Conclusion

Chapter 8: Conclusion

This history provides compelling evidence of the importance of the rules that regulate human subject research today--prior review of risks and potential benefits, requirements of disclosure and consent, and procedures for ensuring equity in the selection of subjects. The history also highlights four issues in the ethics of research with human subjects that are as important today as they were then, issues that are not easily resolved or even addressed by present-day rules. As discussed below, these issues are (1) how to protect the interests of patients when physicians use medical interventions that are not standard care; (2) the effects and attendant obligations of the government when it funds research involving patient-subjects; (3) the impact on patients when research is combined with medical care; and (4) what constitutes fairness in the selection of subjects for research.

The first issue is how best to protect the interests of patients when physicians propose to use medical interventions that are not standard care. Today, when nonstandard interventions are part of a formal research project, the interests of the patient are protected in theory by the institutional review board, which is charged with determining that the risks of the nonstandard intervention are acceptable in light of the available alternatives and the prospect for benefit. Patients are also protected by the requirement of informed consent, which is intended to allow the potential patient-subject to assess whether the balance of risks to potential benefits is acceptable. There is no federally mandated parallel IRB mechanism of review, however, when a medical intervention that is experimental or innovative or even controversial is to be used outside the confines of a research project, although some institutions voluntarily have adopted mechanisms of peer review. The requirement of informed consent remains; the physician is obligated to inform the patient that the proposed intervention is not standard practice, whether it is controversial within the field, and how it compares with alternative approaches, but this requirement provides the patient less protection than would a professional peer review.

At the time of the TBI studies, none of these mechanisms were well developed. During the Cincinnati project, IRBs were in their infancy and the convention of obtaining informed consent from patient-subjects was just emerging. The record is confused and confusing as to whether or when TBI at Cincinnati was viewed as part of a cancer research project and thus properly the subject of IRB review. It is not clear whether the treatment of the Cincinnati patients with TBI was initially intended to be research. In the practice of medicine there has always been a fine boundary between practices or treatments that are accepted as standard, those that are "innovative," and those that are experimental or the subject of research. The use of TBI at Cincinnati is emblematic of the difficulties inherent in sorting through these categories.

By the mid-1960s, TBI without bone marrow protection was a treatment that had been tried and had not been proven effective for patients with radioresistant cancers. By this time, total-body irradiation was not standard treatment for such cases, nor could it be called innovative treatment; some at the time considered its continued use in patients with radioresistant cancers to be controversial. The history of medicine, however, is replete with instances in which failure is followed by success. The continued use of TBI in patients with radioresistant cancers would not have been unethical if the physicians had established clear benchmarks for determining how much additional use was warranted, and if patients had been informed of the speculative nature of the treatment and the gravity of the risks involved. It is not clear that either of these things occurred.

What is clear is that neither the university's IRB nor the funding agency reviewed the appropriateness of continuing to treat patients with radioresistant cancers using TBI without bone marrow protection, despite mounting evidence casting doubt on the utility of TBI treatment for radioresistant tumors in the absence of bone marrow protection. It is also clear that the consent forms did not disclose that it was by this time at best unconventional to treat patients with radioresistant cancers with TBI and that no other medical centers were engaged in this practice at the time; whether physicians told this to their patients is not known. The system of checks and balances that is usually in place today to protect patients’ interests was in its early phase at the University of Cincinnati, and the system did not work well at the time. The responsibility for failure rests at all levels, but it is reasonably clear that patient protection was compromised.

Today, as in the past, there are occasions when nonstandard medical interventions are not subject to human research regulations. In such situations, neither IRB review nor the rigors of scientific design are in place to help determine whether an experimental intervention should continue to be used. Today, for example, many innovations in reproductive technologies and surgery proceed with little oversight and few constraints on the practices of physicians. A physician wishing to use an intervention that other colleagues in the field believe to be ineffective or inferior--as was arguably the case with TBI and radioresistant tumors after several years in the Cincinnati program--will find little standing in his or her way to do so save the fear of malpractice claims and, increasingly, the likelihood that such interventions will not be reimbursed, particularly in managed-care settings. The Cincinnati experience underscores the importance of (1) establishing benchmarks for judging the propriety of continued use and (2) providing for special disclosures to patients in all cases where interventions are not standard--without regard for whether the intervention is deemed "human subject research" or is governed by the Common Rule (see chapter 3).

The question of what role the Department of Defense should have played in reviewing the appropriateness of TBI as medical care for the patient-subjects in its biological dosimetry and radiation-effects research points to the second major issue illustrated by our review of the TBI history. Arguably, the ultimate responsibility for determining that TBI was acceptable medical practice rested with the physicians at Cincinnati and with the university and associated hospitals. At the same time, however, thirty years of government interest in the effects of TBI also arguably had a significant influence on medical practice.

From one vantage, the DOD had little or no obligation to consider the value of TBI to the patients who provided the data it was seeking. The DOD was not paying for the irradiation of the patients. It had reason to assume that the decision about the propriety of the treatment would be made by doctors whose judgment in the matter could be trusted. Yet the TBI experience illustrates that when the government funds research, particularly over a long period, its funding may well have effects beyond the simple conduct of the science and well beyond the confines of the strict terms stated in the contracts or grants authorizing the research.

Over the course of three decades, there was a substantial coincidence between the use of TBI on patients with radioresistant cancers and funding from the Department of Defense and its predecessor. With the exception of work conducted at the City of Hope Hospital, every journal article in the professional literature on the use of TBI with radioresistant tumors during this period was reporting on work supported by the government for military purposes.

In the case of Cincinnati, Dr. Saenger told the Advisory Committee in 1994 that the irradiation of patients might not have been initiated were it not for funding by the DOD and, once initiated, might not have been continued if the objective sought by the DOD (a biological dosimeter) had been realized early on. As Dr. Saenger explained, while the DOD did not directly pay for irradiation, its funding provided for other items---including laboratory equipment and specialists--that facilitated the initiation and maintenance of the TBI program.

Even where the medical care of patients is peripheral to the interests of a funding agency, so long as the research supported by the agency is to be conducted on patient-subjects, it is likely that the research will affect the care patients receive. This is particularly true when agencies support research programs extending over many years, as was the case with the Department of Defense and TBI. Such programs can motivate physician-investigators to alter their practice and can stimulate the adoption of different approaches to the care of patients. Although there is today a greater appreciation of the impact on medical practice of funding patterns in research, it is not clear even now that funding agencies regularly think through the implications for medical care of the research programs they support or that they monitor the impact on patients of their programs over time.

That the joining of research with medical care can alter what happens to a patient is the third issue in research ethics illustrated by the TBI experience. Each purpose introduced into the clinical setting in addition to the treatment of the patient increases the likelihood that the patient will receive more, fewer, or different medical interventions than he or she would otherwise receive. It is naive to think that, either today or thirty years ago, research can be grafted on to the clinical setting without changing the experience for the patient, now turned subject. When the demands of science alter the standard medical practice by increasing the monitoring of physiological indicators, the additional blood tests or bone scans or biopsies are frequently presented as in the interest of patient-subjects. Sometimes this claim is defensible, and the patient-subjects are indeed advantaged by more careful monitoring of their medical condition; at other times, however, this claim is an insupportable rationalization, and there are no offsetting benefits to patients for the risks and discomforts associated with additional monitoring.

In the case of the Cincinnati experiments, the impact of the research protocol on the care of the patient-subjects cannot be construed as beneficial to the patients; in addition, there is evidence of the subordination of the ends of medicine to the ends of research. The decisions to withhold information about possible acute side effects of TBI as well as to forgo pretreatment with antiemetics were irrefutably linked to advancing the research interests of the DOD. To the extent that this deviated from standard care, and caused unnecessary suffering and discomfort, it was morally unconscionable; to the extent that the standard of care in this area is uncertain, it is morally questionable. As troubling as this is, far more troubling is the evidence, including the testimony of the principal investigator, that TBI might not have been employed as treatment for the patients, or once employed continued, in the absence of the government's funding and research requirements.

Whether the ends of research (understood as discovering new knowledge) and the ends of medicine (understood as serving the interests of the patient) necessarily conflict and how the conflict should be resolved when it occurs are still today open and vexing issues. Increasingly, advocates for patients with serious, chronic diseases such as AIDS and breast cancer maintain that it is often in the interests of patients to participate as subjects in clinical research. These advocates are particularly concerned to ensure fair access to participation in research for people who are politically less powerful, such as the poor, minorities, and women. This contemporary perspective upends the traditional way of viewing the fourth issue in research ethics raised by the TBI experiments--fairness in the selection of subjects.

At both M. D. Anderson Hospital and the University of Cincinnati, almost all the patients were drawn from public hospitals, and many were African-Americans. It was common during this period for medical research to be conducted on the poor and the powerless. In part, this practice reflected a general societal insensitivity to questions of justice and equal treatment. In this case, people who were poor disproportionately bore the burdens of questionable research to which their interests as ill people were subordinated. The practice also reflected the view, however, that poor people were better off being patients at hospitals affiliated with research-oriented medical schools where they were likely to become subjects of research (as well as subject matter for clinical teaching). Such institutions, it was thought, offered poor people their best, and perhaps their only, chance to secure quality medical care. Recently, this kind of reasoning has emerged again, as constraints on access to medical care--from the narrowing of entitlement programs to the narrowing of coverage in managed-care medical plans--have made participation in research, as a route to medical care, more attractive. The question of whether the "side benefits" of being a subject should be weighted in the review of the risks and potential benefits of research remains unresolved today.

These findings highlight the contemporary resonance of the TBI story. The issues discussed above are either not now addressed or not addressed adequately by regulation; neither are they covered by clear conventions or rules of professional ethics. Thus, the history of TBI research sponsored by the government is important not only for what it tells us about our past but also for how it illuminates the present.