Result of high-dose imatinib mesylate in patients with Philadelphia chromosome—positive chronic myeloid leukemia after failure of interferon-α

Imatinib mesylate (STI571; Gleevec) is a potent tyrosine kinase inhibitor with selectivity against c-abl and bcr/abl, as well as c-kit and platelet-derived growth factor receptor (PDGF-R).^1,2  Significant activity has been demonstrated in patients with Philadelphia chromosome (Ph)—positive chronic myeloid leukemia (CML) after failure of interferon-alpha (IFN-α) therapy. The standard dose recommended for chronic-phase CML is 400 mg daily. With these doses, more than 90% of patients treated in chronic-phase CML after IFN-α failure achieve a complete hematologic response (CHR),^3,4  and 60% achieve a major cytogenetic (CG) response, which is complete in 40%.^3,4  After a median follow-up of 18 months, 90% of patients have not transformed to the accelerated or blastic phase. As significant as these responses are, most patients still do not achieve a complete CG response, and 8% have lost their major CG response.³  Achievement of a complete CG response has been associated with 10-year survival rates of 70% to 85%^5-8  with IFN-α—based therapy. Thus, achievement of a complete CG remission is the most important short-term objective of CML therapy. Furthermore, molecular remissions have been rare among patients treated with standard-dose imatinib.⁹ 

Data from imatinib trials suggest that higher doses of imatinib may be more effective. The initial dose-finding phase 1 trial reported a clear relationship between dose and response.¹⁰  Complete hematologic responses were achieved in 98% of patients treated at daily doses of 300 mg or higher, compared with 56% in patients treated with 200 or 250 mg daily, and lower responses at doses of less than 200 mg daily. Cytogenetic response occurred in 29 (54%) of 54 patients treated at 300 mg/d or higher, while only 2 patients treated at lower doses had such a response.¹⁰  In the phase 2 studies, patients in accelerated phase were treated with a daily starting dose of either 400 mg or 600 mg.¹¹  The higher dose induced higher response rates and longer response durations and survival compared with the lower dose. In the chronic-phase trials, imatinib dose escalation to 800 mg daily in patients not responding to 400 mg recaptured CHR in 2 of 3 patients and improved cytogenetic response to a major one in 40%. Finally, amplification of the BCR/ABL gene or overexpression of its protein product is one of the known mechanisms of resistance.^12-15 

In this study we investigated the efficacy and toxicity profile of high-dose imatinib in patients with CML after IFN-α failure. We observed a high rate of major and complete cytogenetic responses and, more important, a high rate of complete molecular remission.

Patients with chronic-phase CML were eligible for this study if their disease did not respond to IFN-α therapy (according to hematologic or cytogenetic response criteria) or if they could not tolerate IFN-α therapy. In addition, to be eligible, patients were required to be 18 years or older and to have adequate performance status (Eastern Cooperative Oncology Group [ECOG] 0 to 3), serum creatinine levels lower than 176.8 μM, total bilirubin levels less than 34.2 μM, and normal cardiac function (New York Heart Association grades 3-4 cardiac disease excluded). Patients previously treated with imatinib at any dose for any length of time were not eligible, but patients could have received and failed other therapies for CML besides IFN-α. Women of childbearing age were required to have a negative pregnancy test before starting imatinib, and all patients at risk were required to use barrier contraception during therapy. Patients provided written informed consent before entry into the study; the study was reviewed and approved by the internal review board of the institution and performed in accordance with the Declaration of Helsinki.

Patients with accelerated- or blastic-phase CML, as previously defined,¹⁶  were excluded. Patients with cytogenetic clonal evolution were eligible only if there were no other criteria of accelerated-phase CML. Hematologic and cytogenetic failure to IFN-α, and IFN-α intolerance were as previously defined.^3,4  After initial evaluation of the results of this study, the study was modified to include patients in late chronic phase who had not received and refused therapy with IFN-α, and patients who had not achieved a molecular remission after 2 years or longer of IFN-α therapy.

Imatinib was given as an oral dose of 400 mg twice daily. Hydroxyurea was allowed for debulking during the first 6 weeks of therapy. Anagrelide and leukapheresis (a maximum of 1 procedure per week) were permitted for up to 3 weeks.

Dose reductions of imatinib for nonhematologic or hematologic toxic effects were as follows. For grade 2 persistent nonhematologic toxic effects, therapy was interrupted until recovery to grade 1 or less and resumed at the original dose level. If grade 2 toxicity reappeared, treatment was interrupted again until recovery and resumed at a daily dose of 600 mg. For grades 3 or 4 nonhematologic toxicity, therapy was interrupted until recovery to grade 1 or less and resumed at a daily dose of 600 mg. For grades 3 to 4 hematologic toxicity (granulocyte count of < 0.5 × 10⁹/L or a platelet count of < 40 × 10⁹/L), therapy was interrupted until the neutrophils recovered to 10⁹/L or higher and/or platelets to 60 × 10⁹/L or higher. If the toxicity resolved within 2 weeks, treatment was resumed at the original dose of 400 mg twice daily. If toxicity resolved after more than 2 weeks, or if it recurred after resuming therapy, the dose was reduced to 600 mg. Further dose reductions to 400 mg or 300 mg daily were allowed using the same guidelines for hematologic and nonhematologic toxicity. Patients developing anemia received packed red blood cell transfusions or blood products at the discretion of the investigator, or erythropoietin, 40 000 units given subcutaneously once a week until the hemoglobin level increased to 120 g/L or more.

Complete blood counts and serum chemistry were performed weekly during the first 4 weeks, every other week for the next 3 months, and every 6 weeks thereafter. Bone marrow studies, including morphologic and cytogenetic or fluorescent in situ hybridization (iFISH) analysis, were performed every 3 months. Patients were followed for survival at least every 3 months. Drug safety parameters were evaluated at each visit and graded according to the National Cancer Institute (NCI) Common Toxicity Criteria (CTC), version 2.0.

Response criteria were as previously described.⁸  A CHR was defined as a white blood cell count of less than 10 × 10⁹/L, a platelet count of less than 450 × 10⁹/L, no immature cells (blasts, promyelocytes, myelocytes) in the peripheral blood, and disappearance of all signs and symptoms related to leukemia (including palpable splenomegaly) lasting for at least 4 weeks. Response was further categorized by the best CG response: complete if no Ph-positive cells were present, partial if Ph-positive cells were 1% to 34%, and minor if Ph-positive cells were 35% to 90%. Major CG response included complete plus partial cytogenetic responses (ie, Ph-positive cells 0% to 34%). Cytogenetic response was judged by standard CG analysis of metaphase spreads, not by iFISH. Time to disease progression was calculated from the time treatment began until appearance of accelerated- or blastic-phase disease, discontinuation of therapy for unsatisfactory response, or death. Survival was calculated from the time treatment began until death from any cause.

Cytogenetic analysis was performed by the G-banding technique. For chromosome analysis, bone marrow specimens were examined on direct or short-term (24-hour) cultures. At least 20 metaphases were analyzed. PCR was done by real-time quantitative reverse transcription (RT)—PCR, and negative results (ie, undetectable transcript) were confirmed by nested PCR as previously reported.¹⁷ 

There were 36 patients who were treated and were evaluable. Of them, 2 patients belonged to the categories added late into the study: 1 patient with late chronic-phase CML who had not received prior IFN-α and 1 patient who was not in molecular remission after 3 years of IFN-α. The other 34 patients had late chronic-phase CML that failed to respond to IFN-α, or were intolerant to IFN-α. Their clinical characteristics are summarized in Table 1. Of the patients, 13 (36%) had active disease, and 23 (64%) were in CHR. The median age was 47 years (range, 30-75 years); the median time from diagnosis to start of high-dose imatinib was 25 months (range, 10-135 months). There were 13 (36%) patients who had received other therapies besides IFN-α, including homoharringtonine (n = 10) and R115777 (Zarnestra, a farnesyl transferase inhibitor; n = 3). There were 4 patients (11%) who had cytogenetic clonal evolution. Having received fewer than 3 months of imatinib therapy, 3 patients were inevaluable for response because of poor compliance. Results are described for 33 evaluable patients.

All 11 patients with abnormal peripheral blood counts when treatment was started achieved a CHR (Table 2). At the start of therapy, 12 patients had fewer than 34% Ph-positive metaphases, including 6 who had 0% Ph-positive. A major CG response was achieved in 19 (90%) of 21 other patients (complete in 18, partial in 1). Complete CG response was achieved in 24 (89%) of 27 patients with Ph-positive metaphases at the start of imatinib. Responses occurred early: 19 (73%) of 26 patients evaluable at 3 months had already achieved a complete CG response at this time. Within the small number of patients in each group, there was no significant difference in the response rate according to previous IFN-α response.

Figure 1 shows the values of BCR-ABL/ABL percentage ratio during imatinib therapy. There was a consistent and significant decline of these ratios over time. The median BCR-ABL/ABL percentage ratio decreased from 0.169% (range, .005%-12.281%) after 3 months to .026% (range, 0%-63.454%) after 12 months of therapy (P = .60). Significantly, 18 (56%) of 32 patients evaluable for molecular response have achieved ratios lower than 0.045%, including 13 patients (41%) who had undetectable transcripts (confirmed by nested PCR). There were 16 patients who have had at least 2 consecutive PCR values lower than 0.045%, and 4 patients had at least 2 consecutive analyses with undetectable transcripts.

Of the 33 patients treated for at least 3 months, 31 (94%) remain alive in chronic phase on imatinib therapy. The other 2 patients stopped therapy because of noncompliance (n = 1) and intercurrent illness (arthritis, n = 1). After a median follow-up of 16 months (range, 5-16 months), all patients are alive in chronic phase.

Of the 31 patients with a major cytogenetic response, 30 (97%) continue to have durable major response. One patient had disease relapse (defined as an increase of Ph-positive cells by 30% or more documented on 2 occasions, or an increase to 65% or more documented once) after 6 months from the time the major CG response was achieved. After a median follow-up time of 9 months from best response (range, 1-12 months), 3 patients who achieved a complete CG response have lost their response (complete CG response duration of 3 [1 patient] and 9 [2 patients] months). At their latest follow-up these patients had 20%, 16%, and 5% Ph-positive metaphases, but continue on therapy. All 6 patients who started with 0% Ph-positive continue in complete CG response. The median follow-up time from achievement of BCR-ABL/ABL percentage ratio lower than 0.045% is 3 months (range, 1-9 months); none of the 18 patients who have achieved these ratios has relapsed.

One patient who achieved a minor cytogenetic response has lost the cytogenetic response (defined as an increase of Ph-positive cells to more than 90%) after 9 months from best response.

Common adverse events are listed in Table 3. Fluid retention, skin rash, nausea, diarrhea, and muscle cramps were common, but not more frequent than what has been reported with standard-dose imatinib. Grade 3 or higher nonhematologic adverse events were uncommon, with skin rash (n = 3, 8%), fatigue (n = 1, 3%), liver toxicity (n = 1, 3%), and bone pain (n = 1, 3%) being the most common. Hematologic toxicity is also shown in Table 3. Grade 3 or higher neutropenia occurred in 11 patients (31%) and thrombocytopenia in 8 patients (22%). Anemia was observed in only 3 patients (8%).

After 3 months of therapy, 18 patients (50%) were receiving the planned dose; 3 had discontinued therapy because of lack of compliance and 15 had dose reduced for toxicity. Corresponding values at 6 and 9 months were 44% (14 of 32 evaluable) and 41% (12 of 29 evaluable), respectively. Dose reduction to 600 mg daily resolved the toxicity in most patients; 10 patients (28%) have required further dose reductions to 400 mg daily or lower (300 mg is the lowest acceptable dose). Patients that required dose reductions for toxicity did so early; only 3 patients have required dose reductions after 4 months of therapy (after 20 [2 patients] and 64 [1 patient] weeks of therapy). The most common causes for dose reduction were myelosuppression in 7 patients (thrombocytopenia [n = 2], neutropenia [n = 2], both [n = 3]; 19%), skin rash (n = 5, 14%), fatigue (n = 3, 8%), and bone pain (n = 2 [1 recurrent grade 2, 1 grade 3], 6%).

In this study we report high rates of major and complete CG responses with high-dose imatinib (800 mg daily) in patients whose CML has failed to respond to IFN-α therapy. Of the patients, 56% achieved very low levels of BCR-ABL/ABL percentage ratios, including 41% who achieved complete molecular remissions. Higher-dose imatinib was well tolerated, with infrequent grade 3 or higher toxicity. These results suggest that higher doses of imatinib may be more effective in Ph-positive CML.

The response rate achieved with higher-dose imatinib is remarkable. The major CG response rate was 90% (95% confidence interval [CI], 70%-99%). Most CG responses were complete (89%; 95% CI, 71%-98%). These results compare favorably with those reported with standard-dose imatinib. At 3 months, 85% of the patients have achieved a major CG response, compared with 44% reported with standard-dose imatinib. Early response to imatinib has been shown to have significant prognostic implications.^3,4,11  Kantarjian et al reported a landmark analysis of 343 patients treated with standard-dose imatinib in chronic phase after IFN-α failure with evaluable CG studies after 3 months of therapy.⁴  With a median treatment duration of 18 months, 5 (3%) of 152 patients who had achieved a major CG response at 3 months had progressed, compared with 22 (12%) of 191 without such a response (P = .005). Early response was also associated with a survival advantage.³  Since few patients in our series did not achieve a major CG response even at this early time point, we cannot compare our results with that analysis. However, it is hoped that the early response observed in our patients would have similar implications.

Despite having similar eligibility criteria, the study group included in this study is not necessarily comparable with that reported in studies using standard-dose imatinib. A noticeable difference is the paucity of patients with hematologic failure to IFN-α (n = 1). These patients have been reported to have a low response rate to imatinib.^3,4  Still, within each category of IFN-α failure, the response rate to high-dose imatinib was equally high, and the results appear to be particularly significant in patients with intolerance or CG resistance to IFN-α. A direct comparison with standard-dose imatinib would be required to better define the magnitude of these differences.

The molecular response rates with high-dose imatinib were significant. Of the patients, 56% (95 CI, 38%-74%) achieved BCR-ABL/ABL percentage ratios lower than 0.045%. This threshold was clinically meaningful, as demonstrated by Hochhaus et al,¹⁸  in 54 patients who had achieved complete CG response with IFN-α therapy. Of the 27 patients who did not reach ratios lower than 0.045%, 13 (48%) eventually relapsed, compared with 1 (4%) of 27 patients who achieved these levels or lower.¹⁸  Furthermore, 13 (41%; 95 CI, 24%-59%) patients treated with high-dose imatinib in this study had undetectable transcripts, with a median follow-up of only 14 months. Molecular remissions have been reported infrequently with late standard-dose imatinib. Hochhaus et al analyzed 300 patients with CML treated with imatinib, including 135 patients with late chronic phase.⁹  All 55 patients in complete CG remission remained PCR positive with a median follow-up of 1.3 years. In our experience, 32 (29%; 95 CI, 21%-38%) of 111 patients with chronic-phase CML treated with 400 mg imatinib daily after IFN-α failure have achieved BCR-ABL/ABL ratios lower than 0.045% after a median follow-up of 32 months; 14 patients (13%; 95 CI, 7%-20%) have undetectable transcripts. The percentage of patients who achieve undetectable transcripts appear to be higher with high-dose imatinib, even with the shorter median follow-up time of therapy. It is important however to recognize that the methodology for molecular analysis varies in different studies, and some techniques reported in the literature may be more sensitive.¹⁹  Also, the long-term outcome of patients with undetectable transcripts after imatinib remains to be determined with longer follow-up. In our series, with a median follow-up time of 14 months, all patients who achieved a complete molecular response continue in molecular remission.

High-dose imatinib therapy was not associated with significantly more toxicity compared with standard-dose imatinib. These results are not unexpected. In a phase 1 trial of imatinib in a similar population, patients were treated with escalating doses of 25 mg to 1000 mg daily.¹⁰  A maximum tolerated dose was not defined, and further dose escalations were not pursued because of evidence of biologic and clinical activity. In patients with accelerated-phase CML, there was no significant difference in grade 3 or higher hematologic or nonhematologic toxicity among patients treated with 400 mg or 600 mg daily.¹¹  In our trial, fluid retention was more likely to require therapy (ie, grade 2 according to the NCI CTC), but grade 3 edema was rare. The only nonhematologic grade 3 or higher adverse event that was more common with high-dose imatinib was skin rash. Although myelosuppression was observed, this was not more common with high-dose imatinib.

We conclude that high-dose imatinib may be more effective than standard doses of imatinib in patients with chronic-phase CML and failure on IFN-α therapy. These doses are well tolerated, and molecular remissions can be achieved in a significant percent of patients. A direct comparison of high-dose imatinib with standard-dose imatinib is warranted.