Biostatistical aspects of adjuvant therapy assay.

The specific problems of biostatistical evaluation of adjuvant therapy are discussed. These problems arise from the fact that the effect of adjuvant therapy is usually rather small; in particular, for many kinds of cancer it is smaller than the variability in prognosis among patients due to different prognostic factors. Knowledge about prognostic factors for a particular type of cancer is therefore an essential precondition for the correct planning of controlled clinical trials. The importance of prognostic factor studies is stressed and it is shown how results of such studies can be utilized for the design of efficient trials for adjuvant therapy assay. The results of the retrospective DOSAK study on carcinomas of the oral cavity are reported. Based on a sample of 1021 patients in this study, numerous prognostic factors were analysed by both univariate and multivariate statistical techniques. On the basis of these analyses, a prognostic index was developed which should be a valuable aid for the design of clinical trials.