VSIG4 as a tumor-associated macrophage marker predicting adverse prognosis in diffuse large B-cell lymphoma

Abstract

Introduction: Diffuse large B-cell lymphoma (DLBCL) exhibits heterogeneous tumor microenvironment. However, the role of tumor-associated macrophages (TAMs) in the DLBCL tumor microenvironment remains unclear. This study aims to elucidate the heterogeneity of TAMs in DLBCL to identify critical TAM-associated prognostic biomarkers.

Methods: Transcriptome data from DLBCL patients were obtained from online database. The CIBERSORT algorithm was applied to quantify TAM abundance across samples. Consensus clustering was used to stratify DLBCL into distinct clusters based on TAM subtype enrichment. Differential gene expression analysis, LASSO regression, univariate/multivariate Cox regression and Kaplan-Meier survival analysis were employed to identify key prognostic biomarkers. Validation of VSIG4+TAM subpopulation was performed using flow cytometry and multiplex immunohistochemistry. A local cohort of 375 DLBCL patients was investigated to explore the correlation between VSIG4 expression and various genetic and pathological characteristics including prognostic outcomes.

Results: Four distinct DLBCL clusters, each enriched with specific TAM subtypes were found. The cluster dominated by M2 TAMs exhibited the worst prognosis. Differential analysis identified VSIG4 as a critical prognostic factor, with high expression in the M2 TAM-enriched cluster. Flow cytometry and mIHC confirmed VSIG4+ TAMs as a subpopulation within CD68+/CD163+ M2 macrophages. VSIG4 expression correlated with adverse genetic features (PIM1, ETV6, CD70 mutations) and aggressive pathological characteristics (non-GCB phenotype, MYC+/BCL-2 double-expression). Multivariate Cox regression confirmed VSIG4 as an independent prognostic factor for poor survival. Survival analysis suggested that VSIG4's prognostic impact operates independently of regulating lymphocyte infiltration, highlighting its unique role in DLBCL tumor microenvironment.

Discussion: This study identifies VSIG4 as a TAM-associated marker of adverse prognosis of DLBCL and the expression of VSIG4 is related to high-risk genetic and pathological features. These findings position VSIG4 as a promising therapeutic target for immune checkpoint intervention in DLBCL.

Keywords: VSIG4; diffuse large B-cell lymphoma; prognostic markers; tumor microenvironment; tumor-associated macrophages.

Figure 1

Subclusters of DLBCL classified by the abundance of TAMs derived from CIBERSORT. (A) Heatmap of the consensus clustering of the DLBCL cases by the infiltrating abundance of M0, M1 and M2 macrophages. (B) PCA analysis of the clusters. (C) K-M plot and survival analysis of 4 clusters. (D) Case distribution of clinical stage (left) and IPI risk (right) in each cluster. (E) Abundance of infiltrating macrophages (%) in each cluster. * indicates P-value <0.05; ** indicates P-value <0.01; *** indicates P-value <0.001; ns indicates “not significant”.

Figure 2

Discovering the unfavorable prognostic marker upregulating in the cluster enriched with M2-like TAMs. (A) Flowchart of the analysis process. (B) Volcano plot of the differentially expressed genes in cluster 2. (C) Forest plot of the results of multivariate Cox regression. (D) K-M plot and survival analysis of VSIG4-High group versus VSIG4-Low group. (E) Correlations between the expression of VSIG4 and the infiltration of tumor microenvironmental cells. Numbers below the labels indicate the correlation coefficients; * indicates P-value <0.05; ** indicates P-value <0.01; *** indicates P-value <0.001; N.S indicates “not significant”.

Figure 3

IHC results of representative cases of VSIG4+ and VSIG4- DLBCL. The images of VSIG4, CD68, CD163, CD206 and HE of a representative VSIG4+ case (top) and a representative VSIG4- case (bottom). Original magnification ×400.

Figure 4

The co-expression of VSIG4 with TAM markers. (A) Flowcytometry analysis of a DLBCL cases showing the VSIG4 was expressed mainly in CD68+/CD163+ cells and (B) a part of but not all CD68+/CD163+ cells were VSIG4+. VSIG4+ cells were marked blue in (A) and CD68+/CD163+ cells were marked brown in (B). (C) Multiple IHC images of a representative VISG4+ case showing most majority of VSIG4+ cells co-expressed with CD68+/CD163+ cells (pink arrows) but still some were CD68+/CD163- (green arrows). Original magnification ×400 for panel C.

Figure 5

NGS-based genetic analysis and LymphGen subtyping of VSIG4+ and VSIG4- cases. (A) Heatmap of most frequently variated genes in VSIG4- and VSIG4+ cases and list of significantly variated genes. (B) Comparison of LymphGen subtyping between VSIG4+ and VSIG4- cases. (C) Chromosome instability scores and tumor mutation burden scores of the cases between VSIG4- and VSIG4+ group. * indicates P-value <0.05; ** indicates P-value <0.01; *** indicates P-value <0.001; ns indicates “not significant”.

Figure 6

Overall survival of VSIG4- and VSIG4+ cases in different clinicopathological status. (A) OS of VSIG4- cases versus VSIG4+ cases. (B, C) OS of VSIG4- cases versus VSIG4+ cases in rituximab-based or rituximab-free group. (D, E) OS of VSIG4- cases versus VSIG4+ cases in early stage (I-II) or late stage (III-IV) group. (F, G) OS of VSIG4- cases versus VSIG4+ cases in low (0-2) or high (3-5) IPI group. (H, I) OS of VSIG4- cases versus VSIG4+ cases in non-GCB or GCB group. (J, K) OS of VSIG4- cases versus VSIG4+ cases in non-DE or DE group. (L) OS of cases grouped by VSIG4 expression and CD4+ cell abundance. (M) OS of cases grouped by VSIG4 expression and CD8+ cell abundance.