Improving Risk Assessment: Priorities for Epidemiologic Research

The Epidemiology Work Group at the Workshop on Future Research for Improving Risk Assessment Methods, Of Mice, Men, and Models, held August 16 to 18, 2000, at Snowmass Village, Aspen, Colorado, concluded that in order to improve the utility of epidemiologic studies for risk assessment, methodologic research is needed in the following areas: (1) aspects of epidemiologic study designs that affect doseresponse estimation; (2) alternative methods for estimating dose in human studies; and (3) refined methods for dose-response modeling for epidemiologic data. Needed research in aspects of epidemiologic study design includes recognition and control of study biases, identification of susceptible subpopulations, choice of exposure metrics, and choice of epidemiologic risk parameters. Much of this research can be done with existing data. Research needed to improve determinants of dose in human studies includes additional individual-level data (e.g., diet, co-morbidity), development of more extensive human data for physiologically based pharmacokinetic (PBPK) dose modeling, tissue registries to increase the availability of tissue for studies of exposure/dose and susceptibility biomarkers, and biomarker data to assess exposures in humans and animals. Research needed on dose-response modeling of human studies includes more widespread application of flexible statistical methods (e.g., general additive models), development of methods to compensate for epidemiologic bias in dose-response models, improved biological models using human data, and evaluation of the benchmark dose using human data. There was consensus among the Work Group that, whereas most prior risk assessments have focused on cancer, there is a growing need for applications to other health outcomes. Developmental and reproductive effects, injuries, respiratory disease, and cardiovascular disease were identified as especially high priorities for research. It was also a consensus view that epidemiologists, industrial hygienists, and other scientists focusing on human data need to play a stronger role throughout the risk assessment process. Finally, the group agreed that there was a need to improve risk communication, particularly on uncertainty inherent in risk assessments that use epidemiologic data.     

Health Risk Assessment and Vapor Intrusion: A Review and Australian Perspective

Soil contamination by volatile hydrocarbons is of public health importance due to vapor intrusion and indoor inhalation exposures. These are assessed using measurement or predictive modeling and need to consider the key areas of subsurface partitioning and transport, dwelling ventilation, and receptor inhalation dosimetry. While subsurface partitioning and transport have been subject to intensive international investigation, limited consideration has been given to the latter. Building ventilation research has developed multi-zone airflow and contaminant dispersal models including AccuRate, an Australian model that examines natural ventilation modeling, roof and sub-floor ventilation, and identifies the importance of geometry and thermal factors on ventilation (the most sensitive variable) and indoor pollutant concentrations. Inhalation dosimetry has received recent attention due to concerns over child inhalation susceptibility and dose metrics. Research using coupled computational fluid dynamics (CFD) and physiologically based pharmaco-kinetic (PBPK) models has reported variance from previous animal models’ extrapolation while CFD modeling of transient lung vapor absorption suggests the significance of transient versus steady-state evaluation of volatiles absorption into tissue and blood. The transient nature of sub-surface fate and transport, ventilation, and inhalation uptake thus warrants integrated exploration and application in order to realize improvements in vapor intrusion assessments. These perspectives and Australian modeling initiatives are presented in this article.     

PBPK models in risk assessment--A focus on chloroprene

Mathematical models are increasingly being used to simulate events in the exposure-response continuum, and to support quantitative predictions of risks to human health. Physiologically based pharmacokinetic (PBPK) models address that portion of the continuum from an external chemical exposure to an internal dose at a target site. Essential data needed to develop a PBPK model include values of key physiological parameters (e.g., tissue volumes, blood flow rates) and chemical specific parameters (rate of chemical absorption, distribution, metabolism, and elimination) for the species of interest. PBPK models are commonly used to: (1) predict concentrations of an internal dose over time at a target site following external exposure via different routes and/or durations; (2) predict human internal concentration at a target site based on animal data by accounting for toxicokinetic and physiological differences; and (3) estimate variability in the internal dose within a human population resulting from differences in individual pharmacokinetics. Himmelstein et al. [M.W. Himmelstein, S.C. Carpenter, P.M. Hinderliter, Kinetic modeling of beta-chloroprene metabolism. I. In vitro rates in liver and lung tissue fractions from mice, rats, hamsters, and humans, Toxicol. Sci. 79 (1) (2004) 18-27; M.W. Himmelstein, S.C. Carpenter, M.V. Evans, P.M. Hinderliter, E.M. Kenyon, Kinetic modeling of beta-chloroprene metabolism. II. The application of physiologically based modeling for cancer dose response analysis, Toxicol. Sci. 79 (1) (2004) 28-37] developed a PBPK model for chloroprene (2-chloro-1,3-butadiene; CD) that simulates chloroprene disposition in rats, mice, hamsters, or humans following an inhalation exposure. Values for the CD-PBPK model metabolic parameters were obtained from in vitro studies, and model simulations compared to data from in vivo gas uptake studies in rats, hamsters, and mice. The model estimate for total amount of metabolite in lung correlated better with rodent tumor incidence than did the external dose. Based on this PBPK model analytical approach, Himmelstein et al. [M.W. Himmelstein, S.C. Carpenter, M.V. Evans, P.M. Hinderliter, E.M. Kenyon, Kinetic modeling of beta-chloroprene metabolism. II. The application of physiologically based modeling for cancer dose response analysis, Toxicol. Sci. 79 (1) (2004) 28-37; M.W. Himmelstein, R. Leonard, R. Valentine, Kinetic modeling of beta-chloroprene metabolism: default and physiologically-based modeling approaches for cancer dose response, in: IISRP Symposium on Evaluation of Butadiene & Chloroprene Health Effects, September 21, 2005, TBD--reference in this proceedings issue of Chemical-Biological Interactions] propose that observed species differences in the lung tumor dose-response result from differences in CD metabolic rates. The CD-PBPK model has not yet been submitted to EPA for use in developing the IRIS assessment for chloroprene, but is sufficiently developed to be considered. The process that EPA uses to evaluate PBPK models is discussed, as well as potential applications for the CD-PBPK model in an IRIS assessment.     

Development of a Physiologically Based Model to Describe the Pharmacokinetics of Methylphenidate in Juvenile and Adult Humans and Nonhuman Primates

The widespread usage of methylphenidate (MPH) in the pediatric population has received considerable attention due to its potential effect on child development. For the first time a physiologically based pharmacokinetic (PBPK) model has been developed in juvenile and adult humans and nonhuman primates to quantitatively evaluate species- and age-dependent enantiomer specific pharmacokinetics of MPH and its primary metabolite ritalinic acid. The PBPK model was first calibrated in adult humans using in vitro enzyme kinetic data of MPH enantiomers, together with plasma and urine pharmacokinetic data with MPH in adult humans. Metabolism of MPH in the small intestine was assumed to account for the low oral bioavailability of MPH. Due to lack of information, model development for children and juvenile and adult nonhuman primates primarily relied on intra- and interspecies extrapolation using allometric scaling. The juvenile monkeys appear to metabolize MPH more rapidly than adult monkeys and humans, both adults and children. Model prediction performance is comparable between juvenile monkeys and children, with average root mean squared error values of 4.1 and 2.1, providing scientific basis for interspecies extrapolation of toxicity findings. Model estimated human equivalent doses in children that achieve similar internal dose metrics to those associated with pubertal delays in juvenile monkeys were found to be close to the therapeutic doses of MPH used in pediatric patients. This computational analysis suggests that continued pharmacovigilance assessment is prudent for the safe use of MPH.     

What Can Be Done to Ensure the Usefulness of Epidemiologic Data for Risk Assessment Purposes?

Epidemiologic studies can play a central role in risk assessments. They are used in all risk assessment phases: hazard identification, dose-response, and exposure assessment. Epidemiologic studies have often been the first to show that a particular environmental exposure is a hazard to health. They have numerous advantages with respect to other sources of data which are used in risk assessments, the most important being that they do not require the assumption that they are generalizable to humans. For this reason, fewer and lower uncertainty factors may be appropriate in risk characterization based on epidemiologic studies. Unfortunately, epidemiologic studies have numerous problems, the most important being that the exposures are often not precisely measured. This article presents in detail the advantages of and problems with epidemiologic studies. It discusses two approaches to ensure their usefulness, biomarkers and an ordinance which requires baseline and subsequent surveillance of possible exposures and health effects from newly sited potentially polluting facilities. Biomarkers are biochemical measures of exposure, susceptibility factors, or preclinical pathological changes. Biomarkers are a way of dealing with the problems of poor measures, differential susceptibility and lack of early measures of disease occurrence that inherent in many environmental epidemiologic studies. The advantages of biomarkers is they can provide objective information on exposure days, months or even years later and evidence of pathology perhaps years earlier. The ordinance makes possible the use of a powerful epidemiologic study design, the prospective cohort study, where confounder(s) are best measured, and exposures, pathological changes, and health effects can be detected as soon as possible.     

Linking Pharmacokinetics and Biomarker Data to Mechanism of Action in Risk Assessment

Incorporating information on metabolism, pharmacokinetics, and DNA and protein biomarkers provides a means to integrate these important factors into the risk assessment process. Such data are useful for species to species extrapolation, high- to low-dose extrapolation and PBPK modeling. In addition, these data are critical for understanding the mode of action for chemical carcinogens. Through the use of mass spectrometry, stable isotopes can be used to unequivocally demonstrate pathways of formation of biomarkers and relationships between exogenous and endogenous processes. This paper reviews what has been learned for two carcinogens, vinyl chloride and butadiene. It is clear that such data play major roles in improving the understanding of how chemicals cause cancer, extending the range of data on exposure-response relationships, and examining interspecies differences and inter-individual differences that may affect susceptibility. As such, it is also clear that these data play a critical role in improving the accuracy of risk assessments.     

Separating Pharmacokinetic and Pharmacodynamic Components in Empirical Adjustment Factor Distributions

For a particular chemical, one can treat the chemical-by-chemical variation in relative doses for equal toxicity in experimental animals and humans as a characterization of the likelihoods of extrapolation factors of different magnitudes. An emerging approach to noncancer risk assessment is to use such empirical distributions in place of fixed Uncertainty Factors. This paper discusses dividing the overall variation into two sub-distributions representing pharmacokinetic (PK) and pharmacodynamic (PD) contributions to the variation among chemicals in the animal-to-human toxicologically equivalent dose. If a physiologically based pharmacokinetic model (PBPK model) is used to derive a compound specific adjustment factor (CSAF) for the pharmacokinetic component, the deconvolution of the PK and PD components allows one to remove the PK component (to be replaced with the CSAF), while retaining the uncertainty in pharmacodynamics that PBPK models do not address. One must then add back the uncertainty in the PBPK determination of the CSAF (which may be considerable). A candidate criterion for whether one can use an uncertain PBPK model is whether the generic uncertainty about cross-species pharmacokinetics (reflected in the PK component of the overall empirical distribution) is larger than the chemical-specific uncertainty in the determination of kinetically equivalent doses in experimental animals and humans.     

On the Definition of Internal Dose Metrics

Internal dose metrics, as computed with pharmacokinetic models, are increasingly used as a means for extrapolating animal toxicological data to humans and to extrapolate across routes of administration. These internal dose metrics are thought to provide a more scientific means of comparing toxicological effects across animal species. The use of internal dose metrics is based on the universal assumption that toxic effects are equal across species if and only if the concentration of the toxic moieties in the target tissue is equal across species. Herein it is shown that this assumption is inconsistent with empirical toxicological data. It is shown that measurement of internal dose metrics in chronological time, as is done for AUC (Area under the concentration curve) and rate of metabolite production per kg of target tissue, does not produce equal toxic effects across species. A consequence of this observation is that the application of pharmacokinetics in risk assessments for such important chemicals as trichloroethylene, vinyl chloride, perchloroethylene, and perchlorate may need reassessment.     

Validity of Epidemiological Data in Risk Assessment Applications

Data from epidemiological studies might be seen as superior to data from animal bioassays for risk assessment purposes. Because humans are the population of interest, use of epidemiological data avoids interspecies extrapolation. However, one must not assume that an epidemiological study is necessarily valid at face value. We describe issues of validity that arise in the conduct and interpretation of epidemiological research and that affect the utility of epidemiological data in risk assessment. These issues include choice of study design, size and representativeness of the study sample, measurement of exposures and outcomes, control of confounding and specification of statistical model for analysis of data, all of which affect the accuracy and validity of study results.     

Medical sonography: reproductive effects and risks

While it is clear that the levels and types of medical sonography that have been used in the past have no measurable risks, it would be inaccurate to label the modality of ultrasound as totally safe regardless of exposure. Most agents have reproductive risks and even teratogenic risks if the exposure is raised sufficiently. Thus the prudent use of sonography means that clinicians and designers of equipment have to maintain exposures far below the risks that have been demonstrated in animal studies and from the knowledge obtained about the physical changes that can be produced in humans as the absorbed dose is elevated. The reproductive risks were evaluated using five criteria: 1) human epidemiology, 2) secular trend data, 3) animal experiments, 4) dose response relationships, and 5) biologic plausibility. The analysis reveals that the human epidemiology does not indicate that diagnostic ultrasound presents a measurable risk to the developing embryo or fetus. Animal studies also indicate that diagnostic levels of ultrasound are safe and do not elevate the fetal temperature into the region where deleterious embryonic and fetal effects will occur. Because higher exposures of ultrasound can elevate the temperature of the embryo, the use of diagnostic procedures and the design of sonographic equipment should take into consideration the hyperthermic potential of higher exposures of ultrasound and the hypothetical additional risk of performing sonography on pregnant patients who are febrile. It would appear that if the embryonic temperature never exceeds 39 degrees C, then there is no measurable risk. We suggest that sonography (the field) and sonogram (the procedure) are the most appropriate and least anxiety provoking terms.     

Utilization of juvenile animal studies to determine the human effects and risks of environmental toxicants during postnatal developmental stages

BACKGROUND: Toxicology studies utilizing animals and in vitro cellular or tissue preparations have been used to study the toxic effects and mechanism of action of drugs and chemicals and to determine the effective and safe dose of drugs in humans and the risk of toxicity from chemical exposures. Testing in animals could be improved if animal dosing using the mg/kg basis was abandoned and drugs and chemicals were administered to compare the effects of pharmacokinetically and toxicokinetically equivalent serum levels in the animal model and human. Because alert physicians or epidemiology studies, not animal studies, have discovered most human teratogens and toxicities in children, animal studies play a minor role in discovering teratogens and agents that are deleterious to infants and children. In vitro studies play even a less important role, although they are helpful in describing the cellular or tissue effects of the drugs or chemicals and their mechanism of action. One cannot determine the magnitude of human risks from in vitro studies when they are the only source of toxicology data. METHODS: Toxicology studies on adult animals is carried out by pharmaceutical companies, chemical companies, the Food and Drug Administration (FDA), many laboratories at the National Institutes of Health, and scientific investigators in laboratories throughout the world. Although there is a vast amount of animal toxicology studies carried out on pregnant animals and adult animals, there is a paucity of animal studies utilizing newborn, infant, and juvenile animals. This deficiency is compounded by the fact that there are very few toxicology studies carried out in children. That is one reason why pregnant women and children are referred to as "therapeutic orphans." RESULTS: When animal studies are carried out with newborn and developing animals, the results demonstrate that generalizations are less applicable and less predictable than the toxicology studies in pregnant animals. Although many studies show that infants and developing animals may have difficulty in metabolizing drugs and are more vulnerable to the toxic effects of environmental chemicals, there are exceptions that indicate that infants and developing animals may be less vulnerable and more resilient to some drugs and chemicals. In other words, the generalization indicating that developing animals are always more sensitive to environmental toxicants is not valid. For animal toxicology studies to be useful, animal studies have to utilize modern concepts of pharmacokinetics and toxicokinetics, as well as "mechanism of action" (MOA) studies to determine whether animal data can be utilized for determining human risk. One example is the inability to determine carcinogenic risks in humans for some drugs and chemicals that produce tumors in rodents, When the oncogenesis is the result of peroxisome proliferation, a reaction that is of diminished importance in humans. CONCLUSIONS: Scientists can utilize animal studies to study the toxicokinetic and toxicodynamic aspects of drugs and environmental toxicants. But they have to be carried out with the most modern techniques and interpreted with the highest level of scholarship and objectivity. Threshold exposures, no-adverse-effect level (NOAEL) exposures, and toxic effects can be determined in animals, but have to be interpreted with caution when applying them to the human. Adult problems in growth, endocrine dysfunction, neurobehavioral abnormalities, and oncogenesis may be related to exposures to drugs, chemicals, and physical agents during development and may be fruitful areas for investigation. Maximum permissible exposures have to be based on data, not on generalizations that are applied to all drugs and chemicals. Epidemiology studies are still the best methodology for determining the human risk and the effects of environmental toxicants. Carrying out these focused studies in developing humans will be difficult. Animal studies may be our only alternative for answering many questions with regard to specific postnatal developmental vulnerabilities.     

Developing Parameters for Uncommon Exposures

Common exposure scenarios form the basis for exposure assessments included in human health risk assessments. The quantitative parameters that are used to calculate the receptor's dose associated with these common exposure scenarios are readily available. When humans have uncommon exposures, these often are excluded from the dose calculations because of the lack of parameters or parameters are estimated, which can increase the uncertainty in calculated dose. This study evaluates a third option of performing laboratory experiments to estimate parameters for uncommon exposures. Two exposures, the handling and hand-washing of contaminated work clothes and the drinking of water with sufficient free-phase hydrocarbons to form a visible sheen, were simulated in a laboratory. Parameters were developed for each exposure such that doses could be calculated. Parameters developed include a mass of oily contaminants on skin after washing contaminated clothes and the mass of a visible oily sheen on an aqueous surface. These results demonstrate that developing parameters for uncommon exposure scenarios is a practical alternative to eliminating or estimating exposures and these experiments can be done in compliance with human subject protocols.     

Cancer risk assessment for 1,3-butadiene: data integration opportunities

The US Environmental Protection Agency recently released its new guidelines for carcinogen risk assessment together with supplemental guidance for assessing susceptibility from early-life exposure to carcinogens. In particular, these guidelines encourage the use of mechanistic data in support of dose-response characterization at doses below those at which an increase in tumor frequency over background levels might be detected. In this context of the utility of mechanistic data for human cancer risk assessment, the International Life Sciences Institute (ILSI) has developed a human relevance framework (HRF) that can be used to assess the plausibility of a mode of action (MoA) described for animal models operating in humans. The MoA is described as a sequence of key events and processes that result in an adverse outcome. A key event is a measurable precursor step that is in itself a necessary element of the MoA or is a bioindicator for such an element. A number of cellular and molecular perturbations have been identified as key events whereby DNA-reactive chemicals can produce tumors. These include DNA adducts in target tissues, gene mutations and/or chromosomal alterations in target tissues and enhanced cell proliferation in target tissues. This type of data integration approach to quantitative cancer risk assessment can be applied to 1,3-butadiene, for example, using data on biomarkers in exposed Czech workers [1]. For this study, an extensive range of biomarkers of exposure and response was assessed, including: polymorphisms in metabolizing enzymes; urinary concentrations of several metabolites of 1,3-butadiene; hemoglobin adducts; HPRT mutations in T-lymphocytes; chromosomal aberrations by FISH and conventional staining procedures; sister chromatid exchanges. Exposure levels were monitored in a comprehensive fashion. For risk assessment purposes, these data need to be considered in the context of how they inform the MoA for leukemia, the tumor type reported to be increased in synthetic rubber workers exposed to 1,3-butadiene. Also, for the HRF it is necessary to establish key events for a MoA in rodents for the induction of tumors by 1,3-butadiene. There is clearly a species difference in sensitivity to tumor induction, with mice being much more sensitive than rats; key events need to explain this difference. For butadiene, the MoA is DNA-reactivity and subsequent mutagenicity and so following the EPA's cancer guidelines, a linear extrapolation is used from the point of departure (POD), unless additional data support a non-linear extrapolation. For the present case, the human bioindicator data are not informative as far as dose-response characterization is concerned. Mouse chromosome aberration data for in vivo exposures might be used for establishing a POD, with linear extrapolation from this POD. The available cytogenetic data from rodent studies appear to be sufficiently extensive and consistent for this to be a viable approach. This approach of using MoA and key events to establish the human relevance can lead to the development of specific informative bioindicators of response that can be used as surrogates to predict the shape of the tumor dose response curve at low doses. Truly informative predictors of tumor responses should be able to provide estimates of human tumor frequencies at low, environmental exposures to 1,3-butadiene.     

Axes of Extrapolation in Risk Assessment

Extrapolation in risk assessment involves the use of data and information to estimate or predict something that has not been measured or observed. Reasons for extrapolation include that the number of combinations of environmental stressors and possible receptors is too large to characterize risks comprehensively, that direct characterization is sometimes impossible, and that the power to characterize risk in a particular situation can be enhanced by using information obtained in other similar situations. Three types of extrapolation are common in risk assessments: biological (including between taxa and across levels of biological organization), temporal, and spatial. They can be thought of conceptually as the axes of a 3-dimensional graph defining the state space of biological, temporal, and spatial scales within which extrapolations are made. Each of these types of extrapolation can introduce uncertainties into risk assessments. Such uncertainties may be reduced through synergistic research facilitated by the sharing of methods, models, and data used by human health and ecological scientists     

Benchmark Dose Profiles for Joint‐Action Quantal Data in Quantitative Risk Assessment

Summary Benchmark analysis is a widely used tool in public health risk analysis. Therein, estimation of minimum exposure levels, called Benchmark Doses (BMDs), that induce a prespecified Benchmark Response (BMR) is well understood for the case of an adverse response to a single stimulus. For cases where two agents are studied in tandem, however, the benchmark approach is far less developed. This article demonstrates how the benchmark modeling paradigm can be expanded from the single‐dose setting to joint‐action, two‐agent studies. Focus is on response outcomes expressed as proportions. Extending the single‐exposure setting, representations of risk are based on a joint‐action dose–response model involving both agents. Based on such a model, the concept of a benchmark profile (BMP) – a two‐dimensional analog of the single‐dose BMD at which both agents achieve the specified BMR – is defined for use in quantitative risk characterization and assessment. The resulting, joint, low‐dose guidelines can improve public health planning and risk regulation when dealing with low‐level exposures to combinations of hazardous agents.     

International Commission for Protection against Environmental Mutagens and Carcinogens. ICPEMC publication No. 13. The need for biological risk assessment in reaching decisions about carcinogens

The prudent assumption that carcinogen bioassays in rodents predict for human carcinogenicity is examined. It is suggested that in certain cases, as for example the induction of tumors against a high incidence in controls, or in situations in which high dose toxicity may be a critical factor in the induction of cancer, the probability that animal bioassays predict for humans may be low. The term 'biological risk assessment' is introduced to describe that part of risk assessment concerned with the relevance of specific animal results to the induction of human cancer. Biological risk assessment, which is almost entirely dependent on an understanding of carcinogenesis mechanisms, is an important addition to present mathematical modeling used to predict the effects of animal carcinogens that have been demonstrated after high dose exposure, to the effects of the much smaller doses to which humans are perceived to be exposed. Evidence for the conclusions reached by biological risk assessment may sometimes be supported by a careful review of human epidemiological data.