This meta-analysis found that the most effective of the chemotherapy regimens investigated (based on taxane plus anthracycline or higher cumulative dosage anthracycline) offered on average a one-third reduction in 10-year breast cancer mortality.
Absolute benefit from this reduction depended on absolute risk without chemotherapy. This review of individual patient data was well conducted, thorough and reliable.
Eligible studies needed to provide individual patient data on patients with breast cancer from randomised controlled trials (RCTs) that began between 1973 and 2003. Trials needed to compare taxane-based versus non-taxane-based regimens or any anthracycline-based regimen versus standard or near-standard CMF (cyclophosphamide, methotrexate, fluorouracil) or higher versus lower anthracycline dosage or polychemotherapy versus no adjuvant chemotherapy.
Trials of intensive chemotherapy with stem-cell rescue or of variation only in dose-density were not included.
According to original criteria developed by the Early Breast Cancer Trialists' Collaborative Group, trials conducted in USSR and Japan were excluded due to several lacking detail needed for review.
The main outcomes of interest were recurrence, breast cancer mortality and all-cause mortality. Study selection was conducted in conjunction with trialists. Eligible trials needed to ensure adequate randomisation and concealment of treatment allocation. Trials were excluded if there were imbalances between treatment groups in terms of patient numbers or follow-up. Various consistency checks were conducted in conjunction with principal investigators of trials.
One hundred and twenty-three trials were included (more than 100,000 participants).
A large number of analyses were conducted (the main results as reported by the authors are presented here).
Trials that added four separate cycles of a taxane to a fixed anthracycline control extended treatment duration demonstrated a reduction in breast cancer mortality (RR 0.86, SE 0.04, p = 0.0005).
There were no significant differences when extra cycles of a taxane were counterbalanced in control groups by extra cycles of other cytotoxic drugs. Trials that compared standard 4AC (four cycles of doxorubicin and of cyclophosphamide three-weekly) and standard CMF had similar results. Anthracycline-based regimens with substantially higher cumulative dosage than standard 4AC were found to be superior to standard CMF (RR 0.78, SE 0.06).
Trials versus no chemotherapy suggested greater mortality reductions with CAF (cyclophosphamide, doxorubicin, fluorouracil) (RR 0.64, SE 0.09) than with standard 4AC (RR 0.78, SE 0.09) and standard CMF (RR 0.76, SE 0.05).
In all analyses that involved taxane-based or anthracycline-based regimens, reductions were little affected by age, nodal status, tumour diameter or differentiation and oestrogen receptor status or tamoxifen use. Overall mortality differences were similar to those for breast cancer mortality.
The most effective of the regimens investigated (based on taxane plus anthracycline or higher cumulative dosage anthracycline) offered on average a one-third reduction in 10-year breast cancer mortality compared with no adjuvant chemotherapy. Absolute gain from this reduction depended on absolute risk without chemotherapy.