DOE Openness: Human Radiation Experiments: Roadmap to the Project
Postwar TBI-Effects Experimentation: Continued Reliance on Sick Patients in Place of Healthy "Normals"
Chapter 8: Postwar TBI-Effects Experimentation: Continued Reliance on Sick Patients in Place of Healthy "Normals"In October 1950, the Air Force entered into a contract with the M. D. Anderson Hospital for Cancer Research in Houston, Texas, to provide the DOD with data obtained from TBI studies on cancer patients. Dr. Shields Warren, who seemed to oppose human experimentation on healthy persons during the NEPA debates, did not appear to have any misgivings about this project.(ACHRE Research Project Series, Interview Program File, Targeted Interview Project), 14ff. By the end of that decade, the DOD would have several contracts with TBI researchers. When, in 1959, a DOD newsletter announced the renewal of a TBI contract between the Army and the Sloan-Kettering Institute in New York, readers were told: "It is hoped to make this work [by Sloan-Kettering] as well as the work of Baylor University College of Medicine and University of Cincinnati a complete program to provide us with answers on the human whole body radiation effects." The Navy also conducted TBI-related research in conjunction with patient treatments at the Naval Hospital in Bethesda, Maryland. All five of these studies used TBI on many patients with radioresistant cancers. In contrast, physicians at the AECs hospital in Oak Ridge operated by the Oak Ridge Institute of Nuclear Studies (ORINS), a university-based consortium, chose to perform TBI only on patients with radiosensitive diseases. In each project, the research institutions accepted the dual purposes of treating the patients' illnesses and collecting and analyzing postexposure information for the military.
The DOD-funded experiments would seek to address the three main questions the military wanted answered: How do different doses of radiation correlate with the acute effects? How do different doses of radiation influence psychological effects? And most important, is there a way to find a biological dosimeter to measure how much radiation someone has received? The military was also interested in the diagnosis and treatment of radiation injuries. One reviewer of the initial Cincinnati proposal described the interest in finding a biological dosimeter: if "accurate knowledge of the total dose of radiation received could be determined it would be of inestimable value in case of atomic disaster or nuclear warfare."
When the DOD contracted with medical professionals to perform additional research on their patients receiving TBI, it is not clear whether department officials believed that the TBI itself should be covered by the ethical standards being established for "human experimentation" following the Nuremberg trials. For example, in the NEPA debates, Dr. Robert Stone distinguished experiments, which involved healthy persons such as the prisoners, from studies, which involved sick patients. Did Stone mean by this that patients receiving treatment did not need to give informed consent, while healthy subjects should? The AEC, for example, took the view that the consent standards should apply to patients. Indeed, as we saw in chapter 1, AEC General Manager Carroll Wilson wrote in a 1947 letter to Robert Stone that "the patient [must] give his complete and informed consent in writing."
In 1953, Secretary of Defense Charles Wilson issued a memo establishing the Nuremberg Code as DOD policy. The Wilson memo required that all experimental subjects sign a statement that explained "the nature, duration, and purpose of the experiment; the method and means by which it is to be conducted; all inconveniences and hazards reasonably to be expected; and effects upon his health or person which may possibly come from his participation in the experiment." Unlike the AEC's 1947 pronouncements, the 1953 Wilson memo did not explicitly refer to patients. In addition, if DOD officials believed that the experiments they were sponsoring did not include the administration of the radiation, but only the collection of postradiation biochemical and psychological data, then they might have interpreted the Wilson memo as applying only to the postexposure testing, not to the radiation treatments.
Although the Wilson memo was classified, its requirements were reiterated by the surgeon general of the Army in a 1954 document that was transmitted to at least some university contract researchers. The Committee found no evidence that this memo was transmitted to the TBI contractors in particular. As discussed in chapter 1, in 1952 Congress had passed legislation that provided for Defense Department indemnification of private contract researchers in cases where human experiments resulted in injury to subjects. As we have seen, the DOD appears to have linked the requirements of this statute to contractor adherence to the principles stated by the 1954 Army surgeon general memo, including written consent of the subject. For example, in a March 1957 letter to the University of Pittsburgh, which was proposing to use medical student-volunteers in a (nonradiation) experiment, the Army stated that the indemnification provision in the contract was "contingent upon your adhering to the following [March 1954 Office of the Surgeon General] principles, policies, and rules for the use of human volunteers in performing subject medical research contracts." Although this indemnification provision was in the contract of at least one of the five institutions that conducted DOD-sponsored TBI-effects research, no available information indicates that its inclusion demanded adherence to the principles set forth by the surgeon general. Nonetheless, at least three of the institutions had written forms authorizing the radiation treatment procedure, although the forms did not explicitly spell out all of the risks and benefits of the additional experiments. This chapter will now review what is known about the five DOD-sponsored experiments.
M. D. Anderson Hospital (Houston, Texas)The Air Force's School of Aviation Medicine (SAM) contract with M. D. Anderson Hospital for Cancer Research, in association with the University of Texas Medical School, declared that the Air Force was willing to use sick patients for the needed data because "human experimentation" had been prohibited by the military:
The most direct approach to this information would be by human experiment in specifically designed radiation studies; however, for several important reasons, this has been forbidden by top military authority. Since the need is pressing, it would appear mandatory to take advantage of investigation opportunities that exist in certain radiology centers by conducting special examinations and measures of patients who are undergoing radiation treatment for disease. While the flexibility of experimental design in a radiological clinic will necessarily be limited, the information that may be gained from the studies of patients is considered potentially invaluable; furthermore, this is currently the sole source of human data.
The M. D. Anderson TBI-effects study extended from 1951 to 1956 and involved 263 cancer patients. M. D. Anderson had a well-established and ongoing radiation treatment program. The project began at the same time that M. D. Anderson received the first cobalt 60 teletherapy unit developed by the AEC's Oak Ridge Institute of Nuclear Studies (ORINS). The M. D. Anderson study involved three phases beginning with low doses--15 to 75 R--and gradually increasing to a maximum of 200 R. The patients in the first group were "in such a state that cure or at least definite palliation could still be expected from established methods of treatment [in addition to the TBI]." Based on these results, the researchers then moved to the second phase, which involved doses ranging from 100 to 200 R. The researchers noted that this greater possible risk necessitated "the selection of patients whose disease had advanced to such a state that, in general, significant benefit could not be expected from conventional procedures other than systemic ones."
The final phase involved thirty patients, all of whom had radioresistant carcinomas for which "cure by conventional means was regarded as completely hopeless." These thirty received the highest doses, 200 R, and reportedly "knew about the advanced state of their disease and the experimental nature and possible risks of the proposed radiotherapy." The Advisory Committee has found no written documentation on what types of risks were described to and understood by these patients. Beginning in 1953, patients signed a release form authorizing the physicians to administer "x-ray therapy, radium and radioactive isotopes, . . . which in their judgment they deem necessary or advisable, in the diagnosis and treatment of this patient." This form was designed apparently to waive legal liability, but did not inform the patient of the risks and benefits of treatment and thus did not meet the other requirements established by the 1953 Wilson memo.
With respect to the biomedical findings, a 1954 Air Force review noted that M. D. Anderson had obtained positive preliminary results by finding a biological dosimeter in the blood. However, one of the reviewers commented that because "the patients were not normal people the changes could very well be the effect of the radiation on the abnormal tissue." The review noted that an effort earlier in the study to find a marker in patients who received repetitive small doses of radiation, similar to what might occur on repeated NEPA flights, was not successful; accordingly, the researchers looked for it in patients who received larger doses in single exposures.
An additional aspect of the M. D. Anderson study was the mental and psychomotor tests that most of the patients were subjected to before and after receiving TBI. (The patients reportedly participated "by their own consent and judgment of the hospital staff.") They performed three tests related to the skills required for piloting aircraft. But the value of testing the abilities of extremely ill patients as a measure for the performance of highly fit pilots was doubtful to the Air Force. In an attempt to lessen this problem, the investigators sought outpatients who were in reasonably good physical and mental condition. Nonetheless, because patients received TBI radiation doses according to the severity of their disease rather than from an arbitrary experimental protocol, there was difficulty in determining whether the performance changes noted resulted from the underlying disease or the radiation.
The M. D. Anderson researchers found medical benefit in three of thirty patients who received 200 R: "200 [roentgens] whole-body x-irradiation produced a definite transitory amelioration of the disease in 3 cases, and a questionable improvement in several additional patients." The study concluded that "the threshold dose, beyond which in a small percentage of patients severe complications begin to appear, lies somewhere between 150 and 200 r." This conclusion seems to have moved the threshold tolerance level for acute effects slightly higher than the 1950 level; at that time the AEC's ad hoc NEPA committee had decided that doses above 150 R would pose "grave risks" to troops.
There is very little information concerning subject selection. It appears that many of the patients were indigent members of minorities, although no information is available to determine whether the ratio of minorities differed in relation to the general hospital population. In the context of an Air Force discussion about the costs of the study, one report noted that "there is some racial problems [sic] involved with colored patients and the colored out-patient maintenance facilities were located in another part of the city and, therefore, it would be difficult to have them transported back and forth to the hospital for testing. . . . Colonel McGraw stated that if we are paying for the maintenance of indigents of the State of Texas with research funds, and the State is also paying for the maintenance of those patients, there could be some difficulty. . . ." Another report stated that "language barriers, both of degree and kind," caused problems in the testing of cognitive functions as part of the psychomotor study.
Several years later, researchers at the School of Aviation Medicine and the University of Texas issued a report comparing the effects of radiation based on TBI treatment of eleven patients (most of whom had radiosensitive diseases) with the M. D. Anderson group. The researchers used the data to report on the civil defense implications that would result from mass exposures to doses between 150 and 200 R. They concluded that 60 percent of the people would experience varying degrees of disability from acute radiation sickness that would cause fatigue, nausea, and vomiting for the first twenty-four hours.
TBI-Effects Studies at Baylor University College of Medicine, Memorial Sloan-Kettering Institute for Cancer Research, and the U.S. Naval Hospital in Bethesda
Within a few years after the Air Force's M. D. Anderson program began, the Army funded two TBI-effects programs with leading cancer centers, both of which appear to have been using TBI to treat radioresistant cancers even before receiving the Army contract. The studies began before M. D. Anderson had published any of its findings.
From 1954 to 1963, Baylor University College of Medicine in Houston, Texas, performed TBI on 112 patients (54 of whom had radioresistant cancers) during the military study; doses ranged from 25 to 250 R, and a 2-megavolt (MV) machine was used in place of a 250-KV machine after the first two years. The principal researchers, Drs. Vincent Collins and Kenneth Loeffler, again sought a biological dosimeter and data on the acute effects of radiation. The researchers noted that even though a significant amount of data had been amassed on radiation effects, no one had been able "to establish clear and dependable relationships with precise physical data." There is no discussion of consent or peer review in any of the twelve available reports or published papers currently available.
The Baylor researchers recognized the same problem that confronted M. D. Anderson: that seeking data from sick patients who require therapeutic TBI treatments may be in conflict with an optimal experimental design. They also noted the problems with giving "last-resort" treatment of this kind:
When patients are referred as a "last resort," the radiotherapist does not wish to withhold treatment that may offer possible benefit but he cannot be certain that the benefit will outweigh the risk. The risk is not that the patient will die but that the undesirable effects of radiation [i.e., bone marrow suppression] will appear more severe in the terminal cancer patient and that the time of death may be destined to coincide with the undesirable effects of radiation.
They concluded that for patients, radiation sickness may be avoided for doses up to 200 R by administering proper care (the researchers suggested that nausea and vomiting for some patients may have been caused by the power of suggestion). They then hypothesized that "with correct information and proper preparation, normal healthy individuals could tolerate even higher exposures without undue incapacitation." Efforts to find a biological dosimeter were said to be unsuccessful because the pool of patients was too small and many either died or were unable to tolerate the necessary tests.
From 1954 to 1961, Dr. James J. Nickson of the Memorial Sloan-Kettering Institute for Cancer Research in New York City performed TBI on more than twenty patients with doses ranging from 20 to 150 R and participated in a DOD study on the acute effects of radiation on humans. Again, the military aims were to find a biological dosimeter and better understand the effects of radiation. Sloan-Kettering was a leading U.S. cancer research center and had a long history of using and experimenting with TBI. The patients selected at Sloan-Kettering had a variety of radioresistant and radiosensitive cancers and were in "relatively good condition." However, patients with kidney, liver, or bone marrow impairment were deliberately excluded from the study because their conditions would "contaminate" biological dosimeter data (the record does not indicate whether these patients who were excluded still received TBI). There is no mention in the currently available records regarding consent of the patients or any form of peer review of the protocol or the experiments.
Between 1959 and 1960, the Navy treated seventeen patients using TBI for a variety of radioresistant and radiosensitive disorders at the Naval Hospital in Bethesda, Maryland, with a cobalt 60 teletherapy unit. The report on these treatments concluded that "total-body radiation therapy in a dose range of 100-400 [roentgens] [air dose] appears to offer relatively safe and reasonably effective palliative therapy for advanced radiosensitive disease." There was no equivalent success on the radioresistant tumors. Urine from some of the patients was collected and retained for analysis to see if there was any amino acid change that corresponded to the radiation exposure received by patients, as part of another attempt to identify a biological dosimeter in the urine. The investigators of the urine study could not find any direct correlation between the dose of radiation and biochemicals in the urine, and they acknowledged that the poor state of health of the patients, as well as age, nutritional state, and renal function changes, may have contributed to this problem.
Surviving patient records indicate that the Naval Hospital used an authorization form, which states that the patient "hereby consent[s] to the performance . . . of total body radiation therapy. This procedure has been fully explained to me by a staff physician of the Department of Radiology." There is no information available to determine if patient permission was or was not given for the collection of urine for evaluation as a biological dosimeter or if the biological dosimeter project had any effect on the patients' treatment. Neither is any information currently available on whether the patients were informed about the additional military research interest in the project, or whether there was any form of review of the project as required by Navy procedures.
The early postwar TBI researchers, such as those at M. D. Anderson, may have been enthusiastic to test the new cobalt 60 teletherapy TBI technology on cancers that resisted older TBI techniques, but by the end of the 1950s the new technology did not appear to be producing any more favorable results on radioresistant cancers. Dr. Shields Warren seemed to confirm this view in a 1959 article in Scientific American; he noted that "radioresistant tumors are generally not treated with radiation because the damage to surrounding tissue is too great."
However, in March 1960 the Defense Atomic Support Agency (DASA) sponsored a conference on the effects of whole-body radiation on humans. A DASA summary of the meeting reported: "First, experience at the dosage levels up to 200 r indicates that man is able to tolerate far greater radiation dosages than was predicted in the NEPA report of 1949; second, there is a need for continuation of this work and, more important, investigation and analysis of the radiation syndrome in man up to the 300 r level, is the next logical area of study." Indeed, DASA had just signed a contract with the University of Cincinnati to provide information on the effects from these higher doses of TBI.
The University of Cincinnati College of Medicine--the Last DOD-Sponsored TBI-Effects StudyThe University of Cincinnati (Cincinnati) was the last institution that the Department of Defense contracted with to collect information on radiation effects from patients exposed to total-body irradiation. From 1960 to 1971, Dr. Eugene L. Saenger and a team of medical researchers from the university's College of Medicine (referred to in this chapter as the "Cincinnati doctors") conducted TBI and partial-body irradiation (PBI) on approximately eighty-eight cancer patients. Cincinnati was the only nonmilitary research institution in the DOD program that did not have preexisting clinical experience with TBI therapy. It was also the only institution using TBI to focus almost exclusively on patients with radioresistant cancers (except for three children with Ewings sarcoma, a childhood bone cancer for which widespread irradiation is still considered an accepted form of treatment.) The military contract was, as before, to obtain more information on the acute effects of radiation and to find a biological dosimeter.
The University of Cincinnati experiments came to public attention in 1971, a time in the national debate over the Vietnam War when university associations with the military were being questioned by students, the press, and the public. Research by Roger Rapoport, who subsequently wrote a book entitled The Great American Bomb Machine, led to a story in the Washington Post on October 8, 1971, that described the Department of Defense contract with the University of Cincinnati to measure radiation effects in humans.
It appears that by 1971 the University of Cincinnati was the only remaining institution doing post TBI-effects studies for the Department of Defense. The publicity prompted the University of Cincinnati to hold a news conference on October 11, 1971, to explain its TBI program. The public attention resulted in three investigations of the Cincinnati experiments, all of which reported their findings in January 1972: ( 1) a January 3, 1972, American College of Radiology report in response to a request by Senator Mike Gravel (the ACR report), which was generally supportive of the program; (2) a January 1972 Ad Hoc Review Committee of the University of Cincinnati Report to the Dean of the College of Medicine on "The Whole Body Radiation Study at the University of Cincinnati" (the Suskind report), which probed the facts and supported the overall objectives of the study; and (3) a January 25, 1972, "Report to the Campus Community" of the Junior Faculty Association of the University of Cincinnati (the JFA report),well as the blood counts and survival times of the subjects. The JFA then held a press conference on 25 January 1972, charging that "many patients in the project paid severely for their participation and often without even knowing they were part of an experiment" and asking that the study be terminated by the president of the university (which did in fact take place the following March). The JFA report was covered widely in the press and in a number of subsequent studies. which severely criticized the TBI program. Following these reviews, the president of the University of Cincinnati decided not to renew the DOD contract in the spring of 1972. The use of TBI was suspended after that time, and the effects study was ended. As recently as April 1994 in a congressional hearing, an ACR representative reiterated its belief that the Cincinnati project was reasonably conducted based on the standards of the time, even if "one might judge them harshly from a perspective 20 years later."
Because of this public attention, a substantial number of documents concerning the University of Cincinnati experiments were preserved, including the original application and subsequent progress reports by the researchers for the Department of Defense, records of the Faculty Committee on Research (the Cincinnati IRB) review of a midcourse research protocol, relevant medical literature, and certain patient medical records. In addition to reviewing these documents, the Advisory Committee staff also interviewed Dr. Eugene Saenger, the principal investigator of the study; and the Committee and staff met with and heard from numerous patient family members and other critics of the Cincinnati experiments. The Advisory Committee also held a public hearing in Cincinnati on October 21, 1994, where more than thirty family members and other interested parties related their concerns about what they believed was wrong with the Cincinnati TBI experiments, chiefly that informed consent was inadequate. Other family members have appeared before the Advisory Committee at another public hearing in Knoxville, Tennessee, and at the Advisory Committee's meetings in Washington, D.C. The Committee also heard public testimony from Dr. Bernard Aron, a coresearcher of Dr. Saenger, and heard from counsel for Dr. Saenger and others involved in pending litigation.
What Was the Purpose of the University of Cincinnati TBI Program?The experimenters were supported by the military to find a biological dosimeter and provide additional human performance data of military interest. There is no question that the patients were seriously ill with terminal cancers; indeed, many received other forms of treatment in addition to TBI, including surgery, chemotherapy, and localized radiation. Although there is no indication that the Defense Department had any direct role in patient selection or treatment, there have been questions raised publicly as to whether the military interest influenced or at all compromised the physicians' willingness to objectively present reasonable treatment options (including no treatment at all) to these cancer patients. Thus, the Advisory Committee has sought to determine what effect, if any, the DOD contract requirements had on the actual treatment of patients.
In 1958 Dr. Saenger applied to the Department of the Army for funding of a research proposal entitled "Metabolic Changes in Humans Following Total Body Radiation." (Dr. Saenger had joined the radiology department of the University of Cincinnati College of Medicine in 1949 and became the director of its Radioisotope Laboratory in 1950, serving until 1987. Before and after starting the TBI program, he was a consultant in radiology to the Army, the Air Force, the AEC, DASA, and the PHS.) The primary purpose of his proposal was to determine whether amino acids or other biochemicals in the urine could "serve as an indicator of the biological response of humans to irradiation."
Later, the first of approximately ten progress reports to DASA described the purpose of the research program it was funding as "to obtain new information about the metabolic effects of total body and partial body irradiation so as to have a better understanding of the acute and subacute effects of irradiation in the human." The second progress report added that "this information is necessary to provide knowledge of combat effectiveness of troops and to develop additional methods of diagnosis, prognosis, prophylaxis and treatment of these injuries." The study would focus generally on post-TBI effects in patients with radioresistant carcinomas; those with radiosensitive lymphomas and other hematological diseases were for the most part not included, with the exception of the children with Ewing's sarcoma. Dr. Saenger reported to DASA in 1962 that the further studies would be conducted "so long as the following criteria are fulfilled: 1. There is a reasonable chance of therapeutic benefit to the patient. 2. The likelihood of damage to the patient is no greater than that encountered from comparable therapy of another type. 3. The facilities for support of the patient and complication of treatment offer all possible medical services for successful maintenance of the patients well being."
Midway into the study, the post-TBI-effects researchers added a program of psychological and psychiatric testing, to determine "whether single doses of whole or partial radiation produce any decrement in cognitive or other functions mediated through the central nervous system." They also recorded data on the incidence of nausea and vomiting from radiation. Within the first three years, the Cincinnati doctors reported to the DOD the information that the March 1960 DASA conference had sought, that "[h]uman beings can tolerate doses of 200 rad (300 r) relatively well as far as combat effectiveness is concerned."
In 1973, two years after their work terminated, the Cincinnati doctors published a journal article describing the purpose of their irradiation study as "to improve the treatment and general clinical management and if possible the length of survival of patients with advanced cancer." Unfortunately, no written research protocol now exists for this treatment study, nor did Dr. Saenger state that they had a written protocol while carrying out the TBI palliation treatment study. This lack of a written protocol is consistent with the confusion doctors had at this time (and to a lesser extent today) distinguishing what constituted research from what constituted innovative treatments (see chapter 2).
The clinical objective of the Cincinnati TBI treatments remains difficult to categorize precisely even now. Dr. Saenger stated in a 1994 interview with Advisory Committee staff that there was no need for an experimental treatment protocol because the TBI treatments were given as a palliative cancer therapy for people for whom there was no better alternative. In contrast, the Suskind and ACR reports seem to have assumed that the TBI treatments were experimental; they both describe them as being in "Phase II" of a standard three-phase experimental process.
Because the Cincinnati doctors recognized that higher doses of TBI (150 rad and above) were causing severe bone marrow suppression in some of the patients, beginning in 1963 they sought to develop countermeasures through the use of bone marrow transplants. Over a six-year period, they instituted a program to remove bone marrow from the patient prior to the radiation and to reinfuse it afterward so as to counter any deleterious effect. In 1966, they submitted a protocol on bone marrow transplantation to the College of Medicine's institutional review board, which provisionally approved it in 1967. However, the use of high-dose TBI continued during this time, with the first successful transplant being administered in 1969.
Critics have suggested that the irradiation of the patients who were subjects of the Cincinnati experiments may not have taken place at all in the absence of DOD funding. The TBI regimen did not begin until after the DOD funds were secured in 1960. The DOD provided a total of $651,482 for the TBI-effects study. In addition to the DOD funds, Dr. Saenger has estimated that the hospital spent $483,222 on patient care. Dr. Saenger has stated that he was very careful to separate the DOD work from the patient care and to make sure that the DOD funding was in no way used for the patient therapy. For this reason, he states that he was never personally involved in patient selection or treatment, in order not to influence the judgment of the attending physicians.
When asked if TBI treatment could have begun before the DOD money arrived, Dr. Saenger said: "No, we had to, we hired some people. We had laboratory equipment to set up. . . . It proceeded as one, as really a sort of a two-pronged investigation." Dr. Saenger stated that the work for the DOD "started when we started [administering TBI]--this [DOD] protocol permitted us to get a technique going in trying to look at whole body radiation in comparison with other forms of palliation." Dr. Saenger also said that "if we had found in the first ten or twelve patients a clear biochemical indicator, we possibly would have done something else. We kept being on the edge of finding what we were looking for so we kept on treating the patients."
In the 1969 proposal to renew the DOD contract, Dr. Saenger wrote that, in light of "world tensions from the possibility of nuclear warfare on any scale . . . it is necessary to pursue with increased diligence the scientific investigations of acute radiation effects and the attendant treatment possibilities in the human being." In outlining a plan to compare total-body, partial-body, and trunk and thorax irradiation, the proposal noted that in most cases bidirectional radiation would be used for each of these treatments, but that "whenever possible unidirectional radiation will be attempted since this type of exposure is of military interest." There is no available evidence to show that the Cincinnati doctors ever actually used unidirectional radiation.
The military's interest in the onset level for the acute effects of radiation, such as nausea and vomiting, led the Cincinnati doctors to intentionally withhold from the patients, as discussed later in the chapter, any premedication or information about these effects for the first three days after irradiation in order not to induce them via psychological suggestions. No mention was made of nausea or vomiting in any of the consent forms.
To the extent that palliation of cancer symptoms was the goal of the Cincinnati doctors, TBI presumably would have been given either as part of a planned experimental protocol or as conventional clinical therapy. If the former, then the currently available evidence indicates rather poor scientific design, even by contemporary standards; if the latter, then the TBI treatment administered for the vast majority of patients was nonstandard therapeutic practice for patients with radioresistant carcinomas at that time.
Another reviewer noted that Saenger's study would "augment work being done by Dr. Collins at Baylor and the Sloan-Kettering Institute who are working with humans." This point was reiterated when the contract was approved, at which point the approving officer declared that "diversification is required to achieve adequate results in a field of whole body radiations [sic] in humans." A third reviewer noted that correlating tumor response to total dose of irradiation "would be of great value in the field of cancer . . . [and] in case of atomic disaster or nuclear accident."
There is no indication that the Army reviewers considered whether any therapeutic benefits to the patients outweighed the risks that the TBI treatments might pose. These reviewers seemed to have based their support for funding this proposal on the military's need for collaborative researchers and the reputation of the applicant, rather than on the substance of the science within the application or their knowledge of radiation therapy practice at that time.
There is no evidence that the DOD reviewed the treatment of the patients as the study progressed, even though the University of Cincinnati appears to have been the only federally funded institution in the country that was treating radioresistant carcinomas with total-body irradiation at that time. The Cincinnati doctors administered doses up to 250 rad, and had indicated in their first DASA progress report that they planned to go up to 600 rad, without seeming to raise any concerns within the DOD contract office.
Subsequent internal reviews by Cincinnati committees raised several concerns. In March 1966, Dr. Saenger and a colleague submitted a protocol entitled "Protection of Humans with Stored Autologous Marrow" to the University of Cincinnati FCR. This proposal was considered an adjunct to the TBI treatments, which Saenger said he did not consider an experiment, and was therefore not subject to review. Some members of the FCR, however, raised concerns that attended to the underlying TBI treatments. These questions included whether each patient was advised that "no specific benefit will derive to him," the need for a more detailed description of the potential hazards, and whether the irradiation would "influence the morbidity or the mortality in these patients."
The proposal was revised and resubmitted on March 1967 under the title "The Therapeutic Effect of Total Body Irradiation Followed by Infusion of Stored Autologous Marrow in Humans." A five-person FCR subcommittee reviewed the proposal. One member, Dr. George Shields, recommended that the study be disapproved because "the radiation proposed has been documented in the author's own series to cause a 25% mortality. . . . I believe a 25% mortality is too high, (25% of 36 patients is 9 deaths) but this is of course merely an opinion." Shields added that if the study were to be approved, then his concern could be addressed by improving the consent process--that is, by ensuring that "all patients are informed that a 1 in 4 chance of death within a few weeks due to treatment exists, etc." Another member, Dr. Thomas E. Gaffney, initially recommended disapproval for several reasons, including the "considerable morbidity associated with this high dose radiation," but he subsequently recommended approval along with Dr. Harvey Knowles, Dr. Edward Radford, and R. L. Witt. The proposal was then given "provisional approval" on May 23, 1967. The requirements did not include Shieldss recommendation on mortality, but did stipulate that "the protocol should be modified to indicate that the exclusive purpose of the study is to determine the therapeutic efficacy of whole body irradiation." There is no written evidence as to whether the FCR ever re-reviewed and approved the revised protocol or the new consent forms that the investigators produced in response to this review.
In 1970, the FCR reexamined the bone marrow protocol because it had not been reviewed since it received provisional approval in 1967. Following two protocol revisions intended to meet the committee's concerns, the FCR noted that it still could "not find adequate methods of evaluation in this study protocol. . . . The real problem seems to be how are we going to evaluate the effectiveness of marrow transplants in protecting against the side effects of total body irradiation. Secondly, how are we going to evaluate the effectiveness of total body irradiation." Nonetheless, after yet another revision, the protocol was approved on August 9, 1971.
In April 1972, Dr. Edward Silberstein, a colleague of Dr. Saenger, submitted a protocol to the FCR entitled "Evaluation of the therapeutic effectiveness of total and partial body irradiation as compared to chemotherapy in humans with carcinoma of the lung and colon." This presumably was to be the next experimental phase of the ongoing TBI work, for which an NIH grant was also contemplated. But that same month, the president of the University of Cincinnati refused to allow continued DOD funding for the post-TBI treatment data collection and analysis following the negative public attention brought to the study. TBI was suspended pending FCR review. Dr. Silbersteins protocol was approved by the university's FCR in August 1972 as a grant application to the NIHs National Cancer Institute. However, in February 1973 the NIH elected not to fund the proposal.
Background to Grant Approvals: This research had, of course, been submitted to the DOD initially with yearly reports since the initiation of the project. The entire total body irradiation protocol, including all the therapeutic, metabolic, chemical, hematologic, immunologic, and psychologic studies was incorporated as one of the components for the General Clinical Research Center (GCRC) grant submitted to the NIH in 1966. There was a site visit and counsel visit on this grant application and many aspects of the radiation project were presented to the scientific review committees. This NIH grant was given full approval and was funded. It came up for renewal in 1970, and again all aspects of the radiation study were incorporated. This grant also had full approval by the NIH.
However, a 1969 internal FCR memorandum from the then-chairman of the Cincinnati FCR, Dr. Thomas Gaffney, noted that the NIH had rejected a University of Cincinnati grant application for TBI research "on ethical grounds," even though it had been approved by the FCR: "We learned that two applications received by NIH from this institution have been rejected on ethical grounds. Both had been through this committee. As far as I know, neither of the principal investigators involved were notified of the reason of the rejection by the NIH. One of these grants was the total body radiation study in patients with malignancy. . . ."
In 1974, D. T. Chalkley, then chief of the NIH's Office for Protection of Research Risks, vigorously responded to a magazine article criticizing the Cincinnati experiments. Chalkley stated that "none of the patients involved died from radiation sickness. . . . In all instances, death was clearly attributable to the advance of cancer, or to intercurrent disease associated with advanced cancer."
Risk of TBI on Mortality and MorbidityThe risks associated with total-body irradiation were reported by all of the previous DOD-sponsored TBI research institutions and were known to the Cincinnati doctors. Midway into the program, the Cincinnati doctors acknowledged that the TBI treatments posed a risk of death: "bone marrow suppression was the most life threatening radiation effect at the doses used." In their 1966 report to the DOD, they noted that the general response of the first fifty patients was that their "total white count falls to a low point 25 to 40 days after irradiation. There was lymphopenia [low white blood cell count] which persisted for 40 to 60 days." The same report stated that "severe hematologic depression was found in most patients who expired." Although the efforts by the Cincinnati doctors to employ bone marrow transplantation in response to this problem did not succeed until 1969, they continued to administer TBI without bone marrow transplantation throughout the six-year interim period: thirteen patients received doses of 150 or 200 rad TBI (including five on whom autologous marrow infusion was attempted but did not succeed); nine of these patients died between twenty-five and seventy-four days after being irradiated, and the other four survived longer.
Some relatives of deceased TBI patients contend that their relatives may not have been as seriously ill as the reports claim. While all patients had advanced cancer (indicated either by the presence of metastatic or locally advanced tumors) and thus could be considered "end stage" in terms of unlikely curability, they were clearly not all "near death," in that family members reported some of the patients feeling well enough to carry on normal activities of daily life (e.g., holding down jobs, caring for children) until the day they received the TBI. Patient status reports written by the Cincinnati doctors seem to bear out this view. For example, the first seventeen patients were described as all having "incurable and/or metastatic cancer . . . although in reasonably good clinical condition [emphasis added]." Similarly, the 1969 DASA report states that the patients "have inoperable, metastatic carcinoma but are in relatively good health [emphasis added]." The 1970 DASA report states that the studies conducted during the prior year "were all performed on ambulatory human subjects . . . [who] were all clinically stable, many of them working daily [emphasis added]." This report also noted the "comparatively better physical condition of these new subjects" and went on to state that "only three of our 11 new patients died in less than 100 days following irradiation. This was in sharp contrast to the almost 50 percent low survival rate for earlier years in this study," when lower radiation doses had been administered.
Although the Advisory Committee has received some partial patient hospital records, it has not analyzed the records of every patient, which would be required to determine if any deaths could be attributed to the TBI alone, or if such conclusions could be reached at all from the data currently available. (The Committee did not have the time or resources to review the individual files of every patient from this and the numerous other experiments that it has investigated.) Contemporaneous reports, however, state that TBI treatments may have contributed to the deaths of at least eight and as many as twenty patients. The Suskind report, for example, said that "19 died within 20-60 days and possibly could have died from radiation alone," but noted that bone marrow failure was found in only eight at the time of death. (An additional death occurred six days after irradiation to a patient under anesthesia in the course of a bone marrow transfusion to support the TBI, bringing the number to twenty.) The Suskind report also stated that "there is absolutely no evidence that whole body radiation shortened the period of survival of the treated patients," referring, apparently, to the statistical "survival rate" of the entire group of patients. Similarly, the 1972 ACR report associated the death of eight patients to the fact that "the bone marrow function was subnormal and thus relatable to radiation syndrome." The ACR report also noted that "it is not possible to determine positively that those patients who died within 60 days of the treatment would not have succumbed to their disease within that period, even though the clinical assessment had been that their disease was stable enough to justify their inclusion in the study."
Similarly, following the completion of their study, the Cincinnati doctors wrote that "if one assumes that all severe drops in blood cell count and all instances of hypocellular or acellular marrow at death were due only to radiation and not influenced by the type or extent of cancer and effects of previous therapy, then one can identify 8 cases in which there is a possibility of the therapy contributing to mortality." In 1994 Dr. Saenger wrote:
It is important to realize that in any given patient it is not possible to determine objectively whether death occurred too soon or was prolonged as a consequence of treatment. The only way that an estimate can be made is to compare the length of survival of a group of patients with the same tumor and extent of tumor treated by radiation to a group of patients with the same tumor and extent of tumor treated by different methods.
The attempt by the Cincinnati doctors in their 1973 article to compare survival rates for their TBI patients with the statistics from other published reports is problematic for a number of reasons: first, the comparisons were not controlled for known prognostic factors, such as age, tumor subtype, and stage; second, comparisons with external and historical comparison groups are easily confounded by unmeasured characteristics, such as differences in patient populations and trends in prognosis over time; and third, survival from time of irradiation (some considerable time after diagnosis) in the TBI series is compared to survival from diagnosis in the published series, without using appropriate survival analysis methods. A more meaningful comparison would have been between subgroups of the patients receiving different doses of radiation, preferably including a concurrent unexposed chemotherapy group, adjusting for the interval from diagnosis to the time of death and other relevant prognostic factors. Although limited statistical data were made available to the Advisory Committee, they were not adequate to allow meaningful statistical analysis, and it was not feasible for us to abstract the necessary data from the charts. The Suskind report stated that "before 1966 the design of the study to measure palliation was unstructured and not uniformly applied, particularly as regards uniform definitions and methods of reporting." The report also noted that "it is uncertain whether this study and similar studies reported in the medical literature are truly comparable in all major factors that influence survival, such as selection of patients and ancillary medical management. Therefore, the significance of comparisons of survival rates is doubtful, unless marked differences are found."
The nature of the DOD-sponsored research raises additional concerns as to whether patients were subjected to unnecessary discomfort without full disclosure of experimental purpose or prior consent. In order to collect data on certain side effects of radiation for the military, the patients were not premedicated or informed of potential acute side effects of TBI such as nausea and vomiting so as not to induce these effects psychologically. Patients were to be treated to relieve their symptoms if they affirmatively requested medication. In contrast, researchers at the City of Hope Medical Center conducting a purely clinical TBI study (from 1960 to 1964) gave antinauseant medication to all patients who received 40 or more rad within one hour prior to being irradiated to alleviate possible side effects.
Informed ConsentThere is no indication that the DOD ever informed the Cincinnati doctors about the secretary of defense's 1953 Nuremberg Code directive or any subsequent Army implementation of the directive. It is not clear what patients (or family members) were told about the TBI program in the early years of the experiments, because written consent forms were not standard practice at that time. During the later years of the program, written consent forms were employed, but they have been criticized for not clearly stating all of the risks involved.
Written consent forms were first produced and used in the experiments in 1965, two years before they were required by the University of Cincinnati review committee and NIH; the form was revised twice thereafter. According to the Cincinnati doctors' 1973 article summarizing the study, all patients gave informed consent in accordance with the requirements of the Cincinnati Faculty Committee on Research and the National Institutes of Health. Although by the end of the study the consent forms did describe the TBI procedure and its effects, information about risks associated with TBI--nausea and possible death from bone marrow suppression--was not included in these forms.
The first Cincinnati form, dated May 1, 1965, is entitled "Consent for Special Study and Treatment." It states that the "nature and purpose of this therapy, possible alternative methods of treatment, the risks involved, the possibility of complications, and prognosis have been fully explained to me. The special study and research nature of this treatment has been discussed with me and understood by me." There was no mention in the form about the possible risk of death from bone marrow suppression or of the possible side effects of nausea and vomiting, which the doctors were studying and did not want to induce by suggestion. There is no available documentation on what the patients were told orally about the "risks involved." Because Dr. Saenger was not responsible for recruiting or treating patients, he could not speak to what was actually said to the patients.
In 1981, Dr. Robert Heyssel, director of Johns Hopkins Hospital, discussed the ethical climate before the mid-1960s:
I should say that in the climate of those times . . . that many things were done with human subjects, including the investigator himself, which would no longer be condoned. . . . None of these activities had to be reviewed by anyone else in any formal sense within the institutions. I think this was the situation probably up to and around 1966 in most institutions. I am not suggesting that that was the proper thing; I am simply saying that was the case. In terms of experimental therapeutics, I think an honest effort was made by most investigators to explain to families, to patients, that what was being done was in the range of the untried or experimental, but there were certainly no informed-consent rules that anyone was operating under during that period of time up to the midsixties.
A second consent form went into effect in 1967, following the 1967 FCR review of the bone marrow protocol discussed above. This form listed the risks as these: "The chance of infection or mild bleeding to be treated with marrow transplant, drugs, or transfusion as needed." It also said that consent was for "a scientific investigation which is not directed specifically to my own benefit, but in consideration for the expected advancement of medical knowledge, which may result for the benefit of mankind." One member of the University of Cincinnati FCR had suggested in 1967 that the consent form should inform the patient of a one-in-four risk of death. The 1967 FCR review of the protocol required only that the form make clear "the danger inherent in the method and the steps intended to protect the patient."
In 1971, a third form came into use, following the second Faculty Committee review of the bone marrow protocol. This form expanded on the previous form by explaining that "the bone marrow's ability to make [white] cells will be decreased for four or five weeks after you receive your radiation. If you receive a dose of radiation of 200 rads or more, which your doctor will tell you, your blood counts will fall to levels where infection or bleeding could be a problem." It also refined the previous form by describing the research as "a scientific investigation which is not only directed specifically to my own benefit, but also in consideration for the expected advancement of medical knowledge, which may result for the benefit of mankind." There was no mention of any risk of death.
Beginning in 1968, patient consent was solicited over a two-day period. Dr. Saenger described this process: "Dr. Silberstein was the person who did all this, in that phase. He would explain to somebody the first day what the problems were, what was going to happen, what the risks were, etc. or what the benefits were. Then he had the patient and a representative come back the next day, the representative could have been the patient's mother or cousin, or some family person, or it could have been the patient's minister. And you go through the whole thing with the minister, and the patient and family were all happy with this desperate situation, and the signature was affixed." To the extent it was employed, this procedure appears to have been innovative and above the standard practice of the time.
Family members of some patients testified to the Advisory Committee that neither the patients nor their families were adequately informed about the nature and risks of the radiation treatments. They claim that this occurred despite multiple and persistent requests by family members to meet and discuss their concerns with the doctors involved in administering these treatments. Family members also told the Advisory Committee that patients were not informed about the source of the program funding (although it should be noted that such disclosures are still not mandatory in most institutions). The Suskind report noted that "this information was not withheld if the patient asked about this matter. The procedure follows the custom of every other research project in this University." The ACR report stated that "in the last few years they were told that the information might have military as well as clinical significance."
Subject SelectionAll but five of the patients were referred into the study from either the wards of the Cincinnati General Hospital or its Out-Patient Tumor Clinic. The remaining five were private patients, three of whom were children treated for Ewing's sarcoma. The Suskind report noted that fifty-one of eighty-two patients were black (62 percent) and most were indigent; the report commented that "this distribution reflects the patient population of the Cincinnati General Hospital." Psychological data from the TBI study suggest that some of the subjects may have been of questionable competence or may have been temporarily incapacitated. However, the meaning and importance of these data have been criticized and are in dispute.