Chronic Myeloid Leukemia: Current Status and Controversies

Not so long ago, therapeutic decisions in chronic myeloid leukemia (CML) were rather straightforward: Allogeneic bone marrow transplantation was performed in all patients with a donor who were deemed fit enough for the procedure, interferon-alfa was administered to all others able to tolerate it, and the remainder received hydroxyurea. The advent of imatinib(Drug information on imatinib) mesylate (Gleevec) and, to a lesser degree, the development of reducedintensity conditioning regimens, have radically changed the approach to treating CML and have considerably increased the complexity of treatment decisions. Biology of CML: Unsolved Issues
In the first part of their paper, Mughal and Goldman briefly review some aspects of the biology of CML. A few issues need to be clarified. Bcr- Abl-negative CML is clearly a mixed bag. However, it should be noted that the diseases associated with activation of platelet-derived growth factor receptor-beta or fibroblast growth factor receptor type 1 have clinical features- such as eosinophilia and monocytosis-that are not typical of "standard" CML.[1] Although there is certainly overlap with Bcr-Abl in the use of signaling pathways, major differences exist. Not surprisingly, the transcriptional profile of Bcr-Abl- transformed hematopoietic cells is distinct from that of platelet-derived growth factor receptor-beta-transformed cells.[2] The authors comment that we still do not understand "precisely how [Bcr-Abl] induces the leukemic phenotype." Although this is true, I would argue that it may be time to acknowledge that the transforming network operated by CML is imprecise by its very nature, exhibits extensive redundancy, and is partially driven by stochastic rather than linear processes. In contrast, we have little knowledge about the mechanisms responsible for disease progression. None of the genetic lesions associated with blast crisis, except rearrangements affecting the EVI-1 transcription factor,[3] explain the most striking feature of blast crisis, which is the loss of terminal differentiation. Improving our understanding of disease progression appears crucial, as one could argue that CML would not pose much of a clinical problem if it remained in chronic phase. Treatment Decisions in CML
The larger part of the article nicely reviews the treatment of CML, including imatinib and allogeneic stem cell transplantation. Therapy for CML is a moving target, as the follow-up of patients on imatinib is relatively short, and no direct comparison has been made between conventional and reduced- intensity conditioning regimens. Thus, current recommendations have a limited half-life. At the heart of the dilemma is the question of whether imatinib will eventually replace or only delay allogeneic transplantation. To put this into perspective, one needs to consider the obvious short-, likely medium-, and potential long-term problems associated with imatinib therapy. In the short term, resistance to imatinib after an initial response is the key issue. Although the risk of resistance is much higher in patients with advanced disease, I somewhat disagree with the authors' notion that it is rare in chronic phase. In phase II and III trials in chronic phase CML patients after failure of interferon-alfa as well as in newly diagnosed patients, the rates of disease progression were 20% and 12% at 40 and 31 months, respectively.[4,5] Undoubtedly, these results are an enormous improvement over those achieved with interferon-based therapy. Given that 5-year survival rates in selected patients after allografting approach 75% in some studies,[6,7] this rate of relapse is nonetheless significant. Another important point is that responses induced by dose escalation for resistant disease are frequently not durable.[8] For optimal management of resistant patients, it will be crucial to determine the specific mechanism of resistance and tailor therapy accordingly. Illustrations of this paradigm include the differential sensitivity of various Abl kinase domain mutants to higher doses of imatinib[ 9] or to second-generation Abl inhibitors.[10] Without mutational analysis, rational therapy of such patients will be impossible. Arguably better than treating relapse would be to avoid it in the first place. At present, it is unknown whether disease eradication (defined as neg- ativity by reverse transcription- polymerase chain reaction) is required to avoid relapse or if "operational" cure may be achieved, as discussed by Mughal and Goldman. Given that newly diagnosed patients with a more than 3-log reduction of Bcr-Abl mRNA have excellent progressionfree survival despite the presence of minimal residual disease, one would hope for the latter.[11] Experience with other malignancies however suggests that disease persistence, even at a low level, indicates a continuous risk of relapse. Thus, disease persistence is likely to become the key issue in the medium term. Several large studies are under way in Europe that are testing combinations of imatinib with interferon-alfa or cytarabine(Drug information on cytarabine), as well as high-dose imatinib vs the standard dose, and molecular remission will be a critical end point. The rationale for these combinations is based on in vitro observations in cell lines or committed CML progenitor cells that do not necessarily apply to cells that cause persistence in vivo. Thus, efforts need to be directed toward understanding the mechanisms underlying imatinib's failure to eradicate residual disease, so that rational therapeutic strategies can be developed. Lastly, in the long term, side effects may still become a relevant issue. Generally, imatinib is well tolerated, with minimal toxicity. That said, the perspectives might change if patients are treated for many years, perhaps decades. Examples include the recently reported effects of imatinib on testosterone metabolism with some men developing gynecomastia[12] or lateoccurring liver toxicity.[13] As with persistent disease, the key to solving this problem is disease eradication. However, even after the last CML cell is eradicated, the bone marrow may not become normal, as clonal chromosomal abnormalities, sometimes associated with a myelodysplastic syndrome, have been seen in patients with a cytogenetic response.[14] Advising the Newly Diagnosed Patient
Patients in accelerated phase or blast crisis CML should be offered an allo- geneic transplant, with imatinib used for tumor reduction, as responses in advanced disease are generally not durable. For chronic phase patients, Mughal and Goldman outline two approaches. One calls for a trial of imatinib in all patients, reserving transplant for those who fail to respond optimally or become resistant. In the short term, this strategy will preserve a span of high-quality lifetime, as the early mortality after allografting is avoided. Over the long term, additional factors will become relevant. Much depends on the question of whether imatinib treatment may compromise a subsequent allograft. The available retrospective data are controversial, and the cohorts investigated are high risk, potentially obscuring detrimental effects.[15,16] Prospective studies in standard-risk patients are needed to clarify this crucial issue. Another important point is that the approach requires high-quality follow-up to detect resistance as early as possible. The relative lack of side effects and the ease of administration of imatinib may lead to less stringent follow-up outside of studies and major centers. The second approach outlined by the authors is to offer allografting to patients with a low risk of transplantrelated mortality. Luckily, our ability to identify such patients has improved considerably, although much of the data must be re-validated in patients transplanted with reduced-intensity conditioning regimens. Conclusions
With the many uncertainties described above, either approach is defendable at present. One additional consideration appears crucial: The patient's personal preference is a major factor in the decision-making process, particularly in equivocal situations, and thanks to the Internet, patients are much better informed today than they were not so long ago.