Identification of Post-Induction Minimal Residual Disease by Multidimensional Flow Cytometry Identifies Patients with AML at High Risk of Relapse and Poor Outcome- a Report From the Children's Oncology Group

Abstract 1702

Multidimensional flow cytometry (MDF) is used to identify risk in childhood ALL; however, its utility in AML has been limited. We used 4-color MDF with standard (rather than patient-specific) panels to evaluate diagnostic and postinduction bone marrow specimens from patients treated on Children's Oncology Group (COG) study AAML03P1 for evidence of minimal residual disease (MRD). A total of 254 patients submitted marrow specimens for MRD assessment at the end of induction I, the end of induction II, and the end of therapy. Of the 222 patients with evaluable specimens at the end of induction I, 191 (86%) were in morphologic remission, and 27 (12.2%) had persistence of morphologic disease, 3 had persistent CNS disease and 1 was not evaluable for response. Of those with morphologic disease, 15 were in partial remission (PR, 5%-20% blasts), and 12 had refractory disease (RD, >20% blasts). MDF of specimens showing morphologic disease revealed that 7 (26%) did not have evidence of disease; thus, MDF identified patients with reported morphologic disease who did not have immunophenotypic evidence of disease. Overall, in 222 patients with evaluable marrow at the end of induction I, 69 (31%) had evidence of various levels of MRD by MDF (% blast range, 0.02%-43%, median, 1.5%). For the 208 patients with known cytogenetic data, the presence of MRD was evaluated in the following cytogenetic subgroups: favorable risk, defined as t(8;21) or inv(16) (the Core Binding Factor leukemias); unfavorable risk, defined as –5/del(5q) or –7; and intermediate risk, defined as all other cytogenetic subtypes. MRD prevalence at the end of Induction I in patients with favorable, intermediate-risk, or high-risk cytogenetics was 13%, 36%, and 67%, respectively. Prevalence of MRD at the end of induction I was 50% (10/20) in patients with FLT3/ITD, 40% (4/10) in patients with CEBPA mutations, and 0% (0/5) in patients with NPM1 mutations. Of the 222 patients with evaluable specimens at the end of induction I, 191 had morphologic response to the initial chemotherapy. Of those, 57 (28%) had evidence of disease by MDF. Cumulative relapse risk (RR) and disease-free survival (DFS) was assessed in those with or without MRD. Those with MRD at the end of induction I had a RR at 3 years from the end of induction of 60% vs. 29% for those without MRD (p <0.001). DFS was 32% for those with MRD and 65% for those without MRD (p<0.001). In a multivariate analysis, which included cytogenetic and molecular risk factors, the presence of MRD was highly associated with outcome and was an independent predictor of relapse (p<0.001) and worse survival (p<0.001). We further evaluated the significance of clearance of residual disease. Of the 91 patients evaluated for MRD at the end of therapy, 7 (8%) were MRD-positive (6 of whom relapsed). Of the 84 patients who were MRD-negative, 22 (26%) had previously documented MRD. For those with a history of MRD, RR from the end of therapy was 64%, and for those without previous MRD, it was 25% (p<0.001). Therefore, despite clearance of MRD, patients with previous MRD had a high RR. Given the high correlation of MRD with RR, MDF assessment of post-induction response should be incorporated into AML clinical trials for risk identification and assignment to the appropriate risk-based therapy.