|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2004.10.988 on December 9 2003 © 2004 American Society of Clinical Oncology.
The Phase III Candidate: Can We Improve the Science of Selection?Memorial Sloan-Kettering Cancer Center, New York, NY Oncologists making treatment recommendations for their patients and practicing evidence-based medicine rely heavily on the highest level of data from random assignment trials. Such evidence is clearly needed in advanced bladder cancer, a disease that results in more than 40,000 deaths per year in the United States and Europe. Bladder cancer is sensitive to combination chemotherapy, such as the regimen of methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC), which has been studied and referenced in more than 700 publications during the last 20 years. Despite this plethora of publications, standards of care for advanced bladder cancer are derived from just several randomized trials, underscoring their importance to the field. Based on these trials, we know that classic dose MVAC is superior to single-agent therapy; no other chemotherapy regimen is clearly superior to MVAC in survival, and neoadjuvant MVAC enhances the cure rate for patents undergoing cystectomy for muscle-invasive disease [1-6]. But MVAC also has substantial limitations, among which are the short median survivals (of approximately 1 year) and a 5% 5-year survival rate, morbidities such as mucositis and neutropenic fever, and the 3% to 5% toxic death rate from treatment. Random assignment trials have addressed the morbidity of treatment; both an intensified MVAC regimen with granulocyte colony-stimulating factor (G-CSF) [3] and the doublet of gemcitabine plus cisplatin [5] have significantly less toxicity than classic MVAC. Both regimens are now used in clinical trials in the United States and Europe. We need to ensure that phase III trials comparing new therapies with standard regimens will optimize the chances for improving the standard of care, either by proving that a promising new regimen has a superior survival or by demonstrating equivalent efficacy and less toxicity. Such a phase III study is reported by Bamias et al [7] in this issue of the Journal of Clinical Oncology. The Hellenic Cooperative Oncology Group prospectively compared the combination of docetaxel and cisplatin (DC) to the original MVAC regimen, modified with G-CSF support. Although this study was designed to detect a survival advantage for DC, the investigators instead observed that survival was inferior for patients treated with DC. Based on the conclusion that DC was inferior to MVAC, the authors recommended MVAC plus G-CSF as a standard regimen. A lesson for the use of stratification variables in clinical trials of bladder cancer is provided from the study's secondary analysis; performance status had independent significance for survival. Because performance status was not used in this trial as a prospective stratification variable, the treatment arms were not appropriately balanced. The impact of performance status for advanced bladder cancer patients in the international phase III setting was first reported in 1992 by Loehrer et al [8], in a study comparing single-agent cisplatin with MVAC. Using a logistic regression model assessing pretreatment prognostic factors, the investigators reported that performance status was "the single most important factor with respect to response and survival." These data were available at the inception of the currently reported trial and should have been used in the trial design. The Hellenic Cooperative Oncology Group trial provides insight into the challenges and pitfalls of selecting an investigational arm for random assignment trials, particularly in view of highly variable outcomes inherent in phase II studies. They chose to compare the DC regimen with MVAC, with a statistical design allowing detection of a 20% relative improvement in survival for DC compared with MVAC. Given that MVAC results in a median survival of approximately 1 year, a 20% improvement for DC would therefore correspond to a survival time of approximately 15 months. However, the phase II study performed by the same investigators as the basis for the randomized study showed a median survival of only 8 months. They assumed that the poor survival observed in their phase II study was a consequence of a large number of patients with visceral disease rather than potentially inferior therapy. As evidenced by the results of this phase III study, this assumption can be erroneous and may result in an inferior outcome in the phase III setting. It is the goal for oncologists and patients to participate in randomized trials in which the experimental therapy is likely to be an improvement over current standard practice. The community expects the sponsors of such trials to optimize chances for success by selecting the best possible regimens for comparison and by using all known stratification variables for adequate outcome assessment. We need better approaches in the selection of regimens for comparison with standard therapy in phase III trials. Phase II data for new regimens are typically encouraging, because only patients with good prognostic factors are frequently entered onto a study. Selection of the best regimens for further testing is difficult for multi-institutional studies, which must choose from multiple promising candidates. One approach, therefore, is to consider only candidate regimens from multi-institutional phase II studies, because these studies will likely include a broader spectrum of patients, more representative of the general population. But even this approach may be inadequate. An alternate strategy is to calculate and account for the inherent variability of survival in diverse phase II study populations [9]. Survival prediction is based on the distribution of poor performance status and visceral disease in patients with bladder cancer [9]. Using this approach, the median survival in a typical phase II trial, in which 40% of the patients have two poor-risk features, is calculated to be 10.2 months with standard therapy. Conversely, when 40% of the patients have neither risk factor, median survival with standard therapy is estimated to be 24.6 months. By calculating predicted survival with standard therapy and comparing it to reported survival with a new therapy, investigators could thereby assess whether encouraging survival observed in a phase II study is more likely due to the new therapy or may simply be due to patient selection. Conversely, as occurred in the Hellenic Cooperative Oncology Group trial, it could provide insight into whether the poor survival observed in a phase II study is really due to preponderance of poor-prognosis patients or to inferior therapy. The selection process for new phase III candidate regimens may be enhanced by examining novel analytic approaches to improve the likelihood of choosing successful new regimens from phase II data and reduce the possibility of a negative phase III trial [10]. A limitation of this suggested strategy is that data are frequently derived from a relatively small number of patients from single institution phase II trials. Collaboration at the national and international level could provide a unique opportunity to improve the selection of candidate regimens for future phase III trials. A robust database derived from published randomized trials could potentially provide more refined survival predictions adjusted for relevant prognostic factors. This type of collaboration could make the selection of investigational regimens more rigorous, by comparing observed versus predicted survivals across divergent patient populations, thereby reducing the likelihood of selecting an inferior investigational regimen or of making errors in statistical projections. In the case of advanced germ cell tumors, this collaboration has had an immense impact on patient care and clinical trials [11]. Ultimately, it led to change in the American Joint Committee on Cancer/International Union Against Cancer staging system and standardized eligibility criteria for phase III trials, providing the basis for international collaboration in clinical trials for germ cell tumors. A rigorous approach to trial designs in bladder cancer is required. At a minimum, stratification should include previously identified prognostic variables. Optimally, new methods should be explored to improve the selection of candidate regimens for comparison to standard care. Examples of such strategies include collaborations to identify and use prognostic variables [11], more stringent decision-making rules for deciding which new regimens proceed to large-scale trials [10] and new approaches to account for the inherent variability of phase II studies [9]. Because the number of patients participating in bladder cancer trials is so limited, it is essential to avoid wasting valuable resources, by improving the methodology of clinical trials in this disease. Author's Disclosures of Potential Conflicts of Interest The following author or a member of their immediate family has indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Received more than $2,000 a year from a company for either of the last 2 years: Dean Bajorin, Bristol-Myers Squibb, Eli-Lilly & Company. REFERENCES 1. Loehrer PJ Sr, Einhorn LH, Elson PJ, et al: A randomized comparison of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial carcinoma: A cooperative group study. J Clin Oncol 10:1066-73, 1992[Abstract]
2. Grossman HB, Natale RB, Tangen CM, et al: Neoadjuvant chemotherapy plus cystectomy compared with cystectomy alone for locally advanced bladder cancer. N Engl J Med 349:859-866, 2003
3. Sternberg CN, de Mulder PHM, Schornagel JH, et al: Randomized phase III trial of high-dose-intensity methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) chemotherapy and recombinant human granulocyte colony-stimulating factor versus classic MVAC in advanced urothelial tract tumors: European Organization for Research and Treatment of Cancer Protocol No. 30924. J Clin Oncol 19:2638-2646, 2001 4. Logothetis CJ, Dexeus F, Sella A, et al: A prospective randomized trial comparing CISCA to MVAC chemotherapy in advanced metastatic urothelial tumors. J Clin Oncol 8:1050-1055, 1990[Abstract]
5. von der Maase H, Hansen SW, Roberts JT, et al: Gemcitabine and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in advanced or metastatic bladder cancer: Results of a large randomized multicenter, multinational multicenter, phase III study. J Clin Oncol 18:3068-3077, 2000
6. Siefker-Radtke AO, Millikan RE, Tu S-M, et al: Phase III trial of fluorouracil, interferon alpha-2b, and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in metastatic or unresectable urothelial cancer. J Clin Oncol 20:1361-1367, 2002
7. Bamias A, Aravantinos G, Deliveliotis C, et al: Docetaxel and cisplatin with granulocyte colony-stimulating factor (G-CSF) versus MVAC with G-CSF in advanced urothelial carcinoma: A multicenter, randomized, phase III study from the Hellenic Cooperative Oncology Group. J Clin Oncol 22:220-228, 2004 8. Loehrer P, Einhorn LH, Elson PJ, et al: A randomized comparison of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial carcinoma: A Cooperative Group Study. J Clin Oncol 10:1066-1073, 1992
9. Bajorin DF, Dodd PM, Mazumdar M, et al: Long-term survival in metastatic transitional cell carcinoma and prognostic factors predicting outcome to chemotherapy. J Clin Oncol 17:3173-3181, 1999 10. Verbel DA, Kelly WK, Smaletz O, et al: Estimating survival benefit in castrate metastatic prostate cancer: Decision making in proceeding to a definitive phase III trial. Urology 61:142-144, 2003[CrossRef][Medline]
11. International Germ Cell Consensus classification: A prognostic factor-based staging system for metastatic germ cell cancers. J Clin Oncol 15:594-603, 1997
Related Article
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
