Minimal residual disease assessment following CD19-targeted therapy in B-cell precursor acute lymphoblastic leukemia using standardized 12-color flow cytometry: A EuroFlow study

Abstract

Detection of minimal/measurable residual disease (MRD) is a critical prognostic marker in B-cell precursor acute lymphoblastic leukemia (BCP-ALL). The EuroFlow Consortium previously developed an 8-color flow cytometric MRD protocol, effective for >98% of BCP-ALL patients treated with chemotherapy. This study aimed to enhance MRD detection, particularly for patients treated with CD19-targeted therapies, by expanding the EuroFlow protocol to a 12-color panel. This new panel incorporates additional B-cell markers and exclusion T/NK-cell markers (CD3 and CD7). Through an evaluation of 237 diagnostic BCP-ALL samples, CD22, CD24, and HLA-DR were selected as additional B-cell gating markers. Two 12-color tubes were technically optimized and clinically validated across 101 patient follow-up samples, demonstrating excellent concordance with molecular MRD levels (R ² = 0.88). The 12-color BCP-ALL MRD tubes were compatible with the previously developed 8-color automated gating and identification (AGI) tool and demonstrated good reproducibility. Our findings indicate that the 12-color panel performs comparably to the 8-color BCP-ALL MRD panel, including both CD19-positive and CD19-negative cases. However, it offers an improved definition of the B-cell lineage, particularly for expert-guided manual data analysis, and provides additional information on the expression of the targetable marker CD22.

Figure 1

Marker expression in normal BCPs and BCP‐ALL patients at diagnosis. Marker expression in BCP‐ALL patients stained with the EuroFlow ALOT and the BCP‐ALL Dx panel (n = 237) expressed as: (A) the percentage of positive BCP‐ALL cells (mean fluorescence intensity, MFI > 1000) for each marker and (B) the percentage of patients with more than 80% positive BCP‐ALL cells (MFI > 1000). (C) Expression of CD58 in three B‐cells subsets (CD34+, CD34−/CD10+, CD20+/CD10−) in a representative normal bone marrow, measured using the BCP‐ALL Dx panel. (D) Expression of CD22 and CD24 in BCP‐ALL patients (n = 237). T‐ and NK‐cells are shown as negative control in blue, whereas BCP‐ALL cells are shown in red; data are presented as median MFI of each patient. Curves represent 1 and 2 SD. Using an arbitrary cut‐off for positivity of 1000, the vast majority of BCP‐ALL cells patients were CD22⁺ CD24⁺; some were only positive for CD22 (n = 42; 18%) or CD24 (n = 18; 8%) and 8 (3%) were negative for both markers.

Figure 2

Correlation between MRD data obtained by the 12‐color EuroFlow BCP‐ALL MRD panel and molecular MRD data. (A) Flow cytometric MRD data from the tube with the highest MRD level is shown (Flow MRD). Molecular MRD levels (Mol MRD) are categorized as positive, positive but not quantifiable (PNQ), or negative (Neg). Dashed lines indicate the lower limit of detection (LLOD) and the limit of quantification (LOQ). Black dots represent CD19‐positive MRD or MRD negativity, while red dots represent CD19‐negative MRD cases where enough cells to perform both molecular and flow cytometry analyses were available. The correlation is calculated by Pearson's correlations from log10 transformed MRD levels of the concordant positive files. (B) Bland–Altman analysis of quantitative MRD data obtained by the 12‐color EuroFlow BCP‐ALL MRD panel and molecular MRD data. Original MRD data were log10 transformed and differences (MRD by molecular methods – MRD by flow) versus average of MRD of both methods are shown. The data show a slight difference of 0.34log (MRD by molecular methods 0.34 higher than MRD by flow; in line with previous comparisons (Denys et al., 2013), but do not show any proportional bias. Black dots represent CD19‐positive MRD, while red dots represent CD19‐negative MRD. Dashed lines indicate mean of difference (bias) or 1.96 standard deviation (95% confidence interval).

Figure 3

Correlation between data obtained by the 8‐color BCP‐ALL MRD panel and the 12‐color BCP‐ALL MRD panel, analyzed using the 8‐color AGI tool. Fourteen CD19‐positive cases (with or without parallel molecular MRD analyses) were stained with both the 12‐color tubes and the 8‐color tubes and analyzed by the 8‐color AGI tool by the same expert. Correlations between log10 transformed MRD levels obtained by manual analysis and AGI tool with Pearson R ². Neg, MRD negative.

Figure 4

8‐color AGI‐tool‐based analysis of CD19‐negative BCP‐ALL MRD samples (n  = 14). (A) FCS files of 14 BCP‐ALL patients with CD19‐negative MRD were analyzed manually and by 8‐color AGI tool by the same daily user. (B) Correlations between MRD data, obtained by two 8‐color AGI‐tool daily users. (C) MRD data obtained by different non‐daily AGI tool users (red dots) and two daily users (green dots). The horizontal bar represents the median of the five measurements (%CV of log10 transformed MRD levels: 1%–341%, median: 29%). Neg, MRD negative.

Figure 5

An example of a case with very low levels of CD19‐negative BCP‐ALL cells in the bone marrow. The BCP‐ALL patient was treated with CD19 CAR‐T cells and subsequently developed a CD19‐negative CNS relapse that was confirmed by cytomorphology and RQ‐PCR. Legend population: The dark blue population represents BCP‐ALL cells (64 events; 0.0006% of leukocytes), whereas the light gray population represents normal leukocytes. The BCP‐ALL cells are CD19‐negative, CD34‐negative, dim CD22‐positive, CD24‐positive, CD66c/CD123‐positive, dim CD45‐positive, CD10‐positive, and HLADR‐positive. File number represents the tube number.