Clinical significance of minimal residual disease in childhood acute myeloid leukemia.

Many studies have assessed the clinical significance of the detection of minimal residual disease (MRD) in acute leukemia. Thus far, many studies have suggested that MRD detection to evaluate the response to chemotherapy is useful for predicting the prognosis of childhood acute lymphoblastic leukemia (ALL). However, few studies have reported on the significance of MRD in childhood acute myeloid leukemia (AML), because of small numbers of patients and limited availability of MRD markers. Therefore, we monitored MRD using currently available markers at several points during the treatment for childhood AML and tried to intensify the treatment based on the results of MRD. Thirty-one patients (26 de novo cases and 5 other cases) were examined for MRD between February 1999 and May 2002. After the first consolidation therapy (consolidation 1), the expression of Wilms tumor gene (WT1) and/or leukemia-specific fusion genes such as AML1/MTG8, PML/RAR alpha, and MYH11/CBF beta were analyzed. Patients with positive MRD but in hematological remission at that point were recommended to undergo stem cell transplantation (SCT). Positive WT1 expression (more than 10(3) copies/microgram RNA) was detected in 18 of 31 patients (58.1%) at onset. After consolidation 1 therapy, the WT1 expression became negative in 14 of 18 patients. The AML1/MTG8 fusion gene was expressed in 8 patients, PML/RAR alpha was expressed in 3 patients, and MYH11/CBF beta was expressed in 1 patient. Four of the 8 patients with AML1/MTG8 expression and all 3 with PML/RAR alpha expression also demonstrated positive WT1 expression at onset. Eight (5 de novo cases and 3 other cases) of the 31 patients had no available MRD markers. Four patients who showed pesistently high expression of WT1 after consolidation 1 therapy underwent SCT, and only 1 patient remained in complete remission (CR). Among 14 patients who became negative for WT1 expression, 6 patients received SCT for various reasons. Among 8 patients with the AML1/MTG8 fusion gene, 2 became MRD negative and 6 continued to be positive. Four of these 6 patients underwent SCT, and all but one who underwent syngeneic SCT became MRD negative. On the other hand, 1 of the 2 patients who continued on chemotherapy continued to be MRD positive, suggesting a graft-versus-leukemia effect in allogeneic SCT. All patients with the PML/RAR alpha and MYH11/CBF beta fusion gene continued to be in CR. The 3-year event-free survival in de novo AML was 69.4% +/- 9.8% (n = 26), a result that is encouraging and superior to other reported outcomes. Thus, an MRD-based treatment strategy together with conventional risk factors appears to be required for further improving the outcomes of AML.