Prognostic variables in newly diagnosed children and adolescents with acute myeloid leukemia: Children's Cancer Group Study 213

The objective of this study was to identify biologic parameters that were associated with either exceptionally good or poor outcome in childhood acute myeloid leukemia (AML). Among the children with AML who entered Children's Cancer Group trial 213, 498 patients without Down syndrome or acute promyelocytic leukemia (APL) comprise the basis for this report. Univariate comparisons of the proportion of patients attaining complete remission after induction (CR) indicate that, at diagnosis, male gender, low platelet count (less than or equal to20 000/mul), hepatomegaly, myelodysplastic syndrome (MDS), French-American-British (FAB) category M5, high (>15%) bone marrow BM blasts on day 14 of the first course of induction, and +8 are associated with lower CR rates, while abnormal 16 is associated with a higher CR rate. Multivariate analysis suggests high platelet count at diagnosis (>20 000/mul), absence of hepatomegaly, less than or equal to15% day 14 BM blast percentage, and abnormal 16 are independent prognostic factors associated with better CR. Univariate analysis demonstrated a significant favorable relationship between platelet count at diagnosis (>20 000/mul), absence of hepatomegaly, low percentage of BM blasts (less than or equal to15%), and abnormal 16 with overall survival. Absence of hepatomegaly, less than or equal to15% day 14 BM blast percentage, and abnormal 16 were determined to be independent prognostic factors associated with better survival.