Phase 1 Clinical Trials in Oncology
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Horstmann et al. (March 3 issue)1 assume that a tumor response is of benefit to subjects in phase 1 oncology trials. This assumption is not valid. A complete or partial tumor response in a phase 1 trial is a surrogate end point, which for most agents has not been linked to a clinically meaningful outcome, such as improved survival.2
Informing subjects that they have a 10.6 percent chance of a tumor response is potentially misleading unless accompanied by an explicit discussion of clinical end points and whether any connection exists between a tumor response and clinical end points.3 This discussion should include an explanation that a tumor response is not a cure or a life extender.
Kurzrock and Benjamin's editorial4 serves only to increase the misrepresentation of phase 1 research.5 It is important to know that phase 1 research is essential for the development of future treatments. But it is simply misleading to treat an improvement in the rate of tumor response as an increase in the likelihood of direct clinical benefit to subjects.
Barbra B. Rothschild, M.D.
Nancy M.P. King, J.D.
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599
rothschild@unc.edu
References
Horstmann E, McCabe MS, Grochow L, et al. Risks and benefits of phase 1 oncology trials, 1991 through 2002. N Engl J Med 2005;352:895-904.
Fleming TR, DeMets DL. Surrogate end points in clinical trials: are we being misled? Ann Intern Med 1996;125:605-613.
King NMP, Henderson GE, Churchill LR, et al. Consent forms and the therapeutic misconception: the example of gene transfer research. IRB 2005;27:1-8.
Kurzrock R, Benjamin RS. Risks and benefits of phase 1 oncology trials, revisited. N Engl J Med 2005;352:930-931.
Dresser R. The ubiquity and utility of the therapeutic misconception. Soc Philos Policy 2002;19:271-294.
To the Editor: The article by Horstmann et al. and the accompanying editorial indicate rates of clinical benefit higher than those reported in previous meta-analyses. Horng et al.,1 in a critique of informed consent in phase 1 oncology trials, decried the frequent lack of an explicit statement that efficacy was not to be expected. However, in addition to evidence presented by Horstmann et al., recent phase 1 trials with established drugs have often resulted in high response rates. Among nine trials involving patients with refractory non–small-cell lung cancer that were presented at the meeting of the American Society of Clinical Oncology in May 2002, the reported response rate was 41 percent (range, 0 to 57 percent) in 150 patients, with one drug-related death recorded. Prior estimates of the risks and benefits of phase 1 oncology trials need updating, and insistence on not conveying therapeutic intent in the informed-consent process in all instances is misplaced.
Franco M. Muggia, M.D.
New York University Cancer Institute
New York, NY 10016
muggif01@med.nyu.edu
References
Horng S, Emanuel EJ, Wilfond B, Rackoff J, Martz K, Grady C. Descriptions of benefits and risks in consent forms for phase 1 oncology trials. N Engl J Med 2002;347:2134-2140.
To the Editor: In their review of 460 phase 1 oncology trials sponsored by the Cancer Therapy Evaluation Program between 1991 and 2002, Horstmann et al. report that the overall toxicity-related death rate was 0.49 percent, which suggests that these trials are relatively safe, considering that virtually all participants have a deadly disease and have exhausted the conventional treatments.1
We analyzed the data from 363 trials of investigational new drugs, involving 12,395 adults with solid tumors, that were published between 1976 and 1993.2 A total of 117 toxicity-related deaths (0.94 percent) and 33 early deaths from unknown causes (0.27 percent) were noted. In addition, 36 trials were excluded from the analysis because further clinical development of the drug was not recommended. We found that toxicity-related death occurred in 26 of 1039 patients in these trials (2.5 percent). Thus, the rate of death due to toxic events varies among phase 1 oncology trials and may be higher than suspected.
Ikuo Sekine, M.D., Ph.D.
Tomohide Tamura, M.D.
National Cancer Center Hospital
Tokyo 104-0045, Japan
isekine@ncc.go.jp
References
Kurzrock R, Benjamin RS. Risks and benefits of phase 1 oncology trials, revisited. N Engl J Med 2005;352:930-932.
Sekine I, Yamamoto N, Kunitoh H, et al. Relationship between objective responses in phase I trials and potential efficacy of non-specific cytotoxic investigational new drugs. Ann Oncol 2002;13:1300-1306.
To the Editor: Kurzrock and Benjamin argue that clinical benefit is an objective of phase 1 cancer trials, citing my article as an instance of an opposing "misconception."1 The misconception is theirs, as is evident in authoritative definitions.2,3 Moreover, in failing to distinguish between what phase 1 trials are specifically designed to measure (dose–toxicity profiles) and what is incidental to the design (e.g., the possibility of benefit), Kurzrock and Benjamin ignore the way in which the strictures of protocol constrain the goals of medicine. This misunderstanding, known as the "therapeutic misconception,"4 reinforces the fiction that clinical research is an extension of clinical care, rather than a fundamentally distinct and sometimes contrary enterprise. Patients in early cohorts in these trials who receive, by design, what Kurzrock and Benjamin call "subtherapeutic" doses are not involved in a trial that aims to maximize their clinical benefit. Failure to see this as a conflict between the objectives of science and those of personal care is the reason the therapeutic misconception has been called "the most important threat to the validity of informed consent to research."5
Matthew J. Miller, M.D., M.P.H.
Harvard School of Public Health
Boston, MA 02115
References
Miller M. Phase 1 cancer trials: a crucible of competing priorities. Int Anesthesiol Clin 2001;139:13-33.
National Cancer Institute dictionary of cancer terms. (Accessed May 19, 2005, at http://www.cancer.gov/dictionary/db_alpha.aspx?expand=p#phase%20I%20trial.)
Eisenhauer EA, O'Dwyer PJ, Christian M, Humphrey JS. Phase I clinical trial design in cancer drug development. J Clin Oncol 2000;18:684-692.
Appelbaum PS, Roth LH, Lidz CW, Benson P, Winslade W. False hopes and best data: consent to research and the therapeutic misconception. Hastings Cent Rep 1987;17:20-24.
Ethical and policy issues in international research: clinical trials in developing countries. Bethesda, Md.: National Bioethics Advisory Commission, 2001:48.
The authors reply: The letters from Drs. Rothschild and King and from Dr. Muggia demonstrate the complexity of understanding "benefits" in the context of phase 1 oncology trials. As Drs. Rothschild and King suggest, tumor response, the most common measure of the effect of agents used for the treatment of cancer, is indeed a surrogate marker. Although tumor response does not necessarily correlate with clinical benefit, it is predictive of potential benefit, and there is evidence that tumor response is associated with symptom relief, improved quality of life, and increased survival.1,2,3,4
We agree that information provided to potential participants in phase 1 trials should be comprehensive, contextual, and clear about the uncertain or inconsistent relationship of possible tumor responses to clinically meaningful benefit.
Furthermore, it should be made clear that although some participants in phase 1 trials may benefit clinically, these trials are designed to evaluate safety, not therapeutic effect. There is a difference between the possibility of benefit from an intervention in a trial and the intent of the researchers when designing the trial. In this regard, we disagree with Dr. Muggia and maintain that consent forms should not describe the purpose or intent of phase 1 trials as therapeutic. Nonetheless, we recognize that although institutional review boards, bioethicists, and others might emphasize the intention of a trial, prospective patients may be more interested in possible benefits than in whether or not the trial is intended to be therapeutic. Our data demonstrate that sometimes there is therapeutic benefit, regardless of the intention of the research.
The statement by Drs. Sekine and Tamura that "the rate of death due to toxic events varies among phase 1 oncology trials" is consistent with the findings of our study. The data they cite emphasize two important realities that should be considered with regard to response or toxicity rates in phase 1 trials: first, different subsets of data have strikingly different benefit and toxicity rates, and second, response and toxicity rates based on published data may be biased. Their data support the view that the details of a trial matter in interpreting the data on response and toxicity. Simply labeling a trial phase 1 is not sufficiently informative about risks and benefits; more specific details about the trial and the intervention are necessary.
Elizabeth Horstmann, B.A.
Ezekiel J. Emanuel, M.D., Ph.D.
Christine Grady, R.N., Ph.D.
National Institutes of Health
Bethesda, MD 20892-1156
cgrady@nih.gov
References
Lokich J. Tumor response and survival end points in clinical trials: a clinician's perspective. Am J Clin Oncol 2004;27:494-496.
Pazdur R. Response rates, survival, and chemotherapy trials. J Natl Cancer Inst 2000;92:1552-1553.
Modi S, Panageas KS, Duck E, et al. Prospective exploratory analysis of the association between tumor response, quality of life, and expenditures among patients receiving paclitaxel monotherapy for refractory metastatic breast cancer. J Clin Oncol 2002;20:3665-3673.
Geels P, Eisenhauer E, Bezjak A, Zee B, Day A. Palliative effect of chemotherapy: objective tumor response is associated with symptom improvement in patients with metastatic breast cancer. J Clin Oncol 2000;18:2395-2405.
The editorialists reply: Rothschild and King's allegation that it is "misleading to treat an improvement in the rate of tumor response as an increase in the likelihood of direct clinical benefit to subjects" is at variance with our clinical experience and the oncology literature. Decades ago, Freireich et al.1 established that improvement in survival in leukemia could be attributed directly to the duration of a response. A response to chemotherapy in randomized trials improved the quality of life despite significant side effects.2 Differences in benefit between patients with and those without a response may be obscured, however, by an inadequate definition of a response. For example, patients with gastrointestinal stromal tumors who were treated with imatinib mesylate and who had stable disease according to the criteria of the Response Evaluation Criteria in Solid Tumors group derived a benefit that was indistinguishable from the benefit in those with a partial response.3 Logic dictates that patients with good performance status and intact organ function — the eligibility criteria for most phase 1 studies — will not die of their cancer unless it progresses.
The perception that, in phase 1 studies, drugs are administered to patients solely to reveal drug toxicity is incorrect, since the objectives of phase 1 trials specifically include describing the response. Oncologists refer patients for phase 1 studies because they determine that participation in those studies offers their patients, whose disease has progressed after recognized therapies, their best chance of benefit. Thus, the primary concern of treating physicians and patients is efficacy. Miller's contention that the scientific restrictions of the protocols interfere with patient care is partially valid. For instance, some patients who might benefit are excluded from phase 1 trials by the eligibility criteria. Low initial doses and small dose increases, resulting from excessive caution about patient safety, can detract from benefit to patients. Nonetheless, as Horstmann et al. have demonstrated, phase 1 studies resulted in stable disease or better in up to 44.7 percent of patients, including those treated at the lower doses.
Increased time before the progression of cancer benefits patients unless the therapy has serious toxic effects. The worse "toxicity" is most often that due to progressive disease. We agree with Muggia, who demonstrates that recent phase 1 trials have higher response rates than previously reported and have extraordinarily low death rates. Although participants in any study should be informed that patients who have a response to therapy may not always benefit, it is misleading to tell patients that there is no clinical benefit from a response and that phase 1 trials have no therapeutic aim.
Razelle Kurzrock, M.D.
Robert S. Benjamin, M.D.
University of Texas M.D. Anderson Cancer Center
Houston, TX 77030
References
Freireich EJ, Gehan EA, Sulman D, Boggs DR, Frei E III. The effect of chemotherapy on acute leukemia in the human. J Chronic Dis 1961;14:593-608.
Coates A, Gebski V, Bishop JF, et al. Improving the quality of life during chemotherapy for advanced breast cancer: a comparison of intermittent and continuous treatment strategies. N Engl J Med 1987;317:1490-1495.
Informing subjects that they have a 10.6 percent chance of a tumor response is potentially misleading unless accompanied by an explicit discussion of clinical end points and whether any connection exists between a tumor response and clinical end points.3 This discussion should include an explanation that a tumor response is not a cure or a life extender.
Kurzrock and Benjamin's editorial4 serves only to increase the misrepresentation of phase 1 research.5 It is important to know that phase 1 research is essential for the development of future treatments. But it is simply misleading to treat an improvement in the rate of tumor response as an increase in the likelihood of direct clinical benefit to subjects.
Barbra B. Rothschild, M.D.
Nancy M.P. King, J.D.
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599
rothschild@unc.edu
References
Horstmann E, McCabe MS, Grochow L, et al. Risks and benefits of phase 1 oncology trials, 1991 through 2002. N Engl J Med 2005;352:895-904.
Fleming TR, DeMets DL. Surrogate end points in clinical trials: are we being misled? Ann Intern Med 1996;125:605-613.
King NMP, Henderson GE, Churchill LR, et al. Consent forms and the therapeutic misconception: the example of gene transfer research. IRB 2005;27:1-8.
Kurzrock R, Benjamin RS. Risks and benefits of phase 1 oncology trials, revisited. N Engl J Med 2005;352:930-931.
Dresser R. The ubiquity and utility of the therapeutic misconception. Soc Philos Policy 2002;19:271-294.
To the Editor: The article by Horstmann et al. and the accompanying editorial indicate rates of clinical benefit higher than those reported in previous meta-analyses. Horng et al.,1 in a critique of informed consent in phase 1 oncology trials, decried the frequent lack of an explicit statement that efficacy was not to be expected. However, in addition to evidence presented by Horstmann et al., recent phase 1 trials with established drugs have often resulted in high response rates. Among nine trials involving patients with refractory non–small-cell lung cancer that were presented at the meeting of the American Society of Clinical Oncology in May 2002, the reported response rate was 41 percent (range, 0 to 57 percent) in 150 patients, with one drug-related death recorded. Prior estimates of the risks and benefits of phase 1 oncology trials need updating, and insistence on not conveying therapeutic intent in the informed-consent process in all instances is misplaced.
Franco M. Muggia, M.D.
New York University Cancer Institute
New York, NY 10016
muggif01@med.nyu.edu
References
Horng S, Emanuel EJ, Wilfond B, Rackoff J, Martz K, Grady C. Descriptions of benefits and risks in consent forms for phase 1 oncology trials. N Engl J Med 2002;347:2134-2140.
To the Editor: In their review of 460 phase 1 oncology trials sponsored by the Cancer Therapy Evaluation Program between 1991 and 2002, Horstmann et al. report that the overall toxicity-related death rate was 0.49 percent, which suggests that these trials are relatively safe, considering that virtually all participants have a deadly disease and have exhausted the conventional treatments.1
We analyzed the data from 363 trials of investigational new drugs, involving 12,395 adults with solid tumors, that were published between 1976 and 1993.2 A total of 117 toxicity-related deaths (0.94 percent) and 33 early deaths from unknown causes (0.27 percent) were noted. In addition, 36 trials were excluded from the analysis because further clinical development of the drug was not recommended. We found that toxicity-related death occurred in 26 of 1039 patients in these trials (2.5 percent). Thus, the rate of death due to toxic events varies among phase 1 oncology trials and may be higher than suspected.
Ikuo Sekine, M.D., Ph.D.
Tomohide Tamura, M.D.
National Cancer Center Hospital
Tokyo 104-0045, Japan
isekine@ncc.go.jp
References
Kurzrock R, Benjamin RS. Risks and benefits of phase 1 oncology trials, revisited. N Engl J Med 2005;352:930-932.
Sekine I, Yamamoto N, Kunitoh H, et al. Relationship between objective responses in phase I trials and potential efficacy of non-specific cytotoxic investigational new drugs. Ann Oncol 2002;13:1300-1306.
To the Editor: Kurzrock and Benjamin argue that clinical benefit is an objective of phase 1 cancer trials, citing my article as an instance of an opposing "misconception."1 The misconception is theirs, as is evident in authoritative definitions.2,3 Moreover, in failing to distinguish between what phase 1 trials are specifically designed to measure (dose–toxicity profiles) and what is incidental to the design (e.g., the possibility of benefit), Kurzrock and Benjamin ignore the way in which the strictures of protocol constrain the goals of medicine. This misunderstanding, known as the "therapeutic misconception,"4 reinforces the fiction that clinical research is an extension of clinical care, rather than a fundamentally distinct and sometimes contrary enterprise. Patients in early cohorts in these trials who receive, by design, what Kurzrock and Benjamin call "subtherapeutic" doses are not involved in a trial that aims to maximize their clinical benefit. Failure to see this as a conflict between the objectives of science and those of personal care is the reason the therapeutic misconception has been called "the most important threat to the validity of informed consent to research."5
Matthew J. Miller, M.D., M.P.H.
Harvard School of Public Health
Boston, MA 02115
References
Miller M. Phase 1 cancer trials: a crucible of competing priorities. Int Anesthesiol Clin 2001;139:13-33.
National Cancer Institute dictionary of cancer terms. (Accessed May 19, 2005, at http://www.cancer.gov/dictionary/db_alpha.aspx?expand=p#phase%20I%20trial.)
Eisenhauer EA, O'Dwyer PJ, Christian M, Humphrey JS. Phase I clinical trial design in cancer drug development. J Clin Oncol 2000;18:684-692.
Appelbaum PS, Roth LH, Lidz CW, Benson P, Winslade W. False hopes and best data: consent to research and the therapeutic misconception. Hastings Cent Rep 1987;17:20-24.
Ethical and policy issues in international research: clinical trials in developing countries. Bethesda, Md.: National Bioethics Advisory Commission, 2001:48.
The authors reply: The letters from Drs. Rothschild and King and from Dr. Muggia demonstrate the complexity of understanding "benefits" in the context of phase 1 oncology trials. As Drs. Rothschild and King suggest, tumor response, the most common measure of the effect of agents used for the treatment of cancer, is indeed a surrogate marker. Although tumor response does not necessarily correlate with clinical benefit, it is predictive of potential benefit, and there is evidence that tumor response is associated with symptom relief, improved quality of life, and increased survival.1,2,3,4
We agree that information provided to potential participants in phase 1 trials should be comprehensive, contextual, and clear about the uncertain or inconsistent relationship of possible tumor responses to clinically meaningful benefit.
Furthermore, it should be made clear that although some participants in phase 1 trials may benefit clinically, these trials are designed to evaluate safety, not therapeutic effect. There is a difference between the possibility of benefit from an intervention in a trial and the intent of the researchers when designing the trial. In this regard, we disagree with Dr. Muggia and maintain that consent forms should not describe the purpose or intent of phase 1 trials as therapeutic. Nonetheless, we recognize that although institutional review boards, bioethicists, and others might emphasize the intention of a trial, prospective patients may be more interested in possible benefits than in whether or not the trial is intended to be therapeutic. Our data demonstrate that sometimes there is therapeutic benefit, regardless of the intention of the research.
The statement by Drs. Sekine and Tamura that "the rate of death due to toxic events varies among phase 1 oncology trials" is consistent with the findings of our study. The data they cite emphasize two important realities that should be considered with regard to response or toxicity rates in phase 1 trials: first, different subsets of data have strikingly different benefit and toxicity rates, and second, response and toxicity rates based on published data may be biased. Their data support the view that the details of a trial matter in interpreting the data on response and toxicity. Simply labeling a trial phase 1 is not sufficiently informative about risks and benefits; more specific details about the trial and the intervention are necessary.
Elizabeth Horstmann, B.A.
Ezekiel J. Emanuel, M.D., Ph.D.
Christine Grady, R.N., Ph.D.
National Institutes of Health
Bethesda, MD 20892-1156
cgrady@nih.gov
References
Lokich J. Tumor response and survival end points in clinical trials: a clinician's perspective. Am J Clin Oncol 2004;27:494-496.
Pazdur R. Response rates, survival, and chemotherapy trials. J Natl Cancer Inst 2000;92:1552-1553.
Modi S, Panageas KS, Duck E, et al. Prospective exploratory analysis of the association between tumor response, quality of life, and expenditures among patients receiving paclitaxel monotherapy for refractory metastatic breast cancer. J Clin Oncol 2002;20:3665-3673.
Geels P, Eisenhauer E, Bezjak A, Zee B, Day A. Palliative effect of chemotherapy: objective tumor response is associated with symptom improvement in patients with metastatic breast cancer. J Clin Oncol 2000;18:2395-2405.
The editorialists reply: Rothschild and King's allegation that it is "misleading to treat an improvement in the rate of tumor response as an increase in the likelihood of direct clinical benefit to subjects" is at variance with our clinical experience and the oncology literature. Decades ago, Freireich et al.1 established that improvement in survival in leukemia could be attributed directly to the duration of a response. A response to chemotherapy in randomized trials improved the quality of life despite significant side effects.2 Differences in benefit between patients with and those without a response may be obscured, however, by an inadequate definition of a response. For example, patients with gastrointestinal stromal tumors who were treated with imatinib mesylate and who had stable disease according to the criteria of the Response Evaluation Criteria in Solid Tumors group derived a benefit that was indistinguishable from the benefit in those with a partial response.3 Logic dictates that patients with good performance status and intact organ function — the eligibility criteria for most phase 1 studies — will not die of their cancer unless it progresses.
The perception that, in phase 1 studies, drugs are administered to patients solely to reveal drug toxicity is incorrect, since the objectives of phase 1 trials specifically include describing the response. Oncologists refer patients for phase 1 studies because they determine that participation in those studies offers their patients, whose disease has progressed after recognized therapies, their best chance of benefit. Thus, the primary concern of treating physicians and patients is efficacy. Miller's contention that the scientific restrictions of the protocols interfere with patient care is partially valid. For instance, some patients who might benefit are excluded from phase 1 trials by the eligibility criteria. Low initial doses and small dose increases, resulting from excessive caution about patient safety, can detract from benefit to patients. Nonetheless, as Horstmann et al. have demonstrated, phase 1 studies resulted in stable disease or better in up to 44.7 percent of patients, including those treated at the lower doses.
Increased time before the progression of cancer benefits patients unless the therapy has serious toxic effects. The worse "toxicity" is most often that due to progressive disease. We agree with Muggia, who demonstrates that recent phase 1 trials have higher response rates than previously reported and have extraordinarily low death rates. Although participants in any study should be informed that patients who have a response to therapy may not always benefit, it is misleading to tell patients that there is no clinical benefit from a response and that phase 1 trials have no therapeutic aim.
Razelle Kurzrock, M.D.
Robert S. Benjamin, M.D.
University of Texas M.D. Anderson Cancer Center
Houston, TX 77030
References
Freireich EJ, Gehan EA, Sulman D, Boggs DR, Frei E III. The effect of chemotherapy on acute leukemia in the human. J Chronic Dis 1961;14:593-608.
Coates A, Gebski V, Bishop JF, et al. Improving the quality of life during chemotherapy for advanced breast cancer: a comparison of intermittent and continuous treatment strategies. N Engl J Med 1987;317:1490-1495.