY-
AUTH("diuranates") OR TITLE-ABS-KEY-AUTH("dioxouranium") OR TITLE-ABS-KEY-
AUTH("uranyldifluoride") OR TITLE-ABS-KEY-AUTH("uranyldifluorides") OR TITLE-
ABS-KEY-AUTH("diacetatodioxouranium") OR TITLE-ABS-KEY-
AUTH("difluorodioxouranium") ORTITLE-ABS-KEY-AUTH("dinitratodioxouranium")
OR TITLE-ABS-KEY-AUTH("yellowcake")) AND (((TITLE-ABS-KEY-AUTH("occupational
disease*") OR TITLE-ABS-KEY-AUTH("human") OR TITLE-ABS-KEY-AUTH("humans")
OR TITLE-ABS-KEY-AUTH("mammals") OR TITLE-ABS-KEY-AUTH("mammals")) AND
((TITLE-ABS-KEY-AUTH("Heavy Metals") AND TITLE-ABS-KEY-AUTH("adverse
effects")) OR (TITLE-ABS-KEY-AUTH("Heavy Metals") AND TITLE-ABS-KEY-
AUTH("blood")) OR (TITLE-ABS-KEY-AUTH("Heavy") AND TITLE-ABS-KEY-
AUTH("cerebrospinal fluid")) OR (TITLE-ABS-KEY-AUTH("Heavy Metals") AND TITLE-
ABS-KEY-AUTH("metabolism")) OR (TITLE-ABS-KEY-AUTH("Heavy Metals") AND
TITLE-ABS-KEY-AUTH("pharmacokinetics")) OR (TITLE-ABS-KEY-AUTH("Heavy
Metals") AND TITLE-ABS-KEY-AUTH("poisoning")) OR (TITLE-ABS-KEY-AUTH("Heavy
Metals") AND TITLE-ABS-KEY-AUTH("toxicity")) OR (TITLE-ABS-KEY-AUTH("Heavy
Metals") AND TITLE-ABS-KEY-AUTH("urine")) OR (TITLE-ABS-KEY-AUTH("Metals")
AND TITLE-ABS-KEY-AUTH("adverse effects")) OR (TULE-ABS-KEY-AUTH("Metals")
AND TITLE-ABS-KEY-AUTH("blood")) OR (TITLE-ABS-KEY-AUTH("Metals") AND TITLE-
ABS-KEY-AUTH("metabolism")) OR (TITLE-ABS-KEY-AUTH("Metals") AND TITLE-ABS-
KEY-AUTH("pharmacokinetics")) OR (TITLE-ABS-KEY-AUTH("Metals") AND TITLE-
ABS-KEY-AUTH("poisoning")) OR (TITLE-ABS-KEY-AUTH("Metals") AND TITLE-ABS-

8,119

This document is a draft for review purposes only and does not constitute Agency policy.

A-l	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



KEY-AUTH("toxicity")) OR (TITLE-ABS-KEY-AUTH("Metals") AND TITLE-ABS-KEY-
AUTHf'urine")))) OR (TULE-ABS-KEY-AUTH("chronic") OR TITLE-ABS-KEY-
AUTH("immun*") OR TITLE-ABS-KEY-AUTH ("lymph*") OR TITLE-ABS-KEY-
AUTHf'neurotox*") OR TITLE-ABS-KEY-AUTH("toxicokin*") OR TITLE-ABS-KEY-
AUTHf'pharmacokin*") OR TITLE-ABS-KEY-AUTH ("bio marker*") OR TITLE-ABS-KEY-
AUTH("neurolog*") OR TITLE-ABS-KEY-AUTH("subchronic") OR TITLE-ABS-KEY-
AUTH("epidemiolog*") OR TITLE-ABS-KEY-AUTH ("acute") OR TITLE-ABS-KEY-
AUTH("subacute") OR TITLE-ABS-KEY-AUTH("ld50") OR TITLE-ABS-KEY-AUTH("lc50")
OR TITLE-ABS-KEY-AUTH("inhal*") OR TITLE-ABS-KEY-AUTH("pulmon*") OR TITLE-
ABS-KEY-AUTH ("nasal") OR TITLE-ABS-KEY-AUTH ("lung*") OR TITLE-ABS-KEY-
AUTH("respir*") OR TITLE-ABS-KEY-AUTH("occupation*") OR TITLE-ABS-KEY-
AUTH("workplace") OR TITLE-ABS-KEY-AUTH ("worker*") OR TITLE-ABS-KEY-
AUTHf'oral") OR TITLE-ABS-KEY-AUTH("orally") OR TITLE-ABS-KEY-AUTH ("ingest*")
OR TITLE-ABS-KEY-AUTH("gavage") OR TITLE-ABS-KEY-AUTH ("diet") OR TITLE-ABS-
KEY-AUTH("diets") OR TITLE-ABS-KEY-AUTH("dietary") OR TITLE-ABS-KEY-
AUTH("drinking") OR TITLE-ABS-KEY-AUTH("gastr*") OR TITLE-ABS-KEY-
AUTH("intestin*") OR TITLE-ABS-KEY-AUTH ("gut") OR TITLE-ABS-KEY-
AUTH("sensitiz*") OR TITLE-ABS-KEY-AUTH("abort*") OR TITLE-ABS-KEY-
AUTH("abnormalit*") OR TITLE-ABS-KEY-AUTH ("embryo*") OR TITLE-ABS-KEY-
AUTH("cleft*") OR TITLE-ABS-KEY-AUTH ("fetus*") OR TITLE-ABS-KEY-
AUTH("foetus*") OR TITLE-ABS-KEY-AUTH ("fetal*") OR TITLE-ABS-KEY-
AUTH("foetal*") OR TITLE-ABS-KEY-AUTH("fertilit*") OR TITLE-ABS-KEY-
AUTH("infertil*") OR TITLE-ABS-KEY-AUTH("malform*") OR TITLE-ABS-KEY-
AUTH("ovum") OR TITLE-ABS-KEY-AUTH ("ova") OR TITLE-ABS-KEY-AUTH("ovary")
OR TITLE-ABS-KEY-AUTH("placenta*") OR TITLE-ABS-KEY-AUTH("pregnan*") OR
TITLE-ABS-KEY-AUTH("sperm") OR TITLE-ABS-KEY-AUTH("testic*") OR TITLE-ABS-
KEY-AUTH("testosterone") OR TITLE-ABS-KEY-AUTH("testis") OR TITLE-ABS-KEY-
AUTH("testes") OR TITLE-ABS-KEY-AUTH("epididym*") OR TITLE-ABS-KEY-
AUTH ("seminiferous") OR TITLE-ABS-KEY-AUTH ("cervix") OR TITLE-ABS-KEY-
AUTH("ovaries") OR TITLE-ABS-KEY-AUTH("ovarian") OR TITLE-ABS-KEY-
AUTH("corpora lutea") ORTITLE-ABS-KEY-AUTH("corpus luteum") ORTITLE-ABS-
KEY-AUTH("estrous") OR TITLE-ABS-KEY-AUTH("estrus") OR TITLE-ABS-KEY-
AUTH("dermal*") OR TITLE-ABS-KEY-AUTH ("derm is") OR TITLE-ABS-KEY-
AUTH("skin") OR TITLE-ABS-KEY-AUTH("epiderm*") OR TITLE-ABS-KEY-
AUTH("cutaneous") OR TITLE-ABS-KEY-AUTH("carcinog*") OR TITLE-ABS-KEY-
AUTH("cocarcinog*") OR TITLE-ABS-KEY-AUTH("cancer") OR TITLE-ABS-KEY-
AUTH("precancer") OR TITLE-ABS-KEY-AUTH("neoplas*") OR TITLE-ABS-KEY-
AUTH("tumor*") OR TITLE-ABS-KEY-AUTH ("tumour*") OR TITLE-ABS-KEY-
AUTH("oncogen*") OR TITLE-ABS-KEY-AUTH("lymphoma*") OR TITLE-ABS-KEY-
AUTH("carcinom*") OR TITLE-ABS-KEY-AUTH("genetox*") OR TITLE-ABS-KEY-
AUTH("genotox*") OR TITLE-ABS-KEY-AUTH("mutagen*") OR TITLE-ABS-KEY-
AUTH("nephrotox*") OR TITLE-ABS-KEY-AUTH("hepatotox*") OR TITLE-ABS-KEY-
AUTH("endocrin*") OR TITLE-ABS-KEY-AUTH ("estrogen*") OR TITLE-ABS-KEY-
AUTH("androgen*") OR TITLE-ABS-KEY-AUTH("hormon*") OR TITLE-ABS-KEY-
AUTH("blood") OR TITLE-ABS-KEY-AUTH("serum") OR TITLE-ABS-KEY-AUTH ("urine")
OR TITLE-ABS-KEY-AUTH("bone") OR TITLE-ABS-KEY-AUTH ("bones") OR TITLE-ABS-
KEY-AUTH("skelet*") OR TITLE-ABS-KEY-AUTH("rat") OR TITLE-ABS-KEY-
AUTH("rats") OR TITLE-ABS-KEY-AUTH ("mouse") OR TITLE-ABS-KEY-AUTH ("mice")
OR TITLE-ABS-KEY-AUTH ("guinea") OR TITLE-ABS-KEY-AUTH("muridae") OR TITLE-
ABS-KEY-AUTH ("rabbit*") OR TITLE-ABS-KEY-AUTH("lagomorph*") OR TITLE-ABS-



This document is a draft for review purposes only and does not constitute Agency policy.

A-2	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



KEY-AUTH("hamster*") OR TITLE-ABS-KEY-AUTH("ferret*") OR TITLE-ABS-KEY-
AUTHf'gerbil*") OR TITLE-ABS-KEY-AUTH ("rodent*") OR TITLE-ABS-KEY-
AUTHf'dog") OR TITLE-ABS-KEY-AUTH ("dogs") OR TITLE-ABS-KEY-AUTH ("beagle*")
OR TITLE-ABS-KEY-AUTH ("canine") OR TITLE-ABS-KEY-AUTH ("cats") OR TITLE-ABS-
KEY-AUTH("feline") OR TITLE-ABS-KEY-AUTH ("pig") OR TITLE-ABS-KEY-AUTH ("pigs")
OR TITLE-ABS-KEY-AUTH ("swine") OR TITLE-ABS-KEY-AUTH ("porcine") OR TITLE-
ABS-KEY-AUTH("monkey*") OR TITLE-ABS-KEY-AUTH ("macaque*") OR TITLE-ABS-
KEY-AUTH ("baboon*") OR TITLE-ABS-KEY-AUTH ("marmoset*") OR TITLE-ABS-KEY-
AUTH("toxic*") OR TITLE-ABS-KEY-AUTH ("adverse") OR TITLE-ABS-KEY-
AUTH("poisoning") OR TITLE-ABS-KEY-AUTH ("prenatal") OR TITLE-ABS-KEY-
AUTH("perinatal") OR TITLE-ABS-KEY-AUTH ("postnatal") OR TITLE-ABS-KEY-
AUTH("reproduc*") OR TITLE-ABS-KEY-AUTH("steril*") OR TITLE-ABS-KEY-
AUTH ("teratogen*") OR TITLE-ABS-KEY-AUTH("sperm*") OR TITLE-ABS-KEY-
AUTH("neonat*") OR TITLE-ABS-KEY-AUTH("newborn*") OR TITLE-ABS-KEY-
AUTH ("development*") OR TITLE-ABS-KEY-AUTH ("zygote*") OR TITLE-ABS-KEY-
AUTH("child") OR TITLE-ABS-KEY-AUTH ("children") OR TITLE-ABS-KEY-
AUTH("adolescen*") OR TITLE-ABS-KEY-AUTH ("infant*") OR TITLE-ABS-KEY-
AUTH("wean*") OR TITLE-ABS-KEY-AUTH ("offspring") OR TITLE-ABS-KEY-AUTH ("age
factor") OR TITLE-ABS-KEY-AUTH("age factors") ORTITLE-ABS-KEY-
AUTH("Genomics") OR TITLE-ABS-KEY-AUTH("Proteomics") OR TITLE-ABS-KEY-
AUTH("Metabolic Profile") OR TITLE-ABS-KEY-AUTH("Metabolome") OR TITLE-ABS-
KEY-AUTH ("Metabolomics") OR TITLE-ABS-KEY-AUTH("Microarray") OR TITLE-ABS-
KEY-AUTH ("Nanoarray") OR TITLE-ABS-KEY-AUTH("Gene expression") OR TITLE-ABS-
KEY-AUTH ("Transcript expression") OR TITLE-ABS-KEY-AUTH("transcriptomes") OR
TITLE-ABS-KEY-AUTH("transcriptome") OR TITLE-ABS-KEY-AUTH("Phenotype") OR
TITLE-ABS-KEY-AUTH ("Transcription") OR TITLE-ABS-KEY-AUTH ("Trans-act*") OR
TITLE-ABS-KEY-AUTH ("transact*") OR TITLE-ABS-KEY-AUTH("trans act*") OR TITLE-
ABS-KEY-AUTH ("genetic") OR TITLE-ABS-KEY-AUTH ("genetics") OR TITLE-ABS-KEY-
AUTH("genotype") OR TITLE-ABS-KEY-AUTH ("messenger RNA") ORTITLE-ABS-KEY-
AUTH("transfer RNA") OR TITLE-ABS-KEY-AUTH("peptide biosynthesis") OR TITLE-
ABS-KEY-AUTH ("protein biosynthesis") OR TITLE-ABS-KEY-AUTH ("protein
synthesis") OR TITLE-ABS-KEY-AUTH("RT-PCR") OR TITLE-ABS-KEY-AUTH("RTPCR")
ORTITLE-ABS-KEY-AUTH("Reverse Transcriptase Polymerase Chain Reaction") OR
TITLE-ABS-KEY-AUTH("DNA sequence") OR TITLE-ABS-KEY-AUTH ("renal") OR TITLE-
ABS-KEY-AUTH ("kidney*") OR TITLE-ABS-KEY-AUTH ("urinary") OR TITLE-ABS-KEY-
AUTH("liver") OR TITLE-ABS-KEY-AUTH("hepat*") OR TITLE-ABS-KEY-
AUTH("osseous") OR TITLE-ABS-KEY-AUTH("ossif*") OR TITLE-ABS-KEY-
AUTH("behavioral") OR TITLE-ABS-KEY-AUTH ("behavioural") OR TITLE-ABS-KEY-
AUTH("brain") OR TITLE-ABS-KEY-AUTH("nervous system") OR ((TITLE-ABS-KEY-
AUTH("Genetic transcription") OR TITLE-ABS-KEY-AUTH("Gene transcription") OR
TITLE-ABS-KEY-AUTH ("Gene Activation") OR TITLE-ABS-KEY-AUTH ("Genetic
induction") OR TITLE-ABS-KEY-AUTH ("Reverse transcription") OR TITLE-ABS-KEY-
AUTH ("Transcriptional activation") ORTITLE-ABS-KEY-AUTH("Transcription factors")
OR TITLE-ABS-KEY-AUTH("Biosynthesis"))AND (TITLE-ABS-KEY-AUTH("RNA") OR
TITLE-ABS-KEY-AUTH("DNA") OR TITLE-ABS-KEY-AUTH("mRNA"))) OR ((TITLE-ABS-
KEY-AUTH ("Informatics") OR TITLE-ABS-KEY-AUTH ("Information Science") OR TITLE-
ABS-KEY-AUTH ("Medical") OR TITLE-ABS-KEY-AUTH("Systems biology") OR TITLE-
ABS-KEY-AUTH ("Biological systems"))AND(TITLE-ABS-KEY-AUTH("monit*") OR
TITLE-ABS-KEY-AUTH ("data") OR TITLE-ABS-KEY-AUTH ("analysis"))))))) AND (LIMIT-
TO(SUBJAREA,"BIOC") OR LIMIT-TO(SUBJAREA, "ENVI") OR LIMIT-TO(SUBJAREA,



This document is a draft for review purposes only and does not constitute Agency policy.

A-3	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



"MEDI") OR UMIT-TO(SUBJAREA, "AGRI") OR UMIT-TO(SUBJAREA, "PHAR") OR
LIMIT-TO (SUBJAREA, "IMMU") OR UMIT-TO(SUBJAREA, "NEUR") OR UMIT-
TO(SUBJAREA, "VETE")) AND PUBYEAR AFT 2010



WoS

((TS="Uranium tetrachloride*" ORTS="Uranium chloride*" ORTS="Sodium
diuranate*" ORTS="Sodium uranate*" ORTS="Sodium uranium oxide*" OR
TS="Disodium heptaoxodiuranate*" ORTS="Ammonium uranyl tricarbonate*" OR
TS="Ammonium uranium carbonate*" ORTS="Tetraammonium uranyl
tricarbonate*" OR Tl="uranium*" ORTI="diuranium*" ORTI="triuranium*" OR
Tl="uranic" OR TI="uranous" ORTI="uranyl" ORTI="uranate" OR Tl="uranates" OR
Tl="diuranate" ORTI="diuranates" ORTI="dioxouranium" ORTI="uranyldifluoride*"
ORTI="uranyldifluorides" ORTI="diacetatodioxouranium" OR
TI="difluorodioxouranium" ORTI="dinitratodioxouranium" ORTI="yellowcake" OR
TI="234U" OR TI="235U" ORTI="238U" ORTI="u-234" ORTI="u-235" ORTI="u-
238") OR ((TS="uranium*" OR TS="diuranium*" OR TS="triuranium*" OR
TS="uranic" ORTS="uranous" ORTS="uranyl" OR TS="uranate" ORTS="uranates"
ORTS="diuranate" ORTS="diuranates" ORTS="dioxouranium" OR
TS="uranyldifluoride" OR TS="uranyldifluorides" OR TS="diacetatodioxouranium"
ORTS="difluorodioxouranium" ORTS="dinitratodioxouranium" OR
TS="yellowcake") AND (((TS="occupational disease*" ORTS="humans" OR
TS="human" OR TS="mammals" OR TS="mammal") AND ((TS="Heavy Metals" AND
TS="adverse effects") OR (TS="Heavy Metals" AND TS="blood") OR (TS="Heavy"
AND TS="cerebrospinal fluid") OR (TS="Heavy Metals" AND TS="metabolism") OR
(TS="Heavy Metals" AND TS="pharmacokinetics") OR (TS="Heavy Metals" AND
TS="poisoning") OR (TS="Heavy Metals" AND TS="toxicity") OR (TS="Heavy Metals"
AND TS="urine") OR (TS="Metals" AND TS="adverse effects") OR (TS="Metals" AND
TS="blood") OR (TS="Metals" AND TS="metabolism") OR (TS="Metals" AND
TS="pharmacokinetics") OR (TS="Metals" AND TS="poisoning") OR (TS="Metals"
AND TS="toxicity") OR (TS="Metals" AND TS="urine"))) OR (TS="chronic" OR
TS="immun*" ORTS="lymph*" ORTS="neurotox*" ORTS="toxicokin*" OR
TS="pharmacokin*" ORTS="biomarker*" OR TS="neurolog*" ORTS="subchronic"
ORTS="epidemiolog*" ORTS="acute" ORTS="subacute" ORTS="ld50" OR
TS="lc50" ORTS="inhal*" ORTS="pulmon*" ORTS="nasal" ORTS="lung*" OR
TS="respir*" ORTS="occupation*" ORTS="workplace" OR TS="worker*" OR
TS="oral" ORTS="orally" ORTS="ingest*" ORTS="gavage" ORTS="diet" OR
TS="diets" ORTS="dietary" ORTS="drinking" OR TS="gastr*" OR TS="intestin*" OR
TS="gut" ORTS="sensitiz*" ORTS="abort*" ORTS="abnormalit*" ORTS="embryo*"
ORTS="cleft*" OR TS="fetus*" ORTS="foetus*" ORTS="fetal*" ORTS="foetal*" OR
TS="fertilit*" ORTS="infertil*" ORTS="malform*" ORTS="ovum" ORTS="ova" OR
TS="ovary" ORTS="placenta*" ORTS="pregnan*" ORTS="sperm" ORTS="testic*"
OR TS="testosterone" ORTS="testis" ORTS="testes" ORTS="epididym*" OR
TS="seminiferous" ORTS="cervix" ORTS="ovaries" ORTS="ovarian" OR
TS="corpora lutea" ORTS="corpus luteum" ORTS="estrous" ORTS="estrus" OR
TS="dermal*" ORTS="dermis" ORTS="skin" OR TS="epiderm*" ORTS="cutaneous"
ORTS="carcinog*" ORTS="cocarcinog*" ORTS="cancer" ORTS="precancer" OR
TS="neoplas*" ORTS="tumor*" ORTS="tumour*" ORTS="oncogen*" OR
TS="lymphoma*" ORTS="carcinom*" ORTS="genetox*" OR TS="genotox*" OR
TS="mutagen*" ORTS="nephrotox*" ORTS="hepatotox*" ORTS="endocrin*" OR
TS="estrogen*" ORTS="androgen*" ORTS="hormon*" ORTS="blood" OR
TS="serum" ORTS="urine" ORTS="bone" ORTS="bones" ORTS="skelet*" OR
TS="rat" OR TS="rats" OR TS="mouse" OR TS="mice" OR TS="guinea" OR

18,396

This document is a draft for review purposes only and does not constitute Agency policy.

A-4	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



TS="muridae" ORTS="rabbit*" ORTS="lagomorph*" ORTS="hamster*" OR
TS="ferret*" ORTS="gerbil*" ORTS="rodent*" ORTS="dog" ORTS="dogs" OR
TS="beagle*" ORTS="canine" ORTS="cats" ORTS="feline" ORTS="pig" OR
TS="pigs" OR TS="swine" OR TS="porcine" ORTS="monkey*" ORTS="macaque*"
ORTS="baboon*" ORTS="marmoset*" OR TS="toxic*" OR TS="adverse" OR
TS="poisoning" ORTS="prenatal" ORTS="perinatal" ORTS="postnatal" OR
TS="reproduc*" ORTS="steril*" OR TS="teratogen*" OR TS="sperm*" OR
TS="neonat*" ORTS="newborn*" ORTS="development*" ORTS="zygote*" OR
TS="child" ORTS="children" ORTS="adolescen*" ORTS="infant*" ORTS="wean*"
ORTS="offspring" ORTS="age factor" ORTS="age factors" ORTS="Genomics" OR
TS="Proteomics" OR TS="Metabolic Profile" ORTS="Metabolome" OR
TS="Metabolomics" ORTS="Microarray" ORTS="Nanoarray" ORTS="Gene
expression" ORTS="Transcript expression" ORTS="transcriptomes" OR
TS="transcriptome" ORTS="Phenotype" ORTS="Transcription" ORTS="Trans-act*"
OR TS="transact*" OR TS="trans act*" OR TS="genetic" OR TS="genetics" OR
TS="genotype" ORTS="messenger RNA" OR TS="transfer RNA" ORTS="peptide
biosynthesis" ORTS="protein biosynthesis" ORTS="protein synthesis" ORTS="RT-
PCR" ORTS="RTPCR" OR TS="Reverse Transcriptase Polymerase Chain Reaction" OR
TS="DNA sequence" OR TS="renal" OR TS="kidney*" OR TS="urinary" OR TS="liver"
ORTS="hepat*" ORTS="osseous" ORTS="ossif*" ORTS="behavioral" OR
TS="behavioural" ORTS="brain" ORTS="nervous system" OR ((TS="Genetic
transcription" OR TS="Gene transcription" OR TS="Gene Activation" OR
TS="Genetic induction" OR TS="Reverse transcription" OR TS="Transcriptional
activation" ORTS="Transcription factors" ORTS="Biosynthesis")AND (TS="RNA" OR
TS="DNA" ORTS="mRNA")) OR ((TS="lnformatics" ORTS="lnformation Science" OR
TS="Medical" ORTS="Systems biology" ORTS="Biological
systems")AND(TS="monit*" ORTS="data" OR TS="analysis"))))))AND PY=(2011-
2021)



PubMed

("uranium"[MeSH Terms] OR "Uranyl Nitrate"[mh] OR "uranium
compounds"[MeSH Terms] OR 7440-61-l[rn] OR 1344-57-6[rn] OR 1344-58-
7[EC/RN Number] OR 12036-71-4[EC/RN Number] OR 1344-59-8[EC/RN Number]
OR 10049-14-6[EC/RN Number] OR 7783-81-5[EC/RN Number] OR 13536-84-
0[EC/RN Number] OR 541-09-3[rn] OR 6159-44-0[rn] OR 10102-06-4[rn] OR 7783-
22-4[EC/RN Number] OR 18378-88-6[rn] OR 12179-35-0[rn] OR 23243-55-2[rn])
AND ("Uranium/adverse effects"[Mesh] OR "Uranium/antagonists and
inhibitors"[Mesh] OR "Uranium/blood"[Mesh] OR "Uranium/immunology"[Mesh]
OR "Uranium/metabolism"[Mesh] OR "Uranium/pharmacokinetics"[Mesh] OR
"Uranium/poisoning"[Mesh] OR "Uranium/radiation effects"[Mesh] OR
"Uranium/toxicity"[Mesh] OR "Uranium/urine"[Mesh] OR "Oxides/adverse
effects"[Mesh] OR "Oxides/antagonists and inhibitors"[Mesh] OR
"Oxides/blood"[Mesh] OR "Oxides/cerebrospinal fluid"[Mesh] OR
"Oxides/metabolism"[Mesh] OR "Oxides/pharmacokinetics"[Mesh] OR
"Oxides/poisoning"[Mesh] OR "Oxides/radiation effects"[Mesh] OR
"Oxides/toxicity"[Mesh] OR "Oxides/urine"[Mesh] OR "chemically
induced"[Subheading] OR "environmental exposure"[mh] OR cancer[sb] OR
"endocrine system"[mh] OR "endocrine disruptors"[mh] OR "hormones, hormone
substitutes, and hormone antagonists"[mh] OR endocrine[tw] OR "dose-response
relationship, drug"[mh] OR "risk"[MeSH Terms] OR "toxicity tests"[mh] OR
(("pharmacokinetics"[MeSH Terms] OR "metabolism"[MeSH Terms] OR "metabolic
networks and pathways"[MeSH Terms]) AND "humans"[MeSH Terms] OR

1,666

This document is a draft for review purposes only and does not constitute Agency policy.

AS	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



"animals"[MeSH Terms]) OR "Computational biology"[mh] OR "Medical
lnformatics"[mh] OR "genomics"[MeSH Terms] OR "genome"[MeSH Terms] OR
"proteomics"[MeSH Terms] OR "proteome"[MeSH Terms] OR
"metabolomics"[MeSH Terms] OR "metabolome"[MeSH Terms] OR "genes"[MeSH
Terms] OR "Gene expression"[mh] OR "phenotype"[MeSH Terms] OR
"genetics"[MeSH Terms] OR "genotype"[MeSH Terms] OR "transcriptome"[MeSH
Terms] OR ("Systems Biology"[mh] AND ("Environmental Exposure"[mh] OR
"Epidemiological Monitoring"[mh] OR "analysis"[Subheading])) OR "Transcription,
Genetic "[mh] OR "Reverse transcription"[mh] OR "Transcriptional activation"[mh]
OR "Transcription factors"[mh] OR ("biosynthesis"[sh] AND ("rna"[MeSH Terms] OR
"dna"[MeSH Terms])) OR "RNA, Messenger "[mh] OR "RNA, Transfer"[mh] OR
"peptide biosynthesis"[mh] OR "protein biosynthesis"[mh] OR "Reverse
Transcriptase Polymerase Chain Reaction"[mh] OR "Base Sequence"[mh] OR
"Trans-activators"[mh] OR "Gene Expression Profiling"[mh] OR "Organometallic
Compounds/adverse effects"[Mesh] OR "Organometallic Compounds/antagonists
and inhibitors"[Mesh] OR "Organometallic Compounds/blood"[Mesh] OR
"Organometallic Compounds/cerebrospinal fluid"[Mesh] OR "Organometallic
Compounds/metabolism"[Mesh] OR "Organometallic
Compounds/pharmacokinetics"[Mesh] OR "Organometallic
Compounds/poisoning"[Mesh] OR "Organometallic Compounds/radiation
effects"[Mesh] OR "Organometallic Compounds/toxicity"[Mesh] OR
"Organometallic Compounds/urine"[Mesh]) AND ("2011/01/01"[PDAT] :
"2021/09/01"[PDAT])



Toxnet

@OR+(@term+@rn+7440-61-l+@term+@rn+1344-57-6+@term+@rn+1344-58-

7+@term+@rn+19525-15-6+@term+@rn+12036-71-4+@term+@rn+171236-10-

5+@term+@rn+1344-59-8+@term+@rn+10049-14-6+@term+@rn+7783-81-

5+@term+@rn+10026-10-5+@term+@rn+13536-84-0)+@AND+@org+tscats

@OR+(@term+@rn+7440-61-l+@term+@rn+1344-57-6+@term+@rn+1344-58-

7+@term+@rn+19525-15-6+@term+@rn+12036-71-4+@term+@rn+171236-10-

5+@term+@rn+1344-59-8+@term+@rn+10049-14-6+@term+@rn+7783-81-

5+@term+@rn+10026-10-5+@term+@rn+13536-84-

0)+@NOT+@org+pubmed+pubdart+crisp+tscats

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("Gene+expression"+"Transc

ript+expression"+"transcriptomes"+"transcriptome"+"Phenotype"+"Transcription"+

"transact*"+genetic+"genetics"+"genotype")+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("Gene+expression"+"Transc

ript+expression"+"transcriptomes"+"transcriptome"+"Phenotype"+"Transcription"+

"transact*"+genetic+"genetics"+"genotype")+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("Genomics"+"Proteomics"+"

Metabolic+Profile"+"Metabolome"+"Metabolomics"+"Microarray"+"Nanoarray")+

@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("Genomics"+"Proteomics"+"



This document is a draft for review purposes only and does not constitute Agency policy.

A-6	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



Metabolic+Profile"+"Metabolome"+"Metabolomics"+"Microarray"+"Nanoarray")+
@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("messenger+RNA"+"transfer

+RNA"+"peptide+biosynthesis"+"protein+biosynthesis"+"protein+synthesis"+"RT+P

CR"+"RTPCR"+"Reverse+Transcriptase+Polymerase+Chain+Reaction"+"DNA+seque

nce")+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("messenger+RNA"+"transfer

+RNA"+"peptide+biosynthesis"+"protein+biosynthesis"+"protein+synthesis"+"RT+P

CR"+"RTPCR"+"Reverse+Transcriptase+Polymerase+Chain+Reaction"+"DNA+seque

nce")+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("Transcriptional+activation"

+"Transcription+factors"+RNA+DNA+"mRNA")+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+("Transcriptional+activation"

+"Transcription+factors"+RNA+DNA+"mRNA")+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(chronic+lymph*+neurotox*

+toxicokin*+pharmacokin*+biomarker*+neurolog*+subchronic+pbpk+epidemiolog

*+acute+subacute+ld50)+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(chronic+lymph*+neurotox*

+toxicokin*+pharmacokin*+biomarker*+neurolog*+subchronic+pbpk+epidemiolog

*+acute+subacute+ld50)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(dermal*+dermis+skin+epide

rm*+cutaneous+carcinog*+cocarcinog*+cancer+precancer+neoplas*+tumor*+tum

our*)+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(dermal*+dermis+skin+epide

rm*+cutaneous+carcinog*+cocarcinog*+cancer+precancer+neoplas*+tumor*+tum

our*)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(gut+sensitiz*+abort*+abnor

malit*+embryo*+cleft*+fetus*+foetus*+fetal*+foetal*+fertilit*+infertil*+malform*

+ovum+ova+ovary+placenta*+pregnan*)+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(gut+sensitiz*+abort*+abnor



This document is a draft for review purposes only and does not constitute Agency policy.

A-7	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



malit*+embryo*+cleft*+fetus*+foetus*+fetal*+foetal*+fertilit*+infertil*+malform*
+ovum+ova+ovary+placenta*+pregnan*)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(hamster*+ferret*+gerbil*+r

odent*+dog+dogs+beagle*+canine+cats+feline+pig+pigs+swine+porcine+monkey*

+macaque*+baboon*+marmoset*+toxic*+adverse+poisoning)+@range+yr+2013+2

017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(hamster*+ferret*+gerbil*+r

odent*+dog+dogs+beagle*+canine+cats+feline+pig+pigs+swine+porcine+monkey*

+macaque*+baboon*+marmoset*+toxic*+adverse+poisoning)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(hormon*+blood+serum+uri

ne+bone+bones+skelet*+rat+rats+mouse+mice+guinea+muridae+rabbit*+lagomor

ph* )+@ ra nge+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(hormon*+blood+serum+uri

ne+bone+bones+skelet*+rat+rats+mouse+mice+guinea+muridae+rabbit*+lagomor

ph*)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana
te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu
m+dinitratodioxouranium+yellowcake)+@AND+@OR+(immune+autoimmun*+imm
unosuppress*+immunolog*+immunotox*)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(immune+autoimmun*+imm

unosuppress*+immunolog*+immunotox*)+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(informatics+"systems+biolo

gy"+"biological+systems"+"information+science")+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(lc50+inhal*+pulmon*+nasal

+lung*+respir*+occupation*+workplace+worker*+oral+orally+ingest*+gavage+diet

+diets+dietary+drinking+gastr*+intestin*)+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(lc50+inhal*+pulmon*+nasal

+lung*+respir*+occupation*+workplace+worker*+oral+orally+ingest*+gavage+diet

+diets+dietary+drinking+gastr*+intestin*)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(microarray+"Genetic+transc

ription"+"Gene+transcription"+"Gene+Activation"+"Genetic+induction"+"Reverse+t

ranscription")+@range+yr+2013+2017



This document is a draft for review purposes only and does not constitute Agency policy.

A-8	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Database

Search string

Results3



@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(microarray+"Genetic+transc

ription"+"Gene+transcription"+"Gene+Activation"+"Genetic+induction"+"Reverse+t

ranscription")+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(oncogen*+lymphoma*+carc

inom*+genetox*+genotox*+mutagen*+nephrotox*+hepatotox*+endocrin*+estrog

en*+androgen*)+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(oncogen*+lymphoma*+carc

inom*+genetox*+genotox*+mutagen*+nephrotox*+hepatotox*+endocrin*+estrog

en*+androgen*)+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(prenatal+perinatal+postnata

l+reproduct*+steril*+teratogen*+sperm*+neonat*+newborn*+development*+zyg

ote*+child+children+adolescen*+infant*+wean*+offspring+"age factor"+"age

factors")+@range+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(prenatal+perinatal+postnata

l+reproduct*+steril*+teratogen*+sperm*+neonat*+newborn*+development*+zyg

ote*+child+children+adolescen*+infant*+wean*+offspring+"age factor"+"age

factors")+@AND+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(renal+kidney*+urinary+liver

+hepat*+osseous+ossif*+behavioral+behavioural+brain+"nervous+system")+@ran

ge+yr+2013+2017

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana

te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu

m+dinitratodioxouranium+yellowcake)+@AND+@OR+(renal+kidney*+urinary+liver

+hepat*+osseous+ossif*+behavioral+behavioural+brain+"nervous+system")+@AN

D+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl*+uranate*+diurana
te*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxouraniu
m+dinitratodioxouranium+yellowcake)+@AND+@OR+(sperm+testic*+testosterone
+testis+testes+epididym*+seminiferous+cervix+ovaries+ovarian+corpora
lutea+corpus luteum+estrous+estrus)+@ AN D+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl+uranate*+
diuranate*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxo
uranium+dinitratodioxouranium+yellowcake)+@NOT+@org+pubmed+pubdart+cris
p+tscats+ntis

@OR+(uranium+diuranium+triuranium+uranic+uranous+uranyl+uranate*+

diuranate*+dioxouranium+uranyldifluoride*+diacetatodioxouranium+difluorodioxo

uranium+dinitratodioxouranium+yellowcake)+@range+yr+2013+2017



a Searchesdates covered in this document are current as of November 2022.

This document is a draft for review purposes only and does not constitute Agency policy.

A-9	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

APPENDIX B. SURVEY OF EXISTING TOXICITY
VALUES

1	Table B-l lists websites that are searched for relevant human health reference values. In

2	addition to these sources, the ToxVal database on the Chemicals Dashboard

3	fhttps: //comptox.epa.gov/dashboard/chemical lists/TOXVAL V51 is searched for both reference

4	values and PODs as described in Appendix D. ToxVal is searched in the EPA CompTox Chemicals

5	Dashboard ("U.S. EPA. 2018al.

Table B-l. Sources searched for existing human health reference values

Source3

Query and/or link

ATSDR

http://www.atsdr.cdc.gov/toxprofiles/index.asp

CalEPA

http://www.oehha.ca.gov/tcdb/index.asp

DWSHA

https://www.epa.gov/sites/production/files/2018-03/documents/dwtable2018.pdf

Health
Canada

https://www.canada.ca/en/services/health/publications/healthv-living.html

https://publications.gc.ca/site/archivee-

archived.html?url=http://publications.gc.ca/collections/collection 2012/sc-hc/H 128-1-11-638-
eng.pdf



https://publications.gc.ca/site/archivee-

archived.html?url=https://publications.gc.ca/collections/Collection/H46-2-96-194E.pdf

HEAST

https://epa-heast.ornl.gov/heast.php



https://nepis.epa.gov/Exe/ZvPDF.cgi/200000GZ. PDF?Dockev=200000GZ. PDF

IRIS

https://www.epa.gov/iris

MlEGLE

https://www.michigan.gov/documents/dea/dea-rrd-chem-CleanupCriteriaTSD 527410 7.pdf

MDH

https://www.health.state.mn.us/communities/environment/risk/guidance/gw/table.html

NHMRC

https://www.nhmrc.gov.au/about-us/publications/australian-drinking-water-guidelines

NY DEC

https://www.dec.nv.gov/docs/remediation hudson pdf/techsuppdoc.pdf

OPP

https://iaspub.epa.gov/apex/pesticides/f?p=chemicalsearch:l

PPRTV

https://www.epa.gov/pprtv/provisional-peer-reviewed-toxicitv-values-pprtvs-assessments

RIVM

https://www.rivm.nl/bibliotheek/rapporten/711701092.pdf



https://www.rivm.nl/bibliotheek/rapporten/711701025.pdf

This document is a draft for review purposes only and does not constitute Agency policy.

B-l	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Source3

Query and/or link

TCEQ

https://www.tceq.texas.gov/remediation/trrp/trrppcls.html



WHO

http://www.who.int/ipcs/publications/ehc/en/

aATSDR = Agency for Toxic Substances and Disease Registry; CalEPA = California Environmental Protection Agency;

DWSHA = Drinking Water Standards and Health Advisories; HEAST = Health Effects Assessment Summary Tables;

IRIS = Integrated Risk Information System; MDH = Minnesota Department of Health; Ml EGLE = Michigan Department of
Environment, Great Lakes & Energy; NHMRC = National Health and Medical Research Council; NY DEC = New York State
Department of Environmental Conservation; OPP = Office of Pesticide Programs; PPRTV = Provisional Peer-Reviewed Toxicity
Values; RIVM = Rijksinstituut voor Volksgezondheid en Milieu, the Netherlands Institute for Public Health and the
Environment; TCEQ = Texas Commission on Environmental Quality; WHO = World Health Organization.

This document is a draft for review purposes only and does not constitute Agency policy.

B-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

Protocol for the Uranium IRIS Assessment (Oral)

APPENDIX C. PROCESS FOR SEARCHING AND
COLLECTING EVIDENCE FROM SELECTED OTHER
RESOURCES

C.l. REVIEW OF REFERENCE LISTS FROM EXISTING ASSESSMENTS
(FINAL OR PUBLICLY AVAILABLE DRAFT), JOURNAL REVIEWS
ARTICLES, AND STUDIES CONSIDERED RELEVANT TO PECO BASED
ON FULL-TEXT SCREENING

Review of the citation reference lists is typically done manually because they are not
available in a file format (e.g., RIS) that permits uploading into screening software applications.
Manual review entails scanning the title, study summary, or study details as presented in the
resource for those that appear to meet the populations, exposures, comparators, and outcomes
(PECO) criteria. Any records identified that are not identified from the other sources are annotated
with respect to source and screened as outlined in Section 4.

C.2. EUROPEAN CHEMICALS AGENCY

A search of the European Chemicals Agency registered substances database was conducted
using the chemical names. The registration dossier associated with the chemical name was
retrieved by navigating to and clicking the eye-shaped view icon displayed in the chemical
summary panel. The general information page and all subpages included under the Toxicological
Information tab were reviewed to identify any human or animal health effects information from
2016 onward that would be eligible for inclusion based on PECO criteria.

C.3. EPA CHEMVIEW

The EPA ChemView database (U.S. EPA. 20191 using the chemical CASRN is searched. The
prepopulated CASRN match and the "Information Submitted to EPA" output option filter are
selected before generating results. If results are available, the square-shaped icon under the "Data
Submitted to EPA" column is selected, and the following records are included:

•	High Production Volume Challenge Database (HPVIS)

•	Human Health studies (Substantial Risk Reports)

•	Monitoring (includes environmental, occupational, and general entries)

•	TSCA Section 4 (chemical testing results)

This document is a draft for review purposes only and does not constitute Agency policy.

C-l	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	• TSCA Section 8(d) (health and safety studies)

2	• TSCA Section 8(e) (substantial risk)

3	• FYI (voluntary documents)

4	All records for ecotoxicology and physical and chemical property entries were excluded.

5	When results were available, extractors navigated into each record until a substantial risk report

6	link was identified and saved as a PDF file. If the report could not be saved, due to file corruption or

7	broken links, the record was excluded during full-text review as "unable to obtain record." Most

8	substantial risk reports contained multiple document IDs, so citations were derived by

9	concatenating the unique report numbers (OTS; 8EHD Num; DCN; TSCATS RefID; and CIS)

10	associated with each document along with the typical author organization, year, and title. Once a

11	citation was generated, the study moved forward to DistillerSR with which it was screened

12	according to PECO and supplemental material criteria.

C.4. NTP CHEMICAL EFFECTS IN BIOLOGICAL SYSTEMS

13	This database is searched using the chemical CASRN

14	fhttps://manticore.niehs.nih.gov/cebssearch). All non-NTP data were excluded using the "NTP

15	Data Only" filter. Data tables for reports undergoing peer review are also searched for studies that

16	have not been finalized fhttps: //ntp.niehs.nih.gov/data/tables/index.htmll based on a manual

17	review of chemical names.

C.5. OECD ECHEMPORTAL

18	The OECD eChemPortal fhttps://h pvchemicals.oecd.org/UI/Search.aspx] is searched using

19	the chemical CASRN. Only database entries from the following sources are included and entries

20	from all other databases are excluded in the search. Final assessment reports and other relevant

21	SIDS reports embedded in the links are captured and saved as PDF files.

22	• OECD HPV

23	• OECD SIDS IUCLID

24	• SIDS United Nations Environment Programme (UNEP)

C.6. ECOTOX DATABASE

25	EPA's ECOTOX Knowledgebase fhttps: //cfpub.epa.gov/ecotox/search.cfm] was searched

26	using the chemical names. Results were refined to terrestrial mammalian studies by selecting the

27	terrestrial tab at the top of the search page and sorting the results by species group. Results were

28	reviewed to verify that it was not already identified from the database search (or searches of "other

29	sources consulted") search prior to moving forward to screening.

This document is a draft for review purposes only and does not constitute Agency policy.

C-2	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table C-l. Summary table for other sources search results

Source

Source address

Search terms

Search date

Total unique
number of
results retrieved

Records not
otherwise identified
that were screened
in DistillerSR

Review of reference lists
studies considered
relevant to PECO-based
on full-text screening

NA

NA

NA

67

65

Review of reference lists
from existing
assessments (final or
publicly available draft)
or journal review articles
that focused on human
health

NA

NA

NA

3

0

EPA CompTox
(Computational
Toxicology Program)
Chemicals Dashboard
(ToxVal)

https://comptox.epa.gov/dashboard

/dsstoxdb/results?abbreviation=TOX
VAL V5&search=DTXSID6021793#to
xicitv-values

90-15-3 (results from human health:
POD, toxicity value, lethality effect level)

12/10/2019

21

5

ECHA

https://echa.europa.eu/information-

on-chemicals/information-from-

existing-substances-regulation

90-15-3

10/8/2019

53

24

EPA ChemView

https://chemview.epa.gov/chemvie

w?tf=0&ch=90-15-3&su=2-5-6-7-

37574985&as=3-10-9-8&ac=l-15-16-

6378999&ma=4-ll-

1981377&tds=0&tdl=10&tasl=l&tas

2=asc&tas3=undefined&tss=

90-15-3

9/19/2019

3

1

This document is a draft for review purposes only and does not constitute Agency policy.

C-3	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Source

Source address

Search terms

Search date

Total unique
number of
results retrieved

Records not
otherwise identified
that were screened
in DistillerSR

High Production Volume
Information System
(HPVIS)

httos://ofmoub. eoa.gov/oDDthov/au
icksearch.disDlav?DChem=101850

90-15-3

9/19/2019

4

4

NTPCEBS

httDs://manticore. niehs.nih.gov/cebs
search/search?a =90-15-3

90-15-3

9/19/2019

0

0

OECD eChemPortal

httDs://hDvchemicals.oecd.org/UI/Se
arch.asox

90-15-3

9/19/2019

0

0

ECOTOX database

httDs://cfoub. eoa.gov/ecotox/search
.cfm

90-15-3

9/19/2019

4

3

EPA CompTox Chemicals
Dashboard version to
retrieve a summary of
any ToxCast or Tox21
high-throughput
screening information

httos://comotox. eoa.gov/dashboard
/dsstoxdb/results?search=DTXSID60
21793

90-15-3

9/19/2019

1

1

Comparative
Toxicogenomics
Database (CTDB)

htto://ctdbase.org/

90-15-3

12/9/2019

57

30

ArrayExpress

httos://www. ebi.ac.uk/arravexoress/

90-15-3 and "naphthol"

12/9/2019

1

1

This document is a draft for review purposes only and does not constitute Agency policy.

C-4	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Source

Source address

Search terms

Search date

Total unique
number of
results retrieved

Records not
otherwise identified
that were screened
in DistillerSR

Gene Expression
Omnibus

https://www. ncbi.nlm.nih.gov/geo/

(90-15-3[rn] OR "l-Naphthol"[tw] OR
"Naphthalen-l-ol"[tw] OR "1-
Naphthalenol"[tw] OR "1-
naphthalenol"[tw]) AND ("Expression
profiling by RT-PCR"[Filter] OR
"Expression profiling by MPSS"[Filter]
OR "Expression profiling by
SAGE"[Filter] OR "Expression profiling
by SNP array"[Filter] OR "Expression
profiling by array"[Filter] OR "Expression
profiling by genome tiling array"[Filter]
OR "Expression profiling by high
throughput sequencing"[Filter] OR
"Protein profiling by Mass Spec"[Filter]
OR "Protein profiling by protein
array"[Filter]).

12/9/2019

2

1

CEBS = Chemical Effects in Biological Systems; ECHA = European Chemicals Agency; NA = not applicable; NTP = National Toxicology Program; OECD = Organisation for Economic
Co-operation and Development; PECO = populations, exposures, comparators, and outcomes; POD = point of departure.

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

Protocol for the Uranium IRIS Assessment (Oral)

APPENDIX D. COMPARISON BETWEEN ATSDR 2013
AND IRIS LITERATURE SEARCH INVENTORY

In this appendix, the following is presented for each health effect category:

•	Summary of findings from studies used in ATSDR 2013;

•	Description of newly identified studies, human and animal, from the IRIS literature search, in
both narrative and tabular format;

•	Conclusions of whether the newly available studies identified in the literature search update
provide further support of the evidence considered by ATSDR 2013 and their interpretation;

•	Units of analysis, if applicable.

D.l. BODY WEIGHT EFFECTS

ATSDR Summary

ATSDR 2013 stated that no body weight effects were reported in the available human
studies. ATSDR 2013 also provide a summary of the animal evidence, but state that body weight
"effects are not necessarily the result of systemic toxicity." This is because the observed decreases
in body weight are accompanied by a reduction in food consumption, which in turn could be caused
by the palatability of uranium in the food. ATSDR 2013 also states the same aversive taste issue
may influence water consumption. They cited studies using rats, mice, and dogs exposed to high
doses of uranium for subchronic and chronic durations, which reported no significant changes in
body weight

Newly Identified Human Studies

No new human studies were identified in the IRIS literature search.

Newly Identified Animal Studies

Three studies using mice and seven studies using SD rats were identified in the IRIS
literature search. In adult C57BL/6J mice and ApoE null mice, subchronic exposures to uranium did
not have a significant impact on body weights fMedina etal.. 2020: Bolt etal.. 2019: Souidi etal..
20121. In adult SD rats most of the available studies reported no significant effects on body weight
or food and water consumption fGrison etal.. 2016: Dublineau etal.. 2014: Gueguen et al.. 2014:
Poisson et al.. 2014b: Hao etal.. 2013a: Rouas etal.. 20111. One study reported decreased body
weight after exposure to uranyl nitrate for 11 or 22 weeks, but the study authors also noted that
water consumption was also decreased in exposed animals (Vicente-Vicente etal.. 20131. These
findings are consistent with ATSDR's interpretation.

This document is a draft for review purposes only and does not constitute Agency policy.

D-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

Protocol for the Uranium IRIS Assessment (Oral)

Conclusion

The available toxicological studies identified in the literature search update provide further
support of the evidence considered by ATSDR 2013 and their interpretation. EPA will not consider
body weight effects in sexually mature animals for hazard evaluation or dose-response as the
majority of the available studies report no effects on body weight or food and water consumption
and the study that observed uranium-induced changes in body weight also reported decreased
water consumption, which may be a potential confounder.

Units of Analysis

N/A

D.2. CARDIOVASCULAR EFFECTS

ATSDR Summary

ATSDR 2013 concluded that "cardiovascular effects following intake of uranium are
unlikely." ATSDR cited animal toxicity studies using rats or New Zealand rabbits and two
epidemiological studies (one case study and one cohort study). The animal toxicity studies cited in
ATSDR 2013 measured organ weights and histopathology, and none reported significant uranium-
induced effects. ATSDR examined a case report, which documented a patient who suffered from
myocarditis after ingestion of a large dose uranyl acetate (approximately 15 g), and an
observational study, which reported a small positive association between urinary uranium
concentrations and blood pressure.

Newly Identified Human Studies

Twenty-two (n = 22) epidemiological studies meeting PECO criteria were identified in the
IRIS literature search for cardiovascular outcomes (see Table D-l). Blood pressure was commonly
examined. Some studies reported significant associations: dilated cardiomyopathy (Malamba-Lez et
al.. 2021). and high blood pressure in NHANES (Shiue and Hristova. 2014). using urinary
biomarkers to assess exposure. For a few studies there were potential limitations, including with
exposure assessment, such as judging exposure by job classification with no biomarker or other
exposure measurement (Al Rashida etal.. 2019: Shumate etal.. 2017: Guseva Canu etal.. 2014).
Additionally, some studies only reported exposure averages by outcome group.

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

Protocol for the Uranium IRIS Assessment (Oral)

Newly Identified Animal Studies

Four animal toxicity studies that meet PECO criteria were identified in the IRIS literature
search (see Table D-2). These studies used SD rats, and wild type and ApoE null mice exposed to
uranyl nitrate in drinking water for 11 weeks to 9 months. No effects were observed for markers of
cardiovascular disease including total cholesterol, LDL and HDL, and triglycerides. Exposure to
uranium in drinking water for 11 and 21 weeks increased systolic blood pressure in SD rats
(Vicente-Vicente etal.. 20131. However, these effects may be confounded by apparent palatability
issues causing large decrease in water intake (54% decrease) at the only dose tested fVicente-
Vicente etal.. 20131.

Conclusion

Potentially impactful epidemiological studies report on a potential association with
uranium exposure and high blood pressure and cardiomyopathy. Based on these findings, plus
animal study findings, EPA will perform a hazard evaluation of uranium-induced cardiovascular
effects. This analysis will consider studies cited in ATSDR and studies that met problem formulation
PECO criteria in the IRIS literature search.

Units of Analysis

Humans: blood pressure, cardiovascular disease.

Animals: Heart and vessel morphology and histopathology, blood and arteriole pressure,
peripheral resistance, and other measures of cardiovascular function.

Table D-l. Studies of cardiovascular endpoints in humans identified 2011-
2021

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported findings

Choi etal. (2019)

Korea

Cross-sectional

Hair

Atherosclerotic
cardiovascular disease

Significant inverse association.

Duan et al.
(2020)

U.S.

Cross-sectional

Urine

CVD mortality

No effects observed.

Feng et al.
(2014)

China
Cohort

Urine

Heart rate variability
indices

Significant association.

Harmon et al.
(2018)

Population-
based

U.S.

cross-sectional

Blood, urine

CVD biomarkers (oxLDL,
CRP)

No effects observed.

Long et al. (2019)

China
Cohort

Blood

Incident CVD

No effects observed.

This document is a draft for review purposes only and does not constitute Agency policy.

D-3	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported findings

Malamba-Lez et
al. (2021)

DR Congo
Case-control

Urine

Dilated cardiomyopathy
(DCM)

Significant association.

Mendv et al.
(2012)

U.S. (NHANES)
Cross-sectional

Urine

Heart failure, coronary
heart disease, heart
attack, stroke

No effects observed.

Richardson et al.
(2021)

Occupational
North

America/Europe
Cohort

Occupational

Circulatory disease
mortality

Significant association
(suggesting benefit).

Shiue and
Hristova (2014)

U.S. (NHANES)
cross-sectional

Urine

Blood pressure

Significant association.

Sankar et al.
(2014)

U.S. (NHANES)
cross-sectional

Urine

Blood pressure

Significant association.

Wu et al. (2018a)

China

Cross-sectional

Urine

Systolic and diastolic
blood pressure, diagnosis
of hypertension

No effects observed.

Ass'ad et al.
(2021)

Occupational
U.S.

Cross-sectional

Blood

Biomarkers of
inflammation (soluble
vascular cell adhesion
molecule 1)

Biomarker levels differed
between uranium miners and
non-uranium miners.

Butler-Dawson
et al. (2021)

Occupational

Guatemala

cohort

Urine

Hypertension

No effects observed.

Guseva Canu et
al. (2014)

Occupational

France

cohort

Occupational
history and
employment-
exposure-
matrix

Mortality (diseases of the
circulatory system,
ischemic myocardial
disease, cerebrovascular
diseases)

Significant increased mortality.

Karakls et al.
(2021)

Israel
Cohort

Urine

Pediatric cardiovascular-
related morbidity

No effects observed.

Pavlvushchik et
al. (2017)

Hypertensive
patients

Hair sample

Blood pressure

No effects observed.

Al Rashlda et al.
(2019)

Occupational
U.S.

Cross-sectional

Occupational

Angina

Significant association.

Samson et al.
(2016)

Occupational

France

Cohort

Occupational

Diseases of the
circulatory system

Significant deficits in deaths.

This document is a draft for review purposes only and does not constitute Agency policy.

D-4	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported findings

Shumate et al.
(2017)

Occupational
U.S.

Cross-sectional

Occupational

Angina, heart attack

No effects reported.

Suliburska et al.
(2016)

Poland

Cross-sectional

Amniotic fluid

Maternal systolic blood
pressure, diastolic blood
pressure

No effects reported.

Tret'iakov et al.
(2011)

Occupational
Russia

Occupational

Arterial hypertension,
coronary heart disease

Unclear findings.

Zablotska et al.
(2013)

Occupational

Canada

Cohort

Occupational

Mortality from CVD

No effects reported.

Table D-2. Summary of animal studies reporting on uranium-induced
cardiovascular effects

Reference

Experimental design

Author-reported findings

Vicente-Vicente et al. (2013)

Male SD rats exposed to 5.4 g/L for 11 wk
(243 mg/kg-d)

Increased systolic blood pressure.

Vicente-Vicente et al. (2013)

Male SD rats exposed to 5.4 g/L for 21 wk
(229.5 mg/kg-d)

Increased systolic blood pressure.

Grison et al. (2013)

Male SD rats exposed to 40 mg/L
(2.7 mg/kg-d) for 9 mo

No effect on plasma markers (total
cholesterol, triglycerides,
phospholipids, HDL& LDL
cholesterol).

Lestaevel et al. (2014)

Male wild type & ApoE null mice exposed
to 20 mg/L (4 mg/kg-d) for 14 wk

Dublineau et al. (2014)

Male SD rats exposed to 0, 0.009, 0.09,
0.23, 0.45, 0.9, 7.8, or 5.4 mg/kg-d for 9 mo

Souidi et al. (2012)

Male ApoE null mice exposed to 0, 20 mg/L
(4 mg/kg-d)

D.3. DEVELOPMENTAL EFFECTS

1	ATSDR Summary

2	ATSDR 2013 did not identify human studies reporting on the potential developmental

3	effects caused by uranium exposure. In their hazard evaluation ATSDR considered animal toxicity

4	studies using rats or mice as experimental models and identified developmental effects as a health

5	response to uranium exposure. Experimental designs used in these studies included gestational and

6	early postnatal exposures to uranium and they measured litter size, numbers of resorptions, live

7	fetuses, pup survival, body weight and length, internal and external malformations, and

This document is a draft for review purposes only and does not constitute Agency policy.

D-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

Protocol for the Uranium IRIS Assessment (Oral)

developmental milestones (e.g., tooth eruption, pinnae unfolding, and eye opening). In Swiss mice
gestational exposure to uranium resulted in decreased pup weight, increased neonatal death and
incidence of external malformations, and reduced litter size, viability index and lactation index. In
SD rats gestational treatment with uranium resulted in decreased pup weight, but there were no
effects on tooth eruption, pinna detachment or eye opening. In 7-day-old Wistar rats, uranium
exposure resulted in delayed tooth eruption and elevated bone resorption. ATSDR 2013 considered
the developmental effects reported in (Domingo etal.. 19891 for derivation of an acute minimal risk
level.

Newly Identified Human Studies

Nineteen (n = 19) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search (see Table D-3). Studies examined developmental-related endpoints including
preterm birth, birth weight, neural tube defects, and orofacial cleft For preterm birth, one study
found an association between maternal urinary uranium and preterm birth fZhang etal.. 20201.
whereas a nested case-control study from the U.S. observed no statistically significant associations
between maternal urinary uranium and preterm birth (Kim etal.. 20181. For birth weight, no
association was seen between umbilical cord blood uranium and birth weight in a Chinese cohort
(Yang etal.. 20201 or in toenail uranium levels in mother-infant pairs from the U.S. (Deyssenroth et
al.. 20181. Bloom etal. (20151 found reduced anthropometric measurements, including birth weight
in a U.S. cohort In a case-control study in China, (Yin etal.. 20221 observed increased risk of neural
tube defects associated with placental tissue uranium concentration. For orofacial cleft (OFC), no
association was observed fWei etal.. 20191. but another study did see associations with OFC, and
with cleft lip with cleft palate (Guo etal.. 20201.

Some studies had potential limitations due to deficiencies in analyses by only reporting
exposure averages by outcome group or correlations; deficiencies in participant selection with no
information on recruitment or inclusion criteria, with major concern for selection bias; and lack of
contrast between the low- and high-exposure groups with concerns for study sensitivity.

Newly Identified Animal Studies

Ten rat studies that met PECO criteria were identified in the IRIS literature search. In SD
rats, uranium exposure led to decreases in body weight without changes in food or water
consumption. However, several studies reported no effects on body weight of developing animals
(see Table D-4). In Wistar rats there was a decrease in pregnancy rate, labor rate, and pup survival
rate (from birth to adulthood). The study using Wistar rats also measured pup weights, and
malformations (including incidence of cleft palate, skeletal variations, or hematomas). Overall, the
results from the (Hao etal.. 20121 study are consistent with the studies and evidence summarized
in ATSDR 2013.

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

Protocol for the Uranium IRIS Assessment (Oral)

Conclusion

The available toxicological and epidemiological studies identified in the IRIS literature
search update provide further support of the studies and evidence considered by ATSDR 2013 in its
evaluation of uranium-induced developmental effects. Furthermore, newly identified
epidemiological studies provide evidence that may be considered for dose response. Based on these
findings, EPA will perform a dose-response analysis on uranium-induced developmental effects that
includes epidemiological and toxicological evidence. This will include studies identified in the IRIS
literature search and studies cited in ATSDR 2013.

Units of Analysis

Humans: Pregnancy outcomes, congenital malformations.

Animals: Fetal viability/survival or other birth parameters (e.g., resorptions, number of
pups per litter), fetal/pup growth (e.g., weight or length).

Note: An analysis of dam health (e.g., weight gain, food consumption) is also conducted to
support conclusions of specificity of the effects as being developmental (versus derivative of maternal
toxicity).

Table D-3. Studies of developmental endpoints in humans identified 2011-
2022

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported findings

Bloom et al. (2015)

U.S.
Cohort

Urine

Birth weight, birth
length, head
circumference,
gestational age

Significant associations
reported for paternal uranium
and endpoints.

Devssenroth et al. (2018)

U.S.
Cohort

Nail

Gestational age

No effects reported.

Guo et al. (2020)

China

Case-control

Umbilical cord
tissue

Orofacial clefts, cleft
lip with cleft palate

Significant associations.

Howe et al. (2022)

U.S.
Cohort

Urine

Body weight for
gestational age

No effects reported.

Kim et al. (2018)

U.S.
Cohort

Urine

Pre-term birth

No effects reported.

Wei et al. (2019)

China

Case-control

Hair

Orofacial cleft

No effects observed.

Wu et al. (2020)

China
Cohort

Urine

Tooth eruption

Significant association.

This document is a draft for review purposes only and does not constitute Agency policy.

D-7	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported findings

Yang et al. (2020)

China
Cohort

Umbilical cord
blood

Birth weight

No effects observed.

Yin et al. (2022)

China

Case-control

Placental
tissue

Neural tube defects

Significant association.

Zhang et al. (2020)

China
Cohort

Urine

Preterm birth

Significant association.

Alaani et al. (2011)

Case-

report/series
Iraq

Hair

Congenital anomalies

No effects reported.

Al-Sahlanee et al. (2017)

Cross-

sectional,

Iraq

Blood,

umbilical cord
blood

Birth weight, birth
length, head
circumference

Significant associations.

Karakis et al. (2021)

Cohort,
Israel

Urine

Preterm delivery

Significant association.

Kocvlowski et al. (2019)

Cohort,
Poland

Blood,

amniotic fluid

Birth defects

No effects reported.

Manduca et al. (2014)

Palestine
Cohort

Hair

Neural tube defects,
polycystic kidney
defect, congenital
heart disease, cleft
lift/palate

No effects reported.

Mckeating et al. (2021)

Australia

Cross-
sectional

Blood, urine

Placental weight

No effects reported.

Rhaifal-Sahlanee et al.
(2016)

Iraq
Cohort

Blood,

umbilical cord
blood

"Deformed and dead
infants."

No effects reported.

Savabieasfahani et al.
(2020)

Iraq

Case-control

Hair

Congenital
abnormalities

No effects reported.

Suliburska et al. (2016)

Poland
Cross-
sectional

Amniotic fluid

Biparietal diameter,
abdominal and head
circumference, femur
length

No effects reported.

This document is a draft for review purposes only and does not constitute Agency policy.

D-8	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table D-4. Summary of toxicological studies reporting on uranium-induced
developmental effects

Reference

Experimental design

Author-reported findings

Legendre et al. (2016)

F0 female SD rats exposed to uranyl
nitrate (0, 40,120 mg/L in drinking
water) from GD 1 to PND 168

No effect on body weight or food and water
consumption.

Lestaevel et al. (2015)

Male SD rats exposed to uranyl
nitrate (0,10, 40 mg/L in drinking
water) for 10 wk starting at birth

No effects on bodyweight or food and water
consumption.

Elmhiri etal. (2018)

Male and female SD rats exposed to
uranyl nitrate (0, 40 mg/L in drinking
water) from GD 1 to 9 mo of age

Decreased body weight in F1 male animals, but no
effect on F2 animals. No effect of food or water
consumption.

Grison et al. (2013)

Male rats exposed to uranyl nitrate
(0, 40 mg/L in drinking water) for
9 mo starting at birth

Decreased body weight, but no effect on food and
water consumption.

Grison et al. (2018)

Male and female F0 generation SD
rats exposed to uranyl nitrate (0,
40 mg/L in drinking water) for 9 mo

Increased body weight in F1 generation males;
and 4/ body weight in F2 generation males. No
effects on water consumption & no effects in F1
or F2 females.

Grison et al. (2019)

Male and female SD rats exposed to
uranyl nitrate (0, 40 mg/L in drinking
water) for 9 mo starting at birth

No effect on body weight.

Lestaevel et al. (2016)

Male & female SD rats exposed to
uranyl nitrate (0,10, 40 mg/L in
drinking water) for 9 mo starting at
birth

No effects on bodyweight or food and water
consumption.

Dinocourt et al. (2017)

Pregnant SD rats exposed to uranium
(0, 2, 6 mg/kg-d in drinking water)
during gestation

No effects on bodyweight or food and water
consumption.

Legrand et al. (2016a)

Pregnant SD rats exposed to
depleted uranium (0, 10, 120 mg/L in
drinking water) during gestation

Decreased body weight on PND0 and increased
body weight on PND5 and PND21.

Haoetal. (2012)

Male and Female Wistar rats
exposed to depleted uranyl nitrate
(0, 4, 40 mg/kg-d, in food) for 4 mo
starting at weaning

Decreased pregnancy rate, labor rate, pup survival
rate (at birth and adulthood), and number of pups
produced. No effect on pup weights, incidence of
cleft palate, skeletal variations, or hematomas.

D.4. ENDOCRINE EFFECTS

1	ATSDR Summary

2	ATSDR 2013 did not identify human studies informing potential uranium-induced

3	endocrine effects. ATSDR 2013 did identify several experimental studies in animal models.

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

18

19

20

21

22

23

24

Protocol for the Uranium IRIS Assessment (Oral)

Although two studies using rats report histopathological effects in the thyroid, the majority of the
available evidence from experiments using rats or rabbits did not report an association between
uranium exposure and endocrine effects in the adrenal, pancreas, thyroid, thymus, parathyroid, or
pituitary.

Newly Identified Human Studies

Ten (n = 10) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for endocrine outcomes (see Table D-5). Many studies were conducted using
NHANES data. Significant associations were observed between urinary uranium and measures of
thyroid hormones fKim etal.. 2022: Christensen. 20121: thyroid antibodies fvan Gerwen etal..
20201: and diabetes (Menke etal.. 20161. No effects were reported for thyroid problems (Mendv et
al.. 20121 and diabetes (Yang etal.. 20221. A few studies had potential limitations, including due to
reporting the exposure-outcome association only as exposure averages for outcomes groups.

Newly Identified Animal Studies

No new animal studies informing endocrine effects after oral exposure to uranium were
identified in the literature search update. Studies that evaluated uranium-induced effects on
reproductive hormones are described in the reproductive effects section.

Conclusion

The epidemiological studies identified in the IRIS literature search suggests that uranium
exposure may impact the endocrine system. Based on these findings, EPA will perform a hazard
evaluation of uranium-induced endocrine effects. This analysis will consider studies cited in ATSDR
and studies that met PECO criteria in the IRIS literature search.

Units of Analysis

Humans: Thyroid hormone measures, diabetes.

Animals: Hormone measures, organ weights, organ morphology/histopathology.

Table D-5. Studies of endocrine endpoints in humans identified 2011-2022

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Christensen
(2012)

U.S. (NHANES)
Cross-sectional

Urine

Thyroid hormones

Significant association.

Kim et al. (2022)

U.S. (NHANES)
Cross-sectional

Urine

Thyroid hormones

Significant association.

Mendv et al.
(2012)

U.S. (NHANES)
Cross-sectional

Urine

Thyroid problems

No effects reported.

This document is a draft for review purposes only and does not constitute Agency policy.

D-10	DRAFT-DO NOT CITE OR QUOTE


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Menke et al.
(2016)

U.S. (NHANES)
Cross-sectional

Urine

Diabetes

Significant association.

van Gerwen et al.
(2020)

U.S. (NHANES)
Cross-sectional

Urine

Thyroid antibodies

Significant association.

Yang et al. (2022)

U.S. (NHANES)
Cross-sectional

Urine

Type 2 diabetes

No effects reported.

Samson et al.
(2016)

Occupational

France

Cohort

Occupational

Endocrine, metabolic
disease mortality

Significant deficits in deaths.

Stoisavlievic et al.
(2019)

Serbia

Cross-sectional

Thyroid tissue

Thyroid disease

No effects observed.

Stoisavlievic et al.
(2020b)

Serbia

Case-control

Thyroid tissue

Colloid goiter disease

No effects observed.

Stoisavlievic et al.
(2020a)

Serbia

Cross-sectional

Thyroid tissue,
blood, urine

Hashimoto's
thyroiditis

No effects observed.

D.5. GASTROINTESTINAL EFFECTS

ATSDR Summary

ATSDR cited two case studies where individuals were acutely exposed to uranyl nitrate
(14.3 mg/kg) or uranyl acetate (131 mg/kg) and reported nausea, diarrhea, vomiting, and paralytic
ileus. They also cited animal studies using rats or rabbits that measured organ weight changes and
histopathology of the gastrointestinal system. In male and female SD rats and New Zealand white
rabbits, exposure to uranium up to 91 days did not affect organ weight or histopathology.

Newly Identified Human Studies

Two (n = 2) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for gastrointestinal effects. Both were occupational studies. One (Richardson etal..
20211 examined mortality from noncancer diseases of the digestive system and did not find an
association. The other (Samson et al.. 20161 also examined mortality from noncancer diseases of
the digestive system. The study found a reduced standardized mortality ratio but had a potential
limitation due to selection bias from the healthy worker effect

Newly Identified Animal Studies

No new animal studies were identified in the literature search update.

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

18

19

20

21

22

23

24

25

26

Protocol for the Uranium IRIS Assessment (Oral)

Conclusion

EPA will not consider gastrointestinal effects for hazard evaluation or dose response.

Units of Analysis

N/A

D.6. HEMATOLOGICAL EFFECTS

ATSDR Summary

ATSDR 2013 considered one case study in which an individual was exposed to a large dose
of uranium (15 g) plus benzodiazepine. The study reported anemia over a period of 8 weeks.

ATSDR also identified experimental studies using SD rats or New Zealand white rabbits and
concluded that most animal studies show no uranium-induced effects on hematological parameters.

Newly Identified Human Studies

Two (n = 2) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for hematological effects (see Table D-6). Both had potential limitations due to
reporting the exposure-outcome association only as exposure averages for outcome groups or
concern for selection bias.

Newly Identified Animal Studies

Two animal chronic exposure toxicity studies were identified in the literature search.
(Grison etal.. 20131 and (Dublineau etal.. 20141 used SD rats exposed to UN for 9 months. Both
studies report that uranium exposure had no significant effects on hematological parameters
including platelets, RBC and WBC counts, hemoglobin, lymphocytes hematocrit, granulocytes, or
monocytes. fDublineau et al.. 20141 observed alterations on cytokines indicative of changes in
hematopoiesis, but blood cell production was unaltered in the bone marrow and spleen.

Conclusion

Because of null evidence from experimental and epidemiological studies EPA will not
consider hematological effects for hazard evaluation or dose response.

Units of Analysis

N/A

This document is a draft for review purposes only and does not constitute Agency policy.

D-12	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

Protocol for the Uranium IRIS Assessment (Oral)

Table D-6. Studies of hematological endpoints in humans identified 2011-
2022

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported
findings

Henriauez-
Hernandez et
al. (2017)

Cross-
sectional

Gran Canaria

Blood sample

Anemia

No effects observed.

Samson et al.
(2016)

Occupational

France

cohort

Occupational

Mortality: Diseases of the blood and
blood-forming organs

No effects observed.

D.7. HEPATIC EFFECTS

ATSDR Summary

ATSDR 2013 considered human and animal toxicological study evidence in their evaluation
of uranium-induced liver effects. A case report in which a patient had elevated serum liver enzymes
levels after exposure to a large dose of uranyl acetate (approximately 15 g) was considered. ATSDR
also considered animal toxicity studies performed in dogs, rats, and rabbits. ATSDR 2013 concluded
that "in the available animal studies, the existing data provide evidence that uranium exposure can
damage the liver," and that "few human data are available on the hepatic effects of uranium."

Newly Identified Human Studies

One study meeting PECO criteria was identified in the IRIS literature search (Samson etal..
20161. It had a potential limitation over the ability of the outcome measure to correctly classify
liver disease, as it examined liver disease combined with "psychosis and other diseases due [sic]—
alcohol."

Newly Identified Animal Studies

Ten animal toxicity studies that meet PECO criteria were identified in the IRIS literature
search (see Table D-7). These studies used SD rats, several strains of mice (including C57BL/6J,
Kunming, and CBA), and genetically modified ApoE null mice. Studies using mice exposed animals
to uranium for 30 days to 4 months. Studies using SD rats exposed animals for 1 to 18 months.
Outcomes considered in the available studies included organ weight measures, macroscopic
appearance, serum markers of liver damage, and histology. In mice, uranium exposure did not
affect liver macroscopic appearance, or clinical markers of liver disease, but one study reported
altered hepatic lipid composition. In SD rats several studies reported alterations in serum markers
of liver disease and one study reported increased liver weight However, these effects were not
accompanied by histopathological responses, and there was no increase in severity after chronic
exposures (9 to 18 months).

This document is a draft for review purposes only and does not constitute Agency policy.

D-13	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	Conclusion

2	The available toxicological studies identified in the literature search update provide further

3	support of the studies and evidence considered by ATSDR 2013 in its evaluation of uranium-

4	induced liver effects. Based on these findings, EPA will perform a dose-response analysis on

5	uranium-induced liver effects. This will include studies identified in the IRIS literature search and

6	studies cited in ATSDR 2013 fATSDR. 2013)7

7	Units of Analysis

8	Humans: Liver disease.

9	Animals: Organ weight, organ morphology/histopathology, clinical measures of biliary
10	function, clinical measures of liver function (including liver enzymes).

Table D-7. Summary of toxicological studies reporting on uranium-induced
hepatic effects

Reference

Experimental design

Author-reported findings

Mouse studies

Bolt et al. (2019)

Male & female C57BL/6J mice
exposed to uranyl acetate (0, 5,
50 mg/L in drinking water) for
60 d

No effect on serum markers of liver disease (ALT
and ALP).

Haoetal. (2013b)

Male Kunming mice exposed to
uranyl nitrate (0, 0.4, 4,
40 mg/kg-d in food) for 4 mo

No effect on markers of liver damage (ALT, AST).

Kudvasheva et al. (2020)

Male CBA mice exposed to
uranyl nitrate (0, 2 mg/L in
drinking water) for 60 d

Altered hepatic lipid composition.

Souidi et al. (2012)

Male ApoE null mice exposed to
uranyl nitrate (0, 20 mg/L in
drinking water) for 3 mo

No effects on markers of liver damage (ALT, AST),
liver weight, or macroscopic appearance.

Rat studies

Dublineau et al. (2014)

Male SD rats exposed to uranyl
nitrate (0.009, 0.09, 0.23, 0.45,
0.9, 1.8, 5.4 mg/kg-d in drinking
water) for 9 mo

No macroscopic or histological effects. Increased
ALT and AST at high dose, but effect not
statistically significant. No effects on bilirubin.

Grison et al. (2013)

Male rats exposed to uranyl
nitrate (0, 40 mg/L in drinking
water) for 9 mo starting at birth

Increased AST, but no effect on ALP, ALT, or
bilirubin.

Grison et al. (2019)

Male rats exposed to uranyl
nitrate (0, 40 mg/L in drinking
water) for 9 mo starting at birth

No effect on plasma markers of liver damage.

This document is a draft for review purposes only and does not constitute Agency policy.

D-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

Protocol for the Uranium IRIS Assessment (Oral)

Reference

Experimental design

Author-reported findings

Gueguen et al. (2014)

Male SD rats exposed to uranyl
nitrate (0, 40 mg/L in drinking
water) for 1-18 mo

No effects on liver weight, histopathology, or
markers of liver damage (ALT, AST, or bilirubin).

Male SD rats exposed to uranyl
nitrate (0,10, 40,120 mg/L in
drinking water) for 9 mo

Increased relative liver weight, but no effect on
markers of liver damage (ALT, AST, or bilirubin).

Legendre et al. (2016)

Male SD rats exposed to uranyl
nitrate (0, 40,120 mg/L in
drinking water) from GD 1 to
PND 168

Increased ALT and AST/ALT, but no effect on AST.

Poisson et al. (2014b)

Male SD rats exposed to uranyl
nitrate (0, 40,120, 400 mg/L in
drinking water) for 3 mo

No effects on liver histopathology or markers of
liver disease.

Male SD rats exposed to uranyl
nitrate (0, 40,120, 600 mg/L in
drinking water) for 9 mo

No effects on liver histopathology or markers of
liver disease.

D.8. IMMUNE EFFECTS

ATSDR Summary

ATSDR 2013 did not identify human studies informing potential uranium-induced
immunological effects. ATSDR 2013 did identify experimental studies using rats, mice or New
Zealand white rabbits and concluded that exposure "to uranium had no significant effect on
immune system function."

Newly Identified Human Studies

Eleven (n = 11) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for immunological outcomes (see Table D-8). A number of studies observed
significant associations, including with ankylosing spondylitis, lupus, immunotoxicity, and
rheumatoid arthritis. The remaining studies observed no significant associations with
autoimmunity or arthritis. One study had potential limitations due to reporting exposure-outcome
associations only as exposure averages for outcome groups.

Newly Identified Animal Studies

Five animal toxicity studies (two using rats and three using mice) were identified in the IRIS
literature search. Outcomes considered in these studies include organ weights, histopathology,
hematological endpoints, and immune function measures. In rat studies exposure was associated
with decreased thymus and spleen weight, alterations in immune cell composition and functions,
and bone marrow and spleen histopathology (see Table D-9). In mice uranium treatment altered

This document is a draft for review purposes only and does not constitute Agency policy.

D-15	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	natural killer and macrophage functions, increased cytokine production, changes in immune cell

2	numbers and functions (see Table D-9).

3	Conclusion

4	The toxicological and epidemiological studies identified in the IRIS literature search

5	suggests that uranium exposure may impact the immune system. Based on these findings, EPA will

6	perform a hazard evaluation of uranium-induced immunological effects. This analysis will consider

7	studies cited in ATSDR 2013 and studies that met PECO criteria in the IRIS literature search.

8	Units of Analysis

9	Humans: Autoimmune disease and measures, immunotoxicity.

10	Animals: Organ weights, clinical endpoints (e.g., immune cell counts/responses), immune

11	functional measures, organ morphology/histopathology.

Table D-8. Studies of immunological endpoints in humans identified 2011-
2022

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported
findings

Aung et al. (2019)

U.S.

Cross-sectional

Urine

Immune markers of
inflammation

Significant association.

Chen et al. (2022a)

U.S. (NHANES)
Cross-sectional

Urine

Rheumatoid arthritis

Significant association.

Chen et al. (2022b)

U.S. (NHANES)
Cross-sectional

Urine

Osteoarthritis

No effect reported.

Erdei et al. (2019)

U.S.

Cross-sectional

Urine

Autoimmunity

Significant association.

Greene et al. (2019)

U.S.

Case-control

Blood

Chemokines
(endometriosis cases)

Significant association.

Lourenco et al.
(2013)

Portugal
Cross-sectional

Blood

Immune cell count

Significant association.

Lu-Fritts et al. (2014)

U.S.

Case-control

Air

Lupus

Significant association.

Mendv et al. (2012)

U.S. (NHANES)
Cross-sectional

Urine

Arthritis

No effects reported.

Scammell et al.
(2020)

U.S., Nicaragua
Cross-sectional

Urine

Autoimmunity

No effects reported.

Shiue (2014)

U.S. (NHANES)

Urine

Ankylosing spondylitis

Significant association.

This document is a draft for review purposes only and does not constitute Agency policy.

D-16	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported
findings



Cross-sectional







Denisova et al.
(2020)

Russia

Case-control

Lung tissue

Sarcoidosis

No effects observed.

Table D-9. Summary of toxicological studies reporting on uranium-induced
immunological effects

Reference

Experimental design

Author-reported findings

Mouse studies

Bolt et al. (2019)

Male and female C57BL/6J mice
exposed uranyl acetate (0, 5,
50 ppm in drinking water) for 60 d

Decreased percent macrophages and natural killer
cells in male spleen. No effect on immune tissue
weights, immune cell recoveries or viability, or
immune responses.

Medina et al. (2020)

Male and female C57BL/6J mice
exposed to uranyl acetate (0, 5,
50 ppm in drinking water) for 45 d

Decreased intraepithelial lymphocyte subsets in
small intestine of males but no effect in females.
No effect on innate immune cells.

Haoetal. (2013b)

Male Kunming mice exposed to
uranyl nitrate (0, 0.4, 4,
40 mg/kg-d in food)

Decreased natural killer cell and macrophage
functions; T* IgG and IgE levels; altered splenic T
and B cells proliferation; T* delayed-type
hypersensitivity; altered T cell and B cell subtypes
and cytokine production in splenic cells.

Rat studies

Haoetal. (2013a)

Female SD rats exposed to
depleted uranyl nitrate (0,1.3,13,
130 mg/kg in food) for 4 mo

Decreased thymus and spleen weight. Altered
immune cell composition and functions, and bone
marrow, and spleen histopathology.

Dublineau et al. (2014)

Male SD rats exposed to uranyl
nitrate (0.009, 0.09, 0.23, 0.45,
0.9,1.8, 5.4 mg/kg-d in drinking
water) for 9 mo

Decreased intestinal macrophages by 50% but
effect was not dose-dependent and not statistically
significant.

D.9. METABOLIC EFFECTS

1	ATSDR Summary

2	ATSDR 2013 cited two acute exposure studies that report altered levels of l,25(OH)2D3, the

3	active form of vitamin D, after a single exposure to depleted uranyl nitrate. Vitamin D levels were

4	measured at 1- or 3-days post exposure. No subchronic or chronic experimental studies and no

5	epidemiological studies on metabolic effects were identified in ATSDR 2013.

This document is a draft for review purposes only and does not constitute Agency policy.

D-17	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	Newly Identified Human Studies

2	Six (n = 6) epidemiological studies meeting PECO criteria were identified in the IRIS

3	literature search for metabolic outcomes (see Table D-10). Urinary uranium was significantly

4	associated with increased risk of metabolic syndrome in a cross-sectional study fXu etal.. 20201. No

5	associations were observed in studies examining urinary uranium and diabetes (Wang etal.. 2020:

6	Chafe etal.. 2018: Liu etal.. 20161. T wo studies had potential limitations including concern for

7	exposure assessment misclassification and only reporting the exposure-outcome association as

8	exposure averages for outcome groups.

9	Newly Identified Animal Studies

10	No new animal studies were identified in the literature search update.

11	Conclusion

12	Use of a lack of evidence from experimental studies, and only one epidemiological study

13	that observed an association cross-sectionally, EPA will not consider hematological effects for

14	hazard evaluation or dose response.

15	Units of Analysis

16	N/A

Table D-10. Studies of metabolic endpoints in humans identified 2011-2022

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Liu et al.
(2016)

Occupational
China

Cross-sectional

Urine

Diabetes

No effects reported.

Chafe et al.
(2018)

Canada
Case-control

Drinking water

Type 1 diabetes

No effects reported.

Xu et al.
(2020)

China

Cross-sectional

Urine

Metabolic syndrome

Significant association.

Wang et al.
(2020)

United States
Cohort

Urine

Diabetes

No association observed.

Su et al.
(2012)

China

Case-control

Blood

Gouty arthritis

No effects reported.

Zablotska et
al. (2013)

Occupational

Canada

Cohort

Occupational

Mortality-diabetes

No effects reported.

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

19

20

21

22

23

24

25

26

27

28

29

30

Protocol for the Uranium IRIS Assessment (Oral)

D.10. MUSCULOSKELETAL EFFECTS

ATSDR Summary

ATSDR 2013 considered one case study in which an individual was exposed to a large dose
of uranium plus benzodiazepine, and a case-control study reporting a significant association
between uranium exposure and serum type I collagen carboxy-terminal telopeptide (a marker of
bone resorption). They also cite three animal toxicity studies that include acute, short-term, and
subchronic studies using rats, mice, or rabbits. In mice, uranium exposure resulted in decreased
percent metaphyseal activity in bone formation and increased bone resorption, but in SD rats and
New Zealand rabbits there were no effects in histological measures of bone damage. ATSDR
concluded that "there are limited data on the potential of uranium to induce bone or muscle
damage." ("ATSDR. 2013118

Newly Identified Human Studies

Five (n = 5) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for musculoskeletal outcomes (see Table D-ll). No associations were observed in
studies examining systemic sclerosis, muscle strength, or mortality from diseases of the
musculoskeletal system. Significant findings were seen in an NHANES study examining the
association with bone density (Park and An. 20221. One study had potential limitations including
selection bias.

Newly Identified Animal Studies

Three animal toxicity studies (one short-term and two chronic exposures) were identified
in the literature search. They exposed young SD rats for 3 to 28 days or 9 months and reported
alterations in cortical bone parameters, reduced bone mineral density, and altered mRNA levels of
genes associated with bone development and functions (see Table D-12). One study (Wade-Gueye
etal.. 2012) compared responses in young and sexually mature animals and observed that younger
individuals appear to be more susceptible to uranium-induced bone effects.

Conclusion

The toxicological and epidemiological studies identified in the IRIS literature search suggest
that uranium exposure may impact the skeletal system and that early lifestages may represent a
susceptible population. Based on these findings, EPA will perform a hazard evaluation of uranium-
induced musculoskeletal effects. This analysis will consider studies cited in ATSDR 2013 and
studies that met PECO criteria in the IRIS literature search.

18fATSDR. 20131 also considered uranium-induced skeletal effects after gestational exposure in mice (see
Domingo etal. 1989, and ATSDR 2013 Developmental Effects section 3.2.2.6).

This document is a draft for review purposes only and does not constitute Agency policy.

D-19	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	Units of Analysis

2	Human: Musculoskeletal conditions, muscle, and bone health.

3	Animal: Muscular & skeletal morphology/histopathology, clinical markers of

4	musculoskeletal disease, and parameters/measures of bone development and function.

This document is a draft for review purposes only and does not constitute Agency policy.

D-20	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table D-ll. Studies of musculoskeletal endpoints in humans identified 2011-
2022

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported
findings

Marie et al. (2017)

Case-control
France

Hair

Systemic sclerosis

No associations
observed.

Park and An (2022)

U.S.

(NHANES)

Cross-

sectional

Urine

Bone density

Significant association.

Wu et al. (2022)

Cross-

sectional

U.S.

Urine

Muscle strength

No effects reported.

Shumate et al. (2017)

Occupational
Cross-
sectional
U.S.

Urine

Arthritis/back pain

Significant association.

Samson et al. (2016)

Occupational

France

cohort

Occupational

Diseases of the
musculoskeletal system-
mortality

No effects reported.

Table D-12. Summary of toxicological studies reporting on uranium-induced
musculoskeletal effects

Reference

Experimental design

Author-reported findings

Wade-Gueve et al.
(2012)

Newborn and mature male SD rats
exposed to uranyl nitrate (0, 40 mg/L
in drinking water) for 9 mo

Cortical bone parameters were affected in the
young animals. No effect in adults. No effect on
clinical markers.

Rodrigues et al. (2013)

Weaning female Wistar rats exposed
to uranyl nitrate (0, 50 ppm in food)
for 3,7,11,14, 21, or 28 d

Decreased femoral bone mineral density.

Souidi et al. (2018)

Newborn male SD rats exposed to 0,
1.5, 10, 40 ppm (0, 0.18, 1.2,
4.8 mg/kg-d)

Decreased cortical bone diameter in the femur.
No effect on microarchitecture parameters, bone
mineral density, or serum markers.

This document is a draft for review purposes only and does not constitute Agency policy.

D-21	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

Protocol for the Uranium IRIS Assessment (Oral)

D.ll. NEUROLOGICAL EFFECTS

ATSDR Summary

ATSDR 2013 identified neurobehavioral health effects as a response to uranium exposure.
ATSDR 2013 did not identify human studies reporting on neurological effects, but considered
toxicological studies using several rat strains, mice, or New Zealand rabbits. In SD and Long-Evans
rats and in Swiss mice exposure to uranium lead to altered behaviors such as line crossing and
rearing behaviors, and motor activity. Brain neurotransmitter levels and sleep cycles were also
altered in exposed rats. However, brain histopathology was not affected in rats or rabbits.

Newly Identified Human Studies

Thirteen (n = 13) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for neurological outcomes (see Table D-13). One study observed a significant
association with schizophrenia (Ma etal.. 20181. but the two other studies saw no association with
cognitive performance. Many studies had potential limitations, including due to not accounting for
confounding and reporting the exposure-outcome association only as exposure average for
outcomes groups.

Newly Identified Animal Studies

Nine animal toxicity studies (eight using rats and one using mice) were identified in the IRIS
literature search. Outcomes considered in these subchronic and chronic exposure studies include
behavioral and functional measures, histopathology, and neurotransmitter levels. Experimental
studies using rats report alterations in behaviors (e.g., depressive, and anxiety-like behaviors) and
functions (e.g., decreased locomotor activity), and increased neurocellular damage (e.g., apoptosis,
and reduced spinal motor neurons) after oral exposure to uranium (see Table D-14). In both mice
and rats, uranium exposure was associated with impaired memory.

Conclusion

The available toxicological studies identified in the literature search update provide further
support of the studies and evidence considered by ATSDR 2013. Based on these findings, EPA will
evaluate the available evidence (studies identified in the IRIS literature search and studies cited in
ATSDR 2013) for dose-response analysis on uranium-induced neurological effects.

Units of Analysis

Humans: Cognitive function, brain disorders.

Animals: Learning/memory, brain morphology/histopathology, neurodegenerative disease,
neurotransmitter levels/function, organ weights.

This document is a draft for review purposes only and does not constitute Agency policy.

D-22	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table D-13. Studies of neurological endpoints in humans identified 2011-
2022

Reference

Study design

Exposure
measu rement

Endpoints

Author-reported findings

Ma etal. (2018)

China

Case-control

Blood

Schizophrenia

Statistically significant.

Nozadi et al.
(2021)

U.S.
Cohort

Blood, urine

Gross motor, fine motor,
problem solving, personal-
social

No effects(s) reported.

Wang et al. (2022)

U.S.

(NHANES)

Cross-
sectional

Urine

Cognitive performance

No effect(s) reported.

Adams et al.
(2013)

U.S.

Case-control

Blood, urine

Autism

Significant association.

De Benedetti et al.
(2017)

Italy

Case-control

Blood

Amyotrophic lateral
sclerosis (ALS)

No effect(s) reported.

Fiore et al. (2020)

Italy

Cross-

sectional

Hair

Autism

No effect(s) reported.

Harchaoui et al.
(2020)

Case-control

Hair

Autism

No effect(s) reported.

Karakls et al.
(2021)

Israel
Cohort

Urine

Developmental disorders

No effect(s) reported.

Lin et al. (2022)

Taiwan

Cross-
sectional

Blood

Alzheimer's disease

Statistically significant (suggesting
benefit).

Roos et al. (2013)

Norway
Case-control

Blood

Amyotrophic lateral
sclerosis (ALS)

No effects(s) reported.

Samson et al.
(2016)

Occupational

France

Cohort

Occupational

Non-malignant tumors of
the central nervous system

No effect(s) reported.

Torrente et al.
(2013)

Spain
Cohort

Hair

Motor function, behavioral
outcomes in children

No effect(s) reported.

Tretvakov et al.
(2011)

Occupational
Russia

Unclear

Cognitive function

No effect(s) reported.

Table D-14. Summary of toxicological studies reporting on uranium-induced
neurological effects

Reference

Experimental design

Author-reported findings

Mouse studies

This document is a draft for review purposes only and does not constitute Agency policy.

D-23	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Experimental design

Author-reported findings

Lestaevel et al. (2014)

Male C57BL/6J and ApoE null mice
exposed uranyl nitrate (0, 20 mg/L in
drinking water) for 14 wk

In ApoE null animals, uranium impaired
working memory, but no effect on anxiety-like
behavior or cerebral cortex levels of
acetylcholine.

Rat studies

Dublineau et al. (2014)

Male SD rats exposed to uranyl nitrate
(0.009, 0.09, 0.23, 0.45, 0.9, 1.8,
5.4 mg/kg-d in drinking water) for 9 mo

No effect on brain acetylcholine levels.

Lestaevel et al. (2015)

Male SD rats exposed to uranyl nitrate
(0, 10, 40 mg/L in drinking water) for
10 wk starting at birth

Decreased locomotor activity, but no effect on
rearing movements; increased anxiety-like
behavior and decreased depressive-like
behavior and rotarod.

Lestaevel et al. (2013)

Male SD rats exposed to uranyl nitrate
(0, 10, 40 mg/L in drinking water)
during gestation plus 10 wk

Decreased object recognition memory. No
effect on sleep-wake cycle or spatial working
memory.

Saint-Marc et al. (2016)

Male SD rats exposed to uranyl nitrate
(0,1, 40,120 mg/L in drinking water)
for 9 mo

Decreased in the number of spinal motor
neurons.

Lestaevel et al. (2016)

Male & female SD rats exposed to
uranyl nitrate (0,10, 40 mg/L in
drinking water) from PND 1-250)

Altered behaviors (motor activity, spatial
working memory, anxiety, depressive-like
behavior).

Legrand et al. (2016a)

Pregnant SD rats exposed to depleted
uranium (0,10,120 mg/L in drinking
water) during gestation

Increased cell death and apoptosis and
reduced dividing cells in dentate gyrus.
Increased cell proliferation in dentate
neuroepithelium.

Legrand et al. (2016b)

Pregnant SD rats exposed to uranium
(0, 6 mg/kg-d in drinking water) during
gestation

Altered neuronal cell differentiation in
hippocampal dentate gyrus, and depression
behavior. No effect on locomotor activity,
exploratory activity, or spatial memory.

Dinocourt et al. (2017)

Pregnant SD rats exposed to uranium
(0, 2, 6 mg/kg-d in drinking water)
during gestation

Altered behaviors (depressive-like behavior,
spatial memory) No effect on hippocampal
morphology. Altered pyramidal cells in
hippocampus.

D.12. REPRODUCTIVE EFFECTS

1	ATSDR Summary

2	ATSDR 2013 did not identify human studies reporting on the potential reproductive effects

3	caused by uranium exposure, but they identified and evaluated animal toxicity studies using rats or

4	mice as experimental models and evaluated male and female reproductive outcomes. ATSDR 2013

This document is a draft for review purposes only and does not constitute Agency policy.

D-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

Protocol for the Uranium IRIS Assessment (Oral)

identified reproductive effects as a health response to uranium exposure. Reproductive effects
observed in studies evaluating male mice and rats include reduced pregnancy rates, numbers of
spermatozoa and epididymal weight after uranium treatment. Female reproductive effects were
reported in studies using murine models and include altered ovarian folliculogenesis, increased
percentage of dysmorphic oocytes, reduced mitotoxic index in oocyte supporting cells, and reduced
proportion of healthy oocytes in exposed mice.

Newly Identified Human Studies

Five (n = 5) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for reproductive outcomes (see Table D-15). A cohort study from Lebanon found
uranium in seminal fluid was significantly associated with low progressive motility, low normal
morphology, and low sperm viability (Sukhn etal.. 20181. In the U.S., (Branch etal.. 20211 observed
urinary uranium to be significantly positively associated with DNA fragmentation index, while
fWang etal.. 20161 observed no effects in a Chinese cohort A few studies had potential limitations
due to a limited exposure contrast and reporting the exposure-outcome association only as
exposure averages for outcome groups.

Newly Identified Animal Studies

Six animal toxicity studies that meet PECO criteria were identified in the IRIS literature
search (see Table D-16). These studies used SD or Wistar rats to evaluate potential U-induced male
and female reproductive effects. Two studies evaluated effects in the male reproductive system
after gestational or chronic exposures. Chronic (6- or 12-month) exposures lead to increased
nuclear pyknosis in testis, decreased spermatocytes and spermatids, and reduced serum
testosterone but no effects on follicle-stimulating hormone levels. Gestational plus postnatal
exposures resulted in altered reproductive hormone levels (decreased plasma testosterone and
intratesticular estradiol, and increased plasma luteinizing hormone and follicle-stimulating
hormone) and increased absolute testicular weight (without changes in relative weight).

Four studies evaluated reproductive outcomes after exposing male and female rats and
evaluated effects in F0, Fl, or F2 generation animals (see Table D-16). Effects reported include
uranium-induced changes in reproductive organ weights and alterations in reproductive hormone
levels after exposure. Sperm measures were also measured. Uranium treatment for 9 months
altered sperm morphology in F0, Fl, and F2 SD animals. Finally, pregnancy rates were considered,
and exposure was associated with decreased pregnancy rate in F0 and Fl animals.

Conclusion

The available toxicological and epidemiological studies identified in the IRIS literature
search update provide further support of the studies and evidence considered by ATSDR 2013 in its
evaluation of uranium-induced reproductive effects. Furthermore, newly identified toxicological
and epidemiological studies provide evidence that may be considered for dose response. Based on

This document is a draft for review purposes only and does not constitute Agency policy.

D-25	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	these findings, EPA will perform a dose-response analysis on uranium-induced male and female

2	effects that includes toxicological evidence identified by ATSDR 2013 and epidemiological and

3	toxicological evidence captured in the IRIS literature search.

4	Units of Analysis

5	Humans: Semen quality.

6	Animals: Organ morphology/histopathology, developmental measures, reproductive

7	hormone measures, functional measures.

Table D-15. Studies of reproductive endpoints in humans identified 2011-
2022

Reference

Study
design

Exposure
measurement

Endpoints

Author-reported findings

Branch et al. (2021)

Cohort
U.S.

Urine

Semen quality

Significant association
(suggesting benefit).

Sukhn et al. (2018)

Cohort
Lebanon

Blood, seminal
fluid

Semen quality markers

Significant associations.

Wang et al. (2016)

Cohort
China

Urine

Spermatozoa apoptosis

measures,

Sperm DNA damage

parameters

No effect(s) reported.

McKeating et al.
(2020)

Cohort
Australia

Cord blood

Pregnancy complications

No effect(s) reported.

Wang et al. (2017)

Cohort
China

Seminal plasma

Sperm apoptosis

Uranium not analyzed further
except for exploratory purposes.

This document is a draft for review purposes only and does not constitute Agency policy.

D-26	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Table D-16. Summary of toxicological studies reporting on uranium-induced
reproductive effects

Reference

Experimental design

Author-reported findings

Studies evaluating male repro toxicity

Lu etal. (2021)

Male SD exposed to depleted
uranium (0, 3, 30, 300 ppm in
food) for 60 d

Increased nuclear pyknosis in testis. Decreased
spermatocytes and spermatids, and decreased serum
testosterone.

Legendre et al.
(2016)

Female SD rats exposed to uranyl
nitrate (0, 40,120 mg/L in drinking
water) from GD 1 to PND 168

Increased absolute testis weight, but no effect on
relative weight. No effect on epididymis weight or
sperm measures. Decreased plasma testosterone and
intratesticular estradiol. Increased plasma LH and FSH.

Studies exposing males and females

Haoetal. (2012)

Male and female Wistar rats
exposed to depleted uranyl nitrate
(0, 0.3, 3 mg/kg-d in food) for 4 mo

Decreased pregnancy rate. In F0 and F1 males:
increased serum T and decreased serum FSH. In F0
males: Increased serum LH. In F1 males: decreased
serum LH.

Grison et al. (2022)

Male and female SD rats exposed
to uranyl nitrate (0, 40 mg/L in
drinking water) for 9 mo; animals
mated at 6 mo

Decreased pregnancy rate in F1 generation animals. No
effect on the number of pups per litter or the male
female ratio in F0, Fl, or F2 generation animals.

Elmhiri et al.
(2018)

Male and female SD rats exposed
to uranyl nitrate (0, 40 mg/L in
drinking water) for 9 mo and then
mated

Increased testes and ovaries weights. These effects
were not apparent in F0 and Fl animals.

Legendre et al.
(2019)

Male and female SD rats exposed
to uranyl nitrate (0, 40 mg/L in
drinking water) for 9 mo

Altered sperm morphology in F0, Fl, and F2 generation
animals. Decreased pregnancy rate and epididymis
weight in Fl generation animals only.

LH = luteinizing hormone; FSH = follicle stimulating hormone.

D.13. RESPIRATORY EFFECTS

1

2

3

4

5

6

7

8

This document is a draft for review purposes only and does not constitute Agency policy.

D-27	DRAFT-DO NOT CITE OR QUOTE

ATSDR Summary

ATSDR 2013 considered human and animal toxicological study evidence in their evaluation
of uranium-induced respiratory effects after oral exposure. A case report in which a patient had
elevated serum liver enzymes levels after exposure to a large dose of uranyl acetate (approximately
15 g) was considered. ATSDR also considered animal toxicity studies performed in dogs, rats, and
rabbits. Experimental designs used in these studies included chronic, subchronic, and short-term
exposures and measured histopathological endpoints. ATSDR concluded that respiratory effects
from oral exposure to uranium are unlikely.


-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

Protocol for the Uranium IRIS Assessment (Oral)

Newly Identified Human Studies

Sixteen (n = 16) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for respiratory outcomes (see Table D-17). Three studies observed urinary
uranium to be significantly associated with asthma or emphysema prevalence (Li etal.. 2021:

Huang etal.. 2016: Mendv etal.. 20121: and one occupational study observed increased risk of
breathless and pulmonary symptoms (Shumate etal.. 20171. Several studies had potential
limitations, including concerns over confounding, selection bias, exposure assessment
misclassification, and lack of contrast

Newly Identified Animal Studies

No new animal studies informing respiratory effects after oral exposure to uranium were
identified in the literature search update.

Conclusion

The epidemiological studies identified in the IRIS literature search suggests that uranium
oral exposure may impact the respiratory system. Based on these findings, EPA will perform a
hazard evaluation of uranium-induced respiratory effects. This analysis will consider studies cited
in ATSDR 2013 and studies that met PECO criteria in the IRIS literature search.

Units of Analysis

Humans: Respiratory disease, pulmonary symptoms.

Animals: Organ weights, organ morphology/histopathology, functional measures.

Table D-17. Studies of respiratory endpoints in humans identified 2011-2022

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Feng et al. (2015)

Cross-sectional
China

Urine

Pulmonary function

No effects observed.

Huang et al. (2016)

Case-control
China

Urine

Asthma

Significant association.

Li et al. (2021)

U.S. (NHANES)
Cross-sectional

Urine

Asthma

Significant association.

Mendv et al. (2012)

U.S. (NHANES)
Cross-sectional

Urine

Asthma, emphysema

Significant association.

Rahman et al.
(2022a)

U.S. (NHANES)
Cross-sectional

Urine

COPD

No effects reported.

Rahman et al. (2022c)

U.S. (NHANES)
Cross-sectional

Urine

Emphysema

No effects reported.

This document is a draft for review purposes only and does not constitute Agency policy.

D-28	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Rahman et al.
(2022d)

U.S. (NHANES)
Cross-sectional

Urine

Emphysema

No effects reported.

Rahman et al.
(2022b)

U.S. (NHANES)
Cross-sectional

Urine

Chronic bronchitis

No effects reported.

Richardson et al.
(2021)

Occupational
North

America/Europe
Cohort

Occupational

Noncancer disease of the
respiratory system
(mortality)

Significant association.

Shumate et al. (2017)

Occupational
U.S.

Cross-sectional

Occupational

Pulmonary symptoms

Significant association.

Samson et al. (2016)

Occupational

France

Cohort

Occupational

Respiratory disease
mortality

Significant deficits in
deaths.

Denisova et al. (2018)

Russia

Cross-sectional

Lung tissue

Sarcoidosis

No effects observed.

Karakis et al. (2021)

Cohort
Israel

Urine

Asthma

No effects observed.

Kavembe-Kitenge et
al. (2020)

Occupational
DR Congo
Cross-sectional

Urine

Pulmonary function

No uranium-specific
analyses.

Kocher et al. (2016)

Occupational
United States
Cross-sectional

Occupational

Pneumoconiosis

No effects reported.

Zablotska et al.
(2013)

Occupational

Canada

cohort

Occupational

Mortality from COPD and
asthma

No associations observed.

COPD = chronic obstructive pulmonary disease.

D.14. URINARY EFFECTS

1	ATSDR Summary

2	ATSDR determined there was sufficient information from experimental studies to conclude

3	that uranium is a kidney toxicant ATSDR 2013 reviewed acute and subchronic exposure toxicity

This document is a draft for review purposes only and does not constitute Agency policy.

D-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

Protocol for the Uranium IRIS Assessment (Oral)

studies that report increased incidence of histological effects and alterations in urinary markers of
renal damage in rats, mice, dogs, and rabbits.

Newly Identified Human Studies

Twelve (n = 12) epidemiological studies meeting PECO criteria were identified in the IRIS
literature search for metabolic outcomes (see Table D-18). Some studies observed an association
between uranium exposure and kidney disease (Park and An. 20221: a deficit in some of the
measured kidney filtration measures fShelley et al.. 20141: and a decrease in eGFR (estimated
glomerular filtration rate) (Wu etal.. 2018b). A number of studies had potential limitations,
including selection bias and exposure assessment concerns.

Newly Identified Animal Studies

Eighteen animal toxicity studies (14 studies using rats and 4 studies using mice) were
identified in the date-limited literature search. Outcomes considered in these studies include organ
weights, macroscopic appearance, histopathology, and markers of renal disease. In SD rats,
subchronic and chronic exposure to uranyl nitrate resulted in altered urinary flow and renal
vascular resistance, kidney weight, and markers of renal disease (see Table D-19). The remaining
studies report no effects on kidney weight, histopathology, macroscopic appearance, or markers of
renal disease in exposed SD rats. However, most of the available studies exposed SD rats to uranium
concentrations (40 mg/L) known to be non-toxic to the urinary system (Gueguenetal.. 2007:
Tissandie etal.. 2007: Souidi etal.. 2005). In C57BL/6J and Kunming mice uranium exposure did
not affect markers of renal disease, and in ApoE null mice there were no treatment-related effects
on macroscopic appearance of the kidney or markers of renal disease.

Conclusion

The available toxicological and epidemiological studies identified in the literature search
update provide further support of the studies and evidence considered by ATSDR 2013. Based on
these findings, EPA will evaluate the available evidence (studies identified in the IRIS literature
search and studies cited in ATSDR 2013) for dose-response analysis on uranium-induced
urinary effects.

This document is a draft for review purposes only and does not constitute Agency policy.

D-30	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

1	Units of Analysis

2	Humans: Kidney disease, markers of kidney function.

3	Animals: Urinary and serum markers of renal disease/function, organ weights, organ

4	morphology/histopathology.

Table D-18. Studies of urinary endpoints in humans identified 2011-2022

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Nanavakkara
etal. (2019)

Sri Lanka
Case-control

Urine, hair,

drinking

water

Chronic kidney disease

No effects reported.

Okaneku et
al. (2015)

U.S.

(NHANES)

Cross-

sectional

Urine

Renal function markers

No effects reported.

Park and An
(2022)

U.S.

(NHANES)

Cross-

sectional

Urine

Kidney disease

Significant association.

Rango et al.
(2015)

Sri Lanka

Cross-

sectional

Urine

Chronic kidney disease

No effects reported.

Shellev et al.
(2014)

Occupational

Cross-

sectional

Urine

Kidney function markers

Significant negative association.

Weaver et
al. (2014)

Mexico

Cross-
sectional

Urine

eGFR measures

No significant findings.

Wu et al.
(2018b)

China

Cross-
sectional

Urine

eGFR measures

Significant negative association.

Oruc et al.
(2022)

Turkey
Case-control

Blood

Trace element status in
hemodialysis patients

No effects observed.

Butler-
Dawson et
al. (2021)

Occupational

Guatemala

cohort

Urine

Increase in creatinine as a
marker of kidney injury

No effects observed.



Occupational

Occupational

Renal disease mortality

Significant deficits in deaths.

Samson et
al. (2016)

France
Cohort







This document is a draft for review purposes only and does not constitute Agency policy.

D-31	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Study design

Exposure
measurement

Endpoints

Author-reported findings

Yang et al.
(2019)

China
Cross-
sectional

Urine, blood

eGFR

No effects reported.

Zablotska et
al. (2013)

Occupational

Canada

cohort

Occupational

Mortality from nephritis and
nephrosis

No effects reported.

eGFR = estimated glomerular filtration rate.

Table D-19. Summary of toxicological studies reporting on uranium-induced
urinary effects

Reference

Experimental design

Author-reported findings

Rat studies

Rouas et al. (2011)

Male SD rats exposed to uranyl nitrate
(0, 40 mg/L in drinking water) for 9 mo

No effects on histopathology or histological
markers of renal disease

Wade-Gueve et al.
(2012)

Decreased serum creatinine, no effect on other
markers of renal disease.

Grison et al. (2013)

Increased relative (but not absolute) kidney
weight, plasma creatinine, and urinary potassium
and sodium.

Grison et al. (2019)

No effects on plasma or urine markers of renal
damage

Dublineau et al.
(2014)

Male SD rats exposed to uranyl nitrate
(0.009, 0.09, 0.23, 0.45, 0.9, 1.8,
5.4 mg/kg-d in drinking water) for 9 mo

No macroscopic or organ weight changes, or
effects on markers of renal disease.

Grison et al. (2016)

Male and female SD rats exposed to
uranyl nitrate (0, 0.015, 0. 15,1.5,
40 mg/L in drinking water) for 9 mo

Decreased kidney weight and urine volume.
Decreased urine calcium concentration, protein
levels, and urea concentration.

Poisson et al.
(2014a)

Male SD rats exposed to uranyl nitrate
(0, 40 mg/L in drinking water) for 90 d

No effects on plasma markers of renal disease.

Legendre et al.
(2016)

Male SD rats exposed to uranyl nitrate
(0, 40,120 mg/L in drinking water) from
GD 1 to PND 168

No effects on kidney weight or plasma markers of
renal disease.

Souidi et al. (2018)

Male SD rats exposed to natural
uranium (0, 40,120 mg/L in drinking
water) for 9 mo

Decreased serum urea at low dose and decreased
creatinine at high dose.

Grison et al. (2018)

Male and female F0 generation SD rats
exposed to uranyl nitrate (0, 40 mg/L in
drinking water) for 9 mo

F0 and F1 generation: no effects on kidney weight
or markers of renal disease.

This document is a draft for review purposes only and does not constitute Agency policy.

D-32	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

Reference

Experimental design

Author-reported findings





F2 generation: decreased kidney weight in males.
No effect on markers of renal disease

Lu etal. (2021)

Male SD rats exposed to depleted
uranium (0, 3, 30, 300 mg/kg in food)
for 6 or 12 mo

No effects on kidney weights or plasma markers
of renal disease.

Vicente-Vicente et
al. (2013)

Male SD rats exposed to uranyl nitrate
(0, 5.4 g/L in drinking water) for 11 or
21 wk

11 wk: decreased urinary flow. No change in
plasma creatinine, plasma urea, proteinuria, in
glucosuria.

21 wk: decreased urinary flow and increased renal
vascular resistance. No change in renal blood
flow, plasma.

Gueguen et al.
(2014)

Male SD rats exposed to uranyl nitrate
(0, 40 mg/L in drinking water) for 1-
18 mo

No effects on plasma markers of renal disease,
organ weights, or histopathology.

Male SD rats exposed to uranyl nitrate
(0, 0, 0.2, 2, 5, 10, 20, 40, 120 mg/L in
drinking water) for 9 mo

Poisson et al.
(2014b)

Male SD rats exposed to uranyl nitrate
(0, 40,120, 400 mg/L in drinking water)
for 3 mo

No effects on kidney histopathology or urinary or
plasma markers of renal disease.

Male SD rats exposed to uranyl nitrate
(0, 40,120, 600 mg/L in drinking water)
for 9 mo

Mouse studies

Bolt et al. (2019)

Male & female C57BL/6J mice exposed
to uranyl acetate (0, 5, 50 mg/L in
drinking water) for 60 d

No effects on plasma markers of renal disease.

Haoetal. (2013b)

Male Kunming mice exposed to uranyl
nitrate (0, 0.4, 4, 40 mg/kg-d in food) for
4 mo

No effects on plasma markers of renal disease.

Lestaevel et al.
(2014)

Male ApoE null mice exposed to uranyl
nitrate (0, 20 mg/L) for 14 wk

No effect on plasma markers of renal disease

Souidi et al. (2012)

Male ApoE null mice exposed to uranyl
nitrate (0, 20 mg/L in drinking water) for
3 mo

No effect on macroscopic appearance or plasma
markers of renal disease.

This document is a draft for review purposes only and does not constitute Agency policy.

D-33	DRAFT-DO NOT CITE OR QUOTE


-------
Protocol for the Uranium IRIS Assessment (Oral)

D.15. OTHER EFFECTS

1	EPA also evaluated other outcomes notcaptured in ATSDR 2013 thatwere identified in the

2	IRIS literature search.

3	Newly Identified Human Studies

4	Kim etal. (20191 measured oxidative stress; Shiue (20131 examined vision, hearing, and

5	balance; Bai etal. (20221 examined optic chiasm; Strand etal. (20141 examined all-cause mortality;

6	Shiue f20151 measured self-rated health; Bouetetal. f20181 examined all causes of death (cancer

7	and noncancer); and Lewicka et al. f20191 examined prepregnancy BMI.

8	Newly Identified Animal Studies

9	The were no new animal toxicity studies that evaluated outcomes not already considered in

10	ATSDR 2013.

11

This document is a draft for review purposes only and does not constitute Agency policy.

D-34	DRAFT-DO NOT CITE OR QUOTE


-------