High Frequency of AML1/RUNX1 Mutations in Specific Cytogenetic Subgroups in De Novo Acute Myeloid Leukemia.

Somatic mutations in the DNA-binding domain, the socalled Runt homology domain, of the AML1/RUNX1 gene have been identified to occur in acute myeloid leukaemia (AML) with the highest incidence in AML M0, in therapy-related myelodysplastic syndrome (t-MDS), in therapy-related AML (t-AML) and AML after MDS (s-AML). Cytogenetic aberrations that are associated with RUNX1 mutations (RUNX1mut) have been reported to be trisomy 13 in AML and trisomy 21 in myeloid malignancies, but also loss of chromosome 7q, mainly in t-MDS but rarely in t-AML. So far the majority of RUNX1mut have been described in secondary or therapy-related cases. Thus, we characterized a cohort of 119 patients (pts) with de novo AML and compared these results to 19 MDS and s-AML, 2 t-MDS (n=2) and 8 t-AML. The cohort was selected for specific cytogenetics with high reported frequencies of RUNX1mut: trisomy 13 (n=17), trisomy 21 (n=9), −7/7q- (n=34). In addition pts with normal karyotype (NK) (n=42), inv(3)/t(3;3) (n=12), trisomy 8 (n=11), complex karyotype (n=13) and 10 pts with various other cytogenetic aberrations (other) were analyzed. The incidence of RUNX1mut in the different cytogenetic subgroups was: 94% (16/17) in +13, 56% (5/9) in +21, 29% (10/34) in −7/7q-, 10% (4/42) in NK, 17% (2/12) in inv(3)/t(3;3), 18% (2/11) in +8, 0% (0/13) in complex karyotype and 20% (2/10) in other, respectively. Based on clinical history we observed RUNX1 mutations in: 6/19 (32%) in MDS/s-AML, 1/10 (10%) in t-MDS/t-AML and 34/119 (29%) in de novo AML. Of the 6 RUNXmut cases with MDS/s-AML the karyotypes were heterogeneous NK (n=1), −7 (n=2) +13 (n=1), +21 (n=1), and inv(3) (n=1). The only recurrent cytogenetic aberration in MDS/s-AML was −7, thus the frequency of RUNXmut in the MDS/s-AML group with −7 was 2/8 (25%). Also the only RUNX1mut case with t-AML revealed a −7. These data correspond to those reported in the literature. We further focussed on the analyses of RUNX1 in de novo AML which is rarely reported so far. In the de novo AML group only we detected RUNX1mut with the highest frequency in +13 (16/16; 100%) followed by +21 (4/8; 50%) −7 (7/21; 33%), + 8 (2/10, 20%), inv(3) (1/8; 12.5%), and NK (3/33; 9.1%). In addition, in the group with “other” aberration 2/8 were mutated. Interestingly, these 2 mutated cases displayed a high number of trisomies including +8 and +13. No RUNX1mut were detected in AML with complex karyotype (n=10). These data for the first time show that RUNX1mut are not strongly correlated to MDS, s-AML or t-AML. With almost the same frequency they can be observed in de novo AML if specific cytogenetic groups are considered. Thus the RUNXmut seem to be more related to these cytogenetic subgroups than to the MDS, s-AML or t-AML.