Interference of bone marrow CD56+ mesenchymal stromal cells in minimal residual disease investigation of neuroblastoma and other CD45-/CD56+ pediatric malignancies using flow cytometry

Background Bone marrow (BM) samples obtained from minimal residual disease (MRD)-negative children with B-cell acute lymphoblastic leukemia (B-ALL) were used in our laboratory as negative biological controls for the development of a neuroblastoma (NBL) flow-cytometric (FC) protocol. The accidental, but systematic, identification of rare cell populations (RCP) mimicking NBL cells (CD45(-)/CD56(+)) in these samples indicated the need for their thorough immunophenotypic identification, in order to elucidate their possible interference in NBL-MRD assessment. Procedure RCP observed in BM samples from 14 children recovering from BM aplasia due to intensive chemotherapy for B-ALL were investigated with the following markers: CD81, CD200, CD24, GD2, CD73, CD13, CD90, CD146, CD9, CD117, CD10, CD99, and NG2. BM samples from six newly diagnosed patients with NBL and an NBL cell line were simultaneously investigated as positive controls. Results The frequency of RCP in B-ALL BM samples was < 1/1 x 10(4) cells (bulky lysis), and their immunophenotypic profile was indicative of CD56(+) mesenchymal stromal cells (MSCs) (CD45(-), CD90(+), CD146(+), CD73(+)). Also, RCP expressed CD81 and CD200, simulating NBL cells. The most useful discriminative markers for CD56(+) MSCs were CD13 and CD73. An appropriate protocol consisting of two tubes with seven color combinations was further proposed: SYTO-16, GD2 (first tube) or CD73 (second tube)-PE, CD24-ECD, CD13-PC5.5, CD45-PC7, CD81-APC, and CD56-APC700. Conclusions RCP that were immunophenotypically similar to NBL were identified as CD56(+) MSCs. As these cells might pose an obstacle to accurate NBL disease assessment by FC, especially MRD, an enhanced NBL-FC protocol is proposed for prospective evaluation.