[Detection of isolated tumor cells in bone marrow: methodological aspects. A critical re-appraisal and prospects]

Systemic dissemination of tumor cells occurring prior to the diagnosis or resection of the primary tumor is a prerequisite for relapse at distant sites and a major cause for death from cancer. Although detection of isolated tumor cells by sensitive immunocytochemical techniques is increasingly being considered as a clinically relevant prognostic factor, methodical objections including a questionable specificity of some of the antibodies applied, the heterogenity of the patient cohorts and the wide variation of the reported frequencies suggest that an independent prognostic impact of tumor cell detection in bone marrow at least for most of the tumor types still remains to be proven. When compared to non-morphological techniques immunocytochemistry is characterized by a lower false positive rate. So far a standarized application of this method still holds the most promise for clinical practice. Alternatively the genetic diagnosis of isolated tumor cells is hampered by a rather high false positive rate due to pseudogenes and illegitimate transcription. Appropriate negative controls in combination with a quantification of the PCR product are one possibility to overcome these difficulties. The clinical utility of genetic assays as a tool for the estimation of prognosis and management of cancer patients still remains an area of active investigation.