Colorectal 13–22

Aims: Current methods for performing sentinel lymph node mapping (SLNM) in colorectal cancer require timeconsuming analysis of the specimen both at the time of mapping and during the formal pathological examination. We aimed to streamline this process to facilitate its implementation in the routine hospital setting. Methods: With informed consent and ethical approval we prospectively applied a novel, ex vivo SLNM technique to 23 primary colorectal cancers. Lymphazurin blue dye (1 ml) was injected quadrantically around tumours by a surgeon within 10 minutes of resection. Specimens were then immediately placed into formalin and left to be processed routinely. At processing, blue-stained nodes were placed in individual cassettes and their archival codes were noted. These codes were used to retrieve sentinel nodes and multilevel step sectioning and cytokeratin immunohistochemistry were performed. Results: Sentinel nodes were found in 23/23 patients (100%) with an average of 3·4 (range, 1 to 7) per specimen. They correctly predicted lymphatic bed status in 9/11 (82%) node-positive patients. Micrometastatic disease was found in 6/12 patients deemed node-negative on routine reporting, two of which had T1 tumours. Conclusions: For the first time we have shown that sentinel nodes are still identifiable in formalin-fixed specimens after ex vivo SLNM with a sensitivity and specificity comparable to previously described methods. A single pathologist therefore may now do all necessary processing at the one time. SLNM may now be easily made a routine procedure with no patient risk. Colorectal 14