Flow cytometric minimal residual disease assessment in B-cell precursor acute lymphoblastic leukaemia patients treated with CD19-targeted therapies - a EuroFlow study

Abstract

The standardized EuroFlow protocol, including CD19 as primary B-cell marker, enables highly sensitive and reliable minimal residual disease (MRD) assessment in B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) patients treated with chemotherapy. We developed and validated an alternative gating strategy allowing reliable MRD analysis in BCP-ALL patients treated with CD19-targeting therapies. Concordant data were obtained in 92% of targeted therapy patients who remained CD19-positive, whereas this was 81% in patients that became (partially) CD19-negative. Nevertheless, in both groups median MRD values showed excellent correlation with the original MRD data, indicating that, despite higher interlaboratory variation, the overall MRD analysis was correct.

Keywords: acute leukaemia; diagnostic haematology; flow cytometry; minimal residual disease.

Fig 1

Gating strategy for MRD assessment of B‐cell precursor acute lymphoblastic leukaemia (BCP‐ALL) patients treated with CD19‐targeted therapies. After initial gating of nucleated, single cells and exclusion of debris and aggregates, an initial evaluation is based on the possible presence of CD10⁺ lymphocytes. If present, these cells are further evaluated for abnormal expression patterns (with help of reference images) to define whether they are ALL cells or not. If CD10⁺ lymphocytes were not present, further analysis is first focused on CD34. If CD34⁺ cells are present these can further be evaluated for abnormal expression patterns. Subsequently, also the CD34⁻ lymphocytes are being evaluated for aberrant expression patterns. In all cases, possible ALL cells are ultimately back‐gated on the forward scatter (FSC)‐side scatter (SSC) and CD45‐SSC plot to check that the cells form a uniform cluster.

Fig 2

Comparison of original MRD levels and MRD levels re‐analysed in phase 6. All MRD data were log10‐transformed before plotting and calculating correlation coefficients. Each symbol reflects the original MRD level versus the median of the re‐analysed MRD data by the participating centres (n = 9). In the targeted therapy group (left panel), one sample (5%) scored negative in the original analysis but was scored low level MRD positive in the re‐analysis. Molecular MRD data showed a level of 4 × 10⁻⁴ (Table SVI). In the chemotherapy group (right panel), two samples (11%) were scored negative in the re‐analysis but were originally scored positive (both at 0·01%). Molecular data were 2 × 10⁻⁴ for both samples (Table SVI). The correlation of the samples in which both MRD data were positive is shown in the plots (dotted line with equation and correlation coefficient).