the most
frequent order of descending
plutonium concentrations, were respira-
tory lymph nodes, lungs, liver, and
bone. In the two cases where urine as-
says definitely indicated a significant
positive exposure and analyses of both
lymph nodes and bone were possible,
the lymph nodes had plutonium concen-
trations 50 times higher per gram of tis-
sue than the bone. Thus, inhalation ex-
posures resulted in the entry of
plutonium into the respiratory lymph
nodes, a phenomenon that should obvi-
ously not have been seen (and was not
seen) in the injection studies. (For a
summary of what has been learned
from autopsy studies, see “A True Mea-
sure of Exposure—the Human Tissue
Analysis Program at Los Alamos.”)
Additional Data from the
Plutonium Patients
In 1969, Patricia Durbin, a biophysicist
at the University of California, Berke-
ley, was involved with metabolic work
on various radioisotopes, including
americium, that led her to the published
work on plutonium. Wanting to learn
more, she began investigating the
records and data on the plutonium
human injections and trying to locate
further information about the patients.
In a letter, dated April 23, 1969, to Dr.
John Howard, an administrator at the
University of California Medical Center
in San Francisco, she said:
Most of the patients injected with
Pu were studied at other hospitals
around the country, and although
most were elderly and expected to
have short life expectancies at the
time of injection, some were misdi-
agnosed. Because of this, there
was an understandably great up-
roar when the civilian A.E.C. took
over from the Manhattan Engineer
District. As a result, the human
data thus obtained was classified
“Secret”, and so it remained for
some years. All efforts to follow
up on those persons who had been
The Human Plutonium Injection Experiments
Number 23 1995 Los Alamos Science
217
One of the problems in applying the Langham power-function model to urine
assays for plutonium workers was how to work backwards from the data to an
estimate of the body burden. Urinary excretion data were usually low-level
values with considerable scatter. Was a jump in a person’s excretion rate due
to analytical variations, physiological changes, or the result of a recent expo-
sure? A method was needed that eliminated suspect data and then weighted
all the remaining data in the determination of the effective dose, or body bur-
den, and the effective exposure time for the Langham power-function.
In 1960, James N. P. Lawrence at Los Alamos devised a computer program
(called PUQFUA), based on the plutonium excretion power functions, that at-
tempted to account for multiple or continuous exposures occurring over a peri-
od of time. Basically, the work period was split into intervals between urine
samplings and each interval was treated as a separate exposure incident.
Using the Langham power function, the dose for that interval was calculated
from the observed increase of plutonium in the urine over what was expected
from previous exposures. If there was no increase, the exposure for that in-
terval was set to zero, and if there was a decrease from what was expected,
the previous data point was rejected, which helped eliminate contaminated
samples (later versions of the code rejected data more than 2 standard devia-
tions from the expected value). The total excretion at any given moment was
then effectively the sum of many Langham power functions, one for each in-
terval, each on its own time scale. The retained plutonium at any given time
was the sum of all the original exposures corrected for excretion losses.
One advantage of the PUQFUA method was that essentially all the urine data
were used to calculate a body burden rather than, as previously, using either
a single urine assay or an average over a time interval. Individual assay
points could fluctuate greatly (because of analytical variations, contamination,
or physiological changes). Lawrence’s approach weighted all but the rejected
assays equally and, thus, was more likely to arrive at a reasonable estimate.
It should be emphasized that this approach, or any approach based on the ex-
cretion equations, was pertinent only for plutonium that had entered the blood
stream and could be excreted by the kidneys. The program could, thus, cal-
culate an effective measure of internalized plutonium, but the result did not
give any indication of how much plutonium might be trapped in the lungs.
Only when such plutonium had worked its way into the blood stream would a
fraction of it appear as excreted plutonium.
Calculations with PUQFUA indicated that the body burdens of twenty-six Los
Alamos plutonium workers (occupationally exposed at Los Alamos between
1944 and 1946 and in the UPPU study of Langham, Hempelmann, and Voelz)
were 60 per cent higher than Langham had estimated with his approach,
which suggested that Langham’s power-function method underestimated pluto-
nium retained in the body. However, we now know that the overestimate is
due to long-term urinary excretion that is truly higher than what is predicted by
the Langham model. When a modified version of the PUQFUA code is used
that properly accounts for long-term data (10,000 days), the predicted body
burdens are consistent with the values obtained from tissue analysis studies.
A Computer Analysis of Plutonium Excretion