Unpublished Data Summaries and the design and conduct of clinical trials. The Nutrition Adjuvant Study experience and commentary.

A trend in cancer clinical investigation has been the application of new analytic techniques and reporting forums to summarize developing trial results. Examples include: Consensus Conferences, Meta-Analyses, and most recently (in the breast cancer area), the "Clinical Alert." These Unpublished Data Summaries have been widely disseminated in lay and scientific communities and have frequently engendered debate conducted in the absence of primary information. We now report the impact of this process on a national, cooperative group effort (the Nutrition Adjuvant Study [NAS] ) designed to test a novel hypothesis involving dietary fat reduction as potential adjuvant breast cancer treatment. It is clear that these Unpublished Data Summaries in the breast cancer area directly resulted in changes in the NAS protocol design and may have influenced patient accrual. The challenge for clinical investigators and governmental agencies is to integrate the positive aspects of the new information forums with those of traditional "peer-review" publication into a system where the conduct of clinical investigation in a timely manner can be facilitated.