Refined detection of CD34⁺CD38⁻CD45RA⁺ leukemic stem cells using a single-tube flow cytometry assay and its strong association with measurable residual disease in acute myeloid leukemia: a retrospective cohort study
摘要
Leukemic stem cells (LSCs) are the cellular reservoir most strongly implicated in relapse of acute myeloid leukemia, yet their operational detection by multiparameter flow cytometry remains challenging because of immunophenotypic overlap with normal progenitors and variability across assays. Including CD45RA in the CD34⁺CD38⁻ gating strategy substantially improves discrimination between malignant and normal stem/progenitor populations and thus enables more precise LSC enumeration in a single-tube format. Given the clinical importance of accurately quantifying the LSC compartment, we evaluated a refined single-tube flow cytometry assay that incorporates CD45RA within the CD34 + CD38- gate to increase specificity for the leukemic stem compartment.
MethodsIn a retrospective cohort of 109 AML bone marrow samples, measurable residual disease (MRD) was assessed with a conventional three-tube, 8-color panel and LSCs were enumerated using an adapted single-tube, 8-color panel defining LSCs as CD34 + CD38-CD45RA + . To ensure analytical reliability we applied a formal lower limit of quantification (LLOQ), defined empirically as a cluster of 50 CD45 + events; samples below the sample-specific LLOQ were not called positive. Positivity thresholds were set at ≥ 0.1% for MRD and ≥ 0.004% for LSCs. Group comparisons used the Mann–Whitney U test and associations were quantified by Pearson correlation.
ResultsLSCs were detectable in 28/109 (25.7%) patients, while MRD positivity was observed in 37/109 (33.9%) patients. A robust association was demonstrated between LSC presence and MRD positivity (p = 0.00035). The LSC burden was significantly elevated in MRD-positive patients, and concomitantly, MRD levels were profoundly higher in patients harboring detectable LSCs (p = 2.01 × 10⁻⁹). A strong positive correlation was observed between LSC and MRD levels across the entire cohort (R = 0.66, p = 3.2 × 10⁻15). LSC and MRD status were independent of sex, FLT3, or NPM1 mutation status. Immunophenotypic profiling of the LSC compartment revealed predominant aberrant co-expression of CD33 (89.3%) and a multi-marker cocktail (89.3%), with CD44 (67.9%) and CD123 (53.6%) also frequently observed.
ConclusionsThe implementation of a refined LSC detection assay, leveraging CD45RA gating and a stringent LLOQ, yields a specific and clinically actionable quantification of the LSC reservoir in AML. The strong correlation between the CD34 + CD38-CD45RA + LSC subset and MRD status suggests its potential as a complementary biomarker for residual disease monitoring; however, prospective validation in outcome-annotated cohorts is required to establish its prognostic utility and clinical applicability.