In CD34-positive acute myeloid leukemia (AML), the leukemia-initiating event is thought to occur in the CD34+CD38− stem cell compartment. Survival of these cells after chemotherapy hypothetically results in minimal residual disease (MRD) and leads to relapse.

C-type lectin-like molecule 1 (CLL-1) could serve as a possible target for therapy, as we previously showed that it is expressed on malignant CD34+CD38− cells at diagnosis and not on normal CD34+CD38− cells (

van Rhenen et al., Blood 106: 6a, 2005
). CLL-1 expression is also present on malignant CD34+CD38− cells after chemotherapy.

In the present study we investigated whether other antigens are also aberrantly expressed on CD34+CD38− cells in AML patients at diagnosis which would allow detection of AML CD34+CD38− cells in remission bone marrow. Such would offer opportunities for patient risk stratification and guidance of therapeutic intervention.

Flowcytometry was performed on bone marrow CD34+CD38− cells in AML at diagnosis, after chemotherapy and in normal (NBM) and regenerating (RBM) bone marrow. Antibodies were anti-CD34, anti-CD45 and anti-CD38 together with antibodies against CLL-1, CD2, CD5, CD7, CD11b, CD19 and CD56, apart from CLL-1 all used as leukemia-associated phenotypes (LAPs) in immunophenotypic MRD detection (

Feller et al., Leukemia 8:1380, 2004
; further referred to as “whole blast” MRD). In diagnosis AML marker expression on CD34+CD38- cells was scored as <50% or >50%, because at least 50% expression is needed for accurate measurements of malignant CD34+CD38- cells after chemotherapy.

At diagnosis, 60/77 AML samples were CD34-positive. A reliable number of CD34+CD38- events (>20) could be measured in 56/60 cases. CLL-1 expression >50% was observed in 15/60 cases, LAP expression in 9/60 cases and both CLL-1 and LAP in 8/60 cases. Altogether, in 32/60 CD34-positive cases, detection of malignant CD34+CD38- cells was possible. In NBM (n=4) as well as in RBM, the CD34+CD38- cells did show low CLL-1 (n=6) and low LAP (n=2, for all antigens) expression (all <3%). For three AML samples in which a leukemia-specific translocation was present, FISH analysis showed the translocation in CD34+CD38-LAP+ cells.

Therefore, CLL-1 and/or LAP staining might enable to accurately discriminate between residual normal and malignant CD34+CD38- cells after chemotherapy. Firstly, we found that irrespective of the aberrant phenotype or the time point of analysis the frequency of residual malignant CD34+CD38- cells significantly correlated with “whole blast” MRD (r=0.42, p=0.008, n=38). Secondly the ratio between malignant and normal CD34+CD38- cells (stem cell ratio) was found to parallel both CD34+CD38- cell frequency and “whole blast” MRD. Both parameters, similar to “whole blast” MRD, parallelled clinical outcome with decreases when entering remission, no changes in non-responders and increases with forthcoming relapses.

In 3/60 cases in which “whole blast” MRD could not be used, stem cell MRD could be performed. Since the populations are so well defined [CD34+/CD38−/CD45dim/aberrant marker] stem cell MRD will require less extensive experience than currently used MRD frequency assessment. These advantages are expected to increase as the use of 5 or more fluorescence channels will improve the success of stem cell MRD.

In conclusion it is possible to detect malignant stem cells in AML patients at diagnosis and after chemotherapy, which would offer opportunities for future patient risk stratification and guidance of therapeutic intervention.

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author

Sign in via your Institution