Abstract
Introduction: CEBPA mutations (mut) occur in 5-14% of patients with acute myeloid leukemia (AML). A difference in prognosis between CEBPA single- (sm) and CEBPA double-mutated (dm) cases has been reported with longer overall and event-free survival in dm cases. This may be due to differences in the amount and composition of concomitant mutations.
Aims: 1) Elucidate the cause of the clinical difference between CEBPAsm and dm cases. 2) Evaluate the clonal architecture of CEBPAmut AML in a well characterized cohort (Fasan et al., Leukemia 2014).
Patients: We investigated concomitant mutations in a cohort of 229 CEBPAmut AML cases (129 sm and 100 dm). The cohort was composed of 113 females and 116 males, median age was 67.1 y (range: 15.7-87.6 y). 218 cases showed intermediate risk (n=162 normal and n=56 aberrant karyotype), 11 cases adverse risk cytogenetics.
Results: CEBPAdm (median age 56.8 y, range 15.7–87.6 y) were significantly younger compared to sm (median age 64.7 y, range 20.4–87.0 y, p=0.001). Furthermore, CEBPAdm occurred significantly more often in females than in males (61.0% vs 39.0%; p=0.002). With regard to cytomorphology and cytogenetics, no significant differences between CEBPAsm and dm cases were seen.
Concomitant mutations were present in significantly more CEBPAsm cases compared to CEBPAdm cases (93.0% vs 79.0%; p=0.003). Furthermore, the amount of additional mutations was higher in CEBPAsm (mean: 2.1 mutations, range: 1-6) compared to CEBPAdm (1.3 mutations, range: 1-4). ASXL1 (p=0.009), DNMT3A (p<0.001), FLT3-ITD (p=0.007), IDH2 (p=0.042), NPM1 (p<0.001) and RUNX1 (p=0.004) mutations were significantly more frequent in CEBPAsm compared to dm. In contrast, GATA2 (p<0.001) and WT1 (p=0.039) were more frequently mutated in CEBPAdm. No respective differences were detected in the frequencies of MLL-PTD and of mutations in IDH1, KRAS, NRAS and TET2. We additionally analyzed concomitant mutations according to functional pathways. No differences between CEBPAsm and dm were detected in myeloid transcription factors (GATA2, RUNX1) and oncogenes (KRAS, NRAS). Significantly higher frequencies were found in CEBPAsm compared to dm for the following pathways: mutations affecting DNA-methylation (DNMT3A, IDH1/2, TET2; 64.3% vs 43.0%; p=0.001), nucleophosmin (30.2% vs 2.0%; p<0.001), signaling (FLT3-ITD, FLT3-TKD; 21.7% vs 8.0%; p=0.006) and chromatin modifiers (ASXL1, MLL-PTD; 27.1% vs 13.0%; p=0.009). In contrast, the tumor suppressor WT1 was significantly less frequently mutated in CEBPAsm (3.9% vs 12.0%; p=0.023). Regarding outcome, CEBPAdm cases with DNA methylation affecting mutations as compared to those without had significantly shorter EFS (median 11.3 vs 45.9 months in all others; p=0.052) and OS (median 29.3 months vs n.r.; p=0.009). Furthermore, mutations in chromatin modifiers negatively influenced EFS (median 5.4 vs 26.4 months; p=0.002) and OS (median 15.9 months vs n.r.; p=0.006) in CEBPAdm cases.
We further compared clonal architecture between CEBPAsm and dm cases. As a first step we compared mean mutation loads of concomitant mutations in CEBPAsm and dm and observed no difference between the two entities. As a next step, cases with uniform mutation loads in all genes were considered monoclonal, while cases with heterogeneous mutation loads (range of loads for different mutations spanning >30%) were considered clonally diverse. CEBPAsm were significantly more often clonally diverse compared to dm (52.7% vs. 32.0%; p=0.002). In 63.3% of diverse CEBPAdm cases (n=19/30), the dominant clone comprised CEBPAmut compared to only 32.3% (n=22/68) of diverse CEBPAsm (p=0.009). In 20.0% (n=6/30) of CEBPAdm with subclonal CEBPAmut, the dominant clone was TET2mut and in 10.0% (n=3/30) WT1mut. In 2 cases (7.6%), the dominant clone had mutations in different genes. In 26.5% (n=18/68) of CEBPAsm with subclonal CEBPAmut, the dominant clone carried TET2mut, 8.8% (n=6/68) DNMT3Amut, 4.4% (n=3/68) ASXL1mut. In 27.9% (n=19/68) the dominant clone had mutations in different genes.
Conclusions: 1. CEBPAsm occur preferentially in male patients and are correlated with older age. 2. Biological diversity of CEBPAmut cases is reflected by the amount and composition of additional mutations. 3. Prognosis in CEBPAdm cases is adversely influenced by mutations affecting DNA-methylation and chromatin modifiers. 4. Clonal architecture is more complex in CEBPAsm compared to dm cases.
Fasan:MLL Munich Leukemia Laboratory: Employment. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Perglerová:MLL2 s.r.o.: Employment. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Schnittger:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.
Author notes
Asterisk with author names denotes non-ASH members.