{"doi":"10.1002/hon.7_2880","title":"HIGHLY MULTIPLEX TISSUE IMAGING OF DLBCL IDENTIFIES NOVEL PATHOLOGICAL FEATURES PREDICTIVE OF OVERALL SURVIVAL","abstract":"Introduction: Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma and is known to be a highly heterogenous disease. Several recent studies have identified subtypes based on prominent genetic patterns; however, efforts to identify relationships between these genetic groups and the pathophysiology of the disease remain limited. Characterization of tumor phenotypes and their microenvironments alongside genomic alterations will provide much needed context and may assist in identifying novel therapeutic targets or improvements in patient stratification without the use of genetic testing. Methods: We used imaging mass cytometry (Hyperion imaging system) to simultaneously quantify 2 sets of 37 target antigens across serial sections of tumor tissues. A total of 64 unique antigens were quantified across 1132 high-dimensional images from 370 tumors. Detailed clinical information including treatment outcome after R-CHOP immunotherapy is available for 303 patients, with information on common genetic alterations available for 338. Existing frameworks for cell segmentation (DeepCell) and single-cell clustering (Seurat) were used to process and identify major cell types and functional states in an unsupervised manner. Additionally, a proportional hazards regression model incorporating normalized quantification of all tumor biomarkers was used to identify relationships to treatment outcome and survival. Results: Clustering of patients solely by cellular composition resulted in 8 distinct groups. Notably, a cluster depleted of tumor infiltrating immune cells with high tumoral BCL2 expression was found to be significantly enriched for MYC and BCL2 double-hit status (q = 1.27 × 10-6). In line with previous work, the double-hit associated cluster had significantly worse overall survival (HR = 2.45; p < 0.0001). Additionally, a subset defined by canonical ABC features, a high proliferation score, and a prominent anti-inflammatory immune response was found to have poor OS (HR = 1.45; p < 0.05). As previously reported, high CD20 expression was found to be significantly associated with superior OS (HR = 0.57; p < 0.05). Tumor expression of other markers including BCL2 (HR = 1.3; p < 0.05), phospho-histone H3 s28 (HR = 1.4; p < 0.005), and T-bet (HR = 1.4, p < 0.05) were found to be associated with inferior outcomes. Conclusions: Advances in multiplex tissue imaging have provided a framework to rapidly improve our understanding of tumor pathophysiology and relationships with genomic alterations. Expanded screening of antigen targets will likely yield novel biomarkers that can then be translated to conventional IHC methods for improved patient stratification. The research was funded by: NIH grants UL1TR002384, R01CA194547, LLSSCOR grants 180078-02, 7021-20 Keywords: Bioinformatics; Computational and Systems Biology, Tumor Biology and Heterogeneity, Aggressive B-cell non-Hodgkin lymphoma Conflicts of interests pertinent to the abstract A. Melnick Consultant or advisory role: Epizyme, Constellation, KDAC Pharma, BMS and ExoTherapeutics Research funding: Janssen Pharmaceuticals, Sanofi and Daiichi Sankyo O. Elemento Consultant or advisory role: Freenome, Owkin, Volastra Therapeutics and One Three Biotech Research funding: Janssen, Johnson and Johnson, Volastra Therapeutics, AstraZeneca and Eli Lilly","journal":"Hematological Oncology","year":2021,"id":225522,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":0,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9584,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2021-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":28988,"name":"Hiranmayi Ravichandran","orcid":"0000-0001-8671-7667","position":1,"is_corresponding":false},{"id":308265,"name":"Wayne Tam","orcid":"0000-0003-4283-0005","position":2,"is_corresponding":false},{"id":255172,"name":"Christian Steidl","orcid":"0000-0001-9842-9750","position":3,"is_corresponding":false},{"id":106831,"name":"David W. Scott","orcid":"0000-0002-0435-5947","position":4,"is_corresponding":false},{"id":237422,"name":"Ari Melnick","orcid":"0000-0002-8074-2287","position":5,"is_corresponding":false},{"id":41043,"name":"Olivier Elemento","orcid":"0000-0002-8061-9617","position":6,"is_corresponding":false},{"id":361880,"name":"Dylan McNally","orcid":"0000-0001-7831-0349","position":0,"is_corresponding":true}],"reference_count":0,"raw_metadata":null,"created_at":"2026-07-18T23:54:26.581453Z","pmid":null,"pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}