[Show abstract][Hide abstract] ABSTRACT: Admixture of normal and neoplastic cells is a serious problem in the evaluation of tumor cell kinetic parameters by flow cytometry, in particular for DNA diploid tumors. The admixture of non-neoplastic cells, such as stromal cells and inflammatory cells, can disturb the estimation of the proliferative tumor fraction. This problem has been addressed in fresh tumor samples by applying bivariate flow cytometric analyses for DNA and cytokeratin. We have adapted this approach for formalin-fixed and paraffin-embedded tissue samples of colorectal carcinomas. After preparation of a single cell suspension from paraffin blocks by means of an enzymatic digestion step, the cells of epithelial origin were selectively stained with a panel of subtype specific cytokeratin antibodies. DNA analysis could thus be performed on the cytokeratin-positive cells. The proliferative fractions of the paraffin-embedded samples could be compared with those of the fresh tissue samples and a very good correlation was seen between DNA indices from fresh and paraffin-embedded material. As expected, after gating on the cytokeratin-positive cells an enrichment of the S-phase fraction was seen compared with the ungated cell population. However, this enrichment was more pronounced in the cell suspensions derived from the paraffin-embedded part of the tumor compared with the fresh disaggregated, ethanol-fixed part of the tumor.