Scientific Justification for and Conduct of Intentional Human Dosing Studies
Scientific and ethical issues must be considered whenever intentional human dosing studies are proposed. These issues are, in most respects, interconnected. For example, an intentional human dosing study conducted for Environmental Protection Agency (EPA) regulatory purposes that is designed in such a way that it cannot make a scientifically sound contribution to regulatory decision making cannot be judged as ethical. However, for ease of explication, scientific and ethical issues are discussed separately in this chapter, with scientific issues the principal concern.
Intentional human dosing studies involving potentially toxic substances can, in some circumstances, contribute significant and useful knowledge for regulatory standard setting and other forms of public health protection. In fact, there is a long history of using data from such studies for these purposes, along with data from epidemiological investigations and animal experiments (Faustman and Omenn, 2001; Lippman et al., 2003; Paustenbach, 2002; Rodricks et al., 1997). The committee supports continued use of such information, provided that it is generated in compliance with the criteria and procedures recommended in this report that are designed to ensure ethical and scientific validity. The committee strongly recommends, however, that EPA should introduce much greater scientific care and rigor into its process for considering and relying on intentional human dosing studies by establishing criteria and procedures for deciding when and how they are to be conducted and their results used. Importantly, the same criteria and procedures should apply to both agency-conducted or agency-sponsored and third-party human dosing
studies. Although EPA has in place procedures for ethical and patient protection review of agency-sponsored human studies (EPA, 1999), a more uniform and scientifically rigorous system should be considered for them and for third-party studies (discussed further in Chapter 6). The principal criteria for the scientific review of human dosing studies are briefly described in this chapter.
As with all types of research, proposals to conduct intentional human dosing studies should begin with a discussion of the purpose and value of the study—the study justification. Assuming a study is justified, questions arise regarding study design and conduct and the reporting and evaluation of study results, matters that should be detailed in a study protocol. The protocol also includes information regarding protection of research participants. These two critical elements—study justification and study protocol—are the focus of this chapter.
It is important to recognize some of the critical distinctions between the types of research that are of interest to EPA as it carries out its legislative mandate and research that has a broader purpose. EPA is a regulatory agency that seeks information to fulfill its mission, such as that needed to improve the scientific basis of the risk assessments that are used to set regulatory standards or to fashion other types of health protection goals. Much of the committee’s thinking regarding study justification and study protocols has been developed in recognition of the unique needs of regulatory agencies such as EPA. The committee also recognizes that all human research, whatever the purpose, must be conducted in adherence to the highest scientific standards, and it sought to incorporate such standards, along with those uniquely related to the regulatory process, into its recommendations. In addition, the committee proposes careful, independent review of study justifications and protocols for all intentional human dosing studies within the scope of EPA’s mandate.
Before examining the issues involved in providing scientific justifications of and study protocols for intentional human dosing studies, a brief discussion is presented of the types of scientific investigations involving intentional dosing that are typically considered for possible conduct in human populations.
TYPES OF INTENTIONAL HUMAN DOSING STUDIES
There are three principal types of studies involving intentional dosing of research participants with chemicals that have been conducted for EPA regulatory purposes. The three types of studies seek to elicit (1) phar-
macokinetic (PK) information, (2) effects on a biomarker, but not symptoms, and (3) effects on a symptom. These studies are not intended or expected to cause any irreversible or serious effect, based on previous animal and human experience. This is appropriate, as the committee cannot envision circumstances in which it would be ethical to knowingly harm research participants in order to generate data for EPA regulatory purposes. Although the three types of studies are not considered likely to cause lasting or serious harm to study participants, as will be explained in this chapter, their low levels of risk are not identical.
Studies That Seek to Elicit Pharmacokinetic Information
The goal of studies that seek to elicit PK information, or PK studies, is to delineate the absorption, distribution, metabolism, and excretion of chemicals in the body. Gaining an understanding of these processes can greatly aid in the interpretation of toxicity study findings and in the refinement of risk-assessment practices.
Comprehensive PK data can substantially reduce uncertainties inherent in route-to-route, high-to-low dose, and species-to-species extrapolations (Andersen, 1995; Leung and Paustenbach, 1995; see also Appendix B). In addition, knowledge of the toxicity and pharmacokinetics of a particular pesticide in one species can be useful in predicting and understanding adverse effects in a second species, whether in another laboratory animal or in humans.
A recent development in risk assessment is the use of so-called physiologically based PK models to improve the bases for cross-species extrapolation (Andersen, 2003; Bailer and Dankovic, 1997; Clewell et al., 2002). However, the successful development of such models depends on the availability of PK data in humans, with these and other developments in risk assessment placing increased reliance on human PK data. In addition to informing interspecies comparisons, human PK data can shed light on the appropriateness of the intraspecies uncertainty factor, for example, by showing similar PK activity in a wide range of participants.
Useful PK data typically can be developed in humans using very low doses of chemicals—doses that cannot cause adverse effects and often that cannot cause any detectable biological changes in research participants. PK studies conducted at levels that, based on extensive previous testing in animals, are expected to cause no detectable biological effect in participants, can be considered to pose no identifiable risks to research participants.1
Pharmacodynamic Studies That Examine Low-Dose Effects on Biomarkers
Pharmacodynamic (PD) studies (sometimes called toxicodynamic studies) are designed to measure the effect of a chemical or its metabolites on particular components of the body (e.g., tissues, cells, cell components). In some cases, the measured effect is a short-term biological response that is not thought to be adverse to health at the level studied, but that would cause the expected adverse effect of the chemical if the response were larger or more sustained—that is, it is on the causal pathway of the adverse effect. At the doses and duration used, however, the response would not be expected to cause an adverse effect in study participants. These biological responses often are referred to as biomarkers for the effects of interest (e.g., neurological effects).
Ordinarily, such studies—which involve brief and low exposure to chemicals and which are the majority of third-party studies submitted to EPA to date—present little risk to participants. Examples of such PD studies submitted to EPA include organophosphate (OP) pesticide inhibition of blood cholinesterase and perchlorate inhibition of radioactive iodine uptake by the thyroid gland (Greer et al., 2000; Lawrence et al., 2000; Lawrence et al., 2001). In each case, the inhibition is linked to the mechanism of the serious toxic effects of the chemical, but the effects on the biomarkers are known through other studies to become observable at dosage levels well below those at which adverse effects become clinically apparent. Moreover, these changes in biomarkers are reversible and temporary (whether longer term effects are possible is another consideration). The inhibition studies are valuable because the dose or blood concentration that causes a given degree of inhibition in humans and animals can be compared, which allows for the determination of different sensitivities to the inhibition among species.
In many of these studies, the specific determinations of interest in humans are the doses causing some effect on the biomarker (the lowest observed effect level, or LOEL) and the highest level at which no effect is seen (the NOEL, or no observed effect level) (NRC, 1994). These can then be compared with the LOEL and the NOEL in animals. Importantly, a study in which no effect is seen and no LOEL is defined is generally uninterpretable, because there is no evidence that the study could detect the effect on the biomarker and that the dose that was studied is truly the highest dose that causes no effect.
Pharmacodynamic Studies That Examine Low-Dose Effects That May Be Adverse to Participants
Some PD studies conducted for EPA regulatory purposes involve measuring the effect of an administered substance on a clinically detectable, adverse effect that, if larger and sustained could harm study participants. Such studies can yield a lowest observed adverse effect level (LOAEL), a no observed adverse effect level (NOAEL), and possibly an NOEL, although it is expected that any observed effects will not be sustained once exposure ceases (that is, the change is fully reversible). These studies present a somewhat greater risk to participants than PK studies, but, if the effects are well understood, familiar, closely observed, and reversible, participants should experience no lasting harm. The endpoints studied to date have involved air pollutants and have included changes in lung function or exercise ability and symptom onset (e.g., dyspnea) (Koenig et al., 1994; Langley et al., 2003). Generally, the substances studied are those to which the general population is already exposed, such as air and water contaminants. (See Box 3.1, which presents the committee’s use of risk terminology for data derived from intentional dosing studies.)
In some studies, participants are healthy volunteers. In others, participants have a pre-existing medical condition (e.g., compromised cardiac or pulmonary function), and an exacerbation of the condition is used to assess exposure effects. If the purpose of such a study is to determine the effects of exposure on those who have pre-existing conditions that already put them at risk, it may be appropriate to include these participants. Additionally, in some cases valuable information can be gained from studies that include people with pre-existing conditions that are conducted at exposure levels known to exist in certain geographical areas, as participants would be exposed to levels they might encounter in their normal environments. Experimental studies of transmission dynamics that would include studies to determine infective dose, dose response curves, infectivity, and challenge studies (e.g., for cryptosporidium) are similar to those in this category of studies, as they are often expected to provoke a specific adverse but reversible effect (e.g., diarrhea).
All three of these types of studies can provide the opportunity to produce human data to improve the EPA risk-assessment process. In all cases, it is presumed that thorough animal data concerning the effects of the toxicants have been obtained and considered as the basis for concluding that a study poses no identifiable risk, that there is a reasonable certainty that no harm will occur to participants, or that the risks involved in the study are understood sufficiently that they can be evaluated in relation to the potential benefits. Of particular interest is whether the carcinogenicity and genotoxicity of a particular toxicant have been assessed. Depending
No observed effect level (NOELHU)
A NOELHU is the highest dose or concentration at which no changes of any kind are seen relative to controls. Depending on the number of doses studied and the ability to detect the LOELHU, the NOELHU could underestimate the actual dose that could be given without a response.
Lowest observed effect level (LOELHU)/ No observed adverse effect level (NOAELHU)
A LOELHU is the lowest dose or concentration at which a biological effect that is not adverse is seen. An example of such an effect would be cholinesterase inhibition by pesticides. A small amount of cholinesterase activity has not been demonstrated to have any adverse health effects. If lower doses are not studied, the LOELHU could overestimate the dose that could actually elicit a response. What the committee terms a LOELHU is often referred to by EPA as a no observed adverse effect level (NOAELHU). The committee is careful in its use of the term “NOAELHU” because it is most appropriately used in situations in which a clear LOAELHU has been identified. A NOAELHU is the highest dose or concentration at which no adverse effect is seen relative to controls.
Lowest observed adverse effect level (LOAELHU)
A LOAELHU is the lowest dose or concentration at which an adverse effect is seen. In terms of the committee’s discussion, for intentional human dosing studies there should be high confidence that any anticipated adverse effect is not serious and is reversible.
on the findings, evidence relating to genotoxicity or carcinogenicity in animals could be important in determining whether human studies are safe enough to conduct and could influence the content of the informed consent.
It is important to underscore the difference between the three types of studies described here and clinical trials involving therapeutic doses of experimental drugs. It is well recognized and accepted that even Food and Drug Administration (FDA)-approved drugs can pose significant risks to patients and thus that, in Phase 2 and Phase 3 clinical trials on experimental drugs, research participants may experience adverse side
effects. Indeed, in addition to assessing a drug’s effectiveness, these trials are used to identify and better understand its possible harmful side effects. This possibility of harm is one reason informed consent and independent Institutional Review Board (IRB) review of the risks and benefits of a trial are needed.
Importantly, in therapeutic clinical trials, there may be personal benefits for study participants, sometimes as an immediate consequence of participation, more typically in developing treatments for the condition the participant has. This benefit can, in some cases, be considered in deciding whether the risks are justified. The human dosing studies likely to be conducted (and found ethically acceptable) for EPA regulatory purposes pose much less risk to participants than often is accepted in drug trials, but they also are unlikely to provide any personal benefit to the participants. The different character of both the risks and benefits in human dosing studies conducted for EPA regulatory purposes makes many of the specific issues addressed in this report novel and underlies many of the committee’s recommendations.
This chapter now turns to the issues of how and why such studies may be justified and the types of protocols that are needed to ensure their proper conduct, including the protection of research participants.
JUSTIFICATION FOR INTENTIONAL HUMAN DOSING STUDIES
Criteria for Study Justification
Justification of intentional human dosing studies depends on the importance of the expected results to a regulatory decision that will protect the public health and a demonstration that other means of acquiring the necessary information are substantially deficient. In the case of intentional human dosing studies conducted for EPA regulatory purposes, ethical and scientific standards demand that every effort be made in advance to ensure that the biological endpoints to be measured are important to the assessment of human risk. Whether the data are to be used for determining risks for acute or short-term exposures, or for the derivation of a Reference Concentration (RfC) or a Reference Dose (RfD), every effort should be made to document in advance their critical nature. Data unrelated to or peripheral to regulatory risk assessments should never be sought through intentional human dosing studies, even those involving no identifiable risk to participants (PK studies).
For example, cholinesterase inhibition is generally considered to be the mechanism of action of the neurotoxic effects of many organophosphates (OP) pesticides, and doses that do not inhibit acetylcholinesterase (AChE) do not produce the cholinergic-mediated effects (see IOM, 2003,
for review). The inhibition of cholinesterase that mediates toxicity occurs at the synapses of the central and peripheral nerves, but in human studies only blood cholinesterase activity usually is measured. It is necessary, therefore, to know, in considering whether such a study is justified, if blood cholinesterase is a relevant measure of the state of peripheral nerve and central cholinesterase.
Even if acute blood cholinesterase inhibition were considered a reasonable surrogate marker for acute toxicity, it might not be an adequate marker for all effects of OPs, including possible long-term effects or effects on development. In addition, effects might differ across age groups or developmental stages (Clewell et al., 2002). This issue is sometimes far from straightforward.2
Therefore, even if it were well established that the short- and long-term effects of OPs are mediated through cholinesterase inhibition and that a dose with no effect on blood cholinesterase is very unlikely to cause harm in adults, those data would not necessarily provide information regarding possible effects on the developing nervous system. Because of these issues, careful documentation of the value and relevance of the endpoint to be measured is a critical component of study justification.
A second criterion that should be applied to justify studies in humans pertains to the availability of different ways of acquiring the necessary data. Data from animal models have widespread use in regulatory risk assessment, and many, if not most, standards are derived from such assessments (NRC, 1994). Considerable effort over the past several decades has been devoted to improving and standardizing protocols for animal bioassays (Ashby, 2001; Gaylor, 1996).
It may be asked why human studies are ever justifiable if animal models are available. There are three broad reasons to turn to human studies to supplement animal data (see Appendix A for further discussion of the limitations of animal studies). First, it is well established that animal models are not especially accurate predictors of certain adverse biological effects, particularly those involving immune-mediated responses (e.g., hypersensitivity reactions, other allergic responses) and certain airway responses to hazardous air pollutants (Samet et al., 1994). In some cases, no validated animal models may be available to serve as surrogates for individuals with compromised immune systems or with other medical conditions that may render them especially sensitive to pollutants. Many of the EPA-sponsored short-term air pollutant studies in humans have been motivated by such concerns, and the data derived from some of them have been informative for both setting standards and for gaining critical knowledge about mechanisms of toxicity (EPA, 2003).
Second, animal models have little value for assessing adverse effects that cannot be objectively measured, such as those that can be known only because they can be reported by study participants (headaches are a prime example, as are feelings of nausea and dizziness). Such symptoms can be significant indicators of toxicity, and sometimes efforts must be made to determine whether they can be produced by certain chemicals.
There have been reports that repeated OP exposures, that do not cause inhibition of brain AChE in preweanling rats, result in decreased locomotor activity and impaired spatial learning when the rats become juveniles (Carr et al., 2001; Jett et al., 2001). Efforts are now being made to understand the functional significance of cellular and molecular changes observed in the immature central nervous system and to determine whether household or dietary exposures to pesticides can produce such changes.
A third reason that animal data may be insufficient is that there are, in some instances, quantitative differences in response between average human and animal responses. This is recognized in the 10-fold interspecies uncertainty factor typically applied when animal data are used to set exposure limits (this assumes that humans may be 10 times more sensitive) (see Chapters 2 and 7), but, in fact, human sensitivity may be either greater than or less than 10 times that of animals. Human studies of a relevant endpoint can allow for decisions that are more informed about the risk of any given level of exposure (Dourson et al., 2001).
Even PK studies involving no identifiable risk to participants require scientific justification. As noted, PK data can be relevant to interspecies comparison and to within-human variability. The specific use and value of PK information need to be considered.
Documentation of Study Justification
Written and well-referenced documentation of the justification for intentional human dosing studies is a necessary prerequisite for their conduct. As will be seen in Chapter 6, the committee recommends that, prior to the conduct of both agency-sponsored and third-party studies, EPA should establish an independent board to review such documentation and to review the study protocols (see Box 3.2).
It should be emphasized that although a study may be scientifically justifiable according to the above criteria, it may nonetheless not be undertaken if the protection of research participants cannot be ensured. The committee views ensuring the protection of research participants as an element of the study protocol.
PROTOCOLS FOR INTENTIONAL HUMAN DOSING STUDIES
Along with providing documentation related to the justification of an intentional human dosing study, a study protocol should be provided that sets forth the study’s design and method of conduct and a plan for analyzing, reporting, and evaluating the results. These elements must be described and justified.3 The protocol also should include a demonstration of how participant protection will be assured.
There is an extensive literature on the design, conduct, and analysis of clinical studies (see, for example, FDA, 2003). However, rather than provide a comprehensive treatment here, the committee highlights issues that are especially important to the evaluation of intentional human dosing studies or that were identified as especially problematic in studies that were submitted to EPA and reviewed by the committee.
Overall Study Plan
The specific objectives of the proposed study, as described in the scientific justification document, are used to guide study design. Selection of doses to be used, criteria for participant selection, sizes of individual groups, and clinical measurements to be made are all dictated by the stated objectives of the study. In the end, it must be shown that the proposed study design is capable of yielding results that will satisfy the specified objectives.
A plan for the specific procedures to be followed in the conduct of the study, and for recording all of the relevant data, also is necessary, as is a description of methods to be used in evaluating study results. Finally, the overall plan should include documentation of the adequacy of preclinical data for establishing that study participants are not likely to be harmed at the doses selected and that other appropriate safeguards are in place.
Aspects of Study Design
Five features are critical to designing an intentional human dosing study, including endpoint, dose, and participant selection; study method; and dosing and measurement schedules.
1. Endpoint Selection
The endpoints to be measured should be described and their relation to study objectives explained. It should be asked whether the endpoints are the same as, or human equivalents to, those assessed in animals. The ability to measure the selected endpoints with reliability and precision should be described.
2. Dose Selection
Sufficient preclinical (animal) data relevant to the clinical endpoint of interest, or other human data, should be available to support selection of the doses to be used in humans. Dose selection for PK studies usually is dictated by technical questions related to analytical detection capabilities, rather than by any factors related to clinical response. The highest dose selected should be sufficient to induce the desired response, whether it is a critical biomarker or other endpoint. Doses lower than the highest dose should be selected to characterize the dose-response relationship and, if possible, to identify the maximum dose that represents the NOEL. Failure to see any response raises the question of whether the study was able to detect the response at all—that is, did the study have assay sensitivity? Consideration also must be given to the purity of the test compound, to ensure that it differs in no significant way from that of the test compound used in the preclinical studies that were used as the basis for dose selection. The mode of compound administration also should be described and the relevance of the method of administration justified. Box 3.3 provides two examples of designs used in studies submitted to EPA to identify a NOELHU, with accompanying committee commentary.
3. Participant Selection
The choice of participants is dictated by the objectives of the study. If the objective is to modify uncertainty factors and replace animal data with relevant human data (potentially eliminating the need for the uncertainty factor for animal-to-human extrapolation), healthy adult humans of, for example, similar age and weights might be most appropriate to represent the average human population. Selection of such individuals also would reduce possible variability in biological responses and make more precise estimates of the intraspecies factor possible. Although this study will not capture the full range of human variability, risk-assessment procedures already include an intraspecies uncertainty factor that will accommodate expected variability (see Chapters 2 and 7). Despite the desirability of a reasonably homogeneous population, including participants of both gen-
The investigator selects 10 males: Caucasian, healthy, 20 to 30 years of age. Each participant is dosed with Xmg of the substance, and the relevant endpoint is measured. No effects of interest are seen.
Commentary on Design #1: If no effects are observed, it is not possible to use the Xmg dose as establishing the NOELHU, because in the absence of observed effects, there is no evidence that the study could detect an effect if it were present (no proof of assay sensitivity) and no information about the dose that did have an effect (the LOELHU or LOAELHU). Furthermore, there could be no estimate of the uncertainties that surround any conclusion. This kind of study has been called a “NOEL-only” study and is not useful for formal risk assessment (see Chapter 7).
After careful review of the preclinical data, the investigator expects that the NOELHU for the substance will be near Xmg. Rather than dosing every participant with that dose, dosing proceeds as follows:
Commentary on Design #2: The study produces information that supports estimating a dose-response curve and confidence limits for the curve. The analysis can reasonably and credibly establish the NOELHU, or NOAELHU, depending on the outcome measured for acute exposure.
ders is desirable, unless there is compelling evidence that differences in response are not expected.
If, on the other hand, the goal of the study is to set acceptable levels of an air pollutant, it will be critical to focus on sensitive populations because they represent those most clearly at risk and often include individuals with specific medical conditions. Careful review of these conditions
among potential participants is critical in order to avoid wide variability among members of the study and control groups and to avoid including participants who will not test the question at issue (see Chapter 5).
Study protocols should include justification for participant selection, a description of how potential participants are identified, and a description of the procedures to be used in randomizing participants to dose groups.
4. Study Method
Protocols must provide a carefully delineated justification for the proposed study method. Sample sizes proposed for each group should be justified by a demonstration that there is adequate power to detect a relevant change in the endpoint(s) to be measured given the estimated variability in the response.
5. Dosing and Measurement Schedules
The specific schedule for dosing and measuring the response should be clearly related to the objectives of the study. Scientific support for the schedules should be provided.
Conducting and Recording Statistical Analysis of Results
It is essential to develop the statistical analysis plan as an integrated part of the study design and to ensure that primary statistical analysis is linked to primary study goals. An approach for recording the results also should be provided. All data generated should be thoroughly analyzed and reported, and the protocol should identify a hierarchy of outcomes with a narrowly defined set of primary goals. Confidence intervals surrounding the estimates and other measures of uncertainty should be reported, and the quality of the data should be assessed as they come in so that timely corrections can be made (and documented).
Protection of Research Participants
One section of the protocol should be devoted to a careful and thoroughly documented presentation of the likely risks to participants at the proposed levels and duration of dosing. This documentation should be accompanied by a discussion of other critical elements of study participant protection, as called for in Chapter 5 and 6.
The Protocol Document
As noted, the study protocol should provide a detailed statement on the study objectives and scientific justification for the study design. It also should provide the study analysis or a detailed statement on it and information regarding how the data will be reported. It should contain a thorough guide to participant protections, including any proposed data and safety monitoring plan or committee, assurance that the proposed study will be conducted in compliance with Good Clinical Practice (GCPs) guidelines, and provisions to permit EPA to monitor the conduct of the study (FDA, 2003). It also should include assurance of review by an IRB or an equivalent body as well as assurance that informed consent was obtained. Finally, the protocol should contain a copy of the written consent form, describe the consent procedures, and include an agreement to permit onsite inspection. The committee recommends that an independent review board evaluate the study protocol document, together with the scientific justification (see Chapter 6).
Study Reporting to EPA
For EPA to assess the scientific validity of the results of an intentional human dosing study, study reporting should be comprehensive and should include an assessment of the implications of the study relative to the study objectives and the relationship of study results to existing knowledge. The full protocol and detailed analyses should be submitted to EPA with a narrative interpretation of the results that includes summary tables and graphs, the data codebook, and all data (in a computer analyzable form, e.g., an SAS dataset), so that a reviewer could replicate reported analyses and conduct additional analyses. The report should fully document any problems and any changes in the protocol. Study participant characteristics must be well documented, and all adverse events must be reported and evaluated, regardless of determination of “relatedness” or causality assessment. All relevant studies conducted by the laboratory, clinic, or funding organization should be reported in at least summary form, even if their findings are not in the interest of those sponsoring the study.
Recommendation 3-1: Scientific Validity of Intentional Human Dosing Studies
EPA should issue guidelines for determining whether intentional human dosing studies have been:
justified, in advance of being conducted, as needed and as scientifically appropriate, in that they could contribute to address-
ing an important scientific or policy question that cannot be resolved on the basis of animal data or human observational data;
designed in accordance with current scientific standards and practices to (i) address the research question, (ii) include representative study populations for the endpoint in question, and (iii) meet requirements for adequate statistical power;
conducted in accordance with recognized good clinical practices, including appropriate monitoring for safety; and
reported comprehensively to EPA, including the full study protocol, all data produced in the study (including adverse events), and detailed analyses of the data.
Three principal types of studies involving intentional dosing of research participants with chemicals have been conducted for EPA regulatory purposes: PK studies; PD studies of low-dose, nonadverse effects; and PD studies designed to elicit an adverse but fully reversible effect. The first two types of studies are likely to pose no identifiable risk to study participants or can be scientifically demonstrated to provide a reasonable certainty that no harm will result to them. PD studies eliciting adverse but reversible effects pose a risk, although it should remain low, depending on factors such as the nature of the effect and whether it is fully reversible, whether the study is properly conducted, and the study population.
Prior to its conduct, a study should be deemed justifiable on the basis of existing scientific data from animal and other studies. This justification should include an explanation of the relevance and importance of the endpoint to the potential effects of concern for regulatory purposes and evidence of the lack of ability to obtain the needed information in other ways.
An intentional human dosing study cannot be ethical if it is not designed, conducted, and reported in ways that ensure the highest scientific quality. The need for scientific quality begins at the planning stage and includes the choice of endpoint, exposure conditions, and dose, as well as a consideration of the study power and statistical analysis. The full study results should be reported, including details regarding design, conduct, and outcomes, even if they are not in the interest of those sponsoring the study.
Andersen, M. E. 1995. Development of physiologically based pharmacokinetic and pharmacodynamic models for applications in toxicology and risk assessment. Toxicology Letters 79:35-44.
Andersen, M. E. 2003. Toxicokinetic modeling and its applications in chemical risk assessment. Toxicology Letters 138:9-27.
Ashby, J. 2001. Expectations for transgenic rodent cancer bioassay models. Toxicologic Pathology 29(Suppl.):177-182.
Bailer, A. J., and D. A. Dankovic. 1997. An introduction to the use of physiologically based pharmacokinetic models in risk assessment. Statistical Methods in Medical Research 6:341-358.
Benke, G. M., and S. D. Murphy. 1975. The influence of age on the toxicity and metabolism of methyl parathion and parathion in male and female rats. Toxicology and Applied Pharmacology 31:254-269.
Brimijoin, S., and C. Koenigsberger. 1999. Cholinesterases in neural development: new findings and toxicologic implications. Environmental Health Perspectives 107(Suppl. 1):59-64.
Carr, R. L., H. W. Chambers, J. A. Guarisco, J. R. Richardson, J. Tang, and J. E. Chambers. 2001. Effects of repeated oral postnatal exposure to chlorpyrifos on open-field behavior in juvenile rats. Journal of Toxicological Sciences 59:260-267.
Clewell, H. J., J. Teeguarden, T. McDonald, R. Sarangapani, J. E. Lawrence, T. Covington, R. Gentry, and A. Shipp. 2002. Review and evaluation of the potential impact of age- and specific pharmacokinetic differences on tissue dosimetry. Critical Reviews in Toxicology 32:329-389.
Committee on Toxicity of Chemicals in Food, Consumer Products, and the Environment. 1999. Organophosphates: Report of the Working Group on Organophosphates. Available at www.doh.gov.uk/cot/op.htm.
Dourson, M. L., M. E. Andersen, L. Erdreich, and J. Macgregor. 2001. Using human data to protect the public’s health. Regulatory Toxicology and Pharmacology 33:234-256.
Environmental Protection Agency (EPA). 1999. Policy and Procedures on Protection of Human Research Subjects in EPA Conducted or Supported Research. EPA Order. Available at www.epa.gov/oamrtpnc/forms/1000_17a.pdf.
EPA. Risk Assessment Forum. 2003. Draft Formal Guidelines for Carcinogen Risk Assessment. EPA/630/P-03/001A, NCEA-F-0644A.
Faustman, E. M., and G. S. Omenn. 2001. Risk assessment. In: C. D. Klaassen, ed. Casarett and Doull’s Toxicology: The Basic Science of Poisons. 6th ed. New York: McGraw Hill.
Food and Drug Administration (FDA). 2003. Good Clinical Practices. Available at www.fda.gov/oc/gcp.
Gaylor, D. W. 1996. Risk estimation—an overview: disease risk estimation based on animal bioassays. Drug Metabolism Reviews 28:9-27.
Greer, M. A., G. Goodman, R. Pleus, and S. E. Greer. 2000. Does environmental perchlorate exposure alter human thyroid function? determination of the dose-response for inhibition of radioiodine uptake. In: Abstracts of the 12th International Thyroid Congress, October; Kyoto, Japan. Endocrine Journal 47(suppl.):146.
Institute of Medicine (IOM). 2003. Responsible Research: A Systems Approach to Protecting Research Participants. Washington, D.C.: The National Academies Press.
Jett, D. A., R. V. Navoa, R. A. Beckles, and G. L. McLemore. 2001. Cognitive function and cholinergic neurochemistry in weanling rats exposed to chlorpyrifos. Toxicology and Applied Pharmacology 174:89-98.
Karanth, S., and C. Pope. 2003. In vitro inhibition of blood cholinesterase activities from horse, cow and rat by tetrachlorvinphos. International Journal of Toxicology 22:429-433.
Koenig, J. Q., D. S. Covert, W. E. Pierson, Q. S. Hanley, V. Rebolledo, K. Dumler, and S. E. McKinney. 1994. Oxidant and acid aerosol exposure in healthy subjects and subjects with asthma—part I: effects of oxidants, combined with sulfuric or nitric acid, on the pulmonary function of adolescents with asthma. Research Report (Health Effects Institute) 70:1-36.
Langley, S. J., S. Goldthorpe, M. Craven, J. Morris, A. Woodcock, and A. Custovic. 2003. Exposure and sensitization to indoor allergens: association with lung function, bronchial reactivity, and exhaled nitric oxide measures in asthma. Journal of Allergy and Clinical Immunology 112(2):362-368.
Lawrence, J. E., S. H. Lamm, and L. E. Braverman. 2001. Low dose perchlorate (3mg daily) and thyroid function [Letter]. Thyroid 11:295.
Lawrence, J. E., S. H. Lamm, S. Pino, K. Richman, and L.E. Braverman. 2000. The effect of short-term low-dose perchlorate on various aspects of thyroid function. Thyroid 10:659-663.
Leung, H. W., and D. J. Paustenbach. 1995. Physiologically based pharmacokinetic and pharmacodynamic modeling in health risk assessment and characterization of hazardous substances. Toxicology Letter 79:55-65.
Levin, E. D., N. Addy, A. Nakajima, N. C. Christopher, F. J. Seidler, and T. A. Slotkin. 2001. Persistent behavioral consequences of neonatal chlorpyrifos exposure in rats. Developmental Brain Research 130:83-89.
Lippman, M., B. S. Cohen, and R. B. Schlesinger. 2003. Risk assessment. In: Environmental Health Science: Recognition, Evaluation, and Control of Chemical and Physical Health Hazards. New York: Oxford.
Liu, J., K. Olivier, and C. N. Pope. 1999. Comparative neurochemical effects of repeated methyl parathion or chlorpyrifos exposures in neonatal and adult rats. Toxicology and Applied Pharmacology 158:186-196.
Maurissen, J. P. J., A. M. Hoberman, R. H. Garman, and T. R. Hanley, Jr. 2000. Lack of developmental neurotoxicity in rat pups from dams treated by gavage with chlorpyrifos. Journal of Toxicological Sciences 57:250-264.
Mileson, B. E., J. E. Chambers, W. L. Chen, W. Dettbarn, M. Ehrich, A. T. Eldefrawi, D. W. Gaylor, K. Hamernik, E. Hodgson, A. G. Karczmar, S. Padilla, C. N. Pope, R. J. Richardson, D. R. Saunders, L. P. Sheets, L. G. Sultatos, and K. B. Wallace. 1998. Common mechanism of toxicity: a case study of organophosphorus pesticides. Journal of Toxicological Sciences 41:8-20.
Moser, V. C., S. M. Chanda, S. R. Mortensen, and S. Padilla. 1998. Age- and gender-related differences in sensitivity to chlorpyrifos in the rat reflect developmental profiles of esterase activities. Journal of Toxicological Sciences 46:211-222.
National Research Council (NRC). 1994. Science and Judgment in Risk Assessment. Washington, D.C.: National Academy Press.
Paustenbach, D. 2002. Section A. Theory: human health assessment. In: Human and Ecological Risk Assessment. New York: John Wiley & Sons.
Rodricks, J. V., L. Rudenko, T. B. Starr, and Turnbull, D. 1997. Risk assessment. In: Sipes, I., C. McQueen, and A. Gandolfi, eds. Comprehensive Toxicology 315-338. Pergamon, Elsevier Sciences: Oxford, UK.
Samet, J. M., W. E. Pepelko, B. Sonawane, G. E. Hatch, K. E. Driscoll, and G. Oberdörster. 1994. Risk assessment of oxidant gases and particulate air pollution: uncertainties and research needs. Environmental Health Perspectives 10(Suppl.):209-213.
Steenland, K., R. B. Dick, R. J. Howell, D. W. Chrislip, C. J. Hines, T. M. Reid, E. Lehman, P. Laber, E. F. Krieg Jr., and C. Knott. 2000. Neurologic function among termiticide applicators exposed to cholorpyrifos. Environmental Health Perspectives 108:293-300.
Wessels, D., D. B. Barr, and B. Mendola. 2003. Use of biomarkers to indicate exposure of children to organophosphate pesticides: implications for a longitudinal study of children’s environmental health. Environmental Health Perspectives 111:1939-1946.