In CD34-positive acute myeloid leukemia (AML), the leukemia-initiating event likely takes place in the CD34+CD38- stem cell compartment. Survival of these cells after chemotherapy hypothetically leads to minimal residual disease (MRD) and relapse. We have previously shown that a high CD34+CD38- frequency correlates with both MRD frequency, especially after the third course of chemotherapy and poor survival (Clin Cancer Res, in press). Furthermore, we have shown that a monoclonal antibody against the novel cell surface marker C-type lectin-like molecule-1 (CLL-1), directed against myeloid cells, stains 92% of diagnosis AML (

Bakker et al.,
Cancer Res.
64
:
8443
,
2004
). In the present study we investigated whether this antibody can be used to identify AML stem cells in remission bone marrow. Such would offer opportunities for MRD stem cell detection and stem cell-directed therapy.

We found that anti-CLL-1 antibody homogeneously stained the CD34+CD38- compartment in 77/89 cases (median expression of 33.3% in all 89 cases, range 0–100%). The median stem cell expression of CLL-1 in control bone marrow was 0% ranging from 0–11% (n=11). Furthermore, CLL-1 expression on AML stem cells is highly stable: no differences between paired diagnosis and relapse samples (p=0.9, n=12). Like most antigens CLL-1 is expressed on part of the CD34+CD38+ compartment, but expression is absent on megakaryocytic precursors, which for therapeutic use would circumvent delayed platelet recovery. For antibody-mediated therapy it is crucial that normal stem cells remain negative throughout treatment of the disease. Therefore we tested bone marrow regenerating after high dose chemotherapy, obtained from either non-AML hematological patients or CD34 negative or CLL-1 negative AML patients. In those patients complete absence of CLL-1 expression was found in CD34+CD38− cells (n=4).

Under MRD-conditions CLL-1 staining thus enables to accurately discriminate between normal and malignant CD34+CD38− stem cells. In agreement with this, the different ratios of AML and normal stem cells that were found in a number of patients, paralleled clinical outcome in terms of probability of relapse.

For comparison, the stem cell marker CD123 was studied. Although anti-CD123 antibody homogeneously stained CD34+CD38− cells with high intensity in almost all AML samples studied (35/36 cases) with also no differences between diagnosis and relapse (p=0.6, n=6) and with low expression in normal bone marrow (median 14.9%, range 0–18.8%, n=5), a high expression was found in regenerating bone marrow (median 60%, range 53–84%, n=4). The latter suggests that anti-CD123 antibody is not AML stem cell specific under all conditions of disease.

In conclusion, our data provide strong evidence that a large CD34+CD38− population at diagnosis reflects a higher percentage of chemotherapy-resistant cells, which, in remission, will lead to the outgrowth of MRD, thereby affecting clinical outcome. The specificity of anti-CLL-1 antibody under all conditions of disease enables both reliable detection and quantification of the stem cell compartment for prognostic use under MRD conditions, as well as characterization. Moreover, it shows that AML stem cell targeting using antibody treatment at different stages of disease has now become an option in the treatment of AML patients.

Author notes

Corresponding author

Sign in via your Institution