AML_Driver_Neoantigens

Acute myeloid leukemia (AML) has lagged in benefiting from immunotherapies, primarily due to the scarcity of actionable AML-specific antigens. While driver mutations are considered optimal sources for immunogenic targets, a comprehensive characterization of the AML driver mutational landscape and their resultant candidate neoantigens is lacking. Herein, we extensively integrated somatic mutation data from over 2,800 AML cases, identifying 49 driver genes, notably characterized by a significant proportion of insertions and deletions (indels). Neoantigen prediction analysis revealed that indels triggered a higher abundance of neoantigens both in quantity and quality, compared with single nucleotide variants (SNVs) and gene fusions. We further integrated peptide features pertinent to neoantigen presentation and T cell recognition to establish two robust models of epitope immunogenicity that significantly enriched immunogenic neoepitopes derived from both gene mutations and fusions. The predicted number of neoantigens correlated with decreased survival in AML, partially attributed to the immunosuppressive microenvironment and exhausted CD8+ T cells, as supported by the deconvolution of large-scale bulk transcriptomes and the integration of single-cell RNA sequencing with multiparametric flow cytometry. Altogether, this study not only offers a valuable resource for adoptive cell-based therapy or personalized vaccine development but also underscores the necessity of incorporating strategies to revitalize the immunosuppressive milieu in conjunction with neoantigen-targeted therapies.