Clofarabine, a novel nucleoside analog, is active in pediatric patients with advanced leukemia

Leukemia is the most common malignancy in children and adolescents.¹  Because of advances in therapeutic approaches and supportive care, approximately 80% of children with acute lymphoblastic leukemia (ALL) and 40% of children with acute myeloid leukemia (AML) are disease free 5 years from diagnosis.^2,3  For the remaining children who do not achieve or maintain complete remission (CR), the prognosis remains dismal. Nucleoside analogs are among the most active agents in hematologic malignancies. Cytarabine, a deoxycytidine analog, is the most active drug in AML.^4,5  Guanosine analogs include the widely used 6-mercaptopurine, 6-thioguanine, and nelarabine.⁶  The deoxyadenosine analogs cladribine and fludarabine have shown activity against childhood leukemia and lymphoproliferative disorders, respectively.^7,8 

A hybrid of these drugs, clofarabine (Cl-F-ara-A, 2-chloro-2′-fluoro-deoxy-9-β-d-arabinofuranosyladenine), retains the 2-halogenated aglycone of fludarabine and cladribine, which renders those analogs resistant to deamination by adenosine deaminase (Figure 1). A fluorine molecule at the 2′-position of the carbohydrate inhibits cleavage of the glycosidic linkage by bacterial purine nucleoside phosphorylase and stabilizes the compound in acidic environments. These may be mechanisms of clearance for the parent drugs, which generate potentially toxic halogenated nucleobases.⁹  As do fludarabine and cladribine, clofarabine requires intracellular phosphorylation by deoxycytidine kinase to be metabolized to the triphosphate form necessary for cytotoxic effect.^10,11  The triphosphate of fludarabine (F-ara-adenosine triphosphate [ATP]) primarily inhibits DNA polymerases, whereas cladribine triphosphate (CdATP) particularly inhibits ribonucleotide reductase. Clofarabine triphosphate (Cl-F-ara-ATP) inhibits DNA polymerases and ribonucleotide reductase.^10-13  A phase 1 trial in adult patients with acute leukemia showed a 16% response rate and defined the maximum tolerated dose (MTD) at 40 mg/m² daily for 5 days; the dose-limiting toxicity (DLT) was hepatotoxity.¹⁴  Based on the activity demonstrated in the adult study, a parallel pediatric phase 1 trial was opened to define the DLT and MTD for children with acute leukemia.

Clofarabine for clinical use was initially prepared by Ash Stevens Inc (Detroit, MI) and formulated for injection by the University of Iowa Pharmaceutical Services (Ames, IA). Subsequently, the drug was produced in bulk by Delmar Chemicals (Lasalle, Quebec, ON, Canada). For cellular pharmacokinetics, high-pressure liquid chromatography (HPLC) standards clofarabine 5′-triphosphate (Cl-F-ara-ATP) was synthesized by Sierra Biochemicals (Tucson, AZ). All other chemicals were reagent grade.

Patients younger than 21 years of age with relapsed or refractory leukemia who were not candidates for treatments of higher efficacy or priority were eligible for study. Informed consent indicating that patients, guardians, or both were aware of the investigational nature of this study was signed according to institutional guidelines. Other eligibility criteria required the absence of chemotherapy in the 2 weeks before study entry and the resolution of any toxic effects of prior therapy; adequate liver and renal function values (bilirubin 2 mg/dL or lower and creatinine 1.5 mg/dL or lower), and good performance status (Zubrod 0-2).

The starting dose level of clofarabine was 11.25 mg/m² intravenously over 1 hour daily for 5 days. Courses were repeated every 2 to 6 weeks, depending on toxicity and response. A modified 3 + 3 design was followed, with doses ranging from 11.25 to 70 mg/m² per day. As in the classic 3 + 3 design,¹⁴  at least 3 patients were treated at each dose level unless, because of slow accrual, the parallel adult trial reached 2 dose levels above the pediatric trial without evidence of significant toxicity. If that occurred, pediatric patients could be entered at 1 dose level below a determined safe dose level in adults. Once the adult MTD was reached, dose escalation in pediatric patients occurred in 30% increments following a classic 3 + 3 design until the pediatric MTD was defined.

CR was defined as 5% or less blasts in cellular marrow, an absolute granulocyte count greater than 10⁹/L, and a platelet count greater than 100 × 10⁹/L. Partial response (PR) was defined as 6% to 25% marrow blasts, an absolute granulocyte count greater than 0.5 × 10⁹/L, and a platelet count greater than 25 × 10⁹/L. Patients with marrow CR or PR not associated with peripheral count recovery were considered to have hematologic improvement (HI). Toxicity was graded on a scale of 0 to 5 using the National Cancer Institute Common Toxicity Criteria (NCI-CTC) except for defining myelosuppression.¹⁵  Myelosuppressive toxicity in patients with acute leukemia was defined as marrow with less than 5% cellularity and without evidence of leukemia, lasting for more than 6 weeks from the start of therapy. All patients who received at least 1 dose of clofarabine were considered evaluable for toxicity.

Blood samples were obtained from 12 patients who agreed to have blood drawn for pharmacologic determinations. Two patients (patients 10 and 17) were studied during the first and second courses of therapy. Samples were obtained before therapy for baseline values, at the end of the infusion on the first day, and usually also on days 2 to 4. Blood samples were collected in green stopper Vacutainer tubes containing heparin and 1 μM deoxycoformycin as a precaution against clofarabine deamination during sample processing (obtained from the NCI, Bethesda, MD) and immediately placed in an ice-water bath and processed as previously described.¹⁶  All patients, guardians, or both gave written, informed consent for plasma and cellular pharmacology investigations.

Plasma was removed after centrifugation and stored at –70°C until analyses were performed. Clofarabine plasma levels were analyzed using reverse-phase HPLC with a tandem quadruple mass spectrometer according to a modified, previously described procedure.¹⁴ 

Cell pellets from blood samples were diluted with phosphate-buffered saline, and mononuclear cells were isolated using Ficoll-Hypaque densitygradient step-gradient centrifugation procedures described previously.¹⁶  A Coulter electronics channelizer (Coulter, Hialeah, FL) was used to determine the mean cell volume. Normal nucleotides and Cl-F-ara-ATP were extracted from cells using standard procedures and analyzed using HPLC, as described in detail previously.^14,17 

Leukemia cells were obtained before and at the end of clofarabine infusion on consecutive days of therapy to determine levels of DNA synthesis. The cells (2 × 10⁶ in triplicate) were incubated ex vivo with 1.0 μCi (0.037 MBq) [³H]thymidine for 30 minutes to determine DNA synthesis, as described previously.¹⁸  The cells were collected on a 25-mm glass fiber disc (prewetted with 1% sodium pyrophosphate) by filtration and then washed 2 times with 4 mL ice-cold 0.4 N HClO₄ and twice with 2 mL ethanol. Radioactivity retained on the filter disc was determined by liquid scintillation counting and expressed as a percentage of untreated (control) cells.

Linear regression for plasma clofarabine levels and nonlinear regression analyses for r and rectangular hyperbola curves for clofarabine triphosphate accumulation were obtained using the Prism software program (GraphPad Software, San Diego, CA).

Twenty-five patients were enrolled between August 23, 2000, and May 6, 2002. Their characteristics are summarized in Table 1. Median age was 12 years (range, 1-19 years). Most patients were heavily pretreated, and 36% had undergone prior stem cell transplantation. The first patient was treated at the starting dose level of 11.25 mg/m² daily for 5 days. Because of faster accrual on the parallel adult trial, the second patient was treated at 15 mg/m² without significant toxicity, and the dose level was escalated to 30 mg/m² for a second cycle. Two additional patients were treated at 30 mg/m², resulting in a total of 3 patients at this dose level. Because the adult DLT was hepatotoxicity at the 55 mg/m² dose level, the pediatric dose level was then escalated to 40 mg/m², 52 mg/m², and 70 mg/m² in a classic 3 + 3 design. At 40 mg/m² daily for 5 days, the cohort was expanded because of grade 3 diarrhea observed in 1 patient. No significant mucositis or diarrhea was observed in the expanded cohort. Of the 6 patients treated at 40 mg/m,² 4 had transient grade 3 elevations in transaminase levels that resolved before the initiation of the second cycle. All 4 patients were retreated at the same dose level; 2 received a total of 2 cycles, and 2 completed 3 cycles. Two of the 4 patients then underwent stem cell transplantation and had no hepatic complications. Further escalated doses to 52 mg/m² and 70 mg/m² daily for 5 days were initiated. Three patients were treated at 52 mg/m² and experienced no hepatotoxicity. Two patients treated at 70 mg/m² had DLTs. One patient had grade 4 hyperbilirubinemia and a grade 3 elevation in transaminases, and another patient had a grade 3 skin rash. All 3 conditions ultimately resolved with no sequelae. The patient with hepatotoxicity had septic shock concomitant with elevated bilirubin and transaminase levels. His medical condition could have contributed to the liver toxicity. We expanded the number of patients at the lower dose schedule of 52 mg/m² daily for 5 days. Of the 13 patients treated at this dose level, 3 had grade 2 hyperbilirubinemia and 2 had grade 3 elevation in transaminases. In all 5 patients the elevation in liver enzymes was transient and reversed by day 14. This was the recommended dose schedule for the phase 2 pediatric studies. Some patients experienced irritability associated with the 1-hour infusion, which resolved when the drug was given over 2 hours. Therefore, the infusion time in the ongoing pediatric phase 2 studies is 2 hours. Other toxicities are detailed in Table 2.

Four (24%) of 17 patients with ALL achieved CR. One patient (6%) achieved PR, for an overall response rate for ALL of 30%. Of the 8 AML patients, 1 (13%) achieved CR and 2 (25%) achieved PR for an overall response rate for AML of 38% (Tables 3, 4, and 5). Patient 2 had Philadelphia-positive ALL refractory to 3 consecutive inductions including idarubicin, fludarabine, and cytarabine. She achieved complete morphologic and cytogenetic remission after 2 cycles of clofarabine. She then underwent stem cell transplantation and maintained morphologic and cytogenetic remission for 70 weeks.

At the clofarabine dose level of 40 mg/m², 2 patients achieved CR and 2 achieved PR. Patient 8, with diploid ALL refractory to 2 consecutive induction regimens, achieved CR after the first cycle of clofarabine at 40 mg/m². He received a second cycle followed by stem cell transplantation and maintained remission for 57 weeks. Patient 7, with ALL in second relapse, achieved PR with the first cycle at 40 mg/m² and CR after the second cycle. He received a third cycle followed by stem cell transplantation and remains in remission at 76+ weeks. Two additional patients treated at this dose level achieved PR. Both had AML. Patient 5 died of aspergillosis after bone marrow transplantation (BMT), and patient 10 had progressive disease while awaiting transplantation.

At the clofarabine dose level of 52 mg/m², 2 patients achieved CR and 1 patient achieved PR. Patient 12 had an AML relapse refractory to salvage with fludarabine/cytosine arabinoside (FLAG). The first cycle of clofarabine resulted in HI, the second in PR, and the third in CR. The patient did not undergo stem cell transplantation and was maintained on clofarabine therapy for a total of 8 cycles. He remained in CR for 64 weeks (43 weeks after the discontinuation of clofarabine therapy). Patient 17 had T-cell ALL refractory to 3 induction regimens including compound 506U and achieved CR with clofarabine therapy but died of complications resulting from stem cell transplantation. Patient 19 with ALL in second relapse achieved PR. Two additional patients treated at 52 mg/m² per day for 5 days achieved marrow CR and underwent stem cell transplantation before complete peripheral count recovery. Their responses were defined as HI. Patient 20 has been in remission for more than 42 weeks. Patient 21 died of complications resulting from transplantation. Patients 11 and 14 had HI after 1 cycle at the 52 mg/m² and the 70 mg/m² dose level, respectively. Each acquired a fungal infection and received no further therapy.

As with other nucleoside analogs, the peak level of clofarabine in plasma occurred at the end of the infusion. To evaluate the dose-dependent accumulation of clofarabine, we compared the level of plasma clofarabine at the end of infusion on day 1 (Figure 2). There appeared to be a linear relationship for the plasma clofarabine concentration and the doses administered (r = 0.75, n = 13, P = < .002 for slope). At 52 mg/m², the MTD, the median plasma clofarabine level, was 1.1 μM (range, 0.4-2.3 μM; n = 7). At 24 hours after infusion, plasma in 8 of the 10 patients contained clofarabine; the concentration ranged from 0.05 to 0.10 mM. Generally, the end-of-infusion value remained similar on each day; however, some samples had minor increases in the concentration of clofarabine during subsequent dosing. In 2 patients, the duration of clofarabine infusion was changed because of toxicity during drug administration. Clofarabine levels at the end of infusion in both these patients were similar after a 1-hour or a 2-hour infusion (1.72 vs 1.52 μM [patient 17] and 1.05 and 1.08 μM [patient 23]).

Levels of clofarabine triphosphate were analyzed at the end of clofarabine infusion in the circulating leukemia blasts of 11 patients. As shown in Figure 3, though there seemed to be a trend for a dose-dependent increase in the clofarabine triphosphate concentration, there was wide heterogeneity among patient blasts to accumulate the analog triphosphate. At MTD (n = 5), the variation was between 6 μM and 19 μM.

DNA synthesis activity in leukemia blasts was measured in 4 patients at doses of 11.25, 15, 40, and 52 mg/m². Generally, after each infusion, there was a rapid decline in the DNA synthetic capacity of the leukemia blasts. However, there was 10% to 50% recovery of DNA synthesis at 24 hours (before the second infusion). The recovery of DNA synthesis seemed to be inversely related to the dose of infusion. At the maximum tolerated dose or higher doses, the inhibition of DNA synthesis was maintained throughout therapy (Figure 4).

Relapsed leukemia is the fourth most common pediatric malignancy. Despite the current success in curing children with newly diagnosed leukemia, resistant forms of the disease still represent a leading cause of cancer-related deaths in children. A continuing effort to identify new antileukemic agents and novel therapeutic approaches in childhood leukemia is essential. Clofarabine, a deoxyadenosine nucleoside analog designed to incorporate the antimetabolic properties of fludarabine and cladribine, showed encouraging activity in adult patients with lymphoproliferative disorders and acute leukemias in a phase 1 trial conducted at the University of Texas MD Anderson Cancer Center.¹⁴  A parallel trial was opened for children with advanced leukemia. The MTD for pediatric leukemia was 52 mg/m² per day for 5 days. The DLT in pediatric leukemia was reversible hepatotoxicity and skin rash. Patients who had undergone earlier transplantation did not have significantly greater toxicity. Twenty patients received more than 1 cycle of clofarabine (12 patients had 2, 7 patients had 3, and 1 patient had 8 cycles), and 7 patients underwent stem cell transplantation after clofarabine therapy. Significant hepatotoxicity was not observed with repeated cycles and did not complicate the transplantations that followed. Objective responses (CR, PR, HI) were observed in 48% of children with multiple relapsed refractory ALL and AML, including patients resistant to fludarabine/cytarabine therapy and 4 patients who underwent earlier transplantation. The responses achieved were relatively durable. Four CRs lasted longer than 50 weeks; this extended remission included a patient with AML refractory to fludarabine/cytarabine therapy who declined stem cell transplantation and remained in unmaintained CR for 43 weeks after completing 8 courses of clofarabine. This suggests clofarabine activity in both ALL and AML. Significant responses were observed with nelarabine and cladribine but were restricted to T-cell leukemia⁶  and childhood AML,⁸  respectively. In a phase 1 study, cladribine demonstrated activity in 14% of the 31 children treated,¹⁹  with responses lasting 6 to 16 weeks. The response rate was 59% in a phase 2 cladribine trial, with all CRs occurring in children in first relapse.²⁰  Later studies showed that cladribine was as effective as doxorubicin or cytarabine as a single agent in newly diagnosed pediatric AML.^7,21  The 32% response rate and CR duration in our current study suggests that clofarabine is a more active agent than cladribine in pediatric leukemia. Unlike cladribine, clofarabine is active in both AML and ALL and can be given as a short infusion on an outpatient basis. Although cladribine was not active in adult leukemia and induced neurotoxicity in this population,^22,23  clofarabine showed activity in adult leukemia and did not induce the neurotoxicity associated with other nucleoside analogs.^14,24  One patient with AML who received clofarabine monotherapy remained in complete remission for 43 weeks without therapy.

In light of the promising activity of clofarabine in pediatric AML, it is important to consider combining other agents with clofarabine based on laboratory results and clinical experience. Clofarabine and cytarabine target 2 different sites in DNA.^10,12  The pharmacokinetic profile of clofarabine nucleotides in leukemia blasts is favorable compared with other deoxyadenosine analogs. In preclinical settings, prior treatment with clofarabine increases the ability of leukemia cells to accumulate ara-CTP (1-beta-D-arabinofuranosylcytosine triphosphate) through a mechanism driven by the inhibition of ribonucleotide reductase by clofarabine triphosphate.²⁵  Hence, one promising approach is to sequentially combine clofarabine with cytarabine, as had been done with fludarabine^26,27  and with cladribine.^28,29  We are testing such a biochemical modulation strategy in circulating leukemic cells from adults receiving combination therapy with clofarabine and cytarabine.³⁰  Another approach is to use clofarabine as an inhibitor of excision DNA repair elicited by DNA-damaging agents.³¹  Such a mechanistic interaction could be the basis for the design of treatment plans combining clofarabine with anthracyclines, alkylating agents, or other drugs that initiate excision DNA repair processes.

In summary, clofarabine is the first deoxyadenosine analog with promising activity in pediatric ALL, pediatric AML, and adult acute leukemia at doses that are well tolerated. Phase 2 pediatric multicenter studies are under way. Combinations of clofarabine, Ara-C (1-beta-D-arabinofuranosylcytosine), and anthracyclines are being tested in adults with leukemia based on toxicity profiles and pharmacokinetic properties.²⁴