Multivariate analysis of prognostic factors for operable rectal cancer.

Previous reports have established that several variables are of prognostic importance in rectal cancer. The information from a Medical Research Council trial in seventeen centres in the United Kingdom was used to assess the simultaneous contribution of several of these factors in a multivariate analysis. In the preoperative assessment mobility of the tumour and the number of quadrants involved were predictive of the likelihood of a curative resection being possible. Among patients who had curative resections, prognosis was influenced by Dukes' classification, height of tumour, histological grade, and venous invasion. A scoring system based on these criteria was constructed and applied to this group of patients. It successfully defined five prognostic groups with 5-year survival rates ranging from 80% to 19%.