|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 19 (July 1), 2006: pp. 2987-2990 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.04.9296
Rethinking Risk-Benefit Assessment for Phase I Cancer Trials1 Department of Pediatric Oncology, Dana-Farber Cancer Institute; and Department of Medicine, Children's Hospital, Boston, MA
2 Department of Clinical Bioethics, National Institutes of Health, Bethesda, MD Although vital to developing new cancer treatments, phase I trials in oncology remain ethically controversial. Classically, they involve the initial administration of new antineoplastic agents to patient-participants, based on apparent safety and potential for efficacy in in vitro and animal studies. Participants, though not imminently dying, generally have advanced cancer, have exhausted established anticancer treatments, and are ineligible for later-phase trials.1 The primary objectives of these studies are to clarify the appropriate dose of the agent for subsequent trials, to define its preliminary toxicity profile, and to describe its pharmacokinetics.2 Although phase I trials typically seek evidence of tumor response, a surrogate marker for efficacy, they are not designed to draw conclusions about clinical benefit.3 Ethical controversy regarding phase I trials revolves around the nature of their scientific objectives, the likelihood that participants will experience direct benefits, and the validity of consent from terminally ill patients.4-7 Critics charge that in view of their emphasis on toxicity and dose finding, they "have no therapeutic content"8 and offer participants no reasonable prospect of benefit.9 Others cite evidence that participants commonly misconstrue the purpose of and overestimate the benefits from phase I trials, thus raising doubts about the validity of consent.10-12 Defenders counter that, given the small chance for benefit, phase I trials "should be considered therapeutic research intervention[s],"13 and decry the misconception that "such trials are nontherapeutic toxicology studies."7 They also reject the challenge to informed consent, claiming instead that patients' difficult circumstances do not imply a "lack of capacity to give informed consent" and that "empirical data ... do not support the notion that consent is uninformed."3 This debate has advanced little in the last 25 years.14 In this article, we address the ethical controversy regarding risk-benefit assessment in phase I trials. Progress in resolving this controversy rests on dispelling two fundamental misconceptions. First, both critics and defenders rely on a mistaken distinction between therapeutic and nontherapeutic research.15 Second, they specify erroneous benchmarks, or no benchmark at all, when judging the risks and benefits of these trials for participants. Recognition of these misconceptions can help break the impasse in the long-running debate about the ethics of phase I trials. Existing ethical commentary relies primarily on meta-analyses of phase I trials, especially evaluations of cytotoxic agents, conducted in the 1960s through 1980s.16-18 Several recent studies provide an updated picture of the nature, risks, and direct benefits of these trials.19,20 Thus, reconsideration of the conceptual framework for assessing the risks and benefits of these trials is timely. Phase I trials are heterogeneous.19 Many newer trials evaluate targeted or biologic agents rather than cytotoxic drugs; others involve combinations of previously approved drugs. We confine our attention to "classical" trials evaluating single agents, because debate has centered on this paradigm. However, because the issues we raise are generally relevant to dose-finding trials of new interventions, our argument applies equally to studies of cytotoxic, targeted, and biologic agents commonly evaluated in contemporary phase I trials. ARE PHASE I TRIALS THERAPEUTIC OR NONTHERAPEUTIC RESEARCH? The original Declaration of Helsinki proposed "a fundamental distinction ... between medical research in which the aim is essentially diagnostic or therapeutic for a patient, and medical research the essential object of which is purely scientific ... ."21 As Levine demonstrated many years ago, this taxonomy of research, though still influential, is a flawed account of both the intent and the execution of clinical trials.15 The aim of clinical research is to develop generalizable knowledge by means of scientific investigation involving groups of participants; the aim of medical care is to benefit particular patients.22 It is often reasonable, of course, for an investigator to offer a trial to a patient with the hope or expectation that she will gain from participation; trials may have therapeutic benefits despite their primary scientific intent. Nevertheless, even in the context of a trial that offers a substantial prospect of direct benefit, researchers typically employ procedures or methodological devices to serve the study's scientific aims that would be inappropriate if used in the setting of ordinary medical care. Examples include random assignment, placebo controls (sometimes despite availability of proven effective therapy), double-blinding, and the performance of invasive procedures that offer no benefit to the participant but are necessary to measure study outcomes.23 Thus, because all human-subjects research pursues scientific objectives and involves procedures performed for methodologic rather than benefit-related reasons, dichotomizing research as therapeutic or nontherapeutic is misguided. Why, then, has the phase I debate clung so insistently to the therapeutic/nontherapeutic framework? A review of the literature suggests two reasons. First, meta-analyses demonstrate that the likelihood of direct benefit, at least from classical phase I trials, is low.16-20,24,25 Recognition of this limited benefit engenders legitimate discussion about whether the risk-benefit ratio for participants is acceptable. Second—here is where the logic falters—critics reason that because phase I trials are primarily dose-finding and safety studies, they are necessarily nontherapeutic and therefore do not offer participants a prospect of direct benefit. For example, Oberman and Frader contend that "phase I studies are explicitly nontherapeutic in nature" because they are "intended to reveal the pharmacologic action of the study agent" and seek "to establish a baseline regarding toxicity."26 Similarly, Sankar asserts that "While lack of benefit is the logical outcome of a phase I trial design, its absence remains controversial."27 These criticisms are mistaken. The purpose of a study does not determine whether study interventions offer participants a prospect of direct medical benefit, nor how potential benefits compare with risks. Some safety studies, including phase I trials, offer participants a prospect of direct benefit. In contrast, some efficacy studies, such as placebo-controlled trials, do not offer a prospect of direct benefit, at least for a subset of participants.7 Moreover, attributing a therapeutic or nontherapeutic intent to clinical trials represents a category mistake. Whether or not an activity is therapeutic has to do with its purpose (ie, devoted to personalized care or to a set of protocol-specified scientific objectives) and whether it has a prospect of benefit has to do with its anticipated consequences. Thus, to deny that a phase I trial—or any clinical research study—is a therapeutic activity leaves an open question of whether it offers a prospect of benefit. Bioethicists who rightly see that phase I trials are not therapeutic draw the mistaken inference that they offer no prospect of benefit. Conversely, clinician-investigators who recognize that they offer some, albeit small, prospect of benefit draw the mistaken inference that they must be "therapeutic research." Distinguishing the purpose of phase I oncology trials from their anticipated consequences is a necessary prerequisite to accurate risk-benefit assessment by investigators, institutional review boards, bioethicists, and prospective participants alike. DO PHASE I TRIALS OFFER INDIVIDUALS WITH CANCER A REASONABLE RISK-BENEFIT RATIO? All ethical codes require careful assessment of the risks and benefits, both to participants and to society, of each proposed research project.28,29 It is important to recognize that, like other types of research, phase I trials need not offer participants a favorable direct risk-benefit ratio to be ethical. Rather, the benefits to participants together with the value of the knowledge to be gained must justify the risks.22,23,28,30 Some individuals might choose to participate on altruistic grounds, even if the risk-benefit ratio to themselves were unfavorable.11 Nevertheless, it is appropriate to ask how the risks and direct benefits of study participation compare with those of appropriate nonresearch alternatives. For this reason, and because the literature on the ethics of phase I trials emphasizes risks and benefits for participants themselves, we adopt this perspective here. To judge the risks and benefits of phase I trials for participants, one must assess them against those of some relevant alternative.3 As a rule, neither defenders nor critics specify their comparator when assessing risks and benefits, making their arguments incomplete and difficult to evaluate.31 Where they define a benchmark, commentators generally assess phase I trials against either palliation alone or U.S. Food and Drug Administration (FDA) –approved cancer treatment. Neither comparison is apt. Ackerman contends that "the risk-benefit ratio of participating in a phase I trial must...be compared to the advantages and disadvantages of...palliative care only."13 He argues further that, compared with palliation alone, the small chance of direct benefit from these trials justifies their incremental risks and burdens. Critics conclude instead that the benefits to participants fail to justify the risks.8,9 That observers would come to such different conclusions when considering the same data is unsurprising, as the choice between palliation alone and continued anticancer interventions depends at least as much on value considerations as it does on empirical data about risks and benefits. To be sure, investigators must encourage cancer patients contemplating phase I trial participation to weigh the option of palliation alone, and must clearly describe the limited probability and magnitude of direct benefit that most phase I trials offer. However, comparing or contrasting these options misrepresents the immediate choice that such patients confront. Well-informed patients who elect to enroll in phase I trials have rejected palliation alone in favor of continued anticancer interventions. In light of this choice, it is important to ask how the balance of risks and benefits that phase I trials offer participants compares with that of alternative anticancer treatments. Approved cancer treatments constitute a second possible benchmark for assessing the risks and direct benefits of phase I trials. By definition, patients eligible for classical phase I trials do not have approved cancer treatments available to them. Nevertheless, evidence that the direct benefits of phase I trials approximate those of approved agents would support the conclusion that the risk-benefit ratio for participants is favorable. On the basis of this reasoning, Agrawal and Emanuel suggest that the risk-benefit ratio of phase I trials "is not clearly worse" than that of some FDA-approved treatments.3 However, the examples they cite, including interleukin-2 for metastatic renal cell carcinoma (response rate, 14%; median response duration, > 19 months), topotecan for ovarian cancer (response rate, 16%; median duration, 21.7 weeks), and gemcitabine for metastatic pancreatic cancer (response rate, 5.4% [but with demonstrated quality-of-life and survival benefits in a randomized comparison]) fail to support their contention.32-34 The response rates in the studies they cite are generally higher than those reported in recent meta-analyses of single-agent phase I trials (3.8% to 4.8%), or are accompanied by evidence of efficacy as measured by more clinically meaningful end points such as survival or quality of life.19,20 Finally, approximately 90% of agents studied in phase I trials fail to gain FDA approval, often as a result of insufficient efficacy.20,35 Thus, the assertion that phase I trials offer participants benefits commensurate with those of some approved agents is doubtful. Nevertheless, even if phase I trial participation offers a lesser prospect of benefit than even the most marginal FDA-approved agents, these trials could still be ethical when judged according to standards appropriate to clinical research. Rather than using palliation alone or FDA-approved treatments as benchmarks, it is most appropriate to compare the risks and direct benefits of phase I trials with those of off-label therapy outside a trial. Physicians might offer off-label therapy on the basis of weak evidence such as case series, uncontrolled small studies, or preliminary reports. Use of off-label cancer therapy is common, particularly among patients whose cancers have progressed or recurred despite their receiving accepted therapies, and represents the principal alternative that many phase I trial participants have available after rejecting palliation alone.7,36,37 Although the toxicities of such nonvalidated treatments are generally well understood, their benefits in this context are uncertain. Furthermore, there are no systematic data that permit comparison between the benefits or risks for participants of phase I trials and those of nonvalidated treatments outside a trial. If such data were to demonstrate that, on average, phase I trial participation had an unfavorable risk-benefit ratio for participants when compared with nonvalidated treatments, then only an autonomous decision by participants to forgo the advantages of nonvalidated treatments in favor of contributing altruistically to improved treatment for others could justify trial entry. In contrast, if phase I trial participation were shown to have, on average, an equally or more favorable risk-benefit ratio than that of nonvalidated treatments, then trial entry would be justified on self-interest grounds alone. Collection of such data, which would inform the risk-benefit calculus for phase I trials, should therefore be a high priority. In the interim, it is difficult to argue convincingly that the risks and direct benefits of phase I trials are unfavorable when compared with those of the relevant alternatives, and therefore to conclude that trial entry should not be available to eligible individuals. Some authors claim that the trend among phase I trials to evaluate targeted and biologic agents, rather than cytotoxic drugs, alters the risk-benefit assessment of these trials.20 First, the toxic death rate in trials of newer agents is lower than that in trials of cytotoxic drugs, whereas response rates appear comparable.19,20 Second, phase I trials of newer agents frequently seek biologic evidence of anticancer activity, using surrogate markers such as inhibition of the target receptor in tumor tissue. Third, such trials may seek to define the optimum biologic dose rather than the maximum-tolerated dose, and therefore may not require escalation to dose-limiting toxicity in order to meet their scientific objectives.38 Notwithstanding these important methodologic distinctions, trials of newer agents are generally analogous to trials of cytotoxic drugs with respect to the nature of the study population, the available alternatives, the uncertainty of risk, the limited likelihood of direct benefit, and the essential objective of defining an appropriate dose for use in subsequent research.19,20 Accordingly, the ethical analysis appropriate to trials of newer agents does not differ fundamentally from that appropriate to studies involving cytotoxic agents. OPTIMIZING THE RISK-BENEFIT RATIO OF PHASE I TRIALS Notwithstanding the erroneous ethical description of phase I oncology trials as therapeutic or nontherapeutic research, and the mistaken use of palliation alone or FDA-approved treatments as benchmarks for risk-benefit assessment, these trials raise challenging ethical issues. In particular, phase I investigators (like all investigators) have obligations, within the constraints imposed by the study's objectives, to maximize benefits and minimize risks for participants.28,30 Crucially, these obligations do not derive from their clinical duties as physicians; rather, they are inherent in their roles as researchers. An investigator or institutional review board that neglects to maximize benefits or to minimize risks in designing or approving a study, where doing so would preserve the trial's ability to achieve its objectives, is liable to the charge of unethical research. It is particularly difficult for investigators designing phase I trials to satisfy their obligations to maximize benefits and minimize risks.30 The central ethical dilemma of phase I trials derives from the fact that the optimal dose of a new agent or combination is unknown. Consequently, efforts to enhance benefits and to reduce risks, both for individual participants and for the group of participants as a whole, are intertwined. Tradeoffs between these objectives are unavoidable.39 In the standard phase I design, investigators typically administer a low dose of the agent to an initial cohort of participants. Successive cohorts receive escalating doses until a substantial proportion of subjects experiences dose-limiting toxicities. Because most responses occur between 80% and 120% of the maximum-tolerated dose,40 enhancing benefits generally requires decreasing the proportion of participants exposed to doses below this range.4 Possible strategies include smaller initial cohorts, higher starting doses, more rapid dose-escalation schemes, and dose-escalation within as well as among participants.41,42 These innovations can also decrease the time, resources, and number of participants needed to complete phase I trials, albeit while increasing the risk of serious study-related toxicities.41 Data suggest that participants might be willing to accept an increased risk of toxicity in return for a greater chance of direct benefit.43 At present, such innovative designs are uncommon, despite evidence that they offer participants a higher likelihood of clinical response.20,44 Reasons for this conservatism may include reluctance to part with tradition, emphasis on minimizing risk rather than maximizing benefit, or concerns that intraparticipant dose escalation may make distinguishing acute from cumulative toxicities difficult.39,42 Nevertheless, in designing and approving phase I trials, investigators and IRBs must strive to optimize the risk-benefit ratio for participants, while remaining within the constraints imposed by the trials' scientific objectives. CONCLUSION Risk-benefit analysis of phase I cancer trials requires application of a conceptual model appropriate to clinical research. Further, it requires specification of an appropriate comparator for assessing risks and benefits to participants. The literature on the ethics of these trials has failed to satisfy these requirements. Recognition that these trials are neither more nor less therapeutic than other phases of trials, and that the available data are consistent with the conclusion that their risk-benefit ratio is favorable when compared with that of the relevant alternatives, alleviates several of the major ethical worries associated with these trials. Legitimate concerns about informed consent to participate in phase I oncology trials remain.11,12 These need to be addressed by further empirical research on the quality of patient-participants' understanding, on the voluntariness of their choices, and on optimal methods for seeking informed consent in this setting. Authors' Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. Author Contributions
ACKNOWLEDGMENTS We thank Christine Grady, RN, PhD, Ezekiel Emanuel, MD, PhD, Dan Brock, PhD, and Robert Truog, MD for their thoughtful comments on earlier versions of this manuscript. Supported by a Research Scientist Development Award (CA096872) from the National Cancer Institute (National Institutes of Health, Bethesda, MD; S.J.). The funding source played no role in the conception or articulation of the ideas contained herein, nor in the decision to publish. The opinions expressed are those of the authors and do not necessarily reflect the position or policy of the National Institutes of Health, the Public Health Service, or the Department of Health and Human Services. REFERENCES 1. Freireich EJ: Ethical considerations in cancer chemotherapy. Annu Rev Pharmacol Toxicol 19:547-557, 1979[Medline] 2. Pratt CB: The conduct of phase I-II clinical trials in children with cancer. Med Pediatr Oncol 19:304-309, 1991[Medline] 3. Agrawal M, Emanuel EJ: Ethics of phase 1 oncology studies: Reexamining the arguments and data. JAMA 290:1075-1082, 2003 4. Brody BA: The Ethics of Biomedical Research: An International Perspective. New York, NY, Oxford University Press, 1998 5. Kodish E: Pediatric ethics and early-phase childhood cancer research: Conflicted goals and the prospect of benefit. Account Res 10:17-25, 2003[Medline] 6. Ratain MJ, Mick R, Schilsky RL, et al: Statistical and ethical issues in the design and conduct of phase I and II clinical trials of new anticancer agents. J Natl Cancer Inst 85:1637-1643, 1993 7. Critical role of phase I clinical trials in cancer treatment: American Society of Clinical Oncology. J Clin Oncol 15:853-859, 1997 8. Annas GJ: Questing for grails: Duplicity, betrayal and self-deception in postmodern medical research. J Contemp Health Law Policy 12:297-324, 1996[Medline] 9. King NM: Defining and describing benefit appropriately in clinical trials. J Law Med Ethics 28:332-343, 2000[Medline] 10. Daugherty C, Ratain MJ, Grochowski E, et al: Perceptions of cancer patients and their physicians involved in phase I trials. J Clin Oncol 13:1062-1072, 1995[Abstract] 11. Daugherty CK, Banik DM, Janish L, et al: Quantitative analysis of ethical issues in phase I trials: A survey interview of 144 advanced cancer patients. IRB 22:6-14, 2000[Medline] 12. Weinfurt KP, Castel LD, Li Y, et al: The correlation between patient characteristics and expectations of benefit from phase I clinical trials. Cancer 98:166-175, 2003[CrossRef][Medline] 13. Ackerman TF: The ethics of phase I pediatric oncology trials. IRB 17:1-5, 1995[Medline] 14. Lipsett MB: On the nature and ethics of phase I clinical trials of cancer chemotherapies. JAMA 248:941-942, 1982[CrossRef][Medline] 15. Levine RJ: Ethics and Regulation of Clinical Research (ed 2). New Haven, CT, Yale University Press, 1986 16. Decoster G, Stein G, Holdener EE: Responses and toxic deaths in phase I clinical trials. Ann Oncol 1:175-181, 1990 17. Estey E, Hoth D, Simon R, et al: Therapeutic response in phase I trials of antineoplastic agents. Cancer Treat Rep 70:1105-1115, 1986[Medline] 18. Furman W, Pratt C, Rivera G: Mortality in pediatric phase I clinical trials. J Natl Cancer Inst 81:1193-1194, 1989 19. Horstmann E, McCabe MS, Grochow L, et al: Risks and benefits of phase 1 oncology trials, 1991 through 2002. N Engl J Med 352:895-904, 2005 20. Roberts TG Jr, Goulart BH, Squitieri L, et al: Trends in the risks and benefits to patients with cancer participating in phase 1 clinical trials. JAMA 292:2130-2140, 2004 21. Rickham PP: Human experimentation: Code of Ethics of the World Medical Association—Declaration of Helsinki. BMJ 5402:177, 1964 22. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research: The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Biomedical and Behavioral Research. Washington, DC, US Government Printing Office, 1979 23. Miller FG, Brody H: A critique of clinical equipoise: Therapeutic misconception in the ethics of clinical trials. Hastings Cent Rep 33:19-28, 2003[Medline] 24. Shah S, Weitman S, Langevin AM, et al: Phase I therapy trials in children with cancer. J Pediatr Hematol Oncol 20:431-438, 1998[CrossRef][Medline] 25. Smith TL, Lee JJ, Kantarjian HM, et al: Design and results of phase I cancer clinical trials: Three-year experience at M.D. Anderson Cancer Center. J Clin Oncol 14:287-295, 1996[Abstract] 26. Oberman M, Frader J: Dying children and medical research: Access to clinical trials as benefit and burden. Am J Law Med 29:301-317, 2003[Medline] 27. Sankar P: Communication and miscommunication in informed consent to research. Med Anthropol Q 18:429-446, 2004[Medline] 28. Emanuel EJ, Wendler D, Grady C: What makes clinical research ethical? JAMA 283:2701-2711, 2000 29. World Medical Association: Declaration of Helsinki, 5th revision. http://www.wma.net/e/policy/b3.htm 30. U.S. Department of Health and Human Services: 45 CFR 46. Federal Register 56:28012, 1991 31. Kurzrock R, Benjamin RS: Risks and benefits of phase 1 oncology trials, revisited. N Engl J Med 352:930-932, 2005 32. Fyfe G, Fisher RI, Rosenberg SA, et al: Results of treatment of 255 patients with metastatic renal cell carcinoma who received high-dose recombinant interleukin-2 therapy. J Clin Oncol 13:688-696, 1995 33. Creemers GJ, Bolis G, Gore M, et al: Topotecan, an active drug in the second-line treatment of epithelial ovarian cancer: Results of a large European phase II study. J Clin Oncol 14:3056-3061, 1996[Abstract] 34. Burris HA III, Moore MJ, Andersen J, et al: Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: A randomized trial. J Clin Oncol 15:2403-2413, 1997 35. Von Hoff DD: There are no bad anticancer agents, only bad clinical trial designs: Twenty-first Richard and Hinda Rosenthal Foundation Award lecture. Clin Cancer Res 4:1079-1086, 1998[Abstract] 36. Laetz T, Silberman G: Reimbursement policies constrain the practice of oncology. JAMA 266:2996-2999, 1991[Abstract] 37. Mack JW, Hilden JM, Watterson J, et al: Parent and physician perspectives on quality of care at the end of life in children with cancer. J Clin Oncol 23:9155-9161, 2005 38. Anderson BD, Adamson PC, Weiner SL, et al: Tissue collection for correlative studies in childhood cancer clinical trials: Ethical considerations and special imperatives. J Clin Oncol 22:4846-4850, 2004 39. Miller M: Phase I cancer trials: A collusion of misunderstanding. Hastings Cent Rep 30:34-43, 2000[Medline] 40. Von Hoff DD, Turner J: Response rates, duration of response, and dose response effects in phase I studies of antineoplastics. Invest New Drugs 9:115-122, 1991[Medline] 41. Simon R, Freidlin B, Rubinstein L, et al: Accelerated titration designs for phase I clinical trials in oncology. J Natl Cancer Inst 89:1138-1147, 1997 42. Arbuck SG: Workshop on phase I study design. Ann Oncol 7:567-573, 1996 43. Daugherty CK, Ratain MJ, Minami H, et al: Study of cohort-specific consent and patient control in phase I cancer trials. J Clin Oncol 16:2305-2312, 1998[Abstract] 44. Dent SF, Eisenhauer EA: Phase I trial design: Are new methodologies being put into practice? Ann Oncol 7:561-566, 1996 Related Correspondence
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
