Impact of NPM1 co-mutation on outcomes in RAS-mutated Acute Myeloid Leukemia

Abstract Introduction: Oncogenic Mutations in RAS, most commonly involving NRAS and KRAS, occur in 10–25% of cases of acute myeloid leukemia (AML) and have been implicated in the disease biology of monocytic subtypes of AML. RAS mutations in AML have been recently shown to impart an intermediate prognostic risk when treated with hypomethylating agents (HMA) + venetoclax and may benefit from high-dose cytarabine consolidation in pts treated with intensive chemotherapy (IC). Co-occurring mutations are frequently observed in RAS-mutated AML and may modify prognostic risk in the context of different therapies. The prognostic and therapeutic implications of concurrent RAS and NPM1 mutations and type of therapy remain poorly understood. Methods: We conducted a retrospective analysis of pts with newly diagnosed AML with RAS mutations to analyze outcomes by NPM1 co-mutations and type of therapy. Mutation status was determined by next-generation sequencing, and pts were classified into two groups: (1) RAS and NPM1 co-mutated, and (2) RAS-mutated with NPM1 wild type. Overall survival (OS) was defined from diagnosis to death or last follow-up, and event-free survival (EFS) from treatment start to remission failure, relapse, or death. Results: A total of 418 pts with RAS-mutated AML were included. Of these, 79 (19%) had co-occurring NPM1 mutations (RAS-mut/NPM1-mut) and 339 (81%) were NPM1 wild type (RAS-mut/NPM1-wt). The median age was 68 years (range, 22–94) and was similar across subgroups. Females made up 39% of the pts (n=162) overall, 53% (n=42) in RAS-mut/NPM1-mut and 35% (n=120) in RAS-mut/NPM1-wt (p=0.004). 73% of pts were aged 60 yrs and older (n=304), 70% of the RAS-mut/NPM1-mut (n=55) cohort, and 73% of the RAS-mut/NPM1-wt (n=249) cohort. Median follow-up was 54.7 months (range, 49.3–61.4) overall. RAS isoforms included NRAS (n=268, 64%), KRAS (n=102, 24%), HRAS (n=3, 1%), and dual N/KRAS (n=45, 11%). Involved codons were G12 (47%), G13 (23%), Q61 (15%), and others (15%). Cytogenetics showed a diploid karyotype in 35% (n=147). Other abnormalities included +8 (10%, n=40), KMT2A/MLLr (7%, n=30), −5q/−7q (16%, n=66), MECOMr (2%, n=10), complex karyotype (5%, n=21), and others (18%, n=77). Cytogenetic testing was not done in 7% (n=27). Median OS and EFS for all pts with RAS-mut were 11.6 months (range 9.4–13.6) and 8.3 months (range 7.1–9.9), respectively. The median OS was significantly longer in pts with RAS-mut/NPM1-mut AML (22 months) compared to those with RAS-mut/NPM1-wt AML (9.7 months, p<0.0001). Similarly, the median EFS was 21.5 vs. 7.2 months (p<0.0001) for pts with RAS-mut/NPM1-mut and RAS-mut/NPM1-wt, respectively. Among pts with only diploid cytogenetics (n = 147), median OS was still significantly better for RAS-mut/NPM1-mut compared to RAS-mut/NPM1-wt (46.7 vs. 15.9 months; p = 0.02). Comparison within different treatment groups shows that pts treated with low-intensity chemotherapy (Low-IC) [Cladribine+LDAC or HMA] without Venetoclax (Ven) had a median OS of 17.9 months (CI 3.8–46.6) vs. 7.2 months (CI 4.8–10.02; p = 0.04) for RAS-mut/NPM1-mut vs. RAS-mut/NPM1-wt, respectively. For patients with RAS-mut/NPM1-mut vs. RAS-mut/NPM1-wt AML, treatment with Clad+LDAC+Ven produced a median OS that was not reached (NR) vs. 7.3 months (CI 4.1–17.8; p = 0.02) while pts treated with HMA+Ven had a median OS of 15.4 (CI 7.1–21.5) vs. 7.9 months (CI 5.6–10.5; p = 0.1), respectively. For intensive chemotherapy (IC) alone, median OS was NR for RAS-mut/NPM1-mut vs. 8.9 months (CI 5.9–39.6; p = 0.04) for RAS-mut/NPM1-wt. Addition of Ven to IC appeared to improve OS among pts in both cohorts (median OS: NR for both), and there was no significant difference in OS between RAS-mut/NPM1-mut and RAS-mut/NPM1-wt (2-yr OS of 76% vs. 56%; p=0.3). Conclusions: The prognostic impact of RAS mutations in AML is modified in the setting of NPM1 co-mutations. Pts with RAS-mutated AML without co-occurring NPM1 mutations had significantly inferior OS and EFS than those with RAS-mut/NPM1-mut. Low-IC and IC alone showed a clear survival benefit for RAS-mut/NPM1-mut compared to RAS-mut/NPM1-wt. Addition of Ven appeared to improve outcomes in all subgroups and across therapies, except pts treated with HMA+Ven. These findings highlight the favorable prognostic impact of NPM1 co-mutation within RAS-mutated AML and underscore the importance of optimizing treatment intensity, including the use of Ven, to improve survival in this high-risk population.