439
DO TERATOGENICITY TESTS SERVE THEIR
OBJECTIVES?
S!R,—In a 1981 editorial on thalidomide and children with severe
I The main thrust is still on screening for structural malformations
congenital malformations The Lancet stated that "The
repercussions continue to this day, most conspicuously in the
proliferation of drug regulatory bodies and their regulations, of
birth-defects monitoring or surveillance programmes, and of drug
teratology law-suits. Less conspicuous has been a burgeoning of
animal teratology research".
Animal tests are designed to predict if a chemical is likely to
produce birth defects or interfere with reproduction in man. Some
of the very first recommendations for the extension of animal testing
for the teratogenicity of drugs appeared in the British Medical
Journal in 1962.2,3
The design of the two-litter test for drugs, in use when the
thalidomide disaster occurred, encompassed too many indices of the
reproduction process. These studies were subdivided into three
segments, each pertaining to a specific phase, by the U.S. Food and
Drug Administration guidelines in 1966. Those segments are the
study of fertility and general reproductive performance,
teratological study, and prenatal and postnatal study. Detailed
committees and presented as a W.H.O. report (1967) and as
guidelines issued in the U.K., Canada, Japan, and Sweden in
1973-74.
Peck wrote in 1968 that "The question of the predictability of
extrapolation of results of teratogenic and reproduction studies in
animals needs little comment. The reliability of such tests remains
to be determined".4 This is still so, and it is interesting to note that,
as recently as May, 1981, the Organisation for Economic
Cooperation and Development adopted guidelines very similar to
the ones issued by F.D.A. in 1966.
Since 1962 few chemical teratogens for man have been identified,
though a lot have been tested in animals. In 1978, the number of
chemicals tested, with published data on embryotoxicity and
teratogenicity, had reached 2300. Besides these, an unknown5
number of unpublished studies have been done by industry.
Schardein found that of 1930 chemicals studied, 30% showed
teratogenic and 62% non-teratogenic properties; 8% could not be
definitely classified.5 Staples6 gave the figure of 800 chemicals
identified as teratogens in laboratory animals, but less than 25
known in man. The 1977 edition of the Registry of Toxic Effeèts of
Chemical Substances (published by the U.S. National Institute of
Occupational Safety and Health) contains 2474 citations on 405
compounds related to teratogenesis. What is the value of routine
tests in animals for prediction of chemical teratogens? The
correlation between known effects in laboratory animals and clinical
adverse effects is very low.
The predictability of animal tests could be increased if
information on pharmacokinetics, metabolism, and enzymatic drug
induction was also used.7,8 No attention is paid to the dose to the
embryo of the chemical tested during its period of exposure. Recent
data, for example, have shown that placental transfer of diazepam,
is faster in late pregnancy while the volume of distribution of
1. Editorial. Thalidomide: 20 years on. Lancet 1981; ii: 510-11.
2. Dent CE, Laurence DR, Nixon WCW.
Drugs and foetal abnormalities. Br Med J 1962,
ii: 254.
3. Woollam DHM. Thalidomide disaster considered as an experiment in mammalian
teratology. Br Med J 1962; ii: 236-37.
4. Peck HM. An in animals and the extrapolation of
appraisal of drug safety evaluation
results to man. In: Tadeschi Tadeschi RE, ed Importance of fundamental
DH,
principles in drug evaluation. New York: Raven Press, 1968: 449-71
5. Leuschner F. Zur Relevanz tierexperimenteller Embryotoxizitätsstudien fur den
Menschen. In- Schnieders B, Stille G, Grosdanoff P, ed. Embryotoxikologische
Probleme in der Arzneimittleforschung. AMI Berichte Berlin Dietrich Reimer
Verlag, 1978; 170-75.
6. Staples RE Predictiveness and limitations of test methods in teratology. Overview
Environ Health Perspect 1976; 18: 95-96.
7. Neubert D. Bedeutung pharmakokinetischer Parameter für dir Teratologie. In:
Schnieders B, Stille G, Grosdanoff P, ed AMI Berichte Berlin: Dietrich Reimer
Verlag, 1978; 83-88.
8 Young JF, Holson JF. Utility of pharmacokinetics in designing toxicological protocols
and improving interspecies extrapolation. J Environ Pathol Toxicol 1978; 2:
169-86.
phenytoin is increased. The value of the results from teratogenicity
testsin the risk evaluation of chemical exposure of, for example,
women in certain workplaces is severely limited.
caused during the organogenic period (first trimester). However,
other types of chemical adverse effects on reproduction are just as
important-namely, pre-implantation loss, physiological
impairment of the fetus and newborn, and pharmacological
interference with transmitter
mechanisms and behavioural
development. Prenatally induced cancer and fertility problems in
both man and woman also deserve more attention.
A vast amount of information on physicochemical properties and
on biological effects is, however, available, but scattered in different
sources. Compilation and use of such data on chemicals concerning,
for instance, exposure conditions, rate of transfer to and elimination
from the embryo/fetus, tissue binding, and biotransformation and
reactivity, is in our view an important complement to studies on
chemically impaired reproduction.
We would urge a critical revaluation of teratogenicity tests and
revision of their design.
Department of Toxicology,
University of Uppsala,
S-751 23 Uppsala, Sweden K. SUNE LARSSON
Department of Clinical Pharmacology,
Hospitals of Huddinge and Danderyd Carl-ERIC ELWIN
Department of Biopharmaceutics
and Pharmacokinetics, JOHAN GABRIELSSON
University of Uppsala LENNART PAALZOW
Department of Organic Chemistry,
Wallenberg Laboratory,
University of Stockholm CARL-AXEL WACHTMEISTER
CLINICAL TRIALS IN CANCER
SIR,-All serious participants in controlled clinical trials in
oncology are concerned about the subject of informed consent. The
ethical consideration, however, is surely not whether the patient
gives consent to a controlled trial per se, but whether the control
arm offers the patient the best standard treatment and the
"experimental" arm offers that treatment together with a departure
from the standard regimen which has theoretical laboratory, or
clinically demonstrated, additional benefit.
Zelen produced a new design for randomised clinical trials in
which randomisation was performed before consent was sought. If
the instructions were for control treatment the patient was asked,
after adequate explanation, to give consent for that treatment, the
benefits and dangers of which were fully explained. If the
instructions were for the experimental arm the deviation from
standard treatment was fully explained, as were the potential
benefits and dangers. Only the patients in this group were to be
informed that they were being asked to participate in a trial.
This has now become standard practice in my unit and so far has
not resulted in refusal to accept the experimental treatment. I do not
explain to all patients all possible alternatives, but this surely
applies not only to doctors taking part in controlled trials but to any
. clinician explaining to a patient what treatment he or she
recommends. It is a very rare patient with, for instance, carcinoma
of the breast who wants to know that there are at least a dozen
methods of treatment and that none has been conclusively proved to
be better or worse than the others. If the patient expresses a strong
wish for a particular treatment (such as wide local excision) and
persists in that wish after explanation of the pros and cons, then she
will have the treatment she wants and be excluded from trials in
which total mastectomy is part of the protocol.
Our job as clinicians is to help our patients, not to confuse them by
telling them that part of their treatment will be determined by what
is, to all intents and purposes, the toss of a coin.
Scarborough Hospital,
Scarborough,
North Yorkshire YO12 6QL A. V. POLLOCK
1. Zelen M. A new design for randomised clinical trials. N Engl J Med 1979; 300:
1242-45.