IRIS Public Meeting (Jun 2015)

EPA hosted an IRIS public science meeting to provide an opportunity for the public to give input and participate in an open discussion regarding the scoping and problem formulation materials that were prepared for IRIS chemicals prior to the development of the draft assessment.

This included the following chemical: Polychlorinated Biphenyls (PCBs): Effects Other Than Cancer

Meeting Objective:

IRIS Public Science Meetings allow the public the opportunity to provide input and participate in discussions about problem formulation, preliminary assessment materials, and draft IRIS assessments. The objective of this public meeting was to obtain input from the scientific community and the general public on problem formulation materials that frame the scientific issues that were the focus of the systematic review of potential health hazards of PCBs (effects other than cancer).

In October 2014, EPA’s IRIS Program announced an agreement with the National Academies’ National Research Council (NRC) to arrange for independent experts to attend IRIS public science meetings to provide input on the science underlying the development of IRIS assessments. The NRC will select a limited number of experts to join in the discussion of the key science topics with EPA and those that have registered as discussants on the EPA website. Participation of these experts is meant to be supplementary and stakeholders and the public should continue to register as discussants on the EPA website and are also invited to suggest additional science topics by sending an e-mail to


The meeting was held on June 17-18, 2015 from 8:30 am - 5:30pm Eastern Time.


The meeting was held in the EPA Conference Center at 2777 South Crystal Drive, Arlington, Virginia 22202. The meeting was also available by webinar/teleconference.

Top of Page

Meeting Agenda:

On June 17-18, 2015, EPA hosted a public meeting/webinar in Arlington, VA, to provide an opportunity for the public to give input and participate in an open discussion regarding several IRIS chemical assessments that are under development. See the final agenda:

Meeting Documents:

In March 2014, EPA released the draft literature searches and associated search strategies, evidence tables, and exposure response arrays for this chemical to obtain input from stakeholders and the public prior to developing the draft IRIS assessments for polychlorinated biphenyls (PCBs). Specifically, EPA was interested in comments on the following:

  • Draft literature search strategies
    • The approach for identifying studies
    • The screening process for selecting pertinent studies
    • The resulting list of pertinent studies
  • Preliminary evidence tables
    • The process for selecting studies to include in evidence tables
    • The quality of the studies in the evidence tables

The literature search strategy, which describes the processes for identifying scientific literature, contains the studies that EPA considered and selected to include in the evidence tables. The preliminary evidence tables and exposure-response arrays present the key study data in a standardized format. The evidence tables summarize the available critical scientific literature. The exposure-response figures provide a graphical representation of the responses at different levels of exposure for each study in the evidence table.

Polychlorinated Biphenyls (PCBs): Geniece Lehmann, Assessment Manager

Key Science Topics:

Science Topic 1: Impact of Congener Profile on the Toxicity of PCB Mixtures (including MOA)

PCB congeners differ not only structurally but also qualitatively and quantitatively with respect to biological responses. Humans are environmentally exposed to PCBs as complex mixtures of congeners. EPA guidance on conducting risk assessment of chemical mixtures recommends several quantitative approaches, depending upon the type of available data. The preferred approach is to use toxicity data on the mixture of concern. Alternatively, when toxicity data are not available for the mixture of concern, use of toxicity data on a “sufficiently similar” mixture is recommended.

Out of all the possible congener combinations that may exist, a relatively small subset of complex PCB mixtures has been tested in animal studies. Most animal studies have administered commercial PCB mixtures (e.g., “Aroclors”). One disadvantage of these studies is that the congener profiles of commercial PCB mixtures do not match those that occur in the environment. Prior to human exposure, commercial mixtures in the environment undergo processes such as volatilization and preferential bioaccumulation, which dramatically alter a PCB mixture’s congener profile.

For example, oral exposures to PCBs occur primarily via consumption of contaminated foods that contain mixtures of persistent PCB congeners that have been biomagnified through the food chain. Biomagnification of PCBs roughly increases with higher congener chlorination; the PCB mixtures most often consumed by humans consist largely of PCBs with 5, 6, or 7 chlorine substitutions (e.g., PCBs 138, 153 and 180). Meanwhile, the prominent PCB congeners found in air samples are not determined by biomagnification but rather by volatility and the congener profile of the source material. Volatility is greatest for the lower chlorinated congeners (i.e., 1–4 chlorine substitutions); thus, the proportion of these congeners making up an inhalation exposure to PCBs may be relatively large compared to what might be found for an oral exposure. However, inhalation (as well as dermal and oral) exposure to higher chlorinated congeners bound to dust may also occur.

A few studies have utilized mixtures of PCB congeners formulated to mimic an environmental exposure (e.g., formulations representing the congeners found in human milk or in contaminated fish or soil); for a typical oral exposure, these mixtures may best represent the “mixture of concern.” Thus, these studies may be preferred for human health risk assessment because they minimize the uncertainty that results from using research on one PCB mixture to assess the risk from exposure to a different mixture. However, despite the fact that their congener profiles do not precisely replicate that of an environmental PCB mixture, studies administering commercial PCB mixtures have generally observed toxicological effects within the same dose range as environmental mixtures.

Furthermore, commercial PCB mixtures contain overlapping groups of congeners that, together, span the range of congeners most frequently found in environmental mixtures, including those found in air. Therefore, commercial PCB mixtures may be “sufficiently similar” to environmental mixtures, and studies using commercial PCB mixtures may be useful to support human health hazard identification and dose-response assessment for PCBs in the environment. Based on the available data, EPA intends to evaluate the impact of congener profile on toxicity, (1) determining the relative toxic potencies of complex PCB mixtures (e.g., environmental and commercial) for various non-cancer health effects, (2) evaluating the implications of using toxicological data from a limited set of PCB mixtures for human health risk assessment in a wide variety of exposure contexts (e.g., breastfeeding infant exposure to PCBs in human milk, exposure to PCBs from fish consumption, inhalation exposure to PCBs in contaminated indoor air), and (3) identifying important considerations for using media-specific data (e.g., soil, groundwater, sediment, fish) collected using various analytical techniques (e.g., Aroclor analyses, measures of total PCBs, and congener or isomer analyses) with toxicity information to be provided in the assessment.

EPA sought public comment and discussion on (1) the composition of various PCB mixtures (e.g., commercial mixtures and mixtures found in environmental media), (2) the impact of PCB congener profile on a given mixture’s toxicity, (3) approaches for evaluating toxicological similarity across mixtures, and (4) recommended approaches for using a variety of PCB analytical techniques in risk assessment applications.

Science Topic 1: Presentations

Top of Page

Science Topic 2: Evaluation of Epidemiological Studies for PCB Dose-Response Assessment

Human data are generally preferred over animal data for human health hazard identification and dose-response assessment. However, certain study design and methodologic considerations are important for determining which human studies, if any, are appropriate for use in an assessment: documentation of study design, methods, population characteristics, and results; definition and selection of the study group and comparison group; ascertainment of disease or health effect; duration of exposure and follow-up and adequacy for assessing the occurrence of effects; sample size and statistical power to detect anticipated effects; participation rates and potential for selection bias as a result of the achieved participation rate; potential confounding and other sources of bias addressed in the study design or in the analysis of results; ascertainment of exposure to the chemical or mixture under consideration; and characterization of exposure during critical periods of development.

Of particular concern for epidemiological studies of PCBs is the common practice of characterizing exposure using current measures of body burden, often relying on a limited number of measured congeners. This approach to exposure assessment may be appropriate for some applications, but may be of limited utility for characterizing the extent of human PCB exposure and the relationship between exposure and effect:

  1. Current body burden reflects cumulative exposure to persistent PCB congeners, but only recent exposure to labile congeners. The half-life and elimination characteristics of PCB congeners vary significantly. And, the relative contributions of less-persistent and more-persistent PCB congeners to toxicological outcomes are poorly-defined. In recent years, a better appreciation has been gained for the full scope of human exposure to PCBs in the general environment, including the potential for significant inhalation and dermal exposure to lower-chlorinated, less-persistent congeners. Especially given this new understanding, it seems likely that cross-sectional estimates of body burden capture, at best, only a portion of past exposure levels which may have precipitated observed health effects.
  2. Most current body burden estimates rely on only a subset of PCB congeners selected because of their relative occurrence in biological samples and/or the ability to detect them using a given analytical method – not because of their biological activity or their potential to induce a particular health effect. Again, use of this approach results in an incomplete exposure assessment that may easily miss important relationships between exposure and effect.
  3. Even for persistent congeners that are routinely measured in epidemiological studies (e.g., PCBs 138, 153 and 180), a current, cross-sectional estimate of body burden may not be useful for assessing exposure during a time period critical for the development of a particular toxicological outcome (e.g., developmental outcomes known to be sensitive to PCB exposure). Depending on the endpoint of concern, the timing of exposure could be just as important as the magnitude. It is generally possible to envision several different exposure scenarios that could lead to the same current PCB body burden. And, for each scenario, although the resulting body burden is the same, the toxicological implications could be very different.

Altogether, these issues may lead to a significant potential for exposure misclassification in epidemiological studies of PCBs that rely entirely on measures of body burden for exposure assessment. Despite this potential limitation, most cohorts studied have revealed adverse health effects associated with PCB exposure, including developmental neurobehavioral outcomes, thyroid hormone disruption, immunological effects, and reduced birth weight.

These studies provide important evidence for hazard identification of human health effects. However, to define the quantitative dose-response relationship between PCBs and associated health effects, EPA will consider not only human data, but also data from toxicological studies in animals, where the source, level and timing of exposure are known with greater certainty.

EPA is seeking public comment and discussion on the database of epidemiological studies investigating potential associations between PCB exposure and health effects, including discussion of the appropriate use of these studies to support hazard identification and/or dose-response assessment, considering important study design and methodologic aspects, especially exposure assessment methods.

Science Topic 2: Presentations

Top of Page

Science Topic 3: Potential for Hazard Identification and Dose-Response Assessment for PCB Exposure via Inhalation

There is evidence to suggest that PCB inhalation may pose a hazard to human health. Inhalation exposure to PCBs occurs in both occupational groups and members of the general population, especially those who spend time in indoor settings where PCB sources exist. For example, children can be exposed via inhalation of indoor air in schools where PCB-contaminated caulk or other building materials are present.

However, the database of studies investigating health effects resulting from PCB exposure consists primarily of oral exposure studies. It is not clear whether the existing database of inhalation studies will be sufficient for human health risk assessment for inhalation exposure to PCBs. In cases such as this, data from oral exposure studies may be considered to support the assessment of human health risk from inhalation exposure, using route-to-route extrapolation. This extrapolation is sometimes used for chemicals that are not expected to

  1. have different toxicity by the oral and inhalation routes,
  2. be impacted significantly by first-pass metabolism, nor
  3. cause respiratory (portal of entry) effects.

PCBs may generally meet these criteria; however, the congener content of volatilized PCB mixtures is often, but not always, skewed toward lower-chlorinated congeners (i.e., those with ≤ 4 chlorine substitutions) compared with the congener content of a PCB mixture likely to be present in contaminated fish, human milk, or some of the Aroclor mixtures administered in oral exposure studies.

It is not clear whether such differences in congener profile translate into meaningful differences in toxicity between the two exposure routes (also discussed in the context of Science Topic 1: Impact of Congener Profile on the Toxicity of PCB Mixtures). Because of the potential for widespread human exposure to PCBs by the inhalation route, EPA intends to gather all available information on the health consequences of PCB inhalation and to explore all options for using the available data for hazard identification and dose-response assessment for inhalation exposures.

EPA is seeking public comment and discussion of potential options for conducting a dose-response assessment for PCB inhalation exposure, including the use of data from available PCB inhalation studies, the route-to-route extrapolation from oral PCB exposure data, or additional options.

Science Topic 3: Presentations

Top of Page

Science Topic 4: Suitability of Available Toxicokinetic Models for Reliable Route-to-Route, Interspecies, and/or Intraspecies Extrapolation

Because this assessment will address non-cancer hazards associated with exposure to complex PCB mixtures, EPA intends to evaluate available toxicokinetic models for their ability to predict the dose metrics of such mixtures. Lipophilicity, binding to liver proteins (e.g., cytochromes, AhR), and rate of elimination (due to metabolism or fecal excretion) are the main determinants of PCB pharmacokinetics. Variation of these pharmacokinetic determinants among individual PCBs limits the application of congener-specific models in the assessment of a complex PCB mixture. A single set of parameters to describe these determinants for the complex mixture may not be justifiable because significant individual pharmacokinetic variation has been observed for different PCB congeners. Additionally, possibilities of pharmacokinetic interaction, such as competition at binding sites or synergy in the case of induction of enzymes, may exist between PCB congeners in a complex mixture.

EPA is seeking public comment and discussion on the evaluation and possible use of toxicokinetic models for addressing route-to-route, interspecies, and intraspecies extrapolation of complex PCB mixtures. Such models could be especially useful to support a route-to-route extrapolation approach for dose-response assessment of PCB inhalation exposure (see Science Topic 3) or to estimate an infant lactational PCB exposure resulting from a given maternal exposure (see Science Topic 5).

Science Topic 4: Presentations

Top of Page

Science Topic 5: Potential Toxicokinetic Models or Methods to Inform Evaluations of Human Susceptibility and Estimate the Relationship between Continuous Daily Maternal PCB Intake and Milk PCB Concentrations in Humans

PCBs accumulate in body lipids and can be transferred to infants via breast milk, presenting a critically important challenge for human health risk assessment. This lactational exposure occurs at higher levels and over a shorter time period compared to maternal exposure, which occurs over the long-term prior to and during pregnancy and lactation. In addition, because of the relatively small size of a nursing infant, this high exposure may lead to PCB levels in blood and tissues of the infant that far exceed those in the mother. Furthermore, developmental effects have been observed in humans and animals exposed to PCBs via lactation.

Therefore, breastfeeding infants represent a life stage and population uniquely susceptible to the adverse health effects of PCBs by virtue of both increased exposure and vulnerability to potential disruption of ongoing developmental processes. In recognition of the potential for early life susceptibility to PCB exposure, EPA intends to gather all available information on the lactational transfer of PCBs and to develop methods to estimate the relationship between long-term maternal PCB exposure and consequent exposure in a breastfeeding infant.

EPA is seeking public comment and discussion of biological modeling approaches (including modeling of early life exposures, especially lactational exposures to lipophilic environmental chemicals) that can be used to quantify the relationship between maternal PCB dose and lactational dose in humans.

Science Topic 5: Presentations