High WT1 expression predicted induction chemotherapy failure in acute myeloid leukemia patients with non-favorable cytogenetic risk.

Abstract

The prognostic significance of WT1 expression at diagnosis in acute myeloid leukemia (AML) remains obscure, and subgroup analysis is the way for clarification. We previously reported the results in t(8;21) AML. In this study, 437 consecutive adult AML patients with non-favorable cytogenetic risk were enrolled. All patients were tested WT1 transcript levels using real-time quantitative PCR at diagnosis; AML-related common fusion genes, KMT2A-PTD, FLT3-ITD, NPM1, CEBPA and TP53 mutations were simultaneously tested. 92.4% of patients overexpressed WT1 compared to normal bone marrow. The existence of FLT3-ITD, NPM1 mutation and the absence of CEBPA biallelic mutation were significantly related to higher WT1 expression. The cutoff value for WT1 was determined by performing receiver operating characteristic curve analysis in regard to complete remission (CR) achievement and was used to categorize patients into low-expression (WT1-L) and high-expression (WT1-H) groups. In the entire cohort, WT1-H was significantly associated with a lower 1-course and 2-course CR rate (P < 0.0010 and P = 0.0060) but was not related to relapse-free survival (RFS). Multivariate analysis showed that WT1-H was an independent adverse prognostic factor for both 1-course and 2-course CR achievement. Subgroup analysis was further performed. WT1-H had a significant adverse impact on CR achievement within intermediate-cytogenetic risk, high-cytogenetic risk, ELN-defined-intermediate-risk, normal karyotype, KMT2A rearrangement, FAB-M2, FAB-M5 and NPM1 mutation (+) subgroups, whereas it had no impact within ELN-defined-low-risk, ELN-defined-high-risk, FAB-M4, FLT3-ITD mutation (+) and CEBPA biallelic mutation (+) subgroups. Moreover, WT1-H patients had a significantly lower RFS rate than WT1-L patients within both FAB-M5 and KMT2A rearrangement subgroups (P = 0.010 and 0.028), whereas WT1 had no impact on RFS within other subgroups mentioned above (all P > 0.05). Therefore, high WT1 expression at diagnosis independently predicted induction chemotherapy failure in AML patients with non-favorable cytogenetic risk, and it was related to relapse just within FAB-M5 and KMT2A rearrangement subgroups.