'Missed' micrometastases--the extent of the problem.

The value of detecting micrometastases in patients with breast cancer has been debated for many years. The aim of this study was to determine whether and why such tumour deposits are missed at the time of reporting. The series comprised 272 patients treated surgically for breast carcinoma. For node-negative cases, the haematoxylin and eosin stained slides were re-examined. Those still remaining negative were stained with epithelial membrane antigen marker (EMA). Hilar sections were used in 76% of cases. Micrometastases were found in 35 cases reported as node-negative: 15 being identified on re-examination and 20 after staining with EMA, a gain of 44%, including 20 of embolic type. All were found in hilar sections of the nodes. The patients in whom micrometastases were found on further examination had significantly smaller tumour deposits than those reported as node-positive. In cases with infiltrating ductal carcinoma these presented as embolic growth, while those with infiltrating lobular carcinoma, for example, tended to colonize the nodal parenchyma, giving nodal growth. Differentiation between these growth patterns enables pathologists to distinguish between the dangerous embolic type and the less important nodal growth. In conclusion, many of these micrometastases can be detected if the slides reported as node-negative on first reading are re-examined. In those remaining negative, immunohistochemical staining is recommended.