Immune‐related gene signature predicts clinical outcomes and immunotherapy response in acute myeloid leukemia. (30th March 2022)

• Record Type: Journal Article
• Authors: Xu, Qiang; Cao, Dedong; Fang, Bin; Yan, Siqi; Hu, Yu; Guo, Tao
• Is Part Of: Cancer medicine. Volume 11:Number 17(2022)
• Page Start: 3364
• Page End: 3380
• Publication Date: 2022-03-30
• Subjects: acute myeloid leukemia -- biomarker -- immune microenvironment -- immunotherapy -- prognosis; 616.994005
• DOI: 10.1002/cam4.4687
• Languages: English
• ISSNs: 2045-7634

Abstract

Abstract: Background: The immune response in the bone marrow microenvironment has implications for progression and prognosis in acute myeloid leukemia (AML). However, few immune‐related biomarkers for AML prognosis and immunotherapy response have been identified. We aimed to establish a predictive gene signature and to explore the determinants of prognosis in AML. Methods: Immune‐related genes with clinical significance were screened by a weighted gene co‐expression network analysis. Seven immune‐related genes were used to establish a gene signature by a multivariate Cox regression analysis. Based on the signature, low‐ and high‐risk groups were compared with respect to the immune microenvironment, immune checkpoints, pathway activities, and mutation frequencies. The tumor immune dysfunction and exclusion (TIDE) method was used to predict the response to immune checkpoint blockade (ICB) therapy. The Connectivity Map database was used to explore small‐molecule drugs expected to treat high‐risk populations. Results: A seven‐gene prognostic signature was used to classify patients into high‐ and low‐risk groups. Prognosis was poorer for patients in the former than in the latter. The high‐risk group displayed higher levels of immune checkpoint molecules (LAG3, PD‐1, CTLA4, PD‐L2, and PD‐L1), immune cell infiltration (dendritic cells, T helper 1, and gamma delta T), and somatic mutations ( NPM1 and RUNX1 ). Moreover, hematopoietic stem cell/leukemia stem cell pathways were enriched in the high‐risk phenotype. Compared with that in the low‐risk group, the lower TIDE score for the high‐risk group implied that this group is more likely to benefit from ICB therapy. Finally, some drugs (FLT3 inhibitors and BCL inhibitors) targeting the expression profiles associated with the high‐risk group were generated using Connectivity Map. Conclusion: The newly developed immune‐related gene signature is an effective biomarker for predicting prognosis in AML and provides a basis, from an immunological perspective, for the development of comprehensive therapeutic strategies. Abstract : The progression of bioinformatics tools has accelerated the search for biomarkers associated with acute myeloid leukemia (AML) prognosis. This article aims to originate an immune‐associated risk signature for prognostic analysis of AML and explore the mutation characteristics, immune characteristics, and immunotherapy response defined by risk signature.