Immune cellular patterns of distribution affect outcomes of patients with non-small cell lung cancer

Abstract

Background Study of the geographic distribution of cellular populations and their interaction with malignant cells in non-small cell lung cancer (NSCLC) is essential to understand the roles of cellular populations and potentially design new therapeutic approaches. Material and Methods We studied 225 formalin-fixed, paraffin-embedded tumor tissue samples from patients with stage I-III NSCLC—142 adenocarcinomas and 83 squamous cell carcinomas—placed in tissue microarrays. Twenty-three markers were used, including T-cell, B-cell, immune checkpoint, and myeloid cell markers, placed in five multiplex immunofluorescence panels. We used dimensional data reduction and cellular spatial distribution to identify cell phenotypes and their cellular distribution across panels. In addition, we analyzed associations between cellular spatial distribution patterns and clinicopathologic variables, tumor mutational status, and outcomes. Results Overall, we observed two patterns of cellular distribution—unmixed and mixed—related to T-cells, B-cells, granulocytic cells, and macrophages, and we detected various cellular interactions between malignant and immunosuppressive cells and various associations between cell distribution and clinicopathologic characteristics. Cellular distribution patterns and the distance from various cell phenotypes to malignant cells were associated with recurrence-free survival and overall survival in univariate and multivariable analysis. Conclusions Diverse populations of tumor cells and tumor-associated immune cell phenotypes can be present in NSCLC samples. Spatial distribution is a new tool to better understand the tumor microenvironment and help select therapeutic interventions.

Up-regulated PLA2G10 in cancer impairs T cell infiltration to dampen immunity

“…Moreover, longer distances between CD8 + T cells and tumor cells were recorded in PLA2G10-high tumors (Fig. 7G) (70,71). In summary, our results indicate that the up-regulation of PLA2G10 is prevalent in various types of human cancers and is associated with T cell exclusion.…”

“…This tool creates a single data file with information related to individual cell events, marker abundance, nuclear morphologic features, density in specific compartments (number of positive cells per mm 2 ), and marker-based phenotype and spatial location (X and Y coordinates) (70). For the automated calculation of distances between cells, we used the tool find_nearest_distance function of the R package (Akoya Biosciences) as previously reported (71).…”

Nicheformer: a foundation model for single-cell and spatial omics

“…Beyond cellular niche labels and neighborhood composition, we asked whether local cell density is encoded in a cell's expression profile. It is long known that cell density can strongly affect growth behavior in vivo and in culture; also, increased cell density is a key feature of the formation of the tumor microenvironment, which leads to the creation of a hypoxic environment and depletion of infiltrating immune cell populations 83,84 . For example, in colon cancer, it was shown that the immune cell density is associated with patient survival and can be used for tumor-immune patient stratification for improved anticancer therapy 75 .…”

Spatial insights into immunotherapy response in non-small cell lung cancer (NSCLC) by multiplexed tissue imaging

“…Indeed, meta-analysis of the field across 10 tumour types in 8135 patients has shown that PD-L1 IHC (AUC 0.65) and TMB (AUC 0.69) fall behind multiplex immunofluorescence methods (AUC 0.79) in their ability to accurately predict patient ICI response [9]. Moreover, this suggests largely that the efficacy of the ICIs is not tied directly to PD-L1 directed immune evasion nor tumour antigenicity but rather other phenotypic properties of tumour tissues that are being systematically discovered through multiplexed imaging techniques [10,11]. Spatial metrics describing the geographical associations between cells are providing increasing depth to our ability to quantify and compare tissue composition, and are poised to deeply aid our understanding of the cellular and functional architecture of tumour tissue.…”

“…It requires coordination of both supervised and unsupervised analytical steps to accurately assign genuine cell phenotypes from single cell expression data of cohorts of FFPE samples with varying expression levels [19]. This approach differs significantly from trained thresholds that may be applied to optimized, amplified multispectral data [11,12]. Indeed, no single tool offers a turn-key solution for image QC, cell segmentation, data integration and normalization, parameter optimization, cell phenotyping and spatial analysis to identify features associated with binary and time to event clinical outcomes [20].…”