RESUMEN
OBJECTIVE: To evaluate the efficacy and safety of aromatase inhibitors (AIs) as adjuvant hormonal therapy for postmenopausal patients with hormone receptor positive breast cancer compared with tamoxifen therapy, or as a subsequential treatment to this therapy. METHOD: We carried out a bibliographical search using the Medline database and papers presented at the American Society of Clinical Oncology and the San Antonio Breast Cancer Symposium conferences. The efficacy parameters evaluated were disease-free survival (DFS) and overall survival (OS), measured after the follow-up period for each of the tests analysed. The data selected were subjected to a stratified meta-analysis according to the different strategies used to introduce the AIs. The Mantel-Haenszel odds ratio (MH OR; at 95% CI) was calculated for each of the parameters evaluated. The principal side effects with respect to toxicity, described in the various comparative clinical trials, have been listed in a table. RESULTS: Compared to tamoxifen, AIs have been shown to increase DFS both as a first-line therapy (MH OR = 0.83; 95% CI: 0.76-0.92) and when sequentially administered for 2-3 years following 2-3 years of tamoxifen treatment (MH OR = 0.65; 95% CI: 0.57-0.75). Similarly, their use after 5 years of tamoxifen therapy also produces an increase in DFS (MH OR = 0.63; 95% CI: 0.51-0.77). As for OS, a statistically significant difference is obtained only when the AI is administered following 2-3 years of tamoxifen treatment (MH OR = 0.77; 95% CI: 0.64-0.94). In comparison with tamoxifen, AIs reduce the incidence of thromboembolic and gynaecologic events, although they increase bone toxicity. CONCLUSIONS: The clinical studies evaluated show the consistent benefits of AIs at different adjuvant treatment stages; however, we have been unable to establish the optimum moment for their introduction due to the absence of direct comparisons between the different strategies. We now need to focus on the selection of patient sub-groups which could benefit from their use as a first-line therapy, the long-term toxicity of AIs, and their capacity to increase OS, regardless of the strategy followed, after a longer monitoring period. In light of the evidence available, bearing in mind certain limitations, we propose criteria for the use of AIs in daily clinical practice.