Mutations in the AML1 Gene Define an Unfavorable Subgroup in Acute Myeloid Leukemia with FAB M0.

The AML1/RUNX1 gene is the most frequent target for chromosomal translocations in leukemia. Recently point mutations in the AML1 gene have been demonstrated as another mode of genetic aberration. AML1 mutations have been reported in de novo MDS and AML, as well as in therapy related MDS and AML. The AML M0 subtype has been found to be most frequently affected by sporadic AML1 gene mutations. We analysed AML1 gene mutations in a cohort of 49 M0 patients. Mutation screening was performed either with SSCP (n=21) and/or denaturating High Performance Liquid Chromatography (dHPLC) (n=33), 5 cases were analyzed by both methods. SSCP screening of exons 3–5 of the AML1 gene was carried out at the genomic level. These exons cover the socalled Runt domain, which is most frequently mutated. Fragments with aberrant mobility were sequenced. With this method 5 cases were found to be mutated. Subsequently, to improve the screening efficiency an assay using dHPLC was established. Hereby, we screened the cDNA of patient samples for mutations in amino acid codons 1–277 of the AML1b transcript, where the Runt domain is located between codons 49 and 178. All 5 cases detected by SSCP were confirmed by dHPLC. Nine mutations were detected in the cohort of 28 cases (32%) which had not been analyzed by SSCP. In total, 14 of the 49 samples (29%) tested were identified to be mutated, which is a slightly higher frequency than previously reported. In the cohort of 35 AML1 non-mutated cases 20 (57%) had a normal karyotype and 15 (43%) an aberrant karyotypes, whereas only 6 of the 14 AML1 mutated cases (43%) had a normal karyotype (p=0.001). Three of the AML1 mutated cases (21%) also had FLT3 mutations. One had an FLT3-LM, one an FLT3-TKD mutation, and one case both LM and TKD mutations. Clinical follow up data were available for 33 patients (22 AML1 non- mutated, 11 AML1 mutated). The median OS and EFS of the AML1 non-mutated versus the mutated group was 276 days versus 63 days (p = 0.0679) and 276 vs. 63 days (p=0.0630) respectively. Thus the AML1 mutated cases tend to have a worse clinical outcome. When other AML subtypes were screened for AML1 mutations, i.e. M1 (n=26), M2 (n=21) and M4 (n=3), only 1 additional AML1 mutation was detected, confirming the highest prevalence of AML1 mutations in M0. In conclusion, 1) we established a new assay to screen for AML1 mutations. 2) We confirmed the high incidence of AML1 gene mutations in AML M0, both in cases with normal and aberrant karyotype. 3) For the first time we demonstrated that AML1 mutations define an unfavorable subentity in AML M0.