Persisting Chronic Myeloid Leukemia Stem and Progenitor Cells From Patients in Major Molecular Remission Under Imatinib Are Characterized by Low BCR/ABL Expression.

Treatment with the Abl-kinase specific inhibitor imatinib (IM) is very effective in chronic myeloid leukemia (CML). However, IM presumably fails to eradicate CML stem cells (HSC) leading to disease persistence and relapse after IM-discontinuation. Although causes of CML persistence under imatinib remain ill defined, quiescence and BCR/ABL-overexpression of CML stem and progenitor cells have been suggested as underlying mechanisms. We here set out to identify means to directly study persistence mechanisms in residual BCR/ABL-positive progenitor and stem cell clones from chronic phase CML patients in major molecular remission (mmR) under imatinib.

Bone marrow specimens of twenty-one CML patients in at least major molecular remission (mmR) according to the international scale, first diagnosis (FD) patients (n=5) and healthy donors (n=4) were sorted into HSC, common myeloid progenitors (CMP), granulocyte/macrophage progenitors (GMP) and megakaryocate-erythrocyte progenitors (MEP) and BCR-ABL mRNA expression was directly assessed by quantitative real time (qPCR) and/or nested PCR (mRNA of 4.000 sorted cells). Alternatively, HSC, CMP, GMP and MEP were seeded into soft agar and mRNA was extracted from individual colony forming units (CFU) to assess BCR/ABL-mRNA expression by qPCR. Moreover, CFU of sub-fractions of first diagnosis CML patients were treated in vitro with IM at 3mM and BCR/ABL-expression of surviving CFU was compared with the BCR/ABL expression levels of mock-treated CML-CFU. In total, 595 soft agar colonies were analyzed.

By nested PCR, BCR/ABL-mRNA was readily detectable in the HSC compartments of 7 of 10 (7/10) CML patients in mmR. BCR/ABL was also detected in the CMP-, GMP-, and MEP-compartments in 6, 10 and 8 of the 10 patients, respectively. Real time qRT-PCR suggested only a trend toward stronger BCR/ABL positivity of the HSC compartment when compared to the other progenitor compartments (table 1). A detailed analysis of the BCR/ABL-expression of individual CFU from HSC-, CMP-, GMP-, and MEP-compartments of mmR patients revealed that persisting CML-CFU expressed significantly less BCR/ABL than first diagnosis CML-CFU obtained before imatinib therapy (table 1). This finding could be recapitulated in vitro: primary CML-CD34+ cells of first diagnosis CML patients (n=4) were seeded into soft agar in the presence or absence of 3 uM imatinib. After 14 days BCR/ABL expression only of BCR/ABL-positive CFU was compared. BCR/ABL-positive CML-CFU (n=30) that had survived imatinib exposure expressed significantly less BCR/ABL than mock-treated CML-CFU (n=175) (p<0.001).

Work is in progress providing in vitro evidence that selection/induction of low BCR/ABL expression in immature progenitor and stem cells is a new mechanism of imatinib persistence in mmR patients via reducing oncogenic addiction from BCR/ABL.

We showed that BCR/ABL-persistence is not confined to the quiescent CML-stem cell compartment, but seems to affect also the highly proliferative progenitor compartments. More intriguingly, persisting CML-HSC and -precursor cells express remarkably low levels of BCR/ABL when compared to first diagnosis HSC and progenitors, implying that low BCR/ABL expression reduces imatinib sensitivity in vivo. The simple model of selection / induction of low BCR/ABL expression as mechanism of imatinib persistence in CML would explain the low propensity of disease progression after achieving mmR, and the low genetic instability of CML clones from mmR patients.

No relevant conflicts of interest to declare.