Background & aims: Immunohistochemical detection of bone marrow micrometastases has been reported as a prognostic marker in colorectal cancer. The aims of this study were to evaluate the potential advantage of flow cytometry as an objective method of identifying and quantifying micrometastatic deposits within bone marrow and to determine the prevalence and quantity of micrometastases in patients undergoing surgery for gastrointestinal cancers.
Methods: Flow cytometry was first validated by a controlled "spike" experiment in which varying numbers of neoplastic epithelial cells were added to bone marrow, and cytometry was performed in a blinded fashion. Three neoplastic cell lines (colonic and esophageal) with varying degrees of expression of cytokeratin-18 were used. Epithelial cells were detected by dual staining with fluorescence-labeled, monoclonal anti-cytokeratin, and propidium iodide.
Results: Cytometry reproducibly detected the presence of > or = 10 neoplastic cells per 10(5) marrow cells. Micrometastases were found in 20%-30% of patients undergoing potentially curative resection of colorectal and gastroesophageal adenocarcinomas. There was a trend toward increasing positivity for marrow deposits with advanced Dukes' staging of colorectal cancer.
Conclusions: Flow cytometric assessment of bone marrow is a reliable, objective, and quantitative method of detecting micrometastatic deposits found in a substantial subset of patients undergoing surgery for gastrointestinal adenocarcinomas.