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.
16 Conclusions and Recommendations Before a biologic marker is applied in a toxicologic context, its frequency and distribution in a normal, healthy popu- lation must be established. A logical next step is to validate the marker and judge its utility by using it in a high- exposure, high-risk cohort to verify that it measures what it is meant to measure. Markers of biologic processes rely on subjective or objective assessments. Regardless of the care taken to construct measurement instruments, patient-depen- dent measures are likely to be highly vari- able; one reason is the presence of bias associated with failure of recall, lack of understanding of the question, or lack of recognition of a symptom. Objective measures are often superior, provided that they are reliable and valid. Application of biologic markers in re- productive toxicology depends on inter- disciplinary work by laboratory scien- tists, clinicians, and epidemiologists. Collaborations of this kind impose heavy obligations on investigators to keep abreast of and understand all the disci- plines involved. Because of the rapid pace of change in the development of new tools and bioassays, agency implementation of field studies with biologic markers of human reproduc- tion should be guided by an oversight panel of experts conversant with the reliability 197 and validity of such markers, measurement issues, and design and conduct of field studies. A task force might be convened periodically to reassess the status of biologic markers in use and proposed for future studies. Biologic markers must be fully developed and studied before they are applied in epidemiologic studies as the principal indicators of events. Initially, it might be advisable to use a new biologic marker in conjunction with other assessments. Without adequate laboratory testing and cross - validation with markers having known properties, a potential marker is likely to be misused or misinterpreted. Epidemiologic studies need to be de- signed carefully because of the large ef- fort involved in collecting data, particu- larly if a study entails biologic assays. Nested designs might be appropriate to test questions involving intensive labor- atory assessments or involving only spe- cial exposure conditions (NRC, 1985~. During design of a study, careful consider- ation should be given to the next steps to be taken; this will help to ensure that questions can be addressed retrospective- ly as new information develops. Longitudinal studies will be particu- larly important for the identification of toxic effects. Ways must be found to assess damage during fetal oogenesis, and
198 identification and tracking of cohorts with relevant in utero exposures will be important. Daughters exposed to DES, for instance, soon will enter the perimeno- pausal period, at which time it will be possible to see whether age at menopause is indicative of oocyte damage sustained in utero. For postnatal exposure, contin- ued monitoring of ovarian failure and re- covery among women treated for cancer will be useful. SPECIAL RESEARCH OPPORTUNITIES Many important opportunities for col- lecting data relevant to the evaluation of toxicologic effects on female reproduc- tion are at hand. Environmental health research could - be incorporated into current assessments of clinical populations and studies di- rected at answering biologic questions. Valuable information regarding the poten- tial utility of biologic markers could be developed by integrating assessment of environmental exposures and tests for various biologic markers into existing work with populations undergoing in vitro fertilization, artificial insemination, prenatal diagnosis, and normal obstetric care. Such an approach would be cost effec- tive for research funding, and could be useful in assessing effects of common ex- posures. Hence, the committee recommends that clinics and epidemiologic studies of women of reproductive age be encouraged to collect information relevant to patient exposure history. Treatment of infertility problems could provide extensive information on reproductive physiology. For instance, IVF centers have contributed knowledge about factors influencing oocyte quality and the milieu for continued development of the fertilized egg. Such data are valu- able for assessing and predicting the ef- fects of various toxic exposures. Materi- als from IVF centers would lend themselves to systematic research and study, although ethical concerns regarding their use in research must be considered. Fetal and placental material from spon- taneous and induced abortions might pro- FEMALE REPRODUCTIVE TOXICOLOGY vice valuable information on the pharmaco- kinetic properties and potential effects of common toxic exposures. Such material is not widely used to assess environmental health risks to the fetus. Reproductive function is related to other biologic processes; for instance, ovarian hormones appear to influence the risk of cardiovascular disease and the rate of bone loss (Bush and Barrett-Con- nor, 1985~. Thus, it might be expeditious to use current studies of osteoporosis and heart disease in women as vehicles for addressing questions about the role of toxic insult in ovarian senescence. SPECIFIC RESEARCH RECOMMENDATIONS The preceding chapters reviewed many biologic markers of exposure and effect. Table 16-1 summarizes the status of each marker. Germ Cell Damage Induced aneuploidy in female germ cells is an important subject for basic science, especially because aneuploidy in humans arises largely in maternal germ cells. Fundamental understanding of the mechanisms of chromosomal malsegregation, as well as assays to detect agents that induce aneuploidy, must be improved. Research should be encouraged to clarify the age-related increase in oocyte aneuploidy. Effects on oocytes of substances in the follicular fluid are not well understood; what substances get through the zone pellucida and whether barriers weaken with age are unknown. Valid noninvasive methods to evaluate oocyte stock directly are needed. These important subjects involve fundamental research as well as toxicologic studies. Sexual Differentiation "Critical periods"a unique feature of sex differentiation-have important implications for the subject of markers in reproduction. At the beginning of em- bryonic life, the gonads and other organs destined to become the genital tract and
CONCLUSIONS AND RECOMMENDATIONS 199 TABLE 1~1 Status of Current and Potential Markers in Female Reproductive Toxicology Animal Can Be Used Can Be Used Needs Studies in Limited in Large Scale Further Marker Neededa Human Subsetsb Human StudiesC Development Exposure markersChemical analysis for toxicants or metabolites, or mutagenic analysis of body fluids Blood, urine, saliva Tissues Intact Cytologic specimens Fluids Cerebrospinal fluid Follicular fluid, amniotic fluid Placental tissue Peritoneal fluid Genotoxic markersDNA + + + + + + adducts (chemical specific, generic) Oocytes, ovarian tissue + Placental tissue + Fetal tissues + Maternal serum + Fetal serum + Unscheduled DNA synthesis Maternal lymphocytes + Fetal lymphocytes + SCE (sister-chromatic exchange) Maternal lymphocytes + Fetal cells + Chromosomal aberrations Maternal serum + Abortus tissue + Chorionic villi + Amniotic cells + Fetal serum + Micronuclei Maternal blood Vaginal/cervical cells Fetal liver cells Fetal lymphocytes Specific-locus mutations Development/aging Onset of puberty Clinical observation, breast bud development Blood Melatonin DHEA-S Gonadotropin (pulsatile) Age of first menstrual bleeding Hormones: estrogens, inhibin, LH, FSH, androgens Age of breast development Sexual behavior Neurotransmitters in CSF Menstrual cycle length + + + + + + + + + + + +
200 FEMALE REPRODUCTIVE TOXICOLOGY Animal Can Be Used Can Be Used Needs Studies in Limited in Large Scale Further Marker Neededa Human Subsetsb Human StudiesC Development Ovanan-ooc yte stock Ultrasound for ovarian size IVF Biopsy MRI Periodic ultrasound to monitor follicular development Inhibin Premenopausal hormonal status (estrogens, gonadotropins, inhibin, LH, FSH) CNS reproductive senescence + Menstrual Function Cycle frequency and characteristics Detection of corpus luteum follicular development (ultrasound) Basal body temperature Thermometer + + + + Improved, self-recording, electronic thermometer Cervical mucus Sexual behavior Vaginal cytology Biophysical measurements of vaginal secretions Endometnal histology End o cnnolog;y: go na d o t repins, steroids, ovulatory hormones + In vitro assays LH-FSH Pituitary cells (from cadavers) Granulosa cells Luteal-specific proteins, endometnal cell cultures Mucus production, endocemcal cells Fertilization, Implantation, and Loss hCG EPF PEP + + + + + + + + + + + + + + + a + = not ready for application in humans. b + = too invasive or too demanding of subjects for use on broad scale. c + = sufficiently validated and safe for application in field studies, although might warrant further refinement and additional work with animals. the external genitalia are bipotential. Primordial germ cells induce development of a male or female gonad, and the gonads secrete diffusible organizing substances, or hormones, that complete the process of morphogenesis of sexually relevant tissue. Bipotential tissues are inducible during only a short period; if induction or differentiation is prevented during that critical period, later exposure cannot alter morphology or function. For example, at a critical early time in the human embryo, a functional testis secretes a protein that suppresses the anlage of
CONCLUSIONS AND RECOMMENDATIONS the Mullerian duct. The protein-called Mullerian duct-inhibiting hormone, Mul- lerian-inhibiting substance, and anti- Mullerian hormone-and its early secretion (a critical event) ensures that males do not retain female sex accessory organs. In the neonatal male rat, testosterone secretion on the first postnatal day per- manently alters the size of hypothalamic nuclei; testosterone does not have this effect after postnatal day 5. An environmental event might alter re- productive organs at one time-during a critical period-but not at another time, and an abnormality manifested in adults could be the result of an environmental insult that was effective only during a critical period of reproductive ontogene- sis. The specific relationships between particular developmental periods and particular toxicants are not well under- stood. Moreover, some functional effects might not be manifested until oubertY or menopause. Therefore, in utero exposures need to be documented, and exposed popula- tions should be followed up through many periods of life. Puberty The onset of puberty is affected by nu- tritional status; hence, other environ- mental factors might affect puberty onset, although systematic studies are lacking. The hypothalamus presumably is insensi- tive to some environmental insults early in childhood. However, when hypothalamic secretions begin during puberty, exposure to some toxicants might have an effect. Studies of normal populations and of popu- lations potentially exposed during child- hood or puberty need to be conducted. Ovulation Probably only a small percentage of the healthy population is trying to become pregnant at any point. Consequent- ly, the study of nonconceptive menstrual cycles is important as a way to monitor or to study exposed populations. Changes in menstrual cycles might have long-term consequences for other medical condi- tions, such as cancer and heart disease. 201 Among the components of female fertility, ovulatory function appears to be the most sensitive to environmental effects; therefore, it is of interest to investigate the influence of various agents on menstru- al cycling. This should be studied in ani- mal models, as well as human populations. Short-term changes in the menstrual cycle induced by stress or diet differ from long-term toxicity. Markers are need- ed that can distinguish between these events. A battery of assessments should be developed that could be used to evaluate ovulatory function in current epidemio- logic studies and in special studies of exposed human populations. Because anovu- latory bleeding can be confused with regu- lar ovulatory cycles, a more accurate de- tection of ovulation would be useful. These assessments should be sensitive to alterations induced by low concentrations of toxicants and should be designed so that they can be readily applied in field stud- ies. Field studies to monitor cyclic ovar- ian function with urinary assays of steroid and gonadotropin hormones have been con- ducted successfully, although not in the context of toxicology. These studies sug- gest that corpus luteum formation can be documented in a practical and valid way. Evaluating cellular physiology in vitro allows pathophysiology of naturally occurring diseases that cause anovulatory infertility to be investigated. Whether this in vitro approach will prove profit- able in evaluating biologic effects of environmental chemicals remains to be seen. Future efforts in this field will expand the understanding of biologic markers. Clinically Inapparent Loss Early spontaneous abortions might be related to ovulatory dysfunction or to genetic damage, and studies of clinically inapparent losses should yield important toxicologic information. Such studies must be carefully considered. The assay used must be sufficiently sensitive to detect early pregnancy and loss and must be applied appropriately with regard to timing of specimen collection and choice of fluid to sample. With appropriately
202 sensitive assays, consistent estimates of the frequency of implantation and early loss have been obtained; such estimates are important for use in studies evaluating environmental exposures. Ovarian Senescence An accurate way to assess the oocyte pool, either with imaging techniques or with measures of regulatory factors or gonadotropins, is needed. Presumably, rates of loss vary. Groups with premature senescence should be assessed, to identify differences from normal populations. NATIONAL DATA BASE A national data base should be developed to obtain information on human reproduc- tive biologic markers in normal and exposed populations. Several major clinical cen- ters could supply information to help to develop national reproductive profiles. The National Center for Health Statistics might conduct a survey of a probability sample of women that involved periodic assessments (including biologic assess- ments) of reproductive measures. The Agency for Toxic Substances and Disease Registry might appropriately develop comparable data on exposed populations and coordinate a data base to link informa- tion on the two groups. Infertility and irregular menstruation are neither life- threatening nor rare, and are likely to be undetected in the general population unless they are monitored. EXPERIMENTAL STUDIES The relevance of data obtained in non- human primates or nonprimate mammals often FEA~1LE REPRODUCTIVE TOXICOLOGY is questioned-principally because the 28-day human menstrual cycle does not ap- pear to resemble the 4-day rodent estrous cycle, the spontaneous ovulation of the primate is quite different from coitus- induced ovulation in the rabbit, and the human does not show the clear seasonality of cycles manifested by Rhesus monkeys, sheep, and rodents (such as hamsters). Nevertheless, the basic processes of fe- male reproduction are probably similar in all mammals. Species differences are apparent in some phenomena, such as luteal function; for example, the unmated rat does not have a functional corpus luteum of the cycle, whereas the primate does. The rat and mouse are probably good models for oogenesis in primates, but not for masculinization of the hypothalamus, in that testosterone administered peri- natally in female rats permanently pre- vents estradiol from inducing an LH surge, but does not do so in primates. the Rhesus monkey is an excellent model for human follicular and luteal phases, but is expensive. More work on luteal events probably should be carried out in the guin- ea pig-the only small, readily available laboratory mammal with a spontaneous lute- al phase; the potential for xenobiotics to alter luteal function or cause luteoly- sis might be better assessed in this spe- cies. Careful consideration needs to be given to selecting appropriate laboratory models for different components of female reproduction.
