Relapsed/refractory pediatric AML has a poor prognosis despite salvage therapy including stem-cell transplantation. Chemotherapy using FLAG plus liposomal daunorubicine (FLAG-DNX) is currently considered the standard in 1strelapse in Europe (Kaspers et al, JCO 2013). FLAG is based on potentiation of cytarabine (Ara-C) by fludarabine (Flu) by increasing Ara-CTP levels. Clofarabine (CLO) is a novel purine nucleoside analog, designed to have improved efficacy.


We initiated an ongoing phase 1B dose-escalation study using a ‘3x3 design’ to define the optimal dose of CLO, replacing FLU in FLAG-DNX. Dosages consisted of Ara-C 2gr/m2/day (day 1-5), with escalation of DNX from 40, 60 to 80 mg/m2/day (day 1, 3 and 5), and CLO from 20, 30 to 40 mg/m2/day (day 1-5) in 5 dose-levels, without GCSF priming. At day 6 intrathecal Ara-C was administered. Dose-level 1-4 recruited early 1st relapse, refractory 1st relapse and 2nd relapse of AML, and escalated CLO to 40 mg/m2 with DNX 60 mg/m2. Dose level 5 is open to 1st relapse pts and escalates DNX to 80 mg/m2. Responding patients received a 2ndconsolidation with CLARA-DNX or treated by investigator discretion. Other inclusion criteria consisted of performance status ≥60%, and normal liver and renal function. Maximum tolerated dose was the primary endpoint. Hematologic DLTs were defined as count recovery in responding pts >42 days. Serum and CSF were collected for pharmacokinetics (PK), taken at day 1: predose, T=2hrs and T=5 hrs post-infusion, and repeated on day 5, including a T=24 sample on day 6, together with a CSF sample just prior to intrathecal medication. CLO plasma and CSF concentrations were analyzed using LC-MS/MS. Efficacy data are awaiting central review and not released until the end of the study. The study is registered in the Dutch Trial Registry (nr. 1880).


We here report safety and PK data on the first 4 dose-levels after accrual of 22 AML patients (≥2nd relapse, n=7; refractory 1st relapse, n=9; early 1strelapse, n=6). Patients were treated at dose-level (DL) 1, n=4; 2, n=3; 3, n=12; 4, n=3. Eleven patients had been transplanted prior to enrollment. In total 27 CLARA-DNX courses were administered. Median age was 6.4 years, 14 patients were male. 20 SAEs were reported, consisting of febrile neutropenia (n=11), typhlitis (n=1), lung-infection fungal (n=3), lung-infection bacterial (n=2), candida sepsis (n=1), rash (hand-foot skin reaction) (n=1), and tumor lysis/capillary leak syndrome (n=1). Other grade 3-4 AEs reported in at least 2 patients consisted of abdominal pain, nausea and vomiting, diarrhea, anorexia and allergic reaction. Two patients had bilirubin elevation (max 61 umol/l); six patients had mild transaminase elevation > 100 U/l (max. 330U/l), which was reversible.

At dose-level 3, two patients experienced DLTs (fungal infection) halting dose-escalation. Detailed analysis showed that most patients had evidence of subclinical fungal infections prior to enrollment. Therefore a protocol amendment was issued that required screening patients with CT-scan and serum-galactomannan prior to enrollment. After enrolling another cohort of 6 patients at DL3 this appeared to be safe (1 DLT of Candida sepsis out of 6 patients). No DLTs occurred at dose-level 4 in 3 patients. No hematological DLTs occurred.

PK samples were available from 19 patients; in 2 patients sampling was repeated at course 2. At day 1 the median AUC was 28 ng/ml.mg.hr (range 6-401), with a mean T1/2 of 1.5 hrs. Day 1 and day 5 results were similar. CSF levels were not measurable in most patients and were 0.1-0.2 ng/ml.mg in the 3 patients with detectable levels. Our PK-data in combination are not significantly different from our data with single-agent CLO in relapsed ALL patients (Joel et al, AACR, 2007), and fit the CLO single agent PK model of Bonate et al (J Clin Pharmacol 2004, 44: 1309-32) .


The RP2D of CLO in a CLARA/DNX course in relapsed/refractory pediatric AML is 40 mg/m2, excluding patients with evidence of prior subclinical aspergillus. There is no evidence for PK interaction between CLO and the other drugs. We are currently testing the safety of an augmented dose of DNX (80 mg/m2) in 1strelapse AML patients. When tolerable this dose may be randomized against other induction regimens in front-line therapy in pediatric AML.


This study was financially supported by Sanofi and by the KiKa-foundation grant nr. 26.


Off Label Use: Clofarabine for pediatric AML.

Author notes


Asterisk with author names denotes non-ASH members.