To improve prognosis in children with acute myeloid leukemia (AML) by randomized comparisons of (1) two short consolidation cycles versus the Berlin-Frankfurt-Muenster (BFM) -type biphasic 6-week consolidation and (2) the prophylactic administration of granulocyte colony-stimulating factor (G-CSF) versus no G-CSF. Further, therapy for standard risk patients was intensified by addition of a second induction, HAM (high-dose cytarabine and mitoxantrone).

Four hundred seventy-three patients younger than 18 years with de novo AML were enrolled in trial AML-BFM 98. Patients received five courses of intensive chemotherapy, cranial irradiation, and 1-year maintenance therapy.

Four hundred eighteen patients (88%) achieved remission. Compared with trial AML-BFM 93, early deaths decreased from 7.4 to 3.2% (P = .005), and 5-year overall survival increased from 58% to 62% (log-rank P = .03). Both types of consolidation therapy led to similar outcome (event-free survival, 51% v 50%), but in the two-cycle arm, treatment duration was shorter (median duration, 15 days), and treatment related mortality was lower (five v nine patients). G-CSF shortened neutropenia, but did not reduce the rate of severe infections. Intensification of induction therapy did not improve prognosis of standard-risk patients (event-free survival, 62% v 67%).

Overall results were improved by neither the administration of G-CSF nor by cycle therapy; however, the latter was easier to perform. Compared with study AML-BFM 93, therapy intensification with HAM in standard-risk patients did not result in improved prognosis. Future treatment designs have to balance intensification of treatment with higher toxicity, improve supportive care, and to consider alternative treatment strategies.