BACKGROUNDAxicabtagene ciloleucel is an autologous anti-CD19 chimeric antigen receptor (CAR) T-cell therapy that shows efficacy in patients with refractory diffuse large B-cell lymphoma (DLBCL), primary mediastinal B-cell lymphoma and transformed follicular lymphoma after failure of conventional therapy. However, the exact mechanism of anti-tumor immunity is poorly understood, in part due to the dearth of data on the events in the tumor micro-environment (TME) that occur upon exposure to CAR-T cells. We sought to quantify and characterize both CAR-T cells and non-CAR T cells within the TME of DLBCL using tissue biopsy samples collected in the ZUMA-1 multicenter trial of CAR-T cell therapy for patients with refractory DLBCL. METHODSTumor samples obtained from patients 5-30 days (median 10 days) after CAR-T infusion (“CAR-treated”, n=14) and randomly-selected untreated (“untreated ”, n=15) archival DLBCL tissue samples were analyzed by multiplex immunofluorescence using formalin-fixed, paraffin embedded tissue sections, with successive labeling by the primary antibodies KIP-1 and/or KIP-3 (recognizing separate CD19 CAR epitopes), PAX5, PD-1, CD4, and CD8, followed by secondary amplification and tyramide-conjugated fluorophores. For each case, at least 3 representative 20x fields of view were selected and imaged using a multispectral imaging platform. Two specific image analysis algorithms were designed to accurately identify CD4 and CD8 T cells and PAX5+ DLBCL cells simultaneously, then to threshold PD-1 and KIP-1/-3 by relative fluorescent units (RFU) in each phenotype. RESULTSWe identified CAR T-cells within the fixed biopsy samples of CAR-treated DLBCLs by immunostaining with CAR T-cell specific antibody KIP-1; at the timepoints analyzed, CAR T-cells comprised only a small minority of total T- cells (<2%) and included CD4+ and CD8+ T-cells. Immunostaining with a second antibody, KIP-3, validated the presence of CAR T-cells in these cases and confirmed the KIP-1 results. Expression of the T cell activation marker PD-1 was detected among majority of KIP-1+ cells. Further analysis that included KIP1-negative cells revealed that the percentage of CD8+ cells co-expressing PD-1 across all CD8+ cells was higher in the CAR-treated DLBCLs compared to the untreated DLBCLs (mean 50.1% vs 17.5%, p<0.0001 with unpaired t test ), indicating CD8 T cell activation within the tumor environment. In contrast, PD-1 positivity across CD4+ T cells were equivalent between the two groups (mean 21.8% vs 21.6%, ns with unpaired t test). The percentages of total, CD4+, and CD8+ T-cell populations in the TME were similar between the CAR-treated DLBCL and untreated biopsies. CONCLUSIONSCD4+ and CD8+ CAR-T cells can be detected in CAR-treated DLBCL patient tissue biopsies by multiplex immunofluorescence. At the time points analyzed to date, CAR-T cells comprise only a small percentage of all T-cells (<2%) within the TME. However, the presence of gene marked T cells with downregulated CAR protein expression is also possible. The activation marker PD-1 is preferentially expressed by KIP-1-negative CD8+ T cells compared to CD4+ T cells in CAR-T treated DLBCLs relative to untreated DLBCLs. These data implicate preferential activation of CD8+ non-CAR “by-stander” T-cells in the post CAR-T TME, and the possible benefit of combining PD-1 blockade with CAR-T therapy in DLBCL.*PH.C and M.L share equal contribution. DisclosuresArmand:Otsuka: Research Funding; Affimed: Consultancy, Research Funding; Pfizer: Consultancy; Infinity: Consultancy; Adaptive: Research Funding; Merck: Consultancy, Research Funding; Bristol-Myers Squibb: Consultancy, Research Funding; Roche: Research Funding; Tensha: Research Funding. Roberts:KITE: Employment. Rossi:KITE: Employment. Bot:KITE: Employment. Go:KITE: Employment. Rodig:Merck: Research Funding; Bristol Myers Squibb: Research Funding; Affimed: Research Funding; KITE: Research Funding.