The disease-free interval in breast cancer trials: scientific or spurious?

Reports that stem from the National Surgical Adjuvant Breast and Bowel Project clinical trials have had a major influence on the trend toward adjuvant chemotherapy. These reports use life-table analysis to report survival or disease-free survival (DFS). DFS is reported more frequently than survival, and most studies report an increase in DFS with adjuvant therapy but no difference in survival. This means a shorter time between recurrence (end of DFS) and death. A conceptual model is presented herein with five possible explanations for this observation: (1) Bias exists in randomizing (entrance); (2) bias is introduced by the examining physician (exit end point); (3) the spectrum of micrometastases is selectively affected; (4) the sites of systemic recurrence (macrometastases) are affected; (5) treatment after recurrence (salvage) is less effective in the previously treated group; probably because of the induction of resistance to antineoplastic therapy. Hypothesis 5 seems most probable, and if so, the benefit of chemotherapy should be reserved for those women who need it rather than rendering everyone less treatable. The disease-free interval, with its doctor-determined end point, is insufficiently reliable to be used for major decisions regarding adjuvant therapy.