Reply to the Editor

We thank our colleagues Cheng and associates for their commentary with regard to our recently published article, “The Sentinel Node Concept in Adenocarcinomas of the Distal Esophagus and Gastroesophageal Junction.”1 In this study, we used the widely recognized immunohistochemical assay in which antibodies against epithelial-cell proteins are used to detect small clusters of tumor cells in histologically node-negative lymph nodes of patients with esophageal adenocarcinoma. Yao and Han claim that the monoclonal antibody against cytokeratin 8 and 18 (clone NCL-5D3) that we used in the present study should not be mistaken with the CAM5.2 antibody (clone CAM5.2) that appears to be specific for cytokeratin 7 and 8. In our article, we stated that the used antibody CAM 5.2 is specific for intracellular cytokeratin 8 and 18, referring to the original paper of Makin, Bobrow, and Bodmer,2 in which it is claimed that (at that time newly developed) antibody CAM5.2 identifies the lower molecular weight cytokeratins (cytokeratin 8, 18 and 19). Other studies that investigated the presence and relevance of micrometastases or isolated tumor cells by using antibody CAM5.2 have also referred to this article.3 However, over the last years, companies producing these antibodies have changed their products and further investigated the corresponding specificity in reactivity against certain cytokeratins, as pointed out by Yao and Han. Therefore, their statement that antibody CAM5.2 is not specific for intracellular cytokeratin 8 and 18 is correct. Nevertheless, this finding does not affect the conclusions of our study. The presence of cytokeratin 8 and/or 18 (which are both expressed in esophageal adenocarcinoma4) as detected with the antibody used in our study (clone NCL-5D3) indicates epithelial cell deposits in lymphoid tissue. Therefore, these cytokeratin deposits will still imply the presence of micrometastatic disease.