Abstract
Abstract 1569
Poster Board I-593
Acute myeloid leukemia (AML) is a clonal disorder characterized by acquired genetic abnormalities in hematopoietic progenitors. CCAAT/enhancer binding protein α (C/EBPα) is a basic leucine zipper transcription factor that regulates differentiation-dependent genes during granulocyte differentiation. CEBPA mutations have been observed in several human malignancies. Epigenetic modification of the distal CEBPA promoter region (-1422 to -896) has been reported to result in down-regulation of CEBPA expression in lung cancer, pancreatic cancer, and head and neck squamous cell carcinoma. Apart from solid tumors, epigenetic modification of CEBPA in AML has also been reported. Chim et al. reported aberrant CEBPA core promoter methylation in a small proportion (2.85%) of patients with AML. Wouters et al. identified a subgroup of AML patients (1.4%) in which the CEBPA gene was silenced in association with CEBPA core promoter methylation. Hackanson et al. observed CEBPA methylation in the distal promoter region in 51% of selected AML patients. Due to the selection of differing patient populations and CEBPA promoter regions for analysis in past studies, the clinical significance of CEBPA methylation in AML remains unclear. To evaluate the CEBPA methylation status of patients with AML, 193 patients with de novo AML were evaluated for CEBPA methylation by bisulfite sequencing and MSP methods. A substantial proportion of patients were noted to have some degree of heterogeneous methylation in the distal region (-1423 to -1121), but not in the proximal region (-1121 to -896) or in the core region (-141 to +103), of the CEBPA promoter. Only one patient was noted to have broad methylation from the distal region to the core of the CEBPA promoter, and extending into exon1 (-25 to +103). This patient was a 61-year-old female with M1 subtype AML, a normal karyotype and an absence of CEBPA, NPM1 or FLT3 mutations. We further correlated the CEBPA methylation levels with CEBPA transcript levels in leukemic cells from 12 patients with AML and found a negative correlation (coefficient of correlation = -0.722, P = 0.017). To clarify the correlation between CEBPA methylation and clinical features in AML, quantitative MassARRAY analyses for CEBPA methylation were performed. Among the 193 patients studied, methylation levels ranged from 0.073 to 0.971 with a median value of 0.229, compared to normal bone marrow cells (NBM) with levels ranging from 0.207 to 0.269 (n=5, median: 0.218) and mature leukocytes (NPB) which ranged from 0.144 to 0.201 (n=5, median 0.161). Using complete linkage clustering, we divided the patients into two groups: one with high CEBPA methylation levels (≥ 0.55, n = 28), and the other with low methylation levels (< 0.55, n = 165). Clinical and laboratory features of these two groups are explored. Gene alteration analysis revealed NPM1 mutation was mutually exclusive with CEBPA methylation (P = 0.004), so was similar trend in CEBPA mutation. Of the 28 patients with high CEBPA methylation, only one patient had a CEBPA mutation. immunophenotyping work revealed that leukemic blasts from patients with high CEBPA methylation were predominately negative for HLA-DR (P = 0.005) and CD56 (P = 0.015) expression. Of the 140 patients receiving standard induction therapy, 110 (79.3%) patients achieved CR. With a median follow-up of 17.5 months, patients with high CEBPA methylation demonstrated a greater probability of survival (median >60 months vs. 20 months, P = 0.020) than those with low CEBPA methylation. In addition, in the subgroup of AML patients without favorable karyotypes as well as NPM1 and CEBPA mutations, patients with high CEBPA methylation (n = 14) still had better five-year survival (median 34 months vs. 11 months, P = 0.048) than those with low CEBPA methylation. Intriguingly, among the cytogenetically normal AML patients, except those with CEBPA and NPM1 mutations, the five-year survival of patients with high CEBPA methylation (n=8) was remarkably better than those with low CEBPA methylation (n=19; median >60 months vs. 11 months, P = 0.002). These results suggest that, in addition to current cytogenetic and molecular markers, CEBPA methylation status could emerge as a prognostic factor for survival advantage in AML patients.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.