one rad. However, they were careful to point out that
their analysis did not exclude the possibility of a curved
dose-response relation. The data in their figure five can
also be interpreted to suggest a “threshold” at perhaps
50-100 rads, below which cancers would not be in-
duced. Perhaps roughly half of the radiation-induced
thyroid cancers may appear within thefirst fifteen years
(compare values in Ref. 11 with those in Ref. 12), but
this fraction is uncertain.
There is some experimental evidence that iodine 131
is less effective than X-rays in inducing cancer. Doniach
found that the iodine 131 dose had to be about ten
times greater than an abrupt dose of X-rays to cause
equal effects in adult rats.'* However, adult rats are
not infant humans and the relative effectiveness of
iodine 131 vs. X-rays in children is yet to be established.
Perhaps the iodine 131 exposures in Utah could provide
information on this point.
The dose-response relation is very difficult to estab-
lish at very low doses because the incidence of effects
is so very small. Some radiobiologists feel that cancer
induction is not proportional to dose, but that a cer-
tain “threshold” level must be exceeded before any
cancers can be induced. If such a threshold exists and
if the doses fall below this level, no induced cancers
will appear. It is entirely possible that no cancers will
result from the Utah exposures.
the lower limit—the number of induced cancers cannot
be less than zero! Assumption (B), while between the
upper and lower limits, is not necessarily the best esti-
mate but it yields reasonable values for planning the
experimental search for thyroid cancer. Estimates of the
number of radiation-induced thyroid cancers which
might appear during the lifespan of the 250 thousand
Utah infants for doses of 1.3-10 rads are compared in
Table 3 to their expected natural incidence by age
fifteen.
Similarly, it is instructive to calculate the anticipated
number of radiation-induced thyroid cancers which
might appear during the lifespans of the 565 infants in
Washington County, Utah, exposed to higher thyroid
doses (estimated at 120-440 rads*) following the
“Harry” shot of 19 May 1953. These estimates are
shown in Table 4.
TABLE 4
ESTIMATED THYROID CANCERS
FOR THE 565 INFANTS NEAR ST. GEORGE IN 1953
ASSUMPTION
NUMBER OF CANCERS
(A) lodine 131 effect equals
X-ray effect
(B)
Effect equals 1/10 of
X-ray effect
2
0.2
{C) High Threshold
ASSUMP TION
NUMBER OF CANCERS
{A)
lodine 131 effect equals
X-roy effect
(B)
Effect equals 1/10 of
Xeray effect
(C) High Threshold
1?
-
88
7
-
9
0
ee ee a ee ee ee
“NATURAL” INCIDENCE BY 15
YEARS OF AGE
6
Theoretical estimates of the number of radiationinduced thyroid cancers will nowbe calculated for each
of the following three assumptions: (A) a linear doseresponse with iodine 131! irradiation equally as effective
as X-irradiation, and causing a lifetime incidence of 35
thyroid cancers per million infants each receiving one
rad: (B) a linear dose response, but iodine 131 irradiation only one tenth as effective as X-irradiation: and
(C) a high threshold. Assumption (A) probably sets
an upper Iimit, for it is unlikely that iodine 131 is more
effective than X-rays. Assumption (C) certainly sets
9
— 0.9
0
NATURAL" INCIDENCE BY 15
YEARS OF AGE
TABLE 3
ESTIMATED THYROID CANCERS
FOR THE 250,000 IRRADIATED UTAH INFANTS
-
0.01
If the additional doses received before birth and in
later childhood had been included, the estimated number of radiation-induced cancers would have been even
higher than shown in Tables 3 and 4.
DISCUSSION
1) The St. George study has the advantage that the
“natural” occurrence of childhood thyroid cancers is
extremely unlikely among the 565 infants exposed in
1953. The probability of one natural case is one in one
hundred whereas the chance for two natural cases is
only one in ten thousand. Thus the observation of any
childhood thyroid cancer would seem suggestive of
radiation damage, unless it could be shown that other
factors render these children exceptionally susceptible.
Is it possible that genetic or environmental factors
might increase the “natural’’ incidence in this region
by a factor of one hundred to ten thousand times
greater than for the rest of the USA? I do not know.
But I do knowhowto test this hypothesis. Individual
estimates should be made of the most probable dose to
each St. George child based on source of milk and
DOE ARCHIVES
Scientist and Citizen