Mycotoxins (EHC 11, 1979)

INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY

ENVIRONMENTAL HEALTH CRITERIA 11

MYCOTOXINS

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 World Health Organization or the United Nations
Environment Programme.

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

World Health Organization
Geneva, 1979

ISBN 92 4 154071 0

^(c) World Health Organization 1979

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 impression 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

ENVIRONMENTAL HEALTH CRITERIA FOR MYCOTOXINS

1. SUMMARY AND RECOMMENDATIONS FOR FURTHER RESEARCH

1.1 Summary
1.1.1 Aflatoxins
1.1.1.1 Sources and occurrence
1.1.1.2 Effects and associated exposures
1.1.2 Other mycotoxins
1.1.2.1 Ochratoxins
1.1.2.2 Zearalenone
1.1.2.3 Trichothecenes

1.2 Recommendations for further studies
1.2.1 General recommendations
1.2.2 Recommendations for aflatoxins
1.2.3 Recommendations for other mycotoxins

2. MYCOTOXINS AND HUMAN HEALTH

3. AFLATOXINS

3.1 Properties and analytical methods
3.1.1 Chemical properties
3.1.2 Methods of analysis for aflatoxins in foodstuffs
3.1.2.1 Sampling
3.1.2.2 Methods of analysis
3.2 Sources and occurrence
3.2.1 Formation by fungi
3.2.1.1 Moisture content and temperature
3.2.1.2 Invasion of field crops by A. flavus
3.2.2 Occurrence in foodstuffs
3.2.2.1 Maize
3.2.2.2 Wheat, barley, oats, rye, rice, and
sorghum
3.2.2.3 Groundnuts (peanuts)
3.2.2.4 Soybeans and common beans
3.2.2.5 Tree nuts
3.2.2.6 Copra
3.2.2.7 Cottonseed
3.2.2.8 Spices and condiments
3.2.2.9 Animal feeds
3.2.2.10 Animal products
3.2.3 Fate of aflatoxins during the handling and
processing of food
3.2.4 Pathways and levels of exposure

3.3 Metabolism
3.3.1 Absorption
3.3.2 Tissue distribution
3.3.2.1 Animal studies
3.3.2.2 Studies in man
3.3.3 Metabolic transformation and activation
3.3.4 Excretion
3.3.4.1 Animal studies
3.3.4.2 Studies in man

3.4 Effects in animals
3.4.1 Field observations
3.4.2 Experimental studies
3.4.2.1 Acute and chronic effects:
hepatotoxicity
3.4.2.2 Hepatotoxicity connected with
extrahepatic effects
3.4.2.3 Carcinogenesis
3.4.2.4 Teratogenicity
3.4.2.5 Mutagenicity
3.4.2.6 Biochemical effects and mode of action
3.4.2.7 Factors modifying the effects
and dose-response relationships of
aflatoxins

3.5 Effects in man -- epidemiological and clinical studies
3.5.1 General population studies
3.5.1.1 Liver carcinogenesis
3.5.1.2 Other effects reported to be
associated with aflatoxins
3.5.2 Occupational exposure

3.6 Evaluation of the health risks of exposure to aflatoxins
3.6.1 Human exposure conditions
3.6.1.1 Sources and levels of aflatoxins in
food
3.6.1.2 Dietary intake and levels in human
tissues
3.6.2 Acute effects of exposure
3.6.2.1 Acute liver disease
3.6.2.2 Reye's syndrome
3.6.3 Chronic effects of aflatoxin exposure
3.6.3.1 Cancer of the liver
3.6.3.2 Juvenile cirrhosis in India
3.6.4 Guidelines for health protection

4. OTHER MYCOTOXINS

4.1 Ochratoxins
4.1.1 Properties and analytical methods
4.1.1.1 Chemical properties
4.1.1.2 Methods for the analysis of foodstuffs
4.1.2 Sources and occurrence
4.1.2.1 Fungal formation
4.1.2.2 Occurrence in foodstuffs
4.1.3 Metabolism
4.1.3.1 Absorption
4.1.3.2 Tissue distribution and metabolic
conversion
4.1.3.3 Excretion
4.1.4 Effects in animals
4.1.4.1 Field observations
4.1.4.2 Experimental studies
4.1.5 Effects in man
4.1.5.1 Ochratoxin A and Balkan nephropathy
4.1.6 Conclusions and evaluation of the health risks
to man of ochratoxins
4.1.6.1 Experimental animal studies
4.1.6.2 Studies in man
4.1.6.3 Evaluation of health risks

4.2 Zearalenone
4.2.1 Properties, analytical methods, and sources
4.2.2 Occurrence
4.2.3 Effects in animals
4.2.3.1 Field observations
4.2.3.2 Experimental studies
4.2.4 Conclusions and evaluation of health risks to
man of zearalenone
4.2.4.1 Animal studies
4.2.4.2 Evaluation of health risks

4.3 Trichothecenes
4.3.1 Properties and sources
4.3.2 Occurrence
4.3.3 Effects in animals
4.3.3.1 Field observations
4.3.3.2 Experimental studies
4.3.4 Alimentary toxic aleukia
4.3.5 Conclusions and evaluations of the health risks
to man of trichothecenes

REFERENCES

NOTE TO READERS OF THE CRITERIA DOCUMENTS

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

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

WHO TASK GROUP ON ENVIRONMENTAL HEALTH CRITERIA FOR MYCOTOXINS

Members

Dr B. K. Armstrong, University Department of Medicine, Perth Medical
Centre, Nedlands, Australia^a

Dr A.D. Campbell, Food and Drug Administration, US Department of
Health, Education and Welfare, Washington, DC, USA^b

Dr T. Denizel, Department of Agriculture Microbiology, Faculty of
Agriculture, University of Ankara, Ankara, Turkey^a

Dr M. Jemmali, Service Mycotoxines de l'INRA, Station de Biochimie
et Physico-Chimie des Céréales, Institut National de la
Recherche Agronomique, Paris, France^a

Professor P. Krogh, The University Institute of Pathological
Anatomy, Copenhagen, Denmark,^a,b,c

Professor V. Kusak, Institute of Experimental Medicine,
Czechoslovak Academy of Sciences, Prague, Czechoslovakia
(Vice-Chairman)^a,b

Dr V. Nagarajan, National Institute of Nutrition, Jamai-Osmania,
Hyderabad, India^a

Dr M. F. Nesterin, Institute of Nutrition, Academy of Medical
Sciences of the USSR, Moscow, USSR^b

Professor P. Newberne, Department of Nutrition and Food Science,
Massachusetts Institute of Technology, Cambridge, MA, USA^a

Dr D. S. P. Patterson, Central Veterinary Laboratory, Ministry of
Agriculture, Fisheries and Food, Weybridge, England
(Chairman)^a,b

Dr F. G. Peers, Tropical Products Institute, London, England^a,b

Professor A. C. Sarkisov, Laboratory of Antibiotics and Mycology,
All-Union Institute of Experimental Veterinary Science, Moscow,
USSR^a

Dr P. L. Schuller, Laboratory of Chemical Analysis of Foodstuffs,
National Institute of Public Health, Bilthoven, Netherlands
(Vice-Chairman)^a

Dr A. Rogers, Department of Nutrition and Food Science,
Massachusetts Institute of Technology, Cambridge, MA, USA
(Rapporteur)^b

Professor H. D. Tendon, All-India Institute of Medical Sciences,
New Delhi, India^a

Professor A. Wasunna, Department of Surgery, University of
Nairobi, Kenya^a

Representatives of other International Organizations

Dr O. Alozie, United Nations Environment Programme,
Nairobi, Kenya^a,b

Dr D. Djordjevic, Occupational Safety and Health Branch,
International Labour Office, Geneva, Switzerland^a

Dr G. D. Kouthon, Food and Agriculture Organization of the
United Nations, Rome, Italy^a

Professor D. Reymond, Coordinating Committee on Food Chemistry,
International Union of Pure and Applied Chemistry, La Tour
de Peilz, Switzerland^b

Mrs. M. Th. van der Venne, Commission of the European Communities,
Health Protection Directorate, Luxembourg^a,b

WHO Secretariat

Dr C. Agthe, Environmental Health Criteria arid Standards, Division
of Environmental Health, WHO, Geneva, Switzerland
(Co-Secretary)^a,b

Dr L. Fishbein, US Public Health Service, National Centre for
Toxicological Research, Chemistry Division, Jefferson, AR, USA
(Temporary Adviser)^a

Dr J. Korneev, Environmental Health Criteria and Standards, Division
of Environmental Health, WHO, Geneva, Switzerland^a,b

Professor E. Lillehoj, US Department of Agriculture, Northern
Research Laboratory, Peoria, IL, USA (Temporary Adviser)^a

Dr C. A. Linsell, Interdisciplinary Programme and International
Liaison, IARC, Lyons, France^a,b

R. Lunt, Cancer Unit, Division of Noncommunicable Diseases, WHO,
Geneva, Switzerland^b

Dr Z. Matyas, Veterinary Public Health, Division of Communicable
Diseases, WHO, Geneva, Switzerland^a,b

Professor C. J. Mirocha, Department of Plant Pathology, University
of Minnesota, St Paul, MA, USA (Temporary adviser)^a

Dr J. Parizek, Environmental Health Criteria and Standards,
Division of Environmental Health, WHO, Geneva, Switzerland
(Co-Secretary)^a,b

Dr V. B. Vouk, Environmental Health Criteria and Standards, Division
of Environmental Health, WHO, Geneva, Switzerland^a,b

^a Attended the first meeting of the Task Group.
^b Attended the second meeting of the Task Group.
^c Present address: Department of Veterinary Microbiology,
Pathology and Public Health, School of Veterinary Medicine,
Purdue University, West Layayette, IN, USA.

ENVIRONMENTAL HEALTH CRITERIA FOR MYCOTOXINS

Members of the Task Group on Environmental Health Criteria for
Mycotoxins met in Geneva from 1 to 7 March 1977 and from 19 to 23
June 1978.

The first meeting was opened on behalf of the Director-General
by Dr B. H. Dieterich, Director, Division of Environmental Health,
and the second by Dr C. Agthe, Division of Environmental Health.

The first and second draft criteria documents were prepared by
Professor C. J. Mirocha. The comments on which the second draft was
based were received from the national focal points for the WHO
Environmental Health Criteria Programme in Belgium, Czechoslovakia,
Federal Republic of Germany, India, New Zealand, Poland, Sweden,
Thailand, USSR, and USA, and from the International Agency for
Research on Cancer (IARC), Lyons, the United Nations Industrial
Development Organization (UNIDO), Vienna, and the Food and
Agriculture Organization of the United Nations (FAO), Rome. Comments
were also received from the Tropical Products Institute, London.

Dr P. Krogh, Dr D. S. P. Patterson, and Dr P. L. Schuler
assisted in the preparation of the third draft criteria document,
which was submitted for review to all the members of the Task Group,
and to Dr R. Plestina of the Institute for Medical Research and
Occupational Health, Yugoslav Academy of Sciences and Arts, Zagreb,
Yugoslavia before the second meeting of the Task Group. The final
edited draft was kindly reviewed by Dr D. S. P. Patterson. The
collaboration of these national institutions, international
organizations, WHO collaborating centres, and individual experts is
gratefully acknowledged.

The document is based primarily on original publications listed
in the reference section. However, several recent publications
reviewing the occurrence, health effects, and other aspects of
mycotoxins have also been used including monographs prepared by
Purchase (1974), Pokrovskij et al. (1977) and Wyllie & Morehouse
(1977), and the report on the joint FAO/ WHO/UNEP Conference on
Mycotoxins in Nairobi 1977 (FAO, 1977). In addition, comprehensive
data have been obtained from the proceedings of several symposia and
meetings including the Conference on Mycotoxins in Human and Animal
Health, held in Maryland, USA, in 1976 (Rodricks et al., 1977).

Details of the WHO Environmental Health Criteria Programme
including some of the terms frequently used in the documents, may be
found in the general introduction to the Environmental Health
Criteria Programme published together with the environmental health
criteria document on mercury (Environmental Health Criteria 1,
Mercury, Geneva, World Health Organization, 1976), and now available
as a reprint.

1. SUMMARY AND RECOMMENDATIONS FOR FURTHER RESEARCH

1.1 Summary

The ingestion of food containing mycotoxins, the toxic products
of microscopic fungi (moulds), may have serious adverse health
effects in man. Occasionally, occupational exposure to airborne
mycotoxins may also occur.

The occurrence of mycotoxins in foodstuffs depends on their
formation by specific strains of fungi and is influenced by
environmental factors such as humidity and temperature. Thus,
mycotoxin contamination of foodstuffs may vary with geographical
conditions;, production and storage methods, and also with the type
of food, since some food products are more suitable substrates for
fungal growth than others.

The present document contains an evaluation of health risks
associated with four classes of mycotoxins. Aflatoxins are treated
in most detail because more is known about them than about the other
mycotoxins and because there is epidemiological evidence associating
health effects in man with exposure to aflatoxins.

For the other 3 classes (ochratoxins, zearalenone, and
trichothecenes), toxic effects in animals have been established and
there is well-documented evidence that human exposure may occur, at
least for the first two classes.

1.1.1 Aflatoxins

1.1.1.1 Sources and occurrence

Aflatoxins are produced by certain strains of Aspergillus
flavus and Aspergillus parasiticus. These fungi are ubiquitous
and the potential for contamination of foodstuffs and animal feeds
is widespread. The occurrence and magnitude of aflatoxin
contamination varies with geographical and seasonal factors, and
also with the conditions under which a crop is grown, harvested, and
stored. Crops in tropical and subtropical areas are more subject to
contamination than those in temperate regions, since optimal
conditions for toxin formation are prevalent in areas with high
humidity and temperature. Toxin-producing fungi can infect growing
crops as a consequence of insect or other damage, and may produce
toxins prior to harvest, or during harvesting and storage.

The chemical structures of aflatoxins have been elucidated, and
analytical techniques are available for their identification and
determination in foodstuffs and body tissues at the µg/kg level and
lower. Four aflatoxins (B₁, G₁, B₂, G₂, often occurring
simultaneously, have been detected in foods of plant origin
including maize, groundnuts (peanuts), and tree nuts as well as many
other foodstuffs and feeds.

In animals, ingested aflatoxins may be metabolically degraded.
Aflatoxin B₁ may be converted into aflatoxin M₁ which may occur
in the milk. The concentration of aflatoxin M₁ in the milk of cows
is about 300 times lower than the concentration of aflatoxin B₁
consumed in the feed. In certain experimental animals, only small
amounts of administered aflatoxins have been found in tissues, 24 h
after injection.

In studies on pigs, aflatoxin residues were detected in the
liver, kidney, and muscle tissues of animals given aflatoxins in the
feed for several months. There do not appear to be any published
works on aflatoxin residues in the tissues of slaughtered animals.

The use of resistant varieties of seed and of pesticides, and
careful drying and storing procedures can reduce fungal infestation
and thus diminish food contamination by aflatoxins. The toxin is not
eliminated from foodstuffs or animal feeds by ordinary cooking or
processing practices and, since pre-and post-harvest procedures do
not ensure total protection from aflatoxin contamination, techniques
for decontamination have been developed. The toxin is generally
concentrated in a small proportion of seeds that are often different
in colour. Segregation of discoloured seeds by sorting can
significantly reduce the aflatoxin levels in some crops, such as
groundnuts. Visual inspection for mould growth before processing can
serve as an initial screening technique but toxin-producing fungi
can be present without detectable aflatoxins and vice versa. Because
aflatoxin distribution in a contaminated, unprocessed commodity is
uneven, adequate sampling is essential for effective monitoring. As
aflatoxins can be chemically degraded in vitro by several oxidizing
agents and alkalis, hydrogen peroxide and ammonia are currently used
for the chemical decontamination of animal feeds.

1.1.1.2 Effects and associated exposures

Outbreaks of aflatoxicosis in farm animals have been reported
from many areas of the world. The liver is mainly affected in such
outbreaks and also in experimental studies on animals, including
nonhuman primates. The acute liver lesions are characterized by
necrosis of the hepatocytes and biliary proliferation, and chronic
manifestations may include fibrosis. A feed level of aflatoxin as
low as 300 µg/kg can induce chronic aflatoxicosis in pigs within
3-4 months.

Aflatoxin B₁ is a liver carcinogen in air least 8 species
including nonhuman primates. Dose-response relationships have been
established in studies on rats and rainbow trout, with a 10% tumour
incidence estimated to occur at feed levels of aflatoxin B₁ of
1 µg/kg, and 0.1 µg/kg, respectively. In some studies, carcinomas of
the colon and kidney have been observed in rats treated with
aflatoxins. Aflatoxin B₁ causes chromosomal aberrations and DNA
breakage in plant and animal cells, and, after microsomal
activation, gene mutations in several bacterial test systems. In
high doses, it may be teratogenic.

The acute toxicity and carcinogenicity of aflatoxins are greater
in male than in female rats; hormonal involvement may be responsible
for this sex-linked difference. Nutritional status in animals,
particularly with respect to lipotropes, proteins, vitamin A, and
lipids (including cyclopropenoid fatty acids), can modify the
expression of acute toxicity or carcinogenicity or both.

There is little information on the association of acute
hepatoxicity in man with exposure to aflatoxins but cases of acute
liver damage have been encountered that could possibly be attributed
to acute aflatoxicosis. A recent outbreak of acute hepatitis in
adjacent districts of two neighbouring states in north-west India,
which affected several hundred people, was apparently associated
with the ingestion of heavily contaminated maize, some samples of
which contained aflatoxin levels in the mg/kg range, the highest
reported level being 15 mg/kg.

Liver cancer is more common in some regions of Africa and
southeastern Asia than in other parts of the: world and, when local
epidemiological information is considered together with experimental
animal data, it appears that increased exposure to aflatoxins may
increase the risk of primary liver cancer. Pooled data from Kenya,
Mozambique, Swaziland, and Thailand, show a positive correlation
between daily dietary aflatoxin intake (in the range of 3.5 to
222.4 ng/kg body weight per day) and the crude incidence rate of
primary liver cancer (ranging from 1.2 to 13.0 cases per 100 000
people per year). There is also some evidence of a vital involvement
in the etiology of the disease.

In view of the evidence concerning the effects, particularly the
carcinogenic effects, of aflatoxins in several animal species, and
in view of the association between aflatoxin exposure levels and
human liver cancer incidence observed in some parts of the world,
exposure to aflatoxins should be kept as low as practically
achievable. The tolerance levels for food products established in
several countries should be understood as management tools intended
to facilitate the implementation of aflatoxin control programmes,
and not as exposure limits that necessarily ensure health
protection.

1.1.2 Other mycotoxins

1.1.2.1 Ochratoxins

Ochratoxins are produced by several species of the fungal genera
Aspergillus and Pencillium. These fungi are ubiquitous and the
potential for contamination of foodstuffs and animal feed is
widespread. Ochratoxin A, the major compound, has been found in more
than 10 countries in Europe and the USA. Ochratoxin formation by
Aspergillus species appears to be limited to conditions of high
humidity and temperature, whereas at least some Pencillium species
may produce ochratoxin at temperatures as low as 5°C.

Analytical techniques have been developed for the identification
and quantitative determination of ochratoxin levels in the µg/kg
range.

Ochratoxin A has been found in maize, barley, wheat, and oats, as
well as in many other food products, but the occurrence of
ochratoxin B is rare. Residues of ochratoxin A have been identified
in the tissues of pigs in slaughterhouses, and it has been shown,
under experimental conditions, that residues can still be detected
in pig tissues one month after the termination of exposure.

Field cases of ochratoxicosis in farm animals (pigs, poultry)
have been reported from several areas of the world, the primary
manifestation being chronic nephropathy. The lesions include tubular
atrophy, interstitial fibrosis, and, at later stages, hyalinized
glomeruli. Ochratoxin A has been found to be nephrotoxic in all
species of animals studied so far, even at the lowest level tested
(200 µg/kg feed in rats and pigs). It has also been reported to
produce teratogenic effects in mice, rats, and hamsters.

Human endemic nephropathy is a kidney disease of unknown
etiology that has so far only been encountered in some areas of the
Balkan Peninsula. The renal changes observed with this disease are
comparable to those seen in ochratoxin A-associated nephropathy in
pigs. High ochratoxin A exposure through diet has been found in some
of the areas of the Balkan Peninsula, where endemic nephropathy is
prevalent.

1.1.2.2 Zearalenone

Zearalenone, a metabolite produced by various species of
Fusarium, has been observed as a natural contaminant of cereals,
in particular maize, in many countries in Africa and Europe, and in
the USA.

It has been shown to produce estrogenic effects in animals, and
field cases of a specific estrogenic syndrome in pigs and of
infertility in cattle have been encountered in association with feed
levels of zearalenone of 0.1-6.8 mg/kg and 14 mg/kg, respectively.
The compound has also produced congenital malformations in the rat
skeleton.

In some countries, zearalenone has been found in samples of
cornmeal and cornflakes destined for human consumption, at levels up
to 70 µg/kg, corresponding to doses 400-600 times lower than those
causing effects in monkeys or mice under experimental conditions. In
certain areas of Africa, substantially higher levels have
occasionally been found in beer and sour porridge prepared from
contaminated maize and sorghum.

No adverse effects due to zearalenone intake have been reported
in man, so far, but a possible health hazard connected with the
dally intake of zearalenone at levels such as those reported for
African fermented preparations needs further attention.

1.1.2.3 Trichothecenes

Trichothecene toxins belong to a group of closely related
chemical compounds produced by several species of Fusarium,
Cephalosporium, Myrothecium, Trichoderma, and Stachybotrys.
Four trichothecenes (T-2 toxin, nivalenol, deoxynivalenol, and
diacetoxyscirpenol) have been detected as natural contaminants in a
small number of food samples.

Alimentary toxic aleukia, a disease diagnosed in man about 40
years ago, was apparently associated with the ingestion of grains
invaded by Fusarium species. No cases have been reported since the
end of the Second World War and the disappearance of the disease is
probably due to improved food production and storage conditions.
There is no firm evidence connecting the recently identified
trichothecenes with alimentary toxic aleukia occurring in the past,
or with other human disease.

1.2 Recommendations for Further Studies

1.2.1 General recommendations

(a) There is a need for more information concerning the
occurrence of mycotoxins in various parts of the world and
the possible daily intake of mycotoxins by man.

(b) Further studies should be undertaken on factors affecting
fungal growth and mycotoxin formation in foodstuffs, under
preharvest, postharvest, and storage conditions.

(d) Better methods should be developed for the rapid detection
and measurement of mycotoxin levels in foodcrops.

(e) Sampling has proved to be the most difficult step in the
surveillance of food commodities. The development of
reliable, internationally accepted sampling procedures is
strongly recommended.

(f) Better methods should be developed for the identification
and measurement of mycotoxins in human tissues, body fluids,
and excreta.

(g) A network of reference centres should be established to
assist Member States in confirming the identity of
individual mycotoxins found in human foods and tissues.
These reference centres should also provide mycotoxin
reference samples, upon request, to reinforce the
inter-comparability of analytical results obtained in
different parts of the world.

(h) Better understanding is needed of the role of mycotoxins in
human diseases. Where association between exposure to
mycotoxins and the incidence of certain diseases is
suspected, detailed epidemiological studies should be
carried out.

(i) Improved diagnostic methods for the effects on health of
mycotoxins are needed, particularly methods for the
detection of early changes that occur before the development
of irreversible effects.

(j) Attempts should be made to monitor exposure levels and to
search for effects in workers handling pure mycotoxins or
contaminated materials. This could provide important
information on the effects of chronic exposure to mycotoxins
and also indicate the need for safety measures.

1.2.2 Recommendations for aflatoxins

(a) The validity of the assumption of a causal relationship
between aflatoxin ingestion and primary liver cancer should
be examined further by introducing control measures to
reduce aflatoxin exposure in areas of high liver cancer
incidence and high aflatoxin exposure. This should be
followed by the monitoring of liver cancer incidence in
these areas and in comparable areas where the aflatoxin
exposure has been low.

(b) The prevalence of hepatitis B antigen should be determined
in areas with various levels of aflatoxin exposure and a
high incidence of primary liver cancer.

(c) Suspected outbreaks of acute aflatoxicosis should be
studied in detail. Such studies should include measurements
of the exposure to aflatoxins through foods and other
routes. The presence of aflatoxins and their derivatives in
the tissues and excreta of individuals including both those
affected and those apparently unaffected by the disease,
should be investigated.

(d) A prolonged, continuous surveillance of the health status
of exposed populations is considered essential in
localities, where outbreaks of acute aflatoxicosis have
occurred. Such follow-up studies are important to fill the
gaps in knowledge on the late effects of short-term exposure
to high levels of aflatoxin in man. Recently reported
outbreaks of aflatoxin-associated hepatitis in southeastern
Asia may provide an ideal opportunity for such studies.

(e) Reports from the various countries on the presence of
aflatoxins in human tissues, body fluids, and excreta should
be confirmed using specific assay methods with adequate
limits of detection. The frequency of such events should be
studied in appropriate samples of the general population of
various countries and a search made for the sources of
aflatoxin exposure.

(f) The implication of aflatoxin as a contributing factor in
the development of Reye's syndrome should be further
investigated using the case-control approach. Data should be
obtained on the presence of aflatoxins in tissues, body
fluids, and excretion products for each case and its
control, and attempts should be made to identify dietary
sources of aflatoxins.

(g) More information is needed on the gastrointestinal
absorption of aflatoxins in animals and human subjects, as
well as on the rate of disappearance of aflatoxins from
farm-animal and human tissues. This is important both for
the evaluation of aflatoxin residues in food of animal
origin, and for the evaluation of aflatoxin levels in human
tissues as a means of assessing exposure.

(h) The modifying effect of the dietary intake of lipotropes,
protein, or vitamin A on aflatoxin-related carcinogenesis
should be further studied in experimental animals. This
aspect should also be included in epidemiological studies on
the association between human liver cancer incidence and
aflatoxin intake.

1.2.3 Recommendations for other mycotoxins

(a) The levels of ochratoxin A and possibly citrinin should be
measured in "food-on-the-plate" in areas of the Balkan
Peninsula with different incidence rates of Balkan
nephropathy.

(b) Further systematic investigations are needed on the levels
of ochratoxin in foodstuffs and animal feeds in different
parts of the world, and their association with nephropathy
in farm animals. More work is needed in various parts of the
world to confirm or exclude the strictly localized
occurrence of endemic nephropathy affecting human subjects
and considered so far to be confined to certain areas of the
Balkan Peninsula.

(c) Further studies are required on the mechanisms of ochratoxin
toxicity, and on possible interactions with other nephrotoxic
agents.

(d) Studies should be made in different countries of the levels
of zearalenone in human food and on total daily intake.

(e) More information is needed on the levels of zearalenone in
foods prepared from fermented maize and sorghum, such as
those found in certain parts of Africa, and on the possible
adverse effects of the daily consumption of these products,
particularly in view of the estrogen-like effects of
zearalenone observed in animals.

2. MYCOTOXINS AND HUMAN HEALTH

The toxicity of certain mushrooms has been known for a long
time. However, the potential human hazard of the toxic products of
other fungi was not recognized until the 1850s when an association
between the ingestion of rye infected with Claviceps purpurea and
the clinical features of ergotism was discovered. This was followed
by reports of other mycotoxicoses that affected man such as the
identification of a syndrome associated with the ingestion of bread
infected by Fusarium graminearum, recognition of human
stachybotryotoxicosis, and studies on the association between
alimentary toxic aleukia (ATA) and the ingestion of over-wintered
grains infested with Fusarium poae and Fusarium sporotrichioides
(Sarkisov, 1954).

Recognition of the association of ATA with the consumption of
food contaminated by moulds and the corresponding preventive
measures taken, resulted in the eradication of the disease (Leonov,
1977) showing that, even before the isolation of the first
mycotoxins, fungi-related foodborne diseases could be prevented.

The discovery of the hepatotoxic and hepatocarcinogenic
properties of Aspergillus flavus in the early 1960s quickly followed
by the elucidation of the structure of the aflatoxins changed the
control strategy in the whole field of mycotoxins. A more
quantitative approach is now possible, based primarily upon the
chemical determination of the toxins and on studies of their effects
in relation to dose.

In spite of increasing knowledge concerning human mycotoxicoses,
the majority of data available on mycotoxins and mycotoxicoses have
been obtained from veterinary medicine. Field studies, as well as
studies on experimental animals indicate that the potential toxicity
of mycotoxins is great. Future investigations may well establish a
causal role of mycotoxins in other human diseases besides those
considered so far.

Almost all plant products can serve as substrates for fungal
growth and subsequent mycotoxin formation, thus providing the
potential for direct contamination of human food. When farm animals
used for food production, ingest feed contaminated with mycotoxins,
not only may a direct toxic effect on the animals occur but there
may also be a carry-over of the toxins into milk and meat, thus
creating a further avenue for human exposure to mycotoxins.
Furthermore, occupational exposure may occur through other media
such as air.

In this document, the risks of health effects have been
considered only for those mycotoxins for which there is evidence of
human exposure and of well defined adverse effects, at least in
animals. This category includes the aflatoxins, ochratoxins, and
zearalenone. The trichothecenes have also been included, as these
have been shown, more recently, to be produced by fungi, that were
reported to be associated with outbreaks of human illness several
decades ago (ATA).

During recent years many other mycotoxins have been discovered
such as: citreoviridin; citrinin; cyclochlorotine; luteoskyrin;
maltoryzine; patulin; P R toxin; rubratoxin; rugulosin;
sterigmatocystine; and tremorgens.

Some of these toxins, which are not discussed in this document,
have been suggested to be related to disease outbreaks in farm
animals (Pier et al., 1977). Certain human diseases, suspected of
being associated with mycotoxins (van Rensburg, 1977), have not been
discussed in this document as the causative agents have not been
identified.

Of the four groups of mycotoxins considered only aflatoxins have
been shown to be associated with well recognized human health
effects. For this reason, they are treated separately from the other
mycotoxins.

3. AFLATOXINS

3.1 Properties and Analytical Methods

3.1.1 Chemical properties

Although 17 compounds, all designated aflatoxins, have been
isolated, the term aflatoxins usually refers to 4 compounds of the
group of bis-furanocoumarin metabolites produced by Aspergillus
flavus and A. parasiticus, named B₁, B₂, G₁ and G₂,
which occur naturally in plant products. The 4 substances are
distinguished on the basis of their fluorescent colour, B standing
for blue and G for green with subscripts relating to the relative
chromatographic mobility. Cows fed rations containing aflatoxin B₁
and B₂ excrete metabolites in the milk called aflatoxin M₁ and
aflatoxin M₂ (see section 3.3.4.1); M stands for milk, and again
the subscripts relate to the relative chromatographic mobility.
(Aflatoxin M₁ is also a fungal metabolite.) Of the 4 major
aflatoxins, B₁ is usually found in the highest concentrations,
followed by G₁ while B₂ and G₂ occur in lower concentrations.

The structures of a number of aflatoxins and of aflatoxin
B₁-related metabolites (see section 3.3.3) are illustrated in
Fig. 1. The structure of aflatoxins B₁ and G₁ were determined
by Asao et al. (1963, 1965) and that of B₂ by Chang et al. (1963).
Aflatoxins B₂ and G₂ are dihydroderivatives of the parent
compounds (Hartley et al., 1963). Aflatoxins M₁ and M₂ are the
hydroxylated metabolites of B₁ and B₂, respectively (Holzapfel
et al., 1966; Masri et al., 1967; Buchi & Weinreb, 1969). Chemical
properties of some naturally-occurring aflatoxins and metabolites
are summarized in Table 1.

The aflatoxins are intensely fluorescent, when exposed to
long-wave ultraviolet (UV) light. This makes it possible to detect
these compounds at extremely low levels (ca. 0.5 ng or less per spot
on thin-layer chromatograms) and provides the basis for practically
all the physicochemical methods for their detection and
quantification. A concentration of aflatoxin M₁ of 0.02 µg/litre
can be detected in liquid milk (Schuller et al., 1977).

Aflatoxins are freely soluble in moderately polar solvents
(e.g., chloroform and methanol) and especially in dimethylsulfoxide
(the solvent usually used as a vehicle in the administration of
aflatoxins to experimental animals); the solubility of aflatoxins in
water ranges from 10-20 mg/litre.

As pure substances, the aflatoxins are very stable at high
temperatures, when heated in air. However, they are relatively
unstable, when exposed to light, and particularly to UV radiation,
and air on a TLC plate and especially when dissolved in highly polar
solvents. Chloroform and benzene solutions are stable for years if
kept in the: dark and cold.

Little or no destruction of aflatoxins occurs under ordinary
cooking conditions, and heating for pasteurization. However,
roasting groundnuts appreciably reduces the levels of aflatoxins
(see section 3.2.3) and they can be totally destroyed by drastic
treatment such as autoclaving in the presence of ammonia or by
treatment with hypochlorite.

Table 1. Physical and chemical properties of some aflatoxins and their metabolites

Aflatoxin Molecular Relative Melting Ultraviolet absorption (epsilon)^c Fluorescence Reference
formula molecular point emission
mass °C 265 nm 360-362 nm nm

B₁^a C₁₇H₁₂O₆ 312 268-269 12 400 21 800 425 Asao et al. (1965)
B₂^a C₁₇H₁₄O₆ 314 286-289 12 100 24 000 425 Chang et al. (1963)
G₁^a C₁₇H₁₂O₆ 328 244-246 9 600 17 700 450 Asao et al. (1965)
G₂^a C₁₇H₁₄O₇ 330^d 237-240^d 8 200 17 100 450 Hartley et al. (1963)
M₁^a C₁₇H₁₂O₇ 328 299 14 150 21 250 (357 nm) 425 Holzapfel et al. (1966)
M₂^a C₁₇H₁₄O₇ 330 293 12 100 (264 nm) 22 900 (357 nm) f Holzapfel et al. (1966)
P₁^b C₁₆H₁₀O₆ 298 >320 11 200 (267 nm) 15 400 (362 nm) f Dalezios et al. (1971 ) and
14 900 (342 nm) Buchi et al. (1973)

Q₁ C₁₇H₁₂O₇ 328 e 11 450 (267 nm) 17 500 (366 nm) f Masri et al. (1974a,b)
Aflatoxicol C₁₇H₁₄O₆ 314 230--234^d 10 800 (261 nm) 14 100 (375 nm) 425 Detroy & Hesseltine (1970)

^a Molar absorption coefficient for aflatoxins B₁, B₂, G₁, and G₂ obtained from Rodricks et al. (1970) and those for M₁ and M₂
from Stubblefield et al, (1972).
^b P stands for phenolic products of O-demethylation of aflatoxin B₁.
^c Compounds dissolved in methanol except for aflatoxin P₁ which in this case was dissolved in ethanol. Data on molar absorption
coefficients for other peaks and on the ultraviolet absorption characteristics of aflatoxins in other solvents can be found
in the original papers.
^d Data from Butler (1974).
^e Not available.
^f Violet fluorescence of aflatoxin M₂ and yellow-green fluorescence of aflatoxins P₁ and Q₁ reported in original papers.

The presence of a lactone ring in the aflatoxin molecule makes
them susceptible to alkaline hydrolysis (De Iongh et al., 1962).
This characteristic is important in that any food processing
involving alkali treatment can decrease the contamination of the
products (section 3.2.3) although the presence of protein, the pH,
and the duration of treatment may modify the results (Beckwith et
al., 1975). However, if the alkaline treatment is mild,
acidification will reverse the reaction to reform the original
aflatoxin.

The chemistry of the aflatoxins has recently been reviewed by
Roberts (1974).

3.1.2 Methods of analysis for aflatoxins in foodstuffs

3.1.2.1 Sampling

Sampling is an integral part of the analytical procedure and the
sample drawn should be representative of the lot. The total error
made in an analytical procedure consists of the sampling error, the
subsampling error, and the error in analysis (Whitaker, 1977). The
difficulty in sampling for aflatoxins arises because of the
heterogeneity of aflatoxin distribution in contaminated unprocessed
commodities. On the basis of a large number of analyses, Whitaker et
al. (1974a) were able to calculate the contribution of each error to
the total error. The total variance of the analytical procedure is
primarily caused by sampling variability, whereas the subsampling
variability and the analysis variability are more or less
independent of aflatoxin concentration. The coefficient of variation
associated with sampling is about 115% at a level of contamination
of 20 µg/kg and about 145% at a level of contamination of 10 µg/kg.
Whitaker et al. (1974b) and Whitaker (1977) have summarized a
procedure for sampling and have developed a number of sampling plans
used in the USA for the control of aflatoxin contamination in
shelled groundnuts (peanuts). Other recent publications deal with
aflatoxin-testing programmes for maize (Whitaker et al., 1978) and
cottonseed (Whitaker & Whitten, 1977).

The sampling of small grains, oilseed cakes, foodstuffs, and
feeds is also difficult, although in most cases the aflatoxin
distribution within a batch is not likely to be as uneven as in the
case of groundnuts. Fluids and well-mixed processed products such as
milk and milk products, beer, and cider do not present such a
sampling problem.

Peanut butter, flours, and cornmeal do not present the same
problems as the original raw materials because a finely divided
product is formed during processing from which it is much easier to
obtain a representative analytical sample.

Some practical aspects of sampling are dealt with in Chapter 26
of "Official methods of analysis" of the Association of Official
Analytical Chemists (Horwitz et al., 1975). For survey purposes,
1-5 kg samples are usually taken and the size of the sample for
analysis ranges from 20 to 100 g. A sample of 50 g ensures both a
representative sample and solvent economy.

3.1.2.2 Methods of analysis

Biological and chemical procedure have been developed for the
detection and determination of aflatoxins and other mycotoxins. The
bioassay techniques that are currently available are not suitable
for routine screening purposes and their detection levels are not
low enough. The chemical assay techniques, although more accurate
and faster, are not always specific. The presence of a certain toxin
is usually confirmed by derivative formation and its toxicity
verified by bioassay.

Biological methods. In the original biological test (Carnaghan
et al., 1963), one-day-old ducklings were used as test animals for
determining the presence of aflatoxins in suspect food by measuring
the degree of biliary proliferation as a semiquantitative index (see
section 3.4.2.1). The lowest dose level of 0.4 µg/day administered
for 5 days represents the minimum intake required to induce a
detectable biliary proliferation. The test is also effective for
detecting aflatoxin M₁ in both liquid and powdered milk (Purchase,
1967). Little information is available on the sensitivity of this
test (and other biological methods) to aflatoxins other than B₁.

A commonly used method in regulatory actions is the chicken
embryo bioassay in which 0.1-0.2 µg of aflatoxin B₁ is applied to
the egg membrane and the mortality rate recorded during the 23-day
period of hatching (Horwitz, 1975).

Several other biological procedures have been developed, using
maize seedlings, zebra fish larvae, brine shrimps, bacteria etc.
Detailed descriptions can he found in the reviews by Goldblatt
(1969) and Ciegler et al. (1971).

Chemical methods. Although procedures are continually
changing, the basic steps remain; extraction, lipid removal,
cleanup, separation, and quantification. Since there is considerable
overlap in the various methods, most of the published reviews
(Jones, 1972; Stoloff, 1972) examine the different procedures by
these basic steps. Depending on the nature of the commodity, methods
can sometimes be simplified by omitting unnecessary steps. The
presence of specific interferences such as theobromine in cacao and
gossypol in cottonseed, may require additional steps.

Numerous methods of analysis have been reported for the
determination of aflatoxins in human and animal foodstuffs. Many of
them are minor modifications of the basic steps adapted to special
commodities or problems.

Collaborative studies designed to assess the performances of
different laboratories give information on the accuracy, precision,
and specificity of the method under consideration, as well as on the
occurrence of false negative and false positive results. Only
methods that have been subjected to such studies are reported in
this document.

Chemical methods have mainly been developed for such commodities
as groundnuts. In the first method for the analysis of groundnuts,
the aflatoxins were extracted from the contaminated sample using
methanol; this was later replaced by chloroform. An improvement was
made by Lee (1965) who showed that the addition of water to
hydrophilic plant tissues during extraction with chloroform resulted
in more effective removal of aflatoxin. The combination of
liquid-liquid extraction techniques and partition chromatography led
to a method, which is now one of the most widely used, known as the
Contamination Branch (CB) method (Eppley, 1966). The sample is
extracted with water and the water extracted with chloroform, the
lipids and aflatoxins are transferred to a silica-gel column where
the lipids are selectively eluted with hexane and the pigments and
other interfering material eluted with absolute diethylether;
finally the aflatoxins are eluted from the column with 3% methanol
in chloroform.

Because the CB method is time-consuming, attempts have been made
to simplify it. Waltking et al. (1968) drew attention to the fact
that a separation funnel was simpler and faster for liquid-liquid
partition than the silica-gel column, and that centrifuging was a
faster method of separating a solid than filtration. Thus the Best
Foods (BF) method was developed, which is faster and more economical
in terms of the amounts of solvents used but provides a poorer
cleanup. The sample is extracted and defatted with a two-phase
aqueous methanol-hexane system, the aflatoxins are then partitioned
from the aqueous phase into chloroform, leaving lipids and pigments
in the hexane and aqueous methanol.

In both the CB and BF methods, the aflatoxins are concentrated
by evaporation of the chloroform, and then separated by thin-layer
chromatography (TLC). Aflatoxins are intensely fluorescent when
exposed to long-wave ultraviolet radiation, which makes it possible
to determine these compounds at extremely low levels. An analyst
experienced in this field can detect 0.5 ng aflatoxin B₁ on a TLC
plate. In most methods, the intensity of fluorescence of the sample
is compared with that of a standard. Under ideal conditions this
technique; has a coefficient of variation of about 20% which can be
reduced to 5%-9% by the use of a fluorodensitometer.

It should be pointed out that quantification and confirmation of
identity can only be obtained if pure authentic standards are
available for reference. Current sources of aflatoxin standards and
methods for the determination of mass concentration and purity cart
be found in Chapter 26 of the AOAC "Official methods of analysis"
(Horwitz et al., 1975).

When the CB and BF methods were compared in a collaborative
study (Waltking, 1970), the methods were found to be equivalent in
accuracy and precision with a recovery of about 70% of added
aflatoxin and an overall coefficient of variation of about 35% for
total aflatoxin levels down to about 20 µg/kg. This result has been
confirmed by the latest International Aflatoxin Check Sample Study
(Coon et al., 1972) in which 129 laboratories participated. However,
the coefficient of variation was about twice as high as in the
original collaborative studies. This illustrates the inadequacies of
many laboratories.

Another collaborative study was conducted (Stack, 1974) in which
the CB and BF methods were compared at levels down to the 2-10 µg/kg
range, Again both methods proved to be equally accurate (about 80%
recovery) for total aflatoxins at the 5-10 µg/kg levels. The CB
method, however, was as precise (coefficient of variation = 30%) at
the lowest level of 2 µg/kg as at the highest levels. The BF method
lost precision at these low levels and the coefficient of variation
was of the order of 100%.

In spite of cleanup and separation procedures, there might still
be problems with compounds that have fluorescent and chromatographic
properties similar to those of the aflatoxins. Thus, the presence of
a spot on a TLC plate is only presumptive evidence of identity and
additional confirmatory tests are necessary. Probably the first step
in confirming the presence of aflatoxins is to use additional
solvent systems in the TLC. The developed TLC plate can be sprayed
with 25% sulfuric acid (Schuller et al., 1967), which changes the
fluorescence colour of the aflatoxin spots to yellow. This test, if
negative, would rule out the presence of aflatoxins but does not
provide confirmatory evidence. The formation of chemical derivatives
was first described for aflatoxins B₁ and G₁ (Andrellos & Reid,
1964). The reagents used were formic acid-thionyl chloride, acetic
acid-thionyl chloride, and trifluoroacetic acid, the acid-catalysed
addition products formed were a dimeric acetate and an addition
product with water, respectively. The characteristic mobilities and
fluorescent properties on thin-layer chromatograms can be compared
with those of standard derivatives. Pohland et al. (1970) simplified
the preparation of the derivatives by using a mixture of
hydrochloric acid and acetic anhydride, and hydrochloric acid alone.
Further improvement, by elimination of the preparative
chromatography step in these procedures, has been achieved by
Przybylski (1975). The water adduct is formed directly on a TLC
plate from as little as 0.5 ng of aflatoxin B₁ or G₂.

In addition to the procedures for groundnuts, methods of
analysis have been developed for cottonseed, copra, maize, various
tree nuts (pistachio, walnut, Brazil nuts, etc.) and for animal
feeds. Many of these methods are modifications of the CB and BF
methods. Milk and dairy products require a far greater sensitivity
for the determination of M₁ and M₂ because these animal
metabolites are usually only found at sub µg/kg levels; additional
cleanup to eliminate interferences and sometimes two-dimensional TLC
techniques (Schuller et al., 1973) are necessary to attain
satisfactory performance. These methods are described in Chapter 26
of the AOAC "Official methods of analysis" (Horwitz et al., 1975).

Several analytical methods for the detection of aflatoxin
residues in animal tissues have been developed, and their detection
limits evaluated (Jemmali & Murphy, 1976).

Column detection methods are being used for control purposes in
the field because of their simplicity. The method of Romer (1975) is
of particular value because it combines column detection, TLC
quantification, and TLC plate chemical derivative confirmation in a
method that has a wide application for a number of foodstuffs and
feeds including mixed feeds.

It appears that methods using high-pressure liquid
chromatography will become the methods of choice for mycotoxin
analyses in the near future because of their sensitivity and
improved accuracy, and because they can be applied to a number of
mycotoxins including aflatoxins B₁, B₂, G₁, and G₂ (Panalaks
& Scott, 1977).

3.2 Sources and Occurrence

3.2.1 Formation by fungi

The ability to produce aflatoxins seems to be confined to
strains of the two species Aspergillus flavus Link and
A. parasiticus Speare, both members of the A. flavus group.

Aflatoxin-producing strains of A. flavus are common and
widespread, and have been isolated from a host of different
materials. As indicated in Table 2, a high proportion (from 20% to
98%) of isolated strains of A. flavus is able to produce
aflatoxins.

Table 2. Aflatoxin-producing strains of A. flavus isolated from
four field crops^a

Source Isolates Isolates producing Maximum yield
tested aflatoxin of aflatoxin B₁
(No.) (%) (µg/flask)

groundnut 100 98 3300
cottonseed 59 81 3200
rice 127 20 1100
sorghum 63 24 3300

^a Data from Schroeder & Boller quoted by Hesseltine (1976) in
"Mycotoxins and other fungal related food problems".

3.2.1.1 Moisture content and temperature

The moisture content of the substrate and temperature are the
main factors regulating fungal growth and mycotoxin formation.

Koehler (1938) established that a moisture content of 18.3% on a
wet weight basis, was the lower limit for the growth of A. flavus
in shelled corn. Extensive studies under precisely controlled
conditions (Sanders et al., 1968; Diener & Davis, 1969; Davis &
Diener, 1970) established a moisture content in equilibrium with a
relative humidity of 85% (or water activity (a_w) = 0.85) as the
lower limit for growth of A. flavus and for the production of
aflatoxins. In starchy, cereal grain such as wheat, oats, barley,
rice, sorghum, and maize, the lower limit is a moisture content of
18.3%-18.5% on a wet weight basis and in groundnuts, Brazil nuts,
other nuts, copra, and sunflower and safflower seeds, all of which
have a high oil content, it is a moisture content of 9%-10%.

The minimum, optimum, and maximum temperatures for aflatoxin
production are 12° C, 27° C, and 40-42° C, respectively (Davis &
Diener, 1970). Northolt et al. (1976) studied the effect of water
activity and temperature on the growth and aflatoxin production of
A. parasiticus and came to the conclusion that no detectable
quantities of aflatoxin B₁ were formed at an a_w value below 0.83
and at temperatures below 10°C.

3.2.1.2 Invasion of field crops by A. flavus

Groundnut seeds may be invaded by A. flavus before harvest but
are more likely to be invaded very rapidly after the plants have
been pulled and piled for preliminary drying before the nuts are
removed. This postharvest period is the "high hazard" time for

aflatoxin production. On the other hand, studies of aflatoxin
contamination in North Carolina (USA) (Dickens & Satterwhite, 1973)
under conditions of drought, suggest that drought after groundnuts
are formed but before they are dug is conducive to their infection
with A. flavus. Damage caused by the lesser cornstalk borer (LCB)
might also aid in the infection process because insects may carry
fungal spores, although many drought area fields infested with LCB
did not produce groundnuts with a high aflatoxin content. Drought
alone does not result in high levels of aflatoxin contamination. It
must coincide with, or promote infestation by insects which in turn
infect the groundnut. The LCB may act as a vector for A. flavus.

Pettit et al. (1971) reported that groundnuts grown under dry
land conditions (drought stress) accumulated more aflatoxin before
digging than those grown under irrigation. Dry land fresh-dug
kernels contained a maximum aflatoxin level of 35 800 µg/kg while a
maximum of 50 µg/kg was detected in kernels from irrigated plots.
Apparently, the higher kernel moisture content occurring under
irrigated conditions reduced the aflatoxin production potential,
whereas a moisture content of about 31% under drought conditions was
near optimum. Similar observations have been reported from West
Africa.

In some irrigated regions with moist weather at harvest time,
cottonseed may be invaded by A. flavus while still on the plant
and after the bolls open and may contain large amounts of aflatoxins
(Marsh et al., 1973). Stephenson & Russell (1974) related the high
aflatoxin contamination in the field (in USA) to invasion by insects
that provided a site of injury and served as vectors for A. flavus.

Insect injury in ears of maize in the field may also be
accompanied or followed by infection with A. flavus and by
aflatoxin formation before harvest (Lillehoj et al., 1976). To what
extent this constitutes a contamination problem in many regions of
the world, where maize is an important crop, is not known.
Aflatoxins have also been reported in heads of sorghum heavily
infected with mould in India (Tripathi, 1973). Pistachio nuts can
become contaminated with Aflatoxins prior to harvest but the cause
of infection with aflatoxin-producing strains of fungi has not yet
been found. The contamination of almonds and of walnuts has been
traced to specific types of insect damage in the orchard (Stoloff,
1977).

3.2.2 Occurrence in foodstuffs

This subject has been reviewed recently by Stoloff (1976). Of
the four major aflatoxins (B₁, B₂, G₁, G₂) B₁ is usually
found in the greatest concentrations. Measurements of toxin
concentration are based on the wet weight of the commodity in

question. The four toxins may occur together, although they need
not, and their concentrations in relation to each other and their
occurrence may vary depending on the fungal strain and substrate.
For example, Hesseltine et al. (1970) found that most fungal strains
that produced aflatoxin G₁ also produced aflatoxin B₁, but that
not all strains that produced aflatoxin B₁ produced aflatoxin
G₁. One strain of A. flavus from black pepper produced only
aflatoxin B₂ on 2 natural substrates tested (Schroeder & Carlton,
1973).

Although aflatoxins have been found in a variety of foodstuffs,
the most pronounced contamination has been encountered in groundnuts
and other oilseeds including cottonseed and maize. The most
frequently contaminated tree nuts are Brazil nuts and pistachios.

3.2.2.1 Maize

Surveys in the USA of more than 1500 samples of maize collected
in crop years 1964-67, mainly from commercial channels, revealed
that 2%-3% of the samples contained aflatoxins (total aflatoxin B₁
and G₁) in the range of 3-37 µg/kg (Shotwell et al., 1969a, 1970).
In a subsequent survey of 293 samples, 8 samples (2.7%) contained
aflatoxin B₁ levels in the range of 6-25 µg/kg, one of the samples
containing aflatoxin G₁ (25 µg/kg) as well as aflatoxin B₁
(Shotwell et al., 1971). Aflatoxin B₂ was also detected in some of
the samples. In a further study of 60 samples from south-east USA
(Shotwell et al., 1973), aflatoxin B₁ was found in 21 samples
(35%) at levels ranging from 6-308 µg/kg in the 1969-70 period.
Aflatoxin B₂ at levels ranging from a trace to 40 µg/kg was found
in 15 of these samples and aflatoxin G₁ was found in 5 samples at
levels of a trace to 10 µg/kg. In 2 samples, aflatoxin G₂ (1 and
<2 µg/kg) was detected in addition to aflatoxin B₁. In some of
the southeastern states of the USA a high frequency of aflatoxin
contamination was also found in maize in the field (Anderson et al.,
1975; Lillehoj et al., 1976). In some of the field samples, the
levels of aflatoxin B₁ ranged up to several thousand µg/kg or
more. Most of the field contamination was associated with damage
caused by insects such as the European corn borer, corn ear worms,
or weevils, and it seems likely that wherever such damage is
prevalent, field contamination with aflatoxins will occur. In
Thailand, 35% of maize samples contained aflatoxin B₁ (average
level 400 µg/kg) while 40% contamination by aflatoxin B₁ (average
level 133 µg/kg) was found in Uganda (Stoloff, 1976), and 97% in the
Philippine island of Sebu (average 213 µg/kg) (Alpert et al., 1971;
Campbell & Salamat, 1971; Shank et al., 1972a; Campbell & Stoloff,
1974; Stoloff, 1976).

Aflatoxin levels found in household maize samples in connexion
with the outbreak of acute toxic hepatitis in north-west India
(Krishnamachari et al., 1975a,b; Tandon et al., 1977) are discussed
in section 3.5.1.2.

3.2.2.2 Wheat, barley, oats, rye, rice, and sorghum

Shotwell et al. (1968b) reported the presence of aflatoxins at
levels of less than 19 µg/kg in 9/1368 samples of wheat, sorghum,
and oats in the USA. Shotwell et al. (1976b) did not detect any
aflatoxin B₁ (detection limit 1-3 µg/kg) in 848 samples of wheat
from various districts of the USA. The presence of aflatoxins B₁,
B₂, G₁, and G₂ was reported by Tripathi (1973) in heads of
sorghum heavily infected with mould, in field samples in India, but
he did not apply any confirmatory tests. Aflatoxins were also found
in sorghum in Uganda (Alpert et al., 1971) and, in a survey in the
USA, aflatoxins were detected in 2/66 samples of sorghum grain (13
and 50 µg/kg) (Stoloff, 1976). Aflatoxins have been detected in less
than 2% of more than 400 samples of rice from markets in Africa, the
Philippines, and Thailand (Alpert et al., 1971; Campbell & Salamat,
1971; Shank et al., 1972a). However, Lucas et al. (1970-71) reported
that out of 139 samples of rice obtained from the Ho Chi Minh
(Saigon) area in Viet Nam, 31% were found positive for aflatoxins,
no confirmatory tests were included in this study. In surveys of
wheat and other cereals in the USSR, Lvova et al. (1976) found
aflatoxin B₁ at a level of 100 µg/kg in 1/169 samples (0.6%) in
the 1972 crop. Aflatoxins (aflatoxin B₁ at levels ranging from 20
to 444 µg/kg and aflatoxin G₁ at levels of 10-333 µg/kg) were
found in 24/138 samples (17.4%) in the 1973 crop year. In this year,
the samples to be analysed were specially selected from those that
were mouldy or had undergone heating or both. In a survey of wheat
in southern USSR (Kazakstan), Bucarbaeva & Nikov (1977) found
aflatoxin B₁ in 2/50 samples (4%) from one district and 3/50
samples (6%) from another (levels ranging from 5 to 10 µg/kg).

3.2.2.3 Groundnuts (peanuts)

In the 1973 survey in the USA of shelled consumer groundnuts,
15% of 361 samples contained aflatoxins in the range of trace to
50 µg/kg (Stoloff, 1976). Krogh & Hald (1969) found aflatoxins in
86.5% of 52 samples of groundnut products imported into Denmark for
feed, one sample contained 3465 µg/kg. Aflatoxins were found in 41%
of 173 samples of groundnuts in the Sudan, 16% of the samples
containing more than 250 µg/kg, and 9%, more than 1000 µg/kg (Habish
et al., 1971). In the Philippines, all the samples of peanut butter,
tested in 1967-69, contained aflatoxins with a median value of
155 µg/kg and a mean value of 500 µg/kg. The highest level detected
was 8600 µg/kg (Campbell & Salamat, 1971). In Thailand, 49% of
market samples contained an average level of aflatoxins of
1530 µg/kg (Shank et al., 1972a).

3.2.2.4 Soybeans and common beans

No significant degree of aflatoxin contamination has been found
in soybeans or common beans in commerce in the USA (Stoloff, 1976),
although aflatoxin contamination sufficient to be of public health
concern has been found in various types of edible beans in Thailand
(Shank et al., 1972a) and in Africa (Alpert et al., 1971).

3.2.2.5 Tree nuts

Aflatoxin has been found occasionally in Brazil nuts, almonds,
walnuts, pistachio nuts, pecans, and filberts. In some of these,
contamination occurs when the nuts are still on the tree and is
usually associated with damage of one sort or another. However,
apparently sound, undamaged pecans may contain aflatoxins (Stoloff,
1976). Yndestad & Underdal (1975)in Norway found 66% of Brazil nuts
contaminated with aflatoxin B₁ and Nilsson et al. (1974) found
that all of 23 batches of Brazil nuts intended for importation to
Sweden were contaminated. Fourteen percent of 74 samples of
California almonds were contaminated with aflatoxin B₁ with levels
of less than 20 µg/kg in 90% of the contaminated samples (Schade et
al., 1975).

3.2.2.6 Copra

Aflatoxins were found in 88% of 72 samples of copra and copra
meal (Stoloff, 1976) in amounts ranging from a trace to 30 µg/kg,
and similar contamination was found by Krogh et al. (1970) in copra
imported into Finland.

3.2.2.7 Cottonseed

In 3 successive crop years (1964-67), aflatoxin B₁ was
detected in 6.5%-8.8% of more than 3000 cottonseed samples and in
12.8%-21.5% of more than 3000 samples of cottonseed meal (Stoloff,
1976). In contrast, aflatoxin was not detected in cottonseed hulls
(Whitten, 1969). Relatively high levels of aflatoxin contamination
were found in an area in southern California. Aflatoxin levels
increased from 1735 µg/kg in some samples of seed harvested in
November to 2578 µg/kg in some samples going into storage in late
January. The amount present in the stored seeds did not increase
with time, even though fungi, including A. flavus, could be seen
growing on some of the seeds. Marsh et al. (1973) tested cottonseeds
from 13 locations across the USA cotton belt in 1969 and from 11
locations in 1970. Aflatoxins B₁ and B₂ were found in one or
more samples from 3 regions in areas where boll rot caused by
A. flavus had been repeatedly observed in previous years. Seeds
from individual lots contained aflatoxin B₁ levels ranging from
200 000 to 300 000 µg/kg indicating the high potential hazard that
might occur from cottonseed.

3.2.2.8 Spices and condiments

Scott & Kennedy (1973) did not find any aflatoxins in 24 samples
of ground black or white pepper. Low concentrations (up to 8 µg/kg)
were found in 10/33 samples of cayenne pepper and 6/6 samples of
Indian chili powder, mainly as trace amounts.

3.2.2.9 Animal feeds

In studies by Strzelecki & Gasiorowska (1974), aflatoxins
occurred in 12.7% of 306 samples of animal feed and feed components
in Poland, 4.2% of the samples containing more than 100 µg/kg and
2.6% of the samples containing more than 1000 µg/kg. Feed
components, mainly groundnut meals, were contaminated by aflatoxins
more frequently and with higher levels. On the other hand, aflatoxin
was detected in only one sample (2.7%) of cattle and sheep feeds
(300 µg/kg) and in one sample (1.7%)of poultry feeds (30 µg/kg).
Swine feeds contained aflatoxins in 11.4% of samples, with 6 samples
(5.7%) exceeding 250 µg/kg. Two recent surveys of mixed feeds in the
Federal Republic of Germany revealed that 1 in 60 samples contained
aflatoxin B₁ levels exceeding 20 µg/kg (Seibold & Ruch, 1977); 45
out of another 105 samples contained levels of between 7 and
300 µg/kg (Kiermeier et al., 1977). Similar results were obtained in
the United Kingdom (Patterson, personal communication) where 95/172
samples of dairy feed were contaminated with aflatoxin B₁ levels
of 1-350 µg/kg, and 92.4% contained no more than 30 µg/kg.

3.2.2.10 Animal products

Surveys in several countries have shown that aflatoxin M₁ may
be present in liquid or dried milk (Table 3) and in milk products
(Kiermeier, 1977). In addition, highly exceptional aflatoxin levels
in the range of 50-500 µg/litre were reported by Suzanger et al.
(1976) in half of the samples of cow's milk collected in villages
around Isfahan, Iran (15/30 samples collected in 1973 and 21/37
samples in 1974). In contrast, no aflatoxins were detected in 8
samples of milk obtained from large-scale producers in the same area
in 1974 and only 10% of such samples (2/20) contained aflatoxin M₁
(in the range 8-10 µg/litre) in 1973. Aflatoxin M₁ was identified
in all the positive samples. Eight of the 36 village samples
containing aflatoxin M₁ also contained aflatoxin M₂ and 2
samples contained aflatoxin B₁. Considerable differences in the
handling and storage of animal feeds were thought by the authors to
be responsible for the differences in the aflatoxin M₁ contents of
milk samples from villages and large-scale producers in this area.
However, levels of aflatoxins in animal feeds were not reported in
this paper, but they must have been exceptionally high.

Table 3. Selected surveys of aflatoxin M₁ in cow's milk

Milk Country Total no. No. containing Range of concentrations Reference
samples of samples aflatoxin M₁ in the positive samples
analysed (µg/litre or µg/kg)

Liquid Belgium 68 42 0.02-0.2 Van Pee et al. (1977)
German Democratic Republic 36 4^a 1.7-6.5^c Fritz et al. (1977)
Germany, Federal Republic of 61 28 0.01-0.25 Kiermeier (1973)
Germany, Federal Republic of 419 79 trace-0.54 Kiermeier et al. ( 1977)
Germany, Federal Republic of 260 118^a 0.05-0.33 Polzhofer (1977)
India 21 3 up to 13.3 Paul et al. (1976)
Netherlands 95 74 0.09-0.5 Schuller et al. (1977)
United Kingdom 278 85^a 0.03-0.52^a Patterson et al. (in press)

Dried German Democratic Republic 18 0^f --- Fritz et al. (1977)
Germany, Federal Republic of 166 8 0.67-2.0 Neumann-Kleinpaul & Terplan (1972)
Germany, Federal Republic of 52 35 trace-4.0 Hanssen & Jung (1972)
Germany, Federal Republic of 120 7^a 0.05-0.13 Jung & Hanssen (1974)
South Africa 56 0 --- Luck et al. (1977)
USA 320 24 0.1-0.4 FDA (1977) 1973 survey
USA 302 192 trace-3.9^e FDA (1977) 1977 survey

^a Seasonal effect observed, i.e., concentration obviously, dependent upon level of concentrate feeding.
^b Samples collected in retail outlets in 4 southeast States of the USA; it has been estimated that approximately two-thirds
of the crops in these areas contained aflatoxin concentrations exceeding 20 µg/kg because of unusual drought, insect
damage, and high temperature conditions that occurred in 1977.
^c Values for 4 positive samples collected in winter; aflatoxin was not detected in the other 32 winter milk samples as well as
in 12 milk samples collected in summer (detection limit = 0.1 µg/kg).
^d 92.5% samples contained aflatoxin M₁ concentrations of less than 0.1 µg/litre.
^e Levels ranging from 0.1 to 0.4 µg/kg reported in 158 samples; levels exceeding 0.5 µg/kg reported in 19 samples.
^f Aflatoxin B₁ contamination detected in one sample.

Aflatoxin residues have been found in animal tissues, eggs, and
poultry following the experimental ingestion of aflatoxin-
contaminated feed and this subject has been reviewed by Rodricks &
Stoloff (1977). However, the toxins have not yet been found in these
products on the market.

3.2.3 Fate of aflatoxins during the handling and processing
of food

Aflatoxins are affected by some ordinary food processing
procedures. In the roasting of groundnuts, approximately 50% of the
aflatoxins are altered to such an extent that they can no longer be
detected (Lee et al., 1969b; Waltking, 1971). The chemical nature of
the alteration products has not been fully elucidated.

The usual methods of processing groundnuts to make peanut butter
and some nuts for confections may appreciably reduce aflatoxin
contamination. The removal of undersized nuts (shrivels and pegs);
the removal of nuts that resist splitting and blanching; and the
removal of discoloured nuts by hand or electronic sorting are
effective means of reducing contamination (Rodricks et al., 1977).

In the removal of oil from oilseeds, most of the aflatoxins are
found in the oilseed meal. Small amounts remaining in the crude
vegetable oil are mainly taken out in the soap stock, the byproduct
from the alkali refining step. The remaining traces of aflatoxins
are removed in the bleaching refining steps to give aflatoxin-free
refined oil (Parker & Melnick, 1966).

The normal alkali processing of maize to produce tortilla-type
foods, a common practice in some areas of the world including some
Latin American countries, effectively reduces the levels of
aflatoxins in contaminated maize (Ulloa-Sosa & Schroeder, 1969).
Although the mechanism of this reduction has not been clarified,
some of the aflatoxin is most likely washed out by the initial
soaking of the maize in lye water and some is undoubtedly chemically
changed by the alkali; although this process is reported to produce
a substantial reduction in aflatoxin contamination, it is not enough
to give a safe product, when highly contaminated maize is being
processed.

Jemmali & Lafont (1972) reported only partial destruction of
aflatoxins during bread making, indicating the importance of the
contamination of wheat with A. flavus.

The above treatments are normal steps in the processing of
particular foods. In addition to such steps, procedures have been
developed specifically for the destruction or removal of aflatoxins
from grains, nuts, oilseeds, and oilseed meals (cake) (FAO, 1977).

Treatments with ammonia or hydrogen peroxide (H₂O₂) have been
the most effective procedures developed to date for the
detoxification of foodstuffs and animal feeds (Goldblatt & Dollear,
1977). The treatments with ammonia, developed for the industrial
decontamination of aflatoxin-contaminated groundnut and cottonseed
meal (cake) and maize, are limited to the production of animal
feeds. These procedures have recently been discussed in detail at
the joint FAO/WHO/ UNEP conference on mycotoxins in Nairobi (FAO,
1977). The process of treating groundnut protein isolate with
hydrogen peroxide to obtain a product suitable for use as a human
food supplement was also discussed. This process has been developed
in India and is reported to be operating on a small commercial
scale.

The feasibility of methods combining the physical separation of
contaminated portions of produce (for detailed descriptions of
segregation techniques see, for example, Rodricks et al., 1977) with
chemical decontamination, was considered at the same conference
(FAO, 1977).

3.2.4 Pathways and levels of exposure

From the previous discussion, it can be seen that a range of
commodities may become contaminated with trace amounts of
aflatoxins. In vegetable foods, this contamination results directly
from fungal spoilage, maize and nuts being particularly susceptible.
On the other hand, milk, and possibly meat and eggs can become
indirectly contaminated through the absorption by farm animals of
aflatoxins from contaminated feed resulting in residues of the
parent toxin or its metabolites in body fluids or tissues.

Thus, the level of man's exposure to dietary aflatoxins depends
upon the food available and on eating habits and will vary from
country to country according to the local conditions, including the
traditions of different ethnic groups, and amongst individuals.
Where contaminated groundnuts or maize make a significant
contribution to the diet, the level of exposure will be relatively
higher than where less commonly contaminated commodities take their
place as the staple food or when milk is the sole
aflatoxin-containing constituent of the diet.

In this connexion, the Task Group felt it was important to
identify the infant as being potentially at risk because: (a) baby
food products may be made from dried milk or even maize, commodities
known to be prone to contamination by aflatoxins; and (b) in terms
of larger amounts of food consumed per kg body weight, any level of
aflatoxin contamination is more significant for the child than for
the adult.

Attempts to quantify dietary exposure to aflatoxins are
discussed in detail in section 3.6.1.1.

Occupational exposure (section 3.5.2) to aflatoxins with its
attendant high risks, concerns two groups of individuals: those who
handle grain, animal feedstuffs, groundnuts, groundnut meal etc.
(where exposure could occur largely through inhalation of
contaminated dust), and those who work with toxins in experiments or
with pure toxins as analytical standards.

In one paper (van Nieuwenhuize et al., 1973) available to the
Task Group, an attempt was made to quantify occupational exposure to
airborne aflatoxins in an oil-mill crushing groundnuts and other
oil-seeds. Based on airborne dust determinations (mean aflatoxin
concentrations of 250 and 410 µg/kg airborne dust), the estimated
airborne aflatoxin levels ranged from 0.87 to 72 ng/m³ of air.

3.3 Metabolism

3.3.1 Absorption

Although quantitative data on absorption are not available at
present, there is no doubt that most of the field cases of
aflatoxin-induced diseases in animals and man have been associated
with ingestion of aflatoxin-contaminated foodstuffs and thus with
the absorption of aflatoxins in the alimentary tract. In spite of
reports of respiratory exposure (sections 3.2.4 and 3.5.2), there is
no quantitative information available on aflatoxin resorption from
the respiratory tract or on percutaneous absorption.

3.3.2 Tissue distribution

3.3.2.1 Animal studies

Experiments with ¹⁴C-ring-labelled aflatoxin B₁ have shown
that rats retain about 20% of the ¹⁴C activity 24 h after a single
intraperitoneal dose of 0.07 mg/kg body weight (Wogan et al., 1967).
The highest concentration was found in the liver, which contained
amounts of radioactivity equivalent to the entire remainder of the
carcass (about 5%-8% of the total ¹⁴C recovered).

When poultry were fed rations containing aflatoxins at
concentrations ranging from 25 to 15 000 µg/kg for 8 weeks, residues
of aflatoxin B₁ were found in the liver and in muscle tissue
(Mintzlaff et al., 1974). The liver contained the highest
concentration, with a mean value of 15 µg/kg at the highest exposure
level. Similarly the highest concentration of aflatoxin B₁ was
found in the livers of pigs (range: trace-137 µg/kg) fed rations
containing aflatoxins (both aflatoxin B₁ and B₂) at levels of
300 and 500 µg/kg for 4 months (Krogh et al., 1973a). Aflatoxin
residues were also detected in kidneys, and muscle and adipose
tissue.

3.3.2.2 Studies in man

Levels of aflatoxins in the tissues of children with Reye's
syndrome are discussed in section 3.6.2.2. In a liver biopsy from a
subject with carcinoma of the rectum and liver in the USA, Phillips
et al. (1976) found 520 µg/kg of aflatoxin B₁. In France, Richir
et al. (1976) found aflatoxin B₁ in liver biopsies in 6 out of 100
subjects suffering from various diseases. Concentrations observed
ranged from 1.6-8 µg/kg.

3.3.3 Metabolic transformation and activation

With one exception, all primary biotransformations of aflatoxin
B₁ involve its conversion to hydroxylated metabolites but only one
such derivative, aflatoxin M₁ has appreciable oral toxicity
(Holzapfel et al., 1966). Even so, this metabolite may be detoxified
by conjugation with taurocholic and glucuronic acids prior to
excretion in the bile or urine (Bassir & Osiyemi, 1967). In this
respect, two recently discovered metabolites, P₁ (Dalezios et al.,
1971; Buchi et al., 1973) and Q₁ (Masri et al., 1974a,b) are
similar in that they also undergo this type of detoxification
(Dalezios et al., 1971; Dalezios & Wogan, 1972).

The conversion in the liver (Fig. 2) of aflatoxin B₁ to
aflatoxicol (Patterson & Roberts, 1971) and to aflatoxicol H₁ via
aflatoxin Q₁ (Salhab & Hsieh, 1975) is unusual in that, unlike
other biotransformations that are catalysed by liver microsomal
enzymes, a cytoplasmic NADH-dependent dehydrogenase is involved.
Furthermore, the formation of aflatoxicol can be inhibited by
17-ketosteroid sex hormones (Patterson & Roberts, 1972a) and this is
the only metabolic transformation of aflatoxin in vitro known to
be sensitive to hormones.

Liver homogenates of certain avian and rodent species are
particularly active in converting aflatoxins B₁ and G₁ to their
2-hydroxy, 2,3-dihydro derivatives or hemiacetals called also
aflatoxins B_2a and G_2a (Patterson & Roberts, 1970). These
metabolites bind strongly to protein and are probably sufficiently
reactive, when formed in vivo, to cause many of the acute effects
of aflatoxin poisoning (Patterson & Roberts, 1972b; Patterson, 1973,
1977).

At present, there is only indirect evidence for the formation of
the epoxides of aflatoxins B₁ and G₁ but this is probably the
more important form of metabolic activation. When either of the
parent toxins is incubated with microsomes prepared from the livers
of many animal species including man, a metabolite is formed which
appears to have only a transient existence, is highly reactive,

binds covalently to DNA, and induces mutation in a bacterial
in vitro test system (Garner et al., 1971, 1972; Ames et al.,
1973). The metabolite of B₁ has not been isolated but the
2,3-dihydrodiol has been recovered following mild acid hydrolysis of
an adduct formed when the microsomal metabolite was generated in the
presence of added DNA or RNA (Swenson et al., 1,974) and, more
recently, after in vivo intraperitoneal injection of aflatoxin
B₁ (Swenson et al., 1977). This has been assumed to be indirect
evidence of the formation of the 2,3-epoxide and, in view of the
interaction with DNA, it is now generally accepted that the epoxide
of aflatoxin B₁ is the bacterial mutagen and the proximal
carcinogen.

Certain of these biotransformations are better developed in some
animal species than others (Patterson, 1977) and attempts have been
made to correlate liver metabolism of aflatoxins with toxicity. In
the first such attempt (Patterson, 1973), it was proposed that rapid
in vitro formation of aflatoxin hemiacetal was correlated with
susceptibility to acute aflatoxin poisoning. More recently (Hsieh et
al., 1977), it has been suggested that the reversible formation of
aflatoxicol, which is thought to provide a "metabolic reservoir" of
aflatoxin (Patterson & Roberts, 1972b), is correlated with
susceptibility to liver tumour induction. On the basis of this, it
has been tentatively suggested (Hsieh, 1977; Salhab & Edwards, 1977)
that the human liver might be relatively more resistant to aflatoxin
carcinogenesis than that of some other species, particularly the
rat.

3.3.4 Excretion

3.3.4.1 Animal studies

Excretory pathways. Using aflatoxin B₁, ring-labelled or
methoxy-labelled with ¹⁴C, Wogan et al. (1967) have shown that
rats excrete 7096-80% of a single intraperitoneal dose within 24 h.
A major excretory route of the ring-labelled toxin was through
biliary excretion into the faeces, accounting for about 60% of the
administered dose; approximately 20% of administered radioactivity
was excreted in the urine, and only negligible amounts in expired
air in the form of ¹⁴CO₂. In contrast, approximately 25% of
radioactivity from methoxy-labelled material appeared in expired air
as ¹⁴CO₂ with a concomitant decrease in the faeces, indicating
that O-demethylation is a significant metabolic pathway for
aflatoxin B₁ in the rat.

Excretion in the milk of farm animals. Several reviews deal
with the excretion of aflatoxins in the milk of farm animals
(Allcroft, 1969; Kiermeier, 1973, 1977; Patterson, 1977; Rodricks &
Stoloff, 1977). When cattle (Allcroft et al., 1968), sheep (Nabney
et al., 1967) or goats (Vesely, et al., 1978) are given feed
contaminated with aflatoxin B₁ their milk contains aflatoxin M₁.
In the cow, there is a linear relationship between the amount of
aflatoxin B₁ ingested daily and the level of aflatoxin M₁ in the
milk (Allcroft & Roberts, 1968; Purchase, 1972; Patterson, 1977; see
Fig. 3), indicating that about 1.5% of aflatoxin B₁ is excreted as
the metabolite M₁ (Kiermeier, 1973), and that the concentration of
aflatoxin B₁ in milk is approximately 1/300 of the concentration
of aflatoxin B₁ in the dairy ration (Rodricks & Stoloff, 1977).
Smaller quantities of unmetabolized aflatoxin B₁ have been found
in cow's and sheep's milk (Nabney et al., 1967; Allcroft et al.,
1968; Wogan, 1969).

3.3.4.2 Studies in man

In the Philippines, aflatoxin M₁ has been found (not
measured)in the urine of human subjects known to have ingested
aflatoxin-contaminated peanut butter (Campbell et al., 1970).

Claims concerning an aflatoxin involvement in the etiology of
juvenile cirrhosis in India (section 3.6.3.2) based on a
blue-fluorescent B₁ spot in the breast milk of mothers and the
urine of children with the disease (Robinson, 1967) are largely
discounted by the later studies of Yadgiri et al. (1970) who
produced spectrophotometric evidence that, although such a spot
could be identified in the urine of children with the overt disease,
this was not aflatoxin B₁ For other reports see section 3.5.1.

3.4 Effects in Animals

The effects of aflatoxins in animals have., been reviewed by
Allcroft (1969), Newberne & Butler (1969) and Butler (1974).

3.4.1 Field observations

When foodstuffs are affected by microbial deterioration, man
normally eats the less affected parts, whereas domestic animals may
be exposed to more contaminated rations. This explains why the
discovery of several mycotoxins has been based on field observations
in domestic animals.

The first observation of a disease in animals subsequently
associated with aflatoxins was an acute outbreak of a lethal disease
in turkey poults in England in 1960 causing an estimated loss of at
least 100 000 birds. Extensive research eventually revealed that the
disease was caused by aflatoxins contained in a batch of Brazilian
groundnut meal. The concentration of aflatoxin B₁ in the original
groundnut meal was later estimated to be about 10 mg/kg. The disease
was characterized by rapid deterioration in the condition of the
birds, subcutaneous haemorrhages, and death. At postmortem, the
livers of the birds were pale, fatty, and showed extensive necrosis
and biliary proliferation (Butler, 1974). A similar case of acute
disease was observed in day-old ducklings fed "toxic" groundnut meal
(Asplin & Carnaghan, 1961), where the liver changes described were
followed by cirrhosis. Outbreaks of liver disease in chickens have
also been associated with aflatoxin-contaminated feed (Asplin &
Carnaghan, 1961).

Loosmore & Harding (1961) noted outbreaks in pigs fed groundnut
meal in which the toxic factor was later identified as aflatoxins.
The lesions in the pigs included haemorrhages, and liver damage
characterized by dissecting fibrosis and biliary proliferation.
Calves fed rations containing 15% toxic groundnut meal also
developed liver lesions characterized by fibrosis and biliary
proliferation (Loosmore & Markson, 1961). Outbreaks associated with
"toxic" groundnut meal and characterized by similar liver lesions
have been reported in older cattle even though they are more
resistant (Clegg & Bryston, 1962); there was also a drop in milk
production.

A liver disease "hepatitis X" has been reported in dogs in
southeastern USA (Seibold & Bailey, 1952; Newberne et al., 1955).
Icterus and in some cases ascites were observed and the liver
lesions included fatty changes with centrilobular parenchymal
necrosis and biliary proliferation. Commercial dog food thought to
be the cause of toxicity was later found to contain aflatoxin B₁
(up to 1.75 mg/kg) (Newberne et al., 1966a). Similar lesions have
been reproduced in dogs by peroral administration of aflatoxins, and
it has been suggested that the "hepatitis X" in dogs could be
causally associated with aflatoxins in the diet (Newberne et al.,
1966a). During an outbreak of toxic hepatitis affecting several

hundred people in north-west India and considered to be possibly
associated with the consumption of maize heavily contaminated by
aflatoxins (section 3.5.2 and 3.6.2.1), dogs fed food remnants from
households in affected villages manifested a disease characterized
by jaundice, ascites, and frequently death (Krisnamachari et al.,
1975a, b; Tandon et al., 1977). A nonportal type of micronodular
cirrhosis, with less conspicuous parenchymal and cholangiolar
changes was found on histological examination of the livers of two
dogs (Tandon et al., 1977).

3.4.2 Experimental studies

3.4.2.1 Acute and chronic effects: hepatotoxicity

Different species vary in their susceptibility to acute
poisoning by aflatoxins, with LD₅₀ values ranging from 0.3 to
17.9 mg/kg body weight (Table 4). In all the animals studied, the
liver was the principal target organ (see for example Butler, 1974).

Table 4. Acute toxicity of aflatoxin B₁^a

Species LD₅₀ Zone of
(mg/kg body weight) liver lesion

chick embryo 0.025^d
rabbit 0.3 midzonal
duckling 0.335 periportal
cat 0.55 periportal
pig 0.62 centrilobular
dog 0.5-1.0 centrilobular
sheep 1.0 centrilobular
guineapig 1.4 centrilobular
baboonb 2.0 centrilobular
rat (male) 7.2 periportal
macaque female^c 7.8 centrilobular
mouse 9.0
hamster 10.2
rat (female) 17.9 periportal

^a Adapted from: Newberne & Butler (1969) and Butler (1974).
^b From: Peers & Linsell (1976).
^c From: Shank et al. (1971 b).
^d µg/embryo.

The lesions observed in field cases (section 3.4.1) in poultry,
pigs, cattle, and dogs have all been reproduced in the same animal
species by feeding experiments during periods of time ranging from a
few weeks to a few months, using diets containing aflatoxins, or
pure aflatoxins ranging from 0.3 to several mg/kg (Newberne &
Butler, 1969).

In the study by Carnaghan et al. (1966), chickens were fed a
diet containing aflatoxin B₁ at a level of 1.5 mg/kg. Groups of 3
control and 3 test chicks were killed after 3´ days, 7 days, and
then at weekly intervals for 8 weeks. After 4 weeks, the liver
lesions included fatty change, biliary proliferation, and fibrosis.

In 20 pigs, aflatoxins (aflatoxins B₁ and B₂) at a feed
level as low as 300 µg/kg resulted in the development of
centrilobular necrosis and fibrosis of the liver as well as growth
depression, during a normal feeding period of 3-4 months (Krogh et
al., 1973a). In cattle, the liver lesions (centrilobular
degeneration, fibrosis, biliary proliferation) occurred in all 4
animals after 4 months on a feed containing an aflatoxin level of
2 mg/kg (Allcroft & Lewis, 1963). The hepatic lesions induced in the
duckling by the aflatoxins formed the basis of a bioassay originally
described by Sargeant et al. (1961). At sublethal doses of
aflatoxins, the bioassay depends upon an assessment of the degree of
biliary proliferation. Liver lesions similar to those observed in
farm animals have been experimentally induced by the administration
of aflatoxins in a number of laboratory animals, including the rat,
cat, guinea-pig, and rabbit (Newberne & Butler, 1969).

In a study of Madhavan et al. (1965b), 2 rhesus monkeys
(Macaca mulatta) were given daily oral doses of aflatoxins at
500 µg/animal for 18 days (corresponding approximately to 250 µg/kg
body weight per day) and then 1 mg/animal per day (corresponding
approximately to 500 µg/kg body weight per day) until death occurred
after 32 and 34 days. Three rhesus monkeys were given 1 mg each,
daily, until death occurred after 19, 20, and 27 days respectively.
The liver lesions included fatty infiltration, biliary
proliferation, and portal fibrosis. Death or similar lesions were
not observed in the 2 control monkeys.

Deo et al. (1970) studied the effect on male rhesus monkeys of
repeated administration by gastric tube of 3 different levels of
aflatoxins (B₁ + G₁). At the highest dose level (1 mg/kg body
weight daily for 3 weeks), 35/35 animals died within 22 days with
extensive haemorrhagic necrosis of the liver. A dose level of
0.25 mg/kg body weight, twice a week, for 5 months induced various
degrees of liver changes in 24/24 animals characterized by biliary

proliferation and focal appearance of the liver cells with multiple
nuclei and giant-sized liver cells with enlarged hyperchromatic
nuclei. At the lowest dose, 5 animals were given 62 µg/kg body
weight once a week for periods ranging from a few days to 2 years.
Liver changes were similar to the changes seen in the second group
but in a milder form.

Cynomolgus monkeys (Macaca fascicularis = M. irus) fed a
dietary level of aflatoxin B₁ of 5 mg/kg rapidly developed liver
damage with biliary proliferation and all 6 animals died within 2
months. When fed aflatoxin B₁ at a dietary level of 1.8 mg/kg, 5
animals died within 3 months showing liver damage characterized by
centrilobular necrosis, biliary proliferation, and fibrosis. Two
animals survived and were killed after 3 years; the liver of one
animal had the appearance of nodular cirrhosis. Two groups of 4
animals each were fed lower levels of aflatoxin B₁ (0.07 and
0.36 mg/kg, respectively) for 3 years without showing any signs of
liver lesions (Cuthbertson et al., 1967).

The relationship between the chemical structure of different
aflatoxins and their biological activity (discussed also in section
3.4.2.3) was investigated in a small number of experiments; more
extensive studies were not possible because of the limited
quantifies available of some of the pure aflatoxins. Carnaghan et
al. (1963) compared 6-day mortality following single doses of
different aflatoxins, administered by intubation to one-day-old
Khaki Cambell ducklings, and concluded that both aflatoxins B₂ and
G₂ were less toxic than aflatoxins B₁ and G₁, the ratio of
LD₅₀ values being 1:4.7 for B₁:B₂ and 1:4.4 for G₁:G₂.
Aflatoxins G₁ and G₂ were less toxic than the corresponding
aflatoxins B₁ and B₂, the ratio of the LD₅₀ values being
1:2.15 for B₁:G₁ and 1:2.03 for B₂:G₂. The corresponding
LD₅₀ values for aflatoxins B₁, B₂, G₁, and G₂ were 0.36,
1.70, 0.78, and 3.45 mg/kg, respectively. Comparable results were
obtained by Wogan et al. (1971) who recorded 14 day mortality after
intubation of male Pekin ducklings and reported LD₅₀ values of
0.73, 1.76, 1.18 and 2.83 mg/kg for aflatoxins B₁. B₂, G₁, and
G₂, respectively. In the same paper, a study was reported on the
14-day mortality of male Fischer rats after a single
(intra-peritoneal) dose of aflatoxin. The LD₅₀ value for aflatoxin
B₁ was 1.16 mg/kg body weight (95% confidence interval 0.91 to
1.48 mg/kg) whereas the LD₅₀ for aflatoxin G₁ was between 1.5
and 2.0 mg/kg body weight. On the other hand, no deaths occurred in
20 rats given 12-200 mg of aflatoxin B₂ per kg body weight, and
all 4 rats given 170-200 mg of aflatoxin G₂ per kg body weight
survived. A similar difference in the toxicity of aflatoxins was
observed when male Fischer rats were given repeated doses of

aflatoxins by stomach tube over a 4-week period. The 4-week
mortality in rats given a total dose of 1 mg of aflatoxin B₁ per
rat was 8/10 whereas all 10 rats given the same dose of aflatoxin
G₁ survived and only 4/10 animals given double this dose of
aflatoxin G₁ died. In another trial, all 11 rats survived
intragastric administration of 3.75 mg of aflatoxin B₂ per rat
repeated every second day for 4 weeks to give a total dose of
52.5 mg per rat.

Holzapfel et al. (1966) and Purchase (1967) reported 7-day
mortality alter oral dosing of one-day-old Pekin ducklings with
aflatoxins B₁, M₁, and M₂. Five groups of 2-3 ducklings (body
weight 40-50 g) were used for each of the aflatoxins tested, and the
following LD₅₀s were calculated (with 95% confidence limits given
in brackets): aflatoxin B₁, 12 (3.9-37.2) µg per duckling,
aflatoxin M₁, 16 (5.4-51.5) µg per duckling, and aflatoxin M₂,
61.4 (37-100) µg per duckling. Ducklings receiving aflatoxin M₂
showed characteristic liver lesions indistinguishable from those
observed after a similar dose of aflatoxin B₁. Higher doses of
aflatoxin M₂ produced similar effects (Purchase, 1967). A study
comparing the acute toxicity of synthetic (racetalc) aflatoxins B₁
and M₁ and the natural optical isomer of aflatoxin B₁ was
reported by Pong & Wogan (1971), suggesting that only one isomer of
each synthetized racemic mixture was biologically active.
Fourteen-day mortality rates observed after a single intraperitoneal
dose of 1.5 mg/kg body weight of synthetic aflatoxin B₁ and
synthetic aflatoxin M₁ were 1/1 and 1/2, respectively. However, no
deaths occurred in groups of rats (each consisting of 4 animals)
given these synthetic aflatoxins at doses of 1, 0.8, 0.6, or
0.4 mg/kg body weight. With the natural aflatoxin B₁, the observed
mortalities at these dose levels were 4/4, 2/4, 2/4, and 0/4
respectively.

For information on the toxicity of certain other aflatoxin
metabolites or derivatives, see Wogan et al. (1971) and Patterson
(1976).

3.4.2.2 Hepatotoxicity connected with extrahepatic effects

Many other organs besides the liver are more or less severely
affected in acute experiments with high doses of aflatoxins (Butler,
1964): in male and female rats, a single dose of aflatoxin B₁
proved lethal in half of the animals (7.2 mg/kg body weight in the
male and 17.9 mg/kg body weight in the female, by garage). Frequent
bilateral adrenal haemorrhages, petechial haemorrhages in many
organs, particularly in the congested lungs, and occasionally patchy
necroses in the myocardium and in other organs (kidney, spleen) were
observed during the first few days following administration. These
changes were not detected in male or female rats given aflatoxin B₁

at 3.5 mg/kg body weight. With higher doses, the haemorrhages seen
in the lungs, kidneys, and adrenals were more extensive. Animals
dying within the first few days often had altered blood in the whole
of the small intestine and in the colon. Ascites and oedema of the
omentum were observed in some of the animals a week or more (but not
one month) after aflatoxin administration. After a month, with the
exception of the liver damage, all the other organs appeared normal
in surviving animals. Histologically, certain renal changes were
detected in the loops of Henle at this stage, consisting of a few
cells with large irregular hyperchromatic nuclei, very similar to
those seen in the liver (Butler, 1964).

Congested lungs with small petechial haemorrhages, haemorrhagic
necroses in the adrenals (localized in the inner zone of the
reticularis) and patchy necroses in the kidneys, pancreas, and
spleen were observed in guineapigs 2-3 days after a single
intraperitoneal injection of aflatoxin B₁ at 1.4 mg/kg body weight
(lethal in half of the males and females). Even at this dose, the
small intestine was frequently filled with altered blood. At higher
doses, the haemorrhagic disease was more marked, with pleural,
pericardial, and peritoneal haemorrhages. The only change seen in
the heart of the guineapig, 2-3 days after aflatoxin administration,
was an occasional small area of fatty degeneration of the
myocardium. Many animals showed marked ascites and oedema of the
omentum and subcutaneous tissue during the first week after
injection (Butler, 1966).

Bourgeois et al. (1971) reported a special syndrome induced by
oral administration of aflatoxin B₁ in the macaque (Macaca
fascicularis). In 2 groups of 4 young females, each receiving a
single oral dose of aflatoxin B₁ at 13.5 or 40.5 mg/kg body
weight, all animals died within 149 h. Death occurred in 1 out of 4
other animals receiving a dose of 4.5 mg/kg body weight. Doses of
toxin of 1.5 mg/kg or 0.5 mg/kg (4 animals in each group) did not
result in death or unusual clinical signs. Cough, vomiting,
diarrhoea, and coma were characteristic clinical findings in animals
exposed to toxic doses. Analysis of blood serum revealed a
dose-dependent decrease in serum levels of phospholipids within 24 h
of administration of the aflatoxin. A dose-dependent decrease in
serum levels of glucose and an increase in nonesterified fatty acids
occurred within 72 h of aflatoxin administration. The liver lesions
included centrilobular necrosis, some biliary proliferation, and
massive fatty degeneration which was also observed in the heart and
kidneys. Cerebral oedema with neuronal degeneration was seen. Some
of these findings resemble those associated with Reye's syndrome in
children (see section 3.5.1.2).

3.4.2.3 Carcinogenesis

The carcinogenesis of aflatoxins has been reviewed by Wogan
(1973, 1977) and re-evaluated by IARC (1976).

Hepatic and renal tumours. Orally administered aflatoxins,
mainly B₁, have been hepatocarcinogenic in all species of test
animals studied so far (including nonhuman primates), with the
exception of the mouse, in which carcinogenic effects have been
demonstrated only following intraperitoneal administration of
aflatoxin B₁ to neonates (Tables 5 and 6). These studies were
concerned with repeated or long-term exposure to aflatoxins. In a
study by Carnaghan (1967), 2 groups consisting of 16 and 18 weanling
female Wistar rats, respectively, were given single oral doses of
crystalline aflatoxin B₁ or a mixture of aflatoxins containing
about 40% aflatoxin B₁ and 60% aflatoxin G₁, at the rate of
0.5 mg/rat in 0.1 ml dimethylformamide. These doses corresponded to
averages of 7.65 mg aflatoxin B₁/kg body weight and 2.7 mg
aflatoxin B₁ plus 4.0 mg aflatoxin G₁/kg body weight,
respectively. Within 21-32 months, 7 rats out of each group
developed hepatic tumours with metastases in half the cases. Hepatic
rumours were not observed in 19 control rats given the solvent only.
No hepatocellular carcinomas were found in 22 male Fischer rats
killed successively 16 weeks (3 rats), 25 weeks (5 rats), 38 weeks
(5 rats), 55 weeks (4 rats), and 69 weeks (5 rats) after a single
dose of aflatoxin B₁ at 5.0 mg/kg body weight, administered by
garage (Wogan & Newberne, 1967).

A linear dose-response relationship was observed by Wogan et al.
(1974) for the development of liver-cell carcinomas in male Fischer
rats fed dietary concentrations of aflatoxin B₁ ranging from
1-100 µg/kg (Table 7). At 1 µg/kg, a 10% tumour incidence was found,
compared with no rumours in the control group and at 100 µg/kg the
tumour incidence was 100%. A linear log (dose)-response relationship
has been demonstrated in trout fed dietary levels of aflatoxin B₁
ranging from 0.5 to 20.0 µg/kg, for 20 months. Extrapolating this
relationship to lower exposure levels, an incidence of approximately
10% would be expected with a dietary concentration of 0.1 µg/kg.

Table 5. Hepatocarcinogenicity of aflatoxin B₁ in rodents^a

Species Dosing regimen Duration of Period of Liver Reference
treatment observation tumour
incidence

rat, Fischer 1.0 mg/kg diet 33 weeks 52 weeks 3/6 Svoboda et al. (1966)
rat, Fischer 1.0 mg/kg diet 41-64 weeks 41-64 weeks 18/21 Wogan & Newberne (1967)
rat, Porton 1.0 mg/kg diet 20 weeks 90 weeks 19/30 Butler (1969)
rat, Wistar 1.0 mg/kg diet 21 weeks 87 weeks 12/14 Epstein et al. (1969)
mouse, Swiss 150 mg/kg dietb 80 weeks 80 weeks 0/60 Wogan (1973)
mouse, C57Bl/6NB 1.0 mg/kg diet 80 weeks 80 weeks 0/30 Wogan (1973)
mouse, C3HfB/HEN 1.0 mg/kg diet 80 weeks 80 weeks 0/30 Wogan (1973)
mouse, hybrid F₁, 4 days old 6.0 µg/g body weight 3 doses (i.p.) 80 weeks 16/16 Vesselinovitch et al. (1972)

^a From: Wogan (1977).
^b A mixture of aflatoxins B₁ and G₁ was used in this experiment.

Table 6. Hepatocarcinogenicity of aflatoxin B₁ in nonrodent species^a

Species Dosing regimen Duration of Period of Liver Reference
treatment observation tumour
incidence

monkey, rhesus (M) 1.655 g total^b 5.5 years 8.0 years 1/1 Gopalan et al. (1972)
monkey, rhesus (F) 1.855 g total 5.5 years 10.75 years 1/1 Tilak (1975)
monkey, rhesus (F) 0.504 g total 6.0 years 8.0 years 1/1 Adamson et al. (1973)
marmoset 3.0 mg total 50-55 weeks 50-55 weeks 1/3 Lin et al. (1974)
5.04-5.84 mg total^c 87-94 weeks 87-94 weeks 2/3 Lin et al, (1974)
tree shrew (M & F) 24-66 mg total 74-172 weeks 74-172 weeks 9/12 Reddy et al, (1976)
ferret 0.3-2.0 µg/kg 28-37 months 28-37 months 7/9 Butler (1969)
duck 30 µg/kg 14 months 14 months 8/11 Carnaghan (1965)
rainbow trout 4 µg/kg in diet 12 months 12 months 15% Sinnhuber et al, (1968b)
8 µg/kg in diet 12 months 12 months 40% Sinnhuber et al. (1968b)
rainbow trout embryos 0.5 mg/kg in water 1 h 296-321 days 38% Sinnhuber & Wales (1974)
salmon 12µg/kg in diet^d 20 months 20 months 50% Wales & Sinnhuber (1972)
guppy 6 mg/kg in diet 11 months 11 months 7/11 Sato et al. (1973)

^a Modified from: Wogan (1977).
^b A mixture of aflatoxins B₁ and G₁ was used in this experiment.
^c These animals were infected simultaneously with hepatitis virus.
^d This diet also contained 50 mg/kg cyclopropenoid fatty acids.

Table 7. Dose-response characteristics of aflatoxin B₁ carcinogenesis
in male Fischer strain rats^a

Dietary Duration Liver Time of appearance
aflatoxin of feeding carcinoma of earliest tumour
level (week) incidence^b (week)
(µg/kg)

0 74--109 0/18^c --
1 78--105 2/22 104
5 65--93 1/22 93
15 69--96 4/21 96
50 71--97 20/25^d 82
100 54--88 28/28^e 54

^a From: Wogan et a