Issues in clinical trial design for tumor marker studies.

Scientific inquiry into the discovery, development, and application of tumor markers is proceeding rapidly. Despite this explosion in research and interest, the design of studies to formally assess the value of tumor markers in clinical practice is inconsistent and immature. Indeed, few markers have been widely accepted into standard clinical practice. Many issues must be prospectively considered in a methodical, systematic, and scientific fashion if progress is to be made in the development of validated tests that will have value in the management of patients with cancer. The purpose of this report is to present a discussion of the issues involved in designing clinical studies of putative tumor markers which provide sufficient data to result in the incorporation of the marker into clinical practice. We will focus on the design of studies to demonstrate and validate the clinical utility of both prognostic and predictive markers. Topics to be covered include issues of patient and sample heterogeneity, the prevalence of the marker, the sample capture rate, and the choice of endpoints. This will be followed by explicit consideration of study design, specifically the trial randomization schema for both prognostic and predictive factor studies.