Annals of Oncology Advance Access originally published online on June 9, 2006
Annals of Oncology 2006 17(10):1607-1608; doi:10.1093/annonc/mdl106
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2006 European Society for Medical Oncology
letters to the editor |
Factorial design for randomized clinical trials
1 Department of Medical Oncology, Regina Elena National Cancer Institute, Rome
2 National Cancer Institute, Naples
3 Biostatistics, Regina Elena National Cancer Institute, Rome, Italy
*(E-mail: emiliobria{at}yahoo.it)
The recent publication of the randomized clinical trial performed by the M.D. Anderson Cancer Center raised the issue of the benefit of weekly paclitaxel over classical 3-weekly regimen in breast cancer [1]. Many clinical researchers widely approach the field of chemotherapy delivery optimization by increasing dose-intensity through dose-dense and weekly strategy [2]. The most important trial that demonstrated the benefit of dose-dense chemotherapy was designed as a factorial design, exploring the effect on outcome of both concomitant over sequential approach, and a 2-weekly over a 3-weekly schedule [3]. The sample size of the factorial design was based on a non-interaction assumption between the two aimed questions. The application of factorial design allows two independent questions to be answered using the same patients: in other words, it is a simple way to conduct two trials in one.
A further, still unpublished, ongoing trial, the ECOG 1199 study [4], is testing the weekly schedule in a factorial design. In this study, patients with breast cancer underwent surgery and adjuvant anthracycline-based chemotherapy for early breast cancer, and were subsequently randomized to receive paclitaxel or docetaxel, each drug either in a weekly or 3-weekly schedule. Does a non-interaction hypothesis really exist under this important clinical issue? Are the questions ‘paclitaxel versus docetaxel’ and ‘weekly versus 3-weekly’ really independent between them?
This issue became complicated by the results of important phase III trials. Dr Seidman presented the first results of the CALGB 9840 randomized trial at the 2004 American Society of Clinical Oncology meeting, which randomized patients with metastatic breast cancer to receive paclitaxel in either a weekly or 3-weekly schedule [5]. The weekly administration of paclitaxel provided a significant improvement in response and time to progression over the standard schedule. Indeed, basic and clinical evidence supports two different behaviours and mechanism of paclitaxel when administered weekly or 3-weekly [2, 6–8]. Dr Jones recently published the results of the TAX 311 randomized trial, in which patients with metastatic breast cancer were randomized to receive 3-weekly docetaxel or paclitaxel [9]; docetaxel provided a significant benefit in time to progression and overall survival.
These findings seem to suggest that weekly paclitaxel is better than 3-weekly and that 3-weekly docetaxel is better than 3-weekly paclitaxel. Moreover, to date no evidence is available about the weekly/3-weekly comparison of docetaxel in metastatic breast cancer in a phase III fashion, although evidence in other settings, such as advanced non-small-cell lung cancer, suggests a substantial equivalence of the two schedules. When comparing paclitaxel versus docetaxel in the ECOG 1199 trial, 3-weekly paclitaxel could lower the eventual benefit of weekly paclitaxel. Similarly, when comparing a weekly versus a 3-weekly schedule in the ECOG 1199 trial, the eventual benefit provided by weekly paclitaxel could be diluted, because the weekly docetaxel could not determine a similar benefit. Of course, factorial design allows the demonstration of an interaction between the two experimental questions, but we should remember that, if a significant interaction exists, the planned sample size would not be enough to guarantee the desired statistical power. In other words, if a difference of efficacy exists between weekly and 3-weekly paclitaxel, but not between weekly and 3-weekly docetaxel, the randomized patients that received docetaxel would dilute the difference of efficacy between weekly and 3-weekly schedule existing in the other half of the patients receiving paclitaxel.
This risk seems to be real, according to the first results of the trial, recently presented at the San Antonio Breast Cancer Symposium [4]. No difference in efficacy was shown between a weekly and a 3-weekly schedule of chemotherapy, but a trend favouring weekly over 3-weekly paclitaxel was demonstrated. In our opinion, randomized clinical trials with a factorial design, such as ECOG 1199, represent a good way of demonstrating an interaction between factors. However, when planning such a trial, if a significant interaction is expected, it should be kept in mind that its occurrence could make the primary result of the trial practically useless, and the unique useful information could come from secondary, less powered analyses.
 |
References |
|---|
 Top References |
|---|
1. Green MC, Buzdar AU, Smith T, et al. (2005) Weekly paclitaxel improves pathologic complete remission in operable breast cancer when compared with once-every-3-weeks paclitaxel. J Clin Oncol 23:5983–5992.
2. Seidman A. (2005) ‘Will Weekly Work?’ Seems to be so.... J Clin Oncol 23:5873–5874.
3. Citron ML, Berry DA, Cirrincione C, et al. (2003) Randomized trial of dose-dense versus conventionally scheduled and sequential versus concurrent combination chemotherapy as postoperative adjuvant treatment of node-positive primary breast cancer: first report of Intergroup trial c9741/Cancer and Leukemia Group B trial 9741. J Clin Oncol 21:1431–1439.
4. Sparano JA, Wang M, Martino S, et al. (2005) Phase III study of doxorubicin-cyclophosphamide followed by paclitaxel or docetaxel given every 3 weeks or weekly in patients with axillary node-positive or high-risk node-negative breast cancer: results of North American Breast Cancer Intergroup Trial E1199. Proceedings San Antonio Breast Cancer Symposium.
5. Seidman AD, Berry D, Cirrincione C, et al. (2004) Phase III study of weekly paclitaxel via 1-hour infusion versus standard 3h infusion every third week in the treatment of metastatic breast cancer, with trastuzumab for HER2 positive MBC and randomized for trastuzumab in HER2 normal MBC. Proc Am Soc Clin Oncol 22:6s (Abstr 512).
6. Winer EP, Berry DA, Woolf S, et al. (2004) Failure of higher-dose paclitaxel to improve outcome in patients with metastatic breast cancer: cancer and leukemia group B trial 9342. J Clin Oncol 22:2061–2068.
7. Jones SE, Erban J, Overmoyer B, et al. (2005) Randomized phase III study of docetaxel compared with paclitaxel in metastatic breast cancer. J Clin Oncol 23:5542–5551.
8. Taghian AG, Abi-Raad R, Assaad SI, et al. (2005) Paclitaxel decreases the interstitial fluid pressure and improves oxygenation in breast cancers in patients treated with neoadjuvant chemotherapy: clinical implications. J Clin Oncol 23:1951–1961.
9. Pasquier E, Honore S, Pourroy B, et al. (2005) Anti-angiogenic concentrations of paclitaxel induce an increase in microtubule dynamics in endothelial cells but not in cancer cells. Cancer Res 65:2433–2440.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||