Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
TO Biologic Markers of Reproductive Development and Aging Toxicants and other environmental fac- tors can influence female reproductive function during development, as well as during later age-related changes. This chapter discusses biologic markers of female reproduction across the life span, including markers of neuroendocrine func- tion that are potentially pertinent to reproduction. Some of these markers might be used to assess effects of toxicants. Little information is available on cri- tical periods or ages of particular vul- nerability for effects of toxicants on postnatal development and aging. Some effects of toxicants could be equivalent to accelerated senescence. Cryptic damage-damage not immediately manifestedmight interact with other insults. All-or-none effects might not be associated immediately with a particu- lar exposure. Daughters exposed to DES in utero manifested an increased incidence of genital tract cancer as adults (Herbst et al., 1974~. Exposure of neonatal rats to DDT causes major impairments of female reproductive functions that emerge after , . . puberty (He~nr~chs et al., 1971~. Chronic exposure of mice to endogenous ovarian steroids secreted during young adult life is a cause of age-related estrous cycle lengthening (Felicio et al., 1986) or pre- mature loss of estrous cycles after the 169 estrogen treatment ends (Mobbs et al. 1984; Kohama et al., 1986~. These rodent phenomena have no human analogues. However, the likelihood of hot flushes at menopause when estrogen concentrations decrease appears to depend on exposure to ovarian steroids during puberty. Women with Turnerts syn- drome, whose estrogen concentrations are the same as postmenopausal concentra- tions, do not have hot flushes; however, withdrawal from estrogen treatment will cause hot flushes in these women (Yen, 1977~. That important result indicates that the adult human nervous system has memory mechanisms for exposure to steroids such that the effects can be manifested decades after the exposure (Finch et al., 1984~. In analyzing interactions of estro- genic toxicants with the nervous system, we should anticipate potential cryptic effects on the numerous neurons throughout the brain that contain receptors with high affinity for estrogens and other steroids. Many examples demonstrate cryptic brain damage during early adulthood. The relation between viral encephalitis (don Economo's disease) and parkinsonism is a classic case (Poskanzer and Schwab, 1961; Finch, 1976; Caine et al., 1986~. Parkinsonism induced by ingesting the neurotoxicant methyl~phenyl~tetrahydro-
170 pyridine (MPTP), which often contaminates synthetic heroin, might not occur immedi- ately; some persons without necrologic symptoms have depressed metabolisms in their dopaminergic systems, as determined by positron emission tomography analysis (Calneetal., 1985~. Those examples constitute ample pre- cedent for considering possible long- term adverse effects of toxicants on neuro- endocrine loci that might be incurred oc- cupationally (e.g., by metal workers). Most biologic markers of toxicant-re- lated effects on female reproductive func- tions are physiologic or morphologic. Among the best-characterized markers used in epidemiologic surveys and individual case studies are infertility and length of menstrual cycles, which appear to be sensitive to many of the same environmental influences in humans and rodent models. However, human neuroendocrine repro- ductive functions appear to differ in im- portant ways from those of rodents; the hypothalamus might be less crucial in regu- lating the preovulatory surge in humans than of rodents (Knobil, 1980~. Nonethe- less, human female reproductive functions clearly are susceptible to neurogenic influences from stress (Peyser et al., 1973) and perhaps from pheromones (Russell etal., 1980~. Transgenerational toxic effects can arise in several ways. The oocyte comple- ment of an adult female is attained in utero before the midpoint of gestation. Toxi- cants might reduce the number of oocytes and, if mutagenic, affect later genera- tions. Furthermore, toxicants in the ma- ternal environment that affect fetal brain development could influence the maternal physiology and behavior of female off- spring, which then affects the next genera- tion. For example, environmental influ- ences that extend to the F2 generation have been demonstrated in rats (Zamenhof et al., 1972~. Malnutrition during pregnancy reduces the number of brain cells in rats for at least two generations, despite crossfostering of pups with normal sur- rogate nurses (Zamenhof et al., 1972~. The concept of transgenerational envi- ronmental effects is well known to develop- FEMALE REPRODUCTIVE TOXICOLOGY mental biologists, but has not been dis- cussed widely as an aspect of toxicology. MARKERS OF MATURATION Adrenarche Increases in plasma DHEA-S (a metabolite of DHEA) precede increases in estrogens by several years in humans and indicate maturation of adrenal cortical function. The stimuli for increased secretions of DHEA and other adrenal steroids before puberty are poorly understood and do not seem to involve direct action of ACTH or gonadotropins. Although readily meas- ured by radioimmunoassays, the increase in the plasma metabolite DHEA-S is linked only circumstantially to functional changes in maturation. Ovary and Uterus Size changes in the uterus and ovary can be followed by ultrasonography _ . . . The data base is modest t~or ultrasound measurements of these morphologic changes, but such measurements can be obtained in conjunc- tion with other common clinical markers of puberty. ~rs~n~ et a ., 15~84 . Menarche Onset of menstrual bleeding is the most obvious sign of active ovarian steroid secretion. The onset is triggered by increases in E2 and progesterone, fol- lowed by decreases in progesterone within a few days. Ovulatory cycles usually are established months after menarche and vary considerably in timing and hormonal characteristics. Thelarche and Pubic Hair The ages of thelarche and the appearance of pubic hair are used widely to judge whether puberty is precocious or delayed, and an extensive data base is available. Five morphologic stages of breast develop- ment generally are accepted (Marshall and Tanner, 1969~. The five stages can vary extensively in duration and can revert
REPRODUCTIVE DEVELOPMENT AND AGING to earlier stages. Four pubic hair stages also show extensive individual var- iations. Breast and pubic hair stages often are asynchronous and by themselves do not precisely indicate the rate of development. MENSTRUATION Menstrual cycles constitute the most accessible and noninvasive biologic mark- er of female reproductive function in hu- mans. The cycle lengths can vary (average, 28 days); determination of amenorrhea or other acyclic conditions requires a daily menstrual record for at least some 3 months in humans and higher primates. Because the menstrual cycle can fluctuate as a result of nutrition, stress, use of oral contraceptives, and other influ- ences, detailed personal and medical histories must be collected from study subjects. Laboratory rodents do not have menstrual cycles, but have estrous cycles, and their cycle status usually can be established from cell changes in daily vaginal smears. Acyclicity is defined as 14-30 days without evidence of ovulation in a proestrus smear. The onset of acyclicity can be stud- ied as part of the aging processes or as a response to experimental intervention. Strings of cornified smears for 14 days or more indicate a polyfollicular anovula- tory condition (persistent vaginal cor- nification), which is the most common ini- tial acyclic state during aging (Finch et al., 1984~. Similar vaginal smear pat- terns after acute or chronic exposure of young rodents to suspected environmental toxicants indicate severe disruption of the reproductive neuroendocrine system. Whether acyclicity is reversible depends on the agent, age at exposure, and duration of exposure. Permanent acyclicity has been induced in rats by DDT (Heinrichs et al., 1971~. Menstrual or estrous cycles can be analyzed for cycle length distributions and for lengths of consecutive cycles. The most detailed analyses have been done on rodents during aging. The frequency distribution of estrous-cycle lengths (e.g., 4-day, 5-day, or 6-day cycles) ..` 171 and the frequency of length transitions (e.g., 4-day to 4-day or 4-day to 5-day transitions) are sensitive indicators of maturation and aging as well as for ef- fects of estrogen toxicity in mice (Fig. 13-1 ) (Nelson et al., 1982~. Available longitudinal menstrual rec- ords should be analyzed in more detail (see Treloar et al., 1970~. Many powerful sta- tistical time-series analyses might de- tect random and structured effects of toxi- cants on cycle length. For example, a digital filtering technique that removed atypical frequency variations (Orr and Hoffman, 1974) detected cyclic increases in gonadotropins in premenarchal girls (Hanson et al., 1975~. Several groups have analyzed pulsatile LH in rats (Ellis and Desjardins, 1984) with the iterative ap- proach of the Cycle Detector Program (Clifton and Steiner, 1983~. Phase rela- tionships and couplings among cycle length distributions should be investigated in depth. LOSS OF FERTILITY AND FECUNDITY Using infertility to assess the effects of toxicants is difficult, because of the common use of oral contraceptives and the importance of male-specific factors (e.g., oligospermia). Proper diagnosis of infertility usually requires endocrine and gynecologic data. Resorption of most abnormal rodent fe- tuses is reflected in reduced litter size; the age-related decrease in litter size does not result from the shedding of fewer ova at ovulation (Holinka et al., 1979~. Age-related increases in number of stillborn pups contribute to reductions in litter size and are associated with an increase in length of gestation (Holinka et al., 1979~. In humans, there is also an age-related decrease in number of offspring long before the approach to meno- pause, in which behavioral factors also are important (see Fig. 11-7~. The age- related frequency of increasing fetal malformations causes increased rates of spontaneous abortion.
172 FERTILE REPRODUCTIVE TOXICOLOGY Cohort A Cohort B Cohort C a 0 - CJ) llJ _ co lo O co ~ ~ ~ 10 oh o z 8 o z LL C' LL G LL 50 , 50 us 10 50 I ~ 10 . . . . . . . . . . . . . · · . . . . L 1 l 2 X' Nfl \..W,... \ `~_] .'. ..~....-..., 1 L: I... · . . . 5 10 15 5 10 15 5 10 15 AGE (months) FIGURE 13-1 Frequency profiles of estrous-pycle length transitions (4 4, ~5, and ~5 + 54 days) in three cohorts of aging virgin mice. Phase designations are described in Figure 11-5. Reprinted with permission from Nelson et al., 1982. PRECOCIOUS MENOPAUSE Early menopause can have various causes, including hereditary influences associated with either parent, autoimmune destruction of ovarian tissue, mumps oo- phoritis, and exposure to ionizing radia- tion (Mattison et al., 1983; Gosden, 1985; Finch and Gosden, 1986~. Premature onset of infertility and menopause usually is attributed to premature depletion of the ovarian follicular stock (Golden, 1985~. Rodents are susceptible to premature infertility syndromes in association with neuroendocrine damage, as when steroids are administered to neonates in submas- culinizing doses (Mobbs et al., 1984~. Natural variation of rodent infertility occurs due to in utero factors; female fetuses flanked by males eventually become infertile several months before fetuses flanked by females; such effects probably are limited to a critical period during development (Vom Saal and Moyer, 1985~. OVARIAN OOCYI E DEPLETION Oocytes become depleted throughout the lifetime of an individual female. Assays of the oocyte stock require labor- ious histologic analyses of excised ova- ries. Very few human ovaries have been sectioned serially to determine oocyte numbers. Biopsies or ultrasound examina- tions can show whether the ovaries are cystic or atrophied (Orsini et al., 1984), and continued improvements in image analysis systems could make large-scale histological analyses of human oocyte stock plausible in the near future. More information about ovarian oocyte stock might come from administering ex- ogenous gonadotropins for controlled hyperstimulation in new reproductive technologies. Mouse ovaries produce ova with endogenous or exogenous gonado- tropic stimuli until almost immediately before the stock is exhausted (Gosden et al., 1983), as do human ovaries (Sherman et al., 1976~.
REPRODUCTIVE DEVELOPMENT AND AGING HORMONES Gonadotropins and steroids can be as- sayed accurately in small blood samples- 0.01-5 ml. Most assays can be scaled down for mice and rats with appropriate immuno- reagents. In addition, human urine and saliva are sources for measurements of major changes. Gonadotropins Pregnancy-related increases in chorion- ic gonadotropins in urine can be detected by a highly sensitive assay as early as 8 days after fertilization. The assay can help to identify groups with inapparent spontaneous abortion in early pregnancy (Wilcox et al., 1985) that might be due to effects of toxicants. Daily urine samples are needed, in addition to records of in- tercourse frequency and menstrual cycles. A large epidemiologic study is described in Chapter 15. Measurement of LH and FSH to monitor the complex changes during puberty and menopause is problematic because of in- dividual variations. However, postmeno- pausal LH and FSH increases are detected easily, and their absence after menopause might be an indirect biologic marker of hyper - prolac tine mia , since prolac tin suppresses LH secretion at the pituitary (Cheung, 1983~. Stable, increased LH and FSH constitute a biologic marker of prema- ture ovarian exhaustion or atrophy. More refined analyses of daily fluctuations in women of reproductive age require multi- ple daily blood sampling over several months, which is difficult to do on a large scale. Resolution of toxic effects on high-frequency LH surges is even more dif- ficult, and requires sampling every 5 to 1 5 minutes. Prolactin also might be important for monitoring toxicants and drugs that pro- mote growth of lactotropes (pituitary acidophilic cells that secrete prolac- tin). Rodents are susceptible to prolac- tinemia and the spontaneous lactotrope- containing pituitary tumors that commonly arise during acyclicity; these can be in- duced by chronic exposure to estrogens in some genotypes (Finch et al., 1984~. 173 Although lactotrope adenomas in humans have not been linked to estrogen exposure (E1 Etreby, 1980), the possibility is still being considered. Some widely prescribed drugs, such as haloperidol and reserpine, increase prolactin. Monitoring prolactin may not require serial daily blood sam- pling. Fluctuations in prolactin across menstrual cycles (Guyda and Friesen, 1973) are much smaller than increases often seen in hypersecreting pituitary tumors. Steroids E2, progesterone, and DHEA-S are major age-related biologic markers from men- arche through menopause. All are best measured from blood, but all have small daily fluctuations and can be assayed in small blood samples (5-10 ml). Urine also can be used to assay estrogens (Thijssen et al., 1975), progesterone metabolites (Teitz et al., 1971; Speroff et al., 1983; Rebar, 1986; Shackleton, 1986), and etio- cholanolone, an androgen related to DHEA- S (Bulbrook et al., 1971~. DHEA-S changes little across the menstrual cycle (Guer- rero et al., 1976) or diurnally (Rosenfeld et al., 1975~. Daily sampling for proges- terone is needed to identify changes in corpus luteum functions, but a single mid- luteal sample suffices to document corpus luteum formation. Cortisol is also of interest, because of its close relation to stress, which can affect cyclicity, and therefore, fertil- ity. However, extensive diurnal fluctua- tions and response to activity and ambient temperature contraindicate cortisol as a useful general biologic marker of repro- ductive function. NERVOUS SYSTEM A major issue in interpreting the impact of chronic exposure to toxicants on brain functions is the extent of under- lying age changes. Rodents become more sensitive to some agents with age, as judged by greater depletion of dopamine in 28- versus 4-month-old rats exposed to the same dose of the neurotoxin 6-hy- droxydopamine (Marshall et al., 1983~.
174 These age changes might be caused by al- tered clearance or weakened detoxifica- tion mechanisms, of which there is ample evidence (Vessel and Dawson, 1985~. Some rodent studies have shown age-related decreases in extracellular space, which might increase the efficacy of toxicants (Bondareff, 1973~. The effects of age on toxicant stability and metabolism in viva need more study. Alternatively, toxic effects might be greater with age because of age-related reductions in "neuronal reserve" (Fig. 13-2~. Particular support for that idea is found in Huntington's chorea, an autoso- mal dominant disease that has onset in persons 30 to 60 years old and is associated with the selective neuronal deteriora- tion. Asymptomatic carriers have shown necrologic abnormalities after drug chal- lenge (Klawans et al., 1972), as though their resistance to the drug is diminished or their capacity to function under stress has diminished (Finch, 1980~. Thus, pro- gressive, age-related changes in the brain might approach a dysfunction threshold so that the margin of safety for a range of insults is smaller, through either loss of neurons or erosion of the dendritic arbor (Finch, 1976; Calne et al., 1986~. Those phenomena are suspected to occur in the hypothalamus with advancing age, but few data are available. Dopaminergic losses do not produce neur- - a) 100 z - ~: o ~ 10- c,~ O - FEMALE REPRODUCTIVE TOXICOLOGY ologic symptoms until critical threshold is reached (cross-hatched region) leading to parkinsonian symptoms (see Fig. 13- 2~. Ultimately, lesion summates with normal age trend for loss of nigro-striatal dopaminergic activity to exceed putative threshold (Finch, 1976~. Greater induc- tion of reversible parkinsonian symptoms in older patients by antipsychotic dopa- minergic antagonists (Ayd, 1961 ) is also consistent with this model. Parkinson disease thus could afford a "window" on aging changes. Huntington's chorea (Finch, 1980) and Alzheimer's disease (Caine et al., 1986) are among other age- related necrologic diseases with distinc- tive ages of incidence that might be viewed similarly (from Finch, 1981~. The extent to which age-related manifes- tations are time-dependent and cumulative is also uncertain; for example, the inci- dence of skin cancer in mice is related more closely to duration of exposure to benzo- pyrene than to age (Peso et al., 1975~. A general representation of that concept is the experience space of Figure 13-3, which shows how duration of exposure and strength of the stimulus (e.g., toxicant or drug) could interact to determine dif- ferent trajectories that approach the hypothetical dysfunction threshold at different rates. DELAYED CONSEQUENCES OF VIRAL ENCEPHALITIS OR GENETIC DEFICIENCIES 20 40 60 80 100 FIGURE 13-2 Late-midlife onset of some types of parkinsonism might result from early viral lesion (encephalitis lethargica) that causes loss of substantial nigral dopaminergic neurons \ NORMAL AGING TREND (Poskanzer and Schwab, 1961) or from genetic , Viral V~%Loss/Yr) deficiency (Mjones, 1949; Mynanthopoulos et g_ n e ales \ al., 1969) of nigral dopaminergic function. Genetic ~-_ \^ Dopaminergic losses do not produce necrologic Deficiencies? ~-_~, symptoms until critical threshold is reached //// PARKINSON SYMPToMs if//// (cross-hatched region), leading to parkinsonian symptoms. Ultimately, lesion summates with 420 (years) normal age trend for loss of nigro-striatal dopa- minergic activity to exceed putative threshold (Finch, 1976~. Greater induction of reversible parkinsonian symptoms in older patients by an- tipsychotic dopaminergic antagonists (Ayd, 1961) is also consistent with this model. Par- kinson's disease thus might afford a window on aging changes. Huntington's chores (Finch, 1980) and Alzheimer's disease (Caine et al., 1986) are among other age-related necrologic diseases with distinctive ages of incidence that might be viewed similarly. Source: Finch, 1981.
REPRODUCTIVE DEVELOPMENT AND AGING functional ~ impairment T 100 a) Threshold-- Q 50 LL cryptic ~ damage ~' 10 o 175 -- ~ ~ ~ ~ l iN dose o CHANGE (percent) 100 10 time- ~ lifespan Change = f (dose, time) FIGURE 1~3 Ag~related phenomena might be represented on a 3-d~mensional experience surface whose axes are time, dose (strength of a cause of ag~related change, e.g., estradiol in ova~y-dependent neuroendocrine syn- drome of rodents), and change (impairment as function of dose and timed. In many cases, change might be cryptic functional consequence) until some threshold is reached (stippled background). Three trajectories at different doses are shown (arrows). Source: Finch, 1987. Behavior Rodents have a robust repertoire of sex-steroid-dependent behavior that can be used to assess the effects of environ- mental toxicantssuch as lordosis and open-field activity. Hormonal influences on sexual behavior in women have been dif- ficult to prove, and no markers have been generally accepted. Increasing evidence shows that sexual interest might be linked to concentrations of plasma androgens (Morriset al., 1987~. Cell Populations A few reports suggest age-related loss of hypothalamic neurons in the human female (Sheehan, 1968; Swaab and Fliers, 1985~. The data are scarce, however, and do not establish any change in neuron populations as a suitable marker of reproductive neuro- toxicity. However, studies in rodents indicate that this may be a useful biologic marker. In rodents, at least three markers of age-related neuron damage are available. Glial hyperactivity in the arcuate nucleus increases during female reproductive senescence (Schipper et al., 1981 ~ and can be induced in young rats by estrogen exposure (Brewer et al., 1983~. N-Methyl- d-aspartate causes premature cycle lengthening and also kills 30% of arcuate neurons (May and Kohama, 1986~. Binding of lead-210 by the hypothalamic median eminence in autoradiographic studies (Stumpf et al., 1980) and the effect of lead on neurotransmitters (Silbergeld, 1983) are additional possible CNS markers of toxicity in rodents. Postmortem human brain tissue might be used for similar analyses, but few studies have been attempted. In humans, research is needed to deter- mine whether changes in hypothalamic cell populations can be measured nonin- vasively and whether these changes corre- late with reproductive toxicity. Important information must be obtained by brain-imaging techniques, such as 2- deoxyglucose uptake, which is sensitive
176 to neuronal degeneration in other brain regions. Sex-related differences in size of neuron population and in susceptibility to damage from endogenous hormones or ex- ogenous agents also are of interest. Neuron damage often is viewed as an all-or-none outcome, with cell death as the only end point. However, a spectrum of damage should be considered-from rever- sible to irreversible. Neurotoxicants and their interactions with hypoxia and sugar in hippocampal neurons illustrate the possibilities (Sapolsky, 1985, 1986~. Molecular approaches should be applic- able, e.g., by following changes of mRNA that are induced by hypoxia (Pulsinelli, 1985) and corticosteroids (Nichols et al., 1986~. Other molecular markers of inter- mediate stages of neuron damage during early and late responses to toxicants should be sought. Neuron loss from acute or chronic expo- sures to toxicants cannot be evaluated without a greatly extended data base on possibly age-related neuron loss. Data are available from postmortem material that is meager for ages under 65 years. Most studies have been conducted on fewer than 5 brains from persons within any 20- year age span. Consequently, the normative range of neuronal numbers for different brain regions and pathways in healthy young adults is unknown. The absence of these data severely limits future studies on neurotoxicant-age interactions in the hypothalamus. More information will emerge slowly from studies of nerve cell loss during Alzheimer disease, since a range of control groups is being sought. Feasibility studies are needed to deter- mine which of the toxicant-sensitive neur- on populations can be counted most reliably by automated image analysis to determine loss. Normative age-related profiles could be established later from postmortem specimens of neurologically normal brain donors with known health sta- tus enrolled in the Alzheimer's Disease Research Centers. Such tissues would pro- vide a basis for investigating groups that might be at risk of toxicant-related neuron damage, e.g., from occupational exposure to metals or from dietary expo- sure to phytoestrogens. F~hL4LE REPRODUCTIVE TOO OBSERVATIONS ON THE USE OF THESE MARKERS Need for Multiple Biologic Markers Onset, changes, and loss of menstrual cycles and their duration are major b~o- logic markers of changes during maturation and menopause. Fertility depends on other functions as well; therefore, other mark- ers of reproductive status are needed to assess the impact of environmental toxi- cants. Multiple assays of ovarian and pituitary hormones are needed-especially during transitional periods-to establish the stability of basal values and charac- terize the frequency and amplitude of epi- sodic changes, such as preovulatory LH surges and pulsatile LH release. The ef- fects of lead in reducing progesterone concentrations and lengthening luteal phases and menstrual cycles of young adult monkeys (Franks et al., 1986) illustrate the value of these biologic markers and the need for data on menstrual cycle length and hormonal status. Accounts of irreg- ular menstrual cyclicity in association with phytoestrogens from tulip bulbs eaten in the Netherlands during World War II (Burroughs et al., 1985) cannot be evalu- ated clearly without supporting data on nutritional and endocrine status. Differential Susceptibility of Individuals and Populations Biologic markers are needed to identify individuals or populations whose repro- ductive functions are particularly sus- ceptible to toxic effects. Human indi- viduality might arise from genotypic dif- ferences or from a broad spectrum of en- vironmental influences throughout life. A wide range of examples of genotypic and environmental influences on responses to toxicants has been modeled in labora- tory animal studies, genetic influences on the cytochrome P-450 drug metabolizing enzymes (Gonder et al., 1985; Koizumi et al., 1986~. Human genetic polymorphisms in responses to toxicants have not been identified, but studies of twins and drug clearance indicate greater concor- dance between monozygotic twins than be-
REPRODUCTIVE DEVELOPMENT AND AGING tween dizygotic twins (Vesell et al., 1971~. The well-known prevalence of lac- tase deficiency in adult Orientals (McKus- ick, 1975) that causes intolerance to the lactose in milk and unfermented milk prod- ucts also supports the presence of genetic polymorphisms that could influence re- sponses to environmental toxicants. Analysis of Menstrual Cycle Lengths The frequency of menstrual cycles gener- ally has been characterized in statistical terms. The large longitudinal data bases of cycle lengths from hundreds of women collected by Treloar et al. ( 1970) and others could be used to develop new statis- tical descriptors of changes in cycle fre- quency, for example, the frequency of con- secutive cycles of particular lengths. Many sophisticated approaches might de- tect structured or random toxic effects on cycle length and on lengths of consecu- tive cycles. That information also might bear on the increase in birth defects with maternal age. Pilot studies are needed to evaluate the applicability of existing approaches and needs for further 177 development. Primate and rodent responses to toxicants that influence cycle frequen- cy could be used to evaluate the sensitiv- ity of these approaches. Long-Term Necrologic Consequences of Toxicant Exposure In view of the many examples of cryptic neuron damage that results in necrologic disorders years after toxicant exposure, pilot studies should be established to track subjects exposed to neurotoxicants early in life. The recently established followup studies of MPTP exposure (Caine et al., 1985) are a precedent for this ap- proach that might be extended to long- term effects from occupational exposures to lead, manganese, and other neurotoxic agents. Longitudinal necrologic and psychiatric studies at the Alzheimer Disease Research Centers compare various normal and dis- ease-afflicted groups with dementia pa- tients and might be helpful in determining long-term effects. Additional groups for antemortem or postmortem studies could be added easily.
