Impact and Outcomes of RAS gene Mutations in Core Binding Factor Acute Myeloid Leukemia

Acute myeloid leukemia (AML) with t(8;21) or inv(16) chromosomal rearrangements are distinct heterogeneous disease entities within AML that are classified together as core binding factor AML (CBF-AML). Given the nature of the chromosomes involved, these rearrangements lead to the production of leukemogenic chimeric transcripts (RUNX1-RUNX1T1 and CBFB-MYH11) which disrupt the physiologic activity of the heterodimeric transcription factor CBF complex. Although CBF-AML patients have a favorable prognosis and good response to treatment compared to other AML subtypes, survival outcomes are not uniform. Indeed, 30-50% of patients with CBF-AML eventually relapse, and the 5-10 yr survival ranges between 55-61% for patients < 60 yr in MRC/NCRI AML trials. Studies have analyzed the clonal architecture of CBF-AML patients and identified cooperating mutations independently of receptor tyrosine kinases (FLT3, KIT) mutations while others have found a 30% occurrence of KIT mutations. Studies of murine models of Runx1 and Cbfb have demonstrated that inactivation of both genes does not lead to leukemia, suggesting that other factors are necessary to recapitulate the leukemia phenotype fully. Although mutations in RAS family of genes (NRAS/KRAS) are among the most frequently observed mutations described in CBF-AML [54% in inv(16) and 30% in t(8;21)], no associations between those mutations and survival outcomes have been found. Because of the lack of association between RAS^MT and clinical outcomes, their role in CBF-AML is still unknown.

Here, we focused on dissecting the impact of RAS mutations (NRAS/KRAS; RAS^MT) on the clinical characteristics, survival outcomes, and the molecular associations among CBF-AML patients by evaluating the clonal succession of RAS^MT. In total, 284 CBF-AML patients were identified, in whom inv(16) and t(8;21) represent 61% (n=173) and 39% (n=111) of the cases, respectively. Thirty-five % (99/284) of the patients carried RAS mutations (NRAS=78; KRAS=21) with 8 patients harboring 2 mutations comprising NRAS, KRAS, or both NRAS/KRAS genes. RAS mutations were point mutations affecting known hotspots in NRAS and KRAS. Both RAS^MT and RAS^WT had a median age < 60 years (55 (14-83) vs. 49 (7-83) years, P=0.9) and sex was equally distributed among the two groups. Leukopenia, defined as white blood cell count <3.5 x 10⁹/L, was the only significant different hematological parameter between RAS^MT and RAS^WT. Fewer RAS^MT patients had leukopenia (3 vs. 12%; P=0.02). Compared to RAS^WT patients, RAS^MT patients had more often +8 (14 vs. 6%; P=0.02) and less -Y (7 vs. 16%; P=0.04) abnormalities. The analysis of 30 genes frequently mutated in myeloid malignancies revealed that RAS^MT patients had a significantly lower number of additional mutations compared to RAS^WT patients (n=59 vs. 197; P=0.03). Focusing on molecular associations, there were no differences in the frequency of FLT3 (TKD or ITD) mutations between RAS^MT and RAS^WT (15 vs. 18%). Although RAS^WT were more enriched with ASXL1, DNMT3A, IDH1/2, TET2, and WT1 mutations, no significant difference in frequencies was found when compared to RAS^MT. KIT mutations have been associated with poor outcomes in CBF-AML. In our cohort, KIT mutations (small deletions, points mutations), were significantly less associated with RAS^MT than RAS^WT (8 vs. 59%; P<.0001). RAS^MT were ancestral events in 58% of the cases (with FLT3 mutations the most common secondary lesions), secondary in 38% (with TET2,ASXL1, RUNX1 and IDH2 mutations represented the most common dominant hits), and codominant in 11% (with KIT mutations). RAS^MT in CBF-AML was associated with worse OS (HR: 1.520; P=0.04) compared to RAS^WT. Clonal hierarchy and succession analysis showedno change in OS when RAS^MT were secondary vs. dominant hits in CBF-AML (HR: 0.8; P=0.5).

In sum, our study of the frequency, clonal architecture and impact on survival of RAS^MT in CBF-AML patients points out the unique characteristics of this specific disease subgroup in which sole RAS^MT might represent, in addition to KIT^MT, a mutational lesion cooperating with inv(16) and t(8;21) rearrangements in driving leukemic evolution.