Findings 15-23

Chapter 17: Findings 1-7

Findings for the Period 1944-1974

Biomedical Experiments

Finding 1

The Advisory Committee finds that from 1944 through 1974 the government sponsored (by providing funding, equipment, or radioisotopes) several thousand human radiation experiments. In the great majority of cases, the experiments were conducted to advance biomedical science; some experiments were conducted to advance national interests in defense or space exploration; and some experiments served both biomedical and defense or space exploration purposes.

These experiments were conducted by researchers affiliated with government agencies, universities, hospitals, and other research institutions. Only fragmentary information survives about most experiments.

Finding 2

The Advisory Committee finds that the majority of human radiation experiments in our database involved radioactive tracers administered in amounts that are likely to be similar to those used in research today. Most of these tracer studies involved adult subjects and are unlikely to have caused physical harm. However, in some nontherapeutic tracer studies involving children, radioisotope exposures were associated with increases in the potential lifetime risk for developing cancer that would be considered unacceptable today. The Advisory Committee also identified several studies in which patients died soon after receiving external radiation or radioisotope doses in the therapeutic range that were associated with acute radiation effects.

Review of available information indicates that the majority of the approximately 4,000 human radiation experiments in the Advisory Committee database involved the use of radioisotopes as tracers in research designed to measure physiological processes in either normal or diseased states. These experiments were not typically aimed at measuring the biological effects of radiation itself. However, information on the majority of experiments in our database was fragmentary and thus did not allow for detailed estimates of dosimetry or examination of issues of experimental design and subject selection.

To supplement the information in our database and provide context to our analysis, we independently reviewed archival documents from AEC-mandated institutional local isotope committees. These local use committees were part of a larger AEC program that facilitated the distribution of radioisotopes for use in government-sponsored human subjects research in the 1947-1974 period and involved the review of experimental risk on an individual basis to ensure that human uses of isotopes were within accepted risk standards of the day. We thus used these materials as an indicator of isotope use and regulatory practices at that time.

While we recognize the limitations of the data available to us, our evaluation suggests that most tracer studies conducted during the period 1944-1974 likely involved low doses that did not cause any acute or long-term effects. The Advisory Committee cannot rule out, however, the possibility that some people were or will be harmed as a consequence of their involvement in these experiments.

The Committee did identify some nontherapeutic tracer experiments involving the administration of iodine 131 to children, which may have raised the subsequent risk of developing thyroid cancer to levels that would be considered unacceptable today. Based on the average risk estimate for each experiment, approximatedly 500 individuals were exposed to greater than minimal risk. (The Committee used a threshold of greater than or equal to one excess case of cancer per 1,000 subjects for categorizing experiments as greater than minimal risk.) Combining the average risk estimates for each experiment, this translates into an expected excess of 1.3 incident cases of thyroid cancer for the entire group. Fortunately, unlike many other cancers, thyroid cancer is curable in more than 90 percent of cases; therefore, it is unlikely that, even if cancers developed, these exposures caused any premature deaths. Furthermore, although there is strong scientific evidence that radiation doses delivered over a short period of time from external sources can result in increases in cancer incidence at specific sites, comparable data suggest that the carcinogenic effects of isotope exposures are less than those of external irradiation. The difference in carcinogenic effect is thought to be due to the relatively low dose rate of the isotope exposure, which allows for effective repair of radiation damage.

One additional isotope study involving the administration of radioiron to pregnant women has been linked to a possible increase in cancers in children who were exposed in utero. However, the small number of observed cancers as well as considerable uncertainties in the amount of radioisotope administered have made the determination of causality difficult. Finally, the Committee found some experiments where radioisotope exposures were associated with either acute or chronic physiologic changes of uncertain clinical significance, pathologic evidence of kidney damage secondary to chemical and radiation toxicity in some patients injected with uranium, and radiographic evidence of minimal bone changes in some long-term survivors of plutonium injections.

Studies that involved radiation doses in the therapeutic range were for the most part performed on patient-subjects where there was, at least arguably, a prospect that the subjects might benefit medically from the exposure. However, the TBI and experimental gallium treatments, in which patients suffered symptoms of acute radiation sickness and died soon after treatment, raise the question of whether their deaths were hastened by the radiation treatments. Resolution of this issue requires review of individual medical histories, which the Advisory Committee could not undertake.

Finding 3

The Advisory Committee finds that human radiation experimentation during the period 1944 through 1974 contributed significantly to advances in medicine and thus to the health of the public.

Human radiation research was essential to the development of new therapies such as the use of radioactive iodine to treat thyroid cancer; the use of phosphorus to treat blood diseases such as polycythemia vera; and the use of radioactive strontium as a palliative in prostate and other cancers metastasized to the bone. Diagnostic uses of radionuclides developed during this period include scanning techniques to identify tumors and radiolabeling techniques that help evaluate a variety of cardiac diseases. The quality of images produced by external sources of radiation also improved dramatically between 1944 and 1974, making possible, for example, techniques such as balloon angioplasty to open occluded arteries.

Finding 4

The Advisory Committee finds that some government agencies required the consent of some research subjects well before 1944. These requirements generally did not stipulate what was meant by consent, however, nor did they generally indicate whether investigators were obligated to disclose specific information to potential subjects. The government did not have comprehensive policies requiring the consent of all subjects of research, including both healthy subjects and patient-subjects, until 1974.

4a. Research Involving Healthy Subjects: In the 1920s, the Army promulgated a regulation concerning the use of "volunteers" for medical research. In 1932, the secretary of the Navy required that subjects of a proposed experiment be "informed volunteers." In 1942 the requirement that healthy subjects be informed volunteers was also articulated by the Committee on Medical Research, which oversaw war-related research for the Executive Office of the President. In 1953, the principle of consent articulated in the Nuremberg Code was adopted by the Department of Defense in a Top Secret memorandum from Secretary of Defense Charles Wilson regarding human research related to atomic, biological, and chemical warfare (this document is known as the Wilson memorandum); in 1954, this application of the Nuremberg Code was expanded by the Army Office of the Surgeon General as an unclassified policy for all research with "human volunteers." A policy of requiring researchers to obtain consent was adopted by the Clinical Center, the research hospital of the National Institutes of Health, in 1953; by the Atomic Energy Commission in 1956; and by the Air Force in 1958. In the 1960s, all branches of the Department of Defense promulgated regulations requiring the consent of healthy subjects, and the Isotopes Distribution Division of the AEC included in its guide for researchers a requirement of consent from all subjects. In 1966, the surgeon general of the Public Health Service issued a policy requiring the consent of all subjects of research conducted or funded by PHS; also in the late 1960s, the Veterans Administration codified in its operating manual a requirement of consent from all research subjects. In 1972, the National Aeronautics and Space Administration adopted similar consent requirements, although exceptions were made for certain subject populations, such as astronauts. In 1974, the Public Health Service policy was promulgated as a regulation for all contracts and grants of the Department of Health, Education, and Welfare. The CIA did not formally adopt consent requirements until 1976, when an executive order mandated that it follow the 1974 regulations of DHEW concerning research involving human subjects.

4b. Research Involving Patient-Subjects: In an April 1947 letter, the AEC general manager stated the AEC's understanding that AEC contract researchers would inform patient-subjects of the risks associated with a research intervention and that patient-subjects express a willingness to receive the intervention. In a second letter, written in November 1947, the general manager specifically stipulated that the AEC require researchers to obtain "informed consent in writing" from patient-subjects where "substances known to be or suspected of being poisonous or harmful" were given to human beings. In 1948, the AEC permitted the administration of "larger doses [of radioisotopes] for investigative purposes," but only with the patient-subject's consent. In 1953, the NIH Clinical Center required consent from all patient-subjects and specified that written consent was to be obtained from patient-subjects involved in high- or uncertain-risk experiments. In the early 1960s, several government agencies adopted consent provisions for investigational drugs; these requirements applied to some radioisotope experiments with patients. In 1965, the AEC required that consent be obtained from all subjects, including patient-subjects, who were administered radioisotopes for experimental purposes, except when it appeared "not feasible" or not in the patient's "best interest." By 1967, the VA required the consent of all patient-subjects. As noted in Finding 4a above, in 1965 the AEC required that consent be obtained from all subjects administered radioisotopes for experimental or nonroutine uses. In 1966 the surgeon general of the Public Health Service issued a policy requiring the consent of all subjects of research conducted or funded by PHS, including patient-subjects. Exceptions to this requirement were permitted for only certain kinds of social science research posing minimal risk. A 1972 NASA policy applied to all subjects of research, presumably including patient-subjects. By 1973, all the branches of the military had promulgated regulations requiring the consent of patient-subjects. In 1974, the PHS policy was promulgated as a regulation for all contracts and grants of DHEW.

Finding 5

The Advisory Committee finds that government agencies did not generally take effective measures to implement their requirements and policies on consent to human radiation research.

Evidence of the implementation of the AEC's consent requirements stated in April and November 1947 letters from the general manager is slim. A document suggests that the April 1947 requirement for a signed statement from two physicians testifying to consent was satisfied in at least one case. However, the Advisory Committee did not find evidence that this or other requirements stated in the 1947 letters were embodied as a provision of AEC contracts involving human subject research or otherwise routinely communicated to contract researchers. Further, there was no reference to the requirements stated in these letters or to the letters themselves in the written material disseminated to researchers by the AEC's program for distributing radioisotopes for "human uses." Moreover, requests for guidance concerning human use policies from investigators at AEC-operated research facilities suggest that the 1947 requirements were not routinely disseminated. Subsequent requirements that healthy subjects be informed volunteers and that consent be obtained from seriously ill patients receiving higher doses of radioisotopes were more widely communicated; we have not been able to determine the extent to which they were actually implemented.

Secretary of Defense Wilson's February 1953 Top Secret memorandum detailing requirements for research with human subjects was rewritten as an unclassified June 1953 directive from the secretary of the Army. It is difficult to determine why these requirements were applied to some activities and not to others. For example, elements of some of these requirements appear to have been implemented in some experiments conducted in conjunction with atomic bomb tests and not in others. In 1954, these requirements were adopted by the Army surgeon general as applicable to all research involving "human volunteers." This 1954 statement was transmitted to contractors as a "nonmandatory guide." However, there is some evidence that the Army sought to include this statement as a condition in at least some contracts.

Evidence of implementation of the NIH Clinical Center's 1953 policy requiring that information be provided to and consent obtained from all subjects is difficult to find; in most cases involving patient-subjects, documentation would not have been required in writing. By contrast, the use of healthy subjects in the Clinical Center required written consent and the "normal volunteer program" appears to have involved greater supervision to ensure that consent was obtained from these subjects.

Finding 6

The Advisory Committee finds that from at least 1946 some government agencies had in place procedural mechanisms for reviewing the acceptability of risks to human subjects from exposure to radioisotopes. By 1974, the government had policies requiring review of the acceptability of risks to human subjects in other forms of research, including research involving exposure to external radiation.

Beginning in 1946 the Manhattan Project, and from 1947 onward the AEC, required some investigators seeking to conduct experiments using radioisotopes supplied by the government to have the risks to subjects reviewed by a committee at the institution where the work was to be conducted and in some cases by the AEC's Subcommittee on Human Applications as well. The AEC required that local committees be composed of at least three physicians or researchers with relevant expertise regarding radiation safety and medical applications. By 1949, it was clear that this policy applied to all investigators using radioisotopes supplied by the AEC.

In 1953 prior group review for risk was also begun at the NIH Clinical Center for proposed human research that involved unusual hazard. No such requirement applied to research funded by NIH but conducted at universities and other nongovernmental research facilities until 1966, when the PHS required that all institutions establish a local peer review committee to evaluate the adequacy of the protection provided to human subjects in each proposal. This requirement was promulgated as an institutional policy by the DHEW in 1971.

In 1953, by adopting the Nuremberg Code, the secretary of defense and the Department of the Army endorsed several principles intended to minimize risk in research with human subjects, at least in regard to the atomic, biological, and chemical warfare experiments that were subject to this policy. In the DOD, both the purpose of proposed research and the level of risk were subjected to prior review through the military chain of command. This was previously required by the Navy at least from 1943, and the Air Force from 1952. However, the extent to which these requirements covered particular research activities (such as healthy subjects vs. patients; radioisotopes vs. external radiation) and particular institutions (such as contractors vs. in-house research) differs and is difficult to reconstruct. Also difficult to reconstruct is the extent to which the risk protection principles of the Nuremberg Code were implemented. In the mid-1960s, concurrent with the adoption of regulations related to investigational drug testing, the DOD and each military service adopted provisions requiring the establishment of a "review board" or committee to oversee proposed research projects involving new drugs. In some cases, such as with the Air Force beginning in 1965, this committee also served to evaluate all other proposals involving human subjects. During this period, the VA also established a review board mechanism for research involving new drugs and investigational procedures.

Finding 7

The Advisory Committee finds that the government program of distributing radioisotopes for use in human subject research included procedures for the review of risk. These were widely implemented by researchers and institutions that used isotopes obtained from the AEC for human experimentation. However, there is no evidence that a parallel mechanism for reviewing the risks of research involving external radiation was in place.

From its 1947 birth, the AEC, as part of its policy to promote the peaceful use of radioisotopes, required private institutions that wished to obtain radioisotopes for "human uses" (including human experimentation as well as patient treatment) to establish local review committees. These committees reviewed proposed human uses under guidelines provided by the AEC's own Subcommittee on Human Applications of the Advisory Committee on Isotope Distribution Policy. This AEC subcommittee reviewed these applications, providing a second level of oversight of risk. By 1949, the AEC's own labs were required to establish local committees and to have human use applications, reviewed by the same AEC Subcommittee on Human Applications. The control of risk, and the assurance of safety to all those involved (including doctors and other health care workers), was a primary purpose of the reviews. The Advisory Committee lacked sufficient evidence to determine whether the system was implemented in all of the many institutions that used government-supplied radioisotopes for human subjects research or whether the system was always adhered to in any particular institution.

In addition to providing for the review of research proposals, the AEC dramatically increased the number of qualified personnel by offering training courses in the safe handling and use of radioisotopes. As individual procedures became routine, the degree of review was lessened; as specific institutions became more experienced, more reviewing authority was delegated to them.

The primary function of the system was to reduce the physical hazards of using radioisotopes, not to enforce any policies regarding consent of subjects. (See chapter 6.)