(208e) Spatial Protein Analysis of the Tumor Microenvironment and Biomarkers in Hodgkin’s Lymphoma

Introduction: The tumor microenvironment (TME) is the complex milieu of cells and molecules surrounding the tumor. Single cell methods and systems biology have been used to great effect to identify subpopulations of cells that have pro- or anti-tumor properties, and selectively modulating these has great therapeutic benefits, especially in immunotherapy. However, there is limited information on the spatial organization of these subpopulations, which determines how they signal and their therapeutic potential. Single cell resolved spatial computational analysis is needed to describe the complex interactions of the TME and their effect on patient outcomes. Hodgkin's Lymphoma presents a unique spatial TME due to the sparse tumor distribution of Hodgkin's Reed-Sternberg tumor cells. Yet even though it is difficult to disentangle signaling in Hodgkin's tumor cells - which are immune derived - from the immune TME, Hodgkin's is highly receptive to checkpoint inhibitors among lymphomas and insights gleaned from the Hodgkin's TME could better inform immunotherapies across lymphomas.

Materials and Methods: Here we apply Imaging Mass Cytometry (IMC), a technology to perform ~40-plex protein analysis with 1 micron resolution in tissue, to study a cohort of 260 matched samples at diagnosis and after relapse from 90 patients with relapsed/refractory Hodgkin's Lymphoma. We developed a computational pipeline to perform cell phenotyping, spatial analysis, and biostatistics, to describe tumor architecture and propose putative biomarkers of Hodgkin’s clinical response and relapse. A novel feature of the pipeline is to quantify proteins and spatial analysis on the same numerical scale for each cell, to generate hybrid spatial and protein biomarkers. New methods are used to describe the localized tumor-immune clustering (rosetting) unique to Hodgkin's, and biomarkers to predict patient outcomes.

Results and Discussion: We analyzed over 7 million cells for their phenotype and spatial organization. We use IMC to describe spatial features of the tumor - cell subtypes and their positioning - that correlate to clinical outcomes. We identify proteins such as CXCR5 that correlate to survival in specific spatial contexts, and we describe spatial reorganization from diagnosis to the relapsed tumor as it relates to survival and relevant clinical factors such as MHC expression and EBV infection. A significant conceptual advance was to use spatial metrics to perform “digital biopsies”, a selection of tumor regions comparable across patients. We also describe cell rosetting, a localized recruitment of immune cells to tumor that is characteristic of Hodgkin's, by defining ligand-receptor associations that are unique to different stages of disease severity. We find different cell-specific protein expression biomarkers for different length scales of spatial interaction using IMC data. At whole image scales, we validated multiple existing biomarkers in the literature using our data set. At smaller regions and cell-cell contact spatial scales, we proposed novel biomarkers involving macrophages (TIM3/Galectin9, CD80, PDL1) and CD4 T cells (LAG3/HLAII, VISTA/Galectin9, PD1/PDL1).

Conclusions: Spatial analysis of the HL microenvironment revealed composite features of the TME that predict clinical outcomes. These features cannot be described using single cell tools or low-plexed imaging, and represent a truer picture of HL biology. The pipeline developed here can be universally applied to other spatial protein data for biomarker discovery and analysis, and the biomarkers proposed here will be validated with IHC.

Figure caption: IMC analysis of Hodgkin’s Lymphoma. A. IMC generates a 40-plex protein image that can be segmented into single cells. B, C. Single cell data can be processed to obtain immune phenotypes, shown by heatmap and UMAP. D. Immune cells aggregate to form rosettes around tumor cells. We find specific ligand-receptor interactions associated with rosetting. E. Tumor/Treg ICOSL/ICOS expression is highly correlated with Treg rosettes, while Tumor/CD4 LAG3/HLADPDQDR expression is negatively correlated with CD4 rosettes. F. IMC-generated biomarkers predict overall survival.