Assessment of adjuvant trials in breast cancer.

In this paper, we discuss the analytic problems associated with the evaluation of overall survival for breast cancer adjuvant studies and results of pooling data from the published literature to determine if there is evidence showing an overall survival advantage to adjuvant therapy. An investigation of the effect of competing causes of death shows that trials on older patients have low statistical power. Hence many of the current trials have low statistical power and may fail to find overall survival advantage for adjuvant therapy even when a benefit exists. The implications of the assessment of short-term follow-up are discussed in the context of the heterogeneous distribution of residual disease after primary treatment. Short-term follow-up (less than 5 yr) precludes anyone from making any conclusions about benefit for patients with small residual tumor burdens who constitute the patient subgroup most likely to benefit from adjuvant therapy. We reviewed 15 published randomized trials (each having a control group receiving no systemic therapy) to determine if overall survival has been increased by adjuvant therapy. We conclude that there is a benefit from some chemotherapy regimens given to node-positive premenopausal women. However, the published data for tamoxifen are mixed, with some trials showing benefit and others not. As more follow-up time is accumulated, this matter should be settled.