Environmental Neurotoxicology (1992)

The recognition that exposure to chemicals can cause neurologic injury evolved from studies of acute illnesses in people exposed to high doses of environmental toxicants. These illnesses included encephalopathy in children who ate chips of lead-based paint; blindness in persons who consumed wood alcohol (methanol); and coma, convulsions, and respiratory paralysis after exposure to organophosphorus pesticides. Epidemics of neurotoxic diseases related to environmental exposures have occurred: blindness and ataxia caused by organic mercury in fish from Minamata Bay, Japan, and in fungicide-treated grain in Iraq; spinal-cord degeneration and peripheral neuropathy caused by tri-o-cresylphosphate (TOCP) in cooking oil in Morocco; and tremors, motor disturbance, and anxiety caused by the pesticide Kepone (chlordecone) in Hopewell, Virginia. In all, these epidemics affected thousands of people and established clearly that toxic chemicals in the environment ¹ can cause neurologic and psychiatric illnesses.

In response to a request from the Agency for Toxic Substances and Disease Registry, the National Research Council convened the Committee on Neurotoxicology and Models for Assessing Risk in 1988. The committee, charged to review the biologic principles and mechanisms of neurotoxic action relevant to risk assessment, produced this report, which discusses the magnitude of the problem of neurotoxic effects, testing strategies, surveillance efforts, biologic markers, and risk assessment.

Neurotoxicity caused by environmental toxicants results in a range of neurologic and

psychiatric disorders. Concern over the potential neurotoxic effects of chemical substances is greatest for agents that cause irreversible or progressive changes. In addition to immediate and progressively developing effects, there is increasing evidence that neurotoxic effects can occur after long latent periods. It is postulated that intervals as long as many decades can elapse between exposure to a chemical and the appearance of neurologic illness.

A major unanswered question—indeed, a central issue confronting neurotoxicology today—is whether the causal associations observed in epidemics of neurotoxic diseases reflect isolated events or are merely the most obvious examples of a widespread association between environmental chemicals and nervous system impairment. Concern about subclinical neurotoxicity has brought this issue to its current prominence. Subclinical toxicity refers to exposure-induced adverse effects that are too small to produce signs and symptoms evident in a standard clinical examination. Subclinical neurotoxic effects can include alteration of a wide spectrum of behaviors. Environmental chemicals associated with subclinical neurotoxicity include lead, organophosphorus pesticides, some chlorinated hydrocarbons, some solvent mixtures, and mercury. Those are chemicals to which many thousands of Americans are regularly exposed at work and to which even more are exposed in smaller doses in the general environment. Although often subtle, subclinical neurotoxic effects are not necessarily inconsequential; moreover, even subtle alterations can be irreversible. It has been hypothesized that an undefined fraction of chronic neurologic and psychiatric illness in the human population can be exacerbated or even caused by chronic, low-level exposure to environmental neurotoxicants.

The committee recommends that more accurate estimates of the extent of the problem of neurologic and psychiatric dysfunction attributable to chemical agents in the environment be developed.

The estimates must be based on a combination of clinical, epidemiologic, and toxicologic studies coupled with the techniques of quantitative risk assessment.

The general reason to test substances for neurotoxicity is to identify neurotoxic potential before the occurrence of human exposure. The ultimate goal is to prevent human disease. About 70,000 chemicals are used in commerce, of which several hundred are known to be neurotoxicants. However, except for pharmaceuticals, less than 10% of the chemicals in commerce have been tested at all for neurotoxicity, and only a handful have been evaluated thoroughly. Furthermore, resources are not readily available to undertake across-the-board testing of all chemical substances already in commerce.

New strategies for neurotoxicologic assessment of environmental chemicals will need to include establishing testing priorities among chemicals for hazard identification (with emphases on new chemicals, chemicals considered likely to be hazardous, and chemicals to which large numbers of people are exposed). The new strategies will also require refining existing neurotoxicity test systems and developing sensitive, new testing systems. To contain the labor and resource requirements of the testing strategy, developing quicker and more economical approaches than are currently used must be encouraged, particularly for screening for potential neurotoxicants.

A new strategy for neurotoxicologic assessment will build on and extend currently available test systems. It will have a "tiered" structure—decisions to test chemicals at the higher tiers, as well as decisions concerning types of testing, will be guided by data from the initial, or screening, tier. The first tier, or screen, will be used for hazard identification. The results of the screen and a chemical's exposure pattern would guide further characterization of dose-response relation-

ships (second tier) and mechanisms (third tier).

No existing validated system satisfies all the necessary requirements for a screening program to detect the neurotoxic potential of chemicals. The range of such a program should extend to the detection of neurodevelopmental effects and effects on cognitive function and of neuroendocrine effects. No comprehensive effort has been made to determine the predictive ability of individual screening tests by examining the relationship between test results and data from long-term studies in animals or epidemiologic and clinical studies in humans.

The screening tier will consist of a set of tests to measure multiple end points—including chemical, structural, and functional changes—and a functional observational battery. Determination of the component tests of the first tier is the most crucial aspect of the three-tiered approach, because truly positive substances will not continue to later tiers unless detected at this point. The first tier is the heart of the screening aspect of neurotoxicity testing; the later tiers might produce data of great value in advancing understanding of neurotoxic processes and how the nervous system operates, but the first tier is the first line in preventing neurotoxic disease.

Much of the controversy over proper testing procedures arises from two intrinsically conflicting objectives—minimizing the incidence of false positives (substances incorrectly identified as hazardous) and minimizing the incidence of false negatives (substances incorrectly identified as nonhazardous). Too high an incidence of false positives wastes resources. However, a high incidence of false negatives is potentially dangerous. For the prevention of human disease, testing systems must be highly sensitive. Testing in the second and third tiers of future test systems should expose false positives; the abandonment of an occasional new chemical on the basis of what are actually false-positive screening results is the likely cost of this process.

The committee recommends that a rational, cost-effective neurotoxicity testing strategy be adopted.

It should allow an accurate and efficient progression from the results of hazard-identification studies (screening) to the selection and application of appropriate test methods for defining mechanisms of toxicity and for quantitative characterization of neurotoxic hazards.

A coherent testing strategy would incorporate a series of checks and balances. The results of the later phases of testing would provide the data necessary to evaluate and validate initial screening batteries and thereby help to identify tests that should be excluded from or incorporated into an efficient battery. Information generated by such a strategy would reveal which types of data are most useful for accurate, quantitative prediction of the risks to humans associated with exposure to similar chemical compounds.

Tests of neurotoxicity on chronically exposed animals might be carried out in conjunction with tests of other chronic effects. Such testing might identify toxic effects of substances not previously known to be neurotoxic. There is a particular lack of data on chronic and long-latency neurotoxic effects.

The committee recommends that for reasons of efficiency, integrative studies combining a variety of end points be explored in the development of the neurotoxicity testing strategy.

To maximize detection of toxicity, some toxicity studies encompassing the full life span of experimental animals should be encouraged.

Data are needed on the influence of dose, route of exposure, toxicokinetics, metabolism, and elimination on the effects of a given neurotoxicant, and data are needed on the existence of interspecies differences. More complete understanding of neurotoxic disease at a molecular level should also improve the ability to evaluate new chemicals

on the basis of structure-activity relationships. Existing in vitro test methods should be exploited more efficiently than at present to identify and analyze the mechanisms of neurotoxic action at cellular levels. Development of a mechanistic understanding of neurotoxicity might facilitate the discovery of biologic markers of exposure to toxicants, as well as markers of early, subclinical neurotoxic effects.

The committee recommends that studies to define mechanisms of neurotoxicity in as much detail as possible be encouraged, as well as studies to identify hazards.

A program needs to be undertaken to test the relationship between in vitro and in vivo findings and between animal and human results for a set of well-defined substances.

Epidemiologic and clinical studies of populations exposed to potentially neurotoxic chemicals are needed to provide additional information on the human neurotoxic effects of environmental chemicals and to complement in vitro and animal screening studies. High-risk populations must be identified and monitored. Public-health surveillance systems to identify people who are potentially exposed to environmental neurotoxicants are not well developed, and there is little information on the background incidence and prevalence of the major neurologic diseases in the American population. People with diagnosed neurologic illnesses must be studied to identify possible environmental etiologies and to complement and extend the knowledge gained through in vivo and in vitro laboratory investigations.

Recognition of the neurotoxic effects of exposure to environmental chemicals through epidemiologic and clinical studies is made difficult by the enormous variety of the possible reactions of the nervous system to toxic insult. The changes are often subtle and subclinical, and months or years can elapse between exposure to a neurotoxicant and the appearance of dysfunction or disease. Few attempts have been made to explore the possible relationships between chemical exposures and chronic or progressive neurologic and behavioral disorders. Populations known to be exposed to potential neurotoxicants should be followed for long periods in prospective studies, and retrospective studies of people with neurologic illness must consider the possibility that exposures occurred many years previously. Epidemiologic studies will increasingly need to use biologic markers of exposure, of effects, and of susceptibility.

The committee recommends that exposure-surveillance systems cover a much broader range of chemicals and use improved monitoring techniques for long-term assessment.

Existing disease-surveillance systems, such as those of the Social Security Administration, the Department of Veterans Affairs, and the National Center for Health Statistics, should be modified to provide more useful data on the incidence and prevalence of chronic neurologic and psychologic disorders, some of which are likely to be of occupational and environmental origin. A broader range of neurologic disease end points should be covered by surveillance programs. Anecdotal reports of neurotoxicity in humans need to be pursued vigorously with clinical surveillance and followup. The incorporation in surveillance systems of the concept of sentinel health events (SHEs) specifically for neurotoxic illnesses should be encouraged.

Recognition of the possible environmental origin of neurologic and psychiatric disease is hampered by the inadequate training of most physicians and other health providers

in occupational and environmental medicine. Greater uniformity in disease definition would improve identification of diseases of neurologic interest.

The committee recommends that improved disease reporting be supported by the dissemination of information on neurotoxic illnesses to physicians and other health professionals to increase their awareness of environmental neurotoxicity as a possible explanation of specific illnesses or sets of symptoms.

All physicians should be trained to take a thorough occupational-exposure history and to be aware of other possible sources of toxic exposure, such as hobbies and self-medication. Uniform clinical definitions of neurotoxic disorders are needed to provide a common basis for reporting by physicians. Standardized national reporting systems must be established for physicians to report outbreaks of suspected environmentally and occupationally caused neurologic and psychiatric disorders.

As recently defined by a previous NRC committee, biologic markers are measures of changes or variations in biologic systems or samples. It is useful to classify biologic markers into three types: markers of exposure, of effect, and of susceptibility. A biologic marker of exposure in an organism is a measurable presence of an exogenous substance, its metabolite, or the product of its interaction with some target molecule or cell. A biologic marker of effect is a measurable biochemical, physiologic, or other alteration within an organism that, depending on magnitude, can indicate potential or established disease. A biologic marker of susceptibility is an indicator of an inherent or acquired variation in an organism's ability to respond to the challenge of exposure to a specific substance.

Biologic markers provide a means to explore the relationship between exposures and effects. By virtue of the incorporation of more relevant information about critical events and molecular mechanisms, biologic markers contribute to a more complete scientific understanding of toxic injury. Biologic markers can be used to improve sensitivity, specificity, and predictive value of detection and quantification of adverse effects at low dose and early in exposure. Neuroscience and neurotoxicology can be enhanced by directing more research attention to quantitative questions that information on biologic markers helps to answer.

The committee recommends that putative biologic markers in animal species be evaluated and validated in in vivo and in vitro systems.

Biologic markers should be regularly incorporated into epidemiologic and clinical studies of neurologic disease, particularly prospective studies. The primary goal of the incorporation of biologic markers into such studies should be to validate their predictive accuracy and to test hypothesized quantitative relationships between specific markers related to causal pathways involving neurotoxic outcomes.

Risk-assessment techniques provide a means for estimating the risks to humans associated with exposure to toxic chemicals in the environment. The estimation of the risks most often involves extrapolation from high experimental doses used in animal tests to lower environmental doses. Numerous assumptions are typically made to bridge gaps in the available scientific data. Most

risk-assessment procedures have focused on cancer as an end point, and techniques for assessing other types of risk are relatively undeveloped. Commonly used paradigms for risk assessment do not accurately or adequately model the risks associated with exposure to neurotoxicants. Virtually all neurotoxicologic risk assessment today is limited to qualitative hazard identification and to the early stages of hazard characterization; neither sufficient data nor adequate paradigms are available to permit quantitative evaluation of most neurotoxic risks.

Risk-assessment techniques that incorporate more quantitative information about dose-time-response relationships and mechanisms of neurotoxicity are under development. Advances in this regard will improve not only the assessment of risk, but also assessment of the benefits to be gained by reducing human exposure to specific neurotoxic agents. The construction of new models for neurotoxicologic risk assessment depends on the acquisition of new knowledge of the fundamental mechanisms of action of chemical toxicants on the human nervous system. The molecular and subcellular mechanisms by which environmental neurotoxicants act need to be delineated. Such information will improve prediction and quantification of risks, including risks that become evident only long after exposure.

Neurotoxicologic risk assessment has been largely limited to the application of no-observed-effect levels and uncertainty factors, so it has not generated information on specific risks for given magnitudes of exposure. Experiments should include a range of doses that spans those relevant to expected human exposures. In addition to providing a firmer basis for estimating human risk, such designs permit tests of the assumption that results obtained at high doses predict the pattern of effects at low doses. Such variables as age, sex, duration of exposure, and route of exposure need to be more systematically evaluated. Species-specific effects need to be recognized and investigated. The usefulness of that kind of information for quantitative risk assessment would be greatly amplified by serial measures of both neurotoxic end points and biologic markers. A single model will not be adequate for all conditions of exposure, for all end points, or for all agents. It might be necessary to build risk-assessment models to deal simultaneously with several end points produced by a toxicant. Such models should incorporate biologic markers of neurologic dysfunction and be based on fundamental information on mechanisms derived from experimental test systems and epidemiologic data.

The committee recommends the development of risk-assessment methods that capture the complexities of the neurologic response, including dose-time-response relationships, multiple outcomes, and integrated organ systems.

Experimental designs for studies of neurotoxic agents should provide information needed in the risk-assessment process. Full exploration of relationships and development of risk assessment models would be facilitated by original researchers' making complete data sets available to other investigators.

For the future, one of the major challenges for neurotoxicology will be to use insight from clinical medicine, epidemiology, and toxicology to design effective systems for the prevention of neurotoxic disease in the American public.

Environmental Neurotoxicology