Abstract A051: Spatially dispersed MYC-BCL2 co-expressing cells confer poor survival in DLBCL through functional p53 loss and PGE2-mediated immune evasion

Abstract Intratumoral heterogeneity in the spatial arrangement of phenotypically distinct tumor subpopulations is increasingly recognized, yet whether spatial topology independently determines clinical outcomes and reflects distinct biological programs remains unclear. Using diffuse large B-cell lymphoma (DLBCL) as a model, we reported (AACR 2025) that the spatial distribution of MYC+BCL2+BCL6- double expressor (DE) cells — quantified by point process modeling of multiplex immunohistochemistry (mIHC) across 476 patients in four independent cohorts — independently predicts survival after chemoimmunotherapy, and identified a transcriptional signature of dispersed DE cells associated with inferior outcomes in gene expression cohorts (4,594 patients). The mechanistic basis for why genotypically identical MYC+BCL2+ cells adopt distinct spatial configurations with divergent clinical impact is unknown. Given that MYC-driven ribosome biogenesis (RiBi) normally triggers the impaired RiBi checkpoint (IRBC), wherein the RPL5/RPL11/5S rRNA complex binds and inhibits MDM2 to stabilize p53, we investigated whether this prognostic spatial signal reflects differential IRBC engagement. Using single-cell RNA-sequencing in de novo (n=17) and relapsed/refractory (R/R; n=99) settings and digital spatial profiling (GeoMx DSP-WTA, n=64), we found that although cells with dispersed and clustered signatures both upregulated RPL5 under MYC-driven ribosomal stress, dispersed DE cells selectively evaded IRBC engagement: they exhibited elevated MDM2 expression, reduced p53 stabilization, and diminished p53 transcriptional activity, conserved across de novo and R/R contexts and independent of TP53 mutation status. We postulate that NF-κB pathway activation in dispersed DE cells transcriptionally drives MDM2 overexpression sufficient to overwhelm RPL5/RPL11-mediated inhibition, establishing a feed-forward loop: sustained MDM2 activity inactivates p53, which normally suppresses NF-κB, thereby promoting PTGES3-mediated prostaglandin E2 (PGE2) synthesis. Cell-cell communication analysis confirmed enriched PGE2 signaling in dispersed versus clustered configurations across both disease settings. Critically, hyperplex spatial proteomics (PhenoCycler, n=152) revealed that dispersed DE cells, despite significantly greater spatial proximity to T cells than clustered counterparts, orchestrated a profoundly immunosuppressive niche characterized by regulatory T cell enrichment and terminal CD8+ T cell exhaustion — a spatial immune phenotype consistent with tumor-derived PGE2-mediated suppression of anti-tumor effector responses and conserved across de novo and R/R disease. These findings establish IRBC avoidance as a spatially determined mechanism of non-mutational p53 inactivation in MYC-driven lymphoma and identify the p53-NF-κB-PGE2-T cell exhaustion axis as a potential therapeutic vulnerability in spatially dispersed double expressor DLBCL. Citation Format: Shruti Sridhar, Charmaine Ong, Chartsiam Tipogamut, Qiang Pan Hammarström, Kasthuri Kannan, David W. Scott, Xubin Li, Michael R. Green, Claudio Tripodo, Anand D. Jeyasekharan. Spatially dispersed MYC-BCL2 co-expressing cells confer poor survival in DLBCL through functional p53 loss and PGE2-mediated immune evasion [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A051.