Splicing Gene Mutations Do Not Affect Outcomes in Newly Diagnosed Patients with Acute Myeloid Leukemia

Abstract

Background: Recent studies have shown that splicing factor (SF) gene mutations in AML implicate a negative prognostic significance associated with secondary AML and are now included as adverse in the 2022 ELN guidelines (Dohner et.al., Blood 2022). However, the contribution of these mutations to the prognosis of patients (pts) with AML, relative to their age, prior history of myeloid disorders, cytogenetics (CTG), concurrent mutations and frontline (FL) therapy requires further characterization. Methods: We retrospectively analyzed all pts with newly diagnosed (ND) AML treated at our center over the last 5 years (yrs) and compared outcomes for patients with and without SF mutations (Smut= SF3B1, SRSF2, U2AF1 and ZRSR2). FL therapy was classified as intensive (INT) and low intensity (L-I), with/without venetoclax (VEN). Relapse free survival (RFS) was calculated from the time of response to relapse or death and overall survival (OS) was calculated from the time of therapy initiation to death; both were censored at last follow up (F/U) Results: From January 2017 to April 2022, 994 ND pts with AML were reviewed; 266 (27%) had Smut in ≥ 1 of the SF genes while the rest were wild type (SWT). The most common Smut was SRSF2 (53%). Only 3% pts had >1 Smut (Table 1). The median age of Smut group was 72 yrs compared to 65 yrs in the SWT group, and significantly more pts in the former had secondary/therapy related AML (s/t AML) (51% vs 41%). The Smut group had less complex karyotype (CK), non CK adverse risk CTG and did not include any pts with core-binding factor (CBF) leukemias. Fewer pts in the Smut had a pathogenic TP53mut but had higher frequencies of RUNX1 and ASXL1 mutations. Overall, 71% Smut pts were classified as ELN 2017 adverse risk compared to 54% SWT pts (p<0.0001). 20% pts in the Smut group vs. 44% SWT pts received INT therapy (p<0.0001). Of the Smut pts who received L-I therapy, 62% received a hypomethylating agent (HMA) and 66% received a VEN- based combination, frequencies similar to the SWT group. At a median F/U of 26 months (mos), the median RFS and OS was 9.6 vs. 12.4 and 13.1 vs 13.3 mos for the Smut and SWT groups respectively. Given the significant age heterogeneity between the 2 groups we analyzed by young (<60 yrs) and older (≥60 yrs) pts with non-CBF AML and compared survival outcomes (Fig.1). Both Smut and SWT pts had similar RFS and OS in their respective age groups. In a small subgroup of older pts with de novo AML with CK treated with L-I regimens, the RFS and OS were especially dismal for those with Smut (RFS= 1.5 vs 5.7 mos, p=0.14 and OS= 2.6 vs 9.3 mos, p=0.003). For other subgroups (stratified based on ELN risk stratification, de novo or s/t AML, CK, TP53mut) outcomes of Smut and SWT were similar in both young and older pt groups treated with INT or L-I therapies. Smut pts who received VEN in the L-I arm had better RFS and OS, especially if they had denovo AML (RFS= 19 vs 5.5 mos and OS 25 vs 4 mos). On univariate (UV) analysis using age, CTG [adverse risk (AR) /intermediate risk (IR)], therapy arm (INT/L-I), Smut/SWT status and concurrent mutations (TP53, ASXL1, RAS, RUNX1, NPM1, FLT3-ITD) as covariates for overall response (OR= CR/CRi/MLFS): lower age, IR-CTG, TP53wt, RASwt ,FLT3mut, NPM1mut and use of INT therapy were favorable while all the factors other than FLT3mut , age and therapy arm remained significant on the multivariate (MV) model. Within the Smut cohort, the presence of SF3B1mut alone negatively affected the odds of OR (0.65, 95%CI=0.3-0.98) on UV but not on MV analysis. Cox regression analysis demonstrated adverse risk CTG, TP53mut and s/t AML had higher hazards of relapse, while the above factors along with the presence of RASmut increased the hazards of death. On stratifying by specific Smut, no individual mutation impacted OS. Cox regression analysis of the above covariates affecting RFS and OS in Smut pts who received L-I therapy showed VEN based combinations significantly reduced the hazards for relapse (0.5, 95% CI=0.3-0.8) and death (0.5, 95%CI=0.3-0.7). Overall, the presence of Smut did not affect response rates to therapy, RFS or OS in both young and older pts treated with L-I (majority VEN based) or intensive therapy. Conclusion: Smut are more common in elderly pts with AML and in s/t AML. Smut does not necessarily affect response rates and survival outcomes compared to SWT pts when controlled for these covariates, and with current therapeutic options including HMA + VEN based regimens. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal