Children and the unborn child

4. Testing methodology

4.1 The use of epidemiology
4.1.1 Types of epidemiological studies
4.1.2   Bias
4.1.3 Confounding
4.1.4 Controls
4.1.5 Strength
4.2 Test methods in experimental animals
4.2.1 Reproductive toxicity tests
4.2.2 Comparison to assessment of effects in adult animals
4.3 Test procedure for medicinal products
4.4 References


Well-planned and documented epidemiological studies have a clear advantage over studies in experimental animals in providing the most relevant information on health effects in humans, thus avoiding extrapolation from experimental animals. However, owing to the lack of adequate epidemiological data for most chemical substances, toxicological studies in animal species play an important role in hazard identification and risk assessment.

4.1 The use of epidemiology

4.1.1 Types of epidemiological studies

In general there are two groups of basically different types of studies in epidemiology. One group includes different types of epidemiological experiments e.g. clinical trials (with patients as subjects), field trials (with healthy subjects) and community intervention trials (with the intervention assigned to groups of healthy subjects). For various reasons these types of experiments are hardly ever used in human toxicology (except for the registration of adverse effects in phase 2 and 3 of clinical trials). In the other group are the non-experimental trials of which there are three primary types: the cross-sectional study, the case-control study and the cohort study.

Cross-sectional study

In the cross-sectional study both the exposure and the effect are measured at the same time in a defined population (all persons in a limited area or a random sample). Because of the lack of longitudinal data this type of studies is often insufficient to elucidate a cause-effect relationship especially if there is a time delay before the effects occur. The advantage is that the data are collected within a limited time resulting in a relatively short time interval between generation of the hypothesis and the answer.

Case-control study

In the case-control study different exposures in two selected groups of people, the cases (the ill) and the controls (the reference group of well persons) are compared. This type of study is relatively simple and economical to carry out. The cases are defined by the investigator (case definition). A classic example of a case-control study was the discovery of the relation between thalidomide and unusual limb defects in babies born in the Federal Republic of Germany in 1959 and 1960; the study, undertaken in 1961, compared affected children with normal children. Of 46 mothers whose babies had typical malformations, 41 had taken thalidomide between the fourth and ninth weeks of pregnancy, whereas none of the 300 control mothers, whose children were normal, had taken the drug at these stages. In a case-control study the association of an exposure and an effect is measured by calculation of the odds ratio, which is the ratio of the odds of exposure among the cases to the odds of exposure among the controls.

Ccohort studies

In cohort studies (follow-up studies) the risk factor (exposure) in a defined group is defined and measured. Then the group is followed over time for development of the disease (or another end- point parameter). The ratio between the risk in the exposed group and the risk in the unexposed group is the Relative Risk (RR). In this kind of studies the entire population is known, which is not always the case in the case-control study. The disadvantage of this type of study is that they are usually very big and expensive. On the other hand they provide the best information about the causation of disease and the most direct measurement of the risk of development of disease.

4.1.2 Bias

The quality of the result of epidemiological investigations depends on the validity and accuracy of the data on the exposure as well as the effect. The validity depends on the number of systematic errors when data are collected (bias). The accuracy depends on the number of random errors. The two have a tendency to counterbalance each other. Information- and selection bias give rise to misclassification. Whether the study population is selected in a truly random manner or not is crucial for the results of epidemiological trials. Selection bias (e.g. healthy worker effect) might mask an association between an exposure and a health effect. A small vulnerable group might appear insignificant in a big less vulnerable group (maybe they are already dead due to the exposure) and a small heavily exposed group might appear insignificant in a big less exposed group. A low rate of participation increases the risk for selection bias.

4.1.3 Confounding

In a study of the association between exposure to a cause (or risk factor) and the occurrence of disease, confounding can occur when another exposure exists in the study population and is associated both with the disease and the exposure being studied. A problem arises if this extraneous factor – itself a determinant or risk factor for the health outcome – is unequally distributed between the exposure subgroups. It could be age, gender, smoking habits, socio-economic status etc. One way of reducing the effect of confounding is by matching cases and controls with regard to the potential confounders.

4.1.4 Controls

In case-control studies the controls act as representatives for the background exposure. In order to fulfil this they have to be chosen independently of their exposure and they should have been submitted to the same selection mechanisms as the cases (i.e. as potential cases). When selecting the control group there is a potential for selection bias. Random controls selected from the community often introduce recall bias. They are less prone to recall a certain exposure in the past than cases, resulting in a drift towards a false positive association. Problems of similar kinds arise when using other not randomly selected controls as family, friends or cases with another end-point.

4.1.5 Strength

In general, case descriptions are suitable to show a possible effect of an exposure but they are insufficient to evaluate a risk. Case-control studies may be limited by the small study groups with few cases and rare specific exposures. Cohort studies may be limited by the rare occurrence of effects associated with the exposure in question. Hence, the strength of the statistical calculations are often limited in case-control and cohort studies.Traditional epidemiology is often challenged by incorrect information about exposure and effect, resulting in a low sensitivity. This, and a small gap between the exposure in the exposed and the unexposed group, makes it difficult to study the effects of low dose exposures. To get statistically significant results either large study populations, a strong cause-effect relationship or a very rare effect parameter are required. A small study population may in itself be the reason for overlooking a relevant association. This makes it risky to conclude that there is no cause-effect relationship when the study indicates that the exposure-effect relationship is not existing. In reviewing the literature there is an additional risk of publication bias, as both the scientists and the editors are more prone to publish positive correlations than negative results.

In general epidemiological studies has a number of advantages as compared to animal experiments: they study effects in humans, they are often cheaper, they can be based on already existing data, and they can use intellectual and psychological end-points.

4.2 Test methods in experimental animals

4.2.1 Reproductive toxicity tests

Many different experimental methods for investigating toxic effects of chemicals on reproduction and development are in use. Several tests are standardised and guidelines have been issued by various governmental agencies and international organisations, others are still undergoing scientific evaluation. In the following sections, standardised and regulatory accepted methods are discussed concerning procedures and endpoints with emphasis on effects induced during the prenatal and postnatal period. Table 4.2 summarises the tests discussed. In the last section the effects assessed during the prenatal and postnatal period is compared to the possibility for detection of effects in toxicity testing in adults, e.g. in repeated dose toxicity studies.

Other test than those included can reveal effects which indicate a potential of a chemical to interfere with normal reproduction, e.g. the dominant lethal test, fertility assessment by continuous breeding, and repeated dose toxicity testing where the gonads are subjected to pathological examination. These tests, however, provide only information on effects of dosing adult animals.
During recent years many in vitro test systems have been proposed as alternatives to animal testing for developmental toxicity. These tests may be useful for screening of closely related chemicals and for pinpointing mechanisms underlying developmental effects, but they cannot replace animal testing. Consequently, they are not considered in the following sections.

Table 4.2
Overview of in vivo tests for reproductive toxicity testing.

Test

Exposure period

Endpoints in offspring

Guideline(s)

Generation studies

Continuously over one, two or several generations

Growth, development and viability
Histopathology of sex organs, brain and target organs
Fertility
Proposal: oestrous cyclicity and sperm quality

OECD TG 415 One-generation Study OECD TG 416 Two-generation Study

Prenatal Developmental Toxicity Study (Teratology study)

Usually during organogenesis
Proposal: from implantation to the day before birth

Resorptions
Foetal growth
Morphological variations and malformations.

OECD TG 414

Developmental Neurotoxicity Study (Behavioural teratology studies)

During pregnancy and lactation

Birth and pregnancy length
Physical and functional maturation
Behavioural changes due to CNS and PNS effects
Brain weights and neuropathology

OECD TG 426 Developmental Neurotoxicity Study (draft 1999)

Reproduction/

Developmental toxicity screening test

At least three dose levels from 2 weeks prior to mating until day 4 postnatally

Fertility
Pregnancy length and birth
Foetal and pup growth and survival until day 3

OECD TG 421 and 422

Modified from Hass et al. (1994).

Design considerations in reproductive toxicology studies

When designing experimental in vivo studies in reproductive toxicology some important factors must be considered such as selection of animal species, dose, time of dosing, route of exposure, and housing and handling of the animals.

The animal species usually used for screening studies are rats, mice and rabbits. There are important species differences between humans and rodents especially concerning the timing of brain development in relation to the birth of the offspring. These differences must be taken into account when extrapolating from animal data to the human situation.

The dose levels used in animal studies are often higher than the expected exposure levels in humans. There are several reasons for this. Firstly, the limited number of animals per group limits the sensitivity of the experimental studies; secondly, the human exposure may continue for a long time, even throughout life, while the exposure period in experimental studies is shorter; and thirdly, comparisons of human and animal data seem to indicate that humans are more sensitive than the commonly used laboratory animals when the dose is expressed as mg/kg (Jakobsen & Meyer 1989). The highest dose level should normally be chosen with the aim to induce (slight) general toxicity in adult animals but not death (OECD 1983, OECD 1994).

The time and duration of treatment in reproductive toxicity tests influences the types of effects found. Tests for developmental toxicity are, for example, often restricted to the period of organogenesis, i.e. days 6-15 for mouse and rats, and days 6-18 for rabbits. This period is sensitive to induction of structural malformations while functional effects may be induced at all stages of development. Ideally, treatment in the mouse, rat and rabbit should continue from day 6 of pregnancy through to weaning, though there may be advantages in separating prenatal and postnatal phases in different groups of animals (Barlow & Sullivan 1975).

The route of exposure in studies designed for hazard assessment should ideally be the same as the expected exposure of humans. In the workplaces the main routes of exposure are via inhalation or uptake through the skin. However, oral exposure is often used and recommended in guidelines (OECD 1983, OECD 1994). An important difference between inhalation exposure and e.g. oral exposure is that there is no first-pass effect in the liver for inhaled chemicals, i.e. the chemical in the bloodstream reach the placenta before the liver. Therefore, extrapolation from experimental inhalation studies to inhalation exposure of humans is easier than extrapolation from oral studies. Inhalation studies are, however, not easily performed during the birth period and the early neonatal period. Therefore, inhalation exposure is often postponed a few days before expected birth and only in some cases continued after birth at day 2 or 3.

Standardisation of litter size to 8 pups (culling) was previously commonly used in generation studies, and the procedure is described in the guidelines for generation studies (OECD 1983). An argument in favour of culling is that since pup weight is related to litter size, culling might lead to a more uniform pup weight at weaning. However, in a study of data from approximately 500 litters culling seemed only to increase the mean pup weight in the litters, while the variation remained unaffected (Palmer 1986). Actually culling results in elimination of 25-40% of the offspring and may introduce bias for example by random elimination of runts (Palmer 1986).

Postnatal effects, especially on body weight and behaviour of the pups, may be induced via effect on the mother. For example, lactation or maternal care may be affected and potentially any alteration in maternal physiology and behaviour may in turn influence the offspring. To control for this, cross-fostering techniques may be used where prenatally exposed litters are reared by non-exposed mothers and vice versa. This obviously requires more animals and demands more resources, and is therefore not normally used in screening studies.

Indications of gender differences in response level were among other examples seen in the Collaborative Behavioral Teratology Study (CBTS) in the USA (Kimmel & Buelke-Sam 1985). There were no indications that female behaviour were more variable than male behaviour in the CBTS study. Therefore, there is no reason to exclude females because of possible response variability due to oestrus cycling. Especially in a screening situation, behaviour of both sexes should be monitored (Kimmel & Buelke-Sam 1985).

Housing and handling of animals may influence the results. For example, frequent handling of rats during infancy may alter their physical response to stress and their behaviour in tests for emotionality and learning. To control for environmental influences, the conditions under which the animals are kept must be standardised within experiments with respect to variables such as temperature, humidity, noise level, lighting, cages, handling, and cage cleaning (Barlow & Sullivan 1975).

One-two-, and multigeneration studies

Guidelines for carrying out one-, two-, and multigeneration studies have been published by the US-FDA, OECD (1983), and the EEC, amongst others. In the EEC, the test is requested for new industrial chemicals (i.e. introduced after 1980) with a production volume reaching 100 tons per year.

The purpose of generation studies is to examine successive generations to identify possible increased sensitivity to a chemical, effects on the fertility of male and female animals, pre-, peri-, and postnatal effects on the ovum, foetus and progeny, including teratogenic and mutagenic effects, as well as peri- and postnatal effects on the mother.

In a one-generation study, the test substance is administered in graduated doses to groups of males and females. Males should be dosed during growth and for at least one complete spermatogenic cycle (approx. 56 days in the mouse and 70 days in the rat) in order to elicit any adverse effect on spermatogenesis. Females should be dosed for at least two complete oestrus cycles in order to elicit any adverse effect on oestrus. The test substance should be given continuously during mating and for the females also during pregnancy and the nursing period.

The main idea in two- and multigeneration studies is to incorporate dosing during the time of the organogenesis of the ovaries and testis and to investigate whether a chemical causes increasing toxicity and reproductive problems during subsequent generations. The dose levels and dosing period are similar to the one-generation study but continues until the third generation is weaned. The endpoints assessed in the offspring are viability, growth and histopathology of sex organs, brain and identified target organs. In a revised proposal for the OECD TG 416 Two-generation reproductive toxicity study, assessment of effects on sperm quality and oestrous cyclicity in offspring have been added (OECD 1999).

There are some limitations of generation studies concerning endpoints such as neonatal death and malformations, since the commonly used laboratory animals may eat dead or seriously malformed pups immediately after birth. An effect may, therefore, only be indicated indirectly by a smaller litter size. If only a few pups were malformed or dead, the reduction in litter size will be small compared to the normal variation in litter size and may therefore go undetected or not reach statistical significance. Preimplantation losses and resorptions are indicated in an indirect way as a decreased litter size and the sensitivity for these effects may be rather low.

Birth weight and postnatal growth can be assessed, but it is important to include variations due to different litter sizes or different sex distribution in the litters in the analysis. A change in offspring body weight is a sensitive indicator of developmental toxicity, in part because it is a continuous variable. In some cases, weight reduction in offspring may be the only indicator of developmental toxicity in a generation study. While there is always a question remaining as to whether weight reduction is a permanent or transitory effect, little is known about the long-term consequences of short-term foetal or neonatal weight changes. Therefore, weight reduction should be used to establish the NOAEL (OECD 1989).

Teratology study, prenatal developmental toxicity

This is the most used in vivo method for studying developmental toxicity. Current guidelines include those issued by USFDA, OECD and EEC (Meyer et al. 1989). In the EEC, the test is requested for new industrial chemicals (i.e. introduced after 1980) with a production volume reaching 100 tons per year.

In this protocol, pregnant animals are dosed during the period of organogenesis, since this period is most sensitive to the induction of structural, anatomical malformations. The day before expected birth the foetuses are removed and examined. The main endpoints recorded include resorptions, retarded growth, and visceral and skeletal anomalies.

The teratology test was designed to detect malformations. In the past, there was a tendency to consider only malformations or malformations and death as relevant endpoints in teratology studies. Today it is assumed that all of the four manifestations of developmental toxicity (death, structural abnormalities, growth alterations, and functional deficits) are of concern (OECD 1989). As a consequence, the name of the test is changed to "prenatal developmental toxicity test" and the exposure period is extended to the day before birth in a revised proposal for the OECD guideline 414 (OECD 1999).

The sensitivity of the test for detection of rare events such as malformations is limited, due to the use of a relatively small number of animals. With the normal group sizes of 20 pregnant rats, it is not possible to identify any increase in major malformations unless high dose levels are administered or the substance studied is highly embryo/foetotoxic (Palmer 1981). To assess the developmental toxicity of a chemical, it is therefore important to include information on other developmental effects such as minor anomalies, variations, foetal death and growth. In addition, malformations of organs developing after the period of major organogenesis, e.g. the sex organs and the brain, may at present not be detected in the teratology study. An example is the suspected endocrine disrupter dibutyl phthalate where exposure during the period of male sexual differentiation resulted in major disturbances in the morphological and functional development of the male reproductive system (Mylchreest et al. 1999).

Teratology studies are very suitable for the demonstration of intra-uterine death after implantation (resorptions). In studies where dosing is started before implantation, preimplantation loss may also be assessed.

Foetal weight can be assessed rather exact in a teratology study, but it is important to include variations due to different litter sizes or sex distribution in control vs. exposed groups in the analysis.

Developmental neurotoxicity studies, postnatal studies

A number of chemicals are known to produce developmental neurotoxic effects in humans and other species. A guideline for developmental neurotoxicity study was issued by USEPA in 1991 and a revised US guideline was proposed in 1995 (USEPA 1991). During recent years a proposal for OECD TG 426 Developmental Neurotoxicity Study has been developed based on the US guideline (OECD 1999).

Developmental neurotoxicity studies are designed to develop data on the potential functional and morphological hazards to the nervous system arising in the offspring from exposure of the mother during pregnancy and lactation.

The protocol is designed to be performed as a separate study, however, the observations and measurements can also be incorporated into e.g. a two-generation study.

The evaluation of the offspring consists of observations to detect gross neurological and behavioural abnormalities, assessment of physical development, reflex ontogeny, motor activity, motor and sensory function, and learning and memory; and evaluation of brain weights and neuropathology during postnatal development and adulthood.

The behavioural functions assessed cover many important aspects of the nervous system, however, some functions of relevance for e.g. endocrine disruption such as social interaction and mating behaviour are not included in guidelines at present.

The limitations mentioned in the section on generation studies concerning endpoints such as neonatal death, malformations, preimplantation loss and resorptions apply also to developmental neurotoxicity studies.

Other potentially relevant postnatal endpoints as e.g. kidney function, liver function and immunotoxicity, are not included in guidelines.

Reproduction/developmental toxicity screeing tests

Recently, the OECD introduced guidelines for screening tests for reproductive toxic effects, i.e. the Reproduction/Developmental Toxicity screening tests, as part of the Screening Information Data Sets (SIDS) for high production volume (HPV) chemicals (OECD 1994). The "Combined repeat dose and reproduction/developmental toxicity screening test" is a combination of a 28 days toxicity study and a reduced generation study whereas the "Preliminary reproduction toxicity screening test" is a reduced generation study.

The purpose of the test is to generate limited information concerning the effects of a test substance on male and female reproductive performance such as gonadal function, mating behaviour, conception, development of conceptus and parturition. It is not suggested as an alternative to or as a replacement for the existing test guidelines for generation and teratology studies.

The dosing of the animals is initiated 2 weeks prior to mating and continued until the end of the study on postnatal day 4. The number of animals per group is at least 10 animals of each sex and is expected to provide at least 8 pregnant females per group. Effects on fertility and birth are registered. Live pups are counted and sexed and litters weighed on days 1 and 4 postpartum.

The test does not provide complete information on all aspects of reproduction and development. In particular, it offers only limited means of detecting postnatal manifestations of prenatal exposure or effect induced during postnatal exposure.

The value of a negative study is more limited than data from generation and teratology studies due to the lower number of animals per group, the shorter period of exposure as well as the limited number of endpoints measured.

4.2.2 Comparison to assessment of effects in adult animals

Repeated dose toxicity testing in adult animals provide information on the potential for systemic toxicity by investigations of growth, clinical symptoms, haematology, biochemistry, organ weights, pathology and histopathology of organs.

Reproductive toxicity testing can provide information on a number of developmental effects, such as malformations, growth retardation, foetal and postnatal death, fertility, and functional effects on the CNS. However, the investigations of systemic effects are not similar to the repeated dose toxicity studies in adults, since haematology and biochemistry is not investigated. In addition, the investigations of organ weight, pathology and histopathology are limited to the brain, sexual organs and identified target organs. Consequently, systemic effects induced during pre- or postnatal development on e.g., liver and kidneys may not be investigated.

In most cases, the effects of chemicals have not been assessed in the two-generation study, the prenatal developmental toxicity study as well as the developmental neurotoxicity study. In these cases, some of the developmental effects mentioned above will not be covered. For example, information on developmental effects on fertility and sex organs are only provided in the two-generation study, while effects on brain development is investigated only in the developmental neurotoxicity study.

In order to have a sufficient background to determine the sensitivity of the developmental period compared to adulthood there is a need for studies where end-points are investigated similarly for both age groups. Ideally, this would require a two-generation study incorporating developmental neurotoxicity end points and supplemented with similar investigations of systemic effects in offspring as in repeated dose toxicity studies.

4.3 Test procedure for medicinal products

Guidelines on Clinical Investigation of Medicinal Products in Children have been issued by the European Committee for Proprietary Medicinal Products (CPMP 1997a). According to the guidelines, children should not be given medicines which have not been adequately evaluated for use in that age group. The rationale for this is that adequate evaluation of medicinal products for use in children cannot be achieved in adult studies because there are physiological differences between children and adults, and because children suffer from different diseases from adults, or show a different natural history for the same disease.
Furthermore, the guidelines state the scientific data required before medicinal product testing in children:
When paediatric patients are included in clinical trials, safety data from previous adult human exposure should generally be available before paediatric clinical trials are started.
In all cases, in addition to appropriate repeated dose toxicity studies, all reproductive toxicity studies and the standard battery of genotoxicity tests should be completed prior to the initiation of trials in paediatric populations. Juvenile animal safety studies should be considered on an individual basis.
The need for carcinogenicity testing should be addressed prior to long term exposure in paediatric clinical trials, taking into account the duration of treatment and/or cause for concern.

The ICH Guideline on Non-Clinical Safety Studies for the Conduct of Human Clinical Trials for Pharmaceuticals (CPMP 1997b) also points at juvenile animal safety studies when previous animal data and human safety data are insufficient.

The ICH draft Guideline on Clinical Investigation of Medicinal Products in the Paediatric Population (CPMP 1999) provides an outline of critical issues in paediatric drug development and approaches to the safe, efficient, and ethical study of medicinal products in the paediatric population. The specific clinical study issues addressed include considerations when initiating a paediatric program for a medicinal product, timing of initiation of paediatric studies during medicinal product development, types of studies (pharmacokinetic/pharmacodynamic, efficacy, safety), age categories for studies, and ethics of paediatric clinical investigation.

4.4 References

Barlow SM and Sullivan FM (1975). 6. Behavioural teratology. Teratology - trends and applications, Berry CL, Poswillo DE (eds): Springer Verlag, 103-120.

CPMP (1999). Note for Guidance on Clinical Investigation of Medicinal Products in the Paediatric Population. ICH Topic E 11. The European Agency for the Evaluation of Medicinal Products, Human Medicines Evaluation Unit, CPMP/ICH/2711/99.

CPMP (1997a). Note for Guidance on Clinical Investigation of Medicinal Products in Children. The European Agency for the Evaluation of Medicinal Products, Human Medicines Evaluation Unit, CPMP/EWP/462/95.

CPMP (1997b). Note for guidance on Non-Clinical Safety Studies for the Conduct of Human Clinical Trials for Pharmaceuticals. ICH Topic M 3. The European Agency for the Evaluation of Medicinal Products, Human Medicines Evaluation Unit, CPMP/ICH/286/95.

Hass U, Hansen EV and Østergaard G (1994). Experimental studies in laboratory animals. In Hass U et al. (eds) Occupational reproductive toxicity - Methods and testing strategy for hazard assessment of workplace chemicals. Nordic Council of Ministers and National Institute of Occupational Health, Denmark 1994.

Jakobsen BM and Meyer O. Extrapolation from in vivo/in vitro experiments to human beings. In: Andreasen PB, Brandt NJ, Cohr K-H, Hansen EV, Hass U, Hauge M, Jakobsen BM, Knudsen I, Lauritsen JG, Melchior JC, Meldgaard L, Meyer O, Olsen JH, Palludan B, Poulsen E. Embryo- foetal damage and chemical substances. Working party report, Copenhagen: Levnedsmiddelstyrelsen, 1989. 1-127. (Translated version of Danish report from 1986)

Kimmel CA and Buelke-Sam J (1985). Collaborative behavioral teratology study: Background and overview. Neurobehav Toxicol Teratol 7, 541-546.

Meyer O, Jakobsen BM and Hansen EV (1989). Identification of embryo-foetal toxicity by means of animal studies. In: Andreasen PB, Brandt NJ, Cohr K-H, Hansen EV, Hass U, Hauge M, Jakobsen BM, Knudsen I, Lauritsen JG, Melchior JC, Meldgaard L, Meyer O, Olsen JH, Palludan B, Poulsen E. Embryo-foetal damage and chemical substances. Working party report, Copenhagen: Levnedsmiddelstyrelsen, 1989. 1-127. (Translated version of Danish report from 1986).

Mylchreest E, Sar M, Catley RC and Foster PMD (1999). Disruption of androgen-regulated reproductive development by di(n-butyl) phthalate during late gestation in rats is different from flutamide. Toxicol Appl Pharmacol 156, 81-95.

OECD (1999). Guideline for the Testing of Chemicals. TG 414 Prenatal developmental toxicity study (draft proposal); TG 416 Two-generation reproductive toxicity study (draft proposal); Proposal for a new guideline 426. Developmental neurotoxicity study.

OECD (1994). Guideline for the Testing of Chemicals. TG 421. Reproduction/Developmental Toxicity Screening Test; TG 422. Combined Repeat Dose Toxicity Study with the Reproduction/Developmental Toxicity Screening Test.

OECD (1989). Report on approaches to teratogenicity assessment (draft). Room document no 31.

OECD (1983). Guideline for the Testing of Chemicals. TG 417. One generation reproduction toxicity study. TG 418. Two generation reproduction toxicity study.

Palmer AK (1986). A simpler multigeneration study. International Congress of Pesticide Chemistry 1986 (Abstract).

Palmer AK (1981). Regulatory requirements for reproductive toxicology: theory and practice. In: Developmental toxicology, Kimmel CA and Buelke-Sam J (eds), New York, Raven Press, 259-288.

US-EPA (1991). Guideline 83-6 Developmental neurotoxicity. Washington DC:US Environmental Protection Agency.