Review of the Department of Defense Research Program on Low-Level Exposures to Chemical Warfare Agents (2005)

The overall goal of the U.S. Department of Defense (DOD) Master Research Plan (Research Plan) on low-level exposure to chemical warfare agents (CWAs) is to obtain information that can be used to protect DOD personnel¹ from (1) potential operationally relevant performance decrements and (2) potential delayed adverse health effects from exposures to low levels of CWAs, with initial emphasis on the organophosphorus anticholinesterase nerve agents and sulfur mustard (HD). Other chemicals, such as hydrogen cyanide and tear agents, are given a lower priority. To achieve this goal, the primary objectives of the Research Plan are (1) to obtain the appropriate data for the identification of the most sensitive end point(s) applicable to humans that are indicative of early effects on operationally relevant performance decrements or potential delayed adverse health effects after low-level exposure to anticholinesterase nerve agents and other CWAs, especially during military operations, and (2) to identify strategies for use of the research data for human health risk assessment.

As presented in the Research Plan (Table 5-2, p. 8), the specific objectives are (1) to characterize concentration-time (Ct) relationships for low-level, longer time CWA vapor exposure by conducting in vivo inhalation studies; (2) to identify alternative, but physiologically significant, toxicologic end points resulting from low-level CWA exposures using end points indicative of performance decrements, mechanisms of toxicity, and potential for persistent or delayed health effects (e.g., neurobehavior effects, changes in gene expression); and (3) to conduct appropriate integration studies linking experimental data sets with predictive human health effect assessments using cross-species/cross-route data, toxicokinetic modeling, and biomarkers.

The DOD Research Plan states that the nerve agents (G-series agents tabun [GA], sarin [GB], soman [GD], and cyclosarin [GF], and VX) are being, or are going to be, studied alone and in combination with other CWAs after single and multiple exposures by the inhalation route. The highest research priority, as indicated in presentations made by DOD personnel during the committee meetings, is to evaluate the effects of single nerve agents. DOD notes that the dose-response relationships for nerve agent effects studied to date are extremely steep and therefore require a revision of the default uncertainty-factor approach to identify exposure levels that are just under threshold levels. For example, in typical noncarcinogenic risk assessments, a no-observed-adverse-effect level is divided by an uncertainty factor (generally by a factor of 10) to account for variability in sensitivity for intraspecies or interspecies extrapolations. Thus, considering the markedly steep nature of the dose-response curves for effects elicited by these agents, it may be possible to incorporate smaller uncertainty factors (e.g., 3) that would still adequately account for susceptibility differences. The data obtained from the proposed research are to be evaluated for the ability to contribute materially toward quantitative refinement of the human health risk assessments for exposure to low levels of CWAs.

The Research Plan consists of three major research thrusts, with each of the major thrusts subdivided into subthrusts. Subthrusts are further divided into research tasks.

• Major Thrust I: Characterize Ct relationships (concentration-time response) for low-level, longer-time CWA vapor exposures.

• Major Thrust II: Identify alternative, but physiologically significant, toxicologic end points.

• Major Thrust III: Conduct appropriate integration studies linking experimental data sets with predictive human health effect assessments.

The primary question addressed by the committee was, “Do the studies that DOD proposed in its Research Plan match the objectives stated above?” DOD is particularly interested in generating research data that can be used for human health risk assessment and risk management. Risk of adverse effects might ensue from the first attack or from subsequent attacks with CWA munitions. DOD might accept reasonable risk to complete a mission, but information needs to be provided to the commander of the operation to facilitate choosing the most appropriate action(s) (action that reduces risk to personnel but minimizes interference with the military mission). Acceptable risk is expected to be determined by the commander based on the understanding of the situation and mission requirements at that time. Of most concern to DOD currently is a single CWA attack (acute or short-term exposure) and the potential need for mission-oriented protective posture (MOPP), and a deficiency of information on how long such protection may be needed.

This major research thrust is subdivided into two subthrusts: (1) Subthrust IA, conduct in vivo inhalation studies to define Ct relationships; and (2) Subthrust IB, improve capabilities to establish and maintain low-level vapor exposure systems for in vivo (animal) inhalation studies. Subthrust IA is subdivided into five tasks (Tasks IA1 to IA5), which are essentially the same. Each of the tasks involves studying the five CWAs (GB, GF, GD, VX, and HD) individually. Experimental studies proposed in the DOD plan for those goals (tasks) either have been conducted for some CWAs or are in progress or planned with mice, rats,

guinea pigs, and minipigs. Studies in nonhuman primates are apparently being planned at Edgewood, Maryland and Wright-Patterson Air Force Base (WPAFB). Exposures are primarily by the inhalation route with experimental animals, and it is assumed that results would be qualitatively similar if exposures were by other route. The committee recommends that toxicity end points need to be based on tests other than simple cholinesterase inhibition in blood because this end point has been recognized to correlate poorly with clinical signs. Studies currently proposed by DOD focus on miosis as the most-sensitive operationally significant effect resulting from nerve-agent exposure. Effects on rodent behavior are under investigation as alternative end points indicative of exposure.

This subthrust is subdivided into the following five tasks:

• Task IA1: GB inhalation studies to investigate 50% lethal Ct (LCt₅₀) relationships and miosis (rat and swine).

• Task IA2: GB inhalation studies to investigate LCt₅₀ relationships and miosis (rat and swine).

• Task IA3: VX inhalation studies to investigate LCt₅₀ relationships and miosis (rat and swine).

• Task IA4: GD inhalation studies to investigate LCt₅₀ relationships and miosis (rat and swine).

• Task IA5: HD inhalation studies to investigate nonlethal lung and systemic effects in swine.

These tasks are discussed below.

The goal of this research task for agent GB is to demonstrate a dose-response relationship for unequivocal end points of operational significance and to validate or refute the historical database for short-duration exposures. An extension of this goal is to provide new

data for exposure durations well beyond previous studies using a well-established animal model for toxicity studies (rat) and an animal model that provides more direct extrapolation to human physiology (swine). According to DOD, this research task will provide a clear description of the Ct-effect profile for GB and predict results in humans based on well-established physiologic extrapolation methods for longer, operationally relevant exposure times (6 hours). DOD research efforts for nerve agents will include studying neurophysiology, biomarker, and dose-metric end points.

Recent studies with GB have been conducted and published by Mioduszewski et al. (2000, 2001, 2002a). Studies of acute lethality and mitogenic response have been reviewed in a National Research Council (NRC) report (2003).

Committee’s Evaluation, Conclusions, or Recommendations for Task IA1

As indicated earlier, of note in the data provided by DOD personnel during their presentations at the committee meeting is that miosis is considered an important end point. Much miosis data collected from human studies on GB already exist (NRC 2003, Table 1-7 and Table 1-10, p. 51), and these data should be considered before planning further experiments in animal models to evaluate tissue binding. Evaluation of existing data, identification of data gaps, and sufficient planning would promote appropriate and efficient use of experimental animals. For example, effects of topical exposure on the eye could be discriminated from effects of systemic (inhalation) exposure by conducting experiments with and without eye protection. Also, during experiments examining effects of nerve agents on the eye, it should be possible to examine selected behavioral end points under various ambient light conditions as well. For example, the primate equilibrium platform test, which may serve to measure operationally important performance decrements, has been shown to be sensitive to the effects of anticholinesterase exposures (Hartgraves and Murphy 1992; Murphy et al. 1993; Blick et al. 1994). Therefore, including sophisticated neurobehavioral tests may require the use of nonhuman primates in animal experiments.

The committee further recommends that the research effort also involve studies carried out in mammals other than nonhuman primates. However, the committee recommends that primates be considered the most appropriate species for behavioral studies and for studying the possibility of long-term effects from acute exposures. This recommendation is consistent with the recommendation of the 2003 NRC report. The committee recognizes the issues associated with testing nonhuman pri-

mates (e.g., high cost and restrictive laboratory environments), but such testing is recommended in this case. The committee does not make this recommendation lightly but sees such studies as crucial in light of the high precision needed for the estimation of risk estimates for these CWAs, especially since there are competing risks from wearing protective gear.

Regardless of the animals being studied, blood cholinesterase concentrations should be determined routinely, and tissue cholinesterase concentrations from brain and other tissues (e.g., eye) should routinely be collected for determination of acetylcholinesterase (AChE) activity and/or agent regeneration whenever animals are killed. The committee considered the lack of correlation between cholinesterase inhibition and miosis or other low-dose effects of CWAs. Although this may be the case with blood cholinesterase inhibition and changes such as miosis, the correlation between cholinesterase inhibition in the eye and the development of miosis may be high. Experimental studies in which CWAs are instilled in the eyes, with appropriate dose ranges for correlation with cholinesterase inhibition in the eye, may be informative. While blood cholinesterase inhibition may not always correlate with functional changes after CWA exposure, it is a reasonable expectation that miosis and cholinesterase inhibition in the sphincter muscle controlling pupillary constriction will correlate. If they do not, then some other mechanism may be operating, and a search for alternative macromolecular targets in the ocular tissue may be warranted.

The ongoing study evaluating miotic effects of aerosolized CWAs with minipigs was suggested by DOD to be a state-of-the-art approach for evaluating effects of low-level exposures on the putative most-sensitive end point. One issue of concern for animal studies, however, may be the need for restraining the animals to focus the camera on the pupil. It is clear from the DOD Research Plan that although studies to evaluate coexposures to stress along with CWAs are not a high priority, the design of miosis studies in minipigs may have stress as a confounder. A number of reports show that stress alters a variety of physiologic parameters that could be important in responses to CWAs. Thus, researchers should consider measuring plasma corticosteroid concentrations along with other relevant blood measurements during the experimental period. This testing also should be considered for any studies with the primate equilibrium platform, which was found to be a sensitive tool for assessing subtle behavioral effects of GD (Hartgraves and Murphy 1992).

Miosis occurred at 1% of the lethal dose in 50% (LD₅₀) of the minipigs during inhalation studies. Duration of exposure, however, affected the dose-response curve and the calculation of the effective concentration in 50% of the animals (EC₅₀) for this low-dose effect in this species. Information on interspecies relevance of this end point was not provided in the DOD plan or in DOD presentations to the committee, although it was suggested that the EC₅₀ for miosis in humans exposed to nerve agents is approximately equivalent to that in minipigs. This is supported by information in the 2003 NRC report.

The pupillary size is reactive to levels of ambient light as well as various concentrations of nerve agents, so the levels of ambient light confound the pupillary effects of CWAs. Therefore, the effects of a CWA should be determined under known lighting conditions to quantify such interactions and to inform field commanders about those lighting conditions under which miosis may be an insensitive (or overly sensitive) toxicologic end point. A rapid and accurate sensor of both miosis and ambient light will be needed to provide information necessary for field commanders to make decisions. It will also be neces-sary to know if the range of decrements in pupil size expected with exposure to low concentrations of CWAs would overlap with the pupil size of unexposed individuals. Existing human and animal data should be of value here (NRC 2003). For operational risk management, there is a need to review data on decrements in task performance with changes in pupil size under various conditions of ambient light, including relative magnitude and relative durations of these changes (such as those induced by the Food and Drug Administration [FDA] approved organophosphate [OP] eye drops—echothiophate). Studies with reversible cholinesterase inhibitors, currently used as therapeutic agents, in people as well as in several animal species including nonhuman primates, could have value. The DOD plan indicates studies would be conducted on various CWAs to evaluate the Ct relationship for each agent in eliciting miosis. If each CWA could be demonstrated to elicit miosis by inhibiting AChE in the iris sphincter muscle (as expected), it might be feasible to model concentration-time relationships by studying organophosphorus-AChE inhibition kinetics in that muscle in vitro and in vivo.

In light of the high precision needed for estimating risk under varying exposure conditions, the committee recommends that studies be undertaken to determine whether miosis is the sole cause of operationally relevant performance decrements in humans after low-dose CWA exposure. This research could be done safely in humans using FDA-approved topically applied anticholinesterase agents and in nonhuman primates

using both CWAs and FDA-approved anticholinesterase agents. Furthermore, these studies should be replicated in nonhuman primates using both CWAs and non-CWA agents to determine whether miosis is also the sole cause of performance decrements.

Acute exposure guideline levels (AEGLs) for exposures from 10 minutes to 8 hours are available for the nerve agents GA, GB, GD, GF, and VX (NRC 2003). AEGL-1 (causing nondisabling and reversible effects) values were based on miosis in rats, nonhuman primates, and humans for the G-series compounds. In the NRC (2003) report, the miotogenic response of mammalian eyes was similar across species; therefore, the interspecies uncertainty factor for miosis was considered to be 1. It should be noted that AEGL values are for the general public, including susceptible individuals. Although differences in sensitivity to miosis from direct-acting nerve agent exposures may occur among healthy military personnel, they are not likely to differ substantially from the variability in the general population. It should also be recognized that differences in sensitivity due to differences in metabolic capacity of susceptible subpopulations would likely be of concern only with systemic, rather than local, exposures because differences in biotransformation (or metabolism) would be important only with systemic effects. When establishing AEGL-1 values, NRC (2003) also did not consider other end points that might be more sensitive or more likely to cause operational deficits than miosis.

As noted above, accurate determinations of effect thresholds for CWAs are hindered by the steepness of the dose-response curves. The linearity of these curves at the lower end is still unknown.

The committee concludes that the LCt₅₀ studies will probably not provide useful information for low-level exposures because deaths will not occur in animals exposed at low levels. Instead, DOD should focus on effective Ct in 50% of the subjects (ECt₅₀) for other end points that might be relevant to low-level exposures. The committee concludes that dose-duration-response studies provide more robust information than single toxicity end points, such as the LCt₅₀ and EC₅₀, for specific exposure durations; therefore, such research is appropriate and should be continued.

These three tasks are similar to Task IA1 except different CWAs are to be studied. That is, instead of GB, agents studied are GF, GD, and

VX. The committee’s conclusions and recommendations generally apply to these tasks also.

HD is a potent vesicant and is also carcinogenic. Considerable toxicity data are available in the literature for HD. DOD proposes to study dose-response for nonlethal lung and/or systemic effects for HD and characterize the dose-response-time-effect profile in swine. DOD has assigned a relatively low priority for this task. The committee recommends that such research be continued with a low priority assigned to it, and thus resources involving these studies should be limited. The committee concludes that DOD’s proposed research on HD in swine is appropriate.

This subthrust is subdivided into the following five tasks:

• Task IB1: GB vapor generation and chamber systems.

• Task IB2: GF vapor generation.

• Task IB3: VX generation and sampling.

• Task IB4: GD vapor generation and chamber systems.

• Task IB5: HD vapor generation and chamber systems.

These tasks are discussed below.

The goal of this research task is to overcome technical challenges to generate consistent vapor atmospheres for GB and relate fairly robust historical generation data to modern, validated chamber-exposure methodology. The DOD goal is also to develop an irrefutable technical method for delivering vapor exposure across a range of relevant concentrations and times. Generation and sampling analytical systems have

been developed; comparison of DAAMS (depot area air monitoring system) tubes with Bubbler methodology has been completed by DOD.

Committee’s Evaluation, Conclusions, or Recommendations for Task IB1

Conducting experiments with potent, volatile CWAs having very steep dose-response curves is exceedingly difficult. Of critical importance for studies of low-dose effects is a means to ensure a quantifiable, unvarying concentration over the course of data collection necessary for defining specific conditions under which toxicity will increase. The Research Plan recognizes that there is a need for inhalation chambers that provide constant, predictable delivery of CWAs. The Research Plan states that this research will help considerably in determining what constitutes a low-dose effect and whether such an effect will be the same across species.

The proper design for—and safe execution of—inhalation studies with potent CWAs poses a number of challenges. The testing laboratory must create, confine, and define such inhalation exposures with considerable care. Conventional inhalation exposure systems with comparatively large (e.g., 0.25-1.0 cubic meter [m³]) Hinners-type chamber systems—if operated properly and isolated from the rest of the laboratory—can serve for longer exposure durations. However, studies of short-term exposures (less than 1 hour) might require an alternative approach.

The issue arises from the need to fill and equilibrate the exposure chamber rapidly—ideally, in less than 10% of the total exposure time. Chamber volume, chamber ventilation (flow rate), and exposure duration all need to be balanced, an important requirement for proper execution and evaluation of short-term exposures. If a chamber is operated too slowly, then equilibrium to the targeted exposure concentration is delayed until well into the exposure period. Clearance of the agent or its metabolite postexposure is equally protracted. When evaluating the animals’ response (dose-response relationship, the objective of the study), researchers should consider whether eventual peak concentration or time-weighted average concentration best reflects the protracted equilibration and clearance of the exposure system. These considerations bear on the engineering, design, and operation of the exposure systems.

As a consequence, the protective assumption of response linearity from 8 hours to 24 hours was applied by the U.S. Army Center for Health Promotion and Preventive Medicine (USACHPPM) to develop 24-hour military exposure guidelines (MEGs) for nerve agent and HD from the 8-hour AEGLs (i.e., assuming Haber’s law that Ct = k for expo-

sure duration greater than 8-24 hours; it should be noted that Haber’s law is a special case of the ten Berge expression, when n = 1).

For DOD’s proposed task GF, VX, and GD, the committee’s conclusions apply because these agents are structurally similar. However, there are differences in volatility; VX, for example, is harder to volatilize than GB, GF, or GD. Therefore, volatility needs to be considered. VX also adsorbs into chamber surfaces, which makes it difficult to clean chambers between tests with different agents.

The committee concludes that DOD’s proposed research in response to the technical challenges in generating consistent vapor atmospheres for HD and the development of sampling and analytical systems is appropriate. DOD proposes to study dose-response for nonlethal lung and/or systemic effects for HD and characterize the dose-response-time-effect profile in swine. DOD has assigned a relatively low priority for this research task. The committee recommends that such research be continued with a low priority assigned to it, and resources involving these studies should be limited.

This major thrust is further divided into four subthrusts, subthrusts IIA to IID.

This subthrust is subdivided into the following five tasks:

• Task IIA1: Establish maximum tolerated doses (MTDs) for parenteral rodent models.

• Task IIA2: Neurobehavioral and cognitive changes in rodents.

• Task IIA3: Neurochemical and immunohistochemical changes in rodents.

• Task IIA4: Cardiomyopathy and cardiac arrhythmia in swine and mice.

• Task IIA5: Cognitive tests in nonhuman primates.

These tasks are discussed below.

The goal of this DOD research task is to identify the maximum absorbed dose of each agent that can be tolerated without clinical signs of systemic toxicity and without gross pathologic changes in likely target tissues from a sensitive rodent model. Another goal is to determine a range of doses for nerve agents that could be considered to be low-dose exposures for any relevant duration of exposure up to a maximum of 13 weeks.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIA1 The committee reviewed several presentations from DOD personnel that demonstrated that this research task was completed and that the established parenteral MTD was being used in follow-up studies in rodents. The MTD for the nerve agents of interest and the application of these data to follow-up neurobehavioral, pathophysiologic, and toxicogenomic studies is detailed in the Research Plan (pp. 14-15).

Because this work has been completed, it is of no value to make recommendations about whether the research should be conducted. The most important question for the committee is How do data from this work contribute materially toward a quantitative refinement of the human health risk assessment for low-level CWA exposures? (Research Plan, p. 17.) This concern is based on the fact that “cognitive tests will be performed in nonhuman primates using established low-level CWA exposures that show effects in rodents” (Research Plan, Appendix 3, Task IIA5, p. A3-7). The committee’s recommendation for primate studies suggests a potential reweighing of this task.

The goal of this DOD research task is to study alterations in behavior that are suspected to occur at CWA doses below those causing clinical toxicity. This hypothesis will be tested in rodents exposed singly and repeatedly to low-level CWA nerve agents. Neurobehavioral tests, such as the functional observational battery, identify changes in central and peripheral nervous system function beyond those observable in clinical signs of intoxication. According to DOD, cognitive performance tests are potentially more-sensitive indicators of subtle performance decrements and include operant conditioning, passive avoidance procedures, and radial arm maze learning tasks. DOD believes data from these studies will identify more subtle alterations in central and peripheral nervous system function due to CWA exposure.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIA2 The committee concludes that animal studies focusing on alterations in behaviors are difficult to use when extrapolations must be made across species and to humans. Behavior in a single species can be affected by time of day, environment, age, sex, and so forth. Interspecies extrapolations using behavioral end points will always be difficult, because baseline behavior contributes to whether an effect is seen. Baseline activities even of different species of rodents (rats and mice) are quite different. And although rodents can learn, the importance of their memory compared with that of humans (or nonhuman primates), is unknown, decreasing the significance of rodent behavioral data in risk assessment. These considerations do not negate the potential utility of complimentary neurobehavioral studies in lower mammals; however, nonhuman primates may be better animal models for subtle behavioral changes in humans (Hartgraves and Murphy 1992). The long-term effects reported in some humans exposed to concentrations of OP compounds below those causing toxicity end points indicative of cholinergic poisoning might result from a combination of factors, such as memory, fear, and stress.

Electroencephalogram (EEG) measurements were mentioned during a DOD presentation as a possible sensitive end point indicative of exposure to low concentrations of CWAs, but this has not been correlated with behavioral alterations.

It appears relevant to determine a behavior, if one exists, accompanying miosis that occurs after systemic (inhalation) or topical (ophthal-

mic) exposure to CWAs. Review of existing data collected from experiments in humans or with nonhuman primates should be valuable. For example, although the primate equilibrium platform test provides a means for detecting subtle behavioral effects of GD in nonhuman primates, it was not reported whether this was correlated with miosis (Hartgraves and Murphy 1992). Other behavioral end points that could be important in the evaluation of alternative toxicologic end points after low-level CWA exposures are deficits in attention and vigilance. Experimental tests—for example, a serial reaction time task (Robbins 2002; Ferraro and Gabriel 2003)—could be incorporated into the testing protocols in humans and both lower and higher mammalian species to evaluate cognitive function that could be pertinent to operational performance. Furthermore, data presented by DOD at a committee meeting suggested that unpublished results indicated visual decrements could occur even in the absence of maximal miosis. Because tests for attention, such as the serial reaction time task, depend on visual discrimination, combined measurements of miosis and visual attention/vigilance or other sensitive behavioral end points could provide important information on performance deficits with low-level CWA exposures.

In addition, DOD should collaborate further with TNO (Nederlandse Organisatie voor Toegepast Natuurwetenschappelijik Onderzoek) in the Netherlands. This institute has studied EEG, visual evoked responses, and miosis in guinea pigs and marmosets and has considerable experience in inhalation exposures for nerve agent exposure as well as analytical capabilities.

Regardless of the behavioral tests to be used, the committee recommends that consideration be given to evaluating possible adaptive changes in behavior after repeated low-level exposures to CWAs and that the connection between measured changes and overt clinical and behavioral effects be evaluated. Repeated exposures to relatively high concentrations of OP compounds are well known to elicit tolerance, at least partially because of adaptive changes in cholinergic receptors. For this context and the proposed DOD studies, focus on pharmacological adaptation along with the evaluation with drug challenges is sufficient. However, the concern of DOD does not focus much on repeated exposures. The dosages of CWAs used in preliminary studies presented, such as repeated exposures at 0.2-0.4 MTDs,² were shown to cause substantial AChE in-

hibition in brain and other tissues and therefore may elicit such cellular tolerance. Furthermore, some OP toxicants, including nerve agents, interact directly with some cholinergic receptor subtypes with high affinity. Thus, changes in cholinergic receptor function might occur with repeated agent exposures due to both indirect and direct interactions. In such cases, adaptation can be evaluated by drug challenge with cholinergic agonists or antagonists. For example, if attention or radial arm maze behavior was being evaluated after repeated nerve agent exposures, a muscarinic antagonist (e.g., scopolamine) or agonist (e.g., carbachol) could be used to evaluate the integrity of cholinergic transmission. Classically, persistent AChE inhibition elicits down regulation of cholinergic receptors. Thus, anticholinergic drug challenge under conditions of cholinergic receptor down regulation can often elicit an exaggerated response. In contrast, cholinergic-agonist challenge under conditions of receptor down regulation often leads to reduced responsiveness. Thus, characterization of dose-related responses to agonists or antagonists can be used to determine whether receptor-mediated adaptation has occurred.

The Navy Environmental Health Center has already deployed such a battery of tests, with human data to validate it. While these behavioral studies were focused on smaller mammalian species, extrapolation to humans is an obvious strength. More communication with and involvement of those investigators in the proposed DOD Research Plan regarding neurobehavioral assessments should be facilitated.

For this DOD research task, modern techniques will be used to identify changes in brain neurochemistry and cellular-level alterations in metabolism in rodents exposed to low-level chemical warfare nerve agents. According to DOD, data compiled from these studies will provide quantitative estimates of changes in brain chemistry and/or microscopic anatomy in response to chemical warfare nerve agents and will identify doses and durations of exposure required to induce the changes.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIA3 Data are already available in this area. A portion of this database is summarized in the study matrix references in the Research Plan (Appendix A4). The use of sensitive neurochemical and immunohistochemical methods in the examination of exposed animals is an important component of many of the proposed research efforts. All efforts should be made to include the collection and testing of relevant tissues when studies are designed and to determine the relationship between sensitive measures and more-overt measures of toxicity.

For this research effort, DOD plans to study the origin of cardiac anomalies seen with chemical warfare nerve agents at low levels of exposures. According to DOD, cardiac failure appears to be a major contributor to nonpulmonary toxicity of these agents in susceptible individuals at low doses. DOD believes conclusions derived from these studies will identify the central or peripheral origin of cardiac arrhythmias after nerve agent exposure as well as the need for, and type of, therapeutic interventions required.

Committee’s Evaluations, Conclusions, or Recommendations for Task IIA4 Low-level exposures to CWAs may elicit adverse effects through interaction with other macromolecules in addition to AChE. The research describing interaction of VX with Na⁺/K⁺-ATPase to elicit cardiotoxicity, along with preliminary findings by DOD investigators that brain Na⁺/K⁺-ATPase is affected by low-level nerve agent exposures, suggests this macromolecule may be an important additional site of action with some OP toxicants. It is also of interest that the cardiotoxicity of VX was suggested to be due to inhibition of an isoform of Na⁺/K⁺-ATPase (Robineau et al. 1991). In addition, some OPs bind to muscarinic M2 receptors, including those in the mammalian heart, in a potent manner. For example, recent findings indicate that the OP toxicant chlorpyrifos oxon binds irreversibly to M2 receptors (Bomser and Casida 2001; Howard and Pope 2002). While cardiac muscarinic receptors are prominent in the regulation of cardiac function, ganglionic nicotinic receptors also play a role, and nicotinic receptors have also been reported to be sensitive to direct binding of some OP nerve agents. Thus, there is

sufficient preliminary information to warrant examination of additional sites of action with low-level nerve agent exposure that could contribute to cardiomyopathy and cardiac arrhythmia. The committee recommends continuing this research into possible additional sites of action in the heart. The committee also recommends that preliminary experiments could be initiated using cardiac cell lines—for example, the use of cardiomyocytes prior to any in vivo studies to investigate potential alternate mechanisms of cardiac toxicity.

The goal of this DOD research task is to conduct cognitive testing that will be performed in nonhuman primates with established low-level CWA exposures that show effects in rodents. According to DOD, the use of nonhuman primates will permit more sophisticated, computerized testing and will ensure that any observed effects can be extrapolated to human populations.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIA5 The committee gives this task a high priority because it will provide the most relevant information useful in establishing the levels of exposure causing operationally relevant performance decrements in humans. Performance decrements in humans could be caused by miosis or by more subtle neurophysiological changes unrelated to miosis. Conducting accurate range-finding studies in the nonhuman primate model will be more useful than extrapolating a MTD from a rodent or using a dose of CWA that shows an effect in rodents, the method recommended in the plan. Studies should be conducted that build on previous efforts that used serial probe recognition after CWA exposure in nonhuman primates (Castro et al. 1994; Myers et al. 2002). The committee supports the DOD plan to conduct neurobehavior studies in nonhuman primates. Neurobehavioral studies using primates are currently on-going at WPAFB. When nonhuman primates are involved in this research, good laboratory practice (GLP) standards must be in place. The requirement for GLP is planned for the execution of all research in the DOD plan; however, no evidence of GLP documentation was presented to the committee.

This subthrust is subdivided into the following two tasks:

• Task IIB1: Toxicogenomic/toxicoproteomics.

• Task IIB2: In vitro electrophysiological and biochemical changes.

These tasks are discussed below.

The goal of this DOD research task is to identify changes in gene expression and protein expression induced by exposure to low-level CWAs. According to DOD, the analysis of toxicogenomic data sets will identify doses of CWAs that induce alterations in metabolic homeostasis; it also will establish threshold exposures that have minimal or no effect and will indicate biochemical pathways affected by CWA exposures. For this research task, tissues from animals exposed via inhalation to CWAs are being evaluated to determine whether this end point will correlate with the Ct profile.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIB1 The committee believes this type of research might be of some importance when the results suggest possible involvement of noncholinesterase targets. As noted above, if sensitive end points such as miosis can be dissociated from cholinesterase inhibition in the target tissue (in this case, the eye), then these types of screening approaches may be valuable in determining alternative target sites. In this context, a recent report (Chilcott et al. 2003) indicated that mastication after dermal VX exposure occurred before significant reduction in cholinesterase activity and onset of miosis and muscle fasciculations in domestic pigs. Comparative changes in gene and protein expression or differences in phosphorylation of target proteins could provide insight into mechanisms not involving cholinesterase inhibition, especially long-term, latent, and delayed effects. It is important to recognize that these are nontrivial stud-

ies and that they need to be related to end points already known. Controls and clear end points are important, especially since effects may occur that are not necessarily directly related to exposure to the test compound.

DOD has included in its Research Plan other experiments that would provide new data of academic interest, but the length of time required to obtain validated end points would make them difficult to adapt for field conditions, especially when acute exposure is of interest. These experiments include the research in toxicogenomics and the research on the binding and recovery of nerve agents from exposed tissues. However, the use of the fluoride-regenerated agent has promise for the further development of dose-response relationships for CWA toxicity at low levels of exposure. As suggested before, inhibition of cholinesterase activity in blood may poorly correlate with physiological changes at low-level exposures. Part of this difficulty is undoubtedly due to the problem in detecting small alterations in the activity of a large set of enzymes (a small signal/noise ratio). The detection of regenerated CWA eliminates this difficulty, with essentially zero background or noise. This inherent difference should provide much higher sensitivity for detecting recovery of agent from tissue samples than for measuring loss of enzyme activity. Use of the fluoride-regenerated agent provides an opportunity to study effects of the agent in human tissues in vitro. Comparative studies on the regeneration of agents in red blood cells from different species, including humans, should be performed—for example, stability conditions, correlation with cholinesterase inhibition, and so forth. The possible effects of dietary fluoride on the regeneration of agents from red blood cells also should be evaluated, and as noted above, the regeneration of agent compared with cholinesterase inhibition in other tissues (e.g., the eye) should be studied to shed light on possible alternative targets (e.g., different binding proteins) involved in low-level effects.

In results DOD presented to the committee, only apoptosis measured by the Comet assay in lymphocytes of guinea pigs exposed to GD and changes in cortical Na⁺/K⁺-ATPase (increased [3H]ouabain binding) in guinea pigs exposed to GD appeared at exposures less than the MTD. Studies with other nerve agents are lacking. Furthermore, DOD reported that some changes that did occur disappeared rapidly upon cessation of exposure, reducing their usefulness as biomarkers. Results of the Comet assay for apoptosis in lymphocytes were quite robust, however, even at dosages smaller or equal to MTD, potentially representing an additional adverse response to low-level nerve agents. Toxicogenomic studies are only in survey mode. The relevance of these studies in terms of risk as-

sessment of low-level exposures to CWAs are currently unknown. Therefore, these studies should be assigned a low priority.

For this DOD research task, specific alterations in cellular-electrical excitability, synaptic activity, and metabolic state will be studied in brain tissue culture slices and in isolated cell cultures exposed to low concentrations of CWAs. According to DOD, these studies will identify intracellular and neural network mechanisms for low-dose nerve agent effects, differences in agent effects at the cellular level, and pharmacologic interventions required for therapy.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIB2 The description of this DOD research task is relatively vague. However, these tests might provide some mechanistic information, although extrapolation of results from single cells and tissue slices to whole animals is extremely difficult. Therefore, the likelihood of generating data that will identify pharmacologic interventions for treatment of individuals with low-level CWA exposures seems remote. Given the overall emphasis on protecting soldiers from operationally relevant performance decrements with low-level CWA exposures, this task seems to be a lesser priority than many others in the DOD plan. The committee therefore recommends that less emphasis be placed and less resources be utilized in performing studies in this task.

Overall, DOD presentations to the committee suggested that this area had lower priority than studies on immediate effects that could interfere with military performance.

The current problem, as presented by DOD, is that determining the lowest dose for effect (either for performance decrement or for adverse health effects) has been difficult. With MTD defined by DOD as the upper limit of exposure without clinical signs, the nonlethal, repeated MTD of nerve agents is approximately 0.2-0.4 LD₅₀ per day. Nonlethal effects from exposure to repeated dosages greater than or equal to MTD

included significant changes in gait, increased startle response, alterations in passive avoidance behavior, enhancement of mitogen-activated protein (MAP2) activity, EEG changes, messenger RNA (mRNA) alterations, and histopathologic changes on silver staining. Startle response was inconsistent across test compounds.

This research thrust has the following three tasks:

• Task IIC1: Neurobehavioral effects in rodents up to 1 year postexposure.

• Task IIC2: Synaptic receptor density up to 1 year postexposure.

• Task IIC3: Immunohistochemistry and neuron density up to 1 year postexposure.

These tasks are discussed below.

For this DOD research task, animals exposed either acutely or subacutely will be followed for up to 1 year after exposure, with repeated neurobehavioral testing. According to DOD, low-level CWA expo-sure might produce performance deficits that manifest months after exposure or result only from multiple exposures. Results from animals exposed by inhalation or parenteral routes will be correlated with each CWA profile for the determination of operationally significant effects. According to DOD, findings from these studies will also indicate the potential risk due to low-dose CWA exposures of persistent or delayed-onset neuropathology that could produce long-term disabilities.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIC1 The long-term effects of low-level exposures to agent GB (0.5 milligrams per cubic meter [mg/m³] for 30 minutes) in humans have been studied by Baker and Sedgwick (1996). This exposure caused 60% inhibition of red blood cell AChE activity after 3 hours or 3 days. Human volunteers exhibited miosis and in some cases mild dyspnea. Small changes in single fiber electromyography of the forearm were measured 3 hours and 3 days postexposure and were still detectable at the first follow-up examination 15-30 months postexposure. The authors suggested these electrophysiologic changes “may indicate subclinical onset of non-

polarizing type of neuromuscular block” that is reversible. These findings are subclinical. They are consistent with the research of Senanayake and Karalliedde (1987) with OP insecticides. The committee concludes that such research in laboratory animals by DOD might have limited relevance because, if there are performance decrements, they would be quite subtle and difficult to differentiate from control animals.

For this DOD research plan, rodents previously exposed to CWA nerve agents for short-term exposures will be periodically examined after exposure, and brain levels of cholinesterase, muscarinic acetylcholine receptors, and other targeted proteins will be quantitatively determined. According to DOD, alterations in these protein levels and recovery times to normal levels will indicate the rate of recovery of normal brain chemistry after perturbation by low-dose CWA, thus pointing to either improved prognosis with time or lack of recovery.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIC2 In general, exposures to either OP CWAs or OP insecticides sufficient to inhibit brain AChE activity substantially for prolonged periods elicit adaptive changes (down regulation) in cholinergic receptors. This down regulation of cholinergic responses in response to AChE inhibition is considered a primary mechanism in the development of tolerance. The adaptive changes in receptor regulation are thought to be elicited through significant increases in synaptic acetylcholine levels after inhibition of AChE.

With very low-level exposures to anticholinesterases, adaptive down regulation would not be anticipated because the level of AChE inhibition likely would be insufficient to alter synaptic acetylcholine levels. However, low-level CWA exposures (repeated MTD exposures) as defined here may be sufficiently high to lead to adaptive receptor changes based on the degree of AChE inhibition elicited. Furthermore, as noted before, some OP anticholinesterases have differential direct effects on cholinergic receptors, in particular the muscarinic M2 subtype. Thus, neurochemical evaluations of specific cholinergic receptor changes after low-level CWA exposures could provide important findings. However, the committee believes that relatively short-term exposures would not be expected to lead to long-term (1-year) alterations in these cholinergic

markers. The committee recommends that these studies be considered but that shorter time periods after cessation of dosing be evaluated first before extending to such long-term studies. In addition, the relevance of these changes for extrapolation to whole animals needs to be elucidated. The committee assigns a lower priority for this research task.

For this DOD research task, the characteristics of brain neurons after exposure will be monitored during and after short-term exposure to chemical warfare nerve agents. Alterations in immunohistochemical staining in existing cells and overall cell number will be monitored after exposure to identify whether any cell populations are selectively affected by the CWA and whether these cells degenerate over time or recover function. According to DOD, such results might indicate changes in brain functions in rodents that are too subtle to detect through behavioral testing.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIC3 Henderson et al. (2002) found effects on the muscarinic receptor sites for cholinesterase in the brain only in the combined exposures to GB and heat stress. The committee recommends, based on the investigations of Henderson et al. (2002), that if these tests are conducted, they be done with and without heat stress. The committee, however, concludes that this research task would provide limited information relevant to determining exposure guidelines for low-level exposures.

This research subthrust has the following two tasks:

Task IID1: Identify polymorphisms in human blood esterases.

Task IID2: Computer modeling of cardiac failure to chemical warfare nerve agents.

These tasks are discussed below.

For this DOD research task, differences in levels and subtypes of proteins that bind or degrade OP nerve agents, including chemical warfare nerve agents, will be identified in human blood samples. According to DOD, data sets from these studies will indicate the relative numbers of highly susceptible individuals in the military population. It is possible that testing could identify such individuals.

Committee’s Evaluation, Conclusions, or Recommendations for Task IID1 Differences between individuals in blood cholinesterase activity may affect their susceptibility to the toxic effects of CWAs. It has been shown that a small subpopulation of men and women possess genetically determined variants in their plasma cholinesterase resulting in very low activity levels, which therefore may increase their susceptibility to CWAs (reviewed by NRC 2003, p. 156). Several studies reviewed by NRC (2003) indicate that homozygous individuals have plasma cholinesterase activity reduced to less than 25% of the normal value. For heterozygous individuals, mean plasma cholinesterase activity is 64% of normal (range, 28% to 114%). About 3% of individuals may have genetically determined low levels of plasma cholinesterase and therefore may be unusually sensitive to some anticholinesterase compounds. The frequency of the atypical homozygous phenotype is estimated at 0.025% (reviewed by NRC 2003).

Several studies indicate that plasma and red blood cell cholinesterase activity is significantly lower in women than in men (reviewed by NRC 2003). Plasma cholinesterase activity also may be depressed in pregnant women and in people with liver disease, heart disease, allergic conditions, and neoplasms (reviewed by NRC 2003). Such individuals also may be at greater risk from exposure to OP compounds. Although some investigators consider gender differences in plasma cholinesterase activity to be confined to young persons (reviewed by NRC 2003), data are available suggesting that adult females may be more susceptible than males to nerve agents. Yokoyama et al. (1998b, reviewed by NRC 2003) reported vestibulocerebellar effects (increased postural sway) in a small population of patients tested 6-8 months after being exposed to GB during the Tokyo subway terrorist attack. Both female and male patients (nine of each gender) had similar levels of plasma cholinesterase inhibition after the attack, and, in both genders, postural sway was inversely

correlated with plasma cholinesterase activity; however, only in females was the increase in sway significantly greater than in controls.

NRC (2003) considered females to be part of the susceptible subpopulation. In the Mioduszewski et al. (2000, 2001, 2002a, 2002b; reviewed by NRC 2003) studies on rats, females were statistically more susceptible than males for the lethality end point. For agent GF, LCt₅₀ values generally were lower in adult female rats than in adult male rats (Anthony et al. 2002; reviewed by NRC 2003). The observed increased susceptibility of females was taken into account by applying an intraspecies uncertainty factor for susceptible subpopulations in the estimation of acute exposure guideline levels.

A-esterases (paraoxonase/arylesterase) present in the blood and liver are also capable of hydrolyzing phosphate esters (reviewed by NRC 2003, p. 154). Paraoxonase is also known to be polymorphic in human populations, and individuals express widely different enzyme levels. People expressing certain isomeric forms of the enzyme with low hydrolyzing activity are considered to be more susceptible to OP anticholinesterase poisoning.

Carboxylesterases, another enzyme group capable of binding with certain OP compounds, are present in human blood, monocytes, liver, kidney, and lung. The detoxication potential of carboxylesterases is multifaceted and is an area that requires further experimental characterization.

It should be noted, however, that polymorphisms in blood esterases are important primarily for systemic effects, and they are unlikely to be important for the miotic response that follows topical exposure. If, however, there is concern for susceptibility to systemic effects, DOD’s proposed research for this task is appropriate. DOD should also study gene markers for unusual susceptibility to nerve agents and HD. DOD should try to leverage research being done in genomics in other laboratories.

The goal of this DOD research effort is to develop computer models of excitatory activity in networks of cardiac tissue that might be used to identify ionic mechanisms that undergo failure with chemical warfare nerve agent exposures. According to DOD, this research might be used to identify specific ionic channels/biochemical pathways that are highly susceptible targets for chemical warfare nerve agents at low levels, and

the findings from this research might lead to the development of therapeutic approaches to treating chemical-warfare-nerve-agent-induced arrhythmias.

Committee’s Evaluation, Conclusions, or Recommendations for Task IID2 The committee referred to DOD’s concerns for cardiac toxicity of CWA with respect to Task IIA4. In particular, the committee encourages research on “interactions of different CWA with cardiac muscarinic receptors and changes in cardiac function” and also possible low-dose effects on myocardial Na⁺/K⁺-ATPase.

By contrast, Task IID2 calls for the use of computer models to identify “ionic mechanisms that undergo failure” due to CWA exposure in order to “identify specific ionic channels/biochemical pathways that are highly susceptible targets” for CWA. The task itself refers to the “modeling of cardiac failure.”

The committee has concerns about this task. It is not clear whether the data to be modeled are those to be generated in Task IIA4 or will be obtained elsewhere. It also is not clear that this is intended to address low-level exposures. Task IIA4 is being undertaken to determine whether there are low-level-exposure–related biochemical perturbations leading to cardiac toxicity and dysrhythmias and, if so, to characterize them. Accordingly, Task IID2 seems premature. It is not yet obvious that cardiac effects are important in the context of the principal DOD concerns. It also is not obvious that “cardiac failure” is a relevant end point for low-level exposures.

The committee concludes that Task IID2 has not been adequately described or justified in the Research Plan. It is not clear whether cardiac failure and CWA-induced arrhythmias are appropriate end points for low-level exposures. It is not clear whether cardiac failure refers to myocardial dysfunction or disturbances of cardiac rhythm. The Research Plan refers to agent-induced arrhythmias, but it is not clear whether that actually addresses disordered conduction and/or ectopy, which are potential causes of arrhythmias, or alterations in heart rate (e.g., bradycardia due to muscarinic stimulation, tachycardia due to nicotinic stimulation), which may or may not be sufficient to warrant the term dysrhythmias. The task description does not indicate whether computer modeling will be based on data generated in Task IIA4 or obtained elsewhere; the relationship between Tasks IID2 and IIA4 should be explicitly described. If there are other sources of appropriate data, then their availability might

reduce the need for Task IIA4. If not, then Task IID2 cannot be justified until the results of Task IIA4 are available.

This major research thrust is further divided into two subthrusts: Subthrust IIIA and Subthrust IIIB.

This research thrust is subdivided into two tasks. Both of these tasks are considered together.

The goal of DOD Task IIIA1 is to demonstrate chemical warfare nerve agent equivalents for two or more experimental animal model systems in predicting well-established end points (e.g., lethality) and compare routes and species dose-response relationships to reconcile historical data sets. The goal is also to deliver a validated basis for predicting dose-effect-time profiles for GB and similar agents to refine future studies and minimize duplication of effort.

The goal of DOD research Task IIIA2 is to develop and extend, if necessary, candidate physiologically based pharmacokinetic (PBPK) models applicable to the human exposure response for CWA and to generate criteria for model evaluation based on the ability of the model to reflect known human and animal exposure data. The goal is to critically assess the model’s capability to reflect known human and animal nerve-agent-exposure data and to define any data gaps and, if needed, to determine what parameters and data would be necessary to improve the selected animal model. According to DOD, results from these studies will provide critical operational data to develop guidelines for human activities in a CWA-contaminated environment. Current exposure data

for response due to local effects are lacking. DOD also believes the proposed research would use PBPK modeling to overcome some of the problems with extrapolation from species and from differences in exposure conditions.

Committee’s Evaluation, Conclusions, or Recommendations for Tasks IIIA1 and IIIA2 In the health-risk-assessment area, the DOD plan appropriately focuses on uncertainties regarding concentration-duration-effect relationships and interspecies extrapolation (Tasks IIIA1 and IIIA2). However, as noted earlier in this chapter and in Chapter 4, many of the interspecies differences are well characterized, and further research should (1) focus on minimizing extrapolation, rather than characterizing it; (2) focus on low-dose effects rather than lethality; and (3) be closely tied to determining the critical effects, since different end points might require different extrapolation approaches. In addition, the use of lethality as an end point for low-level exposures is not a good choice; some other end point such as miosis or ocular irritation should be considered. PBPK models can help to address these uncertainties as well as help address differences in breathing rate between experimental and operational conditions; the development of criteria for evaluating PBPK models is important. One of the key aspects of such models would be to ensure that the model provides information on dose metrics relevant to the critical end points.

This subthrust is subdivided into the following four tasks:

• Task IIIB1: Correlation of markers of absorbed CWA dose to physiological effects.

• Task IIIB2: Novel protein/genomic markers for CWA exposures.

• Task IIIB3: Identification of cutaneous biomarkers for HD.

• Task IIIB4: Noninvasive neurological tests for CWA exposure (EEG, magnetic resonance imaging [MRI], conduction velocity [CV], single fiber electromyography [SFEMG]).

These tasks are discussed below.

The goal of this DOD research task is to develop a consistent biomarker in body tissues that is directly related to the absorbed dose of a given chemical agent and its physiological effectiveness. The goal is also to demonstrate that this dose-metric profile for exposures in an animal model provides direct extrapolation to human physiology. The proposed biomarkers/dose metric to be studied includes alkyl phosphonates, regenerated nerve agent, AChE activity, and butyrylcholinesterase activity. Sites monitored would be relevant to the development of pharmacokinetic models that include blood, lungs, kidney, adipose tissue, brain, muscle, and liver. The biomarkers/dose metrics from inhalation and intravenous routes of exposure would be assessed at three levels, ranging from near lethal to a low-level end point such as miosis. Serial blood samples would be drawn during and after exposure and analyzed for each biomarker/dose metric. According to DOD, the data generated from those studies are likely to provide information on which biomarkers/dose metrics should be used for modeling and health risk assessment studies.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIIB1 It is vital to obtain reliable indicators of exposure for kinetic analyses. Of the four types of biomarkers to be studied in this task, two of the four are enzymes (AChE, butyrylcholinesterase). An inherent difficulty in using enzyme inhibition as a biomarker of exposure (in particular with low-level exposures) is that the measurement is “loss” of activity. Thus, the signal/noise ratio for low degrees of enzyme inhibition is small. Against the background of substantial activity, a small loss is statistically difficult to demonstrate. This difficulty makes the use of both esterases for this purpose less than ideal. In contrast, the other two biomarkers (alkyl phosphonates and regenerated nerve agent) are not measured against a background of activity (there should be no alkyl phosphonates or nerve agent to be regenerated if exposure has not occurred). This inherently makes these two biomarkers more sensitive indicators of exposure. Thus, the committee recommends making more efforts to correlate either alkyl phosphonates or regenerated nerve agent with physiological effects and making fewer efforts with the enzymes as markers of exposure. The preliminary findings presented by DOD to the committee also demonstrated that the method for detecting regenerated nerve agent had already been successfully used in DOD laboratories and was suitable

for multiple CWAs. An added strength of the regenerated nerve agent method is that the agent, instead of one of its metabolites, can be identified from tissue extracts. Furthermore, studies can be used to directly compare regeneration of nerve agent in tissues (e.g., whole blood) from humans and laboratory animals. The committee recommends that future efforts on this task be focused on the regenerated nerve agent method for biomonitoring exposures.

The committee supports the intent of research Task IIIB1, which calls for developing a “consistent marker in body tissues that is directly related to the absorbed dose.” On the other hand, it is not clear that such a marker will necessarily correlate with all physiological effects of exposure or at all dose levels, as implied in the task description. Given that underlying DOD concerns are low-level exposures, it is not obvious that assessment of all potential biomarkers at levels of exposure ranging from “near lethal to a low-level end point such as miosis” will be efficient. The committee recommends that studies initially focus on effects from low-level exposure and that particular attention be given to fluoride regeneration of agent. Analytical end points and methods relevant to field concerns and field application should be emphasized. Attention also should be paid to developing markers of absorbed dose for HD.

The goal of the DOD research task is to evaluate differential enzyme and molecular biological markers for CWA exposure that might supplement cholinesterase as an indicator that low-level exposure has occurred. According to DOD, pattern analysis of multiple biomarkers would determine whether those changes are sufficiently unique to act as diagnostic or forensic markers of CWA exposure.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIIB2 In the case of organisms exposed to G agents, the presence of the agent is detectable for only a few hours. Therefore, intact G agents are not good candidates for retrospective detection of exposure. The metabolism of G agents takes place primarily by hydrolysis. In addition to binding to AChE, they bind with the closely related plasma protein butyrylcholinesterase and to carboxylesterase; binding to serum albumin also occurs (Noort et al. 2002).

Polhuijs et al. (1997) developed a method for analyzing phosphorylated binding sites (e.g., butyrylcholinesterase), which is based on reactivation of phosphorylated enzyme with fluoride ions. The OP-inhibited butyrylcholinesterase in human plasma has a half-life of 5-16 days. Another method for detecting exposure to nerve agents involves electrospray tandem mass spectrometric analysis of phosphylated nonapeptides obtained after pepsin digestion of butyrycholinesterase from human serum samples (Fidder et al. 2002).

The DOD research efforts to develop or identify novel protein/genomic markers for CWA exposure are appropriate and might provide some information in identifying exposed people, which would help determine the treatment or management of the exposed personnel. Studies to date, however, suggest that it will be difficult to obtain results that are relevant and useful to DOD. This task would have more utility for HD.

The goal of this DOD research task is to develop antibodies that can recognize DNA or keratin adducts with HD as potential indicators of exposure. HD binds to proteins and DNA to produce adducts that are immunologically distinct. DOD will also explore the utility of proteases and cytokines produced in human keratinocytes in response to exposure to HD as alternatives to existing diagnostic methods. According to DOD, biomarkers of HD exposure identified by those studies might prove to be better diagnostic tools than existing tests of metabolic products of HD.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIIB3 HD is a bifunctional alkylating agent that reacts rapidly with nucleophiles. Its metabolites are excreted in urine within 24 hours, yet a significant portion of the absorbed dose persists in the blood for weeks to months, depending on the species. The use of urinary metabolites as biomarkers offers the advantage of a noninvasive test, but its value is limited to very recent exposure, because the urinary metabolites are detectable for only hours to days after exposure. This limits their use for retrospective detection. HD forms many protein adducts by alkylation. In a group of victims exposed to HD, the amino-terminal valine adducts corresponded with approximately 0.9 micromoles (µmol) of HD. These findings were confirmed by immunochemical analysis of DNA adducts in lymphocytes from the same blood samples (reviewed by Noort et al.

2002). A sensitive method to detect protein adducts has been developed that uses liquid chromatography (Noort et al. 1997; Black et al. 1997). Recently, Van der Schans et al. (2002) showed that most of the radioactivity (80%) of radioactive HD was bound to keratin. Van der Schaus et al. (2002) developed a direct detection method for these adducts in stratum corneum of human skin that used immunofluorescence microscopy. This method has the potential to lead to the development of a rapid detection kit that can be applied to the skin (Noort et al. 2002).

The committee recommends that biomarkers for routes other than cutaneous also be developed because there could be noncutaneous exposures.

The goal of this DOD research task is to determine the diagnostic utility of changes in nerve and muscle activity that previously have been shown to correlate with exposure to a CWA in rodents. If those previous findings can be replicated and if the effects are adequately robust, then their use will be considered for population-based studies. According to DOD, such studies will indicate whether changes in EEG, nerve CV, SFEMG, or MRI can serve as reliable diagnostic markers of CWA exposures.

Committee’s Evaluation, Conclusions, or Recommendations for Task IIIB4 The noninvasive methods listed above for evaluating potential effects of CWAs may have general utility under some conditions. For example, MRI provides a powerful tool for evaluating morphological changes in tissues (e.g., brain) in the same individual over time. Such measurements may be particularly important in evaluating persistent, delayed responses to CWA exposures. However, because the major goal of DOD research is to determine exposure levels that could lead to operationally relevant performance decrements and delayed health effects and the exposures to be modeled are for low doses, which are likely to produce subtle changes at most, those methods might not provide information of highest priority for the objectives of this research program. The committee recommends that, with the limited resources and time available for completion of the proposed research, less emphasis be placed on continuing the research proposed in this task. If this research is pursued, it should be done in nonhuman primates before suggesting it

would be useful for humans. These tests are also difficult to adapt to field conditions.

The DOD Research Plan identifies research needs that include experiments likely to provide data to contribute to the understanding of concentration-time relationships, the identification of biomarkers of exposure, the indicators of susceptibility and variability of response, improved capabilities and methods for assessing the potential to cause delayed adverse health effects, the description of mechanisms of toxicity associated with mixed and combined exposures, relationships between CWAs and other factors, and hazard assessment models. DOD presentations to the committee indicated that the current emphasis is on single exposures to single agents and the immediate effects on operational performance. For DOD, an important task is also to identify the highest level of CWA to which an unprotected person can be intermittently or continuously exposed without immediate or delayed health effects when exposures range from 1 hour to 1 year, focusing on exposures from acute single exposures to those repeated over 2 weeks. The primary objectives of DOD is the concern for the degradation of performance of military personnel from exposures to CWAs, and that concern makes the research on immediate effects more important.

The committee recognizes that a considerable amount of research has been done and much information is available on the acute (short-term, high-level exposures) and subchronic toxicity of nerve agents and HD. Genetic testing, neurotoxicity testing, metabolic studies, and other research studies have been done. The committee recommends that those studies not be repeated. The committee recommends that the Research Plan not attempt to fill in all the data gaps—that is, not investigate every species by various routes at multiple doses. The time, money, and effort could be better used in focusing on the most important and promising animal models and toxicity end points. The operational relevance of the research in terms of duration of exposure and CWA concentrations must be considered in establishing research priorities. The committee makes these recommendations because these studies should have operational relevance and the current interest is for consistent, sensitive detection of adverse responses after low-dose, short-term exposures to CWAs.

The committee recommends that the Research Plan include the development and application of appropriate statistical models for the data being generated. Specifically, statistical models should include concentration of the agent and duration of exposure as predictor variables along with important covariates that would allow for testing various extrapolation methods (e.g., Haber’s law, ten Berge’s law). Studies should include sufficient sample size at each tested exposure concentration and duration to detect changes that are judged to be biologically significant. To adequately determine the appropriate size of a planned study, statistical principles of design should be used. The following issues must be considered when determining animal numbers in an experimental study: (1) the magnitude of a difference that is considered significant (e.g., amount of pupillary constriction), (2) the magnitude of variation in response that is expected, and (3) the sensitivity/power that is desired for detecting this difference. Exposure concentrations, durations, and routes of exposure should be selected to include scenarios that are as realistic as possible.

To obtain the information most valuable in detecting and protecting military personnel from operationally significant decrements or potential delayed adverse health effects after short-term exposures to low levels of CWAs, DOD should ensure that the total database from previous human and animal studies has been fully examined to fill data gaps. These include specific studies done with human subjects (NRC 1982, 1985), studies in primates (Hartgraves and Murphy 1992), toxicokinetic studies (Benschop and De Jong 2001), and the studies used to derive AEGL-1 values (NRC 2003). Then, studies should be designed to correlate the most likely operationally significant decrement (e.g., miosis under different ambient light conditions after topical and inhalation exposures) with subtle behavioral change(s). When possible, actual tasks of military importance (e.g., visual tracking on radar screens) should be evaluated after topical OP exposure. These studies would best be done in humans (by using FDA approved therapeutic OP agents) or nonhuman primates, although the value of appropriately designed, efficient preliminary experiments in lower mammalian species (e.g., rodents and minipigs) is recognized. Wherever possible, associations also should be made with cholinesterase inhibition and/or reactivation or tissue binding of CWAs, especially in target tissues. Although unlikely to provide immediate results to assist with detection of and protection from CWAs in the field, other studies proposed in the DOD Research Plan are likely to have future value (e.g., toxicogenomics, in vitro cellular and biochemical ef-

fects). The committee recommends that some research with potential long-term rewards be continued even in the absence of immediate applications. Such experiments may provide insight into delayed adverse health effects. That research includes studies on the binding of CWAs to and reactivation from target and nontarget tissues, toxicogenomics, and biochemical effects noted at concentrations less than the MTD.

The committee concludes that DOD’s Master Research Plan on low-level exposures to CWAs, in general, is well planned and many of the proposed research tasks are likely to provide valuable information to DOD in protecting military personnel from low-level exposures to CWAs, in terms of avoiding performance decrements and delayed health effects. The Research Plan includes some studies that have some potential to identify delayed adverse health effects, but those studies should be assigned lower priority in the context of DOD’s primary objectives. Available information to date does not, however, provide a sound basis for anticipation of delayed adverse health effects following low-level (in particular, short-term) exposures to nerve agents. However, the committee recommends that a small proportion of the DOD research budget be allocated to some research tasks to rule out the possibility of delayed health effects.