Histopathologic parameters and DNA analysis in colorectal adenocarcinomas.

Human colon adenocarcinomas have histological parameters that are clearly associated with prognosis. These include tumor grade, pattern of invasion, presence of lymphocytes, and vascular involvement by tumor. The latter remains controversial with respect to the relative importance of intramural and extramural vascular involvement. Some studies show a poor prognosis for intramural invasion of capillary size vascular channels by tumor. On the other hand, when veins are involved by tumor, the presence of tumor in large extramural veins appears to have a much more ominous effect than intramural tumor involvement of small veins. The results of DNA analysis of colorectal adenocarcinomas varied greatly depending on study methodology but several important points can be summarized: (1) higher stage tumors have a greater proportion of aneuploid tumors; (2) aneuploid tumors tend to have a higher growth rate (SPF) and poorer survival than diploid tumors; and (3) aneuploid tumors are associated with histological parameters indicative of a poor prognosis such as vascular invasion, but ploidy is not related to tumor grade. One of the major problems in drawing firm conclusions about the relationship of flow cytometric DNA measurements to prognosis is great variability among the reported studies. The types of variation appear to fall into two major categories: patient selection and technical problems. The former are especially relevant in retrospective studies in which there is poor patient definition (site, grade, stage, and other standard tumor definitions) and a bias for selecting early stage tumors which have improved survivals. The latter includes a spectrum of technical problems inherent in this widely but not necessarily uniformly applied laboratory procedure. This is particularly true for DNA analysis of nuclei removed from paraffin tissue blocks. For the most part, quality control measurements including cell yields, the efficiency of extraction or disaggregation of aneuploid cells, exclusion of non-neoplastic cells, and other features characteristic of the scientific method are seldom included in studies of DNA analysis of solid tumors. Hopefully, a consensus regarding optimum technical and analytic methods will evolve, followed by the careful definition of human colon cancer cohorts. The importance of further studies is suggested by the results of this review which indicated that the presence of an aneuploid cell population is associated with a less favorable prognosis for all colorectal tumors.