Immune Landscape and Outcomes of Patients with RNA Splicing Factor-Mutant Acute Myeloid Leukemia and Myelodysplastic Syndromes Treated with Azacitidine +/- the Anti-PD-L1 Antibody Durvalumab

#: AMZ and BF contributed equally

Introduction: Mutations in RNA splicing factor (SF) genes such as SF3B1, SRSF2, U2AF1, and ZRSR2 are common in patients (pts) with myelodysplastic syndromes (MDS) and secondary acute myeloid leukemia (AML) and are associated with lower response rates to currently available therapies. Preclinical data suggest that splicing mutations can lead to the generation of neoantigens, which might render SF-mutant MDS and AML more susceptible to treatment with immune checkpoint inhibitors (Lu X et al. Cell 2021). However, the immunologic landscape of pts with SF-mutant MDS/AML is poorly characterized.

Methods: We performed a subgroup analysis comparing pts with SF-mutant and SF wild-type MDS and AML who were enrolled in the randomized, phase 2 FUSION trial (NCT02775903). Newly diagnosed pts were treated with azacitidine +/- the anti-PD-L1 antibody durvalumab, the addition of which did not improve overall survival (OS; Zeidan et al. Blood Advances 2022). A 38 gene-targeted myeloid mutation analysis from screening bone marrow (BM) was performed at Munich Leukemia Laboratory and pts were classified as being SF-mutant if they had a pathogenic variant in any of the following genes with a variant allele frequency ≥2%: SF3B1, SRSF2, U2AF1, or ZRSR2. Overall response rate (ORR) was defined as complete response (CR), marrow CR, partial response, or hematologic improvement based on modified International Working Group (IWG) 2006 response criteria for MDS pts and as CR or CR with incomplete blood recovery based on modified IWG 2003 response criteria for AML pts, respectively. OS was calculated as the time from randomization to any-cause death.

Gene expression profiles were analyzed by RNA-sequencing from BM aspirates from screening and cycle 3 day 22 (C3D22). Flow cytometry from BM aspirates at baseline were analyzed using a collection of 170 reportable subpopulations in the immune cell and immune checkpoint panel and 68 reportables in the tumor/blast panel.

Results: A total of 126 pts with AML (51 SF-mutant and 75 SF wild-type) and 79 pts with MDS (33 SF-mutant and 46 SF wild-type) were included. There were no differences in terms of baseline disease and demographic characteristics between SF-mutant and SF wild-type pts except for SF-mutant MDS pts being older (75.3 vs. 71.2 years; p=0.02) and a lower rate of adverse cytogenetics (13.7% vs. 33.3%; p=0.04) and IPSS cytogenetics (poor or very poor: 24.3% vs. 76.1%; p<0.001) among SF-mutant AML and MDS pts, respectively. ORR was independent of SF mutation status for both AML (SF-mutant: 35.3% vs. SF wild-type: 33.3%; p=0.47) and MDS pts (51.5% vs. 56.5%; p=0.63). OS was not statistically different between SF-mutant and SF wild-type AML (median 14.9 months [95% confidence interval [CI]: 11.3-20.5 months] vs. 12.2 months [95% CI: 9.7-15.9 months]; p=0.50) and MDS pts (median 23.5 months [95% CI: 12.3 months – not reached] vs. 10.6 months [95% CI: 9.7 months – not reached; p=0.16) and was independent of the addition of durvalumab and TP53 co-mutation status. There were no differences in key cell populations from BM flow cytometry samples including the percentage of CD3+ T-cells, PD-1+/TIM-3+ CD4 T-cells, PD-1+/TIM-3+ CD8 T-cells, PD-L1-positive progenitor cells, and myeloid progenitor cells by SF mutation status in neither the AML nor MDS subset. Finally, gene expression analyses showed an increase in MKI-67 expression in SF-mutant pts, but no differences were observed in the expression of several immune-related genes including IL7R and PD-L1.

Conclusion: The presence of RNA SF mutations did not impact ORR and OS among older or unfit, newly diagnosed AML and MDS pts treated with azacitidine +/- durvalumab. There were no differences in the immune landscape by SF mutation status as assessed by flow cytometry and gene expression analyses, which could be a potential explanation for the lack of ORR and OS benefit with anti-PD-L1 therapy in SF-mutant MDS and AML.