|
Anastrozole (ARIMIDEX) for adjuvant treatment of breast cancer - new meta-analysis unsuitable as proof of a survival benefit: A meta-analysis of three studies (ABCSG trial 8, ARNO 95, ITA*) presented few months ago at a conference on adjuvant breast cancer treatment with the aromatase inhibitor anastrozole (ARIMIDEX) compared to the standard therapy, the anti-oestrogen tamoxifen (NOLVADEX etc.), has now been published in full. "Study shows survival benefit for anastrozole" is the heading used by the publication organ, Lancet Oncology, in a press release. However, because of major deficiencies in the three analysed studies, the meta-analysis financed by the manufacturer is without validity. All three studies examined the change to anastrozole after two or three years of tamoxifen therapy compared to taking tamoxifen alone for five years. In two of the studies, including the largest one, almost 30% out of a total of 4,960 originally randomised women were not analysed because they either had not yet completed the two-year tamoxifen therapy, stopped the therapy during this period, died, have developed recurrence or a second carcinoma or did not meet the inclusion criteria. Among the other women who were included in the meta-analysis, at least 10% were missing in the follow-up observation of overall mortality after three years**. This number is much larger than the difference in deaths between the two treatment groups (3.3% on anastrozole versus 4.5% on tamoxifen with a median follow-up of 30 months). In addition, more women on anastrozole than on tamoxifen had dropped out from follow-up at this time. The result is, therefore, not robust. The intention-to-treat principle is violated. The apparent advantage in survival with anastrozole, moreover, is based to a great extent on the results of the two smaller studies with together 1,400 patients (24 vs. 46 deaths), while there is no difference in the biggest study with almost 2,600 analysed patients (42 vs. 44 deaths). All three studies were open-label and were stopped prematurely because of unplanned interim analyses. This can also lead to overestimation of the observed effect (a-t 2005; 36: 107-8 and 2006; 37: 19-21; JONAT, W. et al.: Lancet Oncology: published online 17. Nov. 2006). Aromatase inhibitors reduce the recurrence rate in the adjuvant therapy of breast cancer compared to tamoxifen. The crucial question of whether they prolong life cannot be answered positively from the available data, -ed.
|
