Spatial Analysis Reveals Impaired Immune Cell Function within the Tumor Microenvironment of HIV-associated Non-small Cell Lung Cancer

Background Among people with HIV (PWH), lung cancer is the leader cause of cancer mortality, with increased risk and poor clinical outcomes compared to people without HIV (PWOH). HIV is known to result in persistent global immune dysfunction despite antiretroviral therapy, but little is known about the lung cancer tumor microenvironment. This study explored whether the tumor microenvironment (TME) of HIV-associated non-small cell lung cancer (NSCLC) is associated with an immunoregulator environment that limits tumor-specific immune responses.

Methods A tissue microarray was constructed with NSCLC tumors from 18 PWH and 19 PWOH (matched for histological subtype, stage, year of diagnosis, age, sex and smoking status), and incubated with metal-conjugated antibodies for evaluation by imaging mass cytometry (IMC). IMC marker scores were extracted by automated cell segmentation and single-cell data was analyzed by Phenograph using unsupervised cell-segmentation and clustering of cells. Evaluation of tumor infiltrating immune cells, CD4+ and CD8+ T cells as well as CD68+ tumor associated macrophages were characterized for marker expression using a linear mixed-effects model. Additionally, a computational strategy based on the PageRank mathematical algorithm was used in order to establish an unsupervised and cell segmentation-independent signature associated with HIV status to discriminate differential expression of immune cell markers within the TME of the two groups. Peripheral blood mononuclear cells (PBMCs) from HLA-A02 donors (PWH and PWOH) were co-incubated with HLA-A02 lung cancer cell lines to quantify tumor killing (by Annexin V staining) and expression of T cell markers Lag-3 and CD25.

Results Within the TME from HIV+ tumors, there is comparable level of infiltration of lymphocytes and tumor associated macrophages (TAMs) compared to non-HIV tumors, with a trend towards increased CD8+ T cells and decreased CD4:CD8 ratio among HIV+ tumors. Using a random effects model of individual markers, HIV+ tumors revealed increased expression of Ki67 and Granzyme B (GRZB) among CD8+ T cells; increased Ki67 and PD-1 among CD4+ T cells; and increased PD-L1, PD-L2, and Ki67 among TAMS. Unsupervised clustering analysis from IMC data demonstrated differential distribution of tumor infiltrating CD8+ T cell clusters between HIV+ and non-HIV tumors, defined by marker expression patterns. Three clusters were significantly elevated in HIV+ tumors (57.1% vs. 21.7% in non-HIV tumors, p<0.0001). All three clusters had comparatively elevated PD-1 and Lag-3 expression with varying expression of activation and proliferation markers CD25 and Ki67. Within tumor-infiltrating CD4+ T cells, a cluster characterized by checkpoint protein expression (PD-1+ and LAG-3) was also highly represented in HIV+ cases (35.2% vs. 9.8% in non-HIV cases, p<0.0001). HIV+ tumor-associated macrophages (TAM) had higher expression of immunoregulatory molecules (PD-L1, PD-L2, B7-H3, B7-H4, IDO1 and VISTA), confirmed by the expansion of three clusters comprising 58.8% of TAMs vs. 17.8% in non-HIV tumors (p<0.0001). Discrimination of cells between HIV+ and HIV-TME was further confirmed by spectral graph theory with 84.6% accuracy, with a combination of markers on TAMs and T cells. Lastly, PBMCs from PWH exhibited decreased tumor killing when exposed to HLA-matched NSCLC cell lines compared to PBMCs from PWOH. CD8+ T cells from PWH additionally had increased expression of immune checkpoint inhibitor Lag-3 upon exposure to tumor cells.

Conclusions Our study demonstrates that the TME of HIV+ patients is characterized by a unique immune landscape, distinct from that of PWOH, with evidence of expansion of immune cells with enhanced immunoregulatory phenotypes and associated with impaired anti-tumor responses.