Abstract 2615 Background: The first EORTC-CLG AML pilot study (58872) demonstrated the efficacy of Mitoxantrone (MTZ), substituted for daunorubicin, in the treatment of childhood AML. The subsequent trial (58921) aimed to compare MTZ and Idarubicin (IDA), an anthracycline with a favorable pharmacokinetic profile (such as good CSF penetration) and suggested to be more efficacious than daunorubicin in adult AML trials performed in the last two decades. Methods: Between March 1993 and December 2002, 227 eligible patients (pts) st intensification course, each at a dose of 10 mg/m 2 /d for 3 days in both courses. Concomitant chemotherapy consisted of standard dose Cytarabine (AraC) and Etoposide (Eto) in induction and high-dose AraC (18–36 g/m 2 ) in 1 st intensification. Allogeneic stem cell transplantation after 1 st intensification was recommended for pts who achieved CR and had an HLA-identical sibling donor. Patients in CR without donor had to receive a 2 nd (DCTER = continuous infusion of standard dose AraC, Daunorubicin 4x 20 mg/m 2 /d, Eto, Dexamethasone, 6-thioguanine) and 3 rd intensification course (high dose AraC 12 g/m 2 + Eto), followed by maintenance therapy (6-Thioguanine daily + AraC 4 days/month) for 12 months. CNS irradiation after the 3 rd intensification for patients with initial WBC counts >=70×10E9/L was abandoned from November 1994 on. Randomization was done centrally. The primary endpoint was EFS; secondary endpoints were OS, CR rate after induction/1 st intensification, DFS and toxicity. Intention-to-treat analysis was used. Results: A total of 112 and 115 eligible pts were randomly assigned to receive IDA and MTZ, respectively. The rate of CR after two courses was 79.5% (IDA) vs 85.2% (MTZ). At an overall median follow-up of 9.9 y (range 0.25–16 y), there were 65 vs 59 events in the IDA vs MTZ group: failure to achieve CR (23 vs 17), relapse (35 vs 40), and death without relapse (7 vs 2). The 5-year EFS rate was 42.0% (SE 4.7%) in the IDA group and 48.4% (SE 4.7%) in the MTZ group (hazard ratio (HR) = 1.20, 95% CI 0.84–1.71, 2-sided log rank p=0.29). The 5-year OS rate was comparable in both treatment arms: 59.8% (SE 4.6%) in the IDA group and 57.5% (SE 4.7%) in the MTZ group (HR = 1.03, 95% CI 0.70–1.54, 2-sided log rank p=0.87). In CR patients with (N=46/187) or without (N=141/187) an HLA-identical sibling donor, the 5-year DFS rate from CR was 65.1% (SE 7.1%) and 51.5% (SE 4.2%) respectively (HR=0.65, 95% CI 0.37–1.11, 2-sided log rank p=0.11) and the 5-year OS rate 78.0% (SE 6.1%) and 60.7% (SE 4.1%) respectively (HR=0.53, 95% CI 0.28–1.01, 2-sided log rank p=0.048). This advantage for patients with a donor remained important regarding both DFS (HR=0.60, p=0.07) and OS (HR=0.49, p=0.03), even after adjustment for WBC count at diagnosis, age, cytogenetic features and randomized arm. The interval between start of induction and start of 1 st intensification was similar in both arms (median 5.2 weeks). Grade 3–4 infection following the induction course was 37.5% (IDA) vs 25.4% (MTZ); incidence of fever grade 3–4 was 25% (IDA) vs 22.8% (MTZ). In this trial, the cumulative anthracycline dosage (conversion factor 5) was 380 mg/m 2 . Acute and late-onset cardiotoxicity was comparable in both treatment arms. Conclusion: There was no significant difference in efficacy and in toxicity between the two randomized treatment groups, IDA versus MTZ, although grade 3–4 infection rate following the induction course was slightly higher in the IDA arm. Patients who reached CR and who had a HLA compatible sibling donor had a longer DFS and OS than pts without a donor. Disclosures: No relevant conflicts of interest to declare.