Detection of residual neuroblastoma cells in bone marrow: Comparison of flow cytometry with immunocytochemistry

BACKGROUND: Because the cytomorphologic examination of bone marrow (BM) aspirates appears not sensitive enough to detect residual neuroblastoma cells, two four-color flow cytometric assays using different combinations of CD9, CD81, CD56, CD45, and anti-GD2 were evaluated. METHODS: The sensitivity of the flow cytometric assays was assessed by spiking experiments in normal peripheral blood samples. Twenty-eight BM samples, 12 biopsies, and 3 peripheral blood stem cell (PBSC) preparations from 22 patients with neuroblastoma were analyzed. The results were compared with those of an anti-GD2 immunocytochemical reference assay. RESULTS: Flow cytometric and immunocytochemical analyses showed residual neuroblastoma cells in four BM samples. One PBSC preparation and 20 BM samples were negative for both assays. Four BM and two PBSC samples scored positive for the immunocytochemical assay but were negative for the flow cytometric tests. This was due to the limited number of cells that were flow cytometrically analyzed. A strong correlation between the flow cytometric and immunocytochemical tests was found (chi2 = 6.4, P = 0.011). CONCLUSIONS: When an equal amount of cells is analyzed, the sensitivity of the flow cytometric assays is to be about 10 times lower than that of the immunocytochemical test. However, the flow cytometric assays can be used to screen for residual cells in clinical samples with a sensitivity of one neuroblastoma cell in 10(4) to 10(5) normal mononuclear cells. Flow cytometry is simple, quick, and cost effective compared with immunocytochemistry. In addition, the flow cytometric assays can be used to screen for residual neuroblastoma cells in case of a GD2-negative primary tumor. Therefore we recommend flow cytometry for the detection of residual neuroblastoma cells.