17th International Conference on Malignant Lymphoma, Palazzo dei Congressi, Lugano, Switzerland, 13 - 17 June, 2023.

Abstract

Background: Previous researches have mainly focused on the discussion of whether cancer stem cells exist in diffuse large B-cell lymphoma (DLBCL). However, some subgroups with dismal prognosis that exhibit stem cell-like characteristics have been overlooked. Methods: Using an innovative one-class logistic regression (OCLR) machine learning algorithm, we calculated stemness indices and systematically assessed their ability to reflect the oncogenic dedifferentiation-like characteristics of DLBCL. Next, we identified signatures associated with oncogenic dedifferentiation-like features and prognosis using LASSO and SVM-RFE algorithms and developed a proportional hazards model (Riskscore) for DLBCL through Cox regression. Then we stratified the risk of DLBCL (n = 2133) and evaluated the prognostic value of the risk model across known clinical and molecular subgroups. We further compared the characteristics of high- and low-risk DLBCL with Burkitt lymphoma. Finally, we investigated the mechanisms of poor prognosis in high-risk DLBCL using transcriptomics, genomics (n = 576) and single-cell RNA sequencing (n = 19) data, as well as cell experiments and internal validation cohorts. Results: In this study, we identified and validated a DLBCL subgroup (25.6% of DLBCL) with stem cell-like characteristics and dismal prognosis. This high-risk group was defined as polyamine metabolism-cold immune tumor with upregulated polyamine metabolic key enzyme (ODC1) and desert-like immune infiltration, and had a poor prognosis with lower 3-year OS rate (54.3% vs. 83.6%, p < 0.0001) and PFS rate (42.8% vs. 75.2%, p < 0.0001) compared to the low-risk group. We found that some patients with MYC rearrangement, double-hit, double-expresser, or complete remission may have either favorable or poor prognosis, which can be accurately identified by our risk stratification model. Additionally, the high-risk group exhibited malignant proliferative phenotypes similar to Burkitt lymphoma. Genomic analysis revealed widespread copy number losses in the chemokine and interferon coding regions 8p23.1 and 9p21.3 in the high-risk group. Bulk and single-cell transcriptome analysis indicated that the upregulation of ODC1 might mediate the cold immune microenvironment of DLBCL, and knocking down ODC1 effectively inhibited DLBCL cell proliferation. Promisingly, our model effectively identified patients who were insensitive to immunotherapy (CAR-T and immune checkpoint antibodies). The research was funded by: This work was supported by grants from the National Natural Science Foundation of China (grant no. 81873450, 82170181), Beijing Hospitals Authority Youth Programme (code: QML20200201), and Beijing Natural Science Foundation (No. 7222027) to Liang Wang. Keywords: Bioinformatics, Computational and Systems Biology, Microenvironment, Risk Models No conflicts of interests pertinent to the abstract.