CDX2 is a recently cloned homeobox gene that encodes an intestine-specific transcription factor, expressed in the nuclei of epithelial cells throughout the intestine, from duodenum to rectum. While expression of CDX2 protein in primary and metastatic colorectal carcinomas has been previously documented, neither the sensitivity nor the specificity of CDX2 expression, as determined by immunohistochemistry, for colorectal adenocarcinoma has been determined. We performed an immunohistochemical survey of 476 tumors with a monoclonal antibody, CDX2-88, including 89 tumors from the colon and duodenum and 95 tumors from other gastrointestinal sites, including the esophagus, stomach, pancreatobiliary system, gastrointestinal carcinoids, and liver. CDX2 was expressed uniformly (that is, in 76-100% of tumor cells) in all but one of the evaluated colorectal and duodenal tumors. High-level expression of CDX2 was also found, however, in mucinous ovarian carcinomas and adenocarcinomas primary to the urinary bladder of which 64% and 100% were positive, respectively. Gastric, gastroesophageal, and pancreatic adenocarcinomas and cholangiocarcinomas all showed similar, heterogeneous patterns of CDX2 expression. Most tumors in each group showed CDX2 expression by a minority of cells, whereas a substantial minority of cases in each group was completely negative and a smaller minority was uniformly positive. Gastrointestinal carcinoids gave similarly varied results, but the majority (58%) was negative. Hepatocellular carcinomas showed no expression of CDX2. Only very rare examples of carcinomas of the genitourinary and gynecologic tracts, breast, lung, and head and neck showed significant levels of CDX2 expression. In this study of primary and metastatic epithelial tumors, uniform CDX2 expression is demonstrated to be an exquisitely sensitive and highly, but incompletely, specific marker of intestinal adenocarcinomas. Compared with villin, a previously described marker of GI adenocarcinomas, CDX2 demonstrated superior sensitivity and comparable specificity. CDX2 expression can be seen, however, in selected non-GI adenocarcinomas such as mucinous ovarian carcinomas and adenocarcinomas of the urinary bladder.

HE HUMAN EPIDERMAL GROWTH factor receptor 2, HER-2 (HER-2/neu or c-erbB-2), oncoprotein is overexpressed in 20% to 25% of invasive breast cancers. 1 Previous studies have supported the importance of HER-2 overexpression as an independent prognostic marker of clinical outcome, 1-5 as a predictor of benefit from doxorubicin-based chemotherapy, 6,7 and for metastatic disease, for benefit from the anti-HER-2 antibody, trastuzumab. 8-11 Determination of HER-2 status is now an integral part of the clinical-pathological workup of breast cancer. The HER-2 protein overexpression is usually a direct consequence of gene amplification. 1,12 This unique geneprotein relationship has spawned several methods for assessing HER-2 status. Currently, immunohistochemistry and fluorescence in situ hybridization (FISH) are the most widely used, both because these technologies are ideally suited for routine and archival paraffinembedded tissue and are evaluated by direct visualization of tumor cells. Published gene-protein correlation studies in general show good-to-excellent concordance. 13-26 Initial studies evaluating which method best predicts clinical outcome have shown better clinical correlation with FISH than with immunohistochemistry assays. 15,27 Some clinical investigators have therefore advocated the use of FISH as the method of choice for HER-2 testing. 22,27,28 However, in many of these assessments, im-munohistochemistry assays were inadequately optimized, rendering comparison between these differing

SummarySpecific antibodies to tissue factor pathway inhibitor (TFPI) were used in immunohistochemical procedures to determine the distribution of TFPI in normal and neoplastic human tissues. TFPI was restricted to megakaryocytes and the endothelium of the microvasculature in normal and abnormal tissues, but was not found in the endothelium of larger vessels or in hepatocytes. TFPI was also detected in macrophages in the villi of term placenta. Tumor-associated macrophages in several types of malignancy that we have shown previously to express a complete tissue factor-initiated pathway of coagulation and thrombin generation also manifested TFPI. By contrast, malignant cells in small cell carcinoma of the lung, renal cell carcinoma, and malignant melanoma that we have shown previously to express coagulation factors together with tumor cell-associated fibrin formation failed to stain for TFPI. We postulate that TFPI may be lacking from the latter malignancies because of the absence of the appropriately configured tissue factor – factor VII a – factor Xa complex required for TFPI binding.

We compared the effectiveness of polymerase chain reaction (PCR) and DNA blot analysis (DBA) for detecting clonal T-cell populations and investigated whether a nonradioactive PCR method could be used in routine clinical diagnosis. We analyzed DNA from 117 cases for T-cell clonality by PCR amplification. DBA was performed on 77 of these cases. Denaturing polyacrylamide gel electrophoresis (PCR-PAGE) of radiolabeled PCR products and capillary electrophoresis (PCR-CE) of fluorescently labeled PCR products were used for PCR product separation and quantitation. Complete agreement was obtained between PCR-PAGE and DBA in 67 of 77 cases. One case was positive by DBA and negative by PCR-PAGE, and 3 cases were positive by PAGE and negative by DBA. Five cases indeterminate by DBA were positive by PCR-PAGE, and 1 indeterminate case was negative by PCR-PAGE. In the comparison of PCR-PAGE and PCR-CE, of 63 cases with height ratios less than 2.0, all were negative by PCR-PAGE. Of 52 cases with height ratios of 2.0 or more, 50 were positive by PCR-PAGE. We conclude that PCR-CE is analytically equivalent to DBA and PCR-PAGE for detecting clonal T-cell populations. The PCR-CE method is semiquantitative and, therefore, may be more objective than gel-based methods.