Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Aromatase inhibitors in advanced breast cancer

Systematic review

Bandolier is dimly aware of the difficulties of evaluating cancer treatments. New treatments are usually evaluated against best current treatment, a moving target that can make trials difficult, let alone trying to get a coherent picture from overviews. Cancer treatments differ from early after diagnosis to when the disease is more advanced. Then there is the difficult issue of outcomes, with survival as the hardest, though not necessarily always the most important; getting good survival data means long trials, and results are neither quick nor easily come by.

Looking at cancer, then, is a bit of a minefield. But when a new meta-analysis of aromatase inhibitors in breast cancer swims into our ken [1], it deserves that at least we stir the neurone to try and understand what it says.

Systematic review

Randomised trials were eligible if they compared an aromatase inhibitor or inactivator with tamoxifen or progestagens (like medroxyprogesterone acetate) in patients with advanced breast cancer, defined as metastatic and inoperable locally advanced or recurrent breast adenocarcinoma. Any line of treatment was considered, whether first line or second or subsequent line in patients who had received such therapy in the past. Excluded were trials in earlier stages or with other histological types of cancer.

Data were analysed according to the generation of the agent: first (aminoglutethimide), second (formestate, fadrazole), and third (anastrazole, examestane, letrazole, vorozole) versus standard hormonal treatment. Also analysed were third generations versus tamoxifen as first line and progestogens as second or subsequent lines of treatment.


Twenty-three trials with survival data with 8,500 women were eligible for the meta-analysis, published between 1982 and 2004; 11 were double blind, and nine investigated treatment as first line therapy. Trial size varied between 40 and about 800 patients. The typical median age of women in the trials was about 65 years.

Only third generation aromatase inhibitors and inactivators demonstrated any significant survival benefit compared with standard hormonal treatment (Table 1).

Table 1: Comparisons of different generations and line of use of aromatase inhibitors, and the relative hazard for mortality in breast cancer trials

Number of comparisons
Relative hazard
Generation (vs standard hormone treatment)
0.96 (0.84 to 1.09)
1.00 (0.89 to 1.13)
0.87 (0.82 to 0.93)
First line vs tamoxifen
0.89 (0.80 to 0.99)
Second line vs progestagens
0.86 (0.79 to 0.84)

In cumulative meta-analysis, statistical significance only occurred in 2000, and remained subsequently, with the addition of the third generation aromatase inhibitors. Only four individual trials showed statistical significance on their own, and three of these were third generation aromatase inhibitors.

The survival benefit for third generation agents was similar in both first line treatment compared with tamoxifen and second or subsequent line versus progestogens (Table 1).


This meta-analysis shows a small but significant benefit of third generation aromatase inhibitors (anastrazole, examestane, letrazole, vorozole) over standard hormonal treatment in advanced breast cancer. It also shows them to have a benefit over tamoxifen and progestogens, and it is of sufficient importance to be considered when making decisions about care pathways in breast cancer. It might well change some of them.

It exemplifies the need for meta-analysis when trials are relatively small (average size of 360 patients in these trials), when the outcome of the trial is survival, when there are relatively few deaths because of relatively short duration of trials (median survival was typically two years or more), and when differences between groups was relatively small (10-15% reduction with third generation aromatase inhibitors).

For a theoretical median survival of 20 months with standard treatment with standard hormone therapy, change to or addition of third generation aromatase inhibitor would confer an additional four months or so of life. By such small steps are improvements in cancer treatment achieved.


  1. D Mauri et al. Survival with aromatase inhibitors and inactivators versus standard hormonal therapy in advanced breast cancer: meta-analysis. Journal of the National Cancer Institute 2006 98:1285-1291.

previous or next story