High Tumor Mutational Burden Associates with Increased CD8 T-Cell Presence, Decreased Tumor Heterogeneity, and Increased AICDA Expression in an HIV-Inclusive Cohort of Diffuse Large B-Cell Lymphoma

Introduction:People with HIV have increased risk of diffuse large B-cell lymphoma (DLBCL) and represent nearly 10% of deaths from DLBCL in the United States, yet HIV+ DLBCL has been severely understudied compared to HIV- DLBCL. As part of our long-standing Kamuzu Central Hospital Lymphoma Study in Lilongwe, Malawi, we have a comprehensively clinically annotated cohort of DLBCLs, with over 60% arising in people with HIV. We have previously reported that HIV infection and antiretroviral therapy exposure associate with distinct immunogenomic features, including TCR repertoire and immune cell presence, in line with studies by other groups. In the era of precision oncology, understanding the relationship between the mutational landscape and immune features is crucial for selecting appropriate therapies that improve outcomes and minimize cytotoxicity for DLBCL patients. Here, we investigate the interplay between genetic complexity, tumor heterogeneity and tumor-immune interactions by stratifying based on a well-studied genomic biomarker, tumor mutational burden (TMB).

Methods: We performed whole exome sequencing of n=24 (n=16 HIV+, n=8 HIV-) formalin-fixed, paraffin-embedded (FFPE) DLBCLs. Only gene variants that met the following criteria were included: tumor and germline read depth >15 reads, germline alternative read count <1, tumor alternative read count >5, tumor VAF >5%. Samples with >100 non-synonymous variants were considered TMB-high. Spatial transcriptomics (Visium, 10x Genomics) was performed on n=6 of the exome-sequenced samples (n=5 HIV+, n=1 HIV-). Spatial transcriptomic tissue preparation, library preparation, and sequencing were performed according to manufacturer's instructions at UNC. The raw data was processed using the spaceranger pipeline, and spots with < 200 genes and/or > 30% mitochondrial gene reads were removed. We performed spot-labelling, dimensionality reduction, and gene expression analyses in R studio using Seurat, clusterprofiler, and msigdbr.

Results: Median TMB was higher in HIV+ vs. HIV- tumors (1.5-fold, p=0.05, pairwise Wilcoxon rank-sum test). Given the spatial relevance of tumor heterogeneity and immune-interaction, we performed Visium spatial transcriptomics on n=6 of the samples to identify tumor-specific gene expression and TME architecture. Three samples were TMB-high (n=3 HIV+, range 110-142 variants) and three samples were TMB-low (n=2 HIV+, n=1 HIV-, range 41-78 variants). TMB-high tumors had a higher percentage of CD8 T cells compared to TMB-low (median 8% vs 1.5%, p=0.023). Further, upon performing unsupervised clustering of B cells, TMB-high tumors had fewer distinct transcriptomic clusters of tumor cells compared to those with low TMB (median 9 vs 14, p=0.0061). TMB-high tumor clusters were characterized by increased expression of AICDA compared to TMB-low tumors (p=0.037) and exhibited distinct tumor immune evasion and DNA repair gene set profiles. TMB-high tumors had increased expression of tumor-promoting autophagy-related genes (ATG101, ATG3), and purine and pyrimidine metabolism genes (POLR1C, CDA, NME1). TMB-low tumors had increased expression of interferon-signaling genes (IRF1, IFNGR1, STAT1), and nucleotide excision repair genes (LIG1, DDB2, POLH).

Conclusions: High TMB was associated with increased CD8 T-cell presence and decreased tumor heterogeneity in an HIV-inclusive cohort of DLBCL. Because high TMB has been shown to correlate with neoantigen presentation and subsequent immune targeting of tumor, increased CD8 T-cell presence in the TMB-high tumors may lead to tumor clone selection that results in lower tumor cell heterogeneity. Additionally, TMB-high tumors exhibited increased AICDA expression, which is known as a major source of mutagenesis and an early driver of lymphomagenesis. The TMB-high and TMB-low tumor cells also expressed differential gene profiles for genes related to tumor immune evasion and DNA damage/repair, suggesting alternative mechanisms of tumorigenesis and tumor maintenance in TMB-high vs. TMB-low DLBCL. This study highlights the interplay between genomic complexity, intrinsic tumor features and immune response in DLBCL. Our findings may lead to new therapeutic avenues for DLBCL patients regardless of HIV status, though further work must be done to elucidate the role of AICDA and systemic immunity on tumorigenesis and anti-tumor response.

No relevant conflicts of interest to declare.