Abstract 417

Mutations in the additional sex combs like-1 (ASXL1) gene have been identified in patients (pts) with myelodysplastic syndromes, myeloproliferative neoplasms, and acute myeloid leukemia (AML). We present here the first study on ASXL1 mutations in adult primary cytogenetically normal (CN-)AML, and report their associations with clinical and molecular characteristics, treatment outcomes, and gene- and microRNA- (miR-)expression profiles. We studied 423 primary CN-AML pts, aged 18–83 years (y) and treated on cytarabine/daunorubicin-based frontline protocols, for ASXL1 exon 12 mutations (frame shift and nonsense changes) and other prognostic gene mutations [FLT3-internal tandem duplications (ITD) and tyrosine kinase domain mutations, MLL partial tandem duplications, and mutations in NPM1, CEBPA, WT1, IDH1, IDH2 and TET2].

ASXL1mutations were 5 times more common in older (≥60y) than in younger (<60y) pts [38/234 (16.2%) vs 6/189 (3.2%); P<.001], and subsequent analyses therefore focused on older CN-AML pts. Compared to ASXL1-wild type (-wt) pts, ASXL1-mutated (-mut) pts very rarely carried NPM1 mutations (P<.001) or FLT3-ITD (P=.002), but more often had CEBPA mutations (P=.01). ASXL1-mut pts also had lower white blood counts (P=.02), lower blast percentages in blood (P<.001) and bone marrow (P=.04), and tended to have higher platelet counts (P=.06) and more frequently be male (P=.08) than ASXL1-wt pts.

Among older primary CN-AML pts, those with mutated ASXL1 had a lower complete remission (CR) rate (53% vs 71%; P=.04) and shorter disease-free survival (DFS; P=.03; 3y rates, 10% vs 19%), overall survival (OS; P=.006; 3y rates, 5% vs 23%) and event-free survival (EFS; P=.002; 3y rates, 5% vs 14%; Fig. A) than ASXL1-wt pts. Due to the strong associations of ASXL1 mutations with NPM1-wt, absent FLT3-ITD and mutated CEBPA, we studied their prognostic impact within the genetic categories defined in the European LeukemiaNet (ELN) guidelines [ELN Favorable (Fav): CN-AML with mutated CEBPA and/or mutated NPM1 without FLT3-ITD; ELN Intermediate-I: all remaining CN-AML pts]. ELN Fav/ASXL1-mut pts had a lower CR rate (50%) compared with ELN Fav/ASXL1-wt pts (82%; P=.04). All 6 ELN Fav/ASXL1-mut pts who achieved CR relapsed within 13 months, while 27% of ELN Fav/ASXL1-wt pts were alive and disease-free at 3y. All ELN Fav/ASXL1-mut pts died within 18 months after enrollment, whereas 34% of ELN Fav/ASXL1-wt pts were alive at 3y (OS, P<.001). EFS of ELN Fav/ASXL1-mut pts also was significantly worse than for ELN Fav/ASXL1-wt pts (P<.001; 3y rates, 0% vs 22%; Fig. B). Multivariable analyses confirmed that ASXL1 mutations associated with lower CR rates (P=.03), shorter DFS (P<.001), OS (P<.001) and EFS (P<.001) only among ELN Fav pts, after adjusting for other risk factors. In contrast, ASXL1 mutations were not associated with outcomes in the ELN Intermediate-I group. Further exploratory analyses in molecular subgroups suggested that ASXL1 mutations may be associated with particularly unfavorable outcomes [ie, shorter OS (P<.001) and EFS (P=.02)] among CEBPA-mut pts.

Gene- and miR-expression profiles were derived using Affymetrix HG-U133 plus 2.0 and custom-made miR microarrays. We identified an ASXL1 mutation-associated gene-expression signature comprising 67 differentially expressed genes (92 probe-sets), including upregulation of WNT pathway co-receptor LRP6, cytochrome P450 enzyme CYP1B1, and GJA1 (connexin 43, mediating stem cell-stroma interactions in the bone marrow). No significant signature of differentially expressed miRs was found.

In conclusion, in this first study of ASXL1 mutations focusing on primary CN-AML, we demonstrate that they associate with inferior outcomes in older pts, particularly within the ELN Fav genetic group. We also report the first ASXL1-mutation associated gene-expression signature in CN-AML that may provide useful insight into the biology of ASXL1-mut AML, and help design novel treatment approaches for this high-risk group of older pts.
Disclosures:

No relevant conflicts of interest to declare.

This icon denotes a clinically relevant abstract

Author notes

*

Asterisk with author names denotes non-ASH members.