A multivariate analysis of pathologic prognostic indicators in large bowel cancer.

A multivariate analysis of the pathologic data of 350 patients with primary colorectal cancer was performed. In addition to conventional parameters such as shape and size of the primary tumor, central node involvement, angioinvasive growth, grade, and stage, new variables such as the immunoreactivity patterns of carcinoembryonic antigen (CEA), CA 19-9, mucin, serotonin, secretory component (SC), and the DNA index were tested for their potential prognostic value. Every variable except CA 19-9, serotonin, and DNA showed significant prognostic information in univariate analysis. However, in the multivariate analysis stage was the predictive factor with the highest hazard ratio, but absence of central node involvement, tumors with diameters between 3.5 cm and 6 cm, exophytic tumor growth, well-differentiated tumors, tumors with CEA immunoreactivity, absence for staining with serotonin, and diploid tumors also were included in the relative risk model. Thus, the aforementioned variables appear to play a role in the establishment of a prognostic index.