Combined immune and DDR pathway classifier: A novel pathway-based classification aimed at tailoring personalized therapies for acute myeloid leukemia patients

Abstract

Acute myeloid leukemia (AML) constitutes a group of lethal hematological malignancies with high heterogeneity, resulting in widely variable outcomes of targeted therapy and immunotherapy. A better basic understanding of the molecular pathways of AML would help greatly in tailoring treatments to patients. Here, we propose a novel subtyping protocol for AML combination therapy. Three datasets, namely, the TCGA-LAML, BeatAML and Leucegene datasets, were used in this study. Single-sample GSEA (ssGSEA) was performed to calculate the expression scores of 15 pathways, including immune-related, stromal-related, DNA damage repair (DDR)-related and oncogenic pathways. The consensus clustering was used to classify AML based on pathway score data. We identified four phenotypic clusters—IM+DDR−, IM−DDR-, IM−DDR+ and IM+DDR+—representing distinct pathway expression profiles. The IM+DDR− subtype exhibited the most robust immune function, and patients of IM+DDR− subtype were likely to derive the greatest benefit from immunotherapy. Patients in IM+DDR+ subtype had the second highest immune scores and the highest DDR scores, suggesting that combination therapy (immune + DDR-targeted therapy) is the optimal treatment. For patients of IM−DDR- subtype, we recommend the combination of venetoclax and PHA-665752. A-674563 and dovitinib could be combined with DDR inhibitors to treat patients in IM−DDR+ subtype. Moreover, single-cell analysis revealed that there are more immune cells clustered in the IM+DDR− subtype and higher number of monocyte-like cells, which exert immunosuppressive effects, in the IM+DDR+ subtype. These findings can be applied for molecular stratification of patients and might contribute to the development of personalized targeted therapies for AML.