Abstract Primary central nervous system lymphoma (PCNSL) is histologically a subtype of diffuse large B-cell lymphoma (DLBCL), sharing genetic and transcriptomic similarity, but with distinct clinical features, particularly its confinement to the CNS and higher relapse risk. The microenvironmental basis for its divergence from systemic DLBCL remains unclear. Using spatial transcriptomic approaches (Xenium/GeoMx digital spatial profiling) in a comparative study of PCNSL (n=17) and DLBCL (n=76), we found that PCNSL, unlike systemic DLBCL, is dominated by an immunosuppressive macrophage compartment enriched for cholesterol-metabolism programs. In independent cohorts of PCNSL profiled by single-cell RNA sequencing, we validated the presence of a recurrent population of lipid-laden macrophages (LLMs): TREM2/GPNMB-expressing, lipid-remodeled cells transcriptionally distinct from resident microglia and consistent with an infiltrating monocyte origin, not previously characterized in CNS lymphoma. LLMs formed immunosuppressive niches with regulatory T cells, and using Cellscape hyperplex proteomic imaging we demonstrate that LLM-Treg spatial interactions are associated with chemotherapy response. To test whether LLMs are lymphoma-driven and functionally important, we developed an immunocompetent syngeneic PCNSL mouse model, driven by Myd88 L252P and Cd79b mutations with Bcl2 overexpression. Monocyte-derived macrophages in lymphoma-bearing brain regions acquired an LLM-like state, not seen in lymphoma-free brain regions or in splenic tumors driven by the same oncogenic lesions. TREM2-SYK signaling sustained this state, and SYK inhibition reversed its tumor-supportive activity ex vivo . These findings identify the LLM program as a targetable immunosuppressive myeloid state in CNS lymphoma. Key Points PCNSL is enriched for TREM2 + lipid-laden macrophages (LLMs) forming immunosuppressive niches enriched relative to systemic DLBCL. The first immunocompetent MCD-like-genotype PCNSL model recapitulates human LLMs; SYK inhibition reverses their tumor-supportive activity.

Introduction Primary central nervous system lymphoma (PCNSL) is an aggressive B-cell lymphoma exhibiting unique central nervous system (CNS) tropism, and high recurrence rates despite sharing morphological and molecular features with systemic Diffuse Large B-Cell Lymphoma (DLBCL). Given the unique immune landscape of the CNS and the critical role of macrophages in neural tissue, we hypothesized that PCNSL may be sustained by a CNS-specific macrophage program distinct from DLBCL, representing a targetable mechanism underlying immune evasion, CNS confinement, and treatment resistance. While tumor-associated macrophages (TAMs), particularly CD163+ M2-like macrophages, are known to be enriched in PCNSL tumor microenvironment (TME), prior spatial studies have focused predominantly on tumor-intrinsic features and T-cell dysfunction, leaving macrophage organization and functional programming poorly characterized. Methods We employed cutting-edge spatial multi-omics using four complementary platforms to comprehensively profile macrophage heterogeneity. Formalin-fixed paraffin-embedded (FFPE) samples in TMA format (26 PCNSL, 89 DLBCL) were analyzed using a Xenium 380-gene immuno-oncology panel. GeoMx digital spatial profiling whole transcriptome analysis (DSP-WTA) was performed in 82 cases (17 PCNSL, 65 DLBCL) using CD3, CD20, and CD68 cell masks. Macrophage signatures identified from DSP were validated across independent single-cell RNA sequencing (scRNA-seq) datasets (PCNSL N=28, DLBCL N=17, reactive lymph nodes N=2), distinguishing microglia from monocyte-derived macrophages. CellScape high-plex imaging was used to confirm phenotypes at the protein level and assess spatial proximity and interactions in 24 PCNSL and 5 tonsil samples using 40 architecture, immune and macrophage markers. Results Xenium spatial profiling revealed significantly higher macrophage abundance in PCNSL versus DLBCL, confirmed by DSP-WTA (p=0.0002). DSP further demonstrated that PCNSL TAMs upregulate immunosuppressive genes (CRYAB, SPP1) while downregulating T-cell recruitment genes (CCL19, IGSF6), with enrichment of glycolysis, cholesterol homeostasis, and peroxisome pathways. The CXCL9:SPP1 expression ratio, a validated macrophage polarity biomarker of prognosis in cancer, was significantly reduced in PCNSL macrophages across both DSP and scRNA-seq datasets (p=0.027). Of specific interest, differentially expressed gene (DEG) projection and BayesPrism deconvolution of CD68+ regions of interest (ROIs) revealed enrichment of TREM2+ macrophages (p=0.00066) and elevated GPNMB+ lipid-laden macrophage (LLM) signatures (p=0.004) in PCNSL. scRNA-seq confirmed higher LLM signatures in PCNSL versus DLBCL (p=0.0052) and lymph nodes (p=1.2e-06), identifying a monocyte-derived subset enriched in cholesterol metabolism and high-density lipoprotein (HDL) binding pathways. Cell-cell communication analysis revealed enhanced interactions between LLMs and regulatory T-cells via secreted phosphoprotein 1 (SPP1), apolipoprotein E (APOE), and intercellular adhesion molecule 1 (ICAM1) signaling axes. CellScape imaging confirmed the presence of GPNMB+CD163+CD274+ macrophages in PCNSL, with spatial analysis showing that LLM-T cell proximity correlated with treatment response. Conclusion We define a CNS-specific macrophage program characterized by TREM2+ GPNMB+ lipid-laden macrophages that may foster CNS tropism in PNCSL, along with immune evasion through metabolic reprogramming and regulatory T-cell activation. This first comprehensive spatial multi-omics characterization of macrophage heterogeneity distinguishing PCNSL from DLBCL identifies TREM2, SPP1 and lipid metabolism as potential therapeutic targets for macrophage-directed immunotherapy in this disease of unmet clinical need.

Supplementary methods. Supplementary Figure 1. Phenotyping of B-cells in non-malignant tissues. A, Quantitation of marker positivity across ten tonsil and two reactive lymph node samples (rLN). Analysis is spatially resolved between the GC and extra-GC zones. B, Spatial map of cellular coordinates based on cell segmentation of images in Figure 1B. Marker-positivity is indicated, and a total proportion of positive and negative cells is depicted as a pie chart. These maps were used to derive sub-population phenotypes depicted in Figure 1C. Scale bar is 100μm. C, Proliferation analysis (i.e., Ki67-positivity) among sub-populations in five tonsil samples. Median with interquartile range, whiskers denote 10th and 90th percentile. Supplementary figure 2. Example pseudo-colored mfIHC images for MYC, BCL2, BCL6 cases in DLBCL. Images of a range of mean fluorescent intensities are shown with equal scaling for reference. Supplementary figure 3. Global distribution of MYC, BCL2 and BCL6 sub-populations within DLBCL cohorts. Heat-maps displaying the percentage extent of individual markers and each sub-population within the DLBCL NUH, CMMC, SGH and MDA cohorts. Hierarchical k-means clustering of patients according to sub-population extent is applied. Positivity shading for single markers ranges between 0-100% positivity, whereas shading for sub-populations reflects 0-50% positivity and remains fully saturated until 100%. IPI Risk Group - International Prognostic Index Risk Group, FISH - fluorescence in situ hybridization. Supplementary figure 4. Intra-tumor heterogeneity of sub-populations. A, Correlation of sub-population extent quantification between two biopsies of the same patient for which at least two tissue microarray (TMA) biopsies are available. Correlation is shown separately for lymph node and extranodal biopsies. Spearman rho is indicated for each correlation. Axes are in exponential and equivalent in all panels. B, Sub-population percentage extent quantification across multiple TMA cores (columns) of the same patient (rows). Pie charts are ordered according to decreasing cell numbers evaluated per core. All patients from the NUH cohort with at least five cores are evaluated. A heterogenous cluster is highlighted by the red box. Supplementary figure 5. Spatial heterogeneity of sub-population interactions. A, Conceptual schematic of pair correlation function (PCF) plots depicting a clustered distribution (left, green) and a random distribution (right, grey). Representative counterpart spatial maps are above each plot. B, PCF analysis for sub-populations to investigate spatial clustering (top). Mean results for two independent cohorts (shading is cohort standard deviation). An example tissue microarray core is shown as physical distance reference for spatial analyses (bottom left). Absolute number of neighboring cells expected within a given radius (data from 3500 randomly selected cells across all images, mean with standard deviation) (bottom right). C, Actual spatial map of sub-populations of an example DLBCL case (top). Extent of all sub-populations within the sample is shown on the left. Simulated, hypothetical random distribution of cells for the same case (middle). PCF analysis for the shown sample and its matched simulated random distribution (bottom). Scale bars in B and C are 100µm. D, Mean deviations from expected neighbor abundance (Δ%) summarizing cell-cell interactions between sub-populations for the sample shown in (C). E, Sub-population interaction matrices from spatially distinct biopsies (cores in tissue microarray) for example DLBCL patients. Biopsies of stable, spatially homogenous, sub-population interaction profiles are grouped (top), whereas biopsies of a differing, heterogenous, interaction profile are grouped separately (bottom). Supplementary figure 6. Global deviations from expected spatial neighbor abundance (Δ%). Hierarchical clustering (minimum variance method) of measured Δ% for all cases in the SGH and MDA cohorts. Extents of sub-populations are indicated for reference (top). For the MDA cohort, multiple biopsies (n = 1-3) from the same patient were included in the analysis to determine spatial interaction similarity across spatially distinct regions (bottom). Supplementary figure 7. Correlation of predicted MYC, BCL2 and BCL6 sub-population percentage extent based on single oncogene positivity and observed percentage extent in DLBCL cohorts. Spearman rho, axes are equivalent in all panels. Supplementary figure 8. Variance of M+2+6- percentage extent in the context of positivity calling across a 15% cut-off. A, M+2+6- scoring variance across multiple pathological imaging fields. All whole-tissue DLBCL sections from University of Palermo (UP), and samples from the NUH TMA with at least four fields scored per patient and a mean M+2+6- score above 5% are shown. Mean with SD. Ordinates between 50-100% are compressed for clarity. Dashed line denotes M+2+6- 15% positivity. B, Stability of M+2+6- case positivity calling across scoring increasing number of imaging fields. All cases from panel A with at least five fields scored in this study are shown. Only one case is called M+2+6- Low (<15%) at the first image scored, and subsequently called M+2+6- High (≥15%) after two or more fields scored. Supplementary figure 9. Mapping of mRNA expression data into percentage extent data. A, Cumulative histogram of MYC, BCL2 and BCL6 protein percentage extent positivity in DLBCL cohorts (data transformed from Figure 4A) (top). B, Aggregated single oncogene cumulative distribution of MYC, BCL2 and BCL6 protein percentage extent positivity across all measured protein cohorts and its smoothed empirical cumulative distribution function (eCDF).C, Distribution of inferred single oncogene percentage extent in GEP cohorts. (see Supplementary table 6 for all values). Supplementary figure 10. Analysis of the GOYA clinical trial. A, Correlation of MYC mRNA with quantitative IHC score. Linear regression (left) and Wilcoxon rank sum test (right). B, Analysis as in (A) for BCL2. C, Kaplan-Meier curves for PFS and OS for patients stratified across the 15% M+2+6- metric (GEP-derived). Multivariate Cox proportional hazards model is available in Supplementary table 9. PFS - progression free survival, OS - overall survival. Supplementary figure 11. Proliferative advantage of cyclin D2 (CCND2) overexpressing B-cells. Representative FACS plots documenting to the expansion over time of the cyclin D2 positive GC B-cell population in cyclin D2 overexpressing GC B-cells (CCND2-Lyt2) and non-cyclin D2 overexpressing GC B-cells (Empty vector Lyt2). All GC B-cells co-overexpress BCL2, BCL6, MYC and GFP. Supplementary table 3. Non-parametric correlation of sub-population percentage extent with clinicopathological features. Supplementary table 4. Pooled univariate analysis for MYC, BCL2 and BCL6 single oncogene and sub-populations percentage extents as a continuous variable at 5% increments as predictors for overall survival (OS) in mfIHC cohorts of DLBCL (Cox proportional hazards model). Supplementary table 5. Univariate analysis of clinicopathological features as a predictor of overall survival (OS) after first-line R-CHOP treatment in the NUH, SGH and MDA cohorts of DLBCL (Cox proportional hazards model). Supplementary table 7. Pooled univariate analysis for sub-population metrics as a continuous variable at 5% increments as predictors for overall survival (OS) in GEP DLBCL cohorts (Cox proportional hazards model). Supplementary table 8. Multivariate analysis of continuous M+2+6- metric at 5% increments as a predictor of overall survival (OS) in cohorts with gene-expression data (Cox proportional hazards model). Supplementary table 9. Univariate and multivariate analysis of continuous M+2+6- metric as a continuous variable at 5% increments as predictor of progression-free survival (PFS) and overall survival (OS) in the GOYA trial cohort (Cox proportional hazards model). Supplementary table 10. Multivariate analysis of M+2+6- metric dichotomized at 15% as a predictor of overall survival (OS) in cohorts with gene-expression data (Cox proportional hazards model). Supplementary table 15. Clinicopathologic characteristics of DLBCL patients evaluated by multiplexed fluorescent immunohistochemistry (mfIHC) in this study. Supplementary table 16. Manual multiplexed fluorescent immunohistochemistry (mfIHC) staining protocol performed on the NUH and CMMC cohort TMA. Supplementary table 17. Automated multiplexed fluorescent immunohistochemistry (mfIHC) staining protocol performed on the SGH, MDA and BCA cohort TMA.

Supplementary table 1. Per-patient mfIHC MYC, BCL2 and BCL6 single oncogene and subpopulation scores for normal tonsil tissue and reactive lymph node tissue. Supplementary table 2. Per-patient mfIHC MYC, BCL2 and BCL6 single oncogene and subpopulation scores for DLBCL tissue (NUH, CMMC, SGH, MDA, BCA and UP). Supplementary table 6. Inferred percentage extents of MYC, BCL2, BCL6 and sub-population metrics in GEP cohorts. Supplementary table 11. Correlation of M+2+6- metric with gene expression in GEP cohorts. Supplementary table 12. Differential gene expression analysis of primary germinal center (GC) B-cells with M+2+ and M+2+6+ overexpression. Supplementary table 13. Differentially expressed genes between M+2+6- and all other malignant cells in scRNA-seq samples of DLBCL. Dichotomized non-parametric comparison, Wilcoxon rank sum test. Supplementary table 14. Analysis of positive enrichment of Wikipathways terms between M+2+6- and all other malignant cells in scRNA-seq samples of DLBCL by gprofiler2.

The NF-κB family comprises five transcription factors (RELA, RELB, C-REL, NF-κB1 (p50) and NF-κB2 (p52)) that form homo- or heterodimers among themselves to regulate gene expression by binding DNA. Here we show that p52 activates transcription without directly binding DNA but as a heterotetrameric complex with ETS1, a transcription factor outside the NF-κB family. By generating a knock-in mouse model (Nfkb2ki/ki) with three mutated residues on p52 required for its interaction with ETS1, but not RELB, we demonstrate that the p52–ETS1 complex regulates the expression of transcription factors OCT1 and OBF1, which are known to be critical for the germinal center program. Consequently, B cell-intrinsic expression of the p52–ETS1 complex was indispensable for splenic germinal center B cell formation and T cell-dependent antibody responses. Functionally, loss of p52–ETS1 interaction led to diminished antigen-specific IgE, thereby protecting mice from allergic responses. Collectively, our findings expand current knowledge of NF-κB signaling and may provide new therapeutic targets for the treatment of allergic diseases. Tergaonkar and colleagues identify a noncanonical interaction between the NF-κB transcription factor family member p52 and the ETS family member ETS1. They find that the p52–ETS1 complex is required for splenic germinal center B cell formation and T cell-dependent antibody responses.

Chemotherapy forms the backbone of treatment for Diffuse Large B Cell Lymphoma (DLBCL); however, approximately 20% of tumors are chemoresistant. Inhibitors of the DNA damage response (DDR) sensitize various tumor types to chemotherapy and show pre-clinical activity in lymphoma. DLBCL with replication stress and genomic instability are particularly sensitive to ATR and WEE1 inhibitors, making DDR inhibition a promising therapeutic avenue to enhance chemosensitivity. We therefore conducted an in-vitro screen to identify optimal chemotherapy-DDR inhibitor (DDRi) combinations in a panel of eight DLBCL cell lines. This screen utilized the Quadratic Phenotypic Optimization Platform (QPOP), an experimental-analytic method designed to identify potent drug combinations. Six DDRis targeting various pathways (ATR, ATM, CHK1/2, DNA-PK, WEE1, PARP) were screened along with six routinely used chemotherapy agents. The ATR inhibitor AZD6738 and replication stress-inducing chemotherapeutic gemcitabine (A+G) emerged as the most effective combination, inducing synergistic cell death via apoptosis across fourteen DLBCL cell lines, including gemcitabine-resistant lines. This efficacy was further demonstrated in vivo, with the A+G combination significantly reducing tumor growth in NSG mouse xenograft models. Contrary to the expected mechanism of the A+G combination causing mitotic catastrophe, flow cytometry revealed that only a small proportion of cells entered mitosis. RNA sequencing of A+G treated cell lines revealed expected suppression of cell-cycle and DNA replication-related pathways. Interestingly, the combination also strongly reversed a poor-prognostic gene expression signature characteristic of dark zone (DZ) biology in the DZ-like and gemcitabine-resistant cell lines HT and SUDHL4. This was accompanied by changes in a variety of chromatin regulation pathways. To further investigate the mechanism underlying this transcriptional shift, we performed chromatin immunoprecipitation sequencing (ChIP-seq) for the enhancer marker H3K27ac. This revealed that A+G treated cells showed a significant loss of super-enhancer marks for BCL6, a master regulatory transcription repressor in DLBCL. BCL6 is a proto-oncogene that represses genes involved in cell cycle arrest and apoptosis, allowing lymphoma cells to survive and proliferate, and its expression was significantly downregulated at the RNA level in A+G treated cells. Additionally, mass-spectrometry cellular thermal shift assays (MS-CETSA) demonstrated that gemcitabine initiated an early destabilization of BCL6 protein, which was sustained by ATR inhibition. However, only the A+G combination led to sustained loss of BCL6 protein levels, explaining improved cell kill with the reversal of the DZ signature. We hypothesize that the replication stress induced by gemcitabine treatment leads to activation of pathways that cause degradation of BCL6, which is then reinforced by transcriptional changes induced by ATR inhibition. In conclusion, we have identified a chemotherapy-DDRi combination, ATR inhibition and gemcitabine, that is highly effective in killing chemoresistant DLBCL cells in vitro and in vivo. The mechanism of synergy involves suppression of a BCL6-regulated transcriptional program driving dark zone biology. Gemcitabine is a routinely used chemotherapeutic for second-line treatment of DLBCL, and neither it nor AZD6738 shows lymphodepletion in humans. As DLBCL with DZ-like signatures show poor outcomes on standard chemo-immunotherapy, our finding that ATR inhibition with gemcitabine reverses DZ signatures suggests the possibility of this being a non-lymphodepleting genotoxic backbone for combinations with T-cell engaging bispecific antibodies.

DNA replication stress (RS) is a widespread phenomenon in carcinogenesis, causing genomic instability and extensive chromatin alterations. DNA damage leads to activation of innate immune signaling, but little is known about transcriptional regulators mediating such signaling upon RS. Using a chemical screen, we identified protein arginine methyltransferase 5 (PRMT5) as a key mediator of RS-dependent induction of interferon-stimulated genes (ISGs). This response is also associated with reactivation of endogenous retroviruses (ERVs). Using quantitative mass spectrometry, we identify proteins with PRMT5-dependent symmetric dimethylarginine (SDMA) modification induced upon RS. Among these, we show that PRMT5 targets and modulates the activity of ZNF326, a zinc finger protein essential for ISG response. Our data demonstrate a role for PRMT5-mediated SDMA in the context of RS-induced transcriptional induction, affecting physiological homeostasis and cancer therapy.

Abstract Diffuse large B-cell lymphoma (DLBCL) is the most common aggressive lymphoma and overexpression of the oncogenes MYC, BCL2 and BCL6 impacts patient outcomes. We demonstrated using single-cell resolved analysis that cells with unique co-expression of high MYC and BCL2 but lacking BCL6 (M+2+6-) were consistently correlated with poor survival compared to other combinations. Here we present a follow-up study evaluating M+2+6- spatial patterns and their relationship to survival, tumour processes, and the immune microenvironment. We measured oncogene co-expression at single-cell resolution through multiplexed fluorescent immunohistochemistry (mfIHC) in four cohorts of DLBCL (n=449). Spatial point patterns were derived from multispectral images, upon which Gibbs modeling was applied. This analysis uncovered a ‘dispersed’ spatial phenotype exhibited by M+2+6- cells, which stratified DLBCL cohorts for survival. Bulk/single-cell transcriptomic analyses of DLBCL samples enriched with the ‘dispersed’ phenotype identified several genes, such as LAG3 and IFI27, that are implicated in migration, cell adhesion, and invasion- suggesting that this spatial phenotype is associated with tumour invasiveness. Single cell analysis revealed unique communication pathways IL10 and SEMA3 between this phenotype and immune cells. IL10 promotes aggressiveness and proliferation in DLBCL and SEMA3 and is a regulator of RAC1 that influences cell migration. Digital Spatial Profiling (DSP) analyses were also performed, applying the protein-based nCounter method (29 immune markers) to 110 DLBCL samples, and the Whole Transcriptome Atlas (WTA) panel (18,000 genes) to CD3+ enriched regions of 47 DLBCL samples. Analysis of patients enriched in the ‘dispersed’ phenotype revealed an immune cold microenvironment, enriched in Tregs and exhausted CD4+ and CD8+ T cells. Taken together, we postulate that the M+2+6- associated ‘dispersed’ spatial phenotype is associated with tumor cell invasiveness and an immune cold microenvironment composition, all contributing towards the poor prognosis associated with this spatial phenotype. Citation Format: Shruti Sridhar, Michal Marek Hoppe, Min Liu, Patrick Jaynes, Yanfen Peng, Sanjay De Mel, Limei Poon, Esther Hian Li Chan, Joanne Lee, Chandramouli Nagarajan, Nicholas F. Grigoropoulos, Soo-Yong Tan, Susan Swee-Shan Hue, Shaoying Li, Joseph D. Khoury, Pedro Farinha, Anja Mottok, David W. Scott, Gayatri Kumar, Kasthuri Kannan, Wee Joo Chng, Yen Lin Chee, Siok-Bian Ng, Claudio Tripodo, Anand D. Jeyasekharan. Single Cell Resolution Spatial Modeling Uncovers Survival-associated Phenotypes in Diffuse Large B Cell Lymphoma [abstract]. In: Proceedings of Frontiers in Cancer Science; 2023 Nov 6-8; Singapore. Philadelphia (PA): AACR; Cancer Res 2024;84(8_Suppl):Abstract nr P21.

Background Malignancies exhibit variable cellular distribution patterns and the relationship between these topographic variations, underlying biological processes, and clinical outcomes remain poorly understood. Point process analyses, widely used in ecology, can elucidate the spatial distribution of points in complex systems but have rarely been applied to tumor heterogeneity. We recently demonstrated (Hoppe et al, Cancer Discovery 2023), that cells co-expressing high MYC and BCL2 but lacking BCL6 (M+2+6-), in Diffuse Large B Cell Lymphoma (DLBCL) are consistently correlated with poor survival compared to other MYC/BCL2/BCL6 combinations. Machine learning approaches can be applied to understand nuances of cellular point patterns and help with correlating with clinicopathological variables. Here, we developed a code frame that can be generalized across tissue regions accounting for heterogeneity to quantitatively study spatial patterns of M+2+6- cells in DLBCL. Methods We developed a scalable automated workflow for generating spatial point patterns. The individual steps of the pipeline have been consolidated into a standalone package that can be executed in a facile manner for any image type, without dependencies on any proprietary software. Using multiplexed fluorescent immunohistochemistry (mfIHC) in four cohorts of DLBCL (n=449), spatial point patterns were derived, and Geyer's point process model was applied. Machine learning classification models were benchmarked for spatial statistics derived from the point process model. A multi-omic analysis, including single-cell transcriptomic analyses of 22 DLBCL samples, was then conducted. Spatial transcriptomic technique, Stereoseq, was also conducted on 2 DLBCL samples. Results The workflow consists of the following parts: 1) The python script using the OpenCV package to manipulate the kernel size and intensity of the spatial coordinates overlaid on the images; 2) A QuPath groovy script to automate the import and export of the images and the parameter thresholds for the pixel classifier; 3) An R script to build the spatial point patterns from coordinates and save different oncogene co-expression as marks within the accurate geojson annotations; and 4) an R script to obtain measures of quality in terms of minimizing the number of points excluded while generating accurate spatial point pattern windows. After applying the pipeline, we see that patients could be divided into two: one group showed “clustered” spatial organization, while the other displayed a “dispersed” M+2+6- cell distribution. We achieved an accuracy of 98% in classifying patients as “dispersed” and “clustered” across four cohorts, through the random forest model. Cases with “dispersed” M+2+6- cells had shorter overall survival across all analyzed cohorts (P < 0.05 in 4/4 cohorts). Patients enriched in the “dispersed” phenotype, predominantly belonged to the ABC cell of origin subtype. We derived a “dispersed” pattern gene signature through multi-omic analyses which expressed genes implicated in cell migration and adhesion. Validation of the dispersed signature was conducted using Stereoseq, where M+2+6- cells enriched in the signature displayed greater values of L function across distances. Patients enriched in the dispersed phenotype displayed lower infiltration of immune subtypes though deconvolution hinting at a possible immunologically cold microenvironment. Conclusion This study demonstrates the clinical relevance studying the spatial distribution of malignant cell subpopulations through point pattern analysis. We anticipate that this machine learning pipeline can be developed for clinical use, enabling the classification of spatial phenotypes in DLBCL biopsies for patient stratification.

Abstract Point process analyses are widely used in ecology and geography to understand the impact and relevance of spatial distribution of points in a complex system. To date, only few studies have explored point process analyses in the context of tumor heterogeneity. Malignancies often show variable patterns of cellular distribution, and the relationship of these topographic variables with underlying biological processes and clinical outcomes is not well understood. Here, we performed spatial point process analyses in diffuse large B-cell lymphoma (DLBCL), leveraging advances in multiplexed immunohistochemistry and cellular phenotyping, with downstream multi-omic and clinicopathological analyses. We have demonstrated that survival in DLBCL is strongly associated with the fraction of malignant cells showing co-expression of the oncogenes MYC and BCL2 in the absence of BCL6 (M+2+6-). These cells display non-random spatial organization within tumor infiltrates, deviating from a poisson distribution. We therefore modeled the spatial organization of M+2+6- cells within DLBCL, using x-y coordinate information from multiplexed fluorescent immunohistochemistry (mfIHC) in four independent cohorts (N=449 patients). We derived spatial point patterns, upon which Geyers point process models were applied. We observed that patients could be divided into two groups based on these: one group showed “clustered” spatial organization, while the other displayed a “dispersed” M+2+6- cell distribution. Cases with predominantly “dispersed” M+2+6- cells showed shorter overall survival in all analyzed cohorts (P < 0.05 in 4/4 cohorts). We performed multi-omic analyses to identify potential biological explanations for this association between topography and clinical outcome. Using digital spatial profiling of the transcriptome from CD20, CD3 and CD68 compartments in 47 DLBCL samples, we observed that lymphomas with an M+2+6- ‘dispersed’ phenotype had an immunologically cold microenvironment enriched in Tregs and exhausted CD4+ and CD8+ T cells. Through an integration with single-cell transcriptomic analyses (N=22 samples) we observed that malignant B-cells from cases with M+2+6- ‘dispersed’ phenotype expressed genes implicated in cell migration and adhesion, but also in potentially targetable immune checkpoints such as LAG3. In summary, by integrating multiplexed immunophenotyping and point process analysis with molecular profiling in the setting of DLBCL, we show that a dispersed pattern of distribution of the M+2+6- subclonal fraction uniquely confers poor prognosis, and associates with a potentially reversible immune cold microenvironment composition. This represents the first demonstration that the spatial distribution of malignant cell subpopulations in lymphoma can embody biological and clinical significance. Citation Format: Shruti Shridar, Michal Marek Hoppe, Patrick Jaynes, Gayatri Kumar, Siddham Jasoria, Ziwei Meng, Vaibhav Rajan, Kasthuri Kannan, Claudio Tripodo, Anand Devaprasath Jeyasekharan. The spatial organization of cells expressing MYC and BCL2 affects immune microenvironment composition and prognosis in DLBCL [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3648.

Primary central nervous system lymphoma (PCNSL) is a rare and aggressive non-Hodgkin lymphoma with a high risk of recurrence, posing significant clinical challenges due to its morbidity. While PD1-based immunotherapy has shown promise in some PCNSL cases, it has been less effective in diffuse large B-cell lymphoma (DLBCL). Understanding the tumor microenvironment factors that distinguish PCNSL from DLBCL and those mediating PCNSL relapse after standard chemotherapy is crucial for integrating immunotherapy into treatment. In this study, we employed digital spatial profiling (DSP), an advanced technique for spatially resolved transcriptomics, to capture the whole transcriptome atlas (WTA) with over 18,000 RNA targets across specific cell types. Using DSP, we profiled the WTA of macrophages, T cells, and B cells in tumor tissue samples from patients with DLBCL (n = 64) and PCNSL (n = 16), encompassing 409 areas of interest (AOIs). Selective collection of UV-cleavable probes from distinct masks, generated by immunofluorescent staining of the morphology markers CD68, CD3, and CD20, accurately captured the WTA of respective cell types in their native tissue environment. We identified many differentially expressed genes (DEGs) significantly upregulated in DLBCL or PCNSL (adjusted P value < 0.05 and |log2FC| > 0.58). In PCNSL tumor cells, immune checkpoints such as LAG-3 and chemokines (CXCL13, CCL3, CCL5) were actively upregulated. Pathway enrichment analysis revealed that inflammatory response, complement, and IL-2/STAT5 signaling were enriched in PCNSL tumor cells, indicating a more pronounced immune response within the tumor microenvironment, potentially contributing to the preferential sensitivity to PD1 therapy. Significant differences were also noted in the macrophage compartment between DLBCL and PCNSL. DEGs such as SPP1 and CD163 were highly expressed in PCNSL macrophages. Metabolic pathways such as hypoxia and glycolysis were enriched in PCNSL macrophages, suggesting that metabolic reprogramming may create a pro-tumorigenic microenvironment that supports tumor progression. Utilizing MoMac-VERSE, the largest single-cell transcriptomic meta-analysis of human monocytes and macrophages, we projected the top DEGs of the CD68 mask and found that PCNSL macrophage DEGs overlapped with the TREM2 macrophage cluster. This cluster has been shown to be associated with a poor prognosis and attenuate responses to PD-1 antibodies in various solid cancers, indicating that this specific macrophage subpopulation in PCNSL may have unique functions in shaping the tumor microenvironment and affecting the response to immune checkpoint inhibitors. In summary, leveraging DSP, we present a comprehensive understanding of the distinct immune landscapes in DLBCL and PCNSL. Our findings highlight the importance of spatial transcriptomics in elucidating the unique characteristics of these lymphoma subtypes and may have implications for novel immunotherapeutic treatment strategies.

Macrophages are abundant immune cells in the microenvironment of diffuse large B-cell lymphoma (DLBCL). Macrophage estimation by immunohistochemistry shows varying prognostic significance across studies in DLBCL, and does not provide a comprehensive analysis of macrophage subtypes. Here, using digital spatial profiling with whole transcriptome analysis of CD68+ cells, we characterize macrophages in distinct spatial niches of reactive lymphoid tissues (RLTs) and DLBCL. We reveal transcriptomic differences between macrophages within RLTs (light zone /dark zone, germinal center/ interfollicular), and between disease states (RLTs/ DLBCL), which we then use to generate six spatially-derived macrophage signatures (MacroSigs). We proceed to interrogate these MacroSigs in macrophage and DLBCL single-cell RNA-sequencing datasets, and in gene-expression data from multiple DLBCL cohorts. We show that specific MacroSigs are associated with cell-of-origin subtypes and overall survival in DLBCL. This study provides a spatially-resolved whole-transcriptome atlas of macrophages in reactive and malignant lymphoid tissues, showing biological and clinical significance.

Background: Cancers often overexpress multiple clinically relevant oncogenes. However, it is not known if multiple oncogenes within a cancer combineuniquely in specific cellular sub-populations to influence clinical outcome.We studied this phenomenon using the prognostically relevantoncogenes MYC, BCL2 and BCL6 in Diffuse Large B-Cell Lymphoma(DLBCL). Methods: Quantitative multispectral imaging simultaneously measured oncogene co-expression at single-cell resolution in reactive lymphoid tissue (n=12)and four independent cohorts (n=409) of DLBCL. Mathematically derived co-expression phenotypes were evaluated in DLBCLs with immunohistochemistry (n=316) and nine DLBCL cohorts with gene expression data (n=3974). Bulk and single-cell RNA sequencing was performed on patient-derived B-cells with induced co-expression of MYC,BCL2 and BCL6. Results: Unlike in non-malignant lymphoid tissue where the co-expression of MYC, BCL2 and BCL6 in a B-cell is limited, DLBCLs show multiple permutations of oncogenic co-expression in malignant B-cells. The percentage of cells with a unique combination MYC+BCL2+BCL6-(M+2+6-) consistently predicts survival in contrast to that of other combinations (including M+2+6+). An estimated percentage of M+2+6-cells can be derived from any quantitative measurement of the component individual oncogenes, and correlates with survival in immunohistochemistry and gene expression datasets. Comparative transcriptomic analysis of DLBCLs and transformed patient-derived B-cells identifies cyclin D2 (CCND2) as a potential BCL6-repressedregulator of proliferation in the M+2+6- population. Conclusions: Unique patterns of oncogene co-expression at single-cell resolution affect clinical outcomes in DLBCL. Similar analyses evaluating oncogenic combinations at the cellular level may impact diagnostics and target discovery in other cancers. Citation Format: Michal M. Hoppe, Patrick Jaynes, Shuangyi Fan, Yanfen Peng, Shruti Sridhar, Phuong Mai Hoang, Xin Liu, Sanjay de Mel, Limei Poon, Esther Chan, Joanne Lee, Choon Kiat Ong, Tiffany Tang, Soon Thye Lim, Chandramouli Nagarajan, Nicholas F. Grigoropoulos, Soo-Yong Tan, Susan Swee-Shan Hue, Sheng-Tsung Chang, Shih-Sung Chuang, Shaoying Li, Joseph D. Khoury, Hyungwon Choi, Pedro Farinha, Anja Mottok, David W. Scott, Carl Harris, Alessia Bottos, Gayatri Kumar, Kasthuri Kannan, Laura J. Gay, Hendrik F. Runge, Ilias Moutsopoulos, Irina Mohorianu, Daniel J. Hodson, Yen-Chee Lin, Wee-Joo Chng, Siok-Bian Ng, Claudio Tripodo, Anand D. Jeyasekharan. Patterns of oncogene co-expression at single cell resolution influence survival in diffuse large B-cell lymphoma. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5557.