With regard to our previous article,1 the author of the current letter is correct in his comments that while the North American trial (0030) showed a significant advantage for anastrozole over tamoxifen for time to progression (TTP),2 the European trial (0027) did not.3 It is our belief that this difference is mainly due to the imbalance in the number of patients with confirmed hormone-receptor positive tumors between these two trials (89% vs. 45%). In trial 0027, treatment with anastrozole was associated with increased TTP (8.9 vs. 7.8 months) in the 45% of patients with confirmed hormone receptor-positive tumors. However, no statistical analysis of these data was performed as both studies were prospectively designed for combined data analysis, and it was considered more appropriate statistically to perform subgroup analyses on the larger combined data set. Regarding the author's concern about the trend toward a lower TTP in the anastrozole group compared with the tamoxifen group in trial 0027, since receptor status was unknown in this subgroup, it is possible that patients with receptor positive disease may be distributed unequally between the two groups. Thus, one cannot compare the unknown subgroup with any meaningful certainty. This combined data set for the hormone-receptor positive subgroup across both trials showed a statistically significant advantage in favor of anastrozole for TTP (P = 0.022). It was clearly stated within our article that the subgroup analysis was retrospective; however, we felt it was essential to emphasize strongly the importance of determining tumor hormonal receptor status before deciding on treatment strategy. In this context, for patients who were appropriate candidates for endocrine therapy, anastrozole was superior to tamoxifen in terms of TTP. In our view, these observations have an important practical clinical use, as they provide guidance in selecting hormonal therapy for receptor-positive patients. The author is correct in his comment that both trials were designed as noninferiority trials. This was clearly identified in both papers. However, it is accepted by many regulatory agencies, including the Food and Drug Administration (FDA), that in some cases, trials designed to test for noninferiority can be used to show superiority, based on the nature of the studies and the larger sample size included in such trials.4 Regarding the author's comments on survival, the data quoted from the FDA website are immature, with approximately 60% of patients in trial 0030 and 70% of patients in trial 0027 still alive at the time of the report. The data contained in the website were submitted to the FDA as part of a routine safety update, and because of the immaturity of the results were not included in the publications discussed here. A mature survival analysis of the combined data from the two trials will be presented later this year. Finally, the article by Bonneterre et al. reporting the results of the combined data from the two trials did not determine TTP in the subgroup of patients who had received prior adjuvant tamoxifen, as there were insufficient patients to allow a meaningful analysis. Only 77 patients in the anastrozole group and 68 patients in the tamoxifen group had received adjuvant tamoxifen therapy, either alone or in combination with chemotherapy. In summary, we believe that the combined analysis of both trials indicates that anastrozole increased TTP compared with tamoxifen in patients with hormone receptor-positive disease. Given its favorable toxicity profile, with significantly fewer thromboembolic events and approximately half the incidence of vaginal bleeding compared with tamoxifen, these data support the use of anastrozole as first-line treatment for hormone-responsive advanced breast carcinoma in postmenopausal women. Furthermore, results from the first planned analysis of the Arimidex, Tamoxifen, Alone or in Combination trial (ATAC) for the adjuvant treatment of early breast carcinoma in postmenopausal women have shown anastrozole to produce significantly prolonged disease-free survival (P = 0.013) compared with tamoxifen, along with a significant reduction (P = 0.007) in the incidence of contralateral breast carcinoma,5 and are, therefore, supportive of our findings in first-line therapy for patients with advanced breast carcinoma. Jacques Bonneterre M.D.*, Aman Buzdar M.D. , Jean-Marc Nabholtz M.D. , John Robertson M.D.?, Beat Thürlimann M.D.?, Mikael von Euler M.D.**, Tarek Sahmoud M.D. , Mark Steinberg M.D. , Alan Webster B.Sc. , * Centre Oscar Lambret, Lille, France, M.D. Anderson Cancer Center, Houston, Texas, University of California at Los Angeles, Los Angeles, California, ? City Hospital, Nottingham, UK, ? Swiss Group for Clinical Cancer Research, St. Gallen, Switzerland, ** AstraZeneca, Osaka, Japan, AstraZeneca, Wilmington, Delaware, AstraZeneca, Alderley Park, Cheshire, UK.
Read full abstract
Apr 29, 2005
G Cartron + 5
Open Access