Frequent Mutation of the Polycomb-Associated Gene ASXL1 In Acute Myeloid Leukemia Secondary to Myelodysplastic Syndrome or Chronic Myelomonocytic Leukemia

Abstract 2940

Recent studies have identified mutations of the ASXL1 gene in MDS and CMML. Mammalian ASXL proteins are believed to play a role in chromatin remodelling. We have previously reported that ASXL1 mutations are frequent in patients with MDS, CMML and AML. We observed that the mutation rate was relatively high in late MDS and AML with a lower frequency across early MDS. We found that the mutation is common in the normal karyotype group (especially AML secondary to MDS), occurring in 40% of all cases. ASXL1 mutations (including the common mutation c.1934dupG;p.Gly646TrpfsX12) were present in myeloid cells, but not in T-cells, indicating that they were acquired in all seven cases examined. The majority of the mutations identified were heterozygous frameshift mutations caused by deletion or duplication of a nucleotide. Given the high frequency of ASXL1 mutations in advanced MDS and AML in our earlier study, we have now screened a larger group of 143 AML samples, comprising 111 de novo AML and 32 AML secondary to either MDS or CMML, for mutations in the ASXL1 gene. In primary AML samples we found only 6/111 patients with ASXL1 mutations (5%), whereas the frequency of mutations was significantly higher in AML secondary to MDS (10/25, 40%) or CMML (5/7, 71%), strongly suggesting an association with disease progression in MDS and CMML. In order to determine the frequency of ASXL1 mutations in de novo and secondary AML patients with a normal karyotype and to identify cooperating mutations, we screened 85 samples for ASXL1 (all coding exons), NPM1 (exon 12), FLT3 (ITD and D835Y), TET2 (all coding exons), IDH1 (R132), IDH2 (R140 and R172) and RUNX1 (exons 3–7) mutations (Table 1). FLT3 and NPM1 mutations were more common in de novo AML (FLT3 51% and NPM1 58%) than in sAML (FLT3 9% and NPM1 12%), consistent with the role of NPM1 mutations as a hallmark of cytogenetically normal de novo AML. TET2 was mutated in 13% of primary AML, and in 28% of sAML, with most TET2 mutations found in AML secondary to CMML (5/7, 71%), as expected according to previous reports on higher mutation frequency in CMML patients. Mutations in IDH1 and in IDH2 were evenly distributed in primary and secondary cases. Similarly, RUNX1 mutation rate showed no significant differences between primary and secondary AML cases. Interestingly, ASXL1 mutations were mutually exclusive with NPM1 mutations, suggesting that they could be markers of different subgroups with a distinct aetiology. In conclusion, we have shown that mutation of ASXL1 is a very frequent event in AML secondary to MDS and CMML, but is much less frequent in de novo AML. Our data support a role for ASXL1 mutations in disease progression in MDS and CMML.