Urothelial dysplasia and carcinoma in situ (CIS) are related to recurrence and progression of urothelial carcinoma. Distinguishing CIS and dysplasia from reactive atypia is often difficult on the basis of histological features alone. Cytokeratin 20 (CK20), p53, and Ki-67 are related either to neoplastic change or prognosis in urothelial proliferations. The objective of the present study was to establish the immunohistochemical pattern of these three antibodies in urothelial dysplasia and CIS. Three groups of patients were evaluated: 40 nonneoplastic urothelial samples, 50 cases with histologically incontrovertible CIS, and 30 samples with nonconclusive atypical changes (atypia of unknown significance). Monoclonal antibodies (MoAb) against CK20, p53, and Ki-67 (MIB-1) were used on paraffin-embedded samples. Nonneoplastic urothelium showed no reactivity to CK20 except for umbrella cells; p53 and Ki-67 were negative or weakly positive in <10% of basal cells. In the CIS group, 42% showed positivity for all three MoAb; 44%, for two; and 14%, only for one. CK20 was positive through the full thickness of the urothelium in 72% of cases, p53 was positive in 80% of cases, and Ki-67, in 94% of cases. In the third group, the suspected dysplastic cells showed strong positivity in scattered cells through the epithelium in 75% of cases. Aberrant CK20 expression in urothelial cells plus overexpression of p53 and Ki-67 are indicators of dysplastic change in urothelial mucosa. Thus, immunohistochemistry is a useful tool to confirm the diagnosis of CIS and could be helpful to distinguish dysplastic changes from reactive atypia.

cdc25 is a family of cell-cycle phosphatases that activate the cyclin-dependent kinases. cdc25A and B, but not C, have oncogenic potential in vitro. In this study, we analyzed the possible implication of cdc25 genes in the progression of colorectal tumors. RNA and DNA were extracted from 34 paired tumor and normal colorectal tissues and examined by Northern blot, RT-PCR, and Southern blot, respectively. Protein expression was analyzed by Western blot in a subset of normal and tumor samples. The expression levels were correlated with the clinicopathologic characteristics and survival of the patients. cdc25B mRNA was overexpressed in 19 carcinomas (56%). A significant correlation was observed between high cdc25B mRNA levels and the relapse-free, overall, and cancer-related survival of the patients. The cdc25B2 splicing variant was detected in 27 carcinomas (79%) but only in 9 normal samples (26%) and was associated with the grade of the differentiation of the tumors. cdc25A mRNA was overexpressed in four tumors (12%) and cdc25C1 mRNA was overexpressed in nine tumors (26%). A new cdc25C2 splicing variant lacking exon 4 and 5 was identified in all of the tumors and in 56% of the normal samples. No amplifications or gene rearrangements of these genes were detected. In conclusion, these findings indicate that cdc25 isoforms and splicing variants are differentially regulated in colorectal carcinomas and may participate in the development of these tumors. Additionally, the correlation between cdc25B mRNA levels and the survival of the patients also suggest that the cdc25B isoform may be involved in the progression of the disease.

Three cases of epithelial-myoepithelial carcinoma (EMC) with coexisting areas of high grade carcinoma are reported. In two of the cases there was a previous recurrence, and in all three patients there had been a sudden increase in size before final surgery. The typical ductal and myoepithelial components of EMC showed the usual biphasic pattern and the expected immunophenotypes, with expression of wide spectrum cytokeratins, Cam 5.2 and EMA in the ductal part, and muscle-specific actin, smooth muscle actin, S-100 protein, vimentin and cytokeratins in the myoepithelial component. These areas also had a low mitotic count and low proliferation rate as measured by immunohistochemistry and by flow cytometry. Conversely, areas of high-grade tumour had the features of a large cell carcinoma, with focal mucin secretion in two cases. This high-grade component showed an epithelial immunophenotype in two cases, and was negative for all tested markers in the third one. The mitotic counts and the proliferation rates were much higher in these anaplastic areas. One of the patients died 3 months after treatment; another developed lymph node metastases 1 year later and was alive after 6 years of follow-up. The third patient was alive without evidence of disease 7 months after wide surgical resection of the tumour. The possibility of anaplastic transformation in EMC makes thorough sampling mandatory in this type of neoplasm.