For patients with CP CML, treatment with IM is associated with improved survival through high rates of sustained cytogenetic and molecular remissions, (16
) while the impact of single agent IM on outcomes of advanced phase CML, despite high initial responses, is less impressive. HSCT therefore remains an essential part of the therapeutic armamentarium for patients with advanced phases of CML. Several papers have addressed the effects of pre-transplant IM on outcomes of transplantation, (6
) and the safety profile of IM is now well established for all phases of the disease. Addressing prognostic factors in the IM era for advanced phase CML has been limited by the relatively small number of patients in single center reports. Additionally, outcomes of those patients who respond well to IM and then proceed in remission with allogeneic transplantation are not well-defined. The CIBMTR database offers the advantage of a large number of patients with extensive data which permits multivariate analyses.
Our analysis included 449 CML patients with advanced phase CML and confirmed that pre-transplant IM was not associated with deleterious or beneficial effects on post-transplant outcomes. Among the studies that have reported an impact of pre-transplant IM on post-transplant outcomes, three have included historical controls.(6
) In a study that included 145 IM+ to 231IM− allograft recipients with CML, Oehler et al (6
) reported comparable 3 year OS for the 73 advanced phase IM+ (60 CP2/AP and 13BP) and the 48 IM− (38 CP2/AP and 10 BP). Deininger et al,(8
) analyzed 70 CML and 21 Ph+ ALL and compared outcomes to historical controls identified in the EBMT database. Pre-transplant IM did not influence overall survival, progression-free survival or non-relapse mortality, while a trend towards higher relapse mortality and significantly less chronic GVHD was observed in the IM+ group (OR=0.44, p
=0.027). Finally, in 30 Philadelphia chromosome positive leukemias including 16 advanced phase CML, outcomes were similar in 48 controls that did not receive prior IM.(7
Previously reported and well-recognized prognostic indicators such as disease phase, age, donor type, donor/recipient sex mismatch, and time from diagnosis to transplant (18
) were found in our analysis to affect post-transplant outcomes in this patient population.. Disappointingly, the outcomes of allografting advanced phase CML remain poor and have not improved over time and with the availability of IM. Indeed, 35–43% of CP2, 26–37% of AP, and 8–16% of BP patients are alive and in remission 3 years post-transplant. However, CP2 patients defined as remission or chronic phase after prior AP or BP, had comparable outcomes to AP and more favorable outcomes than BP patients. Therefore it is possible that newer therapies, and/or more potent tyrosine kinase inhibitors that increase the response rates and the achievement to CP2 may improve outcomes of patients with BP CML. Interestingly, comparable survivals were observed after transplantation using a reduced intensity or a myeloablative conditioning. Our results are comparable to outcomes reported by the EBMT on allogeneic transplantation using reduced intensity conditioning in advanced phase CML.(20
This study, similar to all registry analyses, has its inherent limitations: it is a retrospective study, with a relatively short follow-up (median 3 years). Additionally, requested data from transplant centers was often incompletely reported: 53% and 68% had no information on prior BP in the IM+ and IM− cohorts. Information on patients identified at conditioning as AP or BP with prior transient remission after treatment with IM was not available. Patients were classified as AP according to the criteria previously defined in the CIBMTR case report forms and was not done according to the WHO classification. Although a supplemental data questionnaire was sent to transplant centers to determine the reasons patients proceeded with transplantation (planned, or IM resistance), reasons that often led to delays in transplantation in the IM group were unclear. To circumvent this particular limitation, we analyzed the impact of duration of IM therapy prior to transplantation on outcomes and no effects were found. Of interest, information on post-transplant use of IM in patients who did not receive pre-transplant IM was not available, and could potentially have affected our results. Finally, data on the presence of BCR/ABL mutations was unavailable.
This analysis, in our opinion, is still relevant despite the widespread use of imatinib and other tyrosine kinase inhibitors. Indeed, and in contrast to chronic phase CML patients who are more likely to proceed to allogeneic HSCT after IM-resistance and exposure to second generation kinase inhibitors (21
), patients presenting at diagnosis with advanced phases of CML usually do so after exposure to front-line IM. This is the approach recommended by the European Leukemia Net (23
), and the National Cancer Center Network guidelines (24
). This study provides updated outcomes analysis of allografting for these patients. Additionally, one can speculate that with higher remission rates associated with IM as compared to chemotherapy, (5
) it is possible, albeit impossible to demonstrate, that a higher fraction of patients reach CP2 after treatment with IM and therefore are able to undergo transplantation and achieve better outcomes than if they had remained in BP.
In conclusion, in this largest cohort of patients with advanced phase CML, CP2 and AP patients had similar outcomes following allogeneic HSCT; whereas outcomes of BP patients, were dismal unless CP2 can be achieved. Conventional prognostic indicators remain the major determinants of transplant outcomes in the IM era. Time from diagnosis to allogeneic HSCT less than 12 months is a modifiable variable that is associated with better outcomes and early planning for transplantation in appropriate patients may be beneficial.