Predictive analysis of tumor response

To select the best treatment modality for individual patient with cancer, the oncologist is subjected to consider various factors and parameters predictive of the outcome of therapy. A critical review is presented relating to the prognostic factors in the local control and/or long term survival rate in various kinds of malignant tumors. As predictors of response in cancer patients, it is attributed not only to physical characteristics of ionizing radiation and technical factors, but also tumor- and host-related factors such as gene amplification, DNA pattern, nuclear shape factor, histological grade, growth rate, glucose metabolic state, NMR spectroscopy, tumor marker levels, shape and location of tumor, in vitro sensitivity assay of tumor, age, general immunologic competence and psychological state of the patient. Most of the studies done sofar, are focused on the involvement of only one or two factors for each tumor, but importance of each factor as a predictor of tumor response is not clearly highlighted. It is emphasized that multivariate analysis of all these factors including potential predictors should be performed for each malignant tumor and for each treatment modality.