Phase 1 study of tipifarnib in combination with imatinib for patients with chronic myelogenous leukemia in chronic phase after imatinib failure




The tolerability and efficacy of the combination of tipifarnib, an orally bioavailable nonpeptidomimetic farnesyl transferase inhibitor, and imatinib was investigated in patients with chronic myelogenous leukemia in chronic phase who had failed imatinib.


Twenty-six patients (13 [50%] with Abl kinase domain mutations) were treated. The initial dose level was tipifarnib at a dose of 300 mg twice daily and imatinib at a dose of 300 mg daily. Therapy was escalated following a ‘3 + 3’ phase 1 design and the maximum tolerated dose was defined as tipifarnib at a dose of 400 mg twice daily and imatinib at a dose of 400 mg daily. Therapy was administered for a median of 26 weeks (range, 3–150 weeks).


Adverse events included diarrhea in 21 patients (81%) and nausea in 18 patients (69%), but were generally grade 2 or less (using the revised National Cancer Institute Common Toxicity Criteria). Grade 3–4 neutropenia and thrombocytopenia occurred in 11 patients (42%) and 8 patients (31%), respectively. Sixteen patients discontinued therapy (5 due to toxicity and 11 due to lack of response or disease progression). Hematologic responses were attained by 17 (68%) of 25 assessable patients. Nine patients (36%) also achieved a cytogenetic response (3 complete responses, 4 partial responses, and 2 minimal responses), including 4 patients harboring mutant Bcr-Abl tyrosine kinases. One patient bearing the highly imatinib-resistant T315I mutant achieved a partial cytogenetic response. The median response duration was 3 months (range, 2–30+ months).


The combination of tipifarnib and imatinib is well tolerated and has activity against several Abl kinase domain mutants. Combinations of tipifarnib with more potent tyrosine kinase inhibitors warrant further investigation. © 2007 American Cancer Society.

Imatinib mesylate induces complete hematologic (CHR) and major cytogenetic response (MCyR) in greater than 95% and 85%, respectively, of patients with chronic myelogenous leukemia (CML) in chronic phase (CP).1 However, imatinib fails to eradicate quiescent BCR-ABL-positive stem cells2–4 and a subset of patients develop imatinib resistance, which has been associated in 30% to 50% of cases with point mutations in the kinase domain of BCR-ABL.2, 5–8 One strategy to overcome imatinib resistance is to interfere with Bcr-Abl downstream pathways such as the Ras pathways. Localization to the inner plasma membrane is critical in Ras activation9 and is accomplished through a posttranslational prenylation process by which a 15-carbon isoprenyl (farnesyl) moiety is attached to the C-terminal cysteine residue by the enzyme farnesyltransferase (Ftase) and to a lesser extent by geranylgeranyl-protein transferases (GGPTases).10 Ftase inhibitors (FTIs) were developed as a means of inhibiting Ras activation, although their antineoplastic effects have been proved to be mediated through inhibition of farnesylation of other proteins such as RhoB,10, 11 Rab,12 and CENP-E and -F.13

FTIs have antiproliferative activity against BCR-ABL-positive cell lines and cells from patients, including some resistant to imatinib.14 Tipifarnib is a nonpeptidomimetic FTI that has modest single-agent efficacy in the treatment of patients with CML in all phases after imatinib failure.15, 16 In vitro, FTIs have proven synergistic with imatinib both in imatinib-sensitive and imatinib-insensitive cell lines.17–19 Hence, it is conceivable that a combination of these agents acting through a different mechanism of action could overcome and potentially prevent the development of imatinib resistance and improve clinical outcomes.

In this phase 1 study, we investigated the maximum tolerated dose (MTD) and the toxicity profile of the combination of tipifarnib and imatinib in patients with Philadelphia chromosome (Ph)-positive CML in CP who had previously failed therapy with imatinib. The results suggest that this combination is safe and active in this patient population.



Patients with BCR-ABL-positive CML in CP who had failed therapy with at least the standard dose imatinib (ie, ≥400 mg daily) were eligible. Imatinib failure was defined as: 1) inability to achieve or loss of CHR after 3 months of imatinib, 2) failure to achieve or loss of at least a minimal cytogenetic response after 6 months of imatinib, or 3) failure to achieve or loss of a MCyR after 12 months of imatinib. Other eligibility criteria were: 1) performance status ≤2 by the Zubrod scale; 2) serum creatinine and total bilirubin level <2 mg/dL; 3) a white blood cell count (WBC) <30 × 109/L (hydroxyurea had to be discontinued 24 hours prior study entry); and 4) the practice of effective methods of contraception by women of childbearing potential. All patients provided informed consent approved by the Institutional Review Board.

Treatment Schedule

Patients received tipifarnib initially at a dose of 300 mg orally twice daily for 14 days every 21 days (ie, 14 days on/7 days off) and imatinib at a dose of 300 mg orally daily. Dose escalation to the dose levels shown in Table 1 were performed according to a ‘3 + 3’ phase 1 design. Patients were observed for at least 3 weeks before the next cohort was treated at the next dose level. Tipifarnib and imatinib doses could be decreased to dose level –1 and –2 whenever unacceptable toxicity occurred. Development of an absolute neutrophil count (ANC) <1.0 × 109/L or platelet count <50 × 109/L led to therapy interruption until counts recovered above these levels. Therapy was then resumed at the same dose level if recovery occurred within 2 weeks or at a lower dose level if it occurred after 2 weeks. Patients who experienced grade ≥2 neurotoxicity for more than 48 hours had therapy withheld until resolution to grade ≤1, and then resumed at the same dose level. If neurotoxicity recurred, therapy was stopped until recovery to grade ≤1 and then resumed at the next lower dose level. Patients who experienced other grade ≥3 nonhematologic toxicity had therapy stopped until resolution to grade ≤1 and then resumed at the next lower dose level. Individual tipifarnib or imatinib dose adjustments were allowed if the toxicity wasclearly related to 1 of the drugs. Intrapatient dose escalation was allowed for patients achieving an inadequate response.

Table 1. Dose Escalation Levels
Dose levelImatinib, mg(given orally daily)Tipifarnib, mg (given orally twice daily)

Toxicities were evaluated using the revised National Cancer Institute Common Toxicity Criteria (version 2.0). The MTD and dose-limiting toxicity (DLT) were defined within the first 2 cycles of therapy. DLT was defined as grade ≥3 nonhematologic toxicity. Nausea and vomiting were considered a DLT only if not responsive to antiemetic therapy. Hematologic DLT was defined as ANC <1 × 109/L or platelet count <50 × 109/L lasting for ≥6 weeks after interruption of therapy with a hypocellular bone marrow in the absence of bone marrow blasts. Anemia was not considered a hematologic DLT.

Patient Evaluation

Baseline studies included a complete physical examination, complete blood count (CBC), comprehensive biochemistry panel (including liver function tests), bone marrow aspirate with cytogenetics, and an electrocardiogram. A CBC with differential was obtained weekly for the first 3 weeks and then every 3 weeks. Blood chemistries were repeated weekly during the first 3 weeks and every 6 to 8 weeks thereafter. Bone marrow aspirate and cytogenetics (fluorescence in situ hybridization [FISH], when cytogenetics were unevaluable) were performed at the end of Cycle 3, every 3 months to 4 months for 1 year, and then every 12 months while the patient received therapy.

Response criteria were as previously described.20 CHR was defined as a WBC count <10 × 109/L, a platelet count <450 × 109/L, no immature cells (blasts, promyelocytes, or myelocytes) in the peripheral blood, and the disappearance of all signs and symptoms related to leukemia for at least 4 weeks. A CHR was further categorized by the best cytogenetic response as complete (0% Ph-positive), partial (1–35% Ph-positive), minor (36–65% Ph-positive), and minimal (66–95% Ph-positive). A major cytogenetic response (MCyR) included complete (CCyR) plus partial (PCyR) cytogenetic responses. The criteria for partial hematologic remission (PHR) were similar to those of CHR except for the persistence of a few immature cells in the peripheral blood and/or persistent splenomegaly or thrombocytosis (>450 × 109/L) but at least 50% less than pretreatment.

Study Design

A minimum of 3 patients were entered at each dose level. If no DLT was observed in 3 patients treated at a given dose level after the last patient treated had been observed for a minimum of 3 weeks, dose escalation was indicated in the next cohort of patients. If DLT occurred in 1 of 3 patients, 3 more patients were entered at the same dose level. If grade 3–4 toxicity developed in 2 of 3 or 6 patients, this dose was considered to exceed the MTD. The MTD was thus defined as the dose resulting in DLT in 1 or fewer of 6 patients. Once the MTD was defined, additional patients were treated at this dose to define the toxicity profile more precisely.

Criteria for withdrawing a patient from the study were: 1) continuous increase in WBC counts (WBC >40 × 109/L or ≥50% increase in WBC count for patients beginning with WBC >40 × 109/L) despite repeated cycles, 2) the development of features of accelerated ([AP]; except clonal evolution) or blastic phase (BP) while on study, 3) unacceptable grade 3 to 4 toxicity in the absence of significant antileukemic effect, and 4) at the patient's request.


Patient Characteristics

Between October 2002 and May 2005, 26 patients (12 males) with Ph-positive CML in CP who had failed single-agent imatinib therapy were treated (Table 2). Other prior therapies included interferon (IFN)-α in 23 patients (88%) (14 of whom received it in combination with cytarabine [10 patients], homoharringtonine [2 patients], imatinib [1 patient], or granulocyte‒ macrophage–colony‒stimulating factor [1 patient]), decitabine (3 patients), homoharringtonine (3 patients), and allogeneic stem cell transplantation (3 patients), 1 of whom received transplantation in 3 different occasions and 2 of whom received donor lymphocyte infusions. The best responses obtained with single-agent imatinib were a CHR in 10 patients, CCyR in 1 patient, PCyR in 11 patients, and minor cytogenetic response in 1 patient. Three patients were found to be primary refractory to imatinib. At the initiation of tipifarnib and imatinib therapy, only 11 patients (42%) were in CHR. Of the 15 patients (58%) who were not in CHR, 4 were receiving hydroxyurea and 6 were receiving anagrelide to control their peripheral blood counts. One patient had a variant Ph chromosome and 1 other had clonal evolution (double Ph chromosome). A baseline BCR-ABL/ABL ratio was obtained in 25 patients (96%; 1 suboptimal specimen) and ABL gene sequencing was performed in 25 patients (96%), revealing the presence of mutations in 13 (52%) (M244V and E355G in 3 patients each; and F359C, L364I, G250E, F317L, D276G, E255K, and T315I in 1 patient each).

Table 2. Baseline Clinical Characteristics of Patients Receiving Tipifarnib and Imatinib
  • Ph indicates Philadelphia chromosome; +, positive; NA, not applicable.

  • *

    Assessed in peripheral blood by quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR).

Age, y6229–82
No. of prior therapies31–7
Time from diagnosis, y5910–160
Time receiving imatinib, mo3111–57
Percentage of Ph+ metaphases10035–100
BCR-ABL/ABL ratio*23.842.42–>100
Leukocyte count (×109 μ/L)7.22.2–25.6
Hemoglobin level (g/dL)10.99–15.6
Platelet count (×109/L)327107–1424
No. patients with Abl mutations13NA
No. cycles receiving tipifarnib plus imatinib81–49
No. weeks receiving tipifarnib plus imatinib263–150
No. days off tipifarnib plus imatinib330–99


Analysis of safety was based on data from all 26 enrolled patients (Table 3). The most frequent nonhematologic toxicities were gastrointestinal, fatigue, and skin rash, but these were manageable with symptomatic treatment, and therefore were not considered to constitute a DLT. Nine patients had central nervous system (CNS) toxicity, including 1 who had a grade 3 traumatic subdural hematoma and 1 with grade 3 cerebellar toxicity secondary to benign paroxysmal positional vertigo at dose level 2, but both toxicities were unrelated to the study treatment. Anemia occurred in 17 patients (65%) (grade 3–4 in 5 patients). Twenty episodes of grade 3 to 4 neutropenia were reported during the administration of a total of 299 cycles, for an overall incidence of 7%. Four patients were hospitalized due to grade 3 neutropenic fever. Grade 3 to 4 thrombocytopenia was reported in 8 patients (31%), although bleeding was only reported in 1 patient (grade 2 lower gastrointestinal bleeding). Adverse events led to dose level reduction in 10 patients (38%), 3 of whom required more than 2 dose level reductions. Patients were required to interrupt therapy for a median of 33 days (range, 0 to 99 days), mostly due to hematologic toxicity. Adverse events led to termination of the study treatment in 5 patients (19%). One patient with risk factors for coronary artery disease died of myocardial infarction after 14 days on the study at dose level 2.

Table 3. Adverse Events Related to Therapy With Tipifarnib and Imatinib
ToxicityNo. (%) of Patients by dose level
Level 0Level 1Level 2Level 3Overall
Any gradeGrade 3–4Any gradeGrade 3–4Any gradeGrade 3–4Any gradeGrade 3–4Any gradeGrade 3–4
Neutropenia2 (67)2 (67)2 (67)2 (67)7 (41)6 (35)1 (33)1 (33)12 (46)11 (42)
Thrombocytopenia2 (67)2 (67)007 (41)5 (29)1 (33)1 (33)10 (38)8 (31)
Anemia002 (67)012 (71)4 (24)3 (100)1 (33)17 (65)5 (19)
Nausea/vomiting2 (67)02 (67)015 (88)02 (67)021 (81)0
Diarrhea3 (100)03 (100)012 (71)3 (18)2 (67)020 (77)3 (12)
Neutropenic fever/sepsis1 (33)1 (33)004 (24)2 (12)1 (33)1 (33)6 (23)4 (15)
Fatigue1 (33)02 (67)010 (59)22 (67)015 (58)2 (7.5)
Pain1 (33)02 (67)08 (47)02 (67)013 (50)0
Infection (non-neutropenic)1 (33)02 (67)07 (41)1 (6)2 (67)012 (46)1 (4)
Rash/pruritus001 (33)06 (35)01 (33)08 (31)0

No DLT was identified with the first 2 dose levels. One patient developed grade 3 diarrhea at dose level 2. Three more patients were entered at this dose level without evidence of DLT. Two of 3 patients treated at dose level 3 experienced DLTs (1 with grade 3 fatigue and 1 with grade 3 esophagitis). Accordingly, tipifarnib at a dose of 400 mg twice daily and imatinib at a dose of 400 mg daily (dose level 2) was considered the MTD. Eleven additional patients were treated at this dose level. Of these, 2 patients had grade 3 neutropenia and 1 had grade 3 dehydration/renal insufficiency and fatigue. A median of 8 cycles (range, 1–49 cycles) were administered, with a median of 12 cycles (range, 7–13 cycles) at dose level 0, 12 cycles (range, 4–16 cycles) at dose level 1, 5 cycles (range, 1–49 cycles) at dose level 2, and 8 cycles (range, 3–15 cycles) at dose level 3.


The median duration of therapy was 26 weeks (range, 3–150 weeks). All but 1 patient, who died before completion of the first cycle of therapy, were assessable for efficacy analysis (Table 4). CHR was maintained or achieved by 17 of 25 (68%) patients while receiving the study treatment. Hematologic responses lasting at least 4 weeks occurred in 10 (62%) of 16 patients not in CHR at study entry, including 8 patients (50%) who achieved a CHR (4 ongoing) that was sustained for a median of 11 weeks (range, 7–38+ weeks) and 2 patients (12%) who attained a PHR. Nine patients (36%) achieved a cytogenetic response. No cytogenetic responses were observed among the first 6 patients (3 with ABL mutations) treated at dose levels 0 and 1, whereas 9 (45%) of 20 patients treated at doses equal or higher to the MTD achieved cytogenetic responses: CCyR (n = 3, 12%), PCyR (n = 4, 16%), and minimal cytogenetic response (n = 2, 8%). Hematologic responses were observed in 8 of 13 patients (62%) with ABL kinase domain mutations, including 4 who had a cytogenetic response (1 with T315I, 2 with M244V, and 1 with E255K). Patients with PCyR had the percentage of Ph-positive metaphases reduced from 100% to 30%, 65% to 30%, 55% to 10%, and 95% to 15%, respectively. It is interesting to note that the latter PCyR (15% Ph-positive) occurred in a patient in whom the imatinib-insensitive T315I mutation was predominant. This response lasted for 2 months before the patient was taken off-protocol to undergo matched unrelated stem cell transplantation. The median duration of cytogenetic response was 3 months (range, 2–39+ months). Responses were sustained for more than 6 months in 3 patients, including 2 ongoing CCyR (12+ months and 39+ months, respectively). The median time to MCyR was 3 months, corresponding to the first assessment of response in most patients. However, 1 patient who had achieved a PCyR 16 months after the initiation of therapy obtained a CCyR after 33 months on the study therapy.

Table 4. Hematologic and Cytogenetic Responses to Tipifarnib and Imatinib
  No. of patients (%) with response
All evaluable patients(n = 25)Patients not in CHR at start of study (n = 16)
  1. CHR indicates complete hematologic response; PHR, partial hematologic response; NR, no response; CCGR, complete cytogenetic response; PCGR, partial cytogenetic response; mCGR, minimal cytogenetic response.

Hematologic responseOverall19 (76)10 (62)
 CHR17 (68)8 (50)
 PHR2 (8)2 (12)
 NR6 (24)6 (38)
Cytogenetic responseOverall9 (36)4 (25)
CCGR3 (12)0 (0)
PCGR4 (16)2 (13)
mCGR2 (8)2 (13)
No CGR16 (64)12 (75)

Molecular responses were assessable in 16 patients (62%) (6 [38%] had a 1-log and 1 [6%] patient had a 2-log reduction of their BCR-ABL transcript levels). After a median follow-up of 6.5 months (range, 1–38 months), 2 patients were still in CCyR receiving the study treatment. The remaining 24 had withdrawn from tipifarnib and imatinib therapy due to progression to AP (6 patients), side effects (5 patients), lack of response (5 patients), loss/lack of cytogenetic response (4 patients), lost to follow-up (2 patients), patient's request (1 patient), and death (1 patient).


Single-agent oral tipifarnib has modest activity in CML. In 1 study, tipifarnib was administered to 22 patients with CML in all phases after IFN-α failure (and 70% after imatinib failure).15 Six of 10 patients in CP and 1 of 6 in AP had a hematologic response (CHR in 5).15 In a phase 1 study of tipifarnib, 2 patients in BP achieved a partial hematologic response.16 Importantly, tipifarnib is synergistic with imatinib, both in imatinib-sensitive and imatinib-resistant cell lines,17–19 providing the rationale for the clinical development of this combination.

In this study, the combination of tipifarnib and imatinib was well tolerated at doses of 400 mg twice daily and 400 mg daily, respectively, which defined the MTD. The most frequent adverse events were gastrointestinal but these were generally mild and manageable. Neutropenia and thrombocytopenia were the most frequent grade 3 to 4 toxicities and the main reason for treatment discontinuation. However, treatment-related myelosuppression did not result in fatal outcomes and led to hospitalization in only 15% of patients. There was no routine use of growth factors in this study, an approach that could potentially be used to minimize myelosuppression.

This combination demonstrated significant clinical activity, with hematologic and cytogenetic response rates of 62% and 36%, respectively. Cytogenetic responses were observed in 45% of patients treated at doses at least equal to the MTD. This response rate is slightly lower than the cytogenetic responses observed in phase 1/2 studies of nilotinib (50%)21 and dasatinib (44%)22 after imatinib failure. However, a potential advantage of FTIs over tyrosine kinase inhibitors (TKIs) is their activity against quiescent leukemic stem cells. BCR-ABL-positive CD34+/Lin− leukemic stem cells of patients with CP CML remained viable even in the presence of growth factors and imatinib,3 potentially facilitating recurrence and development of resistance to imatinib.23, 24 Although dasatinib is significantly more potent than imatinib, quiescent stem cells appear innately insensitive to both agents.25 Moreover, several imatinib-resistant ABL kinase domain mutations have been detected in CD34+/BCR-ABL+ progenitors.26 BMS-214662, a nonpeptidomimetic FTI, was 68-fold more potent in killing quiescent (IC50 0.7 μM) than proliferating (IC50 47.5 μM) K562 cells and exhibited supra-additive cytotoxicity with imatinib in murine K562 xenografts.27 Another FTI, lonafarnib, had similar effects against quiescent CD34+ leukemic cells from patients with CML in CP, and inhibited their ability to efflux imatinib.28 Conceivably, a ‘cocktail’ of agents with distinct mechanisms of action might prevent the emergence of resistance to TKIs. Particularly appealing would be to combine FTIs with dasatinib, which target 3 crucial elements of Bcr-Abl signaling: Abl kinase, Src family kinases, and Ras activation. Such an approach could be more efficacious if introduced early (ie, suboptimal responses rather than failures), before the development of TKI resistance.17–19

An interesting observation in this study was the response in 1 patient with T315I. This mutation is completely insensitive to imatinib, nilotinib, and dasatinib. The contribution of imatinib to the response in this patient is uncertain but the remarkable reduction of Ph-positive cells supports the ability of the combination therapy to eliminate mutated clones and leukemic stem cells. It would be important to confirm this observation in additional patients given the lack of effective therapeutic strategies in this setting.

The results of the current study demonstrated that tipifarnib combined with imatinib is safe for patients with CML in CP after imatinib failure and induces cytogenetic responses. Studies combining FTI with new TKIs and exploring earlier intervention with this combination are warranted, with a focus on exploring the potential elimination of the leukemic stem cell.