125 Single cell spatial analysis and biomarker discovery in hodgkin lymphoma

• DOI: 10.1136/jitc-2023-sitc2023.0125
• Publication Date: 2023-10-31T13:37:11Z
• Authors: Alexander Xu; Aixiang Jiang; Tomohiro Aoki; Alicia Gamboa; Lauren Chong; Anthony Colombo; Yifan Yin; Joseph Lownik; Katsuyoshi Takata; Monirath Hav; Christian Steidl; Akil Merchant
• Keywords: Neoplasms. Tumors. Oncology. Including cancer and carcinogens; RC254-282

<h3>Background</h3> The tumor microenvironment (TME) is the complex milieu of cells and molecules surrounding tumor. Single cell methods systems biology have been used to identify with pro- or anti-tumor properties, selectively modulating these has great therapeutic benefits in immunotherapy. However, there limited information on how spatial organization TME determines cells' signaling their potential. resolved computational analysis needed describe interactions effect patient outcomes. Hodgkin's Lymphoma presents a unique due sparse distribution Reed-Sternberg cells. highly receptive checkpoint inhibitors insights from could better inform immunotherapies general. <h3>Methods</h3> Here we apply Imaging Mass Cytometry (IMC), technology perform ~40-plex protein 1 micron resolution tissue (figure 1A), study cohort 260 matched samples at diagnosis after relapse 91 patients relapsed/refractory Lymphoma. We developed pipeline architecture propose putative biomarkers clinical response relapse. New are define based localized ligand receptor driving tumor-immune clustering (rosetting) Hodgkin's. <h3>Results</h3> analyzed over 7 million IMC features – subtypes positioning that correlate In brief, found signatures reorganization relapsed validated existing such as CD68+ macrophage proportion (p=0.018, likelihood ratio test). metrics 'digital biopsies' derive novel, spatially-dependent LAG3+Tregs (p=0.022 for all cells, p=0.66 near tumor) 1B,C). quantified rosetting, recruitment immune characteristic 1D), by defining ligand-receptor associations different stages disease severity. proposed novel involving macrophages (TIM3/Galectin9, CD80, PDL1) CD4 T (LAG3/HLAII, VISTA/Galectin9, PD1/PDL1) (p&lt;0.05) 1E). <h3>Conclusions</h3> Spatial HL revealed composite predict outcomes 1F). These cannot be described using single tools low-plexed imaging, represent truer picture biology. here can universally applied other data biomarker discovery analysis, multi-color IHC an independent cohort.