Multicolor Immunophenotyping of Candidate Leukemic Stem Cell Markers in CD34+CD38- Chronic Myeloid Leukemia Stem Cells

Introduction: Detection of leukemic stem cells (LSCs) may represent a new potential prognostic parameter in chronic myeloid leukemia (CML) and a tool for minimal residual disease monitoring in combination with standard qPCR method. To date CD26 is studied as a most specific marker for CML LSC detection. Several other candidate LSC markers have been reported, such as IL1-RAP, CD25 and CD93, however a side-by-side testing of their specificity is lacking. Recently, we have identified CD69 molecule to be overexpressed in CD34+CD38-CD26+ cells, which makes this antigen another candidate marker for LSC.

Aim: To compare the surface expression of LSC markers CD69, CD26, CD25, CD56, IL1-RAP, CD56, CD93 in bulk CD34+CD38- population at diagnosis using a multicolor phenotypization assay

Methods: In total, 44 patients were analyzed at diagnosis of chronic phase CML before administration of any treatment. Fresh (n=38) or cryopreserved (n=6) leukocytes obtained by erythrolysis were stained with CD45, CD34, CD38, CD25, CD26, CD56, CD69, CD93, IL1-RAP antibodies and 7-AAD for selection of live cells. Analysis was performed on FACSAria Fusion (BD Biosciences). Acquisition of live mononuclear cells ranged from 2×10⁴ to 2.5×10⁶.

Results: Expression of candidate LSC markers CD26, CD25, CD56, IL1-RAP, CD56, CD93 was analyzed in BM of 35 patients who carried at least 30 CD34+CD38- cells. Median percentage of marker positive cells was 34% for IL1-RAP, 31% for CD25, 23% for CD26, 16% for CD56 and 2% for CD93, from the parent CD34+CD38- population. Next we analyzed the overlap and combination for the three best markers - IL1-RAP, CD25 and CD26. The 3-combination (defined as IL1-RAP or CD25 or CD26 expression) identified 40% of CD34+CD38- positive cells, which was more than any of the markers alone. Expression of the three markers showed good overlap and ruled out mutually exclusive expression of the markers. This was demonstrated by median difference of 0.4% of CD34+CD38- cells (range: 0-18%) and a correlation coefficient r2=0.9914, when comparing the 3-combination and the best performing marker in each patient. In contrast, in 12/35 (34%) of patients, one of the three markers failed to identify at least half of the cells positive for another marker.

In 21/35 patients, we also analyzed the expression of CD69 in the CD34+CD38- compartment. The CD69 showed similar performance as the 3-combination of CD26, CD25, and IL1-RAP, 59% vs 54% positive cells, respectively. We observed excellent overlap between CD69 and 3-combination expression in individual patients with median difference of 0% of CD34+CD38- cells (range 0-15%) and correlation coefficient r2=0.9890.

Furthermore, we compared marker positivity in BM vs PB in 19 paired samples. Both, sample types showed similar frequency of CD34+CD38- cells (5×10^-3 in BM, and 2×10^-3 for PB), but PB carried higher percentage of LSCs identified by the 3-combination - median 76 vs 52% cells.

Conclusions: We show an overlap in surface expression of three previously reported CML LSC markers - IL1-RAP, CD25 and CD26. Nevertheless, a combination of these markers can detect more positive cells than any of the markers alone. Moreover, we demonstrate that CD69 identifies the same cells within the CD34+CD38- compartment as the combination of three above mentioned markers, which makes CD69 the best candidate for routine CML LSC quantification.