EPA/600/R-18/224 | February 2018
www.epa.gov/homeland-security-research
United States
Environmental Protection
Agency
oEPA
Standing Operating Procedure (SOP)
for the Development of Provisional
Advisory Levels (PALs) for Hazardous
Chemicals
Office of Research and Development
Homeland Security Research Program
-------
This page left intentionally blank
-------
EPA/600/R-18/224
February 2018
Standing Operating Procedure (SOP)
for the Development of
Provisional Advisory Levels (PALs)
for Hazardous Chemicals
Final
by
John Lipscomb, Ph.D.
Threat and Consequence Assessment Division
National Homeland Security Research Center
Cincinnati, OH 45268
IAGNo. 1824- S870-T1
IAGNo. DW-89-92241401
EP-W-12-026
U.S. Environmental Protection Agency Project Officer
Office of Research and Development
Homeland Security Research Program
Cincinnati, Ohio 45268
-------
DISCLAIMER
The U.S. Environmental Protection Agency, through its Office of Research and
Development, funded and managed the Provisional Advisory Levels project through two
Interagency Agreements (IAGs): IAGNo. 1824- S870-T1 with the U.S. Department of Energy
(Oak Ridge National Laboratory; ORNL) and IAG No. DW-89-92241401 with the U.S.
Environmental Protection Agency. ORNL is managed by UT-Battelle, LLC for the U.S.
Department of Energy under contract DE-AC05-00OR22725. ORNL empaneled and consulted
subject matter experts and was responsible for developing key work products, including the initial
draft of this Standing Operating Procedure (SOP), which ORNL used to guide the development of
Provisional Advisory Levels. An overview of the ORNL SOP document has been published as
Young et al. (Inhalation Toxicology Volume 21, Supplement 3:1-11, 2009). Additional support
was provided by CSS-Dynamac Corporation, through Contract EP-W-12-026.
This document has been subjected to the Agency's review and has been approved for
publication. Note that approval does not signify that the contents necessarily reflect the views of
the Agency. Mention of trade names or commercial products in this document or in the methods
referenced in this document does not constitute endorsement or recommendation for use. The EPA
does not endorse any commercial products, services, or enterprises.
The contractor role did not include establishing Agency policy.
Questions concerning this document or its application should be addressed to:
John C. Lipscomb, PhD, DABT, Fellow ATS
National Homeland Security Research Center
Office of Research and Development
U.S. Environmental Protection Agency, Cincinnati, OH
513-569-7217
-------
Table of Contents
DISCLAIMER ii
ABBREVIATIONS v
ACKNOWLEDGMENTS vi
1 Introduction 1
1 • 1 History/Proj ect Context 1
1.2 General Scope and Application of PALs Program 3
1-3 The Technical Support Document 6
2 Definitions of PALs, PAL Tiers, and PAL Dosing or Exposure Durations 6
2.1 PALs 6
2.2 PAL Exposure Durations 7
3 PALs and Other Standards/Guidelines 8
4 Data supporting PAL Development 8
4.1 Literature Search 8
4.2 Study Evaluation and Selection 9
4.3 Human data 11
5 Derivation of PAL Values 12
5.1 Evaluation of effects 13
5.1.1 Mechanistic considerations 14
5.2 Identification of critical effect 15
5.2.1 Severity of Effects 16
5 • 3 Identification of the Point of Departure (POD) 17
5.3.1 The Threshold Concept 18
5.3.2 PAL-1 Point of Departure 20
5.3.3 PAL-2 Point of Departure 20
5.3.4 PAL-3 Point of Departure 21
5.3.5 Duration Adjustment of the Point of Departure 22
5.4 Uncertainty and Modifying Factors 25
5.4.1 Interspecies dosimetry adjustment uncertainty factor (UFa) 26
5.4.2 Intraspecies variability uncertainty factor (UFh) 28
5.4.3 Temporal Extrapolation Uncertainty Factor (UFt) 29
5.4.4 Severity Adjustment Uncertainty Factor (UFs) 30
5.4.5 Modifying Factor (MF) 31
5.4.6 Multiplication of Uncertainty Factors 31
5.5 Computation of PAL Values 32
5.6 Weight-of-Evidence and Confidence 32
5.7 Presentation of Values 32
5.7.1 Level of Precision 33
iii
-------
5.7.2 No PAL Value Established 33
5.7.3 Values for Susceptible Population Groups 33
5.8 Surrogate Chemicals, Surrogate Data, Surrogate PAL Values 33
5.9 Carcinogenicity as a Critical Effect 34
5.10 Final Adjustment of PAL Values 35
6 Chemical Considerations 36
6.1 Complex Mixtures/Concurrent Exposure Issues 36
6.2 Degradation Products 36
7 Research Needs/Recommendations 36
8 References 37
List of Tables
Table 1. Variability in Response of Asthmatics to Irritant Gases 29
List of Figures
Figure 1. Exposure standards and guideline values 3
iv
-------
ABBREVIATIONS
AD J adjusted
AEGLs Acute Exposure Guideline Levels
ATSDR Agency for Toxic Substances and Disease Registry
BMC benchmark concentration
BMCL lower confidence limit of the BMC
BMD benchmark dose
BMDL lower confidence limit of the BMD
CAS RN® Chemical Abstracts Service Registry Number
CEELs Community Emergency Exposure Levels for Hazardous Substances
CFR Code of Federal Regulations
DOE Department of Energy
DTIC Defense Technical Information Center
DWE Drinking Water Equivalent
ERPG Emergency Response Planning Guidelines™
HA Health Advisory
HSDB Hazardous Substances Data Bank
IDLH immediately dangerous to life or health
IRIS Integrated Risk Information System
LC50 concentration lethal to 50% of the exposed group
LD50 dose lethal to 50% of the exposed group
LOAEL lowest observed adverse effect level
LOEL lowest observed effect level
MF modifying factor
MRL Minimal Risk Level
NAC National Advisory Committee
NHSRC National Homeland Security Research Center (U.S. EPA)
NIOSH National Institute for Occupational Safety and Health
NOAEC no observed adverse effect concentration
NOAEL no observed adverse effect level
NOEL no observed effect level
NR Not Recommended
NRC National Research Council
NTIS National Technical Information Service
OECD Organization for Economic Co-operation and Development
PAL Provisional Advisory Level
v
-------
PBPK physiologically-based pharmacokinetic
PELs permissible exposure limi
POD point-of-departure
p-RfC Provisional Inhalation Reference Concentration
p-RfD Provisional Oral Reference Dose
QSAR quantitative structure activity relationship
RELs recommended exposure limits
RfC Inhalation Reference Concentration
RfD Oral Reference Dose
SOP Standing Operating Procedure
TD Toxicodynamic(s)
TEEL Temporary Emergency Exposure Limit
TK Toxicokinetic(s)
TSCATS Toxic Substances Control Act Submissions
TSD Technical Support Document
U.S. EPA U.S. Environmental Protection Agency
UF uncertainty factor
UFh intraspecies uncertainty factor
UFl LOAEL to NOAEL uncertainty factor
UFs severity adjustment uncertainty factor
UFt temporal extrapolation uncertainty factor
vi
-------
ACKNOWLEDGMENTS
The National Homeland Security Research Center is grateful to the staff and managers of
Argonne National Laboratory and the Oak Ridge National Laboratory for dedicated and effective
work, and to the members of the ONRL-assembled Expert Consultation Panel. Additional expert
review of this document was provided by NHSRC's Cross-Agency Provisional Advisory Levels
Steering Committee, comprising U.S. Environmental Protection Agency staff. Their guidance
regarding technical approaches and strategic direction has been directly responsible for substantial
improvements in these products. Technical and programmatic-minded scientists within and outside
the Agency have provided valuable and far-reaching advice as Internal Technical Reviewers and as
External Peer Reviewers. Their contributions are also gratefully acknowledged.
Members of the ORNL's Expert Consultation Panel include:
Loren Roller, DVM, PhD, Chair David Dorman, DVM, PhD, DABT, DAVBT
Donald Gardner, PhD Gerry Henningsen, DVM, PhD, DABT, DAVBT
Mark McClanahan, PhD Ernest McConnell, DVM, MS (path), DACVP, DABT
Patricia McGinnis, PhD, DABT Coleen Baird, MPH, MD
Members of the U.S. Environmental Protection Agency's Provisional Advisory
Levels Steering Committee include:
Michele Burgess, Office of Land and Emergency Management
Janine Dinan, Office of Land and Emergency Management
Stiven Foster, Office of Land and Emergency Management
Kathleen Raffaele, Office of Land and Emergency Management
Kevin Garrahan, Office of Research and Development
Annette Gatchett, Office of Research and Development
Tonya Nichols, Office of Research and Development
Emily Snyder, Office of Research and Development
Sarah Taft, Office of Research and Development
George Woodall, Office of Research and Development
Steven Kueberuwa, Office of Water
Emily Rogers, Office of Water
Deborah McKean, Region 8
Internal Technical Reviewers:
Kathleen Raffaele
Erin Silvestri
External Peer Reviewers:
Finis L. Cavender
Rogene F. Henderson
Andrew G. Salmon
vii
-------
This page left intentionally blank
-------
1 Introduction
1.1 History/Project Context
Much of the nation's critical infrastructure can be subject to threats and intentional chemical,
biological, or radiological attacks. Additionally, there is always the potential for releases of
hazardous materials from accidents or natural disasters. The U.S. Environmental Protection
Agency (U.S. EPA) aims to protect the public from the potential consequences of such incidents.
Following the tragic results of the accidental release of methyl isocyanate in Bhopal, India, in
1984, increased attention was focused on the issue of anticipating accidental releases and
preparing for such actions. In 1986, Congress passed the Superfund Amendments and
Reauthorization Act (SARA), which contains provisions often referred to as the Emergency
Planning and Community Right to Know Act. In response, the U.S. EPA developed a list of
approximately 400 Extremely Hazardous Substances on the basis of their acute lethality
information (U.S. EPA, 1987). Soon thereafter, the American Industrial Hygiene Association
initiated its Emergency Response Planning Guidelines™ (ERPG) Committee, which began to
develop one-hour inhalation exposure recommendations structured in a three-tiered system of
severity addressing effects in a once in a lifetime exposure described as ranging from mild and
transient effects, through serious or irreversible effects, to lethality (Rusch, 1993).
In 1991, the U.S. EPA and the Agency for Toxic Substances and Disease Registry (ATSDR)
requested that the National Research Council recommend guidelines for developing emergency
exposure guidelines for exposures to hazardous substances. The National Research Council
(NRC) released Guidelines for Developing Community Emergency Exposure Levels for
Hazardous Substances (CEELs; NRC, 1993). The CEELs represent a tiered system of inhalation
risk values targeted to emergency exposure durations of 1 to 8 hours, for exposures representing a
once in a lifetime occurrence. These values address exposures that may produce effects ranging
from discomfort, through those that may impair an ability to escape of produce disability, to life-
threatening effects. The CEEL values were developed with a preference for human response data,
from a relevant exposure duration.
To address the possibility of exposure to chemicals in use at Department of Energy (DOE) sites,
and in the community in general, DOE embarked on a program to develop temporary exposure
guidelines for chemicals not addressed by the CEEL or ERPG system. Early publications on the
development of Temporary Emergency Exposure Limit (TEEL) values (Craig et al., 1992, Craig
and Lux, 1998) reviewed available data and provided an overview of the development process for
TEEL values, respectively. This led ultimately to the release of the DOE Handbook on
Temporary Emergency Exposure Limits for Chemicals: Methods and Application (DOE, 2008).
The TEELs are targeted to a 15-minute inhalation exposure duration and represent a tiered
structure of risk values that are generally similar to those of ERPGs and later CEELs and Acute
Exposure Guideline Levels (AEGLs). However, CEEL values also include a category 0 value,
below which it is anticipated that no health risks will occur.
To reflect an expansion of coverage by these values to emergency planning, emergency response,
incident prevention, and Superfund site remediation, NRC commissioned a new committee to
develop Acute Exposure Guideline Levels (AEGLs; NRC, 2001). The resulting AEGL values
were developed as a tiered system of inhalation risk values, which cover exposure durations from
10 minutes to 8 hours. The AEGL values are based on the tiered system for CEELs and the
AEGL's tier-specific characterization of response was only minimally revised from that of
CEELs. The primary difference is that the AEGL's Tier 2 was revised to specify that effects
-------
would be serious and possibly irreversible, and expanded the definition to include effects that
would impair the ability to escape, compared to the characterization in CEELs guidance as
exposures producing a disability.
Recognizing that unanticipated chemical releases can also contaminate water supplies, the EPA's
Office of Water develops Health Advisory (HA) values for multiple durations of exposure
extending beyond one day. These values represent advisory, not regulatory, exposure values.
Because the HA values are based on response types and response levels similar to those used to
derive the oral reference dose (RfD), HA values approximate the same risk level. Both RfD values
and HA values are based on toxicity data for effects occurring at relevant exposure durations:
subchronic or chronic durations for RfD derivation, and one day, ten days, or lifetime (chronic)
for HA derivation. HA value are disseminated as concentrations in drinking water (Drinking
Water Equivalent Level values).
Because unanticipated chemical exposures may extend beyond 8 hours for inhaled substances and
for durations between ten days and a lifetime for orally-encountered substances, the U.S. EPA's
National Homeland Security Research Center (NHSRC) initiated the development of Provisional
Advisory Levels (PALs) as a tiered system of health-based guideline values for oral and inhalation
exposures of up to 24-hours, 30-days, and 90-days, and for other durations when necessary.
(Figure 1).
Provisional Advisory Level (PAL) values are risk-based exposure guidance values for inhalation
exposure and oral dosing. PALs have been developed for specified durations thought to represent
those relevant to emergency response and emergency management scenarios. Figure 1, below,
illustrates the relationship between PAL values and some other extant exposure standards and
guideline values described above. Inhalation exposure guideline values include Acute Exposure
Guideline Levels (AEGLs; NRC, 2001); Emergency Response Planning Guideline (ERPG; AIHA,
2016) values; and Temporary Emergency Exposure Levels (TEELs; DOE, 2008). These inhalation
exposure guideline values cover exposures up to 8 hours. Inhalation exposure risk values
(standards or legally enforceable limits) only address no-effect levels. Oral values are available
for several durations, but only address no-effect levels. Inhalation Reference Concentration (RfC;
U.S. EPA, 1994) values and Oral Reference Dose (RfD; U.S. EPA, 1993) values (which can be
used to establish enforceable standards) and Minimal Risk Levels (MRL; ATSDR, 2017), an
estimate of the daily human exposure to a hazardous substance that is likely to be without
appreciable risk of adverse non-cancer health effects over a specified duration of exposure,
address no-effect levels. Provisional Peer Reviewed Toxicity Values for Oral Reference Dose (p-
RfD; U.S. EPA, 2017)) values; and Provisional Peer Reviewed Toxicity Values for Inhalation
Reference Concentration (p-RfC; U.S. EPA, 2017) values are used to establish remediation goals
for Superfund sites, and also address no-effect levels. So, unlike these existing guideline values
and standards, PALs address inhalation exposure and oral dosing lasting longer than 8 hours and
include human health effects ranging in severity from mild to lethal.
2
-------
Acute
Short-Term
Subchronic
Chronic
(< 24 H)
(1 - 30 D)
(30 D-
-7 Yr)
(70 Yr)
Lethality
Inhalation
AEGL-3
NHSRC PROVISIONAL ADVISORY
LEVELS (PALs): Oral and Inhalation
ERPG-3*
TEEL-3*
24 H
PAL-3
30 D
PAL-3
90 D
PAL-3
&
VI
More severe,
disabling effects
Inhalation
AEGL-2
ERPG-2*
TEEL-2*
24 H
PAL-2
30 D
PAL-2
90 D
PAL-2
Irritation or mild,
reversible effects
Inhalation
AEGL-1
ERPG-1*
TEEL-1*
24 H
PAL-1
30 D
PAL-1
90 D
PAL-1
NO ADVERSE EFFECTS
"Likely to be without
appreciable risk"
Acute M RL
Health Advisory
Intermediate MRL
Subchronic p-RfD/C
Chronic MRL
Chronic p-RfD/C
Health Advisory
Chronic RfD/C
Figure 1. Exposure standards and guideline values.
1.2 General Scope and Application of PALs Program
Regulatory exposure standards have been developed for many chemicals to which human
exposure can be anticipated or controlled. However, there are many chemicals for which
regulatory exposure limits (standards) are not established, and/or for which exposure guideline
values are not available. Further, regulatory exposure standards applicable to the general human
(e.g., excluding OSHA PEL values for workers) population are seldom established for the
durations of exposure relevant to activities undertaken during and following the unanticipated
release of toxic chemicals as a result of accidental releases of chemicals, natural disasters, or
terrorist activities. These situations may involve the release and subsequent exposure of limited
or targeted segments of the general population to toxic industrial chemicals or chemical warfare
agents. Adequate responses during and following these situations require making health-
protective decisions as quickly as possible, based on the most reliable and appropriate
information available.
PALs provide guidance to emergency response planners and those making response decisions at
the federal, regional, state, and local levels by communicating the types and likelihood of health
effects that are predicted to occur when exposures exceed those considered acceptable for the
general population. Emergency situations are unique and often complex, including exposures
that may vary over short time intervals and with respect to distance from the source, and
including exposures to populations ranging from small groups of people whose composition
may be well-defined or even controlled, to larger populations more broadly representative of the
overall population and whose composition is not known. Management of emergency situations
is not governed by any statutory exposure limits, and responsibility resides with on scene
professionals who determine and direct emergency response operations and actions based on
scenario-specific information. Like guidance values developed for other routes and durations of
exposure, PALs guidance values can be used to inform (not determine) decisions for
evacuation, re-use and re-entry into affected areas for specified durations. PALs values were
3
-------
developed to extend the available suite of exposure guidance values by addressing exposure
routes, durations, and health effects not covered by other risk value systems. Like guidance
values developed for other routes and durations of dosing or exposure, PALs guidance values
can be used to inform (not determine) decisions for evacuation, re-use, and re-entry into
affected areas for specified durations.
PALs are Provisional because they represent conclusions drawn from a set of information that
was current when summarized and because the exposure recommendations should be
considered along with other pertinent information during the decision-making process. They
are Advisory in that they are not regulatory values and are non-enforceable. PAL values are
developed through risk assessment methods that are similar to methods to develop regulatory
dose and exposure limits and dose and exposure values in other risk system that can/may/have
been translated into regulatory limits and enforceable standards.
Like other inhalation guidance values represented by ERPGs, CEELs, TEELS, and AEGLs,
PALs are structured as a tiered system of values, reflecting increasing severity of response in
exposed individuals with increasing dose or concentration. The PALs severity scheme conforms
to the first principle of toxicology - that response is a function of dose. Like AEGLs, PAL
values are additionally structured by duration, with the important difference that PALs extend
beyond the 8-hour duration of AEGLs to a 24-hr continuous exposure as well as up to 30 or 90
days. This aspect addresses another principle of toxicology - that duration of dosing or
exposure also influences response.
Creating these tiers of risk values enables the identification of exposures that result in increased
risks for the expression of more severe effects. Typically, tiers represent changes from baseline
physiological parameters or mild and reversible effects; increasing to effects that are of an
increased severity and are often irreversible, or are of a nature as to impede the ability to escape
a contaminated environment; and increasing to morbidity, mortality, or life-threatening toxicity.
Of course, the identification of irreversible, more severe toxicities involves a characterization of
the responses that accompany higher oral doses or exposure concentrations. The description
and characterization of such responses allows investigators to quantify the health impact
(identify the human health condition and estimate the likelihood of its occurrence) and to
identify medical countermeasures.
PALs are similar to Health Advisory (HA) values developed by the U.S. EPA Office of Water
in that PALs and HAs are both advisory in nature, derived on the basis of toxicity or response
information from relevant dosing or exposure durations, and are developed for less-than-
lifetime exposures of defined durations. HA values "serve as the informal technical guidance
for unregulated drinking water contaminants to assist federal, state, and local officials, and
managers of public or community water systems in protecting public health as needed" (U.S.
EPA, 2012a); their derivation has been described in some detail (Donohue and Lipscomb,
2002). Similarity also extends to RfD values with respect to the derivation and extrapolation of
points of departure (POD). In these systems, it is expected that the effect characterizing the
response under evaluation is not observed at the point of departure, as demonstrated in toxicity
studies employing a dosing or exposure duration representative of the duration for which the
risk value is developed. However, the PALs risk value system differs from these risk value
systems (HA and RfD values) in the same way that PAL values are similar to the inhalation risk
value systems described above (ERPGs, CEELs, TEELS, and AEGLs) - PALs address levels of
severity greater than the no-adverse-effect level that serves as the basis for RfD and HA
derivation.
4
-------
Thus, three distinctions between PALs and RfD/ RfC values should be appreciated: the nature
of the demarcation between response levels, the extent to which effects are characterized, and
the duration of dosing or exposure covered. PALs are defined as doses or exposures above
which the risk of responses characteristic of several levels of severity is expected to increase. In
contrast, the oral Reference Dose or inhalation Reference Concentration is defined as, "an
estimate (with uncertainty spanning perhaps an order of magnitude) of a daily dosing or
exposure to the human population (including sensitive subgroups) that is likely to be without an
appreciable risk of deleterious effects during a lifetime" (Barnes and Dourson, 1988; US EPA,
1994; 2002). Operationally, RfD values are derived based on threshold doses for effects
defined as "adverse," with the point of departure characterized as aNOAEL (no observed
adverse effect level), a LOAEL (lowest observed adverse effect level), or a BMD (benchmark
dose) value. The concept of "adverse" is inherent in the description of the reference dose
relative to the "acceptable daily intake" (Barnes and Dourson, 1988), as well as reliance on a
threshold for adversity (the "A" in NOAEL, LOAEL; e.g., preference for a NOAEL over a
LOAEL as a point of departure value). Thus, oral RfD/inhalation RfC values are developed as
doses or exposures anticipated to be without an appreciable risk of deleterious noncancer
effects. In contrast, the likelihood of experiencing a given tier (severity) of effect is increased
when dosing or exposures increase above a given PAL value. An important aspect of this
circumstance is that PALs separate doses or exposures into those likely to produce three levels
of severity, and identify the type of health effect likely to occur as dosing or exposures increase.
Finally, PALs address durations of dosing or exposure representing durations pertinent to
exposures during human activities during and following unanticipated chemical releases.
Situations that may benefit from the consideration of PAL values include, but are not limited
to, transport/storage accidents, natural disasters, and subversive activities. PALs could be used
in homeland security efforts, and by public health and law enforcement agencies, emergency
response agencies, water utilities, EPA program and regional offices, and states and local
governments. Like risk values for other routes and durations of dosing or exposure, PAL
values represent a source of risk-based dosing or exposure values and describe a continuum of
chemical-specific health effects possible with increasing inhalation concentrations or oral
doses and increasing durations of exposure. This information may be useful to emergency
planners as parameters for inclusion in the development of scenarios for potential events,
where they may help inform the type and likelihood of injury and thereby support estimates of
the consequences of an unanticipated release. The results can aid in decisions regarding
medical countermeasures and evacuation requirements.
However, the primary application of PALs is to represent a source of information to guide
decisions for reentry, resumed use, and protective action for the prescribed durations of
exposure. Reentry is defined as the entry of persons into an affected area following a release.
Resumed use refers to the reutilization of items and infrastructure impacted by a chemical
release. Protective action is an action or measure taken to avoid or reduce dosing or exposure to
a given hazard. In the event of a release of a chemical, recommendations for protective action
could include shelter-in-place, use of personal protective equipment, evacuation, etc.
Additionally, consideration must be given to selective use/reuse of resources; e.g., community
water supplies not considered potable may be suitable for other uses.
This standing operating procedure (SOP) is intended for use by toxicologists and health risk
assessors who are familiar with basic toxicological principles and health risk assessment
methods.
5
-------
1.3 The Technical Support Document
One Technical Support Document (TSD) should be developed per chemical or group of closely
related chemicals (e.g., G nerve agents) for which a set of PAL values is determined. The TSD
may contain introductory material such as that describing the history of emergency exposure value
development, the intent of PAL values relative to the emergency preparedness and homeland
security missions, and other programmatic information. An Executive Summary section should
present information specific to the chemical sufficient to guide an understanding of the importance
of oral dosing and inhalation exposure, primary manifestations of toxicity that may result from
dosing or exposure, and the PAL values for the chemical. The remainder of the document should
be organized into sections addressing the areas presented in the bulleted list below. These areas
should be developed at a level of detail to convey their significance to the development of PAL
values. The primary purpose of the TSD is to present information sufficient to provide an
understanding of the development of a range of effects with respect to both dosing and exposure
magnitude and duration, and to demonstrate a reliable estimation of the risk of those effects based
on the available data, interpreted in accord with this SOP. The following list of topics should be
given attention in the Technical Summary Document:
o Chemical identification
o Chemical/physical properties
o Environmental fate (air and water)
o Absorption, distribution, metabolism, elimination (ADME)
o Toxicokinetics/Toxicodynamics and Mechanism(s) of toxicity
o Susceptibility
o Human toxicity data
o Animal toxicity data
o Recommended PAL values and rationales
o Other guideline values
o Data gaps and needed additional research
o References
2 Definitions of PALs, PAL Tiers, and PAL Dosing or Exposure Durations
2.1 PALs
PALs are advisory levels relevant to concerns regarding possible dosing or exposure of the
general public in emergency situations. As described above (Figure 1) PALs are developed to
describe a tiered set of responses based upon health-based criteria appropriate for each health
effect tier for a range of durations. Three tiers (PAL-1, PAL-2, and PAL-3), distinguished by the
severity of toxic effects and are the same as those developed for AEGLs. The durations for PAL
estimates should be based on emergency response needs. Recommended durations include up to
24 hours, up to 30 days, up to 90 days, for oral dosing or inhalation exposure, but the application
of PALs methods to derive values for other durations may be justified based on emerging needs.
There is no requirement to develop values for specific durations. The health effect for a given
chemical (and route, duration and tier) is the biological effect identified by a specific data set and
for which its relationship to oral dose or inhalation exposure has been defined. Although PALs
6
-------
are developed with considerable attention to sensitive populations (e.g., asthmatics, age-dependent
sensitivities), PALs are not intended to protect hypersensitive individuals or against idiosyncratic
responses.
Definition of PAL Tiers:
PAL-1 represents the assumed, duration-specific continuous dosing level or exposure
concentration of a chemical above which changes from a baseline of specific biomarkers or
physiological responses could have adverse health effects in the general population.
Concentrations at or below PAL-1 are not expected to be associated with adverse health
effects. Increasingly greater concentrations above the PAL-1 value could cause progressively
harmful effects in the general population, including all age groups and sensitive
subpopulations.
PAL-2 represents the assumed, duration-specific continuous dosing level or exposure
concentration of a chemical above which serious, possibly irreversible, or escape-impairing
effects could result. Increasingly greater concentrations above the PAL-2 value could cause
progressively harmful effects in the general population, including all age groups and sensitive
subpopulations.
PAL-3 represents the assumed, duration-specific continuous dosing level or exposure
concentration of a chemical above which lethality in the general population, including all age
groups and sensitive subpopulations, could occur.
2.2 PAL Exposure Durations
The three PAL tiers (PAL-1, PAL-2, and PAL-3), distinguished by the degree of severity of toxic
effects, may be developed for dosing or exposure durations of up to 24-hours, up to 30-days, up to
90-days, or other durations deemed relevant via oral and inhalation routes. Information previously
compiled in other risk value systems should be consulted. For example, for inhalation exposure of
less than 24 hours to chemicals, additional consultation of other risk values such as Acute
Exposure Guideline Levels (AEGLs) or Emergency Response Planning Guideline (ERPG) values
is recommended. AEGL values were developed by the National Advisory Committee for the
Development of Acute Exposure Guideline Levels for Hazardous Substances (NAC/AEGL
Committee) for many chemicals and are based upon a similar three-tier approach with careful
attention being given to the concentration-time relationship for the exposure durations of concern
(10 minutes, 30 minutes, and 1, 4, and 8 hours). ERPG values are developed by the Emergency
Response Planning Committee of the American Industrial Hygiene Association using a similar
three-tiered approach for chemicals in anticipation of a one-hour, once in a lifetime exposure. For
oral exposures of 24 hours or less, information contained in EPA's Health Advisory and Human
Health Benchmarks for Pesticides systems, as well as in ATSDR's Acute MRL values should be
evaluated. Information in other risk systems pertinent to other PALs durations should also be
considered (e.g., information in subchronic and chronic reference value derivations for longer-
term PALs durations). The 30-day PAL will be applicable to durations of greater than 1 day to 30
days. The 90-day PAL will be applicable to durations greater than 30 days to 90 days. Values will
be developed only if data describing the development of a tier-appropriate health effect from a
duration-relevant experimental dosing or exposure duration is available. The relevance of an
experimental dosing or exposure duration to the targeted PAL duration will be justified on the
basis of duration-response information for the chemical, for the lesion(s) observed following the
dosing or exposure to other chemicals, or on the basis of professional judgement. Additional
information on temporal aspects of dosimetry is contained in sections 5.3.5 and 5.4.3.
7
-------
3 PALs and Other Standards/Guidelines
While PAL values are uniquely structured to fill important gaps in available standards or
guidelines for dosing or exposure values, information contained in some existing
standards/guidelines may be useful in determining PALs. These standards/guidelines may
include, but are not necessarily limited to, Drinking Water Health Advisories, p-RfCs and p-
RfDs, RfCs and RfDs, NIOSH [National Institute for Occupational Safety and Health] RELs
[recommended exposure limits], NIOSH immediately dangerous to life or health (IDLH) values,
ATSDR MRLs (minimal risk levels), ERPGs, AEGLs, and military standards/guidelines.
However, consideration must be given to the extent to which these standards and guidelines have
been developed for broad or specific human populations, and the extent to which they are based
on an intent (or mandate) to provide an extent of health protection as complete as possible.
Existing standard or guideline values or points of departure identified for their derivation may be
used in the estimation of a PAL value, however, it is imperative to provide a rationale for their
use. Standard or guideline values derived for relevant exposure durations, and based on relevant
health endpoints will be the most valuable for consideration. Summary or support documents for
other standard or guideline values might also contain data or information pertinent to PAL value
determination (e.g., descriptions of more severe health effects occurring at shorter durations of
dosing or exposure than those pertinent to the particular standard or guideline value being
supported). Any quantitative adjustments and assumptions regarding the use of existing standards
and guidelines should be clearly described on a case-by-case basis.
4 Data supporting PAL Development
4.1 Literature Search
Reliable literature describing key aspects related to chemical oral dosing or inhalation exposure
and toxicity should be evaluated, whether available in the open and peer-reviewed literature, in
reports from contractors or the government, or study summaries as available in some regulatory
files. The focus of this task is on identifying reports describing toxic effects of the chemical, but
should also include identification of important sources describing: chemical/physical properties,
environmental fate (air and water), absorption, distribution, metabolism, elimination (ADME),
toxicokinetics (TK) / toxicodynamics (TD) and mechanism of toxicity, and human susceptibility.
The goal of this task is to produce a comprehensive search of the literature for the researcher to
identify and review. The specific approach is determined by the amount of information obtained
on the chemical of interest during the initial search. The initial search should be based on a
strategy including broad terms such as "toxicity," chemical name, synonyms, and CAS RN®
should be employed, and revised to include more terms descriptive of the effects known to be
associated with the chemical, as necessary. When necessary, additional search iterations may be
conducted based on additional information for the chemical including search terms more
descriptive of the organs, tissues, processes, or systems affected. Ultimately, this process should
identify and prioritize for retrieval studies that demonstrate a dose- or exposure-dependent effect
of the chemical on living mammals exposed via the oral or inhalation routes.
In addition to sources of information that may be chemical- or program-specific, primary
databases include TOXNET®, [TOXNET is a collection of six databases - HSDB, CHEM-ID,
TOXLINE, CCRIS, DART, and GENETOX ] TOXLINE®, Hazardous Substances Data Bank
(HSDB), PubMed®, Web of Science®, National Technical Information Service (NTIS),
-------
Integrated Risk Information System (IRIS), Defense Technical Information Center (DTIC),
NIOSHTIC-2, and the U.S. EPA Toxic Substances Control Act Submissions (TSCATS).
In some cases, reports from the non-peer-reviewed literature may be identified (e.g., industry
internal reports, industry submissions to EPA offices, military research with limited
distribution restrictions, or unpublished reports from other sources), and may present
potentially valuable quantitative (i.e., dose or exposure response) information. In these
instances, the reliability of the reports with respect to characterizing the dose response and
identifying a point of departure for potential critical effects will be evaluated as part of the peer
review for the respective Technical Summary Document. Section 4.2 provides additional
guidance for study evaluation. For some chemicals (e.g., nerve agents, riot-control agents,
lacrimators, and sternutators), data come from military sources. Only unclassified, non-
confidential literature will be incorporated into PAL documentation, which then allows for the
PAL documentation to be unclassified as well. The use of limited distribution information (a
separate issue from classification) will be determined on a case by case basis. Some sources for
such data include the Chemical, Biological, Radiological, and Nuclear Defense Information
Analysis Center's on-line files, and reports from the U.S. Army's Edgewood Chemical
Biological Center.
4.2 Study Evaluation and Selection
Many toxicity studies used as the basis for risk value development are not conducted according
to regulatory guidelines or protocols for endpoint-specific evaluations. Their structure may
demonstrate a basis on protocols developed to support different types of decisions or generate
data useful for another purpose, they may employ different quantitative methods or statistical
models than those prescribed elsewhere, and may communicate conclusions based on the
information available to the authors at the time. Consequently, study reports should not be
judged solely on the basis of currently accepted experimental protocol criteria; neither should
the conclusions reached by study authors be accepted without additional consideration of more
recently available information. A study may be valuable to the derivation of PALs if the study
uses scientifically valid methods or provides sufficient details to determine its quality or
deficiencies, contains adequate and reliable data, or provides data-driven conclusions
applicable to PAL development.
Studies should be evaluated for five general principles, as described by the U.S. EPA (2003,
2012b): When evaluating the quality and relevance of scientific and technical information,
Agency considerations may be characterized by five general assessment factors:
• Soundness - The extent to which the scientific and technical procedures, measures,
methods, or models employed to generate the information are reasonable for, and
consistent with, the intended application.
• Applicability and Utility - The extent to which the information is relevant for the
Agency's intended use.
• Clarity and Completeness - The degree of clarity and completeness with which the
data, assumptions, methods, quality assurance, sponsoring organizations, and analyses
employed to generate the information are documented.
• Uncertainty and Variability - The extent to which the variability and uncertainty
(quantitative and qualitative) in the information or in the procedures, measures,
methods, or models are evaluated and characterized.
• Evaluation and Review - The extent of independent verification, validation, and peer
review of the procedures, measures, methods, or models.
9
-------
The list of guidelines for study evaluation should be established on science-based
methodologies, but not be so restrictive that it removes professional judgment. Some bioassay
guidelines provide a basis for selection of a robust list of study elements that, in concert with
professional experience and scientific judgment, are used to qualify the data that support the
development of PALs. For example, the NRC (1993), the Organization for Economic Co-
operation and Development's OECD Guidelines for Testing of Chemicals (OECD, 2017), and
EPA Health Effects Testing Guidelines found in the Code of Federal Regulations (40 CFR 798)
provide a basis for selection.
Evaluation of specific reports and pertinent data is critical. Similar to the process for
acceptance of manuscripts in peer-reviewed journals, references of interest to PAL
development are subjected to critical review and analysis. The following elements for
evaluating reports (adapted from NRC, 2001) provide guidance and are not intended to be
prescriptive.
1. Route of exposure.
a. Inhalation routes are preferred for inhalation PALs.
b. Preference is given to oral dosing studies in which the test article is dispersed in
the drinking water in developing oral PALs. . . Experiments utilizing gavage or
dietary administration will also be considered. When available, studies using
water as the vehicle are preferred.
c. Intravenous administration represents a worst-case scenario for bioavailability as it
assumes 100% absorption. For some chemicals, especially therapeutic agents,
findings from intravenous administration may be the most robust datasets
available. Inclusion and/or reliance on results from i.v. studies should be justified
on a case-by-case basis.
2. Exposure inhalation concentration or oral dose. Factors to be considered will
include, but not be limited to, nominal vs. target vs. analytical concentration vs. average
concentration, degradation of the test article, exposure/dose range and intervals, the
critical effect associated with the exposure/dose range being studied, etc.
3. Dosing or exposure duration. The experimental dosing or exposure regimen (e.g., #
doses/day or # hr/day and # day/week) should be provided. Professional judgment will
be used to assess relevance of the dosing or exposure duration to development of a
specific PAL. Attention will be given to factors such as mode of action, latency,
reversibility of effects, and metabolism. Dosing or exposure durations closely bounding
the PAL-specific duration will be preferentially examined; e.g., a 2-week or 4-week
experimental dosing or exposure duration would be considered more appropriate for
Short-Term PAL development than would a study having a 48-hour or 10-week dosing
or exposure duration.
4. Analytical methods. The procedures used to determine chamber concentration for
inhalation exposures should be specified or described. For oral dosing, dose may be
determined from the amount of test chemical placed into the vehicle (preferably
drinking water).
5. Number of subjects. For acute studies, 5-10 rodents/sex/test group is generally
acceptable, but as few as 2-3 primates or dogs/test group may be acceptable depending
upon the severity of the effects observed, variability within test groups, and by the
amount of change or intensity of the effect that is considered to be of biological
significance. Smaller group sizes may be acceptable if the number of treatment groups is
increased sufficiently. For subchronic and chronic studies, test groups should contain ten
or more animals of two different species (possibly fewer for primates).
6. Species studied. Humans are most relevant. Historically used experimental species
10
-------
including rats, mice, rabbits, guinea pigs, ferrets, dogs, or nonhuman primates are
acceptable as surrogates for humans. Other species require evaluation on a case-by-case
basis. It is important that the species used have an historical control history and exhibit a
toxic effect that can be established as relevant to humans. The PALs assessment
protocol will be consistent with human research as described in U.S. EPA (2006a).
7. Control groups. There should be a concurrent control group that is treated identically
except for dosing or exposure to the test article and/or vehicle for the test article.
8. Inhalation concentration or oral dose range. Regimens should be selected to
allow for a clear dose-response or concentration-response relationship.
9. Observation period. Duration of the observation period is variable based upon the onset
of the effect and whether or not there is prior knowledge regarding the nature of the
effect expected or the continuum of the toxic response. For rapidly occurring effects (i.e.,
minutes to several hours) with rapid recovery, observation periods of 3-4 days are often
sufficient. For latent and persistent effects (i.e., days to weeks or longer), a minimum
observation period of 14 days is recommended. For some effects, such as pulmonary
damage, assessment at months following the initial effect may be necessary.
10. Signs of toxicity. The nature and frequency thereof should be noted during and
after dosing or exposure. Relationship of effects to gender and concentration/dose
should be recorded.
11. Animal studies should provide body and, where appropriate, organ weights for
critical time points in the study.
12. For an ideal repeated inhalation or oral study, a NOAEL for the endpoint of concern
should be established. Data pertaining to biomarkers of dosing or exposure or pre-
dosing or exposure baseline values (e.g., cholinesterase activity levels) may also be
important.
13. At least 3 dose levels or exposure concentrations should be used to establish a dose-
response or concentration-response relationship.
14. Identifying both a NOAEL and LOAEL for observed effects level strengthens the
confidence in the study.
15. Record of time of death if it occurred and note any significant/consistent latency
periods.
16. For animal studies, necropsy should be conducted with at least gross effects noted.
17. Histopathological changes, clinical chemistry, and hematology data are very useful
and may allow reduction of uncertainty factor(s).
18. Stop-exposure studies. Use of a recovery group is useful for determining the timing and
degree of reversibility.
19. Statistical analysis. Appropriate statistical analysis of data including levels of
significance is required.
4.3 Human data
The use of human data in the development of a dosing or exposure POD for health risk
assessment avoids the uncertainties inherent in the extrapolation of animal data to the human
response. Additionally, human data sets may provide insight into individual variability and
sensitivity in the toxic response to chemical exposures based on the normal distribution of
effects. These data may come from epidemiological and occupational dosing or exposure
studies, case reports, or controlled clinical testing. Conversely, human data sets (especially case
reports and retrospective studies) are often limited by poorly characterized or the lack of data
on dosing or exposure. When human data are considered for use in the development of PALs, it
is imperative that special consideration be given to assure that studies involving intentional
dosing or exposure were conducted ethically with the informed consent of the participants. For
11
-------
early reports, which may lack details regarding participation of human subjects, there must be
assurance that such studies were conducted in accordance with guidelines for ethical standards
(appropriate at the time) and with clinical supervision in place. These ethical standards may
include: Nuremberg Code (1947), U.S. Army Regulation 70-25 (1962; 1990), Declaration of
Helsinki (1964 and amendments up to 2000), National Research Act (1974), Belmont Report
(1979), the "Common Rule," Protection of Human Subjects, (40 CFR 26, 2000), and U.S. EPA
Final Rule (2006a).
5 Derivation of PAL Values
Optimally, PAL values are developed based on reports describing the dose or concentration
dependent changes in the expression of effects that can be assigned to a specified PAL tier, in
humans representing those thought to be generally susceptible to such effects, for durations of
dosing or exposure that are in close agreement with the PAL duration of interest, and which
demonstrate sufficient data to support an estimation of a threshold dose or exposure
concentration (a POD) that demarcates the transition of the severity of effects from one level of
severity (tier) to another. Because data sets such as these are seldom available, adjustments
can be made to less robust yet reliable data, as guided below. This may include both
quantitative adjustments of dosimetry and the application of uncertainty factors.
PAL values are derived according to the threshold method used for other risk values, including
RfD, RfC, and MRL values. Here,
PAL Value = POD (ADJ) / (UF x MF)
Where, the route-specific (inhalation or oral) PAL value is the defined severity category or tier
1, 2, or 3 for the specified duration of 24-hours, 30-days, 90-days, or other; POD(ADJ) is the
point of departure for the critical effect adjusted to a continuous exposure duration; UF
(Uncertainty Factor) is the product of values for four Uncertainty Factors (described later); and
MF is the Modifying Factor (described later).
Data analysis and the development of PAL values should focus on the identification of the
route-duration- and tier-relevant critical effect and point of departure (POD), adjustments for
areas of uncertainty regarding the chemical-specific toxic response, and dosimetry. The
identification of a specific critical effect, especially when available data shows multiple effects
for a given exposure, requires professional judgment. Unique, chemical-specific issues should
be addressed in the PALs documentation, as required.
If data to derive a scientifically defensible PAL value are unavailable, the designation NR (Not
Recommended) will be applied with a specific note on this status. In cases where data are
available to enable the computation of a PAL value but the resulting value is relationally
inconsistent with other PAL values (e.g., a PALI being very close to or higher than a PAL2
value; a 30-day PAL-1 value is lower than a corresponding 90-day PAL-1 value), or where
properties of the chemical make it inadvisable to recommend a value (e.g., notable effects
would be experienced in the absence of detection of the chemical), an explanation will be given
with the NR designation.
Intravenous administration represents a worst-case scenario for bioavailability as it assumes
100% absorption. For some chemicals, especially therapeutic agents, findings from intravenous
administration may be the most robust datasets available. Findings from the i.v. route may be
12
-------
used when necessary. In these cases, procedures to adjusted i.v. doses to oral or inhalation
equivalents must be clearly presented. When conducted, adjustments based on pharmacokinetic
data or models should include a rationale for selection of the dose metric upon which the route
extrapolation was based (U.S. EPA, 2014).
5.1 Evaluation of effects
Because the application of PALs methodology results in the development of dosing or exposure
values corresponding to defined levels of severity, and because empirical observations of toxicity
information are typically used as the basis for the development of these values, the identification
of effects corresponding to the three established categories of severity (which may generally be
characterized as producing mild, reversible effects that are not necessarily adverse; producing
more severe or irreversible or escape-impairing effects; or lethality) is an important activity.
NRC (1993) acknowledged this requirement relative to distinguishing CEEL-1 effects from
CEEL-2 effects, "Two other grades of severity - disability and discomfort - though less well
defined, place distinct demands on emergency and health care services." Given the continuum of
biological responses following dosing or exposure, and the extent to which Tier effects (lethality)
may be relatively easily determined, a higher level of attention toward the distinction between Tier
1 and Tier 2 effects may be expected.
The assignment of "adverse" or "not adverse" as descriptors for changes in the biochemistry or the
physiology of an organism following chemical dosing or exposure, and the determination that an
effect may or may not be reversible upon cessation of exposure, may be complicated. However,
decisions regarding the adverse and reversible nature of chemical effects have direct relevance to
selection of critical effects for derivation of PAL-1 and PAL-2 values. The selection of adverse
effects may best be accomplished on a chemical-by-chemical basis, or even on the basis of an
assessment under development. The National Academy of Science evaluated and described the
development of toxic (adverse) responses in the context of the Toxicity Pathway paradigm (NRC,
2007). In this paradigm, chemical dosing or exposure may initially or at low doses or
concentrations result in biologic perturbations, the consequences of which are related to the dose
or exposure, timing, and duration of exposure. For example, adaptive changes may enable a
maintenance of homeostasis, but the maintenance of homeostasis may be overwhelmed by doses
or exposures of longer duration at higher doses or concentrations, resulting in the development of
additional changes that may represent an adverse condition. In other cases, an observed and
measurable change in a biological system may be a biomarker of dosing or exposure, which is
different from a biomarker of an effect. The relative importance of such changes may also be
better estimated when interpreted in the context of the Adverse Outcome Pathway (Villenueve et
al., 2014). The characterization of a suite of observed and interlinked changes sufficient to
constitute a Mode of Action (Sonich-Mullin et al., 2001), and to determine its relevance to the
humans (Boobis et al., 2006, 2008) may be available. If so, this information can be used to
increase the level of confidence in extrapolating dose or concentration response findings from
experimental animals to humans.
Two recent publications from the pathology community of experts shed additional light on the
adverse and reversible nature of chemical effects. Kerlin et al. (2016) recommend that a
distinction between "markers of toxicity" and toxic/adverse effects should be made, with "markers
of toxicity" observed in one tissue being considered more representative of an adverse condition
when they are observed in conjunction with related changes in target organs, tissues, or functions.
Palazzi et al. (2016) provide additional insight that cautions against binary interpretations of
individual effects as representing adversity or not, and advocate for development of a complete
understanding of the lesion or effect including control incidence, severity, and correlations with
13
-------
other relevant effects. Holsapple and Wallace (2008) reached a similar conclusion regarding
isolated findings, and recommended that changes may not represent an adverse condition when
they are of a magnitude insufficient to result in a functional change. This line of thought seems
consistent with the decision by the AEGL Committee to select a 22% methemoglobin
concentration in humans as an endpoint consistent with a tier 1 effect (NRC, 2001) - representing
an "asymptomatic or nonsensory effect" not rising to the level of a "serious or irreversible health
effect."
The categorization of effects according to severity should be based on the expected health impact
at the level of the intact animal or human, recognizing that the distinction between tier 1 and tier 2
effects may at times be challenging. Because of the chemical-specific nature of toxicity, it may be
difficult to establish default categorizations for effects that may be generally similar among
chemicals. While some information may be gained by considering effect categorizations in other
risk systems (e.g., AEGLs), consideration must also be given to factors such as differences in
exposure duration and differences in the availability of additional or supporting information since
the previous categorization effort. Regarding categorization of health effects by tiers, several
questions can be formulated. Data for the chemical, route and duration of interest can provide
informative answers to questions like these.
• Is the effect morbidity or lethality? If so, it is appropriate for tier 3 consideration.
• Does the effect indicate a decrement of motor, sensory or central nervous system function?
If so, whether reversible or not, it may indicate a decreased ability to escape a contaminated
environment and is appropriate for tier 2 consideration.
• Does the effect represent a non-life threatening decrement of function of organs, tissues or
physiologic or biochemical processes? If so, the extent to which is represents a serious
condition or a reversible condition can contribute to its characterization as tier 1 or tier 2
effect.
• Does the effect represent a non-life threatening alteration of tissue organization or structure
(e.g., histologic changes)? If the changes are not expected to produce a serious decrement
of function (e.g., gas exchange in the alveolus), and may be expected to be reversible or
repairable upon cessation of exposure, they may be classified as tier 1 effects. If the
changes are expected to produce a serious decrement of function, regardless of the
expectation of reversibility, they may be classified as tier 2 effects.
• Does the effect represent a mild effect, not seriously impacting function, not expected to
reduce the ability to escape a contaminated environment, and for which a return to normal
function is expected upon cessation of exposure? If so, the effect may be categorized as a
tier 1 effect.
• Does the effect represent an adaptive change in a baseline value for a physiologic or
biochemical parameter (which is not expected to represent a more serious effect)? If so, it
may be appropriate to classify the response as a tier 1 effect.
5.1.1 Mechanistic considerations
For some chemicals, mechanistic data may inform the selection of the most appropriate approach
to duration extrapolation. This concept has been operationalized by ten Berge and others, where
responses from a series of concentrations exposed for a series of durations have been compared. It
can sometimes be seen that the same cumulative exposure (concentration x time; e.g., expressed
in units of ppm-minutes) do not produce consistent responses. Likewise, some studies have
demonstrated that animals exposed to the same daily dose (more or less, continuously) via
drinking water or by a single bolus gavage do not develop the same types or severity or response.
Compressing the cumulative daily dose into a shorter exposure period or into a bolus dose can
14
-------
alter the toxicokinetics of the compound relative to multiple individual exposures to lower
(fractional) doses encountered over longer time periods. Different toxicokinetic patterns may
include higher maximal concentrations in blood and tissues following short-duration exposures or
bolus doses, and consideration should be given to the possibility of metabolic saturation. Whether
compressing the daily dose into shorter periods increases or decreases toxicity will vary among
chemicals, and with respect to whether the potentially saturated metabolic process(es) represents a
bioactivation or a detoxication process, among other things.
Some toxicities are based on mechanisms that are relatively slow to develop. Particularly true for
the shortest PALs duration (up to 24 hours), lethality may not become evident until hours or days
following the end of the 24-hour period since initiation of the exposure. For these reasons, it is
important to carefully evaluate reports that include observations made following a post-exposure
(recovery) period, during which latent effects may become manifest. This may be particularly for
the evaluation of effects at short exposure durations for chemicals which are poorly metabolized
and slowly cleared from the body.
The lack of a mechanistic understanding induces significant uncertainty in the adjustment of doses
identified on the basis of longer-term exposure durations for shorter exposure durations on the
basis of a mathematical adjustment based on (e.g.) the number of days exposed.
5.2 Identification of critical effect
PALs directly address the progression of severity of toxicity with increasing dose or exposure as
the basis for deriving multiple health-based levels of exposure. None of these tiers of severity
may be directly comparable with the severity (or lack thereof) of effects or levels used as the basis
for risk values intended for broad population applicability, for continuous dosing or exposure over
a lifetime (e.g., chronic IRIS Reference Dose or Reference Concentration values). The evaluation
of chemical-dependent effects requires considerable judgment regarding the nature, severity, and
permanence of their expression. Dose- or concentration-response data for the critical effect is
used to determine the point of departure (POD), which serves as the quantitative basis for PAL
value derivation. The key study is that manuscript or report that describes the dose- or
concentration-response data for the critical effect, and identifies the POD. The selection or
identification of the critical effect and its POD also identifies the key study.
The critical effect is a response that is consistent with the PAL tier levels and is characterized in
the key study (and supporting studies where relevant). The POD for the critical effect serves as
the basis for deriving a specific PAL value. It is imperative that the critical effect be of a severity
consistent with the PAL tier. Selection of effects or POD for effects of notably lesser severity than
the specific PAL tier may result in unrealistically low PAL values and inappropriate emergency
response (e.g., unnecessary evacuation or delayed reentry). When circumstances require reliance
on an effect of higher severity (i.e., when data fail to identify an effect of the appropriate severity),
the POD for the more severe effect will require adjustment via the application of an uncertainty
factor (described below).
Characterization of the critical effect should include:
• Species in which it occurred
• Gender and strain if relevant
• Exposure or dose duration associated with the effect
• Dose- or concentration-response information including an estimate of the threshold for the
effect
• Evaluation of the strength of the association between exposure or dose and effect,
15
-------
including consideration of the quality of the data
• Detailed description of the effect including the actual physiological/toxicological
response
• Target organ or tissue, if data warrant
• Clinical chemistry
• Gross pathological and histopathological correlates, if these are available
• Post-exposure observation period, if applicable
• Data quality issues
5.2.1 Severity of Effects
The evaluation and characterization of effects to serve as the basis for PAL value development is
based on several precedents established by accredited risk assessment bodies deriving values for
emergency response purposes. The NRC provided guidance for evaluating data and selecting
health effects for application in setting exposure levels for three levels of severity, for one-hour
Community Emergency Exposure Levels (CEELs) for Hazardous Substances (NRC, 1993):
• Exposures below CEEL 1 values are unlikely to lead to discomfort
• Exposures above CEEL 1 values result in an increasing likelihood of discomfort, but for
which the direct toxic effects of the chemical are unlikely to lead to disability
• Exposures above CEEL 2 values result in an increased likelihood of disability (disability
becomes increasingly common), but for which the occurrence of death or life-threatening
effects are unlikely
• Exposures above CEEL 3 values result in an increasing likelihood of the occurrence of
death or life-threatening effects
This three-tiered system of exposure values and its characterization of tier-relevant types of
effects was also implemented in guidance that was used to develop Acute Emergency Guideline
Levels (AEGLs) by the NRC (2001) which include durations of exposure from ten minutes to 8
hours, and has been used as the basis for characterization of effects for development of PAL
values.
5.2.1.1 PAL-1
PALI represents a threshold no-effect-level (PAL-1 effects need not necessarily be "adverse")
immediately above which there may be reversible, measurable changes from baseline values of
various biomarkers of dosing or exposure, or subclinical changes in physiologic responses. The
critical effects upon which PAL-1 may be based include biomarkers such as normal
compensatory changes in clinical chemistry values (e.g., threshold for increased activity of
enzymes associated with normal metabolism and detoxification pathways, threshold
methemoglobin formation, or red blood cell cholinesterase activity inhibition) or thresholds for
reversible physiological responses (e.g., slight miosis or lacrimation, slight nasal irritation, or
odor or taste detection). Careful consideration must be given, however, to developing PAL-1
values using data for effects that may be imperceptible or innocuous but which could be
considered precursors of a significant toxicological process (e.g., hemolysis), especially for
chemicals for which the dose- or exposure-response curve is extremely steep (e.g., arsine, some
metal carbonyls). In such cases, the PAL-1 must be sufficiently protective to avoid an
unacceptable response at the PAL-1 level. In addition, the critical effect cannot be more
appropriate for PAL-2. If reversibility following cessation of dosing or exposure is not
demonstrated, some justification of the effect, at the dose level or exposure concentration and the
duration encountered should be provided.
16
-------
5.2.1.2 PAL-2
PAL-2 represents a threshold for effects characterized as serious, possibly irreversible, or escape-
impairing. Multiple effects may be characterized as PAL-2 effects, and the suite of PAL-2 effects
should be expected to vary from chemical to chemical, and they may also vary according to
dosing or exposure duration for a given chemical. Professional judgment should be used to
determine the severity of the effect. Examples of severe or irreversible effects possibly occurring
above PAL-2 might include clinical signs of CNS depression, pulmonary damage (e.g., pulmonary
edema, hemorrhage, congestion, or evidence of pathologic changes), ocular damage,
gastrointestinal bleeding, organ injury, clinically relevant hemolysis, or initiation of asthmatic
episodes. A response that results in an increased potential for prolonging the effects of exposure or
otherwise subjecting one to a more hazardous dose or exposure would also be considered as basis
for PAL-2 development.
Latency in expression of an effect (e.g., arsine and hemolysis-induced renal failure, phosgene
and pulmonary edema) and the possibility of increasing severity in the effect after cessation of
exposure (e.g., sulfur mustard exposure and pulmonary damage) must be considered. Although
reversible, escape-impairing effects such as CNS depression, severe dizziness, vertigo,
convulsions, ataxia or other motor deficiencies, and impaired vision are consistent with PAL2
severity. The evaluation of developmental, reproductive, and endocrine effects may be relevant
to all durations for PAL-2 development.
5.2.1.3 PALS
PAL-3 values represent estimated human lethality thresholds. At exposures equivalent to the PAL-
3, serious toxicity is to be expected. Because PAL-3 levels generally will be developed based
upon lethality data in animals, PAL development must consider the uncertainties inherent in such
animal-to-human extrapolations while not arbitrarily or excessively reducing PAL-3 values.
Relevant experimental findings include early death, morbidity, and the rapid onset of life-
threatening toxicities. The inclusion of a post-exposure observation period will increase
confidence that the results do not over-estimate a threshold for lethality for the experimental
dosing or exposure duration. When available, interim observations (e.g., the time of death) can
also provide important quantitative details. For example, the authors of a 90-day dosing or
exposure study may only describe lethality indirectly by reporting survival to study termination, or
in the form of survival data which may (inadvertently) quantify lethality at only after (e.g.) 1 or 30
days.
5.3 Identification of the Point of Departure (POD)
The POD is the actual dose or exposure concentration and duration associated with a critical
effect used quantitatively as the basis for derivation of the respective PAL. This exposure or
dose serves as the basis for quantifying PAL values. For the inhalation exposure route, the
"3
concentration used for operational derivation of the PAL will be expressed as ppm or mg/m
and exposure duration. For oral dosing, the dose will be expressed as mg/kg/d. Ideally, the
POD comes from human exposure data or a well-conducted animal study in a species whose
anatomy, biochemistry and physiology is representative of the human. When the expression
of the oral dose in terms of mg/kg-d requires translation from values reported in original
reports (e.g., ppm in diet or mg/L in drinking water), conversions should be clearly presented
and to the extent possible include references to previously reviewed or otherwise reliable
standards, methods, or parameter values.
17
-------
NRC (1993) indicates that the derivation of a CEEL value may be based on a POD identified as a
NOAEL or a LOAEL, "An additional 10-fold UF may be introduced when deriving a CEEL-1
from a LOAEL instead of a NOAEL, or a CEEL-2 or CEEL-3 from appropriate LOAELs or
FELs." (FELs are frank effect levels, that is, levels that can cause incapacitation or death.) The
concept of "threshold" pertinent to animal POD values is given some attention in the context of
uncertainty factor development by NRC (1993): "For agents that are considered to produce
threshold effects, select an appropriate uncertainty factor of 1 or greater for endpoints affording
NOELs, NOAELs, LOELs, or thresholds." (NOEL is no observed effect level; LOEL is lowest
observed effect level.) This description of thresholds seems to indicate that a POD value may be
based on non-traditional POD effects.
5.3.1 The Threshold Concept
In presenting the approach to noncarcinogenic effects, NRC (1993) indicates, "The existence of a
so-called threshold dose or concentration is supported by but not limited to the fact that the
toxicity of many agents is manifest only after the depletion of a known physiological reserve."
The threshold concept is also central in most contemporary noncancer risk value systems.
However, in the establishment of CEELs methodology (NRC, 1993), it was also recognized that
the resulting risk levels are not themselves bright lines, specifically: "CEELs should indicate
exposures that would be thresholds for the occurrence of (1) death or life-threatening effects, (2)
disability, or (3) discomfort in the population. At such a threshold concentration [a CEEL value],
a small proportion of the population might exhibit effects." This was based on considerations of
the imprecision relating to demarcation of risk categories, the imprecise nature of underlying
toxicity data as might be derived in various sources, and because the incidence of effects at a
given CEEL level will depend on the proportion of hypersusceptible people in the population.
NRC (1993) reiterates, "[T]he intent of CEELs is not to provide absolute assurance that everyone
at risk will be protected under all circumstances, and thus the uncertainty factors should be chosen
with the understanding that hypersusceptible persons might not be protected."
Discussions regarding the establishment of AEGL values through the identification of tier-specific
endpoints and PODs (NRC, 2001) references previous CEELs guidance (NRC, 1993), pages 10,
12, and 21 for clarity and consistency. In these discussions, NRC (2001) reiterated points
regarding the selection of critical effects and PODs. In framing the AEGL values, NRC (2001)
indicated that,
"Because the data and methodologies used to derive AEGLs or any other short-term exposure
limits are not sufficiently precise to make a distinction between a ceiling value [the highest
no effect level] and a threshold value, no distinction has been made with respect to AEGL
values. No fine line can be drawn to precisely differentiate between a ceiling level, which
represents the highest exposure concentration for which an effect is unlikely to occur, and a
threshold level, which represents the lowest exposure concentration for the likelihood of onset
of a given set of effects. Hence, AEGLs are not true effect levels. Rather, they are considered
threshold levels that represent an estimated point of transition and reflect the best efforts to
establish quantitatively a demarcation between one defined set of symptoms or adverse
effects and another defined set of symptoms or adverse effects. Therefore, in the development
of AEGLs, the NAC/AEGL Committee selects the highest exposure level from animal or
human data where the effects used to define a given AEGL tier are not observed."
Frequently, the POD will not be precisely defined in the available study reports. Depending on
the PAL level being derived, this will necessitate estimation of a threshold for the effect or, for
18
-------
some PAL tiers, a minimal or no effect level. For a threshold estimation, this usually entails
selection of the ceiling, i.e., the highest dose or exposure that does not cause a relevant adverse
effect (a NOEL, a NOAEL) for the specific PAL tier. For an actual effect level (a LOEL, a
LOAEL), the POD generally will be the lowest reliable exposure or dose causing a minimal
relevant effect consistent with the PAL tier definition.
5.3.1.1 Benchmark Dose Modeling (BMD)
Benchmark dose modeling has seen application in many risk assessments including those
based on dichotomous and continuous data, and its application to dose response modeling
should be considered during PALs development. The traditional approach to determining the
point of departure based on NOAEL or LOAEL values has disadvantages including, (1) it is
limited to the doses tested, (2) response levels may not be comparable, standardized or
identified, (3) the response level at a NOAEL is not determined, (4) responses at other tested
doses are not considered, (5) a NOAEL is not always available, and (6) statistical comparisons
are impacted by sample size. The application of benchmark dose modeling imparts a greater
level of confidence to a derived POD (a BMD or BMDL) value for several reasons including
consideration of the entire range of observations across all doses and not restricting the
definition of the POD to one of the studied doses. More specifically, BMD modeling (1) does
not identify a point of departure that is limited to one of the doses tested, (2) removes the
impact of dose spacing on the identification of the point of departure, (3) takes responses at all
doses into account, (4) allows for flexibility in determining the level of biological relevance in
the effect, (5) provides for an increased level of consistency in the comparison of endpoints
across chemicals and studies, (6) allows for the integration of data from comparable studies in
quantitative analysis, and (7) is performed using standardized software and supported by peer
reviewed documentation.
However, its application can also be resource-intensive. While not presuming the intentions of
the NAC/AEGL Committee, some consideration of BMD modeling across all three AEGL tiers
may have resulted in their indication that, "these methods will generally be considered for an
acute lethal endpoint. Their use to set AEGL-1 and AEGL-2 values will be considered on a
chemical-by-chemical basis." Young et al. (2009) described the approach used by Oak Ridge's
Expert Consultation Panel which seems to indicate more broad reliance on BMD modeling
across tiers: "When data are sufficient, the Benchmark Dose (BMD) or benchmark
concentration (BMC) for inhalation exposure terms (U.S. EPA, 1995, 2007 [2015]) should be
used to derive an estimate (95% lower confidence limit) of the exposure expected to cause a
specified response incidence. The benchmark is usually a 1% to 5% response incidence. For
PAL derivation, a BMDoi (BMCoi), or the lower limit of the BMDos or BMCos (BMDLos or
BMC Los), ./o/' a specific biological effect [italics added] may be considered."
Because of the Provisional and Advisory nature of PAL values, because of the recognized
variability based on the normal distribution within the study population, the sometimes
imprecise nature of the characterization of biological effects, and because of the intent of PAL
values to provide information generally describing the increase in biological severity of effects
with respect to dose or concentration, requirements (recommendations, suggestions) for broad
application of BMD modeling for all types and severities of effects should be carefully
considered. The application of BMD modeling to estimate the threshold and identify the point
of departure for PAL-3 effects (e.g., lethality) should be undertaken when the data permit,
BMD modeling may also be used to estimate the threshold (point of departure) for PAL-1 and
PAL-2 effects when it is determined that the uncertainty in the threshold (point of departure)
estimated by other means undermines confidence in the derived PAL value. Citing an analysis
19
-------
of acute inhalation lethality data by Fowles et al. (1999) in which BMC lower confidence limit
values (BMCl values) were compared to No-Observed-Adverse-Effect-Concentration
(NOAEC) values, NRC (2001) seemed to indicate that (for acute lethality data) a BMR of 10%
"may be too high a response rate," compared to BMR values of 5% or 1% (more fully
described in Section 5.3.4). NOAEC is the no observed adverse effect dose or concentration.
In all BMD applications, the benchmark response value (e.g., 1%, 5%, 10% response rate)
should be determined on a case-by-case basis.
5.3.2 PAL-1 Point of Departure
The threshold concept has been specifically clarified with respect to tier-1 effects relative to
AEGL values: "Airborne concentrations below the AEGL-1 represent exposure levels that can
produce mild and progressively increasing but transient and non-disabling odor, taste, or sensory
irritation, or certain asymptomatic, non-sensory effects" (NRC, 2001). NRC (2001) also clarified
the selection of a POD value, indicating, "The highest concentration not producing an AEGL-1
endpoint or effect levels for mild sensory irritation, asymptomatic, or non-sensory effects, such as
methemoglobin formation (22%) or altered pulmonary function (transient changes in clinically
insignificant pulmonary functions of a hypersusceptible individual), have been used as AEGL-1
endpoints." It can thus be inferred that AEGL values were established with the expectation that
such an exposure might result in an increase in a parameter such as methemoglobin concentration
of up to 22% or a decrease in FEV1 of up to 20%.
Tier 1 values, as established for AEGL-1 values do not represent no-effect exposure conditions, as
acknowledged by NRC (2001; 1993). Below AEGL-1 values "there may be specific effects, such
as the perception of a disagreeable odor, taste or other sensations (mild sensory irritation)," which
NRC explains by stating, "Since there is a continuum of discomfort in which it is difficult to judge
the appearance of "discomfort" in animal studies and human experiences, the NAC/AEGL
Committee has used its best judgment on a case-by-case basis to arrive at appropriate and
reasonable AEGL-1 values."
There is a broad range of measurable changes in biological systems (changes in biomarkers,
adaptive changes, adverse but reversible changes, etc.) that may be observed below exposures that
produce more severe possibly irreversible changes or changes that may impair an ability to escape
a contaminated environment (tier 2 effects). The POD chosen for PAL-1 derivations may
represent a no-effect level (dose/exposure) for the critical effect (e.g., mild irritation), but is an
exposure or dose that may be expected to produce measurable changes in some parameters not
chosen as the critical effect (e.g., 5% inhibition of plasma acetylcholinesterase activity).
5.3.3 PAL-2 Point of Departure
The NAC/AEGL Committee addressed two possibilities for establishing the threshold and the
point of departure for Tier 2 effects. On the basis of acceptable data describing a tier 2 effect, the
NAC/AEGL Committee indicated a preference for determining the POD value by "estimate[ing] a
NOAEL for serious or irreversible effects or effects that impair escape" to develop AEGL-2
values (NRC, 2001). If data describing the development of tier 2 effects are lacking, then the
highest exposure that caused tier 1 effect but did not cause a tier 2 effect may be selected as the
POD. If so, this value may be subjected to a severity adjustment (see section 5.4.4). Section 5.8
presents an alternate method to estimate a PAL-2 value on the basis of a derived PAL-3 value.
20
-------
5.3.4 PALS Point of Departure
Lethality may not become evident during the course of dosing or exposure. Optimal confidence
can be placed in estimates of the threshold for lethality when studies include a post-exposure
observation period. Several approaches to determine or estimate the threshold for lethality from
the results of studies in which lethality is observed (as suggested by NRC, 2001) may be
employed. These include: (1) identify the lethality threshold as the highest dose or concentration
that does not produce lethality, (2) estimating a lethality threshold as one-third of the animal LCso
(concentration lethal to 50% of the exposed group) or LD50 (dose lethal to 50% of the dosed
group), or (3) calculating a lethality benchmark response (e.g., 1% or 5% response level) by the
BMD method (U.S. EPA, 1995, 2012c, 2015) or other methods such as Litchfield and Wilcoxon
(1949).
Estimation of a lethality threshold as one-third of the LC50 or LD50 may be considered when data
are insufficient for a benchmark calculation. This approach has been used in the AEGL Program
(NRC, 2001) as based on inhalation exposures by analyses conducted by Fowles et al. (1999). The
approach has been subsequently evaluated and endorsed for application to acute lethality data
from inhalation exposure by Rusch et al. (2009). Fowles et al. (1999) analyzed 120 acute lethality
data sets using BMC (benchmark concentration) methods and reported that the ratio between the
LC50 and a nonlethal exposure averaged about 2, with the 90th and 95th percentiles being 2.9 and
3.5, respectively. The ratios ranged from 1.1 to 6.5. This approach is especially valid for those
chemicals with a steep concentration-response relationship. Increased confidence in this approach
is developed when a weight-of-evidence approach comparing the POD value for lethality (as
estimated above) with dose or exposure response data for non-lethal effects derived for the same
duration, and/or with lethality data for other durations. For some chemicals, estimation of a tier 3
value may necessitate other, even less quantitatively robust options. In the absence of a
statistically determined LC50 or LD50, data from a well-designed and conducted study describing
the dose or exposure dependent frequency for lethality (e.g., doses or exposures producing 25%,
60%), and 80% lethality) may be used as either the basis for a one-third reduction, or as the basis to
estimate a 50% response rate upon which to derive a one-third reduction. The one-third reduction
is more conservative than using the highest non-lethal concentration. (Rusch et al. 2009) It is
recommended that such analyses be supported by other data indicating that the estimated value is
a plausible estimate of the lethality threshold as described above and in Section 5.8.
In cases where data are insufficient for a quantitative estimate (e.g., a benchmark calculation or
a no-effect level) of a human lethality threshold, the PAL-3 may be derived by analysis of the
best available data. This analysis should attempt to estimate the highest exposure/dose that will
not result in human lethality. Data descriptions should include the species, exposure
concentration or dose level, and the associated effects, and how the data relate to the lethal
response (e.g., is the effect part of the continuum of toxicity that would lead to a lethal
response). Because this approach may lack in-depth quantitative analysis, all assumptions and
rationales should be thoroughly described.
Calculation of a benchmark lethality threshold provides a more accurate and defensible POD
(see section 5.3.1.1.) for PAL-3 values but requires sufficient dose- or exposure-response data
(data requirements for these methods are provided in U.S. EPA, 1995; 2007). Regardless of the
specific method used, estimated lethality thresholds should always be compared to available
toxicity data to assure some level of continuity and scientific validity.
In the absence of lethality data, exposures or doses known to cause extreme toxicity may be
21
-------
considered for PAL-3 development with appropriate rationales clearly stated and, where
possible, supported by data from laboratory animal experiments or human doses or exposures.
For example, severe morbidity, an effect frequently noted in lethality studies with laboratory
animals, may be considered a critical effect appropriate for PAL3 development.
Similarly, AEGL-3 values define the highest exposure level that does not cause death or life-
threatening effects. NRC (2001) specifically treated the selection of the POD in the context of a
no-effect level, indicating that when data were amenable, a benchmark dose modeling analysis to
identify the BMCL05 value could be used to determine the POD. Alternately, the point of
departure can be determined as "the highest experimental exposure that did not cause lethality in
an experiment in which death was observed" or by a fractional reduction of the LC50 value. In this
case, AEGL guidance indicates that a point of departure may be estimated by dividing an LC50
value by 3.
5.3.5 Duration Adjustment of the Point of Departure
PALs are developed for doses or exposures assumed to be continuous, for durations of (e.g.) up to
24-hours, up to 30-days, up to 90-days, or other durations as required. Experimental data used as
the basis for PAL derivations are often developed using a discontinuous or intermittent
experimental dosing or exposure protocol (e.g., 6 hours per day, 5 days per week for 4 weeks).
This dosing or exposure pattern differs from the PAL-specific durations (e.g., an exposure
assumed to be continuous for the PAL duration), requiring a duration adjustment to account for
differences in discontinuous (experimental, observational) versus continuous (human, anticipated)
exposures. This section addresses adjustments for intermittent dosing or exposure (e.g.,
inhalation exposures conducted for 6 hours per day, 5 days per week; oral exposures via dosing 5
days per week).
Because of their technical nature, inhalation toxicity studies often include a daily exposure for a
number of hours less than a typical work day (often 6 hours per day), and exposures for often 5
days per week. This requires an initial duration extrapolation to the 24-hour period, as discussed
in section 5.3.5.1. Whether time-normalizing short exposure periods to the 24-hour period via
Haber's Law or the ten Berge approach, consideration should be given to mechanistic data. These
data may indicate that the effects observed from shorter (e.g. 6 hours) periods of exposure to
higher concentrations may not be similar to the effects observed from longer (e.g., 24 hours)
periods of exposures to proportionately lower concentrations. Reasons may include the saturation
of metabolic processes by higher concentrations, not observed at lower concentrations, the
inhibition of repair processes by higher concentrations, etc.
By convention, oral doses are typically expressed in units of mg/kg per day. Oral studies in
animals often involve dosing via drinking water or feed, representing an exposure scenario
consistent with expectations for human exposure conditions. Alternately, animals may be exposed
via oral gavage (neat or in an aqueous or organic vehicle). Exposure via feed or drinking water
allows the exposure to be temporally dispersed over the day, while gavage dosing (often selected
on the basis of experimental convenience or cost) includes a bolus administration, concentrating
the daily dose into a single unit of administration. This difference in dosing produces differences
in the pattern of internal (tissue) exposure, and can serve as the basis for differences in responses,
as well. Doses administered as a bolus have a higher potential to overwhelm detoxication
pathways and repair capacity, resulting in a pattern of toxicity different from that observed from
drinking water or feed exposures.
Available mechanistic information should be considered (see section 5.1) when conducting a
22
-------
duration extrapolation, whether from periods of less than 24 hours to 24 hours, or from periods of
less than one week to one week. When the toxicity data for a chemical indicate that such an
adjustment is not valid, it should not be conducted (e.g., adjusting exposures by hours per day for
irritant gases producing a pulmonary effect).
When experimental exposure durations are not continuous, extrapolation from the reported time-
specific exposure concentration or dose to an equivalent concentration/dose for a PAL-specific
period may be required for some chemicals (e.g., those inhaled toxicants producing effects not
characterized as irritation). When justified on the basis of the type of toxicity demonstrated, the
first step of the duration extrapolation involves duration adjustment to the day (the 24-hour period
of dosing or exposure), and this is done differently for oral and inhalation studies. The second
step adjusts the number of days per week the animals are dosed or exposed adjust the dosing or
exposure to 7 days per week.
5.3.5.1 Inhalation Data
Inhalation points of departure are typically expressed in units of concentration and time (e.g., 5
mg/m3 for 4 hours, 5 days per week). Because PAL values are applicable to continuous human
exposures, intermittent experimental exposure data require adjustment into equivalent measures
of continuous exposure. The RfC methodologies require the use of an exposure value adjusted
(ADJ) for continuous exposure (NOAELadj or LOAELadj). Because the PAL tiers are based on
threshold effect levels, PODadj is a more appropriate term. For example, an experimental
exposure of 5 mg/m identified as the exposure concentration associated with the critical effect
in a study utilizing a 4 hrs/day, 5 days/week protocol may be adjusted to continuous exposure
by:
POD adj = 5 mg/m3 x [(4 hrs/day)/(24 hrs/day)] x [(5 days/week)/(7 days/week)] = 0.59 mg/m3
The two-part inhalation duration extrapolation procedure first accounts for intermittent
exposure with respect to hours per day, then with respect to days per week (when applicable).
However, many chemicals exhibit an exponential relationship between exposure duration and
effect (Cn x t = k). For acute exposures (durations < 24-hr), an attempt should be made to
empirically derive a time scaling factor (i.e., the exponent 'n') if data are available. For mild
irritation effects, it is recommended that no extrapolation from the experimental exposure
duration be performed. If there is reason to believe that continued exposure may potentially
result in greater pulmonary damage (e.g., increased edema), extrapolation may be appropriate.
Haber's Law traditionally has been used to relate exposure concentration and duration to a toxic
effect (Rinehart and Hatch, 1964), but may apply to a small and sometimes undefined range of
exposure concentrations. Specifically, the equation implies that exposure concentration or
duration may be adjusted to attain a cumulative exposure constant {k) which relates to a toxic
response of specific magnitude. Work by ten Berge et al. (1986) affirmed that chemical-specific
relationships between exposure concentration and exposure time may be exponential rather than
linear; i.e., the expression now becomes Cn x t = k, where n represents a chemical-specific
exponent. Upon examining the concentration and time relationship of the lethal response to
approximately 20 chemicals, ten Berge et al. (1986) reported that the empirically derived value
of n varied from 0.8 to 3.5. The n values for some of these values are presented in NRC (2001,
page 95). The value of the exponent (n) is determined by the relationship between exposure
concentration, exposure duration, and cumulative response, such that if n = 1, the toxic response
to the chemical is dependent solely upon the product of concentration times time, (i.e., a linear
-------
relationship, or Haber's Law where concentration and duration are equally important).
Generally, values for n < 1 result when the exposure duration is the primary determinant of the
cumulative response and values for n > 1 result when the exposure concentration is the primary
determinant of the cumulative response.
Although ten Berge et al. (1986) considered only lethality data, where adequate data are
available (i.e., multiple exposure concentration-exposure duration combinations all of which
produce a quantitatively similar response), a chemical-specific exponent (n) for use in
extrapolating available exposure data to specific durations may be calculated for nonlethal end-
points; confidence in this approach is increased when it can be demonstrated (or reasonably
assumed) that progression of the non-lethal effects to lethality is likely. The time scaling
exponent (n) for a given study and PAL duration is derived by linear regression of the log-
transformed exposure concentrations and exposure time data. Although debate continues
regarding the placement of the exponent in the expression Cn x t = k, PALs will be derived
using the exponent for the concentration variable.
The linear regression is of the form
Y = a + bX
where Y = predicted dependent variable and X is the independent variable, a is
the y intercept, and b is the X intercept. The log transformation of Cn x t = k
becomes
log C = (log k)ln + (-1 In) x log t
where C is the predicted exposure concentration for a given exposure duration,
t. The expression (log k)ln is the Y intercept of log C vs. log T plot, and -1 In is
the slope of this plot.
It is important to emphasize that the response associated with the cumulative exposure constant
(k) must be a response that can be definitively quantified. The data must be discontinuous
quantal data (e.g., live/dead such as LCso or LD50 values) or continuous data (e.g., serum enzyme
activity in International Units) for which precise quantitation of the response is possible.
If data are unavailable for empirically deriving the exponent (n), a default value for n may be
applied. In a previous analysis of concentration-time response data for numerous chemicals
(NRC, 2001), values of 1 and 3 for n were identified as the upper and lower bounds of the range
of exponent values. Specifically, it was recommended that for extrapolation from longer to
shorter exposures, n = 3 would be appropriate while n= 1 would be applied to extrapolations
from shorter to longer exposure durations.
Regardless of the exponent used, the resultant extrapolated value should be evaluated in context
of other data to verify its validity. Careful attention must be given to extrapolations over extreme
time intervals (e.g., extrapolating the results of a 1-hour exposure to a 24-hour duration). For
these cases, analysis of relevant and appropriate data and scientific judgment are especially
necessary to assess validity of the temporal extrapolation. With the logical exception of deriving
24-hour inhalation PAL values, once intermittent exposures are adjusted to the continuous daily
exposure (the first step in duration adjustment), exposure concentrations are extrapolated to the
week by accounting for the number of exposure days per week (as described for oral dosing,
section 5.3.5.2).
24
-------
5.3.5.2. Oral Data
When laboratory animal studies involve discontinuous treatment regimens (e.g., 5 days/week for
gastric intubation dosing), adjustment of such discontinuous dose regimens to the targeted human
exposure scenario (continuous, daily exposure) is necessary for the development of (e.g.) a 30-day
oral PAL values. This involves converting the intermittent dosing (e.g., 5 days/week) to a
continuous dosing (e.g., 7 days/week) according to the second step of the duration extrapolation
shown for inhalation exposures (above; U.S. EPA, 1988a) and consideration of what effect the
vehicle (e.g., food vs. solvent) may have on a chemical's toxicity. For example, if a point of
departure dose were 100 mg/kg, administered orally 5 days per week, the duration-adjusted POD
value would be derived:
POD adj = 100 mg/kg x (5 days / 7 days) = 71 mg/kg-day
This completes the duration adjustments of the POD value. When necessary, the extrapolation of
duration-adjusted point of departure values across broader ranges of time is accomplished by the
application of the Time Extrapolation Uncertainty Factor for (UFt), described in section 5.4.3.
5.4 Uncertainty and Modifying Factors
PAL values are intended to represent best estimates of dosing or exposure associated with three
levels of severity. They are both provisional and advisory in nature, and their application in
emergency response activities should be in conjunction with other available information. This
application implies a requirement that their derivation be based on methods that are reliable. As
such, a technical treatment of uncertainty and variability are typically beyond the scope of PAL
values. However, these concepts should be treated at some level.
Oftentimes, the key study and the overall data set are lacking in some critical aspect (e.g., the
study is not conducted for a duration directly relevant to the PAL duration under evaluation
or the study does not identify a NOAEL for the chosen effect) and, therefore, development of
the PAL will necessitate adjustment with uncertainty factors or modifying factors to account
for extrapolations.
The extrapolation of findings from test species (even from humans) is necessary to estimate
cumulative dose or exposure in the general population, encountered for longer durations than
experimental durations that may produce effects of a given type or magnitude (severity). Such an
extrapolation process is possible only when the necessary steps are completed; when specific data
are lacking, the data gap must be filled by the application of default assumptions, hence the
development of the uncertainty factors. Because these factors are multiplicative (e.g., the total
uncertainty factor value is a product of the values of the individual uncertainty factors), two
requirements emerge. First, all gaps require the application of an uncertainty factor, and second,
uncertainty factors must have a numerical value. Even when an extrapolation step is covered by
available data, the uncertainty factor for the step is included, but is given a value of 1. Uncertainty
factors in PAL risk values are generally similar to uncertainty factors used in other risk systems,
as explained below. PAL values include uncertainty factors for extrapolation:
1) from animals to humans (interspecies extrapolation uncertainty factor, UFa),
2) for extrapolation to account for variation in sensitivity among humans (intraspecies
extrapolation, UFh), 3) to account differences between the duration of the experimental exposure
and the duration of the PAL value being estimated (UFt), and 4) to account for differences
-------
between the severity of the effect chosen as the critical effect and the tier category for the PAL
value being estimated (severity adjustment uncertainty factor, UFs).
An additional factor, the Modifying Factor (MF), is used in PAL derivation and should be
distinguished from the Database Uncertainty Factor, as applied in the derivation of RfD, RfC, and
other risk values. This is further explained in section 5.4.5.
Application of each UF has been described in detail for derivation of both oral and inhalation
reference values (i.e., chronic oral RfD and inhalation RfC values), which address
noncarcinogenic effects of substances due to long-term daily exposures (Dourson and Stara, 1983;
Dourson, 1994; Dourson et al., 1992; U.S. EPA, 1989b; US EPA, 1994). Application of some of
the UFs has also been described for derivation of acute values designed for emergency planning
(NRC, 2001). Concepts developed the for use of UFs in chronic and acute scenarios can be used
as a guide to application of UFs in the derivation of PALs as discussed below for each of the
defined extrapolation steps (representing potential sources of uncertainty). It is important to
account for the species in which the effect is quantified, the severity of the effect, the magnitude
of its expression, and the duration of dosing or exposure when deciding which UFs are relevant to
the respective PAL value. In addition, the value of each UF should be specific for each chemical,
route, tier, and duration. As such, these values should not be expected to be entirely consistent for
the range of PAL values determined, but their respective values should be determined by
consistent application of methods and the basis for their values clearly described. This is
discussed in sections 5.4.1-5.4.5.
To allow for optimally reliable assessments, it is necessary to:
• Carefully analyze the uncertainties and incorporate this uncertainty into PAL
development
• Derive PAL values that are scientifically defensible based upon reliable data
• Evaluate PAL values relative to the overall data/weight-of-evidence for the
specific chemical
NRC (1993) cautioned against the selection and application of uncertainty factors without specific
consideration of the intent of CEELs values, indicating, "In the case of CEEL-2, uncertainty
factors must be balanced against the inherent risk associated with actions, such as evacuations,
that might be taken as a result of application of CEELs. Large uncertainty factors, which might be
appropriate with chronic exposure limits, such as PELs [permissible exposure limits], might be
associated with increased risk to the community in the application of a CEEL-2."
5.4.1 Interspecies dosimetry adjustment uncertainty factor (UFa)
One of the fundamental assumptions in human health risk assessment is that the human is more
susceptible to the toxic effects of a chemical than the test animal species unless data can
establish the relative sensitivity between the species in question. Thus, in the absence of data
describing the response in humans, the default value of 10 is used for UFa. Dosimetry
corrections for animal-to-human conversion are frequently an issue of concern in the
development of toxicity values and are a component of interspecies variability. Specifically, it
is important to assess as accurately as possible the relationship between the exposure or dose in
the experimental animal and the exposure or dose predicted to produce an equivalent effect in
humans.
The subdivision of UFa into toxicokinetic and toxicodynamic components has been
accomplished (U.S. EPA, 1994; WHO, 2005), closely followed by the development of methods
26
-------
to quantify interspecies differences in each component. Interspecies adjustment for non-
carcinogenic effects results in an estimation of a Human Equivalent Concentration (for inhaled
toxicants) and a Human Equivalent Dose (for orally encountered toxicants). Various
guidelines and standards have been developed that often incorporate extensive algorithms to
estimate animal-to-human inhalation exposure adjustments. Several methods have been
proposed to adjust for differences in the delivery of an inhaled dose to target tissue in the
respiratory system (U.S. EPA, 1994) and systemic targets (U. S. EPA, 1994; NRC, 1993), on
the application of allometric scaling approaches to complete species extrapolation for orally
encountered toxicants (EPA, 201 la). Consistent with guidance for AEGLs (NRC, 2001),
interspecies adjustment via default methods established for RfC derivation (U.S. EPA, 1994)
and for oral RfD development (U.S. EPA, 201 la) are not recommended for PALs
development.
Physiologically-based pharmacokinetic (PBPK) models have become more useful in
eliminating some aspects of uncertainty regarding dosimetry issues, perhaps most frequently in
quantifying the variability between test species and humans regarding toxicant exposure at the
tissue level (pharmacokinetics, toxicokinetics). Available PBPK models that satisfy
requirements of available guidance regarding their application in quantitative risk assessment
(e.g., WHO, 2010; U.S. EPA, 2006b, 2014) may be used to assess animal to human (species)
variability regarding the uptake of chemicals, doses to target tissues, and time-course for
absorption, distribution, and excretion of a chemical of concern and its metabolites. If PBPK
modeling is applied, the results should be presented and interpreted in a manner consistent with
available guidance (e.g., U.S. EPA, 1994, 2014; WHO, 2010). The results should be presented
as dose (mg/kg) or concentration (ppm or mg/m3) adjusted for animal-to-human toxicokinetic
differences. Quantitation of toxicokinetic variability should result in a reduction of the 10-fold
default value for the uncertainty factor for interspecies extrapolation (UFa).
Recently, the U.S. EPA (2014) published guidance for development of data-derived extrapolation
factors, which addresses both toxicokinetic (TK) and toxicodynamic (TD) components of inter-
and intra-species variability. However, the utility of the more complex approaches for
(particularly) inhalation exposure are limited by lack of data for many of the parameters, or by
lack of methodology validation. Some of these approaches may also require resources of time and
data that preclude their application to chemicals for which emergency response values are a
priority. Thus, in the interest of consistency among values under development, their application in
PAL value derivation is generally discouraged (but not precluded). This position is consistent
with NRC (2001, page 71) regarding UFa, as evidenced in several statements including: "The
guidance (NRC, 1993) suggests that the UF should be based on the quality of the data available,"
"As data are available, the NAC/AEGL Committee uses data-derived interspecies UFs" and "As
always, all information on the chemical, its mechanism of action, structurally related chemical
analogs, and informed professional judgement will be used when determining appropriate UFs and
evaluating the resultant AEGL values."
An interspecies UF of less than 10 (the historical default value) may be applied if a human
equivalent dose or human equivalent concentration can be determined via PBPK modeling, if
data for the most sensitive species are used, if humans are shown to be less sensitive than
animals, or if the mechanism of toxicity or mode of action (e.g., direct-contact irritation) is not
expected to differ between animal species and humans. For direct-contact irritants, a default UF
of 3 is appropriate and justified (based on a mode of action that may be assumed to be similar
between animals and humans; NRC, 2001) whereas for systemic effects, a UF of 10 may be
appropriate. Metabolism and disposition may exhibit interspecies variability indicating that
humans are notably less susceptible than rodents, thereby dictating an interspecies UF that is
27
-------
operationally less than 1. Such is the case for some volatile organic solvents such as chloroform
and carbon tetrachloride (Delic et al., 2000; Paustenbach et al., 1986 a,b, 1988; Gargas et
al.,1989). Regardless of the method(s) used for quantitation or estimation, the value applied for
UFa should be explained.
5.4.2 Intraspecies variability uncertainty factor (UFh)
Another major area of uncertainty involves human individual variability (the distribution of
effects among humans). Extrapolation of findings to cover the population entails considerations of
differences in exposure to chemicals in environmental media and differences in response.
Variability in response development among humans, once exposed, can be based on differences in
internal dosimetry (toxicokinetics) or differences in the development of the biological response
(toxicodynamics). Preexisting physiological conditions (e.g., asthma or other respiratory
disorders), age (elderly and infants may exhibit variability in toxic responses relative to other age
groups), sex, ethnicity/race, nutritional status, lifestyle habits (smoking and alcohol consumption),
and socioeconomic factors may impart sensitivity (susceptibility) through alterations of
toxicokinetic or toxicodynamic mechanisms.
Using the normal distribution type as an example, with dose or concentration increasing across the
X-axis and frequency of the response among the population increasing up the Y-axis, the bell-
shaped profile demonstrates that the frequency or responses is lower at the lower concentrations,
that the bulk of the population demonstrates the response as the dose or concentration increases,
and that a low proportion of the population may not demonstrate the response until much higher
doses or concentrations are attained. Likewise, at a given dose or concentration sufficient to cause
the response at some fraction of the population, it is likely that some fraction of the population
may demonstrate a response of lesser severity (or none at all), while some other segment of the
population may demonstrate a response of greater severity.
UFh is applied to account for toxicokinetic and toxicodynamic variations among the human
population; the default value of 10 is applied when the POD is not determined in humans. When
the POD is determined in humans, values other than 10 may be considered. This UF is applied
when the toxicity database may not include effects in sensitive populations. An intraspecies UF
of less than 10 (the historical default value) may be used if the POD is determined in humans
that are representative of population groups presumed to be sensitive (in individuals sensitive to
particular compounds such as asthmatics exposed to irritant gases); or if the mechanism of
toxicity or mode of action (e.g., ocular, nasal, or pulmonary irritation) is not expected to differ
greatly among individuals. Similar to UFa, a default UF of 3 for UFh is appropriate and justified
for direct-contact irritation (and pulmonary irritation) whereas a UF of 10 for systemic effects
may be appropriate.
Reduction of the default value of 10 for UFh for direct-acting irritants is supported by studies
with healthy individuals and asthmatics that indicate a range in variation of 1-5-fold (Table 1).
This reduction is also more relevant in derivation of acute and short-term PAL-1 values (for
non-irritant chemicals) if uptake and metabolism/disposition differences and tissue
accumulation are not expected to vary among individuals in the population. For example, it is
documented that the degree of CNS depression caused by volatile organic solvents vary little
among age groups in humans (Gregory et al., 1969; de Jong and Eger, 1975). Non-default
values for UFh should be developed and described on a case-by-case basis.
28
-------
Table 1. Variability in Response of Asthmatics to Irritant Gases
Chemical
Estimated
threshold in
healthy subjects
Estimated
threshold in
susceptible
subjects
Estimated
differential
response
factor
Susceptible group
Reference(s)
Chlorine
1.0 ppm (P)
1.0 ppm (S)
0.5 ppm (P)
0.5 ppm (S)
2 (P)
2 (S)
Asthmatics
People with a
history of allergic
rhinitis
D'Alessandro et
al., 1996;
Rotman et al.,
1983
Formaldehyde
>_3.0 ppm (P)
2.0 ppm(S)
>3.0 ppm (P)
<3.0 ppm (S)
Approx. 1 (P)
Approx. 0.67
(S)
Asthmatics
Sander et al.,
1987
Nitrogen
dioxide
1.0 ppm (P)
0.2- 0.3 ppm (P)
3-5
Asthmatics
Hackney et al.,
1978
Ozone
0.25 ppm (P)
0.12 ppm (P)
<0.24 ppm (P)
>0.25 ppm (P)
0.12 ppm (P)
0.25 ppm (P)
2
1
1
2
1
Asthmatics
People with COPD
People with
Ischemic Heart
Disease
African Americans
Gender
Horstman et al.,
1995; Bedi et al.,
1982
Sulfur dioxide
0.75-1.0 ppm
(P)
0.25 ppm (P)
3-4
Asthmatics
Stacy et al.,
1981;
Bethel et al.,
1985
Sulfuric acid
500 ug/m3 (P)
400 ug/m3 (?)
100 ug/m3 (?)
1.3
5
Adult Asthmatics
Adolescents
Asthmatics
Utell et al.,
1982; Koenig et
al., 1993;
Morrow et al.,
1994
P = Pulmonary Function Measurements, S = Symptoms of respiratory distress
PALs are not intended to protect hypersensitive individuals. In the context of PAL development,
hypersensitive individuals are those that may demonstrate responses that are extreme and
unpredictable based on the range of responses expected from normal characteristics of the
general population (e.g., specific age group susceptibilities; common diseases/maladies such as
asthma, liver disease, heart disease). In this regard, the application of UFh is consistent with
application in other risk systems (e.g., development of Oral RfD values). Differences in
susceptibility among humans based on determinants of exposure to environmental media
containing contaminants (e.g., age-dependent variation in rates of drinking water ingestion) can
be used to adjust PALs values, when it can be determined that such individuals may be present
in the populations exposed. This is covered in section 5.7.3.
5.4.3 Temporal Extrapolation Uncertainty Factor (UFt)
Once adjusted for temporal considerations, if necessary (see section 5.3.5), additional
considerations may be necessary to account for uncertainty introduced by differences between the
duration of experimental dosing or exposure and the targeted PAL duration. This can be
addressed by developing a numerical value for UFt. Historically (Dourson, 1994; Dourson et al.,
1992; U.S. EPA, 1994), a default value of 10 was applied for UFs ("Subchronic;" corresponding
to UFt, for "time" within the PALs nomenclature) when extrapolating the results from a
subchronic study to develop a reference value for the chronic, lifetime exposure duration. A value
of greater than 1 for UFt may be relevant to derivation of PAL values where experimental
29
-------
exposure durations are less than the specific PAL duration, e.g. when a critical effect and POD
value are identified from an experimental dosing or exposure duration of three days and used to
derive a 30-day PAL.
Consideration of the biological endpoint is important - guidance elsewhere in this document (and
in NRC, 2001) addresses the dependence of irritant effects (a tier 1 effect) to generally be based
on the oral dose or exposure concentration more so than on dosing or exposure duration, and
advises against duration extrapolations (either by Haber's law or the ten Berge approach) for
irritant effects. This consideration may be extended from consideration of the 8-hour duration for
AEGLs to the 24-hour duration, as well. Empirical evidence for the chemical, as well as data
describing the temporal relationship of the effect for other chemicals should be considered when
deriving a value for UFt. When the endpoint is not anticipated to increase in severity with respect
to duration of exposure (as for irritant effects), a value of 1 for UFt may be justified.
A value of less than 1 can also be applied for UFt when data from a life-time animal study are
used to derive a longer-term PAL value, which is less than a life-time human dosing or exposure.
As discussed by Jarabek (1995), in some instances, where the results from a longer-term study
may identify a POD value used to derive a PAL value for a shorter duration (e.g., a POD derived
for a tier-appropriate effect from a 28-day study used to derive a 24-hour PAL value), the dose
might represent an overly conservative point-of-departure, when applied to a shorter duration.
This may occur when the critical effect observed following the full term of the study would not be
expected during the first day of exposure (or when the effect observed in a 6-month study might
not be expected to occur as a result of a 30-day exposure). For example, gastric lesions resulting
from multiple low-level oral exposures to a corrosive agent would not be as likely following a
single dose. In this case, the value selected for UFt should not result in a dosing or exposure (or
POD) value below that expected to produce the observed effect at the magnitude of expression at
the shorter duration, given the constraints of biological plausibility (or as demonstrated by any
empirical evidence for the chemical).
Values higher than 1 should not be considered for the purposes of an uncertainty factor-based
adjustment of points of departure for experimental dosing or exposure durations longer than the
targeted PAL duration.
5.4.4 Severity Adjustment Uncertainty Factor (UFs)
This factor has a value higher than 1 when the effect chosen to operationally derive the given
POD represents a higher tier; this is primarily applicable in the development of PAL-1 values.
PAL-1 values are doses or exposures, above which changes from baseline of specific biomarkers
of physiological responses could have adverse health effects. PAL-2 and PAL-3 values are doses
or exposures for which adverse health effect, including lethality, may be expected. When dose or
concentration response data for effects pertinent to a PAL's tier are not available, the POD for an
effect of an increased severity tier may be considered, pending the application of some
adjustments. A severity UF may be appropriate if the effects at the POD are more severe (e.g.,
tier 2 effects) than those defining the targeted PAL tier (e.g., tier 1). The default UF of 10 is
generally applied if the effects at the POD are considered severe for the targeted PAL tier. For
example, when a POD value for tremors (a PAL-2 effect) is used to derive a PAL-1 value, the
value for UFs would be 10. If, however, the POD is for an effect that may be considered minor
within the suite of effects for the higher PAL tier (i.e., of minimal severity), a reduced value of 3
would be considered appropriate for UFs. A hypothetical example of this circumstance might be
observations of histologic changes in the nasal passages of inhalation-exposed rats were
characterized using vaguely descriptive terms not conveying an extent of injury rising to the level
30
-------
of a functional impairment and not indicating that grossly observable lesions occurred that, and
for which post-exposure observations necessary to confirm reversibility/irreversibility were not
available. Under these circumstances (effect not quantitatively described, without apparent
physical/gross changes, apparently not impacting function, and for which data for other chemicals
may demonstrate a likelihood of reversibility), such an effect might be justified as a tier 2 effect of
minimal severity.
When the critical effect is consistent with the PAL tier definition (e.g., a nonlethal effect is
selected as the POD for PAL-3 development), a value of 1 should be applied. An analogy to the
contemporary definition and application of UFl (the LOAEL to NOAEL uncertainty factor) as
applied in RfD or RfC derivation may be appropriate
Alternately, PAL values for a given tier may be derived on the basis of values for a lower tier, on
a case-by-case basis. This involve a discussion regarding the dose- or exposure-response
relationship (i.e., steepness of the dose- or exposure-response curve) if such data are available.
5.4.5 Modifying Factor (MF)
The MF value is specific to the PAL value being derived; the same value for the MF is not applied
to all PAL values for the chemical, unless uniquely justified for each value. The modifying factor
(MF, with possible values of 1, 3, or 10) is intended to account for scientific uncertainties in the
study or database for the PAL value being derived that are not accounted-for in the application of
UFs, when those uncertainties exist. When those uncertainties do not exist, or have been
accounted-for by the application of UFs, the value for MF is set to 1. The value for the MF
depends on professional judgment and chemical-specific science-based assessment. For example,
when data deficiencies such as poorly or non-defined critical effects, absence of dose- or
exposure-response information, stand-alone toxicity data (e.g., a single LC50 in only one species),
or toxicity data for an extremely limited dosing or exposure duration (e.g., 5 minutes) are
identified for a targeted PAL value, science-based judgment should be used to evaluate their
significance and assign a MF value of 3 or 10.
In PAL application, the MF should be distinguished from the Database Uncertainty Factor (UFd)
as applied (e.g.) in the derivation of RfD and RfC values. Because RfD and RfC values are
developed to ensure the safety of the entire population under an assumed continuous exposure for
a lifetime, and because these values are often used as the basis for regulatory standards, the
toxicity database should be as complete as possible. Thus, quantitation of the value for UFd
should include a consideration of the completeness of the toxicity database to address specific
health endpoints (e.g., reproductive toxicity, neurotoxicity, developmental toxicity). In contrast,
emergency exposures may involve only a small number of individuals, dosed or exposed for less
than a lifetime. In addition, because PAL values are not intended to serve any regulatory function,
the adoption of nomenclature for another risk system without also adopting the requirements
determining its numerical value would induce confusion.
5.4.6 Multiplication of Uncertainty Factors
UF values are multiplied together such that 10 x 10 = 100. In the development of PAL values,
when a default value of 10 is reduced by half, it is reduced to one-half an order of magnitude,
or to 3.16. This is rounded to a value of "3" for ease of presentation. The multiplication of
two values of one-half order of magnitude, each presented as "3," results in a product of "10,"
the result of multiplying two values of a half order of magnitude. This approach will be used
in the application of UFs for development of PAL values and is documented as an accepted
31
-------
practice in risk assessment (Dourson et al., 1996; Renwick and Lazarus, 1998; NRC, 2001).
Also, for simplicity, 3x10 = 30.
5.5 Computation of PAL Values
PAL values are computed following identification of the critical effect, identification of the point
of departure, duration-adjustment of the point of departure (when necessary), and identification
and application of values for the uncertainty factors and modifying factors. PAL values are
computed according to:
PAL Value = duration-adjusted POD / (UF x MF)
If the duration-adjusted POD value is 71 mg/kg, and if UFa = 3; UFh = 3, UFt = 1; UFs = 1,
and MF = 1, then the PAL value would be derived as:
71 mg/kg-d / (3 x 3 x l x l x l) = 71 mg/kg-d/10 = 7.1 mg/kg-d
5.6 Weight-of-Evidence and Confidence
PALs should be developed within the context of the overall data pertinent to the specific PAL
tier. This will entail, in part, selection of a critical effect appropriate for the given PAL tier as
well as an appropriate POD. The robustness of the data will ultimately affirm
assumptions/judgments utilized in PAL development, and also dictate areas of uncertainty.
Much of the weight-of-evidence analysis will be implicit in the course of PAL development.
Focal areas may include species variability, dose- or exposure-response relationships, relation
of PALs to other dosing or exposure duration data, continuum of effects, mechanism of action,
etc. Such analysis allows assessment of reasonableness of the PALs. For example, overly
conservative acute exposure PALs will be recognizable as such if they are lower than long-
term doses or exposures causing notably less severe effects (see section 5.10).
Supporting studies are used to support the toxicological findings and values obtained from the
key study. Findings from supporting studies may also be incorporated into "weight-of-
evidence" considerations and may be used to justify the values for some uncertainty factors.
The evaluation of data deficiencies and research needs is also a reflection of the confidence in
the PAL values. Those making use of PAL values (emergency planners, emergency responders,
etc.) have expressed a critical need for dosing or exposure values and that, in some cases, values
with low confidence are better than no value. PAL values will be developed only when data and
scientific judgment warrant the derivation of values. Where data sets are limited, scientific
judgment may be used to avoid a no-value situation. Logically, PAL values with no confidence
will not be developed. Confidence in PAL values will generally be reflected by the magnitude of
uncertainty adjustments and the rationales for these adjustments. Assessment and comments on
confidence should accompany the PAL values.
5.7 Presentation of Values
PAL values should be presented in units of dose or concentration pertinent to oral and inhalation
exposures, respectively. Because the only exposure medium for inhalation exposures is air, PAL
values for inhalation exposures should be presented in units of ppm and mg/m3. However,
because oral exposures may be based on the ingestion of environmental media such as soil or
drinking water, oral PAL doses (mg/kg-d) can be translated into units of concentration in
32
-------
environmental media on the basis of reliable data and methods.
Oral PAL values may be applicable to several environmental media, including drinking water and
soil ingestion. The conversion from an oral PAL dose (mg/kg-d) into corresponding media
concentrations (e.g., mg chemical/L drinking water, |ig chemical/kg soil) can be accomplished
using values for key parameters available from reliable sources. Selection and specification of the
data used for oral ingestion parameters should be in accord with general applicability of PAL
values and on the basis of the demographics of the population anticipated to be exposed. Values
for drinking water ingestion rates and soil ingestion rates are available from a variety of sources
(e.g., U.S. EPA, 2011b).
5.7.1 Level of Precision
All PAL values will be expressed as two significant figures. As necessary, rounding toward
[+infinity] (nearest up) will be applied (e.g., 2.657 will become 2.7; 0.00244 will become
0.0024; 0.00245 will become 0.0025). Rounding of value to the required two significant
figures will be limited to the final PAL value. To retain precision, rounding should not be
performed within the calculation cascade. From the example above, the oral PAL value of
248.5 mg/L would be rounded to 250 mg/L.
5.7.2 No PAL Value Established
If pertinent data for a chemical are insufficient and no surrogate chemical, data, or PAL values
can be identified, PAL values will not be derived. In this case, the reason(s) for not developing
a PAL value will be presented and the term "Not Recommended" (NR) will be presented.
5.7.3 Values for Susceptible Population Groups
The critical effect is communicated for PAL values; when the critical effect is demonstrated in
human population groups presumed to be susceptible, or in test animals assumed to represent the
characteristics of the human population group presumed to be susceptible (e.g., developing fetal
animals or young animals), this should be communicated. Should a child-specific oral PAL value
be required for a drinking water exposure scenario, the oral PAL value (in units of mg/kg-d) can
be translated into a drinking water concentration by the user on the basis of reliable information
such as in the Exposure Factors Handbook (U.S. EPA, 201 lb).
5.8 Surrogate Chemicals, Surrogate Data, Surrogate PAL Values
For some chemicals of concern, insufficient data may preclude development of some or
possibly all PAL values. PAL values may be based upon professional judgment with the
assumption that such judgment will be preferable to a no-guidance situation for responders and
end-users. A conscientious effort should be made to develop PALs using structure-activity
relationships or other scientifically defensible analogies such as accepting dose response data
or PAL values for chemicals justified as surrogates (e.g., Wang et al., 2012) while applying
appropriate science-based rigor. Data that may be relevant to PAL development may include
chemical-specific information regarding metabolism and disposition, interaction with other
chemicals, unique physical-chemical properties, and degradation products. Quantitative
structure activity relationship (QSAR) models may also be used to augment sparse data on a
specific chemical. The development of newer predictive toxicity tools and data such as those
available through high throughput screening offers additional benefit in refining the accuracy
of predictive efforts based on structural similarity among chemicals (Berggren et al., 2015).
33
-------
NRC (2001) also suggested quantifying an AEGL-2 value by selecting a one-third reduction of
the AEGL-3 value (not the AEGL-3 POD). For example, if data are unavailable with which to
empirically derive a value for PAL-2 and the dose- or exposure-response relationship for the
chemical of concern is steep (i.e., marked change in response severity with a small change in
exposure concentration or dose), the PAL-2 value may be estimated by a one-third reduction of
the respective PAL-3 (not by basing the PAL-2 on a one-third reduction of the PAL-3 POD;
see NRC, 2001, page 43). Assessment of the steepness of the dose- or exposure-response
curve would require dose or exposure response data for minor effects (less than tier 2 severity)
as well as lethality data. Such relational data would be necessary to justify using a fractional
reduction of a PAL-3 value to estimate the PAL-2 value for the same duration. This approach
may also be justified for derivation of PAL-1 values on a case-by-case basis. Justification
should include consideration of the nature, strength, and consistency of the data describing the
type of responses observed for other PAL tiers at the duration of interest; and the type of
responses observed for other durations at the tier of interest.
5.9 Carcinogenicity as a Critical Effect
Carcinogenicity may be most characteristic of a tier 2 response; the likelihood of a carcinogenic
response decreases and the uncertainty in predictions of cancer risk increase with decreased
dosing or exposure durations (NRC, 2001). However, it is possible that cancer can result from
short or single exposures to some carcinogens. Historically, cancer risk has been evaluated
based upon continuous, life-time dosing or exposure in laboratory animal studies or from clinical
or epidemiologic studies of long-term human exposures. Although the NRC (1993, 2001) has
identified cancer as a potential health effect possibly associated with short-term inhalation
exposures to certain chemicals, no U.S. federal or state regulatory agency has established
regulatory limits for single short-term (<24 hours) exposure based upon carcinogenicity (NRC,
2001). Extrapolating lifetime cancer risks to shorter durations is possible (e.g., for AEGLs;
NRC, 2001) and may be valuable in some emergency response planning situations, but doing so
requires the communication of certain caveats and the acknowledgement of some inherent
uncertainties.
Additional considerations include the possibility of unnecessary risk and resource allocations
associated with some health-protective emergency response actions (e.g., evacuations) that may
be undertaken on the basis of estimations of risk embodying uncertainty and inherent health
conservatism, as well as the potential for realizing a relatively low predicted incidence of effects
among a potentially small exposed population. It should be understood that cancer risk
assessment is an inherently conservative endeavor, usually including linear low-dose
extrapolation methods in the absence of confirmatory data. This is done both for chemicals
whose carcinogenic mode of action is based on a direct interaction with DNA and for chemicals
for which the mode of action is not known to involve a direct interaction with DNA. Also,
because most of the available cancer risk values have been developed on the basis of findings
from chronic dosing or exposure of experimental animals, the carcinogenic risk from shorter
durations of oral dosing or inhalation exposure has received less attention, and so is less certain
than that from chronic dosing or exposure.
While the extrapolation of lifetime cancer risks to durations of dosing or exposure much shorter
than a lifetime can be completed in the estimation of emergency response guidance levels (NRC,
2003), the level of uncertainty in such risk estimates generally increases with decreased dosing
or exposure durations. For PALs, this means that the level of uncertainty in carcinogenic risk
for PAL durations will increase with decreasing PAL duration. In fact, in considering the
34
-------
application of cancer risk assessment approaches in the development of AEGL values, NRC
(2001) seems to advise caution in developing risk estimates that may impart a higher level of
conservatism (e.g., as might occur when based on cancer risk). With respect to sulfur mustard,
NRC (NRC, 2003) indicated, "The use of excess-cancer-risk estimates in setting AEGL values
was precluded by the uncertainties involved in assessing excess cancer risk following a single
acute exposure of 8 hr or less, the relatively small population dosed or exposed during an
accidental or other emergency situations, and the potential risks associated with evacuations."
Whether cancer risk should receive additional attention relative to identifying a POD for PALs
development should be revisited as cancer risk methods advance into this area and/or when data
from carcinogenicity studies reporting the results of acute, single-dose or short-term dosing or
exposure of test species or humans become available which support such an evaluation.
A quantitative carcinogenicity assessment should be performed only when a U.S. EPA cancer
classification and slope factor/unit risk are available. Data from reliable sources including
reports from controlled cancer bioassays in laboratory animals, genotoxicity studies, human
epidemiologic studies, and human accidental dosing or exposure reports will be evaluated.
Criteria typical for evaluation of animal bioassays (e.g., numbers, species, and strains of animals,
dose groups, dosing or exposure duration, tumor type/incidences) and epidemiologic studies
(methods, confounders, statistical analysis of responses, etc.) will be applied to help determine
the relevance of specific studies. To assess carcinogenic risk potential, NRC guidance (NRC,
1986; 1993; 2001) and U.S. EPA guidelines (U.S. EPA, 2005a, 2005b) should be followed.
Although a cancer assessment should be conducted for completeness if a U.S. EPA cancer
classification and slope factor or unit risk value are available, it should be presented in an
appendix to the Technical Summary Document. Any PAL value derived based on
carcinogenicity should not be recommended as the quantitative basis for establishing a PAL
value, given the intended application of PAL values as decision aids during and following
emergency events. Rather, the observed irreversible effect or effect impairing the ability to
escape a contaminated source or atmosphere should be used as the basis for PAL-2 values.
5.10 Final Adjustment of PAL Values
Although the derivation of PAL values follows well-developed guidelines for data analysis,
selection of an appropriate key study, POD, UF application, and dose or exposure
concentration-duration adjustments, resulting PALs may be occasionally inconsistent with
known human experience. This may occur for several reasons including the selection of UF
values; the differential basis of PAL values for a given duration or severity level on
observations in humans versus animals; the extent to which experimental observations from one
duration to another are adjusted, normalized or justified for application to a PAL value of
another duration, etc. Science-based judgment will be applied to justify a quantitative
adjustment in the final PAL value(s). It is not recommended that this adjustment be applied
through manipulation of the UF. Rather it should be clearly stated as a final change in the PAL
value(s) based upon a comparison of the PAL with human experience information and an
overall evaluation of the reasonableness of the original and adjusted PAL value(s).
If the PAL is found to be in conflict with available human data, adjust the PAL and clearly
explain the basis of the adjustment. Essentially, this might be necessary where accepted
methodology is applied but results in unrealistic values. As cautioned by NRC (2001), the
establishment of overly conservative guideline (AEGL) values may result in health protective
activities and actions that may, themselves, have consequences. Overly protective PALs may
35
-------
drain resources and create undue hardships on affected communities such as increased
incidence of traffic accidents and injuries during unnecessary evacuations and delayed reentry.
6 Chemical Considerations
6.1 Complex Mixtures/Concurrent Exposure Issues
It is likely that situations necessitating the use of a PAL may involve concurrent exposure to other
chemicals, exposures to simple mixtures of chemicals (mixtures with few components) or
exposure to complex mixtures (mixtures with many components). It may be assumed that in
situations where exposure involves more than one specific chemical, the most protective PAL
(i.e., chemical with the lowest PAL) will dictate on-scene responses. Considerations may include
the specific mechanism of action of the individual components and the persistence of the
chemicals in the environment. The evaluation of additive, potentiated, or synergistic effects may
be important. While the management of risks associated with concurrent chemicals might depend
more heavily on the expertise of on-scene risk managers, any development of PALs for a chemical
mixture not based on empirical data for the mixture itself should involve personnel with mixtures
toxicology skills and/or mixtures risk assessment skills.
6.2 Degradation Products
For some chemicals, and in some circumstances, it may be valuable to also consider the
toxicological contribution of degradation products in the assessment (Wang et al., 2012).
Consideration should be given to the availability of data for the degradation product regarding
its identification and quantitation, the environmental factors affecting its formation, its rate of
formation, the nature of its effects, and the relationship of those effects to dose. Knowledge of
the release time, rate of formation and toxic properties of degradation products represent
important sources of information when estimating the value of assessing the risks of
degradation products. For example, if degradation products are formed slowly over time,
dosing or exposure to degradation products may not be a priority concern for acute or short-
duration dosing or exposure, but might be a concern when dosing or exposure involves a
previous release (e.g., during remediation of a previously contaminated site).
If it is known that degradation products will possess toxicity similar to or greater than the parent
compound, development of PAL values for these products will be independent of the parent
chemical. PAL documents for degradation products and the parent chemical should be cross-
referenced. In addition to relative toxicity to the parent compound, attention (in the context of a
combined dosing or exposure i.e., a mixtures toxicity issue) should be given to the relative
contribution of any noteworthy degradation product, as well as to its environmental persistence,
to the overall response (e.g., a degradation product of significant toxicity may occur in only
very small quantities).
7 Research Needs/Recommendations
Data deficiencies will be characterized as precisely as possible for every PAL document where it
was not possible to derive scientifically defensible PALs. Based upon the identified deficiencies,
recommendations will be made regarding generation of required data and studies that would
improve the quality and reduce uncertainties in the PAL document. This may include
suggestions/input regarding experimental protocols for specific type of studies.
36
-------
8 References
AIHA (American Industrial Hygiene Association). 2016. The AIHA 2016 Emergency Response
Planning Guidelines (ERPG) and Workplace Environmental Exposure Level (WEEL)
Guidelines Handbook. AIHA, Fairfax, VA.
ATSDR (Agency for Toxic Substances and Disease Registry). Minimal Risk Levels (MRL).
2017. Available at: https://www.atsdr.cdc.gov/mrls/index.asp.
Barnes DG and Dourson ML. 1988. Reference dose (RfD): Description and use in health risk
assessments. Regul. Toxicol. Pharmacol. 8:471-486.
Bedi JF, Horvath SM, and Folinsbee LJ. 1982. Human exposure to sulfur dioxide and
ozone in high temperature-humidity environment. Am. Ind. Hyg. Assoc. J. 43: 26-30.
Berggren E, Amcoff P, Benigni R, Blackburn K, Carney E, Cronin M, Deluyker H, Gautier F,
Judson RS. Kass GEN, Keller D, Knight D, Lilienblum W, Mahoney C, Rusyn I, Schultz
T, Schwarz M, Schuurmann G, White A, Burton J, Losits AM, Munn S, and Worth A.
2015. Chemical Safety Assessment Using Read-Across: Assessing the Use of Novel
Testing Methods to Strengthen the Evidence Base for Decision Making. Environ. Health
Perspect. 123: 1232-1240.
Bethel RA, Sheppard D, Geffroy B, Tam E, Nadel JA, and Boushey HA. 1985. Effect
of 0.25 ppm sulfur dioxide on airway resistance in freely breathing, heavily exercising,
asthmatic subjects. Am. Rev. Respir. Dis. 131:659-661.
Boobis AR, Doe JE, Heinrich-Hirsch B, Meek ME, Munn S, Ruchirawat M,
Schlatter J, Seed J, and Vickers C. 2008. IPCS framework for analyzing the relevance
of a noncancer mode of action for humans. Crit. Rev. Toxicol. 38, 87-96.
Boobis AR, Cohen SM, Dellarco, V, McGregor D, Meek ME, Vickers C, Wilcocks D,
and Farland W. IPCS framework for analyzing the relevance of a cancer mode of
action for humans. Crit. Rev. Toxicol. 36, 781- 796.
Craig DK. 1992. Toxic chemical hazard classification and risk acceptance guidelines for use
in D.O.E. facilities (Unclassified). WSRC-MS-92-206 Rev 2. Westinghouse M&O
Nuclear Facility Safety Committee, Subcommittee on Non-Radiological Risk Acceptance
Guideline Development.
Craig DK and Lux CR. 1998. Methodology for Deriving Temporary Emergency Exposure
Limits (TEELs). WSRC-TR-98-00080. Aiken, SC: Westinghouse Savannah River.
D' Alessandro A, Kuschner W, Wong H, Boushey HA, and Blanc PD. 1996.
Exaggerated responses to chlorine inhalation among persons with nonspecific airway
hyperreactivity. Chest 109:331-337.
de Jong RH and Eger EI. 1975. MAC expanded: AD50 and AD95 values of common
inhalation anesthetics in man. Anesthesiology 42:384-389.
37
-------
Delic JI, Lilly PD, MacDonald AJ, and Loizou GD. 2000. The utility of PBPK in the
safety assessment of chloroform and carbon tetrachloride. Regul. Toxicol. Pharmacol.
32:144-155.
DOE (Department of Energy). 2008. Temporary emergency exposure limits for chemicals:
Methods and practice. DOE-HDBK-1046-2008. Washington, DC.
Donohue JM, and Lipscomb JC. 2002. Health advisory values for drinking water
contaminants and the methodology for determining acute exposure values. Sci. Total
Environ. 288: 43-49.
Dourson ML. 1994. Methods for estimating oral reference doses (RfDs). In Risk Assessment
of Essential Elements (W. Mertz, C.O. Abernathy, and S.S. Olin, Eds.), pp. 51-61.
Washington, DC: ILSI Press.
Dourson ML and Stara JF. 1983. Regulatory history and experimental support of
uncertainty (safety) factors. Regul. Toxicol. Pharmacol. 3(3):224-238.
Dourson ML, Knauf LA, and Swartout JC. 1992. On reference dose (RfD) and its
underlying toxicity data base. Toxicol. Ind. Health 8(3): 171-189.
Dourson ML, Felter SP, and Robinson D. 1996. Evolution of science-based uncertainty
factors in noncancer risk assessment. Regul. Toxicol. Pharmacol. 24:108-120.
Fowles JR, Alexeeff GV, and Dodge D. 1999. The use of benchmark dose methodology
with acute inhalation lethality data. Regul. Toxicol. Pharmacol. 29:262-278.
Gargas ML, Burgess RJ, Voisard DE, Cason GH, and Andersen ME. 1989. Partition
coefficients of low-molecular-weight volatile chemicals in various liquids and tissues.
Toxicol. Appl. Pharmacol. 98: 87-99.
Gregory GA, Eger EI 2nd, and Munson ES. 1969. The relationship between age and
halothane requirement in man. Anesthesiology 30: 488-491.
Hackney JD, Thiede FC, Linn WS, Pedersen EE, Spier CE, Law DC, and Fischer DA.
1978. Experimental studies on human health effects of air pollutants. IV. Short-term
physiological and clinical effects of nitrogen dioxide exposure. Arch. Environ. Health
33:176-181.
Holsapple MP and Wallace KB. 2008. Dose response considerations in risk assessment—
An overview of recent ILSI activities. Toxicol. Lett. 180, 85-92.
Horstman DH, Ball BA, Brown J, Gerrity T, and Folinsbee LJ. 1995. Comparison of
pulmonary responses of asthmatic and nonasthmatic subjects performing light exercise
while exposed to a low level of ozone. Toxicol. Ind. Health 11: 69-85.
Jarabek AM. 1995. Consideration of temporal toxicity challenges current default
assumptions. Inhalation Toxicol. 7:927-946.
Kerlin R, Bolon B, Burkhardt J, Franke S, Greaves P, Meador V, and Popp J. 2016.
Scientific and Regulatory Policy Committee: Recommended ("best") practices for
38
-------
determining, communicating and using adverse effect data from nonclinical studies.
Toxicologic Pathology 44:147-162.
Koenig JQ, Dumler K, Rebolledo V, Williams PV, and Pierson WE. 1993. Respiratory
effects of inhaled sulfuric acid on senior asthmatics and nonasthmatics. Arch. Environ.
Health. 48:171-175.
Litchfield JT Jr, and Wilcoxon F. 1949. A simplified method of evaluating dose-effect
experiments. J. Pharmacol. Exp. Ther. 96:99-113.
Morrow PE, Utell MJ, Bauer MA, Spears DM, and Gibb FR. 1994. Effects of near
ambient levels of sulfuric acid aerosol on lung function in exercising subjects with asthma
and chronic pulmonary disease. In: Dodgson, J., McCallum, R. I., eds., Inhaled particles
VII: Proceedings of an International Symposium. September 1991; Edinburgh, UK. Ann.
Occup. Hyg. 38 (Supplement 1): 933-938.
NRC (National Research Council). 1986. Criteria and methods for preparing Emergency
Exposure Guidance Level (EEGL), Short-Term Public Emergency Guidance Level
(SPEGL), and Continuous Exposure Guidance Level (CEGL) documents. Appendix
F. Washington, DC: National Academy Press.
NRC (National Research Council). 1993. Guidelines for developing community emergency
exposure levels for hazardous substances. Washington, DC: National Academy Press.
NRC (National Research Council). 2001. Standing operating procedures for developing acute
exposure guideline levels for hazardous chemicals. Washington, DC: National Academy
Press.
NRC (National Research Council). 2003. Acute Exposure Guideline Levels for Selected
Airborne Chemicals, Volume 3. Washington, DC: National Academy Press.
NRC (National Research Council). 2007. Toxicity testing in the 21st century: A vision and a
strategy. Washington, DC, National Academy Press.
OECD (Organization for Economic Cooperation and Development). OECD Guidelines for the
testing of chemicals.
http://www.oecd.org/chemicalsafety/testing/oecdguidelinesforthetestingofchemicals.htm.
Accessed June, 2017.
Palazzi X, Burkhardt JE, Caplain H, Dellarco V, Fant P, Foster JR, Francke S, Germann P,
Groters S, Harada T, Harleman J, Inui K, Kaufmann W, Lenz B, Nagai H, Pohlmeyer-
Esch G, Schulte A, Skydsgaard M, Tomlinson L, Wood CE, and Yoshida M. 2016.
Characterizing "adversity" of pathology findings in nonclinical toxicity studies: Results
from the 4th ESTP International Expert Workshop. Toxicol. Pathol. 44:810-824.
Paustenbach DJ, Carslon GP, Christian JE, and Born GS. 1986a. A comparative study of
the pharmacokinetics, of carbon tetrachloride in the rat following repeated inhalation
exposures of 8 and 11.5 hr/day. Fundam. Appl. Toxicol. 6:484-497.
Paustenbach DJ, Christian JE, Carslon GP, and Born GS. 1986b. The effect of an 11.5-hr/day
39
-------
exposure schedule on the distribution and toxicity of inhaled carbon tetrachloride in the
rat. Fundam. Appl. Toxicol. 6: 472-483.
Paustenbach DJ, Clewell HJ III, Gargas ML, and Andersen ME. 1988. A
physiologically based pharmacokinetic model for inhaled carbon tetrachloride. Toxicol.
Appl. Pharmacol. 96:191-211.
Renwick AG and Lazarus NR. 1998. Human variability and noncancer risk assessment - an
analysis of the default uncertainty factor. Regul. Toxicol. Pharmacol. 27(1 Pt 2):3-20.
Rinehart WE and Hatch T. 1964. Concentration-time product (CT) as an expression of dose
in sublethal exposures to phosgene. Am. Ind. Hyg. Assoc. J. 25:545-553.
Rotman HH, Fliegelman MJ, Moore T, Smith RG, Anglen DM, Kowalski CJ, and
Weg JG. 1983. Effects of low concentration of chlorine on pulmonary function in
humans. J. Appl. Physiol. 54:1120-1124.
Rusch GM. 1993. The history and development of Emergency Response Planning
Guidelines. J Hazard. Mater. 33:192-202.
Rusch GM, Bast CB, and Cavender FL. 2009. Establishing a point of departure for risk
assessment using acute inhalation toxicology data. Regul. Toxicol. Pharmacol. 54:247-
255.
Sander LR, Green DL, Chapman MD, and Kulle TJ. 1987. Acute pulmonary response
of asthmatics to 3.0 ppm formaldehyde. Toxicol. Ind. Health. 3: 569-578.
Stacy RW, House D, Friedman M, Hazucha M, Greene J, Raggio L, and Roger LJ. 1981.
Effects of sulfur dioxide on pulmonary function parameters of normal human subjects.
Arch. Environ. Health. 36: 172-178.
Sonich-Mullin C, Fielder R, Wiltse J, Baetke K, Dempsey J, Fenner-Crisp P, Grant D,
Hartley M, Knaap A, Kroese D, Mangelsdorf I, Meek E, Rice JM, and Younes M. 2001.
IPCS Conceptual Framework for evaluating a mode of action for chemical carcinogenesis.
Regul. Toxicol. Pharmacol. 34:146-152.
ten Berge WF, Zwart A, and Appelman LM. 1986. Concentration-time mortality response
relationship of irritant and systemically acting vapours and gases. J. Hazard.
Mater. 13(3):301-309.
U.S. EPA (United States Environmental Protection Agency). 1985. Reference values for risk
assessment. Cincinnati, OH: Environmental Criteria and Assessment Office,
U.S. Environmental Protection Agency. SRC TR 85-300.
U.S. EPA (United States Environmental Protection Agency). 1987. Final Rule: Extremely
hazardous substance list and threshold planning quantities; Emergency planning and release
notification requirements rule (52 FR 13378). Federal Register 52(77): 13378-13410.
U.S. EPA (United States Environmental Protection Agency). 1988a. Interim methods for
development of inhalation reference doses. EPA/600/8-88/066F. Cincinnati, OH; Research
40
-------
Triangle Park, NC: Environmental Criteria and Assessment Office, Office of Health and
Environmental Assessment, U.S. Environmental Protection Agency.
U.S. EPA (United States Environmental Protection Agency). 1988b. General quantitative risk
assessment guidelines for noncancer health effects. Technical Panel on Risk Assessment
Guidelines for Noncancer Health Effects. Draft. ECAO-CIN-538.
U.S. EPA (United States Environmental Protection Agency). 1989b. Risk assessment guidance
for Superfund (RAGS), Vol 1. Human health evaluation manual (part A).
EPA/540/1-89/002. Washington, DC: Office of Emergency and Remedial Response,
U.S. Environmental Protection Agency.
U.S. EPA (United States Environmental Protection Agency). 1993. Reference Dose (RfD):
Description and Use in Health Risk Assessments. Washington, DC: U.S. EPA, National
Center for Environmental Assessment. Available at: https://www.epa.gov/iris/reference-
dose-rfd-description-and-use-health-risk-assessments.
U.S. EPA (United States Environmental Protection Agency). 1994. Methods for derivation of
inhalation reference concentrations and applications of inhalation dosimetry. EPA/600/8-
90/066F. Washington, DC: Office of Health and Environmental Assessment Office, Office
of Research and Development, U.S. Environmental Protection Agency, October.
U.S. EPA (United States Environmental Protection Agency). 1995. The use of the benchmark
dose approach in health risk assessment. EPA/630/R-94/007. Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
U.S. EPA (United States Environmental Protection Agency). 1997. Exposure factors
handbook. EPA/600/8-89/043. Washington, DC: U.S. EPA, Office of Health and
Environmental Assessment.
U.S. EPA (United States Environmental Protection Agency). 2003. A summary of general
assessment factors for evaluating the quality of scientific and technical information.
EPA/100/B-03/001. Washington DC: Science Policy Council, U.S. Environmental
Protection Agency.
U.S. EPA (United States Environmental Protection Agency). 2005a. Guidelines for
carcinogen risk assessment. EPA/630/P-03/001F. Washington, DC: Office of Research
and Development, U.S. Environmental Protection Agency.
U.S. EPA (United States Environmental Protection Agency). 2005b. Supplemental guidance
for assessing susceptibility from early-life exposure to carcinogens. EPA/630/R-03/003F.
Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Development.
U.S. EPA (United States Environmental Protection Agency). 2006a. Protection for subjects in
human research: Final rule. 40CFR Parts 9 and 26. Federal Register, Vol. 71, no. 24:
6137-6176. Monday, February 6, 2006.
U.S. EPA (United States Environmental Protection Agency). 2006b. Approaches for the
application of physiologically based pharmacokinetic (PBPK) models and supporting data
in risk assessment. EPA/600/R-05/043F. Washington, DC: U.S. EPA, National Center for
41
-------
Environmental Assessment.
U.S. EPA (U.S. Environmental Protection Agency). 2007. Benchmark dose software draft beta
version 1.4.1. National Center for Environmental Assessment Office of Research
and Development.
U.S. EPA (U.S. Environmental Protection Agency). 201 la. Recommended use of body weight
3/4 as the default method in derivation of the oral reference dose. EPA/100/R-l 1/001.
Final. Washington, DC: Office of the Science Advisor, Risk Assessment Forum, U.S.
Environmental Protection Agency.
U.S. EPA (U.S. Environmental Protection Agency). 201 lb. Exposure factors handbook 2011
edition (final report). Washington, DC: U.S. Environmental Protection Agency.
EPA/600/R-09/052F.
U.S. EPA (United States Environmental Protection Agency). 2012a. 2012 Edition of the
Drinking Water Standards and Health Advisories. EPA 822-S-12-001. Office of Water,
U.S. Environmental Protection Agency Washington, DC.
https://nepis.epa.gOv/Exe/ZyPDF.cgi/P 100N01H.PDF?Dockey=P 100N01H.PDF
U.S. EPA (United States Environmental Protection Agency). 2012b. Guidance for Evaluating
and Documenting the Quality of Existing Scientific and Technical Information: Addendum
to: A Summary of General Assessment Factors for Evaluating the Quality of Scientific and
Technical Information. Science and Technology Policy Council, U.S. Environmental
Protection Agency, Washington, DC 20460.
https://www.epa.gov/sites/production/files/2015-05/documents/assess3.pdf
U.S. EPA (United States Environmental Protection Agency). 2012c. Benchmark dose
technical guidance. EPA/100/R-l2/001. Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum.
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/R-14/002F. Washington, DC: U.S. EPA, Office of the Science
Advisor, Risk Assessment Forum.
U.S. EPA (United States Environmental Protection Agency). 2015. Benchmark Dose Software
(Version 2.6.0.1.). Washington, DC: Environmental Protection Agency.
https://www.epa.gov/bmds
U.S. EPA (U.S. Environmental Protection Agency). 2017. Provisional Peer Reviewed Toxicity
Values for Superfund (PPRTV). Office of Superfund Remediation and Technology
Innovation, U.S. Environmental Protection Agency, Washington, DC 20460.
http s: //hhpprtv. ornl. gov/pprtvuser. html
Utell MJ, Morrow PE, and Hyde RW. 1982. Comparison of normal and asthmatic
subjects' responses to sulfate pollutant aerosols. Ann. Occup. Hyg. 26: 691-697.
Villeneuve DL, Crump D, Garcia-Reyero N, Hecker M, Hutchinson TH, LaLone CA,
Landesmann B, Lettieri T, Munn S, Nepelska M, Ottinger MA, Vergauwen L, and
42
-------
Whelan M. 2014. Adverse outcome pathway (AOP) development I: strategies and
principles. Toxicol Sci. 142:312-320.
Wang NC, Zhao Q, Wesselkamper SC, Lambert JC, Petersen D, and Hess-Wilson JK. 2012.
Application of computational toxicological approaches in human health risk
assessment. I. A tiered surrogate approach. Regul Toxicol Pharmacol. 63:10-19.
WHO (World Health Organization), International Programme on Chemical Safety. 2005.
Chemical-specific adjustment factors for interspecies differences and human variability:
guidance document for use of data in dose/concentration-response assessment. (IPCS
Harmonization Project Document No. 2). Geneva, Switzerland: World Health
Organization, http://www.inchem.org/documents/harmproj/harmproj/harmproj2.pdf
WHO (World Health Organization), International Programme on Chemical Safety. 2010.
Characterization and application of physiologically based pharmacokinetic models in risk
assessment. (IPCS harmonization project document no. 9.) Geneva, Switzerland: World
Health Organization.
Young RA, Bast CB, Wood CS, and Adeshina F. 2009. Overview of the Standing Operating
Procedure (SOP) for the development of Provisional Advisory Levels (PALs). Inhalation
Toxicol 21(S3): 1-11.
43
-------
This page left intentionally blank
-------
This page left intentionally blank
-------
vvEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |