Approaches used by the US Environmental Protection Agency (EPA) for testing chemicals and for establishing health reference values have been modified over the past two decades in response to a growing awareness that some chemicals interact with endocrine hormone systems. Some changes made by EPA have been mandated legislatively. For example, as discussed in Chapter 1, the 1996 Food Quality Protection Act (FQPA) requires EPA to screen pesticides and other chemicals for their potential to produce estrogenic and other endocrine effects. Passage of the FQPA led to EPA’s Endocrine Disruptor Screening Program and development of a series of in vitro and in vivo screening tests to identify chemicals that interact with the estrogen, androgen, or thyroid hormone systems (74 Fed. Reg. 54416 ). Since the 1990s, EPA, the Organisation for Economic Cooperation and Development (OECD), and the National Toxicology Program (NTP) have modified toxicity-testing guidelines to improve their ability to detect effects that occur later in life after exposure to endocrine active chemicals (EACs) during sensitive windows of development. Toxicity-testing methods developed by OECD that can detect endocrine toxicity in mammals include the rodent two-generation reproduction study (TG 416), the extended one-generation reproductive toxicity study (TG 443), the rodent reproduction/developmental toxicity screening test (TG 421), the rodent chronic toxicity and oncogenicity studies (TG 451, TG 452, and TG 453), and the enhanced 28-day toxicity study (TG 407) (Bars et al. 2012). NTP has also updated its testing protocols: for example, by adding early life exposure to some cancer bioassays, adding endocrine outcomes, and extending follow-up when studying adverse effects on reproduction and development (Foster 2014).
Several aspects of EACs have prompted the need to assess the adequacy of traditional toxicity-testing strategies (see Box 2-1). Some EACs can mimic natural hormones, which affect the endocrine systems at low concentrations. Questions have been raised about how to use information about endocrine activity to understand potential health risks. For example, should a change in hormone concentrations be considered an adverse health effect? If such a change is not necessarily adverse, can it be used to predict an adverse outcome? If so, can one estimate the probability of an adverse outcome given an exposure?
EACs also have the potential to cause long-lasting developmental effects. Organisms can be especially sensitive to EACs because hormones play a critical role during normal development. Dose and timing also can dramatically influence not only the magnitude of an effect but also the type of outcome observed or, in some cases, the direction of the effect (Ankley and Villeneuve 2015). For example, the synthetic estrogen diethylstilbestrol (DES) produces tumors in different tissues depending on what dose is administered and whether the exposure occurs prenatally or neonatally (Newbold and McLachlan 1982; Newbold 2004; Newbold et al. 2004). Thus, evaluation of EAC effects throughout an organism’s lifespan is increasingly recognized as important, and some study designs that assess fertility, reproductive-tract malformations, or tumor incidence in animals have been modified. The modifications—such as dosing of pregnant animals throughout gestation, longer follow-up periods (including evaluation of several generations), assessment of multiple hormone-sensitive end points, and examination of multiple pups per litter—have improved sensitivity to detect endocrine effects (Blystone et al. 2010; Foster 2014). Awareness that EAC exposure during development can program tissues to respond differently to endogenous hormones or exogenous chemical challenges later in life (Newbold et al. 2004; Jenkins et al. 2007) or produce heritable modifications via epigenetic changes (Jefferson et al. 2013) is also growing.
Another concerning aspect of EACs is that effects from EAC exposure have been reported in humans and wildlife (Bernanke and Köhler 2009). Effects have been identified from a range of human exposures from environmental to deliberate administration of pharmaceutical agents. For example, DES administration to pregnant women provides a prime example of unexpected adverse effects, such as increased risk of breast cancer in mothers and various adverse outcomes in their offspring (Hoover et al. 2011; Reed and Fenton 2013).
One additional factor that has prompted increased scrutiny of EACs is the debate about whether thresholds exist for EAC effects. As reviewed by Hass et al. (2013), arguments in support of a threshold cite homeostatic mechanisms that are involved in endocrine regulation and the resiliency of higher order systems to adapt. Arguments against the presence of a threshold note that small fluctuations in endogenous hormones can affect regulation of a variety of biological processes (Hass et al. 2013). Thus, questions have been raised about whether the dose-selection practices used in traditional toxicity testing should be revised.
To protect public health and the environment, EPA and other agencies will need to work proactively to update testing methods as new science emerges. This chapter describes the committee’s proposed strategy to assist EPA with the tasks of developing and revising testing practices in response to expanded knowledge about the potential for low-dose effects of EACs.
The overall strategy envisioned by the committee for evaluating evidence of low-dose adverse human effects consists of three broad phases—surveillance, investigation and analysis, and actions (see Figure 2-1). The strategy recognizes that a toxicity testing program, no matter how sophisticated, cannot provide 100% assurance that all adverse effects will be identified and can be prevented. Even with pharmaceuticals, which are tested in human clinical trials, adverse effects are often not identified until after the drug has been marketed. Therefore, environmental chemicals, which are tested (if at all) in experimental systems only before use, require continued surveillance to protect public health given the expectation that false negatives will occasionally occur in testing.
Once a topic has been identified for additional investigation, the specific details of the investigation and analysis need to be planned so that they will support future agency actions. The results from the investigations and analyses are then used to select specific actions. In some cases, the only further action would be continued surveillance. In other cases where key uncertainties exist, further action could entail the generation of new data or models to address the uncertainties. If the results of an investigation suggest that adverse outcomes in humans are expected or might be occurring at low doses, the conclusions of previous toxicity testing or toxicity assessments for the chemicals that are under investigation might need to
be updated to reflect the new evidence. Additionally, such evidence might support updates to specific toxicity-testing or assessment practices to reduce the false-negative rate in the future.
The first two phases, surveillance and investigation and analysis, are described in more detail below. As noted in Figure 2-1, completion of each phase could involve one or more approaches. Although the descriptions of the components are presented sequentially in the report, there is no requirement that each approach be used or that a specific order be followed. The last phase, actions, involves policy decisions, which are outside of the committee’s charge and therefore are not discussed in detail in this report.
In the context of this report, surveillance refers to the process for detecting signals (indications that an adverse outcome in a human population or animal model might be related to exposure to an EAC at low doses) by searching, retrieving, and evaluating existing data. Surveillance also refers to the process for monitoring the literature for methods that could be used for toxicity testing of EACs. Types of data that could be considered in an active surveillance program include human, animal, or mechanistic data.
Actively Monitor for New Data
A surveillance program for identifying low-dose effects should have a process for actively monitoring for new data to help ensure that effects will be identified and analyzed on a regular basis. Three broad categories—chemical-specific data, information that could lead to modifications of toxicity-testing methods and best practices for EACs, and information on endocrine-related effects in animals and humans—should be considered in the surveillance program. Relevant information to monitor might include scientific literature, various databases, nontraditional information sources, stakeholder input, and human exposure information. Although monitoring scientific literature and databases and human exposure information might provide the most valuable information, nontraditional information sources and stakeholder input have recently been highlighted as potential sources that could lead to valuable insights. Which sources are selected for surveillance will clearly depend on the problem under consideration and the resources available to the agency. Each source is discussed in more detail in the following sections.
Monitoring the Scientific Literature
EPA has many ongoing literature-review activities for evaluating chemical hazards, such as the chemical-specific evaluations of the Office of Research and Development, the Office of Pesticide Programs, the Office of Water, and the Office of Air and Radiation. Chemical-induced effects on endocrine function might be considered in those assessments. Although the assessments typically rely heavily on animal data, human data also contribute to these assessments. Whether human studies are concordant or discordant with the animal data is an important consideration.
In addition to chemical-specific data, the scientific literature should be monitored to identify relevant end points that might need to be included in toxicity testing or risk assessment or to identify other changes to toxicity-testing practices that might be needed to improve assessment of EAC effects. For example, several assessments conducted by EPA, NTP, and international institutions have investigated various methodological issues, including addition of mammary gland assessment in regulatory guideline protocols (Makris 2011); adequacy of rodent models for detecting hormonally related cancers (Thayer and Foster 2007); study design issues for developmental reproductive toxicity (Blystone et al. 2010); and the question of biological thresholds for EACs (Haas et al. 2013). Furthermore, literature reviews of diabetes, breast cancer, and other diseases have identified biological processes and specific chemicals that appear to be involved in their etiology (Rudel et al. 2007, 2011, 2014; Macon and Fenton 2013; Schwarzman et al. 2015; Auerbach et al. 2016; Smith et al. 2016; Bruner-Tran et al. 2017), and these reviews might reveal effects that need to be integrated into toxicity-testing methods.
The study of endocrine-related human diseases, such as prostate cancer and endometriosis, might also identify adverse effects that are not readily detected in rodent studies. For example, endometriosis is a human disease that is difficult to assess in animal models because most nonprimate mammals do not menstruate and consequently do not develop ectopic lesions, which are the pathological hallmarks of endometriosis (Bellofiore et al. 2017; Bruner-Tran et al. 2017). Concerns about the possible role of chemical exposure in the development of endometriosis might therefore be difficult to assess in an animal model (see Box 2-2). Agencies that are involved in health surveillance and that have health registries (e.g., Centers for Disease Control and Prevention, National Center for Health Statistics) could also be a source of information about relevant human diseases. Thus, by reviewing endocrine-related human diseases, one could ensure that signals are not being missed because they are not recapitulated in animal models. Given that there will always be species differences in response to chemical exposure, a surveillance system should always include monitoring epidemiologic literature or research.
The committee notes that automated methods are being developed for monitoring the literature. For example, methods or tools have been developed to extract drug-safety information from the published literature and electronic medical records (Shetty and Dalal et al. 2011; Wang et al. 2011; Duke et al. 2012; Gurulingappa et al. 2012; Avillach et al. 2013; Pontes et al. 2014; Winnenburg et al. 2015), and these approaches might be relevant for monitoring literature on EACs. Other publicly available tools have been developed for supporting systematic reviews (ICASR 2015), and these might assist with searching and retrieving data related to new toxicity-testing methods, outcomes, exposure assessment, and biomonitoring.
Assessing Scientific Data
In addition to the published peer-reviewed scientific literature, data are being generated and made available to the scientific community through various other venues. For example, as described in Chapter 1, EPA’s ToxCast program has generated substantial data that can potentially provide toxicity and mechanistic information on a variety of chemicals. The testing is typically conducted in in vitro assays and uses a broad range of testing concentrations, including ones found in human biological samples; thus the testing could have relevance to environmental or low-dose exposure. The data are made available through
EPA’s website.1 Other information could come from databases developed to track mechanistic pathways for adverse outcomes.2 These types of data could potentially help identify EACs, and their review might be considered important for a surveillance program.
Tracking Nontraditional Sources of Information
Some investigators have explored whether information extracted from such informal media sources as blogs and other various forms of social media could provide useful drug-safety surveillance data (Harpaz et al. 2012; Lardon et al. 2015; Nikfarjam et al. 2015; Sarker and Gonzalez 2015). Those efforts have proven challenging for drug surveillance programs because the language used to describe medical information in social media is often informal or descriptive (Nikfarjam et al. 2015). Applying those methods to surveillance of environmental chemical exposures could prove even more challenging because, unlike with drug surveillance, people are generally unaware of their environmental exposures. Despite the limitations, approaches for evaluating social media could be explored as a means to provide additional surveillance of potential health effects associated with chemical exposure.
Obtaining Stakeholder Input
A surveillance program could also include components from stakeholder input. EPA could evaluate recommendations or policy statements made by scientific societies or organizations or other academic groups. For example, several position papers have been published on the effect of EACs on human health (Diamanti-Kandarakis et al. 2009; Skakkebaek et al. 2011; Gore et al. 2015; Bennett et al. 2016). Another means of identifying a potential public-health problem is to consider input from advocacy groups,
individual scientists, political leaders, chemical manufacturers, and other stakeholders. For example, NTP has a long-standing process that encourages nominations of chemicals of concern regarding human health (Heindel 1988). Similarly, EPA’s Integrated Risk Information System (IRIS) program provides opportunities for EPA program and regional offices and other stakeholders to nominate chemicals for consideration. Nomination processes provide opportunities for stakeholder input and can help focus attention on societal concerns. Nomination processes are not without some limitations, however, including the possibility of reporting bias by the media, community action groups, or industry. Those groups might also be influenced by social activism or corporate product defense and might not be focused on longer-term public-health issues (Mihaylov and Perkins 2015; Zoller 2017). Guidance for engaging stakeholders has been provided in other National Academies reports (e.g., NRC 2009, 2014).
Monitoring Human Exposure Information
Biomonitoring data are an important information source that can help identify whether human exposure to EACs has occurred or changed over time (NRC 2006a, 2007, 2009, 2012) and are useful in defining low-dose exposures. They can also identify potential exposure sources and demographics of highly exposed groups. A previous NRC committee provided the following recommendations regarding the need to use biomonitoring data in surveillance programs: “Develop biomonitoring-based epidemiologic, toxicologic, and exposure-assessment investigations and public-health surveillance to interpret the risks posed by low-level exposure to environmental chemicals. Where possible, enhance existing exposure assessment, epidemiologic, and toxicologic studies with biomonitoring to improve the interpretation of results of such studies” (NRC 2006a, p. 9).
Box 2-3 illustrates how biomonitoring data from the US Centers for Disease Control and Prevention (CDC) National Health and Nutrition Examination Survey (NHANES) can help inform evaluations of EACs. Occupational exposure data might also provide information that could help define the range of human exposures to some chemicals.
In the absence of biomonitoring data, external exposure data can be used to estimate the range of potential human exposure to chemicals. Active air sampling devices have been used for several years to assess personal exposure to air pollutants, such as particulates, ozone, and polycyclic aromatic hydrocarbons (Geyh et al. 1999; Perera et al. 2003; Tsai et al. 2012; Oliveira et al. 2016). More recently, passive sampling devices have been developed to eliminate the need for cumbersome equipment that decreases compliance in human studies. Passive samplers are favored for personal monitoring because they are lighter and smaller and less likely to interfere with daily activities (NRC 1991). For example, silicone wristbands have been developed as personal passive samplers capable of sequestering polycyclic aromatic hydrocarbons, flame retardants, and pesticides as measures of an individual’s external exposure (O’Connell et al. 2014; Donald et al. 2016; Hammel et al. 2016). Several funding agencies are supporting the development of sensor technologies for the 21st century, which includes wearable monitors that can be used in population studies to measure personal exposure in real time with high sensitivity and specificity and low cost.3
Another type of exposure data comes from efforts to measure chemicals in air, water, soil, and other environmental media (such as house dust), food, and consumer products, and extrapolate those measurements to human exposure. For example, EPA’s Particulate Matter (PM) Supersites Program was established to obtain atmospheric measurements to address the research questions and scientific uncertainties about PM relationships between sources, receptors, exposures, and effects (Solomon and Sioutas 2008). A similar EPA program—the Clean Air Status and Trends Network (CASTNET)—is a national monitoring network established to assess trends in pollutant concentrations, atmospheric deposition, and ecological effects due to changes in air-pollutant emissions (Puchalski et al. 2015). Other air-monitoring programs have evaluated changes in urban ozone concentrations that have occurred over several decades (Sather and Cavender 2016).
Surveillance of Endocrine-Mediated Drug Reactions Using Pharmacovigilance
The US Food and Drug Administration (FDA) has developed pharmacovigilance4 programs to monitor for adverse drug reactions (ADRs). Such programs are needed because human clinical trials use relatively small sample sizes; have shorter durations of exposure than usually occur; often lack diversity among study participants; and do not include pediatric and other susceptible subpopulations (McMahon et al. 2015). Those design features limit the ability of clinical trials to detect rare (1:1,000 to 1:10,000) ADRs that occur at therapeutic drug doses (Schotland et al. 2016).
Pharmacovigilance programs often consist of premarketing surveillance that identifies ADRs during preclinical screening and clinical trials and postmarketing surveillance in which data are accumulated throughout a drug’s market life (Ibrahim et al. 2016). Voluntary reporting systems have historically served as the primary data collection system for postmarketing pharmacovigilance. Such passive surveillance systems primarily rely on the collection of reports of suspected ADRs from health care professionals, consumers, and pharmaceutical companies. Today, FDA has established the Office of Surveillance and Epidemiology to help coordinate its efforts in postmarket drug safety surveillance (see Box 2-4).
4 Pharmacovigilance is defined by the World Health Organization (WHO) as “the science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other drug-related problem” (WHO 2002, p. 7). Pharmacovigilance includes postmarketing safety surveillance activities to detect events that were not seen in a clinical trial (“safety signal generation”).
Some drugs have been shown to demonstrate endocrine activity (see, for example, Friedman et al. 2009). In fact, the Institute of Medicine recommended that FDA “engage the pharmaceutical industry and scientific community in postmarketing studies or clinical trials for hormonally active prescription drugs for which the potential impact on breast cancer risk has not been well characterized” (IOM 2012, p. 21). Such an approach is also relevant for other endocrine-related effects. The committee concludes that methods that have been developed for pharmacovigilance programs might be adapted or help inform an EAC surveillance program. The committee notes, however, that EAC surveillance programs that are based solely on voluntary reporting might be limited because people often cannot self-report environmental exposures.
Periodically Identify, Scope, and Prioritize Topics
Active surveillance will likely result in the identification of information on chemical exposures, outcomes, and advances in toxicity testing that will need to be reviewed. Accordingly, the next step in the committee’s strategy (see Figure 2-1) is periodic review of information and identification of topics that might need to be pursued further. That effort could involve a scoping exercise in which EPA would survey the literature and other information to determine the extent, range, and nature of information available on the topic, to identify data gaps, and to consider whether additional research might be needed. The scoping step would assist EPA in setting priorities for topics that deserve further study. Decisions to pursue a topic could be influenced by a number of factors, including the size of the population at risk, public-health significance, and available resources.
Formulate Questions to Address and Develop an Approach for the Investigation
Once a topic is selected for further analysis, the next step is to formulate the questions to address and develop an approach to the investigation, which will involve consideration of the scientific evidence, expert judgment, and relevant stakeholder perspectives. Questions and approaches are often targeted depending on the topic under investigation and potential actions that could be taken by the agency. For example, EPA might consider whether a new outcome measure should be included in regulatory toxicity tests. In that case, questions that might be posed include the following:
- Is the outcome an emerging concern?
- Are appropriate assays available or could they be developed?
- Is the outcome a more sensitive measure of adverse toxicity than currently used outcomes?
- Is the outcome of interest a reproducible effect after chemical exposure?
- Will including the new outcome in regulatory testing improve hazard identification—that is, improve sensitivity?
Box 2-5 provides a retrospective example of how the answers to these questions helped establish the measurement of anogenital distance as an outcome measure in regulatory toxicity tests.
Once the questions have been identified, the next step is to formulate the approach to the investigation and determine the types of data and analyses that are needed to answer the questions and provide the basis for agency actions. Figure 2-1 shows four types of investigation and analysis that could be considered: generation of new data or models to fill data gaps, targeted analysis of existing data, systematic review, and integration of the available evidence. The types of investigations or analyses might not be mutually exclusive because several related investigations or analyses might be needed to address the questions adequately. And, as noted above, they could be influenced by a number of factors, including the size of the population at potential risk, public-health significance, and available resources.
Targeted Analysis of Existing Data
In some cases, a targeted analysis (or reanalysis) of existing data might support agency actions. A targeted analysis might be used when different test systems, methods, analytical approaches, and other experimental design features differ between studies and make the results difficult to compare. For example, outcome measures that are analyzed as continuous variables (such as changes in mean response) in some studies are not directly comparable to outcome measures that are analyzed as dichotomous variables (such as numbers of individuals beyond a specified cut point) in other studies. Another example is the summary of dose-response data in terms of pair-wise significance. In that case, the need for additional analyses could arise when there is a finding of a no-observed-adverse-effect level (NOAEL) in one study and a lowest-observed-adverse-effect level (LOAEL) in another study at the same dose. A targeted analysis of these seemingly discordant data might be able to strengthen the interpretation of the evidence. Specifically, an analysis of those data should consider whether they are statistically significantly different from each other. Additionally, trend analyses might also be useful as an alternative to pair-wise comparisons, because trend tests might have greater statistical power. Other contextual factors—including experimental design and conduct, mechanistic data, and prior evidence—should also be considered (Goodman 2016). Consideration of those factors will help reduce the tendency of some regulators to incorrectly perceive a NOAEL as a threshold (Scholze and Kortenkamp 2007) and reduce reliance on the use of a pvalue (most often set at 0.05) as a bright line in evaluating whether an effect has occurred (Goodman 2016).
It is also common that studies from different investigators will not be performed at the same doses, and this makes it difficult to compare studies and evaluate consistency. To interpolate between dose groups, one can typically use parametric or nonparametric curve-fitting approaches. For epidemiological data, replacing pair-wise comparisons with regression analyses allows comparison of regression coefficients that can address this issue. For both toxicological and epidemiological data, the benchmark dose (BMD) approach can be well suited for comparing evidence of adverse effects among both toxicologic and epidemiologic studies. The BMD is the estimated dose associated with a specific level of response, called the benchmark response (BMR), along with its confidence interval. The use of a common BMR forces a transparent definition of the size of a biologically significant effect that is common among the studies being compared, and the resulting confidence intervals can be compared to evaluate study consistency or inconsistency. That approach has been extended to end points that are not strictly identical by using categorical regression, in which disparate end points are grouped into “bins” of severity categories (EPA 2000). The approach then estimates the BMD associated with a specific severity. Categorical regression, however, requires judgment to determine which end points and magnitudes of effect are to be grouped together at each level of severity.
There are cases in which data might be reanalyzed to better account for uncertainties. For example, in its report Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment, NRC (2006b) recommended conducting a quantitative uncertainty analysis as part of EPA’s dose-response assessment. There have been numerous other examples in the literature in which data have been reanalyzed to address previously unaddressed uncertainties, particularly at low doses (see, for example, Subramaniam et al. 2007; Crump et al. 2016).
Systematic reviews provide a method for evaluating evidence in a transparent and consistent manner that reduces bias. As described in several National Academies reports (NRC 2011, 2014), systematic review provides a rigorous approach to evaluating evidence that, although not removing the role of expert judgment, aims to make such judgments more transparent and less susceptible to biases. Guidelines for performing systematic reviews relevant to environmental or public-health assessments include the Navi-
A systematic review is guided by a study question that should be carefully crafted to address the problem under consideration. If the question is too broad in scope, the studies included in the analysis might be too heterogeneous for effectively integrating and drawing a conclusion. On the other hand, if it is too narrow in scope, the results are less generalizable and might not be relevant to the underlying public-health concern. The methods for identifying, screening, and analyzing the scientific literature are planned in advance to ensure that the evidence is selected and evaluated in an objective and consistent manner. The committee notes that the systematic review method alone does not lend itself to answering a question about whether a specific EAC has low-dose adverse effects. At a minimum, answering that question requires the completion of hazard identification and dose-response assessment for the effects of interest.
Generate New Data or Models
A focused research question could be addressed by generating new data or models. In some cases, that activity could involve conducting animal toxicity studies, pharmacokinetic studies, or epidemiologic investigations. At other times, new in vitro data could shed light on the mechanisms involved in an observed response. Other types of research questions might be answered through computational model development. For example, physiologically based pharmacokinetic (PBPK) or other quantitative dosimetry models for a chemical of interest could be developed to compare dose-response relationships in human studies with those in rodent studies. Quantitative dosimetry models could also help define the relationship between external dose and internal tissue concentrations and could be used to support cross-species and route-to-route extrapolations. Other types of dosimetry models might be needed to facilitate interpretation of in vitro data. For example, reverse-toxicokinetic modeling and in vitro–in vivo extrapolations are used to compare in vitro data with estimated or measured human exposure data (Wetmore et al. 2012; Yoon et al. 2012).
Questions about the coherence of findings in rodents and humans might be addressed through evaluation of mechanistic data. One way of improving our understanding of mechanistic data involves the development of conceptual models that facilitate the organization of information about biological interactions and toxicity mechanisms. A conceptual model can reflect the initial interactions of a chemical with the biological system and the resulting events that can lead to a specific adverse outcome (Ankley et al. 2010). Tools are being developed that can predict associations between key events and thus lead to the development of quantitative or computationally predicted adverse outcome pathways (qAOPs or cpAOPs) (Bell et al. 2016; Connolly et al. 2017). The committee notes that endocrine-related effects probably do not result solely from one isolated or linear pathway but involve multiple pathways or networks, and disease manifestation likely involves multiple components or stressors (NASEM 2017).
Determining the types of new data and models that are needed should be tailored to the research questions, and specific recommendations are beyond the scope of this report. Nevertheless, Box 2-6 provides an example of generating data to address uncertainty.
Integrate Available Evidence
This committee—as have other National Academies committees before it—emphasizes the need for evidence integration to be both transparent and standardized in its approach. Thus far, evidence integration has focused on the purpose of hazard identification: that is, determining whether a causal relationship exists. Causal frameworks, such as those developed by the International Agency for Research on Cancer, NTP, and EPA, can be adopted or adapted to provide transparency and consistency in conducting causal evaluations. See previous National Academies reports for additional guidance on such approaches (NRC 2014; NASEM 2015, 2017).
The committee recognizes, however, that for addressing low-dose adverse effects of chemical exposure, the question will often be more explicitly quantitative: that is, it specifically concerns the nature of the dose-response relationship at low doses. Although the evidence of causality is still important, the problem is that the causal evidence might include studies that exclusively include high exposures, such as experimental doses near the maximum tolerated dose. High-dose data alone are usually not useful for making inferences about response to exposures at low doses because of uncertainties in the shape of dose-response curves below the range of observation.
Therefore, it might be necessary to integrate the subset of evidence that includes low-dose toxicity data separately. Although not very informative for causal inferences, environmental exposure data, such as biomonitoring data, might nevertheless be useful for defining what subset of the data can be considered as low dose (see discussion earlier in this chapter). Some additional considerations include the following:
- In vitro–in vivo extrapolation or reverse toxicokinetics can help to determine what in vitro mechanistic data could be considered low-dose data.
- Incorporation of modeled dose-response relationships, including BMD estimates.
- Addressing toxicokinetic and toxicodynamic differences between species and populations.
- Biological plausibility given mechanistic data.
- Co-exposures that might act on the same end point.
Data integration can also be used to consider questions that are not about specific EACs but are broader, such as whether a new end point or new exposure or assessment window is relevant to determining low-dose effects. As noted earlier, some end points have been added to regulatory testing protocols in response to growing evidence that they are indicators of toxicity, and the duration of some tests has been extended to capture effects that might occur later in life. Signals identified during the surveillance step that have those types of implications about toxicity testing could be evaluated by integrating the available evidence. One example of such a signal is the growing concern about evaluations of mammary gland toxicity. Makris (2011) evaluated how the effects of environmental chemicals on the mammary gland are
assessed in guideline studies of EPA, OECD, and NTP and made a number of recommendations for enhancing how the end point is assessed. She identified data gaps, issues, and challenges and noted that “to address these issues, a paradigm shift would be needed for the evaluation of [mammary gland] in guideline studies” (Makris 2011, p. 1050). Challenges identified with implementing such a shift were issues of species and strain sensitivity, the timing of exposure, and when the end point is evaluated. Evidence integration could help address such issues.
The remaining step in the committee’s strategy is to select the types of actions that are needed. As shown in Figure 2-1, several types of actions could result, including the need to update chemical assessments, to continue to monitor for new data, to require new data or models to reduce uncertainties, or to update toxicity-testing designs and practices. The type of actions that EPA takes could be influenced by a number of factors, including the size of the population at risk, the public-health significance of the investigation, and available resources. Specific recommendations on exactly what actions to take are beyond the scope of this report.
- To ensure adequate understanding of hazards and to inform regulatory decision making, EPA needs a general strategy for ongoing evaluation of evidence of low-dose effects from exposure to EACs. The committee proposes a strategy involving three phases: surveillance, investigation and analysis, and actions. EPA is already conducting many activities consistent with the proposed strategy, though not necessarily in the specific context of assessing low-dose exposure to EACs.
Recommendation: EPA should develop an active surveillance program focused specifically on low-dose exposures to EACs. This program could include regularly monitoring published research and other information sources, gathering input from stakeholders, and collecting human exposure information. It might also involve data collection in collaboration with other agencies and outside parties. The surveillance program should periodically identify, scope, and prioritize potential areas of focus related to low-dose effects, such as particular chemicals and end points. Some approaches discussed in this chapter will require methods and tool development, such as automated methods for monitoring the literature.
Recommendation: After a topic is selected for further evaluation, the agency should plan its investigation by identifying key questions to be addressed and determining the types of data and analyses needed to answer the questions and to support future agency actions.
- The four main approaches for investigation and analysis are targeted analysis of existing data, systematic review, generation of new data or models, and integration of evidence. The types of analyses used to investigate the questions are not mutually exclusive, and several approaches might be needed to address the questions adequately. Integration of evidence for low-dose adverse human effects of EACs involves consideration of both hazard identification and dose response.
Recommendation: Environmental exposure data should be used, if available, to define what subset of the data should be considered as low dose.
A robust strategy will provide the agency with a range of options to address questions of concern.
Recommendation: The specific approaches and tools used to implement the strategy to address issues related to low-dose endocrine effects will need to be considered on a case-by-case basis and should be guided by the specific questions under study.
- The proposed strategy in this chapter will facilitate more regular consideration of the adequacy of toxicity testing. However, the agency will also be faced with questions about the amount of evidence needed to change traditional test methods, and these questions might be more appropriately addressed through policy decisions.
Ankley, G.T., and D.L. Villeneuve. 2015. Temporal changes in biological responses and uncertainty in assessing risks of endocrine-disrupting chemicals: Insights from intensive time-course studies with fish. Toxicol. Sci. 144(2):259-275.
Ankley, G.T., R.S. Bennett, R.J. Erickson, D.J. Hoff, M.W. Hornung, R.D. Johnson, D.R. Mount, J.W. Nichols, C.L. Russom, P.K. Schmieder, J.A. Serrano, J.E. Tietge, and D.L. Villeneuve. 2010. Adverse outcome pathways: A conceptual framework to support ecotoxicology research and risk assessment. Environ. Toxicol. Chem. 29(3):730-741.
Auerbach, S., D. Filer, D. Reif, V. Walker, A.C. Holloway, J. Schlezinger, S. Srinivasan, D. Svoboda, R. Judson, J.R. Bucher, and K.A. Thayer. 2016. Prioritizing environmental chemicals for obesity and diabetes outcomes research: A screening approach using ToxCast™ high-throughput data. Environ. Health Perspect. 124(8):1141-1154.
Avillach, P., J.C. Dufour, G. Diallo, F. Salvo, M. Joubert, F. Thiessard, F. Mougin, G. Trifirò, A. Fourrier-Réglat, A. Pariente, and M. Fieschi. 2013. Design and validation of an automated method to detect known adverse drug reactions in MEDLINE: A contribution from the EU-ADR project. J. Am. Med. Inform. Assoc. 20(3):446-452.
Aylward, L.L., R. Kirman, R. Schoeny, C.J. Portier, and S.M. Hays. 2013. Evaluation of biomonitoring data from the CDC National Exposure Report in a risk assessment context: Perspectives across chemicals. Environ. Health Perspect. 121(3):287-294.
Aylward, L.L., S.M. Hays, R. Smolders, H.M. Koch, J. Cocker, K. Jones, N. Warren, L. Levy, and R. Bevan. 2014. Sources of variability in biomarker concentrations. J. Toxicol. Environ. Health B Crit. Rev. 17(1):45-61.
Bars, R., I. Fegert, M. Gross, D. Lewis, L. Weltje, A. Weyers, J.R. Wheeler, and M. Galay-Burgos. 2012. Risk assessment of endocrine active chemicals: identifyingchemicals of regulatory concern. Regul. Toxicol. Pharmacol. 64(1):143-154.
Beekhuijzen, M., F. van Otterdijk, W. Wieland, M. van Tuyl, R.P. Rijcken, B. Peter, and H. Emmen. 2016. Update of OECD DART guidelines with endocrine disruptor relevant endpoints: Practical considerations. Reprod. Toxicol. 64:64-71.
Bell, S.M., M.M. Angrish, C.E. Wood, and S.W. Edwards. 2016. Integrating publicly available data to generate computationally predicted adverse outcome pathways for fatty liver. Toxicol. Sci. 150(2):510-520.
Bellofiore, N., S.J. Ellery, J. Mamrot, D.W. Walker, P. Temple-Smith, and H. Dickinson. 2017. First evidence of a menstruating rodent: The spiny mouse (Acomys cahirinus). Am. J. Obstet. Gynecol. 216(1):40e1-40e11.
Bennett, D., D.C. Bellinger, L.S. Birnbaum, A. Bradman, A. Chen, D.A. Cory-Slechta, S.M. Engel, M.D. Fallin, A. Halladay, R. Hauser, I. Hertz-Picciotto, C.F. Kwiatkowski, B.P. Lanphear, E. Marquez, M. Marty, J. McPartland, C.J. Newschaffer, D. Payne-Sturges, H.B. Patisaul, F.P. Perera, B. Ritz, J. Sass, S.L. Schantz, T.F. Webster, R.M. Whyatt, T.J. Woodruff, R.T. Zoeller, L. Anderko, C. Campbell, J.A. Conry, N. DeNicola, R.M. Gould, D. Hirtz, K. Huffling, P.J. Landrigan, A. Lavin, M. Miller, M.A. Mitchell, L. Rubin, T. Schettler, H.L. Tran, A. Acosta, C. Brody, E. Miller, P. Miller, M. Swanson, and N.O. Witherspoon. 2016. Project TENDR: Targeting environmental neuro-developmental risks the TENDR consensus statement. Environ. Health Perspect. 124(7):A118-A122.
Bernanke, J., and H.R. Köhler. 2009. The impact of environmental chemicals on wildlife vertebrates. Rev. Environ. Contam. Toxicol. 198:1-47.
Blystone, C.R., G.E. Kissling, J.B. Bishop, R.E. Chapin, G.W. Wolfe, and P.M. Foster. 2010. Determination of the di-(2-ethylhexyl) phthalate NOAEL for reproductive development in the rat: Importance of the retention of extra animals to adulthood. Toxicol. Sci. 116(2):640-646.
Bradman, A., K. Kogut, E.A. Eisen, N.P. Jewell, L. Quirós-Alcalá, R. Castorina, J. Chevrier, N.T. Holland, D.B. Barr, G. Kavanagh-Baird, and B. Eskenazi. 2013. Variability of organophosphorous pesticide metabolite levels in spot and 24-hr urine samples collected from young children during 1 week. Environ. Health Perspect. 121(1):118-124.
Bruner-Tran, K.L., J. Gnecco, T. Ding, D.R. Glore, V. Pensabene, and K.G. Osteen. 2017. Exposure to the environmental endocrine disruptor TCDD and human reproductive dysfunction: Translating lessons from murine models. Reprod. Toxicol. 68:59-71
Burney, R.O., and L.C. Giudice. 2012. Pathogenesis and pathophysiology of endometriosis. Fertil. Steril. 98(3):511-519.
CDC (Centers for Disease Control and Prevention). 2009. Fourth National Report on Human Exposure to Environmental Chemicals. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Atlanta, GA [online]. Available: https://www.cdc.gov/exposurereport/pdf/fourthreport.pdf/ [accessed July 21, 2016].
CDC. 2015. Fourth National Report on Human Exposure to Environmental Chemicals (Updated Tables) [online]. Available: http://www.cdc.gov/biomonitoring/pdf/FourthReport_UpdatedTables_Feb2015.pdf [accessed July 1, 2016].
Connolly, R.B., G.T. Ankley, W.Y. Cheng, M.L. Mayo, D.H. Miller, E.J. Perkins, D.L. Villeneuve, and K.H. Watanabe. 2017. Quantitative adverse outcome pathways and their application to predictive toxicology. Environ. Sci. Toxicol. 51:4661-4672.
Crump, K.S., C. Van Landingham, and R.O. McClellan. 2016. Influence of alternative exposure estimates in the diesel exhaust miners study: Diesel exhaust and lung cancer. Risk Anal. 36(9):1803-1812.
Diamanti-Kandarakis, E., J.P. Bourguignon, L.C. Giudice, R. Hauser, G.S. Prins, M. Soto, R.T. Zoeller, and A.C. Gore. 2009. Endocrine-disrupting chemicals: An Endocrine Society scientific statement. Endocr. Rev. 30(4):293-342.
Donald, C.E., R.P. Scott, K.L. Blaustein, M.L. Halbleib, M. Sarr, P.C. Jepson, and K.A. Anderson. 2016. Silicone wristbands detect individuals' pesticide exposures in West Africa. R. Soc. Open Sci. 3(8):160433.
Duke, J.D., X. Han, Z. Wang, A. Subhadarshini, S.D. Karnik, X. Li, S.D. Hall, Y. Jin, J.T. Callaghan, M.J. Overhage, D.A. Flockhart, R.M. Strother, S.K. Quinney, and L. Li. 2012. Literature based drug interaction prediction with clinical assessment using electronic medical records: Novel myopathy associated drug interactions. PLoS Comput. Biol. 8(8):e1002614.
EPA (U.S. Environmental Protection Agency). 2000. CatReg Software Documentation. EPA/600/R-98/053F. Office of Research and Development, National Center for Environmental Assessment, EPA, Research Triangle Park, NC.
Eskenazi, B., P. Mocarelli, M. Warner, S. Samuels, P. Vercellini, D. Olive, L.L. Needham, D.G. Patterson Jr., P. Brambilla, N. Gavoni, S. Casalini, S. Panazza, W. Turner, and P.M. Gerthoux. 2002. Serum dioxin concentrations and endometriosis: A cohort study in Seveso, Italy. Environ. Health Perspect. 110(7):629-634.
Foster, P.M. 2014. Regulatory Forum opinion piece: New testing paradigms for reproductive and developmental toxicity—the NTP modified one generation study and OECD 443. Toxicol. Pathol. 42(8):1165-1167.
Friedman, G.D., S.F. Jiang, N. Udaltsova, J. Chan, C.P. Quesenberry, Jr., and L.A. Habel. 2009. Pharmaceuticals that cause mammary gland tumors in animals: Findings in women. Brest Cancer Res. Treat. 116(1):187-194.
Gagne, J.J., X. Han, S. Hennessy, C.E. Leonard, E.A. Chrischilles, R.M. Carnahan, S.V. Wang, C. Fuller, A. Iyer, H. Katcoff, T.S. Woodworth, P. Archdeacon, T.E. Meyer, S. Schneeweiss, and S. Toh. 2016. Successful comparison of U.S. Food and Drug Administration Sentinel analysis tools to traditional approaches in quantifying a known drug-adverse event association. Clin. Pharmacol. Ther. 100(5):558-564.
Geyh, A.S., P.T. Roberts, F.W. Lurmann, B.M. Schoell, and E.L. Avol. 1999. Initial field evaluation of the Harvard active ozone sampler for personal ozone monitoring. J. Expo. Anal. Environ. Epidemiol. 9(2):143-149.
Goodman, S.N. 2016. Aligning statistical and scientific reasoning. Science 352(6290):1180-1181.
Gore, A.C., V.A. Chappell, S.E. Fenton, J.A. Flaws, A. Nadal, G.S. Prins, J. Toppari, and R.T. Zoeller. 2015. EDC-2: The Endocrine Society’s second scientific statement on endocrine-disrupting chemicals. Endocr. Rev. 36(6):E1-E150.
Gurulingappa, H., A. Rajput, and L. Toldo. 2012. Extraction of adverse drug effects from medical case reports. J. Biomed. Semantics 3(1):15.
Hammel, S.C., K. Hoffman, T.F. Webster, K.A. Anderson, and H.M. Stapleton. 2016. Measuring personal exposure to organophosphate flame retardants using silicone wristbands and hand wipes. Environ. Sci. Technol. 50(8):4483-4491.
Harpaz, R., W. DuMouchel, N.H. Shah, D. Madigan, P. Ryan, and C. Friedman. 2012. Novel data-mining methodologies for adverse drug event discovery and analysis. Clin. Pharmacol. Ther. 91(6):1010-1021.
Hartle, J.C., A. Navas-Acien, and R.S. Lawrence. 2016. The consumption of canned food and beverages and urinary Bisphenol A concentrations in NHANES 2003-2008. Environ. Res. 150:375-382.
Hass, U., S. Christiansen, M. Axelstad, K.D. Sorensen, and J. Boberg. 2013. Input for the REACH-review in 2013 on Endocrine Disruptors. Danish Centre on Endocrine Disrupters [online]. Available: http://mst.dk/media/mst/9106721/rapport_input_for_the_reach-review.pdf [accessed November 3, 2016].
Hauben, M., S. Horn, and L. Reich. 2007. Potential use of data-mining algorithms for the detection of ‘surprise’ adverse drug reactions. Drug Saf. 30(2):143-155.
Heindel, J.J. 1988. The National Toxicology Program chemical nomination selection and testing process. Reprod. Toxicol. 2(3-4):273-279.
Hoover, R.N., M. Hyer, R.M. Pfeiffer, E. Adam, B. Bond, A.L. Cheville, T. Colton, P. Hartge, E.E. Hatch, A.L. Herbst, B.Y. Karlan, R. Kaufman, K.L. Noller, J.R. Palmer, S.J. Robboy, R.C. Saal, W. Strohsnitter, L. Titus-Ernstoff, and R. Troisi. 2011. Adverse health outcomes in women exposed in utero to diethylstilbestrol. N. Engl. J. Med. 365(14):1304-1314.
Ibrahim, H., A. Saad, A. Abdo, and A. Sharaf Eldin. 2016. Mining association patterns of drug-interactions using post marketing FDA’s spontaneous reporting data. J. Biomed. Inform 60:294-308.ICASR (International Collaboration for the Automation of System Reviews). 2015. The Vienna Principles: General Principles of Collaboration and Development in the Automation of Systematic Reviews. The First ICASR Meeting, October 2, 2015, Vienna, Austria [online]. Available: http://ebrnetwork.org/wp-content/uploads/2015/04/The-Vienna-Principles.pdf [accessed March 6, 2017].
IOM (Institute of Medicine). 2012. Breast Cancer and the Environment: A Life Course Approach. Washington, DC: The National Academies Press.
Jefferson, W.N., D.M. Chevalier, J.Y. Phelps, A.M. Cantor, E. Padilla-Banks, R.R. Newbold, T.K. Archer, H.K. Kinyamu, and C.J. Williams. 2013. Persistently altered epigenetic marks in the mouse uterus after neonatal estrogen exposure. Mol. Endocrinol. 27(10):1666-1677.
Jenkins, S., C. Rowell, J. Wang, and C.A. Lamartiniere. 2007. Prenatal TCDD exposure predisposes for mammary cancer in rats. Reprod. Toxicol. 23(3):391-396.
Kim, K.N., Y.H. Choi, Y.H. Lim, and Y.C. Hong. 2016. Urinary phthalate metabolites and depression in an elderly population: National Health and Nutrition Examination Survey 2005-2012. Environ. Res. 145:61-67.
Kim, T.S., K.K. Jung, S.S. Kim, I.H. Kang, J.H. Baek, H.S. Nam, S.K. Hong, B.M. Lee, J.T. Hong, K.W. Oh, H.S. Kim, S.Y. Han, and T.S. Kang. 2010. Effects of in utero exposure to di(n-butyl)phthalate on development of male reproductive tracts in Sprague-Dawley rats. J. Toxicol. Environ. Health A 73(21-22):1544-1559.
King, C.M., C. Barbara, A. Prentice, J.D. Brenton, and D.S. Charnock-Jones. 2016. Models of endometriosis and their utility in studying progression to ovarian clear cell carcinoma. J. Pathol. 238(2):185-196.
Lardon, J., R. Abdellaoui, F. Bellet, H. Asfari, J. Souvignet, N. Texier, M.C. Jaulent, M.N. Beyens, A. Burgun, and C. Bousquet. 2015. Adverse drug reaction identification and extraction in social media: A scoping review. J. Med. Internet Res. 17(7):e171.
Lewis, R.C., L.E. Johns, and J.D. Meeker. 2015. Serum biomarkers of exposure to perfluoroalkyl substances in relation to serum testosterone and measures of thyroid function among adults and adolescents from NHANES 2011-2012. Int. J. Environ. Res. Public Health 12(6):6098-6114.
Li, Y., H. Salmasian, S. Vilar, H. Chase, C. Friedman, and Y. Wei. 2014. A method for controlling complex confounding effects in the detection of adverse drug reactions using electronic health records. J. Am. Med. Inform. Assoc. 21(2):308-314.
Liu, C., X. Xu, and X. Huo. 2014. Anogenital distance and its application in environmental health research. Environ. Sci. Pollut. Res. Int. 21(8):5457-5464.
Macon, M.B., and S.E. Fenton. 2013. Endocrine disruptors and the breast: Early life effects and later life disease. J. Mammary Gland Biol. Neoplasia 18(1):43-61.
Makris, S.L. 2011. Current assessment of the effects of environmental chemicals on the mammary gland in guideline rodent studies by the U.S. Environmental Protection Agency (U.S. EPA), Organisation for Economic Co-operation and Development (OECD), and National Toxicology Program (NTP). Environ. Health Perspect. 119(8):1047-1052.
Marty, M.S., B. Allen, R.E. Chapin, R. Cooper, G.P. Daston, J.A. Flaws, P.N. Foster, S.L. Makris, E. Mylchreest, D. Sandler, and R.W. Tyl. 2009. Inter-laboratory control data for reproductive endpoints required in the OPPTS 870.3800/OECD 416 reproduction and fertility test. Birth Defects Res. B Dev. Reprod. Toxicol. 86(6):470-489.
McMahon, A.W., G.T. Wharton, R. Bonnel, M. DeCelle, K. Swank, D. Testoni, J.U. Cope, P.B. Smith, E. Wu, and M.D. Murphy. 2015. Pediatric post-marketing safety systems in North America: Assessment of the current status. Pharmacoepidemiol. Drug Saf. 24(8):785-792.
Mihaylov, N.L., and D.D. Perkins. 2015. Local environmental grassroots activism: Contributions from environmental psychology, sociology and politics. Behav. Sci. 5(1):121-153.
NASEM (National Academies of Sciences, Engineering, and Medicine). 2015. Application of Modern Toxicology for Predicting Acute Toxicity for Chemical Defense. Washington, DC: The National Academies Press.
NASEM. 2017. Using 21st Century Science to Improve Risk-Related Evaluations. Washington, DC: The National Academies Press.
Nayyar, T., K.L. Bruner-Tran, D. Piestrzeniewicz-Ulanska, and K.G. Osteen. 2007. Developmental exposure of mice to TCDD elicits a similar uterine phenotype in adult animals as observed in women with endometriosis. Reprod Toxicol. 23(3):326-336.
Newbold, R.R. 2004. Lessons learned from perinatal exposure to diethylstilbestrol. Toxicol. Appl. Pharmacol. 199(2):142-150.
Newbold, R.R., and J.A. McLachlan. 1982. Vaginal adenosis and adenocarcinoma in mice exposed prenatally or neonatally to diethylstilbestrol. Cancer Res 42(5):2003-2011.
Newbold, R.R., W.N. Jefferson, E. Padilla-Banks, and J. Haseman. 2004. Developmental exposure to diethylstilbestrol (DES) alters uterine response to estrogens in prepubescent mice: Low versus high dose effects. Reprod. Toxicol. 18(3):399-406.
Nikfarjam, A., A. Sarker, K. O’Connor, R. Ginn, and G. Gonzalez. 2015. Pharmacovigilance from social media: Mining adverse drug reaction mentions using sequence labeling with word embedding cluster features. J. Am. Med. Inform. Assoc. 22(3):671-681.
NRC (National Research Council). 1991. Human Exposure Assessment for Airborne Pollutants: Advances and Opportunities. Washington, DC: National Academy Press.
NRC. 2006a. Human Biomonitoring for Environmental Chemicals. Washington, DC: The National Academies Press.
NRC. 2006b. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. Washington, DC: The National Academies Press.
NRC. 2007. Toxicity Testing in the 21st Century: A Vision and a Strategy. Washington, DC: The National Academies Press.
NRC. 2009. Science and Decisions: Advancing Risk Assessment. Washington DC: The National Academies Press.
NRC. 2011. Review of the Environmental Protection Agency’s Draft IRIS Assessment of Formaldehyde. Washington, DC: The National Academies Press.
NRC. 2012. Exposure Science in the 21st Century: A Vision and a Strategy. Washington, DC: The National Academies Press.
NRC. 2014. Review of EPA’s Integrated Risk Information System (IRIS) Process. Washington, DC: The National Academies Press.
NTP (National Toxicology Program). 2015. Handbook for Conducting a Literature-Based Health Assessment Using OHAT Approach for Systematic Review and Evidence Integration. Office of Health Assessment and Translation, Division of the National Toxicology Program, National Institute of Environmental Health Sciences. January 9, 2015 [online]. Available: http://ntp.niehs.nih.gov/ntp/ohat/pubs/handbookjan2015_508.pdf [accessed September 21, 2015].
O’Connell, S.G., L.D. Kincl, and K.A. Anderson. 2014. Silicone wristbands as personal passive samplers. Environ. Sci. Technol. 48(6):3327-3335.
Oliveira, M., K. Slezakova, C. Delerue-Matos, M. do Carmo Pereira, and S. Morais. 2016. Assessment of polycyclic aromatic hydrocarbons in indoor and outdoor air of preschool environments (3-5 years old children). Environ. Pollut. 208 (Part B):382-394.
Omoike, O.E., R.C. Lewis, and J.D. Meeker. 2015. Association between urinary biomarkers of exposure to organophosphate insecticides and serum reproductive hormones in men from NHANES 1999-2002. Reprod. Toxicol. 53:99-104.
Patel, C.J., J. Bhattacharya, and A.J. Butte. 2010. An Environment-Wide Association Study (EWAS) on type 2 diabetes mellitus. PLoS ONE 5(5):e10746.
Patel, C.J., T. Yang, Z. Hu, Q. Wen, J. Sung, Y.Y. El-Sayed, H. Cohen, J. Gould, D.K. Stevenson, G.M. Shaw, X.B. Ling, and A.J. Butte. 2014. Investigation of maternal environmental exposures in association with self-reported preterm birth. Reprod. Toxicol. 45:1-7.
Perera, F.P., V. Rauh, W.Y. Tsai, P. Kinney, D. Camann, D. Barr, T. Bernet, R. Garfinkel, Y.H. Tu, D. Diaz, J. Dietrich, and R.M. Whyatt. 2003. Effects of transplacental exposure to environmental pollutants on birth outcomes in a multi-ethnic population. Environ. Health Perspect.. 111(2):201-205.
Peretz, J., L. Vrooman, W.A. Ricke, P.A. Hunt, S. Ehrlich, R. Hauser, V. Padmanabhan, H.S. Taylor, S.H. Swan, C.A. VandeVoort, and J.A. Flaws. 2014. Bisphenol A and reproductive health: Update of experimental and human evidence, 2007-2013. Environ. Health Perspect. 122(8):775-786.
Platt, R., R.M. Carnahan, J.S. Brown, E. Chrischilles, L.H. Curtis, S. Hennessy, J.C. Nelson, J.A. Racoosin, M. Robb, S. Schneeweiss, S. Toh, and M.G. Weiner. 2012. The U.S. Food and Drug Administration's Mini-Sentinel program: Status and direction. Pharmacoepidemiol. Drug Saf. 21 (Suppl. 1):1-8.
Pontes, H., M. Clement, and V. Rollason. 2014. Safety signal detection: The relevance of literature review. Drug Saf. 37(7):471-479.
Puchalski, M.A., C.M. Rogers, R. Baumgardner, K.P. Mishoe, G. Price, M.J. Smith, N. Watkins, and C.M. Lehmann. 2015. A statistical comparison of active and passive ammonia measurements collected at Clean Air Status and Trends Network (CASTNET) sites. Environ. Sci. Process. Impacts 17(2):358-369.
Reed, C.E., and S.E. Fenton. 2013. Exposure to diethylstilbestrol during sensitive life stages: A legacy of heritable health effects. Birth Defects Res. C Embryo Today 99(2):134-146.
Revesz, C., C.I. Chappel, and R. Gaudry. 1960. Masculinization of female fetuses in the rat by progestational compounds. Endocrinology 66(1):140-144.
Rier, S.E., D.C. Martin, R.E. Bowman, W.P. Dmowski, and J.L. Becker. 1993. Endometriosis in rhesus monkeys (Macaca mulatta) following chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Fundam Appl Toxicol. 21(4):433-441.
Rooney, A.A., A.L. Boyles, M.S. Wolfe, J.R. Bucher, and K.A. Thayer. 2014. Systematic review and evidence integration for literature-based environmental health science assessments. Environ. Health Perspect. 122(7):711-718.
Rudel, R.A., K.R. Attfield, J.N. Schifano, and J.G. Brody. 2007. Chemicals causing mammary gland tumors in animals signal new directions for epidemiology, chemicals testing, and risk assessment for breast cancer prevention. Cancer 109(12 Suppl.):2635-2666.
Rudel, R.A., S.E. Fenton, J.M. Ackerman, S.Y. Euling, and S.L. Makris. 2011. Environmental exposures and mammary gland development: State of the science, public health implications, and research recommendations. Environ. Health Perspect. 119(8):1053-1061.
Rudel, R.A., J.M. Ackerman, K.R. Attfield, and J.G. Brody. 2014. New exposure biomarkers as tools for breast cancer epidemiology, biomonitoring, and prevention: A systematic approach based on animal evidence. Environ. Health Perspect. 122(9):881-895.
Saillenfait, A.M., J.P. Sabaté, and F. Gallissot. 2008. Diisobutyl phthalate impairs the androgen-dependent reproductive development of the male rat. Reprod Toxicol. 26(2):107-115.
Salazar-Martinez, E., P. Romano-Riquer, E. Yanez-Marquez, M.P. Longnecker, and M. Hernandez-Avila. 2004. Anogenital distance in human male and female newborns: A descriptive, cross-sectional study. Environ. Health 3(1):8.
Sarkar, U., A. Lopez, J.H. Maselli, and R. Gonzalez. 2011. Adverse drug events in U.S. adult ambulatory medical care. Health. Serv. Res 46(5):1517-1533.
Sarker, A., and G. Gonzalez. 2015. Portable automatic text classification for adverse drug reaction detection via multi-corpus training. J. Biomed. Inform. 53:196-207.
Sather, M.E., and K. Cavender. 2016. Trends analyses of 30 years of ambient 8 hour ozone and precursor monitoring data in the South Central U.S.: Progress and challenges. Environ. Sci. Process. Impacts 18(7):819-831.
Sathyanarayana, S., R. Grady, J.B. Redmon, K. Ivicek, E. Barrett, S. Janssen, R. Nguyen, and S.H. Swan. 2015. Anogenital distance and penile width measurements in The Infant Development and the Environment Study (TIDES): Methods and predictors. J. Pediatr. Urol. 11(2):76.e1-6.
Scholze, M., and A. Kortenkamp. 2007. Statistical power considerations show the endocrine disruptor low-dose issue in a new light. Environ. Health Perspect. 115(Suppl. 1):84-90.
Schotland, P., N. Bojunga, A. Zien, M.N. Trame, and L.J. Lesko. 2016. Improving drug safety with a systems pharmacology approach. Eur. J. Pharm. Sci. 94:84-92.
Schwarzman, M.R., J.M. Ackerman, S.H. Dairkee, S.E. Fenton, D. Johnson, K.M. Navarro, G. Osborne, R.A. Rudel, G.M. Solomon, L. Zeise, and S. Janssen. 2015. Screening for chemical contributions to breast cancer risk: A case study for chemical safety evaluation. Environ. Health Perspect. 123(12):1255-1264.
Shetty, K.D., and S.R. Dalal. 2011. Using information mining of the medical literature to improve drug safety. J. Am. Med. Inform. Assoc. 18(5):668-674.
Skakkebaek, N.E., J. Toppari, O. Söder, C.M. Gordon, S. Divall, and M. Draznin. 2011. The exposure of fetuses and children to endocrine disrupting chemicals: A European Society for Paediatric Endocrinology (ESPE) and Pediatric Endocrine Society (PES) call to action statement. J. Clin. Endocrinol. Metab. 96(10):3056-3058.
Slayden, O.D. 2016. Translational in vivo models for women's health: The nonhuman primate endometrium—A predictive model for assessing steroid receptor modulators. Handb. Exp. Pharmacol. 232:191-202.
Smith, M.T., K.Z. Guyton, C.F. Gibbons, J.M. Fritz, C.J. Portier, I. Rusyn, D.M. DeMarini, J.C. Caldwell, R.J. Kavlock, P.F. Lambert, S.S. Hecht, J.R. Bucher, B.W. Stewart, R.A. Baan, V.J. Cogliano, and K. Straif. 2016. Key characteristics of carcinogens as a basis for organizing data on mechanisms of carcinogenesis. Environ. Health Perspect. 124(6):713-721.
Sobus, J.R., R.S. DeWoskin, Y.M. Tan, J.D. Pleil, M.B. Phillips, B.J. George, K. Christensen, D.M. Schreinemachers, M.A. Williams, E.A. Cohen Hubal, and S.W. Edwards. 2015. Uses of NHANES biomarker data for chemical risk assessment: Trends, challenges, and opportunities. Environ. Health Perspect. 123(10):919-927.
Solomon, P.A., and C. Sioutas. 2008. Continuous and semicontinuous monitoring techniques for particulate matter mass and chemical components: A synthesis of findings from EPA's Particulate Matter Supersites Program and related studies. J. Air Waste Manage. Assoc.58(2):164-195.
Subramaniam, R.P., K.S. Crump, C. Van Landingham, P. Whiate, C. Cehn, and P.M. Schlosser. 2007. Uncertainties in the CIIT model for formaldehyde-induced carcinogenicity in the rat: A limited sensitivity analysis-I. Risk Anal. 27(5):1237-1254.
Swan, S.H., K.M. Main, F. Liu, S.L. Stewart, R.L. Kruse, A.M. Calafat, C.S. Mao, J.B. Redmon, C.L. Ternand, S. Sullivan, and J.L. Teague. 2005. Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ. Health Perspect. 113(8):1056-1061.
Tan, Y.M., J. Sobus, D. Chang, R. Tornero-Velez, M. Goldsmith, J. Pleil, and C. Dary. 2012. Reconstructing human exposures using biomarkers and other “clues.” J. Toxicol. Environ. Health B Crit. Rev.15(1):22-38.
Tatonetti, N.P., G.H. Fernald, and R.B. Altman. 2012. A novel signal detection algorithm for identifying hidden drug-drug interactions in adverse event reports. J. Am. Med. Inform. Assoc. 19(1):79-85.
Taylor, R.N., M.A. Kane, and N. Sidell. 2015. Pathogenesis of endometriosis: Roles of retinoids and inflammatory pathways. Semin. Reprod. Med. 33(4):246-256.
Teeguarden, J., S. Hanson-Drury, J.W. Fisher, and D.R. Doerge. 2013. Are typical human serum BPA concentrations measurable and sufficient to be estrogenic in the general population? Food. Chem. Toxicol. 62:949-963.
Teeguarden, J.G., N.C. Twaddle, M.I. Churchwell, X. Yang, J.W. Fisher, L.M. Seryak, and D.R. Doerge. 2015. 24-hour human urine and serum profiles of bisphenol A: Evidence against sublingual absorption following ingestion in soup. Toxicol. Appl. Pharmacol. 288(2):131-142.
Thayer, K.A., and P.M. Foster. 2007. Workgroup report: National Toxicology Program Workshop on Hormonally Induced Reproductive Tumors—Relevance of Rodent Bioassays. Environ. Health Perspect.115(9):1351-1356.
Thayer, K.A., D.R. Doerge, D. Hunt, S.H. Schurman, N.C. Twaddle, M.I. Churchwell, S. Garantziotis, G.E. Kissling, M.R. Easterling, J.R. Bucher, and L.S. Birnbaum. 2015. Pharmacokinetics of bisphenol A in humans following a single oral administration. Environ. Int. 83:107-115.
Trifirò, G., A. Pariente, P.M. Coloma, J.A. Kors, G. Polimeni, G. Miremont-Salamé, M.A. Catania, F. Salvo, A. David, N. Moore, A.P. Caputi, M. Sturkenboom, M. Molokhia, J. Hippisley-Cox, C.D. Acedo, J. van der Lei, and A. Fourrier-Reglat. EU-ADR group. 2009. Data mining on electronic health record databases for signal detection in pharmacovigilance: Which events to monitor? Pharmacoepidemiol. Drug Saf. 18(12):1176-1184.
Tsai, D.H., N. Amyai, P. Marques-Vidal, J.L. Wang, M. Riediker, V. Mooser, F. Paccaud, G. Waeber, P. Vollenweider and M. Bochud. 2012. Effects of particulate matter on inflammatory markers in the general adult population. Part Fibre Toxicol. 9:24.
Vandenberg, L.N., S. Ehrlich, S.M. Belcher, N. Ben-Jonathan, D.C. Dolinoy, E.R. Hugo, P.A. Hunt, R.R. Newbold, B.S. Rubin, K.S. Saili, A.M. Soto, H.-S. Wang, and F.S. vom Saal. 2013. Low dose effects of bisphenol A. Endocr. Disruptors 1(1):e26490.
Vercellini, P., P. Viganò, E. Somigliana, and L. Fedele. 2014. Endometriosis: Pathogenesis and treatment. Nat. Rev. Endocrinol. 10(5):261-275.
Wang, W., K. Haerian, H. Salmasian, R. Harpaz, H. Chase, and C. Friedman. 2011. A drug-adverse event extraction algorithm to support pharmacovigilance knowledge mining from PubMed citations. AMIA Annu. Symp. Proc. 2011:1464-1470.
Wetmore, B.A., J.F. Wambaugh, S.S. Ferguson, M.A. Sochaski, D.M. Rotroff, K. Freeman, H.J. Clewell III, D.J. Dix, M.E. Andersen, K.A. Houck, B. Allen, R.S. Judson, R. Singh, R.J. Kavlock, A.M. Richard, and R.S. Thomas. 2012. Integration of dosimetry, exposure, and high-throughput screening data in chemical toxicity assessment. Toxicol. Sci. 125(1):157-174.
WHO (World Health Organization). 2002. The Importance of Pharmacovigilance—Safety Monitoring of Medicinal Products [online]. Available: http://apps.who.int/medicinedocs/en/d/Js4893e/1.html [accessed July 21, 2016].
Winnenburg, R., A. Sorbello, A. Ripple, R. Harpaz, J. Tonning, A. Szarfman, H. Francis, and O. Bodenreider. 2015. Leveraging MEDLINE indexing for pharmacovigilance—Inherent limitations and mitigation strategies. J. Biomed. Inform. 57:425-435.
Woodruff, T.J., and P. Sutton. 2014. The Navigation Guide systematic review methodology: A rigorous and transparent method for translating environmental health science into better health outcomes. Environ. Health Perspect. 122(10):1007-1014.
Yang, X., D.R. Doerge, J.G. Teeguarden, and J.W. Fisher. 2015. Development of a physiologically based pharmacokinetic model for assessment of human exposure to bisphenol A. Toxicol. Appl. Pharmacol. 289(3):442-456.
Yoon, M., J.L. Campbell, M.E. Andersen, and H.J. Clewell. 2012. Quantitative in vitro to in vivo extrapolation of cell-based toxicity assay results. Crit. Rev. Toxicol. 42(8):633-652.
Zoller, H.M. 2017. Health activism targeting corporations: A critical health communication perspective. Health Commun. 32(2):219-229.