Detection and characterization of circulating microsatellite-DNA and tumor cells in bone marrow in non-metastatic breast

3611 Background and purpose of the study: Recent studies on breast cancer (BC) have demonstrated the prognostic significance of disseminated tumor cells (DTC) in the bone marrow (BM). However, not every patient with DTC in the BM will relapse. In this study, we compared the presence of DTC in BM and lymph node staging with genetic aberrations in circulating free DNA derived from blood samples of primary, non-metastatic BC patients. Furthermore, we evaluated, whether the frequency and pattern of allelic losses in the plasma DNA might be suitable for the identification of BC patients who might have a higher risk for relapse. Methods: Bilateral BM (10 ml) aspirations of 60 BC patients at the time of diagnosis were assessed for DTC by a Ficoll density gradient centrifugation and an immunocytochemical cytokeratin (CK) assay with the anti-CK antibody A45-B/B3. For LOH (loss of heterozygosity) analysis, genomic DNA was extracted from cell-free plasma samples and amplified using a panel of 7 polymorphic microsatellite markers mapping to known tumor suppressor genes. Tumor staging was performed by the pathologists. Results: LOH aberrations were found in 14/60 (23%) blood samples. Among the informative cases, the frequency of LOH was highest for the chromosomal loci 3p23 (14%). Further frequent LOHs were detected at 13q12-13 (D13S218, 9%), 9p21 (D9S171, 8%), 10q23.3 (D10S1765, 8%) and 17q21-22 (D17S250, 7%). At the markers 17q21 (D17S855, 4%) and 16q22-23 (D16S421, 3%) LOHs were rarely observed. In 8/14 LOH positive cases, we also detected DTC in the BM. In 3 of these 8 patients, lymph nodes were staged N 1-2 and 5 patients were staged N 0 . In the other 6/14 LOH positive cases no DTC in BM samples were found and tumor spread to the lymph nodes could only be documented in one patient. In 11/46 LOH negative cases, no tumor load was found in the lymph nodes or in the BM whereas tumor spread to the lymph nodes (n=12) or to the BM (n=16) or to both compartments (n=5) was documented in 33/46 of these cases. Conclusions: Our results show that information on tumor specific genomic aberrations can be obtained by analysis of cell-free plasma DNA. Furthermore, these chromosomal loci harboring suppressor genes might be involved in tumor development. Thus, besides the prognostic value of DTC in BM samples, the detection and characterization of tumor-specific genomic aberrations in blood could be an additional tool to detect minimal residual disease in BC patients.