Collaborative clinical trials: Quality or quantity?

Purpose: To review the merits of using the limited available resources - patients, money, clinical scientists, and ideas - in various types of clinical trial. Conclusions: Two types of trial represent a poor use of resources: (a) nonrandomized trials that provide no insight into biologic mechanisms and are not precursors to testing new strategies in comparison with standard treatment in randomized trials; and (b) small randomized trials that are difficult to interpret because of a high rate of false-positive and false-negative trials. Very large trials that can detect small differences in survival for patients with common tumors are appropriate, but a similar design to detect transient improvements due to palliative therapy represent a poor use of resources. Larger gains in therapeutic index will require the recognition of tumor heterogeneity and the conduct of small trials that are based on biologic hypotheses, and which provide mechanistic information in patients; two examples are provided. Ultimately, strategies that may be individualized among a group of patients with histologically similar tumors will need to be evaluated against the current standard (and homogeneous) treatment. (C) 2001 Elsevier Science Inc.