Identification of residual leukemic cells by flow cytometry in childhood B-cell precursor acute lymphoblastic leukemia: verification of leukemic state by flow-sorting and molecular/cytogenetic methods

Abstract

Reduction in minimal residual disease, measured by real-time quantitative PCR or flow cytometry, predicts prognosis in childhood B-cell precursor acute lymphoblastic leukemia. We explored whether cells reported as minimal residual disease by flow cytometry represent the malignant clone harboring clone-specific genomic markers (53 follow-up bone marrow samples from 28 children with B-cell precursor acute lymphoblastic leukemia). Cell populations (presumed leukemic and non-leukemic) were flow-sorted during standard flow cytometry-based minimal residual disease monitoring and explored by PCR and/or fluorescence in situ hybridization. We found good concordance between flow cytometry and genomic analyses in the individual flow-sorted leukemic (93% true positive) and normal (93% true negative) cell populations. Four cases with discrepant results had plausible explanations (e.g. partly informative immunophenotype and antigen modulation) that highlight important methodological pitfalls. These findings demonstrate that with sufficient experience, flow cytometry is reliable for minimal residual disease monitoring in B-cell precursor acute lymphoblastic leukemia, although rare cases require supplementary PCR-based monitoring.

Figure 1.

Schematic presentation of the BM samples and flow-sorted cell populations studied as well as the results from PCR/FISH-analyses in sorted cell populations. Two or more cell populations were sorted from each BM sample. Numbers in the boxes are: number of BM samples from which the flow-sorted cell populations derived (number of patients)/number of flow-sorted cell populations. E.g. in total 255 cell populations, deriving from 53 BM samples from 28 patients, were investigated.

Figure 2.

FC dot plots from discrepant cases with unexpected PCR/FISH-results in flow-sorted cell populations. Left hand plots are from diagnostic BM samples, while right hand plots are from follow-up BM samples with sort gates shown. All plots are gated on CD19^pos cells after excluding dead cells and debris based on FSC/SSC characteristics. Leukemic cells are shown in blue and non-leukemic B-cells in purple. The dashed circles indicate the plausible location of MRD cells (cases A and B), and areas with potential occurrence of MRD cells (cases C and D), respectively. (A) Risk of MRD underestimation due to partly informative LAIP at diagnosis and persistence of CD34^neg cells (or downmodulation of CD34) (Case #1). (B) Risk of MRD underestimation due to signification antigen modulation (CD10 downmodula-tion and CD20 upmodulation) (Case #2). (C) Risk of MRD overestimation in CD10^neg/CD20^neg BCP-ALL patient due to persistence of plasma cells (Case #3). (D) Risk of MRD overestimation due to dead cells in dot plot diagonal region (Case #4) (Online Supplementary Appendix).