[Minimal residual tumor in gastrointestinal carcinoma. Relevance to prognosis and oncologic surgical consequences].

Isolated tumor cells as a consequence of minimal residual disease are often not detectable by routine diagnostic procedures. However, before or after surgery, isolated tumor cells in lymph nodes, the peritoneal cavity, blood, or bone marrow can frequently be identified by immunohistochemical or molecular methods. Failure to reveal the presence of such cells results in under-staging of tumor patients and may constitute the source of unexpected tumor recurrence after radical surgery. These facts emphasize the importance of isolated tumor cells at least as a surrogate marker. The frequency of appearance of isolated tumor cells in different organ systems also depends on the type of primary tumor. Developments in modern detection methods have led to increasing sensitivity but at the expense of specificity. Isolated tumor cells demonstrate remarkable heterogeneity with respect to proliferative potential and tumorigenicity. This characteristic is also reflected by a striking variability in the expression of various genes conditioning the aforementioned biological behavior. Unfortunately there is also remarkable heterogeneity in methods used for sampling and processing patient material as well as for the enrichment and detection of isolated tumor cells. Despite the ongoing controversies concerning detection methods and biological significance of isolated tumor cells, several clinical trials providing data supporting the prognostic relevance of minimal residual disease should also be considered for gastrointestinal carcinoma. In future this finding should be integrated in the planning of trials in surgical oncology, and "minimal residual disease" should receive stronger attention as a stratification criterion in such clinical studies.