Revealing Six Phases of CML Stem Cell Development to Explain Clinical Phenomena Seen in TKI-Treated Patients.

Abstract 4263

Chronic myeloid leukemia (CML) is a stem cell (SC) disease defined by the BCR/ABL oncoprotein that is considered essential for abnormal growth and accumulation of neoplastic cells. Based on in vitro studies and mathematical models, CML clones are considered to be organized hierarchically similar to normal hematopoiesis. More recent data suggest, that CML cells grow in subclones that usually exhibit SC function but vary in their leukemia-initiating potential. The situation is even more complex in patients treated with TKI. In these patients, intrinsic as well as acquired resistance against TKI have been described and recognized as an emerging problem and challenge in practice and research. Most SC concepts focus on imatinib-resistant mutants of BCR/ABL, that are detectable in subclones. However, several questions and phenomena that occur in TKI-treated patients remain to be solved. Based on laboratory and clinical observations, we propose the existence of 6 distinct phases of CML SC development (SCD): a) a Ph-negative phase, b) an early Ph-positive preleukemic SCD-phase in which subclones are (very) small and usually undetectable, c) a pre/leukemic SCD-phase in which one or more subclones expand(s) and replace(s) normal myelopoiesis but still produce(s) normal WBC, d) chronic phase (CP), e) accelerated phase (AP), and f) a blast phase (BP). The latency period until progression into a next SCD phase is variable and may depend on several different factors including subclone inhibition by chalones like lipocalin and other factors, different growth kinetics produced by various BCR/ABL mutants, and TKI-induced subclone-selection. Phase a) may explain the rare occurrence of Ph-negative subclones (+OCA) during TKI treatment. Phase b) may explain why BCR/ABL mutants are not detectable before TKI therapy is initiated, why mutant- and ACA subclones “appear” in CCyR patients after a certain latency period, and why one patient can develop two or more BCR/ABL mutants in different subclones. Phase b) may also explain the rare detection of very small quantities of BCR/ABL in healthy individuals and constant low level-MRD in a few CML patients in whom therapy was stopped. Phase c) can explain early CML patients in whom WBC are normal in repeated tests; and explain relapsed TKI-resistant patients in whom under TKI therapy normal hematopoiesis is replaced by the mutant subclone but WBC remain normal for weeks to several months. The SCD phases a) through c) are not accessible in any of the conventional xenotransplant models available. Even in SCD phases d) through f), it is quite difficult to demonstrate stable long term engraftment in xenotransplant mouse models, although stem cell subclones obtained from patients in f) may grow in NOG- or NSG mice similar to AML SC. However, even if this may be a reproducible approach, analysis of all relevant subclones in these patients will probably not be achievable unless additional oncogenes are introduced in these subclones. In summary, our new concept extends the hydra model of cancer stem cell development for CML by introducing a step wise progression-model with 6 defined phases of SCD. This model may have theoretical and clinical implications for patients with freshly diagnosed and TKI-resistant CML and for cancer stem cell research in general.