Abstract 754

Background:

The key hematopoietic transcription factor RUNX1 is inactivated by several mechanisms in patients with myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML), including translocations and mutations. Mutations have been reported in 10–20% of patients with AML, and appear to be associated with an inferior outcome. However, the analyzed patient populations were either small, contained highly selected patients or the spectrum of mutations investigated was limited. Thus, the information on the prevalence and the prognostic relevance of RUNX1 mutations is still incomplete.

Methods:

In order to study this abnormality in an unselected cohort including older patients and all karyotype abnormalities, we retrospectively analyzed DNA from 1649 newly diagnosed AML patients. The prognostic impact was studied in those 1303 patients treated within the AML96 protocol of the DSIL/SAL. Screening was performed for exons 3–8 using denaturing HPLC (dHPLC). All patients with aberrant dHPLC-profile were sequenced.

Results:

Overall 217 mutations were identified in 186/1649 patients (11.3%); 31 patients had two different mutations. The mutations affected the entire coding region, with a cluster in exons 3–5 coding for the RUNT-homology-Domain (RhD). The alterations consisted predominantly of point mutations, but small insertions or deletions were also found, especially in exons 7 and 8. In addition, several patients showed mutations affecting splice donor or acceptor sites. Correlation with clinical data revealed that patients with RUNX1 mutations had a higher median age than patients with wild-type RUNX1 (64 vs. 59 years; P<.001). No significant differences were found regarding other parameters, including the percentage of BM-blasts, gender, platelet or WBC counts. Mutations were observed in almost all FAB-subtypes besides M3 and M7, but as reported, they clustered in FAB M0 and M1. Consistent with the reported role of RUNX1 mutations in MDS, patients with a history of prior MDS had a rate of 20.1% compared to 9.4% in de novo AML and 12.2% in tAML (P<.001). A similar prevalence of mutations was found in patients with cytogenetic abnormalities (86/765; 11.2%) and patients with normal karyotype (84/766; 11%). A high prevalence of RUNX1-mutations was found in patients with trisomy 13, trisomy 21 and also in patients with t(9;22), whereas significantly fewer mutations were observed in good risk cytogenetics t(8;21) and inv(16). Patients with RUNX1-mutations displayed a similar prevalence of FLT3-ITD mutations but significantly less NPM1-mutations (2.2%; P<.001) and a trend for less CEBPA mutations (2.8%; P=.064). When treatment response was investigated, RUNX1-mutant patients had a significantly lower rate of complete remission after double induction, even if restricted to patients <60 yrs (67.5% vs. 49.2%; P=.006). RUNX1 mutations were also associated with inferior overall survival (OS) in all age groups analyzed, e.g. in patients age 60 and younger with normal karyotype, the median OS was 0.96 years in RUNX1 mutant compared to 1.78 years in wt patients (P=.009). In patients achieving CR, treatment failure was due to a significantly increased relapse rate.

Conclusions:

RUNX1 mutations can be found in a significant proportion of patients with AML. They are associated with prior MDS and older age and appear to characterize a patient group with an increased rate of treatment failure.

Disclosures:

Thiede:AgenDix GmbH: Equity Ownership.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution