Pesticide residues in food, principles for the toxicological assessment of (EHC 104, 1990)

INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY

ENVIRONMENTAL HEALTH CRITERIA 104

PRINCIPLES FOR THE TOXICOLOGICAL
ASSESSMENT OF PESTICIDE RESIDUES IN FOOD

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 Organization
Geneva, 1990

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
chemicals.

WHO Library Cataloguing in Publication Data

Prinicples for the toxicological assessment of pesticide residues
in food.

(Environmental health criteria ; 104)

1.Pesticide residues - analysis - toxicity 2. Food contamination
I.Series

ISBN 92 4 157104 7 (NLM Classification: WA 240)
ISSN 0250-863X

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 1990

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
letters.

CONTENTS

PRINCIPLES FOR THE TOXICOLOGICAL ASSESSMENT OF PESTICIDE RESIDUES IN FOOD

FOREWORD

PREFACE

1. INTRODUCTION

2. GENERAL HISTORICAL BACKGROUND

3. JMPR ASSESSMENT PROCESS

4. CONSIDERATIONS OF IDENTITY, PURITY, AND STABILITY
4.1. Background
4.2. Principles
4.2.1. Identity
4.2.2. Purity
4.2.3. Stability

5. AVAILABILITY AND QUALITY OF DATA
5.1. Background
5.2. Principles

6. HUMAN DATA
6.1. Background
6.2. Current position
6.3. Principles

7. STRUCTURE-ACTIVITY RELATIONSHIPS
7.1. Principle

8. TEST METHODOLOGIES
8.1. Background
8.2. General considerations
8.2.1. Choice of species and strain
8.2.2. Group size
8.2.3. Selection of dose levels
8.2.4. Test duration
8.2.5. Pathological procedures
8.2.6. Historical control data
8.3. Conduct and evaluation of studies
8.3.1. Short- and long-term toxicity studies
8.3.2. Carcinogenicity studies
8.3.2.1 Background
8.3.2.2 Routes of exposure
8.3.2.3 Commonly occurring tumours
8.3.2.4 Pathological classification of neoplasms
8.3.2.5 Evaluation of carcinogenicity studies
8.3.2.6 Extrapolation from animals to man
8.3.2.7 Principles
8.3.3. Reproduction studies
8.3.3.1 Multigeneration reproduction studies
8.3.3.2 Teratology studies
8.3.3.3 Screening studies in teratology
8.3.3.4 Principles

8.3.4. Neurotoxicity studies
8.3.4.1 Delayed neurotoxicity
8.3.4.2 Acute neurotoxicity
8.3.4.3 Chronic neurotoxicity
8.3.4.4 Pyrethroid-induced neurotoxicity
8.3.4.5 Neurobehavioural toxicity
8.3.4.6 Principles
8.3.5. Genotoxicity studies
8.3.5.1 Principles
8.3.6. Immunotoxicity studies
8.3.6.1 Background
8.3.6.2 Current position
8.3.6.3 Principles
8.3.7. Absorption, distribution, metabolism, and excretion
8.3.7.1 Background
8.3.7.2 Current position
8.3.7.3 Principles

9. EVALUATION OF DATA
9.1. Extrapolation of animal data to humans
9.2. Safety factors
9.2.1. Background
9.2.2. Principles
9.3. Allocating the ADI
9.3.1. Background
9.3.2. Temporary ADIs
9.3.3. Present position

10. EVALUATION OF MIXTURES
10.1. Introduction
10.2. Background
10.3. Principle

11. RE-EVALUATION OF PESTICIDES

12. BIOTECHNOLOGY

13. SPECIAL CONSIDERATIONS FOR INDIVIDUAL CLASSES OF PESTICIDES
13.1. Organophosphates - ophthalmological effects
13.2. Organophosphates - aliesterase (carboxylesterase) inhibition
13.3. The need for carcinogenicity testing of organophosphates
13.4. Ocular toxicity of bipyridilium compounds
13.5. Goitrogenic carcinogens

REFERENCES

ANNEX I: GLOSSARY

ANNEX II: APPROXIMATE RELATION OF PARTS PER MILLION IN THE DIET TO MG/KG
BODY WEIGHT PER DAY

INDEX

WHO TASK GROUP ON PRINCIPLES FOR THE TOXICOLOGICAL ASSESSMENT OF PESTICIDE
RESIDUES IN FOOD

Dr N. Aldridge, The Robens Institute of Industrial & Environmental
Health & Safety, University of Surrey, Guildford, Surrey, United
Kingdom^a

Dr G. Becking, International Programme on Chemical Safety, World Health
Organization, Research Triangle Park, North Carolina, USA^a,e

Professor C.L. Berry, Department of Morbid Anatomy, The London Hospital
Medical College, London, United Kingdom^a,e

Dr A.L. Black, Department of Community Services and Health, Woden,
Australia^b,d,e

Professor J.F. Borzelleca, Department of Pharmacology and Toxicology,
Medical College of Virginia, Virginia Commonwealth University,
Richmond, USA^d,e

Dr G. Burin, Health Effects Division, Office of Pesticide Programs, US
Environmental Protection Agency, Washington, DC, USA^b,c,e

Dr J.R.P. Cabral, Unit of Mechanisms of Carcinogenesis, International
Agency for Research on Cancer, Lyon, France^a

Dr D.B. Clayson, Toxicology Research Division, Bureau of Chemical
Safety, Food Directorate, Health and Welfare Canada, Ottawa,
Ontario, Canada^e

Mr D.J. Clegg, Agricultural Chemicals Section, Toxicological Evaluation
Division, Food Directorate, Health Protection Branch, Ottawa,
Canada^a,b,c,d,e

Professor B. Goldstein, Rutgers Medical College, Busch Campus,
Pescataway, New Jersey, USA^a

Dr J.L. Herrman, International Programme on Chemical Safety, World
Health Organization, Geneva, Switzerland^a,b,c

Professor M. Ikeda, Department of Environmental Health, Tohoku
University School of Medicine, Sendai, Japan^a

Dr S.E. Jaggers, ICI Central Toxicology Laboratory, Cheshire, United
Kingdom^e

Dr F.-W. Kopisch-Obuch, Plant Protection Service, Food and Agriculture
Organization, Rome, Italy^e

Dr R. Kroes, National Institute of Public Health and Environmental
Hygiene, Bilthoven, The Netherlands^a

Professor M. Lotti, Istituto di Medicina del Lavoro, Università di
Padova, Padova, Italy^a,b,d,e

Dr L. Magos, Toxicology Unit, Medical Research Council Laboratories,
Woodmansterne Road, Carshalton, Surrey, United Kingdom^a

Dr K. Miller, Immunotoxicology Department, British Industrial Biologi-
cal Research Association, Surrey, United Kingdom^e

Professor R. Nilsson, The National Swedish Chemicals Inspectorate, De-
partment for Scientific Documentation and Research, Solna, Sweden^e

Dr A.K. Palmer, Reproductive Studies, Huntingdon Research Centre Ltd.,
Huntingdon, Cambridgeshire, United Kingdom^b,e

Professor D.V. Parke, Department of Biochemistry, University of Surrey,
Guildford, United Kingdom^a,e

Dr O.E. Paynter, Health Effects Division, US Environmental Protection
Agency, Washington, DC, USA^a,b,c,e

Dr R. Plestina, Division of Vector Biology and Control, World Health
Organization, Geneva, Switzerland^b

Dr F.R. Puga, Section of Toxicology, Instituto Biològico, Sao Paulo,
Brazil^d

Professor A. Rico, Ecole Nationale Vétérinaire, Toulouse, France^e

Dr L. Shuker, Unit of Carcinogen Identification and Evaluation, Inter-
national Agency for Research on Cancer, Lyon, France^b,e

Dr J. Steadman, Department of Health & Social Security, Hannibal House,
Elephant and Castle, London, United Kingdom^a

Dr E.M. den Tonkelaar, Toxicology Advisory Center, National Institute
of Public Health and Environmental Protection, Bilthoven, The
Netherlands^e

Dr G. Ungvary, Section of Toxicology, National Institute of Occu-
pational Health, Budapest, Hungary^d,e

Dr G. Vettorazzi, International Toxicology Information Centre, San
Sebastian, Spain^a,b,c,e

-------------------------------
^a Present at strategy meeting, Carshalton, United Kingdom, 2-6 March
1987
^b Present at consultation, Geneva, Switzerland, 14-16 September 1988
^c Member of editorial committee
^d Member of WHO Expert Group on Pesticide Residues (1989 JMPR)
^e Submitter of written comments

NOTE TO READERS OF THE CRITERIA DOCUMENTS

Every effort has been made to present information in the criteria
documents as accurately as possible without unduly delaying their pub-
lication. In the interest of all users of the environmental health cri-
teria documents, readers are kindly requested to communicate any errors
that may have occurred 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.

* * *

FOREWORD

The WHO activities concerned with the safety assessment of food
chemicals were incorporated into the International Programme on Chemi-
cal Safety (IPCS) in 1980. These activities include administering the
WHO Expert Group on Pesticide Residues, which meets regularly with the
FAO Panel of Experts on Pesticide Residues in Food and the Environment
in the well-known Joint FAO/WHO Meeting on Pesticide Residues, or JMPR.
The objectives of the WHO Expert Group are consistent with those of
IPCS, which include the formulation of "guiding principles for
exposure limits, such as acceptable daily intakes for food additives
and pesticide residues, and tolerances for toxic substances in food,
air, water, soil, and the working environment." The inclusion of the
present publication as a methodology document in the Environmental
Health Criteria series will make it readily available to all of those
who have an interest in the toxicological assessment of pesticide
residues in food.

The IPCS gratefully acknowledges the financial and other support
of the Canadian Health Protection Branch, the US Environmental Protec-
tion Agency, and the United Kingdom Department of Health. This support
was indispensable for the completion of the project.

Dr M. Mercier
Manager
International Programme on
Chemical Safety

PREFACE

Since the early 1960s the Joint FAO/WHO Meeting on Pesticide
Residues, usually known as the JMPR, has evaluated a large number of
pesticides. The WHO component of these Joint Meetings, the WHO Expert
Group on Pesticide Residues, has, during that time, relied upon pro-
cedures developed by other expert groups, such as the Joint FAO/WHO
Expert Committee on Food Additives (JECFA), and developed specific
principles for evaluating the various classes of pesticides that are
used on food crops and may leave residues on them. The publication of
WHO Environmental Health Criteria 70: Principles for the safety assess-
ment of food additives and contaminants in food, which summarizes the
assessment procedures used by JECFA, has been used by the WHO Expert
Group on Pesticide Residues since its publication. Other principles
specific to pesticides, however, have until now been scattered among
the various JMPR reports, which has made it difficult for the WHO
Expert Groups to use them in a consistent manner during their evalu-
ations. In addition, many of the reports date back many years, and some
of the advice given in earlier reports is no longer valid.

Recognizing the importance of maintaining consistency, an inter-
country meeting that was held in 1985 in Ottawa, Ontario, Canada to
consider ways of strengthening the role of JMPR in its evaluation of
pesticide residues in food recommended that the principles that have
been elaborated by JMPR through the years be codified and updated where
appropriate and consolidated in a single publication. The 1985 JMPR
supported such an effort and recommended "that this international
meeting be requested to consider the toxicological basis and data
requirements for the estimation of an ADI or temporary ADI, and to
provide general guidance on relevant toxicological methodology."

An IPCS planning meeting was held in March 1987 in Carshalton,
Surrey, UK in response to these recommendations, at which time areas
were identified for consideration, which were incorporated into the
first draft. This draft was reviewed at a task group meeting in Geneva
in September 1988, after which extensive revisions were made. An edi-
torial group meeting in Geneva in June 1989 produced the final draft,
which was considered by the WHO Expert Group at the 1989 JMPR. Drafts
were widely distributed at several stages, and the comments which were
received from a wide range of international experts have been
incorporated into the final publication.

The present publication therefore reflects the views of a large
number of international experts who are involved with the toxicological
assessment of pesticides. In addition, by concentrating on the pro-
cedures used by the WHO Expert Group on Pesticide Residues, it faith-
fully reflects the principles used in the evaluation of pesticide resi-
dues by JMPR. It is expected, therefore, that the future use of this
publication by the WHO Expert Group will ensure consistent decision-
making using up-to-date principles. Those involved in the production
of this publication also hope that it will be of significant value to
government officials responsible for establishing safe levels of pesti-
cide residues on food commodities and by companies producing safety
data on pesticides.

The WHO Expert Group on Pesticide Residues has been the responsi-
bility of the International Programme on Chemical Safety (IPCS) since
the inception of the Programme in 1980. The preparation of this publi-
cation provides an indication of the importance that IPCS places on the
work of the WHO Expert Group in particular and on the toxicological
assessment of pesticides in general. I am confident that those of us
responsible for the toxicological assessment of pesticides will find
the resources that have been put into the production of this publi-
cation to have been well-spent, and that the publication will be of
enormous value in our work.

Dr J.L. Herrman
WHO Joint Secretary
Joint FAO/WHO Meeting on
Pesticide Residues

1. INTRODUCTION

The World Health Assembly noted in 1953 that the increasing use of
various chemicals in the food industry had in recent decades created a
new public health problem. In response to this, the World Health
Organization, in conjunction with the Food and Agriculture Organiz-
ation, initiated two series of annual meetings on food additives (Joint
FAO/WHO Expert Committee on Food Additives, JECFA) and on pesticide
residues. The first meeting on food additives was held in 1956 and that
on pesticide residues in 1963.

Joint Meetings of the FAO Panel of Experts on Pesticide Residues in
Food and the Environment and the WHO Expert Group on Pesticide Residues
(usually referred to as the Joint FAO/WHO Meeting on Pesticide Resi-
dues, or JMPR) have provided an authoritative voice on the levels of
pesticides that can be ingested daily by man without appreciable risk;
this has been accomplished through the establishment of acceptable
daily intakes (ADIs). Since 1966, JMPR has been establishing maximum
residue limits (MRLs) of pesticides in food commodities.

This monograph has been prepared on behalf of the Central Unit of
the International Programme on Chemical Safety (IPCS) and its aim is to
provide an update of the principles utilized by the WHO Expert Group on
Pesticide Residues. It does not address the work of the FAO Panel.

Certain toxicological principles pertinent to JMPR have previously
been discussed in the JMPR reports. These principles usually relate to
advances in scientific knowledge, which have modified both test pro-
cedures and the evaluation of test results. The contents of the JMPR
reports were collated in 1977 [164]. Many of the basic principles were
initially adopted from the deliberations of JECFA and are detailed in
"Principles for the Safety Assessment of Food Additives and Contami-
nants in Food" [176]. In some of the areas where the principles used
by JECFA and JMPR are identical, direct quotes from that publication
have been included in this monograph. In this context, however, it
should be recognized that the two committees, while utilizing data from
similar types of studies, differ in their approach to the evaluation of
the available data. This difference in approach arises because JECFA
usually evaluates compounds intended for addition to food, which are
usually of low toxic potential, whereas JMPR deals with residues of
compounds that are toxic to at least some groups of living organisms.

The types of data that are evaluated when assessing the safety of a
pesticide include those from biochemical and toxicological studies and,
when available, observations in humans. The recent JMPR viewpoint is
that an understanding of the pharmacokinetic and pharmacodynamic
characteristics of a pesticide is extremely important and that such an
understanding will even compensate for inadequacies in the available
data base. On the other hand, this approach will sometimes lead to the
requirement for either additional parameters to be investigated in rou-
tine studies or for additional specific studies not routinely required
for the particular class of pesticide.

It has been recognized that data from studies using routes of
exposure other than oral are of value in the overall evaluation of the
safety of pesticides. However, these studies are not directly relevant
for the calculation of the ADI, so this monograph will not consider
these other routes of exposure in detail.

The more recent additions to the battery of toxicity tests avail-
able for use in safety assessment are discussed in this monograph. Some
of these tests, especially in the fields of immunotoxicity and behav-
ioural toxicity, are not yet at the stage of development where results
are consistently reproducible and therefore readily utilizable in
safety assessment. In addition, criteria for interpretation of such
studies have not yet been sufficiently developed to be of value in rou-
tine safety assessment. Therefore, only the potential of these studies
is discussed in this document.

The development of knowledge in the field of toxicology in recent
years has been quite remarkable. The history of JMPR and the changes
in its principles of safety assessment reflect this development. Thus
decisions taken by JMPR are always provisional and ADIs are subject to
re-evaluation as new significant data become available.

Each chapter in this monograph provides a rational background to a
specific area, describes the history of relevant changes in principles
according to the development of scientific knowledge, and offers a
short summary of the current position of JMPR. It also indicates the
principles being followed at present by the WHO Expert Group in their
evaluations of pesticide residues in food and recommendations on how
the studies may be performed to provide meaningful results.

2. GENERAL HISTORICAL BACKGROUND

The concept of JMPR was first proposed in 1959, when an FAO Panel
of Experts on the Use of Pesticides in Agriculture [29], recommended
that FAO and WHO should jointly study:

* the hazard to consumers arising from pesticide residues in and on
food and feedstuffs;
* the establishment of principles governing the setting up of pesti-
cide tolerances; and
* the feasibility of preparing an International Code for toxicologi-
cal and residue data required in achieving the safe use of a pesti-
cide.

Consequently, in 1961, a Joint Meeting of the FAO Panel of Experts
on the Use of Pesticides in Agriculture and the WHO Expert Committee on
Pesticide Residues was convened. The report of the 1961 Meeting [32]
recommended to the Directors-General of FAO and of WHO the evaluation
of "toxicological and other pertinent data . . . on those pesticides
known to leave residues in food when used according to good agricul-
tural practice". The evaluations would include the estimate of an
acceptable daily intake and an explanation of its derivation.

To implement this recommendation the first Joint Meeting of the FAO
Committee on Pesticide Residues in Agriculture and the WHO Expert Com-
mittee on Pesticide Residues was convened in September, 1963 [35]. This
Meeting adopted the concept of the acceptable daily intake, which was
based on:

* the chemical nature of the residue,
* the toxicity of the chemical based on data from acute, short-term,
and long-term studies, and knowledge of metabolism, mechanism of
action, and possible carcinogenicity of residue chemicals when con-
sumed (usually determined in animals);
* knowledge of the effects of these chemicals on humans.

The 1963 JMPR [35] adopted the use of "safety factors" for ex-
trapolating animal data to humans and to allow for variability within
the human population. It also noted other points to be considered when
establishing ADIs, including additive effects of multiple pesticides in
the diet, potentiation between pesticide residues, and genetic differ-
ences (especially in enzyme composition) within the exposed human popu-
lation.

The 1963 and 1965 Joint Meetings [35; 36] were concerned solely
with the acceptable daily intake and did not consider tolerances. Sep-
arate meetings of an FAO Working Party on Pesticide Residues examined
the issue of tolerances approximately two months after the Joint Meet-
ings and issued separate reports. The first report considered prin-
ciples [34] and the second proposed tolerances for pesticides on raw
cereals [37].

In 1966, the JMPR report [38] considered both ADIs and tolerances
for the first time. Joint Meetings have since been held yearly, and,
after each one, reports and evaluations have been published. The JMPR

has evolved principles consistent with the changing state of knowledge
in toxicology and chemistry, and the evaluation of new data has often
prompted adjustments in previous conclusions on various chemicals.
However, the products of the Joint Meetings (which include ADIs, tem-
porary ADIs, MRLs (MRL replacing the term "tolerance"), temporary
MRLs, guideline levels, and extraneous residue limits) have remained
essentially unchanged.

3. JMPR ASSESSMENT PROCESS

JMPR comprises two separate groups of scientists. The FAO Panel has
responsibility for reviewing pesticide use patterns, data on the chem-
istry and composition of pesticides, and methods of analysis of pesti-
cide residues, and for recommending MRLs that might occur in food com-
modities following the use of pesticides according to good agricultural
practices. The WHO Group has responsibility for reviewing toxicological
and related data and for estimating (where possible) an ADI for humans.
During the Joint Meetings, the two groups coordinate activities and
issue a joint report. However, the present section on interpretation
of data is limited to the procedures used by the WHO Expert Group.

The data used in the assessment of the toxicity of pesticide resi-
dues generally comprise acute studies, short-term feeding studies,
long-term feeding studies, and biochemical studies (including absorp-
tion, tissue distribution, excretion, metabolism, biological half-life,
and effects on enzymes). In addition, studies on specific effects,
e.g., carcinogenicity, reproduction, teratogenicity, and, for some com-
pounds, neurotoxicity, are usually necessary. Human data and other
information, e.g., SAR (structure-activity relationships), are also
considered when available.

The overall objective of the evaluation is to determine a no-
observed-adverse-effect level (NOAEL), based upon consideration of the
total toxicology data base, which will be utilized in conjunction with
an appropriate safety factor to determine the ADI. The initial stage
of the evaluation has to be a critical examination of the individual
studies. In some cases, a study initially considered to be of marginal
value may, in fact, be useful when considered in the context of the
entire data base. Integration of the results from all studies can then
permit an appraisal of the toxicity of the compound.

Data from acute oral studies are rarely considered to be relevant
to the establishment of the ADI. However, such data may provide infor-
mation that permits a ranking of the sensitivity of different species,
may assist in the selection of dose levels in subsequent studies, and
may indicate types of pharmacological activity, degree of absorption,
or potential target organs. JMPR has, on occasion, required additional
acute data to determine the relative toxicity of salts of a pesticide
(e.g., imazalil [57]) or required further metabolic studies to deter-
mine species differences in acute toxicity (e.g., triazophos [67]).

Historically, short-term feeding studies have provided the basis
for the determination of ADIs for a number of compounds evaluated by
JMPR. Prior to 1971 [48], before long-term toxicity studies were
indicated to be an essential part of the toxicological data base for
evaluating the safety of pesticides, ADIs were established for several
pesticides based on short-term toxicity studies (e.g., demeton [45],
parathion-methyl [182], dimethoate [182; 41], diazinon [182], azinphos-
methyl [182], methyl bromide [39], and dichlorvos [39]). Temporary
ADIs (TADIs) based on such studies have also been established, e.g.,
omethoate [49], fenthion [49]. Since long-term feeding studies have
become an essential part of the toxicological data base, the major use
of the data from short-term toxicity studies has been to determine

suitable dose levels to be utilized in long-term and reproduction
studies. However, studies lasting more than two years are rarely avail-
able in dogs. Thus, in situations where this species is more sensitive
to a toxic effect or more appropriate for use in extrapolation to
humans than are the rodent species, the ADI is usually based on data
generated in studies covering less than 50% of the lifespan of this
species, e.g., methamidophos [182], diflubenzuron [182], and phenthoate
[63; 73]. Occasionally, an ADI may also be based on short-term studies
in rodent species, e.g., diphenylamine [56; 68; 73], even though long-
term studies may exist which indicate higher NOAELs. This situation may
arise from adaptation, resulting in the disappearance of an effect
after long-term exposure.

Multigeneration reproduction and teratogenicity studies have also
been used for establishing ADIs for certain compounds, e.g., chlor-
mequat [51] and dinocap [183].

Delayed neurotoxicity has been identified as a potential problem
for a number of compounds evaluated by JMPR. To date, only leptophos
has had an ADI withdrawn because of this effect. However, withdrawal of
the ADI was not because of possible hazards from exposure to food resi-
dues but because of withdrawal of the product from the market as a
result of effects on heavily exposed individuals during its manufacture
[59]. ADIs have been allocated for other delayed neurotoxicants, e.g.,
isofenphos [76]. JMPR has indicated that the exaggerated doses (ex-
ceeding the LD50) used in the standard hen studies for delayed neuro-
toxicity are not necessarily applicable to the assessment of human
hazard arising from the intake of residues in food.

For the effects discussed above, the basic interpretation of the
available data depends upon the identification of a toxic effect, the
establishment of incremental increases in the incidence of this effect
with increasing exposure (i.e. a dose-effect relationship), and the
establishment of a threshold.

The primary consideration in determining whether a compound can
induce a toxic effect is the dose of test material to which the test
organism is exposed. A basic concept of toxicology is the statement
made by Paracelsus in 1538, which translated indicates: "Only the dose
decides that a thing is not poisonous" [1]. Thus, if a series of doses
used in a study fails to elicit a toxic response, then an insuf-
ficiently high dose level has been used. This philosophy is applicable
to all toxicity studies. However, provided an adequate margin of safety
exists between the highest dose tested and the possible human exposure
from pesticide residues, then a study in which a toxic effect was not
observed may be considered to be acceptable for the purpose of as-
sessing the safety of such residues. The major difficulty in the
absence of toxicity is the determination of the safety factor to be
applied to the highest dose tested, since there would be no indication
of the type, importance, or severity of the effect that might be in-
duced with increased dose levels.

A second factor is the determination of which effects should be
considered to be toxic. A judgment must be made, based on the nature
of the effect, on whether it should be considered adverse. As indicated

in section 8.3.6.2, plasma cholinesterase depression should not be
considered to be a toxic phenomenon since, although it is an effect, it
is not apparently a toxic effect. A reversible increase in liver weight
may be an adaptive response rather than a toxic effect. However, ancil-
lary studies may be required.

The determination of dose-response relationships in an experimental
population is based on the concept that the incidence or severity of an
adverse effect is related to dose. A time-response relationship may
also occur, i.e. where the incidence of an induced effect increases
with the duration of dosing at a constant level. Comparison of the
results of experiments with differing durations (performed on the same
species and strain, preferably in the same laboratory, and under simi-
lar environmental conditions) may be necessary to demonstrate time-
effect relationships if interim kills have not been performed. In the
absence of interim kills or shorter-term experiments for comparative
examination, it is sometimes possible to calculate the approximate time
of appearance of a major lesion based on findings in dying animals or
those sacrificed in moribund condition during the course of the study,
particularly if the lesion is associated with the cause of death.

The integrated nature of mammalian reproductive processes may
complicate the establishment of dose/time/response relationships for
reproductive effects. Events that are initiated during early develop-
ment may be moderated considerably in subsequent developmental stages.
The defense mechanisms which have evolved to minimize the consequences
of insult may repair minimal damage or discard that which is damaged
beyond effective repair (e.g., resorptions or abortions). Consequently,
in reproduction studies, the demonstrated dose response is often a
reflection of a progressive involvement of multiple variables rather
than a temporal change in a single variable.

When evaluating toxicological data, relevant parameters are evalu-
ated statistically so that their significance is established on the
basis of predetermined criteria. A statistically significant difference
between experimental and control groups should be considered in the
light of its biological relevance. Thus, an increased incidence of a
rare tumour type in treated animals may be of concern, even if the
incidence is not significantly different statistically from its inci-
dence in the concurrent control animals. Conversely, a statistically
significant change in an isolated parameter, e.g., erythrocyte count,
would usually not be considered to be biologically relevant unless
supported by changes in other parameters, e.g., bone marrow or spleen
histopathology or methaemoglobin formation.

A high background incidence of a specific lesion frequently compli-
cates the interpretation of data generated in toxicity studies. To some
extent, especially in the case of a specific tumour type, this problem
can be avoided by judicious choice of species or strains of animals.
For example, if the target organ is known to be the kidney, the
interpretation of results would be difficult in long-term studies in a
rat strain with a high incidence of geriatric nephropathy.

The use of more than one species for the same type of toxicity
study may complicate interpretation in those cases where an effect
occurs in one species but not in a second species, or where one species
is much more sensitive to the agent than the other species. In such
cases it is often difficult to determine the most appropriate species
for extrapolation to man. Generally, unless adequate data are available
to indicate the most appropriate species (usually comparative pharmaco-
kinetic or pharmacodynamic data), the most sensitive species (i.e., the
species in which the adverse effect occurred at the lower dose) is used
in determining the NOAEL and allocating the ADI.

In interpreting carcinogenicity data, JMPR bases its evaluations on
the threshold concept, which is the basis for evaluating most other
toxicological effects [170]. In assessing tumour incidence, benign and
malignant tumours have been considered as separate entities in the
majority of cases [176]. For further discussion of the determination
of NOAELs in carcinogenicity studies, see section 8.3.4.5.

The 1986 Joint Meeting stressed the importance of understanding the
mechanism of action that results in the expression of toxicity. It
noted that: "Current knowledge of mechanisms of toxicity is limited,
but there is already a sufficient understanding in some cases to permit
better design, performance, and interpretation of toxicological
studies. Mechanistic studies are therefore encouraged, since a knowl-
edge of mechanism of action is likely to result in a more rational
assessment of the risk to man." [76, p. 2].

4. CONSIDERATIONS OF IDENTITY, PURITY, AND STABILITY

4.1. Background

The report of the WHO Scientific Group on Procedures for Investi-
gating Intentional and Unintentional Food Additives indicated in 1967
that: "adequate specifications for identity and purity should be
available before toxicological work is initiated. Toxicologists and
regulatory bodies need assurance that the material to be tested corre-
sponds to that to be used in practice." It also stated that: "levels
of impurities that, according to current knowledge, are considered to
be toxicologically significant . . . must appear in the specifi-
cations." [169, p. 8].

The need for accurate specifications for pesticides was stressed by
the 1968 JMPR [42] during its deliberations on toxaphene and on techni-
cal grades of benzene hexachloride (BHC). Because of the unknown or
variable composition of these compounds, the JMPR was unable to relate
the existing toxicological data to the products in actual agricultural
use. Consequently, ADIs could not be allocated. Attention was also
drawn to the likelihood of variability between similar chemicals pro-
duced by different manufacturers.

The possible influence of known or unknown impurities on the tox-
icity of technical grade chemicals and of residues resulting from their
use was discussed by the 1974 JMPR [53]. This JMPR noted that toxicity
studies are generally performed on technical grade materials produced
by commercial-scale processes and that the resulting toxicological data
normally, therefore, take into account the presence of impurities (pro-
vided that the manufacturing process remains the same). However, it
noted the problems encountered with trace amounts of biologically ac-
tive materials, e.g., 2,3,7,8-tetrachloro-dibenzo- p -dioxin in 2,4,5-
trichloroacetic acid. It further noted that: "specifications such as
those issued by FAO and WHO are seldom designed to take note of trace-
level impurities, unless the importance of such impurities has already
been revealed by biological studies." [53, p. 15].

The 1977 JMPR [57] noted that data on the nature and level of
impurities, intermediates, and by-products in technical pesticides were
often available, but, because such data could provide valuable infor-
mation to competitors they were normally considered to be a "trade
secret". The Joint Meeting, therefore, agreed that such data would not
normally be published in the JMPR reports or monographs.

In considering the applicability of recommendations to pesticides
from different feedstocks produced by different manufacturers, the 1978
JMPR [59] indicated that evaluations and recommendations are valid only
for the specific technical grade of pesticide examined. Considerable
care and knowledge of the detailed specifications are required to
extrapolate evaluations and recommendations to products of differing
quality or composition.

Subsequent Joint Meetings [60; 62; 72] have stressed the importance
of information on the presence of impurities in technical pesticide
products (e.g., the presence of hexachlorobenzene in various pesti-
cides, impurities in phenthoate, and dimethyl hydrazine in maleic
hydrazide). The need for technical grade pesticides to meet FAO speci-
fications has also been stressed. It was noted by the 1984 JMPR that
occasionally data have been rejected because the test material failed
to comply with these specifications [72].

The 1987 JMPR [78] noted that ADIs based on studies using compounds
of specific purity can be relevant to products of different origin or
purity but that there are examples of changes in the amount or type of
impurity in the technical material that markedly influence the toxicity
of a compound.

Toxicity tests should normally be performed on the technical grade
of the pesticide (except for acute toxicity, for which both formu-
lations and technical materials must be tested to assess the risk to
the applicator). However, the percentage of active ingredient and
impurities in the technical grade material may vary among production
batches and may differ at various stages of product development.
Furthermore, since some toxicity testing is likely to be performed
with the product in the early stages of development, the preliminary
studies (usually designed to assess potential acute hazards to individ-
uals who will be involved in the development of the material) may be
performed on batches of material produced within the laboratory. Sub-
sequent studies may be performed on material produced in a pilot plant,
while other toxicity studies may be performed on the marketed product,
which will be produced in a full-scale manufacturing plant. At each
step in this sequence, there is a potential for change in the percent-
age of the active ingredient in the "technical-grade" material and a
potential for change in the quantity and identity of the impurities
that constitute the remainder of the "technical-grade" product. It
is, therefore, essential that detailed specifications should be pro-
vided for the test material utilized in each study.

In certain cases the pesticidally active ingredient may exist in
two or more forms, e.g., as a diastereoisomeric mixture. In the case
of the synthetic pyrethroids, this is normally the case. Under such
conditions, the ratio of isomers in the test material must be clearly
specified since it has been documented that different isomers fre-
quently have different toxicological activities [60; 87]. For example,
an ADI for permethrin (40% cis : 60% trans) was allocated in 1982 [67],
whereas the ADI for permethrin (25% cis : 75% trans) was not allocated
until 1987 [78].

Data on the stability of the test material is also of importance.
The percentage of the active material will decrease and that of break-
down products will increase with time if a test compound is unstable
under the conditions of storage. This will be of major importance in
the evaluation of the results of studies where a single batch of tech-
nical material is utilized for a long-term study or a multigeneration
study. Variation in the amount of degradation occurring in different
batches (i.e., batches of different post-manufacturing age) may compli-
cate the interpretation of a study. Finally, reaction of the test

compound with components of the test diet may result in the production
of toxic reaction products in the diet, which may affect the nutritive
value of the diet and will result in a decreased concentration of test
compound. The NOAEL may well be overestimated if the percentage of
active ingredient decreases with time. Conversely, if a breakdown
product is more toxic than the parent active material, then the NOAEL
of the parent compound may be underestimated. Either situation would
result in the establishment of an inaccurate ADI.

Up to now, JMPR has evaluated only the active ingredients of pesti-
cide formulations. The toxicity of "inert ingredients" (e.g., sol-
vents, emulsifiers, preservatives) that may occur as residues in food
has not been considered.

4.2. Principles

4.2.1. Identity

(a) A detailed specification of the test material used in each individ-
ual study must be provided.

(b) Where isomeric mixtures exist, the ratio of isomers in the test
material must be clearly specified, since it has been amply docu-
mented that different isomers frequently have different toxicologi-
cal activities.

(c) JMPR recommendations relate to a specific technical grade of a
pesticide. They will not necessarily be applicable to similar
materials produced by different manufacturers or where specifi-
cations of new material used in the manufacturing process are not
consistent.

4.2.2. Purity

(a) The percentage of the active ingredient in any technical material
used in a toxicity test or proposed for marketing must be speci-
fied.

(b) Levels of all identifiable impurities should be specified.

(c) Data on manufacturing processes may be required to permit determi-
nation of potential impurities. However, because of confidentiality
and industrial secrecy, such data will not be published in JMPR
monographs.

4.2.3. Stability

(a) Stability of the test material during storage and in the diet must
be adequately investigated and reported.

(b) Where instability in diets is observed, possible reaction products
and the nutritional quality of the diet should be investigated.

5. AVAILABILITY AND QUALITY OF DATA

5.1. Background

Most of the data utilized by JMPR consists of unpublished pro-
prietary data, as well as information submitted by governments and
other interested parties. When available, relevant reports from the
open literature are considered. However, published data on major
studies must provide sufficient information to permit evaluation [38;
57]. This precludes reliance on summary or abstract publications. Such
information must include complete descriptions of experimental tech-
niques and data adequate to permit assessment of the validity of the
results [38]. It is preferable that published reports be from refereed
journals. Although all available relevant data are considered [46;
54], unpublished data must meet certain criteria, i.e. reports must be
complete, study supervisors should be qualified to perform the study
and should be identified, and the time at which the study was performed
must be identified [38; 46].

Only those data which are available to all members can be con-
sidered during a Joint Meeting [55]. This requirement applies to all
supporting data and cited material. This has been a subject of ongoing
concern and has been addressed frequently in JMPR reports [53; 54; 55;
57; 59; 60; 62; 65; 67; 70; 78; 172].

On some occasions, important information is omitted from the report
of a study. Examples of such information include the identity and
specifications of test material (section 4), information on the quality
of experimental animal diets, and information on their nutritional
composition.

Present-day standards generally require that data should be subject
to quality control and that the study should conform to the standards
specified under codes of good laboratory practice (GLP) [160; 161].
Studies performed in compliance with GLP codes help assure that the
quality of unpublished data is acceptable. However, compliance with GLP
codes does not provide a substitute for scientific quality. An inappro-
priate study is still unacceptable even though it may have been con-
ducted according to GLP standards.

The validity of data submitted for evaluation has always been of
concern. Recent scandals reported in the scientific literature, in-
volving inaccurate or falsified data [12], highlight this problem of
data validity. However, the application of "good laboratory practice"
and "quality assurance" techniques should reduce, but probably will
not eliminate, the problems of data validation.

JMPR does not have the resources to validate studies [59]. There-
fore, it accepts submitted data as being valid unless there is evidence
to the contrary [57]. The 1977 JMPR [57] was informed of the suspicion
that serious deficiencies existed in several studies that had been
utilized by JMPR in allocating some ADIs and TADIs. In 1982, the
Meeting re-evaluated a number of pesticides that had been supported by

data from Industrial Bio-Test Laboratories. As a general principle,
where studies supporting the ADI could not be validated, and where
alternative studies were unavailable, the ADI was converted to a tem-
porary ADI. Furthermore, if the studies were not validated or replaced,
the ADI was withdrawn.

The format for data presentation requires that a summary of all
pertinent studies be prepared [57], together with reports of each study
with complete supporting data. Complete supporting data are usually
considered to be individual animal data, although occasionally, if GLP
codes have been followed and quality control assurance is available,
this requirement has been waived. The submission of an evaluation of
the compound by the sponsor is encouraged. Since the working language
of JMPR is English [65], translations of reports into English are
appreciated.

5.2. Principles

1. To evaluate the safety of pesticide residues, JMPR is dependent
upon the receipt of acceptable data. Data for major studies should
not be in abstract or summary format and should be of good scien-
tific quality from laboratories utilizing acceptable laboratory
practices [46; 54; 57; 59; 60; 62].

2. Compliance with recognized GLP codes (e.g., those of OECD) is
encouraged.

3. Submitted data should be in such a form that the integrity of the
study can be ascertained.

6. HUMAN DATA

6.1. Background

Human data on pesticides are collected from a variety of sources
including accidental, occupational, and experimental exposures. Data
from experimental exposures of human volunteers can provide quantitat-
ive information on dose-effect and dose-response relationships which
may be applied directly in establishing an ADI. Data on accidental and
occupational exposure can serve as supporting information.

A Joint FAO/WHO Meeting in 1961 highlighted the relevance of human
data for toxicological evaluations and the need to study occupational
exposures during production, handling, and uses of pesticides, since
exposure is generally higher than that of the general population [32].

In 1967, a WHO Scientific Group was convened to provide guidance
for the review of intentional and unintentional food additives. This
group addressed the general problem of investigations in human subjects
and recommended the conduct of human metabolic studies. It was recog-
nized that adequate preliminary tests in animals are necessary before
in vivo human studies can be performed [169]. In addition, studies in
volunteers might be required to confirm the predicted safety margin.
However, several conditions were listed which should be fulfilled
before such studies can be undertaken, including the demonstration of
need and full information on toxicity in experimental animals and the
reversibility of toxic effects. The Scientific Group indicated that
experimental studies on the toxic effects of pesticides in humans are
not acceptable.

At the 1968 and 1969 Joint Meetings [42; 44], it was stated that
the availability of adequate human data might justify the use of lower
safety factors in setting the ADI [46].

The use of modern quantitative analytical toxicology concepts was
introduced at the 1973 JMPR [52], with suggestions of the analysis of
tissue and body fluids for a given pesticide. This is of particular
importance in accidental poisonings. This Joint Meeting also suggested
the follow-up of workers exposed to pesticides. Observations in such
studies may reveal effects specific to humans. The 1975 Joint Meeting
recommended to WHO that cooperation should be sought with Poison
Control Centres and other organizations to develop appropriate data
banks [54].

Data from humans continued to be required in relation to a number
of pesticides until the time of the 1976 JMPR [55]. After that time,
because of ethical problems and the increasing difficulties of per-
forming studies in humans, JMPR reports indicated that data on humans
were "desirable". Since 1982 [67], JMPR has generally, when toxico-
logical assessments have been performed, indicated the desirability of
data on observations in humans. When considering again the need for
comparative biotransformation data, the 1987 JMPR [62] stated that
these might also be obtained with in vitro experiments. It should be

noted that there are limitations in the use of in vitro data, in that
absorption and subsequent distribution as well as possible activation
mechanisms must be considered before extrapolating such data to the in
vivo situation and the subsequent establishment of an ADI.

The 1989 JMPR re-emphasized the necessity of obtaining human data.
It indicated that human data may confirm a common mechanism of toxicity
between humans and the test species or may be used to compare doses and
effects between species [183].

6.2. Current Position

All human data (accidental, occupational, and experimental ex-
posures) are fundamental for the overall toxicological evaluation of
pesticides and their residues in food. Data on accidental poisonings
may identify target organs, dose-effect and dose-response relation-
ships, and the reversibility of toxic effects, provided that modern
standards of analytical toxicology (e.g., identity and purity of the
pesticide, blood levels of the parent compound and/or breakdown prod-
ucts, gastric lavage content, and urinary metabolites) have been
applied to the study. A careful assessment of the dose and perhaps of
the effects (e.g., plasma and erythrocyte cholinesterase inhibition)
may enable comparison with animal data. Unfortunately, available data
rarely permit such comparison. Follow-up studies in workers may enable
the validity of extrapolations from animal data to humans to be con-
firmed and unexpected adverse effects specific to humans to be
detected.

The JMPR mandate is to consider the safety of pesticide residues in
food. Dietary exposure on a daily basis is almost always relatively low
compared to occupational exposure, and therefore it might be expected
that an effect on the exposed worker would be more easily detected.
Unfortunately, there are limitations in attempting to extrapolate
observations in the occupational setting to dietary exposure. The major
route of exposure to pesticides for workers is generally dermal. The
extent and rate of absorption via the dermal route usually differs
markedly from that observed after oral exposure. Ingested compounds may
be metabolized by intestinal microflora and may be subject to metab-
olism within the liver directly after absorption from the gastrointes-
tinal tract and transport by the hepatic portal system. Thus target
tissues may be exposed to a different pattern of parent compound and
metabolites after dermal or inhalation administration than after oral
administration. Data on the identity and levels of parent compound and
metabolite(s), following administration by the different routes, are
desirable to assist in the interpretation of the observed toxic
effects.

When large groups of individuals are exposed to pesticides, epi-
demiological studies can be of considerable value. Often, however,
workers in manufacturing plants and pesticide mixer/loaders and appli-
cators are also exposed to several other compounds and it may be diffi-
cult to determine a cause-effect relationship for a given pesticide.
Results are also often confounded by the difficulty in finding suitable
control populations, the large number of other variables involved, the
long latency period for certain effects such as cancer, and small study

populations, especially in manufacturing facilities. Exposure levels
may also be difficult to quantify. Further guidance in the conduct and
interpretation of epidemiological studies is given in Environmental
Health Criteria 27 [173].

Studies on human volunteers are sometimes of considerable value in
allocating ADIs. However, before human in vivo studies are considered,
ethical considerations must be taken into account. The Proposed
Guidelines on Biomedical Research Involving Human Subjects, issued as a
joint project by WHO and the Council for International Organizations of
Medical Sciences [21], have been endorsed by JMPR.

The value of human data was expressed cogently by Paget [127], when
he wrote:

"The question is not whether or not human subjects should be used in
toxicity experiments but rather whether such chemicals, deemed from animal
toxicity studies to be relatively safe, should be released first to
controlled, carefully monitored groups of human subjects, instead of being
released indiscriminately to large populations with no monitoring and with
little or no opportunity to observe adverse effects."

6.3. Principles

1. The submission of human data, with the aim of establishing dose-
effect and dose-response relationships in humans, is strongly
encouraged.

2. Studies on volunteers are of key relevance for extrapolating animal
data to humans. However, attention to ethical issues is necessary.

3. The use of comparative metabolic data between humans and other
animal species for the purpose of extrapolation is recommended.

7. STRUCTURE-ACTIVITY RELATIONSHIPS

The Joint FAO/WHO Meeting on Principles Governing Consumer Safety
in Relation to Pesticide Residues recognized that toxicological "pro-
cedures must be determined by the chemical and physical properties of
the pesticide . . . " [1, p. 8]. A subsequent WHO Scientific Group
stated that: "If a series of chemical analogues can be shown to give
rise to the same main metabolic product and other compounds which are
already present in the organism in greater quantities, or that can be
readily and safely metabolized, it may be sufficient to carry out toxi-
cological studies on a suitable representative of the series." [169,
p. 7]. The same Meeting, in considering the duration of studies, also
indicated that: "Where adequate biochemical and toxicological data on
closely related compounds are available, the objective becomes the
detection of any deviation from the established pattern" [169, p. 13].
This latter principle has been exemplified by some evaluations of
dithiocarbamate pesticides, where related compounds were considered as
a group.

Structure-activity considerations can influence the testing needs
of a pesticide. Thus the organophosphorus compounds, especially those
with the P-S configuration, are routinely tested for delayed neurotox-
icity, while the majority of other pesticides are not. Similarly,
neurotoxicity is carefully considered in assessing the safety of the
synthetic pyrethroid compounds.

The limitations of the use of structure-activity relationships has
been discussed in the recent document on Principles for the Assessment
of Food Additives and Contaminants in Food:

"Structure-activity relationships appear to provide a reasonably
good basis for predicting toxicity for some categories of compounds,
primarily carcinogens, which are characterized by specific functional
groups (e.g., nitrosamines, carbamates, epoxides, and aromatic amines)
or by structural features and specific atomic arrangements (e.g., poly-
cyclic aromatic hydrocarbons and aflatoxins). However, all these chemi-
cal groups have some members that do not seem to be carcinogenic or are
only weakly so." [176, p. 27-28].

For detailed information on the various chemical classes associated
with carcinogenesis, the reader is referred to published review
articles [146; 178].

7.1. Principle

For the determination of ADIs, JMPR relies primarily on data gener-
ated on individual chemicals. Structure-activity considerations are
used only as ancillary information.

8. TEST METHODOLOGIES

The design and conduct of toxicological investigations has always
been, and still remains, the responsibility of competent experts in the
field. Therefore, the following sections and subsections should be
considered only as guidelines unless stated otherwise.

8.1. Background

The second and fifth reports of JECFA addressed the conduct and
uses of acute, short-term, long-term, biochemical, and carcinogenicity
studies in the safety evaluation of food additives [31; 33]. While many
of the proposals included in these documents have changed with advanc-
ing knowledge in toxicology, some are still deemed to be valid. These
include:

* the need for short-term studies in rodents and non-rodents (defined
as studies comprising repeated doses over a period of up to 10% of
the expected lifespan of the animal, i.e., usually 90 days in rats
and 1 year in dogs);

* the non-requirement for determining LD50 values when no mortality
occurs at doses of 5 g/kg body weight or more;

* the need to initiate short- and long-term studies in young (post-
weaning) animals;

* the need for uniform distribution of the test compound in the diet
when feeding studies are utilized;

* the requirement to use both sexes in acute, short-term, long-term
(chronic), and carcinogenicity studies;

* the need for initiating studies with sufficient animals to ensure
adequate numbers of survivors to provide data for proper statisti-
cal analysis;

* the need to restrict the amount of test compound to less than 10%
of the diet when performing feeding studies (although today it is
generally recommended not to exceed a dietary level of 1% for a
pesticide);

* the need, on a routine basis, for data on absorption, distribution,
and excretion, and, where possible, identification of the major
metabolites;

* investigation of the effects of dose level and duration on the
metabolism of the test material;

* the need to test contaminants in food for carcinogenicity by oral
administration;

* the requirement to maintain an adequate nutritional status of the
test animal in feeding studies, especially in carcinogenicity
studies. Information on the quality and composition of the diets
used in toxicology studies should be provided.

The first JMPR [35] indicated that the biological data required for
allocation of an ADI should include biochemical, acute, short-term
(defined as repeated administration for less than half the lifespan),
and long-term studies. The 1976 JMPR outlined the data necessary for
the evaluation of pesticides. These included short- and long-term
studies, special studies on carcinogenicity, mutagenicity, repro-
duction, and teratology, observations in humans, and information on
metabolism, pharmacokinetics, and biochemical effects [55, p. 8-9].
These are the studies that are now generally available for pesticides
used on food items. Salient aspects of the toxicological tests most
often used in determining the safety of pesticide residues in food are
discussed in the following sections.

8.2. General Considerations

8.2.1. Choice of species and strain

Limitations are inherent in the selection of laboratory animal
species. The most readily available test species are the rat, mouse,
hamster, guinea-pig, rabbit, cat, dog, pig, and monkey. More exotic
animals (e.g., Tupia) are also utilized but only rarely. The major
reasons for the use of such a limited number of species include econ-
omics (cost of obtaining and maintaining animals), lifespan, behaviour
and survival in captivity, handling, and, perhaps most importantly,
knowledge of the "normal" physiology and pathology of the species
(see section 8.2.6).

In 1967, a WHO Scientific Group indicated the need to utilize the
most appropriate species in extrapolating to man, i.e., "the species
most similar to man with regard to its metabolic, biochemical, and
toxicological characteristics in relation to the subject under test"
[169, p. 9]. The choice of an ideal test species requires considerable
knowledge of the absorption and biotransformation of the test material,
not only in the experimental animal species, but also in humans. Unfor-
tunately, other considerations (e.g., cost or availability of test
species, duration of the study) must also be considered, and it is not
always practical to use the optimum test species.

It is necessary to consider both quantitative and qualitative
responses in laboratory animals when establishing the ADI. For example,
it is recognized that compound-induced peroxisome proliferation is con-
siderably greater in mice, rats, and hamsters than in humans [9; 176;
180]. Thus, these species may be inappropriate for investigating this
effect in man. Since specific knowledge of comparative metabolism and
the basis for differences in species sensitivity are often unavailable,
the effects noted in the most sensitive species usually provides the
basis for the ADI assessment.

JMPR, recognizing the difficulties of obtaining in vivo human data,
has proposed as a compromise the generation of in vitro data using
human tissues or cultured human cell lines [78]. Comparison can then
be made (a) between in vitro data generated in a number of species and
(b) between the in vitro and in vivo data in the test species. Such a
procedure would markedly assist in the selection of the most appropri-
ate species for studies involving multiple daily administrations and in
the extrapolation of data. A comparison of this nature for methylene
chloride has recently generated a great deal of interest and has been
proposed for use in safety assessment [3].

The choice of species should also depend upon the susceptibility of
the species (or strain) to the toxic effect being investigated. Thus,
in teratogenic studies, the test species or strain should be known to
be susceptible to teratogenic agents. As new strains of rabbits have
been introduced for teratology studies, JMPR has had to request evi-
dence (from exposure to known teratogens) of their sensitivity and
hence their appropriateness for such studies, e.g., methacrifos [67].
In addition, the time of specific embryological events in different
mouse strains may result in the absence of insult at crucial times in
teratology studies [120].

The normal incidence of a pathological lesion may also influence
the choice of test species or strain. For instance, the use of a strain
in which the incidence of tumours in a particular organ is excessively
high in untreated animals (e.g., the incidence of pituitary tumours in
most strains of rat) would be contra-indicated if there was information
indicating that elements of the endocrine system could be among the
target organs (i.e., if hormonal imbalance were suspected). Similarly,
the high incidence of liver tumours in control male B6C3F₁ mice may
also mask a neoplastic response in treated animals. Thus, a thorough
knowledge of the strain being considered for the study is essential to
determine its suitability for a specific type of experiment.

8.2.2. Group size

Group size in toxicity studies is dependent upon a number of fac-
tors, including the purpose of the experiment, the required sensitivity
of the study, the lifespan of the species under test, the design of the
study, the reproductive capacity and the fertility of the species,
economic aspects, and the availability of animals. This section con-
tains a brief discussion of group sizes acceptable for various toxicity
tests followed by a more detailed discussion of numbers of animals to
be utilized in long term/oncogenicity studies.

In acute oral toxicity tests in rodent species, the number of ani-
mals utilized depends upon the degree of accuracy required. LD50 de-
terminations (as indicated by 95% confidence limits) are approximate,
rather than accurate. To obtain these approximations, five animals per
sex per dose level are usually used. Because of the problems of avail-
ability and because of economic factors involved in utilizing non-
rodent species, smaller numbers of non-rodents (resulting in reduced
accuracy) are frequently utilized in acute toxicity studies, especially
when the objective of the study is to examine the comparative toxicity
between species.

In teratology studies, because the objective is to obtain adequate
numbers of litters from treated females, the actual number of animals
required is dependent on fertility and the difficulties encountered in
breeding. Most protocols for studies with rodent species specify 20-25
pregnant females per dose level. When other species are used, such as
the rabbit, smaller group sizes (usually producing a minimum of 12 lit-
ters) are utilized. However, when equivocal data are obtained from
such studies (e.g., an incidence of congenital malformations, which,
although not statistically significant, shows positive trend analysis),
increased group size or the provision of adequate historical control
data may be necessary.

In multigeneration studies in rats, a minimum of 20 pregnant fe-
males per dose level per mating are usually used. As with teratology
studies, fertility and breeding ability in captivity must be considered
when determining group size at each dose level. In addition, sufficient
litters are required from the mating of the generation that is used as
the source of parent animals for the next generation. Ideally, suf-
ficient litters should be available at each dose level to permit selec-
tion of future parental animals for the next generation on the basis of
1 male and 1 female per litter. Again, this factor must be considered
in initial determinations of group size. This ideal is not always
achievable, since, if some females do not produce offspring, or a
litter contains animals of only one sex, then group size will dimin-
ish as the study proceeds. Under these circumstances, the selection of
parental animals for the next generation should be based on the widest
distribution permissible from the available litters. It should also be
noted that, if closely inbred strains are being used, the distribution
of future parents becomes less critical. The limiting factor in multi-
generation studies is usually the logistics of the study which, since
animals do not mate or deliver to order, become increasingly complex
with each mating and with each generation.

Appropriate group sizes in short-term studies depend upon the pur-
pose of the study. These studies are often designed to provide infor-
mation useful for the selection of dose levels to be used in subsequent
long-term studies. They are, however, sometimes used as the basis for
the ADI. In these cases, increased group size is desirable. The short-
term study utilized for selection of doses in future studies requires a
minimum of 10 animals of each sex per dose level in rodent species.
Smaller group sizes (e.g., 4-6 of each sex per dose level) are gener-
ally accepted for non-rodent species such as the dog.

In considering long-term/oncogenicity studies, the protocol fre-
quently separates the two components of the study. The basic group
size is based on the oncogenicity study, with ancillary groups for
intermediate sacrifices and for investigation of haematological, clini-
cal chemistry, and urinalysis parameters. Group sizes must be suf-
ficient to ensure that adequate numbers of animals survive to the ter-
mination of the study. Furthermore, the study design must be such that
the sensitivity of the study, i.e., its ability to detect an adverse
effect, is acceptable. A recent publication by the International Agency
for Research on Cancer (IARC) [100] has addressed the sensitivity of
carcinogenic studies. Tables 1 and 2, reproduced from this publication,

indicate the numbers of animals of each sex per dose level required to
attain specified sensitivities in a two-dose-level study. (It should be
noted that three dose levels are generally required for safety assess-
ments; see section 8.2.3).

Table 1. Minimum group sizes required to ensure a false-negative
rate of 10% or less^a
--------------------------------------------------------
Excess tumour Tumour incidence in control group
incidence in test ------------------------------------
group (%)^b 0% 1% 5% 10% 20%
--------------------------------------------------------
1 819 2611 9084 16 287 28 110
5 162 243 503 783 1232
10 80 100 166 233 339
15 53 61 90 119 163
20 39 44 59 75 98
25 31 34 43 53 67
--------------------------------------------------------
^a Based on Fischer exact test (p < 0.05) with n animals in each of a con-
trol and a test group, and assuming that all animals respond indepen-
dently.
^b Difference between the response rates in the test and the control
groups, respectively.

As can be seen from these data, test sensitivity is a major factor
in determining group size. Furthermore, these data emphasize the
importance of the background incidence of tumours in untreated animals,
which in turn underlines the importance of species or strain selection
for oncogenicity studies (see section 8.2.1 and 8.2.6).

Group sizes utilized in oncogenicity studies are usually in the
range of 50 to 100 animals of each sex per dose level. For additional
information on group sizes in oncogenicity studies, Annex 2 of refer-
ence 176 should be consulted.

Table 2. Number of animals per group required to obtain false-positive
rates of 5% and false-negative rates of 10% based on tests for
linear trend with three equally spaced doses
-----------------------------------------------------
Tumour response rates
---------------------- Number of
Control Low Dose High Dose animals/group
-----------------------------------------------------
0.02 0.04 0.06 420
0.02 0.07 0.12 112
0.02 0.12 0.22 44
0.10 0.12 0.14 1150
0.10 0.15 0.20 224
0.10 0.20 0.30 70
0.20 0.22 0.24 1860
0.20 0.25 0.30 328
0.20 0.30 0.40 93
-----------------------------------------------------

The size of ancillary groups depends upon the basic study protocol.
The utilization of a procedure for interim kills requires sufficient
animals to be sacrificed at each kill to provide adequate numbers for
histological analysis. Groups designated for haematological, clinical
chemistry, and urine analyses must be of adequate size for proper
statistical analysis of the data that are generated and to allow for
anticipated mortality as the study proceeds. In general, a minimum of
10 animals of each sex per dose level should be available for each sub-
group required.

8.2.3. Selection of dose levels

Data obtained from acute toxicity studies can sometimes assist in
the selection of appropriate dose levels for use in short-term feeding
studies. Thus, when acute toxicity data are available, it is not
unusual for some fraction of the LD₅₀ or of the LD₀₁ determined from
acute toxicity studies to be employed. When available, data on pharma-
cokinetics or metabolism can be helpful in determining dose levels for
short-term toxicity studies, particularly if there is evidence of bio-
accumulation of the test compound or of its metabolites, or if there is
evidence of dose-dependent changes in detoxification. Since the deter-
mination of a dose-response curve is one of the objectives of short-
term studies, at least three dose levels are normally required, as well
as a control.

The selection of dose levels in long-term or oncogenicity studies
should be based on the information derived from pharmacokinetic, phar-
macodynamic, and short-term toxicity studies. Frequently, the highest
dose level selected is the maximum tolerated dose (MTD), estimated from
short-term feeding studies. However, there are problems in attempting
to extrapolate data obtained at high dose levels in experimental ani-
mals to probable human exposure levels. This concept was discussed by
the 1987 JMPR, which made the following statement:

"The Meeting was concerned at (sic) the difficulties of inter-
pretation of the results of long-term studies in which high doses had
been used. In reproduction and teratology studies the use of maternally
toxic doses has also caused concern. The Meeting discussed the maximum
tolerated dose (MTD), which has been defined `as a dose that does not
shorten life expectancy nor produce signs of toxicity other than those
due to cancer' and `operationally, as the maximum dose level at which a
substance induces a decrement in weight gain of no greater than 10% in
a subchronic toxicity test' [176]. To identify agents with particularly
low orders of toxicity, exposure conditions are often maximized. These
may include the use of very high doses and gavage administration. A
number of assumptions are implicit in the use of the MTD: (i) the ab-
sorption, distribution, biotransformation, and excretion of a chemical
are not dose-dependent (that is, their kinetics are the same at low and
high doses); (ii) both the rate and extent of reparative processes (for
example, DNA repair) are independent of dose and of the extent of
damage; (iii) the response to a chemical is not age-dependent; (iv) the
dose-dependent response is linear; (v) doses tested in animals need not
bear any relationship to human exposure levels." [78, p. 3].

At the 1987 JMPR meeting, these assumptions were questioned. Thus:

(i) absorption, distribution, biotransformation, and excretion of a
compound are dependent on several factors, e.g., physicochemical
properties, degree of protein binding, bioavailability, and satu-
ration of routes of metabolism (resulting in variations in the
proportions of different metabolites or complete changes in meta-
bolic pathways with dose (e.g., 2-phenylphenol));

(ii) DNA repair is dependent on dose and/or degree of damage both in
vivo and in vitro [7; 139];

(iii) the response to many chemicals is age-dependent (e.g., acute tox-
icity of DDT or malathion [116]);

(iv) the US NCTR study on 2-acetylaminofluorene (the megamouse study)
did not demonstrate a linear response for bladder tumours [99];

(v) results of studies at dose levels many orders of magnitude above
the level of human exposure to pesticide residues in food have
little relevance to the safety assessment of pesticide residues
in the diet (e.g., 2-phenylphenol [75]).

The JMPR has indicated that, instead of using the MTD to select the
top-dose level, the use of properly designed biotransformation studies
over a range of doses (including human exposure levels) may provide a
more rational basis for dose selection in long-term animal studies.

8.2.4. Test duration

In certain studies, e.g., teratology and multigeneration studies,
the duration of the study is determined by the biological character-
istics of the test species or strain. The duration of these types of
studies is considered in sections 8.3.5.1 and 8.3.5.2.

The duration of other studies is determined, to some extent, by
definition. Thus, an acute study was originally defined as a single-
dose study, observation of the treated animal continuing for 2 to 4
weeks following dosing [37]. The concept of an acute study has changed
slightly through the years; it is now considered to be a study of the
effects of a dose administered either singly or on several occasions
over a period of 24 hours. The observation period is usually 14 days
[124].

A short-term study has been defined as having a duration lasting up
to 10% of the animal's lifespan [31], 90 days in rats and mice, or 1
year in dogs. It has also been defined as a study covering less than
half the animal's lifespan [37].

Long-term/oncogenicity studies are usually defined as studies last-
ing for the greater part of the lifespan of the animal [176, p. 113].
Studies of this type usually fall into one of two categories: (a) a
specific duration; (b) until mortality in the most susceptible group
attains a fixed level, usually 80%. Fixed-term studies vary in duration

with species and strain, depending on lifespan. The late development
of many types of tumours requires that the study be permitted to
continue as long as possible. In addition, reduced liver or kidney
function with increasing age and a consequent increase in the plasma
levels of toxins in older animals may result in manifestations of
toxicity not otherwise seen. However, low survival rates and normal
geriatric changes may complicate study interpretation and limit the
sensitivity of comparison between groups. Thus, the goal of a long-term
oncogenicity study is to determine the optimum balance between these
factors.

The report of the 1967 WHO Scientific Group [169] concluded that it
is better to terminate toxicity studies before the complications of
senescence arise in the test animals. Although many effects of sen-
escence are now recognized, further data are still required before
scientifically supportable generalizations on the duration of long-term
studies are possible. If a finite mortality is the definitive endpoint
of the study, then care must be taken:

* to ensure that mortality does not exceed the predetermined limit in
any group (including the control);

* to consider whether the mortality arises because of tumour develop-
ment;

* that autopsies are performed as soon as possible on animals dying
during the study, thereby avoiding loss of information due to
autolysis or cannibalism.

8.2.5. Pathological procedures

Three steps are involved in the pathological examination of exper-
imental animals:

* gross pathological examination at the time of post-mortem;

* histopathological examination of the tissues;

* a review of these data by an independent pathologist.

For the last of these steps, JMPR has recommended to WHO that a mech-
anism should be established to permit independent pathological assess-
ment of questionable or disputed findings that are brought forward for
review [65].

Pathological examinations and the way in which they are reported
can give rise to a number of problems.

In acute toxicity studies, gross pathological examination of ani-
mals both dying during the study and killed at the termination of the
observation period is desirable, because one of the objectives of an
acute oral study is to obtain information on potential target organs
and on possible dose levels to be used in subsequent repeated adminis-
tration studies. Such information should, therefore, be as comprehen-
sive as possible and should include gross pathology examination. Unfor-
tunately, such examinations are not always performed or reported.

In short- and long-term studies, pathology is a major endpoint.
However, the presentation of pathological data is often confusing.
Gross pathological data (frequently reported separately from the data
on histopathological examinations) are difficult to correlate with
histopathological findings. It is not unusual to find gross pathologi-
cal notations of "lumps and bumps", petechial haemorrhages, etc., in
an organ, for which the histopathological notation is "normal". A
high frequency of such apparent discrepancies in the absence of any
comment is unsatisfactory. The explanation may be either a mix-up in
specimens, or that the sections cut for histopathological examination
failed to intersect a "lump or bump". Either way, the study probably
has not achieved its objectives. Partial resolution of these problems
can sometimes be achieved by cutting multiple sections throughout the
area of the gross lesion.

Pathological terminology is also confusing since several different
names may be used for the same lesion. Therefore, an adequate descrip-
tion of the lesion and an indication of its size and frequency is
essential in pathological reports. Furthermore, a standard classifi-
cation of lesions should always be used in reports, e.g., the Inter-
national Agency for Research on Cancer (IARC) Tumour Register [183].

A high incidence of tissue autolysis is occasionally noted in the
histopathological reports. Even fairly advanced autolysis does not
necessarily preclude the identification of a tumour, despite the fact
that the specific cellular characteristics are obscured by autolytic
activity. Although such tumours cannot always be reported in adequate
detail, their presence can be recorded.

The percentage incidence of tumours is of importance in the evalu-
ation, but data are often such that it is extremely difficult to deter-
mine how many animals were actually examined with respect to a specific
tissue. In the absence of such information, although the number of
diagnosed tumours is known, percentage incidence cannot be determined.

The precise site of tumours may be of major importance. In a recent
evaluation of folpet, tumours in the duodenum and jejunum of the exper-
imental animals were noted and a probable mechanism for the induction
of these tumours was proposed [77]. The data were inadequate to deter-
mine whether these tumours were a "spill-over" (related to the irri-
tant properties of the compound) or whether they were induced indepen-
dently of the postulated mechanism. Additional data were required to
resolve this problem and, hence, to arrive at a valid evaluation of the
safety of the compound [76].

The increasing emphasis on mechanism of action in evaluating tox-
icity studies may be supported by histopathological examinations
utilizing special stains for identification of cell elements (e.g.,
Sudan III for fat droplets) or involving histochemical techniques.
Electron microscopic examination should also be considered when bio-
chemical or other data indicate the need to examine cell organelles or
membrane structures.

Many protocols for multigeneration studies require histopathologi-
cal examination of a representative selection of pups at one or more
points in the study. The need for such examination is questionable
(see section 8.3.5.1).

8.2.6. Historical control data

In almost all toxicity studies, quantitative and qualitative data
from several treated animal groups are compared with data from one or
more concurrent untreated or vehicle-treated control groups. The appli-
cation of appropriate statistical procedures will indicate, with some
predetermined probability, which of the observed differences are not
likely to be attributable to chance. In such procedures, the data from
untreated animals become the standard reference. Yet it is known that,
even with random assignment of individual animals to each group and
strict adherence to GLP, the incidence of spontaneous neoplastic and
other morphologic lesions is often highly variable among control groups
of the same species and strain in different studies conducted within a
single laboratory, as well as in different laboratories [119; 152; 157;
167; 168].

To be indicative of a treatment-induced change, the differences
between control and treatment groups should show a dose-response
relationship and delineate a trend away from the expected norm for the
particular species and strain of experimental animal used. Since data
from the concurrent control group are used as the standard reference
for treatment group responses, and since control data in any particular
study may be unpredictably variable, qualitative and quantitative cri-
teria must be used to evaluate whether the concurrent control data con-
stitute the typical species/strain pattern, i.e. whether they corre-
spond to an expected norm. Historical control data relating to the
specific species/strain used in the study provides such evaluation
criteria [23; 126; 154; 155]. This type of information must be viewed
as an auxiliary aid to interpretation of data from the study. It should
not be used as a complete substitution for concurrent control data.

The following have been proposed for use in the evaluation of car-
cinogenicity data by a Task Force of Past Presidents of the Society Of
Toxicology [155] and may have utility for the evaluation of other forms
of toxicity as well:

* If the incidence rate or other observed effect in the concurrent
control group is lower than in the historical control groups but
these same effects in the treated groups are within the historical
control range, the differences between treated and control groups
are not biologically relevant.

* If the incidence rates or other observed effects in the treated
groups are higher than the historical control range but not stat-
istically significantly greater than the concurrent control inci-
dence, the conclusion would be that there is no relation to treat-
ment (but with the reservation that this result could have arisen
by chance or because of flaws in the assay and may therefore be a
false negative).

* If the incidence rates or other observed effects in the treated
groups are significantly greater than in the concurrent controls
and greater than the historical control range, a treatment effect
is probably present which is unlikely to be a false positive
result.

The best historical control data are obtained using the same
species and strain, from the same supplier, maintained under the same
routine conditions in the same laboratory that generated the study data
being evaluated. The data should be from control animals from contem-
poraneous studies. Statistical procedures can be used to relate the
overall historical incidence to that in a specific study. However, this
leaves much to be desired since the incidence of spontaneous lesions
and the averages of quantitative data can vary considerably between
groups of animals. This type of variation is not apparent if the inci-
dence in combined historical control animals is used [157].

To assess variability, historical control data should be presented
as discrete group incidences, segregated by sex and age and updated
with each new study that is performed [135]. It is also highly desir-
able that additional information on each discrete control group be made
available. This information should include the following:

* identification of species, strain, name of the supplier, and
specific colony identification if the supplier has more than one
geographical location;

* name of the laboratory and time during which the study was per-
formed;

* description of general conditions under which the animals were
maintained, including the type or brand of diet and, where poss-
ible, the amount consumed;

* the approximate age, in days, of the control animals at the begin-
ning of the study and at the time of killing or death;

* description of the control group mortality pattern observed during
or at the end of the study and of any other pertinent observations
(e.g., diseases, infections);

* name of the pathology laboratory and the examining pathologist who
was responsible for gathering and interpreting the pathological
data from the study;

* what tumours may have been combined to produce any of the incidence
data.

8.3. Conduct and Evaluation of Different Types of Studies

8.3.1. Short-term and long-term toxicity studies

Both short- and long-term feeding studies utilize the same method-
ologies and differ only in the duration of the test. The parameters
investigated usually include body and organ weights, clinical chemis-
try and haematological effects, and gross and histopathological exam-
inations.

Short- and long-term toxicity studies are designed to determine the
NOAEL for the test substance and to provide information relevant to the
determination of the safety factor to be applied in extrapolating to
humans (see section 2.2).

The majority of protocols available for toxicity testing are
intended as guidelines, thus leaving the final study design to the
individual investigator. It is usually the case that by the time the
long-term toxicity studies are initiated, the investigator will have
access to the information from earlier studies (acute, short-term, and
metabolic studies) and hence will be able to judge the most suitable
design for long-term studies.

The selection of species and of dose levels have been discussed in
sections 8.2.1 and 8.2.3.

In long-term oral toxicity studies, the test substance is normally
incorporated in the diet and administered for the majority of the life-
time (see section 8.2.4), on a daily basis (7 days per week). Lifetime
exposure is required due to the fact that, during the aging process,
factors such as altered tissue sensitivity, changing metabolic and
physiological capability, and spontaneous disease may alter the nature
of the toxic response [171]. Spontaneous diseases include age-related
increases in the incidence of heart disease, chronic renal failure, and
neoplasia, which are observed in most mammalian species.

To ensure that the objectives of the long-term toxicity study are
achieved, statistical principles must be used to determine adequate
group sizes for reducing the incidence of false positives and false
negatives to a minimum (section 8.2.2). Similarly, the use of random
numbers or comparable statistical techniques, both for allocating ani-
mals to experimental groups and for ensuring that the distribution of
cages of animals within housing racks is random, is essential to mini-
mize bias in selecting animals and minimize possible environmental
effects (e.g., temperature, humidity, light) within the animal house
[176, Annex II].

In conducting these studies, the principles of GLP [161; 160]
should be followed to ensure both acceptable conditions of animal hus-
bandry and adequate conduct of the experiment. Full records on all
animals must be kept, detailing all observations, results of any lab-
oratory techniques (e.g., bleeding and subsequent haematological or
clinical chemistry studies), and information on pathological examin-
ations at the end of the study.

Since one of the objectives of a feeding study is to determine
changes in toxic signs and manifestations, it is axiomatic that
periodic detailed examinations be performed on at least a proportion of
the experimental animals. Non-invasive procedures such as the measure-
ment of body weight and food consumption, palpation, behavioural
observations, and assessment of general condition of the experimental
animals (both control and exposed) can be performed regularly. The
frequency of handling may be limited by the potential for creating
stress in the experimental animal, particularly if the frequency is
increased towards termination of the study (i.e. in oncogenicity
studies). Urinalysis, the remaining routine non-invasive technique,
should also be performed regularly, but at longer time intervals
(usually at 3, 6, 12, 18, and 24 months in rat studies). The process of
collecting urine may cause stress, depending upon the type of caging
used. Thus, if animals are housed one per cage, the use of metabolism
cages for single animals will induce minimal stress. However, where

multiple caging is the norm, sudden isolation can induce a stress
condition, with consequent physiological changes in the animal. This
should be considered when the results of urinalysis are interpreted.

In general, urinalysis utilizes insufficient animals (often as few
as five of each sex per dose level) or an insufficient acclimatization
period in the metabolism cage(s) to be very useful, since variability,
even in the same individual, can be high [150]. Even if the numbers of
animals or acclimatization time is adequate, further problems may be
encountered. Dissolved carbon dioxide may dissipate and thus alter pH,
the appearance of the urine may vary according to the time of day at
which sampling takes place, and bacterial concentration and composition
may change even if preservatives are used. However, useful data can be
obtained in clinical chemistry studies on urine, such as concentrations
of proteins, ketones (elevated in starvation or with low carbohydrate
diets), glucose (diabetes, hypoglycaemia), and porphyrins (elevated
with liver disorders), osmolality (reflecting kidney function, but data
on water consumption is needed to interpret kidney concentration ef-
fects), urinary haemoglobin (often elevated in toxic situations), and
high crystal content (possibly predictive of kidney or bladder stones).
In addition, periodic urine collection and analysis for metabolites of
the test substance may yield data on age-related changes in metab-
olism.

Invasive techniques (usually involving blood sampling) normally
utilize a pre-designated ancillary group of animals identified for that
purpose prior to the onset of the study. Thus the effects of repeated
bleeding at specific intervals (the intervals usually being similar to
those delineated for urinalysis) on terminal pathological manifes-
tations are recognizable in animals in the ancillary group. The ancil-
lary groups (which must allow for mortality with increasing duration of
the study) should comprise at least 12 animals of each sex per dose
level for each group to provide groups of at least 10 animals of each
sex per dose level for haematological and other clinical chemistry
examinations.

End-points normally measured in haematological examinations include
erythrocyte counts, leucocyte counts, differential leucocyte counts,
haemoglobin, haematocrit, and platelet and reticulocyte counts. In
addition, erythrocyte fragility, sedimentation rate, and coagulation
factors are frequently measured and bone marrow is examined.

End-points traditionally examined by clinical chemistry measure-
ments include:

* serum bilirubin (liver and haematological effects);

* serum glucose;

* lactate dehydrogenase (a non-specific indicator of tissue damage
seen in myocardial infarction, renal toxicity, pulmonary embolism,
and pernicious anaemia);

* serum alkaline phosphatase (which, it should be noted, decreases
with age and with nutritional status, and cannot be regarded as
specifically indicative of a disease process because of its wide
distribution in many organs);

* alanine aminotransferase (previously serum glutamic-pyruvic trans-
aminase) and aspartate aminotransferase (previously serum glutamic-
oxalic transaminase) (both indicators of liver toxicity);

* amylase (increased in renal insufficiency and pancreatitis, de-
creased with hepatobiliary toxicity);

* creatinine (renal failure);

* creatinine phosphorylase (elevated with myocardial infarction and
lung disorders);

* cholinesterase (decreased by organophosphates and carbamates);

* serum protein;

* blood urea nitrogen (elevated with renal toxicity, depressed with
liver toxicity);

* serum electrolytes (see reference [93] for a comprehensive dis-
cussion of the interpretation of clinical chemistry measurements).

It has been proposed that clinical chemistry studies be aimed
mainly at known target organs that are identified in short-term tox-
icity studies [150]. However, long-term toxicity studies may result in
changes in the degree of toxicity to specific organs (e.g., adaptation
of initial target organs, secondary effects arising from the initial
effects noted in short-term studies, and changes in circulating enzyme
or hormone levels due to tumour development). Consequently, limiting
clinical chemistry studies to parameters suggested by short-term
studies is not encouraged.

In certain cases, clinical chemistry studies may be necessary to
investigate endocrine organ function. For example, delayed growth or
metabolic dysfunction may be the result of thyroid dysfunction, induced
either by direct toxic action of the test material on the thyroid or by
decreased thyrotropin release by the pituitary. Similarly, altered
liver carbohydrate metabolism may be due to pancreatic dysfunction,
adrenal dysfunction may result in disturbed kidney function, changes in
fertility or reproductive performance may be mediated by gonadal hor-
monal changes, and tumour formation may arise due to enhanced hormonal
stimulation, either in endocrine organs or in non-endocrine organs
(e.g., the mammary gland). A recent publication [163] discusses the
practical problems and describes methods of investigating endocrine
toxicity.

While clinical chemistry data are often non-specific, they do
permit the progress of an effect to be followed in vivo. When histo-
pathological data are available (usually only at the times of interim
and terminal sacrifices), they may supersede clinical chemistry
findings.

The pathological data derived from feeding studies are of paramount
importance. Such data in long-term feeding studies should be obtained
from at least two specified sacrifice periods, one (usually a minimum
of 10 rats of each sex per dose) at a point in time prior to the onset
of senescence and the second at termination of the study. All animals

(including all non-scheduled deaths, or animals sacrificed in a mori-
bund condition) should be examined at least grossly, and tissues should
be preserved where possible for histological examination. To avoid
undue loss of tissues due to autolysis, animals should be checked at
least 2 or 3 times daily. A high incidence of autolyzed animals results
in loss of data and raises concerns about the quality of the animal
husbandry and standard of laboratory expertise.

Histopathological examination should cover a wide range of organs
and tissues. However, recognizing the economics of histopathological
examinations, examination of tissues from mid- and low-dose groups may
be limited to those tissues where differences occurred between those
from control and high-dose groups.

The NOAEL is frequently based on the results of the pathological
examination of the test animals. The initial (gross) examination notes
any abnormalities in the tissues (e.g., masses, discoloration,
necrosis). This is followed by removal and weighing of specific organs.
Because of the high rate of autolysis of some organs (e.g., the kid-
ney), removal, weighing, and preservation should be performed as rap-
idly as is consistent with accurate work. Paired organs should be
weighed separately to avoid inaccuracies arising from unilateral
lesions (e.g., tumours) that are not grossly visible. Organs normally
weighed include the liver, kidneys, heart, adrenals, gonads, spleen,
and brain. Results of such weighings should be reported as absolute
weights, and also as a ratio to body weight and to brain weight.

In assessing data from short- and long-term toxicity studies, the
following factors should be considered:

* Comparison of mean values of body weights for specific groups of
animals may not necessarily be the most appropriate method of
detecting potentially toxic effects. The use of body weight gain
differences should also be considered, as should changes in food
intake.

* Clinical chemistry data can provide a useful indicator of toxico-
logical effects. However, they are limited in sensitivity and frank
pathological changes are often observed at dose levels less than or
equal to those resulting in significant clinical chemistry effects.
When studies include a post-treatment recovery period, clinical
chemistry data are frequently of value in assessing the progress of
recovery. In many cases, the specificity of the test system, e.g.,
serum alkaline phosphatase, is insufficient to permit precise
identification of target tissues or organs. In other cases, e.g.,
acetylcholinesterase measurements, clinical chemistry data may be
the major toxicological effect measured.

* Changes in a single haematological parameter unsupported by further
changes in other haematological parameters or by pathological
changes in bone marrow or spleen are rarely of toxicological sig-
nificance.

* Organ weight changes should always be examined on an absolute and
organ/body weight ratio basis. Organ/body weight ratios can be

misleading when a change in body weight occurs. Mathematical pro-
cedures for correcting for this situation exist. When the body
weight per se is affected, there is a tendency to place greater
reliance on organ/brain weight ratios.

* Gross and histopathological examinations should be carefully
checked for correspondence. A detailed description of the lesion(s)
or photomicrographs may be necessary since the terminology used for
certain lesions is variable and there is some degree of subjec-
tivity in the interpretation of lesions (see also section 8.2.5).

No discussion of toxicity studies would be complete without some
consideration of the dose actually ingested. Dose-level selection is
discussed in section 8.2.3 and stability of the test material in the
diet in section 4. Assuming that the stability is acceptable and that
the homogeneity of the test material in the diet has been measured on a
number of occasions during the study, one major variable remains, i.e.
the food consumption per unit of body weight. This varies with age,
being highest in the young animal and decreasing as the animal ages
(the special case of lactating females is discussed in section
8.3.3.1). When data are available, the actual dose ingested is calcu-
lated from the concentration of test substance in the diet and the food
consumption. Under these circumstances, the JMPR evaluation indicates
the NOAEL as X ppm equal to Y mg/kg body weight per day, and is usually
based on the mean intake of the test substance over the lifespan. In
other cases, when the calculation of intake in mg/kg body weight per
day is not feasible because of inadequate food intake data, the JMPR
evaluation uses the standard conversion factors for ppm to mg/kg body
weight per day ([114], reproduced as Annex II in this monograph), this
being reported as X ppm equivalent to Y mg/kg body weight per day. The
former method is preferable.

8.3.2. Carcinogenicity studies

From its inception, JMPR has recognized the need to evaluate the
carcinogenicity of pesticide residues in food [32]. It has adopted the
principle that carcinogenicity testing should utilize adequate numbers
of animals, generally of two or more species (e.g., rat and mouse), and
a suitably high dose level of the substance should be fed for the life-
time of the animals [33].

8.3.2.1 Background

The 1977 Joint Meeting noted that an evaluation of carcinogenicity
should be undertaken routinely for:

* pesticides whose use results in substantial residues in crops di-
rectly or indirectly used for human food;

* pesticides with structural similarity to known carcinogens;

* pesticides that are metabolized to, or leave residues that are,
known carcinogens or closely related to such compounds;

* pesticides that give rise to early pathology suggestive of poten-
tial tumorigenicity;

* pesticides with pharmacokinetic properties "suggestive of covalent
binding to tissues" or bioaccumulation [57].

JMPR has sometimes recommended that certain compounds should not be
used where residues may occur in food, due to their potential carcino-
genicity (e.g., hexachlorobenzene, captafol). At other times, either
TADIs or ADIs have been set, even though there was limited evidence of
carcinogenicity in animals (e.g., several chlorinated organic insecti-
cides). Overall, JMPR has maintained the philosophy that a pesticide
for which there is limited evidence of carcinogenicity should not
necessarily be prohibited (see section 8.3.2.7).

8.3.2.2 Routes of exposure

The oral route of administration is the most appropriate one for
determining in experimental animals the carcinogenic potential of
pesticides leaving residues in food.

The 1966 JMPR [38] noted the comments of the report of a WHO Scien-
tific Group [169] concerning experimentally induced local sarcomas that
apparently result from the physical characteristics of the test
material. This Joint Meeting concluded that for the routine testing of
pesticide residues, the subcutaneous route is not generally appropri-
ate. The occurrence of local sarcomas following subcutaneous injection
should not alone be considered sufficient evidence of a carcinogenic
hazard following ingestion [169]. It does, however, indicate that
further studies would be desirable.

The 1989 JMPR noted that severe local effects may interfere with
the interpretation of data, e.g., the production of forestomach epi-
thelial hyperplasia and papilloma formation following the adminis-
tration of gastric irritants. It was recommended that methods of
administration other than feeding be justified.

8.3.2.3 Commonly occurring tumours and factors influencing tumour
incidence in different species

Some rodents commonly used for in vivo bioassays exhibit high inci-
dences of some tumours. In evaluating toxicological data, it is import-
ant to determine whether an increased incidence of tumours and/or a
decreased time to tumour in exposed animals are related to treatment.
The incidence of such tumours in control animals may vary considerably
with time. As an example, ten years ago the occurrence of Leydig cell
tumours in rats was rarely reported. By 1987, some laboratories were
reporting that the occurrence of such tumours sometimes reached 50% in
control rats. It is not known whether this change in incidence is due
to a genetic shift in certain rat strains or to more careful pathologi-
cal examinations of the rat testes. The importance of such factors is
discussed in section 8.2.6.

As noted in Environmental Health Criteria 70 [176]:

"The evaluation of studies in which commonly-occurring tumours are
a complicating factor needs careful individual assessment. The tumours
that have given rise to the most controversy in recent years are hepa-
tomas (particularly in the mouse), pheochromocytomas in the rat (see
below), lymphomas and lung adenomas in the mouse, pancreatic adenomas

and gastric papillomas in the rat, and certain endocrine-associated
tumours, including pituitary, mammary, and thyroid tumours in both rats
and mice. Some of these tumours, such as hepatomas and lung adenomas,
may occur in the majority of untreated animals.

"With the exception of lymphomas, some of which are virally
associated, the endocrine-associated tumours, and possibly hepatomas in
high-incidence strains of mice, which may involve oncogenes [82], there
is no clue as to the origin of tumours that occur commonly in exper-
imentally-used rodents. Indeed, there is not even any cogent specu-
lation as to the mechanisms by which these tumour incidences are
increased." [176, p. 44].

Since the publication of Environmental Health Criteria 70, a great
deal of additional research has been carried out on the etiology of
cancer, particularly with respect to the important role of oncogenes in
neoplasia. Nevertheless, additional investigation into the initiation,
promotion, and progression of cancer is necessary to assist the incor-
poration of such mechanistic considerations into human hazard assess-
ment for carcinogens.

JMPR has generally considered it unwise to classify a compound as a
carcinogen solely on the basis of an increased incidence of tumours of
a kind that commonly occur spontaneously in the species and strain
under study and at a frequency that may seriously reduce the statisti-
cal power of the study. Data are usually required in one or preferably
two alternate species, and the overall evidence is then considered.

The significance of mouse liver tumours was first considered by the
1970 JMPR [46]. These tumours were then becoming more frequently
observed in carcinogenicity studies, especially following exposure to
the chlorinated organic pesticides. Subsequent Joint Meetings [46; 48;
52; 53; 57; 72; 74] have also considered the problem of pesticides that
induce mouse hepatic tumours. A number of hypotheses concerning the
etiology of mouse liver tumours have been considered by JMPR [2; 132;
133; 162; 170]. Biochemical differences between the mouse and many
other species, including humans, are highly pertinent to mouse hepa-
tomas [72]. In addition, degranulation of endoplasmic reticulum is
known to be associated with carcinogenesis in the mouse [130; 131;
134]. Both dieldrin and phenobarbitone degranulate the hepatic endo-
plasmic reticulum of CF1 mice, a strain susceptible to dieldrin-induced
tumorigenesis, but do not degranulate the endoplasmic reticulum of LACG
mice, a non-susceptible strain, nor that of rats or humans. The current
position of JMPR is that mouse liver tumours are of little relevance in
predicting human cancer risk. It is inadvisable to classify a substance
as likely to be a carcinogen to humans solely on the basis of an
increased incidence of mouse liver tumours [72].

Other tumours occurring with a high relative frequency are adrenal
medullary lesions in rats. As noted in Environmental Health Criteria 70
[176]:

"An overview of the literature indicates that untreated rats of
various strains may exhibit widely differing incidences of lesions
described as `pheochromocytomas' [69; 141; 142]. There are no clear
criteria for distinguishing between prominent foci of hyperplasia and

benign neoplasms, and pathologists differ in the criteria that they use
for distinguishing between benign and malignant adrenal medullary
tumours.

"Rats fed ad libitum on highly nutritious diets tend to develop a
wide variety of neoplasms, particularly of the endocrine glands, in
much higher incidences than animals provided with enough food to meet
their nutritional needs but not enough to render them obese. The
adrenal medulla is just one of the sites affected by overfeeding. Con-
trolled feeding . . . reduces the life-time expectation of developing
either hyperplasia or neoplasia of the adrenal medulla in rats." [176,
p. 44].

Thus, food intake can be a major factor in experimental carcino-
genesis. Restricted food intake in rodents is known to increase life
expectancy and to reduce the incidence of naturally occurring and some
induced tumours. However, restricted dietary intake may also require
other considerations (e.g., study duration) be taken into account in
designing protocols for carcinogenicity studies.

8.3.2.4 Pathological classification of neoplasms

The need for guidelines leading to consistency in pathological
diagnosis is apparent. As noted in Environmental Health Criteria 70
[176], tumours should be classified and analyzed on the basis of their
histogenic origin in order to prevent different malignant tumours,
occurring in the same organ, from being grouped inappropriately for
statistical analysis. This is particularly important when brain tumour
incidences are being considered, since different tumour types are fre-
quently, but incorrectly, grouped together for analysis.

Accurate determination of histogenic origin is clearly important in
determining the significance of benign tumours, since this is often a
complicating factor in assessing carcinogenicity studies. As noted in
Environmental Health Criteria 70:

"If benign and malignant tumours are observed in an animal tissue
and there is evidence of progression from the benign to the malignant
state, then it is appropriate to combine the tumour types before per-
forming statistical analysis. It is, however, still advisable to exam-
ine incidences of benign and malignant tumours separately. Assessment
of the relative numbers of benign and malignant tumours at the various
dose levels in the study can help determine the dose response of the
animal to the compound under test. On the other hand, if only benign
tumours are observed and there is no indication that they progress to
malignancy, then, in most cases, it is not appropriate to consider the
compound to be a frank carcinogen, under the conditions of the test
(this finding may suggest further study)." [176, pp. 44-45].

The 1983 JMPR [70] indicated possible approaches (e.g., interim
sacrifice of satellite groups, morphometric measurement of tumours) to
the problem of latency, which is an important component of the evalu-
ation of carcinogenic potential.

8.3.2.5 Evaluation of carcinogenicity studies

Various classification schemes have been proposed for potential
chemical carcinogens. For example, IARC Working Groups evaluate evi-
dence on the carcinogenicity of agents in humans and describe them in
standard terms of "sufficient", "limited", or "inadequate" evi-
dence of carcinogenicity or "evidence suggesting lack of carcino-
genicity". These categories refer only to the strength of the evidence
that an agent is carcinogenic and not to the extent of its carcinogenic
activity (potency) nor to the mechanisms involved. Finally the total
body of evidence (including, where relevant, supporting evidence of
carcinogenicity from other data such as genetic and related effects)
from humans and experimental systems is taken into account and an agent
is categorized into one of four groups [101]:

Group 1: carcinogenic to humans,

Group 2A: probably carcinogenic to humans,

Group 2B: possibly carcinogenic to humans,

Group 3: not classifiable as to carcinogenicity to humans, and

Group 4: probably not carcinogenic to humans.

JMPR considers, where possible, both carcinogenic potency and bio-
logical relevance in its evaluations. It does not utilize a classifi-
cation system for carcinogenic pesticides, preferring to evaluate com-
pounds on a case-by-case basis, rather than allocating a compound to
"the best fit" position in existing classification systems.

8.3.2.6 Extrapolation from animals to man

Different approaches to the extrapolation of animal carcinogenicity
data to humans have been utilized. One of these approaches relies on a
knowledge of the comparative metabolism in the test species and in
humans. If data are available indicating that a crucial metabolic path-
way is overloaded, an increase in tumour incidence occurring only at
dose levels exceeding those resulting in the overload, then confidence
in the NOAEL is increased. If comparative metabolic data indicate a
similar situation in humans, the task of extrapolation is simplified.

Another approach is based on pathological considerations. When data
are available to demonstrate a fixed pattern of tumour development
(e.g., progression from hyperplasia, through nodular hyperplasia and
benign tumour, to malignant tumour), then a dose level below that
resulting in the initial pathological change is unlikely to be carcino-
genic (see also section 13.5).

In 1969, JMPR [44] urged the consideration of dose-response
relationships and possible NOAELs for carcinogens. The 1974 JMPR [53]
adopted several of the principles put forth by a WHO Scientific Group
[170] concerning preliminary changes such as hyperplasia, the effects
of hormonal compounds, and tumours apparently induced by the physical

character of the carcinogen. The 1974 Joint Meeting noted that prelimi-
nary changes such as hyperplasia are associated with a number of car-
cinogenic compounds. Furthermore, some chemicals apparently give rise
to neoplasms only after the induction of a particular pathological
effect [19].

The 1983 JMPR recognized that most of the mechanisms of chemical
carcinogenesis were not fully understood. In view of the uncertainty
surrounding the use of various mathematical models for carcinogenicity
assessment, the Meeting decided that the use of safety factors remained
a reasonable approach. It also recognized the importance of taking
into account all biological activities of such agents in arriving at a
safety assessment. This pragmatic approach is used by JMPR in the
absence of satisfactory alternatives (see section 9.2).

In determining the acceptable level of pesticide residues for
humans, the safety factor utilized reflects the confidence in the data
base and the degree of concern for the toxic effect. This is especially
true for carcinogenic effects. Where there is the need for a very high
safety factor due to concern about the safety of the pesticide, it may
be prudent to recommend that the pesticide should not be used where
residues in food may occur.

8.3.2.7 Principles

1. An evaluation of carcinogenicity should be undertaken for those
pesticides that:

* may give rise to substantial residues in crops used directly
or indirectly for human food;

* have a chemical structure similar to known carcinogens or give
rise to metabolites or residues that are known carcinogens or
closely related compounds;

* give rise to histological changes that are suggestive of po-
tential neoplasia.

2. The oral route of administration to experimental animals is the
most appropriate route for determining the carcinogenicity of
pesticide residues in food.

3. All available data should be considered in the evaluation and as-
sessment of carcinogenic activity.

4. A pesticide for which there is limited evidence of carcinogenicity
in animals should not necessarily be prohibited for use.

5. Mechanistic considerations are of major importance in the extrapol-
ation of animal carcinogenicity data to humans.

8.3.3. Reproduction studies

Multigeneration reproduction studies and teratology studies are
routinely required for pesticides. Although experimental designs exist
that combine teratology studies with reproduction studies, these two
types of study will be considered separately in this monograph.

8.3.3.1 Multigeneration reproduction studies

The 1961 FAO/WHO Meeting on Consumer Safety in Relation to Pesti-
cide Residues stated that one of the aims of toxicological investi-
gations of a pesticide is to ascertain "the amount of pesticide to
which man and farm animals can be exposed daily for a lifetime" [32,
p. 10]. With respect to the effect of age on toxicity, a WHO Scien-
tific Group stated: "In general, but not invariably, the young animal
is more sensitive to the toxic effects of exposure to chemicals" [169,
p. 10]. It also pointed out the effects of different gut flora and
changes in enzymes with age (e.g., poorly developed mixed-function
oxidase enzymes in newborn rodents). The Group indicated that "perti-
nent information observed from reproductive (multigeneration) studies
provides some assurance on the safety of compounds which might be
present in the diet of babies" [169, p. 12] and concluded that "use-
ful information may be obtained from studies in newborn or young ani-
mals, from reproduction studies and biochemical studies" [169, p. 23].
It also indicated the need for further studies on "the development of
enzyme systems in the human young, with particular emphasis on those
enzymes responsible for dealing with foreign chemicals" [169, p. 25].

JMPR addressed the problem of toxicity to juveniles indirectly in
1963 when it stated in its report that "the Meeting considered that
foods, such as milk, which figure largely in the diets of babies and
invalids, should be essentially free from pesticide residues" [35,
p. 6]. However, it was not until 1976 that JMPR indicated that repro-
duction studies should be available as part of the basic toxicology
data package required for allocating an ADI [55]. The need for such
studies was repeated in subsequent Meetings [59; 60; 62; 65; 67; 70;
72; 74]. It should be noted, however, that multigeneration studies on a
number of compounds had been submitted and evaluated before that time,
some as early as 1963 (e.g., aldrin, dieldrin, heptachlor epoxide).

In evaluating a multigeneration study, there is a tendency to focus
on the conceptus, the neonate, and the immature animal, because of the
known variations in toxicity in these stages of development compared to
those observed in adult animals. It must also be recognized that pro-
found physical, physiological, and psychological changes occur during
pregnancy, which may affect the susceptibility of the dam to the tox-
icity of a specific chemical. Attention must therefore be given to
maternal toxicity during pregnancy and lactation.

A number of basic protocols for the conduct of multigeneration
studies have been developed [92; 125; 159; 162; 174]. None, however,
have gained unanimous approval and proposals for alternatives continue
to be suggested [16; 112; 128; 129].

The multigeneration study may best be viewed as a screening test
for toxicity in reproducing animals because, although the emphasis is
on detecting effects specific to reproduction, it is also useful for
detecting the enhancement of general toxic effects that may occur as a
consequence of physiological changes associated with reproduction and
development.

The major asset of a multigeneration study that is well designed,
conducted, and interpreted is that it has the ability to detect a wide
range of indirect or direct effects on reproduction. This ability
arises from the complex integration of reproductive processes, so that
minimal effects that may be difficult to demonstrate in isolation may
combine and cascade to generate a more notable deviation in a more
distal end-point (e.g., litter weight). Observations in the premating
period provide a setting for assessing subsequent observations; initial
observations during mating can identify lack of libido or a disturbance
of hormone (oestrous) cycles. Subsequent data are generated to indicate
effects on fertility, fecundity, prenatal toxicity, parturition,
lactation, weaning, and postnatal growth and development of offspring
through puberty to maturity. However, those features that enhance the
ability of the study to detect an effect have the disadvantage of
making it difficult to ascertain the primary cause when an effect is
obtained. Where multigeneration studies provide an indication of an
effect on reproduction, it is usually advisable, or even mandatory, to
perform follow-up studies for further elucidation. Recent proposals
[112; 128; 129] seek to alleviate this limitation of protocols cur-
rently in use by allowing flexibility of operation once an effect has
been detected or is suspected. A wide range of options is available for
follow-up studies, including separate male and female studies, use of
the three segment designs applied in drug testing, and use of the
dominant lethal assay as a male fertility study.

A number of factors in the experimental design of multigeneration
studies have been, or are, points of controversy. The following
examples may be cited:

(a) The duration of the pre-treatment period of the first generation
(F₀) has been the subject of much discussion. A period equal to
one spermatic cycle plus epididymal transit time is generally used
for males and a period of five estrous cycles is advised for
females. A period of 100 days prior to pairing was originally pro-
posed. However, in some rat strains, such a prolonged treatment
period results in the test animals having passed peak reproductive
capacity by the time mating is initiated. At present, a 70-day pre-
mating treatment period is generally used. If two breeding gener-
ations are employed, the problem becomes largely academic, since
the second (F₁) generation cannot reproduce until it reaches
maturity and it will have been exposed continuously throughout
development to sexual maturity.

(b) The need for second litters in each generation has also been a sub-
ject of controversy. Two recent studies [20; 113] indicate that
second litters are more sensitive, with respect to certain par-
ameters, than are first litters. However, although the sensitivity
of the second litters is increased in some areas, there are no
recorded cases where effects been observed that were not present in
the first litters. Thus, provided adequate dose levels are utilized
and no adverse effects are recorded in the first litter, a second
litter should not be necessary. Exceptions to this generalization
apply to studies in which findings in the first litter are equivo-
cal. They also apply when compounds with long biological half-lives
are being tested and plateau levels have not been attained at the
time of first mating.

(c) Litter size in multigeneration studies is often "standardized"
(i.e. culled, usually to eight pups on day 4 post partum). Culling
may introduce bias and reduce sensitivity. Surveys of routine
studies show that some supposed advantages of culling (e.g., pre-
vention of high mortality in large litters and reduced variability
of pup weight) are more imagined than real. If first litters are
culled and it is suspected that this may have masked the detection
of an effect, then production of a second unculled litter in the
same generation is recommended.

(d) The requirement for histopathological examination of pups at
weaning has also been a point of discussion. The majority of histo-
pathological changes will normally be similar to those seen in
routine short- or long-term studies. The need for histological
examination should be on a case-by-case basis, depending upon the
results of the other available studies and gross observations.

(e) It has been proposed that, since the food intake in the female rat
during lactation may be as much as 2.5-fold higher than in the non-
lactating female rat, it would be reasonable to reduce the level of
test substance in the diet during this period, in order to give a
more constant exposure in terms of mg/kg body weight. However, this
is not considered advisable as a routine procedure because the
study would no longer model the human situation, in which maternal
exposure to pesticide residues increases during lactation. Such a
procedure adds complexity to the study and complicates extrapol-
ation of the results to humans.

The evaluation of the data from the multigeneration study starts
with a scan of the entire study for effects and then focuses in more
detail on specific areas, bearing in mind the following points:

(a) Data from the premating period (and from other toxicity studies)
provide the baseline information against which effects on repro-
duction per se are compared. These data provide a check on whether
the exposures were too high or too low or whether the interval
between dosages were too wide to establish dose-related responses
of any effect. The better the baseline information, the more
reliable the judgements on subsequent reproductive effects.

(b) Data from the initiation of mating to parturition provide, with
respect to adults, information on libido, precoital time, fer-
tility, fecundity, duration of gestation, parturition and toxicity
to the pregnant female. With respect to the offspring, litter
values (size, number of live pups, and pup weight) at birth provide
information that would indicate prenatal toxicity.

(c) Data from birth to weaning provide information on the potential
susceptibility of the lactating female to the test substance and
its effect on her nursing ability. Pup weight and survival allow
the assessment of effects on postnatal growth, toxins transmitted
via the milk, and development of the offspring. Effects seen during
this period could be delayed responses to earlier (prenatal) insult
or a combination of post- and pre-natal effects.

(d) Data on the offspring during the period from weaning to puberty
provide information on the persistence or permanence of earlier
effects and on direct effects on the still-immature animal. The
profound changes associated with puberty provide a stress point for
the detection of delayed manifestation of earlier, latent effects
or enhancement of direct effects.

Comparison of results obtained from the two parental generations
may be informative. The F₀ parental animals (first parental gener-
ation) have not been exposed to the test material in utero during
lactation or during early post-weaning development, whereas the F₁
parents (second parental generation), have been exposed throughout
development. Consequently, if an effect is observed (on fertility,
libido, parental body weight, or general condition), comparison between
the F₀ and F₁ parental animal data may yield useful information on
the time at which the effect is initiated. Thus, because oocyte devel-
opment is completed in the female prior to birth, adverse effects on
fertility observed only in female F₁ parents would suggest that an
area for further investigation would be the in utero oogenesis. Simi-
larly, disturbance in the development of male hormonal systems in utero
may be produced (e.g., fenarimol [99]), resulting in reduced libido of
the F₁ males.

In considering fertility, the protocol chosen often has a marked
influence on the ability to determine which sex is involved. Protocols
vary with regard to methods of pairing (e.g., one male to one female,
one male to two females), duration of pairing (1-3 oestrus cycles), use
of replacement males in non-successful pairings, follow-up of appar-
ently infertile males, and use of proven males. Cross-matings of
untreated males with treated females and vice versa may be required to
ascertain the sex of the infertile partner. Once this is determined,
histopathological examination of the reproductive organs may yield
information indicating the type of effect. Studies may also be per-
formed on circulating hormone levels. Further details on these types
of studies are given in reference [166].

The initial data on newborn pups are usually limited to the number
of pups born alive. Data on stillbirths and the number of malformed
pups may be inaccurate because of cannibalism. Thus, although a multi-
generation study may give indications of high prenatal losses and
developmental toxicity, it cannot be considered to be a definitive
teratology study. Often the only indicator of prenatal developmental
toxicity in the reproduction study is reduced litter size at the time
of the first observation, usually several hours after parturition.
However, if dose levels administered in the multigeneration study are
sufficiently high, then the lack of any effect provides some reassur-
ance regarding potential teratogenicity. Furthermore, the continuous
exposure to the test substance over a long period of time in the multi-
generation feeding study may lead to changes in metabolism. It may also
lead to changes in the dose of parent compound or metabolite reaching
the placenta or fetus or to higher blood plasma levels in the case of
chemicals with long half-lives. This can lead to divergence, in either
direction, of the results of multigeneration and prenatal toxicity
(teratogenicity) studies.

The rate of growth and survival of post-partum pups may be affected
by a number of factors including general maternal care, effects
initiated in utero, reduced lactation by the mother, or the presence of
toxicants in the milk. When the need to determine the cause of pup mor-
tality or reduced pup weight gain arises, the initial step is usually
histopathological examination of pups failing to survive. If lactation
has been affected (either in terms of quantity or quality of the milk),
the normal routine for investigation is cross-fostering of pups, in
which pups from treated mothers are weaned by untreated maternal ani-
mals and vice versa.

It needs to be borne in mind that the sensitivity of the repro-
duction study is low for specific end-points. This is particularly true
for discrete end-points such as infertility and total litter loss.
Where such end-points are concerned, more sensitive indicators must be
found or the dimensions of the study must be greatly increased. For
example, if male infertility is suspected, studies of sperm motility,
mobility, and morphology may be undertaken. In the case of male-
mediated reproductive toxicity, the commonly used multigeneration study
design is particularly insensitive and specialized studies are necess-
ary if male fertility is believed to be effected. Sperm measurements
are now being conducted in conjunction with some short- and long-term
studies and this may partially alleviate this problem.

In addition, a commonly available, and sometimes neglected, source
of supporting information may be provided by histopathological examin-
ation of the reproductive organs (after proper fixation) in the chronic
toxicity studies.

8.3.3.2 Teratology studies

In 1967, the WHO Scientific Group on Procedures for Investigating
Intentional and Unintentional Food Additives [169] stated that "at
present, no specific tests can be recommended for the detection of
teratogens, but some safeguard can be provided by multigeneration
studies" [169, p. 24]. In 1976 [57] and more recently [59; 60; 62; 64;
65; 67; 70; 72; 74], JMPR has stated that teratology studies should be
an integral part of the toxicology data base required for evaluation
and for allocation of the ADI.

The basic teratology study (also defined by IPCS as an embryo/feto-
toxicity study) involves the treatment of the pregnant animal through-
out the period of organogenesis. Since this begins at or around implan-
tation of the blastocyst into the endometrium, pre-implantation losses
are not usually of concern. However, an "apparent" pre-implantation
loss could be a failure to detect blastocyst implantation losses.

The route of exposure, in teratology as in other studies, can mark-
edly influence results. The most frequent routes of administration for
pesticides are diet, drinking water, or gavage. The latter, however,
may result in marked differences in kinetics following the bolus admin-
istration of a high dose relative to more frequent intakes of small
amounts. Thus benomyl is teratogenic when administered by gavage but
not when administered via the diet [70]. This is believed to be due to
the short-term high plasma levels resulting from gavage administration,
compared to the much lower sustained levels which result from dietary

administration. Effects following gavage administration are not always
more severe than those resulting from dietary inclusion. For example,
thalidomide administered to rats by gavage provides essentially nega-
tive results whereas administration in the diet in a reproduction study
induces almost complete embryolethality. Comparative pharmacokinetic
studies are useful and often essential for relating the findings in
teratology studies to human dietary exposure.

The species most commonly used in teratology studies are the rat,
the mouse, and the rabbit. More than one species is generally utilized
in attempting to assess teratogenic potential because of the varia-
bility in species sensitivity. Species differences arise because of
variations in metabolism, types of placentation, and in the rates and
patterns of fetal development. As with other toxicity studies, more
weight should be given to results in species that give the closest
approximation to humans in terms of kinetics, dynamics, and other
relevant parameters.

The choice of dose levels in teratology studies has recently become
a point of major concern. In several publications [108; 109; 110], it
has been stated that maternal toxicity is associated with species-
specific patterns of malformations. These associations have often led
to false presumptions of cause and effect and, further, to the presump-
tion or implication that embryonic effects associated with maternal
toxicity are unimportant. However, in such associations it is more
probable that effects on the embryo and dam are independent or mutually
interactive. In practical terms the conceptus and dam are indivisible
and are best considered as a unit. A presumed and even proven cause
and effect relationship provides only an explanation of the mechanism
of action, it does not necessarily preclude the risk. For example,
effects induced by alcohol, lead, or methylmercury show that, even
though these effects occur at doses that induce maternal toxicity, they
remain of relevance for making decisions regarding safety. In consider-
ing the choice of the highest dose level in teratology studies, it is
important to note that: (a) maternal toxicity can and does occur with-
out inducing malformations, and (b) malformations can occur without
maternal toxicity being induced. Thus, at the present state of knowl-
edge, it may be prudent to continue to utilize high dose levels which
induce minimal maternal toxicity. Further research regarding the role
of maternal stress in the induction of developmental toxicity is
recommended.

In interpreting the significance of malformations and other struc-
tural variants, it is important to consider the stage of development of
the fetus at examination. Under routine experimental conditions, the
offspring are removed from the mother 12 to 24 hours before anticipated
parturition to avoid the possibility of cannibalization. However, the
accuracy of estimating the age of the offspring at the time of removal
is questionable, since vaginal smears are normally taken only once per
day, thereby reducing the accuracy of the estimation of the onset of
pregnancy. Furthermore, delays in the rate of development may occur.
For most malformations this is relatively unimportant. The incidence
of minor variants (e.g., ossification variants) may, however, be mark-
edly altered, especially if the compound affects the rate of develop-
ment.

Dose-related minor changes should not be ignored, since they are of
considerable value in assessing whether a low incidence of malfor-
mation is compound-related or coincidental. The association between
changes in the pattern of minor anomalies and malformations has been
amply illustrated in the past. However, considerable variability exists
among laboratories in both the reporting and the assessment of these
minor structural deficiencies, which renders interpretation of some
studies extremely difficult. A consistently higher standard of
reporting of minor anomalies is encouraged. Minor anomalies are not
necessarily of great concern, however, in the absence of other manifes-
tations of developmental toxicity.

Hydroureter and hydronephrosis are frequently associated with
delayed opening of the ureter at the point of entry into the bladder,
with a subsequent hydrostatic effect. Even when the incidence of these
conditions is high in pre-partum fetuses, they may not be apparent in
4-day-old post-partum pups [179]. Further research is encouraged in the
development of protocols for the postnatal assessment of developmental
toxicity.

8.3.3.3 Screening studies in teratology

Studies in which non-mammalian species or mammalian organs and
tissue cultures are used to attempt to predict teratogenicity in mam-
malian systems are not generally of value in safety assessments at
present. Although some tests may be useful as preliminary screens to
prioritize compounds for further investigation, none of the available
techniques can be considered definitive studies. These techniques,
however, are of value in follow-up studies to determine the mechanism
of action of compounds demonstrating positive effects in standard in
vivo studies, or as support studies.

Further discussion of screening teratology studies will be found in
references [91] and [174].

8.3.3.4 Principles

1. Tests in reproducing animals are essential for the complete safety
evaluation of a pesticide.

2. A well designed and conducted multigeneration study providing no
evidence that the pesticide exerts a selective effect on repro-
duction or enhancement of general toxic effects should be given a
high weighting towards establishing its safety.

3. The detection of effects in a multigeneration study may require
further studies if the protocol has a limited ability for charac-
terizing a specific effect.

4. The limited sensitivity of some end-points of a reproduction study
needs to be borne in mind. Discrete responses such as pregnant or
non-pregnant, "fertile" or "infertile" are quite insensitive,
since objective discrimination between groups is governed by the
same laws of statistical probability as are applicable to other low
frequency events such as malformations.

5. Support studies such as examination of sperm motility and mor-
phology may provide a sensitive end-point that can allow further
characterization of effects observed in reproduction studies.

6. Histopathological examination of gonads performed in chronic tox-
icity studies may also provide valuable supplementary information.

8.3.4. Neurotoxicity studies

8.3.4.1 Delayed neurotoxicity

JMPR first reviewed feeding studies in hens, with subsequent exam-
ination of brain, spinal cord and sciatic nerves, in 1967 when
dimethoate was evaluated [41]. In 1968, the first study using single
doses in hens was reviewed, when dioxathion was evaluated [43]. The
question of delayed neurotoxicity was first considered in 1974 [53] in
response to requests from various countries for guidance on the intro-
duction and use of the organophosphate leptophos. This was considered
further at the 1975 JMPR, the report stating:

"A major toxicological problem long recognized to be associated
with such organophosphate esters as tri- O -cresylphosphate (TOCP), and
more recently brought to the attention of the Meeting in the evaluation
of leptophos is that known commonly as `delayed neurotoxicity' . . .
The delayed neurotoxicity syndrome affects only certain animal species,
including man. The most susceptible animal for laboratory bioassay
procedures, the adult hen, is not susceptible before 3-4 months of age.
While the adult hen is the animal of choice for laboratory testing,
cats, dogs, calves, and sheep have been shown to be susceptible. Some
sub-human primates and rodents are resistant to both the clinical and
the histological lesions. In contrast, man has been shown to be highly
susceptible to the syndrome, as suggested by studies where occurrences
of paralysis have been reported . . . There are no known antidotes to
delayed neurotoxicity, and recovery from ataxia is predominantly
through development of collateral nerve pathways and physical therapy
to develop muscles not served by affected nerves." [54, p. 11-12].

A certain degree of peripheral nerve regeneration also occurs, but
regeneration is not observed in the central nervous system (CNS) axons.
Therefore, the ataxia is clinically "irreversible" although the pic-
ture changes from a flaccid paralysis (peripheral nerve plus CNS
lesions) to a spastic paralysis (CNS lesions only).

Reference has been made in some studies [14; 22; 104] to the induc-
tion of neurotoxicity by certain organophosphorus compounds used as
pesticides and drugs. The dose administered in most experimental
studies is high, and atropine has been used to protect animals from
acute signs of poisoning to allow time for the neurotoxicity syndrome
to develop. While atropine protects against the short-term acute
cholinergic signs of poisoning,it is ineffective against delayed neuro-
toxicity occurring 8-14 days after treatment.

A key factor in the problem of delayed neurotoxicity, discussed by
the 1975 Joint Meeting and endorsed by the 1976 [55] and 1978 [59]
Meetings, is the dose response. "The Meeting concluded that delayed

neurotoxicity appears to follow a dose-response relationship and that
it is therefore possible to estimate a no-effect level following acute
or chronic exposure in a susceptible species. With an adequate margin
of safety an ADI for man can be allocated with a sufficient degree of
assurance as far as pesticide residues in food are concerned." [54,
p. 13].

The 1982 JMPR, in discussing acceptable protocols for pesticide
toxicology studies, indicated that "multiple dosing (single oral
doses, 21 days apart) was required for studies of delayed neurotoxicity
of organophosphorus compounds." [67, p. 2].

The 1984 Meeting noted that:

"Some OPs (organophosphates) induce both acute reversible and
delayed irreversible neurotoxicity; the latter relates to inhibition of
another enzyme called neuropathy target esterase (NTE) [115]. Organo-
phosphorus-delayed neurotoxicity is believed to be initiated by a two-
step mechanism: a high level of NTE inhibition and `aging' of the phos-
phoryl enzyme complex [105]."

"Inhibition of NTE within 24-48 hours after dosing correlates with
the clinical and morphological effects of delayed neurotoxicity seen
10-20 days later. This test model was found to be valid for all OPs
known to cause delayed neuropathy in man." [147].

The JMPR recommended that delayed neurotoxicity testing need not be
done for monomethylcarbamates, phosphinates, or sulfonates. They also
recommended that TOCP be used as a positive control only for OPs and
that the NTE assay be included in the assessment of OPs [72].

The most recent comments on delayed neurotoxicity relate to the
optical isomers of organophosphorus esters. The 1987 JMPR stated:

"Recent evidence suggests that when racemic mixtures of phos-
phonates are used in test animals, the optical isomers might show the
same phosphonylating ability for NTE but the rates of aging might
differ. Consequently only the optical isomer which forms an ageable
protein-phosphonyl complex will cause delayed polyneuropathy. This was
the case for EPN, an OP no longer in use. Therefore, whenever OPs are
mixtures of optical isomers the delayed neurotoxic potential might
depend on the chirality." [78, p. 11].

Two types of studies are generally conducted on chemicals suspected
of being neurotoxic. The first is the use of a suitable sensitive
species (usually the adult hen), where test substance is administered
at two acute exposures (separated by 21 days) to atropine-protected
animals at a level at or above the LD₅₀ of the compound. Observations
on body weight, ataxia, and signs of delayed neurotoxicity are made
while the animals are alive. At termination, usually 42 days after the
first dose, histopathological examination of the brain, spinal cord and
proximal and distal sections of (usually) the sciatic nerve is per-
formed. Data from this type of test suffer from two major drawbacks:
the evaluation is often subjective, and a negative result cannot be
graded. The second type of test is the determination of NTE activity
[13; 105]. In its simplest form, this involves treatment of the adult

hen with a single maximum tolerated dose of the test substance and
subsequent assay of the brain enzyme after the time of peak inhibition
but before substantial re-synthesis of new enzyme has occurred. The
time of peak inhibition, which can be from 3 to 48 h post-dosing (and
is determined by the pharmacokinetics of the compound), can often be
assessed by observation of the time of onset of cholinergic signs. The
threshold level of NTE inhibition at this early stage, which correlates
with delayed neurotoxicity, is approximately 80%. No clinical signs are
associated with an inhibition of 60% or less. When multiple determi-
nations of NTE are made during chronic exposures, plateau levels are
observed in 2-3 weeks. If inhibition of NTE in the brain and spinal
cord is less than 50%, delayed neuropathy does not occur. However,
inhibition of 60-70% in such studies might result in neuropathic
sequelae as reported by some authors, while others state that the same
threshold of NTE inhibition (70-80%) has to be reached in single and
repeated exposures.

8.3.4.2 Acute neurotoxicity (acetylcholinesterase inhibition)

In 1967, the WHO Scientific Group on Procedures for Investigating
Intentional and Unintentional Food Additives [169] noted that plasma
and erythrocyte cholinesterase activities were markedly reduced by
organophosphorus and carbamate pesticides. This Group also noted the
absence of a correlation between blood cholinesterase levels and the
signs and symptoms of toxicity. Thus cholinesterase levels in blood
"may be useful as an indication of exposure to a substance with anti-
cholinesterase activity, but not as an invariable guide to the degree
of intoxication present or predicted" [169, p. 17-18]. The Group
indicated that "although changes in blood cholinesterase levels may
be helpful in toxicological studies, it is important that further
research should be done to relate the indices used as closely as
possible to the biochemical changes concerned in bringing about the
toxic effects . . . " [169, p. 18].

Cholinesterase-inhibiting compounds have been evaluated at vir-
tually every Joint Meeting. Until 1982, JMPR used inhibition of plasma
cholinesterase, as well as erythrocyte and brain cholinesterase, for
the purpose of establishing NOELs. In 1982, the status of cholinester-
ase activity as an indicator of anticholinesterase compound toxicity
was reconsidered:

"In reviewing some organophosphorus and carbamate pesticides, the
Meeting noted that previous JMPR reports have commented on, and made
recommendations on the basis of, inhibition of plasma cholinesterase as
a major criterion in the evaluation of some of these compounds. The
present Meeting recognized that most organophosphorus compounds inhibit
butyrylcholinesterase, known also as plasma cholinesterase or pseudo-
cholinesterase, at concentrations lower than those at which they in-
hibit acetylcholinesterase found in erythrocytes and in nerve synap-
ses.

"The function of plasma cholinesterase is not understood but it is
known that it plays no role in cholinergic transmission, the physio-
logical function which is impaired by anticholinesterases. On the other
hand, acetylcholinesterase in erythrocytes, although playing no role in
cholinergic transmission itself, reflects the acetylcholinesterase

activity in nerve synapses, since the two enzymes are considered bio-
chemically identical. Therefore, erythrocyte cholinesterase activity
may be taken as an indicator of the biochemical effect of anticholin-
esterase pesticides." [67, p. 6].

A biologically significant reduction in erythrocyte cholinesterase
is normally considered to be a reduction of >20% of pretest levels in
the same animals in short-duration studies, or in concurrent controls
in longer studies.

The 1988 Joint Meeting further considered the utility of plasma
cholinesterase and erythrocyte and brain acetylcholinesterase measure-
ments [79]. It noted that "the correlation between acetylcholinester-
ase inhibition in erythrocytes and in the nervous system is usually
unknown" and indicated that "data on brain acetylcholinesterase
inhibition are considered to be of greater value than those on erythro-
cytes in assessing the cholinergic effects of cholinesterases." The
Meeting also noted, however, that in the absence of measurements of
brain acetylcholinesterase, those of erythrocyte acetylcholinesterase
serve as a better indicator of toxicity than those of plasma cholin-
esterase activity. It was noted that in vitro kinetic studies may be
necessary for pesticides with anti-esterase activity. Results of these
studies in different species may be combined with in vivo study
findings to establish ADIs for these compounds.

JMPR has drawn attention to the methodology for measuring cholin-
esterase inhibition, stating that "the currently used methods for the
determination of cholinesterase activity may lead to erroneous con-
clusions when applied to rapidly reversible cholinesterase inhibitions
(e.g., N -methyl- and N,N -dimethylcarbamates). in vitro kinetic studies
should be made to elucidate the nature of reversible inhibition reac-
tions. The results obtained in in vivo studies should be interpreted
cautiously until more satisfactory methods are available." [55,
p. 11].

In 1983 [70], the problem of measurement of cholinesterase inhi-
bition by carbamate pesticides was again addressed by JMPR. The report
of this Meeting states "the Meeting noted that in the reports of
several studies on carbamate pesticides, the method of determination of
cholinesterase inhibition was inadequately reported and occasionally
data were inconsistent with respect to dose and the degree of cholin-
esterase inhibition. Carbamates are considered to be reversible inhibi-
tors of cholinesterase with a short duration of action. Because of the
reversible inhibition of the enzyme by dilution, as would occur during
the preparation of the assay, inhibition cannot be accurately measured.
The Meeting stressed that in order to permit evaluation of cholinester-
ase inhibition by carbamates in vivo, special care is required in
reporting all details of such studies." [70, p. 10]. Carbamate cholin-
esterase inhibition studies should utilize minimal dilution during the
preparation of the assay, minimal incubation times and minimal times
between blood sampling and assay (e.g., the Ellman method [28]).

8.3.4.3 Chronic neurotoxicity

The 1972 JMPR [50] noted the work of Murphy & Cheever [144], which
reported modification of the electroencephalographic patterns in cer-
tain experimental animals following long-term exposure to low levels of
cholinesterase-inhibiting compounds. The Meeting indicated that
"insufficient information was available to permit any conclusion to be
reached on the relationship of these studies to the toxicological
assessment of cholinesterase-inhibiting compounds." [50, p. 8].

The 1974 Meeting [53] reiterated the desirability of determining
the usefulness of electroencephalographic criteria for assessing the
effects of cholinesterase-inhibiting pesticides. However, no further
information or verification of this aspect of cholinesterase-inhibiting
pesticide toxicity has become available to JMPR.

8.3.4.4 Pyrethroid-induced neurotoxicity

JMPR has evaluated data on pyrethroids during many meetings since
1965 [39; 47; 51; 61; 63; 66; 182]. Most pyrethroids can be divided
into two classes: the T-syndrome (tremor) and the CS-syndrome
(coreoathetosis-seizures). In general, alpha-cyanopyrethroids cause
CS-syndrome neurotoxic effects, and other pyrethroids cause T-syndrome
effects. The 1984 Meeting [73] noted that the neurotoxicity of
pyrethroids originates from their primary action on the sodium channels
of nerve membranes [123]. This interaction is reversible, as are the
clinical signs of toxicity.

Morphological changes in peripheral nerves are produced as a
secondary effect of the primary interaction only at doses close to the
LD₅₀. Therefore, considering the reversibility of pyrethroid neuro-
toxicity and the high doses required to cause permanent secondary
effects, the neurotoxicity of pyrethroids is not considered to be of
great concern in the evaluation of pesticide residues in food.

8.3.4.5 Neurobehavioural toxicity

A recent WHO publication [175] on the Principles and Methods for
the Assessment of Neurotoxicity Associated with Exposure to Chemicals
stated the following:

"There is ample evidence of real and potential hazards of environ-
mental chemicals for nervous system function. Changes or disturbances
in central nervous function, many times manifest by vague complaints
and alterations in behaviour, reflect on the quality of life; however,
they have not yet received attention. Neurotoxicological assessment is
therefore an important area for toxicological research. It has become
evident, particularly in the last decade, that low-level exposure to
certain toxic agents can produce deleterious neural effects that may be
discovered only when appropriate procedures are used. While there are
still episodes of large-scale poisoning, concern has shifted to the
more subtle deficits that reduce functioning of the nervous system in
less obvious, but still important ways, so that intelligence, memory,
emotion, and other complex neural functions are affected. Information
on neurobehaviour, neurochemistry, neurophysiology, neuroendocrinology,

and neuropathology is vital for understanding the mechanisms of neuro-
toxicity. One of the major objectives of a multifaceted approach to
toxicological studies is to understand effects across all levels of
neural organization. Such a multifaceted approach is necessary for
confirmation that the nervous system is the target organ for the
effect. Interdisciplinary studies are also necessary to understand the
significance of any behavioural changes observed and thus to aid in
extrapolation to human beings by providing specific neurotoxic pro-
files. Concomitant measurements at different levels of neural organiz-
ation can improve the validity of results."

Since the publication of this monograph, a number of protocols for
neurobehavioural toxicity have been proposed for use [11]. However,
the 1989 JMPR noted that the use of behavioural tests in laboratory
animals has not been validated [183]. The meeting concluded: "This
failure relates both to the inter-individual and intra-individual vari-
ations in behaviour and the difficulty in quantifying these changes.
In addition, the biochemistry, electrophysiological and morphological
correlates of observed changes are often lacking." Although much has
been written on behavioural teratology [158; 159], no data on this
aspect of toxicology has been reviewed by JMPR. A discussion of the
utility of these tests will be found in reference [174].

8.3.4.6 Principles

1. Delayed neurotoxicity appears to follow a dose-response relation-
ship. Thus, with an adequate margin of safety an ADI can be allo-
cated.

2. Delayed neurotoxicity testing should be conducted routinely for
organophosphates. However, it need not be done for monomethyl-
carbamates, phosphinates, or sulfonates.

3. TOCP is recommended as a positive control substance only for
organophosphates.

4. The NTE assay should be included in the data base for organophos-
phate evaluations.

5. Data on brain acetylcholinesterase are of greater value in safety
assessment than are data on erythrocyte acetylcholinesterase.

6. Plasma cholinesterase (butyrylcholinesterase) inhibition is not
considered to be an adverse toxicological effect.

8.3.5. Genotoxicity studies

The topic of mutagenicity (now generally referred to by the broader
term, genotoxicity) and its relevance to the evaluation of the safety
of pesticide residues has been repeatedly considered by JMPR. Most
recently, the 1983 Meeting recognized the uncertainty of the associ-
ation between mutagenic and carcinogenic activity, and indicated that
data from long-term carcinogenicity studies must override any possible
concerns raised by mutagenicity studies. In considering mutagenicity
tests per se, the 1983 JMPR was unable to determine the relevance of

the results of such tests to possible human health hazards. It there-
fore indicated such data cannot be utilized directly in the assessment
of the ADI [70].

A recent publication [5] surveying 222 chemicals tested in mice and
rats (NCI/NTP bioassays) has indicated a strong association between
structure/activity, mutagenicity in Salmonella strains, and the extent
and sites of rodent tumourigenicity. When structure/activity and
Salmonella tests were considered and utilized as an index of genotox-
icity, the use of such an index indicated two groups of carcinogens:
those that are genotoxic and those that are apparently non-genotoxic.
In examining sites of action, some 16 tissues were susceptible to car-
cinogenic effects with genotoxins only (accounting for 31% of the indi-
vidual chemical/tissue reports), whereas the remaining 13 tissues were
affected by both groups of carcinogens, the most frequently affected
tissue being the mouse liver (24% of all individual chemical/tissue
reports). Furthermore, chemicals active as carcinogens in both rats
and mice, or in two or more tissues, showed a 70% correlation with
positive Salmonella tests, whereas single species or single tissue
carcinogens showed only 39% correlation. The study also confirmed that
many in vitro genotoxins were not carcinogenic (possibly due to malab-
sorption, metabolism in vivo, or the supposedly greater sensitivity of
the in vitro tests). Mouse liver-specific carcinogens were also
Salmonella positive in only 30% of the cases, indicating that mouse
liver tumour induction may be mechanistically independent of inter-
action of the test chemical with DNA.

These results support the position that rodent carcinogenicity
tests are required for all pesticide evaluations (see section 8.3.4.1),
since without such studies it cannot be determined that a pesticide is
a trans-species, multiple-tissue rodent carcinogen.

8.3.5.1 Principles

1. Mutagenicity is utilized only as supplementary information in the
weight-of-the-evidence determination for carcinogenicity.

2. Mutagenicity tests, especially mammalian in vivo tests, which are
indicative of compound-induced alterations in DNA are of value in
assisting in the determination of the mechanism of action of some
carcinogens.

3. Genotoxicity testing is also potentially useful in the prediction
of the risk of heritable defects.

4. Protocols that are sensitive, practical, and predictive of heri-
table human risk remain to be developed.

8.3.6. Immunotoxicity studies

8.3.6.1 Background

In 1967, progressive haemolytic anaemia was observed in monkeys
exposed to dieldrin [41]. However, it was not recognized at the time
that this anaemia resulted from antibodies produced in the animal which
were directed against dieldrin bound to the erythrocytes [89]. The 1976

JMPR "noted the first observation in a group of animals of a pesticide
(pirimicarb) causing a haemolytical reaction which might be of an
immuno-reactive nature. In the case observed, the phenomenon occurred
only with relatively high doses in a closed, inbred colony of dogs.
However, it is possible that, by prolonged and constant use of such a
pesticide, hypersensitivity may be built up which could eventually lead
to an immunological reaction of a haematological or other nature."
[55, p. 14]. In 1978, JMPR [59] again considered pirimicarb, and noted
that haemolytic changes occurred in a second strain of dogs but not in
monkeys or in rodent species. The effect was therefore considered to
be species specific.

8.3.6.2 Current position

Immunotoxicology has been defined as the discipline concerned with
the study of events that can lead to undesired effects as a result of
the interaction of test substances with the immune system. These unde-
sired effects may be a consequence of:

* direct and/or indirect action of the test substance (and/or its
biotransformation product) on the immune system;

* an immunologically-based host response to the compound and/or its
metabolites;

* host antigens modified by the compound or its metabolite(s).

Zbinden [181] has indicated that chemicals may affect the immune
system immediately and preferentially, but they may also act either by
injury to other organs or by creating a general deterioration of the
health of the animal, resulting in a secondary effect on the immune
system. Consequently, as with any aspect of toxicology, immunotoxi-
cology must be considered in the light of all available toxicity data
and not as an entity independent of other factors.

In mammals the primary lymphoid tissues comprise the thymus,
spleen, lymph nodes, bone marrow and diffuse lymphoid tissues associ-
ated with the gastrointestinal and respiratory systems [117; 165]. Pro-
genitor cells produced in the bone marrow and other lymphoid tissues
undergo maturation in early life via residence in the thymus to produce
the T-cell series (which are mainly responsible for cell-mediated
immunity), and via development in peripheral lymphoid tissues to become
members of the B-cell series, which form the basis of humoral (anti-
body-mediated) immunity. Throughout life, the development of immune
reactions and defenses involves interactions between several types of
T- and B-cells and soluble factors produced by early stages of these
cells, phagocytic cells, and polymorphs.

Chemically-induced immune alterations may be detectable from patho-
logical changes (quantitative and qualitative) in lymphoid organs.
Thus, changes in the weight of the thymus, spleen, and lymph nodes,
combined with histopathological changes in these organs can be import-
ant in assessing the potential immunotoxicity of a chemical. Further-
more, examination of mucosa-associated lymphoid tissue (e.g., Peyer's
patches) may indicate immunotoxic potential. Examination of bone marrow
is essential in any immunotoxic assessment, as is consideration of the
resistance to infection of the living animal.

Atrophy and lymphocytic depletion in the thymic cortex, hypoplasia
or hypercellularity of the paracortical areas of the lymph node,
changes in the numbers of lymphoid follicles, changes in germinal
centres and plasma cells in lymph nodes and the spleen, and the size
and cellularity of the marginal zone of the spleen may all be indica-
tive of immunotoxicity. However, other factors also induce some of
these effects (e.g., thymic atrophy due to stress or weight loss)
[165].

Haematological studies of serial blood samples for total and dif-
ferential leucocyte counts and platelet numbers can provide a potential
indicator of certain autoimmune processes. Similarly, measurements of
body temperature and serum chemistry to determine cortisol and fibrino-
gen levels may suggest consequences of certain types of immunotoxicity
[121].

The recognition that an increased tumour incidence (especially
lymphomas) can be associated with immunosuppression indicates that the
immune system may be involved in controlling neoplastic changes. This
involvement is supported by in vivo evidence of tumour immunogenicity
(e.g., transplant rejection; lymphoid cell transfer experiments), by
the promising use of monoclonal antibodies as therapeutic agents in
cancer therapy, and by many laboratory demonstrations of cellular and
humoral responses to neoplasms.

A number of agents (e.g., tricothecene mycotoxins), known to occur
as contaminants in food, can be shown to affect the immune system of
laboratory animals. These mycotoxins (nivalenol, deoxynivalenol, etc.),
which are unaffected by heating or baking, occur on cereal crops grown
in temperate climates. Information on the potential of pesticide resi-
dues to interact with such immunosuppressive agents would be of value
in the safety assessment of pesticides.

It is becoming apparent that immune dysfunctions induced by test
substances sometimes have severe and diverse health effects ranging
from autoimmune diseases or hypersensitivity reactions to the possible
induction of cancer. In the past, this area has received little atten-
tion because of the lack of basic knowledge of suitable test methods.
The complexity of the mechanisms of action of the immune system makes
it difficult to decide on appropriate studies. Some potential probably
exists for the general identification of immunotoxicants from standard
toxicological protocols, but full identification of immunotoxicity is
likely to require further ancillary studies. The development of
additional methods relevant to the safety assessment of pesticide resi-
dues is to be encouraged in the hope that sets of tests, suitably vali-
dated, will permit evaluation of this important aspect of toxicology.
A collaborative study, sponsored by the IPCS and CEC, is currently
underway to examine and validate test methodologies for the assessment
of immunotoxicity.

8.3.6.3 Principles

1. Immune dysfunctions induced by test substances can result in
serious health effects and should be considered in the evaluation
of pesticide residues in food.

2. Validation of a tiered approach to immunotoxicity tests relevant to
safety assessment is to be encouraged.

8.3.7. Absorption, distribution, metabolism, and excretion

8.3.7.1 Background

The meeting held in 1961 to consider Principles Governing Consumer
Safety in Relation to Pesticide Residues [32] indicated that the pro-
cedures to be followed in generating data for the safety evaluation of
a pesticide "must be determined by . . . its toxicological and bio-
chemical actions, as they are discovered during the progress of the
investigation." The report also cited the second and fifth JECFA
reports [31; 33], indicating that the procedures detailed in these
reports should be followed when a new pesticide is being investigated.

The second JECFA report addressed biochemical and other special
investigations. It indicates that "the aspects of metabolic and bio-
chemical activity that might be profitably studied include the route
and rate of absorption of the test material, the levels of storage in
tissues and the subsequent fate of the stored material. Studies of the
metabolism of the material, together with the identification of the
metabolites, might be extended to include balance experiments, in which
an attempt is made to account for the administered dose as metabolites
excreted or material stored in the body" [31, p. 13]. The report indi-
cated that studies should be performed initially at high dose levels
and later they should be extended to investigate lower dose levels. It
also indicated that examination of enzyme processes and studies using
pharmacodynamic techniques may be useful in specific cases.

The 1963 JMPR stated that "it is important to know whether a sub-
stance is absorbed, its distribution in the body after absorption, its
mechanism of action including its influence on enzyme systems, how it
is metabolized, and the routes of final elimination. The toxicity of a
pesticide may be altered at any of these stages." [35, p. 8].

A WHO Scientific Group in 1967 also indicated the importance of
metabolism studies, stating that:

"The detailed study of metabolism at the molecular level has been
applied to many problems and this has special relevance to toxicology.
Modification of substances in the course of their metabolism may sig-
nificantly affect their toxicity; chemicals may alter enzyme activity
and some substances may stimulate the production of metabolizing
enzymes. Hence for a full understanding of the effects of a chemical on
biological systems, it is necessary to have as much knowledge as poss-
ible about the relationship between the chemical (and its derivatives)
and the complex pattern of enzymes in living organisms." [169, p. 4].

In the section of the report that addressed enzyme studies, the
Scientific Group stated:

"It has become more and more apparent that, among the mechanisms
of action of toxic substances, those of a biochemical nature are of
prime importance. In this connection, the basic enzyme systems are
certainly among the first sites of action to merit careful study, since

their inhibition often constitutes the causal biochemical lesion that
determines, at least in part, the nature of toxic effects." [169,
p. 14].

In 1975, JMPR [54] re-emphasized the principle that tissue distri-
bution and the mode and rate of metabolism and excretion can profoundly
influence the toxicity of a compound. It noted, however, that such data
were usually based on single-dose studies. In proposing the need for
multiple-dose studies, the Meeting noted that biliary excretion, with
the potential for enterohepatic circulation, and the problems of dis-
tribution and storage of highly lipophilic substances in fat deposits,
as well as potential accumulation of slowly metabolized compounds,
would not be adequately addressed by single-dose studies.

8.3.7.2 Current position

In discussing doses in toxicity studies and extrapolation to
humans, the 1987 JMPR indicated that comparative metabolism of the test
material in the experimental animal and man were basic to the choice of
dose levels [78]. The Meeting recognized the rarity of such data and
the ethical problems involved in obtaining the required data in the
required sequence (i.e., experiments in man prior to completion of all
animal studies). In addition, the following points were made:

1. "The processes involved in absorption, distribution, biotransform-
ation, and excretion are dependent upon many factors, including
physico-chemical properties, extent of protein binding, bioavail-
ability, and dose. Some of these processes are saturable. Products
of biotransformation may be formed at different rates and in dif-
ferent quantities, or by different pathways at high doses (e.g., 2-
phenylphenol) . . . It is valid to extrapolate animal data to man
only if the biotransformation pathways of the chemical are ident-
ical or very similar between species, and if the doses do not ex-
ceed the capacity of the pathways being compared. If this capacity
is exceeded, different metabolites may be produced.

2. "Kinetic data are useful in the design of studies and in the
interpretation and extrapolation of the data. For example, if the
test material is not absorbed, the need for one or more long-term
studies would be obviated.

3. "Extrapolation of animal data to man may be compromised by differ-
ences between species in the movement of the chemical after absorp-
tion. For example, the administration of high doses of certain
chemicals may result in increased enterohepatic circulation of the
chemical and/or its metabolites. This is an important system in the
rat, but less so in man.

4. "The proper design of definitive long-term studies should be based
on comparative data on absorption, distribution, biotransformation,
excretion, and appropriate kinetic considerations of the test sub-
stance." [78, p. 3-4].

Some explanation of specific points in the above quote are required
to clarify the intent of JMPR. Point 1 emphasizes the importance of
obtaining comparative metabolism and pharmacokinetic data in humans and
the species in which a toxic effect is observed. Although in the

absence of such data it is assumed that biotransformation in humans and
the test species is similar, only comparative metabolic studies can
confirm the validity of the extrapolation. In point 2, species differ-
ences regarding absorption should be considered. The variability in
gut microflora between species and the possible effects of intestinal
breakdown products require consideration. Similarly, the use of kinetic
data to determine whether a "steady state" has been achieved (i.e.,
the achievement of a state of equilibrium between intake and excretion)
is important in protocol design.

From the above discussion, it is apparent that data on pharmaco-
kinetics, pharmacodynamics, biotransformation, and studies on enzymes
are basic to many considerations in toxicology. Since toxic activity
depends on the interaction of a chemical and a target site (or sites)
in the intact animal, some knowledge of the identity and quantity of
the material and/or its metabolites reaching the target site is needed.
The 1986 JMPR stressed the importance of understanding the mechanisms
that result in the expression of toxicity. It noted that: "Current
knowledge of mechanisms of toxicity is limited, but there is already a
sufficient understanding in some cases to permit better design, per-
formance, and interpretation of toxicological studies. Mechanistic
studies are therefore encouraged, since a knowledge of mechanism of
action is likely to result in a more rational assessment of the risk to
man." [76, p. 2].

The material absorbed may be the administered chemical(s), or it
may be metabolites and/or reaction products of the administered chemi-
cal. Variations in absorption occur because of species differences
(especially when specialized transport mechanisms are involved in
absorption, such as those encountered with metals), differences in
intestinal flora (discussed extensively in reference [176]), age,
nutritional status, dietary fibre content, and factors affecting
motility. The identity of the absorbed material may also differ mark-
edly from that administered, due to acid-mediated hydrolysis in the
stomach, breakdown by gastrointestinal enzymes (e.g., splitting of pep-
tides), chemical reactions between food components (e.g., nitrosamine
formation by reaction between nitrite and secondary amines in the
stomach), and the activity of the intestinal flora. Secondary absorp-
tion may also occur, arising from biliary excretion and subsequent
reabsorption of the excreted material, either in its original excreted
form or following hydrolysis in the intestine.

Information about the site of absorption of the test material is
also important, since this may alter the overall metabolism and thus
the toxicological profile of the test substance. If absorption occurs
in the buccal cavity, the oesophagus, or the stomach, it is likely to
be distributed widely throughout the body in the form in which it is
absorbed. If absorption is from the small intestine, the transportation
of the absorbed material will be via the hepatic portal system to the
liver. Within the liver, it may be metabolized, resulting in distri-
bution of metabolites rather than of parent compound. This factor is
of major importance when considering routes of exposure other than
those by the oral route. Resolution of these potential problems can be
achieved by adequate pharmacokinetic and metabolic data.

Once absorbed from the gastrointestinal tract, distribution depends
on a variety of factors, which may differ between and within species.
For example, the age of the animal, the rate of metabolism, the degree
of previous exposure, and the amount and rate of blood flow through
different organs may all affect the eventual distribution of the
absorbed material. The distribution and, ultimately, the concentration
at the receptor level, is greatly influenced by the ability of the
chemical to penetrate biological membranes such as the placenta, glom-
erular membrane, and the blood/brain barrier. This, in turn, is primar-
ily a function of lipophilicity, molecular size, and extent of ioniz-
ation (pK_a).

The metabolism of the absorbed material depends on an equally wide
range of variables:

* the degree of enzyme development is dependent on age;

* enzymes may vary between species, both qualitatively and quantitat-
ively;

* Michaelis-Menten kinetics indicate that saturation of enzyme sys-
tems may occur at some level, either increasing the importance of
secondary mechanisms of metabolism or resulting in greater plasma
levels of parent compound;

* the site of metabolic activity may differ among species (e.g.,
microbial metabolism in the rodent stomach, which is not observed
in humans, primates, or dogs);

* the rate of metabolism may differ within and between species and
between different tissues and cells;

* interaction among test substances may occur, or metabolism may be
affected by other test substances (e.g., enzyme inhibition, stimu-
lation, or induction);

* duration of exposure (acute or chronic) may modify the rate and
pathways of metabolism.

Both the route and rate of excretion of a test substance may vary
between species. The pharmacokinetic parameters of clearance and bio-
logical half-life are considered to be indicators of the potential for
accumulation. However, rapid elimination of a chemical and its metab-
olites clearly does not necessarily equate to a lack of toxicity.

Use of radioactive labelling or heavy isotope techniques provides
data on absorption, distribution, and excretion. These studies assist
in the identification of sites of covalent binding, and are virtually
indispensable in the study of metabolism and pharmacokinetics. Data
from such studies, in conjunction with analytical determinations of
excreted products, provide the basis for determining the probable
metabolic pathways for administered compounds. It must be remembered
that in interpreting studies involving radiolabelling techniques, con-
sideration must be given to the site of the label on the molecule and

the stability (mobility) of the radiolabel. Thus, an organic molecule
containing several different ring structures may require multiple
studies, with radiolabelling at different sites in the molecule, to
ensure the determination of all metabolic products.

The above list of factors is incomplete, but nevertheless serves to
indicate the complexity of the problems associated with studies of ab-
sorption, distribution, metabolism, and excretion of a test substance.
For useful publications covering these issues, the reader is referred
to the comprehensive texts which have been published on the subject
(e.g., reference [107]).

8.3.7.3 Principles

1. Studies on absorption, distribution, metabolism, and excretion are
essential in the evaluation of the safety of a pesticide. These
studies provide a foundation for the interpretation of all other
toxicology studies.

2. Ethically conducted comparative metabolic and pharmacokinetic
studies in humans and animal test species may permit more accurate
extrapolation of animal data to humans.

9. EVALUATION OF DATA

9.1. Extrapolation of Animal Data to Humans

The objective of the safety evaluation of pesticide residues in
food is to determine the maximum daily intake of the pesticide that
will not result in adverse effects at any stage in the human lifespan.
Since, in the majority of cases, data on humans are inadequate to
permit such a determination, effects observed in other species must be
extrapolated to humans. Ideally, data on comparative pharmacokinetics,
metabolism, and mechanism of action should be utilized in the extrapol-
ation. However, such data are not available in the majority of cases.
The use of relevant biomarkers of exposure and effect such as the for-
mation of adducts to DNA or blood proteins like haemoglobin in humans
and test animals may also be useful in the extrapolation across
species. Further research in this area is to be encouraged.

Three basic approaches are now generally used in the extrapolation
of the results of studies in experimental animals to humans: the use of
safety factors, the use of pharmacokinetic extrapolation (widely used
in the safety evaluation of pharmaceuticals), or the use of linear low-
dose extrapolation models.

JMPR has not utilized the third approach (the use of linear low-
dose extrapolation models). A number of these models have been used to
determine the "virtually safe dose" (VSD) of carcinogens for humans.
One major drawback of these models is the lack of consideration of many
of the biological factors which should be taken into account. Further-
more, the various mathematical models available (Probit, Wiebel, etc.),
when applied to the same data, can result in VSD values which vary by
orders of magnitude. There is no agreement among toxicologists on the
"best" mathematical model available today, nor on whether these math-
ematical models have any biological meaning at all.

Pharmacokinetic extrapolation requires human pharmacokinetic data,
which are rarely available for pesticides. The method involves a com-
parison of pharmacokinetics in human and experimental animals. The
relative sensitivity of receptor sites must also be taken into con-
sideration.

The JMPR approach has generally been limited to the first of the
three approaches, that is the use of safety factors. These are applied
to the NOAEL determined from the experimental animal data, or prefer-
ably, from data in humans, if available.

9.2. Safety Factors

9.2.1. Background

The 1963 JMPR adopted the commonly used empirical approach for the
extrapolation of data to man, i.e. "the maximum no-effect dietary
level obtained in animal experiments, expressed in mg/kg body weight
per day, was divided by a `factor', generally 100." [35, p. 11]. This
concept appears to have been adopted from the report of the second

JECFA Meeting which states that ". . . a dosage level can be estab-
lished that causes no demonstrable effects in the animals used. In the
extrapolation of this figure to man, some margin of safety is desirable
to allow for any species differences in susceptibility, the numerical
differences between the test animals and the human population exposed
to the hazard, the greater variety of complicating disease processes in
the human population, the difficulty of estimating the human intake,
and the possibility of synergistic action among food additives." [31,
p. 17]. The Committee then stated that the 100-fold margin of safety
applied to the maximum ineffective dose (expressed in mg/kg body
weight per day) was believed to be an adequate factor.

The 1965 JMPR [36] discussed the concept of the acceptable daily
intake and safety factors. It noted that the 100-fold factor could be
modified according to circumstances (e.g., reduction to 10 or 20 fold
when human data are available or in the case of well-studied organo-
phosphates).

The 1966 JMPR indicated that when a temporary ADI was allocated,
the margin of safety applied to the NOAEL derived from experimental
animal data should be increased [38]. These principles were applied by
the 1966 Joint Meeting when establishing a temporary ADI for pyrethrin
(safety factor of 250) [39].

A WHO Scientific Group considered safety factors in 1967 [169].
This Group noted that safety factors could be varied and described
circumstances where increased safety factors should be used. These
included toxicological data gaps and when it was necessary to establish
temporary ADIs. Decreasing the margin of safety was proposed when
pertinent biological data indicates uniform species response, when the
initial effect is clear-cut and reversible, or when cholinesterase
inhibition or adaptive liver enlargement is the initial effect. Other-
wise a 100-fold safety factor was considered to be a useful guide.

The 1968 JMPR [42] indicated that, where human data comprised the
basis for the NOAEL used in determining the ADI, a smaller safety
factor might be utilized. This statement was amplified by the 1969
JMPR [44] to include human biochemical as well as toxicological data as
justification for reducing safety factors.

The 1975 JMPR, in addressing the question of safety factors in
toxicological evaluation, stated that:

"It should be emphasized that the magnitude of the margin of
safety applied in each individual case is based on the evaluation of
all available data. In consideration of any information that gives rise
to particular concern, the magnitude of the margin of safety will be
increased. Where the data provide an assurance of safety, the magnitude
may be decreased. Therefore, it is impossible to recommend fixed rules
for the margin of safety to be applied in all instances." [54, p. 9].

In 1977, the JMPR "wished to clarify the situation regarding
safety factors in arriving at ADIs for man. The establishment of the
ADI for man is not a simple arithmetic exercise based on the no-effect

level, as the safety factor may vary widely from one compound to
another. Although safety factors are determined empirically, they are
dependent on the nature of the compound, the amount, nature and quality
of the toxicological data available, the nature of the toxic effects of
the compound, whether the ADI or TADI for man is established, and the
nature of any further data required." [57, p. 4].

During a discussion on general principles used by the JMPR, the
1984 Meeting [72] stressed the degree of uncertainty that accompanies a
toxicological evaluation, and stated:

"The use of variable safety factors by the JMPR in the estimation
of ADI values reflects this uncertainty, and underlines the complexity
of assessing the human health hazards of pesticides. No hard and fast
rules can be made with regard to the magnitude of this safety factor,
since many aspects have to be considered, such as species differences,
individual variations, incompleteness of available data, and a number
of other matters such as considerations of the fact that pesticide
residues may be ingested by people of all ages throughout the whole
life-span, that they are eaten by the sick and the healthy as well as
children, and that there are wide variations in individual dietary
patterns." [72, p. 3].

The original concept of the use of 100-fold safety factors was
based on interspecies and intraspecies variations [114]. Included in
this consideration were variations between strains, provision for sen-
sitive human population sub-groups, and possible synergistic effects
due to exposure to more than one chemical.

The 100-fold safety factor can be viewed as two 10-fold factors,
one for inter- and one for intra-species variability [111]. While these
safety factors appear, on the basis of experience, to provide adequate
margins of safety in the extrapolation of data to man, they may, of
course, be questioned. Some experimental support for safety factors was
published by Dourson & Stara [26] in 1983. This paper also proposed an
additional 10-fold factor for extrapolating sub-chronic data, and for
converting lowest-observed-adverse-effect levels to NOAELs (factors of
1-10, depending upon the severity and concern raised by the observed
effect). Additional clinical and epidemiological research may improve
the characterization of the variation in response within the human
population to various pesticides and may allow a more accurate deter-
mination of safety factors.

9.2.2. Principles

When determining ADIs, the 100-fold safety factor is used as the
starting point for extrapolating animal data to man and may be modified
in the light of the data that are available and the various concerns
that arise when considering these data. Some of these are given below:

1. When relevant human data are available, the 10-fold factor for
inter-species variability may not be necessary. However, relatively
few parameters are studied in man in the assessment of pesticide
safety, and data on oncogenicity, reproduction, and chronic effects

are rarely available. Thus, even if the parameter measured in
humans is the same as the most sensitive adverse effects measured
in the experimental animal (e.g., erythrocyte cholinesterase de-
pression), uncertainty still remains with respect to the potential
effects on other parameters. This usually necessitates an increased
safety factor. Consequently, JMPR rarely utilizes safety factors as
low as 10-fold.

2. The quality of the data supporting the NOAELs determined in the
animal experiments (and also in human experiments) influences the
choice of the safety factor. Unfortunately, toxicity studies are
rarely perfect in all respects. While a study may serve to answer
a basic question, the degree of certainty with which the question
is answered may be reduced by, for example, increased mortality in
all groups in an oncogenicity study, resulting in marginally-
acceptable data being available at the termination of the study.
When a request for a repeat study is not fully justified, an
increased safety factor may be utilized under such circumstances.

3. The quality of the total data base may affect the choice of safety
factor. Significant data deficits may warrant an increased safety
factor due to increased uncertainty.

4. The type and significance of the initial toxic response may alter
the safety factor. Thus a response which is reversible may result
in a reduced safety factor.

5. The limited numbers of animals used in oncogenicity studies limits
the sensitivity of the study in the identification of a threshold
dose. When evidence of neoplasia has been identified, safety fac-
tors may be increased depending on the available ancillary data and
the establishment of an NOAEL.

6. The shape of the dose/response curve (in those cases where data are
adequate to permit derivation of such a curve) may also be con-
sidered in assessing safety factors.

7. Metabolic considerations may influence the choice of the safety
factor. Thus, saturation of metabolic pathways resulting in toxic
manifestations, biphasic metabolic patterns, and data on compara-
tive metabolism may all affect the magnitude of the safety factor.

8. Knowledge of the comparative mechanism of toxic action in exper-
imental animals and man may influence the choice of safety factor.

Several of the factors cited above may apply in the consideration
of any one compound. Certain factors may serve to increase and others
to decrease the choice of the final safety factor. Therefore, it must
be stressed that the total weight of evidence has to be considered in
determining the appropriate safety factor to be used and that the
determination of safety factors must be considered on a case-by-case
basis.

9.3. Allocating the ADI

9.3.1. Background

The FAO/WHO Joint Meeting on Principles Governing Consumer Safety
in Relation to Pesticide Residues indicated that the assessment of the
amount of pesticide to which man can be exposed daily for a lifetime,
without injury, was the primary aim of toxicological investigations.
The Meeting indicated that "when the (toxicological) investigations
are completed, it is possible, by the use of scientific judgement, to
name the acceptable daily intake." [32, p. 9]. The meeting also
defined the ADI as follows:

"The daily dosage of a chemical which, during an entire lifetime,
appears to be without appreciable risk on the basis of all the facts
known at the time. `Without appreciable risk' is taken to mean the
practical certainty that injury will not result even after a lifetime
of exposure. The acceptable daily intake is expressed in milligrams of
the chemical, as it appears in the food, per kilogram of body weight
(mg/kg)." [32, p. 5].

The first JMPR adopted this definition and discussed the concept of
the ADI. The Meeting stated that the following information should be
available in order to arrive at an ADI:

(a) "the chemical nature of the residue. Pesticides may undergo chemi-
cal changes and are frequently metabolized by the tissues of plants
and animals which have been treated with them. Even when a single
chemical has been applied, the residues may consist of a number of
derivatives with distinct properties, the exact nature of which may
differ in animals and plants and in different crops and products.

(b) the toxicities of the chemicals forming the residues from acute,
short-term and long-term studies in animals. In addition, knowledge
is required of the metabolism, mechanism of action and possible
carcinogenicity of residue chemicals where consumed.

The Meeting also noted that the identity of the food bearing the
chemical should theoretically be immaterial; that the ADI was an ex-
pression of opinion, which carried no guarantee of "absolute" safety;
that new knowledge or data could always lead to re-evaluation of an
ADI; and that JMPR would confine itself to proposing a single set of
ADI figures for pesticides. Finally, the Meeting stated that "The
proposed levels (of ADIs) could normally be regarded as acceptable
throughout life; they are not set with such precision that they cannot
be exceeded for short periods of time." [35, p. 7] (see section
9.3.3).

Although the ADI can be exceeded for short periods of time, it is
not possible to make generalization on the duration of the time frame
which may cause concern. The induction of detrimental effects will

depend upon factors which vary from pesticide to pesticide. The bio-
logical half-life of the pesticide, the nature of the toxicity, and the
amount by which the exposure exceeds the ADI are all crucial.

The large safety factors generally involved in establishing an ADI
also serve to provide assurance that exposure exceeding the ADI for
short time periods is unlikely to result in any deleterious effects
upon health. However, consideration should be given to the potentially
acute toxic effects that are not normally considered in the assessment
of an ADI.

The principles discussed above were adopted by subsequent Joint
Meetings but, as would be expected, have been further developed with
time. Thus the 1968 JMPR [42] indicated that metabolites would, under
certain conditions, be considered to be included in the ADI. Generally,
if the metabolites in food commodities are qualitatively and quantitat-
ively the same as those observed in laboratory test species, the ADI
would apply to the parent compound as well as to metabolites. If the
metabolites are not identical or not present at the same order of
magnitude, separate studies on the metabolites may be necessary. When
one or several pesticides are degradation products of another
pesticide, a single ADI may be appropriate for the pesticide and its
metabolites, e.g., oxydemeton-methyl, demeton- S -methyl sulfone and
demeton- S -methyl [183].

In 1973, when considering the accuracy with which ADIs or TADIs
could be estimated, JMPR recommended that ADIs should be expressed
numerically using only one significant figure [52]. The use of more
than one significant figure might be taken to imply a greater degree of
accuracy than that which can be achieved when assessing the hazard from
the wide range of factors that influence toxicity.

9.3.2. Temporary ADIs

Use of the TADI, first proposed by the Scientific Group on Pro-
cedures for Investigating Intentional & Unintentional Food Additives
[169], was adopted by JMPR in 1966. Criteria were set that had to be
met prior to the establishment of the TADI. These included the con-
sideration of each chemical on its own merits, the establishment of the
TADI for a fixed period (usually 3-5 years), and the subsequent review
of original and new data prior to the expiration of the provisional
period.

The establishment of a TADI has always been accompanied by a
requirement for further work by a specified date and by the application
of an increased safety factor. The 1972 JMPR considered the course of
action to be taken if requested data were not forthcoming and indicated
that, under these circumstances, the TADI would be withdrawn. It empha-
sized, however, that such an action "did not necessarily indicate a
potential health hazard, but only that insufficient information is
available at the time of review to permit the Meeting to state with
reasonable certainty that there is no likelihood of adverse effects on
health resulting from ingestion over a prolonged period." [50, p. 7].

In 1986 [76], JMPR indicated that the previously utilized terms
"Further work or information required" or "Further work or infor-
mation desirable" were being replaced, the former by the statement
"Studies without which the determination of an ADI is impracticable",
and the latter by the statement "Studies which will provide infor-
mation valuable to the continued evaluation of the compound." These
new statements not only reflect the actual work performed by JMPR much
more clearly than the previous terms "Required" and "Desirable",
but they also reflect the Meeting's increasing reluctance to allocate
temporary ADIs as well as the desire to continue the evaluation of a
compound even after an ADI has been allocated.

In 1988 [79], JMPR recommended that TADIs should not be allocated
for new compounds and that an ADI should not be allocated in the
absence of an adequate data base. The Meeting intended that monographs
be published for all chemicals which are reviewed, regardless of
whether an ADI is allocated, and that data requirements will be clearly
specified for those chemicals with an inadequate data base.

The concept of the "conditional acceptable daily intake", adopted
by the 1969 JMPR [44], was limited to those compounds for which the use
was at that time considered essential but for which the toxicological
data base was incomplete. This concept, which is unacceptable, has been
abandoned.

9.3.3. Present position

The minimum data base normally utilized in determining an ADI com-
prises short-term feeding studies, long-term feeding studies, carcino-
genicity studies, multigeneration reproduction studies, teratogenicity
studies, and acute and repeated exposure metabolic, toxicokinetic, and
toxicodynamic data. Where deemed necessary, additional special studies
may also be required, e.g., genotoxicity studies.

The NOAEL from the most appropriate study divided by the appropri-
ate safety factor determines the ADI. The lowest NOAEL is not necess-
arily the basis for the ADI (see section 8.2.1). Thus, even though the
NOAEL from a chronic toxicity study may be less than that from a repro-
duction study, the latter may serve as the basis for assessing the ADI,
because of the potential use of a higher safety factor (see section
9.2). On this basis, the entire age range of the population is normally
covered by the ADI. The present procedure therefore provides an accept-
able margin of safety to the entire population for those pesticides
with complete data bases. The advantage of providing separate ADIs for
different age (or physiological) groups of the population, would there-
fore be limited to indicating those groups who may be in a reduced-risk
category, rather than indicating those at increased risk.

A document entitled "Guidelines for Predicting Dietary Intake of
Pesticide Residues" was published by WHO in 1989 [177]. This document
provides guidance on the prediction of the dietary intake of residues
of a pesticide for the purpose of comparison with the ADI allocated by

JMPR. The document recommends a step-wise approach to predicting in-
take, considering average consumption of the treated commodities and a
number of factors (such as processing, variations in residues level
with time and the percentage of a given commodity that is treated) that
usually have the effect of providing a more accurate prediction of real
pesticide residue intake. An example of dietary intake calculations
for a hypothetical pesticide is given in Chapter 3 of "Guidelines for
Predicting Dietary Intake of Pesticide Residues."

10. EVALUATION OF MIXTURES

10.1. Introduction

Survey data indicate that residues of more than one pesticide may
be detected in food. This gives rise to concern over the possibility
of unanticipated interactions between such residues leading to adverse
toxicological effects. There is, of course, a virtually unlimited
number of combinations of pesticides on various crops. There is also a
very large number of combinations of foods containing pesticide resi-
dues.

10.2. Background

The possibility of pesticide interaction was recognized as early as
1961 when the FAO/WHO Meeting on Principles Governing Consumer Safety
in Relation to Pesticide Residues recognized that "different pesti-
cides and other chemicals are often absorbed simultaneously during
occupational use, or in food, by man or animals" [32, p. 10]. The
first JMPR [35] also noted the possibility of interactions between
chemicals in discussions on the shortcomings of the ADI. It was indi-
cated that ADI values were calculated on the assumption that the diet
was contaminated by a single residue, hence additive and synergistic
effects were not considered. An extensive review of the significance
of interactions of pesticides was performed by the 1967 JMPR [40]. The
1981 Joint Meeting gave further consideration to interaction between
pesticide residues and concluded that:

1. "Not only could pesticides interact, but so could all compounds
(including those in food) to which man could be exposed. This leads
to unlimited possibilities, and there is no special reason why the
interactions of pesticide residues (which are at very low levels)
should be highlighted as being of particular concern;

2. "Very little data on these interactions are available;

3. "The data obtained from acute potentiation studies are of little
value in assessing ADIs for man." [62, p. 12].

10.3. Principle

The consideration of mixtures of residues does not require any
change in the general principles for estimating ADIs. However, there is
a need for further data on interactions of pesticides with each other
and with other common contaminants of food (e.g., metals, mycotoxins)
to ensure that, at the very low levels of pesticide exposure likely to
occur via dietary residues, and over the prolonged time periods in-
volved in such human exposure, no adverse effects are likely to occur.

11. RE-EVALUATION OF PESTICIDES

In 1961, the Meeting on Consumer Safety in Relation to Pesticide
Residues stated that "of necessity early views of the amount (ADI)
will be estimated and subject to revision as experience accumulates"
[32, p. 9]. Thus, from its inception, the provisional nature of the ADI
has been recognized [35]. The 1965 Meeting [36] re-examined the 37
pesticides reviewed in 1963 [35]. Changes in the ADIs were instituted
for 16 of these pesticides, based on additional information that had
become available.

The need for a full re-evaluation of the toxicity data base on some
pesticides was identified by the 1981 JMPR [65], based on concerns over
the validity of previously submitted data (see section 5.1). The first
of these re-evaluations was undertaken in 1982 [64]. The development of
new methods for investigating toxicity has also caused concern in
relation to pesticides for which ADIs have been established [52].

The use of the TADI [40] ensures re-evaluation of the data base
pertaining to specific compounds, since one of the criteria for setting
a TADI is that identified data are required for evaluation by a
specific time. However, a more systematic method of re-evaluation has
been suggested such as the automatic re-evaluation of chemicals
reviewed more than 10 years previously [72, p. 8].

Establishing a priority order for the re-evaluation of compounds
requires input from a number of sources including the Codex Committee
on Pesticide Residues (CCPR). This Committee has initiated this process
for pesticides evaluated prior to 1976 [30].

12. BIOTECHNOLOGY

Biotechnology comprises a number of different approaches to pest
control. Three areas are of emerging concern: the production of chemi-
cals of biological origin with pest-control activity (e.g., hydro-
prene); the use of microbial pest control agents (e.g., bacteria,
fungi, viruses, and protozoa); and the development and use of geneti-
cally altered (bioengineered) organisms for specific purposes.

At the present time, JMPR has no experience with these types of
pest-control products other than limited experience with biologically
derived chemicals (e.g., pyrethrin). The following comments are there-
fore proposals leading to approaches which may be feasible in assessing
the safety of such products.

First, with regard to the so-called "biorational" products, these
chemicals are derived from or are synthesized to be identical to nat-
urally occurring pesticidal agents. The fact that they are naturally
occurring does not necessarily mean that they are safe. Thus, such
chemicals should be investigated in the same way as other synthetic
chemicals used as pesticides. In certain instances, it is possible that
justification for reducing the necessary toxicological data base may
exist.

In dealing with microbial pest-control agents and bioengineered
organisms, two factors are of primary importance to human health - the
infectivity of the residual organism and the ability of the organism to
produce toxins which occur as residues. In the case of viruses, their
ability to incorporate into the cell genome should also be considered.

The determination of the safety of microorganisms should follow a
tiered approach, tier 1 being the determination of infectivity and tox-
icity based on acute administration. If measurable survival of the
microorganisms in the test animal is still apparent several days after
administration, short-term feeding studies may be deemed to be appro-
priate. If exotoxins are produced by the microorganisms, then the toxin
should be isolated, identified, and subjected to tests similar to those
for any other chemical utilized as a pesticide. Similarly, if an endo-
toxin is produced and there is evidence that this material could be
released, the endotoxin should also be subjected to standard toxicology
testing, as required for other chemical pesticides. In the event of
both endo- and exo-toxin having potential access to humans or to dom-
estic animals, consideration should be given to simultaneous adminis-
tration of the two compounds in toxicity studies. If circumstances
exist that would indicate the possibility of waiving any of the routine
toxicity tests, scientifically supported evidence indicating the
absence of need for such tests must be provided.

13. SPECIAL CONSIDERATIONS FOR INDIVIDUAL CLASSES OF PESTICIDES

13.1. Organophosphates - Ophthalmological Effects

In 1972, JMPR noted published reports suggesting that certain oph-
thalmological effects may be induced by exposure to some organophos-
phate insecticides [50]. Insufficient information was available at that
time to permit a toxicological assessment of the significance of the
reports. In 1979, additional reports were considered by JMPR [52]. The
Meeting again concluded that insufficient information was available to
permit an evaluation. No additional information has been considered by
JMPR.

13.2. Organophosphates - Aliesterase (carboxylesterase) Inhibition

The 1967 JMPR [40], in considering interactions between pesticides,
noted that "some of the aliesterases are more sensitive than the
cholinesterases to inhibition by certain organophosphorus compounds
[83]. Furthermore, those organophosphates that are more active as
aliesterase inhibitors than as cholinesterase inhibitors appear to be
the most effective in potentiating the toxicity of other organophos-
phates [27]. Also, the aliesterases participate in the detoxication of
many of the organophosphates and probably other chemicals to which
humans may be exposed. For these reasons, it is suggested that con-
sideration be given to the use of no-effect levels for aliesterase
inhibition rather than no-effect levels for cholinesterase inhibition,
as a basis for estimating the daily acceptable intakes of those organo-
phosphorus insecticides to which the aliesterase systems are more sen-
sitive than are the cholinesterases" [40, p. 38].

In 1972, JMPR noted that short-term feeding studies demonstrated
the fact that in the case of many organophosphate pesticides, inhi-
bition of liver and serum carboxylesterases was a more sensitive par-
ameter than inhibition of cholinesterases [122; 151]. However, it noted
that "the physiological significance of carboxylesterase inhibition is
still unknown" [50, p. 8]. Since carboxylesterase inhibition appears
to be a factor in potentiation, JMPR indicated the desirability of
further work in this area. The 1974 Meeting [53] reiterated the need
for information to determine the usefulness of aliesterase inhibition
in assessing the safety of organophosphate compounds.

Carboxylesterases mainly hydrolyze aliphatic esters, but their
substrate specificity is not absolute. They can also hydrolyze aromatic
esters at measurable rates. There is evidence for marked variation in
humans, which is genetically determined. Since there is such varia-
bility, yet no data on the toxicological significance of these rela-
tively non-specific enzymes, they are unlikely to be used to determine
NOAELs in the evaluation of organophosphorus compounds. It should,
however, be noted that some organophosphate impurities are potent
carboxylesterase inhibitors and hence markedly potentiate the toxicity
of pesticides that are detoxified by these enzymes (e.g., malathion).

13.3. The Need for Carcinogenicity Testing of Organophosphates

In 1986, JMPR noted that organophosphate compounds tend not to show
genotoxicity in vivo or to induce carcinogenic responses in laboratory
animals [76]. It was recommended that careful evaluation of all avail-
able data should be performed to determine whether carcinogenicity
tests are required for individual organophosphate pesticides. It is
also recommended that the possible structure-activity relationships of
the non-phosphate ester moiety of the pesticide should be considered.

13.4. Ocular Toxicity of Bipyridilium Compounds

The pyridilium herbicides diquat and paraquat were first reviewed
by JMPR in 1970 [77]. The studies on diquat demonstrated the induction
of lens opacities in rats, dogs, and cows. Studies on paraquat, for
the same duration and at the same dose levels as for diquat, did not
demonstrate ocular effects in any species tested.

Additional data on both compounds were evaluated in 1972 [51]. At
that time, it was demonstrated that prolonged administration of diquat
was required to induce cataracts. The type of cataract induced differed
structurally from those observed due to physical or disease processes.
Again, no evidence of compound-related ocular damage was noted in rats
or mice treated with paraquat.

In 1977, additional data on diquat showed that although the inci-
dence of cataracts was no higher than that of control animals, an
earlier appearance was observed [58]. Because the data base for
paraquat had been generated by Industrial Biotest Laboratories, most
studies of this chemical were repeated. These repeat studies were
evaluated by the 1986 JMPR [77]. No cataracts were induced in a one
year study in dogs or in a long-term feeding study in mice. Cataracts
were observed in Fisher (but not Wistar) rats in long-term studies.
Microscopic examination of these cataracts showed that, in contrast to
diquat-induced cataracts, there was a close similarity to age-related
cataracts in control animals. There is, therefore, some evidence that
paraquat may cause some ophthalmological toxicity, even though this has
only been observed in one strain of one species, and even then the
lesions noted are similar in type to age-related lesions. However, it
would be advisable to perform careful ophthalmological studies on any
bipyridilium compounds that may be developed as future herbicides.

13.5. Goitrogenic Carcinogens

A probable mechanism for this class of compounds is described
below.

Diets low in iodine, causing chronic iodine deficiency in exper-
imental animals, lead to hypertrophy, hyperplasia, and follicular cell
neoplasia of the thyroid gland and pituitary gland adenomas [6; 8; 80;
81; 145]. These effects have also been observed with subtotal thyroid-
ectomy [24], splenic transplantation of thyroid tissue [10], and the
transplantation of pituitary tumours that secrete thyroid-stimulating
hormone (TSH) [24; 90; 148]. The fact that none of these experimental

techniques introduced exogenous agents, other than transplanted tissue,
into the animal's internal environment indicated that the causative
oncogenic mechanism must reside within the animal and must be mediated
through the intimate interrelationship of the pituitary and thyroid
glands.

This led to a concept, subsequently supported with experimental
data, of a negative feedback system which maintained a homeostatic
balance between the pituitary and thyroid gland secretions. Later, the
hypothalamus was added to this system when it was discovered that it
exerted some control over pituitary gland secretions. Subsequent
research indicated that, while the hypothalamus is essential to normal
pituitary and thyroid gland functioning, the receptors residing within
the pituitary/thyroid axis are of primary importance in controlling
thyroid and pituitary hormonal balance [86; 97]. Disturbance of this
balance has significant physiological and morphological effects on the
glands as well as on the well-being of the animal [118].

Exogenous physical and chemical agents can also induce
thyroid/pituitary hypertrophy, hyperplasia, and neoplasia by causing
hormonal imbalance [9; 102; 180]. The chemical goitrogens were first
discovered in animal and human food items [15; 88; 149]. Since then
many chemically defined substances have been reported to induce
thyroid/pituitary hypertrophy, hyperplasia, and, after prolonged
exposure, neoplasia. Radioactive iodine and x-rays can produce the same
effects [17; 25; 96; 98]. The mechanisms by which these substances and
the non-agent experimental techniques produce their pharmacological
(goitrogenic) and neoplastic effects are well known, even though the
precise triggering event for the transformation from hyperplasia to
neoplasia is still uncertain [9; 85; 87]. The most common mechanisms
are interference with the thyroid iodide transport system or inter-
ference with peroxidases essential to the synthesis and secretion of
competent thyroid hormone [15; 18; 95; 138; 149; 156].

The sequence of events triggered by this interference is also well
understood. As the circulation of competent thyroid hormone (TH) is
reduced, the receptors in the hypothalamus and pituitary gland receive
a signal for secretion of TSH. Receptors within the thyroid receive,
through TSH, a signal for increased TH production and secretion and the
gland responds, at first, with functional hypertrophy [86; 97]. Thus
far these events may remain within the normal operation of the feedback
system. At this stage, if the cause of the thyroid hormone deficiency
is removed or corrected, the circulation of competent TH increases to a
critical level and TSH secretion is reduced. As homeostasis is reestab-
lished, the thyroid gland returns to normal. However, under conditions
of chronic TH deficiency and the failure of the feedback mechanism to
restore hormonal balance, both glands continue to respond to their
respective signals and enter hypertrophic and hyperplastic states. The
pituitary continues to secrete TSH, to which the thyroid responds, but
the thyroid cannot signal for TSH shut off because of its inability to
secrete competent TH. Eventually this relationship results in hormonal
imbalance that induces thyroid gland follicular cell neoplasia and fre-
quently pituitary gland neoplasia.

The hypothesis that thyroid/pituitary hormonal imbalance is the
oncogenic mechanism is supported by evidence that follicular cell neo-
plasia can be prevented by the simultaneous administration of goitro-
gens and thyroid hormone. This has been demonstrated with thiouracil
[9] and thiourea [143], two members of a class of potent goitrogens.
The pharmacological effects, hypertrophy and hyperplasia, are revers-
ible upon removal of the goitrogenic stimulus [4; 9; 85; 106; 140;
153]. Furthermore, NOELs have been demonstrated in several species, for
both the goitrogenic and neoplastic effects of thyroid function inhibi-
tors. These facts coupled with evidence that treatment of human hypo-
thyroidism with goitrogens is without appreciable risk of thyroid neo-
plasia [92], support the concept of a threshold for goitrogen-induced
thyroid follicular cell and pituitary neoplasia [137; 138].

The weight-of-evidence indicates that goitrogens occupy an unusual
nitch in oncogenesis in that:

* their pharmacological effects and mechanisms of action are reason-
ably well understood;

* their pharmacological effects are reversible;

* thresholds, NOELs, and NOAELs can be established for their pharma-
cological and neoplastic effects;

* pituitary and thyroid neoplasia potentially induced by thyroid
inhibitors can be prevented by supplying experimental animals with
competent thyroid hormones during treatment with goitrogens;

* a certain degree of thyroid inhibition is accommodated for pro-
longed periods within the homeostatic control limits of the nor-
mally functioning feedback mechanism;

* long-term exposure to excessive TSH is required before hormonal
imbalance induces thyroid follicular cell neoplasia.

Increased TSH secretion is the ultimate common mediator of thyroid
follicular proliferative lesions induced by goitrogens, its level is
moderated by a feedback mechanism, and its neoplasm-inducing potential
is subject to mechanisms demonstrating threshold effects. Laboratory
animal data demonstrate that there is an ordered linkage of steps: thy-
roid blockade, continuous TSH release, thyroid hypertrophy/hyperplasia,
modularity, and adenoma/carcinoma. They also show that a threshold for
an early step automatically becomes a threshold for the whole chain of
steps. These characteristics should be a major consideration when as-
sessing the human oncogenic potential of thyroid-function inhibitors.

REFERENCES

1. ALDRIDGE, W.N. (1986) The biological bases and measurement of
thresholds. Annu. Rev. Pharmacol. Toxicol., 26: 39-58.

2. AMES, B.N. (1983) Dietary carcinogens and anticarcinogens.
Oxygen radicals and degenerative diseases. Science, 221: 1256.

3. ANDERSON, M.E., CLEWALL, H.J., GARGAS, M.L., SMITH, F.A.,
& REITZ, R.H. (1987) Physiologic based pharmacokinetics and the
risk assessment process for methylene chloride. Toxicol. appl.
Pharmacol., 87: 185-206.

4. ARNOLD, D.L., KREWSKI, D.R., JUNKINS, D.B., MCGUIRE, P.F.,
MOODIE, C.A., & MUNRO., I.C. (1983) Reversibility of ethyl-
enethiourea-induced thyroid lesions. Toxicol. appl. Pharmacol.,
67: 264-273.

5. ASHBY, J. & TENNANT, R.W. (1988) Chemical structure, Salmonella
mutagenicity and extent of carcinogenicity as indicators of
genotoxic carcinogenesis amongst 222 chemicals tested in rodents
by the US NCI/NTP. Mutat. Res., 204: 17-215.

6. AXELROD, A.A. & LEBLOND, C.P. (1955) Induction of thyroid
tumors in rats by a low iodine diet. Cancer, 8: 339-367.

7. BARSHNECHT, T.G., NAISMITH, R.W., & KORNBULL, D.J. (1987)
Variations in the standard protocol design of the rat hepatocyte
DNA repair assay. Cell Biol. Toxicol., 3: 193-208.

8. BIELSCHOWSKY, F. (1953) Chronic iodine deficiency as a cause of
neoplasia in thyroid and pituitary of aged rats. Br. J. Cancer,
7(2): 203-213.

9. BIELSCHOWSKY, F. (1955) Neoplasia and internal environment. Br.
J. Cancer, 9: 80-116.

10. BRACHETTO-BRIAN, D. & GRINBERG, R. (1951) Histological
development of intrasplenic thyroid autografts in thyroidectomized
rats. Rev. Soc. Argent. Biol., 27: 199-204.

11. BURIN, G.J. (1988) Report of FIFRA Scientific Advisory Panel on
Neurotoxicity, Washington, DC, US Environmental Protection Agency,
Office of Pesticide Programmes.

12. BYRNE, G. (1987) Conviction of Bruening. Science, 24: 1004.

13. CAROLDI, S. & LOTTI, M. (1982) Neurotoxic esterase in peripheral
nerve: assay, inhibition and rate of resynthesis. Toxicol. appl.
Pharmacol., 62: 498-501.

14. CAVANAGH, J.B. (1973) Peripheral neuropathy caused by toxic
agents. CRC Crit. Rev. Toxicol., 2: 365.

15. CHESNEY, A.M., CLAWSON, T.A., & WEBSTER, B. (1928) Endemic
goitre in rabbits. I. Incidence and characteristics. Bull.
John Hopkins Hosp., 43(5): 261-277.

16. CHRISTIAN, M.S. (1986) A review of multigeneration studies. J.
Am. Coll. Toxicol., 5: 161-180.

17. CHRISTOV, K. (1978) Radiation-induced thyroid tumours in infant
rats. Radiat. Res., 73(2): 330-339.

18. CHRISTOV, K. & STOICHKOVA, N. (1977) Cytochemical localization
of peroxidase activity in normal proliferating and neoplastic
thyroid tissue of rats. An ultrastructural study. Acta
histochem., 58: 5275-5289.

19. CLAYSON, D.B. (1974) Bladder carcinogenesis in rats and mice:
Possibility of artifacts. J. Natl Cancer Inst., 52: 1685-1689.

20. CLEGG, D.J. & WANDELMAIER, F. (1987) An investigation into the
desirability of the inclusion of a second litter per generation in
rat multigeneration studies. In: Greenhalgh, R. & Roberts, T.R.,
ed. Pesticide science and biotechnology, Boston, Blackwell
Scientific Publications, pp. 549-552.

21. CIOMS (1982) Proposed international guidelines for biomedical
research involving human subjects, Geneva, Council for
International Organizations of Medical Sciences.

22. DAVIES, D.R. (1963) Neurotoxicity of organophosphorus compounds.
In: Koelle, G.B., ed. [Handbook of experimental pharmacology],
Supplement 15, Berlin, Springer Verlag, pp. 860-882 (in German).

23. DEMPSTER, A.P., SELWYN, M.D., & WEEKS, B.J. (1983) Combining
historical and randomized controls for assessing trends in
proportions. J. Am. Stat. Assoc., 78: 221-227.

24. DENT, J.N., GODSDEN, E.L., & FURTH, J. (1956) Further studies on
induction and growth of thyrotropic pituitary tumors in mice.
Cancer Res., 16: 171-174.

25. DONIACH, I. (1950) The effect of radioactive iodine alone and in
combination with methylthiouracil and acetylaminofluorene upon
tumor production in rat thyroid gland. Br. J. Cancer, 4(2): 223-
234.

26. DOURSON, M.L. & STARA, J.F. (1983) Regulatory history and
experimental support of uncertainty (safety) factors. Regul.
Toxicol. Pharmacol., 3: 224-238.

27. DUBOIS, K.P. (1987) The comparative inhibitory action of organic
phosphates on several esterases and arudases and the relationship
of esterase inhibition to phosphate potentiation. Unpublished
paper presented to the Gordon Research Conference on Toxicology
and Safety Evaluation, Meriden, New Hampshire, August 1967.

28. ELLMAN, G.I., COURTNEY, K.D., ANDRES, V., Jr, &
FEATHERSTONE, R.M. (1967) A new and rapid colorimetric
determination of acetylcholinesterase activity. Biochem.
Pharmacol., 7: 88-95.

29. FAO (1959) Report on the Meeting of the FAO Panel of Experts on
the Use of Pesticides in Agriculture held in Rome, 6-13 April
1959, Rome, Food and Agriculture Organization of the United
Nations. (Meeting Report No. 1959/3).

30. FAO (1989) Report of the Twentieth Session of the Codex
Committee on Pesticide Residues, Rome, Food and Agriculture
Organization of the United Nations, Codex Alimentarius Commission
(Alinorm 89/24).

31. FAO/WHO (1958) Procedures for the testing of intentional food
additives to establish their safety for use. Second Report of the
Joint FAO/WHO Expert Committee on Food Additives, Geneva, World
Health Organization (FAO Nutrition Meeting Report Series, No. 17;
WHO Technical Report Series, No. 144).

32. FAO/WHO (1962a) Principles governing consumer safety in relation
to pesticide residues, Geneva, World Health Organization (FAO
Plant Production and Protection Division Report, No. PL/1961/11;
WHO Technical Report Series, No. 240).

33. FAO/WHO (1962b) Evaluation of the carcinogenic hazard of food
additives. Fifth Report of the Joint FAO/WHO Expert Committee on
Food Additives, Geneva, World Health Organization (FAO Nutrition
Meeting Report Series, No. 29; WHO Technical Report Series, No.
220).

34. FAO/WHO (1964a) Report of the First Meeting of the FAO Working
Party on Pesticide Residues, Rome, Food and Agriculture
Organization (Meeting Report No. PL/1963/16).

35. FAO/WHO (1964b) Evaluation of the toxicity of pesticide residues
in food. Report of a Joint FAO/WHO Meeting of the FAO Committee
on Pesticides in Agriculture and the WHO Expert Committee on
Pesticide Residues, Geneva, World Health Organization (FAO Meeting
Report No. PL/1963/13; WHO/Food Add./23).

36. FAO/WHO (1965a) Evaluation of the toxicity of pesticide residues
in food. Report of the Second Joint Meeting of the FAO Committee
on Pesticides in Agriculture and the WHO Expert Committee on
Pesticide Residues, Geneva, World Health Organization (FAO Meeting
Report No. PL/1965/10; WHO/Food Add./26.65).

37. FAO/WHO (1966) Report on the Second Meeting of the FAO Working
Party on Pesticide Residues, Rome, Food and Agriculture
Organization of the United Nations (Meeting Report No.
PL/1965/12).

38. FAO/WHO (1967a) Pesticide Residues in Food. Joint Report of the
FAO Working Party on Pesticide Residues and the WHO Expert
Committee on Pesticide Residues, Geneva, World Health Organization
(FAO Agricultural Studies No. 73; WHO Technical Report Series, No.
370).

39. FAO/WHO (1967b) Evaluation of some pesticide residues in Food,
Geneva, World Health Organization (FAO/PL:CP/15; WHO/Food
Add./67.32).

40. FAO/WHO (1968a) Pesticide residues. Report of the 1967 Joint
Meeting of the FAO Working Party and the WHO Expert Committee,
Geneva, World Health Organization (FAO Meeting Report No.
PL:1967/M/11; WHO Technical Report Series, No. 391).

41. FAO/WHO (1968b) 1967 Evaluations of some pesticide residues in
food: The monographs, Geneva, World Health Organization
(FAO/PL:1967/M/11/1; WHO/Food Add./68.30).

42. FAO/WHO (1969a) Pesticide residues in food. Report of the 1968
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Agricultural Studies No.
78; WHO Technical Report Series, No. 417).

43. FAO/WHO (1969b) 1968 Evaluations of some pesticide residues in
food: The monographs, Geneva, World Health Organization
(FAO/PL:1968/M/9/a; WHO/Food Add./69.35).

44. FAO/WHO (1970a) Pesticide residues in food. Report of the 1969
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Group on Pesticide Residues, Geneva,
World Health Organization (FAO Agricultural Series, No. 84; WHO
Technical Report Series, No. 458).

45. FAO/WHO (1970b) 1969 Evaluations of some pesticide residues in
food: The monographs, Geneva, World Health Organization (FAO/PL:
1969/M/17/1.

46. FAO/WHO (1971a) Pesticide residues in food. Report of the 1970
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Agricultural Studies No.
87; WHO Technical Report Series, No. 474).

47. FAO/WHO (1971b) 1970 Evaluations of some pesticide residues in
food: The monographs, Geneva, World Health Organization
(AGP:1970/M/12/1; WHO/Food Add./71.42).

48. FAO/WHO (1972a) Pesticide residues in food. Report of the 1971
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Agricultural Studies No.
88; WHO Technical Report Series, No. 502).

49. FAO/WHO (1972b) 1971 Evaluations of some pesticide residues in
food: The monographs, World Health Organization (WHO Pesticide
Residues Series, No. 1).

50. FAO/WHO (1973a) Pesticide residues in food. Report of the 1972
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Agricultural Studies No.
90; WHO Technical Report Series, No. 525).

51. FAO/WHO (1973b) 1972 Evaluations of some pesticide residues in
food: The monographs, Geneva, World Health Organization (WHO
Pesticide Residues Series, No. 2).

52. FAO/WHO (1974) Pesticide residues in food. Report of the 1973
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Agricultural Studies No.
92; WHO Technical Report Series, No. 545).

53. FAO/WHO (1975) Pesticide residues in food. Report of the 1974
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Agricultural Studies No.
97; WHO Technical Report Series, No. 574).

54. FAO/WHO (1976) Pesticide residues in food. Report of the 1975
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues,
Geneva, World Health Organization (FAO Plant Production and
Protection Series, No. 1; WHO Technical Report Series, No. 592).

55. FAO/WHO (1977a) Pesticide residues in food. Report of the 1976
Joint Meeting of the FAO Panel of Experts on Pesticide Residues
and the Environment and the WHO Expert Group on Pesticide
Residues, Geneva, World Health Organization (FAO Food and
Nutrition Series, No. 9; FAO Plant Production and Protection
Series, No. 8; WHO Technical Report Series, No. 612).

56. FAO/WHO (1977b) 1976 Evaluations of some pesticide residues in
food: The monographs, Rome, Food and Agriculture Organization of
the United Nations (AGP:1976/M/14).

57. FAO/WHO (1978a) Pesticide residues in food - 1977. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues and Environment and the WHO Expert Committee on Pesticide
Residues, Rome, Food and Agriculture Organization of the United
Nations (FAO Plant Production and Protection Paper 10 Rev.).

58. FAO/WHO (1978b) Pesticide residues in food - 1977 Evaluations:
The monographs, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 15
Sup.).

59. FAO/WHO (1979) Pesticide residues in food - 1978. Report of the
Joint Meeting of the FAO Panel of Experts on Pesticide Residues in
Food and Environment and the WHO Expert Group on Pesticide
Residues, Rome, Food and Agriculture Organization of the United
Nations (FAO Plant Production and Protection Paper 15).

60. FAO/WHO (1980a) Pesticide residues in food - 1979. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and the WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 20).

61. FAO/WHO (1980b) Pesticide residues in food - 1979 Evaluations:
The monograph, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper No. 20
Sup.).

62. FAO/WHO (1981a) Pesticide residues in food - 1980. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and the WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 26).

63. FAO/WHO (1981b) Pesticide residues in food - 1980 Evaluations:
The monographs, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 26
Sup.).

64. FAO/WHO (1981c) Evaluation of certain food additives. Twenty-
fifth Report of the Joint FAO/WHO Expert Committee on Food
Additives, Geneva, World Health Organization (WHO Technical Report
Series, No. 669).

65. FAO/WHO (1982a) Pesticide residues in food - 1981. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and the WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 37).

66. FAO/WHO (1982b) Pesticide residues in food - 1981 Evaluations:
The monographs, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 42).

67. FAO/WHO (1983a) Pesticide residues in food - 1982. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and the WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 46).

68. FAO/WHO (1983b) Pesticide residues in food - 1982 Evaluations:
The monographs, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 49).

69. FAO/WHO (1983c) Evaluation of certain food additives and
contaminants. Twenty-seventh Report of the Joint FAO/WHO Expert
Committee on Food Additives, Geneva, World Health Organization
(WHO Technical Report Series, No. 696 and corrigenda).

70. FAO/WHO (1984) Pesticide residues in food - 1983. Report of the
Joint Meeting of the FAO Panel of Experts on Pesticide Residues in
Food and the Environment and the WHO Expert Group on Pesticide
Residues, Rome, Food and Agriculture Organization of the United
Nations (FAO Plant Production and Protection Paper 56).

71. FAO/WHO (1985a) Pesticide residues in food - 1983 Evaluations:
The monographs, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 61).

72. FAO/WHO (1985b) Pesticide residues in food - 1984. Report of
the Joint Meeting on Pesticide Residues, Rome, Food and
Agriculture Organization of the United Nations (FAO Plant
Production and Protection Paper 62).

73. FAO/WHO (1985c) Pesticide residues in food - 1984 Evaluations:
The monographs, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 67).

74. FAO/WHO (1986a) Pesticide residues in food - 1985. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and a WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 68).

75. FAO/WHO (1986b) Pesticide residues in food - Evaluations 1985:
Part II - Toxicology, Rome, Food and Agriculture Organization of
the United Nations (FAO Plant Production and Protection Paper
72/2).

76. FAO/WHO (1987a) Pesticide residues in food - 1986. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and a WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper 77).

77. FAO/WHO (1987b) Pesticide residues in food - 1986 Evaluations:
Part II - Toxicology, Rome, Food and Agriculture Organization of
the United Nations (FAO Plant Production and Protection Paper
78/2).

78. FAO/WHO (1987c) Pesticide residues in food - 1987. Report of
the Joint Meeting of the FAO Panel of Experts on Pesticide
Residues in Food and the Environment and a WHO Expert Group on
Pesticide Residues, Rome, Food and Agriculture Organization of the
United Nations (FAO Plant Production and Protection Paper No.
84).

79. FAO/WHO (1988) Pesticide residues in food - 1988. Report of the
Joint Meeting of the FAO Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues, Rome,
Food and Agriculture Organization of the United Nations (FAO Plant
Production and Protection Paper 92).

80. FORTNER, J.G., GEORGE, P.A., & STERNBERG, S.S. (1958) The
development of thyroid cancer and other abnormalities in Syrian
hamsters maintained on an iodine deficient diet. Surg. Forum, 9:
646-650.

81. FORTNER, J.G., GEORGE, P.A., & STERNBERG, S.S. (1960) Induced
and spontaneous thyroid cancer in the Syrian (Golden) hamster.
Endocrinology, 66: 364-376.

82. FOX, T.R. & WATENABE, P.G. (1985) Detection of a cellular
oncogene in spontaneous liver tumors of B6C1F₁ mice. Science,
228: 596-597.

83. FRAWLEY, J.P. (1965) Synergism and antagonism in research in
pesticides, New York, Academic Press, p. 73.

84. GEMS (1985) Guidelines for the study of dietary intakes of
chemical contaminants, Geneva, World Health Organization, Global
Environmental Monitoring System (WHO Offset Publication No. 87).

85. GILMAN, A.G. & MURAD, F. (1975) Thyroid and antithyroid drugs.
In: Goodman, L.S. & Gilman, A.G., ed. The pharmacological basis
of therapeutics, 5th ed., New York, MacMillan Publishing Company,
pp. 1398-1422.

86. GOODMAN, H.M. & VAN MIDDLESWORTH, L. (1980) The thyroid
gland. In: Mountcastle, V.B., ed. Medical physiology, Saint-
Louis, Missouri, Mosby, Vol. 2, pp. 1495-1518.

87. GREEN, W.I. (1978) Mechanisms of action of anti-thyroid
compounds. In: Werner, S.C. & Ingbar, S.H., ed. The thyroid, New
York, Harper and Row, pp. 77-87.

88. GRIESBACH, W.E., KENNEDY, T.H., & PURVES, H.D. (1945) Studies
on experimental goitre. VI. Thyroid adenomata in rats on brasse
seed diets. Br. J. exp. Pathol., 26: 18.

89. HAMILTON, H.E., MORGAN, D.P., & SIMMONS, A. (1978) A
pesticide (dieldrin) induced immunohaemolytic anemia. Environ.
Res., 17: 155-164.

90. HARAN-GHERA, N., PULLAR, P., & FURTH, J. (1960) Induction of
thyrotropin-dependent thyroid tumours by thyrotropes. J.
Endocrinol., 66: 694-701.

91. HARDEN, M.S., SCHULER, R.L., BURG, J.R., BOOTH, G.M.,
HAZELDEN, K.P., MACKENZIE, K.M., PICCIRILLO, V.J., & SMITH,
K.N. (1987) Evaluation of the Chernoff/Kavlock test for
developmental toxicity. Teratog. Carcinog. Mutagen., 7(1): 1-
128.

92. HAYES, W.J. (1982) Fungicides and related compounds. In:
Pesticides studies in man, Baltimore, Maryland, Williams and
Wilkins, p. 603.

93. HENRY, J.B., ed (1979) Todd-Sanford-Davidson's clinical
diagnosis and management, Philadelphia, Pennsylvania, W.B.
Saunders Co.

94. HEALTH AND WELFARE CANADA (1973) The testing of chemicals
for carcinogenicity, mutagenicity and teratogenicity, Ottawa,
Health and Welfare Canada, pp. 137-185.

95. HILL, R.N., ERDREICH, L.S., PAYNTER, O.E., ROBERTS, P.A.,
ROSENTHAL, S.L., & WILKINSON, C.F. (1989) Review - thyroid
follicular cell carcinogenesis. Fundam. appl. Toxicol., 12: 629-
657.

96. HINDAWI, A.I. & WILSON, C.M. (1965) The effects of irradiation
on the function and survival of rat thyroid. Clin. Sci., 28: 555-
571.

97. HINKLE, P.M. & GOH, K.B.C. (1982) Regulation of thyrotropin-
releasing hormone receptors and response by L-triiodothyronine in
dispersed rat pituitary cell cultures. Endocrinology, 110: 1725-
1731.

98. HOLM, L.E., (1980) Thyroid treatment and its possible influence
on occurrence of malignant tumors after diagnostic 131I. Acta
radiol. oncol., 19(6): 455-459.

99. HUGHES, D.H., BRUCE, R.C., HART, R.W., FISHBEIN, L.,
GAYLOR, D.W., SMITH, J.M., & CARLTON, W.W. (1983) A report on
the Workshop on Biological and Statistical Implications of the
EDO1 Study and Related Databases. Fundam. appl. Toxicol., 3: 129-
136.

100. IARC (1986) In: Gart, J.J., Krewski, D., Lee, P.N., Tarone,
R.E., & Wahrendorf, J., ed. The design and analyses of long-term
animal experiments. Statistical methods in cancer research,
Volume III, Lyon, International Agency for Research on Cancer
(IARC Scientific Publications No. 79).

101. IARC (1987) Preamble to IARC Monograph Programme on the
Evaluation of Carcinogenic Risk to Humans, Lyon, International
Agency for Research on Cancer.

102. ISRAEL, M.S. & ELLIS, I.R. (1960) The neoplastic potentialities
of mouse thyroid under extreme stimulation. Br. J. Cancer, 14:
206-211.

103. IOANNIDES, C.V., LEEM, P.Y., & PARKE, D.V. (1984) Cytochrome P
448 and the activation of toxic chemicals and carcinogens.
Xenobiotica, 14: 119.

104. JOHNSON, M.K. (1975) The delayed neuropathy caused by some
organophosphorous esters: Mechanism and challenge. CRC Crit. Rev.
Toxicol., 3: 289-316.

105. JOHNSON, M.K. (1982) The target for initiation of delayed
neurotoxicity by organophosphorus esters: biochemical studies and
toxicological applications. Rev. Biochem. Toxicol., 4: 141-212.

106. JULL, J.W. (1976) Endocrine aspects of carcinogenesis. In:
Chemical carcinogens, Washington, DC, American Chemical Society,
pp. 52-82 (ACS Monograph No. 173).

107. KATZUNG, B.G. (1987) Basic and clinical pharmacology, 3rd ed.,
Norwalk, Connecticut, Appicton and Lange.

108. KHERA, S. (1984) Maternal toxicity: a possible factor in foetal
toxicity in mice. Teratology, 29: 411-416.

109. KHERA, K.S. (1985) Maternal toxicity: a possible etiological
factor in embryofetal death and foetal malformations of rodent,
rabbit species. Teratology, 31: 129-153.

110. KHERA, K.S. (1987) Maternal toxicity in humans and animals:
effects on foetal development and criteria for their detection.
Teratog. Carcinog. Mutagen., 7: 287-295.

111. KLAASSEN, C.D. (1986) Principles of toxicology. In: Klaassen,
C.D., Amdur, M.O., & Doull, J., ed. Casarett and Doull's
toxicology, 3rd ed., New York, MacMillan Publishing Company, p.29.

112. LAMB, J.C. & CHAPIN, R.E. (1985) Fertility assessment by
continuous breeding. J. Am. Coll. Toxicol., 4: 173-184.

113. LEEMING, N.M., COZENS, D.D., & PALMER, A.K. (1987) Points for
consideration in the design of multigeneration studies,
Huntingdon, United Kingdom, Huntingdon Research Centre.

114. LEHMAN, A.J. & FITZHUGH, O.G. (1954) 100-fold margin of safety.
Food Drug Off. US Q. Bull., 13: 51-58.

115. LOTTI, M., BECKER, C.E., & AMINOFF, M.J. (1984) Organophosphate
polyneuropathy: pathogenesis and prevention. Neurology, 34: 658-
662.

116. LU, F.C., JESSOP, D.C., & LAVALLE, A. (1965) Toxicity of
pesticides in young versus adult rats. Food Cosmet. Toxicol., 3:
591-596.

117. LUSTER, M.I., DEAN, J.H., & MOORE, J.A. (1982) Evaluations of
immune functions in toxicology. In: Hayes, W.A., ed. Principles
and methods of toxicology, New York, Raven Press, pp. 561-586.

118. MACKENZIE, J.M., ZAKARIJA, M., & BONNYNS, M. (1979)
Hyperthyroidism. Endocrinology, 1: 429-459.

119. MACKENZIE, W.F. & GARNER, F.M. (1973) Comparison of neoplasms
in six sources of rats. J. Natl Cancer Inst., 50: 1243-1257.

120. MANSON, J.M. (1986) Teratogens. In: Klaassen, C.D., Amdur,
M.O., & Doull, J., ed. Casarett and Doull's toxicology, 3rd ed.,
New York, MacMillan Publishing Company, pp. 195-222.

121. MIHICH, E. & KANTER, P.M. (1987) The toxicology of biological
response modifiers. In: Berlin, A., Dean, J., Draper, M.H.,
Smith, E.M.B., & Spreafico, F., ed. Immunotoxicology, Hingham,
Maryland, Martinus Nijhoff Publishers, pp. 208-218.

122. MURPHY, S.D., & CHEEVER, K.L. (1968) Effect of feeding
insecticides. Inhibition of carboxyesterase and cholinesterase
activities in rats. Arch. environ. Health, 17: 749-759.

123. NARAHASHI, T. (1982) Cellular and molecular mechanisms of action
of insecticides: neurophysiological approach. Neurobehav.
Toxicol. Teratol., 4: 753-758.

124. OECD (1981) OECD Test guidelines 1981. Report from the OECD
Expert Groups on Short-term and Long-term Toxicity, Paris,
Organization of Economic Cooperation and Development.

125. OECD (1983) OECD Guidelines for testing of chemicals. Section
4: Health effects, Paris, Organization for Economic Cooperation
and Development (Guideline 416).

126. OFFICE OF SCIENCE AND TECHNOLOGY POLICY (1985) Chemical
carcinogens: A review of the science and its associated
principles. Fed. Reg., 49: 10371-10442, 50: 10372-10442.

127. PAGET, G.E. (1970) The design and interpretation of toxicity
tests. In: Methods in toxicology, Oxford, London, Blackwell
Scientific Publishers, pp. 1-10.

128. PALMER, A.K. (1981) Regulatory requirements for reproductive
toxicology. In: Kimmel, C.A. & Buelke-Sam, J., ed. Developmental
toxicology, New York, Raven Press, pp. 259-287.

129. PALMER, A.K. (1986) A simpler multigeneration study (poster
presentation). In: Greenhalgh, R. & Robert, T.R., ed. Proceeding
of the 6th IUPAC Congress on Pesticide Chemistry, 10-15 August,
Oxford, Blackwell Scientific Publications.

130. PARKE, D.V. (1977) Biochemical aspects. In: Raven, R.W., ed.
Principles of surgical oncology, New York, London, Plenum Medical
Press, pp. 113-156.

131. PARKE, D.V. (1979) The role of the endoplasmic reticulum in
carcinogenesis. In: Coulston, F., ed. Regulatory aspects of
carcinogenesis and food additives: The Delaney Clause, New York,
Academic Press, pp. 173-187.

132. PARKE, D.V. (1982) Survey of drugs-metabolizing enzymes.
Biochem. Soc. Trans., 11: 457-548.

133. PARKE, D.V. & IOANNEDES, C.L. (1984) Reactive intermediates and
oxygen toxicity in liver injury. In: Keppler, D., Popper, H.,
Bianchi, L., & Reutter, W., ed. Mechanisms of hepatocyte injury
and death, Lancaster, MTP Press, pp. 37-48.

134. PARKE, D.V. & SYMONS, A.M. (1977) The biochemical pharmacology
of mucus. In: Elstein, M. & Parke, D.V., ed. Mucus in health and
diseases, New York, London, Plenum Medical Press, pp. 423-441.

135. PAYNTER, O.E. (1984) Oncogenic potential guidance for analysis
and evaluation of long-term rodent studies, Washington, DC, US
Environmental Protection Agency, Office of Pesticides and Toxic
Substances (Evaluation Procedure No. 1000.1).

136. PAYNTER, O.E. & SCHMITT, R. (1979) The acceptable daily intake
as a quantified expression of the acceptability of pesticide
residues. In: Giessbuhler, H., ed. Advances in pesticide science,
Oxford, Pergamon Press, pp. 674-679.

137. PAYNTER, O.E., BURIN, G.J., JAEGER, R.B., & GREGORIO, C.A.
(1986) Neoplasia induced by inhibition of thyroid gland function
(guidance for analysis and evaluation), Washington, DC, US
Environmental Protection Agency, Standard Evaluation Procedure.

138. PAYNTER, O.E., BURIN, G.J., JAEGER, R.B., & GREGORIO, C.A.
(1988) Goitrogens and thyroid follicular cell neoplasia: Evidence
for a threshold process. Regul. Toxicol. Pharmacol., 8: 102-119.

139. PROBST, G.S. & HILL, L.E. (1987) Influence of age, sex and
strain on the in vitro induction of unscheduled DNA synthesis in
rat hepatocyte primary cultures. Cell Biol. Toxicol., 3: 113-126.

140. PURVES, H.D. & GRIESBACH, W.E. (1947) Studies on experimental
goitre. VIII. Thyroid tumors in rats treated with thiourea. Br.
J. exp. Pathol., 28: 46-53.

141. RIBELIN, W.E., ROLOFF, M.V., & HOUSER, R.M. (1984) Minimally
functional rat adrenal medullary pheochromocytomas. Vet. Pathol.,
21: 281-285.

142. ROE, F.J.C. & BAR, A. (1985) Enzootic and epizootic adrenal
medullary proliferative disease of rats: influence of dietary
factors which affect calcium absorption. Hum. Toxicol., 4: 27-52.

143. ROSIN, A. & UNGAR, H. (1957) Malignant tumors in the eyelids and
the auricular region of thiourea-treated rats. Cancer Res., 17:
302-305.

144. SANTOLUCITO, J.A. & MORRISON, G. (1971) EEG of rhesus monkeys
following prolonged low-level feeding of pesticides. Toxicol.
appl. Pharmacol., 19: 147-154.

145. SCHALLER, R.T. & STEVENSON, J.K. (1966) Development of
carcinoma of the thyroid in iodine-deficient mice. Cancer, 19:
1063-1080.

146. SEARLE, C.E. (1984) Chemical carcinogens, Washington, DC,
American Chemical Society, Vol. 2 (ACS Monograph No. 182).

147. SENANAYAKE, N. & JOHNSON, M.K. (1982) Acute polyneuropathy
after poisoning by an organophosphate insecticide. New Engl. J.
Med., 306: 155-157.

148. SINHA, D., PASCAL, R., & FURTH, J. (1965) Transplantable thyroid
carcinoma induced by thyrotropin. Arch. Pathol., 79: 192-198.

149. SRINIVASAN, V., MOUDGAL, N.R., & SHARMA, P.S. (1957) Studies
on goitrogenic agents in food. I. Goitrogenic action of
groundnut. J. Nutr., 61: 87-95.

150. STEVENS, K.R. & GALLO, M.A. (1986) Practical considerations in
the conduct of chronic toxicity tests. In: Hayes, A.W., ed.
Principles and methods of toxicology, New York, Raven Press, pp.
53-77.

151. SU, M., KINOSHITA, F.K., FRAWLEY, J.P., & DUBOIS, K.P. (1971)
Comparative inhibition of aliesterases and cholinesterases in rats
fed eighteen organophosphorus insecticides. Toxicol. appl.
Pharmacol., 20: 241-249.

152. SUMI, N., STAVROV, D., FROHBERG, H., & JOCHMANN, G. (1976)
The incidence of spontaneous tumors of the central nervous system
of Wistar rats. Arch. Toxicol., 35: 1-13.

153. SWARM, R.L., ROBERTS, G.K.S., LEVY, A.C., & HINES, L.R. (1973)
Observations of the thyroid gland in rats following the
administration of sulfamethoxazole and trimethoprim. Toxicol.
appl. Pharmacol., 24: 351-363.

154. TARONE, R.E. (1982) The use of historical control information in
testing for a trend in proportions. Biometrics, 38: 215-220.

155. TASK FORCE OF PAST PRESIDENTS (1982) Animal data in hazard
evaluation: Paths and pitfalls. Fundam. appl. Toxicol., 2: 101-
107.

156. TAUROG, A. (1979) Hormone synthesis. Endocrinology, 1: 331-
342.

157. THOMPSON, S.W. & HUNT, R.D. (1963) Spontaneous tumors in the
Sprague-Dawley rat: Incidence rates of some types of neoplasms as
determined by serial section versus single section technics. Ann.
NY Acad. Sci., 108: 832-845.

158. TILSON, H.A. & CRANMER, J.M., ed (1985) Neurotoxicology in the
fetus and the child. Proceedings of the 4th International
Neurotoxicology Conference, Little Rock, Arkansas, 9-13 September
1985. Neurotoxicology, 7: 1-669.

159. US EPA (1982) Pesticide assessment guidelines. Subdivision F -
Hazard evaluations: humans and domestic animals, Washington, DC,
US Environmental Protection Agency, pp. 130-137.

160. US EPA (1983) Good laboratory practice standards: Toxicology
testing. Fed. Reg., 48: 53922.

161. US FDA (1978) Good laboratory practice regulations. Fed. Reg.,
43(274): 49986-60020.

162. US FDA (1982) Toxicological principles for the safety assessment
of direct food additives and color additives used in food,
Washington, DC, US Food and Drug Administration, National
Toxicology Information Service, pp. 80-107.

163. VAN LEEUWEN, F.X.R., FRANKEN, M.A.M., & LAEBER, J.G. (1987)
The endocrine system as the target in experimental toxicology.
Adv. vet. Sci. comp. Med., 31: 121-149.

164. VETTORAZZI, G., ed. (1977) General principles in the
toxicological evaluation of pesticide residues in food. In:
Handbook of international food regulatory toxicology, Volume 1:
Evaluations, New York, London, SP Medical and Scientific Books,
pp. 93-142.

165. VOS, J.G. (1987) The role of histopathology assessment of
immunotoxicity. In: Berlin, A., Dean, J., Draper, M.H., Smith,
E.M.B., & Spreafico, F., ed. Immunotoxicology, Hingham, Maryland,
Martinus Nijhoff Publishers, pp. 125-134.

166. VOUK, V.B. & SHEEHAN, P.J., ed. (1983) Methods for assessing the
effects of chemicals on reproductive functions, New York, John
Wiley and Sons.

167. WARD, J.M. (1983) Background data and variations in tumor rates
of control rats and mice. Prog. exp. Tumor Res., 26: 241-258.

168. WARD, J.M. & RICE, J.M. (1982) Naturally occurring and
chemically induced brain tumours of rats and mice in
carcinogenesis bioassays. Ann. NY Acad. Sci., 26: 304-319.

169. WHO (1967) Procedures for investigating intentional and
unintentional food additives. Report of a WHO Scientific Group,
Geneva, World Health Organization (WHO Technical Report Series,
No. 348).

170. WHO (1974) Assessment of the carcinogenicity and mutagenicity of
chemicals. Report of a WHO Scientific Group, Geneva, World Health
Organization (WHO Technical Report Series, No. 546).

171. WHO (1978) IPCS Environmental Health Criteria 6: Principles and
methods for evaluating the toxicity of chemicals, Part 1, Geneva,
World Health Organization.

172. WHO (1983a) Report of a Strategy Meeting on Updating Principles
of Methodology for Testing and Assessing Chemicals in Food,
Oxford, United Kingdom, 19-25 September, Geneva, World Health
Organization (ICS(Food)/83) (Unpublished report).

173. WHO (1983b) IPCS Environmental Health Criteria 27: Guidelines on
studies in environmental epidemiology, Geneva, World Health
Organization.

174. WHO (1984) IPCS Environmental Health Criteria 30: Principles for
evaluating health risks to progeny associated with exposure to
chemicals during pregnancy, Geneva, World Health Organization.

175. WHO (1986) IPCS Environmental Health Criteria 60: Principles and
methods for the assessment of neurotoxicity associated with
exposure to chemicals, Geneva, World Health Organization.

176. WHO (1987) IPCS Environmental Health Criteria 70: Principles for
the safety assessment of food additives and contaminants in food,
Geneva, World Health Organization.

177. WHO (1989) Guidelines for predicting dietary intake of pesticide
residues, Geneva, World Health Organization.

178. WILLIAMS, G.M. & WEISBURGER, J.H. (1986) Chemical carcinogens.
In: Klaassen, C.D., Amdur, M.O., & Doull, J., ed. Casarett and
Doull's Toxicology, 3rd ed. New York, MacMillan Publishing
Company, pp. 99-173.

179. WOO, D.C. & HOAR, R.M. (1972) "Apparent hydronephrosis" as a
normal aspect of renal development in late gestation of rats: the
effect of methyl salicylate. Teratology, 6: 191-196.

180. WOO, Y., LAI, D.Y., ARCOS, J.C., & ARGUS, M.F. (1985) Chemical
induction of cancer - Structural bases and biological mechanisms,
New York, Academic Press, Inc., pp. 357-394.

181. ZBINDEN, L.C. (1987) A toxicologist's view of immunotoxicology.
In: Berlin, A., Dean, J., Draper, M.H., Smith, E.M.B., &
Spreafico, F., ed. Immunotoxicology, Hingham, Maryland, Martinus
Nijhoff Publishers, pp. 1-11.

182. FAO/WHO (1965b) Evaluation of the toxicity of pesticide residues
the joint meeting of the FOA Working Party of Experts on Pesticide
Residues and the WHO Expert Committee on Pesticide Residues, Rome,
Food and Agriculture Organization of the United Nations.

ANNEX I. GLOSSARY

I.1 Abbreviations Used in this Document

ADI Acceptable Daily Intake

CEC Commission of the European Communities

CNS Central Nervous System

FAO Food and Agriculture Organization of the United Nations

GLP Good Laboratory Practice

IARC International Agency for Research on Cancer

IPCS International Programme on Chemical Safety

JECFA Joint FAO/WHO Expert Committee on Food Additives

JMPR Joint FAO/WHO Meeting on Pesticide Residues

LD₀₁ Lethal Dose, 1%

LD₅₀ Lethal Dose, median

MRL Maximum Residue Level

MTD Maximum Tolerated Dose

NCI National Cancer Institute (USA)

NOAEL No-Observed-Adverse-Effect Level

NOEL No-Observed-Effect Level

NTE Neurotoxic Esterase

NTP National Toxicology Program (USA)

OECD Organization for Economic Cooperation and Development

OP Organophosphate

SAR Structure/Activity Relationship

TADI Temporary Acceptable Daily Intake

TH Thyroid Hormone

TOCP Tri- O -Cresyl Phosphate

VSD Virtually Safe Dose

TSH Thyrotropin

WHO World Health Organization

I.2 Definitions of Terms Used in this Document

Acceptable Daily Intake (ADI): An estimate by JMPR of the amount of a
pesticide, expressed on a body weight basis, that can be ingested daily
over a lifetime without appreciable health risk (standard man = 60
kg).

Codex Alimentarius Commission: The Commission was formed in 1962 to
implement the Joint FAO/WHO Food Standards Programme. The Commission
is an intergovernmental body made up of more than 130 Member Nations,
the delegates of whom represent their own countries. The Commission's
work of harmonizing food standards is carried out through various
committees, one of which is the Codex Committee on Pesticide Residues.
JMPR serves as the advisory body to the Codex Alimentarius Commission
on all scientific matters concerning pesticide residues.

Effect: A biological change in an organism, organ, or tissue.

Elimination (in metabolism): The expelling of a substance or other
material from the body (or a defined part thereof), usually by a
process of extrusion or exclusion, but sometimes through metabolic
transformation.

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

JMPR: JMPR is a technical committee of JMPR specialists acting in their
individual capacities. Each is a separately-constituted committee, and
when either the term "JMPR" or "the Meeting" is used, it is meant
to imply the common policy or combined output of the separate Meetings
over the years.

Long-term toxicity study: A study in which animals are observed during
the whole life span (or the major part of the life span) and in which
exposure to the test material takes place over the whole observation
time or a substantial part thereof. The term chronic toxicity study is
used sometimes as a synonym for "long-term toxicity study".

Lowest-observed-effect level (LOEL): The lowest dose of a substance
which causes changes distinguishable from those observed in normal
(control) animals.

No-observed-adverse-effect level (NOAEL): The highest dose of a sub-
stance at which no toxic effects are observed.

No-observed-effect level (NOEL): The highest dose of a substance which
causes no changes distinguishable from those observed in normal (con-
trol) animals.

Safety factor: A factor applied by JMPR to the no-observed-effect
level to derive an acceptable daily intake (the no-observed-adverse-
effect level is divided by the safety factor to calculate the ADI).
The value of the safety factor depends on the nature of the toxic
effect, and the quality of the toxicological information available.

Short-term toxicity study: An animal study (sometimes called a sub-
acute or subchronic study) in which the effects produced by the test
material, when administered in repeated doses (or continuously in food
or drinking-water) over a period of about 90 days, are studied.

Temporary ADI: Used by JMPR as an administrative procedure to permit
the continued acceptance of the pesticide pending submission of new
toxicological data.

Teratogen: An agent which, when administered prenatally, induces per-
manent abnormalities in structure.

Teratogenicity: The property (or potential) to produce structural
malformations or defects in an embryo or fetus.

Threshold dose: The dose at which an effect just begins to occur, that
is, at a dose immediately below the threshold dose the effect will not
occur, and immediately above the threshold dose the effect will occur.
For a given chemical there can be multiple threshold doses, in essence
one for each definable effect. For a given effect there may be differ-
ent threshold doses in different individuals. Further, the same indi-
vidual may vary from time to time as to his or her threshold dose for
any effect. However, given the present state in the development of
science, for certain chemicals and certain toxic effects, a threshold
dose may not be demonstrable.

The threshold dose will fall between the experimentally determined
no-observed-effect level (NOEL) and the lowest-observed-effect level
(LOEL). Of importance is that when using the NOEL or LOEL, it should
be specified which effect is being measured, in what population, and
what is the route of administration. In situations for which the effect
of concern is considered to be adverse, the terminology often used is
that of a no-observed-adverse-effect level (NOAEL) or lowest-observed-
adverse-effect level (LOAEL), again specifying the effect, the popu-
lation, and the route of administration. Both the NOEL and LOEL (as
well as the NOAEL and LOAEL) have been used by different scientific
groups as a surrogate for the threshold dose in the performance of risk
assessments.

Toxicity: The toxicity of a compound is its potential to cause injury
(adverse reaction) to a living organism.

ANNEX II. APPROXIMATE RELATION OF PARTS PER MILLION IN THE
DIET TO MG/KG BODY WEIGHT PER DAY^a
----------------------------------------------------------------------------------------------------
Weight Food consumed Type of 1 ppm in food 1 mg/kg body
(kg) per day (g) diet = (mg/kg body weight per day
Animal (liquids omitted) weight per day) = (ppm of diet)
----------------------------------------------------------------------------------------------------
Mouse 0.02 3 0.150 7

Chick 0.40 50 0.125 8

Rat (young) 0.10 10 Dry 0.100 10
laboratory
Rat (old) 0.40 20 chow 0.050 20
diets
Guinea-pig 0.75 30 0.040 25

Rabbit 2.0 60 0.030 33

Dog 10.0 250 0.025 40
----------------------------------------------------------------------------------------------------
Cat 2 100 0.050 20

Monkey 5 250 Moist, 0.050 20
semi-solid
Dog 10 750 diets 0.075 13

Man 60 1500 0.025 40
----------------------------------------------------------------------------------------------------
Pig or sheep 60 2400 0.040 25

Cow 500 7500 Relatively 0.015 65
(maintenance) dry grain
forage
Cow 500 15 000 mixtures 0.030 33
(fattening)

Horse 500 10 000 0.020 50
----------------------------------------------------------------------------------------------------
^a Lehman, A.J. (1954) Association of Food and Drug Officials Quarterly Bulletin, 18: 66. The
values in this table are average figures, derived from numerous sources.

Example: What is the value in ppm and mg/kg body weight per day of
0.5% substance X mixed in the diet of a rat?

Solution: I. 0.5% corresponds to 5000 ppm.

II. From the table, 1 ppm in the diet of a rat is equivalent
to 0.050 mg/kg body weight per day. Consequently, 5000
ppm is equivalent to 250 mg/kg body weight per day
(5000 x 0.050).

INDEX

Absorption
Acetylcholinesterase inhibition
Acute studies
ADI,
conditional
temporary
Autolysis

Behavioural toxicity
Biochemical studies
Bioengineered organisms
Biological half-life
Biomarkers
Biorational products
Bipyridilium compounds
Body weights

Carbamates
Carboxylesterases
Carcinogenic pesticides
Carcinogenicity
classification schemes
genotoxic
limited evidence of
organophosphates
principles
studies
testing
Clearance
Clinical chemistry
Codex Committee on Pesticide Residues
Commonly occurring tumours
Comparative metabolic data
Comparative pharmacokinetic data

Delayed neurotoxicity
Dietary intakes
Dose/response
and safety factors
from accidental poisonings
in carcinogenicity
in delayed neuropathy
in human volunteers
relationships

Electron microscopic examination
Enterohepatic circulation
Environmental Health Criteria
Epidemiological studies

FAO
Food intake

Goitrogenic carcinogens
Good Laboratory Practices (GLP)

Haematological examinations
Half-life
Histopathological examinations
Historical control data
Human
Cell lines
Volunteers

Immunotoxicity
Impurities
In utero
In vitro
In vivo
Industrial Bio-Test Laboratories
Inert ingredients
Ingested dose
Intermediates
International Agency for Research on Cancer (IARC)
IPCS
Isomers

JECFA

Lactation
LD₀₁
LD₅₀
Long-term studies
and ADI
and reproduction
conduct of
interpretation of

Maternal toxicity
Maximum Tolerated Dose (MTD)
Mechanisms of toxicity
Metabolites
(animal)
(plant)
Michaelis-Menten kinetics
Mixtures
Mouse liver tumours
MRLs
Mutagenicity

Neuropathy target esterase (NTE)
Neurotoxicity
Nitrosamines
No-observed-adverse-effect-level (NOAEL)

Occupational exposure
Oncogenes
Ophthalmological effects
Organophosphates

Peroxisome proliferation
Pharmacokinetic data
Plasma cholinesterase
Poison Control Centres
Proprietary data
Protein binding
Pyrethroids

Radioactive labelling techniques
Re-evaluation of pesticides
Reproduction
and ADI
dose response
follow-up studies
maternally toxic doses
(multigeneration) study
Routes of exposure

Safety factors
and ADI
and TADI
determination of
in absence of toxicity
in carcinogenicity
Satellite groups
Screening teratology studies
Short-term studies
Special stains
Sperm measurements
Stability
Statistical analysis
Structure-activity relationships

Task Force of Past Presidents of the Society of Toxicology
Technical grade
Teratogenicity
Tetrachloro-dibenzo-p-dioxin (TCDD)
Threshold
Thyroid
Hormone
Neoplasia
Tolerance
Tumours
benign
malignant

Urinalysis

Validity of data

    See Also:
       Toxicological Abbreviations