Purpose: Imatinib mesylate, a specific Bcr-Abl tyrosine kinase inhibitor, has shown encouraging activity in chronic myelogenous leukemia (CML).

Experimental Design: We treated 237 patients (median age, 50 years; age range, 18–82 years) with Philadelphia chromosome (Ph)-positive accelerated-phase CML with oral imatinib mesylate at daily doses of 400 mg (26 patients) or 600 mg (211 patients) and evaluated response and survival characteristics in univariate and multivariate analyses.

Results: Among the 200 patients with accelerated-phase CML for whom follow-up was 3 months or more, rates of complete and partial hematological response were 80% and 10%. Cytogenetic responses were evident in 90 patients [45%; complete response in 47 patients (24%) and partial response (Ph 1–34%) in 21 patients (11%)]. The estimated 18-month survival rate was 73%. The estimated complete hematological response rate at 18 months was 68%; that for cytogenetic response was 82%. In multivariate analyses, a diagnosis-to-treatment interval of 3 years or more, splenomegaly, and peripheral blasts predicted poor major cytogenetic response; age >60 years, marrow basophilia, and clonal evolution predicted poor survival. The 600-mg drug dose was associated with better cytogenetic response and survival in univariate analysis (P < 0.01) but not in multivariate analysis. Landmark analysis showed that achieving a cytogenetic response at 3 months or a major cytogenetic response (Ph < 35%) at 6 months was associated with better long-term survival. Seven of 15 patients who were in a second chronic phase achieved major cytogenetic response. The incidence of severe nonhematological toxic effects was 23%; drug discontinuation for severe toxicity was needed in 3% of patients.

Conclusions: Imatinib mesylate was active against Ph-positive, accelerated-phase CML, and the prognostic factors identified in this study could aid in tailoring treatment strategies to specific risk groups.

The accelerated phase of CML4 is characterized by increased disease aggressiveness and resistance to therapy and cellular transformation (1, 2, 3). Other features include increased cellular proliferation, with correspondingly high percentages of blasts, promyelocytes, and basophils; thrombocytopenia or thrombocytosis; anemia; splenomegaly; the appearance of chromosomal abnormalities other than the Ph; and marrow fibrosis (4, 5, 6, 7, 8). These characteristics are associated with poor prognosis and shortened survival time (4, 5). Expanded definitions of accelerated-phase CML may allow the inclusion of patients who are in chronic phase rather than accelerated phase; hence apparent prolongation of survival among such patients could be erroneously attributed to therapy (1, 6, 8). For example, cytogenetic clonal evolution may be associated with a median survival of 7–56 months, depending on the particular cytogenetic abnormality, the percentage of metaphases involved, the presence of other accelerated-phase features, and the therapy offered (8).

To improve the prognosis of accelerated-phase CML, several new treatment modalities have been investigated, including IFN-α, allogeneic SCT, intensive chemotherapy, and novel agents such as HHT, decitabine, troxacitabine, and others (9, 10, 11, 12, 13, 14, 15, 16). Imatinib mesylate (STI571, Gleevec) is a phenylaminopyrimidine specifically designed to selectively inhibit Bcr-Abl tyrosine kinase activity. By tightly incorporating into the Abl activation sites within Bcr-Abl, imatinib mesylate prevents the transfer of ATP and its binding to the tyrosine moiety, a necessary step for activation of Bcr-Abl. Imatinib mesylate suppresses p210 and p190 Bcr-Abl protein kinases, as well as other kinases such as c-kit and platelet-derived growth factor receptor (17, 18, 19). In Phase I-II clinical trials, imatinib mesylate has shown encouraging activity against CML in all phases, and it has become a cornerstone of therapy for CML (20, 21, 22, 23, 24, 25). Preliminary findings from a multinational, multi-institutional study of oral imatinib mesylate given to 181 patients with confirmed accelerated-phase CML showed a CHR rate of 82% (a response that lasted for at least 4 weeks in 69% of patients) and an overall cytogenetic response rate of 48% [major response (Ph-positive cells <35%) in 24% of patients and complete response (Ph-positive cells = 0%) in 17% of patients]. The estimated 12-month survival rate was 74% (23). In the multinational, multi-institutional studies, the large number of participating institutions, with diverse sources of clinical and laboratory information and different patterns of evaluation and follow-up, has made precise analysis of response profiles, prognostic factors, and details of therapy somewhat difficult.

The purpose of this study was to analyze long-term outcome after imatinib mesylate therapy in a large number of patients with accelerated-phase CML treated at a single institution. A related goal was to evaluate the relevance of known patient and disease features in accelerated-phase disease on response and outcome and use the results to propose treatment-tailored approaches based on the analysis.

Patients.

Adults (≥18 years) with a confirmed diagnosis of Ph-positive CML in accelerated phase were eligible to participate. Informed consent was obtained according to institutional guidelines. Eligibility criteria included (a) adequate performance status (level of 0–2 on the Eastern Cooperative Oncology Group scale); (b) adequate renal function (creatinine level ≤2.0 mg/dl); and (c) adequate hepatic function (bilirubin, aspartate aminotransferase, and alanine aminotransferase levels no more than twice the upper limits of normal).

Patients participated in one of two consecutive protocols, both sponsored by Novartis. The first (Novartis Protocol 109) was the pivotal Food and Drug Administration trial of imatinib mesylate for accelerated-phase CML. Accelerated-phase disease was characterized by the following criteria: blast count of ≥15%; basophil count of ≥20%; blast + promyelocyte count of ≥30%; the presence of chromosomal abnormalities in addition to Ph; and thrombocytopenia (platelet count of <100 × 109/liter) that was unrelated to therapy (4). Constitutional cytogenetic abnormalities, loss of the Y chromosome, and Ph variant translocations were not considered evidence of clonal evolution. In contrast, the second, “expanded access” trial (Novartis Protocol 110) involved the following criteria for accelerated-phase disease: blast count of ≥10%, blast + promyelocyte count of ≥20%, and splenomegaly (spleen palpable at least 10 cm below the costal margin or a 50% increase in spleen size documented over a 1-month period).

Therapy.

The first 26 patients were given imatinib mesylate p.o. at a dose of 400 mg daily. The next 32 patients on the first study and all 179 patients treated on the subsequent expanded access study were given oral doses of 600 mg of imatinib mesylate daily. The appearance of grade 3 or higher nonhematological toxic effects prompted treatment interruption until the toxicity resolved to grade 1 or less, at which time imatinib mesylate was resumed at a lower dose (600 mg was reduced to 400 mg, and 400 mg was reduced to 300 mg). The presence of persistent grade 2 nonhematological toxic effects also led to a reduction of drug dose. Patients experiencing grade 3 or higher hematological toxic effects were evaluated for persistent disease.

If myelosuppression was attributed to disease, then imatinib mesylate treatment was continued at the same dose; otherwise, the dose was reduced as indicated above. In general, imatinib mesylate was withheld for thrombocytopenia (<30 × 109/liter) or granulocytopenia (<0.5 × 109/liter) until the respective counts rose above those levels. If recovery occurred within 2 weeks of drug cessation, the drug was resumed at the same dose; if recovery occurred >2 weeks after drug cessation, the drug was resumed at the lower dose. If myelosuppression recurred after drug resumption at the same dose, the drug was given at the reduced dose. No reductions below 300 mg/day were allowed.

Response Criteria.

Response criteria have been described previously (1, 2). Briefly, a CHR required disappearance of all signs and symptoms related to disease, including normalization of blast counts in the peripheral blood and bone marrow (≤5% marrow blasts), WBC counts of less than 10 × 109/liter, normal differential counts without peripheral blasts, promyelocytes, or myelocytes, and platelet counts of <450 × 109/liter. If thrombocytopenia (<100 × 109/liter) was present before treatment, a CHR required normalization of platelet counts above 100 × 109/liter. Patients who began treatment with normal platelet counts and achieved CHR but had platelet counts of <100 × 109/liter attributable to imatinib mesylate were considered to have achieved CHR but were noted to have low platelet counts. The definition of PHR was the same as that for CHR but allowed for persistence of immature cells (no blasts or promyelocytes, <5% myelocytes + metamyelocytes) in the peripheral blood, splenomegaly, or thrombocytosis (as long as spleen size and platelet count were reduced by 50% or more from pretreatment levels). A further requirement for both CHR and PHR was that the response last for >1 month. Those who achieved CHR were evaluated further by the degree of their cytogenetic response. A cytogenetic response was considered complete if no Ph-positive cells were present, partial if the percentage of Ph-positive cells was reduced to between 1% and 34%, and minor if the percentage of Ph-positive cells was reduced to between 35% and 90%. The term major cytogenetic response was used to refer to complete and partial cytogenetic response (i.e., <35% Ph-positive cells). Patients were evaluated for cytogenetic response every 3 months during the first year after treatment and every 3–6 months during the second year. Cytogenetic response was based on standard cytogenetic analysis of 20 metaphase spreads. Interphase fluorescence in situ hybridization was conducted on marrow samples when the numbers of metaphases were insufficient for cytogenetic analysis. The interphase fluorescence in situ hybridization results were analyzed separately and were not included in the analyses for cytogenetic response.

Statistical Considerations.

Differences between patient subgroups in categorical characteristics and response rates were evaluated by χ2 or Fisher’s exact tests. Rank correlation coefficients were computed for pairs of interval-scaled variables, and distributions were compared between categories by using a Wilcoxon test. Survival distributions were estimated by the method of Kaplan and Meier and compared by the log-rank test. Survival was measured from date of study registration to death from any cause. Duration of cytogenetic response was calculated from date of first evidence of cytogenetic response to recurrence of >90% Ph-positive metaphases. The association of cytogenetic response with survival was assessed using a landmark of 3 months, by which time 75% of total responses were expected to have been achieved (26).

In evaluating the association of individual patient characteristics with cytogenetic response, variables measured on an interval scale were considered as categorical variables using standard cutpoints and also considered as continuous variables in a logistic regression model (27). A graphical smoothing technique was also used to visually assess the association of interval-scaled covariates with response outcomes (28). In a multivariate assessment of covariates and cytogenetic response status, variables were considered for inclusion in a logistic regression model based on historical importance, inspection of findings when covariates were evaluated singly, and correlations among covariates.

A total of 237 patients were treated between December 15, 1999 and May 31, 2001. Their characteristics are shown in Table 1. The median age was 50 years (age range, 18–82); 74 patients (31%) were ≥60 years, and 110 patients (46%) were women. Fifty-eight patients were treated in the first trial, and 179 patients were treated in the subsequent expanded access trial. One hundred and eighty patients had accelerated-phase CML according to the original multivariate analysis-derived criteria (4); disease in the remaining 57 patients met the expanded criteria for participation. Fifteen patients in the latter group were in a second chronic phase. Among patients in accelerated phase, response data were analyzed for the 200 patients for whom follow-up of at least 3 months was available. Survival data were analyzed for all patients who participated in the study. Results from the 15 patients who were in a second chronic phase are also described separately.

Response, Duration of Response, and Survival.

Among the 200 patients for whom follow-up for at least 3 months was available, 160 (80%) achieved CHR, and another 19 (10%) achieved PHR, for an overall response rate of 90%. Ninety patients (45%) showed a cytogenetic response, and that response was complete in 47 patients (24%) and partial in 21 patients (11%; Table 2). The median time to CHR was 1 month, and the median time to cytogenetic response was 2.8 months.

At a median follow-up duration of at least 8.5 months (range, 0.5–24 months), 206 patients (87%) were alive, and 177 patients (75%) were continuing to take imatinib mesylate. Thirty-one patients (13%) died: 15 died after developing blastic-phase disease; 5 died from infection (pneumonia) associated with accelerated-phase disease; 2 died with hypoplastic bone marrow and associated complications; and 9 died of unrecorded causes (death in these 9 patients, all of whom died outside the institution, was most likely of disease progression). Eight patients were able to undergo allogeneic SCT (four of them from a related donor). Disease status at transplant was as follows: (a) second chronic phase, five patients; (b) accelerated phase, two patients; and (c) blastic phase, one patient. No grade 3–4 acute graft-versus-host disease was observed. Chronic graft-versus-host disease was noted in four patients but resolved in all 4 of them with therapy. The median number of days to recovery of granulocytes above 1 × 109/liter was 14 (range, 11–21 days), and the median number of days to recovery of platelets above 30 × 109/liter was 24 (range, 11–28 days). Seven patients were alive (one patient transplanted in second chronic phase died from posttransplantation complications), six without evidence of disease (the patient transplanted in blastic phase relapsed in blastic phase), when this report was written (median, 17 months; range, 10–24 months).

At the time of last follow-up (August 1, 2001), the estimated survival rate for the total population at 18 months was 73% (Fig. 1). The estimated disease control rate at 18 months among the 160 patients who had achieved CHR was 68%, and the estimated rate of durable cytogenetic response at 18 months was 82% (Fig. 2, A and B).

The prevalence of CHR and cytogenetic responses at 3, 6, and 12 months is shown in Table 3. Patients were then grouped by drug dose and their response at 3 and 6 months and evaluated for subsequent survival. Patients who achieved a cytogenetic response had significantly better subsequent estimated survival rates.

The higher imatinib mesylate dose (600 mg) seemed to produce major cytogenetic responses more often than did the lower dose (P = 0.07 among patients who met the standard criteria for accelerated-phase CML; P = 0.13 for total study groups; Tables 4 and 5), despite the fact that follow-up was longer for the lower-dose group (Table 4; Fig. 3). Both response to imatinib mesylate therapy and survival were similar among patients whose disease met the standard criteria for accelerated-phase disease versus those whose disease met the expanded criteria (Table 5).

Among the 15 patients who were treated in a second chronic phase, 12 underwent therapy for 3 months or more, and 7 of these 12 patients achieved a cytogenetic response (major response in 7 patients and complete response in 5 patients). All 15 patients were alive at a median follow-up of 3.5 months.

Factors Predictive of Response and Survival.

Factors of recognized prognostic importance in traditionally defined accelerated-phase CML were evaluated for possible association with achievement of a major cytogenetic response. Patients treated during a second chronic phase were not included in these analyses. Because patients were treated with two different doses of imatinib mesylate, dose was also considered in the analysis. Conventional cutpoints were used to group patients according to individual covariates, although other functional forms of associations between covariates and response were also evaluated. Results are summarized in Table 5. With the exception of platelet count, differences in response were present in the expected direction of effect for all covariates, and most of these differences were statistically significant, regardless of whether factors were considered as categorical data or as linear terms in a logistic model. Because patients given the lower dose of imatinib mesylate were treated in the earliest phase of the trial, these patients tended to have longer diagnosis-to-treatment intervals and higher blast and promyelocyte counts. Several pretreatment factors were also compared with data stratified according to imatinib mesylate dose. Results were similar to the unstratified results and are not presented here.

The major cytogenetic response rate is plotted against the diagnosis-to-treatment interval in Fig. 4. Patients who began imatinib mesylate treatment soon after the initial diagnosis of CML had a higher rate of response than those who began treatment later, but after roughly 3 years, additional time from diagnosis seems to have had little further effect. Similar plots were examined for each factor.

As an initial step in jointly considering the associations of covariates with response, correlations among all interval-scaled covariates were computed (results not shown), and some factors were omitted from subsequent consideration because of their close correlation with retained variables. Factors considered in a multivariate logistic model were diagnosis-to-treatment interval, age, spleen size, hemoglobin level, WBC count, percentages of blasts in the peripheral blood and bone marrow, percentage of basophils in the marrow, the presence or absence of clonal evolution, and drug dose. Terms were generally defined for inclusion in the model according to categories determined by conventional cutpoints, with some adjustments made after inspection of results. Factors were removed from the model in stepwise fashion, except for a term for low or high drug dose, if the significance level was <0.05. Linear terms were also considered for some covariates, and multiple models were evaluated. A model in which the term for peripheral blood blasts was replaced with a term for marrow blasts resulted in similar statistical significance, regardless of whether measurements were considered as categorical or interval-scaled variables.

A model with the four terms remaining at the end of the stepwise process (diagnosis-to-treatment interval, spleen size, percentage of peripheral blood blasts, and drug dose) is summarized in Table 6. The term accounting for the association of dose did not approach statistical significance but was retained in the model to allow assessment of terms for other factors for which corresponding response outcomes may have been influenced by which dose a patient received. Table 6B shows the incidence of major cytogenetic response according to whether patients had zero, one, two, or three factors indicating poor prognosis. Because of some overfitting due to the particular choice of expression for the covariates in this data set, it is unlikely that application of the model to future groups of patients would result in as strong a pattern of association.

Associations between patient variables and survival at 12 and 18 months are summarized in Table 7. These results must be considered preliminary because they are based on 31 deaths at the time of analysis. Factors associated with worse survival included older age, marrow basophilia, cytogenetic evidence of clonal evolution, and lower imatinib mesylate dose. Trends in survival duration according to categories of characteristics tended to parallel the findings regarding achievement of cytogenetic response, although fewer comparisons achieved statistical significance at this time. The generally shorter survival of patients treated with imatinib mesylate at the lower dose may be attributable in part to this group having a worse prognosis. This supposition was confirmed in a comparison between the two dose levels stratified for percentage of peripheral blasts (P = 0.15 for stratified comparison). The median percentage of peripheral blood blasts in the low-dose group was 6.0; the corresponding value in the high-dose group was 1.0. A multivariate analysis showed that age (P < 0.01), marrow basophilia (P < 0.01), and clonal evolution (P = 0.02) remained important in prognosis. Higher drug dose (600 versus 400 mg) tended to remain associated with better survival but was not statistically significant (P = 0.15).

Response to imatinib mesylate also seems to confer prognostic significance. Among the 74 patients who achieved a cytogenetic response after 3 months of therapy, only 3 (4%) have died as compared with 18 deaths (17%) among the 105 patients who did not achieve a cytogenetic response. The estimated 18-month survival rates were 85% versus 70% (Fig. 5,A). At analysis at the 6 month time point, none of the 31 patients who achieved a major cytogenetic response had died, compared with 14 deaths among 68 patients who had not achieved such a response (0% versus 21%). The estimated 18-month survival rates were 100% versus 79% (Fig. 5 B).

Side Effects.

The incidence, type, and severity of side effects from imatinib mesylate therapy were similar to those described in previous reports. Grade 3 or higher nonhematological side effects were unusual and included nausea and vomiting (1%), diarrhea (2%), skin rashes (4%), muscle cramps and/or bone or joint aches (7%), fluid retention (3%), weight gain of 10 kg or more (1%), hepatotoxicity (5%), and fatigue (7%). The overall incidence of such severe toxicities was 23%; however, most improved with dose adjustments, and discontinuation of therapy for persistent toxicity was needed in only seven patients (3%). No drug-associated deaths from nonhematological side effects were observed. Suppression of platelet counts to <30 × 109/liter occurred in 30 of 184 patients (16%) whose pretreatment platelet counts were at least 100 × 109/liter. Granulocytopenia (<0.5 × 109/liter) occurred in 40 of 207 patients (19%) who had pretreatment granulocyte counts of at least 109/liter. Febrile episodes associated with myelosuppression were observed in 30 patients (13%), including documented infections in 17 patients (7%). Two deaths were attributable to imatinib mesylate-associated myelosuppression. One patient who died of a central nervous system hemorrhage had a platelet count of less than 10 × 109/liter; her pretreatment platelet count was 808 × 109/liter. The second patient died of myelosuppression-associated septic shock after 3 months of therapy; no organism was recovered.

Imatinib mesylate produced extremely encouraging results in accelerated-phase CML. Among the 200 patients evaluable for response in this study, the CHR rate was 80%, and the cytogenetic response rate was 45% (complete response, 24%; major response, 34%). At a median follow-up of 8.5 months, the estimated 18-month survival rate was 73%, the disease control rate was 68%, and the cytogenetic disease control rate was 82%. These results are superior to those in published investigational and standard approaches involving IFN-α, HHT, decitabine, and intensive chemotherapy combinations. For example, the CHR rate in response to IFN-α therapy for accelerated-phase CML was 20%, and the cytogenetic response rate was 5% (14). Although experience with HHT in accelerated-phase CML is limited, the CHR and major cytogenetic response rates are, at best, 40% and 5% (15). Decitabine for accelerated-phase disease has produced a CHR rate of 62%, but the complete cytogenetic response rate was only 5%, and the estimated 18-month survival rate was 50% (29). Finally, intensive combination chemotherapy for accelerated-phase CML produced a CHR rate of 42%, a complete cytogenetic response rate of 4%, and an 18-month projected survival rate of 30% (13). Thus, with the exception of allogeneic SCT, imatinib mesylate has produced the best results in treating accelerated-phase CML. Imatinib mesylate, given either as a single agent or in combination with other agents such as IFN-α, cytarabine, HHT, or decitabine, should be considered the new standard of care.

We found that several known prognostic factors were associated with differences in major cytogenetic response rates, a valuable surrogate end point for long-term prognosis (Table 5). By multivariate analysis, a longer interval from diagnosis to therapy (3 years or more), splenomegaly (palpable 10 cm or more below the costal margin), and the presence of blasts in the peripheral blood were associated with worse major cytogenetic response rates (Table 6). The expected major cytogenetic response rates by the presence of zero, one, two, or all three of these factors were 68%, 47%, 16%, and 3%, respectively. These findings suggest that imatinib mesylate should be offered, perhaps in combination, to patients with a poor prognosis as defined by these factors. In our study, the higher dose of imatinib mesylate tended to produce higher major cytogenetic response rates (36% versus 19%; P = not significant), but drug dose did not retain significance as an independent prognostic factor in multivariate analysis, perhaps because of the small number of patients treated with the lower dose (n = 26) and the apparently less favorable prognosis for those patients.

Despite the small number of deaths at this follow-up (31 deaths among 222 patients evaluated by univariate and multivariate analyses for survival), negative associations were already apparent between some pretreatment characteristics and survival. These characteristics, confirmed by multivariate analysis, included older age, increased marrow basophilia, and the presence of cytogenetic clonal evolution. The 600-mg dose seemed to be associated with better survival but did not remain significant by multivariate analysis. Although the role of imatinib mesylate dose on outcome could not be assessed thoroughly due to confounding with prognostic factors, short follow-up, and the small number of patients treated at the lower dose, the results for better trends with higher doses are consistent with other reports demonstrating the higher drug dose to be independently associated with better time to progression and indicating that it should be offered as the standard dose for all such patients (23).

Showing a treatment response, regardless of whether therapy is with hydroxyurea, IFN, or imatinib mesylate, is emerging as a significant prognostic factor that may be useful in selecting subsequent therapy. In this study, response to imatinib mesylate at 3 and 6 months was valuable in predicting outcome (Fig. 5, A and B); this response could be used to select either continued therapy with imatinib mesylate or a shift in treatment.

In this study, 15 patients were in a second chronic phase of CML; all were given the 600-mg daily dose of imatinib mesylate. Among these patients, 58% achieved a major cytogenetic response, and none had died at this follow-up. These results suggest that imatinib mesylate may also be effective for treating recurrent chronic-phase CML.

In summary, positive results from imatinib mesylate therapy for accelerated-phase CML have established it as the new standard of care for most such patients. The multivariate analysis identified a subset of patients who might best benefit from this therapy; it also identified a subset of patients for whom the outcome after single-agent imatinib mesylate therapy was poor. Patients in this latter group should be considered for treatment with imatinib mesylate combinations, allogeneic SCT, or alternative investigational approaches.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by Novartis Pharmaceutical Corp. (East Hanover, NJ).

4

The abbreviations used are: CML, chronic myelogenous leukemia; Ph, Philadelphia chromosome; SCT, stem cell transplantation; HHT, homoharringtonine; CHR, complete hematological response; PHR, partial hematological response.

Fig. 1.

Survival of patients with accelerated-phase CML undergoing imatinib mesylate therapy.

Fig. 1.

Survival of patients with accelerated-phase CML undergoing imatinib mesylate therapy.

Close modal
Fig. 2.

A, time to treatment failure in 160 patients achieving complete hematological response. B, duration of cytogenetic response in 90 patients who achieved a cytogenetic response with imatinib mesylate therapy.

Fig. 2.

A, time to treatment failure in 160 patients achieving complete hematological response. B, duration of cytogenetic response in 90 patients who achieved a cytogenetic response with imatinib mesylate therapy.

Close modal
Fig. 3.

Survival of patients according to imatinib mesylate dose.

Fig. 3.

Survival of patients according to imatinib mesylate dose.

Close modal
Fig. 4.

Martingale residual plot of the incidence of major cytogenetic response by duration of chronic-phase disease.

Fig. 4.

Martingale residual plot of the incidence of major cytogenetic response by duration of chronic-phase disease.

Close modal
Fig. 5.

Landmark analysis of survival based on response evaluation at 3 months (A) and 6 months (B).

Fig. 5.

Landmark analysis of survival based on response evaluation at 3 months (A) and 6 months (B).

Close modal
Table 1

Characteristics of the study group (237 patients)

FeatureaNo. (%)
Gender  
 Female 110 (46) 
Age (yrs)  
 ≥60 (median, 50 yrs) 74 (31) 
Splenomegaly  
 0–9 cm bcmb 157 (67) 
 ≥10 cm bcm 79 (33) 
Hemoglobin (g/dl)  
 <10 85 (36) 
WBC count (× 109/liter)  
 ≥50 60 (25) 
Platelet counts (× 109/liter)  
 <100 54 (23) 
 100–450 119 (50) 
 >450 64 (27) 
Cytogenetic evidence of clonal evolution 96 (41) 
Peripheral blasts (%)  
 0–9 201 (85) 
 10–14 21 (9) 
 ≥15 15 (6) 
Peripheral blasts+ promyelocytes (%)  
 20–29 13 (5) 
 ≥30 3 (1) 
Peripheral basophils (%)  
 7–19 49 (21) 
 ≥20 26 (11) 
Marrow blasts (%)  
 5–9 31 (13) 
 10–14 26 (11) 
 ≥15 45 (19) 
Marrow blasts+ promyelocytes (%)  
 20–29 32 (14) 
 ≥30 4 (2) 
Accelerated-phase criteria  
 Standard 180 (76) 
 Expanded 42 (18) 
 Second chronic phase 15 (6) 
Drug dose (mg/day)  
 400 26 (11) 
 600 211 (89) 
Prior therapyc  
 IFN-α with or without cytarabine 169 (71) 
 Allogeneic stem cell transplant 16 (7) 
 HHT 28 (12) 
 Hydroxyurea 23 (10) 
 Cytarabine with or without other chemotherapy combinations 28 (12) 
FeatureaNo. (%)
Gender  
 Female 110 (46) 
Age (yrs)  
 ≥60 (median, 50 yrs) 74 (31) 
Splenomegaly  
 0–9 cm bcmb 157 (67) 
 ≥10 cm bcm 79 (33) 
Hemoglobin (g/dl)  
 <10 85 (36) 
WBC count (× 109/liter)  
 ≥50 60 (25) 
Platelet counts (× 109/liter)  
 <100 54 (23) 
 100–450 119 (50) 
 >450 64 (27) 
Cytogenetic evidence of clonal evolution 96 (41) 
Peripheral blasts (%)  
 0–9 201 (85) 
 10–14 21 (9) 
 ≥15 15 (6) 
Peripheral blasts+ promyelocytes (%)  
 20–29 13 (5) 
 ≥30 3 (1) 
Peripheral basophils (%)  
 7–19 49 (21) 
 ≥20 26 (11) 
Marrow blasts (%)  
 5–9 31 (13) 
 10–14 26 (11) 
 ≥15 45 (19) 
Marrow blasts+ promyelocytes (%)  
 20–29 32 (14) 
 ≥30 4 (2) 
Accelerated-phase criteria  
 Standard 180 (76) 
 Expanded 42 (18) 
 Second chronic phase 15 (6) 
Drug dose (mg/day)  
 400 26 (11) 
 600 211 (89) 
Prior therapyc  
 IFN-α with or without cytarabine 169 (71) 
 Allogeneic stem cell transplant 16 (7) 
 HHT 28 (12) 
 Hydroxyurea 23 (10) 
 Cytarabine with or without other chemotherapy combinations 28 (12) 
a

Some patients did not have value coded.

b

bcm, below the costal margin.

c

Some patients had more than one regimen.

Table 2

Response to imatinib mesylate therapy

ResponseNo. (% of 200 evaluable patients)
CHR 160 (80) 
Cytogenetic response 90 (45) 
 Complete (Ph 0%) 47 (24) 
 Partial (Ph 1–34%) 21 (11) 
 Minor (Ph 35–90%) 22 (14) 
Insufficient metaphases 11 (6) 
PHR 19 (10) 
Resistant disease 16 (8) 
Withdrawal within 3 months/early death/treatment stopped for toxicity 3/1/1 
ResponseNo. (% of 200 evaluable patients)
CHR 160 (80) 
Cytogenetic response 90 (45) 
 Complete (Ph 0%) 47 (24) 
 Partial (Ph 1–34%) 21 (11) 
 Minor (Ph 35–90%) 22 (14) 
Insufficient metaphases 11 (6) 
PHR 19 (10) 
Resistant disease 16 (8) 
Withdrawal within 3 months/early death/treatment stopped for toxicity 3/1/1 
Table 3

Response to imatinib mesylate therapy according to duration of therapy

ResponseNo. (%)
3 months (200a)6 months (108a)12 months (36a)Overall (200a)
CHR 150 (75) 91 (84) 34 (94) 160 (80) 
Cytogenetic response 73 (37) 42 (39) 14 (39) 90 (45) 
 Complete 32 (16) 23 (21) 10 (28) 47 (24) 
 Partial 19 (10) 13 (12) 2 (6) 21 (11) 
 Minor 22 (11) 6 (6) 2 (6) 22 (11) 
Insufficient metaphases 17 (9) 12 (11) 7 (19) 11 (6) 
ResponseNo. (%)
3 months (200a)6 months (108a)12 months (36a)Overall (200a)
CHR 150 (75) 91 (84) 34 (94) 160 (80) 
Cytogenetic response 73 (37) 42 (39) 14 (39) 90 (45) 
 Complete 32 (16) 23 (21) 10 (28) 47 (24) 
 Partial 19 (10) 13 (12) 2 (6) 21 (11) 
 Minor 22 (11) 6 (6) 2 (6) 22 (11) 
Insufficient metaphases 17 (9) 12 (11) 7 (19) 11 (6) 
a

Number evaluated.

Table 4

Response to imatinib mesylate therapy according to the dose schedule received and accelerated-phase criteria (200 evaluable patients)

ParameterImatinib mesylate dose (mg daily)
Standard accelerated criteriaModified criteria
400 mg600 mg600 mg
No. evaluable for response 26 136 38 
No. (%) CHR 20 (77) 110 (81) 30 (79) 
No. (%) cytogenetic response 8 (31) 68 (50) 14 (37) 
 Complete 3 (12) 38 (28) 6 (16) 
 Partial 2 (8) 16 (12) 3 (8) 
 Minimal 3 (12) 14 (10) 5 (13) 
ParameterImatinib mesylate dose (mg daily)
Standard accelerated criteriaModified criteria
400 mg600 mg600 mg
No. evaluable for response 26 136 38 
No. (%) CHR 20 (77) 110 (81) 30 (79) 
No. (%) cytogenetic response 8 (31) 68 (50) 14 (37) 
 Complete 3 (12) 38 (28) 6 (16) 
 Partial 2 (8) 16 (12) 3 (8) 
 Minimal 3 (12) 14 (10) 5 (13) 
Table 5

Association of prognostic factors with major cytogenetic response

CharacteristicTotal no. of patientsNo. of patients with major response (%)P (categorical)P (continuous)
Total 200 68 (34)   
Interval between diagnosis and therapy (yrs)   0.01 <0.01 
 <1 50 24 (48)   
 1–3 51 20 (39)   
 ≥3 99 24 (24)   
Age (yrs)   0.01 0.04 
 <60 136 54 (40)   
 ≥60 64 14 (22)   
Spleen (cm bcm)a   <0.01 <0.01 
 0 105 51 (49)   
 1–9 23 7 (30)   
 10–19 60 10 (17)   
 ≥20 12 0 (0)   
Hemoglobin (g/dl)   0.11 0.01 
 <10 75 20 (27)   
 10–12 81 28 (35)   
 >12 44 20 (45)   
WBC count (× 109/liter)   <0.01 0.01 
 <10 72 36 (50)   
 10–50 76 21 (28)   
 >50 52 11 (21)   
Platelet count (× 109/liter)   0.35 0.47 
 <100 48 18 (38)   
 100–450 95 35 (37)   
 >450 57 15 (26)   
Peripheral blasts (%)   <0.01 <0.01 
 0 80 47 (59)   
 1–4 63 12 (19)   
 5–14 43 5 (12)   
 ≥15 14 4 (29)   
Marrow blasts (%)   <0.01 <0.01 
 <5 105 51 (49)   
 5–14 51 8 (16)   
 ≥15 44 9 (20)   
Peripheral basophils (%)   0.05 0.12 
 <5 102 42 (41)   
 5–19 73 17 (23)   
 ≥20 25 9 (36)   
Marrow basophils (%)   0.20 0.06 
 <5 109 43 (39)   
 5–20 63 18 (29)   
 ≥20 28 7 (25)   
Peripheral blasts+ pros (%)   0.01 <0.01 
 <10 154 60 (39)   
 10–20 31 4 (13)   
 ≥20 15 4 (27)   
Marrow blasts+ pros (%)   <0.01 <0.01 
 <10 117 52 (44)   
 10–20 49 10 (20)   
 ≥20 34 6 (18)   
Clonal evolution   0.23 0.23 
 No 114 35 (31)   
 Yes 85 33 (39)   
Dose (mg/day)   0.12 0.08 
 400 26 5 (19)   
 600 174 63 (36)   
Accelerated-phase criteria   0.18 0.13 
 Standard 162 59 (36)   
 Expanded 38 9 (24)   
CharacteristicTotal no. of patientsNo. of patients with major response (%)P (categorical)P (continuous)
Total 200 68 (34)   
Interval between diagnosis and therapy (yrs)   0.01 <0.01 
 <1 50 24 (48)   
 1–3 51 20 (39)   
 ≥3 99 24 (24)   
Age (yrs)   0.01 0.04 
 <60 136 54 (40)   
 ≥60 64 14 (22)   
Spleen (cm bcm)a   <0.01 <0.01 
 0 105 51 (49)   
 1–9 23 7 (30)   
 10–19 60 10 (17)   
 ≥20 12 0 (0)   
Hemoglobin (g/dl)   0.11 0.01 
 <10 75 20 (27)   
 10–12 81 28 (35)   
 >12 44 20 (45)   
WBC count (× 109/liter)   <0.01 0.01 
 <10 72 36 (50)   
 10–50 76 21 (28)   
 >50 52 11 (21)   
Platelet count (× 109/liter)   0.35 0.47 
 <100 48 18 (38)   
 100–450 95 35 (37)   
 >450 57 15 (26)   
Peripheral blasts (%)   <0.01 <0.01 
 0 80 47 (59)   
 1–4 63 12 (19)   
 5–14 43 5 (12)   
 ≥15 14 4 (29)   
Marrow blasts (%)   <0.01 <0.01 
 <5 105 51 (49)   
 5–14 51 8 (16)   
 ≥15 44 9 (20)   
Peripheral basophils (%)   0.05 0.12 
 <5 102 42 (41)   
 5–19 73 17 (23)   
 ≥20 25 9 (36)   
Marrow basophils (%)   0.20 0.06 
 <5 109 43 (39)   
 5–20 63 18 (29)   
 ≥20 28 7 (25)   
Peripheral blasts+ pros (%)   0.01 <0.01 
 <10 154 60 (39)   
 10–20 31 4 (13)   
 ≥20 15 4 (27)   
Marrow blasts+ pros (%)   <0.01 <0.01 
 <10 117 52 (44)   
 10–20 49 10 (20)   
 ≥20 34 6 (18)   
Clonal evolution   0.23 0.23 
 No 114 35 (31)   
 Yes 85 33 (39)   
Dose (mg/day)   0.12 0.08 
 400 26 5 (19)   
 600 174 63 (36)   
Accelerated-phase criteria   0.18 0.13 
 Standard 162 59 (36)   
 Expanded 38 9 (24)   
a

bcm, below the costal margin; pros, promyelocytes.

Table 6

Summary of logistic regression model

A. Model
VariableOdds ratio estimate95% CIP
Interval between diagnosis and therapy (baseline, 0–1 yr)    
 1–3 yrs 0.59 (0.23–1.49) 0.26 
 ≥3 yrs 0.29 (0.12–0.69) 0.005 
Spleen (baseline, 0 cm bcm)a    
 1–9 cm bcm 0.61 (0.21–1.78) 0.36 
 ≥10 cm bcm 0.24 (0.10–0.58) 0.002 
Peripheral blasts (baseline, 0%)    
 1–5% 0.22 (0.10–0.49) <0.001 
 >5% 0.29 (0.11–0.78) 0.01 
Dose (baseline, 400 mg)    
 600 mg 1.22 (0.38–3.92) 0.73 
A. Model
VariableOdds ratio estimate95% CIP
Interval between diagnosis and therapy (baseline, 0–1 yr)    
 1–3 yrs 0.59 (0.23–1.49) 0.26 
 ≥3 yrs 0.29 (0.12–0.69) 0.005 
Spleen (baseline, 0 cm bcm)a    
 1–9 cm bcm 0.61 (0.21–1.78) 0.36 
 ≥10 cm bcm 0.24 (0.10–0.58) 0.002 
Peripheral blasts (baseline, 0%)    
 1–5% 0.22 (0.10–0.49) <0.001 
 >5% 0.29 (0.11–0.78) 0.01 
Dose (baseline, 400 mg)    
 600 mg 1.22 (0.38–3.92) 0.73 
B. Application model
No. of unfavorable factorsbNo. of patientsNo. of patients with major cytogenetic response (%)
38 26 (68) 
66 31 (47) 
64 10 (16) 
32 1 (3) 
B. Application model
No. of unfavorable factorsbNo. of patientsNo. of patients with major cytogenetic response (%)
38 26 (68) 
66 31 (47) 
64 10 (16) 
32 1 (3) 
a

bcm, below the costal margin.

b

Unfavorable factors were interval between diagnosis and therapy of 3 years or more; splenomegaly ≥10 cm bcm; and peripheral blasts >0%.

Table 7

Survival by patient characteristics (univariate analysis)

CharacteristicTotalDeathEstimated survival proportionP
12 months18 months
Total222310.830.70
      
Interval between diagnosis and therapy (yrs)     0.17 
 <1 56 0.98 0.79  
 1–3 61 0.78 0.78  
 >3 105 20 0.80 0.66  
Age (yrs)     0.02 
 <60 154 16 0.87 0.75  
 ≥60 68 15 0.73 0.60  
Spleen (cm bcm)a     0.09 
 0 115 13 0.87 0.76  
 1–9 29 0.94 0.75  
 >9 78 15 0.69 0.53  
Hemoglobin (g/dl)     0.10 
 <10 82 17 0.76 0.63  
 10–12 88 12 0.82 0.72  
 >12 52 0.98 0.86  
WBC count (× 109/liter)     0.74 
 <10 76 13 0.81 0.65  
 10–50 88 13 0.84 0.75  
 >50 58 0.83 0.69  
Platelet count (× 109/liter)     0.24 
 <100 54 11 0.75 0.58  
 100–450 108 0.89 0.75  
 >450 60 11 0.81 0.73  
Peripheral blasts (%)     0.33 
 0 91 0.92 0.84  
 1–5 80 13 0.76 0.62  
 6–14 36 0.84 0.73  
 ≥15 15 0.82 0.62  
Marrow blasts (%)     0.17 
 0 31 0.97 0.97  
 1–5 101 0.86 0.69  
 6–14 45 0.75 0.75  
 ≥15 45 15 0.79 0.65  
Peripheral basophils (%)     0.68 
 0 40 0.85 0.68  
 1–5 94 12 0.87 0.76  
 6–19 62 10 0.84 0.68  
 ≥20 26 0.65 0.65  
Marrow basophils (%)     0.04 
 0 41 0.80 0.53  
 1–5 89 0.94 0.87  
 6–19 62 10 0.82 0.71  
 ≥20 30 0.64 0.51  
Peripheral blasts+ pros (%)     0.28 
 ≤5 146 15 0.86 0.73  
 6–10 26 0.87 0.87  
 ≥10 50 13 0.77 0.59  
Marrow blasts+ pros (%)     0.37 
 <10 132 11 0.86 0.76  
 10–19 54 10 0.85 0.72  
 ≥20 36 10 0.72 0.6  
Clonal evolution     0.01 
 No 124 11 0.86 0.80  
 Yes 96 19 0.81 0.59  
Accelerated criteria     0.58 
 Standard 180 29 0.83 0.71  
 Expanded 42 NA NA  
Dose (mg daily)     0.04 
 400 26 14 0.73 0.58  
 600 196 17 0.85 0.78  
CharacteristicTotalDeathEstimated survival proportionP
12 months18 months
Total222310.830.70
      
Interval between diagnosis and therapy (yrs)     0.17 
 <1 56 0.98 0.79  
 1–3 61 0.78 0.78  
 >3 105 20 0.80 0.66  
Age (yrs)     0.02 
 <60 154 16 0.87 0.75  
 ≥60 68 15 0.73 0.60  
Spleen (cm bcm)a     0.09 
 0 115 13 0.87 0.76  
 1–9 29 0.94 0.75  
 >9 78 15 0.69 0.53  
Hemoglobin (g/dl)     0.10 
 <10 82 17 0.76 0.63  
 10–12 88 12 0.82 0.72  
 >12 52 0.98 0.86  
WBC count (× 109/liter)     0.74 
 <10 76 13 0.81 0.65  
 10–50 88 13 0.84 0.75  
 >50 58 0.83 0.69  
Platelet count (× 109/liter)     0.24 
 <100 54 11 0.75 0.58  
 100–450 108 0.89 0.75  
 >450 60 11 0.81 0.73  
Peripheral blasts (%)     0.33 
 0 91 0.92 0.84  
 1–5 80 13 0.76 0.62  
 6–14 36 0.84 0.73  
 ≥15 15 0.82 0.62  
Marrow blasts (%)     0.17 
 0 31 0.97 0.97  
 1–5 101 0.86 0.69  
 6–14 45 0.75 0.75  
 ≥15 45 15 0.79 0.65  
Peripheral basophils (%)     0.68 
 0 40 0.85 0.68  
 1–5 94 12 0.87 0.76  
 6–19 62 10 0.84 0.68  
 ≥20 26 0.65 0.65  
Marrow basophils (%)     0.04 
 0 41 0.80 0.53  
 1–5 89 0.94 0.87  
 6–19 62 10 0.82 0.71  
 ≥20 30 0.64 0.51  
Peripheral blasts+ pros (%)     0.28 
 ≤5 146 15 0.86 0.73  
 6–10 26 0.87 0.87  
 ≥10 50 13 0.77 0.59  
Marrow blasts+ pros (%)     0.37 
 <10 132 11 0.86 0.76  
 10–19 54 10 0.85 0.72  
 ≥20 36 10 0.72 0.6  
Clonal evolution     0.01 
 No 124 11 0.86 0.80  
 Yes 96 19 0.81 0.59  
Accelerated criteria     0.58 
 Standard 180 29 0.83 0.71  
 Expanded 42 NA NA  
Dose (mg daily)     0.04 
 400 26 14 0.73 0.58  
 600 196 17 0.85 0.78  
a

bcm, below the costal margin, pros, promyelocytes; NA, not applicable.

1
Kantarjian H. M., Deisseroth A., Kurzrock R., Estrov Z., Talpaz M. Chronic myelogenous leukemia: a concise update.
Blood
,
82
:
691
-703,  
1993
.
2
Kantarjian H. M., O’Brien S., Anderlini P., Talpaz M. Treatment of chronic myelogenous leukemia: current status and investigational options.
Blood
,
87
:
3069
-3081,  
1996
.
3
Faderl S., Talpaz M., Estrov Z., Kantarjian H. Chronic myelogenous leukemia: biology and therapy.
Ann. Intern. Med.
,
131
:
207
-219,  
1999
.
4
Kantarjian H. M., Dixon D., Keating M., Talpaz M., Walters R. S., McCredie K. B., Freireich E. J. Characteristics of accelerated disease in chronic myelogenous leukemia.
Cancer (Phila.)
,
61
:
1441
-1446,  
1988
.
5
Karanas A., Silver R. T. Characteristics of the terminal phase of chronic granulocytic leukemia.
Blood
,
32
:
445
-459,  
1968
.
6
Savage D. G., Szydlo R. M., Chase A., Apperley J. F., Goldman J. M. Bone marrow transplantation for chronic myeloid leukaemia: the effects of differing criteria for defining chronic phase on probabilities of survival and relapse.
Br. J. Haematol.
,
99
:
30
-35,  
1997
.
7
Arlin Z. A., Silver R. T., Bennet J. M. Blastic phase of chronic myeloid leukemia (blVML): a proposal for standardization of diagnostic and response criteria.
Leukemia (Baltimore)
,
4
:
756
-757,  
1990
.
8
Majlis A., Smith T., Talpaz M., O’Brien S., Rios M. B., Kantarjian H. M. Significance of cytogenetic clonal evolution in chronic myelogenous leukemia.
J. Clin. Oncol.
,
14
:
196
-203,  
1996
.
9
Gratwohl A., Hermans J., Niederwieser D., Frassoni F., Arcese W., Gahrton G., Bandini G., Carreras E., Vernant J. P., Bosi A., for the Chronic Myeloid Leukemia Working Party of the European Group for Bone Marrow Transplantation. Bone marrow transplantation for chronic myeloid leukemia: long term results.
Bone Marrow Transplant.
,
12
:
509
-516,  
1993
.
10
McGlave P. Therapy of chronic myelogenous leukemia with related or unrelated donor bone marrow transplantation.
Leukemia (Baltimore)
,
6 (Suppl.)
:
115
-117,  
1992
.
11
Clift R. A., Buckner C. D., Thomas E. D., Bryant E., Anasetti C., Bensinger W. I., Bowden R., Deeg H. J., Doney K. C., Fisher L. D. Marrow transplantation for patients in accelerated phase of chronic myeloid leukemia.
Blood
,
84
:
4368
-4373,  
1994
.
12
Sacchi S., Kantarjian H. M., O’Brien S., Cortes J., Rios M. B., Giles F. J., Beran M., Koller C. A., Keating M. J., Talpaz M. Chronic myelogenous leukemia in nonlymphoid blastic phase. Analysis of the results of first salvage therapy with three different treatment approaches for 162 patients.
Cancer (Phila.)
,
86
:
2632
-2641,  
1999
.
13
Kantarjian H., Talpaz M., Kontayiannis D., Gutterman J., Keating M. J., Estey E. H., O’Brien S., Rios M. B., Beran M., Deisseroth A. Treatment of chronic myelogenous leukemia in accelerated and blastic phase with daunorubicin, high-dose cytarabine and granulocyte macrophage colony stimulating factor.
J. Clin. Oncol.
,
10
:
398
-405,  
1992
.
14
Kantarjian H. M., Keating M., Estey E. H., O’Brien S., Pierce S., Beran M., Koller C., Feldman E., Talpaz M. Treatment of advanced stages of Philadelphia chromosome positive chronic myelogenous leukemia with interferon-α and low dose cytarabine.
J. Clin. Oncol.
,
10
:
772
-778,  
1992
.
15
Kantarjian H. M., Talpaz M., Smith T. L., Cortes J., Giles F. J., Rios M. B., Mallard S., Gajewski J., Murgo A., Cheson B., O’Brien S. Homoharringtonine and low-dose cytarabine in the management of late chronic-phase chronic myelogenous leukemia.
J. Clin. Oncol.
,
18
:
3513
-3521,  
2000
.
16
Giles F. J., Cortes J. E., Baker S. D., Thomas D. A., O’Brien S., Smith T. L., Beran M., Bivins C., Jolivet J., Kantarjian H. M. Troxacitabine, a novel dioxolane nucleoside analog, has activity in patients with advanced leukemia.
J. Clin. Oncol.
,
19
:
762
-771,  
2001
.
17
Druker B. J., Tamura S., Buchdunger E., Ohno S., Segal G. M., Fanning S., Zimmermann J., Lydon N. B. Effects of a selective inhibitor of the ABL tyrosine kinase on the growth of BCR-ABL positive cells.
Nat. Med.
,
2
:
561
-566,  
1996
.
18
Deininger M. W., Goldman J. M., Lydon N., Melo J. V. The tyrosine kinase inhibitor CGP57148B selectively inhibits the growth of BCR-ABL-positive cells.
Blood
,
90
:
3691
-3698,  
1997
.
19
Beran M., Cao X., Estrov Z., Jeha S., Jin G., O’Brien S., Talpaz M., Arlinghaus R. B., Lydon N. B., Kantarjian H. Selective inhibition of cell proliferation and BCR-ABL phosphorylation in acute lymphoblastic leukemia cells expressing Mr 190,000 BCR-ABL protein by a tyrosine kinase inhibitor (CGP-57148).
Clin. Cancer Res.
,
4
:
1661
-1672,  
1998
.
20
Druker B. J., Sawyers C. L., Kantarjian H., Resta D. J., Reese S. F., Ford J. M., Capdeville R., Talpaz M. Efficacy and safety of an abl specific tyrosine kinase inhibitor as targeted therapy for chronic myeloid leukemia.
N. Engl. J. Med.
,
344
:
1031
-1037,  
2001
.
21
Druker B. J., Sawyers C. L., Kantarjian H., Resta D. J., Reese S. F., Ford J. M., Capdeville R., Talpaz M. Activity of an abl specific tyrosine kinase inhibitor in patients with chronic myeloid leukemia in blast crisis and other Philadelphia chromosome positive acute leukemias.
N. Engl. J. Med.
,
344
:
1038
-1042,  
2001
.
22
Kantarjian H., Sawyers C., Hocchaus A. Phase II study of STI571, a tyrosine kinase inhibitor, in patients (pts) with resistant or refractory Philadelphia chromosome-positive chronic myeloid leukemia (Ph+CML).
Blood
,
96
:
470a
2000
.
23
Talpaz M., Silver R. T., Druker B. A Phase II study of STI 571 in adult patients with Philadelphia chromosome positive chronic myeloid leukemia in accelerated phase.
Blood
,
96
:
496a
2000
.
24
Sawyers C., Hochhaus A., Feldman E. A Phase II study to determine the safety and anti-leukemic effects of STI571 in patients with Philadelphia chromosome positive chronic myeloid leukemia in myeloid blast crisis.
Blood
,
96
:
503a
2000
.
25
Ottman O. G., Sawyers C., Druker B. A Phase II study to determine the safety and anti-leukemic effects of STI571 in adult patients with Philadelphia chromosome positive acute leukemias.
Blood
,
96
:
828a
2000
.
26
Anderson J. R., Cain K. C., Gelber R. D. Analysis of survival by tumor response.
J. Clin. Oncol.
,
1
:
710
-719,  
1983
.
27
Cox D. R., Snell E. J. Logistic regression for binary response variables.
Analysis of Binary Data
,
33
-43, Chapman and Hall London  
1989
.
28
Cleveland W. S. Robust locally weighted regression and smoothing scatterplots.
J. Am. Stat. Assoc.
,
74
:
829
-836,  
1979
.
29
Kantarjian H. M., O’Brien S., Keating M., Beran M., Estey E., Giralt S., Kornblau S., Rios M. B., de Vos D., Talpaz M. Results of decitabine therapy in accelerated and blastic phase of chronic myelogenous leukemia.
Leukemia (Baltimore)
,
11
:
1617
-1620,  
1997
.