Summary

The US Environmental Protection Agency (EPA) Integrated Risk Information System (IRIS) program develops toxicologic assessments of environmental contaminants. IRIS assessments provide hazard identification and dose–response assessment information. The information is then used in conjunction with exposure information to characterize risks to public health and may be used in risk-based decision-making, in regulatory actions, and for other risk-management purposes. Since the middle 1990s, EPA has been in the process of updating the IRIS assessment of inorganic arsenic. Arsenic is a naturally occurring element in the environment that can occur in different forms. Inorganic arsenic refers to a set of arsenic-containing molecules that includes elemental arsenic and many other molecular structures that involve arsenic in combination with elements other than carbon. Organic arsenic consists of a broad set of arsenical compounds that include carbon in their structures; they include the relatively simple monomethyl arsenic and dimethyl arsenic that are metabolites of inorganic arsenic in humans and more complex molecules that are used as herbicides, pesticides, or animal feed additives. EPA’s toxicologic assessment will focus on inorganic arsenic.

The development of the toxicologic assessment of inorganic arsenic is a considerable undertaking for the IRIS program. Much research has been performed on the chemical, and improved approaches to performing assessments have been recommended. Although the multiple challenges are clear given the history of the assessment, there are also opportunities to take advantage of the rich data on inorganic arsenic to apply evaluation tools and integrative approaches so as to conduct the assessment in a transparent manner. Many of the improved approaches were recommended by previous National Research Council committees that evaluated other chemicals, and it is clear from EPA’s draft conceptual model and analysis plans that the agency plans to incorporate the recommendations. The task of the present committee was to focus on recommendations specific to the assessment of inorganic arsenic. (A different National Research Council committee is reviewing EPA’s initiatives to improve the overall process and quality of IRIS assessments.)

In response to a congressional mandate for an independent review of the IRIS assessment of inorganic arsenic, EPA requested that the National Research Council convene a committee to conduct a two-phase study. In the first phase (the focus of the present report), the committee was to organize a workshop to evaluate critical scientific issues in assessing cancer and noncancer effects of oral exposure to inorganic arsenic and offer recommendations on how the issues could be addressed in EPA’s IRIS assessment. The second phase of the study will begin after EPA completes its draft assessment; it will involve a review of the draft assessment to determine whether the committee’s recommendations were appropriately addressed and also reflect recommendations for improving IRIS assessments made in other National Research Council reports. As part of the first phase, a workshop was held on April 4-5, 2013, to obtain input on critical aspects of interpreting and applying the scientific information on inorganic arsenic for the purposes of hazard identification and dose–response analysis. Having been informed by the workshop and its own preliminary survey of the scientific literature, the committee offers the following recommendations for performing the IRIS assessment as outlined in Figure 1 and described in Box 1. These recommendations are specific to inorganic arsenic and may not be applicable to the assessment of other chemicals.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 3
Summary The US Environmental Protection Agency (EPA) Integrated Risk Information System (IRIS) pro- gram develops toxicologic assessments of environmental contaminants. IRIS assessments provide hazard identification and dose–response assessment information. The information is then used in conjunction with exposure information to characterize risks to public health and may be used in risk-based decision- making, in regulatory actions, and for other risk-management purposes. Since the middle 1990s, EPA has been in the process of updating the IRIS assessment of inorganic arsenic. Arsenic is a naturally occurring element in the environment that can occur in different forms. Inorganic arsenic refers to a set of arsenic- containing molecules that includes elemental arsenic and many other molecular structures that involve arsenic in combination with elements other than carbon. Organic arsenic consists of a broad set of arseni- cal compounds that include carbon in their structures; they include the relatively simple monomethyl ar- senic and dimethyl arsenic that are metabolites of inorganic arsenic in humans and more complex mole- cules that are used as herbicides, pesticides, or animal feed additives. EPA’s toxicologic assessment will focus on inorganic arsenic. The development of the toxicologic assessment of inorganic arsenic is a considerable undertaking for the IRIS program. Much research has been performed on the chemical, and improved approaches to performing assessments have been recommended. Although the multiple challenges are clear given the history of the assessment, there are also opportunities to take advantage of the rich data on inorganic arse- nic to apply evaluation tools and integrative approaches so as to conduct the assessment in a transparent manner. Many of the improved approaches were recommended by previous National Research Council committees that evaluated other chemicals, and it is clear from EPA’s draft conceptual model and analysis plans that the agency plans to incorporate the recommendations. The task of the present committee was to focus on recommendations specific to the assessment of inorganic arsenic. (A different National Research Council committee is reviewing EPA’s initiatives to improve the overall process and quality of IRIS as- sessments.) In response to a congressional mandate for an independent review of the IRIS assessment of inor- ganic arsenic, EPA requested that the National Research Council convene a committee to conduct a two- phase study. In the first phase (the focus of the present report), the committee was to organize a workshop to evaluate critical scientific issues in assessing cancer and noncancer effects of oral exposure to inorgan- ic arsenic and offer recommendations on how the issues could be addressed in EPA’s IRIS assessment. The second phase of the study will begin after EPA completes its draft assessment; it will involve a re- view of the draft assessment to determine whether the committee’s recommendations were appropriately addressed and also reflect recommendations for improving IRIS assessments made in other National Re- search Council reports. As part of the first phase, a workshop was held on April 4-5, 2013, to obtain input on critical aspects of interpreting and applying the scientific information on inorganic arsenic for the pur- poses of hazard identification and dose–response analysis. Having been informed by the workshop and its own preliminary survey of the scientific literature, the committee offers the following recommendations for performing the IRIS assessment as outlined in Figure 1 and described in Box 1. These recommenda- tions are specific to inorganic arsenic and may not be applicable to the assessment of other chemicals. 3

OCR for page 3
4 Critical Aspects of EPA’s IRIS Assessment of Inorganic Arsenic A. Hazard  B. Evidence Evaluation  C. Assessment of  F. Dose‐ Final    D. Mode‐of‐Action  E. Susceptible  Identification and Systematic  Causality Response  arsenic  Analyses Subpopulations    Reviews  Analysis assessment FIGURE 1 Steps of the toxicologic assessment of inorganic arsenic. BOX 1 Committee’s Guidance and Recommendations for Improving Steps of the Toxicologic Assessment of Inorganic Arsenic Illustrated in Figure 1 A. Hazard Identification. A broad literature search and screening process will be used by EPA to identify health effects that have been studied in relation to inorganic arsenic, and a preliminary draft of these efforts was provided to the committee. For the purposes of its review, the committee conducted a preliminary survey of the literature and identified categories of health outcomes that, in the end, overlapped with those identified by EPA. As a starting point, the committee attempted to prioritize these health end points on the basis of the perceived strength of evidence and the importance to public health (see Box 2). Chapter 4 provides some guidance on end point specific issues EPA should consider as it conducts a more comprehensive and systematic evaluation of the literature. B. Evidence Evaluation and Systematic Reviews. EPA has indicated that systematic review will be used to support its toxicologic assessment of inorganic arsenic. To perform such reviews, the committee recommends searching for studies on specific outcomes (see Box 2) that meet the following criteria: individual measures of arsenic exposure, measurement of arsenic that precedes the outcome, and low to moderate exposure to inorganic arsenic (less than 100 µg/L in drinking water). It will also be important to organize the data from the individual studies into evidence tables. The example tables provided by EPA in its draft plans appear to capture the salient categories of information with respect to epidemiologic data. Meta-analysis should also be considered for priority end points if there are three or more peer reviewed studies on the outcome of interest. For dose-response meta-analysis, studies will need to have characterized three or more exposure levels. Chapter 3 provides general guidance for performing systematic reviews and meta-analyses, as well as for developing evidence tables on animal and in vitro data to inform causality determinations and mode-of-action analyses in the low exposure range. C. Assessment of Causality. EPA’s draft plans provide a causal determination framework that describes how it will categorize the evidence on different end points into five possible categories. The committee supports this five-category approach, and recommends that judgments are characterized with respect to the modified Bradford Hill criteria for causality. The assessment of causality will help EPA prioritize end points for subsequent analysis of mode of action and dose response. D. Mode-of-Action Analyses. The committee supports EPA’s plans to conduct mode-of-action analyses for end points it classifies as having a causal or likely to be causal relationship with arsenic. Consideration might also be given to performing mode-of-action analyses for end points with suggestive evidence for the purpose of determining whether there is sufficient evidence to support a stronger causal association and, if so, informing dose-response analyses. Guidance on performing mode of action analyses is presented in Chapter 6. E. Susceptible Subpopulations. The committee’s survey of the literature has identified several factors that could contribute to susceptibility to arsenic. Chapter 5 outlines the importance of considering these factors when interpreting the epidemiologic evidence. Although it is unclear that any of these factors can be evaluated quantitatively in the assessment, their existence is an important consideration when evaluating population risk. F. Dose-Response Analysis. Epidemiologic data are expected to serve as the basis for the dose- response analyses performed for most end points. As outlined in Chapter 7, efforts should be directed at performing dose-response analyses in the range of epidemiologic observations. For some end points it may be possible to perform dose-response meta-analyses. Should the data in the range of observation be inadequate for developing risk estimates that meet EPA’s needs, mode-of-action data should be used to the extent possible to extrapolate below the observed range. The committee concurs with EPA’s draft plan that even if a mode of action cannot be determined with reasonable certainty, dose-response analyses should be performed on health end points deemed to have a causal or likely causal relationship with arsenic. In the absence of mode-of-action data, alternative statistical approaches, described in Chapter 7, are recommended.

OCR for page 3
Summary 5  Hazard Identification: Overall, the committee agrees that documents provided by EPA have out- lined an improved approach to determining the cancer and noncancer effects that may be associated with exposure to inorganic arsenic. The process involves screening the literature, evaluating studies, organizing the evidence, and using a causal determination framework to document how decisions are made about which health end points are linked to inorganic arsenic. Evidence tables are a useful means of organizing the epidemiologic and toxicologic information, and it will be important for the narrative of the IRIS assessment to explain clearly how judgments are made with respect to the causal determination framework. The com- mittee conducted a preliminary survey of the literature on many of the health end points that EPA will con- sider, and key observations about those end points and the implications for dose–response modeling are provided later in this report. In Box 2, the committee has attempted to set priorities among the health end points. These categorizations will be refined by EPA after it conducts a more comprehensive analysis.  Systematic Reviews: EPA has indicated that several systematic reviews of the scientific litera- ture on inorganic arsenic will be used to support its toxicologic assessment of inorganic arsenic. The sys- tematic reviews will be used to address questions raised during the hazard-identification process with re- spect to guiding the dose–response analysis. To facilitate and focus the systematic reviews with respect to epidemiologic studies, the committee recommends that in selecting the studies to be included in meta- analyses, EPA should consider proceeding as follows: o Systematically search for studies on specific outcomes (see suggested list in Box 2) that meet the following criteria: individual measures of arsenic exposure (preferably with biomarker meas- urements), measurement of arsenic that precedes the outcome, and low to moderate exposure to inorganic arsenic (less than 100 µg/L in drinking water). o Consider meta-analyses if there are at least three or more peer-reviewed studies available on the outcome of interest. For dose-response meta-analysis, studies will need to have characterized at least three or more exposure levels. o Review the quality of the evidence (risk of bias) using established guidelines for epidemiologic studies. Elements in the evaluation of study quality include the assessment of study outcome based on standardized definitions, participation rate, adjustment of associations for relevant con- founders, and other considerations that will depend on study design. BOX 2 Hierarchy of Health End Points of Concern for Inorganic Arsenic  Tier 1: Evidence of a causal association determined by other agencies and/or in published systematic reviews o Lung, skin, and bladder cancer o Ischemic heart disease o Skin lesions  Tier 2: Other priority outcomes o Prostate and renal cancer o Diabetes o Nonmalignant respiratory disease o Pregnancy outcomes (infant morbidity) o Neurodevelopmental toxicity o Immune effects  Tier 3: Other end points to consider o Liver and pancreatic cancer o Renal disease o Hypertension o Stroke o Pregnancy outcomes (fetal loss, stillbirth, and neonatal mortality)

OCR for page 3
6 Critical Aspects of EPA’s IRIS Assessment of Inorganic Arsenic  Mode of Action: Mode-of-action analyses should be used to inform dose–response modeling with respect to the shape of the curve, particularly in the low dose region, and the understanding of inter- human variability. Even if the mode of action cannot be firmly established, the exercise can be used to guide modeling qualitatively in the low dose region and used in considering susceptibility. EPA has out- lined a process of delineating a pathway that connects a molecular initiating event to an adverse health end point at a higher biologic level of organization. The process will be used to organize mechanistic in- formation to determine how mode-of-action information supports low-dose extrapolation and to inform how dose–response analyses account for the uncertainty associated with susceptibility. The committee supports EPA’s plans to perform mode-of-action analyses for health outcomes on which there is evidence to infer a causal or likely to be causal relationship with inorganic arsenic. The analyses should follow a stepwise approach, starting with the questions to be answered. All the existing data should be explored, supportive and conflicting evidence should be examined, and it should be determined whether an exposure–response continuum for sequential progression and time dependence of the proposed key events can be established. Biologic plausibility and concordance of evidence from in vitro, animal, and human data should be explored. Consideration could also be given to how the mode of action could be modulated by other potentially causal agents.  Susceptibility: Multiple factors can affect susceptibility to inorganic arsenic, including life stage, genetic factors, sex, nutritional deficiencies, health status, lifestyle (for example, smoking and alcohol con- sumption), and coexposures. The committee agrees with EPA’s proposal to use probabilistic approaches in considering the uncertainty and variability associated with those factors. Susceptibility due to pre-existing disease is an important consideration, as arsenic has been shown to increase the risk of several major diseas- es prevalent in the United States (e.g., diabetes, cardiovascular disease). On the basis of the degree of evi- dence on a vulnerable population, consideration should be given to whether dose–response assessment should focus on the population as a whole or should involve separate assessments for the general population and susceptible groups. With respect to smoking—which appears to increase the risk of skin lesions, lung and bladder cancer, and maybe cardiovascular disease—it might be possible to apply sensitivity approaches to the dose–response relationship for the various end points to determine the degree to which smoking would change the potency calculation. Consideration should also be given to the growing evidence from human and animal studies that sug- gests that early-life exposure to arsenic may increase the risk of adverse health effects and the risk of im- paired development in infancy and childhood and later in life. Thus, the timing of exposure should be con- sidered in evaluating epidemiologic studies for dose–response assessment. EPA’s current approach of assessing less than lifetime exposures for cancer risk from nonmutagenic carcinogens by prorating the risk equally regardless of age needs to be critically evaluated to determine whether it is appropriate for inorganic arsenic. There is clear evidence of sex differences in the metabolism of inorganic arsenic. Because arsenic me- tabolism is a recognized susceptibility factor, it seems likely that the toxicity of arsenic could differ between men and women. Indeed, evidence of such differential risks is growing.  Dose–Response Analyses: Because many epidemiologic studies of cancer and noncancer end points now characterize risk associated with low to moderate concentrations of inorganic arsenic, in some cases approaching or including background concentrations, it may be possible to estimate risk directly from the range of observations in the literature. Background concentrations of arsenic vary, but the com- mittee judged that urinary arsenic concentrations of 1–5 µg/L (summing inorganic, monomethyl, and di- methyl arsenic forms) is a reasonable estimate for the US population. The committee does not assume that those background concentrations are with or without health effects; rather, it assumes that the needs of assessing health risks can be facilitated by characterizing dose–response relationships down to the back- ground concentrations by using observed data. The committee recommends that EPA develop risk estimates across the array of health effects on which there is adequate epidemiologic evidence and then derive risk-specific doses to address the needs

OCR for page 3
Summary 7 of analyses that would typically use a reference dose (RfD). That approach would facilitate efforts to evaluate cumulative risks posed by exposure to multiple chemicals, conduct risk–benefit assessments, or to conduct other comparative analyses. An RfD might be selected from among the risk-specific doses on the basis of such factors as end-point severity, interhuman variability, and policy considerations so that it may be used by stakeholders until guidance on how to use risk-specific doses for noncancer end points is established. If the health-assessment needs cannot be fully met by using modeling of the data in the range of ob- servation, extrapolation to below the observed range may be necessary. The preferred approach would be to use data that define the individual and population risks and their associated uncertainty on the basis of analyses of adverse-outcome pathways or mode of action and human variability in susceptibility. That would describe human pharmacokinetics, biomarkers of exposure, tissue doses of relevant arsenic forms, and the multiple toxicodynamic processes that lead to the adverse outcomes. However, it is not clear whether such an approach would be feasible without further research and modeling. In the absence of ad- equate mode-of-action information, EPA could use modeled curve shapes that approach the low range of observation to extrapolate below that range. Fitting of alternative models could inform how far below the range of observation the extrapolation is essentially independent of model choice and thus provide greater confidence in the extrapolation to that point. Consideration should be given to whether mechanistic in- formation can inform the process. Extrapolations will become increasingly uncertain as they go further below the observed range, so it would be reasonable for the IRIS assessment to stipulate the range over which the dose–response relationship derived for cancer and noncancer end points is useful for risk as- sessment. To assess uncertainties in such an extrapolation, EPA could use a form of sensitivity analysis, fit multiple models to the data in the range of observation, and assess how sensitive extrapolation below the range of observation was to the choice of model. The IRIS program should explore the possibility of performing dose–response meta-analyses for inorganic arsenic with standard methods that have been de- scribed in the environmental-health and public-health literature. Methods are available to evaluate both linear and nonlinear dose–response curves.