Abstract
Several mutations and aberrant gene expression have been proposed as prognostic factors in cytogenetically normal acute myeloid leukemia (CN-AML), but few studies have analyzed the clinical value of several genes in a large group of CN-AML patients. In 121 de novo CN-AML included into the Spanish PETHEMA therapeutic protocols, we evaluated FLT3 and NPM1 mutations and BAALC, ERG, EVI1, MLL-PTD, MN1, PRAME, RHAMM and WT1 RNA levels. Ninety-one patients achieved CR with standard induction therapy while 20/111 (16.5%) were refractory to treatment. This latter subgroup showed higher ERG (media 1.0±0.8 vs. 0.6±0.6;p=0.012) and lower PRAME (media 28.8±52.9 vs. 1640.6±6102.0;p=0.014) levels than no-refractory patients. Moreover, high ERG expression (defined by median value) was associated with shorter OS at 2-years (38% vs. 61%;p=0.012) and RFS at 2-years (38% vs. 60%;p=0.005). By contrast, high PRAME levels (defined by 75th percentile) was related to longer OS (61% vs. 48%;p=0.029) and RFS (74% vs. 43%;p=0.016). Additionally, high EVI expression (defined by 75th percentile) was associated with a poorer OS (37% vs. 56%;p=0.049). The same results in OS were observed in patients with high BAALC RNA levels (median value; 41% vs. 62%; p=0.049). Interestingly, ERG, EVI1 and PRAME markers improved the current prognostic classification of CN-AML based on FLT3 and NPM1 status. Therefore, intermediate risk patients (FLT3-ITD/NPM1mut and FLT3wt/NMP1wt) with low ERG, low EVI1 or high PRAME achieved long OS and RFS, similar to the favorable FLT3wt/NPM1mut subgroup. By contrast, the opposite expression subgroup (high ERG, high EVI1 or low PRAME) showed OS and RFS similar to the adverse FLT3-ITD/NPM1wt subset. These results allowed identify clearly two risk-subgroups for OS and RFS, based on three molecular markers (FLT3, NPM and one of the 3 proposed genes). In conclusion, we demonstrated that ERG, EVI1 and PRAME are relevant prognostics markers in CN-AML and they could improve their risk-stratification.
Disclosures: No relevant conflicts of interest to declare.
Author notes
Corresponding author