|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 28 (October 1), 2008: pp. 4694-4695 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2008.18.8292
In ReplyDepartment of Medicine, Memorial Sloan-Kettering Cancer Center, Gastrointestinal Oncology Service, New York, NY Dr Booth's comments regarding the report of the NO16966 trial are appreciated. We are in agreement that this large, randomized, double-blind, placebo-controlled trial adds new—and to a degree—unexpected information regarding bevacizumab in colorectal cancer (CRC) management, and that new information should be added to—and interpreted in—the context of the other trial results and information available to us on the topic. However, Dr Booth states in his letter, "If this was the first large randomized controlled trial [RCT] to address the role of bevacizumab in advanced CRC, I suspect the authors’ conclusions would be quite different." Here I respectfully disagree. The conclusion of the article reads as follows: "In conclusion, this trial reached its primary objective by showing a statistically significant increase in progression-free survival [PFS] through the addition of bevacizumab to oxaliplatin-based chemotherapy in first-line metastatic CRC. No increase in response rate [RR] was seen. The observed difference in overall survival [OS] did not reach statistical significance. Continuation of bevacizumab, and most likely fluoropyrimidine therapy as well, until disease progression [DP] appears to be critical with regards to the magnitude of clinical benefit derived from bevacizumab."1 The statements in the conclusion are an appropriate summation of the trial results. It is outside of the scope of the article to establish what criteria should be used in defining standard treatment paradigms. However, it is well within the purview of Dr Booth and others to consider the information presented, and to challenge their assumptions regarding what should or should not be standard practice. Are the benefits seen with the addition of bevacizumab as substantial as those seen in prior studies? Clearly not. Is the earlier discontinuation of bevacizumab a satisfactory explanation for these differences? Perhaps; it is a reasonable hypothesis, but one that is unlikely to be interrogated by another large RCT given the expense and complexity of such an undertaking. Could other factors account for these differences? Certainly. Should the findings of the NO16966 trial change the paradigm for treatment of CRC? That is a more complex question, without a clear and simple answer. In our current drug development paradigm, success or failure is predicated on the achievement of the prespecified primary end point, in this case a statistically significant improvement in progression-free survival. That objective end point was achieved in the NO16966 trial. Dr Booth raises the important subjective question of what is a clinically meaningful improvement in PFS. This is a challenging question to answer, a challenge to which we as a cancer community of doctors, patients, industry, government, and society as a whole, have not yet risen. Dr Booth writes, "As clinicians and investigators, we need to have an open dialogue about the true benefit of these novel anticancer therapies in the general population and distinguish results that are clinically significant from those that are only statistically significant." I could not agree more. How to do this, however, is not at all clear. The incorporation of bevacizaumab into current treatment paradigms is based on the trial reported by Hurwitz et al,2 which showed a PFS advantage of 4.4 months and an OS advantage of 4.7 months. To those of us who treat patients and conduct clinical research, these findings may seem robust and meaningful. To a patient with CRC, however, or to family member or friend of that patient, they might seem considerably less so. In the NO16966 trial, the PFS advantage was 1.4 months. The OS benefit was also 1.4 months; however, the P value was .078, which did not meet the prespecified level of statistical significance, and so is a negative finding as defined in the trial (note that this does not mean there was no survival advantage; it means that the survival difference seen was only 1.4 months, and we cannot say with 95% confidence that this modest difference is due to a real drug effect, and not just due to chance; we can only say it with 92% confidence). How then are we to interpret the findings of the NO16966 trial? Do they mean that the prior studies were wrong? No. Do they mean that there is no benefit to adding bevacizumab to the fluorouracil-leucovorin-oxaliplatin or capecitabine-oxaliplatin regimens? No, but the benefits seem to be more modest than we had hoped for; specifically, the data are most compelling that the addition of bevacizumab to front-line oxaliplatin-based therapy does not improve the response rate at all. How, then, do we reach consensus on what is clinically meaningful? Within the context of how difficult it has been to show any benefit for any new interventions in CRC, both the NO16966 trial and the Hurwitz et al trial might be regarded as clinically meaningful; within the context of what we want for our patients and what they need, arguably no colorectal trial to date fulfills the clinically meaningful criterion. Much discussion will be required to reach agreement on what a clinically meaningful improvement in a given end point is. It would be useful for such criteria to be defined in clinical trials a priori, but I fear that reaching societal consensus on the definition of clinically meaningful, if it is to be agreed that the drug will not be used if it fails to meet that prespecified clinically meaningful threshold, will be as challenging as the drug development process itself. In the mean time, one can hardly argue with the call to develop meaningful predictive markers to allow for appropriate patient selection. However, here too we must be careful to recognize that what a predictive marker would offer in the setting of bevacizumab in CRC is to tell us who not to treat. It will not open up a new option for any CRC patient; only the development of a new effective drug would do that. What a useful predictive marker, if found, would offer to the CRC community is a more precise delineation of for whom the use of bevacizumab will not provide a benefit that is clinically meaningful. AUTHOR'S 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: Leonard B. Saltz, Pfizer (C), Roche (C), Genentech (C), ImClone (C), Merck (C), YM Bioscience (C) Stock Ownership: None Honoraria: None Research Funding: Leonard B. Saltz, Pfizer, Bristol-Myers Squibb, ImClone, Roche, Genetech, Bayer, Merck, Taiho, Amgen Expert Testimony: None Other Remuneration: None REFERENCES
1. Saltz LB, Clarke S, Diaz-Rubio E, et al: Bevacizumab in combination with oxaliplatin-based chemotherapy as first-line therapy in metastatic colorectal cancer: A randomized phase III study. J Clin Oncol 26:2013-2019, 2008 2. Hurwitz H, Fehrenbacher L, Novotny W, et al: Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 350:2335-2342, 2004
Related Correspondence
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
