A legitimate model of disease progression in CML predicts that BCR-ABL activity promotes the accumulation of genetic and epigenetic alterations directly or indirectly responsible for the reduced apoptosis susceptibility and the enhanced proliferative potential and differentiation arrest of CML blast crisis cells. Indeed, there is evidence for a mechanism of CML disease progression where alterations in DNA repair processes coupled with enhanced survival of BCR-ABL-expressing cells may allow the propagation of secondary genetic changes that favor the emergence and persistence of increasingly malignant cells. Among the secondary changes, those directly or indirectly affecting the p53 tumor suppressor gene seem to have a central role for frequency and biological consequences.
At the molecular level, CML blast crisis remains a heterogeneous disease and yet only few pathways are commonly affected. Of the many issues that need to be investigated for a better understanding of the pathogenic mechanisms in CML disease progression, some should attract considerable attention.
First, is BCR-ABL overexpression a common feature of blast crisis CML and is determined stochastically or by genetic/epigenetic mechanisms? Recent evidence suggests that expression of BCR-ABL is more abundant in early than in late CML chronic phase progenitors and that CML blast crisis undifferentiated and committed myeloid progenitors express higher BCR-ABL levels of the corresponding chronic phase progenitors (1,15,17,23). If this is confirmed by additional studies, an obvious question will be whether BCR-ABL overexpression is causally linked or secondary to the expansion of homogeneous populations of blast cells. A molecular characterization of apparently identical subpopulations of CML-chronic phase and blast crisis progenitors and their comparison with normal progenitors may prove important in addressing these possibilities. Overexpression of BCR-ABL during disease progression may be the result of stochastic forces and/or facilitated by genetic/epigenetic mechanisms. We favor the second hypothesis, as restoration of PP2A activity in myeloid CML blast crisis cells leads to SHP-1-mediated dephos-phorylation and proteasome-dependent degradation of BCR-ABL, suggesting that functional inactivation of PP2A tumor suppressor by BCR-ABL represents an auto-regulatory mechanism allowing increased and sustained BCR-ABL activity and expression in early CML blast crisis progenitors (23). However, a more detailed analysis of the mechanisms responsible for increased BCR-ABL expression in CML blast crisis patients with and without chromosomal/molecular abnormalities could be informative in supporting or disproving either possibility.
Second, are there specific molecular determinants of lymphoid CML blast crisis and do they interact with BCR-ABL in selectively promoting the expansion of B-cell type blasts? The high frequency of homozygous deletions at the INK4A/ ARF locus suggests that inactivation of the INK4A or ARF gene or both serves as the important molecular determinant of lymphoid blast crisis. Since specific knockout models (INK4A-/-, ARF-/-, and INK4A/ARF-/-) are available, it will be important to test whether INK4A, ARF, or the combination of both mutants cooperates with BCR-ABL to induce the selective expansion of a pool of progenitors committed to B-cell development. Perhaps, loss of the INK4A or ARF gene or both favors cell cycle entry of B-cell type rather than myeloid progenitors. Loss of p16 and ARF function is required for immortalization of human cells, whereas loss of ARF is sufficient for immortalization and RAS-dependent transformation of mouse cells (98); thus, the role of ARF and p16 should also be independently tested in in vitro models of BCR-ABL transformation of multipotent and unipotent human progenitor cells.
Third, will disease progression and development of secondary genetic abnormalities be affected by imatinib monotherapy? If disease progression depends on the constitutive activity of the BCR-ABL tyrosine kinase, suppressing its activity will postpone, if not prevent, the development of CML blast crisis. Although there is no proof for it, the secondary changes of CML blast crisis may reflect, in part, the effects of treatment with DNA damaging agents on the genetically unstable background of BCR-ABL-expressing cells. if so, imatinib-resistant CML blast crisis might be characterized by a distinct pattern of secondary changes primarily caused by the constitutive activity of BCR-ABL. It is also possible that the molecular mechanisms of disease progression will be totally different in those patients resistant to imatinib as well as to AMN107 and dasatinib (11); however, it is unclear whether the T315I is a gain-of-function mutation that confers increased kinase activity to BCR-ABL (99). The ultimate goal of BCR-ABL kinase inhibitor-based CML therapy is disease eradication and prevention of transition to blast crisis. Since it is unlikely that this will be achieved in each patient, as it appears that both imatinib and dasatinib are not effective in killing the so-called quiescent CML stem cell (8), understanding disease progression in the imatinib-resistant group will be essential for the development of "rational" therapeutic strategies [e.g., PP2A activating compounds (23) or drugs able to restore C/EBPa expression and its tumor suppressor and pro-differentiation activities (100)] to be used in conjunction with BCR-ABL tyrosine kinase inhibitors.
Was this article helpful?