Combination of CD34/CD38 Immunophenotypes and Side Population (SP) Reveals the Putative Leukemia Stem Cell/Leukemia Initiating Cell In Acute Myeloid Leukemia.

Abstract 1582

Although the originally defined CD34+CD38- leukemia stem cell (LSC)/leukemia initiating cell (LIC) in acute myeloid leukemia (AML) still serves as a lead in studies on the characterization of LSC/LIC, in more immuno-compromised mouse models, other stem cell immunophenotypes, i.e. CD34-CD38- and CD34-CD38+ turn out to have LSC/LIC activity (Taussig et al, Blood 2010; 115: 1976). Two functional LSC phenotypes offer the ability not only to restrict the LSC/LIC compartment to a lower frequency compartment, but also to include such CD34- negative immunophenotypes. These concern high activity of aldehyde dehydrogenase (ALDH) or high efflux of Hoechst 44432 (resulting in side population, SP). The present study deals with the possible overlap between CD34/CD38 and SP defined stem cell compartments. Based on our previous results (van der Pol, Haematologica 2003; 88: 983), we defined truly CD34 negative AML as AML with a very small CD34+ population only (usually≤1%), with particular scatter properties, and proven to be of normal origin. Using expression of cell surface markers present on LSC/LIC, but absent on HSC (van Rhenen et al, Leukemia 2007; 21: 1700; Blood 2007; 110: 2659), the LSC/LIC compartment in this type of leukemia (5 cases studied) was shown to be present in the CD34-CD38+ compartment, while the HSC was found in the CD34+CD38- compartment. When determining aberrant cell surface markers within the SP, the only malignant cells present in the SP population were CD34-CD38+. To assess the primitive character of the putative LSC/LIC sub-populations, cell sorting experiments were performed in which all aberrant marker positive SP cells along with marker positive non-SP cells (n=2) and all aberrant marker negative SP cells and marker negative non-SP cells (n=2), were assessed in the liquid culture stem cell assay. SP marker positive cells had 280- and 725-fold more stem cell activity compared to non-SP cells (14,500 versus 20 and 8,400 versus 30, respectively), while for marker negative cells only the SP fraction contained stem cell activity (8,000 and 8,200 colonies per million input cells). Even though the frequency of SP cells is far below that of non-SP cells, the absolute number of colonies was higher in SP than in non-SP cells. In CD34 positive leukemia, in contrast to CD34 negative AML, the SP fraction was largely filled with cells with all combinations of CD34 and CD38, which all fulfill the criteria (marker positivity) for malignancy. Liquid culture assay results were: for marker positive cells, colonies were formed only in the SP compartment in 2/3 cases studied (6,500 and 4,400 colonies per million input cells). In the remaining case colony formation was 195-fold higher in SP versus non-SP (4,100 versus 21 colonies per million input cells). For marker negative cells, in all three cases only the SP fraction produced colonies (5,300, 1,4500 and 90 per million input cells). Malignant character of marker positive cells was confirmed using FISH analysis, immediately after cell sorting prior to cell culturing, and at the end of the liquid culture on plucked colonies. These results show that both in CD34 negative and CD34 positive leukemia the SP fraction contains most primitive cells and has much higher clonogenic potency compared to non-SP cells. In CD34 negative AML, SP HSC are CD34+CD38- and SP LSC/LIC are CD34-CD38+. In CD34 positive AML, SP HSC are CD34+CD38- and SP LSC/LIC may be CD34+CD38-, but also CD34+CD38+, CD34-CD38+ and CD34-CD38-. The absence of CD34 negative engraftment, seen in earlier studies, likely results from relatively high immune competence of the NOD/SCID mice in those studies. The results suggest that CD34 negative leukemia may contain more immune sensitive LSC/LIC. Above that, CD34 negative AML is characterized by lower multidrug resistance (abstract Schuurhuis, Kelder et al, subm for this meeting). From these characteristics favorable outcome for the particular class of patients, defined as truly CD34 negative, can be predicted. This was indeed the case in our large study (394 patients): median overall survival not reached (>41 months) for CD34 negative patients (n=85) and 19 months for CD34 positive patients (n=309) (p=0.006). Our results warrant to consider CD34 negative patients as a completely different entity of AML patients, in terms of stem cell characteristics, which may evoke different therapeutic strategies compared to CD34 positive patients.