[Relationship between immunological characteristics and prognosis in children with acute myeloid leukemia].

ABSTRACT

OBJECTIVE: The prognostic significance of immunophenotyping in acute myeloid leukemia (AML) has been controversial. This study investigated the relationship of immunophenotypes with French-American-British (FAB) subtypes and chromosomal abnormalities and assessed the prognostic value of immunophenotyping in children with AML. METHODS: From January 1998 to May 2003, 75 children with newly diagnosed AML were enrolled on protocol AML-XH-99. Immunophenotypes were measured with the flow cytometry. According to the McAbs used, the patients were classified into five groups panmyeloid antigens (CD13, CD33, and MPO), myeloid-lineage associated antigens (CD14, CD15), lineage-specific antigens (CD41, GlyA), progenitor-associated antigens (CD34, HLA-DR) and lymphoid-associated antigens (CD19, CD7). The probability of event-free survival (EFS) was estimated by Kaplan-Meier analysis. The distributions of EFS were compared using the log-rank test. Chi-square analysis or Fisher exact test was used to compare the differences in the distribution of biologic presenting features. A Cox proportional hazards model was used to identify independent prognostic factors. RESULTS: At least one of panmyeloid antigens CD13, CD33 and MPO was expressed in 72 patents (97.3%). Two or more panmyeloid antigens were expressed in 45 patients (60.8%). The proportion of children with AML expressing one or more of the lymphoid-associated antigens was 24.3%. Lymphoid-associated antigen CD19 was expressed by blast cells in most of FAB M2 patients. The patients with acute promyelocytic leukemia were characterized by the absence of HLA-DR and lymphoid-associated antigens CD19 and CD7. Monovariate analysis showed immunophenotypes were not related to the complete remission rate after the first induction course and the 5-year-EFS. Multivariate analysis suggested immunophenotyping had no independent prognostic value in AML. CONCLUSIONS: Immunophenotyping can not be used independently in the evaluation of risk classification in children with AML. However, it is useful in the reorganization of special types of AML.