I 
ANIMAL RADIATION DATA AND THEIR RELEVANCE TO MAN 
Harry E. Walburg, Jr.* 
The relevance of animal data to human radiobiology 
is discussed in terms of whole-body exposure to external 
radiation, late somatic effects, and induction and accel- 
eration of neoplasms. It appears that the implantation 
of nuclear-powered medical devices in a few nonhuman 
primates for less than five years is an inadequate test 
of the radiation risk to man. A reasonable estimate of 
risk can be made by determining the quality of radia- 
tion, its dose-rate, and the dose to the susceptible tissue 
and applying to these data the entire body of knowledge 
on radiation carcinogenesis in man and animals. 
INTRODUCTION** 
The broad topic of the relevance of animal 
data to human radiobiology has received re- 
peated treatment in the literature over the past 
30 years. It will suffice to outline in general 
terms the conclusions dravi^n from examinations 
of experimental animal data without extensive 
documentation which is readily available in 
the literature.i-^ More recent treatments of this 
topic have been published by Storer and Bond 
and by the United Nations Scientific Committee 
on the Effects of Atomic Radiation.^'^ 
In addition to limiting the amount of docu- 
mentation, I will restrict the discussion to the 
following: (1) whole-body exposure to external 
radiation, where there is a relatively uniform 
dose to all cells of the body, permitting an ac- 
curate determination of the dose received by the 
susceptible tissue or cell populations, (2) late 
somatic effects, excluding genetic effects and 
the hematologic disorders seen within approxi- 
mately 90 days following irradiation, and (3) 
induction and acceleration of neoplasms, which 
is the most significant late effect following ex- 
posure to low-level radiation. 
* Biology Division, Oak Ridge National Laboratory, Oak Ridge. 
Tennessee 37830. 
** Research sponsored by the U.S. Atomic Energy Commission 
under contract with the Union Carbide Corporation. 
METHODS 
One of the principal problems we face in 
extrapolating experimental data to man is the 
variability in type and sensitivity of tumors 
induced by radiation in different species and 
even among different strains of the same spe- 
cies. For example, it is well known that mice 
are highly susceptible to the induction of cer- 
tain types of hemopoietic neoplasms (principally 
thymic lymphoma), as well as female endocrine 
organ tumors (i.e., ovarian, pituitary, and 
breast).^ Swine, on the other hand, are suscep- 
tible to induction of liver, intestinal, and uterine 
tumors, while burros show no evidence of 
tumor induction by whole-body radiation.^^ 
Man appears to be most susceptible to induction 
of certain forms of leukemia (i.e., acute and 
chronic myelogenous and acute lymphatic) and 
thyroid carcinoma.^ Such variation is common 
even among strains of the same species. While 
mice of the RFM strain are highly susceptible 
to the induction of thymic lymphoma and 
myeloid leukemia, C3H mice rarely develop 
these diseases following whole-body radiation, 
and C57BL mice are intermediate between the 
other two strains in incidence of thymic lym- 
phoma but do not develop myeloid leukemia 
(H. E. Walburg, Jr., unpublished data). With 
such variation evident, it seems highly unlikely 
that a single adequate model exists which will 
permit direct quantitative extrapolation to man 
of experimental data on radiation carcinogene- 
sis. 
How, then, can such experimental data be 
useful in estimating man's risk to neoplasia 
from radiation? In general there are three pos- 
sibilities, as previously outlined by Mole:^^ (1) 
directly applicable models, (2) useful qualita- 
tive generalizations, and (3) useful quantitative 
generalizations. While directly applicable mod- 
939 
