Stefan Faderl receives research support from Genesys.
Clofarabine combinations as acute myeloid leukemia salvage therapy†
Article first published online: 28 AUG 2008
Copyright © 2008 American Cancer Society
Volume 113, Issue 8, pages 2090–2096, 15 October 2008
How to Cite
Faderl, S., Ferrajoli, A., Wierda, W., Huang, X., Verstovsek, S., Ravandi, F., Estrov, Z., Borthakur, G., Kwari, M. and Kantarjian, H. M. (2008), Clofarabine combinations as acute myeloid leukemia salvage therapy. Cancer, 113: 2090–2096. doi: 10.1002/cncr.23816
See editorial on pages 1995–8, this issue.
- Issue published online: 3 OCT 2008
- Article first published online: 28 AUG 2008
- Manuscript Accepted: 27 MAY 2008
- Manuscript Revised: 6 DEC 2007
- Manuscript Received: 30 OCT 2007
- nucleoside analogs;
- acute myeloid leukemia;
Outcome of patients with relapsed acute myeloid leukemia (AML) remains unsatisfactory. Clofarabine is a nucleoside analog with activity in adult AML. Combinations with cytarabine in AML are feasible and effective. Idarubicin is another active AML drug, which has not yet been tested with clofarabine.
The authors therefore designed a phase I study of clofarabine ± cytarabine, plus idarubicin. Patients with primary refractory or first-relapse AML were assigned to either clofarabine plus idarubicin (CI) if previously exposed to cytarabine with a response lasting <12 months, or clofarabine and idarubicin plus cytarabine (CIA) for responses ≥12 months, or if never exposed to cytarabine. A standard “3 + 3” phase 1 design was followed to define maximum tolerated dose (MTD). Forty-four patients were treated (23 CI; 21 CIA).
Dose-limiting toxicities were hyperbilirubinemia and hepatic transaminase elevations for CI-treated patients in addition to mucositis and diarrhea for CIA-treated patients. MTD for CI was clofarabine 22.5 mg/m2 intravenously daily × 5 and idarubicin 10 mg/m2 intravenously daily × 3. MTD for CIA was clofarabine 22.5 mg/m2 intravenously × 5, idarubicin 6 mg/m2 intravenously × 3, and cytarabine 0.75 g/m2 intravenously × 5 days.
A phase 2 randomized trial is in process to compare activity between treatment arms. Cancer 2008. © 2008 American Cancer Society.
Most patients with acute myeloid leukemia (AML) relapse and will require some form of salvage therapy. Whereas response to salvage therapy depends on duration of a previous remission and number of preceding salvage attempts, cure rates remain dismally low, and only few patients enjoy long-term disease-free survival. There is therefore a continued need to identify and test new agents in clinical trials so that they can be incorporated into effective salvage regimens for patients.1–3
Clofarabine (2-chloro-2′-fluoro-deoxy-9-β-D-arabinofuranosyladenosine) is a second-generation nucleoside analog, which was developed as a hybrid molecule to combine the most favorable pharmacokinetic properties of both fludarabine and cladribine.4 Clofarabine acts as an inhibitor of both DNA synthesis and the enzyme ribonucleotide reductase (RNR). Its structural characteristics render it resistant to deamination by adenosine deaminase and phosphorolytic cleavage by bacterial purine nucleoside phosphorylase.5–8 Clofarabine has been active against a broad range of human tumor cell lines and human tumor xenograft models.9 A phase 1 trial in adult patients with leukemias and solid tumors showed a 16% response rate in acute leukemia patients.10 Activity in children with acute lymphoblastic leukemia (ALL) led to approval by the Food and Drug Administration in December 2004 for the treatment of pediatric patients with relapsed or refractory ALL who have failed to respond to at least 2 prior regimens.11 The focus of clinical development of clofarabine in adults has shifted to AML. In a phase 2 trial of single-agent clofarabine in relapsed and refractory acute leukemias, response rates in 31 patients with AML reached 55% (42% complete response [CR]) and were reported to be as high as 87% for patients with first remission durations lasting longer than 12 months.12
Encouraging results from the single-agent experience and preclinical data lending support to biochemical modulation between clofarabine and intracellular cytarabine-5′-triphosphate (ara-CTP) levels led to trials of clofarabine combinations with cytarabine.13, 14 Anthracyclines are active in AML therapy and are the class of agents most commonly combined with cytarabine. To investigate clofarabine combinations in AML further, we designed 2 phase 1 trials of clofarabine with idarubicin and idarubicin plus cytarabine.
MATERIALS AND METHODS
Eligible were adults (age, ≥18 years) with 1) AML in first relapse or with primary refractory disease, who were receiving the study drugs as first salvage therapy; and 2) high-risk myelodysplastic syndrome (≥10% marrow blasts) and not more than 1 prior chemotherapy regimen. Patients could have received additional therapy with biological or targeted therapies (eg, hypomethylating agents, tyrosine kinase inhibitors, or monoclonal antibodies such as gemtuzumab ozogamicin). Additional eligibility criteria included adequate hepatic (serum total bilirubin, ≤2 mg/dL; serum glutamic pyruvic transaminase [SGPT], ≤4 × upper limit of normal), renal (serum creatinine, ≤2 mg/dL), and cardiac (ejection fraction >30% by echocardiography or gated cardiac scan) function, and a performance status of at least 2 (Eastern Cooperative Oncology Group). Patients were excluded if they had received previous treatment with clofarabine, had symptomatic central nervous system (CNS) involvement, or had an active uncontrolled systemic infection.
The study was approved by the institutional review board (IRB) of the University of Texas M. D. Anderson Cancer Center and was conducted in accordance with the basic principles of the Declaration of Helsinki.
The study consisted of 2 phase 1 trials: clofarabine plus idarubicin (CI), and clofarabine, idarubicin, plus cytarabine (Ara-C) (CIA). For criteria for allocation to treatment arm (CI vs CIA) see the section on “Statistical Methods.” The starting doses (cohort 1) of the CI combination consisted of clofarabine 22.5 mg/m2 administered as a 1-hour intravenous infusion daily for 5 consecutive days (Days 1 to 5) plus idarubicin 12 mg/m2 intravenously over 30 minutes daily for 3 consecutive days (Days 1 to 3). Starting doses for CIA were as follows: clofarabine 22.5 mg/m2 intravenously on Days 2 to 6, idarubicin 8 mg/m2 intravenously on Days 1 to 3, and cytarabine 1 g/m2 intravenously over 2 hours on Days 1 to 5. Patients could receive a maximum of 2 induction cycles or treatment until a clinical response was documented. Patients who did not achieve at least a partial response after a second induction cycle were removed from the study. All other patients were eligible to receive up to 4 consolidation cycles. Consolidation therapy consisted of the last induction regimen at a 25% dose reduction for all drugs alternating with standard-dose cytarabine (100 mg/m2 as a continuous intravenous infusion over 24 hours over 5 consecutive days). Cycles were repeated every 4 to 6 weeks as indicated by leukemia recurrence and hematopoietic recovery.
Supportive measures for optimal medical care were provided throughout the study as determined by the treating physicians and the patients' medical needs. Use of colony-stimulating factors was permitted, but not mandated. Prophylactic antibiotics, antifungals, and antiviral agents (eg, levofloxacin, itraconazole, valacyclovir) were recommended.
The pretreatment evaluation included history, physical examination, complete blood counts (CBC) with differential and platelet count, a complete chemistry survey, and marrow aspiration with cytogenetic analysis. No marrow was requested if the diagnosis of relapse could be made unequivocally from a peripheral blood specimen, including immunophenotyping by flow cytometry. An echocardiogram or gated cardiac nuclear scan (“MUGA”) to evaluate the left ventricular ejection fraction was performed before therapy. Follow-up studies included CBC, differential, and platelet count at least weekly. Marrow aspirations were repeated starting on Day 21 of therapy and then every 2 weeks until remission or nonresponse.
A CR was defined as disappearance of all clinical and/or radiologic evidence of disease with ≤5% marrow blasts, neutrophil count ≥1 × 109/L, and platelet count ≥100 × 109/L. A complete remission without platelet recovery (CRp) had identical marrow results and neutrophil recovery as for CR, but with platelets <100 × 109/L and ≥20 × 10/L. Partial remission consisted of a peripheral blood recovery as for CR, but with a decrease in marrow blasts of ≥50% compared with baseline before therapy and not more than 6% to 25% blasts in the marrow. All other responses were considered failures.
The study consisted of 2 open-label phase 1 trials with allocation to the respective treatment arm depending on assumed responsiveness to cytarabine. Patients who received a prior cytarabine-based induction regimen and either were primary refractory or experienced a CR of <12 months were assigned to CI, whereas all others received CIA. However, if only 1 treatment arm was open for accrual, all eligible patients could be accrued on that particular arm irrespective of pretreatment characteristics. The primary objective of the study was to determine the maximum tolerated dose (MTD) and dose-limiting toxicity (DLT) of the clofarabine salvage combinations. A standard 3 + 3 design was used to determine cohort expansions and dose escalations. All ≥grade 3 drug-related, nonhematologic toxicities that occurred within 14 days of the start of treatment were considered DLTs with the exception of nausea and vomiting (if manageable with supportive care measures), alopecia, drug-related fevers, asymptomatic abnormalities of lactate dehydrogenase, alkaline phosphatase, and disturbances of electrolytes, as these are common events in patients with relapsed leukemias.
A total of 44 patients have been treated, 23 with CI and 21 with CIA (Table 1). The median age of the CI group was 58 years (range, 24 years-71 years). One patient had refractory anemia with excess of blasts (RAEB-2). Nine (39%) patients had primary refractory disease (ie, lack of response to initial induction regimen, which in all cases except 1 [treatment with tipifarnib] indicated prior cytotoxic chemotherapy), and 14 (61%) were in their first relapse with a preceding median first remission duration of 7 months (range, 1.9 months-9.3 months). Sixteen (70%) patients had received at least intermediate doses (≥1 g/m2/day) of cytarabine before, and 2 (9%) relapsed from an allogeneic stem cell transplant. The remainder received standard dose cytarabine inductions or noncytarabine-containing regimens (tipifarnib in 1 patient and the multitargeted tyrosine kinase inhibitor PTK787 in a second patient). Fifteen (65%) patients had abnormal cytogenetics, which included abnormalities of chromosome 5 and/or 7, and 11q23 in 3 patients each. For patients in the CIA group, the median age was 56 years (range, 23 years-78 years). One patient had a diagnosis of RAEB-2. Eight (38%) patients had primary refractory disease and 13 (62%) patients were in their first relapse, with a median first remission duration of 12.4 months (4.2 to 59.5). Eleven (52%) patients were exposed to intermediate dose cytarabine-containing regimens, and 2 failed matched unrelated donor transplants before the clofarabine salvage. Abnormal cytogenetics occurred in 12 (57%) patients, including abnormalities of −5/-7 in 5 patients and of 11q23 in 1. The incidence of prior malignancies (excluding squamous or basal cell skin cancer) was 22% for the CI group and 19% for CIA. All but 1 of these patients received prior chemotherapy and/or radiotherapy.
|Clofarabine Plus Idarubicin||Clofarabine, Idarubicin, Plus Cytarabine|
|Median age, y (range)||58 (24-71)||56 (23-78)|
|First CR duration, m (range)||7 (1.9-9.3)||12.4 (4.2-59.5)|
|First CR duration ≥12 m, n||—||8|
|-5 and/or -7/11q23||3/3||5/1|
|At least IDAC||16||11|
|No prior anthracyclines||3||6|
|HSCT||2 allo||2 MUD|
Determination of MTD, DLTs, and Toxicities
The process of dose escalations and occurrence of DLTs by dose level for the CI group are detailed in Table 2 and for the CIA group in Table 3. A summary of all toxicities (including <grade 3) is provided in Table 4. In general, skin rashes occurred more frequently with the 3-drug combination, as did ≥grade 3 hyperbilirubinemia. Mucositis (<grade 3) was more frequent with clofarabine plus idarubicin.
|C (mg/m2)||I (mg/m2)||n||DLTs During First Cycle|
|22.5 × 5||12 × 3||6||Diarrhea, mucositis, esophagitis, skin rash||3|
|15 × 5||8 × 3||6||Bilirubin||1|
|18 × 5||10 × 3||3||—||-|
|22.5 × 5||10 × 3||3||—||-|
|30 × 5||10 × 3||5||SGPT, bilirubin, headaches||2|
|C (mg/m2)||I (mg/m2)||A (g/m2)||n||DLTs During First Cycle|
|22.5 × 5||8 × 3||1 × 5||3||Diarrhea, bilirubin, acute renal failure||3|
|15 × 5||6 × 3||0.75 × 5||6||Bilirubin, skin rash||1|
|22.5 × 5||6 × 3||0.75 × 5||6||Bilirubin||1|
|30 × 5||6 × 3||0.75 × 5||6||Bilirubin, diarrhea, mucositis||2|
|Toxicity||CI (%)||CIA (%)|
|<Grade 3||≥Grade 3||<Grade 3||≥Grade 3|
|Increase in SGPT and/or SGOT||13||4||38||—|
|Increase in creatinine/acute renal failure||13||—||14||5|
The first dose level in the CI group consisted of clofarabine 22.5 mg/m2 intravenously daily × 5 days and idarubicin 12 mg/m2 intravenously daily for 3 days. The combination of the drugs at these doses proved to be too toxic. One patient of the first cohort of 3 developed ≥grade 3 gastrointestinal-related adverse events (diarrhea and mucositis). Another 2 patients in the expansion group developed grade 3 skin rash and diarrhea, respectively. The occurrence of 3/6 patients with DLTs therefore required a reassessment of the study and redesign of the dose escalation plan. After IRB approval of the study amendment, the new starting dose levels were clofarabine 15 mg/m2 daily × 5 days and idarubicin 8 mg/m2 daily × 3 days. One patient developed grade 3 hyperbilirubinemia, but no further DLTs occurred in the expanded group of 6 patients within the 14-day observation period. No DLTs were observed in the next 2 dose levels until 2 of 5 patients developed grade 3 SGPT elevations, and grade 3 hyperbilirubinemia and headaches, respectively. The MTD and recommended phase 2 doses are therefore the next lower dose level of clofarabine 22.5 mg/m2 daily × 5 days and idarubicin 10 mg/m2 daily × 3 days.
A similar picture was seen in the 3-arm combination of CIA. The initial starting dose level of clofarabine 22.5 mg/m2 daily × 5 days, idarubicin 8 mg/m2 daily × 3 days, and cytarabine 1 g/m2 daily × 5 days had to be abandoned. One patient aged 78 years developed grade 4 hyperbilirubinemia between Days 10 and 17 of therapy, with a maximum bilirubin of 25.2 mg/dL. He died on Day 19 with signs of liver failure and in acute anuric renal failure requiring hemodialysis. Autopsy findings included nonbacterial thrombotic endocarditis, acute tubular necrosis and vascular thrombi in the kidneys, hepatomegaly with cholestasis and vascular thrombi, pulmonary congestion with intra-alveolar hemorrhage and focal microthrombi, and no evidence of leukemia in a <5% cellular marrow. Post autopsy cultures revealed CNS in the blood and CNS plus Lactobacillus acidophilus in the right lower lung. Another patient of 78 years developed bilirubinemia of up to 11.4 mg/dL between Days 14 and 22 of therapy. He had ongoing febrile episodes in the absence of positive cultures requiring multiple changes in antibiotics. He died on Day 34 of therapy after suffering bilateral intracerebral hemorrhages, which were considered unrelated to therapy. The marrow was aplastic at the time of death. The third patient of 29 years developed grade 3 hyperbilirubinemia starting on Day 7, which resolved around Day 18 without any further adverse sequelae. After amendment of the study, the dose levels could be gradually re-escalated from clofarabine 15 mg/m2 daily × 5, idarubicin 6 mg/m2 daily × 3, and cytarabine 0.75 g/m2 daily × 5 to a maximum of clofarabine 30 mg/m2 daily × 5, idarubicin 6 mg/m2 daily × 3, and cytarabine 0.75 g/m2 daily × 5. Two patients experienced DLTs (skin rash/diarrhea and diarrhea/hyperbilirubinemia, respectively) at the highest dose levels, so that the MTD and recommended phase 2 doses are the next lower dose levels of clofarabine 22.5 mg/m2 daily × 5, idarubicin 6 mg/m2 daily × 3, and cytarabine 0.75 g/m2 daily × 5.
Myelosuppression was ubiquitous. Prolonged myelosuppression (absolute neutrophil count [ANC] <1 × 109/L and/or platelets <100 × 109/L for >42 days) of responding patients occurred in 3 patients (13%) treated with clofarabine plus idarubicin. The median time to CR in this group was 56 days (range, 53 days-58 days). Prolonged myelosuppression with clofarabine, idarubicin, and cytarabine were observed in only 1 patient (5%). The median time to CR for CIA patients was 31 days (range, 24 days-67 days).
Response rates and characteristics of responders are summarized in Table 5. In the CI group, 3 (13%) patients achieved CR and 2 (9%) achieved CRp, which resulted in an overall response rate of 22%. All responses occurred in patients with AML. Ten patients (48%) achieved CR in the CIA group, which included the only patient with a diagnosis of RAEB-2. In either group, responses occurred after 1 course of therapy, and were observed at all dose levels. Median response duration was 4.4 months (range, 2.3 months-8.1+ months) for CI and 15.5 months (range, 3 months-27.8+ months) for CIA responders. Thirty percent of patients could be referred for a stem cell transplant.
|Clofarabine Plus Idarubicin||Clofarabine, Idarubicin, Plus Cytarabine|
|Complete remission, n (%)||3 (13)||10 (48)|
|Median age, y (range)||52 (27-68)||50 (24-76)|
|By dose of clofarabine, n (%)|
|15 mg/m2||—||4 (19)|
|22.5 mg/m2||1 (4)||4 (19)|
|30 mg/m2||2 (9)||2 (10)|
|Median CRD1, mo (range)||4.5 (0-9)||15 (0-60)|
|Median CRD2, mo (range)||4.4 (2.3-6)||10.6 (3-29.9+)|
Clofarabine is a second-generation nucleoside analog with activity in adult AML. As is a common treatment strategy in AML, experience with single-agent activity quickly led to combination trials in an effort to improve clinical efficacy. The first combinations tested were with cytarabine. Cytarabine is 1 of the most active antileukemic agents and forms the backbone of many combination regimens for patients with AML. Furthermore, preclinical evidence suggested a biochemical interaction between the RNR-inhibiting properties of clofarabine and the augmentation of intracellular levels of ara-CTP (“biochemical modulation”).15, 16 The theory was put to test in clinical trials of clofarabine plus cytarabine in patients with relapsed AML as well as in specifically older patients with previously untreated AML. The clinical experience is summarized in several original articles and presentations.12–14, 17–19 Whereas the contribution of biochemical modulation between clofarabine and cytarabine remains unclear, at least 1 study demonstrated a better response rate with the combination.18 In this study, 70 patients aged ≥60 years with frontline AML were adaptively randomized to low-dose clofarabine (30 mg/m2/dose) versus low-dose clofarabine plus low-dose cytarabine (20 mg/m2 subcutaneously daily for 14 days during the induction). The randomization significantly favored the combination arm, with a CR rate of 63% versus 31% (P = .025), and more favorable event-free survival (P = .04).
In this study, we tried to evaluate the feasibility of further combining anthracyclines with clofarabine ± cytarabine. Anthracyclines are the most common combination partner of cytarabine. Anthracycline activity results in DNA strand breaks, and the combination with nucleoside analogs may synergize with anthracycline activity by inhibiting DNA repair damage. Interactions of 2-chlorodeoxyadenosine (2-CdA, cladribine) with 3 anthracyclines (doxorubicin, idarubicin, and mitoxantrone) have been evaluated on murine leukemias P388 and L1210.20 Combinations of all 3 anthracyclines potentiated the antileukemic activity of cladribine. In case of the combination of 2-CdA plus idarubicin, an increase in the life span of the leukemic mice of 60% was observed, which was significantly higher than any increase in life span gained on monotherapy schedules. With this background, and as idarubicin has been the preferred anthracycline for induction and salvage regimens of patients with AML in our institution, we chose idarubicin as part of the salvage combination program with clofarabine.
This has been the first study to combine idarubicin with clofarabine. DLTs were similar in both treatment groups and included hyperbilirubinemia, elevation of transaminases, and headaches in the CI group, and hyperbilirubinemia, diarrhea, and mucositis in the CIA group. The suggested phase 2 doses for CI are clofarabine 22.5 mg/m2 daily × 5 plus idarubicin 10 mg/m2 daily × 3, and for CIA clofarabine 22.5 mg/m2 daily × 5, idarubicin 6 mg/m2 daily × 3, and cytarabine 0.75 g/m2 daily × 5. The idarubicin combinations proved difficult in the first cohort of patients, where the dose levels in both treatment groups proved too toxic. This was especially obvious with the CIA combination, where the first 2 patients died on study. These events raise an important issue, which is not commonly addressed in phase 1 studies. Phase 1 trials do not usually establish an MTD based on age. It is very conceivable that DLTs may differ between younger and older patients, and that a dose that is effective and feasible in patients younger than eg, 60 years, may prove too toxic in those who are older than 70 years. In our example, both of the patients in the first CIA cohort were 78 years old. Although both of them had failed only biologically targeted therapies before (tipifarnib in 1 case and PTK787 in the other), they were not able to tolerate intensive chemotherapy combinations. Although the deaths may be attributed to causes not directly related to the drug combinations (septic/infectious complications in the first patient and cerebral hemorrhage in the second), the observed liver toxicities are almost certainly drug-related. Fewer complications with less serious consequences occurred in subsequent patient cohorts, who were in general younger than 70 years, with 1 exception of a 76-year-old patient enrolled in the third dose level of CIA and who did not experience ≥grade 3 toxicities. In general our study population is too small to enable going beyond anecdotal evidence of a difference in toxicities in patients older than 70 years. Only 2 patients exceeded 70 years in the CI group, and 6 patients who received the 3-drug combination were at least 70 years old. Comparing the incidence of ≥grade 3 toxicities in these patients with those of the whole group (Table 4), we arrive at the following results: CI, 1 patient with grade 3 skin rash (incidence 50% vs 9% overall); CIA, 1 patient with diarrhea (17% vs 14%), 3 patients with hyperbilirubinemia (50% vs 33%), and 1 patient with an increase in creatinine levels/acute renal failure (17% vs 5%). A larger database of phase I studies would need to be reviewed with regard to difference in toxicities by age.
As patients were allocated to treatment combination based on pretreatment characteristics, notably likelihood of response to cytarabine-containing regimens, it might be criticized that this allocation created patient groups with an inherently skewed likelihood of developing toxicities. Patients who were treated with CIA had on average longer first remission durations, which translated into longer treatment-free periods before salvage for some patients. However, no such effect was evident in our study. The toxicity profile was similar between the groups with respect to both type and severity of adverse events.
No conclusions can be drawn from the differences in response rates between CI and CIA. Although CR rates of up to 48% in a phase 1 study are encouraging, the study did not aim to establish and far less to compare efficacy of the proposed combinations. Differences in patient selection with respect to presumed cytarabine responsiveness or prior exposure to anthracyclines have to be considered when comparing responses between the treatment groups. To answer this question, we are currently conducting an adaptively randomized 3-arm study comparing CI and CIA with clofarabine plus cytarabine as established in a previous trial.13
- 17Clofarabine as first-line treatment of elderly (=65 yrs) AML patients with an unfavorable cytogenetic profile who are unsuitable for standard treatment. Haematologica. 2006; 91( suppl 1): 45., , , et al.
- 18Clofarabine and clofarabine plus low-dose cytarabine (ara-C) as induction therapy for patients (pts) ≥60 years with newly diagnosed acute myeloid leukemia (AML). Blood. 2005; 106: 786a., , , et al.
- 19Phase II study of clofarabine and cytosine arabinoside in adult patients with relapsed AML and in elderly patients with untreated AML who are at high risk of anthracycline toxicity. J Clin Oncol. 2007; 25: 372s., , , et al.