Diffuse large B-cell lymphoma (DLBCL) is the most common malignant lymphoma and exhibits significant clinical and molecular heterogeneity. Hepatitis B virus (HBV) infection has been associated with distinct clinicopathological features and an unfavorable prognosis in DLBCL. However, the molecular mechanisms underlying HBV-associated DLBCL and the remodeling of its immune microenvironment remain poorly understood. Through in silico analysis on transcriptomic data from the GEO and TCGA databases, we identified a set of prognostic HBV-associated differentially expressed genes (DEGs) in DLBCL and constructed a prognostic model based on these genes. A total of 72 HBV-associated DEGs were identified. From these, a 13-gene risk signature was constructed from univariate Cox regression and stepwise Akaike Information Criterion (stepAIC) optimization. This model successfully stratified patients into high- and low-risk groups, with differing survival analyses (GSE10846: risk HR = 2.38, p < 0.001; 3-year AUC = 0.753). A combined nomogram integrating clinical data with the HBV-derived signature maintained excellent discriminative ability (3-year AUC = 0.794) and calibration in the external datasets. Immune infiltration analysis revealed that compared to high-risk patients, low-risk patients had a greater total stromal and total immune cells, lower tumor purity, and a higher proportion of cytotoxic T and NK cells. In contrast, high-risk patients exhibited higher counts of M2 macrophages and immune checkpoint-expressing suppressive cells. The 13-gene model, linked to HBV biomarkers, establishes a novel computational framework for risk stratification in DLBCL and highlights key alterations in the tumor immune microenvironment associated with HBV infection. These findings provide a rationale for exploring immunotherapeutic strategies and elucidate the pathogenesis of HBV-related lymphomas.
Zhao et al. (Sat,) studied this question.