Myelodysplastic Syndrome Stem Cells Evaluated By Multiparameter Flow Cytometry Are Associated with Posttreatment Relapse: A Prospective Study

Background: We previously showed that leukemia stem cell (LSC)-based measurable residual disease (MRD) assay (Leukemia,2016;30:439-446), mainly including a cocktail panel of CD7, CD11b, CD22, CD56, CD366, and CD371 on CD34⁺CD38^- cells, is superior to traditional multiparameter flow cytometry (MFC) method for outcome prediction in acute myeloid leukemia patients who received allograft (Blood,2022;140:516-520). However, whether the cocktail panel for LSC could be applied to patients with myelodysplastic syndrome (MDS) for MDS stem cell detection ? What about the transcriptomic features of MDS stem cells ? Whether the MDS stem cell-based MRD assay is superior the traditional MFC approach for forcasting disease recurrence ? To answer these questions, we first investigated the single-cell transcriptome characteristic of MDS stem cell subgroups defined by the expression of membrane aberrancies (CD7, CD11b, CD22, CD33, CD44, CD45RA, CD56, CD123, CD366 or CD371) on CD34⁺CD38^- cells. We also prospectively compared the traditional MFC method and the MDS stem cell-based MRD assay in allograft modality for MDS cases.

Methods: In this study, a total of 143 MDS patients who received allogeneic stem cell transplantation (allo-SCT) between November 2021 and December 2023 were prospectively enrolled. The transcriptomic features of MDS stem cells were investigated using the BD Rhapsody™ single-cell analysis system. The MDS stem cells were evaluated by MFC according to the LSC-based assay before and after transplantation.

Findings: In the present study, we found that the stem cell signature was highly enriched in CD34⁺CD38^-cocktail⁺ MDS stem cells compared with CD34⁺CD38^- donor stem cells. The patients were classified as MDS stem cell positive group (group A) and MDS stem cell negative group (group B) according to a cutoff value of CD34⁺CD38^-cocktail⁺ MDS stem cells determined pre-SCT as 0.1064%. Patients in group A had a higher 1-year cumulative incidence of relapse (CIR, 18.3% vs. 3.6%, P=0.005) and a trend of inferior 1-year disease-free survival (DFS, 78.1% vs. 87.9%, P=0.087) than those of group B. Multivariate analysis showed that MDS stem cell positivity before transplantation could independently predict CIR (HR=6.173, P=0.015). However, we did not found the association of MDS cells evaluated by traditional MFC method with posttransplantation relapse. The subjects were classified as MDS stem cell positive group (group C) and MDS stem cell negative group (group D) according to a cutoff value of CD34⁺CD38^-cocktail⁺ MDS stem cells detected post-SCT as 0.0081%. Subjects in group C had a higher 1-year CIR (37.7% vs. 1.0%, P<0.001) and an inferior 1-year DFS (59.6% vs. 91.9%, P<0.001) than those of group D. Multivariate analysis showed that MDS stem cell positivity after transplantation could independently predict CIR (HR=47.455, P<0.001) and DFS (HR=6.365, P<0.001). The results also showed that the 1-year CIR of patients in the traditional MRD positive group post-SCT was significantly higher than that of patients in the traditional MRD negative group (65.7% vs. 6.28%, P<0.001), leading to lower 1-year DFS (34.3% vs. 87.4%, P<0.001). Compared with MRD detected by the traditional MFC method, the CD34⁺CD38⁻cocktail⁺ MDS stem cell-based residual disease assay for post-transplantation MRD evaluation had a high sensitivity (94.4% vs. 50%), C-index (0.834 vs. 0.710) and Youden index (0.728 vs. 0.452).

Conclusion: Our results for the first time suggest that a cocktail panel of LSC can be used for determining MDS stem cells to evaluate residual disease. Our data also indicate the superiority of MDS stem cell-based MRD assay to traditional MFC MRD method, such as a high sensitivity and C-index, for relapse prediction of patients who received allograft.