Domain-based analysis of RUNX1somatic mutations in myeloid neoplasms.

Abstract

6542 Background: RUNX1is a master regulator of hematopoiesis. ELN2022 risk stratification categorizes RUNX1 as an AML/MDS gene associated with poor prognosis. Functional domains in RUNX1are the Runt homology domain (RHD) and a transactivation domain (TAD), with DNA binding and co-activator recruitment functions, respectively. Domain-based analysis can elucidate the functional and biological impact of somatic RUNX1mutations, which have not yet been reported in the literature. Methods: Sequencing data from Karmanos Cancer Institute and a publicly available meta-analytic cohort [Awada et al., Blood 2021, Kewan et al., Nature Communications 2023, cBioPortal (Cerami et al., 2012) and AACR GENIE (v 13.1)] were used to construct a larger dataset of 16,565 patients with myeloid neoplasms. Baseline clinical and molecular characteristics were noted. The chi-square test was used to study various parameters described, and Kaplan-Meier curves were used to estimate survival. Results: RUNX1was mutated in 10.8% (1804/16,565) of patients. Median OS (mOS) for RUNX1MT vs. RUNX1WT was significantly worse at 16.4 vs 23.4 mos. (p < 0.0001). 990 patients with somatic mutations were studied; germline variants were screened and excluded if VAF was between 40-60%. Among somatic RUNX1MT patients, primary AML was the most frequent at 52.8%, followed by secondary AML-35.7% and MDS-9%. Most patients had mutations in the RHD (53.8%), followed by TAD (22%), and the remaining in non-RHD, non-TAD zones (NRNT). There was no statistically significant difference in OS by domain stratification; TAD had the numerically least mOS at 13.6 mos, RHD at 16.1 mos, and NRNT at 20.4 mos. The median age of patients was 70.9 (62-76) years, similar across domains. TAD patients were less likely to be male when compared to RHD (47% vs. 56%, p=0.04). No specific enrichment of disease type or abnormal karyotype was observed across domains. Co-mutational profiles across RHD and TAD domains among all diseases showed spliceosome co-signature enriched in RHD vs. TAD (59% vs. 46%, p=0.01). NPM1was noted to be inversely correlated with RUNX1 in general, with 1% and 3.5% in RHD and TAD. ASXL1was the most frequently distributed co-mutation in both groups, but not statistically different (41% vs. 33%, p=0.11). TAD patients with spliceosome co-mutation had worse mOS (11.4 mos) than those without spliceosome mutation (16.2 mos, p=0.02) or RHD patients with spliceosome co-mutation (16 mos, p=0.001). Missense mutations were strongly associated with RHD (55.3% vs. 7.9%); p <0.00001, while frameshift mutations were enriched in TAD (24.1% vs. 69.8%); p<0.0001. Conclusions: Domain-based analysis of somatic RUNX1MT myeloid neoplasms reveals TAD mutation to be common in females, frequently a frameshift mutation, and numerically has the worst survival. There appears to be a unique spliceosome signature enrichment in RHD compared to TAD, with worse OS in TAD patients with spliceosome co-mutations overall.