FISH+CD34+CD38- cells detected in newly diagnosed acute myeloid leukemia patients can predict the clinical outcome
Institute of Hematology, Changhai Hospital, Second Military Medical University, Shanghai 200433, China
Journal of Hematology & Oncology 2013, 6:85 doi:10.1186/1756-8722-6-85Published: 7 November 2013
In acute myeloid leukemia (AML), the leukemia initiating cells (LICs) or leukemia stem cells (LSCs) is found within the CD34+CD38- cell compartment. The LICs subpopulation survives chemotherapy and is most probable the cause of minimal residual disease (MRD), which in turn is thought to cause relapse. The aim of this study was to determine the prognostic value of the percentage of LICs in blasts at diagnosis.
Design and methods
The percentage of LICs in the blast population was determined at diagnosis using a unique Flow-FISH analysis, which applies fluorescent in situ hybridization (FISH) analysis on flow cytometry sorted cells to distinguish LICs within the CD34+CD38- cell compartment. Fourty-five AML patients with FISH-detectable cytogenetic abnormalities treated with standardized treatment program were retrospectively included in the study. Correlations with overall survival (OS), events-free survival (EFS) and cumulative incidence of relapse (CIR) were evaluated with univariate and multivariate analysis.
The percentage of LICs is highly variable in patients with acute myeloid leukemia, ranged from 0.01% to 52.8% (median, 2.1%). High LIC load (≥1%) negatively affected overall survival (2-year OS: 72.57% vs. 16.75%; P = 0.0037) and events-free survival (2-year EFS: 67.23% vs. 16.33%; P = 0.0018), which was due to an increased cumulative incidence of relapse (2-year CIR: 56.7% vs. 18.0%; P = 0.021). By multivariate analysis, high LIC load retained prognostic significance for OS and EFS.
In the present study, we established the Flow-FISH protocol as a useful method to distinguish normal and leukemic cells within the CD34+CD38- cell subpopulation. The high percentage of LICs at diagnosis was significantly correlated with increased risk of poor clinical outcome.