This report contains the collective views of an international group of
    experts and does not necessarily represent the decisions or the stated
    policy of the United Nations Environment Programme, the International
    Labour Organisation, or the World Health Organization.

    Published under the joint sponsorship of
    the United Nations Environment Programme,
    the International Labour Organisation,
    and the World Health Organization

    World Health Orgnization
    Geneva, 1984

         The International Programme on Chemical Safety (IPCS) is a
    joint venture of the United Nations Environment Programme, the
    International Labour Organisation, and the World Health
    Organization. The main objective of the IPCS is to carry out and
    disseminate evaluations of the effects of chemicals on human health
    and the quality of the environment. Supporting activities include
    the development of epidemiological, experimental laboratory, and
    risk-assessment methods that could produce internationally
    comparable results, and the development of manpower in the field of
    toxicology. Other activities carried out by the IPCS include the
    development of know-how for coping with chemical accidents,
    coordination of laboratory testing and epidemiological studies, and
    promotion of research on the mechanisms of the biological action of

        ISBN 92 4 154090 7  

    The World Health Organization welcomes requests for permission
    to reproduce or translate its publications, in part or in full.
    Applications and enquiries should be addressed to the Office of
    Publications, World Health Organization, Geneva, Switzerland, which
    will be glad to provide the latest information on any changes made
    to the text, plans for new editions, and reprints and translations
    already available.

    (c) World Health Organization 1984

         Publications of the World Health Organization enjoy copyright
    protection in accordance with the provisions of Protocol 2 of the
    Universal Copyright Convention. All rights reserved.

         The designations employed and the presentation of the material
    in this publication do not imply the expression of any opinion
    whatsoever on the part of the Secretariat of the World Health
    Organization concerning the legal status of any country, territory,
    city or area or of its authorities, or concerning the delimitation
    of its frontiers or boundaries.

         The mention of specific companies or of certain manufacturers'
    products does not imply that they are endorsed or recommended by the
    World Health Organization in preference to others of a similar
    nature that are not mentioned. Errors and omissions excepted, the
    names of proprietary products are distinguished by initial capital





    2.1.  Basic events underlying normal development
          2.1.1.  DNA and chromosomes
          2.1.2.  Transcriptional control
          2.1.3.  Translational control
          2.1.4.  Post-translational control and
                  significance of membrane proteins
          2.1.5.  Placentation
    2.2.  Abnormal development
          2.2.1.  Genetic influences
          2.2.2.  Nutrition
          2.2.3.  Critical and sensitivity periods during
          2.2.4.  Abnormalities of placental development
          2.2.5.  Toxicokinetics and toxicodynamics
          Placental transfer
          2.2.6.  Structure-activity relationship


    3.1.  Human studies
          3.1.1.  Measures of reproductive outcome
          Stillbirths and neonatal deaths
          Birth weight
          Congenital malformations
          3.1.2.  Prenatal diagnostic procedures
          Invasive intrauterine techniques
          Non-invasive techniques
                           (a) Ultrasound investigations
                               (i)  Visualisation
                               (ii) Fetal heart rate monitoring
                           (b) X-rays
                           (c) Analysis of maternal blood
                           (d) Analysis of maternal urine
         3.1.3.  Epidemiological methods
         Hypothesis generating (descriptive) studies
                           (a)  Case reports
                           (b)  Surveillance
                           (c)  Correlation studies
         Hypothesis testing (analytical) studies
                           (a)  Cross-sectional studies
                           (b)  Case control studies
                           (c)  Cohort studies
                           (d)  Intervention studies
         Populations at special risk
         Confounding and complicating factors

         Statistical power
         Quantifying studies
    3.2.  Experimental animal studies
          3.2.1.  Species
          3.2.2.  Dosages and dose levels
          3.2.3.  Positive control groups
          3.2.4.  Historical controls
          3.2.5.  Dosing regimen
          3.2.6.  Route of administration
          3.2.7.  Numbers of animals and statistical analyses
          3.2.8.  Observations on pregnant animals
          3.2.9.  Observations on the progeny
          3.2.10. Animal husbandry, and laboratory practices


    4.1.  Introduction
    4.2.  Behaviour
          4.2.1.  Strategy of testing
          Behavioural functions to be assessed
          Tests to be used in behavioural studies
          Choice of species
          Variables to be controlled
          4.2.2.  Methods of assessment of specific functions
          Physical development
          Reflex development
          Sensory functions
          Motor function
          Activity, reactivity, and emotionality
          Cognitive development
          Social behaviour
          4.2.3.  Relevance of behavioural studies for human risk
    4.3.  Reproduction
          4.3.1.  Normal gametogenesis and development of the genital tract
          4.3.2.  Mechanisms of abnormal development
          4.3.3.  Testing procedures
          "Fertility test" on progeny
                            following prenatal exposure
          Multigeneration studies
          Choice of species
          Doses, route, and duration of treatment
          Presentation of results
    4.4.  Transplacental carcinogenesis
          4.4.1.  Principles and mechanisms of action
          Comparative sensitivity of the adult
                            and fetal organism to carcinogens
          Dependence on the stage of
                            prenatal development
          Species and strain-specificity
          Mechanisms of organotropism
          Metabolism of chemical carcinogens in the
                            maternal organism, the placenta, and the

          4.4.2.  Relationship between teratogenesis and carcinogenesis
          4.4.3.  General principles of transplacental carcinogenicity tests
          Stages of pregnancy
          Dose and route of administration
          Evaluation of results
          4.4.4.  Methods with potential for the future
          Transplacental host-mediated cell culture
          Pre- and postnatal exposure to carcinogens
                            and promoting factors
          Intraamniotic and intrafetal injection


    5.1.  Scope of  in vitro developmental systems
    5.2.  Essential and desirable features of short-term selection tests
    5.3.  Validation of short-term selection tests
    5.4.  Available developmental systems
          5.4.1.  Whole-embryo culture (warm-blooded animals)
          Ovum maturation and preimplantation stages
          Postimplantation mammalian embryos
          Chick embryo in culture
          Chick embryo  in ova
          5.4.2.  Organ cultures
          Organ culture of limb buds
          Organ culture of the pancreas
          Organ culture of palatal shelves
          Organ culture of tooth anlagen
          Organ culture of the embryonic lens
          Organ culture of the embryonic kidney
          Organ culture of embryonic gonads and
                            accessory duct systems
          Organ culture of thyroid tissue
          Other organ culture systems
          5.4.3.  Culture of non-mammalian or non-avian embryos
          Studies on the development of fish
          Studies on the development of amphibia,
                            lurchae, sea-urchins, and other
          5.4.4.  Tissue culture systems
          5.4.5.  "Micromass" culture of dispersed cells from
                  the embryonic limb and lung
          5.4.6.  Studies with non-embryonic tissues


    6.1.  Introduction
    6.2.  Interpreting in laboratory animal studies
          6.2.1.  Endpoints
          6.2.2.  Interspecies variations
          6.2.3.  Statistical limitations
          6.2.4.  Quantitative risk assessment
    6.3.  Interpretation of human data for risk assessment






    While every effort has been made to present information in the 
criteria documents as accurately as possible without unduly 
delaying their publication, mistakes might have occurred and are 
likely to occur in the future.  In the interest of all users of the 
environmental health criteria documents, readers are kindly 
requested to communicate any errors found to the Manager of the 
International Programme on Chemical Safety, World Health 
Organization, Geneva, Switzerland, in order that they may be 
included in corrigenda, which will appear in subsequent volumes. 

    In addition, experts in any particular field dealt with in the 
criteria documents are kindly requested to make available to the 
WHO Secretariat any important published information that may have 
inadvertently been omitted and which may change the evaluation of 
health risks from exposure to the environmental agent under 
examination, so that the information may be considered in the event 
of updating and re-evaluation of the conclusions contained in the 
criteria documents. 



a,b,e  Dr V.A. Alexandrov, N.N. Petrov Research Institute
           Oncology of the USSR Ministry of Health,
           Leningrad, USSR  (Vice-Chairmanb,e)

b,e,f   Dr E.1. Anderson, Office of Health and
           Environmental Assessment, RD-689, US Environmental
           Protection Agency (EPA), Washington, DC, USA

a,b,e   Dr V.N. Anisimov, N.N. Petrov Research Institute of
           Oncology of the USSR Ministry of Health, Leningrad, USSR 

b,c,e   Professor C.L. Berry, Department of Morbid Anatomy,
           The London Hospital Medical College, London, England

e       Dr C. Cortinas de Nava, Institute of Biomedical
           Investigations, National Autonomous Universtiy of
           Mexico, Mexico

a,b,c   Dr R.L. Dixon, Laboratory of Reproductive and
           Developmental Toxicology, Department of Health and
           Human Services, National Institute of
           Environmental Health Sciences, Research Triangle
           Park, NC, USA  (Chairmanb )

e       Professor A.P. Dyban, Department of Embryology,
           Institute of Experimental Medicine, Leningrad, USSR

b,e     Dr S. Fabro, Colombia Hospital for Women, Washington,
           DC, USA

b,e     Dr W.G. Flamm, Division of Toxicology, Food and Drug
           Administration, Washington, DC, USA

b       Dr S. Frankova, Institute of Psychology, Czechoslovak
           Academy of Sciences, Prague, Czechoslovakia

b,e     Dr E. Guerrero, National Institute of Health, Bogota,

b       Dr O.P. Heinonen, Public Health Laboratories,
           Helsinki, Finland

b,e     Dr R. Jelinek, Institute of Experimental Medicine,
           Czechoslovak Academy of Sciences, Prague,

b,e     Dr J.Ev. Jirasek, Research Institute for the Care of
           Mother and Child (UPMD), Prague, Czechoslovakia

 Members (contd.)

b,c,e,f  Dr K.S. Khera, Health Protection Branch,
            Department of Health and Welfare, Tunney's
            Pasture, Ottawa, Canada  (Rapporteur)

b,e     Professor E. Klika, Department of Histology, Faculty
           of General Medicine, Charles University, Prague,
           Czechoslovakia  (Vice-Chairmanb )

e       Dr J. Ku?cera, Research Institute for the Care of
           Mother and Child (UPMD), Praque, Czechoslovakia

a,b     Professor V. Kusak, Institute of Experimental
           Medicine, Czechoslovak Academy of Sciences,
           Prahue, Czechoslovakia

b,d,e   Dr B.V. Leonov, Laboratory for Experimental
           Embryology, All-Union Research Institute of
           Obstetrics and Gynecology, Moscow, USSR

b       Professor M. Marois, Faculté de Médecine
           Saint-Antoine, Laboratoire d'Histologie, Paris,
           France  (Vice-Chairman)

b,e     Dr L. Martson, All-Union Scientific Research
           Institute of the Hygiene and Toxicology of
           Pesticides, Polymers, and Plastics, Kiev, USSR

e       Dr N. Matsumoto, Department of Public Health, Faculty
           of Medicine, University of Tokyo, Tokyo, Japan

b,e     Dr R.V. Merkurjeva, A.N. Sysin Institute of General
           and Community Hygiene, Academy of Medical Sciences
           of the USSR, Moscow, USSR

a,c,e,  Professor N.P. Napalkov, N.N. Petrov Research
           Institute of Oncology of the USSR Ministry of
           Health, Leningrad, USSR  (Chairmana )

b,c,e,f  Professor D. Neubert, Institute of Toxicology and
            Embryo-Pharmacology of the Free University of
            Berlin, Berlin 33 (West)  (Chairmane,f )

b,e     Dr A.I. Nikitin, Laboratory of Early Human
           Embryogenesis, Institute of Obstetrics and
           Gynecology of the USSR Academy of Medical
           Sciences, Leningrad, USSR

b,e     Dr A.K. Palmer, Department of Reproductive
           Toxicology, Huntingdon Research Laboratories,
           Huntingdon, England

b,e     Dr A. Pavlik, Institute of Physiology, Czechoslovak
           Academy of Sciences, Prague, Czechoslovakia

 Members (contd.)

b,e     Dr R.M. Pratt, Experimental Teratogenesis Section,
           Laboratory of Reproductive and Developmental
           Toxicology, NIEHS, Research Triangle Park, NC, USA

b       Dr J.M. Rice, Laboratory Comparative Carcinogenesis,
           National Cancer Institute, Frederick Cancer
           Research Facility, Frederick, MD, USA

b       Dr L. Rosival, Centre of Hygiene, Research Institute
           of Preventive Medicine, Bratislava, Czechoslovakia

b,c,e   Dr L. Rossi, Institute of Oncology, Genoa, Italy

b       Dr Z. Rychter, Institute of Physiology, Czechoslovak
           Academy of Sciences, Prague, Czechoslovakia

b,e     Dr J.G. Scandalios, Department of Genetics, North
           Carolina State University, Raleigh, NC, USA

b,e     Professor T.H. Shepard, Central Laboratory for Human
           Embryology, University of Washington, Seattle,
           Washington, USA  (Vice-Chairmane )

a,b,c,  Dr S. Tabacova, Institute of Hygiene and
e,f     Occupational Health, Medical Academy, Sofia,
           Bulgaria  (Rapporteur)

e       Dr H.A. Tilson, Laboratory of Behavioural and
           Neurological Toxicology, National Institute of
           Environmental Health Sciences, Research Triangle
           Park, NC, USA

b       Dr M. Vargova, Research Institute of Preventive
           Medicine, Bratislava, Czechoslovakia

b,e     Dr T. Vergieva, Institute of Hygiene and Occupational
           Health, Medical Academy, Sofia, Bulgaria

e       Dr B.R. Vitvitskaya, I.M. Sechenov First Moscow
           Medical Institute, Chair of Communal Hygiene,
           Moscow, USSR


a,e     Dr M.M. Buslajeva, Group of International Research
           Programmes, N.N. Petrov Research Institute of Oncology
           of the USSR Ministry of Health, Leningrad, USSR

a,b     Dr M. Draper, Medical Officer, International
           Programme on Chemical Safety, Division of
           Environmental Health, WHO, Geneva, Switzerland

 Secretariat (contd.)

e       Dr M. Gounar, Scientist, International Programme on
           Chemical Safety, Division of Environmental Health,
           WHO, Geneva, Switzerland

e       Dr Z.P. Grigorevskaja, Centre of International
           Projects, USSR State Committee for Science and
           Technology, Moscow, USSR

b       Dr V. Kodat, Department of Hygiene and Epidemiology,
           Ministry of Health of the Czech Socialist Republic,
           Prague, Czechoslovakia (National Focal Point for IPCS)

b,e     Dr A.I. Kucherenko, Scientific Officer, UNEP/IRPTC,
           Geneva, Switzerland

b,e,f   Dr A. Lihachev, Programme on Mechanisms of
           Carcinogenesis, IARC, Lyons, France

b       Dr M. Mercier, Manager, International Programme on
           Chemical Safety, Division of Environmental Health,
           WHO, Geneva, Switzerland

e       Dr M. Mikheev, Regional Officer for Workers' Health,
           WHO Regional Office for Europe, Copenhagen, Denmark

e       Dr L.A. Moustafa, IPCS Interregional Research Unit,
           WHO, Research Triangle Park, NC, USA

a,b,c,  Dr J. Parizek, Scientist, International Programme
e,f     on Chemical Safety, Division of Environmental
           Health, WHO, Geneva, Switzerland  (Secretary)

e       Mr I.A. Rozov, Programme Support Service, Division
           of Public Information and Education for Health,
           WHO, Geneva, Switzerland

a   Preparatory meeting, Leningrad, 28-30 June 1981.
b   First Task Group meeting, Prague, 30 November - 4 December
c   Steering Committee.
d   Alternate for Professor N.P. Napalkov at the Steering
e   Second Task Group meeting, Leningrad, 8-15 June 1983.
f   Editorial Committee.


    This monograph was prepared by an international Task Group of 
experts from thirteen countries.  The generous contributions and
the personal involvement of every member of the group during
successive stages of the preparation of the document are
gratefully acknowledged. 

    The aim of the document and the plans for its development were 
discussed at a preparatory meeting, chaired by Professor N.P. 
Napalkov with Dr R.L. Dixon, as Vice-Chairman.  The meeting, hosted 
in Leningrad on 28-30 June 1981 by the N.N. Petrov Institute of 
Oncology of the USSR Ministry of Health, was attended by 
representatives from IPCS institutions that had expressed their 
particular interest in the project.  These included, in addition to 
the Petrov Institute, the Institute of Experimental Medicine, 
Czechoslovak Academy of Sciences, Prague, the National Institute of 
Environmental Health Sciences (NIEHS), Research Triangle Park, and 
the Institute of Hygiene and Occupational Health, Medical Academy, 
Sofia.  It was agreed that these Institutions would prepare four 
background papers for the first plenary meeting of the Task Group. 

    The first meeting of the Task Group took place in Prague, 
Czechoslovakia, hosted by the Ministry of Health of the Czech 
Socialist Republic, on 30 November - 4 December 1981.  Dr R.L. 
Dixon was elected Chairman of the meeting and Dr V. A. Alexandrov, 
Professor M. Marois, and Professor E. Klika, Vice-Chairmen.  Dr K. 
S. Khera and Dr S. Tabacova were designated Co-rapporteurs for this 
and subsequent Task Group meetings.  The experts identified and 
discussed in detail the problems to be covered in the monograph and 
established three working groups for further development of the 
draft sections dealing with prenatal and postnatal manifestations 
of embryotoxic effects, and with non-mammalian and  in vitro tests.  
Professor C.L. Berry, Dr L. Rossi, and Professor D. Neubert were 
elected leaders of the respective working-groups.  The Steering 
Committee, consisting of the Chairmen of the preparatory and first 
Task Group meetings, leaders of the working groups, and the Co-
rapporteurs, co-ordinated the work of the working groups, after the 
first Task Group meeting, and prepared a consolidated draft, which 
was sent for comments to all the members of the Task Group.  
Comments were also received from the OECD Chemicals Division. 

    A revised draft was prepared by Professor C.L. Berry, with the 
assistance of the Steering Committee, and submitted to the Task 
Group before its second meeting, held in Leningrad on 8-15 June 
1983.  Professor D. Neubert was elected Chairman of the Leningrad 
meeting, Dr V. A. Alexandrov and Professor T.H. Shepard, Vice-
Chairmen, and Dr K. S. Khera and Dr S. Tabacova again acted as Co-

    The second Task Group meeting, hosted by the Petrov Institute 
and the USSR Commission for UNEP, completed the document, including 
those sections dealing with human risk assessment (on the basis of 
a draft prepared by Dr R.L. Dixon and Professor D. Neubert), the 
conclusions and recommendations of the Task Group, and explanation 
of terms used in the document.  A small Editorial Committee 
consisting of Professor D. Neubert, Chairman, Dr E.L. Anderson, 
leader of the working group responsible at the Leningrad meeting 
for the human risk assessment section, and both Co-rapporteurs, was 
designated by the Task Group to finalize the document. 

    The International Programme on Chemical Safety would like
to express deep gratitude to all the members of the Task Group
for their work.


    This document is intended to aid in the design and assessment 
of studies concerned with exploring the association between 
exposure to chemicals during pregnancy and defective development.  
Personal tragedies are involved when defects are present at birth 
or appear later in life and the need for improved methods for the 
detection of embryo/fetal toxic agents, the assessment of health 
risks, and the prevention of unfavourable outcomes of pregnancy, is 

    It is not possible, at the moment, to give a precise estimate 
of the extent to which chemicals contribute to the induction of 
abnormalities in human development.  Though hundreds of chemicals, 
which have been listed by Shepard (1983), have induced embryotoxic 
effects in laboratory animals, very few have been shown to be 
teratogenic (for definition see section 9) in human beings.  
Differences between animals and man in metabolism, inherent 
sensitivity, and levels of exposure may be responsible for this 
apparent discrepancy.  Among the problems involved in conducting 
epidemiological studies is the detection of chemically-induced 
anomalies against a background of sporadic defects.  The size of an 
epidemiological study necessary to detect changes in the rates of 
some abnormalities, is so great that such studies are seldom 
conducted.  Smaller studies may result in inaccurate risk 

    There are many difficulties involved in extrapolating human 
risk from animal studies.  For example, if a chemical, tested in 
animals at high doses, alters maternal homeostasis, it is possible 
that the conclusions drawn will be inappropriate for human beings 
exposed to lower levels of the same chemical and that such studies 
may over-predict the incidence of teratogenic events. 

    Gross structural defects are an obvious area of concern. 
However, in recent years, it has become clear that consideration of 
developmental toxicity following prenatal exposure must be expanded 
to include chemically-induced embryolethality, reduced fetal 
growth, and functional alterations that may not be expressed until 
late after delivery.  Some chemicals may decrease fertility by 
causing a loss of the early conceptus, even before pregnancy is 

    The processes of cellular and morphogenetic differentiation, 
though poorly understood, offer many potential targets for toxic 
chemicals.  The chemicals that can alter these processes have the 
potential to induce adverse developmental effects.  Thus, increased 
understanding of the underlying processes will aid in using 
laboratory studies to identify potential human health effects and 
to quantitatively estimate risks resulting from exposure to 

    Current test methods are generally carried out using whole 
animals.  Although the number of animals tested might be reduced 
and the studies designed more efficiently, there is no substitute 

for animal tests, at present.  In the future, the potential of 
biochemical, cellular, and other  in vitro approaches might be more 
effectively exploited. 

    This document includes a description of the use of laboratory 
data in defining the potential embryotoxic hazards of chemicals and 
methods of assessing the human risks associated with occupational 
or environmental exposure to such chemicals.  Three major areas are 
covered including: prenatal toxic manifestations, postnatal 
manifestations, and short-term tests.  Prenatal toxic 
manifestations are discussed in relation to human epidemiological 
studies and laboratory studies on experimental animals.  In the 
discussion on postnatal manifestations, the emphasis is placed on 
test methods for the assessment of alterations in behaviour, 
reproduction, and of transplacental carcinogenesis following 
gestational chemical exposure.  Short-term tests are discussed in 
the context of understanding the mechanisms of developmental 
toxicity and their potential for future application in 
developmental toxicity testing. 

    In each area, basic data and processes were considered by the 
Task Group in relation to three questions, i.e., 

    (a)  What is the value of currently used toxicity tests
         for predicting human risk?

    (b)  What are the difficulties involved in applying test
         results to the estimation of human health risks and

    (c)  How can testing strategies be improved?

    The document is divided into nine sections, which differ in 
structure.  The Introduction is followed by Section 2, which 
provides background information on the mechanisms of development.  
Section 3 is devoted to methods of assessing prenatal toxic 
manifestations and represents the overall views of the Working 
Group, in this complex field.  Sections 4 and 5 include detailed 
discussions on postnatal events and short-term tests, respectively, 
since these sections present a new field and the Task Group wished 
to review the potential of available tests.  Section 6 on human 
risk assessment, was considered by the Task Group to be 
fundamental.  The Task Group's conclusions are given in Section 7 
and their recommendations in Section 8.  Section 9 comprises an 
explanation of terms used in the document. 

    The Task Group, aware of the existence of several guidelines 
published by national or international bodies (FDA, 1966, 1970; 
NHW, 1973; NAS, 1977; US EPA, 1978; OECD, 198l; CMEA, 1982), wish 
to state that the present document is not intended to replace, 
complement, or discuss these guidelines. 

    For detailed information on methods of studying prenatal 
toxicity, reference can be made to a number of publications 
including those of Nishimura et al. (1968), Shepard et al. (1975), 

Ebert & Marois (1976), Neubert et al. (1977b), Wilson & Fraser, 
(1977-78), Benesova et al. (1979), Klingberg et al. (1979), Porter 
& Hook (1980), and Kimmel & Buelke-Sam (198l). 

    This document primarily deals with chemical exposures during 
pregnancy and should be useful to all those concerned with the 
evaluation of chemical safety.  Evaluation of the effects of 
exposure to drugs is not within the scope of this document and has 
been addressed in other publications (WHO, 1967; CIOMS, 1983). 

    There are a number of reference sources dealing with chemicals 
reported to be teratogenic for man and laboratory animals 
(Schardein, 1976; Heinonen et al., 1977; Shepard, 1983).  The 
embryotoxic effects of chemicals are regularly assessed in the IPCS 
Environmental Health Criteria documents, available from the World 
Health Organization.  The Environmental Teratology Information 
Center (ETIC) operated by the Oak Ridge National Laboratory (ORNL) 
maintains computer-stored teratological scientific literature on a 
large number of chemicals, listed according to CAS (Chemical 
Abstract Service) registry numbers.  Advice on access to this 
information is available from the International Register of 
Potentially Toxic Chemicals, United Nations Environment Programme 
(UNEP), Geneva. 

    The International Clearinghouse for Birth Defects Monitoring 
Systems covers nineteen monitoring programmes from various 
countries.  The results are published, annually, by the Swedish 
National Board of Health and Welfare, Stockholm and, quarterly, by 
the March of Dimes Birth Defects Foundation, White Plains, New 


    The purpose of this section is to give an outline so that those 
working in different fields can see the background against which 
new developments might occur.  For more detailed discussions of 
processes involved in normal and abnormal development, see, for 
example, Hicks & D'Amato (1966), Thomas (1968), Langman et al. 
(1975), Scandalios (1979), Page et al. (198l), and Krowke & Neubert 

2.1.  Basic Events Underlying Normal Development

    Embryogenesis is a finely balanced programme of cellular 
events, including proliferation, migration, association, 
differentiation, and cell death, precisely arranged to produce 
tissues and organs from genetic information, present in each 
conceptus.  Underlying the morphological development of the embryo 
is a progressive unfolding of biochemical potentialities, 
determined by temporally- and spatially-regulated transcription and 
translation of genetic messages. 

    Embryogenesis involves complex interactions occurring in both 
time and space.  In the earliest stages, rapid cell multiplication 
is the rule.  The cells of the early embryo have vast developmental 
potential, depending on their relative positions in the embryonic 
mass.  The development of primordial tissue depends strongly on the 
interactions of adjacent cell groups, apparently mediated by 
endogenous chemicals or growth modifiers.  Complex processes of 
cell migration, pattern formation, and the penetration of one cell 
group by another, characterize the later stages of organogenesis 
(Johnston & Pratt, 1975).  Final morphological and functional 
development occurs at different times in different species and is 
sometimes completed after birth. 

2.1.1.  DNA and chromosomes

    DNA is the target of several mutagenic, growth-inhibiting, and 
carcinogenic chemicals.  The eukaryotic chromosome is a complex 
structure mainly composed of DNA and protein (Lewin, 1980).  A 
large part of the chromosomal protein is made up of a small group 
of basic proteins known collectively as histones, which may be 
important in controlling the functional state of DNA.  All 
eukaryotic cells contain essentially the same types of histones.  A 
chromatin fibre contains a double strand of DNA and associated 
proteins.  Between cell divisions, chromatin is normally dispersed.  
Most chromatin can undergo the transition between dispersed 
(euchromatic) and condensed (heterochromatic) states.  The 
information encoded in DNA is transferred by a multistage process 
to RNA (see below). 

    The amount of a particular mRNA, and hence the rate of specific 
protein synthesis, can be regulated by controlling the number of 
genes specific for its production.  Gene amplification is one means 
by which a cell can produce large quantities of a specific gene 

    Microscopically visible excesses, deficiencies, or 
rearrangement of chromosomes and chromatids, occurring 
"spontaneously" or induced by chemicals, have been found to be 
associated with human developmental defects.  Similar cytogenetic 
changes may result from viral infections or irradiation.  Many 
cytotoxic agents act by inhibiting the synthesis of DNA, thereby 
slowing or preventing mitosis. Other chemicals interfere with the 
polymerization of tubulin into the microtubules of the spindles, 
which prevents cell division by arresting the formation of the 
mitotic spindle. 

2.1.2.  Transcriptional control

    Each of the two steps (transcription and translation) involved 
in the synthesis of a protein is controlled.  In eukaryotic cells, 
transcription occurs in the nucleus and translation outside the 
nucleus.  These processes may also occur in mitochondria.  
Differential gene activation or transcriptional control is the 
principal mechanism for gene control in prokaryotic cells.  The 
finding that three distinct forms of RNA polymerase (EC 
are present in eukaryotic cells, transcribing different sets of 
genes, points to transcriptional control as a predominant mechanism 
regulating gene expression in eukaryotic cells. 

    Transcriptional controls are well documented for highly 
specialized single-copy genes, the products of which make up a 
large proportion of cellular mRNA and protein.  Examples are genes 
for globin, ovalbumin, and metallothionein. 

    Eukaryotic genes seem to be transcribed from large regions of 
DNA, producing RNAs with up to about 20 000 nucleotides, which are 
then processed into smaller mRNAs, before leaving the nucleus.  The 
process is not a conservative one, however, for a considerable 
fraction of the RNA is degraded, without ever leaving the nucleus.  
This degraded RNA represents linkage regions between structural 
genes.  The extent of gene expression emerges as the controlling 
factor in differentiation, as different gene products are produced 
at specific developmental stages or in response to hormones. 

2.1.3.  Translational control

    The generally long lifetime of eukaryotic mRNA might imply that 
translational control is especially important; merely turning off a 
gene does not stop the synthesis of its enzyme, as long as the 
appropriate mRNA is intact and available for translation. 

    The translation of mRNA after the fertilization of an oocyte is 
an example of translational control that seems to involve 
activation of a stable mRNA.  The advantage of translational 
control is the speed with which protein synthesis can be turned on 
and off.  The longer an mRNA can survive intact, the more important 
it becomes in controlling the rate of its translation.  The 
lifetimes of mRNAs are highly variable.  Enzymes that increase or 
decrease in quantity in response to external signals have mRNA with 
shorter lifetimes than enzymes that are not closely regulated.  The 

half-lives of most mRNAs in eukaryotic cells range from about an 
hour to several days.  The extreme case of stability of mRNA is 
found in an unfertilized egg, in which masked mRNA may remain 
untranslated for very long periods, even years, until fertilization 

    The amount of translation of a given message can be controlled 
by regulating the lifetime of the mRNA.  However, little is known 
of the reason why some mRNAs are more stable than others or how 
exogenous chemicals affect this stability. Complementary sequences 
within the same mRNA molecule result in a specific folding, 
consisting of loops and hairpins.  An alternative scheme suggests 
that the number of times some mRNAs are translated, is 

2.1.4.  Post-translational control and significance of 
membrane proteins

    Post-translational control involves the modification of 
synthesized proteins (e.g., sub unit interactions, 
compartmentation, turnover, conjugation of prosthetic groups, 

    The plasma membrane of eukaryotic cells is composed of lipid 
glycoproteins.  Controlled membrane permeability is essential to 
cell maintenance and survival.  The plasma membrane also serves as 
a site for receptors with a high affinity for specific hormones.  
Interactions of hormones with receptors trigger the chain of 
molecular events necessary for gene expression and protein 
synthesis.  Furthermore, plasma membranes provide cell-specific 
antigens, and differentiated cells within an organ provide organ-
specific antigens. Antigenic specificity is essential for the cell 
recognition required in morphogenesis and embryogenesis.  
Therefore, any alteration of plasma membranes by toxic chemicals 
can be expected to alter permeability, receptors, and immunological 
cell specificity, both qualitatively and quantitatively, as well as 
cell-cell interactions. 

2.1.5.  Placentation

    The placenta is a temporary organ that establishes a functional 
union between mother and fetus.  This unique organ transfers all 
the nutrients needed for development, eliminates fetal metabolic 
waste, synthesizes hormones essential to pregnancy, and carries out 
many other anabolic and catabolic functions.  Placental morphology 
varies with species and with the stage of fetal development. 

    During and immediately after implantation, before the placenta 
has developed, the embryo depends on histiotrophic nutrition, 
consisting of digested endometrial cells mixed with a secretion of 
the endometrial glands.  The histiotrophic material resorbed by the 
trophoblast reaches the embryonal anlage by diffusion.  Another 
source of early embryonal nutrition is the content of the yolk sac.  
In mammals, the significance of the yolk sac varies considerably.  
It usually functions as a non-selective structure allowing transfer 

to the fetus of most chemicals resorbed by the trophoblast from the 
chorion and chorionic cavity until allantoic (embryochorionic, 
feto-placental) circulation is established. The yolk sac cavity, 
enclosed by endodermal cells and visceral mesoderm, is of 
nutritional importance for the embryo for up to 9 1/2 days in the 
hamster (Boyer, 1953), up to 10 days in the rat, and for 
approximately one month in man (Langman, 1969).  Though its role 
becomes less significant with the progress of pregnancy, the yolk 
sac in the rabbit and rodent species continues to function in the 
maternal-fetal transfer of chemicals until later in pregnancy 
(Lambson, 1966; Seibel, 1974), when the yolk sac is separated from 
the gut. 

    The embryonal yolk sac mesoderm is the source of the first 
haematopoietic tissues supplying the blood cells of the embryo-
chorionic circulation.  During implantation, as the trophoblast 
penetrates into the endometrium, maternal capillaries come into 
contact with the trophoblastic shell, and consequently open into 
the trophoblastic lacunae.  As the trophoblastic shell changes into 
the villous chorion, maternal blood begins to circulate within the 
intervillous space. 

    The chorion is vascularized by allantoic vessels.  As reported 
by Jirasek (1980), the human embryochorionic circulation begins 
about day 28-30, there being a gap of about 14 days between the 
time when the maternochorionic (maternal) circulation and the 
embryochorionic (fetal) circulation are established. 

    The allantois is an endodermal diverticulum from the yolk sac 
(later hind gut) into the connecting stalk located between the yolk 
sac, the amnion, and anchored to the chorion.  The allantois is 
accompanied by umbilical vessels connecting the intraembryonal 
vessels with the chorionic vessels.  As the amnion expands, the 
yolk sac duct, and the connecting stalk containing the allantois 
and umbilical vessels are pushed together, giving rise to the 
umbilical cord.  After establishment of the embryochorionic 
circulation, embryonal tissues are nourished from the embryonal 

    Placentation is highly species-specific with important 
physiological and anatomical peculiarities in different mammals.  
The relationships between the structure and function of placental 
membranes have not been satisfactorily clarified. 

2.2.  Abnormal Development

    The complicated series of developmental events offers a variety 
of time-specific targets for toxicity.  Even temporary retardation 
in the growth of one group of cells may have serious consequences 
for overall development, because each step in embryogenesis may 
depend on a previous one and the development of numerous tissues 
and organs is interrelated. 

    The development of an organism concerns several levels of 
biological organization.  The targets for chemical agents occur at 
the molecular level; chemicals may interact with DNA, RNA, membrane 
constituents, enzymes, receptors, etc.  It does not mean, however, 
that the primary events resulting from these interactions need 
either be specific for the cells of developing systems, or lead 
inevitably to some kind of embryotoxicity manifestation.  In 
teratogenicity studies, the outcome of the primary events depends 
on the properties of the affected morphogenetic systems (Jelinek & 
Rychter, 1979). Occurrence of defects depends, in principle, on the 
coincidence of the critical and sensitivity periods (defined in 
Section 9).  Some factors can cause embryonic death directly by 
interrupting the basic life functions of the conceptus, e.g., blood 

    Teratogenic research was stimulated by the observation that 
congenital malformations in human beings could be induced by 
maternal exposure to thalidomide, pelvic irradiation, or by 
maternal infection with rubella. 

    Teratogenesis, mutagenesis, or carcinogenesis may occur 
spontaneously or may be induced by external physical or chemical 

    Teratogenic mechanisms have mainly been studied in laboratory 
animals; human data have been obtained from either case reports or 
epidemiological studies.  The mechanisms of action of embryotoxic 
agents are still poorly understood, even though rapid advances are 
being made.  Only by increasing knowledge concerning the 
biochemical and biological events in normal embryogenesis can the 
processes of teratogenesis be better understood. 

    Chemicals that interfere directly with embryonic or fetal 
development at exposures lower than those that cause apparent 
maternal toxicity are most likely to be of concern.  Many of the 
most widely-recognised embryotoxic or teratogenic chemicals meet 
this criterion of selective action; the difference between doses 
that induce malformation and doses that cause maternal death or 
serious intoxication, may be considerable (Khera, 1983).  
Embryotoxic evaluation of such chemicals requires more serious 
attention than evaluation of chemicals that induce toxic effects in 
the embryo only at doses toxic to the mother.  With the latter, 
recognition of maternal toxicity aids in controlling the risk of 
embryotoxic dosages.  There are cases in which both embryonic and 
maternal effects in experimental animals are elicited within the 
range of human exposure levels (for instance with alcohol or methyl 
mercury) and the dual danger to mother and fetus should be 

    Many chemicals, including teratogens, when administered at high 
doses or very early in embryonic development, can cause embryonal 
or fetal death followed by resorption or abortion of the fetus.  
However, substances that can kill fetuses selectively in 
experimental animals do not necessarily induce teratogenic effects 
in the survivors.  The association between intrauterine death, 
spontaneous abortion, still birth, and congenital malformation in 
human beings needs to be explored further. 

    Although very detailed studies of the teratogenic effects of 
drugs have been carried out on developing mammals, birds, and other 
submammalian species, relatively little information is available 
regarding biochemical changes in the human embryo.  Any knowledge 
of these effects is the result of limited clinical observations 
made some time after accidental or incidental ingestion of 
chemotherapeutic or other agents. As yet, there is not sufficient 
knowledge of the associations between the various manifestations of 
developmental toxicity or, more precisely, the reasons for the lack 
of consistency in the associations.  This may partly be because 
information on the biochemical changes in the human embryo is 
lacking, thus leaving a gap in the link with the apparent final 
common pathway that results in a marked reduction in the number of 
cells, as suggested by Connors (1975), or cell products, which then 
fail to support either full morphogenesis or complete functional 

2.2.1  Genetic influences

    As previously mentioned, morphogenesis involves precisely 
coordinated interactions of cell groups.  Each embryonic cell 
depends entirely on its genome and signals coming from the 
exterior.  Both of these can be disturbed by chemical agents 
transported and transformed between entry into the maternal 
organism and arrival at the target cells.  Thus, the response to 
embryotoxic agents strongly depends on embryonic and maternal 

    Genetic alterations can be the result either of hereditary 
defects, already present in the zygote, or of an acute effect on 
the embryonic genome.  However, the importance of somatic mutations 
in development is not clear.  Chromosomal aberrations contribute 
substantially to early embryonic loss and may persist in the form 
of proliferation mosaics. 

    The mechanisms that control the timing of gene expression or 
repression during development and result in the orderly development 
of the embryo, are not understood.  It is not known how the 
products of gene expression control development, though it is 
obvious that an altered gene product may affect developmental 
selectivity.  The achondroplastic gene induces dwarfism of a 
particular type; abnormal meiotic divisions with the production of 
aneuploid cells may result in embryonic death or phenotypic 
anomalies such as Down's syndrome or Patau's syndrome.  If the 
genetic defect is present in the germ cells (as with defects 
characterized by Mendelian inheritance such as fibrocystic 
disease), then it will be transmitted to future generations.  
However, the majority of human malformations are not inherited in 
this way. 

    The considerable variation in the susceptibility of individuals 
to teratogens, is well documented for animals. For example, some 
strains of mice are markedly more susceptible to cleft palate 
induction by glucocorticoids, than others.  Mice, in general, are 
more susceptible than other species, but, results of studies with 

highly potent fluorinated corticosteroids indicate that it would be 
unwise to presume that other species including rats and primates 
are completely resistant to cleft palate induction.  The 
differences in susceptibility could be related to such 
toxicokinetic features as the rate at which the corticosteroid is 
absorbed, distributed, eliminated, or transformed by the maternal 
animal or its rate of passage across, or biotransformation by, the 
placenta.  The nature of the chemical's interaction within the 
cells and tissues of the embryo (toxicodynamics) is also an 
important determinant of toxicity.  The basic rate of embryonic 
development may be important.  There is evidence that the number of 
glucocorticoid receptors in the developing palate of sensitive mice 
is higher than the number found in resistant strains. However, 
further studies are needed to establish a cause and effect 

    In man, epidemiological studies on the families and relatives 
of the malformed children have shown evidence of the influence of 
genetic factors in the human response to teratogens.  The 
pathogenetic mechanism appears to be the interaction of 
environmental factors with several genes.  This hypothesis, the 
polygenic or "multi-factorial" theory, has been described in detail 
in the human setting by Edwards (1969) and Carter (1976).  For 
discussion concerning other factors causing malformations see 
Neubert et al. (1980) and Berry (198l). 

    In animals, the interaction of both genetic and environmental 
factors determines the "liability" of a given congenital 
malformation to become overt.  Genetic predisposition is usually 
due to the additive effects of many genes.  Thus, the various genes 
regulate the expression of heritability and in doing so, are 
influenced by the environment. 

2.2.2.  Nutrition

    Malformations can be induced experimentally by nutritional 
imbalance.  Thus, the relationship between nutrition and human 
congenital malformations needs further study.  Contrary to the 
long-held view that the fetus has first claim on available 
nutrients, vitamin deficiencies or a reduced availability of trace 
metals, such as zinc (Hurley & Shrader, 1972), appear to 
selectively damage the animal fetus.  Teratogenic effects may be 
caused by deficiencies that are too slight to harm the mother and 
may be a reflection of the nutritional requirements of the growing 
fetus and its lack of stored reserves.  On the other hand, general 
malnutrition has not been found to induce teratogenic effects. 

    The results of recent studies in the United Kingdom have 
suggested that occult vitamin deficiency may be a factor in the 
pathogenesis of neural tube anomalies.  This finding tends to 
emphasize that the nutritional status of the mother is an important 
consideration in developmental toxicity (Smithells et al., 1982, 

2.2.3.  Critical and sensitive periods during development

    It is widely accepted that the developing conceptus is 
particularly sensitive to toxic agents during certain periods, 
generally related to the development of particular organ systems or 
types of cells.  A critical phase for the induction of structural 
malformations usually occurs during the period of organogenesis.  
In the rat, for example, this critical period extends from about 
day 6-15 (day of finding sperms in vaginal smear counted as first 
day of pregnancy) of the 22-day gestation period.  In man, 20-70 
days after conception is perhaps the comparable period.  It may be 
unwise to rely absolutely on these time periods. 

    Before implantation, specific malformations cannot be induced 
in the conceptus.  Malformations induced with actinomycin D or 
cycasin administered during the preimplantation periods are 
probably due to the persistance of these compounds in the maternal 
or fetal organism until a later stage of development. 

    With physical agents such as X-rays, exposure can be limited 
exactly to a period of seconds or minutes.  However, with chemical 
teratogens, the situation is more complicated because of the time 
courses of absorption, metabolism, and excretion.  In addition, the 
actual (proximate) teratogen may be a metabolite rather than the 
compound administered. 

    If the time in gestation, when the differentiation of a 
particular organ is complete, is known with certainty, then a 
teratogen must be present prior to or at that time, if it is to be 
a causative agent of malformation. 

    It might be thought that all chemicals operating at a 
particular critical time in organogenesis would induce the same 
pattern of abnormalities, but this is not the case. Differences 
between species in fetal development, toxicokinetics, and 
toxicodynamics account partly for this variability in response 
caused by interplay of critical and sensitive periods. 

    Other forms of developmental toxicity may have the same or 
different periods of peak susceptibility.  For example, functional 
(behavioural) deficits and other latent defects may originate more 
commonly during the period of cellular differentiation that follows 
organogenesis.  This susceptibility extends into the period of 
postnatal development to a greater or lesser extent, according to 

2.2.4.  Abnormalities of placental development

    Certain diseases (syphilis, toxoplasmosis, or viral infections) 
may affect the embryo indirectly by damaging the placenta.  Certain 
placental anomalies (placenta praevia, accreta) are associated with 
a relatively poor pregnancy outcome, but few cases of abnormal 
development, entirely due to placental causes, can be singled out 
(Benirschke, 1975). However, the placenta is a crucial organ in 
development, and abnormal fetal growth and development may be 

related to placental insufficiency.  It is also active 
metabolically and, for this reason, it is important, both in the 
biotransformation of chemicals and in determining how chemicals are 
distributed between fetus and mother. 

2.2.5.  Toxicokinetics and toxicodynamics

    The term toxicokinetics is used in this document to emphasize 
the fact that the toxicity of environmental chemicals is the 
subject of concern in this document. 

    The rates of absorption, distribution, and elimination of any 
chemical are interrelated and collectively determine the level and 
duration of the presence of the compound or its metabolites in the 
maternal and fetal body compartments (Mirkin & Singh, 1976).  The 
toxicokinetics of a chemical in the maternal organism determine, to 
a large extent, the availability of the chemical for the embryo, 
though the placenta may exert modifying effects. 

    The kinetics of a chemical vary with species, the route of 
administration, length of exposure, the physiological status of the 
animal, etc.  Differences in kinetics, especially in 
biotransformation, account for a considerable number of species 
differences in response to chemicals.  It has been, and continues 
to be, suggested that prior performance of toxicokinetic studies 
would make it possible to choose the most appropriate species for 
toxicity testing.  But toxicokinetic similarity is only one of 
several aspects that have to be considered in the choice of a model 

    Biological responses in tissues are initiated by an interaction 
between a chemical and a receptor, as, for example, interactions 
between enzymes and substrates, oxygen and haemoglobin, 
catecholamines and neuroreceptors.  Agents such as pharmaceutical 
compounds and pesticides are often designed to compete with the 
natural substance for the receptor or to prevent a response by 
altering or blocking the receptor.  This process may be labelled 
pharmacodynamic, if desirable, and toxicodynamic if undesirable.  
It is generally assumed that pharmacodynamics are similar for most 
species and while this may be the case for individual receptors, 
species-and strain-differences can arise, because of the different 
numbers of receptors present during development.  Placental transfer

    Concentration gradients and physicochemical factors, such as 
lipid solubility, relative molecular mass, and protein-binding, 
determine the rate at which chemicals cross biological membranes 
and also largely determine the rate of placental transport.  The 
binding of a substantial fraction of a chemical to maternal plasma 
proteins will reduce the concentration gradient of free (unbound) 
molecules and thereby diminish the rate of their passage across the 
placenta.  The fetal plasma protein may act as a sink for the 
chemical molecules after they cross the placenta; therefore, a 
relatively large amount of a chemical may have to be transferred, 

against a relatively low concentration gradient, in order to 
establish maternal-fetal equilibrium.  If the protein-binding 
capacity in the fetal compartment is smaller than that in the 
maternal organism, the free concentration of the xenobiotic may be 
higher in the fetus. 

    Substances that are very lipid-soluble (e.g., the anaesthetic 
gases) diffuse across the placental membranes so rapidly that their 
overall rates of equilibration are probably limited only by organ 
blood flow.  Though some steroids cross the placenta readily, their 
glucuronide conjugates penetrate at much slower rates.  In general, 
the placenta does not seem to represent a barrier to chemicals with 
relative molecular masses of approximately 300 or less, but 
chemicals of higher relative molecular mass appear more slowly in 
fetal blood after maternal exposure, because the rate of transfer 
is lower.  However, even substances of comparatively high relative 
molecular mass, such as antibody globulins, viruses, cellular 
pathogens, and erythrocytes, pass from the maternal circulation 
into the fetus.  Biotransformation

    A number of chemicals are biotransformed into metabolites by 
the enzyme systems in the mother, placenta, or fetus.  The exact 
role of genetically-determined biotransformation enzymes in 
influencing the expression of teratogenicity has not been 
ascertained for most substances.  Important enzymes, in this 
respect, are localized in the endoplasmic reticulum and 
constitute a family of cytochrome  P-450-dependent
mono-oxygenases.  Environmental chemicals that enhance or
inhibit the activities of these mono-oxygenases (and thus the
rate of metabolism of certain chemicals), are referred to,
respectively, as inducers or inhibitors.  The inducers include
phenobarbital, chlordane, DDT, polycyclic aromatic
hydrocarbons, flavons, dioxins, indoles, and polyhalogenated
biphenyls, many of which are combustion products, industrial
chemicals, and pesticides.  The inhibitors include carbon
monoxide ( in vitro), imidazole, methylene blue, aniline,
amines, and some more or less specific inhibitors.  The
ability of chemicals to induce or inhibit the activity of the
microsomal enzyme systems is an important factor to be taken
into account in devising experimental protocols.

2.2.6.  Structure-activity relationship

    The invention of new pharmaceutical compounds or pesticides is 
based on the knowledge of chemical groups that will interfere with 
specific enzyme reactions or act on specific receptors.  The 
development of new chemicals is intended to enhance the desired 
activity and reduce undesired effects.  However, a similar strategy 
cannot yet be extended to developmental toxicity, since it is not 
possible to predict, with certainty, the effects of chemicals on 
developmental processes, on the basis of their structure alone. 


3.1.  Human Studies

    Human data are considered to be of great potential value. The 
present casual attitude to documenting abortions and examining the 
products of conception is a serious handicap to acquiring knowledge 
concerning the influence of environmental chemicals on human 
development.  Epidemiological studies are important for identifying 
environmental hazards and for providing human data as a basis for 
validating animal models that might predict human disease before it 
occurs.  Therefore, well-designed studies are extremely valuable 
and should be encouraged. 

    When it is suspected that the environment is contaminated from
a known or unknown source, the exposed area and population should 
be monitored.  The chemical should be identified and the levels of
exposure quantified; contamination of the soil, air, water, food, 
animals, and people should be assessed.  Specimens from these 
sources should be retained for future study.  Information on
exposure levels and/or the body burden of exposed persons can
sometimes be obtained by analysis of body fluids, such as blood,
urine, or breast milk, biopsy or autopsy tissues, placental
tissues, hair, or extracted teeth. 

3.1.1.  Measures of reproductive outcome

    Because of the complex integrated nature of reproduction in all 
mammalian species, abnormal reproductive performance provides a 
sensitive indicator of chemical effects. Widely-used measures of 
reproductive performance include abortion, still birth, birth 
weight, and peri- or neonatal death.  These measures are discussed 
below.  Abortion

    An area of concern for the Task Group was the lack of general 
awareness of present data on human reproductive wastage, even among 
those who deal with the effects of chemicals on development.  In 
addition, the Task Group considered that the acquisition of new 
data in this field would be of major importance in estimating 
putative risks from chemicals. 

    "Spontaneous" or induced impairment in development often 
results in embryonic death during early pregnancy.  A spontaneous 
wastage of early conceptuses (15-20% abortion rate) is documented 
by clinical observations of embryonic loss after implantation.  
However, the results of studies using the Beta-HCG radioimmunoassay 
for detecting trophoblast differentiation and implantation, as 
well as observations on the development of transplanted pre-
implanted human embryos, have indicated that, even in a group of 
women who might be expected to have good reproductive performance, 
more than half of early conceptuses are lost within the first weeks 
of gestation (Miller et al., 1980; Edmonds et al., 1982; Lopata et 
al., 1982). 

    Important in its own right, the death of the conceptus becomes 
even more relevant, because of its frequent association with 
congenital malformations.  Thus, while 2.5-3% of live births show 
major malformation, numerous studies have shown more than 90% of 
conceptuses with anomalies as diverse as cebocephaly, Turner's 
syndrome, and neural tube defects are aborted in early pregnancy 
(Table 1, Berry, 1981).  It is becoming clear that many abnormal 
embryos are aborted.  In spontaneous abortions occurring in the 
first trimester, about 60% exhibit chromosomal aberrations, mostly 
trisomy, monosomy, or triploidy (Boue et al., 1975). 

    Nishimura (1969) reported that 3.7 - 4.7% of nearly a thousand 
late embryos and early fetuses obtained from surgical abortions had 
externally visible malformations.  This relatively high background 
of "spontaneous" abnormalities makes it difficult to detect a small 
increase in the overall incidence that might be caused by a single 

    Curetted materials of 5-12 weeks gestation were collected by 
Watanabe (1979) from a private hospital in Nigata city, in 
collaboration with attending physicians, and the chromosomes of 
embryonic cells and chorionic villi analysed.  All of the material 
was taken from women whose pregnancies had been terminated for 
socio-economic reasons only.  During May 1977, 1250 cases were 
analysed from which 80 chromosome abnormalities were detected, 
giving an incidence of 6.4%.  An increased incidence was found with 
increasing maternal age, particularly in the occurrence of complete 

Table 1.  Malformations in aborted fetuses and newborn
                                   Prevalence per 1000  
Malformation                 At abortion  At birth  Loss %
Neural tube                  13.1         1.0       92
Cleft lip and palate         24.4         2.7       89
Polydactyly                  9.0          0.9       90
Cyclopia and cebocephaly     6.2          0.1       98
From: Nishimura (1969) and Berry (1981).

    In order to ascertain the frequency of chromosome aberrations 
among newborn infants in Japan, Higurashi (1979) screened 12 319 
newborn babies, 6382 male and 5937 females, for clinical 
manifestations of autosomal aberrations and for sex chromatin and 
sex chromosome aberrations.  The incidence of chromosomal 
aberrations was 1/725 (0.14%) for autosomal trisomy and 1/766 
(0.13%) for sex chromosome aberrations.  Any potential human 
teratogen must be evaluated against the above background.  
Hereditary factors, irradiation, certain drugs and chemicals, and 
some viruses are all known to cause malformations.  However, few 
data are available to define the incidence of malformations induced 
by them.  Precise evaluation of relative risks is difficult in any 
group with a high background incidence of abortion and a wide 
variation in malformations.  Still births and neonatal deaths

    Though still births and neonatal deaths are well recorded in 
some countries, their low frequencies make them weak indicators of 
developmental toxicity related to environmental chemicals.  
Clusters can be identified and studied only by using statistical 
data on prenatal mortality.  Studies of cohorts after exposure to a 
chemical may include laboratory examinations to determine the body-
burden of the chemical, cytogenetic studies, and examination of 
tissues obtained by biopsy or at autopsy.  Birth weight

    Birth weight is an appealing variable for study, because it is 
sometimes obtainable from birth certificates and is known to be 
markedly affected by several factors such as cigarette-smoking 
during pregnancy (Abel, 1980).  Certain malformations are 
associated with low birth weight, and a few others with high birth 
weight (Ciba Foundation Symposium, 1974).  Some agents, such as 
aminopterin or methotrexate, cause low birth weight, when 
administered early in pregnancy (Milunsky et al., 1968).  Before 
lower birth weight for gestational age can be associated with 
exposure to a chemical, alternative explanations, such as 
differences between cases and controls in prepregnancy weight, 
social class, diet, and ethnicity must be excluded.  With care, 
birth weight can be analysed as a continuous variable.  Thus, 
subtle effects of chemicals may be detected by differences in the 
mean and variance of birth weight, which would not be noticed by 
measuring the frequency of birth weights under a fixed value (e.g., 
2.5 kg).  Congenital malformations

    Congenital malformations are recorded to a certain extent on 
some birth certificates, but generally the omissions are too 
numerous for useful study.  To be detected from registry data, a 
teratogen or embryotoxic chemical would, of course, have to be 
present in the area covered by the registry. Furthermore, the coded 
classification of anomalies in the registry must be of sufficient 
detail to reveal an excess of the particular birth defect observed; 
e.g., the diagnosis should be listed as "limb-reduction deformity" 
rather than "other skeletal anomalies".  Often, the use of plain 
language giving accurate detail can provide more reliable 
information than the use of strictly scientific terms.  Also, the 
registry staff must review its data on a regular basis to determine 
if clustering has occurred.  If a cluster is observed, the 
diagnoses should be verified, and a case-control study conducted to 
seek the cause.  Rarely, if ever, do all of these events occur in a 
way that would make it possible for a new teratogen to be detected 
from a registry. 

    In a geographical area where chemical contamination has 
occurred, surveillance for congenital malformations and other 
effects on development may be of value.  If surveillance is deemed 
necessary, a full description of the malformations observed should 
be recorded.  Photographs should be taken of patients with unusual 
findings, to assist in documenting the abnormality. 

    When a potential teratogen is studied, exposed persons should, 
if possible, be compared with a similar group of people who were 
not exposed.  If circumstances prevent the selection or study of 
such a cohort group, internal comparisons, such as high versus low 
exposure, may be possible. 

    Another approach is to study all the children of exposed 
mothers in an effort to identify clustering of specific types of 
malformations.  If a cluster is found, knowing the nature of the 
exposure may suggest a relationship to a particular chemical.  The 
malformation observed should be classified according to other 
possible aetiologies such as infection, trauma, or heredity.  For 
example, webbed toes in a family with a history of this disorder is 
obviously hereditary rather than due to environmental exposure. 

    There are observations that suggest that offspring vary 
markedly in their susceptibility to embryotoxic chemicals. This is 
demonstrated by the occurrence of the "fetal hydantoin syndrome" in 
only one of four siblings whose mother continued to take 
phenylhydantoin during each of her four pregnancies (Allen et al., 

    When classifying single and multiple malformations, each child 
should appear in a tabulation only once.  Patterns of multiple 
malformation should be noted, since they may provide evidence of 
the time at which developmental disturbances occurred. 

3.1.2  Prenatal diagnostic procedures

    These methods can be invasive or non-invasive.  The detection 
of malformation by these methods can only be made after conception 
and, since many patients and health personnel do not approve of 
planned termination of pregnancy, their value in the prevention of 
malformations is limited.  Invasive intrauterine techniques

    Though invasive intrauterine techniques are valuable in well-
defined clinical situations for diagnosis, they are not suitable 
for use in population studies.  Invasive techniques include: (a) 
amniocentesis, by which it is possible to sample the amniotic fluid 
and obtain fetal cells for cytogenetic studies and enzyme assays as 
well as fluid for chemical analysis; (b) fetoscopy by means of 
which the fetus can be inspected directly; (c) combined fetoscopy 
and biopsy or d) ultrasound guided biopsy for obtaining samples of 
fetal blood, skin, muscle, or other tissue (such as choriomic 
villi) for special analysis.  These methods may be combined with 
molecular biological techniques in the future.  Such techniques 
include restriction enzyme fragmentation of genomic DNA, in which 
altered DNA restriction sites can be detected when a large 
percentage of the cells have the altered genotype.  Specific probes 
can also be derived from structural gene sequences or from highly 
repetitive genomic DNA sequences.  cDNA probes for the globin have 
been used to detect defects in the structural gene regions, coding 

for the haemoglobin proteins.  Deletion of specific globin gene 
fragments has been found in the case of thalassaemia.  Defects have 
been found in genes coding for the Beta-globin protein. Some of 
these approaches have already been applied in clinical prenatal 
diagnosis.  Non-invasive techniques

  (a)  Ultrasound investigation

    (i) Visualisation

    Visualisation of the conceptus is possible as early as 15-20 
days after conception, and  heartbeat can be detected at around 35 
days.  Subsequently, the conceptus can be "inspected" 
morphologically, its size and growth determined, and some 
malformations diagnosed.  This is a valuable technique for 
determining fetal growth in the mid and final trimesters.  When 
three or four subsequent measurements of standard variables, such 
as the biparietal distance, are performed around weeks 18, 22, 26, 
and 32, a reliable estimation of the prenatal growth is obtained.  
Gross abnormalities, such as anencephaly, omphalocoele, and cystic 
kidneys can be diagnosed long before term. 

    (ii) Fetal heart rate monitoring

    Fetal heart rate can be determined with ultrasound, and 
recorded in the prenatal period with special monitors.  The 
technique permits the follow-up of acute changes, produced when the 
mother has been exposed to a toxic agent. 

  (b)  X-rays

    X-rays are generally only used to confirm fetal death.

  (c)  Analysis of maternal blood

    By measuring alpha fetoprotein (AFP) in maternal plasma, it is 
possible to detect certain malformations, especially CNS 
dysraphias, without amniocentesis.  The advantages and 
disadvantages of maternal serum AFP monitoring are being studied in 
several countries. 

  (d)  Analysis of maternal urine

    Various indicators of development are found in maternal urine.   
Levels of HCG, or its B-subunit, monitor trophoblastic (placental) 
development.  Levels of estriol are quantified to estimate fetal 
"well-being" towards the end of pregnancy.  The quantity of 
maternal urinary estriol reflects the capacity of hydroxylation of 
C-19 steroids at the 16 position in the fetal adrenals and liver 
and of the transport of these steroids into the placenta, where 
they are aromatized.  Improved chemical tests involving analysis of 
maternal urine are being developed for the detection of fetal 
metabolic abnormalities. 

3.1.3.  Epidemiological methodsa

    The epidemiological approach contributes to the attainment of 
three main goals: 

    (a) identification of environmental teratogens;

    (b) monitoring of poor pregnancy outcome in populations
        exposed to chemicals; and

    (c) evaluation of the usefulness of preventive measures.

    The following approaches can be used to assess the toxic 
effects of chemicals (Tables 2 and 3).  Hypothesis-generating (descriptive) studies

    Descriptive studies are useful for generating a hypothesis, 
but, alone, they are not a means of testing an aetiological 
hypothesis.  The main aspects considered in these studies include: 

    (a)  Case reports

    The description of cases of malformation, in which a birth 
defect coincides with known chemical exposure, rouses suspicion 
that the chemical is a potential human teratogen. 

    (b)  Surveillance

    Data collection systems for the surveillance of congenital 
malformations have been established at both regional and national 
levels, e.g., the International Clearinghouse for Birth Defects 
Monitoring Systems, already mentioned in section 1.  The value of 
these monitoring activities, which should be continuous programmes 
with rapid reporting and periodic reviews and analyses of data, is 
to provide background for testing hypotheses. 

    In the Scandinavian countries, workers are registered according 
to occupation (Erickson et al., 1979; Hemminki et al., 1981).  This 
computerized registry can be linked to the occurrence of 
developmental hazards such as abortions and malformations, to 
identify populations at risk.  Once a registry has been 
established, such data analysis is rapid and relatively 
inexpensive.  The shortcomings of this approach are that not all 
women go out to work and some may have been in their recorded 
occupation for only a short time. 

a  See also IPCS Environmental Health Criteria 27: "Guidelines on 
   Studies in Environmental Epidemiology" Geneva, World Health 
   Organization (1983) 

Table 2.  Hypothesis-generating studies and observations
Methods        Advantages                     Limitations
Case report    No special resources needed;   Relationships with
               calls attention to possible    causal agent cannot be
               potential teratogens           established; lack of
                                              information on
                                              variables of interest

Surveillance   Documentation on background    Controls not included;
registers      and time trends                information sparse;
                                              occupational exposures
                                              missing; slow;
                                              limitations on further

Correlation    Useful pointers to the         Non-causal correlations
studies        etiological factor;            cannot be ruled out;
               Easy to design                 confounding variables
                                              difficult to control

   With standardized records, it is possible to compare data 
collected over a certain period.  Such a registry would be useful 
for the chemical industry.  The technique can be supplemented by 
special epidemiological methods including analytical studies.  
These methods will produce more reliable results if population-wide 
surveillance systems for monitoring  congenital  malformations are 
established.  The longer these systems operate, the more reliable 
will be the information concerning both the incidence of 
spontaneous malformations in a population (Kucera, 1961a,b, 1977), 
and more will also be known of the temporal and spatial variations 
in the frequency of congenital malformations (Kucera, 1971a).  The 
occurrence of unusual malformations (Kucera, 1968); and possible 
environmental impacts on fetal formation (e.g., maternal 
occupation) can be identified (Kucera, 1968b, 1971b). 

    A surveillance system is generally considered to be a 
descriptive method, that has been formally justified and is 
generally accepted.  However, surveillance should be regarded as an 
essential element in further epidemiological investigation rather 
than a separate entity for evaluating chemical hazard and/or the 
subsequent risk. 

    (c)  Correlation studies

    Correlation studies are concerned with the pattern of 
distribution of any health condition or malformation, as well as 
variations in the occurrence of a malformation in relation to time, 
space, and personal characteristics, in populations with various 
levels of exposure to chemicals.  The advantages and disadvantages 
of hypothesis-generating studies are summarized in Table 2. 

Table 3.  Hypothesis-testing studies
Type of study    Advantages                 Limitations
Cross-sectional  Fast to execute;           No incidence rate is
studies          Relative low cost          possible, only a
                                            prevalence rate

                 Control of confounding     Poor for cause-effect
                 variables possible         relationships

Case control     Specific;                  Difficult design;
studies          fast to execute;           difficult
                 inexpensive; reasonable    interpretation;
                 data base; control of      exposure data usually
                 confounding variables      obtained after birth

Cohort           Exposure data prior        Large, slow to execute;
studies          to birth; less difficult   expensive; less specific
                 to design and interpret;   information; many
                 control of confounding     confounding variables to
                 variables possible         be controlled
-----------------------------------------------------------------------  Hypothesis-testing (analytical) studies

    These studies help in establishing cause-effect relationships 
and in the estimation of their magnitude. 

    (a)  Cross-sectional studies

    The objective of cross-sectional or prevalence studies is to 
determine the prevalence rate of a malformation in a given 
population.  The burden and distribution of a malformation are 
described and its association with a potential causal factor is 
determined.  Often a prevalence study constitutes the first phase 
of a prospective study and is usually conducted on a representative 
sample of the population, within a short period of time, to provide 
an "instantaneous" image about diseased and healthy groups within 
that population.  The population and the variables to be considered 
must be very well defined. 

    (b)  Case-control studies

    In case-control studies, which supplement hypothesis-generating 
studies, the frequency of malformations and extent of exposure to a 
potential teratogen are compared with those in an unaffected 
control group.  The affected individual is the starting point, 
while the control group may be selected from either unaffected 
children or those affected with another type of malformation.  The 
advantage of such a control group is that the so-called memory bias 
in pregnancies with abnormal outcomes is reduced.  This approach is 
of great value when rare outcomes, such as a specific malformation, 
are examined. Case-control studies are feasible, only if previous 

exposure is ascertained.  A major problem is lack of adequate 
exposure data.  Other advantages and disadvantages are shown in 
Table 3. 

    (c)  Cohort studies

    These are prospective studies in which two or more groups of 
people exposed to different levels (including no exposure) of a 
potential teratogen are followed up, and the pregnancy outcomes 
recorded.  The prospective method tends to be more informative 
about the quantitative risk associated with exposure to a potential 
agent, and is a very valuable method for analysing the association 
between exposure to chemicals and subsequent developmental effects.  
The value of epidemiological studies in associating environmental 
agents with health effects is greatly increased by carefully 
defining the study group and its cohort.  Women of child-bearing 
age may be grouped according to occupational exposure, prescribed 
drugs, age, diseases, smoking and drinking habits, and other 
variables of interest, but the difficulties involved in the 
evaluation remain.  The size of study populations is a major 
problem.  The use of epidemiological studies for risk estimation 
can be improved by combining study populations from several 
countries, an approach used successfully in the Scandinavian 
countries.  Other advantages and disadvantages of the method are 
shown in Table 3. 

    (d)  Intervention studies

    In these studies, the frequency of a specific disorder in 
groups where corrective measures have been applied is compared with 
that in controls.  Intervention studies can be conclusive, if they 
are carefully designed and there are adequate controls.  Populations at special risk

    As with carcinogens, most human teratogens have been identified 
in studies on therapeutically- or occupationally-exposed 
populations (Sokal & Lessmann, 1960; Shepard & Fantel, 1981).  This 
emphasizes the necessity for identifying sub-populations at high 
risk of producing abnormally developed offspring. 

    Both men and women exposed to chemicals at the workplace should 
be followed up to evaluate the health of their progeny.  
Occupational health programmes, in particular, should include 
reproductive data as an index of workers' health.  Special 
attention should be paid to pregnant women who have been exposed 
accidentally or over a long period to high levels of chemicals, 
even if they appear to be in good health.  Confounding and complicating factors

    All of the previously described methods have their limitations, 
if used as the only source of data for risk assessment.  Of 
particular interest, in the design of epidemiological studies, is 
the consideration of confounding variables.  One of the most 

troublesome factors is lack of agreement about what constitutes a 
developmental defect. Structural defects are emphasized more than 
abnormalities that develop later in life.  As mentioned earlier, 
endpoints include growth retardation, functional disorders, 
malformation, and death.  Ascertainment of malformation is 
incomplete, even by the end of the first year of life, while 
examination for minor structural changes may be haphazard or 
virtually non-existent.  In addition, one case of malformation 
could be reported several times by different medical doctors, 
sometimes with a different diagnosis. 

    The accuracy of incidence rates is further complicated by 
differences in the frequency of malformations known to occur in 
different populations.  The rate for the same malformation may vary 
significantly according to state or country.  Some of this can be 
attributed to ethnic differences, but probably other more subtle 
factors are also involved.  Little is known regarding the 
susceptibility of women of different ethnic groups to chemicals 
that induce birth defects.  Thus, it is difficult to extrapolate 
epidemiological findings from one geographical area to another. 

    Age and parity of the mother and sex of the offspring are 
variables that need to be recorded.  Smoking and drinking habits 
may play a role with regard to induction of embryotoxic effects.  
Thus, many variables appear capable of influencing the rates of 
abnormal development.  Other variables include the time and method 
of evaluation and the interpretation of manifestations considered 
to be adverse.  Furthermore, no woman is exposed to only a single 
agent, and the effects of exposure to other environmental agents 
and other factors such as eating and drinking habits, drug use, and 
diseases acquired during pregnancy should be taken into 
consideration.  Statistical power

    Statistical power is largely determined by the frequency of 
unfavourable pregnancy outcomes and the size of the population 
under study.  To define the probability that a recently introduced 
chemical will double the frequency of normally observed birth 
defects, large numbers of mother/child pairs will have to be 
studied, and even larger numbers will be needed, if a specific type 
of relatively common abnormality (e.g., ventricular septal defect) 
is to be investigated. Conversely, slightly lower numbers may 
suffice, if an increased incidence of an extremely rare abnormality 
coincides with exposure to a specific agent (e.g., phocomelia and 
thalidomide), though, in such cases, a consensus of opinion 
accepting the association, usually precedes completion of formal 
statistical analysis.  Quantifying studies

    As already mentioned, the main drawback of most epidemiological 
studies is the lack of a quantitative assessment of exposure. 

    The environment contains numerous chemicals to which people are 
continuously exposed.  It is therefore difficult to isolate an 
exposure to a single agent, and a dose-response relationship is 
rarely established.  The best conditions appear to be in the use of 
prescribed drugs, where use, dose, and dosage regimen are recorded 
on the prescription.  Computerized linkage between prescribed drugs 
and pregnancy outcome has been used to establish associations 
between drug exposure and pregnancy outcome (Jick et al., 1982). 
Occupationally-exposed groups constitute another appropriate model 
for observational health risk studies, in particular, if 
complemented by analysis of the potentially toxic chemical(s) or 
the metabolites in body fluids. 

3.2.  Experimental Animal Studies

    Developmental toxicity studies should be designed and performed 
in such a way as to be compatible with other types of toxicological 
investigation.  Designing, performing, interpreting, and 
extrapolating specific studies in isolation rather than as part of 
this wider background makes it more difficult to estimate risk from 
extraneous agents. 

    For historical rather than objective reasons, the main focus of 
a range of reproductive toxicity tests (section -, 
necessary for investigating new chemicals, is the teratogenic 
study, which may be more aptly designated as a test for selective 
embryo/fetotoxicity.  These initial investigations should not be 
termed "teratogenicity tests" since they aim to identify a variety 
of effects on the mother and developing conceptus in addition to 
malformations (Palmer, 1976).  The tests are conducted to 
determine: (a) deviation in prenatal variables (resorptions, fetal 
deaths, fetal anomalies, and fetal weight); and (b) a minimal or 
no-observed-adverse-effect dose that does not induce any 
significant deviation from normal development.  The period of 
administration should cover the period of organogenesis, which is 
usually equated with daily dosing on pregnancy days 6-15 and 6-18 
for rats and rabbits, respectively.  During pregnancy, maternal 
responses should be recorded at least daily and body weight 
recorded regularly; a record of food consumption may also be 
required.  Just prior to parturition, dams should be killed, litter 
values (number of live and dead young, fetal weights) recorded, and 
fetuses examined for external, visceral, and skeletal 
abnormalities.  There are variations employing different dosing 
periods and some designs allow for some litters to be reared so 
that offspring can be examined for postnatal manifestations of 
prenatal toxicity. 

    Basically, experimental animal studies should be subdivided 
into two types (Palmer, 1967): (a) studies in which an endpoint is 
sought with a material of uncertain potential; and (b) studies in 
which a defined endpoint has already been established.  The latter 
are usually second-stage investigations that might be regarded as 
follow-up studies after initial tests and most of the literature 
published in this field deals with this type of investigation.  The 
aims and approaches of the two types of examination are quite 
different (Table 4). 

    When used in the initial testing of materials of unknown 
potential, the aims and approaches of the developmental toxicity 
study are quite different from the subsequent study in which the 
activity of a known and potent agent needs to be defined.  For 
initial testing, the main priorities are: (a) to detect the lowest 
dosage causing an adverse effect; and (b) to determine whether the 
effects on the developing conceptus are selective, occurring at 
doses lower than those causing maternal effects. 

    The purpose of an initial test is accomplished if results 
indicate that the probability of selective effects is unlikely.  
Elucidating the nature of the selective action, if necessary, is 
reserved for second-stage studies, because, with a more precise 
target to investigate more appropriate studies, because, with a 
more precise target to investigate, more appropriate studies can be 
designed.  The design of such secondary studies must be flexible 
and will vary according to circumstances, to such an extent that a 
general account would be inadequate.  The following comments are 
therefore confined to initial tests. 

Table 4.  Types of experimental animal studies
                 Initial testing             Second stage
Aims:            To determine the lowest     To determine optimum
                 dosage causing any adverse  conditions of dosage
                 embryo/fetal response       and timing for maximizing
                                             the frequency of specific
                                             malformations or other

Means:           Chemicals of unknown but    Chemicals of known or
                 expectedly low terato-      expected high embryo-
                 genic potential, adminis-   toxic or teratogenic
                 tered in repeated doses     potential administered
                 to determine a broad        as a single dose
                 range of possible effects

Usual results:   Low rates of malformations  High rates of embryotoxic
                 not clearly                 effects and malformations,
                 distinguishable from        clearly distinguishable
                 control values              from control incidences

3.2.1.  Species

    Animal studies have mainly been carried out on mice, rats, 
hamsters, and rabbits because of high fertility, short gestation 
period, ease of determining the onset of pregnancy, and economy of 
cost and housing of animals.  Other species such as ferrets, cats, 
dogs, pigs, and non-human primates have also been used when 
toxicokinetic similarities between the test species and man or a 
longer gestation period was thought important (Palmer, 1978). 

    The suitability of a test species is determined by a multitude 
of factors including cost, availability, ease of handling and 
housing, life span, and similarity to human beings in respect of 
both developmental processes and toxicokinetics.  However, the 
choice is too often based on a limited and unbalanced set of 
factors, for example, on the assumption that sub-human primates are 
more similar to man than other species, because of evolutionary and 
toxicokinetic factors.  This reasoning is valid only when the test 
is used to confirm a positive response to a potent material such as 

    The use of primates in prenatal studies has frequently been 
advocated by various authors, stressing that these animals can be 
used to good effect to provide a closer approximation to the likely 
human response to a material with a known or strongly suspected 
teratogenic effect (Delahunt & Lassen, 1964).  As a means of 
demonstrating or confirming the absence of an effect, they are 
impractical, even when toxicokinetic similarity to man can be 
defined.  The reason for this deficiency is that malformations 
occur at a low frequency, so that to detect a ten-fold increase in 
malformations occurring naturally at a rate of 0.1% would require 
100 or more monkeys per group.  Moreover, primates tend to abort 
defective fetuses fairly readily, so that, particularly with 
monotocous species, the number of endpoints that can be used may 
often be restricted to the two "all or none" responses of abortion 
or malformation. 

    It can be expected that, if alternative species are used, 
investigators will naturally exploit species that they already use 
in other studies because of the valuable background data already 

3.2.2.  Dosages and dose levels

    The selection of doses and dose regimens, and prediction 
concerning the types of effect most likely to occur, are difficult 
issues.  Errors in choosing dose levels are most often made because 
data from other studies are not used (Palmer, 1978).  Cooperation 
between the general toxicologist and reproductive toxicologist is 
essential.  The usual requirement is for 3 test groups and 1 
control (vehicle) group; the highest dosage should induce signs of 
minimal maternal toxicity.  It is rarely necessary to use more than 
3 dose levels of a test compound, when it is administered during 
organogenesis to females of 2 (or more) mammalian species via a 
route similar to that in human exposure. 

    Very high doses may alter the toxicokinetics.  An effective 
alternative is to use a single "limit" dose (equivalent to a high 
dosage) for chemicals with low toxicity.  The high dose is best 
arrived at following a preliminary study in which females of the 
same strain and species are administered the same dose of the test 
material, with the same dosage regimen under conditions similar to 
those to be used for the main study (Palmer, 1978).  Conversely, 
with many industrial chemicals, food additives, and pesticides, as 
well as materials administered by unusual routes, it may be 

impractical to seek a maternally toxic dose.  Under these 
circumstances, the maximum practical dosage or a limit dosage is 
usually employed.  The extent to which the vehicle affects 
absorption of the test compound or to which it could exert an 
independent toxic effect on the fetus or dam should be known. 

3.2.3.  Positive control groups

    Inclusion of a group of animals dosed with a known teratogen is 
sometimes recommended.  However, it would be better to use these 
animals for increasing the size of test groups or a negative 
control group rather than for creating a positive control group. 

3.2.4.  Historical controls

    Control values should be collected and permanently recorded.  
They provide qualitative assurance of the nature of spontaneous 
malformations that occur in control populations. Such records also 
monitor the ability of the investigator to detect various subtle 
structural changes that occur in a variety of organ systems 
(Palmer, 1977b). 

    Historical control values for spontaneous abnormalities are 
difficult to use in assessing the relevance of test group 
observations, because they show marked temporal fluctuations and 
clustering, particularly for malformations, anomalies, and variants 
(John et al., 1982).  Thus, it is often more appropriate to express 
historical values in terms of different blocks denoting the 
frequency of various effects in control groups of different average 
size.  However, extensive historical control observations may be of 
value in the case of very rare (low frequency) malformations. 

3.2.5.  Dosing regimen

    As mentioned, the dosing regimen employed is usually daily from 
days 6 to 15 of pregnancy in studies on rats, and from days 6 to 
18, in studies on rabbits.  The day of mating or of finding 
spermatozoa in vaginal smears is designated as day 0 of pregnancy.  
For routine testing, specified dosing periods need not be 
considered sacrosanct; as long as the dosing period includes the 
entire duration of organogenesis.  The dosing periods selected 
represent a compromise between a short period of exposure (one day 
or less), which with a potent agent will usually provoke a specific 
response, and a longer dosing period which, because it covers a 
greater number of potential responses, induces a less precise 
response.  When testing agents of unknown toxic potential, the 
latter is recommended as a first step. 

    Sometimes, the test agent may be administered more than once a 
day, if, for example, the results of toxicokinetic studies indicate 
that it has an extremely short half-life. Minipumps have been used 
to maintain constant blood levels (Nau et al., 1981).  Conversely, 
for materials with a long half-life, a dosing interval of more than 
one day might be needed to avoid accumulation of the agent in the 
animal. Appropriate intervals should be based on the biological 
half-life of the compound. 

    Two or more complementary studies, each employing a short 
overall dosing period of 4 or 5 days, should be considered when 
testing compounds with either pronounced cumulative toxicity or 
ability to induce tolerance associated with enzyme induction or 
inhibition.  Such refinement of the dosage regimen is particularly 
important, when there is only a slight margin between human 
exposure levels and animal test dosages (Palmer, 1978). 

3.2.6.  Route of administration

    In routine tests, administration should be by the anticipated 
route(s) of human exposure.  This is logical, since the amount and 
rate of a chemical that reaches the embryo varies according to the 
route of administration. However, it is not always true that routes 
allowing administration of the greatest amount of material will 
induce the greatest effects.  For example, application of 
hexachlorophene to the vagina or mucous membranes results in a 
greater systemic reaction in mother and fetus than larger doses 
administered dermally or by incorporation in the diet (Thorpe, 
1967; Kimmel et al., 1974). 

    Though administration by the intended human exposure route(s) 
is usually best for the prediction of the human situation, it may 
not always be possible to achieve other important objectives such 
as the administration of adequate doses or the avoidance of 
unnecessary stress, which in itself may induce embryo/fetal 
toxicity.  In this respect, inhalation studies with "face only" 
exposure are more stressful than whole body exposure and should be 
used only when there is no alternative.  A study in which a more 
practical route of administration is coupled with an awareness of 
species differences in toxicokinetics will often provide more 
meaningful data.  An understanding of comparative toxicokinetics 
may reduce the need to examine a large number of routes of 

3.2.7.  Numbers of animals and statistical analyses

    Group sizes generally considered acceptable have been 
determined by tradition and empiricism, rather than by sound 
reasoning.  Usually, the larger and more expensive the species, the 
smaller the group, in spite of the fact that, on a purely 
scientific basis, this can be questioned.  For example, the number 
of primates per group should be higher than that of rabbits because 
the primate is monotocous and offers fewer endpoints, all of which 
are of an "all or none" type such as abortion or malformation.  
Similarly, the number of rabbits per group should be higher than 
that of rats, because of greater variations in the incidences of 
maternal and fetal death, abortion, malformations, and anomalies in 
the rabbit.  The scientific principle holds, even though it is 
somewhat compensated for by the fact that it is easier to examine 
primate and rabbit fetuses for both visceral and skeletal 

    The advantage of using polytocous species such as mice, rats, 
or rabbits is not because there are more sample units (litters) for 
analysis, but because a higher number of endpoints can be recorded 

and a series of responses can be obtained within litters.  This 
makes it possible to use statistical methods, other than 
contingency analysis, to detect significant differences with 
smaller sample (group) sizes.  Any comprehensive investigation of 
values employed in screening tests (e.g., litter size, incidences 
of embryonic death, malformation, anomaly, or variation) shows a 
litter based, non-normal distribution for which non-parametric 
analysis is safest (Weil, 1970; Palmer, 1974; Staples & Haseman, 
1974; Haseman & Hogan, 1975; Palmer, 1976; Palmer, 1977a; Cooke, 
1976, 1981; Woo & Hoar, 1979, 1982; Shirley & Hickling, 1981).  
There has been and remains great resistance to accepting the 
viewpoint that the litter, not the fetus, represents the valid 
sample unit and that litter values are not normally distributed. 

3.2.8.  Observations on pregnant animals

    Observations on dams should include examination for signs of 
toxicity and regular measurement of body weight. Examination of the 
dam contributes to the study in two important ways: (a) by 
providing a reference point so that any effects on the conceptus 
can be judged as selective and specific, rather than secondary to 
toxic effects on the dam; and (b) by making it possible to 
ascertain whether the agent is more toxic for pregnant animals.  
Such a phenomenon has been observed for anti-inflammatory agents 
and iron dextrans (Flodh et al., 1977). 

    Records of food and water consumption are often requested, but 
rarely provide more precise information than body weights and/or 
clinical signs, provided dosages were correctly chosen in the first 

    In initial testing, the emphasis is logically placed on using 
observations on the pregnant animal to determine the selectivity of 
action on the conceptus.  However, in second stage investigations 
there may be, in specific situations, a need to pay greater 
attention to maternal toxicity as a cause of embryo/fetal toxicity.  
Maternal ill-health and disease conditions in the mother have been 
associated with malformations in human beings (Kalter & Warkany, 
1983) and toxic effects in the dam with malformations in the mouse 
(Khera, 1983).  Thus, where exposure to chemicals at, or near, 
maternally toxic levels is unavoidable, greater attention should be 
paid to the possible consequences, for the conceptus, of maternal 

3.2.9.  Observations on the progeny

    When examining the litter, fundamental requirements include the 
recording of the number of live young, and embryonic and fetal 
deaths; the number of implantations can be recorded directly or 
deduced.  It may be useful to count corpora lutea of pregnancy 
(except with mice, because of the difficulty of counting) to insure 
against incorrectly attributing reduced litter size to treatment.  
With studies in which dosing is initiated at, or before 
implantation, or when a species, such as the pig or the sheep, that 
has a long gestation period is used, corpora lutea counts can 
contribute directly to determining embryonic loss. 

    Viable fetuses should be examined for external malformations, 
and skeletal and visceral anomalies and variations, and their sex 
should be determined.  Other suggestions, some of which are of 
value while others are less useful, include: 

    (a) The weighing and examination of dead fetuses for 
abnormalities.  This usually provides data that are difficult to 
interpret, because dead fetuses undergo decomposition. 

    (b) Distinguishing between early and late embryonic deaths 
(resorptions).  This is generally a useful exercise as observations 
of dose-related effects can help in determining the sensitive 

    (c) The recording of individual fetal weights (or crown-rump 
length) of rats or mice.  This is generally of little value in 
initial testing, unless individual fetal weight can be associated 
with altered fetal development, such as delayed ossification or 
malformation.  Most investigators who record individual fetal 
weights rarely attempt to associate these with fetal anomalies or 
variations.  In routine tests, it is just as effective and less 
tedious to weigh the entire litter and calculate the mean fetal 
weight. The occasional abnormal or obviously oversized or 
undersized fetus may be weighed individually and the result related 
to abnormalities. 

    Another disadvantage of weighing fetuses individually is that 
it may lead to inappropriate statistical analysis, which, in turn, 
would provide "statistically significant" differences between 
groups, which would be of little biological relevance (Palmer, 
1978).  Similar difficulties occur when individual crown-rump 
length is measured in rodents or lagomorphs.  This variable is 
tedious to record and has little value on its own, though it may 
complement measurements of fetal weight and skeletal development.  
However, this level of effort and specificity is best reserved for 
second-stage studies rather than initial testing. 

    (d) Comparison of litter weights.  This is a particularly 
useful parameter in initial tests, because it reflects both litter 
size and fetal weights.  A non-specific toxic response to an agent 
may be seen as reduced fetal weight, at low doses, and as reduced 
litter size through increased deaths, at higher doses.  In such 
circumstances, a dose-related reduction in litter weight provides 
useful information. 

    (e) The weighing of the uterus and contents (gravid uterus).  
Some investigators weigh the gravid uterus and subtract this from 
the total maternal body weight to obtain a corrected maternal body 
weight in the belief that this corrected weight, together with 
maternal weights measured during pregnancy, facilitate distinction 
between the maternally-mediated and direct effects of chemicals on 
the fetus.  Other workers are not convinced that such data are 
enough to determine the direct effect of a chemical on the fetus. 

    (f) Allocation of a percentage of fetuses for either visceral 
or skeletal examinations.  It has been suggested that 66%, 50%, or 
33% of fetuses should be allocated for one or other type of 
examination, but this applies only to rats, mice, and hamsters.  It 
is quite feasible to examine the fetuses of rabbits and most other 
species for both skeletal and visceral anomalies, thus, 
considerably increasing the amount of information obtained from a 
single study.  With fetuses of the smaller species, the safest 
procedure is to allocate half of the fetuses in each litter for 
visceral examination and the other half for skeletal examination. 
However, the most important criterion is to ensure that sufficient 
fetuses and litters are examined for each technique, to make a 
valid statistical analysis. 

    Two suggested methods of visceral examination are hand 
sectioning (Wilson technique) and microdissection.  When performed 
well, the Wilson technique (Wilson, 1965; Beck, 1977) is 
acceptable.  However, good hand-sectioning is difficult to achieve 
and high quality is rarely attained.  For the vast majority of 
investigators, microdissection is a safer technique for small 
fetuses and a better technique for large fetuses. 

    (g) Reporting units.  Data should be reported on each endpoint 
in the form of a test group value (mean, median, or incidence) and 
an individual litter value.  Reporting of malformations and 
anomalies should clearly indicate the specific fetuses and litters 
affected.  Accurate interpretation and correct analysis is not 
possible, if anomalies are listed by type, without indicating their 
distribution among fetuses and litters.  If developmental toxicity 
is recorded at dosages considerably lower than those eliciting 
maternal toxicity, the effect may be considered selective, and a 
direct effect on the embryo or a specific result of interference 
with its nutrition. 

    In concluding whether or not selective action has occurred, the 
answer should be positive, negative, or inconclusive.  When the 
answer is positive or negative, the process of extrapolation can 
begin.  When the results are inconclusive, further investigations 
should be carried out. 

3.2.10.  Animal husbandry and laboratory practices

    Animal studies should be performed according to the highest 
standards of care, welfare, and husbandry, for both ethical and 
humane reasons.  It is equally important from the practical point 
of view, as animals in poor condition will provide unreliable 
results.  Furthermore, all animal studies are conducted under 
conditions with a large number of potential confounding variables 
(Chance, 1957; Biggers et al., 1958; Russell & Burch, 1959; Kalter, 
1965; Laroche, 1965; Dearborn, 1967; Ellis, 1967; Kallai, 1967; 
Palmer, 1969). Studies involving pregnant animals may be especially 
susceptible to these variables because of the rapid changes in 
maternal physiology and in interactions between the maternal 
organism and the developing conceptus. 

    Laboratory practices necessary for scientific acceptability of 
any work are, honesty, care, consistency, and responsibility, in 
addition to those mentioned or implied in this and other sections 
of this report.  Principles of good laboratory practice (GLP) are 
detailed in various regulatory requirements and presented for 
instance, with the guidelines mentioned in Section 1. 


4.1.  Introduction

    The production of offspring able to adapt to postnatal life 
requires the development of functionally normal organs and systems.  
Since organs and systems may often be at greater risk of damage 
from the adverse effects of a chemical during the period of their 
development and growth, functional abnormalities in various organs 
and systems have at times been linked with exposure to toxic 
chemicals during the prenatal period.  The consequences of such 
damage may not be readily evident, unless functional or behavioural 
tests are performed. 

    The period of organ and system maturation extends beyond the 
period of organogenesis and even beyond the prenatal period.  
Therefore, the susceptible period for the induction of insults that 
may lead to functional deficits is much longer than that for the 
induction of gross structural defects. 

    Functions that have been shown to be affected by prenatal and 
early postnatal exposure to chemicals include behaviour, 
reproduction, endocrine function, immune competence, xenobiotic 
metabolism, and various other physiological functions.  Late 
manifestations of toxicity may include neoplasia and shortened life 

    Fetal damage by chemicals has been shown to modify enzyme 
ontogeny.  Examples of active compounds are PCBs, TCDD, 
organomercurial compounds, and formaldehyde and its derivatives 
(Andrew & Lytz, 1981; Merkurjewa et al., 1983). Since organ-
specific profiles of enzymes continue to evolve and undergo 
pronounced changes, long after the organs have become 
morphologically distinct, enzyme systems are an easy target for 
chemical insults.  Alterations in enzyme activity indicate 
disturbance in the fundamental biochemical processes (Battaglia & 
Meschia, 1978) and in the maturation of organ functions (Stave, 

    It has been reported that the level of drug-metabolising 
microsomal monooxygenases is altered following prenatal exposure to 
chemical inducers, and such an effect may persist in adult life 
(Illsey & Lamartinière, 1979; Yanai, 1979; Salganik et al., 1980).  
As a result, the metabolism of chemical carcinogens, e.g., 
aflotoxin B1, may be altered (Faris & Campbel, 1983). 

    Toxic agents might exert their effects by causing alterations 
in endocrine function and hormone levels. Hormones are the natural 
stimuli that evoke the synthesis of many enzymes, especially those 
required for adaptation to extrauterine life (Greengard, 1973; 
1975a,b).  Major disorders may result from impairment or imbalances 
in developmental endocrinology, if they occur at critical stage, 
when hormones normally exert discrete effects on maturation.  So 
far, few studies have dealt with the effects of hormonally-active 
xenobiotics on the developmental programming processes (Illsley & 
Lamartinière, 1979; Lamartinière & Lucier, 1981). 

    Deviations in functional development may also be mediated by 
induced adverse effects on liver function.  Prenatal rat liver does 
not acquire full competence until some weeks after weaning, which 
makes it a suitable model for studying developmental delays.  There 
are data suggesting that  in utero exposure to environmental agents 
may result in altered liver function (Lucier et al., 1975; Chahoud 
& Eggert, 1978; Lucier & McDaniel, 1979). 

    There are indications that compounds that interfere with adult 
immune systemsa may have effects on developing immune systems 
(Roberts & Chapman, 1981).  It has been reported that exposure to 
toxic substances during immune ontogenesis may induce functional 
defects that remain dormant until adulthood (Spyker & Fernandes, 
1973), and that the stage of immune development at the time of 
exposure may affect the nature, magnitude, and duration of the 
immune defect (Pinto-Machado, 1970). 

    Methods of assessing organ function in progeny do not differ 
from those conventionally used in toxicology and are not discussed 
in this document.  Evaluation of behaviour, reproduction, and 
transplacental carcinogenicity requires more specific methods and 
these are discussed in the following subsections.  The evaluation 
of postnatal growth and survival has been incorporated in the 
subsection on behaviour (section 4.2). 

4.2.  Behaviour

    The argument has been advanced that behavioural tests may be 
useful in the risk evaluation of prenatal exposure to chemicals, 
because behaviour is a functional indicator of the net sensory, 
motor, and integrative  processes occurring in the central and 
peripheral nervous system (Mello, 1975). Thus, alterations in 
behaviour following exposure to environmental agents might be a 
relatively sensitive indicator of nervous system toxicity (Spyker, 
1975).  Because behaviour can also be influenced by the functioning 
of other organ systems, specific organ toxicity might also be 
reflected as a change in behaviour (Weiss, 1975; Tilson & Mitchell, 

    Two separate phenomena underly behavioural deviations that 
appear as a manifestation of embryotoxicity.  The first, an 
afferent phenomenon, involves the interaction of a chemical with a 
molecular entity of the developing organism, such as a receptor or 
a specific enzyme.  The second, an efferent phenomenon, represents 
the consequences of this interaction at the molecular, subcellular, 
cellular, or cell population leve1.  Malformations are the outcome 
of altered development at the cell population level and are due to 
disturbances in basic morphogenic processes. 

a  Principles for assessing immunotoxic effects will be dealt
   with in a separate IPCS document.

    An embryotoxic agent may induce behavioural deviations by:

    (a)  disturbing the morphogenesis of the central nervous
         system and consequently influencing its functional
         development (Langman et al., 1975; Pellegrino &
         Altman, 1979; Rodier et al., 1979); and

    (b)  affecting functional development by interfering with
         the processes of axogenesis, dendrogenesis,
         synaptogenesis, and myelination (Jacobson, 1978) or
         with the function of already differentiated cell
         populations in the brain.

    The relationship between morphogenesis and functional 
development is far from clear.  Specific components of the nervous 
system may be affected to induce neurobehavioural deviations, in 
the absence of any apparent morphological changes in the brain 
(Reinisch & Karow, 1977; Lundborg & Engel, 1978; Rosengarten & 
Friedhoff, 1979).  However, this view may be too simple, since an 
altered rate of cell acquisition in the developing brain and other 
morphological changes may escape detection (Patel et al., 1977; 
Rodier, 1978; Bendek & Hahn, 1981).  Altered structural and/or 
biochemical components must be present to account for functional 

    In some cases, structural defects of the central nervous system 
have been found not to be accompanied by behavioural changes 
(Brunner et al., 1978).  However, the manifestations of behavioural 
defects may be minimized by plasticity and compensatory processes 
or may escape detection, because of inadequate testing. 

    If behavioural deviations are considered to be embryotoxic 
phenomena, then dose-response, stage-response, and stage-effect 
relationships should be demonstrated.  The latter characteristics 
are determined by a pattern of behavioural deviations that may 
change in a day-to-day fashion, according to the time of chemical 
exposure (Rodier et al., 1979).  The sensitive period for inducing 
behavioural deviations in animals appears broader than the period 
for inducing malformations of the central nervous system (Vorhees 
et al., 1979).  Behavioural changes can be induced at certain times 
in the early postnatal life of mice and rats that correspond to the 
late prenatal and perinatal periods of human brain development 
(Langman et al., 1975; Lundborg & Engel, 1978; Pavlik & Jelinek, 
1979; Pellegrino & Altman, 1979; Pavlik et al., 1980).  The 
development of the human brain is characterized by intensive 
morphogenic and functional developmental processes during the 
perinatal period, which continue, even during the first and second 
years of life (Lewis & Patel, 1980).  Thus, the relevant period of 
brain development in the subject must be considered in a 
behavioural study. 

    Testing procedures to be employed for assessing the mechanisms 
of behavioural effects should be decided on the basis of the 
mechanisms underlying embryotoxicity.  One possible approach, may 
include the use of the primary interaction of chemicals with 

molecular entities of the developing brain, such as receptors or 
specific enzymes. Affinities of chemicals for the following classes 
of receptors could also be assessed  in vitro: steroid hormone 
receptors, classical neurotransmitter receptors, peptidergic 
receptors, neuromodulator receptors.  Furthermore, the rate-
limiting enzymes in the synthesis and/or degradation of different 
mediators could be investigated.  The disturbances of morphogenic 
and functional developmental processes that are triggered by 
exposure to test chemicals during brain development can also be 
quantified.  A third approach deals with assessing behavioural 

4.2.1.  Strategy of testing

    Study of the behavioural effects induced by prenatal exposure 
to chemicals is a new field which has, so far, dealt mainly with 
the detection and description of behavioural deviations induced by 
environmental agents in experimental animals rather than with the 
elucidation of the mechanisms of their induction (Coyle et al., 
1976; Kimmel, 1976; Rodier et al., 1979; Vorhees et al., 1979).  As 
stated by Zbinden (1981), "there are no direct and unambiguous 
measures of behaviour; ... even the simplest behavioural response 
results from an interplay of many parts of the whole nervous 
system, and it is always modulated by a unique set of information 
stored in the memory of the experimental subject". 

    The incomplete knowledge of normal behaviour and its underlying 
mechanisms, together with the fact that the study of altered 
behaviour has only developed very recently, account for the lack of 
uniformly established and accepted strategies and methods for the 
evaluation of behavioural effects.  No single approach or 
behavioural test battery has been identified as the most reliable, 
sensitive, and economic means of detecting behavioural disfunction 
following developmental insult, and, with the current state of 
knowledge, it is not possible to select a battery of tests from the 
procedures available.  Thus, it is considered premature to decide 
on a fixed protocol to be included in routine experimental animal 
studies for testing chemical substances (Mitchell, 1978; Tilson & 
Cabe, 1978; Adams & Buelke-Sam, 1981; Zbinden, 1981). 

    Behavioural testing is performed after something is known about 
the teratological and embryotoxic effects of the agent and at 
exposure levels lower than those inducing structural abnormalities.  
It is suggested that tests be performed at two levels of 
sensitivity/complexity: initial and secondary. Initial testing of 
chemicals of unknown potential should ensure that behavioural 
effects are not missed.  It should be broad and comprehensive with 
regard to the types of functions assessed.  This is necessary 
because many types of functions may be adversely affected by an 
agent and these effects cannot be reliably predicted.  Using test 
methods that monitor a number of functions simultaneously is 
appropriate at this stage, in order to save time, cost, and effort 
(Butcher, 1976; Dews & Wenger, 1979).  The behavioural study should 
preferably be conducted in combination with a developmental 
toxicity study on the same progeny, so that data from the toxicity 
study can be used to support and clarify the behavioural findings. 

    Secondary testing should be performed if behavioural and/or 
neurological alterations are observed in initial testing, or 
suggested by human data, or by structure-activity relationships.  
More sophisticated and selective behavioural techniques, perhaps in 
other species, should be used to help delineate the type(s) and 
extent of effects produced and the possible mechanisms involved.  
Also, at this stage, more selective exposure regimens and testing 
schedules should be employed, as well as procedural controls to 
distinguish a direct neurotoxic action from an indirect effect.  
Secondary behavioural evaluations should be flexibly designed with 
agent specificity in mind (Adams & Buelke-Sam, 1981). 

    In behavioural testing, several aspects should be taken into 
account including the choice of functions to be assessed, the tests 
to be employed, the species to be used, and the factors to be 
controlled.  Behavioural functions to be assessed

    The following functional categories should be evaluated:

    (a) physical growth and maturation;a

    (b) reflex and motor functions;

    (c) sensory functions;

    (d) affective functions (activity/reactivity);

    (e) cognitive functions (learning and memory);

    (f) social behaviour; and

    (g) reproductive behaviour.

    Ideally, behavioural functions should be assessed as they are 
initiated, as well as during their development, maturation, and 
decline, since behavioural deficits might be expressed by: (a) 
retardation or abnormal development of certain functions; (b) 
alterations in adult function; (c) premature decline in functional 
capacity or induction of premature senescence (Grant, 1976).  At 
present, most behavioural testing is done on adolescent or young 
adult animals, with the exception of reflex and motor development 
tests, which are performed in early neonatal life (Buelke-Sam & 
Kimmel, 1979). 

a  Although not behavioural parameters, physical growth and
   maturation are so integrated with behavioural development that
   they must be taken into account in any interpretation of
   alterations in behaviour.  Tests to be used in behavioural studies

    Tests used in behavioural studies are basically derived from 
experimental ethology, psychology, neurology, neurophysiology, and 
neurotoxicology.  The decision regarding the tests to be used 
depends on several factors related to the characteristics of the 
animal model, the availability of technology for measuring a 
specific function, and on cost-effectiveness/time-efficiency 
factors (Tilson et al., 1980).  Within a particular study, the 
choice of tests depends basically on the desired level of analysis.  
At the initial level of behavioural analysis, procedures are sought 
that sample a wide range of functions (e.g., apical tests), are 
inexpensive to perform, require little or no training of 
experimental animals, and make possible the assessment of large 
numbers of animals under study.  These tests are used to provide a 
tentative assessment of  behavioural deficits  and to quantify them 
as precisely as possible.  However, such procedures frequently 
require subjective measurements, often yield quantal data, and may 
not be sensitive enough to detect subtle neurobehavioural 

    When a more detailed evaluation is necessary, tests that 
require extended or special training, special behavioural 
instrumentation, and refined procedures are used.  At this stage of 
analysis, behavioural tests can be supplemented with other methods 
to determine the mechanism(s) of action, such as neurochemical 
assays and sophisticated morphological examinations. 

    The present state of development of behavioural tests requires 
that these methods be validated and standardized to be maximally 
used in any neurobehavioural-testing programme. However, before 
recommending a particular testing method for routine use, it must 
be certain that the test actually measures the desired parameters.  
In this context, very few of the present behavioural test methods 
have been validated (Tilson et al., 1979).  The choice of a test 
method can be greatly facilitated by knowing more about its 
inherent variability in normal populations.  As yet, such data are 
scarce (Phillips et al., 1980). 

    Another related problem is the choice or determination of the 
testing sequence.  In general, the testing sequence should progress 
from the least stressful to the more stressful procedures, for 
instance, from natural to manipulated behaviour (Vorhees et al., 
1979).  However, the possible "carry-over" effect from one test to 
another presents problems, for which there is no easy solution.  
Possibly, in initial testing, the risk of carry-over should be 
accepted, in order to reap the benefits of being able to correlate 
the responses of an individual through different procedures.  If 
there is a suggestion of a real problem associated with carry-over, 
this may be resolved in second-state studies by allocating separate 
sub-sets of animals to each test procedure.  Choice of species

    The selection of suitable species obviously involves practical 
and economic factors.  Animals with pregnancies of short duration 
and high fecundity are preferred.  It is also important that 
adequate embryological, toxicological, and behavioural data are 
available for the species chosen.  As it is preferable to study 
species with well-investigated normal behavioural profiles, rats 
and mice would be the first choice.  Other species or strains may 
be used to investigate specific behavioural deficits only at the 
secondary testing leve1.  The choice of test species at this level 
will greatly depend on the specific function(s) to be measured.  
For instance, albino rats are inappropriate for assessing visual 
function.  The use of species that are phylogenetically close to 
man, such as primates, may sometimes be justified.  Variables to be controlled

    Control of variables is essential in behavioural studies, since 
the integrated and subtle nature of behaviour may make it more 
susceptible to factors that have little impact on other endpoints 
of developmental toxicity. 

    In the current context, this control must be exercised at the 
time of exposure, at the time of testing, and in between exposure 
and testing.  The list of potential influential factors is so 
extensive that it cannot be dealt with adequately in this document.  
As brief examples, psychosocial factors such as litter size, number 
of animals per cage, type of cage, and frequency of contact 
(handling by the attendant technician or investigator) can affect 
performance considerably.  The circadian rhythm may markedly affect 
performance in activity tests, so it is important not to test 
animals from different groups at different times of the day or 
night (assessment of possible effects of chemicals on circadian 
rythm should also be considered) (Chahoud et al., 1975). 

    There are pronounced maternal and litter influences operating 
from conception to weaning that can exert a lasting impression on 
maturity.  Fostering and cross-fostering techniques can be applied 
in second-stage investigations to separate direct treatment effects 
from postnatal maternal influences (Spyker & Spyker, 1977).  Litter 
size is an important factor that determines not only the level of 
nutrition but also the amount of social stimulation (Frankova, 
1968).  Variations in the litter size have been shown to affect the 
responses to open field and T-maze tests (Robinson, 1976; Akuta, 
1979).  Reducing litters to a certain standard number, in order to 
avoid nutritional differences, has been widely practised, but there 
are discrepancies in the methods of standardizing litter size for 
postnatal behavioural stud-ies.  However, according to some 
authors, standardization of litter size could introduce bias into 
the study (Tesh, 1977). 

    As in other aspects of developmental toxicity, time of exposure 
may be critical in the determination of the type of effect induced.  
Timing in the performance of behavioural tests may also be critical 

in the determination of the effects detected.  As well as being 
delayed in appearance, behavioural manifestations may be 
multiphasic in character (Sobrian, 1977; Tabacova, 1981).  
Furthermore, the integrated nature of behaviour is such that the 
consequences of damage may be overcome by the development of 
adaptive or compensatory mechanisms.  Ideally, therefore, behaviour 
should be investigated as a dynamic progression.  This could be 
uneconomic and impractical on a routine basis, but it should be 
borne in mind that differences in behaviour recorded at any one 
time may be misleading.  The research worker should also be aware 
that the "carry-over" effect of repeated testing may interfere with 
the evaluation of borderline differences. 

4.2.2.  Methods of assessment of specific functions  Physical development

    Growth and survival are considered to be among the most 
important indicators of functional normality (Wilson, 1973a; 
Palmer, 1976; Zbinden, 1981).  From the point of view of 
screening, a difference in body weight gain may be a significant 
observation of developmental toxicity, since most physical 
landmarks correlate well with body weight (Adams & Buelke-Sam, 

    Physical development is routinely assessed by recording body 
weight and survival at weekly intervals from birth until weaning, 
as well as by registering the time of appearance of certain 
physical landmarks (Table 5), which should be looked for daily, 
starting at least one day prior to their expected appearance, until 
prevalent in all the test animals.  However, behavioural parameters 
may be affected by the excessive handling of pups. 

Table 5.  Physical signs and the approximate time of their 
postnatal appearance in ratsa
Physical sign                                  Postnatal age
Pinna detachment (unfolding of external ear)   day 2

Primary coat of downy hair                     day 5

Incisor eruption                               day 8

Development of fur                             day 9

Ear opening                                    day 11

Eye opening                                    day 14

Testes descent                                 day 25

Vaginal opening                                day 30
a  From: Alder & Zbinden (1977).

    Variations in the times of appearance listed in Table 5 can 
result from differences in environment, animal strain, and 
investigators recognition level (Hughes & Palmer, 1980). Physical 
development may also be influenced by the duration of pregnancy, 
maternal health, and litter size.  Reflex development

    The reflexes most frequently evaluated are the righting 
reflexes, negative geotaxic response, auditory startle reflex, and 
grasping and placing reflexes (Table 6).  Variations from the 
quoted values can be expected.  Detailed descriptions and suggested 
procedures for measuring these reflexes are available in the 
literature (Bolles & Woods, 1964; Fox, 1965; Altman & Sudarshan, 
1975; Alder & Zbinden, 1977; Vergieva et al., 1981a).  The normal 
disappearance of certain reflexes may also be important, indicating 
that the reflex has been integrated into more complex structures of 
behaviour (Lapointe & Nosal, 1979). 

Table 6.  Reflex development in the rat
Test                       Postnatal age of 
                           (mean value)
Surface righting reflex    day 1.8
Negative geotaxis          day 4.8

Cliff avoidance            day 4.8

Palmar grasp               day 6.3 (waned)

Auditory startleb          day 11.4

Vibrissae placing          day 12.5

Free fall righting         day 17.5

Visual placingc            day 17.6
a  From: Smart & Dobbing (1971).  For reflex ontogeny in mice, see 
   Fox (1965). 
b  At the indicated age, the test reflects the opening of external 
   meatus of the ear.
c  The test may be unreliable, since the visual acuity of many 
   albino strains is so limited that confusion with tactile 
   (vibrissae) placing can readily occur.  Sensory functions

    Tests to assess sensory functions are based on the localization 
or orientation responsiveness to stimuli.  These tests are 
influenced by a wide range of other variables as they are based on 
indirect measurements (Evans, 1978). Procedures to evaluate sensory 

functions are simple to perform and require little investment in 
time and equipment, but the results are limited to quantal data 
(the response is scored as present or absent) and the scores are 
subjective (Tilson & Mitchell, 1980).  Thus, these tests provide 
weak evidence of sensory effects.  More precise details regarding 
the extent of the impairment require other, more sophisticated 
methods, based on operant conditioning (Evans, 1978; Stebbins & 
Rudy, 1978).  However, these methods are not applicable in the pre-
weaning period.  The ages when specific sensory functions can be 
assessed are presented in Table 7.  For more details concerning 
early sensory assessment, the reader is referred to Alder & 
Zbinden, (1977), Grauwiler & Leist (1977), Tesh (1977), and 
Zbinden, (1981).  Motor function

    Spontaneous movements can be observed in familiar (home cage, 
activity cage) or unfamiliar (open field) environments, and provide 
information about the time of appearance and type of different 
motor functions.  Open field studies provide more precise data and 
additional information on the latency and organization of motor 
behaviour (Zbinden, 1981). 

Table 7.  Specific sensory functions (rat)a
Function          Postnatal age at which      Test
                  assessment usually starts
                  to be possibleb
Vision            18 days                     pupillary reflex;
                                              visual placing response;
                                              visual cliff avoidance;

Hearing           11-13 days                  auditory startlec

Olfaction         3-8 days                    odour aversion
                  3-12 days                   odour preference

Ultrasonic        17-20 days                  failure of vocalisation
vocalization                                  when removed from nest
a  Adapted from Zbinden (1981).
b  For all these tests, different authors have established their own
   ranges, a practice recommended to investigators.
c  See footnote b to Table 6.

    The procedure of normal development of spontaneous motility in 
rat is indicated in Table 8. 

    In the rat, a peak of motor activity (as an integral part of 
overall activity) usually originates between 14 and 16 days of age 
and falls to adult levels at about four weeks of age (Campbell et 
al., 1969; Melberg et al., 1976; Randall & Campbell, 1976; Campbell 
& Raskin, 1978).  This seems to be correlated with the concurrent 
development of the brainstem catecholaminergic system (Campbell & 
Mabry, 1973). 

Table 8.  Development of spontaneous motility in rata
Type of movement                        Postnatal age 
                                        of appearance
Pivoting (early ambulatory movements,   day 1
  showing no coordination between
  fore- and hindlimbs, head and pelvis)

Head raising                            day 8

Elevation of forelimbs and shoulders    day 9

Crawling substituted by walking         days 9-11

Elevation of hindlimbs and pelvis       days 13-17
a  From: Alder & Zbinden (1977).

    An open-field procedure can be used to quantify the development 
of motor function in the pre-weaning period.  The open-field test 
has been successfully used in early postnatal life for detecting 
motor disturbances after prenatal exposure to environmental toxic 
substances (Kavlock et al., 1980; Tabacova et al., 1981, 1983).  
When used periodically during the first three postnatal weeks, 
disturbances in the developmental sequence of normal motor patterns 
can be revealed. 

    Another group of test methods measures movements elicited by 
the research worker or by some kind of mechanical device. Tests 
frequently employed (Table 9) assess body posture, muscle tone, 
motor coordination, equilibrium, and gait, and provide information 
on the time of appearance and achievement scores.  Most of the test 
procedures are described in detail by Fox (1965), Altman & 
Sudarshan (1975), and Alder & Zbinden (1977). 

Table 9.  Elicited neuromuscular activities
Activity                      Reference
Hindlimb support              Alder & Abinden (1977)

Hanging-grip strength         Werboff et al. (1961); Altman
                              et al. (1971)

Clinging to inclined plane    Cabe et al. (1978); Sobotka et
                              al. (1974); Rivlin & Tator (1977)

Climbing vertical rod         Altman et al. (1971)

Climbing inclined screen      Werboff et al. (1962)

Homing response               Altman et al. (1971)

Crossing narrow path          Grauwiler & Leist (1977)

Rotating rod performance      Jones & Roberts (1968)

    Most of these tests can be performed by the end of the second 
and during the third postnatal week.  Performance on the rotating 
rod is best assessed during the fourth to sixth postnatal week.  
The optimum time for testing may vary according to the strain of 
rat, the design of the rotating rod, and the operating procedure.  
If performed early during development, the animals have neither 
sufficient coordination nor strength to provide meaningful scores; 
if conducted too late, the individual variation may be so 
pronounced that intergroup comparisons are not possible.  As an 
alternative to the rotating rod, the inclined plane test requires 
simpler and less expensive apparatus and can be applied over a 
wider age range.  In the pre-weaning stages, the same apparatus can 
be used for assessing negative geotaxis.  Generally, the inclined 
plane test shows less variation between individual scores and is 
less demanding on animals' coordination than the rotating rod. 

    Physical strength is assessed using activity wheels or by 
swimming ability.  However, these tests are not practised during 
the pre-weaning period. 

    It should not be inferred that all of the tests listed must be 
used in the same study.  A small battery of adequately chosen tests 
is quite sufficient for any level of analysis (Tilson et al., 
1979), allows better control, and creates less disturbance for the 
animals.  Activity, reactivity, and emotionality

    Activity is a complex measure of behaviour.  It might indicate 
various functions of the CNS and of the whole organism, for 
example, motivational, affective, cognitive, and physiological 
responses.  These various aspects of spontaneous activity 
demonstrate clearly the necessity for a more precise definition of 
what is meant by activity measurements. Prenatal exposure to 
toxicants may result in shifting the curve of activity development 
to a later period (Michaelson et al., 1974) or in producing changes 
in activity that disappear as the animal matures (Culver & Norton, 
1976; Reiter, 1977; Tabacova et al., 1981).  Longitudinal testing 
to uncover possible age-dependent changes in activity would 
therefore be advisable. 

    The most widely-used test for assessing activity is the open-
field procedure, of which there are various modifications (Spyker 
et al., 1972; Grauwiler & Leist, 1977; Kavlock et al., 1980, etc.).  
For screening, automated methods are preferable to direct 
observation by the investigator (Ljungberg & Ungerstedt, 1976; 
Norton et al., 1976; Marsden, 1979). 

    Reactivity is defined as the animal's responsiveness to various 
external stimuli.  Changes in the excitability of the nervous 
system can be measured experimentally as changes in overall 
responsiveness to environmental stimulation. Reactivity is measured 
as avoidance of aversive stimuli, escape responses, orienting 
reflexes, and startle responses. The startle response is one of the 
most frequently used procedures to assess the responsiveness of 

animals to external stimuli (Adams & Buelke-Sam, 1981).  This 
reflex can be elicited by a variety of environmental stimuli such 
as auditory startle or air puff startle and has a reproducible 
short latency.  The startle response can be influenced by changing 
the variables of the eliciting stimulus, the level of background 
stimulation, or the general responsiveness of the animal (Tilson & 
Mitchell, 1980).  Toxicant-induced alterations in startle 
responsiveness can be readily interpreted, because the anatomical 
pathways and neurotransmitter systems that mediate this response 
have been reasonably well studied (Fechter, 1974; Davis, 1980). 
Measurement of the startle response has been highly recommended as 
a tool for the screening of potential neurotoxic agents (Adams & 
Buelke-Sam, 1981).  As mentioned earlier, the startle response can 
also be used as a test for the development of the ear (in the 
preweaning period) and as a specific test for hearing. 

    Emotionality is directly related to spontaneous activity. In 
psychology, the concept persists that emotionality and exploration 
are inversely related.  High emotionality is felt to inhibit 
exploration while low emotionality facilitates it. Urination and 
defecation in the open field are often used as criteria of 
emotionality (Rodier, 1978), but other criteria may also be 
assessed (Archer, 1973).  Cognitive development

    Cognitive development is essentially defined as the ability to 
learn or respond appropriately to a changing environment.  
Associative learning involves the acquisition of a response to a 
previously neutral stimulus, as a result of temporal association 
and/or reinforcement contingencies (Thompson & Glanzman, 1976).  
Non-associative learning is expressed by various other kinds of 
behaviour such as habituation, latent learning, play, and 
manipulation of objects. 

    Several experimental designs have been developed to assess 
associative learning in young animals (Misanin et al., 1971; Amsel 
et al., 1976; Blozovski & Cudennec, 1980; Gemberling et al., 1980).  
Although it had previously been thought that associative learning 
could not take place until a certain level of postnatal maturation 
had been reached (Campbell & Coulter, 1976), subsequent research 
has demonstrated associative learning in rat pups as young as l day 
old (Johanson & Hall, 1979). 

    Numerous methods are available for the study of learning and 
memory in older animals.  Different kinds of learning such as 
simple associative learning (Pavlovian conditioning) and more 
complex forms of learning (such as problem solving in a Hebb-
Williams maze, puzzle boxes, or tests of reasoning activity) have 
been studied (Rodier, 1978).  The conventional criteria for 
learning include speed of learning, number of correct responses, 
and duration of latency.  Data from such studies are interpreted 
more precisely, if behaviour in the course of learning is well 
analysed and spontaneous activity occurring during the testing 
period is carefully recorded. Analysis of the inter-trial interval 

on freely occurring, non-contingent responses can also provide 
valuable information about the presence of a neurobehavioural 
deficit (Frankova & Barnes, 1968). 

    Habituation, an example of non-associative learning, is defined 
as a decrease in response following repeated presentation of a 
stimulus.  The development of habituation reflects the functional 
maturation of structures responsible for the development of 
inhibitory processes (Bronstein et al., 1974).  The ontogeny of 
habituation has been studied by Campbell & Stehouwer (1980), who 
demonstrated ontogenic changes in habituation, learning, and 
retention during the suckling period.  Methods for the assessment 
of learning and memory in neonatal animals have been discussed by 
Adams & Buelke-Sam (1981).  Social behaviour

    The development of social behaviour has mainly been studied in 
primates.  Although rodents have not been used traditionally, the 
development of social behaviour can be assessed in this species, 
even during the preweaning period. The type and duration of 
contacts in the home cage or in a novel environment of a group (or 
pair) of pups can be recorded.  Play activities, early sexual 
manifestations, and agonistic activities, are other endpoints that 
might prove useful (Frankova, 1973). 

    Social behaviour in older animals has not been used extensively 
to indicate developmental toxicity.  Basic information on the type 
and characteristics of social interactions may be obtained by 
analysing contacts between a pair of animals of the same sex placed 
in the novel environment.  Over 40 elements (acts and postures) 
have been described by Grant & Mackintosh (1963).  The two most 
common social behaviours investigated are dominance and aggression. 

    Sex-dependent behaviour represents a complex set of 
motivationally different activities, such as courtship, pair 
formation, reproduction, and parental activities (Gerall & McCrady, 
1970; Frankova, 1977).  This category of behaviour depends on the 
stage of development, on environmental stimuli, as well as on the 
hormonal status of the organism. 

4.2.3.  Relevance of behavioural studies for human risk assessment

    There is a current requirement to screen agents for 
developmental neurotoxicity in some countries; similar rules are 
also being considered in other countries.  In view of the already 
enormous cost of safety evaluations (Johnson, 1981), great care 
must be taken to ensure that the studies provide meaningful 
information regarding potential hazard.  The following requirements 
must be fulfilled, before routine assessment of the effects of 
prenatal exposure to chemicals on behaviour can be generally 

    (a)  the test(s) should be sensitive to neurobehavioural
         alterations produced by a wide range of test agents
         or conditions;

    (b)  the test(s) should be standardized and validated so
         that accurate and reproducible results are provided
         within and between laboratories;

    (c)  the test(s) should be cost effective;

    (d)  the test(s) should be able to recognize chemical
         agents known to be neurotoxic to man.

    There are several instances in which the neurobehavioural 
function of human progeny has been reported to be altered following 
 in utero exposure to environmental agents (Tanimura, 1980).  The 
number of examples in which there is a correspondence between 
observed neurobehavioural defects in human beings and effects 
predicted from animal studies is limited to a few agents (i.e., 
methylmercury, lead, psychoactive drugs).  In the case of 
methylmercury, similar types of neurobehavioural deficits were 
detected in human beings and experimental animals after prenatal 
exposure (Reuhl & Chang, 1979). 

    The general lack of association reflects the need for more 
careful epidemiological examinations and additional basic research 
using animal models and agents suspected of being neurotoxic. 

    As developmental neurobehavioural toxicology is a recently 
developed field of research, not enough evidence has accumulated to
convincingly demonstrate a parallel between effects in animals and
human beings.  Manifestation of similarities may often be obscured
by inadequate testing or interpretation of data, by interspecies 
differences in behavioural patterns or in the plasticity of the 
central nervous system.  Behavioural techniques have been widely
used to detect and quantify neurobehavioural alterations in adult 
animals exposed to many environmental agents.  There is good reason
to believe that these techniques can also be used to assess 
neurobehavioural changes following prenatal exposure. The fact that
the relevance of animal data to human data is less questionable in
the field of adult neurotoxicity is at least partly due to the fact
that more information has been accumulated in this more traditional

4.3.  Reproduction

    Reproduction depends on structural and functional integration 
of various systems.  Consequently, there are many ways in which 
reproduction may be impaired directly or indirectly.  The 
development of the reproductive organs takes place over a period 
extending beyond birth, and structural as well as functional 
maturation does not occur until puberty. Because of the long 
quiescent phase between initial formation of the basic structures 
and adoption of their final functional form, defects may not become 
apparent until a considerable time after induction, unless they are 
so gross as to be considered malformations.  Even then, certain 
changes may be so extreme that they lead to incorrect determination 
of sex rather than to identification of a developmental error.  The 
error is only discovered later; in man, this can be 14 - 20 years 
after induction. 

4.3.1.  Normal gametogenesis and development of the genital tract

    In both sexes of mammalian species, the reproductive organs 
consist of gonads, an internal duct system, and external genitalia.  
Since these components arise from diverse primordia, and 
contributions from these primordia vary qualitatively and 
quantitatively in the two sexes, the development of the 
reproductive system is particularly complex.  A detailed 
description of the stages of development is beyond the scope of 
this monograph and discussion is restricted to the stages most 
vulnerable to chemical action. 

    Primordial germ cells of both male and female, are clones of 
special cells present in the ectoblast at the early bilaminar 
stage.  In later stages of development, they are transferred into 
the endoderm of the yolk sac.  From there, the primordial germ 
cells migrate into the mesoblastic undifferentiated primordia of 
gonads located on the medioventral surface of the urogenital 
ridges.  In individuals with normal Y chromosome expressing H-Y 
antigen, gonads differentiate into testes.  In individuals without 
the Y-chromosome and H-Y antigen, the gonadal anlagen give rise to 
fetal ovaries.  The primordial germ cells in the ovaries develop 
into oogonia and undergo repeated mitosis differentiating into 
oocytes.  The latter enter meiotic prophase and become invested by 
granulosa cells to produce primary follicles.  Thus, the oocytes in 
postnatal mammals are at an arrested stage of meiotic division; 
diffuse diplotene, dictyate, dictyotene, or resting stage 
(Erickson, 1967).  A complete oocytic envelopment by granulosa 
cells separates the oocytes from the ovarian stroma. 

    The first meiotic metaphase is completed much later after birth 
shortly before ovulation.  In human beings, the number of oogonia 
and oocytes between fetal weeks 18-22 is estimated at 5-7 million, 
in the newborn there are approximately 2 million oocytes, and at 
the age of 18-20 years about 200 000. From the whole population of 
human oocytes only approximately 400 undergo ovulation, assuming 
regular ovulation; all the others degenerate. 

    The embryonal testis contains testicular cords composed of 
primitive Sertoli cells and spermatogonia.  The embryonal 
spermatogonia originate from the primordial germ cells while 
Sertoli-cells are mesodermal in origin.  In the prenatal period, 
spermatogonia, in contact with Sertoli-cells, divide slowly by 
repeated mitoses, they do not differentiate into spermatocytes, and 
do not undergo meiosis.  Spermatogenesis and the onset of meiosis 
in the testis follows maturation of the Sertoli-cells, a step under 
hormonal control (follicle-stimulating hormone (FSH), luteinizing 
hormone (LH), and testosterone), which begins at puberty. 

    Synthesis of male hormones by the fetal testis is critical for 
the growth and differentiation of male accessory organs during the 
prenatal as well as postnatal periods.  The male phenotype is 
mainly imposed on the embryo by two types of hormones, produced in 
the testis, i.e., androgenic steroids and the Müllerian-inhibiting 
factor (MIH).  Androgens, produced by Leydig cells, stimulate the 

development of the male reproductive ducts from the epigenital 
portion of the mesonephros and from the Wolfian ducts, and the 
development of the male external genitalia from structures located 
around the urogenital sinus.  MIH, a non-steroid factor, produced 
by the embryonal testis causes the Müllerian ducts to regress.  The 
epigenital mesonephric nephrons and Wolfian duct are forerunners of 
the epididymides, vasa deferentia, seminal vesicles, and 
ejaculatory ducts.  The urogenital sinus differentiates into the 
urethra and gives rise to the prostate; derivatives of the genital 
tubercle and of the urethral plate become the penis; the 
labioscrotal swelling turns into the scrotum.  Testosterone plays 
an important role in the sexual differentiation of the Central 
nervous system (CNS) during the perinatal period. 

    Testosterone is synthesized in the testis from pregnenolone.  A 
series of steps mediated by enzymes are involved in the conversion 
of pregnenolone to testosterone and 5a-dihydrotestoterone, the 
active derivative of testosterone in some organs.  Interruption of 
any of these steps during the critical period in pregnancy would 
lead to a deficiency of androgens which, in turn, might lead to the 
development of equivocal secondary sex characteristics 
(pseudohermaphroditism) with normal sex chromosomes.  Thus, a 
decreased amount of androgen can fail to masculinize a male fetus, 
which might have malformed ambiguous external genitalia, hypospadia 
(pseudovaginal pseudoscrotal in some cases), but normal testes and 
the XY karyotype.  In severe cases of testicular dysgenesis, if the 
testes fail to produce MIH, the Mullerian ducts may persist and 
develop into a uterus and vagina. In addition, male efferent ducts 
may be present in some cases. 

    Feminization of a male may be related to the absence of, or 
deficiencies in, testosterone receptors.  In human males exhibiting 
errors in testosterone receptor, the phenotype is female despite a 
normal male 46,XY karyotype, and the presence of testes and 
regressed Müllerian ducts. 

    In the female, the Wolfian ducts regress and the Müllerian 
ducts differentiate into the oviducts, uterus, and the upper 4/5 of 
the vagina.  The urogenital sinus gives rise to the vestibule and 
the lower l/5 of the vagina, while the genital tubercle and the 
urethral plate develop into the clitoris and labia minora.  The 
labioscrotal swelling turns into the labia majora.  Female fetuses 
can be virilized or masculinized as a consequence of congenital 
defects in the synthesis of glucocorticoids or by the 
administration of androgens.  In the normal female fetus, gonadal 
synthesis of testosterone or estrogens is not apparent and there is 
no evidence that the interstitial cells or follicular apparatus of 
the fetal ovary play any role in female genital differentiation. 

    In the absence of male gonadal function, the internal and 
external genital anlagen differentiate according to the female 
pattern.  MIH, a protein of relatively high relative molecular 
mass, prevents the development of the uterus and the fallopian 
tubes in males.  The function of testicular testosterone is (i) to 
stimulate the growth of the epididymis, vasa deferentia, and 

seminal vesicles, and (ii) to promote the stabilization/
transformation of the genital sinus and urogenital tubercle, 
urethral plate and labioscrotal swellings into the prostate, male 
urethra, penis, and scrotum, respectively, and (iii) to promote the 
descent of the testes. Prior to the male differentiation, both the 
urogenital sinus and external genital anlagen acquire the capacity 
to convert testosterone to 5a-dihydrotestosterone (DHT) which, in 
these tissues, is thought to be the effective androgen (Siiteri & 
Wilson, 1974). 

    Postnatally, in both sexes, normal pituitary gonadotrophin 
secretion appears to be required for the maturation of germ cells 
and their supporting cells.  The pattern of FSH and LH secretion 
reflects the gradual maturation of a functional hypothalamic 
pituitary unit, responsive to feedback inhibition by sex steroids.  
Sexual differences in the pattern of gonadotrophin secretion are 
well recognized; cyclic pulsatile in females, tonic in males. 

    Spermatogenesis continues throughout reproductive life. It 
starts with stem cells (type A spermatogonia) which, in the rat, 
undergo 5 mitotic divisions followed by two meiotic divisions to 
produce spermatids.  The latter change into spermatozoa, which move 
to, and mature in, the epididymis. 

4.3.2.  Mechanisms of abnormal development

    Adverse effects on the reproductive function of offspring, 
resulting from prenatal exposure to environmental chemicals, have 
not received sufficient attention.  Examples of reproductive 
effects that have been related to maternal chemical exposure, are 
given in Table 10.  Chemicals may affect either the gonadal tissues 
directly or antagonize the action of growth-promoting or growth 
inhibiting factors. However, the precise mechanisms of action are 
not known. 

    When the complexity of sexual development, gametogenesis, and 
reproductive tract differentiation is considered, it becomes 
apparent that prenatal and early postnatal chemical exposure is a 
special toxicological problem.  The gonads, the reproductive 
accessory organs, and the neuroendocrine system are all vulnerable 
to toxicity induced by chemicals acting through unique mechanisms.  
The specificity of male and female developmental processes also 
accounts for differences in susceptibility to toxic agents between 
the sexes. 

    Dixon (1982) has recently pointed out that there are a number 
of periods during which the developing reproductive system is 
susceptible to chemical insults.  Interference by xenobiotics can 
lead to morphological, biochemical, physiological, and/or 
behavioural disorders as well as to the ineffective integration of 
the biological processes required for successful reproductive 
performance.  Though reduced fertility in the offspring may be the 
most obvious consequence of prenatal exposure to toxic 
environmental chemicals, more subtle effects, such as those 
involving secondary sexual characteristics and behaviour, are also 

    In mammals, the female fetus is especially vulnerable to agents 
that damage the germ cells, because, in most species, all oocytes 
develop prenatally; no (or few) primordial germ cells are formed 
after birth.  Therefore, chemicals that affect oogenesis may be 
expected to have a lasting effect on the fertility of the female.  
Particular periods of susceptibility during oogenesis with long-
term toxicological effects on fertility have been shown to be 
connected with peak oocyte DNA synthesis periods (McLachlan & 
Dixon, 1973).  Fetal ovaries have been reported to be a target site 
for the action of some environmental pollutants, particularly 
polycyclic aromatic hydrocarbons (MacKenzie et al., 1979a,b). 

    Unlike germ cell production in the mammalian female, 
spermatogenesis is initiated in the pubertal male, and germ cells 
are produced continuously throughout adulthood. Prenatal exposure 
to several compounds, including benzo( a)pyrene, 7,12-
dimethylbenz( a)anthracene (DMBA), and diethylstilboestrol (DES), 
has been reported to alter male fertility (McLachlan et al., 1975; 
Davis et al., 1978; Gill et al., 1979; MacKenzie et al., 1979a).  
Some compounds act directly on the germ cells, probably at the 
level of DNA, RNA, and protein synthesis; other chemicals act by 
interfering with the normal hormonal regulation of spermatogenesis. 
Table 10.  Examples of reported abnormalities of the reproductive system
in the offspring of mothers treated with chemicals during pregnancy
Type of abnormality             Species  Chemical agent        Reference
Degeneration and                rabbit   cyclophosphamide      Gerlinger & Clavert (1964)
reduction of gonocytes
in male and female fetuses

Reduced fertility in males      mouse    cyclophosphamide      Sotomayor & Cumming (1975)
and females

Gonadal dysplasia in            rat      busulfan              Forsberg & Olivecrona (1966)
males and females

Clitoral hypertrophy and        rat      metyrapone            Goldman (1967)
adrenal hyperplasia in female
Decreased incidence in          rat      2,3,7,8-tetrachloro-  Khera & Ruddick (1973)
pregnancy and reduced                    dibenzo- p
litter size                              -dioxin (TCDD)

Reduced fertility in females    mouse    procarbazine          McLachlan & Dixon (1973)

Reduced fertility in males and  mouse    benzo( a)pyrene       Mackenzie et al. (1979a)
females and gonadal dysplasia

    Though the exact mechanisms by which chemicals can alter 
neuroendocrine function are not clearly understood, two possible 
neuroendocrine effects have been suggested that may result in 
impaired reproduction (McLachlan et al., 1981).  The first involves 
an action on endogenous catecholamines which, secondarily, 
influences the ability of the sympathetic nervous system to 
stimulate gonadotrophin release; the second involves the more 
direct action of a chemical at specific neuroendocrine centres. 

    Chemicals may also affect hormone action by altering steroid 
synthesis, by affecting the activity of steroid-metabolizing 
enzymes, or by interfering with hormone-receptor interactions. 

    Since reproduction is a highly integrated function that 
involves a behavioural component, compounds that alter behaviour 
may result in infertile animals, even in the absence of direct 
toxic effects on the genital tract or the endocrine system (see 
section 4.2 on behaviour).  Development of the integrative 
behaviour necessary to reproduction may be susceptible to 
disruption by psychotropic agents or hormonally-active xenobiotics. 

    Effects on reproduction may also be mediated by prenatally- or 
perinatally-induced disturbances in liver function.  Compounds that 
alter the hepatic metabolism of gonadal hormones can alter the 
reproductive characteristics of the affected individual.  The long-
lasting effects of neonatal exposure to polychlorinated biphenyls 
(PCB) and 2,3,7,8-tetrachlorodibenzo- p-dioxin (TCDD) on hepatic 
function (Lucier et al., 1975, 1978; Lucier & McDaniel, 1979) 
suggest that  in utero exposure to environmental agents may result 
in altered enzyme activity. 

4.3.3.  Testing procedures

    A detailed discussion on methods for assessing effects of 
chemicals on reproductive function has recently been published by 
the Scientific Group on Methodologies for the Safety Evaluation of 
Chemicals (SGOMSEC) established jointly by WHO and SCOPE/ICSU and 
supported by IPCS (Vouk & Sheehan, 1983). 

    Potentially all of the processes in reproduction may be 
affected by chemicals.  Fecundity tests assess spermatogenesis, 
sperm maturation, sperm transport, semen production, and mating, in 
the male, and ovulation and oviductal transport, fertilization, and 
implantation, in the female.  These are examined collectively by 
measuring reproductive outcome.  Results of chemically-induced 
reproductive failure must take into account the health of the 
animals after chemical exposure.  When evaluating chemicals for 
effects on reproduction, it is critical to determine the place of 
observed reproductive effects within the chemical's general 
toxicity profile.  For transplacental toxicity studies, effects on 
reproductive performance in offspring are most meaningful in the 
absence of maternal toxicity and gross malformations in the young. 

    Testing of chemicals for specific endpoints is likely to 
improve the long-term understanding of the processes affected by 
toxicity.  This is an important area for further development.  
However, one of the easiest measures of reproductive capacity is 
the ability of male and females to mate and produce normal 

    The following is an outline of a few of the reproductive tests 
that involve exposure to chemicals during either the prenatal or 
pre- and postnatal periods.  Many of the comments made in section 
3.2 are relevant here and will not be repeated.  "Fertility test" on progeny following prenatal exposure

    If a "fertility test" is to be conducted in the progeny of mice 
or rats following prenatal exposure, the chemical is administered 
over a period extending from implantation until the end of 
pregnancy.  The progeny are allowed to litter normally and their 
postnatal development is observed until sexual maturity.  At 
adulthood, the males and females of the test progeny are mated with 
animals that have not been previously exposed to the test chemical.  
All pregnant females are killed at term.  Data on incidence of 
pregnancy, and prenatal values are evaluated for toxic effects.  Up 
to the present, such tests have only been used for second-stage 
testing and have not been performed routinely.  This approach is 
included in multigeneration studies.  Multigeneration studies

    There is no general agreement about the most satisfactory 
design for multigeneration tests or even the number of generations 
to be studied (Palmer, 1981).  The FDA Advisory Committee (1970) 
reported that with a range of substances, including 
organophosphorus anticholinesterases, herbicides, and pesticides, 
adverse effects were occasionally seen after two, three, or even 
four generations that were not clearly evident earlier.  However, 
it is unlikely that genuine effects would be recorded in later 
generations that were not evident in the first or second 
generations (Clegg, 1979; Leeming et al., 1982).  Until recently, a 
minimum of two litters per generation was usually recommended, 
though some workers feel that the maximum possible number of 
litters should be obtained in one of the generations, to determine 
whether there is any overall effect on the maximum reproductive 
capacity.  However, supportive data for such large studies have not 
been published.  In some designs, some of the pregnant dams are 
killed in each generation so that the offspring can be examined, at 
term, for teratogenic effects.  Some of the offspring of the third 
generation are subjected to full histopathological examination. 

    No single ideal test exists for the wide range of possible 
compounds that will be tested.  For substances that show absolutely 
no adverse reproductive effects at doses up to those toxic for the 
parent, a two-generation, one litter per generation test, would be 
satisfactory.  For other compounds that show a small, but 
increasing effect with time, more litters and more generations may 

be required to establish a no-observed-adverse-effect leve1.  In 
practice, if only one litter is produced per generation, it is 
often very difficult to be certain that no effect at all is 
present.  The preferred basic study design is therefore a two-
generation, two-litter study with three dose levels of the test 
compound, plus a control group.  The parental generation should be 
treated for at least 9 weeks before the second mating, and 
treatment continued throughout the study.  One advantage of 
deriving the F2 generation from the F1a, rather than from the 
traditional F1b generation, is that an opportunity is available to 
investigate in detail any effects that may be observed during the 
mating of the F1a animals.  Reproductive performance from mating 
first litters should be normal, if the animals are not too young at 
first mating.  Normally, not less than 20 males and 20 females 
should be treated and mated per group.  Mating brothers and sisters 
is not recommended in the last generation.  For the earlier 
generations, larger group sizes should be used to reduce the risk 
of a concurrent inbreeding bias (Palmer, 1981).  All offspring 
should be reared to weaning, though it is acceptable to reduce the 
litters to a standard size if desired.  At weaning, offspring not 
involved in producing subsequent generations should be killed and 
examined macroscopically.  In the absence of any gross changes, it 
is doubtful if there is much to be gained by random measurements of 
organ weights and histological examination.  However, these 
procedures must be performed if any toxic effect is suspected.  It 
may be advantageous to fix and store the tissues from one male and 
one female from each litter, at each killing stage, until the study 
is completed, so that these tissues can be more carefully examined 
if unexpected toxicity is revealed in the final stages of the 
study.  Choice of species

    For economic and temporal reasons, it is usual to perform the 
above reproduction studies on mice or rats.  Ideally, in assessing 
reproductive toxicity, it should be shown that the species chosen 
for the reproductive studies metabolize the substances under test 
by the same routes as man.  In practice, this is not always 
possible (especially with pesticides), but efforts should be made 
to obtain this information and perform the tests in this way.  
Where equivocal results are found in rodent studies, tests in other 
species may be of value.  There is no evidence that the results 
from one particular species or strain extrapolate best to man.  
Thus, there are no rigid requirements regarding choice of species, 
but it is very important that adequate control data on reproductive 
performance should exist for the species used.  Doses, route, and duration of treatment

    The choice of suitable dose levels is one of the most difficult 
problems in the design of reproduction tests.  The highest dose 
should induce some signs of general toxicity such as reduced gain 
in body weight.  The lowest dose should be a no-observed-adverse-
effect level.  Since, usually, only a small percentage of animals 

are adversely affected, a dose-response relationship is very 
important in assessing results.  Three dose levels are normally 
required, unless the highest practical dose does not induce any 
adverse reproductive effect. 

    The route of administration should be relevant to that in the 
human exposure.  Inhalation exposure for reproductive toxicology 
has not been widely used and can give rise to misleading results, 
if undue stress is involved.  In such circumstances, alternative 
routes of administration may be used. 

    For environmental chemicals such as pesticides, administration 
throughout the whole period of gestation seems to be the most 
appropriate, but other treatment regimens are also acceptable.  Presentation of results

    There is no general agreement on the best statistical method 
for the analysis of the results of reproductive studies.  In the 
presentation of results, appropriate statistical methods, depending 
on the distribution of particular endpoint measured, can be used 
(Snedecor & Cochran, 1967; Sokal & Rohlf, 1969; Hollander & Wolfe, 
1973), but it is essential that all of the data for both individual 
fetuses and litters are presented, so that other methods of 
assessment can be used if necessary.  In all studies, it should be 
possible to trace the complete outcome of pregnancy in each female. 

4.4.  Transplacental Carcinogenesis

    Transplacental carcinogenesis is defined as the appearance of 
neoplasia in the progeny of females exposed to chemical agents 
during pregnancy.  This phenomenon has been demonstrated in rats, 
mice, hamsters, rabbits, pigs, dogs, and monkeys with approximately 
60 chemicals (Table 11) including representatives of the major 
classes of chemical carcinogens such as nitroso compounds, 
polycyclic aromatic hydrocarbons, aminoazo compounds, mycotoxins, 
and others.  While practically all sites of the developing fetuses 
can be affected by carcinogens administered during pregnancy, the 
tissues most susceptible to neoplastic responses are those of the 
nervous system, kidney, and lung.  Usually, prenatal susceptibility 
to carcinogens is greater in the advanced stages of organogenesis; 
in rodents, for example, it is confined to the last third of the 
gestational period.  With some exceptions specific for certain 
species or strain (lung adenomas in Swiss mice and some tumours in 
patas monkeys), tumours do not usually appear before adult life.  
In at least some tissues, effects of transplacental carcinogens may 
be enhanced by the application of a carcinogen or promotor to the 
progeny.  In addition, Tomatis (1979) has reported that increased 
tendency to develop neoplasia may persist without additional 
exposure, although greatly diminished, in subsequent generations of 
animals.  The same group (Tomatis et al., 1981) has suggested that 
this transmission may occur via the germ cells, and this is 
supported by the data of Nomura (1982). 

    Twenty three compounds and groups of chemicals and 7 industrial 
processes have been shown to induce carcinogenic effects in human 
beings (IARC, 1982).  However, there is convincing epidemiological 
evidence of transplacental tumour induction in man for only one 
compound, i.e., diethylstilboestrol (Herbst et al., 1979a; Rice, 

    The following is a summary of the main features of 
transplacental carcinogenesis in laboratory animals, the research 
perspectives offered by models under development, and advice on the 
need for testing compounds in this field. 
Table 11.  Substances reported to be transplacental carcinogens 
in animalsa
Polycyclic aromatic hydrocarbons   7,12-dimethylbenz[ a]anthracene
                                   benzo[ a]pyrene

Aminoazo compounds                 2-fluorenylacetamide

Carbamates                         urethane

Natural products of plant and      aflatoxins
animal origin                      pyrrolizidine alkaloids

Nitroso compounds                  nitrosodimethylamine
                                   nitrosobis(2-hydroxypropyl) =

Table 11 (contd).
Nitroso compounds resulting from   amidopyrine + sodium nitrite
interaction of the following       diethylamine + sodium nitrite
precursors                         methylurea + sodium nitrite
                                   ethylurea + sodium nitrite
                                   butylurea + sodium nitrite
                                   isopropylurea + sodium nitrite
                                   methylbenzylimidazolcarbamate +
                                     sodium nitrite
                                   citrulline + sodium nitrite
                                   propylhexedrine + sodium nitrite

Other alkylating compounds         azoxymethane
                                   dimethyl sulfate
                                   diethyl sulfate
                                   5-(3,3-dimethyl-1-triazeno) =
                                   3,3 dimethyl-1-pyridyltriazene
                                   propane sultone

Drugs                              diethylstilbestrol

Miscellaneous                      tobacco smoke condensate
                                   4-oxyphenyllactic acid
                                   vinyl chloride
                                   4-nitroquinoline  N-oxide
a  Data taken from the review by Rice (1981) with additional references
   for following compounds: nitrosoethylurethane, 3,3-dimethyl-1-
   pyridyltriazene (Druckrey, 1973), 2-fluorenylacetamide (Armuth & 
   Berenblum, 1977), benzidine (Vesselinovitch et al., 1979), zineb 
   (Kvitnizkaja & Kolesnicenko, 1971), pyrrolizidine alkaloids 
   (Schoental & Cavanagh, 1972), dinitrosopiperazine (Börzsönyi et 
   al., 1980), nitrosobutylurea (Maekawa & Odashima, 1975), 
   nitrosobutylurethane (Maekawa et al., 1980), 5-(3,3-dimethyl-1-
   triazeno)imdazol-4-carboxamide (Zeller, 1980), amidopyrine + 
   sodium nitrite (Alexandrov & Napalkov, 1979), diethylamine + 
   sodium nitrite (Vesselinovitch & Rao, 1974), methylurea + sodium 
   nitrite (Alexandrov, 1973), butylurea + sodium nitrite (Maekawa 
Table 11 (contd.)

a   et al., 1977), isopropylurea + sodium nitrite (Schneider et al., 
   1977a), methylbenzylimidazolcarbamate + sodium nitrite (Börzönyi 
   et al., 1976), citrulline + sodium nitrite (Ivankovic, 1979), 
   propylhexedrine + sodium nitrite (Schneider et al., 1977b), 
   ethylmethanesulfonate (Schneider et al., 1978), 
   diethylstilboestrol (Rustia & Shubik, 1976; Napalkov & Anisimov, 
   1979), carcinolipin (Sabad et al., 1973), cyclophosphamide 
   (Roschlau & Justus, 1971), tobacco smoke condensate (Nicolov & 
   Chernozemsky, 1979). 

4.4.1.  Principles and mechanisms of action  Comparative sensitivity of the adult and fetal
          organism to carcinogens

    All the transplacental carcinogens that have been studied were 
already known to be carcinogenic in adult animals. However, the 
ability of a substance to induce tumours transplacentally can 
differ in a quantitative way from that observed in weanling or 
adult animals.  For example, it has been established in rats that 
fetal sensitivity to the action of nitrosoethylurea (NEU) is 20-50 
times higher than that of adult animals (Ivankovic & Druckrey, 
1968).  On the other hand, both fetal and adult rats (Alexandrov, 
1969) showed similar sensitivity to nitrosomethylurea (NMU), while 
sensitivity to the action of nitrosodimethylamine (NDMA) proved to 
be 6-10 times lower in the fetus than in the adult animal 
(Alexandrov, 1968).  Thus, it appears that the action of chemical 
carcinogens may be stronger or weaker, after passing through the 
placenta.  A possible explanation is that physico-chemical 
characteristics of chemical agents affect the rate of transport 
across the placenta and the biotransformation in the mother and 

    Negative results in testing for transplacental carcinogenic 
effects should be evaluated with caution.  It should be kept in 
mind that the period of development when the rodent embryo is 
sensitive to carcinogenic stimuli is rather short (10-12 days) and 
that some compounds require more prolonged application in order to 
induce tumours.  Dependance on the stage of prenatal development

    The period of development sensitive to carcinogenesis starts in 
the advanced stages of organogenesis and lasts until the end of 
intrauterine life.  This has been observed in at least 3 animal 
species, i.e., the rat (Ivankovic & Druckrey, 1968; Napalkov & 
Alexandrov, 1968; Alexandrov, 1974), the mouse (Nomura, 1974), and 
the hamster (Mohr et al., 1975). However, some carcinogens 
administered to pregnant animals during the early period of 
embryogenesis, when the placenta has not yet formed, may also 
induce tumours in the progeny (Spatz & Laqueur, 1967; Nomura, 1974; 
Börzsönyi et al., 1976; Stavrou et al., 1977). 

    Generally, transplacental carcinogenesis does not reveal stage 
specificity for any site in terms of the development and 
histological appearance of tumours.  For example, many chemicals 
almost exclusively induce tumours of the nervous system and kidney 
in the rat and of the lung in the mouse, irrespective of the day of 

    Susceptibility to transplacental carcinogenesis is influenced 
by genetic predisposition and appears to be time-specific during 
pregnancy.  Examples of this phenomenon are to be found in rat and 
mouse studies (Druckrey, 1973; Rice, 1973a; Vesselinovitch, 1973; 
Alexandrov & Napalkov, 1981).  There may be periods of maximal 
susceptibility that vary from one organ system to another.  In the 
rabbit, for example, the progeny of animals given NEU early in 
gestation developed brain tumours whereas exposure later in 
gestation caused only nephroblastoma (Stavrou & Lübbe, 1974; 
Stavrou et al., 1977).  Species and strain specificity

    Studies with carcinogenic substances have shown that the type 
of tumours induced in the progeny differs from species to species.  
The same agent may induce tumours at different sites, depending on 
the animal species or strain (Druckery, 1973; Napalkov, 1973; Rice, 
1973b; Rice et al., 1979; Ivankovic, 1979).  The type of tumours 
induced transplacentally may or may not necessarily be similar to 
that induced in adult animals.  Among animal species, the most 
pronounced similarity has been demonstrated in mice (Rice, 1969; 
Vesselinovitch, 1973; Diwan & Meier, 1974). 

    Strain dependence has been revealed, even when testing highly 
organotropic substances.  Urethane, which displayed pulmonotropic 
carcinogenic properties in adult mice, induced transplacentally 
both lung cancer and tumours at other sites, typical of the strain 
tested (Vesselinovitch, 1973).  The hepatocarcinogen 
orthoaminoazotoluene, which induced liver tumours transplacentally 
in certain mouse strains, also often induced tumours of the breast 
and lung in others (Gelstein, 1961; Golub et al., 1974; 
Kolesnicenko et al., 1978).  Thus organotropism in mice is not only 
determined by the carcinogenic properties of the compound, but also 
by strain-specific susceptibility factors.  The stage of 
embryogenesis is also sometimes a determining factor. 

    In rats, the great majority of substances tested transplacentally, 
with the exception of symmetrical dialkylnitrosamines, tend to 
display a pronounced neurotropic carcinogenic effect (Alexandrov, 
1976).  This great susceptibility of the fetal rat nervous system 
to transplacental tumour induction by different carcinogenic 
substances is noteworthy.  Kidney tumours also often appeared in 
rat progeny as a result of transplacental exposure to a number of 
compounds including 7,12-dimethylbenz(a)anthracene (DMBA), NDMA, 
 N-nitrosodiethylamine, NMU, azoxymethane, methylazoxymethanol, 
1-methyl-2-benzylhydrazine, procarbazine, and 3,3-dimethyl-1-
phenyltriazene (Alexandrov, 1976).  It should be noted that all 
these carcinogenic substances, with the exception of DMBA, are also 
able to induce kidney tumours in adult animals. 

    Results of a few studies on rabbits have indicated that the 
nervous system and kidney of rabbits are sensitive to develop 
tumours (Güthert et al., 1973; Stavrou & Lubbe, 1974; Beniasvili, 
1978).  Following transplacental exposure to NEU, monkeys suffered 
from kidney, soft connective tissue, lung, and nervous system 
tumours (Rice et al., 1977).  Hamsters have been used for assessing 
the transplacental effects of cigarette condensates and hormones 
including diethylstilboestrol (Nicolov & Chernozemsky, 1979; 
Rustia & Shubik, 1976).  Mechanisms of organotropism

    The mechanism of organotropism displayed by the majority of 
transplacental carcinogens is poorly understood.  Some factors that 
are believed to contribute to organotropism are: 

    (a)   The predisposition of somatic cells: Little is known
         about the causes of variation in predisposition and
         susceptibility among various cell types (Kleihues
         et al., 1979a);

    (b)   Metabolic conversion: The extent of metabolic
         conversion of indirect carcinogens in the whole 
         organism or in the target cells may influence the 
         carcinogenic outcome (Alexandrov & Napalkov, 1981);

    (c)   Carcinogen-DNA interaction: Cellular DNA is a
         principle target for carcinogenic chemicals. Studies
         in transplacental carcinogenesis with ENU, NMU, DMBA,
         and 3,3-dimethyl-1-phenyltriazene suggest that
         neoplastic transformation may depend on the degree
         and persistance of DNA damage in the target tissues
         (Goth & Rajewsky, 1974; Cooper et al., 1978; Doerjer
         et al., 1978; Kleihues et al., 1979a; Lihachev et
         al., 1983).  The state of differentiation of target
         cells at the time of exposure may be important, since
         the capacity for DNA repair has been shown to change
         gradually as cells approach the final stages of
         differentiation (Counis et al., 1977);

    (d)   Proliferative activity: A relationship has been
         established between the stage-dependence of the
         number of lung adenomas induced in fetal mice by
         exposure to NEU and changes in the number of
         epithelial cells within the population of cycling
         cells (Kauffman, 1976).  Cells with a high rate of
         proliferation may be more vulnerable to the action of
         carcinogenic stimuli as the induced damage is, thus,
         quickly amplified.

    All these factors are closely interrelated since the levels of 
the enzymes, cell differentiation, and proliferative activity in 
the target organs are genetically controlled. Altogether, these 
factors determine both the general sensitivity of the growing 
organism and the stage specificity of target organs to the insults 
by carcinogenic agents (Alexandrov & Napalkov, 1981).  Metabolism of chemical carcinogens in the maternal
organism, the placenta, and the embryo

    Tumour development in the offspring is generally determined by 
the rate of placental transfer, and the concentration of the 
carcinogen or its active metabolites in the embryonic tissues.  It 
is widely acknowledged that almost all non-ionic organic compounds 
with a relative molecular mass of less than 600, can cross the 
placenta (Ginsburg, 1971). There are almost no carcinogenic 
substances with a relative molecular mass greater than 300 and many 
are soluble in both lipids and water, which facilitates their 
transfer across the placenta. 

    There is no absolute correlation between the accumulation of a 
given carcinogen in the fetus and the neoplastic response.  For 
example, following administration to pregnant rats, NDMA, a 
metabolism-dependent carcinogen in adult rats, reaches a high 
concentration in the fetal tissues, where it remains for almost 24 
h (Shendrikova & Alexandrov, 1978). Despite this, however, NDMA is 
a weak transplacental carcinogen, which suggests that the compound 
is poorly metabolized by embryonal tissues, probably because of the 
immaturity of the relevant enzymes.  Studies with other carcinogens 
(polycyclic hydrocarbons and nitroso compounds) have also shown 
that the concentration of parent chemicals in various embryonic 
tissues did not correspond with the subsequent incidence of 
neoplasia in these tissues (Baranova & Alexandrov, 1978). 

    From experimental data, it appears likely that a transplacental 
carcinogen acts on the target fetal cells in the following ways 
(Alexandrov & Napalkov, 1981): 

    (a)  Directly, by reaching the fetus either in its original 
form or as a spontaneously-formed proximal carcinogen.  Examples 
are nitrosamides, methylazoxymethanol, dimethyl- and 
diethylsulfate, propane sultone, methyl-methanesulphonate.  The 
stability of the initial compounds or their proximal derivatives is 
the limiting factor. Methylnitrosourethane (MNU) when given to 
pregnant rats failed to induce carcinogenic effects in the progeny, 
because of its rapid degradation in the mother.  However, the 
carcinogenic effect was readily induced following intra-amniotic 
administration and direct injection into fetal rats (Alexandrov, 
1972; Napalkov, 1973); 

    (b)  Indirectly, by generating proximal carcinogens in target 
fetal cells through metabolic activation.  Examples are 
dialkylnitrosamines, l,2-dialkylhydrazines, azo- and azoxy-alkanes, 
and procarbazine.  In this case, the metabolic capacity of the 
target cells is a limiting factor.  In particular, NDMA, when 
administered to pregnant rats, induced no teratogenic effects and 
only a weak transplacental carcinogenic effect.  These findings 
were ascribed to the incomplete metabolic development of the rat 
fetuses (Alexandrov, 1968; Napalkov, 1973).  When the active 
metabolites of enzyme-dependent carcinogens are short-lived, the 
enzymatic capability of the target fetal cells acquires primary 

    (c)  Indirectly, following maternal metabolism.  A stable 
proximal carcinogen may be formed and pass through the placenta.  
Examples are cyclophosphamide (Hales, 1982), and 3,3-dimethyl-1-
phenytriazine (Kleihues, 1979b), which are metabolized in the 
maternal liver, and cycasin, which is metabolized by maternal 
intestinal flora (Laqueur & Spatz, 1973); 

    (d)  Indirectly, by the production of proximal and ultimate 
carcinogenic metabolites in the placenta.  This mode of action is 
not well documented though polycyclic hydrocarbons are known to 
undergo metabolic activation in the placenta (Welch et al., 1969; 
Pelkonen et al., 1972; Sehgal & Hutton, 1977), and may belong to 
this class. 

    Transplacental carcinogenesis can also occur as a result of the 
formation of nitroso compounds from their precursors in the stomach 
of pregnant females (Table 11).  The carcinogenic action on the 
fetus occurs via (c) and (a), in the case of ethylurea and nitrite 
administration, and according to types (c) and (b), in the case of 
amidopyrine and nitrite administration. 

4.4.2.  Relationship between teratogenesis and carcinogenesis

    There are a number of compounds that have both carcinogenic and 
teratogenic properties (Neubert, 1980b). This is not surprising 
when considering electrophilic substances such as alkylating 
agents, which react with many cell components.  However, 
teratogenic activity has not been observed for every carcinogen, 
nor would this be expected. Though the possibility of common steps 
in teratogenesis and carcinogenesis exists, there are probably 
specific mechanisms pertinent to only one or other of these 
phenomena.  In studies in which rats were tested with a combination 
of NMU and NEU, it was established that the development of 
malformations of the hypoplastic type (microcephaly) did not 
prevent the appearance of brain tumours, and that the phenomena of 
teratogenesis and carcinogenesis could coexist, but were apparently 
independent (Alexandrov & Napalkov, 1976). 

    In most cases, since malformations and tumours are induced in 
different organs, there is no reason to assume that the two 
processes are related.  In clinical/epidemiological observations 
(Miller, 1977), diethylstilboestrol is the only compound shown to 
have caused malformations (adenosis), and malignant tumours (clear 
cell adenocarcinomas) in the same organs, the cervix and vagina.  
The sensitive period for the induction of these carcinogenic and 
teratogenic effects was the same (Herbst et al., 1979b).  The 
immature opossum is the only animal model developed to date that 
has provided clear-cut indications that malformations can be 
produced together with embryonal tumours in the same organs 
(Jurgelski et al., 1976). 

4.4.3.  General principles of transplacental carcinogenicity tests

    A compound found to be carcinogenic in adult carcinogenicity 
studies does not usually need testing for transplacental 
carcinogenicity.  However, it may be worth investigating the 
relative adult versus fetal sensitivity for chemicals to which many 
pregnant women have been, or are, unavoidably exposed.  As a 
primary task, compounds that are suspected of being human 
carcinogens (IARC, 1982), should be tested for transplacental 

    The general guidelines for conducting transplacental 
carcinogenicity studies do not differ considerably from those 
adopted for adult animal studies (IARC, 1980).  Some substances may 
prove inactive in conventional test systems. This could be because 
the period of development during which the rodent embryo is 
sensitive to carcinogenic stimuli is rather short (during last 10-
12 days of pregnancy) and these compounds may require prolonged 
application to induce tumours.  Therefore, care should be taken in 
extrapolating negative results from animal studies to man, and 
appropriate models with non-human primates or other suitable 
mammalian species should be developed.  Variables that should be 
taken into account include: species and strain, stage of pregnancy, 
dose and route of administration, period of observation, 
registration of biological variables (e.g., body weight) and 
statistical evaluation of the results.  The following experimental 
conditions are peculiar to transplacental studies.  Species

    It is advisable to use at least two laboratory species, 
preferably mice and rats.  The sensitivity of fetal rats to direct 
and indirect carcinogens varies and is manifested with a rather 
peculiar organ specificity that is not typical of adult animals.  
On the other hand, the sensitivity of fetal mice to the 
transplacental action of carcinogens does not differ significantly 
from the sensitivity of adult mice, and neoplasms usually appear at 
sites typical for spontaneous tumours. 

    In some studies, other species, including hamsters, have been 
used.  For economic reasons, non-human primates should be reserved 
for studies on the mechanisms of prenatal carcinogenesis related to 
human beings.  Stages of pregnancy

    The period of optimum susceptibility during pregnancy of a test 
species should be chosen as the duration of treatment with the test 
agent.  The following two experimental designs could be recommended 
for mice, hamsters, and rats: (a) single exposure, 1-5 days before 
parturition; and (b) multiple exposures, daily from day 11 to the 
end of pregnancy (Alexandrov & Napalkov, 1981).  In monkeys 
 (Erythrocebus patas),the period between days 30-50 of pregnancy 
may be selected, since this period has been shown to be quite 
effective in transplacental studies (Rice et al., 1977).  Dose and route of administration

    To select an appropriate route, toxicokinetic peculiarities 
should be given particular consideration.  For example, substances 
that are slowly absorbed from the site of administration should be 
tested by means of a course of treatment.  However, for practical 
purposes, it is advisable to use two different routes, namely 
intravenous and oral. 

    Doses should be selected from pilot tests in which mortality, 
body weight gain, and other pathological alterations in mothers and 
offspring, have been evaluated. Administration of the maximum 
tolerable dose may provide early and pronounced carcinogenic 
effects, while a series of lower doses will help to reveal a "dose-
effect" relationship, which is important for extrapolating and 
evaluating the carcinogenic risk for human beings. 

    The number of pregnant females in each group should be large 
enough to provide for not less than 50 offspring per sex.  The 
offspring of treated and control mothers should be cross-fostered 
to ascertain that the carcinogenic effect of a chemical is 
transplacental, unless postnatal transfer from mother to offspring 
can be excluded.  Evaluation of results

    The progeny should be followed up for the entire life-span of 
the species under consideration, neoplastic changes being analysed 
by histological examination of the affected organs and tissues, at 
the end of the study.  A reliable source for classifying the 
tumours has been provided by IARC (Turusov, 1973, 1976, 1979, 
1982).  Attention should be paid to other toxic manifestations 
including dystrophy, necrosis, atrophy, malformations, dishormonal 
status, etc.  Methods for the statistical evaluation of results 
have been described (IARC, 1980). 

4.4.4.  Methods with potential for the future  Transplacental host-mediated cell culture

    This system offers a means for the rapid screening of 
compounds, including those requiring metabolic activation, for 
possible transplacental carcinogenicity.  In cells subcultured from 
embryonal tissues, following administration of chemicals to animals 
during pregnancy, signs of morphological transformation may be 
detected several weeks after  in vitro growth (Quarles, 1981).  In 
another model, the activity of a chemical may be determined 
according to the marked induction of mutants resistant to 8-
azaguanine in a culture of embryo cells derived from pregnant 
animals treated with the test compound (Endo et al., 1980).  Pre- and postnatal exposure to carcinogens and promoting factors

    A few studies have shown an increased carcinogenic effect when 
an initial prenatal exposure to a carcinogen was followed by 
postnatal exposure to either a carcinogen (Lichachev, 1971) or a 
promoting agent (Goerttler & Loehrke, 1976), or by hormonal 
imbalance (Alexandrov & Anisimov, 1976).  This approach could be 
useful for determining the carcinogenic potential of chemicals with 
marginal effects in adult animals (Napalkov, 1973).  Intraamniotic and intrafetal injection technique

    This procedure distinguishes between the direct effects of a 
test compound on the fetus and the effects induced following 
maternal treatment.  The high susceptibility of rodent fetal cells 
to some of the carcinogens administered directly into the fetus 
suggests that intraamniotic or intrafetal injections might find 
useful applications in detecting potentially hazardous chemicals 
(Alexandrov & Napalkov, 1981; Rossi et al., 1983). 


    Various  in vitro developmental systems (cell, organ, and whole 
embryo culture) have been used to investigate the morphological and 
biochemical basis for normal and abnormal development.  Such 
systems are extremely useful for this sort of basic and applied 
research.  The current demands on whole animal testing systems may 
prevent testing of all the rapidly increasing number of chemical 
agents being introduced into the environment. 

    Short-term tests using various  in vitro and  in ovo systems 
might offer a future means of selecting chemical agents for 
subsequent teratology testing in animals. Short-term  in vitro
methods and their applicability have been reviewed by Rajan (1974), 
Kochhar (1975a), Saxén & Saxén (1975), Ebert & Marois, ed. (1976), 
Wilson (1978a,b), Barrach & Neubert (1980), Neubert (1980a), 
Neubert & Merker, ed. (1981), Kimmel et al. (1982), Neubert (1982), 
Neubert (1983), and Shepard et al. (1983). 

5.1.  Scope of  In Vitro Developmental Systems

    These studies are carried out using a variety of biological 
systems (Table 13), capable of reflecting normal developmental 

    It is evident that, within the last decade, several culture 
techniques, especially those using mammalian tissues, have been 
successfully applied to the elucidation of basic mechanisms 
occurring at the biochemical and cellular levels during embryonic 
development, and to a closer analysis of the modes of teratogenic 
action.  There is no doubt that  in vitro methods have greatly 
broadened the possibilities in this field of research.   In vitro
systems can be considered from three points of view (Neubert, 1981; 
Kimmel et al., 1982; Neubert, 1982): (a) as foods for the 
elucidation of mechanisms relevant to normal and abnormal 
development; (b) as tools for obtaining information on dose 
responses and specific organ toxicity; (c) as foods for the 
selection of chemicals for possible future developmental toxicity 
testing.   In vitrodevelopmental systems are attractive for 
toxicological studies as many are relatively inexpensive, and they 
are easier to perform and less time-consuming than whole-animal 
studies.  However, a number of problems need to be solved before 
using Short-Term Selection Tests (STST) routinely.   In vitro
developmental systems, applicable to studies or problems concerning 
basic developmental biology, must be able to mimic certain 
processes characteristic of prenatal or perinatal development. 

5.2.  Essential and Desirable Features of Short-Term Selection 

    Essential and desirable features of STST, summarized in Table 
12, have been proposed by several investigators (Wilson, 1977; 
Kimmel et al., 1982; Neubert, 1982; Shepard et al., 1983). 

    The injudicious use of  in vitro test data could be even more 
dangerous than extrapolation from routine  in vivo tests. Excessive 
false positives could lead to an undesirable situation in which  in 
 vivo test systems in use, would be overburdened.  Some false 
negatives are bound to occur, but they must be small in number, 
otherwise there is little purpose in the STST. 

    A simpler system is preferable to a whole organism, only if the 
underlying toxic mechanism is largely understood, occurs 
universally, and can be completely mimicked.  In the short-term 
tests used in mutagenicity studies, these prerequisites seem to 
have been fulfilled to some extent but, even here, no single test 
system has been found to be fully adequate.  The situation in 
prenatal toxicity is more complex than in mutagenicity, since 
numerous mechanisms are known to lead to abnormal development.  
Thus, no one test system is likely to identify all the abnormal 
reactions that may occur. 

    The culture technique, if chosen, will largely depend on the 
problem to be solved.  If, for example, cell interactions and their
disturbance by certain agents are to be studied, cell cultures or 
techniques using cell suspensions may be the method of choice.  
Specific  embryonic differentiation processes (palate closure, 
limb development, kidney development) may be best studied with  
organ culture techniques, whereas processes of cell migration over
long distances within the embryo may be best studied in whole 
embryo cultures.  In many instances, the results obtained are only
valid within the limits of the model and for the species used. 

Table 12.  Features of a short-term selection test (STST)
Essential    Results must be predictive of  in vivo effects
             Few false positives
             Very few false negatives
             Inexpensive and readily-available biological unit
             Easily quantificated endpoint
             Reproducibility of results between laboratories

Desirable    Incorporating as many developmental processes as
               possible by using a series or battery of tests
             Ability to test water-soluble and
               water-insoluble chemicals and gaseous agents
             Ability to determine dose-response relationships
             Ability to incorporate biotransforming systems,
               especially hepatic monooxygenase systems, from
               different species including man

    In short-term tests, it is difficult or impossible to establish 
the relationship between embryo/fetal toxicity and maternal 
toxicity, which is essential for determining whether the effects 
are selective or not.  In addition, many compounds are 
metabolically activated to electrophilic substances that react with 
cell constituents and thereby cause toxicity.  This activation may 
occur in the embryo/fetus or the mother, or in both.  The role of 

enzyme induction in the mother and the conceptus is another 
important issue.  The addition of metabolizing enzyme systems from 
different species may make it possible for investigators to obtain 
information on inter-species differences (Fantel et al., 1979; 
Neubert, 1982; Shepard et al., 1983).  This method might be 
especially valuable in extrapolating from experimental animals to 
human beings, even though this approach will not fully reproduce 
metabolism  in vivo. 

    The influence of the mother on the toxic effects in the fetus 
is difficult to reproduce in an  in vitro system.  This is also 
true of chick embryos  in ovo, which are used for the assessment of 
the embryotoxic potential of chemicals.  The maternal factor is 
highly relevant if little is known of the toxicokinetics of a given 
chemical - as is the case with many chemicals in the environment. 

5.3.  Validation of Short-Term Selection Tests

    To be accepted as part of a routine testing process, STSTs will 
have to be validated for their ability to predict the teratogenic 
potential of an agent.  For the purpose of validation, necessary 
criteria will have to be developed, the number and type of 
chemicals (representative of diverse structures and biological 
activities) required for testing will have to be decided, and an 
acceptable percentage of false positives and false negatives will 
need to be agreed upon.  To carry out validation in a controlled 
fashion and to facilitate interlaboratory comparisons, a list of 
chemicals to be used should be developed jointly by scientists 
experienced in teratology, toxicokinetics, and biochemistry.  An 
attempt to choose agents that are teratogenic or non-teratogenic 
has been undertaken by Smith et al. (in press). 

    Final validation of tests to be used must be based on the 
results of the  in vitro tests of known human teratogens.  Up to 
the present, no STSTs have been fully validated and no 
recommendations are made in this document as to which batteries of 
STSTs are likely to emerge for possible use in the future.  The 
STSTs are not in any way intended to replace animal tests, but 
merely to supplement them so that they can be made more efficient 
and productive. 

5.4.  Available Developmental Systems

    Some developmental systems that are in use at present for basic 
mechanism studies, or proposed as STSTs (Table 13) are discussed in 
section 5.4.1.  None of these developmental systems involves the 
chorioallantoic placenta or an intact maternal metabolic system.  
In some cases, a mono-oxygenase or other type of bioactivating 
system may be added.  A summary of many of these developmental 
systems has been published by Neubert (1982). 

5.4.1.  Whole-embryo culture (warm-blooded animals)

    There are three types of cultures that have been used 
extensively in various laboratories: 

    (a)  culture of mammalian preimplantation stages;

    (b)  culture of mammalian postimplantation embryos; and

    (c)  culture of non-mammalian (avian) embryos.

    Though not a culture method, the evaluation of the chick embryo 
 in ovo will also be discussed.  Ovum maturation and preimplantation stages

     In vitromaturation of the ovum of different mammalian species 
(especially rodents) has been intensively studied (Edwards, 1980; 
Edwards & Purdy, 1982).  With this system it is possible to 
investigate the mechanism of action of various chemicals on 
chromosomal disjunction, ovum maturation, and activation.  It is 
possible to use  in vitro fertilization of human ova not only as a 
method of treating human infertility, but also to obtain a better 
understanding of the basic mechanism of human oogenesis (Edwards, 
1980; Kurilo et al., 1982, 1983). 

    Culture of cleaving mammalian ova is being used to study the 
mechanism of embryotoxic action of some chemicals, especially those 
that are expected to influence proliferation and induce chromosomal 
aberrations (Dyban et al., 1976, 1977a,b; Spielmann et al., 1981; 
Matsumoto & Spindle, 1982). 

    Hsu (1979, 1980, 1981) published results of a system in which 
it is possible to study mammalian development  in vitro from the 
preimplantation phase to the late embryonic stages. However, the 
success rate of this model has not yet exceeded 10%.  In general, 
preimplantation embryos cannot be cultured beyond stages 
corresponding to implantation or early postimplantation. 

Table 13.  Developmental systems
                          Period of              Developmental
   System                 development            process studied
                          covered                and endpoint
I.  Whole-embryo culture

   Mammalian              preimplantation        fertilization,
   embryos                                       cleavage, and
   (including human)                             blastocyst formation

   Rodent embryos         post-implantation      organogenesis
                          organogenesis          over 2 (or 3) days

Table 13 (contd).
                          Period of              Developmental
   System                 development            process studied
                          covered                and endpoint
   Chick embryo culture   organogenesis          organogenesis

   Chick embryo  in ovo   entire                 whole development
   Amphibian embryos      entire                 whole development

   Zebra fish             entire                 whole development

II.  Organ cultures

   Limb buds (avian,      late organogenesis     morphogenesis
   rodents)                                      cartilage formation
                                                 muscle formation

   Palatal shelves        organogenesis          epithelial fusion and
                                                 cell death, palate

   Tooth bud              organogenesis          dental development,
                                                 histogenesis of tooth

   Lens                   organogenesis          lens differentiation,
                                                 histogenesis, and
                                                 protein production

   Pancreas               organogenesis          acinar development,

   Sex organs             late organogenesis     gonadal and organ
                                                 development, germ cell
                                                 maturation, accessory
                                                 sex gland,

   Kidney                 late organogenesis     nephrogenesis,

   Thyroid                late organogenesis     thyroid,
                                                 histogenesis, and
                                                 functional development

Table 13 (contd).
                          Period of              Developmental
   System                 development            process studied
                          covered                and endpoint
III.  Tissue culture

   Chick                  organogenesis          migration and pigmen-
     neural crest                                tation, melanin for-
                                                 mation or acethyl-
                                                 choline transferase

   Chick limb bud         organogenesis          growth and
     mesenchyme                                  chondrogenesis

   Lung                   organogenesis          pattern formation,
     micromass culture                           cell recognition, and

   Embryonic heart        organogenesis          histogenesis and

   Skeletal muscle        organogenesis          myogenesis, cell
                                                 and histogenesis

   Teratocarcinoma        organogenesis          histogenesis

   Mouse tumour cells     none                   cell attachment

   Human embryonic        organogenesis          growth of patatal
     cell culture                                mesenchymal cells

IV.  Invertebrates

      Hydra attenuae       adult and embryonic    aggregation and
                                                 movement regeneration
                                                 of damaged and
                                                 disassociated adults

      Drosophila           entire                 myogenesis, nerve
      melanogaster                                  attachment

      Planaria             regeneration of adult  neurogenesis, eye

     Sea urchins          cleavage               cell replications
                                                 and contacts

      Dictyostelium       entire                 aggregation,

Table 13 (contd).
                          Period of              Developmental
   System                 development            process studied
                          covered                and endpoint
V.  Other approaches

   Amphibian limb         differentiation        myogenesis,
     adult regeneration                          neurogenesis,
                                                 and chondrogenesis

   Virus                  unknown                (viral replication)

   Lipophilic or          unknown                (cell membrane
     electrophilic                               penetration,
     chemical property                           toxic potential)

    It is now possible to culture mammalian embryos from the one- 
or two-cell stage to the blastocyst stage.  Most of the work has 
been performed with the preimplantation stages of mice (McLaren & 
Biggers, 1958; Brinster, 1963; Runner, 1965; Biggers et al., 1971; 
Whitten, 1971; Whittingham, 1971; Eibs & Spielmann, 1977; Speilmann 
& Eibs, 1977), but cultivation of preimplanted rat, rabbit, and 
human embryos has also been carried out successfully (Brachet, 
1913; Chang, 1949; Daniel, 1965; Daniel, Jr, 1968; Mauer et al., 
1968;  Onuma et al., 1968; Edwards et al., 1969; Brinster, 1970; 
Steptoe & Edwards, 1978; Edwards & Purdy, 1982).  A chemically-
defined, serum-free, culture medium is generally used for this 

    Biophysical and biochemical techniques have been used in 
addition to morphological and biochemical methods to analyse the 
normal and abnormal development of preimplantation stages.  The 
measuring of membrane potentials has been used as an indication of 
functional stage in the eggs of sea urchins and mammals (Crose et 
al., 1973; Leonov et al., 1972, 1975; Persianinov et al., 1973). 

    Most investigators agree that the preimplantation period is not 
sensitive to the induction of gross structural abnormalities 
(Austin, 1973).  Preimplantation embryos, therefore, have not been 
used much in teratology.  However, abnormal development associated 
with early or delayed mortality can be induced before implantation 
(cf. Snow, 1973, 1975; Goldstein et al., 1975; Spielmann et al., 
1981; Pedersen, 1981). 

    Many drugs may reach the preimplantation embryo via the tubal 
or uterine fluids, and agents, such as nicotine, may even 
accumulate within the blastocyst (Fabro, 1973; Sieber & Fabro, 
1974).  Lutwak-Mann (1973), has suggested the use of "flat mounts" 
of rabbit blastocysts to detect abnormalities induced by 
embryotoxic agents, but this technique has not gained acceptance.  Postimplantation mammalian embryos

    Developed by New and co-workers (New, 1966a,b, 1976, 1978; New 
et al., 1976a,b), the culture of whole rodent embryos  in vitro is 
now an established procedure in many laboratories. Mouse or rat 
embryos are cultured with their embryonic membranes intact by the 
rotating bottle technique (New et al., 1973; Cockroft & Coppola, 
1977), which is generally preferred to more sophisticated 
experimental procedures.  Rat embryos (8.5 - 10.5-day-old), can be 
cultured over a 24-h or even a 48-h period and embryonic stages 
corresponding to about 35-somites may be reached.  Up to the 30-
somite stage, the development proceeds at roughly the same rate  in 
 vitro as  in vivo.  Abnormal development may occur in culture, if 
conditions are inadequate. 

    Some studies have been conducted on mice (Clarkson et al., 
1969; Sadler, 1979a,b; Davis et al., 1981) and on hamster embryos 
(Givelber & Dipaolo, 1968).  Rat embryos have been traditionally 
cultured in a medium containing rat serum, but some investigators 
have successfully grown them in human serum (Shepard et al., 1969; 
Chatot et al, 1980; Klein et al., 1980, 1981; Herken & Anschütz, 
1981), dialysed serum (Clarkson et al., 1969; Gunberg, 1976), and 
in a chemically-defined serum-free medium (Klee-Trieschmann & 
Neubert, 1981).  In several studies, the direct embryotoxic 
potential of teratogens for the developing embryo was elucidated 
following their addition to the culture medium (for bibliography 
see Neubert, 1982).  Popov et al. (1981) and Fantel (1982) used a 
liver microsomal system added to the culture to activate 
cyclophosphamide.  Dyban as well as Klein and his co-workers 
studied the effects on embryos in a culture serum containing a 
variety of compounds (e.g., cyclophosphamide), which had been 
obtained from animals treated  in vivo or from patients (Dyban, 
1977; Dyban et al., 1979; Klein et al., 1980, 1981).  Only gross 
morphological evaluations were performed in most of the studies. A 
morphological scoring system (Brown et al., 1980) may be useful for 
evaluation, but, for more detailed analysis, histological 
examination is essential (Herken & Anschütz, 1981).  A technique of 
culturing whole embryos in which it is possible to the record the 
embryonic heart beat has been developed (Robkin et al., 1972).  An 
attempt has also been made to use a sequence of whole-embryo 
cultures with subsequent organ culturing of, for example, limb buds 
(Kochhar, 1976).  The applicability of the whole-embryo technique 
as a routine method is limited, because it is expensive, too 
sensitive to disturbances, often of unknown origin, and quite prone 
to abnormal development, even under normal conditions. Furthermore, 
a maximum culture period of 48 h is often too short compared with 
the entire period of organogenesis.  The method is especially 
useful for studying certain aspects of developmental biology, such 
as the nutrient requirements of embryos during organogenesis, or 
for performing biochemical analyses or studies with radioactively-
labelled precursors (Shepard et al., 1970).  Chick embryo in culture

    Explantation of the early chick embryo is technically more 
difficult than culturing embryonic mammals at the preimplantation 
stages.  The problems involved, however, have mostly been solved in 
the technique of New (1955; 1966a,b). It consists of supporting the 
explanted blastoderm on a piece of vitelline membrane stretched 
across a glass ring.  The ring is placed on a watch glass serving 
as a reservoir for nutrient medium - thin albumen, obtained perhaps 
from the same donor egg.  The watch glass is kept in a covered 
Petri dish, which forms a moist chamber.  Primitive streak 
blastoderms explanted by this technique develop at a normal rate 
for approximately 48 h, until after the 20-somite stage.  Agents 
can be tested by direct addition to the medium, and sera from 
individuals exposed to the agent can be added to the culture.  
Embryos examined for toxic effects are examined for changes in size 
and gross morphology, histology, and biochemical variables such as 
DNA and protein content, glucose consumption and lactate 
production, cellular damage, etc. (New, 1976).  This method is 
inexpensive and can be used routinely with standard laboratory 
equipment.  Large numbers of embryos can be processed in a short 

    Since the embryos can only be maintained for a relatively short 
period of time, the usefulness of this technique is limited to the 
early developmental stages.  Effects caused by substances acting 
mainly or entirely at later developmental stages cannot be detected 
by this procedure.  Chick embryo  in ova

    Chemicals have frequently been tested in chick embryos, at all 
stages of development, by administration into fertile eggs.  A 
chemical, in solution or suspension, is usually deposited on the 
air-sac membrane or injected into the yolk and the chicks examined 
before or after hatching.  This method has often failed to supply 
reliable results as it has been difficult to standardize the blind 
injection technique. Moreover, it is doubtful whether the form and 
concentration of a chemical, when it reaches the embryo, is known.  
A standard window technique for delivering the chemical to the 
chick embryo has been developed for embryotoxicity and 
teratogenicity studies in which better control of experimental 
conditions is possible (Jelinek et al., 1976; Jelinek, 1979; 
Jelinek & Peterka, 1981).  The test substance is administered in a 
single application to the subgerminal portion of the yolk of the 
embryo on day 2, and intra-amniotically on days 3 or 4.  The number 
of dead, malformed, and growth-retarded embryos is counted on the 
eighth day of incubation, when the study is terminated.  The method 
is inexpensive and requires a moderate degree of skill.  The effect 
of a large number of variables can be assessed and the amount of 
the test chemical needed is small.  With this test it is possible 
to determine dose-response and stage-response relationships and the 
affinity of a substance for particular organ systems.  The capacity 
of the system for the metabolic activation of xenobiotics is ill-
defined.  Testing of compounds that are insoluble or soluble in 

toxic solvents is difficult and false-positives are common with 
surface-active compounds. Though many compounds have so far been 
tested in chick ova, the predictive value of these tests has not 
yet been established. 

5.4.2.  Organ cultures

    The term organ culture should be used when a whole-organ 
anlagen, or its representative part (e.g., a lobe of a lung or 
liver), is kept in culture and growth, particularly organ-specific 
differentiation, is observed.  Two approaches have been used: 

(a)  the explant, is kept on a filter or other support, and
     differentiates at the interface of the culture medium and
     the gas phase (Trowell, 1961); and

(b)  the explant is submerged in the culture medium, and
     shaking or rotation is used to guarantee gas exchange and

    Both of these techniques show a number of advantages and 
certain disadvantages depending on the embryonic organ or organ 
anlage used.  The successful use of organ culture systems in 
developmental biology and toxicology has recently been reviewed 
(Neubert, 1982).  Serum-containing medium has been employed in the 
majority of cases, but successful attempts have been made to 
culture certain tissues in chemically-defined media. 

    The technique described by Trowell (1961) has been used in the 
culture of a variety of tissues and organs from mammalian or avian 
embryos or fetuses.  It is comparatively easy to perform and 100-
200 explants can be prepared by one person in one day.  A 
modification of the method, the trans-filter technique, is used for 
induction studies (Grobstein, 1956; Saxén et al., 1968; Wartiovaara 
et al., 1974; Saxén & Lehtonen, 1978).  Organ culture of limb buds

    The Trowell technique is used for the culture of limb buds from 
rat or mouse embryos (Shepard & Bass, 1970; Kochhar, 1970; 
Aydelotte & Kochhar, 1972) but it may also be used for limb buds 
from embryos of other species, e.g., rabbit, chick, or ferret 
(Lessmöllmann et al., 1975; Beck & Gulamhusein, 1980a).  Growth of 
limb buds in culture, is retarded compared with  in vivo 
development, but morphological differentiation of the cartilaginous 
bone anlagen and of muscle can be observed. With any of these 
techniques, the extent of the morphogenetic differentiation of the 
cartilaginous bone anlagen depends largely on the stage of 
development at which the cultures are initiated.  In earlier 
studies, serum-containing medium was used, but more recently a 
chemically-defined medium has been preferred (Lessmöllmann et al., 
1975, 1976). 

    The submerged culture system has also been successfully used 
for the study of limb bud differentiation  in vitro (Neubert & 
Barrach, 1977b; Blankenburg et al., 1981), and it shows many 

advantages over the Trowell technique.  Dissected limb buds can be 
stored at low temperatures for several weeks (Neubert & Bluth, 
1981a) and some studies may be facilitated by the use of such an 
organ bank.  This organ culture technique is one of the best and 
most extensively studied. Normal development and differentiation, 
predominantly of the cartilaginous bone anlagen, have been studied 
and a variety of morphological (e.g., Merker, 1975; Merker et al., 
1978b; Neubert et al., 1974a, 1976) and biochemical (e.g., 
Rautenberg & Neubert, 1975; Neubert & Rautenberg, 1976; Barrach et 
al., 1975, 1980, 1981; Blankenburg et al., 1980, 1981, Dusemund & 
Barrach, 1981) variables have been evaluated.  Although the amount 
of protein increases within the explants, there is little or no 
increase in the total DNA (Neubert et al., 1976). 

    It has been possible to induce abnormal development  in vitro by 
adding a variety of teratogens to the culture medium (Kochhar, 
1970, 1975b; Aydelotte & Kochhar, 1972; Neubert et al., 1974b, 
1976, 1977a; Barrach et al., 1978; Merker et al., 1978a; Welsch et 
al., 1978; Merker & Günther, 1979; Beck & Gulamhusein, 1980b; 
Bochert et al., 1981; Neubert & Bluth, 1981b; Stahlmann et al., 
1981, 1983). 

    It has also been possible to quantify the extent of 
differentiation achieved in culture and to establish dose-response 
relationships for abnormal development using the organ culture 
system.  Biochemical variables, such as levels of DNA, RNA, 
protein, collagen, etc. can be measured (Neubert et al., 1974a); or 
a score system may be used for morphogenetic differentiation 
(Neubert et al., 1977a; 1978b). Image analysis (Kwasigroch et al., 
1981) can also be used to quantify abnormal development.  Organ culture of the pancreas

    Organ culture has been used by Rutter and his colleagues 
(Pictet & Rutter, 1972; Rutter et al., 1975) to study the 
development of the embryonic rat pancreas.  The information gained 
using this system has provided the most extensive biochemical 
background for any prenatally developing mammalian tissue 
investigated so far.  However, the pancreas is not considered a 
major target for embryo- or fetotoxic agents.  Organ culture of palatal shelves

    Growth and differentiation of the palatal shelves has also been 
achieved in organ culture.  This system has been extensively used 
to study cleft palate-inducing agents (Lahti & Saxén, 1967, 1972; 
Saxén, 1973; Fairbanks & Kollar, 1974; Pratt et al., 1980; 
Brinkley, 1980; Pratt, 1983).  Organ culture of tooth anlagen

    Normal and abnormal development of tooth anlagen has been 
studied, both morphologically and biochemically, using the organ 
culture technique.  Important information on connective tissue 
development and on the induction of abnormalities has been obtained 

(Ruch & Karcher-Djuricic, 1971; Kollar, 1973; Galbraith & Kollar, 
1974; Koch, 1974; Hetem et al., 1975; Kerley & Kollar, 1978; 
Thesleff & Pratt, 1980).  Organ culture of the embryonic lens

    Lens development can be followed using human embryonic 
material.  The induction of abnormal development by rubella virus 
can be mimicked  in vitro.  Cataract formation was observed by 
Karkinen-Jääskeläinen et al. (1975).  Organ culture of the embryonic kidney

    Organ culture techniques have been developed for the study of 
explanted mouse or human embryonic kidney (Saxén, 1965; Saxén et 
al., 1968; Crocker & Vernier, 1970; Crocker, 1973; Saxén & 
Lehtonen, 1978; Saxén & Ekblom, 1981).  This system may prove 
useful for investigating the normal mechanisms of renal development 
and also testing suspected nephrotoxic agents.  The expense and 
time involved are both fairly high, since the endpoints require 
histological analysis.  Organ culture of embryonic gonads and accessory duct systems

    Organ culture of embryonic gonads together with assessory ducts 
has contributed to the understanding of how hormones and some 
hormone inhibitors modify sexual development (Martinovitch, 1938; 
Jost & Bergerard, 1949; Price & Ortiz, 1965; Josso, 1974; Lasnitzki 
& Mizuno, 1981).  Both mammalian and avian embryos have been used.  
These techniques could be used to investigate new chemicals that 
might have specific effects on the genital system.  The method is 
moderately expensive and time-consuming.  Organ culture of thyroid tissue

    Tissue from the thyroid glands or their embryonic primordia 
from chick, rodents, and human beings may be explanted into organ 
cultures, which may then be exposed to hormones and various 
xenobiotics.  The effects can be studied biochemically or 
histochemically (Shepard, 1965, 1967, 1974). Appearance of colloid 
within thyroid follicles, histoautoradiography of incorporated 
radioiodine, or biochemical analysis of various labelled organic 
compounds are used for evaluation.  The expense and time required 
are moderate.  This test system would be of use in investigating 
very specific questions about the effects of new chemicals on gland 
maturation and function.  Other organ culture systems

    The use of organ culture is not restricted to the examples 
provided in the preceeding paragraphs and other organ anlagen have 
been cultured.  Examples include the use of bone from rat or mouse 
fetuses (to demonstrate the pathological developments induced by 
substances such as glucocorticoids or tetracycline) (Fell & Weiss, 
1965; Saxén, 1966; Saxén & Kaitila, 1972), digits (Rajan, 1969, 

1974; Rajan et al., 1980), lung anlagen (Riso & Zimmermann, 1981), 
intestine (Silano et al., 1981), and liver (Räihä & Schwartz, 

5.4.3.  Culture of non-mammalian or non-avian embryos

    Embryos of amphibia, lurchae, or fish have been extensively 
used in research (New, 1966b), but only a few of these systems have 
been suggested as models for the evaluation of embryotoxic risks.  
Some of these animal models are listed below, though these species 
are much less favourable for the routine testing of potential 
embryotoxic agents, as their response to such agents has not been 
properly investigated, so far.  Studies on the development of fish

    Several species of fish have been studied embryologically and 
at least two species, Oryzias and zebra fish, have been used for 
studies on abnormal development (Smithberg, 1962; Schreiweis & 
Murray, 1976; Leung & Bulkley, 1979). Streisinger (1975) proposed 
using zebra fish for the study of teratogens.  By fertilizing eggs 
with genetically-inactive sperm and blocking their first mitotic 
division, homozygous diploid embryos can be obtained, some of which 
develop into adult forms.  Clones of the fish can be established, 
which may be useful in dose-response analysis.  The raising and 
study of the wild form of zebra fish might prove to be inexpensive 
and not too time-consuming.  So far, not much information on 
teratogenicity has been obtained from these studies, but they may 
be valuable for the evaluation of the effects of environmental 
chemicals on wild life.  Studies on the development of amphibia, lurchae,
sea-urchins, and other invertebrates

    None of these systems has been adequately analysed for 
responsiveness to toxic substances.  The species observed in 
developmental studies include: amphibia (Schultz et al., 1982), 
 Drosophila(Demerec, 1950), sea urchins (Lallier, 1965) and worms 
(Deppe et al., 1978).  Sea urchin eggs have been the subject of 
reproduction studies, because synchronically developing embryos can 
be obtained in large numbers (Harvey, 1956; Costello et al., 1957; 
Mateyko, 1967) and experimental conditions (fertilization and 
incubation) are simple (Tyler & Tyler, 1966; Hinegarduer, 1967; 
Reverby, 1971; Hörstadium, 1973).  It is possible to study 
embryotoxic and teratogenic effects (Hagström & Lönning, 1973; 
Leonov, 1979) and many chemical substances can easily penetrate the 
external coats of eggs and embryos. 

     Hydra attenuaehas been recommended as a model for
developmental toxicity testing but, in this case, it is not the 
development that is studied.   Hydra attenuata are grown in trays 
and are fed  Artemia nauplii(Johnson, 1980; Johnson et al., 1982).  
Two endpoints can be measured after addition of a chemical agent to 
the liquid medium.  One endpoint, representing the adult toxic 
dose, is an irreversible morphological alteration called the 

"tulip" stage.  With the second, artificial "embryos" are formed by 
compacting dissociated adult cells (Gierer et al., 1972) and 
dissolution of the aggregates is the endpoint of the test.  The 
expense is small and little time is required.  If the system can be 
validated, it might prove useful in evaluating large numbers of 

5.4.4.  Tissue culture systems

    The use of tissue culture systems in developmental toxicology 
is limited, but primary cultures from various embryonic tissues 
have been extensively used in developmental studies including the 
chick neural crest (Greenberg, 1982), heart mesenchymal cells, and 
skeletal muscle cells (Holtzer et al., 1958).  Cell lines can be 
used in developmental toxicology.  They may be employed in testing 
for the general cytotoxicity of chemical.  However, it must be 
remembered that some cells may change their characteristics in 
culture.  Pratt et al. (1980) have used an established cell line 
from the human embryonic palatal mesenchyme (HEPM) to aid in 
understanding the biochemical basis for the way in which 
glucocorticoids inhibit palatal mesenchymal cell growth and cause 
cleft palate in the rodent.  Furthermore, Pratt et al. (1982) have 
suggested that teratogen-induced growth inhibition in these HEPM 
cells might be used as part of a battery of assays to screen for 
potential environmental teratogens. 

    It seems possible that, in some instances, primary cell 
cultures of embryonic or fetal cells - including those from human 
embryos or fetuses - may be used for studies on the drug-
metabolizing capacity of cells during prenatal, and especially 
perinatal, development (Nau et al., 1977; Egger et al., 1978; 
Liddiard et al., 1978; Merker et al., 1978a; Kremers et al., 1981).  
Comparative studies of this kind can be helpful when potential 
differences between the capacity of embryonic or fetal cells and 
adult cells of various species for metabolizing xenobiotics are 

5.4.5.  "Micromass" culture of dispersed cells from 
the embryonic limb and lung

    On the basis of work performed with cells of chick embryos 
(Caplan, 1970; Schacter, 1970; Levitt et al., 1975), a method was 
developed for the study of cartilage formation from limb blastema 
cells grown at high cell density (Umansky, 1966; Ahrens et al., 
1977; Pennypacker et al., 1978b; Solursh et al., 1978).  This 
technique is referred to as micromass, spot, or high density cell 
culture.  Dispersed cells from a number of organ anlagen can be 
used for this purpose. 

    Blastema cells from the limbs of 11-day-old mouse embryos are 
disintegrated with trypsin and the dispersed cells cultured in a 
drop of culture medium in a Petri dish.  At the time of initiation 
of the culture, there is no or little indication of the formation 
of cartilaginous structures in the anlage.  After two days in 
culture, nodules of cartilage form  in vitro.  This system has been 

used extensively in several laboratories, and the effect of a 
variety of chemical agents on cartilage formation has been 
investigated (Hassel et al., 1978; Pennypacker et al., 1978a,b; 
Lewis et al., 1978; Flint, 1980; Merker et al., 1980a,b).  The 
culture system has a major disadvantage; it is not possible to 
study typical morphogenetic differentiations, only differentiation 
at the cellular level. However, the advantage of this system is 
that the technique is easy and results are obtained after a 
comparatively short period. 

    Merker and his group established a system in which it is 
possible to observe the morphogenetic differentiation processes in 
lung cells (Merker et al., 1981).  The tissue of a lung anlage from 
a 14-day-old mouse embryo is disintegrated with trypsin and 
collagenase and the cells cultured at high density in a drop of 
culture medium in a Petri dish.  Budding with structures resembling 
embryonic lung tissues, can be obtained in culture.  Abnormal 
development has also been studied using this system (Zimmermann et 
al., 1981). 

5.4.6.  Studies with non-embryonic tissues

    It may be asked whether it is feasible to use non-embryonic 
tissues for evaluating the possible embryotoxic effects of 
chemicals.  Basic processes such as cell-cell interactions, cell 
migrations, differentiation processes, and proliferation processes, 
can be used as end points, as these are certainly involved in 
embryonic development.  Systems suggested so far include the 
investigation of aggregation phenomena with cells of mammalia 
(Moscona & Moscona, 1952) or lower invertebrates such as hydra 
(Johnson, 1981), the investigation of cell-cell interactions as 
studied by the attachment of cells to lectin-coated surfaces (Braun 
et al., 1979, 1982), and the possibility of studying induction 
processes using hormones or drugs.  The use of teratocarcinoma 
cells, which are pluripotent cells capable of many types of 
embryonic differentiation, might also be considered (Stevens, 1967; 
Graham, 1977; Martin, 1978). 

    The regeneration of the tail of the  Xenopuslarvae or of the 
crested newt has been suggested as a model for the evaluation of 
the teratogenic potential of drugs.  The biology of this 
regeneration has been well investigated (e.g., Lüscher, 1946, 1955; 
Gebhardt & Faber, 1966; Thornton, 1968; Neukomm, 1969; Tsonis & 
Eguchi, 1980) and possible means of interfering with this 
regeneration have been explored.  It has been found possible to 
induce abnormal regeneration in this system with thalidomide and 
some of its teratogenic and non-teratogenic derivatives (Bazzoli et 
al., 1977).  There are no systematic studies on the use of such 
systems in routine testing for embryotoxicity. 


6.1.  Introduction

    Human risk assessment is intended to provide an estimate of the 
potential for human disease, and of exposure levels for protecting 
human health.  The dose-effect data from human epidemiological 
and/or environmental animal studies are essential for human risk 
evaluation.  When coupled with estimates of actual exposure levels, 
they can be used to estimate the probability (magnitude of the 
likelihood) of an effect in the exposed population.  Such a hazard 
assessmenta requires a quantitative evaluation of exposure for each 
sub-population at risk.  Since the assessment of hazard for sub-
populations necessarily relies on the level and type of exposure 
specific to each circumstance, hazard assessment is not discussed 
in this document. 

    Direct observations in human beings obviously provide the best 
information for assessing human risk.  These data, however, are 
rarely available and may often be difficult to interpret, 
particularly with respect to the timing and level of exposure.  
Given these factors, scientists most often rely on observations 
from laboratory animal studies to estimate potential human risks 
associated with environmental exposures. 

    Most experience with risk assessment has been in the area of 
carcinogenesis, although methods are being developed for assessing 
risk for other health endpoints (EPA, 1976; Albert et al., 1977; 
IRLG, 1979; EPA, 1980; EPA, 1982; NAS, 1983). The assessment of 
human risk from prenatally-induced toxicity presents particular 
problems, which are described in this chapter.  In general, the 
risk assessment comprises two steps: (a) a qualitative assessment 
of the biomedical data to determine the likelihood that the 
chemical may pose human risks of prenatally-induced toxicity, and 
(b) on the assumption that the chemical is capable of causing such 
adverse effects, a quantitative estimate or prediction of the dose-
effect relationships in human beings, to provide the basis for 
recommending safe exposure levels and for further assessing the 
potential hazard to sub-populations exposed to the chemical at 
various levels. 

    The first step (i.e., qualitative assessment) relies on 
observations in human beings and  laboratory  animals  together 
with other pertinent supporting data.  Taking a weight-of-evidence 
approach, observations in human beings, backed up by animal 
studies, provide the most convincing evidence that a chemical is 
capable of causing adverse effects in developing progeny.  In the 
absence of human data, however, observations in laboratory animals 
must be relied on.  The best animal evidence is provided by 


a  In this context, the term "hazard assessment" is used to 
   indicate the likelihood that a chemical will cause an adverse 
   health effect (injury) under the conditions in which it is 
   produced or used (WHO, 1978). 

replicated results in laboratory animals and results in multiple 
species. Experimental animal data of a more limited nature may 
contribute some evidence of potential adverse human effects 
provided the design and conduct of the study is adequate; although 
the data for each chemical must be evaluated independently.  In an 
assessment of these data, the nature, severity, and extent of the 
biological responses should be considered together with other data 
that may have a bearing on the overall evaluation.  These include 
pharmacokinetic parameters, cytotoxicity, other toxic endpoints, 
structure-activity relationships, maternal toxicity, and 
information from short-term non-mammalian and  in vitro tests. 

    The second step, that of quantitatively estimating the dose-
effect relationship for human beings, generally requires 
extrapolation from higher doses, where effects have been observed 
in human beings or in laboratory animals, to lower doses to which 
human beings may be exposed in the environment.  A dose-response 
relationship, observed in human beings is more readily 
interpretable for low-dose extrapolation than animal data.  Very 
often, however, observations in human beings are not accompanied by 
very precise estimates of the magnitude and duration of exposure. 
In the absence of such data, experimental animal data may be used 
to extrapolate dose-effect relations for human beings. For this, it 
is usually necessary to fit some shape to the dose-effect curve 
outside the observed range, with a view to extrapolating adverse 
effects observed at higher doses in experimental studies to 
possible effects that might occur at lower human exposure.  The 
shape of the dose-effect curve should be justified, if possible, on 
the basis of mechanisms of action.  It is also necessary to 
compensate for the biological, especially pharmacokinetic, 
differences between animals and human beings.  Although a variety 
of mathematical models have been suggested for extrapolation 
involving other toxic manifestations, e.g., carcinogenicity (Brown, 
1980; Anderson & CAG, in press), their applicability for predicting 
adverse health effects in progeny following exposure during 
pregnancy has not been justified.  In the absence of a biological 
justification for applying mathematical extrapolation models to 
these biological data, safety factors are being used to recommend 
safe levels of exposure.  Various factors involved in data 
interpretation for the purpose of human risk assessments are 
discussed in sections 6.2 and 6.3. 

6.2.  Interpreting Laboratory Animal Studies

    For qualitative risk assessment purposes, positive results in 
laboratory animals may provide substantial evidence that the agent 
may be a potential human teratogen.  However, animal studies must 
be interpreted in the light of the adequacy of the study and the 
nature and extent of the responses, together with all other 
biological information that might have a bearing on the weight-of-
evidence to establish potential human teratogenicity. 

    The primary basis for using animal bioassay systems to predict 
prenatally-induced toxicity in human beings is that almost all of 
the 15-20 known or suspected human chemical teratogens give 

positive results in animal bioassay systems (Schardein, 1983; 
Shepard, 1983).  However, many agents shown to induce developmental 
toxicity in laboratory animals have not been clearly associated 
with human effects.  Factors such as interspecies differences, and 
the use of very high doses, often causing toxic effects in the dam, 
in experimental animal studies compared with the low levels to 
which human beings are exposed, may account for this.  It is 
possible that most agents that produce teratogenic effects in 
laboratory animals would also be shown to affect human beings, if 
the methods for detection were sufficiently sensitive and the human 
exposures were high enough. 

    Several approaches have been proposed for classifying the 
qualitative evidence for developmental toxicity.  For example, 
Neubert et al. (1980) have proposed that agents that cause 
developmental toxicity should be classified as inducing non-
specific or specific effects.  Non-specific effects can be induced 
by almost any agent, if the doses are high enough. These effects 
include weight reduction, developmental retardation, and embryo-
mortality.  Such effects may often be caused by maternal toxicity 
and may, therefore, warrant different quantitative approaches.  If 
maternal toxicity is the cause of the fetal effect, then the 
extrapolation of experimental data is carried out according to the 
methods used in adult toxicity.  Also lower safety factors are 
often used for such agents.  On the other hand, agents that cause 
specific effects do so by interfering with some stage of fetal 
development, independent of maternal toxicity.  For specific 
effects, any extrapolation must be related to the fetal effect and 
be based on the mechanisms of action, if possible, and safety 
factors higher than those for non-specific effects should be used.  
Similar use of different safety factors for different classes of 
effects has been proposed by Khera (in press).  One of the 
difficulties associated with using any categorization scheme is 
that some chemicals may not fall clearly into any one category. 

    The adequacy of animal data depends in part on experimental 
conditions.  The route, temporal pattern, and duration of exposure 
for experimental studies should simulate those of man for use in 
performing risk assessments. Sometimes it might be advisable to 
test impure chemicals or chemical formulations to which man is 

    Despite widespread interest in short-term non-mammalian and  in 
 vitrotechniques, pregnant mammals are today the only accepted test 
systems for predicting human disease. Currently, some short-term 
test systems offer promise for elucidating the mechanisms of action 
of embryotoxic agents and some may also be useful for providing 
supporting information for interpreting animal bioassay studies. 

6.2.1.  Endpoints

    A wide variety of responses characterize developmental 
toxicity.  Spontaneous abortion, intrauterine death, prematurity, 
low birth weight, birth defects, and postnatal manifestation 
following prenatal insults are effects of abnormal human and animal 

reproduction.  Even offspring that seem perfectly normal at birth 
may be afflicted with some degree of mental retardation or with 
transplacentally-induced infertility (Khera & Ruddick, 1973) and 
cancer later in life. At present, most of the evidence regarding 
these types of delayed toxicity is derived from studies on animals, 
though a few examples have come to light in human beings.  These 
are that prenatal exposure to diethyl stilboestrol impaired 
fertility in male and female offspring and induced vaginal cancer 
in female offspring (Herbst et al., 1979b) and prenatal exposure to 
methyl mercury induced cerebral palsy in offspring (Reuhl & Chang, 
1979).  Thus, risk assessment is complicated by the various 
outcomes of altered development that need to be considered.  
Mechanisms of action may differ for these different endpoints; 
therefore characterization of dose-effect curves, particularly at 
low levels of exposure, to quantitatively assess human risk for 
different disease endpoints is complex. 

    The various endpoints of developmental toxicity can occur 
spontaneously or be induced.  Altered growth consists of reduced 
fetal weight, which is widely recognized, and increased fetal 
weight which is less recognized.  Structural variations include 
changes of a permanent nature i.e., extra rib, absent vertebral 
segment, retarded ossification, precocious ossification, and open 
ductus arteriosus.  The role of minor structural deviations 
relative to more obvious forms of developmental toxicity in animals 
and human beings has not received sufficient attention to determine 
their relative significance in human risk assessment (Khera, 1981). 

6.2.2.  Interspecies variations

    Even among laboratory species, the embryotoxic and teratogenic 
responses to the same chemical may vary markedly. Such variations 
in response complicate the extrapolation of laboratory animal data 
to man, because it is not known which is the best animal model.  
Interspecies differences are presumably due to differences in the 
intrinsic sensitivity of embryonic and fetal processes to chemical 
perturbation, the rate and stage of embryogenesis, and various 
toxicokinetic factors (Neubert & Chahoud, 1983; Khera, in press). 

    There is also a difference in the background for developmental 
abnormalities between some laboratory animal species and human 
beings.  For example, the preimplantation and early 
postimplantation losses in rats are low, less than 10% (Perraud, 
D., 1976; Fritz, H. et al., 1978; Frohberg, 1977), while the fetal 
wastage in human beings may be as high as 50% (section  
Human beings have an apparent mechanism to naturally abort, 
particularly if the fetus is abnormal.  At present, not enough is 
known to account for these differences in predicting potential 
human disease. 

    If a teratogen acts by means of the same biochemical mechanism 
in the animal model and in man, then the concentration of the 
chemical at the target site is an important determinant of the 
susceptibility of the species. In turn, the concentration of the 
chemical at the target cell is determined by pharmacokinetic 

variables.  By modelling pharmacokinetic constants in the test 
species and man, it may be possible to predict the incidence of 
teratogenicity based on maternal blood concentrations. 

    The differences in toxicokinetics make it almost impossible to 
extrapolate data from rodents to man with any certainty (Khera, 
1976).  Nau et al. (1981) demonstrated that, with the use of 
"minipumps", concentrations similar to the therapeutic 
concentrations obtained in human beings can be obtained in mice for 
valproic acid.  Such methods may aid an evaluation of teratogenic 

    Pharmacokinetic models may greatly aid in assessing the human 
risk associated with exposure to potential teratogens. These 
procedures, however, must be perfected and validated. In the 
future, computer models based on well-validated mechanisms, 
structure-activity relationships, pharmacokinetic data,  in vitro 
data, and whole animal models should greatly enhance the 
reliability of data extrapolation from laboratory animals to man. 

6.2.3.  Statistical limitations

    Although many variables are controlled in well-conducted animal 
tests, there remains the major problem of the statistical 
limitation of the study.  A basic aspect of toxicology is the dose-
response relationship relating the incidence of a specific effect 
induced by a chemical to increasing exposure levels.  The incidence 
of an effect and the size of the study population are important in 
determining the sensitivity of the study.  Because the underlying 
mechanisms may be different, it is advisable, in principle, to 
evaluate dose-effect relationships for all prenatally-induced toxic 
effects, separately.  When considering effects as completely 
separate entities, it must be recognized that the relevant dose-
range for that effect may be severely restricted.  If the endpoints 
studied are restricted to "all or none", such as low frequency 
events of malformations or abortions, then the power of statistical 
tests in studies of conventional dimensions is often extremely 
poor.  Because a single effect in prenatal toxicity may often be 
modulated and obscured by other effects of toxicity, such an 
approach can frequently prove unsatisfactory.  In such a case, the 
evaluation of compiled data concerning several effects may be 
necessary, although not completely justified.  For the purposes of 
interpretation of animal data for predicting potential human 
disease, the dose-effect relationship of the most sensitive effect 
has the greatest significance. 

    Statistical tests should be used as a guide and not as a 
substitute for biological judgement.  Before statistical 
significance can be transformed to toxicological relevance, the 
result must make biological sense and the same applies to "non-
significant" results. 

    Since by using fetus or litter as the statistical unit, 
important data may be lost, it is advisable to report effects based 
both on the number of litters and on the total number of fetuses 

observed (Neubert & Dilmann, 1972).  Statistical methods for 
combining fetal and litter data for evaluating chemically-induced 
effects have been described (Snedecor & Cochran, 1967; Sokal & 
Rohlf, 1969; Haseman & Kupper, 1979; Haseman & Hogan, 1975). 

6.2.4.  Quantitative risk assessment

    Since there is no universally-accepted mathematical model for 
extrapolating embryotoxic including teratogenic effects below the 
observed dose-effect range, safe levels of exposure for human 
beings have been estimated by applying a safety factor to the no-
observed-teratogenic-effect level (Lehman et al., 1954; FDA, 1974; 
EPA, 1975; NAS, 1977; Faber, 1980; ACGIH, 1981; Khera, 1984).  
Usually, large safety factors are used, because it is not possible 
to accurately characterize synergistic or antagonistic factors that 
may occur in human beings through exposure to a combination of 
chemicals in the ambient environment.  Khera (in press) compared 
the quantitative dose-effect data available in the literature for 
embryotoxic effects in human beings and animals.  Using crude 
estimates, the ratios between the lowest doses of several drugs 
reported in literature to have produced adverse effects in human 
progeny and the lowest reported embryotoxic doses in various animal 
species varied between 1 and 400. 

    A number of statistical (tolerance) models, such as probit, 
logit, and Weibul, which were developed for other toxicological 
endpoints, might also be used to describe teratogenic dose-response 
relationships.  Other mechanistic models such as the multi-hit and 
multi-stage models have been specifically developed for the 
carcinogenic response (Mantel & Schreiderman, 1975; Chaud & Hoel, 
1974; Brown, 1980; Krewski & Van Ryzin, 1981).  The tolerance 
models were developed on the basis of the notion that each 
individual in the population possesses a tolerance to the test 
compound.  If dose D does not exceed the tolerance for an 
individual, then there will be no effect on that individual; while 
a dose exceeding the tolerance will result in a positive response.  
The probability that the individual selected at random will respond 
at a dose D is given by, 

    P = f (D) = Pr [tolerance  <  D]

    The probit model is obtained when the tolerance is considered 
as a random variable, and follows a log normal distribution.  Logit 
and Weibul models are obtained when the tolerance is assumed to 
follow logistic or extreme value distribution, respectively.  The 
logit, like the probit model, has an S-shaped dose-response curve 
that is symmetric about the 50% response point.  The logit curve 
approaches the zero response more slowly than the probit curve, 
leading to a higher response rate for a given dose in the low-dose 
region (Brown, 1980). 

    All these statistical (tolerance) models and the mechanistic 
models will probably fit teratological dose-response data in the 
observed range reasonably closely; but would be expected to diverge 
substantially in the low-dose region.  Since the risk estimates 

obtained at low doses from both the statistical (tolerance) models 
and the mechanistic models are expected to cover a wide spectrum at 
low doses, they could be used to define a range of risk estimates 
at a given exposure level.  The primary reason for believing that 
such a risk range is plausible is based on the hypothesis that most 
biological effects have a dose-response curve in the low-dose range 
that is concave upwards.  If this is also true for teratogens, then 
the above models would establish a broad risk range for a given low 
dose exposure.  However, low-dose extrapolation for teratogens is 
largely an uninvestigated field.  Only with knowledge of the 
mechanisms of action, can an appropriate extrapolation model be 
selected for better describing risk at low doses for embryotoxic 
agents including teratogens. 

    Probit models have been proposed as an extrapolation model to 
determine safe human exposure levels (Mantel & Bryan, 1961; Weil, 
1972).  Probit models have also been used to compare the relative 
teratogenicity of different chemicals in a specific animal species 
(Fabro et al., 1981) and to fit experimental data in an observed 
dose-effect range (Neubert et al., 1973). 

    Biddle (1978) studied the potential of cortisone and 6-
aminonicotinamide to induce cleft palate in a variety of inbred and 
hybrid mouse strains using a probit model.  For both compounds, a 
separate family of parallel dose-effect curves was derived from the 
experimental data generated by studying the different genotypes.  
Parallel dose-effect curves were assumed to suggest a common 
mechanism of teratogenicity. Biddle concluded that when dose-effect 
curves for the different strains have the same slope, the ratios of 
estimated ED50s provide the most reasonable measure of relative 
strain sensitivity.  Hogan & Hoel (1982) pointed out the 
limitations of this approach.  Their main criticism was that, 
because approximately parallel dose-response curves for different 
strains of mice were restricted to the experimental dose range, it 
was not possible to know whether they would persist in the low-dose 
region, near the origin of the curve. 

    An attempt to ascertain whether a "threshold" exists in 
prenatal toxicity and whether embryotoxic agents may be grouped in 
different categories in this respect was made by Jusko (1972).  
Using a mathematical model and a graphing procedure, he suggested 
that embryotoxic substances might be grouped in two mechanists 
categories: agents that show a "threshold" and others that do not, 
with implications for low-dose extrapolation.  Although the method 
suffers from the common difficulty of having to project data in the 
low-dose range, the approach seems worthy of consideration with 
appropriate experimental data. 

    In any attempt to extrapolate animal data, the possible 
contribution of maternal toxicity to the toxic effects in the fetus 
must be considered, because most chemicals could damage the fetus 
by this mechanism, if exposure were high enough (Martson & 
Shepelskaja, 1982, 1983).  In recent studies, attempts have been 
made to quantify the toxic effects on the mother and the embryo and 
to associate dose-response relationships of the various parameters.  

From the data, it is possible to calculate indices, derived by 
comparing doses for the maternal toxic responses with the dose 
producing embryotoxic effects, including teratogenic effects 
(Murphy, 1965; Chaube & Murphy, 1968; Johnson, 1981; Vergieva, 
1981, 1982; Bass et al., 1982; Fabro et al., 1981, 1982; Platzek et 
al., 1982).  In some situations, these indices are being used to 
classify the relative hazard of environmental agents such as 
pesticides (Kalojanova, 1982) and to estimate safe levels for water 
pollutants (Vitvitskaja et al., 1980; Korolev & Agareva, 1983). 

6.3.  Interpretation of Human Data for Risk Assessment

    Human data are the most reliable basis for estimating risks of 
human disease.  These data, however, are difficult to obtain and 
may be of limited value for risk assessment because of confounding 
variables common to all epidemiology studies, including 
difficulties in exposure estimation, high background rates of birth 
defects in the human population, and wide variation in diagnosing 
and reporting defects.  These factors were discussed in detail in 
section 3.1 (Human studies).  For the purposes of quantitative risk 
extrapolation, the model that best fits the human dose-effect data 
in the observed range should be used and extrapolated to lower 
doses, if possible, on the basis of plausible mechanisms of action. 
Even negative human data can be useful for comparison with dose-
effect relationships and risk levels estimated from animal data, in 
determining an upper level for possible human risk. 


    In concluding this report, it is worthwhile to return to the 
questions posed in the introduction to the document: 

(a)  What is the value of currently used toxicity tests for
     predicting human risk?

(b)  What are the difficulties involved in applying these test
     results to estimate human health risks and hazards?

(c)  Can better testing strategies be developed?

1.   Animal tests are valuable in identifying chemicals that
can alter morphological and functional development. Extrapolation 
from such data to man is the principal method of predicting health 
risks.  Prenatally-induced toxicity is identified in animals and 
its transposition to man depends on essential elements in the 
experimental design.  The species (and strain) of the test animal 
must be carefully selected, bearing in mind availability, 
fecundity, and the incidence of spontaneous malformations or 
variations.  A knowledge of toxicokinetic variables, homeostatic 
mechanism, and ability to adapt to a chemical insult, relative to 
man, greatly aid in extrapolating laboratory data to man.  If a 
study is properly executed, it is possible to compare fetal and 
maternal levels of toxicity. 

2.  A number of difficulties are involved in applying laboratory 
animal test data to  estimate human risks.They become all the more 
pronounced, when attempting a quantitative risk assessment.  Animal 
studies usually involve healthy animals; food and water and other 
environmental factors are controlled; adequate group sizes are 
possible; and exposure is usually to a single agent.  The offspring 
can be carefully examined at any time.  In contrast, an exposed 
human population varies in genetic constitution, health, diet, and 
physical environment.  In addition, the subjects under study may 
elect to smoke, consume alcoholic beverages, or take drugs. 

    Human studies are perhaps the weakest link in the chain of 
events, necessary to develop and validate means of predicting human 
risk to a toxic chemical.  In epidemiology studies, it is rarely 
possible to identify the ideal test population and its control; the 
numbers of individuals in the test group are usually small and the 
incidence of the effect low and difficult to detect.  Toxic effects 
may be apparent at birth or manifest later during postnatal life.  
Therefore, few environmental chemicals have been shown to be toxic 
to the human fetus.  However, when collecting human epidemiological 
data, data on accidental exposure and other evidence of human 
hazards is emphasized.  Toxicokinetic data can sometimes be 
obtained on human beings occupationally or accidentally exposed to 
chemicals.  Quantitative data on human exposures are essential, 
both in conducting human studies to establish dose-response 
relationships and for the estimation of risks to sub-populations 
exposed to various levels of environmental agents. 

3.  There is no question that better  testing strategies could be 
developed in both laboratory animals and human studies. Topics that 
deserve emphasis include comparative toxicokinetic studies to 
enable interspecies comparison of test results and extrapolation to 
man; a better understanding of the genetic, molecular, and cellular 
processes that are the targets for toxic chemicals, which disturb 
them and cause dysfunction; short-term tests in which enzymes, 
cells, organs, and embryos in culture are used as well as sub-
mammalian species; and studies on effects of prenatal chemical 
exposure on neurobehavioural development and on cancer incidence. 

4.  Though data on the  molecular biology of development and 
morphogenetics is also increasing, information and understanding of 
crucial development processes is still rudimentary.  Nevertheless, 
available data seem to suggest that the biological mechanisms 
underlying normal development are similar and that interspecies 
differences are often only temporal in nature.  The disciplines of 
genetics, and molecular and cellular biology provide the 
experimental and clinical scientist with tools to detect altered 
genes or genetic regulation.  A molecular explanation should be 
sought for every endpoint induced by chemicals in experimental 

5.  Considerable scientific activity is concerned with the 
development of  short-term tests.They are useful research tools for 
investigating the processes underlying normal and abnormal 
development.  However, the Task Group was of the opinion that, with 
the present state of knowledge, short-term tests cannot replace 
whole animal-testing procedures for assessing risk to human 
subjects.  Short-term tests should be further developed to provide 
information to support data obtained in whole experimental animal 
studies, to provide a basis for selecting chemicals for further 
testing, and to develop models for possible correlation with 
results from whole animal tests. 

6.  Morphological defects at birth represent only a part of the 
effects produced by prenatal toxicity.  The process of development 
during its various stages is also susceptible to chemically-induced 
 adverse effects, manifested during postnatal life.These could 
involve every tissue and organ system and function, such as the 
immune system, reproductive capacity, neurobehavioural function.  
Transplacental carcinogenesis belongs to this category. 

    Although the importance of the potential effects of prenatal 
exposure to chemicals on postnatal development is acknowledged, 
there are, at present, no generally agreed testing strategies for 
the evaluation of the postnatal manifestation of such effects.  
However, it may be worth investigating the relative adult versus 
fetal sensitivity to chemicals to which many pregnant women have 
been, or are, unavoidably exposed.  As a primary task, compounds 
that are suspected of being human carcinogens (IARC, 1982), should 
be tested for transplacental carcinogenicity. 

7.  For  qualitative risk assessment,results from human studies, 
backed up by animal studies, provide the strongest evidence that a 

chemical is toxic to the developing progeny, if the mother is 
exposed during pregnancy.  In the absence of human studies, 
positive results in animal studies provide evidence that the agent 
is potentially toxic for the developing fetus, although the 
importance of carefully evaluating each individual study and all 
related biomedical evidence in completing the qualitative risk 
assessment is essential.  Categorizing of embryotoxic effects into 
specific fetal effects and non-specific effects is useful for 
establishing safety guidance. 

8.  At present, the statistical (tolerance) and mechanistic 
 extrapolation modelsthat have been designed for extrapolating 
high dose-effect data to lower doses for cancer and other 
toxicological endpoints, remain largely uninvestigated for 
teratogens.  Use of these models for teratogenic extrapolation, as 
a general matter, can only establish broad risk ranges associated 
with low-dose levels of exposure, unless there is some indication 
of the mode of action and some rational basis for extrapolating 
from the test species to man. 


    The Task Group was of the opinion that, though progress has 
been made in the evaluation of risks for progeny following prenatal 
exposure to chemicals, very little information on the mechanisms by 
which adverse effects are produced has been forthcoming.  The Group 
believed that the following recommendations, if carried out, would 
enhance the predictive value of the present approaches: 

1.  In the areas of genetics, and cellular and molecular biology, a 
    number of new techniques have been developed. These involve: 
    cell fusion, cellular migration, interaction and pattern 
    formation, monoclonal antibodies, embryo transfer, whole 
    chromosome and/or gene transfer, production and use of chimeric 
    animals, gene isolation,  in vitro mutagenesis, and general 
    recombinant-DNA methods. Such techniques should be applied to 
    determine the underlying  mechanisms in normal and abnormal 
     developmentwith a view to understanding and controlling 
    chemically-induced congenital abnormalities. 

2.  The following improvements in the design and conduct of  human 
     studies are recommended: 

        (a)  Collaborative international epidemiological
        studies, using standard techniques and quantitative
        assessments of exposure, should be performed.  The
        information will be useful in the study of the
        responses of different sub-populations to particular
        chemicals.  These studies can also provide
        dose-response data to validate animal models for
        human risk assessment.

        (b)  Registries monitoring congenital malformations
        and pediatric tumours should be expanded to increase
        the number of participating countries.  Data on the
        frequency and geographical distribution of
        malformations and infant neoplasms should be
        published regularly.

        (c)  Monitoring of occupational health should include
        complete reproductive and exposure data of workers.
        Special risk groups should be studied for pregnancy
        outcomes.  Pregnancy wastage, including spontaneous
        abortions, intrauterine deaths, and the more subtle
        effects that occur prenatally or develop postnatally,
        must also be considered.

        (d)  Relationships, if any, between structural defects
        and biochemical and/or functional defects, and fetal
        loss should be investigated.

        (e)  Methods should be established to diagnose
        so-called "minor abnormalities" in the general
        population and in risk groups.

3.  In order to make the  animal embryotoxicity studiesmore 
    meaningful for extrapolation to human safety, the following 
    suggestions are made: 

        (a)  It has been suggested that 3.5% of all human
        malformations are associated with the ill health and
        disease conditions of the mother (Kalter & Warkany,
        1983).  However, in animal studies, the role of
        maternal toxicity in the induction of fetal anomalies
        and embryotoxicity needs clarification.  Such a
        clarification would be extremely helpful in the
        embryo/fetal evaluation of chemicals and
        extrapolation of human safety from animal data.

        (b)  The role of maternal, placental, and fetal
        factors in activating or detoxifying environmental
        chemicals in different animal species and man should
        be clarified.

        (c)  The significance of different kinds of
        experimentally-induced gross structural abnormalities
        for a human risk assessment must be determined by
        well-designed studies.

        (d)  Studies should be designed with a view to
        defining target cells or receptors in the fetus that
        are specific for teratogens.

        (e)  Since human beings are usually exposed to a
        variety of chemicals, studies on the treatment of
        pregnant animals with a combination of two or more
        chemicals are suggested.

        (f)  In view of increasing knowledge, the design and
        interpretation of animal studies should be
        periodically appraised by an international group of

4.  Attention must be directed to the entire spectrum of 
    developmental toxicity including  postnatal manifestations in 
    human and animal studies.  Test methods must be developed to 
    monitor neonates and young children of mothers exposed to 
    chemicals during pregnancy, for functional abnormalities.  This 
    should include evaluation of neurobehavioural effects, the 
    immune and endocrine systems, and of xenobiotic metabolism, 
    preferably with non-invasive methods.  The background data for 
    these functions must be established. 

5.  Additional effort is needed to evaluate  neurobehavioural tests 
    for routine use in developmental toxicology. Attention should 
    be given to: 

        (a)  selection of chemicals that can be used as
        positive controls;

        (b)  identification of the most suitable tests and
        periods during postnatal development;

        (c)  development of procedures for the early detection
        of latent neurotoxic effects; and

        (d)  better understanding of basic mechanisms
        underlying neurobehavioural effects.

6.  The following research needs were recognized in the area of 
    transplacental carcinogenicity: 

        (a)  Fetal exposure to carcinogens in animal studies
        has been reported to increase susceptibility to
        tumour induction following further exposure,
        postnatally, to various carcinogens or promoting
        agents (section 4.4).  Verification of these results
        is needed.

        (b)  Results of a few animal studies have suggested
        that chemical exposure of the parent generation can
        be followed by an increased incidence of tumours in
        subsequent generations (section 4.4).  This
        phenomenon should be explored further.

        (c)  Animal models with a comparatively long pregnancy
        should be investigated to estimate the susceptibility
        of the embryo/fetus to transplacental carcinogens at
        various stages of development.

7.      (a)   Short-term testsshould be further developed to
        provide information on embryotoxic mechanisms and to
        determine their use for selecting chemicals for
        further animal testing;

        (b)  Short-term tests that show promise for chemical
        selection should be validated using the principles
        provided in section 5.  An international meeting to
        discuss validation methods is needed.  The
        sensitivity and susceptibility of any technique
        considered for use in short-term selection tests must
        be established taking into account results from  in

8.      (a)   Toxicokinetic modelsshould be developed in order
        to compare toxic effects among species, based on the
        amount of chemical present in the maternal blood,
        various tissues, or even at receptor sites of the dam
        or embryo.  These data are needed because a
        satisfactory risk assessment for human beings can
        never be based solely on the doses administered to
        animals and the human exposure levels.  Toxicokinetics, 
        and homeostatic and adaptive factors largely account for 
        differences in response. Finally, these data are needed to 
        establish a better understanding of the mechanisms of 
        action as a basis for qualitative risk extrapolation. 

        (b)  Research is needed to (i) generate basic
        biological information that will provide a more
        rational scientific basis for the proper application
        of safety factors for estimating "safe levels" of
        exposure; and (ii) develop the biological basis for
        selecting appropriate mathematical models.  These
        methods are germane to  extrapolation from high doses
         in animal studies to low exposure levels in human
         beings, in order to estimate the magnitude of effect
        in human beings.

9.   Priority for testingshould be high for chemicals: (i)
    that show an increased potential for altering development by 
    affecting genetic, molecular, or cellular processes; (ii) that 
    show structural similarities to known human teratogens; and 
    (iii) to which women of child-bearing age will probably be 


    This section explains the meanings of certain terms as used in 
this document.  These meanings are not necessarily valid for other 


    All products of conception derived from and including the 
fertilized ovum at any time during pregnancy, including the embryo 
or fetus and embryonic membranes. 

     Congenital malformation

    A permanent structural abnormality present at birth. 

     Critical period

    A particular developmental phase during which a morphogenetic 
system is especially vulnerable (e.g., neurulation, fusion of the 
facial swellings, elevation of palatal shelves). 

     Developmental toxicity

    Any adverse effect induced prior to attainment of adult life.  
This is a new term and includes effects induced or manifested in 
the embryonic or fetal period and those induced or manifested 
postnatally (before sexual maturity). 


    Any toxic effect on the conceptus resulting from prenatal 
exposure, including structural or functional abnormalities or 
postnatal manifestation of such effects. 

     Embryonic period

    The period from conception to the end of major organogenesis.  
Generally, the organ systems are identifiable at the end of this 

     Fetal period

    The period from the end of embryogenesis to the completion of 

     Poor pregnancy outcome

    A term used to describe the failure of a pregnancy to produce 
viable, biologically normal offspring.  It covers all adverse 
effects: fetal death (including abortion), intra-uterine growth 
retardation and abnormal development. 

     Pregnancy dating and conceptional age

    Throughout the text, Day 0 is considered to be the first day of 
pregnancy.  In human beings, gestational age (date from first day 
of last menstrual period) and conceptional age (date from 
conception) may be used.  In this document, conceptional age has 
been adopted. 

     Prenatally-induced abnormality

    Any structural, functional, or biochemical deviation from the 
norm, initiated prenatally, that can be detected during prenatal or 
postnatal life. 

     Sensitive period

    A developmental phase during which differentiating cells become 
susceptible to a given toxic agent.  The period may not be related 
to critical morphogenetic periods, but may be related to the 
appearance of specific receptors (e.g., for glucocorticoids). 


    An agent which when administered prenatally induces permanent 
abnormalities in structure. 


    A type of embryo/fetotoxicity that is restricted to permanent 
structural abnormalities produced by prenatal exposure to toxic 

a  "Teratogen" and "Teratogenicity" as agreed by the Task Group
are used here in the traditional sense derived from Greek word 
"teratas".  They refer to a particular case of embryotoxic effect 
demonstrated by structural malformation detectable by the present 
routinely used methods (Section 3.2).  The Group recognized that 
these meanings differ from those used by other bodies. 


ABEL, E.L.  (1980)  Smoking during pregnancy; a review of
effects on growth and development of offspring.   Hum. Bio.,
52: 593-625.

ACGIH  (1981)   American Conference on Governmental Industrial
 Hygienists, TLD for chemical substances and physical agents
 and the work room environment with intended changes for 1981,
Cincinnati, Ohio, ACGIH (Publication Office).

ADAMS, J. & BUELKE-SAM, J.  (1981)  Behavioral assessment of
the postnatal animal: testing and methods development.  In:
Kimmel, C. A. & Buelke-Sam, J., ed.   Developmental toxicology,
New York, Raven Press, pp. 233-258.

AHRENS, P.B., SOLURSH, M., & REITER, R.S.  (1977)
Stage-related capacity for limb chondrogenesis in cell
culture.   Dev. Biol., 60: 69-82.

AKUTA, T.  (1979)  Effects of rearing conditions on the
behavior of mothers and on the later open-field behavior of
the offspring in the mouse.   Jpn. J. Psychol., 50: 73-81.

ALBERT, R.E., TRAIN, R.E., & ANDERSON, E.  (1977)  Rationale
developed by the Environmental Protection Agency for the
assessment of carcinogenic risks.   J. Natl. Cancer Inst., 58:

ALDER, S. & ZBINDEN, G.  (1977)  Methods for the evaluation of
physical, neuromuscular and behavioral development of rats in
early postnatal life.  In: Neubert, D., Merker, H.-J., &
Kwasigroch, T. E., ed.   Methods in prenatal toxicology,
Stuttgart, georg Thieme Publ., pp. 175-185.

ALEXANDROV, V.A.  (1968) Blastomogenic effect of
dimethylnitrosamine on pregnant rats and their offspring.
 Nature (Lond.), 218: 280-281.

ALEXANDROV, V.A.  (1969)  [Transplacental blastomogenic effect
of  N-nitrosomethylurea on rat offspring.]   Vop. Onkol., N4:
55-61 (in Russian).

ALEXANDROV, V.A.  (1972)  [Effects of  N-nitrosomethylurethane
on rat embryos.]   Vop. Onkol., N7: 59-64 (in Russian).

ALEXANDROV, V.A.  (1973)  [Teratogenic and transplacental
carcinogenic blastomogenic effect of combined treatment with
alkylurea and nitrite (endogenous formation of
alkylnitrosourea).]   [Proceedings of the First Symposium
 "Carcinogenic N-nitrosocompounds action, synthesis and
 detection",Tallin,] pp. 84-86 (in Russian).

ALEXANDROV, V.A.  (1974)  [Embryotoxic, teratogenic and
carcinogenic effects of  N-nitrosomethylurea in rats.]   Vop.
 Onkol., 12: 76-82 (in Russian).

ALEXANDROV, V.A.  (1976)  Some results and prospects of
transplacental carcinogenesis studies.   Neoplasma, 
23: 285-299.

ALEXANDROV, V.A. & ANISIMOV, V.N.  (1976)  [Peculiarities of
transplacental effect of  N-nitrosomethylurea in persistent
oestrus rats.]   Vop. Onkol., 11: 98-102 (in Russian).

ALEXANDROV, V.A. & NAPALKOV, N.P.  (1976)  Experimental study
of relationship between teratogenesis and carcinogenesis in
the brain of the rat.   Cancer Lett., 1: 345-350.

ALEXANDROV, V.A. & NAPALKOV, N.P.  (1979)  [Transplacental
carcinogenic effect on mice after combined administration of
amidopyrine and nitrite.]   Vop. Onkol., 7: 48-52 (in Russian).

ALEXANDROV, V.A. & NAPALKOV, N.P.  (1981)  General
regularities of transplacaental (prenatal) carcinogenesis.
 Biol. Res. Pregnancy, 2: 181-187.

(1980)  Fetal hydantoin syndrome, neuroblastoma, and
hemorrhagic disease in a neonate.   J. Am. med. Assoc., 244:

ALTMAN, J. & SUDARSHAN, K.  (1975)  Postnatal development of
locomotion in the laboratory rat.   Anim. Behav., 23: 896-920.

D.  (1971)  The influence of nutrition on neural and
behavioral development:  III. Development of some motor,
particularly locomotor patterns during infancy.   Dev.
 Psychobiol., 4: 97-114.

AMSEL, A., BURDETTE, D.R., & LETZ, R.  (1976)  Appetitive
learning, patterned alternation and extinction in 10-d-old
rats with non-lactating suckling as reward.   Nature (Lond.),
262: 816-818.

 methods in use to assess cancer risks.  Risk analyses.

ANDREW, F.D. & LYTZ, P.S.  (1981)  Biochemical disturbances
associated with developmental toxicity.  In: Kimmel, C.A. &
Buelke-Sam, J., ed.   Developmental toxicology, New York, Raven
Press, pp. 145-165.

ARCHER, J.  (1973)  Tests for emotionality in rats and mice: a
review.  Anim. Behav., 21: 205-235.

ARMUTH, V. & BERENBLUM, I.  (1977)  Possible two-stage
transplacental liver carcinogenesis in C57BL/6 mice.   Int. J.
 Cancer, 20: 292-295.

AUSTIN, C.R.  (1973)  Embryo transfer and sensitivity to
teratogenesis.   Nature (Lond.), 244: 333-334.

AYDELOTTE, M.B. & KOCHHAR, D.M.  (1972)  Development of mouse
limb buds in organ culture: chondrogenesis in the presence of
a proline analog, L-azetidine-2-carboxylic acid.   Dev. Biol.,
28: 191-201.

BARANOVA, L.N. & ALEXANDROV, V.A.  (1978)  [Peculiarities of
transplacental penetration of 7,12-dimethylbenz( a)anthracene
in rats, and its distribution in fetus organs as studied by
spectrofluorescent method.]   Bull. exp. Biol., 10: 477-479 (in

BARRACH, H.J. & NEUBERT, D.  (1980)  Significance of organ
culture techniques for evaluation of prenatal toxicity.   Arch.
 Toxicol., 45: 161-187.

(1975)  Some biochemical characteristics of mouse limb buds
differentiating in organ culture.  In: Neubert, D. & Merker,
H.-J., ed.   New approaches to the evaluation of abnormal
 embryonic development,Stuttgart, Georg Thieme Publ., pp.

BARRACH, H.J., BAUMANN, I., & NEUBERT, D.  (1978)  The
applicability of  in vitro systems for the evaluation of the
significance of pharmacokinetic parameters for the induction
of an embryotoxic effect.  In: Neubert, D., Merker, H.-J.,
Nau, H., & Langman, J., ed.   Role of pharmacokinetics in
 prenatal and perinatal toxicology,Stuttgart, Georg Thieme
Publ., pp. 323-336.

Immunofluorescent microscopic investigations of intercellular
substances during limb development.  In: Merker, H.-J., Nau,
H., & Neubert, D., ed.   Teratology of the limbs, Berlin,
Walter de Gruyter Publ., pp. 272-293.

Comparison of the differentiation of muscle and connective
tissue of mouse limb buds in culture and  in vivo: a
morphological study by indirect immunofluorescence.  In:
Neubert, D. & Merker, H.-J., ed.   Culture techniques.
 Applicability for studies on prenatal differentiation and
 toxicity, Berlin, New York, Walter de Gruyter Publ., pp.

Effects on reproduction and other toxicological endpoints.
Quantitative dose-response models in teratogenesis.  Presented
at the SGOMSEC Workshop on:  "Quantitative estimation of risk
 to human health from chemicals",Rome, 12-16 July, 1982.

BATTAGLIA, F.C. & MESCHIA, G.  (1978)  Principal substrates of
fetal metabolism.   Physiol. Rev., 58: 499-527.

(1977)  The effects of thalidomide and two analogues on the
regenerating forelimb of the newt.   J. Embryol. exp. Morphol.,
41: 125-135.

BECK, F.  (1977)  Evaluation of organs (gross organ
pathology).  In: Neubert, D., Merker, H.-J., & Kwasigroch, T.
E., ed.   Methods in prenatal toxicology, Georg Thieme Publ.,
Stuttgart, pp. 103-112.

BECK, F. & GULAMHUSEIN, A.P.  (1980a)  The contrast between
mouse and ferret limb buds in culture - possible advantages of
comparing results from a limb culture system with whole embryo
explantation.  In: Merker, H.-J., Nau, H., Neubert, D., ed.
 Teratology of the limbs, Berlin, Walter de Gruyter Publ., pp.

BECK, F. & GULAMHUSEIN, A.P.  (1980b)  The effect of sodium
salicylate on limb development.  In: Merker, H.-J., Nau, H., &
Neubert, D., ed.   Teratology of the limbs, Berlin, Walter de
Gruyter Publ., pp. 393-401.

BENDEK, G. & HAHN, Z.  (1981)  Effect of amphetamine on the
metabolism and incorporation of 3H-thymidine into DNA of
developing rat brain.   Dev. Neurosci., 4: 55-65.

BENESOVA, O., RYCHTER, Z., & JELINEK, R., ed.  (1979)
 Evaluation of embryotoxicity, mutagenicity, and
 carcinogenicity risks in new drugs, Prague, Universita Karlova.

BENIASVILI, D.S.  (1978)  [Comparative study of action of some
carcinogenic compounds inducing transplacental blastomogenesis
in rabbits.]   Vop. Onkol., 3: 77-83 (in Russian).

BENIRSCHKE, K.  (1975)  Placental causes of maldevelopment.
In: Berry, C.L. & Poswillo, D.E., ed.   Teratology: trends and
 applications, Berlin, Heidelberg, New York, Springer-Verlag,
pp. 148-164.

BERRY, C.L.  (1981)  Congenital malformations.  In: Berry,
C.L., ed.   Paediatric pathology, Berlin, Heidelberg, New York,
Springer-Verlag, pp. 67-86.

BERRY, C.L. & POSWILLO, D.E., ed.  (1975)   Teratology trends
 and applications, Berlin, Heidelberg, New York, Springer-

BIDDLE, F.G.  (1978)  Use of dose-response relationship to
discriminate between the mechanisms of cleft palate induction
by different teratogens:  an argument for discussion.
 Teratology, 18: 247-252.

BIGGERS, J.D., MCLAREN, A., & MICHIE, D.  (1958)  Variance
control in the animal house.   Nature (Lond.), 182: 77-80.

culture of mouse embryos  in vitro. in: Daniel, J. C., ed.
 Methods in mammalian embryology, San Francisco, Freeman, W.
H., pp. 86-116.

acid content of limb buds at different stages of mammalian
embryonic development.  In: Merker, H.-J., Nau, H., Neubert,
D., ed.   Teratology of the limbs, Berlin, Walter de Gruyter
Publ., pp. 183-190.

BLANKENBURG, G., BLUTH, U., & NEUBERT, D.  (1981)  On the
significance of ascorbate and of cysteine on differentiation
of limb buds in organ culture.  In: Neubert, D. & Merker,
H.-J., ed.   Culture techniques.  Applicability for studies on
 prenatal differentiation and toxicity, Berlin, New York,
Walter de Gruyter Publ., pp. 197-206.

BLOZOVSKI, D. & CUDENNEC, A.  (1980)  Passive avoidance
learning in the young rat.   Dev. Psychobiol., 13: 513-518.

BOCHERT, G., PLATZEK, T., & WIESSLER, M.  (1981)  Comparison
of effects on limb development  in vivo and  in vitro using
methyl(acetoxymethyl)nitrosamine.  In: Neubert, D. & Merker,
H.-J., ed.   Culture techniques.  Applicability for studies on
 prenatal differentiation and toxicity, Berlin, New York,
Walter de Gruyter Publ., pp. 223-235.

BOLLES, R.C. & WOODS, P.J.  (1964)  The ontogeny of behaviour
in the albino rat.   Anim. Behav., 12: 427-441.

Transplacental induction of lymphomas in Swiss mice by
carbendazim and sodium nitrite.   Int. J. Cancer, 17: 742-747.

L., & ROLLER, P.  (1980)  Carcinogenic effect of
dinitrosopiperazine in adult Swiss mice and after
transplacental or translactional exposure.   Cancer Res., 40:

BOUE, J., BOUE, A., & LAZAR, P.  (1975)  Retrospective  and
prospective epidemiological studies of 1500 karyotyped
spontaneous human abortions.   Teratology, 12: 11-26.

 Drosophila embryon cell cultures as an  in vitro teratogen
assay.   Teratog. Carcinog. Mutag., 2: 333-341.

BOYER, C.C.  (1953)  Chronology of development for the golden
hamster.   J. Morphol., 92: l-38.

BRACHET, A.  (1913)  Recherches sur le déterminisme
héréditaire de l'oeuf des mammifères. Développement  in vitro
de jeunes vésicules blastodermiques du lapin.   Arch. Biol.,
 Paris, 28: 447-503.

Teratogenic drugs inhibit tumour cell attachment to
lectin-coated surfaces.   Nature (Lond.), 282: 507-509.

Inhibition of tumour cell attachment to concavalin A-coated
surfaces as an essay for teratogenic agents: approaches to
validation.   Teratog. Carcinog. Mutag., 2: 343-354.

BRINKLEY, L.L.  (1980)   In vitrostudies of palatal shelf
elevation.  In: Pratt, R. M. & Christiansen, R. L., ed.
 Current research trends in prenatal craniofacial development,
New York, Amsterdam, Oxford, Elsevier/North-Holland Publ., pp.

BRINSTER, R.L.  (1963)  A method for  in vitro cultivation of
mouse ova from two-cell to blastocyst.   Exp. Cell Res., 32:

BRINSTER, R.L.  (1970)   In vitrocultivation of mammalian
ova.  In: Raspé, G., ed.   Advances in the biosciences, oxford,
London, Edinburgh, New York, Toronto, Sidney, Paris,
Braunschweig, Pergamon Press - Vieweg, Vol. 4, pp. 199-233.

(1974)  The development of habituation in the rat.   Anim.
 Learn. Behav., 2: 92-96.

BROWN, C.C.  (1980)  High dose to low dose extrapolation in
animals.  Presented at:  The Food and Drug Law Institute
 Symposium on the Practical Applications of Risk Assessment,
 June 14(Proceedings in print).

(1978)  A comparison of behavioral and anatomical measures of
hydroxyurea induced abnormalities.   Teratology, 18: 379-384.

BUELKE-SAM, J. & KIMMEL, C.A.  (1979)  Development and
standardization of screening methods for behavioural
teratology.   Teratology, 20: 17-30.

BUTCHER, R.E.  (1976)  Behavioral testing as a method for
assessing risk.   Environ. Health Perspect., 18: 75-78.

(1978)  A simple recording grip strength device.   Pharmacol.
 Biochem. Behav., 8: 10l-102.

CAMPBELL, B.A. & COULTER, X.  (1976)  Neural and physiological
preocesses underlying the development of learning and memory.
In: Tighe, T. J. & Leaton, R. N., ed.   Habituation:
 perspectives from child development, animal behaviour and
 neurophysiology, Hillsdale, New Jersey, Lawrence Erlbaum
Associates, pp. 129-157.

CAMPBELL, B.A. & MABRY, P.D.  (1973)  The role of
catecholamines in behavioural arousal during ontogenesis.
 Psychopharmacology, 31: 253-264.

CAMPBELL, B.A. & RASKIN, L.A.  (1978)  Ontogeny of habituation
and sensitization in the rat.  In: Spear, N. A. & Campbell, B.
A., ed.   Ontogeny of learning and memory, Hillsdale, New
Jersey, Lawrence Erlbaum, pp. 67-100.

CAMPBELL, B.A. & STEHOUWER, D.J. (1980)  Retention of
habituation and sensitization in neonatal rats.   Behav.
 Neurol. Biol., 29: 190-202.

CAMPBELL, B.A., LYTLE, L.D., & FIBIGER, H.C.  (1969)  Ontogeny
of adrenergic arousal and cholinergic inhibitory mechanisms in
the rat.   Science, 166: 635-637.

CANADA, NATIONAL HEALTH AND WELFARE  (1973)   Method of testing
 for carcinogenicity, mutagenicity, teratogenicity, ottawa, NHW
Publishing Department, 139 pp.

CAPLAN, A.I.  (1970)  Effects of the nicotinamide-sensitive
teratogen 3-acetylpyridine on chick limb cells in culture.
 Exp. Cell Res., 62: 341-355.

CARTER, C.O.  (1976)  Genetics of common single
malformations.   Br. med. Bull., 32: 21-26.

CHAHOUD, I. & EGGERT, S.  (1978)  Drug-induced impairment of
liver function in the prenatal period.  In: Neubert, D.,
Merker, J.-H., Nau, H., & Langman, J., ed.   Role of
 pharmacokinetics in prenatal and perinatal toxicology,
Stuttgart, Georg Thieme Publ., pp. 263-271.

CHAHOUD, I., MICO-GINER, L., & NEUBERT, D.  (1975)  Some
evaluations of effects of prenatal drug treatment on postnatal
development.  In: Neubert, D. & Merker, H.-J., ed.   New
 approaches to the evaluation of abnormal embryonic
 development,Stuttgart, Georg Thieme Publ., pp. 502-516.

CHANCE, M.R.A.  (1957)  The contribution of environment to
uniformity.   Coll. Pap. Lab. Anim. Bur., 6: 59.

CHANG, M.C.  (1949)  Effects of heterologous sera on
fertilized rabbit ova.   J. gen. Physiol., 32: 291-300.

CHATOT, C.L., KLEIN, N.W., PIATEK, J., & PIERRO, L.J.  (1980)
Successful culture of rat embryos on human serum: use in the
detection of teratogens.   Science, 207: 1471-1473.

CHAUBE, S. & MURPHY, M.L.  (1968)  The teratogenic effects of
the recent drugs active in cancer chemotherapy.  In: Wollam,
D. H. M., ed.   Advances in teratology, New York, London,
Academic Press, pp. 181-237.

CHAUD, N. & HOEL, D.G.  (1974)  A comparison of models for
determining safe levels of environmental agents.  In:
 Reliability and biometry, statistical analysis of life length,
Philadelphia, Penn., SIAM, pp. 681-700.

CIBA FOUNDATION SYMPOSIUM 27 (new series)  (1974)   Size at
 birth, amsterdam, Oxford, New York, Elsevier, Excerpta Medica,
North-Holland, Associated Scientific Publishers.

CIOMS  (1983)   Safety requirements for the first use of new
 drugs and diagnostic agents in man, Geneva, Council for
International Organizations of Medical Sciences.

CLARKSON, S.G., DOERING, J.V., & RUNNER, M.N.  (1969)  Growth
of postimplantation mouse embryos cultured in a
serum-supplemented, chemically defined medium.   Teratology, 2:

CLEGG, D.J.  (1979)  Animal reproduction and carcinogenicity
studies in relation to human safety evaluation.  In:
Deichmann, W.B., ed.   Toxicology and occupational medicine,
New York, Amsterdam, Oxford, Elsevier North Holland, pp. 45-59.

CMEA  (1982)  [Methodological recommendations, protocols,
experimental studies for establishment threshold levels of the
effect of industrial compounds on reproductive function].  In:
 Problemi promislenoi toksikologii, SEV, Postoiannaja
 komissija po sotrudnicestvu v oblasti zdravoohranenija,
Moskva, 1982, pp. 58-75 (in Russian).

COCKROFT, D.L. & COPPOLA, P.T.  (1977)  Teratogenic effects of
excess glucose on head-fold rat embryos in culture.
 Teratology, 16: 141-146.

CONNORS, T.A.  (1975)  Cytotoxic agents in teratogenic
research.  In: Berry, C.L. & Poswillo, D.E., ed.   Teratology,
 trends and applications, New York, Springer-Verlag, pp. 49-79.

COOKE, J.  (1976)   Mathematical models of biological
 parameters,Thesis, Hatfield Polytechnic, England.

COOKE, J.  (1981)   Evaluation of data recording methods for
 statistical testing of behavioural responses in the rat,
Thesis, Cranfield Institute, England.

(1978)  DNA alkylation and neuro-oncogenesis by
3,3-dimethyl-1-phenyltriazene.   Acta neuropathol. (Berlin),
43: 105-109.

HENLEY, C.  (1957)   Methods for obtaining and handling marine
 eggs or embryos,Wood Hool, USA, Marine Biology Laboratory.

synthesis and repair in terminally differentiating embryonic
lens cells.  Dev. Biol., 57: 47-55.

COYLE, I., WAYNER, M.J., & SINGER, G.  (1976)  Behavioral
teratogenesis: a critical evaluation.   Pharmacol. Biochem.
 Behav., 4: 191-200.

CROCKER, J.F.S.  (1973)  Human embryonic kidneys in organ
culture: abnormalities of development induced by decreased
potassium.   Science, 181: 1178-1179.

CROCKER, J.F.S. & VERNIER, R.L.  (1970)  Fetal kidney in organ
culture: abnormalities of development induced by decreased
amounts of potassium.   Science, 169: 485-487.

CULVER, B. & NORTON, S.  (1976)  Juvenile hyperactivity in
rats after acute exposure to carbon monoxide.   Exp. Neurol.,
50: 80-98.

DANIEL, J.C., Jr  (1965)  Studies on the growth of 5-day old
rabbit blastocysts  in vitro.  J. Embryol. exp. Morph., 13:

DANIEL, J.C., Jr  (1968)  Oxygen concentrations for culture of
rabbit blastocysts.   J. Reprod. Fertil., 17: 187-190.

DAVIS, G.J., MCLACHLAN, J.A., & LUCIER, G.W.  (1978)  The
effect of 7,12-dimethylbenz-( a)-anthracene (DMBA) on the
prenatal development of gonads in mice.   Teratology, 17: 33A.

DAVIS, L.A., SADLER, T.W., & LANGMAN, J.  (1981)   In vitro
development of the heart under influence of retinoic acid.
In: Neubert, D. & Merker, H.-J., ed.   Culture techniques.
 Applicability for studies on prenatal differentiation and
 toxicity, Berlin, New York, Walter de Gruyter Publ., pp.

DAVIS, M.  (1980)  Neurochemical modulation of sensory motor
reactivity: acoustic and tactile startle reflexes.   Neurosci.
 Biobehav. Rev., 4: 241-263.

DELAHUNT, C.S. & LASSEN, L.J.  (1964)  Thalidomide syndrome in
monkeys.   Science, 146: 1300-1305.

DEARBORN, E.H.  (1967)  Comparative toxicity of drugs.   Fed.
 Proc., 26: 1075-1077.

DEMEREC, M., ed.  (1950)   Biology of Drosophila, New York,
John Wiley.

YODER, B., & VON EHRENSTEIN, G.  (1978)  Cell lineages of the
embryo of the nematode  Caenorhabditis elegans.  Proc. Natl.
 Acad. Sci. USA, 75: 376-380.

DEWS, P. & WENGER, G.  (1979)  Testing for behavioural effects
of agents.   Neurobehav. Toxicol., 1: 119-127.

DIWAN, B.A. & MEIER, H.  (1974)  Strain and age-dependent
transplacental carcinogenesis by 1-ethyl-1-nitrosourea in
inbred strains of mice.   Cancer Res., 34: 764-770.

DIXON, R.L.  (1982)  Potential of environmental factors to
affect development in reproductive systems.   Fundam. appl.
 Toxicol., 2: 5-12.

(1978)  Reaction of 7,12-dimethylbenz[ a]anthracene with DNA of
fetal and maternal rat tissues  in vivo.  Int. J. Cancer, 22:

DRUCKREY, H.  (1973)  Chemical structure and action in
transplacental carcinogenesis and teratogenesis.  In: Tomatis,
L., Mohr, U., & Davis, W., ed.   Transplacental carcinogenesis,
Lyons, pp. 45-58, IARC Sci. Publ., No. 4.

DUSEMUND, B. & BARRACH, H.J.  (1981)  Quantification of
collagen types I and II in mouse limbs during differentiation
 in vitro and  in vivo. in: Neubert, D. & Merker, H.-J., ed.
 Culture techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 161-169.

DYBAN, A.P.  (1977)  [Theoretical and applied aspects of
experimental teratology].   Ontogenesis (Sov. J. Devel. Biol.),
8(6): 582-589 (in Russian).

DYBAN, A.P., SEKIRINA,  G.G., & GOLINSKY, G.F.  (1976)
[Sensitivity to antifolic drug pyrimethamine early mice
embryos  in vitro and  in vivo.]   Bull. exp. Biol. Med., 10:
47-51 (in Russian).

DYBAN, A.P., SEKIRINA, G.G., & GOLINSKY, G.F.  (1977b)  [The
effect of aminopterin on the preimplantation rat embryos
cultivated  in vitro.]   Ontogenesis (Sov. J. Devel. Biol.), 8:
121-127 (in Russian).

(1979)  Testing of some chemical environmental pollutants to
teratogenicity using mammalian embryos cultivated  in vitro.
 Proceedings of the US-USSR third Joint Symposium on Problems
 on Environmental Health,Suzdal, USSR, pp. 300-322.

EBERT, J.D. & MAROIS, M., ed.  (1976)   Tests of teratogenicity
 in vitro, Amsterdam, New York, Oxford, North Holland
Publishing Co.

WOOD, P.J.  (1982)  Early embryonic mortality in women.
 Fertil. Steril., 38: 447-453.

EDWARDS, J.H.  (1969)  Familial predisposition in man.   Br.
 med. Bull., 25: 58-64.

EDWARDS, R.G.  (1980)   Conception in the human female.
London, New York, Academic Press, pp. 573-667, 668-766.

EDWARDS, R.G., BAVISTER, B.D., & STEPTOE, P.C.  (1969)  Early
stages of fertilization  in vitro of human oocytes matured  in
 vitro.  Nature (Lond.), 221: 632-635.

EGGER, H.J., WITTFOHT, W., & NAU, H.  (1978)  Identification
of diphenylhydantoin and its metabolites, including the
dihydrodiol and the catechols in maternal plasma, placenta and
fetal tissues of man.  In: Neubert, D., Merker, H.-J., Nau,
H., & Langman, J., ed.   Role of pharmacokinetics in prenatal
 and perinatal toxicology,Stuttgart, Georg Thieme Publ., pp.

EIBS, H.G. & SPIELMANN, H.  (1977)  Preimplantation embryos.
Part II: Culture and transplantation.  In: Neubert, D.,
Merker, H.-J., & Kwasigroch, T.E., ed.   Methods in prenatal
 toxicology,Stuttgart, Georg Thieme Publ., pp. 221-230.

ELLIS, T.M.  (1967)  Environmental influences on drug
responses in laboratory animals.  In: M.L. Conalty, Husbandry
of laboratory animals:  Proceedings of the 3rd International
 Symposium,London, Academic Press, 650 pp.

(1980)  Formation of a transplacental mutagen,
l,3-di(4-sulfamoylphenyl)triazene, from sodium nitrite and
sulfanilamide in human gastric juice and in the stomachs of
hamsters.   J. Natl. Cancer Inst., 65: 547-551.

ERICKSON, B.H.  (1967)  Radioresponse of the pre-puberal
porcine ovary.   Intern. J. radiat. Biol., 13: 57-67.

Parental occupation and birth defects.  A preliminary report.
In: Klingberg, M.A. & Weatherall, J.A.C., ed.   Epidemiologic
 methods for detection of teratogens. Contributions to
 epidemiology and biostatistics. Basel, München, Paris, London,
New York, Sydney, S. Karger, pp. 107-117.

EVANS, H.L.  (1978)  Behavioral assessment of visual
toxicity.   Environ. Health Perspect., 26: 53-57.

FABER, E.  (1980)  Problems in the safety evaluation of tissue
residues.   J. environ. Pathol. Toxicol., 3: 73-79.

FABRO, S., SHULL, G., & BROWN, N.A.  (1982)  The relative
teratogenic index and teratogenic potency: proposed components
of developmental toxicity risk assessment.   Teratog. Carcinog.
 Mutag., 2: 61-76.

FABRO, S.  (1973)  Passage of drugs and other chemicals into
the uterine fluids and preimplantation blastocyst.  In:
Boréus, L.O., ed.   Fetal pharmacology, New York, Raven Press
Publ., pp. 443-461.

FABRO, S., SHULL, G., & BROWN, N.A.  (1981)   The relative
 teratogenic index: an approach to the estimation of
 teratogenic potency,Research Triangle Park, TCM, Laboratory
of Reproductive and Developmental Toxicology, NIEHS.

FAIRBANKS, M.B. & KOLLAR, E.J.  (1974)  Inhibition of palatal
fusion  in vitro by hadacidin.   Teratology, 9: 169-178.

FANTEL, A.G.  (1982)  Culture of whole rodent embryos in
teratogen screening.   Teratog. Carcinog. Mutag., 2: 231-242.

(1979)  Teratogenic bioactivation of cyclophosphamide  in
 vitro. Life Sci., 25: 67-72.

FARIS, R.A. & CAMPBELL, T.C.  (1983)  Long-term effects of
neonatal phenobarbital exposure on aflatoxin B1 disposition in
adult rats.  Cancer Res., 43: 2576-2583.

FDA  (1966)   Food and Drug Administration: Guidelines for
 reproduction studies for safety evaluation of drugs for human
 use,January 1966, US FDA.

FDA  (1970)  Food and Drug Administration. Advisory Committee
on Protocols for Safety Evaluations: Panel on reproduction
Report on Reproduction Studies in the Safety Evaluation of
Food Additives and Pesticide Residues.   Toxicol. appl.
 Pharmacol., 16: 264-296.

FDA  (1974)  Food and Drug Administration, Poisonous or
deleterious substances in peanuts, evaporated milk, fish, and
shellfish.   Fed. Reg., 39: 42738-42752.

FECHTER, L.D.  (1974)  The effects of L-dopa, clonidine, and
apomorphine on the acoustic startle reaction in rats.
 Psychopharmacology, 39: 331-344.

FELL, H.B. & WEISS, L.  (1965)  The effect of antiserum, alone
and with hydrocortisone, on foetal mouse bones in culture.   J.
 exp. Med., 121: 551-560.

FLINT, O.P.  (1980)  The effects of sodium salicylate,
cytosine arabinoside, and eserine sulphate on rat limb buds in
culture.  In: Merker, H.-J., Nau, H., & Neubert, D., ed.
 Teratology of the limbs, Berlin, New York, Walter de Gruyter
Publ., pp. 325-338.

FLODH, H., MAGNUSSON, G., & MALMFORS, T.  (1977)  Toxicological 
studies on ferastralR, an iron-poly (sorbitol-gluconic acid) 
complex.   Scand. J. Haematol. Suppl., 32: 58-68.

FOX, W.M.  (1965)  Reflex ontogeny and behavioural development
of the mouse.   Anim. Behav., 13: 234-241.

FRANKOVA, S.  (1968)  Nutritional and psychological factors in
the development of spontaneous behaviour in the rat.  In:
Scrimshaw, N.S. & Gordon, J.E., ed.   Malnutrition, learning
 and behaviour, Cambridge, Massachusetts, MIT Press, pp.

FRANKOVA, S.  (1973)  Effect of protein-calorie malnutrition
on the development of social behavior in rat.   Dev.
 Psychobiol., 6: 33-43.

FRANKOVA, S.  (1977)  Drug-induced changes in the maternal
behavior of rats.   Psychopharmacology, 53: 83-87.

FRANKOVA, S. & BARNES, R.H.  (1968)  Effect of malnutrition in
early life on avoidance conditioning and behaviour of adult
rat.   J. Nutr., 96: 485-493.

H.  (1978)  Collection of control data from teratologic
experiments on mice, rats, and rabbits.   Arzneim.-Forsch., 28:

FROHBERG, H.  (1977)  An introduction to research in
teratology.  In: Neubert, D., Merker, H.-J., & Kwasigroch, T.
E., ed.   Methods in prenatal toxicology,Stuttgart, Georg
Thieme Publ., pp. 1-13.

GALBRAITH, D.B. & KOLLAR, E.J.  (1974)  Effects of
L-azetidine-2-carboxylic acid, a proline analogue, on the in
vitro development of mouse tooth germs.   Arch. oral Biol., 19:

GEBHARDT, D.O.E. & FABER, J.  (1966)  The influence of
aminopterin on limb regeneration in Ambystoma mexicanum.   J.
 Embryol. exp. Morphol., 16: 143-158.

GELSTEIN, V. I.  (1961)  [Development of tumours in mouse
offspring after exposure to orthoaminoazotoluene.]   Vop.
 Onkol., N10: 58-64 (in Russian).

GEMBERLING, G.A., DOMJAN, M., & AMSEL, A.  (1980)  Aversion
learning in 5-day-old rats: taste-toxicosis and texture-shock
associations.   J. comp. Physiol. Psychol., 94: 734-745.

GERALL, A.A. & MCCRADY, R.E.  (1970)  Receptivity scores of
female rats stimulated either manually or by males.   J.
 Endocrinol., 46: 55-59.

Regeneration of hydra from reaggregated cells.   Nature, New
 Biol., 239: 98-101.

SCHOENBERG, H.W.  (1979)  Association of diethylstilbestrol
exposure in utero with cryptorchidism, testicular hypoplasia
and semen abnormalities.   J. Urol., 122: 36-39.

GINSBURG, J.  (1971)  Placental drug transfer.   Ann. Rev.
 Pharmacol., 11: 387-408.

GIVELBER, H.M. & DIPAOLO, J.A.  (1968)  Growth of explanted
eight day hamster embryos in circulating medium.   Nature
 (Lond.), 220: l131-1132.

GOERTTLER, K. & LOEHRKE, H.  (1976)  Diaplacental carcino-
genesis: initiation with the carcinogens dimethylbenz-
anthracene (DMBA) and urethane during fetal life and postnatal
promotion with the phorbol ester TPA in a modified 2-stage
Berenblum-Mottram experiment.   Virchow's Arch., A 372(1):

X-ray sensitivity of the preimplantation mouse embryo  in
 vitro.  Radiat. Res., 62: 276-287.

[Blastomogenic effect of some nitrogen compounds on offspring
of mice.]   Bull. exp. Biol., 2: 62-65 (in Russian).

GOTH, R. & RAJEWSKY, M.F.  (1974)  Persistence of
06-ethylguanine in rat-brain DNA: correlation with nervous
system-specific carcinogenesis by ethylnitrosourea.   Proc.
 Natl Acad. Sci., USA, 71: 639-643.

GRAHAM, C.F.  (1977)  Teratocarcinoma cells and normal mouse
embryogenesis.  In: Sherman, M.J., ed.   Concepts in mammalian
 embryogenesis,USA, MIT Press, pp. 315-395.

GRANT, E.C. & MACKINTOSH, J.H.  (1963)  Comparison of the
social postures of some laboratory rodents.   Behaviour, 21:

GRANT, L.D.  (1976)  Research strategies for behavioral
teratology studies.   Environ. Health Perspect., 18: 85-94.

GRAUWILER, J. & LEIST, K.H.  (1977)  A schedule for
examination of postnatal development in rats.  In: Neubert,
D., Merker, H.-J., & Kwasigroch, T.E., ed.  Methods in prenatal
 toxicology,Stuttgart, Georg Thieme Publ., pp. 166-174.

GREENBERG, J.H.  (1982)  Detection of teratogens by
differentiating embryonic neural crest cells in culture;
evaluation as a screening system.   Teratog. Carcinog. Mutag.,
2: 319-323.

GREENGARD, O.  (1973)  Effects of hormones on development of
fetal enzymes.   Clin. Pharmacol., Ther., 14: 721-726.

GREENGARD, O.  (1975a)  Cortisol treatment of neonatal rats:
Effects on enzymes in kidney, liver and heart.   Biol. Neonat.,
27: 352-360.

GREENGARD, O.  (1975b)  Steroids and the maturation of rat
tissues.   J. steroid Biochem., 6: 639-672.

GROBSTEIN, C.  (1956)  Trans-filter induction of tubules in
mouse metanephrogenic mesenchyme.   Exp. Cell Res., 10: 424-440.

GUNBERG, D.L.  (1976)   In vitrodevelopment of
postimplantation rat embryos cultured on dialyzed rat serum.
 Teratology, 14: 65-69.

[Carcinogenic action of ethylnitroso urea in rabbits.]   Zbl.
 allg. Pathol. U. Pathol. Anat., 117: 461-471 (in German).

HAGSTROM, B.C., LONNING, S.  (1973)  The sea urchin egg as a
testing object in toxicology.   Acta pharmacol. toxicol., 32
(Suppl. 1): 49 pp.

HALES, B.F.  (1982)  Comparison of the mutagenicity and
teratogenicity of cyclophosphamide and its active metabolites,
4-hydroxycyclophosphamide, phosphoramide mustard, and
acrolein.   Cancer Res., 42: 3016-3021.

HARVEY, E.G.  (1956)   The American Arbacia and other sea
 urchins,Princeton University.

HASEMAN, J.K. & HOGAN, M.D.  (1975)  Selection of the
experimental unit in teratology studies.   Teratology, 12:

HASEMAN, J.K. & KUPPER, L.L.  (1979)  Analysis of dichotomous
response data from certain toxicological experiments.
 Biometrics, 35: 281-293.

Chondrogenesis and cell proliferation in limb bud cell
cultures treated with cytosine arabinoside and vitamin A.
 Exp. Cell Res., 112: 409-417.

HEINONEN, O.P.  (1977)   Birth defects and drugs in pregnancy,
Littleton, Mass., Publishing Sciences Group.

HELLER, C.G. & CLERMONT, Y.  (1964)  Kinetics of the germinal
epithelium in man.   Recent Progr. hormone Res., 20: 545-575.

Congenital malformation and maternal occupation in Finland:
multivariate analysis.   J. Epid. commun. Health, 35: 5-10.

R.E.  (1979a)  Epidemiologic aspects and factors related to
survival in 384 registry cases of clear cell adenocarcinoma of
the vagina and cervix.   Am. J. Obstet. Gynecol., 135: 876-886.

HERBST, A.L., SCULLY, R.E., & ROBBOY, S.J.  (1979b)  Prenatal
diethylstilbestrol exposure and human genital tract
abnormalities.  In:  Perinatal carcinogenesis, pp. 25-35 ( Natl
 Cancer Inst. Monogr.,Vol. 51).

HERKEN, R. & ANSCHUETZ, M.  (1981)  Differentiation of
embryonic tissues in whole-embryo cultures as compared to the
development  in vivo. in: Neubert, D. & Merker, H.-J., ed.
 Culture techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 19-35.

HETEM, S., KOLLAR, E.J., CUTLER, L.S., & YAGER, J.A.  (1975)
Effect of alpha,alpha-dipyridyl on the basement membrane of tooth
germs  in vitro.  J. dent. Res., 54: 783-787.

HICKS, S.P. & D'AMATO, C.J.  (1966)  Effects of ionizing
radiations on mammalian development.  In: Woollam, D.H.M.,
ed.   Advances in teratology,London, Lago Press, pp. 195-250.

WATANABE, N.  (1979)  Incidence of major chromosome
aberrations in 12 319 newborn infants in Tokyo.   Hum. Genet.,
46: 163-172.

HINEGARDUER, R.T.  (1967)  Echinoderms.  In: Wilt, F.H. &
Wessels, N.K., ed.   Methods in developmental biology,T.Y.
Cromwell, pp. 139-155.

HOGAN, M.D. & HOEL, D.G.  (1982)  Extrapolation to man.  In:
Hayes, A.W., ed.   Principles and methods of toxicology, New
York, Raven Press, pp. 71l-731.

HOLTZER, H., ABBOTT, J., & LASH, J.  (1958)  On the formation
of multinucleated myotubes.   Anat. Rec., 131: 567-568.

HORSTADIUM, S.  (1973)   Experimental embryology of
 echinoderms, Oxford, Clarendon Press.

HSU, Y.C.  (1979)   In vitro development of individually
cultured whole mouse embryos from blastocyst to early somite
stage.   Dev. Biol., 68: 453-461.

HSU, Y.C.  (1980)  Embryo growth and differentiation factors
in embryonic sera of mammals.   Dev. Biol., 76: 465-474.

HSU, Y.C.  (1981)  Regulation of mouse embryogenesis by
exogenous growth factors.  In: Neubert, D. & Merker, H.-J.,
ed.   Culture techniques. Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 83-96.

HUGHES, E.W. & PALMER, A.K.  (1980)   An assessment of
 pre-weaning development and behaviour in safety evaluation
 studies, Annual Meeting of the ETS, Münster.

HURLEY, L.S. & SHRADER, R.E.  (1972)  Congenital malformations
of the nervous system in zinc deficient rats.  In: Pfeiffer,
C.C., ed.   Neurobiology of the trace metals zinc and copper.
pp. 7-51 (Int. Rev. Neurol., Suppl., 1).

IARC  (1980)   Long-term and short-term screening assays for
 carcinogens: a critical appraisal,Lyons, International Agency
for Research on Cancer (IARC Monographs on the evaluation of
the carcinogenic risk of chemicals to humans, Suppl. 2).

IARC  (1982)   Chemicals, industrial processes, and industries
 associated wih cancer in humans,Lyons, International Agency
for Research on Cancer (IARC Monographs on the evaluation of
the carcinogenic risk of chemicals to humans, Suppl. 4).

ILLSLEY, N.P. & LAMARTINIERE, C.A.  (1979)  Prenatal
programming of hepatic monoamine oxidase by
5,5-diphenylhydantoin.   Biochem. Pharmacol., 28: 2585-2590.

IRLG  (1979)  Interagency Regulatory Liaison Group, Scientific
basis for identification of potential carcinogens and risk
estimation.   J. Natl Cancer Inst., 63: 243-268.

IVANKOVIC, S.  (1979)  Teratogenic and carcinogenic effects of
some chemicals during prenatal life in rats, Syrian golden
hamsters, and minipigs.  In:  Perinatal carcinogenesis.  Natl
 Cancer Inst. Monogr., 51: 103-115.

IVANKOVIC, S. & DRUCKREY, H.  (1968)  [Transplacental
induction of malignant tumours of the nervous system. I.
Ethylnitroso urea with BD IX-rats.]   Z. Krebsforsch, 71:
320-360 (in German).

JACOBSON, M.  (1978)   Developmental neurobiology, New York,
Plenum Press.

JELINEK, R.  (1979)  Embryotoxicity assay on morphogenetic
systems.  In: Benesova, O., Rychter, Z., & Jelinek, R., ed.
 Evaluation of embryotoxicity, mutagenicity and carcinogenicity
 risks in new drugs,Prague, Univerzita Karlova, pp. 195-205.

JELINEK, R. & PETERKA, M.  (1981)  Morphogenetic systems and
 in vitro techniques in teratology.  In: Neubert, D. & Merker,
H.-J., ed.  Culture techniques. Applicability for studies on
 prenatal differentiation and toxicity.  Berlin, New York,
Walter de Gruyter Publ., pp. 553-557.

JELINEK, R. & RYCHTER, Z.  (1979)  Morphogenetic systems and
the central phenomena of teratology.  In: Persand, T.-V.N.,
ed.   New trends in experimental teratology,Lancaster, MTP
Press, pp. 41-67.

JELINEK, R., RYCHTER, Z., & PETERKA, M.  (1976)  Cs. Authors
Certificate No. 191, 501, Praha.

A., MADSEN, S., & PORTER, J.B.  (1982)  Vaginal spermicides
and miscarriage seen primarily in the emergency room.
 Teratog. Carcinog. Mutag., 2: 205-210.

JOHANSON, I.B. & HALL, W.G.  (1979)  Appetitive learning in
1-day-old rat pups.   Science, 205: 419-421.

JOHN, D.M., CLARK, S.M., & CLARK, R.  (1982)  Temporal shifts
in the incidence of skeletal variations in the CD rat.  In:
 Abstracts of the 10th Meeting of the European Teratology
 Society, Norwich, England.

JOHNSON, E.M.  (1980)  A subvertebrate system for rapid
determination of potential teratogenic hazards.   J. environ.
 Pathol. Toxicol., 4: 153-156.

JOHNSON, E.M.  (1981)  Screening for teratogenic hazards:
Nature of the problems.   Ann. Rev. Pharmacol. Toxicol., 21:

(1982)  The  hydra attentuatasystem for detection of
teratogenic hazards.   Teratog. Carcinog. Mutag. (in press).

JOHNSTON, M.C. & PRATT, R.M.  (1975)  A developmental approach
to teratology.  In: Berry, C.L. & Poswillo, D.E., ed.
 Teratology, trends and applications, New York, Springer-
Verlag, pp. 2-16.

JONES, B.J. & ROBERTS, D.J.  (1968)  The quantitative
measurement of motor inco-ordination in naive mice using an
accelerating rotarod.   J. Pharm. Pharmacol., 20: 302-304.

JOSSO, N.  (1974)  Effect of PGA2 and PGB1 on the rat foetal
müllerian duct < in vitro> absence of relationship to
testicular anti-müllerian hormone.   Biomedicine, 21: 225-229.

JOST, A. & BERGERARD, Y.  (1949)  Culture  in vitro d'ébauches
du tractus génital du foetus de rat.   C.R. Soc. Biol., 143:

(1976)  Embryonal neoplasms in the opossum: a new model for
solid tumors of infancy and childhood.   Science, 193: 328-332.

JUSKO, W.J.  (1972)  Pharmacodynamic principles in chemical
teratology: dose effect relationships.   J. Pharmacol. exp.
 Therap., 183: 469-480.

KALLAI, L.  (1967)  Pharmaco-ecology and some nutritional
paradoxes.  In: M.L. Conalty, ed.   Husbandry of laboratory
 animals,London, Academic Press.

KALOJANOVA, F.  (1982)  Evaluation of pesticide toxicity for
sanitary registration.  In:  Toxicology of pesticides,
 Proceedings of a seminar, Sofia, Bulgaria, 31 August-12
 September, 1981.  Interim document 9,pp. 265-274.

KALTER, H.  (1965)  Interplay of intrinsic and extrinsic
factors.  In: Wilson, J.G. & Warkany, J., ed.   Teratology
 Principals & Techniques, Chicago, University of Chicago Press,
279 pp.

KALTER, H. & WARKANY, J.  (1983)  Congenital malformations.
Etiology factors and their role in prevention. Part I & II.
 New Engl. J. Med., 308: 424-431, 491-497.

LEINIKKI, P.  (1975)  Rubella cataract  in vitro: sensitive
period of the developing human lens.   J. exp. Med., 141:

KAUFFMAN, S.L.  (1976)  Susceptibility of fetal lung to
transplacental 1-ethyl-1-nitrosourea: its relation to
epithelial proliferation.   J. Natl Cancer Inst., 57: 821-825.

KAVLOCK, R.J., MEYER, E., & GRABOWSKI, C.T.  (1980)  Studies
on the developmental toxicity to ozone: postnatal effects.
 Toxicol. Lett., 5: 3-9.

KERLEY, M.A. & KOLLAR, E.J.  (1978)   In vitro inhibition of
mouse dental development by tetracycline.   J. exp. Zool., 203:

KHERA, K.S.  (1976)  Significance of metabolic patterns in
teratogenic testing for food safety.   Clin. Toxicol., 9:

KHERA, K.S.  (1981)  Common fetal aberrations and their
teratologic significance: a review.   Fundam. appl. Toxicol.,
1: 13-18.

KHERA, K.S.  (1983)  Materno-fetal toxicity - a possible cause
of fetal malformations in the mouse.   Teratology, 27: 56A-57A.

KHERA, K.S.  (in press)  Adverse effects in humans and animals
of prenatal exposure to selected therapeutic drugs and
estimation of embryo-fetal sensitivity of animals for human
risk assessment: a review.   Issues and Reviews in Teratology,

KHERA, K.S. & RUDDICK, J.A.  (1973)  Polychlorodibenzo-
 p-dioxins: perinatal effects and the dominant lethal test in
Wistar rats.  In: Blaire, E.H., ed.  Chlorodioxins: origin and
 fate.  Adv. Chem. Ser., 120: 70-85.

KIMMEL, C.A.  (1976)  Behavioural teratology: overview.
 Environ. Health Perspect., 18: 73.

KIMMEL, C.A. & BUELKE-SAM, J., ed.  (1981)   Developmental
 toxicology, New York, Raven Press.

(1974)  Teratogenicity of hexachlorophene in rats.   Arch.
 environ. Health, 28: 43-48.

KIMMEL, G.L., SMITH, K., KOCHHAR, D.M., & PRATT, R.M.  (1982)
Overview of  in vitro teratogenicity testing; aspects of
validation and application to screening.   Teratog. Carcinog.
Mutag., 2: 221-229.

KLEE-TRIESCHMANN, U. & NEUBERT, D.  (1981)  Preliminary
communication on the feasibility of culturing whole embryos in
a chemically defined medium.  In: Neubert, D. & Merker, H.-J.,
ed.   Culture techniques.  Applicability for studies on
 prenatal differentiation and toxicity, Berlin, New York,
Walter de Gruyter Publ., pp. 97-100.

BUECHELER, J., & COOPER, H.K.  (1979a)  DNA repair as
regulatory factor in the organotropy of alkylating
carcinogens.   Arch. Toxicol., Suppl., 2: 253-261.

DIESSNER, H.  (1979b)  Mechanism of perinatal tumour induction
by neurooncogenic alkylnitrosoureas and dialkylarytriazenes.
Perinatal Carcinogenesis,  NCI Monograph, 51: 227.

CLAPPER, M.L.  (1981)  Evaluation of serum teratogenic
activity using rat embryo cultures.  In: Neubert, D. & Merker,
H.-J., ed.   Culture techniques.  Applicability for studies on
 prenatal differentiation and toxicity, Berlin, New York,
Walter de Gruyter Publ., pp. 67-81.

(1980)  The use of cultured rat embryos to evaluate the
teratogenic activity of serum: cadmium and cyclophosphamide.
 Teratology, 21: 199-208.

KOCH, W.  (1974)  The effects of azetidine-2-carboxylic acid
on cultured incisors of embryonic mice.   Anat. Rec., 178:

KOCHHAR, D.M.  (1970)  Effect of azetidine-2-carboxylic acid,
a proline analog, on chondrogenesis in cultured limb buds.
In: Bass, R., Beck, F., Merker, H.-J., Neubert, D., &
Randhahn, B., ed.   Metabolic pathways in mammalian embryos
 during organogenesis and its modification by drugs, Berlin,
Free University Press, pp. 475-482.

KOCHHAR, D.M.  (1975a)  The use of  in vitro procedures in
teratology.   Teratology, 1l: 273-288.

KOCHHAR, D.M.  (1975b)  Assessment of teratogenic response in
cultured postimplantation mouse embryos: effects of
hydroxyurea.  In: Neubert, D. & Merker, H.-J., ed.   New
 approaches to the evaluation of abnormal embryonic
 development,Stuttgart, Georg Thieme Publ., pp. 250-277.

KOCHHAR, D.M.  (1976)  Elucidation of mechanisms underlying
experimental mammalian teratogenesis through a combination of
whole embryo, organ culture, and cell culture methods.  In:
Ebert, J.D. & Marois, M., ed.   Tests of teratogenicity in
 vitro, amsterdam, North-Holland Publ., pp. 485-494.

[Liver tumours in mice induced by prenatal and postnatal
exposure to orthoaminoasotoluene.]  Bull. exp. Biol., 2:
193-202 (in Russian).

KOLLAR, E.J.  (1973)  Effect of alpha,alpha-dipyridyl on organ
cultures of embryonic mouse tooth germs.   J. dent. Res., 52:

KOROLEV, A.A. & AGAREVA, E.N.  (1983)  [Sanitary and
toxicological assessment of certain asodyes and their
transformation products.]   Hig. i. Sanit., 1: 15-18 (in

GIELEN, J.E.  (1981)  Multiplicity of cytochrome P-450 in
primary fetal hepatocytes in culture.   Eur. J. Biochem.,116:

KREWSKI, D. & VAN RYZIN, J.  (1981)  Dose response models for
quantal response toxicity data.  In: Scorgo, M., Dawson, D.,
Rao, J.N.K., & Salch, E., ed.   Statistics and related topics,
New York, North-Holland, pp. 201-231.

KROWKE, R. & NEUBERT, D.  (1977)  Embryonic intermediary
metabolism under normal and pathological conditions.  In:
Wilson, J.G. & Fraser, F.C., ed.   Handbook of Teratology, New
York, London, Plenum Press, Vol. II, pp. 117-151.

KUCERA, J.  (1961a)  [Registration of congenital
malformations in the CSR.]   Prakt. Lek. (Praha), 41: 487-490
(in Czech).

KUCERA, J.  (1961b)  The frequency of congenital
malformations in the CSR, and factors influencing it.   I. Cas.
 Lek. Ces., 100: 1609-1614.

KUCERA, J.  (1968a)  [Syndrome of caudal regression and its
significance in the study of etiology and/or prevention of
congenital malformations.]   Cs. Pediat., 23: 17-25 (in Czech).

KUCERA, J.  (1968b)  Exposure to fat solvents: a possible
cause of sacral agenesis in man,   J. Pediat., 72: 857-859.

KUCERA, J.  (1971a)  Movement of the frequencies of
congenital malformations in time and space.   Soc. Biol., 18:

KUCERA, J.  (1971b)  Patterns of congenital anomalies in the
offspring of women exposed to different drugs and/or chemicals
during pregnancy.   Teratology, 4: 492.

KUCERA, J.  (1977)  Relation between population genetics and
population teratology.  In: Szabo, L. & Papp. S., ed.
 Proceedings of Medical Genetics, Budapest, Amsterdam, Akad.
Kiado, pp. 449-473.

Z.S., & HILKEVIC, L.H.  (1982)  [Some morpho-functional data
about the follicular system of the human ovary.]  In: [Current
aspects of studies of women's reproductive functions.] Moscow,
pp. 41-46 (in Russian).

Z.S., & HIULKEVIC, L.B.  (in press)  [Oxitetracillinum and
17-B-estradiol inhibition of oogonia proliferation in human
fetal ovary in culture.]   Obstetrics and Gynaecology(in

transplacental blastomogenic action zineb on mice offspring.]
 Vop. Onkol., 1: 49-50 (in Russian).

Development of limb buds in organ culture: examination of
hydroxyurea enhancement of bromodeoxyuridine toxicity using
image analysis.  In: Neubert, D. & Merker, H.-J., ed.   Culture
 techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 237-253.

LAHTI, A. & SAXEN, L.  (1967)  Effect of hydrocortisone on the
closure of palatal shelves  in vivo and  in vitro.  Nature
 (Lond.), 216: 1217-1218.

LAHTI, A., ANTILA, E., & SAXEN, L.  (1972)  The effect of
hydrocortisone on the closure of the palatal shelves in two
inbred strains of mice  in vivo and  in vitro.  Teratology, 6:

LALLIER, R.  (1965)  Effets du 5-fluoro-uracile et de la
6-méthylpurine sur le développement de l'oeuf de l'oursin
Paracentrotus lividus.   J. Embryol. exp. Morphol., 14: 181-189.

LAMARTINIERE, C.A. & LUCIER, G.W.  (1981)  Role of xenobiotic
conjugation reactions in organ-specific toxicity (carcino-
genesis).   Proceedings of Organ and Species Specificity in
 Chemical Carcinogenesis Symposium, 2-4 March 1981.

LAMBSON, R.O.  (1966)  An electron microscopic visualization
of transport across rat visceral yolk sac.   Am. J. Anat., 118:

LANGMAN, J.  (1969)   Medical embryology,2nd edition,
Baltimore, Williams and Wilkins.

LANGMAN, J., WEBSTER, W., & RODIER, P. (1975)  Morphological
and behavioural abnormalities caused by insults to the CNS in
the perinatal period.  In: Berry, C.L. & Poswillo, D.E., ed.
 Teratology, trends and applications, New York, Springer-
Verlag, pp. 182-200.

LAPOINTE, G. & NOSAL, G.  (1979)  A rat model of
neurobehavioral development.   Experientia, Basel, 35: 205-207.

LAQUEUR, G.L. & SPATZ, M.  (1973)  Transplacental induction of
tumours and malformations in rats with cycasin and
methylazoxymethanol.  In: Tomatis, L., Mohr, U., & Davis, W.,
ed.   Transplacental carcinogenesis,Lyons, International
Agency for Research on Cancer, pp. 59-64 (IARC Scientific
Publications No. 4).

LAROCHE, M.J.  (1965)  Influence of environment on drug
activity in laboratory animals.   Food Cosmet. Toxicol., 3:

LASNITZKI, I. & MIZUNO, T.  (1981)  Interaction of epithelium
and mesenchyme in the induction of foetal rat and mouse
prostate glands by androgens in organ culture.  In: Neubert,
D. & Merker, H.-J., ed.   Culture techniques.  Applicability
 for studies on prenatal differentiation and toxicity, Berlin,
New York, Walter de Gruyter Publ., pp. 359-369.

LEBLOND, C.P. & CLERMONT, Y.  (1952)  Definition of the stages
of the cycle of the seminiferous epithelium in the rat.   Ann.
 NY Acad. Sci., 55: 548-573.

LEEMING, N.M., COZENS, D.D., & PALMER, A.K.  (1982)  Points
for consideration in the design of multigeneration studies.
In:  Abstracts of the 10th Meeting of the European Teratology
 Society, Norwich.

LEONOV, B.V.  (1979)   [Sexual hormones and hormone-like
 substances in early embryogenesis.]Moscow, Medicina, p. 160
(in Russian).

Mammalian limb buds differentiating  in vitro as a test system
for the evaluation of embryotoxic effects.  In: Neubert, D. &
Merker, H.-J., ed.   New approaches to the evaluation of
 abnormal embryonic development,Stuttgart, Georg Thieme Publ.,
pp. 99-113.

LESSMOELLMANN, U., HINZ, N., & NEUBERT, D.  (1976)  In vitro
system for toxicological studies on the development of
mammalian limb buds in a chemically defined medium.   Arch.
 Toxicol.,36: 169-176.

LEUNG, T.S. & BULKLEY, R.V.  (1979)  Effects of petroleum
hydrocarbons on length of incubation and hatching success in
the Japanese medaka.   Bull. environ. Contam. Toxicol., 23:

LEVITT, D., HO, P.L., & DORFMAN, A.  (1975)  Effect of
5-bromodeoxyuridine on ultrastructure of developing limb bud
cells  in vitro.  Dev. Biol., 43: 75-90.

LEWIN, B.  (1980)   Gene expression, Vol. 2.  Eukaryotic
 chromosome,John Wiley & Sons.

(1978)  Inhibition of limb chondrogenesis  in vitro by vitamin
A: alterations in cell surface characteristics.   Dev. Biol.,
64: 31-47.

LEWIS, P.D. & PATEL, A.J.  (1980)  Psychtropic drugs and brain
development: effects on cell acquisition.  In: di Benedetta,
C., ed.   Multidisciplinary approach to brain development,
Amsterdam, Elsevier, North Holland, pp. 509-517.

LIDDIARD, C., BRENDEL, K., & NAU, H.  (1978)  Drug metabolism
in cultures of isolated hepatocytes of the human fetus, the
newborn pig and the adult rat.  In: Neubert, D., Merker,
H.-J., Nau, H., Langman, J., ed.   Role of pharmacokinetics in
 prenatal and perinatal toxicology,Stuttgart, Georg Thieme
Publ., pp. 91-108.

LIKHACHEV, A.J.  (1971)  [Blastomogenesis induced by the
combined transplacental and postnatal exposure to
 N-nitrosodiethyl-amine (NDEA) in mice].   Vopr. Onkol., 17(4):
64-69 (in Russian).

NAPALKOV, N.P., & TOMATIS, L.  (1983)  Persistence of
methylated purines in the DNA of various rat fetal and
maternal tissues and carcinogenesis in the offspring following
a single transplacental dose of N-methyl-N-nitrosourea.   Int.
 J. Cancer, 31: 779-784.

LJUNGBERG, T. & UNGERSTEDT, U.  (1976)  Automatic registration
of behaviour related to dopamine and noradrenaline
transmission.   Eur. J. Pharmacol., 36: 181-188.

Embryonic development and blastocyst implantation following in
vitro fertilization and embryo transfer.   Fertil. Steril., 38:

LUCIER, G.W. & MCDANIEL, O.S.  (1979)  Developmental
toxicology of the halogenated aromatics: Effects on enzyme
development.   Ann. NY Acad. Sci., 77: 449-457.

(1975)  Postnatal stimulation of hepatic microsomal enzymes
following administration of TCDD to pregnant rats.   Chem.
 Biol. Interact., 1l: 15-26.

LUNDBORG, P. & ENGEL, J.  (1978)  Neurochemical brain changes
associated with behavioural disturbances after early treatment
with psychotropic drugs.  In: Vernadakis, ed.   Maturation of
 neurotransmission, Basel, S. Karger, pp. 226-235.

LUSCHER, M.  (1946)  [Inhibition of regeneration of the tail
of xenopus larvae by colchicine and analysis of its action
during development.]   Helv. Physiol. Pharmacol. Acta, 4: 465
(in German).

LUSCHER, M.  (1955)  [The regeneration in Zoology.]  In:
 Handbuch der allgemeinen Pathologie, Berlin, Springer-Verlag,
Vol. 6, Part I, pp. 406-440 (in German).

LUTWAK-MANN, C.  (1973)  Drugs and the blastocyst.  In:
Boréus, L.O., ed.   Fetal pharmacology, New York, Raven Press
Publ., pp. 419-442.

Infertility in mice exposed prenatally to benzo-[ a]-pyrene.
(BP)  Teratology, 19: 37A.

Infertility in mice following prenatal exposure to
9,10-dimethyl-1,2-benzanthracene (DMBA).  In:  Proceedings of
 the 12th Annual Meeting of the Society for the Study of
 Reproduction,Quebec, p. 30A.

MAEKAWA, A. & ODASHIMA, S.  (1975)  Induction of tumours of
the nervous system in the ACI/N rat with 1-butyl-1-nitrosourea
administered transplacentally, neonatally, or via maternal
milk.   Gann. Jpn. J. Cancer Res., 66(2): 175-183.

Transplacental carcinogenesis and chemical determination of
1-butyl-1-nitrosourea in stomach content after simultaneous
oral administration of l-butylurea and sodium nitrite to ACI/N
rats.   Gann. Jpn. J. Cancer Res., 68(l): 81-87.

Transplacental and neonatal carcinogenesis by 1-butyl-1-
nitrosourethan in the ACI/N rat.   Gann. Jpn. J. Cancer Res.,
71: 81l-815.

MANTEL, N. & BRYAN, W.R.  (1961)  "Safety"-testing of
carcinogenic agents.   J. Natl Cancer Inst., 27: 455-470.

MANTEL, N. & SCHNEIDERMAN, M.A.  (1975)  Estimating "safe"
levels, a hazardous undertaking.   Cancer Res., 35: 1379-1386.

MARSDEN, C.A.  (1979)  The use of doppler shift radar to
monitor physiological and drug induced activity patterns in
the rat.   Pharmacol. Biochem. Behav., 10: 631-635.

MARTIN, G.R.  (1978)  Advantages and limitations of
teratocarcinoma stem cells as models of development.  In:
Hohnsonn, M.M., ed.   Development in mammals, elsevier, North
Holland, Biomedical Press, Amsterdam, pp. 225-261.

MARTINOVITCH, P.N.  (1938)  The development  in vitro of the
mammalian gonad.  Ovary and ovogenesis.   Proc. Roy. Soc.
 Lond., 125: 232-249.

MATEYKO, J.M.  (1967)  Developmental modification in Arbacia
punctulate by various metabolic substances.   Biol. Bull.,
13(l): 184-228.

MATSUMOTO, N. & SPINDLE, A.  (1982)  Sensitivity of early
mouse embryos to methylmercury toxicity.   Toxicol. appl.
 Pharmacol., 64: 108-117.

(1968)  Culture of two cell rabbit eggs in chemically defined
media.   Exp. Cell Res., 52: 293-300.

MCLACHLAN, J.A. & DIXON, R.L.  (1973)  Reduced fertility in
female mice exposed prenatally to procarbazine.   Fed. Proc.,
32: 2982.

Reproductive tract lesions in male mice exposed prenatally to
diethylstilbestrol.   Science, 190: 991-992.

SUZUKI, Y.  (1981)  Transplacental toxicology: prenatal
factors influencing postnatal fetility.  In: Kimmel, C.A. &
Buelke-Sam, J., ed.   Developmental Toxicology, New York, Raven
Press, pp. 213-232.

MCLAREN, A. & BIGGERS, J.D.  (1958)  Successful development
and birth of mice cultivated  in vitro as early embryos.
 Nature (Lond.): 182: 877.

(1976)  Ontogenetic development of locomotor activity and rate
of tyrosine hydroxylation.   Psychopharmacology, 49: l19-123.

MELLO, N.K.  (1975)  Behavioral toxicology: a developing
discipline.   Fed. Proc., 34: 1832-1834.

MERKER, H.-J.  (1975)  Significance of the limb bud culture
system for investigations of teratogenic mechanisms.  In:
Neubert, D. & Merker, H.-J., ed.   New approaches to the
 evaluation of abnormal embryonic development,Stuttgart, Georg
Thieme Publ., pp. 161-199.

MERKER, H.-J. & GUENTHER, TH.  (1979)  The influence of
insulin, cAMP and the calcium ionophore X 537 A on the growth
of the cartilage anlagen of limb buds  in vitro.  Experientia,
 Basel:35: 1307-1308.

MERKER, H.-J., LILJA, S., & GUNTHER, TH.  (1978a)  Formation
of an atypical collagen and cartilage pattern in limb bud
cultures by highly sulfated GAG.   Virchow's Arch., 380: 11-30.

MERKER, H.-J., GUENTHER, TH., & KRUEGER, U.  (1978b)  Effect
of 4-methylumbelliferyl-phi-D-xylopyranoside on the morphology
of embryonic cartilage in limb bud cultures.   Teratology, 18:

& EBEL, H.  (1980a)  Simulation of steps of limb
skeletogenesis  in vitro. in: Merker, H.-J., Nau, H., &
Neubert, D., ed.   Teratology of the limbs, Berlin, Walter de
Gruyter Publ., pp. 137-151.

Differentiation of isolated blastemal cells from limb buds
into chondroblasts.  In: Richards, R.J. & Rajan, K.T., ed.
 Tissue culture in medical research (II), oxford, New York,
Pergamon Press, pp. 31-39.

MERKER, H.-J., ZIMMERMANN, B., & RISO, J.  (1981)  The
embryonic lung as  in vitro model for testing teratogenic
substances.  In: Neubert, D. & Merker, H.-J., ed.   Culture
 techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 30l-317.

influence of orally administered formaldehyde on embryogenesis
and enzyme activity.]   Z. Ges. Hyg., 29(2): 66-68 (in German).

Studies of lead encephalopathy in the developing rat.  In:
 Proceedings of the CEC-EPA-WHO International Symposium -
 Recent advances in the assessment of the health effects of
 environmental pollution, Vol. II, Paris, June, CEC Directorate
General, Scientific and Technical Information and Information

GRUDZINSKAS, J.G., & SYKES, A.  (1980)  Fetal loss after
implantation. A prospective study.  Lancet, I: 554-556.

MILLER, R.W.  (1977)  Relationship between human teratogens
and carcinogens.   J. Natl Cancer Inst., 58: 471-474.

MILUNSKY, A., GRAEF, J.W., & GAYNOR, M.F.  (1968)
Methotrexate-induced congenital malformations.   J. Pediatr.,
72: 790-795.

MIRKIN, B.L. & SINGH, S.  (1976)  Placental transfer of
pharmacologically active molecules.  In: Mirkin, B.L., ed.
 Perinatal pharmacology and therapeutics, New York, Academic
Press, pp. l-69.

MISANIN, J.R., NAGY, Z.M., KEISER, E.F., & BOWEN, W.  (1971)
Emergence of long-term memory in the neonatal rat.   J. comp.
 Physiol. Psychol., 77: 188-199.

MITCHELL, C.L.  (1978)  Target organ toxicity: nervous
system.   Environ. Health Perspect., 26: 3-4.

(1975)  Transplacental effects of diethylnitrosamine in Syrian
hamsters as related to different days of administration during
pregnancy.   J. Natl Cancer Inst., 55: 681-683.

MOSCONA, A. & MOSCONA, H.  (1952)  The dissociation and
aggregation of cells from organ rudiments of the early chick
embryo.   J. Anat., 86: 287-301.

MURPHY, M.L.  (1965)  Factors influencing teratogenic response
to drugs.  In: Wilson, J.G. & Warkany, J., ed.   Teratology,
Chicago, London, The University of Chicago Press, pp. 145-184.

NAPALKOV, N.P.  (1973)  Some general considerations on the
problem of transplacental carcinogenesis.  In: Tomatis, L.,
Mohr, U., & Davis, W., ed.   Transplacental carcinogenesis,
Lyons, International Agency for Research on Cancer, pp. 1-13
(IARC Scientific Publications No. 4).

NAPALKOV, N.P. & ALEXANDROV, V.A.  (1968)  On the effects of
blastomogenic substances on the organism during embryo-
genesis.   Z. Krebsforsch., 71: 32-50.

NAPALKOV, N.P. & ANISIMOV, V.N.  (1979)  Transplacental effect
of diethylstilbestrol in female rats.   Cancer Lett., 6:

NAS  (1977)   Principles and procedures for evaluating the
 toxicology of household substances.  7. Reproduction and
 teratogenicity tests,Washington, DC, US National Academy of
Sciences, pp. 99-110.

NAS  (1977)  National Academy of Sciences,  Drinking water and
 health,Washington, DC, Printing and Publishing Office,
National Academy of Sciences, pp. 24-54 (prepared for the US
Environmental Protection Agency under contract no. 68-0l-3139).

NAS  (1983)  National Academy of Sciences, Committee on the
institutional means or assessment of risk to public health,
commission on life sciences, National Research Council,  Risk
 assessment in the Federal Government: Managing the process,
Washington, DC, National Academy Press, ISBN D-309-03349-7.

(1977)  Benzodiazepine metabolism studies in organ and
isolated cell cultures of human fetal liver by a gas
chromatography mass spectrometer computer system.  In:
Eggstein, M. & Liebich, H.M., ed.   Mass spectrometry and
 combined techniques in medicine, clinical chemistry and
 clinical biochemistry,Tübingen University Press, pp. 346-358.

(1981)  A new model for embryotoxicity testing: teratogenicity
and pharmacokinetics of valproic acid following constant-rate
administration in the mouse using human therapeutic drug and
metabolite concentrations.   Life Sci., 29: 2803-2814.

NEUBERT, D.  (1980a)  [Extrapolations from animal studies.
Prospective  in vitro models as substitute for long-term
tests.]  In: Schnieders, B. & Grosdanoff, P., ed.   Zur
 Problematik von chronischen Toxizitätsprüfungen, aMI-Berichte
1/1980, Berlin, Dietrich Reimer Verlag, pp. 277-287 (in

NEUBERT, D.  (1980b)  Teratogenicity: any relationship to
carcinogenicity?  In: Montesano, R., Bartsch, H., Tomatis, L.,
& Davis, W., ed.  Molecular and cellular aspects of carcinogen
 screening tests,Lyons, International Agency for Research on
Cancer, pp. 169-178 (IARC Scientific Publications No. 27).

NEUBERT, D.  (1982)  The use of culture techniques in studies
on prenatal toxicity.   Pharmac. Ther., 18: 397-434.

NEUBERT, D.  (1983)  [Can mutagenicity tests or other  in vitro
methods replace carcinogenicity or other long-term  in vivo
studies?]  In: Grosdanoff, P., Schneiders, B., & UEberla, K.,
ed.  Arzneimittel Sicherheit,MMV, bga Berichte 1/1983, pp.
83-88 (in German).

NEUBERT, D. & BARRACH, H.J.  (1977b)  Techniques applicable to
study morphogenetic differentiation of limb buds in organ
culture.  In: Neubert, D., Merker, H.-J., & Kwasigroch, T.E.,
ed.   Methods in prenatal toxicology.  Evaluation of
 embryotoxic effects in experimental animals,Stuttgart, Georg
Thieme Publ., pp. 241-251.

NEUBERT, D. & BLUTH, U.  (1981a)  Feasibility of storing
embryonic tissues for subsequent use in organ culture.  In:
Neubert, D. & Merker, H.-J., ed.   Culture techniques.
 Applicability for studies on prenatal differentiation and
 toxicity, Berlin, New York, Walter de Gruyter Publ., pp.

NEUBERT, D. & BLUTH, U.  (1981b)  Limb bud organ cultures from
mouse embryos after apparent induction of monooxygenases  in
 utero. effects of cyclophosphamide,, dimethylnitrosamine and
some thalidomide derivatives.  In: Neubert, D. & Merker,
H.-J., ed.   Culture techniques.  Applicability for studies on
 prenatal differentiation and toxicity, Berlin, New York,
Walter de Gruyter Publ., pp. 175-195.

NEUBERT, D. & CHAHOUD, I.  (in press)  Significance of species
differences in pre- and perinatal toxicity.   Z. Histochem.

NEUBERT, D. & DILLMANN, I.  (1972)  Embryotoxic effects in
mice treated with 2,4,5-trichlorophenoxyacetic acid and
2,3,7,8-tetrachlorodibenzo-p-dioxin.   Naunyn-Schmiedebergs
 Arch. Pharmacol., 272: 243-264.

NEUBERT, D. & MERKER, H.-J., ed.  (1981)   Culture techniques.
 Applicability for studies on prenatal differentiation and
 toxicity, Berlin, New York, Walter de Gruyter Publ., pp. 1-607.

NEUBERT, D. & RAUTENBERG, M.  (1976)  Activity of nuclear
DNA-dependent RNA polymerases in mouse limb buds
differentiating  in vivo or in organ culture.   Hoppe-Seyler's
 Z. Physiol. Chem., 357: 1623-1635.

H.-J.  (1971)  Studies on the problem of "aerobic glycolysis"
occuring in mammalian embryos.  Naunyn-Schmiedebergs Arch.
 Pharmak, 268: 235-241.

(1973)  A survey of the embryotoxic effects of TCDD in
mammalian species.   Environ. Health Perspect., 5: 67-79.

NEUBERT, D., MERKER, H.-J., & TAPKEN, S.  (1974)  Comparative
studies on the prenatal development of mouse extremities  in
 vivo and in organ culture.   Naunyn-Schmiedeberg's Arch.
 Pharmacol., 286: 251-270.

U.  (1976)  Biochemical and teratological aspects of mammalian
limb bud development  in vitro. in: Ebert, J.D. & Marois, M.,
ed.   Tests of teratogenicity in vitro, Amsterdam, New York,
Oxford, North-Holland Publ. Comp., pp. 335-365.

FUCHS, G.  (1977a)  Interference of 6-mercaptopurine riboside,
6-methylmercaptopurine riboside and azathioprine with the
morphogenetic differentiation of mouse extremities  in vivo and
in organ culture.   Naunyn-Schmiedeberg's Arch. Pharmacol.,
298: 93-105.

NEUBERT, D., MERKER, H.-J., & KWASIGROCH, T.E., ed.  (1977b)
 Methods in prenatal toxicology,Stuttgart, Georg Thieme Publ.

(1978a)   Role of pharmacokinetics in prenatal and perinatal
 toxicology,Stuttgart, Georg Thieme Publ.

NEUBERT, D., TAPKEN, S., & BAUMANN, I.  (1978b)  Influence of
potential thalidomide metabolites and hydrolysis products on
limb development in organ culture and on the activity of
proline hydroxylase.  Further data on our hypothesis on the
thalidomide embryopathy.  In: Neubert, D., Merker, H.-J., Nau,
H., & Langman, J., ed.   Role of pharmacokinetics in prenatal
 and perinatal toxicology,Stuttgart, Georg Thieme Publ., pp.

NEUBERT, D., BARRACH, H.J., & MERKER, H.-J.  (1980)
Drug-induced damage to the embryo or fetus.  In: Grundmann,
E., ed.   Current topics in pathology, Berlin, Heidelberg, New
York, Springer-Verlag, pp. 242-324.

NEUKOMM, S.  (1969)  Action of a carcinogenic tar on the
regeneration of the tail of the crested newt.  A new possible
test for evaluation of substances with teratogenic action.
In: Bertelli, A. & Donati, L., ed.   Teratology, amsterdam,
Excerpta Medica Found. pp. 11-22.

NEW, D.A.T.  (1955)  A new technique for the cultivation of
the chick embryo  in vitro.  J. Embryol. exp. Morphol., 3:

NEW, D.A.T.  (1966a)  Development of rat embryos cultured in
blood sera.   J. Reprod. Fertil., 12: 509-524.

NEW, D.A.T.  (1966b)   The culture of vertebrate embryos,
London, New York, Logos Press - Academic Press, pp. 1-245.

NEW, D.A.T.  (1976)  Techniques for assessment of teratologic
effects: embryo culture.   Environ. Health Perspect., 18:

NEW, D.A.T.  (1978)  Whole embryo culture and the study of
mammalian embryos during organogenesis.   Biol. Rev., 53:

NEW, D.A.T., COPPOLA, P.T., & TERRY, S.  (1973)  Culture of
explanted rat embryos in rotating tubes.   J. Reprod. Fertil.,
35: 135-138.

NEW, D.A.T., COPPOLA, P.T., & COCKROFT, D.L.  (1976a)
Comparison of growth  in vitro and  in vivo of post-implantation
rat embryos.   J. Embryol. exp. Morphol., 36: 133-144.

NEW, D.A.T., COPPOLA, P.T., & COCKROFT, D.L.  (1976b)
Improved development of head-fold rat embryos in culture
resulting from low oxygen and modifications of the culture
serum.   J. Reprod. Fertil., 48: 219-222.

NICOLOV, I.G. & CHERNOZEMSKY, I.N.  (1979)  Tumours and
hyperplastic lesions in Syrian hamsters following
transplacental and neonatal treatment with cigarette smoke
condensate.   J. Cancer Res. clin. Oncol., 94: 249-256.

NISHIMURA, H.  (1971)  Incidence of malformations in
abortions.  In: Fraser, F.C. & McKusick, V.A., ed.   Congenital
 malformations.  Proceedings of the Third International
 Congress, amsterdam, Princeton, Exerpta Medica, pp. 275-283.

NOMURA, T.  (1974)  An analysis of the changing urethan
response of the developing mouse embryo in relation to
mortality, malformation, and neoplasm.   Cancer Res., 34:

NOMURA, T.  (1982)  Parental exposure to X rays and chemicals
induces heritable  tumours and anomalies in mice.   Nature
 (Lond.), 296: 575-577.

NORTON, S., MULLENIX, P., & CULVER, B.  (1976)  Comparison of
the structure of hyperactive behaviour in rats after brain
damage from X-irradiation, carbon monoxide and pallidal
lesions.   Brain Res., 116: 49-67.

OAKBERG, E.F.  (1956)  Duration of spermatogenesis in the
mouse and timing of stages of the cycle of the seminiferous
epithelium.   Amer. J. Anat., 99: 507-516.

OECD  (1981)   Teratogenicity,Paris, Organization for Economic
Co-operation and Development (OECD guidelines for testing of
chemicals No. 414).

ONUMA, H., MAURER, R.R., & FOOTE, R.H.  (1968)   In vitro
culture of rabbit ova from early cleavage stages to the
blastocyst stage.   J. Reprod. Fertil., 16: 491-493.

PAGE, E.W., VILLEE, C.A., & VILLEE, D.B.  (1981)   Human
 Reproduction,London, Saunders.

PALMER, A.K.  (1967)  The relationship between screening tests
for drug safety and other teratological investigations.  In:
 Proceedings of a Symposium organized by the Italian Society of
 Experimental Teratology, Como, Italy (Excerpta Medica
International Congress Series No. 173, Teratology).

PALMER, A.K.  (1969)  The concept of the uniform animal
relative to teratogenicity.   Carworth Europe Collected Papers,
3: 10l-113.

PALMER, A.K.  (1974)  Statistical analysis and choice of
sample units.   Teratology, 10: 30l-302.

PALMER, A.K.  (1976)  Assessment of current test procedures.
 Environ. Health Persp., 18: 97-104.

PALMER, A.K.  (1977a)  Reproductive toxicology studies and
their evaluation.  In: Ballantyne, B., ed.   Current approaches
 in toxicology, Bristol, Publ. J. Wright, pp. 54-67.

PALMER, A.K.  (1977b)  Incidence of sporadic malformations,
anomalies and variations in random bred laboratory animals.
In: Neubert, D., Merker, H.-J., & Kwasigroch, T.E., ed.
 Methods in prenatal toxicology,Stuttgart, Georg Thieme Publ.,
pp. 52-71.

PALMER, A.K.  (1978)  The design of subprimate animal
studies.  In: Wilson, J.G. & Fraser, F.C., ed.   Handbook of
 teratology, New York, London, Plenum Press, Vol. 4, pp.

PALMER, A.K.  (1981)  Regulatory requirements for reproductive
toxicology: theory and practice. In: Kimmel, C.A. & Buelke-
Sam, J., ed.   Developmental toxicology, New York, Raven Press,
pp. 259-287.

PATEL, A.J., BENDEK, G., BALAZS, R., & LEWIS, P.D.  (1977)
Effect of reserpine on cell proliferation in the developing
rat brain: a biochemical study.   Brain Res., 129: 283-297.

PAVLIK, A. & JELINEK, R.  (1979)  Effect of cycloheximide
administered to rats in early postnatal life: correlation of
brain changes with behaviour in adulthood.   Brain Res., 167:

PAVLIK, A. JELINEK, R., & TEISINGER, J.  (1980)  Short-term
inhibition of protein synthesis applied at brain growth spurt:
behavioural teratology model.  In: Trojan, S. & Stastny, F.,
ed.   Ontogenesis of brain,Prague, Univ. Carol. Prag., Vol. 3,
pp. 397-404.

PEDERSEN, R.A.  (1981)  Benzo[a]pyrene metabolism in early
mouse embryos.  In: Neubert, D. & Merker, H.-J., ed.   Culture
 techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 447-454.

PELKONEN, O., JOUPPILA, P., & KAERKI, N.T.  (1972)  Effect of
maternal cigarette smoking on 3,4-benzpyrene and
 N-methylaniline metabolism in human fetal liver and placenta.
 Toxicol. appl. Pharmacol., 23: 399-407.

PELLEGRINO, L. & ALTMAN, J.  (1979)  Effects of differential
interference with postnatal cerebellar neurogenesis on motor
performance, activity level and maze learning of rats: a
development study.  J. comp. Physiol. Psychol., 93: 1-33.

Altered proteoglycan metabolism in mouse limb mesenchyme cell
cultures treated with vitamin A.   Arch. Biochem. Biophys.,
186: 351-358

PENNYPACKER, J.P., WILK, A.L., & MARTIN, G.R.  (1978b)  In
vitro differentiation of mesenchyme cells into chondrocytes:
Effect of teratogens in the presence or absence of
drug-metabolizing liver preparations.  In: Neubert, D.,
Merker, H.-J., Nau, H., & Langman, J., ed.   Role of
 pharmacokinetics in prenatal and perinatal toxicology,
Stuttgart, Georg Thieme Publ., pp. 41l-419.

PERRAUD, J.  (1976)  Levels of spontaneous malformations in
the CD rat and CD-1 mouse.   Lab. anim. Sci., 26: 293-300.

PHILLIPS, M., ADAMS, J., & BUELKE-SAM, J.  (1980)  Righting
vs. negative geotaxis: a methodological evaluation.
 Teratology, 21: 60-61A.

PINTO-MACHADO, J.  (1970)  Influence of prenatal
administration of busulfan on the postnatal development of
mice.  Production of a syndrome including hypoplasia of the
thymus.   Teratology, 3: 363-370.

Embryotoxicity induced by alkylating agents: 1. Ethymethane-
sulfonate as a teratogen in mice.   Arch. Toxicol., 51: l-25.

POPOV, V.B., VAISMAN, B.L., & PUCHKOV, V.F.  (1981)
[Embryotoxic action of cyclophosphamide after
biotransformation  in vitro on development of rat embryos.]
 Bull. exp. Biol. Med., 91(5): 613-615 (in Russian, English

PRATT, R.M.  (1983)  Mechanisms of chemically-induced cleft
palate.   Trends pharmacol. Sci., 4: 160-162.

(1980)  Involvement of glucocorticoids and epidermal growth
factor in secondary palate development.  In: Pratt, R.M. &
Christiansen, R.L., ed.   Current research trends in prenatal
 craniofacial development, New York, Amsterdam, Oxford,
Elsevier, North-Holland Publ., pp. 235-252.

PRATT, R.M., GROVE, R.I., & WILLIS, W.D.  (1982)  Prescreening
for environmental teratogens using cultured mesenchymal cells
from the human embryonic palate.   Teratog. Carcinog. Mutag.,
2: 313-318.

PRICE, D. & ORTIZ, E.  (1965)  The role of fetal androgen in
sex differentiation in mammals.  In: De Haan, R.L. & Ursprung,
H., ed.   Organogenesis, New York, Rinehart and Winston, pp.

QUARLES, J.M.  (1981)  Transplacental host-mediated assay in
the hamster as a rapid screening test for chemical
carcinogens.   Biol. Res. Pregnancy, 2: 188-194.

RAEIHAE, N.C.R. & SCHWARTZ, A.L.  (1973)  Enzyme induction in
human fetal liver in organ culture.   Enzyme, 15: 330-339.

RAJAN, K.T.  (1969)  The cultivation  in vitro of postfoetal
mammalian cartilage and its response to hypervitaminosis A.
 Exp. Cell Res., 55: 419-423.

RAJAN, K.T.  (1974)  Human organ culture: applications in the
screening of teratogenic drugs.  In: Woollam, D.H.M. &
Morriss, G.M., ed.   Experimental embryology and teratology,
London, Elek Science, Vol. 1, pp. 65-89.

RAJAN, K.T., MERKER, H.-J., & WILKINS, M.  (1980)
Observations on human digits  'in vitro' and its possible role
in evaluating teratogens.  In: Merker, H.-J., Nau, H.,
Neubert, D., ed.   Teratology of the limbs, Berlin, Walter de
Gruyter Publ., pp. 307-324.

RANDALL, P.K. & CAMPBELL, B.A.  (1976)  Ontogeny of behavioral
arousal in rats: effect of maternal and sibling presence.   J.
 comp. Physiol. Psychol., 90: 453-459.

RAUTENBERG, P. & NEUBERT, D.  (1975)  Methods for the
evaluation of mRNA fractions synthesized in limb buds
differentiating in organ culture.  In: Neubert, D. & Merker,
H.-J., ed.   New approaches to the evaluation of abnormal
 embryonic development,Stuttgart, Georg Thieme Publ., pp.

REINISCH, J.M. & KAROW, W.G.  (1977)  Prenatal exposure to
synthetic progestins and estrogens: effects on human
development.   Arch. sex. Behav., 6: 257-288.

REITER, L.  (1977)  Behavioral toxicology: effects of early
postnatal exposure to neurotoxins on development of locomotor
activity in the rat.   J. occup. Med., 19: 20l-204.

REUHL, K.R. & CHANG, L.W.  (1979)  Effects of methylmercury on
the development of the nervous system:  a review.
 Neurotoxicology, 1: 21-55.

REVERBERY, J., ed.  (1971)   Experimental embryology of marine
 and fresh-water invertebrates, New York, American Elsevier, p.

RICE, J.M  (1969)  Transplacental carcinogenesis in mice by
1-ethyl-1-nitrosourea.   Ann. NY Acad. Sci., 163: 813-826.

RICE, J.M.  (1973a)  An overview of transplacental chemical
carcinogenesis.   Teratology, 8: l13-126.

RICE, J.M.  (1973b)  The biological behaviour of
transplacentally induced tumours in mice.  In: Tomatis, L.,
Mohr, U., & Davis, W., ed.   Transplacental carcinogenesis,
Lyons, New York, International Agency for Cancer Research, p.
71-83 (IARC Scientific Publications).

RICE, J.M.  (1981)  Effects of prenatal exposure to chemical
carcinogens and methods for their detection.  In: Kimmel, C.A.
& Buelke-Sam, J., ed.   Developmental toxicology, New York,
Raven Press, pp. 191-212.

G.M.  (1977)  Direct and transplacental carcinogenesis by
ethylnitrosourea in the patas monkey  (Erythrocebus patas).
In:  Proceedings of the American Association on Cancer
 Research., 18 (Mar.): 53.

RISO, J. & ZIMMERMANN, B.  (1981)  Effects of substances
influencing glycosaminoglycan synthesis on lung development in
culture.  In: Neubert, D. & Merker, H.-J., ed.   Culture
 techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 331-339.

RITTER, E.J.  (1977)  Altered biosynthesis.  In: Wilson, J.G.
& Fraser, F.C., ed.   Handbook of teratology,Vol. 2, New York,
London, Plenum Press, pp. 99-116.

RIVLIN, A.S. & TATOR, C.H.  (1977)  Objective clinical
assessment of motor function after experimental spinal cord
injury in the rat.   J. Neurosurg.,47: 577-581.

ROBERTS, D.W. & CHAPMAN, J.R.  (1981)  Concepts essential to
the assessment of toxicity to the developing immune system.
In: Kimmel, C.A. & Buelke-Sam, J., ed.   Developmental
 toxicology, New York, Raven Press, pp. 167-189.

ROBINSON, E.  (1976)  The effects of litter size and crowding
on position learning by male and female albino rats.   Psychol.
 Res., 26: 61-66.

ROBKIN, M.A., SHEPARD, T.H., & TANIMURA, T.  (1972)  A new in
vitro culture technique for rat embryos.   Teratology, 5:

RODIER, P.M.  (1976)  Critical periods for behavioral
anomalies in mice.   Environm. Health Perspect., 18: 79-83.

RODIER, P.M.  (1978)  Behavioural teratology.  In: Wilson,
J.G. & Fraser, F.C., ed.   Handbook of teratology, New York,
London, Plenum Press, Vol. 4, pp. 397-428.

Behavioral consequences of interference with CNS development
in the early fetal period.   Teratology, 19: 327-336.

ROSCHLAU, G. & JUSTUS, J.  (1971)  [Carcinogenic action of
methotrexate and cyclophosphamide in ammrial experiments.]
 Dtsch. Gesundheitsw., 26: 219-222 (in German).

ROSENGARTEN, H. & FRIEDHOFF, A.J.  (1979)  Enduring changes in
dopamine receptor cells of pups from drug administration to
pregnant and nursing rats.   Science, 203: l133-1135.

RUCH, J.V. & KARCHER-DJURICIC, V.  (1971)  Action de la
5-fluoro-deoxyuridine sur la différenciation  in vitro de
molaires d'embryons de souris.   Arch. biol., 82: l15-129.

RUNNER, M.N.  (1965)  Transplantation of mammalian ova:
blastocyst and earlier stages.  In: Wilson, J.G. & Warkany,
J., ed.   Teratology: principles and techniques, Chicago,
London, The Univ. of Chicago Press, pp., 104-111.

RUSSELL, W.M.S. & BURCH, R.L.  (1959)   The principles of human
 experimental technique,London, Metheun, 238 pp.

RUSTIA, M. & SHUBIK, P.  (1976)  Transplacental effects of
diethylstilbestrol on the genital tract of hamster offspring.
 Cancer Lett., 1: 139-146.

(1973)  Transplacental effect of carcinolipin in mice.
 Neoplasma, 20: 347-348.

SADLER, T.W.  (1979a)  Culture of early somite mouse embryos
during organogenesis.   J. Embryol. exp. Morphol., 49: 17-25.

SADLER, T.W.  (1979b)  Effects of hyperglycemic and diabetic
rat serum on neural tube closure in early somite mouse embryos
maintained in whole embryo culture.   Anat. Rec., 193: 672

N.M., & SOLOBYEVA, N.A.  (1980)  Enzymic "imprinting" as the
result of early postnatal administration of enzyme inducers to
animals.  Exper. Basel, 36: 43-45.

SAXEN, I.  (1965)  Ontogenesis of virus resistance.  In:
Bergstrand, H. & Hellström, K.E., ed.   Symposium on Virus and
 Cancer, Unio Nordica Contra Cancrum,Stockholm, Tryckeri
Balder AB, pp. 150-161.

SAXEN, L.  (1966)  Effect of tetracycline on osteogenesis  in
 vitro.  J. exp. Zool., 162: 269-294.

SAXEN, I.  (1973)  Effects of hydrocortisone on the
development  in vitro of secondary palate in two inbred strains
of mice.   Arch. Oral Biol., 18: 1469-1479.

SAXEN, L. & EKBLOM, P.  (1981)  The developing kidney as a
model system for normal and impaired organogenesis.  In:
Neubert, D. & Merker, H.-J., ed.   Culture techniques.
 Applicability for studies on prenatal differentiation and
 toxicity, Berlin, New York, Walter de Gruyter Publ., pp.

SAXEN, L. & KAITILA, I.  (1972)  The effect and mode of action
of tetracycline on bone development  in vitro. in: Klingberg,
M.A., Abramovic, A., & Ckhemke, J., ed.  Drugs and fetal
 development, New York, London, Plenum Press, pp. 205-218.

SAXEN, L. & LEHTONEN, E.  (1978)  Transfilter induction of
kidney tubules as a function of the extent and duration of
intercellular contacts.   J. Embryol. exp. Morphol., 47: 97-109.

SAXEN, I. & SAXEN, L.  (1975)  Organ culture in teratology:
Closure of the palatal shelves as a model system.  In:
Neubert, D. & Merker, H.-J., ed.   New approaches to the
 evaluation of abnormal embryonic development,Stuttgart, Georg
Thieme Publ., pp. 84-98.

J., & WARTIOVAARA, J.  (1968)  Differentiation of kidney
mesenchyme in an experimental model system.   Adv. Morphol., 7:

SCANDALIOS, J.G.  (1979)   Physiological genetics, New York,
Academic Press.

SCHACTER, L.P.  (1970)  Effect of conditioned media on
differentiation in mass cultures of chick limb bud cells.  I.
Morphological effects.   Ext. Cell Res., 63: 19-32.

SCHARDEIN, J.L.  (1976)  Drugs as teratogens, Cleveland, CRC

SCHARDEIN, J.L.  (1983)  Teratogenic risk assessment - past,
present, and future.  In: Kalter, H., ed.   Issues and reviews
 in teratology, New York, Plenum Publishing Corporation, pp.

(1977a)  [Tumor induction by various endogenously formed
 N-nitroso ureas in the rat.]   Zbl. allg. Pathol. u. pathol.
 Anat., 121: 61-75 (in German).

SCHNEIDER, J., WARZOK, R., & SCHWARZ, H.  (1977b)  [Endogenous
formation of carcinogenic N-nitroso compounds in rats after
application of drugs and nitrite.]   Exp. Pathol. (Jena),
13(l): 32-43 (in German).

(1978)  [Postnatal and transplacental carcinogenic action of
methylmethane sulfonate (MMS) and ethylmethane sulfonate (EMS)
in hooded rats.]   Exp. Pathol. (Jena), 16(l/6): 157-167 (in

SCHOENTAL, R. & CAVANAGH, J. B.  (1972)  Brain and spinal cord
tumors in rats treated with pyrrolizidine alkaloids.   J. Natl
 Cancer Inst., 49(3): 665-671.

SCHREIWEIS, D.O. & MURRAY, G.J.  (1976)  Cardiovascular
malformations in  Oryzias latipes embryos treated with
2,4,5-trichlorophenoxyacetic acid (2,4,5-T).   Teratology, 14:

(1982)  Embryotoxic and teratogenic effects of aqueous
extracts of tar from coal gasification electrostatic precip-
itator.   Teratog. Carcinog. Mutag., 2: 1-11.

SEHGAL, C.B. & HUTTON, J.J.  (1977)  Homogenates of pregnant
rat and fetal tissues metabolize carcinogens to mutagens
detected by Salmonella typhimurium TA98 and TA100.   Mutat.
 Res., 46: 325-344.

SEIBEL, W.  (1974)  An ultrastructural comparison of the
uptake and transport of horseradish peroxidase by the rat
visceral yolk sac placenta during mid- and late gestation.
 Am. J. Anat., 140: 213-236.

SHENDRIKOVA, I.A. & ALEXANDROV, V.A.  (1978)  [Transplacental
passage and carcinogenic action of nitrosodimethylamine in
rats.]  In:  Conference on Carcinogenic N-nitrosocompounds -
 action, formation, detection.Tallin, pp. 89-90 (in Russian).

SHEPARD, T.H.  (1965)  The thyroid.  In: De Haan, R.L. &
Ursprung, H., ed.   Organogenesis, New York, Holt, Rinehart and
Winston, pp. 493-512.

SHEPARD, T.H.  (1967)  Onset of function in the human fetal
thyroid: Biochemical and radioautographic studies from organ
culture.   J. clin. Endocrin., 27: 945-958.

SHEPARD, T.H.  (1974)  Embryonic and fetal thyroid
development.  In: Kelley, V., ed.   Endocrine and metabolic
 disorders in children, Harper and Row, pp. 375-382.

SHEPARD, T.H.  (1983)   Catalog of teratogenic agents, 4th
 edition, Baltimore, London, Johns Hopkins University Press.

SHEPARD, T.H. & BASS, G.L.  (1970)  Organ culture of limb buds
from riboflavin-deficient and normal rat embryos in normal and
riboflavin-deficient media.   Teratology, 3: 163-168.

SHEPARD, T.H. & FANTEL, A.G.  (1981)  Teratology of
therapeutic agents.  In: Iffy, L. & Kaminetzky, H.A., ed.
 Principles and practice of obstetrics and perinatology, New
York, Wiley, J. and Sons, pp. 461-481.

SHEPARD, T.H., TANIMURA, T., & ROBKIN, M.  (1969)   In vitro
study of rat embryos.  I. Effects of decreased oxygen on
embryonic heart rate.   Teratology, 2: 107-110.

SHEPERD, T.H., TANIMURA, T., & ROBKIN, M.A.  (1970)  Energy
metabolism in early mammalian embryos.  Dev. Biol., Suppl. 4:

SHEPARD, T.H., MILLER, J.R., & MAROIS, M., ed.  (1975)
 Methods for Detection of Environmental Agents that Produce
 Congenital Defects, amsterdam, Oxford, North-Holland Publ.
Comp., New York, American Elsevier Publ. Comp. Inc., 263 pp.

I.  Teratology testing development and status of short-term
prescreens. II. Biotransformation of teratogens as studied in
whole embryo culture. In Developmental Pharmacology.

SHIRLEY, E.A.C. & HICKLING, R.  (1981)  An evaluation of some
statistical methods for analysing numbers of abnormalities
found amongst litters in teratology studies.   Biometrics, 37:

SIEBER, S.M., & FABRO, S.  (1971)  Identification of drugs in
the preimplantation blastocyst and in the plasma, uterine
secretion and urine of the pregnant rabbit.   J. Pharm. exp.
 Ther., 176: 65-75.

SIITERI, P.K. & WILSON, J.D.  (1974)  Testosterone formation
and metabolism during male sexual differentiation in the human
embryo.   J. clin. Endocrinol. Metab., 38: l13-125.

OCCORSIO, P.  (1981)  Effects of gliadin-derived peptides from
bread and durum wheat on small intestine cultures from rat
fetus and coeliac children.  In: Neubert, D. & Merker, H.-J.,
ed.   Culture techniques.  Applicability for studies on
 prenatal differentiation and toxicity, Berlin, New York,
Walter de Gruyter Publ., pp. 385-394.

SMART, J.L. & DOBBING, J.  (1971)  Vulnerability of developing
brain.  II. Effects of early nutritional deprivation on reflex
ontogeny and development of behaviour in the rat.   Brain Res.,
28: 85-95.

STIELBERG, S., & WILSON, J.  (in press)  Selection of chemical
compounds for  in vitro teratogen test validation. Teratog.
Carcinog. Mutag.,   :  -

SMITHBERG, M.  (1962)  Teratogenic effects of tolbutamide on
the early development of the fish,  Oryzias latipes.  Am. J.
 Anat., 110: 205-213.

SNEDECOR, G.W. & COCHRAN, W.G.  (1967)   Statistical methods,
 6th Edition, ames, Iowa, Iowa State University Press, pp.

SNOW, M.H.L.  (1973)  Abnormal development of pre-implantation
mouse embryos grown  in vitro with 3H-thymidine.   J. Embryol.
 exp. Morph., 29: 601-615.

SNOW, M.H.L.  (1975)  The functional competence of the inner
cell mass.  In: Neubert, D. & Merker, H.-J., ed.   New
 approaches to the evaluation of abnormal embryonic
 development,Stuttgart, Georg Thieme Publ., pp. 394-407.

SOBOTKA, T.J., COOK, M.P., & BRODIE, R.E.  (1974)  Effects of
perinatal exposure to methylmercury on functional brain
development and neurochemistry.   Biol. Psychiat., 8: 307-320.

SOBRIAN, S.K.  (1977)  Prenatal morphine administration alters
behavioural development in the rat.   Biochem. Behav., 7:

SOKAL, J.E. & LESSMANN, E.M.  (1960)  Effects of cancer
chemotherapeutic agents on the human fetus.   J. Am. Med.
 Assoc., 172: 1765-1771.

SOKAL, R. R. & ROHLF, F. J.  (1969)   Biometry.  Assumptions of
 the analysis of variance,San Francisco, Freeman, pp. 367-402.

SOLURSH, M., AHRENS, P.B., & REITER, R.S.  (1978)  A tissue
culture analysis of the steps in limb chondrogenesis.   In
 Vitro, 14: 51-61.

SOTOMAYOR, R.E. & CUMMING, R.B.  (1975)  Induction of
translocations by cyclophosphamide in different germ cell
stages in male mice: cytological characterization and
transmission.   Mutat. Res., 27: 375-388.

SPATZ, M. & LAQUEUR, G.L.  (1967)  Transplacental induction of
tumours in Sprague-Dawley rats with crude cycad material.   J.
 Natl Cancer. Inst., 38: 233-245.

SPIELMANN, H. & EIBS, H.-G.  (1977)  Preimplantation embryos.
Part I: Laboratory equipment, preparation of media, sampling
and handling of the embryos.  In: Neubert, D, Merker, H.-J., &
Kwasigroch, T.E., ed.   Methods in prenatal toxicology,
Stuttgart, Georg Thieme Publ., pp. 210-220.

& SCHIMMEL, A.  (1981)  Investigations on the mechanism of
action and on the pharmacokinetics of cyclophosphamide
treatment during the preimplantation period in the mouse.  In:
Neubert, D. & Merker, H.-J., ed.   Culture techniques.
 Applicability for studies on prenatal differentiation and
 toxicity, Berlin, New York, Walter de Gruyter Publ., pp.

SPYKER, D.A. & SPYKER, J.M.  (1977)  Response model analysis
for cross-fostering studies: prenatal versus postnatal effects
on offspring exposed to methylmercury dicyandiamide.   Toxicol.
 appl. Pharmacol., 40: 51l-527.

SPYKER, J.M.  (1975)  Assessing the impact of low level
chemicals on development: behavioral and latent effects.   Fed.
 Proc., 34: 1835-1844.

SPYKER, J.M. & FERNANDES, G.  (1973)  Impaired immune function
in offspring from methylmercury treated mice.   Teratology, 7:

SPYKER, J.M., SPARBER, S.B., & GOLDBERG, A.M.  (1972)  Subtle
consequences of methylmercury exposure: behavioral deviations
in offspring of treated mothers.  Science, 177: 621-623.

STAHLMANN, R., BLUTH, U., & NEUBERT, D.  (1981)  Effects of
some "indirectly" alkylating agents on differentiation of limb
buds in organ culture.  In: Neubert, D. & Merker, H.-J., ed.
 Culture techniques.  Applicability for studies on prenatal
 differentiation and toxicity, Berlin, New York, Walter de
Gruyter Publ., pp. 207-221.

STAPLES, R.E. & HASEMAN, J.K.  (1974)  Selection of
appropriate experimental units in teratology.   Teratology,
9(3): 259-260.

STAVE, U.  (1975)  Perinatal changes of interorgan differences
in cell metabolism.   Biol. Neonat., 26: 318-332.

STAVROU, D. & LUEBBE, I.,  (1974)  [Transplacental induction
of kidney tumours in rabbits by ethylnitrososourea.]   Verh.
 Dtsch. Ges. Path., 58: 545 (in German).

STAVROU, D., DAHME, E., & SCHROEDER, B.  (1977)  [Trans-
placental neuro-oncogenic effect of ethylnitroso urea in
rabbits during early embryogenesis.]   Z. Krebsforsch., 89:
331-339 (in German).

STEBBINS, W.C. & RUDY, M.C.  (1978)  Behavioral
ototoxicology.   Environ. Health Perspect., 26: 43-51.

STEPTOE, P.C. & EDWARDS, R.G.  (1978)  Birth after the
reimplantation of a human embryo.   Lancet, II: 366.

STEVENS, L.C. (1967)  The biology of teratomas.   Adv.
 Morphog., 6: 1-31.

STREISINGER, G.  (1975)  Invited discussion. On the possible use
of zebra fish for the screening of teratogens.  In: Shepard, T.H.,
Miller, J.R., & Marois, M., ed.   Methods for detection of
 environmental agents that produce congenital defects,  Amsterdam,
Oxford, New York, North-Holland Publ., pp. 59-64.

TABACOVA, S.  (1983)  Behavioural effects of prenatal exposure
to nitrogen dioxide.   11th Conference of European Teratology
 Society,Paris, 29-31 August 1983 (abstracts).

(1981)  Hazards for the progeny after maternal exposure to low
carbon disulphide concentrations.   G. Ital. med. Lav., 3:

TANIMURA, T.  (1980)  Introductory remarks on behavioural
teratology.   Cong. Anom. 20: 30l-318.

TESH, J.M.  (1977)  An approach to the assessment of postnatal
development in laboratory animals.  In: Neubert, D., Merker,
H.-J., & Kwasigroch, T.E., ed.   Methods in prenatal
 toxicology,Stuttgart, Georg Theime Publ., pp. 186-195.

THESLEFF, I. & PRATT, R.M.  (1980)  Tunicamycin inhibits mouse
tooth morphogenesis and odontoblast differentiation  in vitro.
 J. Embryol. exp. Morphol.,195-208.

THOMAS, J.B.  (1968)   Introduction to human embryology,
Philadelphia, Lea & Febigert, 348 pp.

THORNTON, C.S.  (1968)  Amphibian limb regeneration.   Adv.
 Morphog., 7: 205-249.

THORPE, E.  (1967)  Some pathological effects of
hexachlorophene in the rat.   J. comp. Pathol., 77: 137-142.

TILSON, H. & MITCHELL, C.  (1980)  Models for neurotoxicity.
In: Hodgson, E., Bend, J., & Philpot, R., ed.   Reviews in
 biomedical toxicology, New York, Elsevier, North Holland, pp.

TILSON, H.A. & CABE. P.A.  (1978)  Strategy for the assessment
of neurobehavioural consequences of environmental factors.
 Environ. Health Perspect., 26: 287-299.

TILSON, H., MITCHELL, O., & CABE, P.  (1979)  Screening for
neurobehavioural toxicity: the need for and examples of
validation of testing procedures.   Neurobehav. Toxicol., 1
(Suppl.),: 137-148.

TILSON, H., CABE, P., & BURNE, T.  (1980)  Behavioural
procedures for the assessment of neurotoxicity.  In: Spencer,
P.S. & Schaumburg, H.H., ed.   Experimental and clinical
 neurotoxicology, Baltimore, Maryland, the Williams and Wilkins
Co., pp. 758-766.

TOMATIS, L.  (1979)  Prenatal exposure to chemical carcinogens
and its effect on subsequent generations.  In:  Prenatal
 carcinogenesis, Natl Cancer Inst. Monogr., 51: 159-184.

TOMATIS, L. & MOHR, U., ed.  (1973)   Transplacental
 Carcinogenesis,Lyon, IARC Sci. Publ., No. 4.

V.  (1981)  Increased cancer incidence in the progeny of male
rats exposed to ethylnitrosourea before mating.   Int. J.
 Cancer, 28: 475-478.

TROWELL, O.A.  (1961)  Problems in the maintenance of mature
organs  in vitro. in:  La culture organotypique,Paris,
Editions du Centre National de la Recherche Scientifique, pp.

TSONIS, P.A. & EGUCHI, G.  (1980)  Teratogenic effects of
carcinogens on limb regeneration in adult newts.   Dev. Growth
 Differ., 22: 708 (abstract).

DAVIS, W., ed.  (1973)   Pathology of tumours in laboratory
 animals, Vol. 1: Tumours of the rat, Part 1,Lyons,
International Agency for Cancer Research, 201 pp. (IARC
Scientific Publications No. 5).

(1976)   Pathology of tumours in laboratory animals, Vol. 1:
 Tumours of the rat, Part 2,Lyons, International Agency for
Cancer Research, 315 pp. (IARC Scientific Publications No. 5).

TURUSOV, V.S., ed.  (1979)  Pathology of tumours in laboratory
animals, Vol. 2:  Tumours of the mouse,Lyons, International
Agency for Cancer Research, 669 pp. (IARC Scientific
Publications No. 23).

TURUSOV, V.S., ed.  (1982)  Pathology of tumours in laboratory
animals, Vol. 3:  Tumours of the hamster,Lyons, International
Agency for Cancer Research (IARC Scientific Publications No.

TYLER, A. & TYLER, B.S.  (1966)  The gametes; some procedures
and properties.  Physiology of fertilization and early
development.  In: Bvolootian, R.A., ed.   Physiology of
 echinoderma, New York, Willy, I. & Sons Inc., Interscience
Publ., pp. 639-741.

UMANSKY, R.  (1966)  The effect of cell population density on
the developmental fate of reaggregating mouse limb bud
mesenchyme.   Dev. Biol., 13: 31-56.

US EPA  (1980)  Environmental Protection Agency, Mutagenicity
risk assessment: proposed guidelines.   Fed. Reg., 45(221):

US EPA  (1982)  Environmental Protection Agency, Assessment of
risk to human reproductive system and to development of the
human conceptus from exposure to environment substances. In:
 Proceedings of the US EPA sponsored conference, Oakridge,
 Tennessee, Oakridge National Laboratory, 1982.Springfield,
Virginia, National Technical Information Service (NTIS No. DE

US EPA  (1975)  Environmental Protection Agency. Pesticide
programs. Registration, re-registration, and classification
procedures.   Fed. Reg., 40: 28242-28286.

US EPA  (1976)  Environmental Protection Agency.  Health risk
and economic impact assessments of suspected carcinogens.
Interim procedures and guidelines.  Fed. Reg., 41: 21402-21405.

US EPA  (1978)  Environmental Protection Agency, chemical
substances inventory (Cumulative supplement (TS-799), July).

VERGIEVA, T.  (1981)  [Teratogenicity and embryotoxicity].
In: Kaloyanova, F., ed.   Hygienna Toksikologia. Med. i Fizkul.
(Sofia), pp. 148-155 (in Bulgarian).

VERGIEVA, T.  (1982)  [Karbendazim: teratogenic effects in
rats].   Hig. Zdr., 4: 333-339 (in Bulgarian).

VERGIEVA, T., ZAJKOV, CH., & ANTOV, G.  (1981a)  [Prenatal
alcohol exposure-changes in behaviour and activity of some
brain enzymes.]   Hig. Zdr., 6: 532-537 (in Bulgarian).

VESSELINOVITCH, S.D.  (1973)  Comparative studies on prenatal
carcinogenesis.  In: Tomatis, L., Mohr, U., & Davis, W., ed.
 Transplacental carcinogenesis,Lyons, International Agency for
Cancer Research, pp. 14-22 (IARC Scientific Publications No.

VESSELINOVITCH, S.D. & RAO, K.V.N.  (1974)  Transplacental and
neonatal carcinogenesis experimental approach.  In:
 Characterization of human tumours, amsterdam, New York,
Elsevier, pp. 135-150.

Neoplastic response of mouse tissues during perinatal age
periods and its significance in chemical carcinogenesis.  In:
Perinatal carcinogenesis.   Natl Cancer Inst. Monogr., 51:

SKACHKOVA, V.N.  (1980)  [Hygienic evaluation of the major
components of waste water from the production of antimicrobial
materials.]   Hig. i. Sanit., 10: 70-73 (in Russian).

Psychotropic drugs as behavioural teratogens.   Science, 205:

VOUK, V.B. & SHEEHAN, P.J.  (1983)   Methods for assessing the
 effects of chemicals on reproductive functions, Chichester,
Sussex, England, John Wiley & Sons, 480 pp.

(1974)  Transfilter induction of kidney tubules: correlation
with cytoplasmic penetration into nuclepore filters.   J.
 Embryol. exp. Morphol., 31: 667-682.

WATANABE, N.  (1979)  Environmental determinants of birth
defects prevalence.  In: Klingberg, M.A., Weatherall, J.A.C.,
ed.   Epidemiological methods for detection of teratogens,
 Contributions to epidemiology and biostatistics, Basel,
Munchen, Paris, London, New York, Sidney, Karger, S., pp.

WEIL, C.S.  (1970)  Selection of the valid number of sampling
units and a consideration of their combination in
toxicological studies involving reproduction, teratogenesis,
or carcinogenesis.   Food Cosmet. Toxicol., 8: 177-182.

WEIL, C.S.  (1972)  Guidelines for experiments to predict the
degree of safety of a material for man.   Toxicol. appl.
 Pharmacol., 21: 194-199.

WEISS, B.  (1975)  Effects on behaviour.  In:  Principles for
 evaluating chemicals in the environment,Washington, DC,
National Academy of Sciences National Research Council, pp.

FINSTER, M., & CONNEY, A.H.  (1969)  Stimulatory effect of
cigarette smoking on the hydroxylation of 3,4-benzpyrene and
the N-dimethylation of 3-methyl-4-monomethylaminoazobenzene by
enzymes in human placenta.   Clin. Pharmacol. Ther., 10:

WELSCH, F., BAUMANN, I., & NEUBERT, D.  (1978)  Effects of
methyl-parathion and methyl-paraoxon on morphogenetic
differentiation of mouse limb buds in organ culture.  In:
Neubert, D., Merker, H.-J., Nau, H., & Langman, J., ed.   Role
 of pharmacokinetics in prenatal and perinatal toxicology,
Stuttgart, Georg Thieme Publ., pp. 351-358.

(1961)  Behavioural effects of prenatal drug administration in
the white rat.   Pediatrics, 27: 318-324.

WERBOFF, J., HAVLENA, J., & SIKOV, M.R.  (1962)  Effects of
prenatal X-irradiation on activity, emotionality and
maze-learning ability in the rat.   Radiat. Res., 16: 441-452.

WHITTEN, W.K.  (1971)  Nutrient requirements for the culture
of preimplantation embryos  in vitro. in: Raspé, G., ed.
Advances in the biosciences, Oxford - Braunschweig, Pergamon
Press - Vieweg, Vol. 6, pp. 129-141.

WHITTINGHAM, D.G.  (1971)  Culture of mouse ova.   J. Reprod.
 Fertil. Suppl., 14: 7-21.

WHO  (1967)   Principles for the Testing of Drugs for
 Teratogenicity,World Health Organization, Geneva, 18 pp. (WHO
Technical Reports Series No. 364).

WHO  (1978)   Principles and Methods for Evaluating the
 Toxicity of Chemicals, Part 1,World Health Organization,
Geneva, p. 19 (Environmental Health Criteria No. 6).

WILSON, J.G.  (1965)  Methods for administering agents and
detecting malformations in experimental animals.  In: Wilson,
J.G. & Warkany, J., ed.   Teratology: Principles and
 techniques, Chicago, University of Chicago Press, pp. 262-277.

WILSON, J.G.  (1973a)   Environment and birth defects, New
York, San Francisco, London, Academic Press, p. 180.

WILSON, J.G.  (1977)  Current status of teratology.  General
principles and mechanisms derived from animal studies.  In:
Wilson, J.G. & Fraser, F.C., ed.   Handbook of teratology:
 General principles and etiology, New York, Plenum Press, Vol.
1, pp. 47-74.

WILSON, J.G.  (1978a)  Survey of  in vitro systems: Their
potential use in teratogenicity screening.  In: Wilson, J.G. &
Fraser, F.C., ed.   Handbook of teratology, New York, London,
Plenum Press, Vol. 4, pp. 135-153.

WILSON, J.G.  (1978b)  Review of  in vitro systems with
potential for use in teratogenicity screening.   J. environ.
 Pathol. Toxicol., 2: 149-167.

WILSON, J.G. & FRASER, F.C., ed.  (1977)  Handbook of
 teratology, New York, London, Plenum Press, Vol. 1, 1476 pp.

WOO, D.C. & HOAR, R.M.  (1979)  Reproductive performance and
spontaneous malformations in control Charles River CD rats: A
joint study by MARTA. In:  Annual Meeting of Teratology Society.

WOO, D.C. & HOAR, R.M.  (1982)  Reproductive performance and
spontaneous malformations in control New Zealand White
rabbits: A joint study by Middle Atlantic Reproduction and
Teratology Association.  In:  Annual meeting of Teratology

YANAL, J.  (1979)  Long-term induction of microsomal drug
oxidizing system in mice following prenatal exposure to
barbiturate.  Biochem. Pharmacol., 28: 1429-1430.

ZBINDEN, G.  (1981)  Experimental methods in behavioral
teratology.   Arch. Toxicol., 48: 69-88.

ZELLER, W.J.  (1980)  [Prenatal carcinogenic action of
5-(3,3-dimethyl-1-triazeno)-imidazol-4-carboxamide (DTIC) on
the offspring of BD lX Rats.]   Arch. Geschwulstforsch., 50:
306-308 (in German).

    See Also:
       Toxicological Abbreviations