The international conference results from a continuing collaboration between the British Breast Trials Co-ordinating Committee, the Clinical Trial Service Unit, the Union Internationale Contre le Cancer, and the WHO Cancer Office. Mortality is such a serious endpoint that even moderate reductions in it may, especially if produced by widely practicable treatments for common neoplasms, be worth knowing about. Unfortunately, such moderate differences may not be reliably detectable in particular trials, simply because the randomized controlled trials (RCTs) that are currently available may be too small - for example, in early breast cancer (EBC) a typical RCT may involve only a few hundred EBCs. But, when many RCTs have addressed related questions, a properly unbiased overview of all their results may yield evidence far more reliable than that from any single one of the several separate studies would be. Last year, a conference of the principal investigators of 43 RCTs of tamoxifen (TMX) on a total of 16,000 EBCs, of 37 RCTs of cytotoxic chemotherapy (CTX) on a total of 10,000 EBCs and of several RCTs of radiotherapy (XRT) on a total of 10,000 EBCs showed vastly significant (3.3 SD & 5.2 SD) mortality reductions from the first two (Lancet, 24 Nov, p.1205), but a non-significant adverse effect for the latter. This exercise needs repeating, and publishing in a form that can be assessed critically, with more careful scrutiny (and separate publication) of the methods and results of each separate RCT. For this to take place, another meeting of trialists is absolutely essential. The ideal time for this is on Sep 7-8 in Washington, just prior to the NIH consensus development conference on Sep 9-11 on adjuvant treatment for EBC. Trialists will be asked to send (as they did in 1984) their interim mortality data - in strict confidence - to the CTSU at Oxford University, to be checked for completeness of patient lists and analyzed before the 1985 meeting. Publication of each study in special issues of the Lancet and JNCI has been offered. To judge by the 1984 results, this conference is likely to demonstrate, with overwhelming statistical significance, that for EBC less than age 50 CTX with regimes such as CMF may reduce early mortality by about 1/3, while for EBC more greater than 50 TMX may reduce it by about 1/6. If so, this will be the first conclusive proof that systemic therapy can reduce mortality by a potentially worthwhile amount in any common neoplasm.
