Environmental Neurotoxicology (1992)

The human nervous system coordinates behavior; in perceiving and responding to external stimuli, it is responsible for mediating communication with the external environment; and it coordinates the activities of all other organ systems and thus plays an essential role in maintaining metabolic balance. The consequences of damage to the nervous system can be profound. Massive injury can result in coma, convulsions, paralysis, dementia, and incoordination. Even slight nervous system damage can impair reasoning ability, cause loss of memory, disturb communication, interfere with motor function, and impair health indirectly by reducing functions, such as attention and alertness, that ensure safety in the performance of daily activities.

Despite the nervous system's compensatory and adaptive mechanisms, many kinds of injury to the nervous system are irreversible, because, after initial development, new nerve cells are not formed; resulting losses of function can be permanent, as well as debilitating. Effects that are difficult to detect in an individual, such as a 5-point decline in intelligence quotient (IQ), are of great concern if they occur throughout large portions of a population. Prevention of damage to the nervous system is a major objective of social policy, medicine, and public health.

"Neurotoxicity" is the capacity of chemical, biologic, or physical agents to cause adverse functional or structural change in the nervous system. We use the term "environmental neurotoxicity" to refer broadly to adverse neural responses to exposures to all external, extragenetic factors (e.g., occupational exposures, lifestyle factors, and exposures to pharmaceuticals, foods, and radiation); it does not refer merely to the toxic effects of chemicals that are present in the environment as contaminants of air, water, and soil. Table 1-1 lists compounds for which there is evidence of neurotoxicity.

Possible effects of chemical toxicants on the nervous system are varied. For example, Table 1-2 lists neurobehavioral symptoms that—according to clinical reports, epidemiologic investigations, and experimental studies—are caused in humans or animals by at least 25 chemicals (Anger and Johnson, 1985; Anger, 1986). Neurotoxicity can occur at any time in the life cycle, from gestation through senescence, and its manifestations can change with age. The developing nervous system appears to be particularly vul-

nerable to some kinds of damage (Cushner, 1981; Blair et al., 1984; Pearson and Dietrich, 1985; Annau and Eccles, 1986; Hill and Tennyson, 1986; Silbergeld, 1986), but the results of some early injuries may become evident only as the nervous system matures and ages (Rodier et al., 1975).

The observation that some neurologic and psychiatric disorders are of environmental origin is not new (see Table 1-3). Initially, only striking manifestations of neurotoxicity were recognized; e.g., the association of coma, convulsion, and colic with high-dose exposure to lead has been recognized at least since Roman times (Waldron, 1973; Nriagu, 1978). Other long-recognized associations include an increased frequency of depression and suicide among workers in contact with carbon disulfide (Vigliani, 1954); spastic paraparesis (lathyrism) due to

ingestion of Lathyrus species during famines (Denny-Brown, 1947); paralysis following consumption of products contaminated with tri-o-cresylphosphate (TOCP), such as the patent medicine Ginger Jake in the United States (Smith et al., 1930) and cooking oil in Morocco (Smith and Spalding, 1959); toxic psychoses in people who inhale tetraethyl lead (Cassells and Dodds, 1946); and erethism (a syndrome with such neurologic features as tremor and such behavioral symptoms as anxiety, irritability, and pathologic shyness) in people exposed to elemental mercury (Bidstrup, 1964).

Numerous other associations between neurologic impairment and environmental exposures have been noted more recently: Exposure to low concentrations of environmental lead is correlated with reduced scores on tests of mental development (Bellinger et al., 1987a,b; Needleman, 1989); the pesticide Kepone (chlordecone) is associated with nervousness, tremors, and other signs of nervous system dysfunction (Cannon et al., 1978); 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), a byproduct of illicit synthetic-heroin production, causes an irreversible syndrome that closely resembles parkinsonism (Langston et al., 1983); and exposure to manganese can produce parkinsonism and dyskinesia (Cook et al., 1974). In the realm of developmental neurotoxicity, early gestational exposure to ionizing radiation is associated with microcephaly and mental retardation (Otake and Schull, 1984); infants born to women who use cocaine during pregnancy display significant depression of nervous system response to environmental stimuli and other congenital anomalies (Shepard, 1989); heavy ingestion of alcohol by pregnant women produces a syndrome of craniofacial abnormalities and mental retardation in their children (Jones and Smith, 1973), and it is possible that there are persistent adverse effects on mental and motor development in children whose mothers drink more moderately during pregnancy and lactation (Little et al., 1989; Streissguth et al., 1989; West, 1990).

It has been suggested that exposures to environmental chemicals can contribute to clinical neurodegenerative disorders seen later in life. Calne et al. (1986) have hypothesized that various environmental agents contribute to Alzheimer's disease, Parkinson's disease, or amyotrophic lateral sclerosis (ALS) by depleting neuronal reserves to an extent that becomes observable in the context of natural aging. For example, a syndrome combining the symptoms of ALS and of parkinsonian dementia is prevalent among people from Guam and has been postulated to result from earlier ingestion of one or more chemicals found in the seed of the false sago palm, or cycad (Cycas circinalis L.) (Kurland, 1963; Spencer et al., 1987). Some epidemiologic studies have found evidence that various environmental factors are involved in the etiology of ALS, Parkinson's disease, and Alzheimer's syndrome (e.g., Barbeau et al., 1987; Freed and Kandel, 1988; Lilienfeld et al., 1989; Yokel et al., 1988), but other investigations have not detected such relationships (e.g., French et al., 1985; Shalat et al., 1988). The extent to which environmental neurotoxicants contribute to chronic neurologic and psychiatric disease is not yet known.

The nervous system is composed of cells of several types, each with its own functions and characteristic vulnerabilities. Several unique features of the nervous system influence its reactions to toxic agents, including its poor regenerative capacity, its unusual anatomy (especially, long axons), its multiple functions, the extensive interconnections among its cells, its dependence on glucose as an energy source, the existence of highly specialized cellular subsystems, and the wide variety of highly localized neurotransmitter and neuromodulation systems. In addition, environmental toxicants can disturb the complex interactions between the nervous system and other organs. (Chapter 2 discusses the

structure and function of the nervous system In some detail, including consideration of the complex neural arrangements involved in information processing and storage.)

The accessibility of a particular part of the nervous system to a specific chemical is a function of both the tissue and the chemical. Many chemicals are kept from entering the brain by the ''blood-brain barrier''—the tight junctions formed by endothelial cells surrounding capillaries that supply brain tissue and by endothelial cell-astrocyte interactions. That barrier can, however, be crossed by lipophilic substances and has a series of specific transport mechanisms through which required nutrients, hormones, amino acids, peptides, proteins, fatty acids, etc., reach the brain (Pardridge, 1988). Thus, toxicants can gain entry to the brain if they are lipid-soluble or if they structurally resemble substances that are normally taken up by the nervous system. Moreover, the blood-brain barrier might be less effective in immature than in mature organisms (Klaassen, 1986). It is absent in some parts of the brain, such as in the circumventricular area, and around the olfactory nerve, which runs directly from the nose to the frontal cortex (Broadwell, 1989).

Some cells in the nervous system cannot reproduce themselves; once damaged, they cannot be replaced. Many of those cells ordinarily are present in excess, so there is a buffer against damage, and substantial loss need not affect function or behavior. However, the degree of redundancy, particularly in specific regions, is not known and might vary with age. The presence of such excess can result in a threshold or nonlinear dose-response relationship. Also, the interaction between exposures to neurotoxicants and age-related cell loss might explain delays in the manifestations of toxic consequences. If several processes (e.g., chemical damage, normal aging, and death of cells) are proceeding simultaneously, it might be impossible to isolate a single cause of functional impairment. A long latent period before toxicity is manifested makes it difficult to associate an exposure and a response causally and even harder to derive a quantitative dose-response relationship.

Manifestations of neurotoxic response can be progressive, with small functional deficits becoming more serious. In those cases, it is almost impossible to define the onset of impairment. For many biologic variables that respond to exogenous agents, the demarcation between an unimportant change and a health-damaging change is unclear. A small change might be a marker of exposure, a moderate change might signal preclinical disease, and a large change might indicate advanced disease.

Some materials, particularly pharmaceuticals, produce different responses in the nervous system at different doses or have adverse side effects at therapeutic doses. For example, the tricyclic antidepressants have a desired therapeutic activity at a low dose, but produce life-threatening anticholinergic effects at higher doses; the antineoplastic drug cis-platinum is a valuable chemotherapeutic agent, but can cause toxic neuropathies; antipsychotic drugs can produce disabling movement disorders; and some antibacterial agents can trigger loss of hearing and balance (Sterman and Schaumburg, 1980). Some substances valued for their relatively selective neuroactivity, such as ethanol, are particularly likely to have simultaneous neurotoxicity (Goldstein and Kalant, 1990).

Exposure to combinations of chemicals can produce interactive effects. Examples include exacerbated hearing loss in persons exposed to some antibiotics (Bhattacharyya and Dayal, 1984; Lim, 1986; Boettcher et al., 1987) and cumulative toxic effects of occupational and environmental exposures to mixtures of solvents (Cranmer and Goldberg, 1986). The general population is exposed to chemicals with neurotoxic properties in foods, cosmetics, household products, air, water, and drugs used therapeutically or

recreationally. Exposure to naturally occurring neurotoxins, such as some fish and plant toxins, is yet another aspect of the problem. The multitude of voluntary and unintentional exposures to neuroactive substances that characterize the daily life of ordinary people complicates the task of identifying undesirable neurologic outcomes and attributing them to their proper causes.

The number of people with neurotoxic disorders and the extent of neurologic disease and dysfunction that result from exposure to toxic chemicals in the environment are not known. However, it is known that an enormous number of people are exposed to environmental materials that in sufficient doses could pose a neurotoxic hazard. The National Institute for Occupational Safety and Health (NIOSH, 1977) estimated that in 1972–1974 there were 197 chemicals to which a million or more American workers were exposed in the occupational setting for all or part of each working day. Anger (1986), in a review of secondary sources in the research literature, found that more than one-third of those 197 chemicals had demonstrated the potential, if the doses were large enough, to produce adverse effects on the nervous system.

The data needed to estimate the overall magnitude of the problem of environmental neurotoxicity do not exist. A National Research Council committee found a few years ago that few of the 60,000–70,000 chemicals in commercial use had been tested for neurotoxicity (NRC, 1984); no information was available on any aspect of the toxicity of approximately 80% of the chemical substances in commercial use. Even among the most extensively regulated classes of chemicals (pesticides, drugs, and food additives), the information needed for a thorough health-hazard assessment was available on only about 5–18%. Few compounds have been assessed for selective toxicity to vulnerable groups within the population, such as the very young and very old. Moreover, there is little information on the nature and extent of human exposure to even the materials that have been tested and identified as neurotoxic. Finally, not all neurotoxic effects are equivalently damaging or debilitating; scales of relative injury are needed, if the effects of exposures to different materials are to be evaluated in comparable terms.

Given the overall low proportion of substances tested for any type of toxicity and the even smaller fraction that have been assessed for possible neurotoxicity, it is difficult to estimate how many commercial and industrial chemicals are neurotoxic. However, there is evidence that many toxic chemicals in wide use have neurotoxic potential. Of 91 criteria documents produced by NIOSH, 36 (40%) cite neurotoxicity as a reason for recommending limits on occupational exposure (Anger, 1989). Likewise, of 588 chemicals listed by the American Conference of Governmental Industrial Hygienists (ACGIH) in 1982 as both widely used in industry and having toxicologic significance, 167 (28%) had neurologic effects as one basis for recommendations on maximal exposure concentrations (Anger, 1984). However, O'Donoghue (1986) calculated, on the basis of tests of a small, unselected sample of chemicals, that 5% of all industrial chemicals, excluding pesticides, are likely to be neurotoxic. Diener (1987) noted that O'Donoghue's calculation could be an underestimate and should not be considered a firm basis for extrapolation of risk until broader and more systematic surveys of the neurotoxicity of chemicals in commercial use have been undertaken.

In addition to unwanted contact with pollutants or contaminants and inadvertent exposures to industrial products, voluntary exposures to many legal and illegal materials can cause neurotoxic effects (OTA, 1984). Overuse of the legal stimulant caffeine can lead to tremors, irritability, and sleeplessness. Exposure to the depressant ethanol is

responsible for many accidents on the highway and elsewhere, fetal injuries, chronic disease, and antisocial behavior. The devastating personal and societal consequences of "substance abuse," engaged in at least initially for psychoactive effects, have become obvious. The list of self-administered neuroactive substances is extensive, and exposure to them makes epidemiologic identification of additional neurotoxicants difficult.

The Health Care Financing Administration of the U.S. Department of Health and Human Services has estimated that $23 billion was spent in 1980 for the care of people with diagnosed neurologic diseases; in many cases, these illnesses are due to accidental exposures to neurotoxicants or to use of legal drugs, rather than to use of illegal substances (Rice et al., 1985). As has been suggested regarding lead toxicity (Provenzano, 1980; EPA, 1985), it would not be surprising if the direct and indirect societal costs of subclinical neurologic losses or deficits—such as reduction in intelligence, diminution in achievement, and waste of opportunity—might equal or exceed those of neurotoxic effects that are clinically recognized. Modifications of behavior induced by neurotoxic substances can also adversely affect people other than those directly experiencing the toxic effects, as is the case for victims of drunken drivers.

Prevention is the key to dealing with neurologic diseases of toxic environmental origin. Such diseases result from exposures to synthetic and naturally occurring chemical toxicants encountered in the ambient environment, ingested in foods, or administered as pharmaceutical agents. Many of the diseases are not curable, so they must be prevented. They can be prevented by eliminating or reducing exposures at the source (primary prevention), or they can be controlled by early detection and diagnosis of neurotoxic effects while they are still reversible or while they are at a relatively early stage of evolution when their progression can still be halted (secondary prevention).

In light of the difficulty of regulating potentially neurotoxic compounds found in illicit substances of abuse or in certain foods, particularly in plants, the most effective approach to preventing neurotoxic disease of environmental origin consists of identification of neurotoxicity through routine pre-market testing of all new chemicals before they are released and before human exposure has occurred. Disease is prevented by restricting or banning the use of chemicals found to be neurotoxic or by instituting engineering controls and imposing protective devices at points of environmental release and potential human exposure.

Given the current absence of data on neurotoxicity of most chemicals in commerce, an extensive program of primary prevention through toxicologic evaluation is needed. The Environmental Protection Agency (EPA) has regulatory authority to screen chemicals coming into commerce, but some have expressed the opinion that the authority is insufficiently exercised and inadequately supported. Therefore, people might be overexposed to environmental neurotoxicants and might develop neurotoxic responses to chemicals that slip through the regulatory net. For example, important side effects of pharmaceutical agents continue to surface, such as the 3 million cases of tardive dyskinesia in patients on chronic regimens of antipsychotic drugs. Prudence dictates that all chemical substances—both those already on the market and new ones—be considered potential neurotoxicants; a chemical cannot be regarded as free of neurotoxicity in the absence of data on its toxicity. Prudent public policy therefore dictates that all chemicals, both old and new, be subjected to at least basic screening for neurotoxicity when use and exposure warrant.

Neurotoxicants can be identified by several means, ranging from clinical case reports and observations by alert physicians to for-

mal in vivo and in vitro screening. Analysis of structure-activity relationships (SARs) has been used extensively to predict neurotoxicity, but SARs as currently used can provide only minimal guidance for hazard identification. EPA, however, uses SARs as a basis for decisions on testing under the Toxic Substances Control Act. Testing procedures designed for in vivo neurotoxicologic evaluation of new chemicals have been developed and are reasonably effective, but they are tedious and labor-intensive. Additional mechanistically based in vitro test systems are badly needed, and they must be correlated with in vivo test systems. In summary, a rational and efficient strategy for neurotoxicity testing is needed to close the gap in neurotoxicity testing. Development of a blueprint for such a strategy was a major objective of the study reported here.

The Committee on Neurotoxicology and Models for Assessing Risk was formed at the request of the Agency for Toxic Substances and Disease Registry to review the biologic principles and mechanisms of neurotoxic action for the purpose of using them in risk assessment. This report deals with environmental neurotoxicology—with the current state of the field, its implications for human health, and the prospects for its rational future development. The committee first reviewed current knowledge of environmental neurotoxicity, paying special attention to the types of neurologic injury caused by chemicals in the environment, to determine the overall extent of the problem of environmental neurotoxicity. We investigated ways in which improved biologic markers of neurotoxicity (markers of subtle and subclinical effects, of exposure to neurotoxicants, and of susceptibility to their effects) could be developed, could improve laboratory tests, and could facilitate the early recognition of neurotoxic injury in exposed human populations. We then considered available methods of testing for neurotoxicity and evaluated strategies for recognizing and characterizing the neurotoxic potential of chemicals and for detecting neurotoxic effects among environmentally or occupationally exposed humans. Finally, we examined approaches to neurotoxicologic risk assessment.

This chapter has provided an overview of the problem of environmental neurotoxicity, described the level of existing knowledge, and suggested broad issues for scientific research. Chapter 2 illustrates the biologic basis of neurotoxicity by reviewing the aspects of the structure and function of the nervous system that are subject to disruption by exogenous agents. Chapter 3 discusses biologic markers, which provide a conceptual thread connecting human, animal, and in vitro studies. Chapter 4 considers approaches to identifying neurotoxicants and describes available in vitro and in vivo tests that could be used to assess toxicity before human exposure has occurred. It develops a blueprint for a multitiered program to test chemical substances for neurotoxicity, and it lays out a plan for systematic evaluation of the results of neurotoxicity testing. Chapter 5 describes approaches to the epidemiologic surveillance of human populations at risk of environmental neurotoxicity and outlines the uses to which human data can be put in bringing neurotoxic exposures under control. And Chapter 6 describes current approaches to risk assessment and discusses modifications in the risk-assessment paradigm that will be needed to accommodate neurotoxic end points. And Chapter 7 presents the committee's conclusions and recommendations.