Administration of the tyrosine kinase inhibitor imatinib mesylate (STI-571) leads to major hematological responses in 90% of CML patients with chronic phase who have failed interferon and in 40-60% of those with advanced disease. The drug is well-tolerated with Grade III-IV non-hematological toxicities occurring in under 10%. Skin toxicity has been an infrequently reported side effect. However, of the 145 patients treated with imatinib mesylate at the Dana-Farber Cancer Institute on one of seven protocols [Phase II trials in chronic phase CML (73), accelerated phase CML (35), or blast crisis (24); or Phase III interferon/ara-C 'v' imatinib mesylate in newly diagnosed chronic phase (13)] 18 (12%) developed a grade I-III skin rash, of whom 12 underwent skin biopsies. The rash was generally pruritic and of an erythematous maculo-papular quality. Rashes occurred at imatinib mesylate doses of both 300 mg (2), 400 mg (8) or 600 mg (8) orally daily after a median duration of therapy of 2 months (range 1-11 months). Skin biopsies revealed changes consistent with drug-induced hypersensitvity reactions. The biopsies showed a variable perivascular lymphoid infiltrate with eosinophils that ranged from rare to prominent. In eight patients rashes were severe enough to warrant discontination of therapy with imatinib mesylate. In each case patients were rechallenged after the rash resolved; the same (2) or a lower dose (6) was tolerated except for one patient who required removal from the study. In 10 cases the rash resolved despite continuation of imatinib mesylate. Topical steroids were employed in nine patients, oral anti-pruritics in three. Characteristics of patients who developed a rash on imatinib mesylate included median age 60 years (range: 54-78), sex (12/18 female), CML stage (chronic phase in 9, accelerated phase in 8 and blast crisis in 1), median WBC at entry of 60K/ul (range 10-200K/ul), median platelet count at study entry 12 K/ul (range 3-120 K/ul), prior exposure to interferon in 14/18, allopurinol use in 2 and more than 2 concomittant medicines in 8. We compared the age, sex, disease stage, and starting dose of imatinib mesylate in those with and without rashes and found no predictive factors for this toxicity except that rashes were more likely in females (12/59) than in males (6/86) [p=0.02]. In summary we found that rashes are not uncommon in patients with CML on imatinib mesylate; although usually self-limited, in some cases the rash is severe enough to limit drug administration, although succesful re-challenge is possible. The rash appears to be on the basis of a hypersensitivity reaction, but the mechanism (related directly to tyrosine kinase inhibition 'v' idiosyncratic immunological effect) remains unknown.

Imatinib (STI571, Gleevec, Novartis) is a potent and selective inhibitor of the BCR-ABL tyrosine kinase, the molecular abnormalitiy which causes chronic myeloid leukemia (CML) and Philadelphia (Ph) chromosome positive acute lymphoblastic leukemia (ALL). A phase I trial of imatinib in patients with CML in blast crisis and Ph+ ALL showed responses in 55% of patients with myeloid blast crisis and 70% with lymphoid blast crisis or Ph+ ALL [Druker et al NEJM 344, 1038-42, 2001]. Approximately 33% of patients with myeloid blast crisis remained in remission on therapy for up to one year however nearly all patients with lymphoid blast crisis or Ph+ ALL relapsed. Central nervous system (CNS) relapse occurred in 5 out of 24 patients with CML lymphoid blast crisis (2), Ph+ ALL (2) or biphenotypic markers (1) treated on imatinib protocols at our institution. Relapse occurred despite having had complete peripheral blood (5) and bone marrow responses (3). Median time to CNS relapse was on day 32 [range day 23 to day 100]. Four of the five patients had additional cytogenetic abnormalities. CNS relapse in the setting of complete hematologic remission raised the possibility that imatinib may not penetrate into the cerebral spinal fluid (CSF). Simultaneous plasma and CSF imatinib levels were determined in four patients by liquid chromatography and mass spectrophotometric assay. Levels of imatinib were found to be 74 fold lower in CSF than in plasma with mean concentrations of 0.044 mg/ml (0.088mM) +/- 0.029 in the CSF and 3.27 mg/ml (6.54mM) +/- 0.93 in the plasma. Imatinib CSF levels were substantially below the concentration required for inhibition of cellular phosphorylation by BCR-ABL (concentration required for 50 percent inhibition, 0.25mM) and killing of BCR-ABL cell lines in vitro. These results suggest that imatinib may not penetrate the blood/brain barrier. CNS prophylaxis should be given for all CML lymphoid blast crisis and Ph+ ALL patients being treated with imatinib.

Gleevec (imatinib mesylate, formerly STI571) is a selective inhibitor of the Bcr-Abl tyrosine kinase and recently received approval for use in chronic myelogenous leukemia (CML) patients in blast crisis, accelerated, and chronic phase after failing interferon. Ara-C has well described modest activity in CML. In vitro studies of imatinib plus Ara-C using CML cell lines and colony forming assays of CML patient samples have shown synergistic anti-proliferative effects. As part of a Phase I trial, 62 patients were treated with a combination of imatinib plus low dose Ara-C. Imatinib was administered to patients with chronic phase CML either resistant or intolerant to interferon (n=22) at a dose of 400 mg daily for 28 days with Ara-C at escalating doses of 5, 10, or 20 mg/m² on days 15 through 28 of each cycle. An additional cohort of patients received imatinib at a dose of 600 mg daily for 28 days and Ara-C at 20 mg/m² (days 15-28) of each cycle. Of the 22 patients enrolled 91% (n=20) remain on study. The median duration of treatment is 299 days (range 79 to 376). Overall, the rate of complete hematologic responses is 82% (n=18). Major cytogenetic responses (complete plus partial, Ph+ metaphases < 35%) were seen in 32% (n=7) of patients. These response rates are similar to the rates observed in Phase I clinical trials of imatinib alone. Grade 3 or 4 hematologic toxicities occurred in 68% (n=15) of patients. Grade 3 or 4 non-hematologic toxicities occurred in 45% (n=10) of patients and consisted of elevated liver enzymes (n=3), hypophosphatemia (n=1), increased weight (n=1), rash (n=1) and systemic symptoms (n=7), including fevers, fatigue, and abdominal pain. Although no clear increase in Grade 3 or 4 hematologic or non-hematologic toxicities occurred between dose cohorts, no patients on the 600 mg dose of imatinib and 20 mg/m² of Ara-C cohort were able to maintain the original dose, suggesting this results in dose limiting toxicity. Imatinib was also administered to myeloid (n=14) or lymphoid (n=14) blast crisis patients, or Ph+ ALL (n=4) patients at escalating doses of 400, 600, or 1000 mg daily with Ara-C at a dose of 20 mg/m² on days 15 through 28 of each cycle. Response rates, durability of response and toxicities were similar to those reported in Phase I studies of imatinib alone. A total of 11 patients (34%) went on to receive a transplant. A final cohort of patients with myeloid blast crisis (n=8) relapsing while on imatinib were continued on treatment with imatinib daily and Ara-C was added at 20 mg/m² (days 1-14) of each cycle. All patients discontinued for disease progression. These data indicate that low dose Ara-C can be safely administered to blast crisis patients, but is unlikely to provide substantial benefit or salvage relapses. Additional studies in chronic phase patients are required. Based on these limited data we conclude that imatinib can be administered safely in combination with low dose Ara-C and that for chronic patients the maximum tolerated dose is 400 mg of imatinib daily with Ara-C 20 mg/m² given on days 15-28 of each cycle.

Although imatinib mesylate (Gleevec, formerly STI571) has demonstrated remarkable response rates in chronic phase CML patients who have failed therapy with interferon-alpha (IFN-a) approximately 50% of these patients fail to achieve a major cytogenetic response. Even in patients who have achieved cytogenetic remissions, the majority of these patients have detectable Bcr-Abl transcripts by RT-PCR. In both patients populations, there is concern that over time resistance may develop to imatinib as a single agent. Based on in-vitro data showing an additive or synergistic effect using IFN-ain combination with imatinib, we have designed a phase I/II study to determine the safety, tolerability, and anti-leukemic effects of this combination in chronic phase CML. In phase I the primary objective is to establish the maximum tolerated dose (MTD) of IFN-a when administered in combination with imatinib at a dose of 400-600mg per day. For the purposes of this study, the MTD is defined as the dose that produces grade 3 or 4 non-hematological toxicity in less than 33% of patients and patients are excluded if they have a history of > Grade 3 non-hematologic toxicity due to interferon. Hematologic and cytogenetic response rates are secondary endpoints. Accrual to this protocol began in June 2001. Thus far we have treated patients with 400mg of imatinib with IFN-a 3MIU, three times per week (9MIU/wk) and 3 MIU daily (21MIU/wk). Interferon is added 2 weeks after initiation of imatinib. Patients are currently being enrolled in a cohort using IFN-a 5MIU daily (35 MIU/wk) and a 4th cohort is planned using imatinib at a dose of 600mg per day with 5 MIU daily of IFN-alpha. Thus far, the combination of imatinib with low dose interferon does not appear to yield any increase in non-hematological toxicity as compared to imatinib alone. However the combination appears to be more myelosuppressive than imatinib alone with several patients developing prolonged grade 3 hematologic toxicity requiring dose reduction. As expected, the majority of patients have achieved a complete hematologic response. However, it is too early to report on cytogenetic responses at this time. Data collection and patient accrual are ongoing and results with a maximum duration of follow up of 6 months will be presented.

CML invariably progresses into fatal acute leukemia. So far, only 2 treatments were offered: IFN or allogenic SCT. Recently, tyrosine kinase inhibitor STI571 (imatinib) has become available. Conventional karyotyping usually appreciates response to treatment but needs bone marrow aspirate and has low sensitivity. Bcr-Abl mRNA, found in most CML patients, allows minimal residual disease detection by RT-PCR which needs to be quantitative to predict prognosis. We developed a real-time RT-PCR assay to quantitate Bcr-Abl transcript. Our method, assessed in IFN-treated and allografted patients, was applied to imatinib-treated patients. Methods: Bcr-Abl expression was quantitated in blood samples using Taqman chemistry and a LightCycler (Roche). Copy number of target sequence (Bcr-Abl cDNA) or reference sequence (housekeeping gene PBGD cDNA) were calcutated using standard curves. The results were expressed as ratio of Bcr-Abl on PBGD. Sensitivity reached 10-5. Patients: 40 IFN-treated, 30 allografted and 52 imatinib-treated patients were molecularly and cytogenetically serially investigated. In the imatinib group were: 38 IFN resistant-intolerant and 14 accelerated phases. Results: In IFN group, Bcr-Abl transcript levels correlated cytogenetic results in 142 matching samples (r=0.92) and cytogenetic categories after 1 yr treatment (P<0.001). In 30 allografted patients, Bcr-Abl level predicted relapse in 10 patients and response in 7 who received DLI. In the imatinib group, Bcr-Abl transcript decrease was very rapid. Time length for 2 log (100x) decrease was 4 mths (3-6) in 20 imatinib-responder patients vrs 16 mths (7-40) in 20 IFN-responder patients. Thus, short term response (3 mths) was evaluated in 52 imatinib patients which were divided in 4 categories of transcript level decrease. A major response was a superior to 1 log (>10x) decrease. A partial response was an inferior to 1 log (<10x) but superior to twice (>2x) decrease. A "no response" was an inferior to 2x increase or decrease. An escape was a superior to 2x increase. In IFN resistant-intolerant group, 18/38 (48%) had a major response, 12 (32%) had a partial response, 6 (15%) did not respond and 2 escaped. In accelerated phase group, 7/14 (50%) had a major response, 5/14 had a partial response, 1 did not respond and 1 escaped. For 2 patients, Bcr-Abl level went below the assay sensitivity level (10-5). Patients who responded after 3 mths stayed in good response. Cytogenetic responses, available in 34 cases, correlated to real time PCR results. For 12 patients in complete cytogenetic response after 3 mths, real time RT-PCR appears as the only evaluation means. Conclusion: Molecular monitoring of imatinib treated patients is a reliable alternative to cytogenetics. A Bcr-Abl transcript level major decrease was observed in 25/52 patients (48%) after 3 mths of treatment suggesting a dramatic improvement in CML therapy. However, patient numbers are low and follow up is short [9 mths (6-12)]. Larger studies are needed to warrant these results.

Imatinib mesylate (Gleevec, formerly STI571), an ABL tyrosine kinase inhibitor, has shown significant activity and minimal toxicity in CML patients, including those in the accelerated phase. Single agent imatinib at 400-600mg/day has yielded sustained (>4 weeks) hematologic responses in 69% in patients with accelerated CML, with a 34% complete hematologic response rate (CHR) and as well a 24% major cytogenetic response (MCR) rate in phase II clinical trials. Likely a result of disease severity, therapy-associated myelosuppression has been seen in a high proportion of accelerated phase patients; grade 4 leukopenia occurred in 13-18% of patients in these studies. In order to reverse severe neutropenia while maintaining therapy, we employed the use of myeloid growth factors G-CSF and GM-CSF in 13 patients at Oregon Health & Sciences University with accelerated phase CML, enrolled in multi-center phase II trials, who experienced grade 4 (ANC<500/mm3) neutropenia during Imaninib therapy. 12 patients received G-CSF (Neupogen) 300-480mg/day and 2 received GM-CSF (Leukine) 500 mg/day. A total of 21 courses of these agents were given, with 5 patients receiving repeat courses and one patient receiving courses of both agents. Guidelines for myeloid growth factor use required concurrent imatinib therapy and discontinuation of growth factors once ANC rose beyond or was sustained between 1500 and 2500/mm3. G- or GM-CSF courses ranged from 2-60 days (mean 16 days). 62% (8/13) of patients responded, having resolution of neutropenia to NCI grade 1 level (>1500/mm3); ANC values of >2500/ mm3 were reached in 5/13 (38%). Concurrent imatinib and G-CSF or GM-CSF therapy was well tolerated in all patients. With a median follow-up of 8 months, the rate of sustained hematologic response in this cohort is 77%, and the major cytogenetic response rate is 38% which is comparable to the rates seen in the overall Phase II study population as noted above. However, given that the cohort was defined by increased myelosuppression, it is not surprising that the rate of CHR with peripheral blood recovery in this group is lower than seen in the phase II accelerated trial (15% versus 34%). These results suggest no adverse effect on leukemia response to imatinib with the use of myeloid growth factors for grade 4 leukopenia. The use of myeloid growth factors in a small subset of accelerated CML patients treated on a phase II study with imatinib demonstrates the feasibility of restoring peripheral blood neutrophil counts in the majority of patients without adversely the affect anti-leukemic effects of imatinib. Further follow-up of such patients is warranted to determine the durability of response and detection of any differential effects of this combination.

Imatinib is a selective inhibitor of BCR-ABL tyrosine kinase which has recently been approved for the treatment of interferon-refractory or intolerant chronic myelogenous leukemia (CML). The pivotal studies of the agent did not allow patients who previously underwent SCT to participate; however, the expanded access protocols did not exclude these patients. Between May, 2000 and May 2001, we enrolled 166 patients into two multicenter protocols for patients with interferon-refractory or intolerant chronic phase (CP) or accelerated phase (AP) sponsored by Novartis. 12 patients had undergone a SCT (6 autologous, 6 allogeneic) at a median of 44 months prior to beginning imatinib. 10 of 12 patients had failed or been intolerant to interferon therapy after relapse (with donor lymphocyte infusions (DLI) in the allogeneic patients)and 2 patients with AP were treated immediately. Patients had relapsed a median of 21 months after transplant. 8 patients were in CP and received 400 mg imatinib orally each day and 4 patients were in AP (based on additional chromosomal abnormalities) and received 600 mg imatinib orally each day. Median follow up on study at time of report is 8 months. 3 (2 CP, 1 AP) of the 6 autologous SCT patients had complete cytogenetic responses (CCR) at 6 months (1 patient who was interferon intolerant was only 35% Philadelphia positive at trial start). One additional patient in CP had a major cytogenetic response (MCR) at 9 mos; 1 CP and 1 AP patient had no response at 3 and 12 months. 4 patients did not require dose adjustment, 1 patient had grade 3 neutropenia that resolved, and 1 patient required dose reduction to 300 mg a day due to recurrent neutropenia. 5 matched sibling allogeneic SCT patients have been treated, and 4 had CCR at 6 months without requirement for dose reduction; and 1 at 300 mg a day for recurrent thrombocytopenia. 2 of these patients had received DLI within 6 months which may have contributed to the responses; however, both were progressing hematologically at initiation of imatinib therapy. One patient developed aplasia associated with a MCR and evidence of donor cells in marrow; however, a stem cell boost was given prior to count recovery. One patient who relapsed with additional clonal chromosomal abnormalities after unrelated DLI had a CCR at 3 months without hematologic toxicity. These data support the safety and activity of imatinib in patients relapsed after SCT.

In chronic phase CML patients treated with imatinib mesylate (Gleevec, formerly STI571), the durability of responses remains one of the major questions. As part of multi-institutional Phase I and Phase II clinical trials of imatinib in chronic phase CML patients who have failed interferon, we have followed a total of 66 patients treated with doses of imatinib greater than 300 mg. 97% (64/66) of these patients have achieved a complete hematological response and 50% have achieved a major cytogenetic response (<35% Ph+ metaphases) with 39% a complete cytogenetic response. This data is comparable to that reported in the Phase I and II clinical trials with imatinib. With a median duration of follow-up of 18 months (range 2.5 to 40.1), 46/64 (72%) remain as complete hematological responders but 18/64 (28%) have lost a complete hematological response, defined as persistently elevated WBC or plts or progression to accelerated or blast phases. A variety of pre-treatment characteristics that might correlate with the likelihood of relapse were analyzed, including age, sex, body surface area, LDH, Hgb, WBC, Plts, % blasts in peripheral blood or marrow, % basophils, % Ph positivity, and presence of additional cytogenetic abnormalities. Data were analyzed by univariate Cox regression analysis with all pre-treatment characteristics treated as independent variables. Characteristics including Hgb, Hct, WBC, and Plts were further analyzed by dichotomizing them into normal/abnormal according to the corresponding normal ranges. From this analysis, the only pretreatment characteristics that correlated with relapse were an elevated platelet count (p=0.014) and a low Hct (p = 0.013). Thus, 45% (14/31) of patients with elevated platelet counts have relapsed, whereas only 12% (4 of 33) patients with normal platelet counts have relapsed. 41% (14/34) of patients with low Hcts have relapsed, whereas only 10% (3/30) of patients with normal Hcts have relapsed. We next analyzed whether there was a difference in hematological relapse between patients with no cytogenetic response (>65% Ph+) and those with a major cytogenetic response (<35% Ph +) at any time on study. Overall, 48% (15/31) of patients with no cytogenetic response have relapsed, whereas only 9.0% (3 of 33) of patients with major cytogenetic responses have relapsed (p = 0.0136). The three patients with major cytogenetic responses who relapsed had discontinued therapy, one due to recurrent Grade 3 elevations in LFTs, the other two voluntarily. One of these patients has regained a major cytogenetic response after restarting imatinib therapy. We next sought to determine whether 6 months was sufficient to determine the probability of obtaining a cytogenetic response. At 6 months, 30 of 66 patients were >65% Ph+. Of these 30 patients, 7 (23%, 95% CI=10-42%) later achieved a major cytogenetic response (Ph+ < 35%). Our data demonstrate that the major prognostic factor for relapse of chronic phase CML patients treated with imatinib is the lack of a cytogenetic response. Additional data is required to determine if there is a time after which achievement of a major cytogenetic response becomes unlikely.

G-CSF start

G-CSF dose incr.

G-CSF stop

<-------------Leukapheresis----------->

Date

5/7

6/15

6/19

6/20

6/21

6/22

8/6

Sample

marrow

PB

PB

PBSC

PB

PBSC

PB

PBSC

PB

PBSC

marrow

Test

Q-RT- PCRa

Cyto.b

Q-RT- PCRa

Q-RT-PCRa

Q-RT-PCRa

Q-RT-PCRa

Q-RT-PCRa

Q-RT- PCRa

Cyto.b

Results

0.00

NL

2.73

64.3

0.00

1.45

0.00

37.45

13.89

46.82

5.11

47.9

NL

The study aims were to summarize our single institutional experience with imatinib mesylate as a uniform treatment of patients (pts) with Ph+ CML BP, and to compare this experience with previous studies using standard cytarabine-based combination regimens. 61 pts with Ph+ CML BP were treated at our institution with imatinib mesylate 400 mg to 1000 mg orally daily as part of the multiinstitutional Novartis sponsored trials. Their median age was 53 years; 55 pts had non-lymphoid blastic phase; 42 pts received imatinib mesylate as first salvage. Overall, 29 pts (48%) achieved objective responses, including complete hematologic response (CHR) in 16 (26%), partial hematologic response in 1 (2%), hematologic improvement in 9 (15%), and return to second chronic phase in 3 (5%). 13 (21%) patients achieved cytogenetic responses: complete in 5 (8%), partial (Ph < 35%) in 3 (5%) and minor (Ph 34% to 90%) in 5 (8%). Objective responses were 47% with non-lymphoid blastic phase and 50% with lymphoid blastic phase. The estimated median overall survival was 8 months; the estimated one-year survival rate was 32%. Duration of response was similar with or without cytogenetic response. Survival by landmark analysis at 8 weeks showed significantly better survival with response (28 pts) versus no response (23 pts): 1-yr survival rates 45% vs 15% (p=0.01). No prognostic factors predicted for response. Platelet counts and response to therapy (landmark analysis at 8 weeks) were associated with differences in survival. Compared with standard cytarabine combinations given as first salvage to 133 pts with non-lymphoid CML BP, imatinib mesylate therapy was associated with a higher response rate (43% versus 29%), longer median survival (9 versus 4 months; p=0.004), and lower 4-week induction mortality (3% versus 15%; p=0.05). We conclude that imatinib mesylate therapy is effective in Ph+ CML BP, and is superior to and less toxic than standard cytarabine-containing regimens. Imatinib mesylate has now become the new therapeutic cornerstone for combined regimens to improve prognosis in CML BP.

237 adult patients (pts) with Ph+ CML AP were treated with imatinib mesylate 400-600 mg P.O. daily at our institution as part of 2 Novartis sponsored multi-institutional multinational studies: Novartis 109 the pivotal study (N=58) and Novartis 114 the expanded access study (N=179). 193 pts are evaluable with more than 3 months of follow-up. 156 pts had the classical CML AP criteria (Cancer 61:1441, 1988); 33 pts were treated for blasts 10-14%, blasts + pros 20-29%, or spleen 10 cm bcm or 50% increase over 4 weeks (modified CML-AP criteria);4 pts had second chronic phase. 26 received imatinib mesylate 400 mg/D, and 167 pts had imatinib mesylate 600 mg/D. Their median age was 50 years. Overall, 162 pts (84%) achieved CHR, 107 (55%) had a cytogenetic response (Ph < 90%): major (Ph < 35%) in 79 (41%); complete (Ph 0%) in 57 (30%). With a median follow up of 8.4 months, 167 patients (87%) are alive. The estimated 1.5-year survival rate was 75%, and remission duration rate 61%. Prognostic factors associated with lower major CG response rates (p 0.02) were: age 60 yrs, splenomegaly 10 cm bcm, longer duration of chronic phase > 3 yrs, WBC > 10 x 109/L, marrow blasts 15%, and STI dose 400 mg daily. Prognostic factors associated with worse survival (p < 0.02) were: age 60 yrs, hemoglobin < 10 g/dl marrow blasts 15%, cytogenetic clonal evolution and STI dose 400 mg daily and failure to achieve major CG response. Patients treated with 600 vs 400 mg had significantly better major (44% vs 19%, p= 0.02) and complete (32% vs 15%, p= 0.11) CG response rates, and 1.5 yr survival rates (78% vs 67%, p <0.01). Patients with "modified" CML AP criteria had similar major CG response and survival rates. By multivariate analysis, factors independently predictive negatively for major CG response were (p < 0.05): diagnosis to therapy > 3 years, and spleen size > 10 cm bcm. Those associated with worse survival were (p < 0.05): older age, failure to achieve major cytogenetic response, and cytogenetic clonal evolution. In summary imatinib mesylate is the most active single agent therapy in accelerated phase. Imatinib mesylate combinations with interferon alpha, cytarabine, homoharringtonine, decitabine or others are warranted in CML AP.

		No. (%) Response at Higher Dose of STI
STI resistance status	No. pts	CHR	CG response
No CHR post 3 months	6	4(67%)	1-CR
Ph 100% post 12 months	6/8 evaluable	NA	1-minor (Ph 45%)
Loss of CG Response
-Back to Ph 100%	4	NA	2-1 CR, 1 minor (Ph 85%)
-Ph 35-90%	1	NA	1 PR (Ph 55% 10%)
Ph 30-90% post 12 months	1	NA	1-minor (Ph 90% 75%

Overall 4/6 pts with hematologic resistance regained CHR, and 6/20 (30%) improved their CG response (2 CR, 1 PR, 3 minor). We conclude that some patients who do not have a good response to imatinib mesylate 400 mg daily may respond to increasing the imatinib mesylate dose. High-dose imatinib mesylate in newly diagnosed Ph + CML CP may increase the rates of CG and molecular CR, and is worthwhile investigating.

Pts with CML in relapse post allo SCT are generally treated with donor lymphocyte infusions (DLI), interferon alpha (IFN), or second transplants. Imatinib mesylate has significant anti-CML activity, and may be effective in inducing responses post allo SCT relapse without the serious complications associated with DLI. 28 pts with CML in relapse post allo SCT in chronic (5 pts) accelerated (15 pts), or blastic phase (8 pts) received imatinib mesylate 400 to 1000 mg orally daily. Their median age was 43 years. 13 pts had received DLI; the median time from allo SCT to relapse was 9 months (1 to 137 months) and from DLI to imatinib mesylate therapy 4 months (2 to 39 months). 22 pts (79%) had objective responses. A complete hematologic response (CHR) was observed in 18 of 24 evaluable pts (75%) and a cytogenetic response in 14 of 27 evaluable pts (52%), which was complete (Ph-positive cells 0%) in 9 (33%). The CHR rate was 100% in chronic phase, 83% in accelerated phase, and 50% in blastic phase. Cytogenetic responses were noted in 10 of 19 pts in chronic-accelerated phases (53%), and were complete in 7 (37%). Two of four responders in blastic phase had cytogenetic responses. 11 of 13 pts with prior DLI responded. With a median follow-up time of 11 months, 20 patients were still alive, 8 without evidence of disease for 1+ to 17+ months. The one-year estimated survival rate was 68%: 90% in chronic-accelerated versus 25% in blastic phase (p=0.001). 5 pts had recurrence of grade 3 (3 pts) or grade 1-2 (2 pts) GVHD, including skin rashes in 5 and diarrhea in 1. We conclude that imatinib mesylate offers effective disease control in patients with CML relapse post allo SCT. Comparison of its efficacy versus DLI is indicated. In blastic phase, studies of combinations of imatinib mesylate with DLI, IFN, cytarabine, or others are indicated.

We studied 28 pts with accelerated phase (AP) CML enrolled on Novartis expanded access study 114. Diagnosis of AP CML was based on progressive cytogenetic abnormalities (n=9), blasts >10% but <30% (n=9), refractory cythemias (n=5), splenomegaly (n=3), or thrombocytopenia (TP) (n=2). Median age was 59 years (34-84) and median duration of disease was 51 months. All patients received 600 mg/day of imatinib. Dose reductions to 400 mg/day and then 300 mg/day were prescribed for an ANC of < 500/ul or a platelet (plt) count of < 10,000/ul once progressive disease was ruled out. Eleven of 28 patients developed grade IV TP (<20,000/ul) following a median of 7 weeks of therapy. None (0/11) of these patients had progressive disease with the onset of TP. With 20 weeks median follow-up, seven patients recovered to >20,000 plts after an average of 11.3 weeks from diagnosis of TP and 5.3 weeks from discontinuation of imatinib. No TP patients had a complete hematologic recovery and only one (1/11) had no evidence of leukemia (NEL) while continuing imatinib at a lower dose. One died of complications from TP. Patients were transfusion dependent for an average of 9.1 weeks. A comparison of patients with and without TP showed no differences in age, pre-study blood counts, splenomegaly, bone marrow histology, the number of cytogenetic abnormalities, or previous treatment regimens. Patients with TP had a longer history of CML than those without (6.28 vs 4.46 years p = 0.009) and were more likely to be diagnosed with AP-CML by hematologic criteria than by cytogenetic progression (10/20 vs 1/8; p = 0.035). Of the 25 pts with evaluable cytogenetics, 5 complete responses and one major response occurred in the non TP group (6/14) while no (0/11) major responses occurred in the TP patients. Five of the 6 major responses occurred in the 8 pts enrolled for cytogenetic progression of disease as compared to 1 in the 20 patients enrolled for other reasons. No difference was observed in the average duration of imatinib treatment prior to developing a major cytogenetic response (15.9 weeks) or to developing TP (14.7 weeks, p = .77). These results suggest that severe TP can develop soon after the initiation of imatinib for the treatment of AP CML. TP is more common in those diagnosed by hematologic criteria and those with disease duration of greater than 5 years. TP is usually not due to leukemic progression, rarely improves with continued therapy, and is associated with a low likelihood of cytogenetic response.

Myelosuppression is one of the most common side effects of imatinib mesylate (Gleevec, formerly STI571) therapy. In Phase II clinical trials of this agent in chronic phase CML, rates of NCI Grade 3 neutropenia (500-<1000/mm³) and thrombocytopenia (10,000-<50,000/mm³) were 25% and 16%, respectively. Higher rates of cytopenias were observed in trials for patients with more advanced disease. During phase I and II multi-institutional clinical trials of imatinib in chronic phase CML patients who failed therapy with interferon, 66 patients were treated at Oregon Health & Science University with imatinib doses of 300 mg or higher. Data from these patients was analyzed for pretreatment characteristics predictive for Grade 3 or higher myelosuppression. Characteristics analyzed included age, sex, BSA, LDH, Hgb, WBC, Plts, % blasts in peripheral blood or marrow, % basophils in peripheral blood or marrow, % Philadelphia chromosome positivity by metaphase or interphase FISH analysis, marrow cellularity, time from diagnosis of CML, length of interferon therapy, reason for interferon failure, myelosuppression from prior interferon therapy, and presence of additional cytogenetic abnormalities. 35% (23/66) of patients developed Grade 3/4 neutropenia, and 24% (16/66) of patients developed Grade 3/4 thrombocytopenia. Data was analyzed by univariate logistic regression with all characteristics treated as independent variables. From this analysis we found that the only significant variable for the development of Grade 3/4 neutropenia was % of peripheral blood or marrow blasts (p = 0.0115). With each increase in blasts by 1%, the odds of developing Grade 3/4 neutropenia increased by 2.2. In contrast, a variety of pretreatment factors predicted for the development of thrombocytopenia. These included higher LDH (p=0.0481), lower Hgb (p=0.0068) and prior myelosuppression from interferon (p=0.0089). In patients with either Grade 3/4 neutropenia or thrombocytopenia, a statistically significant lower probability of cytogenetic response was observed (p= 0.0229 for Grade 3 neutropenia and p= 0.0274 for Grade 3 thrombocytopenia). However, 33% (10/33) of patients with Grade 3/4 myelosuppression had major cytogenetic responses indicating that myelosupression does not preclude the possibility of a major cytogenetic response. Our data suggests that in chronic phase CML patients with prior myelosuppression from interferon, or advanced features such as higher blast %, higher LDH, or lower Hgb, closer monitoring for the development of myelosuppression during therapy with imatinib is warranted. Differential cytogenetic response rates seen in patients with myelosuppression could either be due to more advanced stage of disease, lower dose intensity of therapy, or both. The issue of dose intensity is currently being analyzed as this may affect the recommendations for management of this common side effect of imatinib therapy.

At 3 months of therapy, imatinib mesylate produced significantly higher complete (36% vs 0 to 4 %), and major [Ph <35%] cytogenetic responses (77% vs 2 to 24%). These have been historically good surrogate endpoints for long-term survival. The complete and major cytogenetic response rates of 36% and 77% are unprecedented with any previous non-transplant therapy in CML. The 6-9 month results will be presented. A multi-national phase III study evaluating upfront therapy with imatinib mesylate versus interferon + ara-C is ongoing.

Pulmonary leukostasis is a rare but serious and often fatal complication of CML in blast crisis and acute myeloid leukemia. Imatinib (STI571, Gleevec, Novartis) is a potent and selective inhibitor of the BCR-ABL tyrosine kinase, the molecular abnormality which causes chronic myeloid leukemia (CML). Recent studies of imatinib in patients with CML in blast crisis have yielded encouraging results. We report the case of a 74-year-old man with a nine year history of CML previously treated with hydroxyurea and interferon-alpha, who presented to our institution with CML in myeloid blast crisis with a rapidly rising white blood cell count (WBC). Over the next 24 hours developed marked increasing dyspnea, hypoxemia (oxygen saturation of 88%), and low-grade fever. On physical examination the patient was noted to be in respiratory distress with labored breathing, to have clear lung fields to auscultation and percussion, and to have hepatomegaly and marked splenomegaly. Laboratory studies showed a WBC of 125, 000/ml with 28% myeloblasts and 18% basophils, and elevated uric acid, phosphate and LDH levels. Chest X-ray showed increased interstitial markings but no focal consolidation. The patient was diagnosed with impending respiratory failure and admitted to our leukemia service. A high-resolution CT scan of the chest showed scattered bilateral thickening of the intralobular septa with patchy areas of ground-glass opacity but no focal lung consolidation. These findings were considered to be compatible with pulmonary leukostasis. A bone marrow examination showed a markedly hypercellular marrow with 43% myeloblasts and 18% basophilia. Due to his low-grade fever the patient had multiple cultures and a bronchoalveolar lavage which were all negative. The patient was started on imatinib at a dose of 500mg bid. Within 48 hours there was a significant decrease in his WBC with marked improvement in his respiratory status. No evidence of tumor lysis syndrome occurred. Followup CT scan of the chest showed interval resolution of the interlobular septal thickening and vascular congestion. imatinib may be an effective single agent treatment of pulmonary leukostasis in patients with CML blast crisis without risk of tumor lysis syndrome.

In chronic myelogenous leukemia (CML), the presence of cytogenetic abnormalities in addition to the Philadelphia (Ph) chromosome in cells already harboring the Ph chromosome, is considered clonal evolution and thus a feature of accelerated phase (AP) disease. While its presence along with other AP features likely confers a poor prognosis, clonal evolution in the absence of other AP features has not adversely impacted outcome following allogeneic stem cell transplantation. Similarly, a previous study utilizing interferon and low dose ara-C to treat AP patients (Kantarjian et al) demonstrated superior survival for patients in whom clonal evolution was the only criterion for disease acceleration (3-year survival rate 67% v 22%; P less than .01). Given this background, we analyzed the impact of clonal evolution on the therapeutic response to the BCR-ABL tyrosine kinase inhibitor, imatinib mesylate (Gleevec, formerly STI571), in the treatment of AP CML. Between November 1st 1999 and May 31st 2001, 70 AP CML patients were treated at OHSU with imatinib (600mg) on Novartis studies 109 and 114. Patients were divided into 3 groups. One group had clonal evolution only (CE, n=17), a second group had AP but no evidence of clonal evolution (AP, n=33) and a third group had AP plus clonal evolution (AP + CE, n=20). We evaluated the rate of complete and major cytogenetic responses using metaphase analysis, the decline in % of Bcr-Abl positive bone marrow cells, using fluorescent in-situ hybridization (FISH) and progression free survival for the 3 different groups. As determined by FISH, Bcr-Abl positive cells of < 5% were seen in 12/17 (71%) CE patients compared to 8/33 (24%) AP patients (p = 0.0015) and 1/20 (5%) AP + CE patients (p < 0.001). At this time, 13 CE, 22 AP and 11 AP + CE patients are evaluable for cytogenetic response. 11/13 (85%) CE patients had a major cytogenetic response compared to 9/22 (41%) AP patients (p < 0.025) and 2/11 (18%) AP + CE patients (p < 0.01). Complete cytogenetic responses were seen in 8/13 (62%) of CE patients compared to 9/22 (41%) of AP patients (p > 0.05) and 0/11 AP + CE patients (p < 0.01). With a median follow up of 1 year (range 3-22 months), 69% of all patients remain progression-free. The lowest rate of progression free survival was seen in AP + CE patients at 35% (7/20). In comparison, 94% (16/17) of CE and 76% (25/33) of CE + AP, remain progression free (p = 0.002, 0.003). There was no statistical difference in progression free survival between CE and AP (p=0.14). Conclusions: In accelerated phase CML, superior outcomes are seen for patients in whom the only criterion for disease acceleration is clonal evolution. Inferior outcomes are seen when clonal evolution is present in addition to typical accelerated phase features. Although this is a relatively small group of patients, the high rate of major and complete cytogenetic responses in CE patients suggests that studies comparing 600mg of imatinib as compared to the currently recommended dose of 400mg for chronic phase patients should be considered.

Multivariate analysis selected % Ph+ 90% (p<.01), hematologic resistance to IFN-a (p=.047) and Dx to Rx > 3 years (p=.03) as factors associated with lower major CG response rates. 16 pts are now off therapy: blastic phase 5, resistant 7, sudden unrelated death (cardiac) 1, pregnancy 1, toxicity 2. 4 pts have died. Serious grade nonhematologic 3-4 toxicities were seen in 12(5%) requiring stopping STI in 2 pts. In summary, imatinib mesylate is the most effective single agent therapy in CML.

No	Additional abnormalities	Months to CE	% Ph	% metaphases with additional abnormalities	Outcome (time since starting STI571)
1	del(5)(q15q33)	6	7	3	Stable at 9 mo
2	+8	9	20	70	Stable at 18 mo
3	+Ph	12	23	4	Stable at 16 mo
4	+Ph	3	100	80	Stable at 9 mo
5	-14, -17, +der(14;17)(q10;q10), +Ph	3	100	28	Stable. at 15 mo
6	t(3;16)(q11;p11), +Ph	12	100	25	Stable at 16 mo
7	+8, +13, i(17)(p11.2), +Ph	6	100	100	Progress. at 7 mo
8	+8	6	100	5	Stable at 12 mo
9	del20(q11)	3	100	5	Stable at 12 mo
10	add(17)(p13)	6	100	79	Progression at 12 mo
11	inv(3)(q21q26), del(5)(q1?5), -21, + mar, +Ph	7	100	76	Progression at 13 mo