Chronic-phase (CP) chronic myelogenous leukemia (CML) once was a terminal cancer with a median survival of about 20 months if left untreated.1 The use of nonspecific cytoreductive agents, such as hydroxyurea or busulfan, was associated temporally with an improved median survival of approximately 4 years, but the 8-year survival rate remained <15%.2 Allogenic stem cell transplantation (allo-SCT) was the first treatment modality in CP-CML that was capable of inducing cytogenetic and molecular remissions and resulted in superior long-term, leukemia-free survival in approximately 50% of patients.3 Consequently, for a time, allo-SCT was considered the first-line treatment of choice for patients with CP-CML despite its association with a >50% incidence of transplantation-related mortality or morbidity from chronic graft-versus-host disease.3
In the early 1980s, it was demonstrated that interferon-α (IFN-α) induced complete cytogenetic remission (CCyR) or partial cytogenetic remission (PCyR) in approximately 20% of patients with CP-CML, and the 10-year survival rate in such patients was >50% versus <30% for all patients who received IFN-α.4-10 The long-term disease progression rate in patients who achieved an IFN-α–induced CCyR was <15%.10 Therefore, some investigators appropriately argued at the time that patients with CP-CML could undergo a treatment trial with IFN-α first and could consider transplantation only if they failed to achieve CCyR.11 Such an approach enabled some patients to bypass the ill effects of allo-SCT without compromising its value as salvage therapy.12-14
Today, imatinib has replaced both allo-SCT and IFN-α as first-line treatment for CP-CML.15 Imatinib is a first-generation tyrosine kinase inhibitor (TKI) of c-abl oncogene 1 receptor tyrosine kinase (ABL1), arginase (ARG), platelet-derived growth factor receptor (PDGFR), and v-kit Hardy-Zuckerman 4 feline sarcoma viral oncogene homolog (KIT).15 Imatinib targets the adenine triphosphate (ATP) binding site within the breakpoint cluster region (BCR)-ABL1 fusion protein, which is the disease-causing mutant kinase in CML,16 and disrupts oncogenic signal by forcing an enzymatically inactive conformation.15, 17, 18 In 2003, results from the International Randomized Study of Interferon Versus STI571 (IRIS) trial were reported; 1106 patients with CP-CML were randomized to receive either imatinib or IFN-α plus cytarabine.19 At a median follow-up of 19 months, treatment with imatinib was superior in terms of both the CCyR rate (76.2% vs 14.5%) and the progression-free survival (PFS) rate (96.7% vs 91.5%).19 Those results led to the December 2002 US Food and Drug Administration approval of imatinib for first-line treatment of adult patients with BCR-ABL1–positive CP-CML; the drug previously had received approval for the treatment of advanced-phase CML (AP-CML) or IFN-α–refractory CP-CML in May, 2001.
The 6-year follow-up of patients who were randomized to the imatinib arm (n = 553) of the IRIS study was published in 2009.20 The overall survival rate at 6 years was 88%, and the PFS and event-free survival (EFS) rates were 93% and 83%, respectively. In patients who achieved CCyR or PCyR at 6 months of therapy, the PFS rate was 97% or 94%, respectively.20 The annual rates of progression to AP-CML in the first 6 years were 1.5%, 2.8%, 1.6%, 0.9%, 0.5%, and 0%, respectively. Approximately 83% of patients achieved CCyR at some point during their treatment; of these, only 16% lost their CCyR, and 34% of them regained their CCyR during continued imatinib therapy. Overall, at 6 years, only 12% of patients discontinued imatinib therapy because of suboptimal drug efficacy.20
In this issue of Cancer, Al-Kali et al describe The University of Texas M. D. Anderson Cancer Center (MDACC) experience with imatinib therapy in patients with CP-CML and underscore the importance of accounting for salvage therapy in reporting first-line treatment outcome.21 In their report, the best CCyR rate among patients who received imatinib as front-line therapy was 88%. After a median follow-up of 68 months, 73% of the patients remained on imatinib and in CCyR, and an additional 7% stopped treatment while still in CCyR. The 7-year EFS rate was 81%. Approximately 20% of patients either failed to achieve CCyR or lost their CCyR. The majority of patients who failed on imatinib and received a second-generation TKI (SG-TKI) regained or achieved CCyR, raising the “current EFS” rate to 88%. This observation is consistent with the established value of SG-TKI in restoring CCyR in imatinib failures and preventing disease transformation into AP-CML.22
The above-described IRIS and MDACC studies clearly demonstrate the value of imatinib in securing long-term PFS in >90% of patients with CP-CML and in >95% of those who achieve either CCyR or PCyR at 6 months of treatment initiation. Can one do better with an SG-TKI as first-line therapy? Two phase 3 studies comparing imatinib with either dasatinib (n = 519)23 or nilotinib (n = 846)24 recently were published. Nilotinib at a dose of either 300 mg or 400 mg twice daily was superior to imatinib 400 mg once daily in inducing a major molecular response (MMR) and CCyR at 12 months (43%-44% vs 22% and 78%-80% vs 65%, respectively).24 In addition, a significant difference in the rate of progression to AP-CML was documented (<1% vs 4%, respectively). In the second randomized study,23 dasatinib 100 mg once daily was superior to imatinib 400 mg once daily in terms of both CCyR (77% vs 66%) and MMR (46% vs 28%) at 12 months; and the rate of progression to AP-CML was similar (0.9% and 3.5%).
There are several issues that I find problematic in the above-described imatinib versus SG-TKI randomized studies. First, only “CCyR” has been correlated with improved survival in CP-CML, and it has not been demonstrated that achieving “MMR” provides an additional survival advantage.25, 26 Therefore, using MMR as a surrogate marker for superior treatment outcome in CP-CML is questionable. Second, comparing CCyR at 12 months from treatment initiation creates a potential bias against standard-dose imatinib, which requires a longer time to express its full potential in that regard25, 26; in the IRIS study, for example, the CCyR rate at 12 months was 69% but increased to 87% by 5 years.26 It is noteworthy that the survival outcome did not appear to be affected by whether or not CCyR occurred before or after 12 months of imatinib therapy.27 The third and most concerning issue is the finding that neither study allowed patient crossover, and the investigators did not account for the contribution from salvage therapy with a SG-TKI.23, 24 Herein lies the main flaw in study design, because what is being compared with SG-TKIs does not reflect current practice.
It is also important to note that some of the findings in the studies mentioned above23, 24 are inconsistent with those reported from other randomized studies of patients with newly diagnosed CP-CML.20, 28 For example, the 3.5% to 4% first-year disease progression rates reported in the imatinib arms of these studies are substantially higher than the 1.5% rate reported in the IRIS study.20 Similarly, the 12-month MMR rate in the IRIS study was estimated at a much higher 39%.29 In another randomized study that compared 400 mg and 800 mg daily imatinib doses, the 12-month MMR rate (46% vs 40%, respectively) and CCyR rate (70% vs 66%, respectively) were similar to those reported above for nilotinib and dasatinib,23, 24 although they did not differ significantly from each other.28 It is possible that some of these inconsistencies will be clarified further with more follow-up information, but it may not be adequate to rectify the deficiencies in the overall study design.
The finding that both dasatinib and nilotinib, as opposed to standard-dose imatinib, led to a more rapid and deeper response rate at 12 months was not unexpected, because both drugs are several-fold more potent than imatinib. Also, both drugs are active against most imatinib-resistant ABL kinase domain mutations,30 which is consistent with their ability to induce approximately 40% CCyR and approximately 70% PFS in imatinib-resistant/intolerant CP-CML.22, 31 Therefore, it is conceivable that a comparative trial between imatinib and SG-TKI, in the absence of early intervention with salvage therapy to overcome primary resistance or suboptimal response to imatinib, might favor SG-TKIs.20, 30 However, the practically relevant question is not whether we can do better with an SG-TKI as first-line therapy but whether the premise holds true in the context of early salvage intervention, which has demonstrated a positive impact on both CCyR at 12 months and “current PFS.”32, 33
We have come a long way since the 19th century therapeutic attempts in CML with arsenic trioxide.34, 35 The accidental sighting of the Philadelphia chromosome in 196036 laid the foundation for the seminal discovery of the disease-causing BCR-ABL1 mutation and successful application of the concept of molecularly targeted therapy.19 Our patients have benefited greatly from this series of events and stand to benefit even more when the market exclusivity of Gleevec (imatinib mesylate) ends around 2015 and a generic version becomes potentially available. This is not necessarily good news for all interested parties, and it is fully expected that drug companies and some indirectly compensated “opinion leaders” will try to make the case for ousting imatinib as front-line therapy for CP-CML. I, for one, am not convinced yet and will continue to use imatinib as front-line therapy with close monitoring and switching to an SG-TKI at the earliest sign of resistance or suboptimal response to imatinib.
TKIs do not cure CML,37 and faster and deeper-than-CCyR treatment responses do not necessarily translate into superior overall survival or PFS.27 Current long-term safety and efficacy data for imatinib are too strong to be undermined by the results of a single phase 3 study with questionable study design. The ideal design would have been to compare long-term outcomes with imatinib followed by an SG-TKI for suboptimal response/resistance versus an upfront SG-TKI.