The Facts

Several clinical trials that changed the treatment paradigm and outlook for patients with CML were reported over approximately 10 years. This article reviews some of the most relevant trials evaluating the use of TKIs in first-line therapy.

IRIS

The International Randomized Study of Interferon and STI571 (IRIS) study established imatinib as the standard of care for CP-CML.^[3] This study randomly assigned 1106 patients to receive either the combination of interferon-α and low-dose cytarabine (IFN-α/Ara-C; the standard of care at that time) or imatinib at a dose of 400 mg daily. At a median follow-up of 19 months, the rates of complete cytogenetic response (CCyR) at 18 months were 76.2% and 14.5% in the imatinib and IFN-α/Ara-C arms, respectively. At 18 months, the rates of freedom from disease progression were estimated to be 96.7% and 91.5%, respectively. Imatinib was significantly better tolerated, with only 14.3% of patients discontinuing therapy or crossing over to the alternative arm, as opposed to 89.2% for IFN-α/Ara-C.

A 6-year update on the patients randomized to the imatinib arm confirmed these outstanding results.^[4] The cumulative best CCyR rate was 82%, with 63% of patients who were still on imatinib in CCyR. The estimated event-free survival was 83%, with an overall survival rate of 88%. The rate of freedom from progression to AP or BP was estimated to be 93%. Follow-up of the imatinib arm at 8 years showed that outcomes continued to be very good, with 82% of patients in CCyR and 92% of patients still free from progression to AP or BP.^[5] The highest risk of loss of response or transformation was during the first 3 years of treatment (3.3%–7.5%), and decreased thereafter.

Hammersmith Hospital

Subsequently, investigators from Hammersmith Hospital reported their experience using imatinib in the first-line treatment of 204 patients with CP-CML.^[6] At 5 years, the cumulative incidence of CCyR was 82.7%, with an estimated overall survival rate of 83.2%. The progression-free survival rate was estimated to be 82.7%. However, 25% of patients had discontinued therapy because of unsatisfactory response or side effects. As a result, the probability of remaining on therapy at 5 years while maintaining a major cytogenetic response was only 62.7%.

The Real World

In a population-based study from the United Kingdom, Lucas et al.^[7] reported their experience using imatinib, 400 mg daily, as first-line therapy in the treatment of 68 patients with CP-CML. Of these, 3 progressed to BP within the first 12 months of therapy. During that same period, 2 patients experienced toxicity serious enough to warrant alternative therapy, and 1 patient lost complete hematologic response. The CCyR rates were 41%, 49%, and 51% at 12, 18, and 24 months, respectively. Remarkably, at 24 months, imatinib treatment failed in 49% of patients because of progression to BP, failure to experience or maintain CCyR, or intolerance. These results suggest that, in a general practice setting, outcomes with imatinib therapy may not be as good as what was observed in the IRIS study.

High-dose Imatinib

Two randomized trials were conducted comparing high-dose and standard-dose imatinib in the treatment of CP-CML. The Tyrosine Kinase Inhibitor Optimization and Selectivity (TOPS) study randomized 476 patients with newly diagnosed CML to either 400 or 800 mg of imatinib daily. At 6 months of therapy, rates of major molecular response (MMR; defined as a 3-log reduction in copy number of BCR-ABL transcripts, 0.1% or lower on the international scale) and CCyR were higher in the high-dose arm.^[8] However, at 12 months they were comparable (46% vs. 40% and 70% vs. 66% for high-dose and standard-dose imatinib, respectively). Furthermore, no difference was seen in progression-free and overall survival rates between the arms at 18 months (97.4% vs. 95.0% and 98.7% vs. 98.2%, respectively).

The German CMLIV trial also reported faster responses with 800 mg of imatinib (administered on a tolerability-adapted schedule) than with 400 mg of imatinib or the combination of imatinib and IFN-α.^[9] However, no difference was seen in overall and progression-free survival rates (92% and 88%, respectively) at 5 years between the treatment arms. Notably, in the 800-mg arm, the median dose that was administered on the tolerability-adapted schedule was only 628 mg/d.

Imatinib-based Drug Combinations

It was expected that imatinib would ultimately be combined with the older agents in an attempt to improve outcomes, and 2 large randomized trials of this strategy were conducted. The German CMLIV study randomized patients with CP-CML to either imatinib at 400 mg/d, imatinib at 800 mg/d, or the combination of imatinib and IFN-α. Responses in the high-dose imatinib arm were higher than in both of the other arms with no differences in progression-free or overall survival.^[9] However, no differences were seen in CCyR and MMR between the standard-dose imatinib and the imatinib/IFN-α arms. At 12 months, the rates of CCyR and MMR were 49% and 31%, respectively, for imatinib at 400 mg, and 50% and 35%, respectively, for imatinib/IFN-α. At 24 months, these rates were 74% and 63%, and 77% and 63%, respectively.

The French SPIRIT trial randomized patients with CP-CML to 1 of 4 arms, namely imatinib at 400 mg or 600 mg daily, or the combination of imatinib (400 mg daily) with cytarabine or pegylated IFN. At 12 months, the combination of imatinib and pegylated IFN led to a higher rate of MMR than was seen in the other arms.^[10] However, no difference was seen in CCyR. At that time point, rates of CCyR were 58%, 65%, 70%, and 66% for the imatinib, 400 mg; imatinib, 600 mg; imatinib/cytarabine; and imatinib/IFN arms, respectively. Rates of MMR were 38%, 49%, 46%, and 57%, respectively. Adverse events were more common with IFN than with imatinib, 400 mg. In both the CMLIV and the SPIRIT trials, the rates of transformation-free and overall survival were not improved by the addition of IFN.

ENESTnd

After its approval as a second-line agent, and with promising results from phase II studies in the first-line setting,^[11,12] nilotinib was compared with imatinib in a randomized fashion for the treatment of patients with newly diagnosed CP-CML. The ENESTnd trial (Evaluating Nilotinib Efficacy and Safety in Clinical Trials–Newly Diagnosed Patients) was a randomized study that enrolled 846 patients to 1 of 3 arms: nilotinib at 300 mg or 400 mg twice daily, or imatinib at 400 mg once daily.^[13] The primary end point of the study was MMR at 12 months. The rates of MMR at that time point were 44%, 43%, and 22% for the nilotinib, 300 mg; nilotinib, 400 mg; and imatinib arms, respectively. The rates of CCyR were 80%, 78%, and 65%, respectively. The superiority of nilotinib over imatinib was observed in all risk groups, including patients with high Sokal risk scores.

The trends were confirmed at 24 months when the rates of BCR-ABL reduction to 0.0032% or less were 26%, 21%, and 10% for the 3 arms, respectively.^[14] Importantly, a statistically significant lower rate of progression to AP/BP was seen in the nilotinib arms as opposed to the imatinib arm. After 24 months of follow-up, these rates were 0.7%, 1.1%, and 4.2% in the nilotinib, 300 mg; nilotinib, 400 mg; and imatinib arms, respectively.^[14] Nilotinib was generally better tolerated, particularly in terms of hematopoietic, musculoskeletal, and gastrointestinal toxicity. However, grade 3/4 abnormalities in alanine aminotransferase, bilirubin, lipase, and glucose were more frequent with nilotinib treatment. Dermatologic events and headaches were also more frequent with nilotinib. Despite previous concerns regarding the effect of nilotinib on the QT interval, no patient on nilotinib developed QTc prolongation to an interval greater than 500 ms at the 24-month data cutoff point.^[15] Nor was a decrease lower than 45% observed in the mean ventricular ejection fraction. After these results, the FDA approved nilotinib for the treatment of patients with newly diagnosed CP-CML.

DASISION

Similar to nilotinib, phase II data showed promising results for dasatinib in the first-line setting.^[16] It was therefore subsequently compared with imatinib in a randomized trial. The DASISION trial (Dasatinib Versus Imatinib Study In Treatment-naïve CML) randomized 519 patients to either dasatinib at 100 mg daily, or imatinib at 400 mg daily. The primary end point was a confirmed CCyR (2 measurements) at 12 months. The CCyR rates at 12 months were 66% and 77% for imatinib and dasatinib, respectively.^[17,18] When compared based on at least 1 measurement, the CCyR rates were 72% and 83% for imatinib and dasatinib, respectively. At 12 months, the rates of MMR were 28% and 46%, respectively, and at 24 months were 46% and 64%, respectively. Development of AP/BP at 12 months occurred in 1.9% of patients treated with dasatinib and 3.5% in patients treated with imatinib. However, the latter differences were not statistically significant. At 24 months, the rates of disease progression were 2.4% and 5.0% for dasatinib and imatinib, respectively. Again these differences were not statistically significant. At 24 months, 17% of patients treated with dasatinib had achieved a BCR-ABL transcript level below 0.0032%, compared with 8% of patients treated with imatinib. Thrombocytopenia was more common with dasatinib, whereas neutropenia occurred at equal frequency. Edema, nausea, vomiting, and rashes were more common with imatinib. In earlier studies, pleural effusions were seen in patients treated with dasatinib, particularly at a dose of 70 mg twice daily.^[19,20] In the DASISION study, no pleural effusions were observed in the imatinib arm. In the dasatinib arm, pleural effusions occurred in 14% of patients. These were mainly grade 1 and 2, with fewer than 1% of patients developing grade 3 pleural effusions.^[21] Notably, a randomized phase II intergroup study was subsequently reported in which patients with CP-CML were assigned to either imatinib 400 mg/d or dasatinib 100 mg/d.^[22] At 12 months, the rates of CCyR and MMR were 69% and 32% in the imatinib arm and 82% and 47% in the dasatinib arm, respectively.

Early Responses and Long-term Outcomes

As could be expected, most of the data on the impact of early responses on long-term outcomes are from the IRIS study. Patients who fared the best in the IRIS study were those who had experienced both a CCyR and an MMR at 12 months. None of those patients’ disease had progressed to AP or BP by 5 years of therapy, compared with a 5% rate of disease progression in patients who had experienced a CCyR but not an MMR at 12 months.^[23,24] Another analysis of the IRIS data showed that the earlier patients experienced a CCyR (3, 6, 12, or 18 months), the better their long-term outcomes.^[5] Most patients in the IRIS trial who experienced disease progression while on imatinib did so during the first 3 years of therapy.^[5]

Similar trends were seen in the long-term analysis of the Hammersmith data. In that study, patients who experienced a CCyR at 12 months showed substantially better transformation-free and overall survival rates at 5 years than patients who did not (96% vs. 74%, and 98% vs. 74%, respectively).^[6] However, in this series, experiencing an MMR at 12 months did not convey a long-term advantage.

Peculiarities of Each Approved TKI

Although imatinib, nilotinib, and dasatinib share a common mechanism of action, namely tyrosine kinase inhibition, they have important differences. From a mechanistic standpoint, both imatinib and nilotinib bind to an inactive conformation of the ABL kinase, whereas dasatinib binds to multiple conformations.^[25,26] Nilotinib is up to 50 times more potent than imatinib at inhibiting BCR-ABL in cell proliferation assays, whereas dasatinib is up to 325 times more potent.^[23,27] Of the 3 agents, nilotinib is the least promiscuous in its ability to inhibit different tyrosine kinases. Dasatinib is a dual BCR-ABL and SRC kinase inhibitor.

ABL kinase domain mutations constitute the most common mechanism of resistance to imatinib. More than 50 of these mutations have been described.^[28] Both dasatinib and nilotinib are active against all known imatinib-resistant mutations, except the T315I mutation, which imparts resistance to all 3 agents. However, dasatinib and nilotinib have differences in their potency against specific non-T315I mutations.^[27]

From a toxicity standpoint, all 3 agents are generally well tolerated. However, a few points are significant regarding nonhematologic toxicity. Diarrhea, fluid retention, and muscle cramps are more common with imatinib than with nilotinib or dasatinib.^[13,17] Although these are generally tolerable (mainly grade 1/2), they can become problematic in the long run from a quality-of-life standpoint. This is particularly relevant in view of the expected long-term survival of patients in the TKI era.

In the early development of nilotinib, unexplained sudden deaths were reported, which led to a black box warning on its label. However, in the ENESTnd trial, no sudden deaths occurred in the nilotinib arms, and no patient on nilotinib had a prolongation of the QTc interval greater than 500 ms.^[17] Furthermore, QTc prolongation seems to be a class effect for these agents.

Again, in its early development, dasatinib was associated with a substantial incidence of pleural effusions. Clinically relevant pleural effusions were seen more commonly at doses of 70 mg twice daily than with the now-adopted regimen of 100 mg daily.^[19,20] In the DASISION study, patients were closely monitored for the development of pleural effusions. Although 14% of patients in the dasatinib arm developed pleural effusions, fewer than 1% of these were grade 3.^[21]

Expansion of clonal cytotoxic T or natural killer (NK) cells has been described in patients treated with dasatinib.^[29–34] In the DASISION trial, lymphocytosis occurred in 5% and 24% of patients treated with imatinib or dasatinib, respectively. Patients on dasatinib who developed lymphocytosis had a higher rate of CCyR at 12 months than those who did not develop lymphocytosis (84% vs. 75%). Lymphocytosis in patients treated with dasatinib was also associated with a higher incidence of pleural effusions (18% vs. 8%). Whether lymphocytosis contributes to dasatinib’s antileukemic effects through an immunomodulatory mechanism is yet to be proven.