The tissue residue approach for toxicity assessment: Findings and critical reviews from a Society of Environmental Toxicology and Chemistry Pellston Workshop

Abstract

Over the past few years, the “critical body residue” approach for assessing toxicity based on bioaccumulated chemicals has evolved into a more expansive consideration of tissue residues as the dose metric when defining dose–response relationships, evaluating mixtures, developing protective guidelines, and conducting risk assessments. Hence, scientists refer to “tissue residue approach for toxicity assessment” or “tissue residue-effects approach” (TRA) when addressing ecotoxicology issues pertaining to tissue (or internal) concentrations. This introduction provides an overview of a SETAC Pellston Workshop held in 2007 to review the state of the science for using tissue residues as the dose metric in environmental toxicology. The key findings of the workshop are presented, along with recommendations for research to enhance understanding of toxic responses within and between species, and to advance the use of the TRA in assessment and management of chemicals in the environment. Integr Environ Assess Manag 2011;7:2–6. © 2010 SETAC

INTRODUCTION

In June 2007, a group of 39 scientists from 9 countries convened at the Sleeping Lady Resort in Leavenworth, Washington, USA to evaluate the utility of tissue concentrations (internal concentrations) as the dose metric for characterizing toxic effects. The invited participants were selected from several countries that support active research programs for tissue residue toxicity and have regulatory interests in the subject area. This area of research and application has colloquially been known as the “critical body residue” approach; however, recently it has evolved into a more general concept that considers tissue residues as the dose metric when characterizing dose–response relationships, evaluating mixture toxicity, developing guidelines to protect organisms, and conducting risk assessments. We now use the phrase “tissue residue approach for toxicity assessment” or “tissue residue-effects approach” (TRA) to group all subtopics in toxicology that consider tissue or internal concentrations as the dose and its subsequent interpretation, evaluation, and application for environmental protection. Over the last 20 y, the determination of adverse effects based on tissue residues of toxicants has advanced significantly in theory and application. Because researchers continue to compile data, form theories, test hypotheses, and produce generalizations, we felt that a critical evaluation of the field was necessary.

At this workshop, conducted under the auspices of the Society of Environmental Toxicology and Chemistry, the participants were divided into 5 work groups to examine the state of the science surrounding tissue residue toxicity. These groups addressed the following aspects of the TRA:

• Scientific underpinnings (Group 1)

• Mechanisms and mixtures (Group 2)

• TRA applications (Group 3)

• Organic and organometallic compounds (Group 4)

• Metals and metalloids (Group 5)

Due to the strong overlap among groups, interwork-group exchange was highly encouraged. Most of the focus at the workshop was on aquatic species; however, terrestrial species were also considered. As noted at the workshop, many of the concepts and applications are broad-based and apply to all taxa.

The fundamental objective for the workshop was to clarify what is known about this underutilized tool for characterizing toxic responses and managing chemical contamination. One impetus for the workshop was a belief that monitoring and assessing chemical risks could be improved by applying residue-based concepts to enhance existing approaches. In many cases, traditional toxicity tests, criteria development, and toxicity assessments will benefit from having tissue residue metrics as the primary focus. Our goal was to demonstrate how data on tissue concentrations enhance predictions and reduce the limitations of traditional approaches for defining ecological risks from chemical contamination. We also strove to identify deficiencies in the tissue residue approach that may be overcome with additional research, theoretical advancement, and model development. Additional charges to workshop participants were to discuss and develop guidance for applying the TRA and to suggest new directions that could provide more sophisticated assessment tools. Another important outcome for the workshop was to provide guidance on how the TRA could be applied to enhance and bridge the gap between science and environmental management and regulation.

In this issue of Integrated Environmental Assessment and Management, we present the results from our collaborative review. The 6 review articles that follow each address a major aspect of the TRA and provide a stand-alone assessment, evaluation, synthesis, and specific recommendations. Also, we provide numerous examples of toxicants or classes of compounds that improve toxicity assessment under the TRA and some examples of chemicals that do not. A large number of important environmental contaminants have not been examined yet; therefore, no decision regarding their appropriateness for the TRA can be made. We also considered chemical classes and their shared properties, a strength of the TRA vis-à-vis the enormous and growing list of chemicals found in the environment. Based on this advantageous feature, we expect a more streamlined assessment of the numerous compounds being detected in environmental samples when the TRA is applied. Knowledge regarding toxicant mechanism of action and comparison to similar compounds with existing tissue residue data should allow a more rapid evaluation and, hopefully, should expedite policy determinations for regulatory control.

A conceptual advantage of the TRA is that tissue residue toxicity metrics are likely to be less variable among species and environmental conditions compared to those responses expressed as a function of an ambient exposure concentration (water, sediment-soil, or prey). When toxicity is defined in terms of tissue concentrations, the variability is often reduced substantially because the toxicokinetics and bioavailability characteristics for that compound are incorporated in the tissue residue determinations. This feature specifically has been demonstrated for a few toxicants; however, general applicability for this approach needs to be explored further.

One important example is the lethal toxicity induced by chemicals exhibiting the baseline mode of toxic action. A large number, possibly hundreds, of nonpolar organic compounds cause mortality within a very narrow range of whole-body tissue concentrations (≈2–8 mmol/g wet weight or about 50 mmol/g lipid) in small aquatic organisms, while exposure-based metrics (e.g., lethal concentration, 50% [LC50] values) span approximately 6 orders of magnitude. The range for polar organic compounds acting by the same mode of action is also narrow, but lower (≈0.6–2 mmol/g). This large number of compounds acting by the (ostensibly) common toxic mechanism of membrane disruption is a testament to the power of using the tissue residue approach to reduce variability across numerous species, chemicals, and environmental conditions. Finding similar patterns, especially among specific-acting chemicals, is an important and continuing endeavor for the TRA.

In addition to reducing variability in toxicity metrics, the TRA has several other advantageous features. Namely, this approach provides information and procedures that are useful for defining causal effects, examining mixture toxicity, quantifying interspecific responses, assessing pulse-exposure toxicity, characterizing delayed responses, evaluating organismal health in the field, performing forensic analysis, and developing scientifically defensible tissue-based guidelines or criteria that also can be translated to water and sediment-soil concentrations. These topics are addressed in greater depth in the 6 review articles presented in this issue.

Anecdotal observations on the effects of contaminants in organisms are found throughout the ages; however, the most significant advances occurred in the early 1900s, starting with the Meyer-Overton hypothesis that addressed the narcotic effects of organic compounds in tissues (Lipnick 1995). Additional advances in the 1940s and 1950s focused mainly on narcosis toxicity; however, these studies helped lay the groundwork for future theoretical advances in the field for other types of toxicants. Throughout most of the 20th century researchers measured various organic compounds and metals in myriad organisms, but these were often considered in terms of food web accumulations. Although wildlife and aquatic toxicologists shared the same goal, they followed somewhat dissimilar paths regarding tissue residue toxicity. Often, dead birds and other species were on hand; hence, measuring tissue concentrations became an important diagnostic tool to determine how they died. Wildlife toxicologists were far ahead of those working in aquatic systems because they began to correlate adverse effects and concentrations of organochlorine pesticides, Hg, and Pb in the organs of birds starting in the 1950s. Subsequently, it became common to conduct controlled laboratory studies with wildlife to determine the magnitude of adverse concentrations that could then be compared to concentrations in species collected from the field (Rattner et al. 2011). During this period, sporadic articles were published by aquatic toxicologists promoting the virtues of tissue residue toxicity; however, their approach was not widely embraced, mainly because of the emphasis on exposure to contaminants in water and sediment.

In the early 1990s a more in-depth analysis of tissue residue toxicity was considered for a variety of chemicals and modes of action (McCarty 1991; McCarty and Mackay 1993). After these and other reviews were published, the scientific literature addressing tissue residue toxicity increased considerably, especially in a broader context that included examinations for multiple species and specific-acting chemicals. Several studies on individual chemicals or classes of compounds have been published, and most of these have been cited in review articles published over the last 15 y, each exploring various theoretical, mechanistic, and application aspects of tissue residue toxicity (van Wezel and Opperhuizen 1995; Keith 1996; Barron et al. 2002; Escher et al. 2002, 2004; Traas et al. 2004; Meador 2006; Meador et al. 2008; Rattner et al. 2011).

The main applications for the TRA include assessing mixture toxicity and ecological risk, evaluating species sensitivity, and determining guidelines and criteria. Recently, the TRA has been explored more widely and applied in various situations, such as Superfund site evaluations and endangered species consultations in the United States. In other countries, scientists are considering the TRA as a means to enhance regulatory action. The deliberations from the Pellston workshop presented in this issue will add significantly to the knowledge base on this subject, especially in the areas of theory, scientific rationale, and application, which were ripe for review and critical evaluation.

KEY FINDINGS

Table 1 summarizes the key themes and concepts developed at the workshop and presented in the 6 review articles.

Table 1. Summary of findings from the SETAC Pellston Workshop on Tissue Residue Approach for Toxicity Assessment

• Information on tissue residues, especially the biologically effective dose, can reduce the variability in toxicity metrics by orders of magnitude for some environmental contaminants.

• The number of factors and characteristics that influence the interpretation of tissue residue toxicity metrics are far reduced over that seen for exposure-based metrics. The exceptions are lipid dynamics, intrinsic potency, processes that alter toxicant potency, metabolites, phototoxicity, organism health, and species and life-stage specific sensitivities.

• All currently available toxicity metrics are based on dose-surrogate models. These are based on the implicit assumption that dose (ambient exposure, whole-body, organ, or cellular concentrations) is proportional to the toxicant–receptor interaction. The TRA enhances our focus on selecting the appropriate and least influenced surrogate for toxicity assessments.

• A general, broadly applicable classification scheme for modes and mechanisms of toxic action is a crucial component for understanding tissue residue toxicity, addressing mixture toxicity, and reducing the observed variability among species and similar toxicants.

• Assessment of chemical mixtures may be simplified when considering tissue residues; however, we are just beginning to collect the necessary data needed to advance our understanding of toxic interactions.

• Because bioaccumulation and bioavailability exhibit a massive influence on exposure-based toxicity metrics for solutions of multiple chemicals, the TRA will be an important approach to reducing the enormous complexity inherent in toxicity assessment of chemical mixtures.

• Separating nontarget, nontoxic chemical fractions within the organism from the biologically active fractions for specific-acting organics and metals and metalloids will improve accuracy in toxicity evaluations.

• Existing tissue residue toxicity databases provide a workable foundation; however, improvements are needed to ensure consistency and quality control in the compilation and application of such data.

• A key technical and practical issue that must be resolved concerns definition and quantification of relationships between ambient exposures, total body and organ tissue residues, and the toxicologically effective dose.

• In some cases, an exposure-based toxicity metric (e.g. LC50 or EC50) can be reliably paired (time- and response-matched) with a bioaccumulation factor (e.g., BCF, BAF, or BSAF) to derive a comparable tissue residue toxicity metric.

• Time should always be considered when characterizing tissue residue toxicity; however for some species and toxicants, a critical body residue can be time-independent.

• Comparison of tissue residue toxicity metrics across species should be examined in terms of taxa, response, and endpoint-specific values and parameters (e.g., ER10 for growth in fish). Mixing any of these may introduce unnecessary variability and bias.

• Normalizing concentrations of neutral organic compounds to lipid content reduces variability within and between species and can improve toxicity relationships for some lipid soluble compounds. Accurate quantification of various lipid classes and appropriate application is an important aspect of this procedure.

• Focused research is needed on genetic adaptation and physiological acclimation, 2 processes that may be responsible for differences in toxic potency, toxicokinetics, and variable tissue residue toxicity values for some species and chemicals.

• Whole-body and organ tissue concentrations for most metals are not an appropriate dose surrogate; however tissue concentrations are often informative for exposure and can be related to site-specific toxic effects for certain metal–taxa combinations.

• Proportionality among exposure dose, tissue residues, and the biologically effective dose for metals often breaks down resulting in disparity in species sensitivity distributions for toxicity metrics based on exposure concentrations and those based on tissue residues.

• In many cases, the rate of accumulation and route of exposure (ventilation or dietary) can influence tissue residue toxicity metrics for metals. Therefore, it is important to compare the results for all relevant routes of exposure in laboratory experiments when characterizing tissue residue toxicity.

• Some of the best TRA examples for specific-acting chemicals are found with the organometals: methylmercury, organoselenium, and organotins. Recent compilations of tissue residue data have resulted in tissue quality guidelines or threshold toxicity values for all 3 groups.

• The TRA should not be viewed as a replacement for conventional exposure or dose-based methods for risk assessment. On the contrary, application of TRA concepts can be particularly beneficial when used to complement conventional assessment approaches, because each approach contains different strengths and limitations.

• A review of the tissue residue-based toxicity literature indicates that a variety of extrapolations will likely be required for application of the TRA, such as extrapolations between species and tissues. This is also the case for translating tissue residues to corresponding concentrations in environmental media.

• Even though the TRA has the potential to greatly enhance ecological risk assessments, a single overarching “decision tree” for deciding when to use this tool across all types of assessments was not indentified. Therefore, the decision to use the TRA should include an evaluation of the relative strengths, limitations, and uncertainties among exposure and residue-based methods for characterizing toxicological effects in the application of interest.

The overall consensus of experts who attended the workshop is that the tissue residue approach has the potential to improve assessment and management of risks from chemical contamination by improving the scientific understanding of the interaction between exposure and consequent effects. Appropriate application of the TRA can reduce variability, diminish uncertainties, and improve interpretation of causality compared to both traditional toxicity testing and assessments based solely on concentrations in exposure media.

This confidence is accompanied by several caveats that should be considered because they create some limitations for application of the tissue residue tool. As outlined in our reviews, important research needs must be addressed before this approach can advance and gain wider acceptance among researchers and regulators. Given the state of the science, in addition to some key advancements in experience and knowledge, the TRA likely will inform the next generation of testing protocols and provide improved data for risk assessment and risk management.

Acknowledgements

We thank the staff from SETAC (Greg Schiefer and Nikki Turman) for their excellent organizational skills and for keeping us free of logistical concerns, allowing increased focus on the topic at hand. We also thank the staff of the Sleeping Lady Resort for maintaining a stress-free environment in a stunning location in which to conduct our deliberations. The authors of this editorial (the steering committee) extend a very special thanks to the workshop attendees who volunteered their time and expertise during the considerable scientific deliberations and contributed to the articles published in this issue. Their commitments assured the highest quality review and evaluation of the state of the science and also ensured the soundness of the recommendations needed to improve the TRA. The body of work presented here is a testament to their hard work. The names of the workshop participants can be found in this issue—because each is a coauthor on one or more of the reviews. We also acknowledge the generous financial support we received from a number of organizations and individuals that allowed this workshop to evolve from an idea to reality, particularly the US Environmental Protection Agency (USEPA, Office of Research and Development and Office of Water); US Department of Commerce–National Oceanic and Atmospheric Administration (Northwest Fisheries Science Center and the National Ocean Service Coastal Response Research Center–University of New Hampshire Partnership); the US Army Corps of Engineers Environmental Research and Development Center; and the numerous workshop participants who covered their own expenses for travel, lodging, and meals. We also express our gratitude to Environment Canada, Rio Tinto, the Nickel Producers Research Association, International Copper Association, International Zinc Association, ExxonMobil, and the US Geological Survey for their support. The opinions and views expressed in this paper are those of the authors and do not necessarily reflect the views and policies of their affiliated institutions.

APPENDIX 1. COMMON TERMINOLOGY USED IN THE TISSUE RESIDUE APPROACH FOR TOXICITY ASSESSMENT

Because scientists often get mired in terminology and acronyms, especially with multiple terms that refer to the same concept or metric, we felt the need to address this topic. Here we provide general descriptions for TRA doses, their metrics, and a number of the more commonly used terms and equivalents employed by our colleagues. We hope this alleviates some of the confusion that often surrounds nomenclature and terms in this subject area. Some of the definitions provided here may not be consistent with past or current usage because many definitions continue to evolve. The intent here is to provide clarity in how these concepts are defined and to invoke some discussion of their meaning.

A major distinction for the TRA concerns the amount of a toxicant that is delivered or administered to the organism and the actual tissue concentrations associated with the response. The administered dose is usually expressed as a rate (e.g., µg or µmol toxicant gram body wt⁻¹ d⁻¹) or as a single dose (µg/g or µmol/g) and is usually provided by feeding, injection, gavage, or bolus to determine the lethal dose, 50% (LD50) or other measures of toxicity. The acquired dose (tissue residue) is used to characterize adverse effects as a function of the measured or predicted tissue concentration. The administered dose may be very different from the tissue concentration associated with toxicity due to possible metabolism and excretion. It is the acquired dose that forms the foundation for the TRA.

Acquired dose—That amount of a toxicant accumulated by an organism expressed as a tissue concentration in mass or molar units (e.g., µmol/g). This dose can be determined by direct observation, quantitative structure–activity relationship, or toxicokinetics. The internal dose can be based on whole-body, organ-specific, or receptor-specific concentrations. Commonly used synonyms: internal dose, received dose.

Administered dose—The dose of a toxicant that is external to the organism that can be used to quantify adverse effects (for example, LD50 or LC50), expressed as a rate or as a single dose provided to the test animal. The external dose can be similar or unrelated to the actual tissue concentration (acquired or internal dose) that is associated with the biological response owing to factors such as metabolism and toxicokinetics. Commonly used synonym: external dose.

Baseline toxicity—Toxicity that some organic compounds produce by partitioning into biological membranes causing a nonspecific disturbance of the integrity and function of the cell. If a compound exhibits a specific mode of action, partitions into the cell membrane, and its tissue concentration is below the range at which biological alterations occur, it may contribute to mixture toxicity as a baseline toxicant. Commonly used synonyms: nonspecific toxicity, narcosis.

Biologically effective dose (BED)—That portion of a toxicant that reaches cells, sites, or membranes where biological alterations are initiated. The BED may represent only a fraction of the administered or acquired dose; however, in some cases, it is the best dose for predicting adverse effects. Also known as the biologically active metal (BAM) for elements.

Body burden—Total mass of a toxicant associated with an individual organism (for example, ng/fish).

Critical area under the curve—Time integral of internal concentration, which is a measure of the internal dose for toxicants that bind irreversibly with the target.

Critical body residue—Any one of several statistics that describe an adverse biological response (LR50, LA_p, ILC50, IEC_p, ER10, LOER) as a function of a whole-body, organ, or target tissue concentration expressed in mass or molar units (acquired dose). Most terms are similar to the traditional toxicological expressions for ambient exposure (for example, LC_p or LD_p [lethal] and EC_p or ED_p [sublethal] values), where L = lethal; R = residue; I = internal; C = external concentration; E = effective; O = observed; D = administered dose; p = the percentage responding. LA_p is the lethal accumulation at the receptor (gill) in the biotic ligand model. LOER is the lowest observed effective residue.

Mechanism of toxic action (MeOA)—The crucial biochemical process and/or interaction between chemical and target site underlying a given mode of action (for example, a specific interaction with an enzyme or receptor). Commonly used synonyms: primary mechanism, receptor interaction.

Mode of toxic action (MoOA)—A set of physico-chemical, physiological, or biochemical pathway alterations (e.g., uncoupling of oxidative phosphorylation, acetylcholine esterase inhibition, baseline toxicity). The alterations may be the result of one or more mechanisms of toxic action.

Target concentration—An internal dose that is expressed in terms of the target tissue concentration. Indicated by a subscript (e.g., LR_{p membrane lipid,} LR_{p liver,} or IEC_{p brain}) that refers to the dose in the target resulting in a p (percent) response.

Mixture toxicity—A mixture is a combination of 2 or more component chemicals to which living organisms may be exposed, either simultaneously or sequentially. The biological response to mixtures can be additive, less than additive, or more than additive when compared to the toxicity produced by individual components. Toxicants that act by the same mechanism of toxic action may be dose (or concentration) additive. Multiple toxicants that act by the same mode of toxic action but different mechanism of toxic actions may be response-additive.

Toxicodynamics—The phase of toxic action that consists of the biological response resulting from the interaction of the chemical at the site of toxic action. The toxicant's potency is a function of its physico-chemical characteristics, the number of chemical–receptor interactions, properties of the receptor, and chemical–receptor affinity. Commonly used synonym: potency.

Toxicokinetics—Rates of uptake, elimination, and internal distribution. Uptake can occur by several routes (for example, absorption, ventilation, and ingestion). Elimination is characterized by the processes of excretion, metabolism, and passive diffusion. Internal distribution involves the rates of chemical transfer between internal compartments (e.g., organs, plasma, and bile).

TRA toxicity metrics (regression based)—Statistics that estimate a population parameter based on the acquired (internal) dose. Values are determined using one of several regression algorithms, such as probit, logit, or generalized linear models. Commonly used terms: LR_p or ER_p—lethal (LR_p) or effective residue (ER_p) for a given proportion or percentage (p) of the test population. Values are based on quantal data (LR_p) or continuous data for sublethal responses (ER_p). Commonly used synonyms: LC50_i, internal lethal concentration (ILC_p), lethal body burden (LBB), and internal effective concentration (IEC_p).

TRA toxicity (analysis of variance [ANOVA] based)—Values are determined with ANOVA and one of several available posthoc tests. Commonly used terms: LOER is the lowest observed effect tissue residue (acquired dose) associated with an adverse toxic effect. NOER is the no observed effect residue, which is considered indistinguishable from the control value. Both the LOER and NOER are determined statistically with posthoc testing and designated p-values by comparing mean values for control and treatments.