Abstract 1605

Poster Board I-631

CCAAT/enhancer binding protein alpha (C/EBPá) function is frequently disrupted in acute myeloid leukemia (AML). This can be caused by different mechanisms, including mutations in CEBPA, the gene encoding for C/EBPá. Recently, promoter hypermethylation resulting in CEBPA silencing has been described. CEBPA mutations are associated with a favorable outcome in adult AML. Recent studies however suggested that the favorable outcome is uniquely associated with CEBPA double-mutated AML (CEBPAdoublemut) (presence of >1 CEBPA mutation), and not with CEBPA single-mutated AML (CEBPAsinglemut). In pediatric AML, data on outcome of CEBPA-mutated AML is limited to one study, showing favorable outcome for CEBPAdoublemut as well as CEBPAsinglemut, which is in contrast with the adult data. So far, data on CEBPA hypermethylation in pediatric AML is lacking.

We therefore studied a large pediatric AML cohort (n=252) to characterize CEBPA mutations by sequencing the entire coding region, and CEBPA promoter hypermethylation by methylation-specific PCR (MSP). Survival analyses were performed in 185 patients with de novo AML (excluding t(15;17) and secondary AML) treated on uniform DCOG and BFM protocols. Furthermore, we generated gene expression profiles using the Affymetrix HGU133 plus 2.0 microarrays to compare subgroups with CEBPA aberrations.

Thirty four CEBPA mutations were identified in 20/252 diagnostic samples (7.9%). In 14 cases double mutations were present, which combined an N-terminal frame shift mutation with an in-frame mutation in the bZIP region (n=13) or with a frame shift-causing insertion before bZIP (n=1). In 6 cases a single mutation was present; i.e. in-frame bZIP mutation (n=4), or frame shift mutation respectively in the TAD2 domain (n=1) or before the bZIP domain (n=1). CEBPAdoublemut were only present in children above 3 years of age and in FAB M1/2 subtypes, in contrast to CEBPAsinglemut, which presented also in children <3 years of age (1/6) and in other FAB subtypes (3/6). CEBPAdoublemut and CEBPAsinglemut were both exclusively found in cases with a normal karyotype (57% and 33%, respectively) and in cases with ‘other’ karyotypes (defined as 'other than t(8;21), inv(16), t(15;17) and MLL-rearrangements'), in 36% and 50%, respectively. However, in both subgroups additional molecular aberrations, i.e. in RAS, FLT3/ITD and WT1, were equally distributed. CEBPAdoublemut patients (n=10) had a significantly better overall survival compared with CEBPAsinglemut (n=5) (5-years pOS 79±13% vs. 25±22%, p=0.04; pEFs 58±16% vs. 30±24%, p=0.16). Furthermore, they showed a trend for favorable outcome compared with CEBPA wild-type AML patients, after excluding CBF-AML cases (n=120; pOS 79±13% vs. 47±5%, p=0.07; pEFS 58±16% vs. 34±4%; p=0.06). Their survival was comparable to the CBF-AML subgroup (n=50) (pOS 91±4%, pEFS 61±8%). Multivariate analysis, including age, WBC, CBF-AML, NPM1 mutations and FLT3/ITD, showed that CEBPAdoublemut were an independent favorable prognostic factor for pOS (HR 0.23; p=0.04) and pEFS (HR 0.32; p=0.03). CEBPA promoter hypermethylation was detected in 3/237 cases, which resulted in CEBPA silencing. Using an unsupervised clustering analysis as previously published by Valk et al. (NEJM 2004) of our de novo AML cases (n=237), CEBPA-mutated cases predominantly aggregated in 1 cluster with the CEBPAsil cases, revealing a common underlying gene expression profile. Two additional cases with silenced CEBPA (CEBPAsil) were identified in this cluster, resulting in 5/237 (2.1%) CEBPAsil cases. All CEBPAsil showed T-lymphoid characteristics, (e.g. high expression of CD7 and high LCK expression). However, NOTCH1 mutations were not found. Three of 5 patients relapsed within 1 year of diagnosis, but the other 2 are in continuous complete remission for 4.8 and 8.5 yrs.

In conclusion, CEBPAdoublemut were identified as an independent predictor of good clinical outcome. Hence, if these results could be confirmed in a prospective serie, CEBPAdoublemut may be used for further refinement of risk-group stratification. Of interest, the other cases with CEBPA aberrations, i.e. CEBPAsinglemut and CEBPAsil, seem to predict for poor outcome, in line with data presented in adults. The subgroup with CEBPAsil due to hypermethylation may potentially benefit from the use of demethylating agents.

Disclosures

No relevant conflicts of interest to declare.

This icon denotes an abstract that is clinically relevant.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution