GATA3 expression has been reported in urothelial and breast carcinomas; however, the published data on GATA3 expression in tumors from other organs are limited. Immunohistochemical evaluation of GATA3 expression in 1,110 carcinomas and 310 cases of normal tissue using tissue microarray sections, 48 breast and bladder biopsy specimens, and 53 breast fine-needle aspiration biopsy specimens was performed. Sixty-two of 72 urothelial carcinomas (86%) and 138 of 147 breast carcinomas (94%) tested positive for GATA3. All other cases, except for 2 of 96 endometrial carcinomas, tested negative for GATA3. On fine-needle aspiration biopsy samples, 88% of primary breast carcinomas and 82% of metastatic breast carcinomas tested positive for GATA3. Our study revealed that GATA3 is a sensitive and specific marker for the diagnosis of breast and urothelial carcinomas. When working on a tumor of unknown origin, GATA3 should be routinely included in the initial screening panel if either a breast or urothelial primary tumor is suspected.

Prostate cancer is a biologically heterogeneous disease with considerable variation in clinical aggressiveness. The behavior of prostate cancer can be considered a direct or indirect result of aberrant alterations of gene expression in prostate epithelial cells. Identification of the patterns of gene-expression alterations that are related to the aggressiveness of prostate cancers will greatly assist the development of tools for early detection of prostate cancers with poor clinical outcome and identification of targets for future therapeutic intervention. To detect the patterns of gene-expression alterations of prostate cancers, we performed a comprehensive gene-expression analysis on 30 prostate tissues of various levels of invasiveness (ranging from those confined to the organ to distant metastases) and Gleason grades (combined scores 4-9), using the Affymetrix chip set Hu35k (A-D) and U95a. Following three sequential selection screens, we identified 84 largely novel genes and expressed sequence tag (EST) sequences whose expression levels were altered significantly in prostate cancer samples compared with control normal tissues. In addition, the expression levels of a group of 12 genes and EST sequences was found to be altered significantly in aggressive type of prostate cancers but not in organ-confined prostate cancers. Cluster analysis using the 84-gene list showed that the highly aggressive prostate cancers contained gene-expression patterns that were distinct from organ-confined prostate cancers.

Context.-Distinction of medullary carcinoma of the large intestine from other cytokeratin (CK) 7 À /CK20 À carcinomas can be challenging when working on a tumor of unknown primary because the majority of medullary carcinomas are negative for CK7, CK20, and CDX2.Objective.-To investigate the expression of cadherin-17 and SATB-2 and other markers in medullary carcinomas of the large intestine and cadherin-17 and SATB2 in a large number of carcinomas and normal tissues from various organs to further test their diagnostic specificity.Design.-This study evaluated cadherin-17 and SATB2 expression in 18 medullary carcinoma cases and 1941 tumors and 358 normal tissues from various organs. Other immunomarkers, including MLH1, PMS2, MSH2, MSH6, CDX2, CK7, CK20, TFF3, MUC4, calretinin, p504S, villin, and synaptophysin, were also tested on the 18 medullary carcinoma cases.Results.-The results demonstrated (1) loss of MLH1 and PMS2 in more than 80% of medullary carcinomas; (2) expression of cadherin-17 and SATB2 in 89% of medullary carcinomas; (3) focal expression of TFF3, MUC4, calretinin, CDX2, CK20, and synaptophysin in 72%, 72%, 67%, 67%, 28%, and 17% of 18 medullary carcinoma cases, respectively; and (4) expression of SATB2 and cadherin-17 in 97% and 98% of the colorectal adenocarcinomas, respectively, whereas their expression was seen in 3.6% and 3.3% of nongastrointestinal tumors, respectively.Conclusions.-We concluded that SATB2 and cadherin-17 were highly sensitive and specific markers for colorectal carcinomas and propose including MLH1, cadherin-17, and SATB2 in a routine immunostaining panel when working on a tumor of unknown primary, especially in an elderly patient with a CK7 À /CK20À carcinoma.