2
Developmental Defects and Their Causes

Major developmental defects, also referred to as major congenital anomalies, occur in approximately 3% of live births, that is, in 120,000 of the approximately 4 million births per year in the United States (ICBD 1991; CDC 1995; Holmes 1997; March of Dimes 1999; NCHS 1998). These anomalies are defined as ones that are life threatening, require major surgery, or present a significant disability (Marden et al. 1964).

In 1995, major developmental defects accounted for approximately 70% of neonatal deaths (occurring before 1 month of age) and 22% of the 6,500 deaths of infants (before 15 months of age) in the United States (March of Dimes 1999). Approximately 30% of admissions to pediatric hospitals are for health problems associated with such defects.

For more than 20 years, major developmental defects have been the leading single cause of infant mortality in the United States (Petrini et al. 1997). Although infant mortality in the United States has declined by approximately 40% from 1968 to 1995, infant mortality attributable to major developmental defects has declined slightly less, by 34%, and, thus, the overall proportion of infant mortality due to developmental defects has increased from 14% to 22% from 1968 to 1995 (Ventura et al. 1997). In 1995, the leading defects associated with infant death were heart defects (31.4%), respiratory defects (14.5%), nervous system defects (13.1%), multiple anomalies associated with chromosomal aberrations (13.4%), and musculoskeletal anomalies (7.2%) (Petrini et al. 1997).

The personal costs of developmental defects, including emotional and mental stress, are impossible to measure. In national health considerations, it is standard practice to compare dollar costs. The 1992 estimated lifetime cost for 18 of the most significant developmental defects in the United States was $8 billion



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment 2 Developmental Defects and Their Causes Major developmental defects, also referred to as major congenital anomalies, occur in approximately 3% of live births, that is, in 120,000 of the approximately 4 million births per year in the United States (ICBD 1991; CDC 1995; Holmes 1997; March of Dimes 1999; NCHS 1998). These anomalies are defined as ones that are life threatening, require major surgery, or present a significant disability (Marden et al. 1964). In 1995, major developmental defects accounted for approximately 70% of neonatal deaths (occurring before 1 month of age) and 22% of the 6,500 deaths of infants (before 15 months of age) in the United States (March of Dimes 1999). Approximately 30% of admissions to pediatric hospitals are for health problems associated with such defects. For more than 20 years, major developmental defects have been the leading single cause of infant mortality in the United States (Petrini et al. 1997). Although infant mortality in the United States has declined by approximately 40% from 1968 to 1995, infant mortality attributable to major developmental defects has declined slightly less, by 34%, and, thus, the overall proportion of infant mortality due to developmental defects has increased from 14% to 22% from 1968 to 1995 (Ventura et al. 1997). In 1995, the leading defects associated with infant death were heart defects (31.4%), respiratory defects (14.5%), nervous system defects (13.1%), multiple anomalies associated with chromosomal aberrations (13.4%), and musculoskeletal anomalies (7.2%) (Petrini et al. 1997). The personal costs of developmental defects, including emotional and mental stress, are impossible to measure. In national health considerations, it is standard practice to compare dollar costs. The 1992 estimated lifetime cost for 18 of the most significant developmental defects in the United States was $8 billion

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment (CDC 1995; Waitzman et al. 1994). The lifetime per-patient cost for spina bifida alone was estimated at $250,000, and the total annual cost for all surviving infants with spina bifida in the United States was $200 million (Sever et al. 1993). A recent study reported that the total lifetime costs for persons born in 1996 with mental retardation, autism, or cerebral palsy will be $47 billion, $4.9 billion, and $12 billion, respectively (Honeycutt et al. 1999). Major developmental defects are the fifth leading cause of years of potential life lost (YPLL) (CDC 1987). For comparison, loss attributed to heart disease before age 65 is 1,600,265 YPLL, loss attributed to cancer is 1,813,245 YPLL, and loss attributed to major congenital anomalies is 694,715 YPLL (CDC 1987). Those major developmental defects represent only one class of the most socially and medically recognized developmental defects. Several other classes are identified below. Their prevalence has been harder to estimate. To begin with, at least one minor structural defect (e.g., preauricular sinus and syndactyly for toes 2-3) has been identified in 14.1% to 22.3% of live-born infants, a frequency that is 5 to 7 times higher than that for major defects (Leppig et al. 1987). The less-recognized defects are of lesser clinical and cosmetic importance, and the estimate of their birth prevalence varies considerably because of substantial differences in definition and detection and the lack of a national systematic database for this information. Another class is made up of functional deficits—that is, deficits that are not accompanied by an overt structural defect but are expressed in a variety of ways ranging from delays in growth to deficits in behavioral and neurological development. Many of these deficits are only recognized in infancy or later in childhood (e.g., attention deficit hyperactivity disorder and dyslexia). Developmental defects with reproductive consequences might not be detected until much later. Finally, there is evidence of some mid-life health conditions (e.g., heart conditions) correlating with abnormal birth status (e.g., low birth weight) (Barker 1999). The costs and years of life lost have not been estimated for these more subtle developmental defects among live-born infants. A further expanded view of developmental defects is gained by examining all pregnancy outcomes (Table 2-1), not only live-birth outcomes. The most common type of outcome in humans is early-pregnancy loss shortly after implantation (Zinaman et al. 1996; Wilcox et al. 1999). That occurs in 20-30% of pregnancies. Many of those losses are difficult to detect and enumerate because they occur prior to clinical recognition of the pregnancy. Spontaneous abortions of clinically recognized pregnancies (generally starting in the 8th week after the last menstrual period) occur in 10-20% of pregnancies (Hatasaka 1994), also a high frequency. Thus, these two categories dominate all other defects. In 40-50% of the spontaneous abortions examined, some type of chromosomal aberration was found, most frequently an extra or missing chromosome (Jacobs and Hassold 1995). Many chromosomally abnormal embryos have anatomical malformations. Fetal deaths (after 20 weeks of gestation) and stillbirths occur in 1-4% of preg-

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment TABLE 2-1 Frequency of a Variety of Developmental Outcomes Outcome Frequency Reference Early pregnancy loss (before 8 weeks) 20-30% of implantations Zinaman et al. 1996; Wilcox et al. 1999 Spontaneous abortion (8-20 weeks) 10-20% of clinically recognized pregnancies Hatasaka 1994 Chromosomal aberrations in spontaneous abortions (8-12 weeks) 40-50% of spontaneous abortions Jacobs and Hassold 1995 Late fetal deaths after 20 weeks and stillbirths 1-4% of the sum of live births and late fetal deaths Fretts et al. 1995 Major congenital anomalies at birth 2-3% of live births Oakley 1986 Minor developmental defects at birth 14-22% of live births Leppig et al. 1987 Major developmental defects leading to infant death (before age 15 months) 0.016% of live births March of Dimes 1999 Chromosomal aberrations in live births 1% of live births Oakley 1986 Severe mental retardation 0.4% of children to age 15 Mastroianni et al. 1994 Neural tube defects 0.001% of live births Velie and Shaw 1996 nancies (Fretts et al. 1995). Chromosomal aberrations occur in approximately 1% of live births (Oakley 1986). Severe mental retardation is an example of a functional deficit that might not be recognized at birth but is recognized in approximately 0.4% of children before 15 years of age (Velie and Shaw 1996). Developmental defects are often defined as those originating in the embryo and fetus, that is, in the prenatal period. A developmental toxicant is then a toxic agent or condition to which the pregnant mother is exposed. However, development goes on throughout the life cycle and includes for example, the continued growth and differentiation of the nervous, skeletal, and reproductive systems in the juvenile and adolescent, and the continuous renewal of cells of the skin, gut lining, and hematopoetic system of the adult. Thus, it is arbitrary to define developmental toxicants only as those affecting the embryo or fetus through maternal exposure in the period of pregnancy. In this report, the committee emphasizes developmental toxicants to which the mother may be exposed in the prenatal period. However, the division is not sharp, and it is to be expected that toxicant

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment exposures at other periods could affect juvenile, adolescent, and adult development, and might affect gametes and reproductive organs in ways that are only expressed much later in the period of pregnancy. TOXICANT EXPOSURE AND DEVELOPMENTAL DEFECTS The current understanding of the various causes of developmental defects is incomplete. A crude distinction can be made between intrinsic and extrinsic causes. Intrinsic causes include genetic defects (mutations), endogenous chromosomal imbalances (e.g., meiotic nondisjunctions), endogenous metabolism (e.g., phenylketonurea), and perhaps failures in the complex developmental processes themselves. Extrinsic causes include the enormous variety of environmental inputs such as infection, nutritional deficiencies and excesses, life-style factors (e.g., alcohol), and closer to the concerns of this committee, the myriad agents—pharmaceuticals, synthetic chemicals, solvents, pesticides, fungicides, herbicides, cosmetics, food additives, natural plant and animal toxins and products, and other environmental chemicals—encountered by humans. Other environmental factors, such as hyperthermia, ultraviolet irradiation, and X-rays, should be included. As noted before, developmental defects comprise all structural and functional deficits detected in the implanted embryo, fetus, neonate, infant, or child. The committee was asked to consider environmental agents that might cause developmental defects. Such agents include mercury, lead, and polychlorinated biphenyls. Natural plant and animal products and toxins have long been recognized as agents that can cause toxicity. They were some of the first environmental agents to be identified as teratogens. Agents can enter the environment by either deliberate (e.g., pesticide residues on food) or accidental (e.g., chemical spills) releases, and humans can be exposed through food, drinking water, or air. Pharmaceuticals and food additives generally would not be considered environmental agents; however, many of the issues under consideration for environmental agents can also apply to these agents. Additionally, it is possible that they incidentally enter the environment at significant concentrations and become environmental agents. What fraction of developmental defects can be attributed to extrinisic or intrinsic causes? Wilson (1973) estimated that 25% of congenital anomalies in humans are attributable to genetic causes. Then, the author estimated that 65-75% of developmental defects are of unknown causation and attributed fewer than 10% of the anomalies to known environmental causes, including maternal diseases (e.g., diabetes and hypertension), infectious agents (e.g., rubella and syphilis), and mechanical problems (e.g., uterine deformations). Approximately 1% are known to be due to environmental toxicant exposures, including ionizing radiation and hyperthermia (Wilson 1973).

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment In a more recent evaluation, Nelson and Holmes (1989; also see Holmes 1997) collected data on 69,277 infants, of which 2.24% had at least one major congenital anomaly. The infants were in a surveillance program at a university hospital and were not from the population at large and, therefore, these percent-ages should be viewed cautiously. Nelson and Holmes estimated the causes of congenital anomalies to be genetic, 28%; multifactorial inheritance, 23%; uterine factors and twinning, 3%; toxicants, 3%; and unknown, 43%. Multifactorial inheritance (23%), which is a category not distinguished by Wilson, has a genetic and an environmental component. The term is used when geneological studies indicate that a physical trait, disease, or developmental defect occurs at a higher rate within families than expected in the general population, but the patterns of inheritance do not follow strict Mendelian segregation rules. To explain the departure from Mendelian rules, the genetic variant of a gene is said to predispose the individual, but further circumstances, either environmental or other genetic factors, are needed for the production of the disease. An example of multifactorial inheritance is the relationship between maternal smoking, transforming growth factor (TGF) polymorphisms, and oral cleft (Hwang et al. 1995). This example is described in detail in Chapter 5. Such a departure from Mendelian rules might be attributable to environmental factors, but the departure could as well be due to the requirement for a combination of particular alleles of two or more genes to produce the trait (a polygenic trait) or to genomic imprinting. Specific genes and environmental exposures have been associated in multifactorial inheritance in only a few instances, but increased information is becoming available. With the identification of the multifactorial inheritance category in the Nelson and Holmes study, Wilson’s unknown-cause category of 65-75% is reduced to 42-52%, equaling the 43% unknown-cause category of Nelson and Holmes. As the Nelson and Holmes figures indicate, however, the knowledge about the causes and prevention of developmental defects continues to be limited (Mattison 1997). Today, about 3% of the major developmental defects are estimated to be attributable to toxicant exposure (Oakley 1986; Kimmel 1997; March of Dimes 1999), but that figure is a rough approximation. It is generally recognized that 40-50 extrinsic agents probably have acted as human developmental toxicants and that more than 1,200 chemical and physical agents produce developmental defects in experimental animals (Shepard 1998; Schardein 2000). It should be noted that much of the developmental toxicity testing on experimental animals was conducted at up to maternally toxic doses and, therefore, observed effects at those doses might not be the same as effects observed after exposure to environmentally relevant doses. It is not known how many of the 1,200 agents actually produce developmental defects in humans, and the figure is not obtainable by direct testing in humans. In light of the experimental animal results, many of the agents have never entered the marketplace or environment, and others are handled with great caution according to preventive public-health and workplace-safety guidelines.

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment In all, only about 50 chemical and physical agents are known to cause developmental defects in humans (Friedman and Polifka 1994; Shepard 1998). They include so-called “life-style” chemicals, such as alcohol, accounting for 0.1-0.2% of defects in live-born infants and cocaine, a variety of pharmaceuticals, and several environmental agents (e.g., mercury, lead, and polychlorinated biphenyls). Table 2-2 lists several representative human developmental toxicants. There is information available on proposed mechanisms of action for these toxicants, however, it is infrequently synthesized into a cohesive and comprehensive mechanistic explanation. Over 80% of agents known to produce developmental defects in humans also cause developmental defects in at least one test animal (rat, mouse, or rabbit) (Shepard1998). The actual percentage of environmental agents that are developmental toxicants in humans could be higher or lower than 3% for several reasons. For example, the epidemiological methods for identifying toxicants are inherently insensitive and depend on the systematic examination of large human populations. Such large-scale examination is difficult to do (Selevan 1985). Thalidomide was an exception. Because it caused such a distinctive outcome (bilateral limb shortening) of a rare human malformation, its effects were recognized in small patient groups. Even though the frequency of fetal alcohol syndrome is high compared with other developmental disorders, it took many years to identify alcohol as a human teratogen because the physical alterations are subtle, and the learning and social adjustment problems are sometimes not detectable until several years after birth. When a human exposure problem is suspected, epidemiological testing can be performed to assess toxicity, but few of the 1,200 agents have been so examined. Also, the number of agents that cause developmental toxicity might be higher if the multifactorial inheritance category of birth defects contains, as indeed is suspected, cases of human variants who are genetically more susceptible (predisposed) to particular environmental conditions than are others. Finally, the number might be higher if some toxicants (extrinsic causes) produce malformations as a consequence of their primary effect in causing genetic damage (intrinsic cause). In conclusion, although it is recognized that environmental agents can, and some do, act as developmental toxicants, it is still unclear how large a role these agents play in producing human congenital anomalies relative to other sources of developmental toxicants such as pharmaceuticals and food additives, and relative to intrinsic causes such as genetic differences. THE CHEMICAL UNIVERSE The “chemical universe” refers to the collective variety of chemicals that humans encounter. This variety is theoretically infinite if no limit is set on the molecular size of chemicals, because new and more complex compounds can always be made by coupling together simpler chemical units. In practice, how-

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment TABLE 2-2 Representative Human Developmental Toxicants Agent Use Developmentally Toxic Dosage Adverse Effects 13-cis retinoic acid Treatment of cystic acne 0.4-1.5 mg/kg/d Craniofacial and cardiovascular malformations and intellectual deficits Aminopterin Folate antagonist 1-2 mg/kg/d Abortion, central nervous system and craniofacial defects, and growth retardation Angiotensin converting enzyme inhibitors Antihypertensive Therapeutic dose (differs for each) Fetal death, stillbirth, oligohydramnios, growth retardation, hypotension, and renal failure Cigarette smoke Stimulant >20/day Growth retardation and facial defects Coumarin derivatives Anticoagulants Therapeutic dose (differs for each) Facial defects, limb anomalies, growth retardation, and neonatal respiratory distress Cyclophosphamide Antineoplastic 4 mg/kg/d Limb and facial defects Diethylstilbestrol Synthetic estrogen 0.1-3 mg/kg/d Reproductive tract malformations and vaginal cancer Diphenylhydantoin Anticonvulsant 8 mg/kg/d Craniofacial defects, growth retardation, fetal loss, and intellectual deficit Etretinate Treatment of psoriasis 0.5-1 mg/kg/d Limb, ear, cardiac, and thymic defects Lead Environmental contaminant 10-15 µg/dl/blooda Abortion, growth retardation, and neurobehavioral deficits Lithium Bipolar disorder 3-5 mg/kg/d Cardiac defects Methylmercury Environmental contaminant 10 µg/kg/db Central nervous system defects Penicillamine Chelator 20 mg/kg/d Connective tissue defects Polychlorinated biphenyls Environmental contaminant   Growth retardation, hyperpigmentation, and neurobehavioral deficitb Thalidomide Sedative/ hypnotic 0.7-3 mg/kg/d Structural malformations (particularly reduction defects of the limbs and ears) Valproic acid Anticonvulsant 5-10 mg/kg/d Neural tube closure defects a The dosage of lead needed to reach these blood levels is dependent on route of exposure. Typically, humans are exposed to lead through a combination of inhalation and oral ingestion. b There is uncertainty about the dosage associated with adverse effects.

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment ever, chemists tend to produce synthetic chemicals or analyze natural chemicals that are between a molecular weight of a few hundred to a few thousand. Technological advances in chemistry will continue to increase the number of synthetic chemicals. Some of these chemicals will undoubtedly be of benefit to society; however, potentially harmful chemicals need to be identified so that human exposure is controlled or prevented. For risk assessment, a key goal is to determine which exposures to chemicals may be harmful to humans before exposure occurs. Millions of synthetic chemicals are registered with the American Chemical Society, but fewer than 100,000 are currently in commercial or industrial use and, therefore, available for introduction into the environment (EPA 1997). Most of these chemicals have not been tested for developmental toxicity. For example, EPA (1998a) conducted a study assessing data availability on close to 3,000 chemicals that the United States produces or imports at more than 1 million pounds per year and concluded that only 23% of those chemicals had been tested for reproductive and developmental toxicity. Test data were considered available if any studies relevant to reproductive and developmental toxicity were located. The number of synthetic chemicals is likely to increase greatly in the near future. Recent advances in combinatorial chemistry have made it possible for chemists to synthesize in parallel small amounts of a large number of chemicals (a “library,” on the order of 104-106 kinds per application). Biologically active members of the library are selected by their performance in specific biological assays (usually in vitro assays rather than animal tests) based on recent insights into the workings of cellular, developmental, and pathological processes. These new synthetic and selective methods are expected to lead to the development of drugs that are more efficacious and have specific pharmacological activities. Although all but the most promising chemical candidates will be restricted to the laboratory, it will be a challenge to gain toxicity information about many of these so that not only the most efficacious but also the safest can advance to the next phase of drug development. As drug discovery and development approaches become more sophisticated, toxicity testing approaches must also become more sophisticated. Toxins, such as chemicals from microorganisms, fungi, plants, and animals (e.g., sponges, coelenterates, and bryozoa), have not been analyzed systematically. Systematic analysis has shown that the variety of chemical constituents is known to be great in some naturally occurring substances. For example, over 400 chemicals have been identified in red wine, and over 1,000 chemicals have been found in tobacco or tobacco smoke. Naturally occurring substances sometimes have significant pharmacological and toxic properties (see, e.g., NRC 1996). Catalogs of known naturally occurring plant toxins, for example, include more than 2,000 entries, and the number with pharmacological activity is larger (Keeler and Tu 1991; Harborne and Baxter 1996). Animals, including humans, have evolved enzymes and ligand-binding proteins to metabolize and eliminate many natural environmental chemicals. They

OCR for page 17
Scientific Frontiers in Developmental Toxicology and Risk Assessment have also evolved adaptive mechanisms, stress responses, and checkpoint pathways to prevent or correct damage from various environmental chemicals and physical conditions (e.g., heat shock), and hence to survive in their environment. Thanks to the broad specificities of these proteins and adaptive processes, animals also detoxify and adaptively respond to many synthetic chemicals as well, even though the animal has never seen these chemicals before in its evolution. Understanding detoxification and adaptation processes in animals and especially in humans has widespread implications for the field of developmental toxicology (for reviews, see Juchau 1980; Juchau et al. 1980; Shepard et al. 1983). Still, some small fraction of old and new chemicals, synthetic and natural, can elude the animal’s defenses enough to impact components of its developmental processes, thereby leading to developmental defects. SUMMARY The frequency at which all classes of developmental defects occur is thought to be very high, perhaps exceeding half of initial pregnancies. However, the total frequency of developmental defects is only vaguely known, and the means of surveillance for defects are only approximate. It is thought that among newborns with major developmental defects, genetic transmission accounts for perhaps 25% of the cases. Lesser genetic defects, which are insufficient on their own to cause major defects but are sufficient in combination with environmental factors or other genetic factors, account for perhaps another 25% of major defects. Genotype combined with environmental causes is a class of developmental defect that is expected to receive incisive attention in the near future. Genotype is an important class because of its implications that some environmental agents might act as toxicants for some people (predisposed individuals) but not for others, making risk assessment a process requiring information about human diversity as well as toxicant action. A few percent (approximately 3%) of developmental defects are probably attributable to chemicals and physical agents alone and have no known genetic contribution. An unknown fraction of those are due to environmental toxicants. Finally, the causes of nearly half of the major defects immediately detected at birth are so poorly understood that they cannot even be classified as being caused by intrinsic or extrinsic factors or both. Presumably, some fraction of them have a complex environmental component. At the same time, there is a steadily expanding universe of chemicals and combinations of chemicals to which humans are exposed. Most of these chemicals have never been tested for developmental toxicity.