|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 12 (April 20), 2008: pp. 1922-1923 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.14.8064
Assessing the Measure of a New Drug: Is Survival the Only Thing That Matters?Division of Biostatistics, Mayo Clinic Cancer Center, Mayo Clinic, Rochester, MN
Breast Oncology Program, University of Michigan Comprehensive Cancer Center, Ann Arbor, MI Clinical research to improve breast cancer treatment has produced multiple, meaningful advances in the last decade. New antiestrogen and chemotherapeutic agents—novel therapies directed towards HER2—such as trastuzumab and lapatinib, and antiangiogenic agents, such as bevacizumab, have all been added to the clinician's armamentarium to treat metastatic disease.1 Several of these advances have now demonstrated efficacy in the adjuvant setting, which raises the important question: What exactly has improved as these agents have been introduced into the metastatic setting? Given the current limited possibility of cure for these patients, there are two clinically important goals of treatment: make the patient live longer and delay or relieve symptoms. The first of these end points, codified as overall survival (OS), is simple to measure, unambiguous, and of unquestionable clinical relevance. However, symptom relief and palliation (or quality of life), despite recent progress, is notoriously difficult to measure.2 Ideally, a newly introduced antineoplastic agent will demonstrate a significant and clinically meaningful prolongation of overall survival in the setting of a randomized controlled trial. However, the use of overall survival as the primary end point for a clinical trial has several major drawbacks. First, because mortality occurs after a relatively long time for most patients, statistically significant differences in OS require large numbers of patients and several years to reliably detect. For example, in a recently reported trial in which women with metastatic breast cancer were randomly assigned to receive either bevacizumab or no bevacizumab, in addition to first-line paclitaxel, the median survival for the trial population was not reached until nearly 8 years after the trial was opened, a full 2 years after initial results (based on an end point of progression-free survival [PFS]) were made public.3 Second, for many new agents, patients assigned to the control arm are either allowed to cross over to the investigational agent or to receive the investigational agent off-study on disease progression. This strategy dilutes effects of an agent on OS, since patients in both arms will receive the new agent at some point. Although a prohibition of such cross-over is scientifically appealing, ethical considerations often preclude it, particularly if outcomes on earlier end points are found to be promising or the drug is approved in later line settings. Finally, for addition of a new agent, a prolongation of OS may be more difficult to detect now than in the past due to the beneficial effects of the many other antineoplastic therapies now available for breast cancer.1 Indeed, over the last two decades, OS has increased substantially for patients with metastatic breast cancer. For example, in retrospective analysis of patients with metastatic breast cancer treated in British Columbia, OS improved from approximately 14.5 months for those diagnosed in 1991 to approximately 22 months for women diagnosed in 2001.4 As median survival continues to increase, the sample size necessary to conclusively identify the same absolute improvement in survival (for example, 3 months) increases considerably, since the same 3-month absolute improvement results in a hazard ratio closer to 1.0 as the control arm survival increases. While here we have focused on metastatic breast cancer, these considerations are also true for metastatic colon cancer, and likely for additional disease sites in the future. Concurrent with these considerable challenges to OS as an end point in metastatic breast (and other) cancers, the pressure from payers, regulators, patients, and the public is ever increasing to develop new effective treatments more quickly and cost-effectively. Based on these considerations, we must constantly seek innovative strategies to test new therapies. Multiple such novel approaches are gaining momentum in the clinical trials community, including selectively limiting enrollment of patients to trials based on biomarkers,5,6 the expanded use of randomized phase II trials as precursors to phase III trials,7 and the use of adaptive, data-driven random assignment probabilities in phase III trials.8 In this context of innovative clinical research methods, surrogate end points offer a further approach to improve the speed with which clinical trials may deliver definitive results. A surrogate end point is a measure that is obtained more quickly, less invasively, or at lower cost than the true clinical end point of interest.9 The key characteristic that allows the use of a surrogate end point is that the inference from the trial (that is, the trial's conclusion regarding efficacy or lack thereof of the experimental agent under investigation) will be the same, regardless of whether the true or the surrogate end point is measured. Multiple possible pitfalls are present in any attempts to validate surrogate markers10; however, several key principles have begun to be widely adopted. These principles11 state that validation should be conducted via a meta-analytic approach, analyses based on individual patient data are highly preferred, and the validation meta-analyses must include trials demonstrating a difference or not in outcomes among treatments. In the setting of first-line metastatic breast cancer, and in the context of improved survival outcomes, additional therapeutic options in second- and third-line treatment, increasing pressure for timely results, and maturing methodology for surrogate end point validation, Burzykowski et al12 present a pooled analysis based on individual patient data. It explores the potential value of multiple possible end points—most promisingly, PFS—in advanced breast cancer trials addressing the relative worth of taxane-based therapy. They conclude that none of the possible measures explored—tumor response, disease control, PFS, or time to progression—is a suitably reliable end point to predict OS in advanced breast cancer. We believe that this conclusion is well justified. Their approach is methodologically strong and clearly meets two of the key principles necessary for validation of a surrogate marker. However, the meta-analysis only marginally included trials with heterogeneity of treatment effect, and cross-over to a second-line taxane was allowed in several trials, potentially diluting an OS observation. Regardless, the findings of Burzykowski et al are similar to those of Bruzzi et al,13 who explored response rate as a surrogate end point for survival by analyzing individual patient data from trials containing epirubicin or not. Taken together, these two analyses indicate that there are no presently available early end points in advanced breast cancer to predict for an improvement in OS. Are there novel end points that might be better surrogates of OS than response or progression? One approach would be to explore older trials, in which second-line and later therapies have not obscured a possible surrogate relationship between PFS and OS, as done recently by Buyse et al14 in the setting of metastatic colon cancer. However, these studies are difficult to find in metastatic breast cancer, and their power will be limited due to the generally smaller sample sizes of older studies. Another approach is to search for novel surrogate end points, such as innovative imaging techniques or new biomarkers. For example, a recent report has suggested that circulating tumor cell levels are a more powerful predictor of survival in patients starting a new therapy for metastatic disease than either tumor response or time to disease progression.15 Do the findings by Burzykowski et al12 imply that PFS is inappropriate as an end point in clinical trials in first-line metastatic breast cancer? In our opinion, they do not. We propose that the application of numerous second-, third-, and later line therapies, including cross-over, is the likely reason that surrogacy for OS could not be demonstrated by Burzykowski et al. Insisting that OS is the only acceptable end point in advanced disease trials would require trials that are double the size of current investigations, and may likely result in the movement of clinical research to settings in which second-line treatment is limited, outside of the United States and Western Europe. In metastatic colorectal cancer, the US Food and Drug Administration's Oncologic Drug Advisory Committee has accepted PFS as an end point indicative of clinical benefit, regardless of its surrogacy with OS. We suggest that similar considerations are likely relevant in metastatic breast cancer. Trials must undoubtedly continue to assess OS, as it remains the most desired goal of most patients with metastatic cancer. However, given the biologic plausibility of PFS as a measure of efficacy and the pace of development of promising new drugs, and assuming that a secondary important goal of advanced disease trials is to improve both quality and quantity of life, we believe that PFS is appropriate as a primary end point regardless of formal surrogacy for OS. AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a "U" are those for which no compensation was received; those relationships marked with a "C" were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: Daniel J. Sargent, Genentech (C) Stock Ownership: None Honoraria: Daniel J Sargent, Genentech Research Funding: None Expert Testimony: None Other Remuneration: None AUTHOR CONTRIBUTIONS Conception and design: Daniel J. Sargent, Daniel F. Hayes Financial support: Daniel J. Sargent, Daniel F. Hayes Administrative support: Daniel J. Sargent Provision of study materials or patients: Daniel J. Sargent Collection and assembly of data: Daniel J. Sargent Data analysis and interpretation: Daniel J. Sargent Manuscript writing: Daniel J. Sargent, Daniel F. Hayes Final approval of manuscript: Daniel J. Sargent, Daniel F. Hayes REFERENCES 1. Ellis M, Hayes DF, Lippman ME: Treatment of metastatic breast cancer, in Harris J, Lippman M, Morrow M, et al (eds): Diseases of the Breast (ed 3). Philadelphia, PA, Lippincott Williams & Wilkins, 2004, pp 1101-1162 2. Sloan JA, Dueck A, Frost MH, et al: Applying QOL assessments: Solutions for oncology clinical practice and research, Part 2. Curr Probl Cancer 30: 233-332, 2006[CrossRef] 3. Miller K, Wang M, Gralow J, et al: E2100: A randomized phase III trial of paclitaxel versus paclitaxel plus bevacizumab as first-line therapy for locally recurrent or metastatic breast cancer. Proceedings of the 41st American Society of Clinical Oncology Annual Meeting, Orlando, FL, May 13-17, 2005a 4. Chia SK, Speers CH, D’yachkova Y, et al: The impact of new chemotherapeutic and hormone agents on survival in a population-based cohort of women with metastatic breast cancer. Cancer 110: 973-979, 2007[CrossRef][Medline] 5. Romond EH, Perez EA, Bryant J, et al: Trastuzumab plus adjuvant chemotherapy for operable HER2-positive breast cancer. N Engl J Med 353: 1673-1684, 2005 6. Simon R, Maitournam A: Evaluating the efficiency of targeted designs for randomized clinical trials. Clin Cancer Res 10: 6759-6763, 2004 7. Rubinstein RV, Korn EL, Freidlin B, et al: Design issues of randomized phase II trials and a proposal for phase II screening trials. J Clin Oncol 23: 7199-7206, 2005 8. Berry DA: Bayesian statistics and the efficiency and ethics of clinical trials. Stat Sci 19:175-187, 2004[CrossRef] 9. Prentice RL: Surrogate endpoints in clinical trials: Definition and operational criteria. Stat Med 8:431-440, 1989[Medline] 10. Fleming TR, DeMets DL: Surrogate end points in clinical trials: Are we being misled? Ann Intern Med 125:605-613, 1996 11. Burzykowski T, Molenberghs G, Buyse M (eds): The Evaluation of Surrogate Endpoints. New York, NY, Springer, 2005 12. Burzykowski T, Buyse M, Piccart-Gebhart MJ, et al: Evaluation of tumor response, disease control, progression-free survival, and time to progression as potential surrogate endpoints in metastatic breast cancer. J Clin Oncol 26:1987-1992, 2008 13. Bruzzi P, Del Mastro L, Sormani MP, et al: Objective Response to Chemotherapy As a Potential Surrogate End Point of Survival in Metastatic Breast Cancer Patients. J Clin Oncol 23:5117-5125, 2005 14. Buyse M, Burzykowski T, Carroll K, et al: Progression-free survival as a surrogate for survival in advanced colorectal cancer. J Clin Oncol 25:5218-5224, 2007 15. Budd GT, Cristofanilli M, Ellis MJ, et al: Circulating tumor cells versus imaging-predicting overall survival in metastatic breast cancer. Clin Cancer Res 12:6403-6409, 2006 Related Article
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
