Histopathological detection of lymph node metastases from colorectal carcinoma

To evaluate whether the assessment of multiple sections from retrieved nodes yields an increased number of metastases compared with the number that would be detected by the commonly applied method of microscopy of a single section of lymph node only.

A prospective study of 72 colorectal carcinoma resection specimens. Lymph node sampling was based on the current guidelines for the detection of breast cancer metastases in axillary nodes. Lymph nodes up to approximately 5 mm in maximum extent were processed in entirety, without prior sectioning, and assessed histologically at three levels; larger lymph nodes were processed in entirety as multiple sections and histologically assessed at one level.

From a total of 72 carcinomas, eight were Dukes's A, 26 were Dukes's B, and 38 were Dukes's C. The mean and median numbers of nodes identified were 13 and 12, respectively (range, three to 44). Of the Dukes's C cases, four contained lymph node metastases identified by our method that might have gone undetected by the current, generally applied method. In one case, this led to the detection of the only nodal metastasis present and therefore "upstaged" the tumour from Dukes's B to C. On average, six extra tissue blocks were processed for each case in applying this method.

The assessment of multiple sections of lymph nodes from colorectal specimens leads to the detection of only a small number of additional nodal metastases. The method involves increased workload for pathologists and laboratory staff.