Specific Detection of Aberrant and Normal Stem Cells in Acute Myeloid Leukemia Patients Opens the Way for Defining Highly Specific Targets for Stem Cell Therapy.

In acute myeloid leukemia (AML), a small fraction of blast cells contains the tumor initiating cells, further referred to as leukemic stem cells (LSCs). LSC resemble hematopoietic stem cells (HSCs) with respect to self renewal capacity and quiescence. Therefore, LSCs are proposed to be therapy resistant. In order to optimally target LSCs and sparing HSC and to monitor therapy, discrimination between LSC is HSC is required. We showed that within the CD34+CD38− stem cell compartment, LSCs can be discriminated from HSC by aberrant expression of markers, including lineage markers like CD7, CD19 and CD56 and the novel LSC marker CLL-1 (van Rhenen et al., Leukemia 2007 and Blood 2007). Too low aberrant marker expression, however, hampers discrimination in part of the cases. Therefore, we investigated additional parameters that would allow to distinguish LSCs from HSCs in CD34 positive AML patients. In 14 out of 48 cases studied, flow cytometry revealed a double population within the CD34+CD38− compartment, characterized by a small but clear difference in forward scatter (FSC, reflecting cell size) and sideward scatter (SSC, reflecting granularity). In 7/14 cases with high marker expression, FSC^highSSC^high population coincided completely with the population with aberrant marker expression. In the other cases, marker expression was too moderate to show a complete overlap. The FSC^lowSSC^low population within the CD34+CD38− stem cell compartment is the minor population at diagnosis (median 16%, range 0.2%–92%; n=14), had no expression of aberrant markers and, moreover, closely resembled the FSC/SSC characteristics of normal BM HSCs. In addition, in these patients, the normality of the FSC^lowSSC^low population was also supported by the fact that the CD34 and CD45 antigen density was similar to that of normal BM HSCs. Altogether, this enabled to use FSC/SSC characteristics together with aberrant CD34 and CD45 expression to discriminate between LSC and HSC in cases with low or absent aberrant marker expression (8/48). In addition, the malignant character of the FSC^highSSC^high population and the normal character of the FSC^lowSSC^low population could be proven in three AML patients with cytogenetic aberrancies. Patient 1 had a t(8;21) translocation and presented with a CD34+CD38−- population that was CD19 positive (81% of the stem cell compartment) and had FSC^highSSC^high properties. FACSsorted cells contained the translocation in 90% of the cells. The CD19 negative population (19% of the stem cell compartment) had FSC^lowSSC^low characteristics and contained 0% t(8;21) cells. In two other AML cases with a cytogenetic aberrancy (t(8;21) and t(15;17), respectively), FSC/SSC characteristics, CD34/CD45 antigen density and aberrant marker expression (CD56 in one case and CLL-1 in the other) were partly overlapping (estimated LSC contribution to the CD34+CD38− compartment was 85% in both cases). Cell sorting on the highest FSC/SSC and marker expression nevertheless resulted in enrichment of cytogenetically aberrant cells (63% and 73%, respectively), while the corresponding FSC^lowSSC^low cells, which missed CD56 and CLL-1 expression, were enriched for cytogenetically normal HSCs (87% and 67%, respectively). The marker and scatter parameters discussed above have generated the possibility to discriminate between LSCs and HSCs and now allows specific detection of LSC in >75 % of the patients. Discrimination between LSCs and HSCs in AML might not only facilitate to establish the therapeutic window of current therapies in terms of LSC specificity, but also allow the identification of new highly AML stem cells specific therapeutic targets. This should ultimately result in more selective therapies, which would be highly effective for AML stem cells, while leaving the normal HSC intact. This work was supported by Netherlands Cancer Foundation KWF.