GeoMx Digital Spatial Profiling Research Articles

Abstract 1151: Reconstruction of the spatial ecosystem of glioblastoma reveals relationships between tumor cell states and microenvironment

Abstract Glioblastoma multiforme (GBM) is the most aggressive form of primary brain tumor, with no curative treatment options. Multiple studies have characterized at single cell resolution the GBM as being composed of transcriptional cell states interconnected with components in the tumor immune microenvironment (TME). Our group proposed and validated the first single cell guided functional classification of GBM in four tumor-intrinsic cell states which informed clinical outcome and delivered therapeutic options. However, single cell technologies are unable to unravel the spatial relationships among the cell states of GBM and between GBM cell states and TME. Spatially resolved transcriptomic technologies are emerging as powerful tools to reconstruct the spatial architecture of a tissue. We performed spatial transcriptomics of multicellular regions of interest (ROI) in 6 primary IDH wild-type GBM and 2 recurrent GBM with both CosMx Spatial Molecular Imager, which analyzes 1,000 RNA probes and 64 proteins at single cell resolution, and GeoMx Digital Spatial Profiler which profiles the whole transcriptome (~18,000 genes) at ROI resolution. The development of computational tools aimed to integrate spatial proximity and CosMx derived single-cell transcriptomics revealed spatial segregation of the tumor cell clones and cellular states and highlighted recurrent patterns of cell states, distinct TME cell types associated with coherent histopathological features across multiple samples. The development of a spatial informed intercellular communication algorithm and the reconstruction of ligand-receptor-target networks will allow the discovery of tumor cell states-TME cross-talks and the biological signaling regulated by these interactions that are driving the heterogeneity of GBM and therefore potentially therapeutically targetable. Analysis of matched regions of interest profiled by GeoMx and spatial proteomics with CosMx further cross-validated the spatial ecosystem of glioblastoma as reconstructed at single-cell resolution. Our studies established a scalable approach to resolve the transcriptional heterogeneity of GBM and reconstruct the architecture of GBM cell states and tumor microenvironment. Citation Format: Bruno Adabbo, Simona Migliozzi, Luciano Garofano, Fulvio D'Angelo, Pedro Davila, Sakir H. Gultekin, Daniel Bilbao Cortes, Benjamin Currall, Sion L. Williams, Marc Sanson, Franck Bielle, Anna Luisa Di Stefano, Michele Ceccarelli, Anna Lasorella, Antonio Iavarone. Reconstruction of the spatial ecosystem of glioblastoma reveals relationships between tumor cell states and microenvironment [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 1151.

Abstract 1388: Spatial transcriptome analysis of EMT trait and corresponding immune composition study in HNSCC

Abstract Head and neck squamous cell carcinoma (HNSCC) is a prevalent malignant cancer worldwide. Conventional treatments for this type of cancer involve surgical resection, radiation therapy, and chemotherapy. However, in cases where the tumor recurs or metastasizes distantly, rendering surgical resection or radiation therapy impractical, the only viable option is the utilization of targeted drugs in conjunction with chemotherapy. Despite such interventions, the prognosis remains grim. Hence, the identification of genetic characteristics or biomarkers associated with easy metastasis or recurrence in the early stages of the tumor has become imperative for effective head and neck cancer treatment. The objective of this study is to analyze tumor specimens from the primary site and lymph node metastasis post-surgical resection in HNSCC patients. We employed the GeoMx Digital Spatial Profiler (DSP) for imaging-based digitized gene expression analysis. The DSP provides gene expression information at specific locations we selected, such as the primary tumor center core, primary tumor invasion borders, and lymph node metastasis specimens. The GeoMx DSP spatial analysis tool illustrates the cancer transcriptome atlas (CTA), encompassing the expression of 1825 genes related to cancer and immunity from the primary site to the metastatic site. We identified tissue-specific EMT gene expression patterns in paired tumor regions of interest (ROIs). ITGB1 emerged as the most significant co-expression marker in the primary outer tumor and metastatic tumor. Distinct EMT signature genes expressed in the HNSCC primary site and lymph node metastatic site were also discovered. Additionally, we conducted CTA profiling in correlated stroma ROIs. Through myeloid activity analysis in HNSCC paired samples, we identified distinct myeloid activation signatures in primary and metastatic site stroma cells. These results will provide crucial insights into head and neck cancer treatment strategies based on different tumor site characteristics. Citation Format: Chih-Hung Chung, Muh-Hwa Yang. Spatial transcriptome analysis of EMT trait and corresponding immune composition study in HNSCC [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 1388.

Abstract 5481: Spatial multiplex proteins profiling of tumor microenvironment to decipher de novo metastasis in nasopharyngeal carcinoma

Abstract Background: Nasopharyngeal carcinoma (NPC) has a high incidence rate in Southeast Asia, where more than 90% of NPC cases are associated with Epstein-Barr virus (EBV). Although NPC patients are generally sensitive to radiotherapy and chemotherapy, about 20-30% of patients have treatment failure due to the development of distant metastasis, i.e. de novo metastasis. In this study, we utilized GeoMx Digital Spatial Profiler (DSP), a multiplex and high-throughput sequencing platform designed for spatially- and morphologically resolved protein expression profiling to investigate the intra-tumoral and inter-tumoral heterogeneity of tumor microenvironment (TME) and identify the mechanisms underlying distant failure in NPC. Methods: We utilized GeoMx Protein Assays consisting of 147 proteins from 15 functional modules to compare the difference between NPC cases developing distant failure (DF) (n=21) and stage-, age-, gender- and treatment-matched NPC cases without developing distant failure (noDF) (n=15) within five years of treatment. Two fluorescent morphological markers including PanCK and CD45, were used for delineating the structural architecture of the tissues, identifying cell types, and selecting regions of interest (ROIs), followed by sequencing for protein expression quantification. NanoString GeoMx DSP Analysis Suite was used for quality control and data normalization. The NPC TME heterogeneity was further investigated by the 10X Visum platform for spatial transcriptome. The public single-cell RNA sequencing (scRNA-Seq) and bulk RNA sequencing datasets were obtained for the downstream analyses. Results: By analyzing protein expression profiles, the study revealed the intratumoral and intertumoral heterogeneity associated with EBV latent infection in the TME of NPC. Regions with high EBV latent gene expression exhibited a noticeable reduction in T cell infiltration, indicating an "immune cold" phenotype, whereas regions with low EBV latent gene expression displayed heavy T cell infiltration, indicating an "immune hot" phenotype. Moreover, the study identified a panel of autophagic and anti-apoptotic proteins that were significantly upregulated in the cancer cells of NPC cases developing distant failure after treatment. This observation, supported by the single-cell RNA sequencing datasets, further implicated EBV latent infection for the upregulation of autophagic and anti-apoptotic genes. Notably, the study also demonstrated that the genes upregulated in the immune cells in NPC cases with distant failure were associated with poor progression-free survival in an independent NPC cohort. Conclusions: These findings shed light on the crucial role of EBV latent infection in TME heterogeneity and highlight the potential involvement of autophagy and anti-apoptosis in de novo metastasis in NPC. Citation Format: Larry Ka-Yue Chow, Ying Wang, Lanqi Gong, Siqi Xu, Zhaozheng Hou, Xinyuan Guan, Dora Lai-Wan Kwong, Victor Ho-Fun Lee, Wai To ng, Anne Wing-Mui Lee, Zhonghua Liu, Yunfei Xia, Wei Dai. Spatial multiplex proteins profiling of tumor microenvironment to decipher de novo metastasis in nasopharyngeal carcinoma [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 5481.

Abstract 5072: Multi-omic, multi-scale characterisation of colorectal cancer defines spatiotemporal patterns of recurrence

Abstract Aim: Patients who undergo intended curative resection of colorectal cancer (CRC) have a 20-25% chance of metachronous recurrence, the site and timing of which are unpredictable and resistant to current treatment. The biological basis for such heterogeneous disease behavior remains to be elucidated. Bulk transcriptomic profiling and subsequently single-cell RNAseq have provided insight in the epithelial subtypes and immune microenvironment. Recently spatially resolved transcriptomic assessment allow molecular profiling of tissue while preserving tissue architecture. Utilizing ST technology we sought to perform deep characterization of a large cohort of patients with primary CRC with the intention to decipher the biological determinants of spatio-temporal patterns of recurrence. Methods: 750 patients who underwent resection of CRC with mature follow up were studied. Bulk transcriptomic, genomic and multiplex immune characterization was performed using a TMA format. Of these, 28 patients were assessed using single cell spatial transcriptomics (Nanostring CosMx Spatial Molecular Imager (SMI, 1000plex gene panel)), 120 tumors underwent regional whole transcriptome profiling of epithelium and TME (Nanostring GeoMx Digital Spatial Profiler using FFPE tissue). We used image analysis and bioinformatics to integrate these complex datasets in over 120000 single-cells in the context of their spatial tissue architecture and clinicopatholgical outcome and recurrence data. Results: Using the CosMx SMI we characterized cells with complete topographic detail and defined 2 unique epithelial cell states defined. Each epithelial state had distinct spatial properties including cell size and morphology, average distance to nearest lymphocyte, neighboring cell types and stromal neighborhood. This epithelial signature was integrated in GeoMx samples and were found to predict recurrence (p<0.05). The Epithelial, Fibroblast and Immune GeoMx transcriptomic compartments were expanded and grouped using unsupervised clustering. Each compartment demonstrated groups of patients where specific spatially derived signatures could predict site and time of recurrence (p<0.05). Conclusions: By maintaining the tissue structure, we have directly measured cellular interactions and captured cells commonly missed during dissociative studies whilst defining novel molecular subtypes of CRC with clinical relevance. Novel platforms such as CosMx allow deeper characterization of unique cell types and cellular interactions that may pave the way for novel therapeutics and precision medicine for patients with CRC. Citation Format: Colin Stuart Wood, Joao Da Silva Filho, Andrew Cameron, Assya Legrini, Holly Leslie, Tengyu Zhang, Yoana Doncheva, Claire Kennedy-Dietrich, Matthias Marti, Joanne Edwards, Paul Horgan, Campbell Roxburgh, Colin Steele, Nigel Jamieson. Multi-omic, multi-scale characterisation of colorectal cancer defines spatiotemporal patterns of recurrence [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 5072.

Abstract 4738: Overcoming therapy resistance in rhabdomyosarcoma by targeting the PI3K-MAPK pathway

Abstract Rhabdomyosarcoma (RMS) is a devastating pediatric cancer. There has not been any significant change in the treatment options for the last 4 decades, especially for relapsed or advanced disease. There is an urgent need to identify alternative treatment options to improve survival outcomes of RMS patients. RMS is comprised of two major subtypes as defined by the presence or absence of PAX3/PAX7-FOXO1 gene fusion. Fusion-negative RMS (FN RMS) is characterized by alterations in the RAS-PI3K-MAPK pathway in >90% of the cases. However, the role of the PI3K-MAPK pathway in therapy resistance and disease relapse remains poorly characterized. We have generated two FN-RMS lines, RD and SMS-CTR, with resistance to the standard-of-care chemotherapeutic agent, vincristine, through long-term culturing in incrementally increased concentrations of the drug. By RNA sequencing, vincristine-resistant RD and SMS-CTR cells demonstrated increased expression of genes in the RAS-PI3K-MAPK pathway compared to the parental lines. We hypothesize that increased activity of the PI3K-MAPK pathway contributes to therapy resistance in FN RMS. To test the hypothesis, we first showed that targeted disruption of PIK3CA by shRNA inhibited the cell growth of vincristine-resistant FN RMS cells. The same cells also showed increased sensitivity to the MEK inhibitor, trametinib, as well as the PIK3Cα inhibitor, alpelisib, compared to the parental cells. We then showed that zebrafish FN-RMS tumors treated with the combination of vincristine and alpelisib significantly inhibited tumor growth compared to the treatment with each agent alone. Through integrative analysis of RNA sequencing and ATAC-seq data, we identified MYOD1 and NFATC2 as among the top candidate upstream regulators. Targeted disruption of either MYOD1 or NFATC2 by CRISPR/Cas9 reduced cell growth of vincristine-resistant FN RMS cells and altered expression of the genes in the PI3K-MAPK pathway, implicating MYOD1 and NFATC2 in driving abnormal activity of this pathway. To assess the relevance of our findings in the clinical settings, we showed that selective components in the PI3K-MAPK pathways are up-regulated in 3 matched primary-recurrence/metastasis paraffinized tissue samples of human FN RMS by the NanoString GeoMx Digital Spatial Profiler (DSP) platform. Our findings indicate that targeting the PI3K-MAPK pathway is a promising targeted therapy option for overcoming therapy resistance in RMS. Citation Format: Yadong Wang, Eleanor Chen. Overcoming therapy resistance in rhabdomyosarcoma by targeting the PI3K-MAPK pathway [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 4738.

Abstract 4172: Pathology supervised approaches for ROI placement and segmentation to overcome caveats and pitfalls in high-plex spatial profiling

Abstract Background: Tissue-based high-plex platforms evaluate numerous biomarkers in limited tissue sections. Certain tissue features and assay technical limitations may jeopardize the accuracy of this analysis. In this study, we aim to determine tissue specific features that influence image-based region of interest (ROI) selection and segmentation for spatial profiling of carcinomas, and to develop pathology approaches to overcome them. Methodology: A total of 317 samples from different carcinoma types processed with the GeoMx Digital spatial profiling protein assay protocol were assessed to identify deviations from original strategies for ROI placement and segmentation (Table 1). ROI were selected by a pathologist. The initial strategy for all cases was to place rectangles ROIs of 660*785 µm in tumor area and to segment based on morphology biomarkers as indicated in Table 1. Strategies to overcome technical and tissue specific features were recorded and quantified. Results: A total of 2068 ROIs and 4422 AOIs were evaluated. In all the cases, tissue displayed specific features that limited planned ROI placement and segmentation. Table 1 describes these features by tissue type and profiling strategies. A total of 970 ROIs were solved using different annotation type or changes in segmentation strategy, 584 bypassed the limitations by changing the processing protocol; and 322 were not solved due to non-compliance with assay requirements. Conclusions: We identified tissue features that limit the planned ROI selection and segmentation in spatial high-plex profiling, mainly due to tissue processing, distinct histology, type of sample, and presence of adjacent non-neoplastic tissue. A deep understanding of tumor architecture, biology and platform related technical limitations is needed to select areas of analysis for spatial profiling. Table 1. ROI and segmentation strategy methodology and tissue specific features with pathology appro Histological type (Profiling segments) Anal Carcinoma (Tu/TME) Lung NSCLC (Tu/TME) Lung NSCLC TMA (Tu/Tcells/Mϕ) Lung NSCLC TMA (Tu/TME) CRC MT Tu/Immune Breast Ca Tu/Immune/else Pathology Approach (ROIs) Methodology Samples 48 33 80 80 6 70 Biopsy type Incisional (Core/other), Whole tissue Incisional (Core/other)Whole tissue TMA TMA Whole tissue Core Morphology markers/mIF Channels Syto13: FITC/525 PanCK: Cy3/568 CD45: Texas red/615 CD68: Cy5/666 Syto13: FITC/525PanCK: Cy3/568CD20: Texas red/615CD3: Cy5/666 Syto13: FITC/525 PanCK: Cy3/568 CD68: Texas red/615 CD3: Cy5/666 Syto13: FITC/525 PanCK: Cy3/568 CD68: Texas red/615 CD3: Cy5/666 Syto13: FITC/525 PanCK: CD45: Texas red/615 Syto13: FITC/525 PanCK: Cy3/568 CD45: Texas red/615 Strategy for segmentation Tumor (PanCK+) TME (PanCK-) Tumor (PanCK+)TME (PanCK-) Tumor (PanCK+) T cell (CD3+) Mϕ (CD68+) Tumor (PanCK+) TME (PanCK-) Tumor (PanCK+) Immune (CD45+) Tumor (PanCK+) Immune (CD45+) Else (PanCK-CD45-) Total ROIs/AOIs 224/435 144/252 80/218 82/162 944/1573 594/1782 Tissue specific features (ROI number/percentage) Folds (ROIs) 10 1 0 0 9 5 Draw Polygons (25) Adjacent* Non-Tumoral tissue 22 7 9 7 388 40 Draw Polygons (473) Tissue Derived Autofluorescence 3 105 0 1 0 45 Draw Polygons (13). New segment to exclude autofluorescence (141) Sample Size 26 19 1 2 5 138 Polygons (191) Necrosis 13 0 1 1 217 6 Polygons (238) Biological absence of morphology marker 10 0 0 0 0 5 Polygons non segmented for Tu and TME (15) Solid tumors with scant TME 0 21 0 0 0 9 Polygons non segmented for Tu and TME (30) Eosin autofluorescence 0 0 0 0 0 584 Stained new section with Validated PanCK in Cy5/666 channel (584) Failure to reach the minimum number of cell requirements 13 36 22 2 249 43 Excluded segments (322) New scan to select more areas (43) Citation Format: Sharia Hernandez, Claudio J. Arrechedera, Pedro Rocha, Larisa Kostousov, Sean Barnes, Khan Khaja, Luisa M. Solis-Soto. Pathology supervised approaches for ROI placement and segmentation to overcome caveats and pitfalls in high-plex spatial profiling [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 4172.

Abstract 1154: Spatial profiling of the tumor microenvironment in head and neck squamous cell carcinoma revealed immune cell type infiltration into the TME as a predictor of immunotherapy response

Abstract Background: Immunotherapy has shown promising results in advanced head and neck cancer patients, however, only a subset of patients are responsive to immune checkpoint blockade therapy. The molecular and cellular characterization of the tumor microenvironment (TME) may reveal clues as to why some head and neck tumors are responsive or resistant to the therapy. Methods: Formalin-fixed paraffin-embedded (FFPE) tissue samples were collected from 34 patients at two major Queensland Hospitals with advanced head and neck squamous cell carcinoma (HNSCC), prior to immunotherapy. To investigate the cellular and molecular characteristics of the HNSCC TME, we used the Nanostring GeoMx Digital Spatial Profiler (DSP) spanning modules in an immune-oncology panel, which included immune cell typing and profiling, cell death, PI3K/AKT signaling, drug target, immune activation, and pan tumor protein biomarkers. A deeper structural analysis was performed using the Nanostring whole transcriptome atlas (WTA) panel of tertiary lymphoid structures (TLSs) and germinal centers (GCs). Results: Out of 34 patients, 3 had a complete response (CR), 7 had a partial response (PR), 5 had stable disease (SD), and 19 had progressive disease (PD), based on the Response Evaluation Criteria in Solid Tumors (RECIST). When we compared patient responders (patients with CR, PR, and SD) to patient non-responders (patients with PD), we found that the former had a higher infiltration of immune cell type markers representatives of CD8+ T cells, dendritic cells, and macrophages, into their tumor microenvironment than the latter. Then, to investigate differentially expressed genes between patients with complete response and patients with partial response to immunotherapy, we performed DE analyses and we found that patients with PR had higher Ki-67 expression than those with CR. Spatial transcriptomic mapping of TLS’s (peripheral, intra-tumoral) and normal germinal centers revealed interferon pathway enrichment in the TLS’s. In a subset, these appear to be linked with HPV viral infection. Conclusion: In this study, we found that immune cell type infiltration into the tumor microenvironment of head and neck cancers could be predictive of response to immunotherapy. Furthermore, it was revealed that patients with a high expression of Ki-67 might be less likely to respond completely to immunotherapy. Spatially resolved whole transcriptome analysis of HNSCC tissues revealed the presence of tertiary lymphoid structures with a higher enrichment of interferon pathway genes. Citation Format: Habib Sadeghirad, Chin Wee Tan, Ning Liu, James Monkman, Caroline Cooper, Ken O’Byrne, Melissa Davis, Brett Hughes, Arutha Kulasinghe. Spatial profiling of the tumor microenvironment in head and neck squamous cell carcinoma revealed immune cell type infiltration into the TME as a predictor of immunotherapy response [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 1154.

Abstract 3652: Spatial transcriptomic study of the tumor microenvironment in HNSCC

Abstract Background: The complex and dynamic nature of the tumor-immune microenvironment (TME) presents challenges for identification of robust and predictive biomarkers in immuno-oncology (IO). Standard multiplex immunohistochemistry (mIHC) or multiplex immuno-fluorescent assays (mIF) allow for the phenotyping on the cellular level but are limited by the number of markers available and the detection systems. Gene expression or next generation sequencing (NGS) platforms provide high-plex bulk data sets about the TME, but the spatial context of the tumor and immune cell interactions is lost. The NanoString GeoMx® Digital Spatial Profiler (DSP) combines spatial context, utilizing fluorescent morphological markers, with molecular profiling capabilities. This combination allows for the quantitative analysis of high-plex analyte abundance which can be traced back to a region of interest within a sample, providing biological inference in the region of interest. Methods: Formalin-fixed paraffin-embedded (FFPE) specimens from HNSCC patients were cut in 5µm sections for all analyses. Whole section tissue analysis was previously performed utilizing the NanoString nCounter® PanCancer IO 360™ Panel. Tumor inflammation signatures (TIS), an 18-gene signature measuring the adaptive immune response within tumors, was generated for each specimen and categorized each as immune-hot or immune-cold based on predefined parameters. Four specimens were selected based on their TIS score and two sets of cut sections were processed by either Labcorp lab in Morrisville, NC or Nanostring lab in Seattle, WA for precision evaluation, using the GeoMx Human Whole Transcriptome Atlas (WTA) panel (18,000 genes) with the NanoString GeoMx digital spatial profiling technology. The TME regions of interest (ROI) were identified using a set of morphology markers consisting of panCK, CD45 and Syto13; 12 ROIs were selected from each sectioned tumor material. The same ROIs were selected at the two testing sites with matching tissue location and morphology characters. Results: GeoMx WTA panel provided an unbiased view of 18,000 protein-coding genes at specific TME regions. These counts were successfully mapped to the 12 ROIs on the tissue. A pseudo-TIS score was generated using the same 18 gene signature from the WTA panel for each ROI. The pseudo-TIS scores from ROIs from the four specimens showed consistency with the specimen-wise TIS score, yet variations were observed across different TME/ROI within each specimen. Whole transcriptome expression results showed high correlation between the two testing sites, providing confidence in the reproducibility of the technology. High resolution fluorescent staining images were able to be fed to downstream quantitative digital analysis to generate complementary spatial data. Citation Format: Caitlin Schroyer, Shanshan Zhou. Spatial transcriptomic study of the tumor microenvironment in HNSCC [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 3652.

Abstract 5486: Spatial profiling reveals unique immune-transcriptomic features of tertiary lymphoid structures in melanoma

Abstract Background:Previous studies have shown that intratumoral immune aggregates called tertiary lymphoid structures (TLS) are associated with exceptional response to immune checkpoint blockade (ICB) in several cancer types. However, the specific mechanisms through which TLS contribute to ICB-response remain poorly characterized in part due to technical limitations with isolating TLS. Here, we use the NanoString GeoMx digital spatial profiler to interrogate TLS with deep spatial and transcriptional resolution and identify cellular and molecular features of response-associated TLS in melanoma. Methods:Twelve neoadjuvant ICB-treated high-risk resectable melanoma patients [NCT02519322] with formalin-fixed paraffin embedded (FFPE) surgical resections were included in this study. Major pathologic response (MPR) defined as <10% viable tumor at time of surgery was used to evaluate response (R – responder; NR – non-responder). TLS were sampled (via H&E) using 1mm core punches and assembled into a tissue microarray (TMA). The TLS TMA was probed using the human whole transcriptome atlas (huWTA) and the human T-cell receptor (huTCR) atlas. Melanoma (S100B) and immune markers (CD45 and CD20) were used to guide region-of-interest (ROI) selection of classical TLS. Differential gene expression was calculated using a linear mixed model and corrected with Benjamini-Hochberg. Immune deconvolution was performed using CIBERSORTx with FDR cutoff of < 0.25. Results:Fifty percent (6/12) of patients included in this study demonstrated MPR following treatment with neoadjuvant ICB. Principal component analysis of huWTA probes revealed distinct transcriptional states of R TLS vs NR TLS, with PC1 strongly correlating with response status (r=0.88, p<0.0001). Differential immune analysis showed that R TLS harbored more CD8+ T-cells (p<0.001), NK cells (p=0.004), γδ T-cells (p=0.017) and memory B-cells (p=0.026) and preferentially upregulated inflammatory genes such as pattern recognition receptors (TLR8 and KLRC2) and immune-stimulatory markers (CD27 and CD103). Immune-repertoire analysis revealed a trend towards increased TCR diversity in R TLS (p = 0.3) and distinct TCR signatures in R TLS underscored by elevated expression of TRDJ4 (p<0.001), TRGJ1/2(p<0.001), TRGV4(p<0.001) and TRAV1-2(p=0.018) genes amongst others. Conclusions:These findings demonstrate that, although melanomas from both R and NR patients contain TLS, the immune-phenotype of R TLS is more cytotoxic and characterized by a more diverse TLS-resident T-cell population. These findings present new insights into the therapeutic underpinnings of TLS in cancer and will inform future investigations evaluating their formation/function in solid tumors. Citation Format: Manoj Chelvanambi, Brenda Melendez, Elise F. Nassif, Rossana N. Lazcano, Matthew J. Lastrapes, Bharat S. Bhushan, Sarah B. Johnson, Khalida Wani, Davis R. Ingram, Y David Seo, Beth A. Helmink, Michael G. White, Russell G. Witt, Laurence P. Diggs, Golnaz Morad, Monika Zelazowska, Somnath Paul, Florentia Dimitriou, Ashish V. Damania, Matthew C. Wong, Neeta Somaiah, Nadim J. Ajami, Emily Z. Keung, Wolf H. Fridman, James P. Allison, Padmanee Sharma, Kevin M. McBride, Tina Cascone, Christina L. Roland, Alexander J. Lazar, Jennifer A. Wargo. Spatial profiling reveals unique immune-transcriptomic features of tertiary lymphoid structures in melanoma [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 5486.

Abstract 1134: Dissecting pancreatic cancer tumor-immune microenvironment crosstalk using spatial transcriptomics

Abstract Background: Pancreatic ductal adenocarcinoma (PDAC) often fails to respond to immune therapies due to various factors, including the role of Epithelial to Mesenchymal Transition (EMT) plasticity in conferring broad resistance to diverse therapies. However, the relationship between cancer cell heterogeneity and the tumor immune microenvironment remains unclear. To address this, we utilized single cell RNA-seq and spatial transcriptomics to uncover the landscape of these cell-cell interactions in human PDAC. Here, we applied to specimens from a clinical trial testing Losartan, an indirect TGF-beta inhibitor with FOLFIRINOX chemotherapy and nivolumab (anti-PD1) in a randomized multi-institutional clinical trial for PDAC (NCT03563248). Methods: Using the NanoString GeoMx Digital Spatial Profiler, we selected multiple regions of interest in formalin-fixed paraffin-embedded (FFPE) human PDAC specimens. Immunofluorescent antibody-guided isolation of RNA and protein from cancer cells (pan-cytokeratin), cancer-associated fibroblasts (alpha-SMA), and immune cells (CD45) were performed. Utilizing the whole transcriptome assay (WTA; 18,000+ protein-coding genes) and a new IO Proteome Atlas (IPA; 500+ plex proteins), we ventured to understand the relationship between tumor cells and the surrounding microenvironment. Results: PDAC cells, CAFs, and immune cells were successfully characterized using NanoString GeoMx in clinical trial specimens. Analysis revealed associations between cancer cell plasticity, TGF-beta signaling mesenchymal subtypes in PDAC cells. These TGF-beta enriched mesenchymal PDAC cells were enriched in Arm 1 (FOLFIRINOX) compared to Arm 2 (FOLFIRINOX+losartan). Immune deconvolution analysis identified variations in immune infiltrates, showing an anti-correlation between macrophages and T-cells. The IPA analysis revealed lower MHC I protein levels in short compared to long term survivors in Arm 3. Conclusions: Spatial transcriptomics and proteomics reveal insights into the spatial relationship between PDAC tumor cell EMT plasticity, CAFs, and immune infiltrates. This enables the discovery of novel immune response biomarkers and potential therapeutic avenues to target tumor and microenvironment interactions. Trial Registration: DF/HCC protocol 18-179: Losartan and Nivolumab in Combination With FOLFIRINOX and SBRT in Localized Pancreatic Cancer. NCT03563248 Ethics Approval: All studies presented were approved by the Dana-Farber/Harvard Cancer Center IRB protocols 18-179. Citation Format: Amaya Pankaj, Michael J. Raabe, Bidish K. Patel, Evan R. Lang, Joshua R. Kocher, Katherine H. Xu, Linda T. Nieman, William L. Hwang, Alec C. Kimmelman, David P. Ryan, Theodore S. Hong, Martin J. Aryee, David T. Ting. Dissecting pancreatic cancer tumor-immune microenvironment crosstalk using spatial transcriptomics [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 1134.

Abstract 1153: Using spatial transcriptomics to dissect cell to cell cooperation in lung adenocarcinoma

Abstract Lung cancer is the second most common cancer in the United States and the leading cause of cancer related deaths in this country. Unfortunately, efficacious treatments for lung cancer remain suboptimal. Lung cancer is characterized by several distinct subtypes of which lung adenocarcinoma is the most common, comprising about 40% of this malignancy. The hallmark gene mutations of lung adenocarcinoma include TP53, RAS, and STK11. LKB1 is a serine-threonine kinase (coded by the gene STK11) that largely functions as a tumor suppressor, and is mutated in 20-30% of non-small cell lung cancers (NSCLCs). The diversity between and within lung cancer subtypes, as well as within a patient (i.e. metastases vs primary tumor), makes treating this cancer very challenging. Lung adenocarcinoma, in particular, has collective invasion packs of cells adjacent to the primary tumor that correlate with metastatic disease in mouse models. We hypothesize that the transcriptomic profile of the collective invasion packs in lung adenocarcinoma patients varies significantly from the adjacent primary tumor and represents a targetable metastatic sub-population. This work will help to identify specific cell signaling pathways that have the translational potential to develop novel therapeutics for metastatic disease, ultimately improving patient outcomes through precision medicine. Utilizing patient lung adenocarcinoma samples with KRAS and KRAS+LKB1 mutations, we identified regions of interest including bulk tumor and surrounding invasion packs. Then, using GeoMx digital spatial profiling technology (by Nanostring) and next-gen sequencing, we generated transcriptomic profiles from bulk tumor and invasion packs. Preliminary results indicate region specific transcriptomic differences, highlighting the heterogeneity of these cell populations. More specifically, collective invasion packs exhibit upregulation of gene networks involving immune cell differentiation, function, oxidative phosphorylation, tumor invasiveness, and mitochondrial structure. To discern the metastatic potential, transcriptomic profiles and biological functions will be compared between invasion packs, tumor bulk vs invasion packs, and finally, inter-patient differences. Successful completion of this project will characterize transcriptomes of metastatic lung cancer and has the potential to ultimately identify biomarkers of aggressive disease. Lastly, this foundational study will pave the way for future studies on the transcriptomic landscape of metastatic pediatric cancers and cancer predisposition syndromes. Citation Format: Raehannah Jamshidi, Lyra Griffiths, Rich Johnston, Vaunita Parihar, Frank Schneider, Adam Marcus. Using spatial transcriptomics to dissect cell to cell cooperation in lung adenocarcinoma [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 1153.

Potential Role of Lymphocyte CD44 in Determining Treatment Selection Between Stereotactic Body Radiation Therapy and Surgery for Early-Stage Non-Small Cell Lung Cancer

Background: Stereotactic body radiotherapy (SBRT) versus surgery for operable early-stage non-small-cell lung cancer (ES-NSCLC) remains highly debated. Herein, we utilized spatial proteomics to identify whether any molecular biomarker(s) associate with the efficacy of either modality, in efforts to optimize treatment selection between surgery or SBRT for this population. Methods: We evaluated biopsy tissue samples from 44 ES-NSCLC patients treated with first-line SBRT (cohort 1) by GeoMx Digital Spatial Profiling (DSP) with a panel of 70 proteins in five spatial molecular compartments: tumor (panCK+), leukocyte (CD45+), lymphocyte (CD3+), macrophage (CD68+), and stroma (α-SMA+). To validate the findings in cohort 1, biopsy samples from 52 ES-NSCLC patients who received SBRT (cohort 2) and 62 ES-NSCLC patients who underwent surgery (cohort 3) were collected and analyzed by multiplex immunofluorescence (mIF). Results: In cohort 1, higher CD44 expression in the lymphocyte compartment was associated with poorer recurrence-free survival (RFS) (DSP: P<0.001; mIF: P<0.001) and higher recurrence rate (DSP: P=0.001; mIF: P=0.004). mIF data from cohort 2 validated these findings (P<0.05 for all). From cohort 3, higher lymphocyte CD44 predicted higher RFS after surgery (P=0.003). Inter-modality comparisons demonstrated that SBRT was associated with significantly higher RFS over surgery in CD44-low patients (P<0.001), but surgery was superior to SBRT in CD44-high cases (P=0.016). Conclusions: Lymphocyte CD44 may not only be a predictor of SBRT efficacy in this population, but also an important biomarker (pending validation by large prospective data) that could better sharpen selection for SBRT vs. surgery in ES-NSCLC.

Abstract B041: Modeling the malignant transformation of fallopian tube epithelium driven by extracellular vesicles cargos in an organ-on-chip microphysiological system

Abstract Epithelial ovarian cancer (EOC) remains the leading cause of death from gynecological malignancy and the fifth leading cause of all cancer-related deaths among American women. The most common and lethal subtype, high-grade serous ovarian cancer (HGSOC) tends to originate from the fallopian tube. Precursor lesions of HGSOC such as secretory cell expansion (SCE), the loss of PAX2, PTEN, p53 mutation and serous tubal intraepithelial carcinoma (STIC) are often found at the fallopian tube epithelium (FTE). However, the mechanism of how neoplastic lesions disseminate in FTE and eventually colonize the ovary is unclear. We are interested in understanding if ovarian cancer initiation in the FTE is mediated by extracellular vesicles (EVs), which are nanosized biovesicles secreted by all living cells. EVs serve as vehicles for nucleic acids, proteins, and lipids. Cancer cells commonly secrete more EVs than healthy cells and shift their EV cargos to promote tumorigenesis. The alteration of molecular content of tumor-derived EV can potentially indicate disease states and serve as a new class of cancer biomarkers. Hence, our hypothesis is that EV cargos secreted by tumor cells contain factors promoting preneoplastic lesions in the FTE. Our goal is to interrogate the role of EVs in the initiation of ovarian cancer in the FTE at the early stage, for which there is no adequate in vitro model. We created a microfluidic model using a 3D dynamic culture system termed PREDICT-MOS to enable the direct interaction of hFTE with EVs and the long-term culture of the fallopian tube explant. We are interested to 1) test the role of EVs to drive preneoplastic transformation of hFTE progenitor cells and 2) to test whether tumor derived-EVs shift EV proteins cargos secreted by hFTE. The hFTE cultured on PREDICT-MOS is directly exposed to OVCAR3-derived EVs and embedded for transcriptomic profiling utilizing GeoMx Digital Spatial Profiler to identify differentially expressed genes in both ciliated and secretory cells within the epithelium. In parallel, we are characterizing the hFTE EV protein content with or without OVCAR3 EV treatment to detect the alteration in hFTE cargos using mass spectrometry proteomics. Our preliminary transcriptomic data from 8 patient fallopian tube samples demonstrates the deregulation of pathways associated with inflammation (TNF and NF-kappa B signaling), immune suppression (PD-1 checkpoint signaling) and epithelium morphogenesis. In addition, we have identified 295 proteins as the results of the first proteomic characterization of hFTE-derived EV proteins, which can serve as a “baseline” to evaluate whether the fallopian tube shifts its sEV cargo during ovarian cancer carcinogenesis. Further validation of transcriptomic findings on translational level using immunohistochemistry and identification of OVCAR3 EV protein cargos that induce the transcriptomic changes are needed to establish the role of EVs in the initiation of HGSOC in the fallopian tube. This project is supported by the Ovarian Cancer Research Fund Alliance Program Grant. Citation Format: Didi Zha, Jared Sipes, Sagar Rayamajhi, Leonidas E. Bantis, Angela Russo, Harsh Pathak, Andrew K. Godwin, Joanna E. Burdette. Modeling the malignant transformation of fallopian tube epithelium driven by extracellular vesicles cargos in an organ-on-chip microphysiological system [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2):Abstract nr B041.

Digital spatial profiling of segmental outflow regions in trabecular meshwork reveals a role for ADAM15.

In this study we used a spatial transcriptomics approach to identify genes specifically associated with either high or low outflow regions in the trabecular meshwork (TM) that could potentially affect aqueous humor outflow in vivo. High and low outflow regions were identified and isolated from organ cultured human anterior segments perfused with fluorescently-labeled 200 nm FluoSpheres. The NanoString GeoMx Digital Spatial Profiler (DSP) platform was then used to identified genes in the paraffin embedded tissue sections from within those regions. These transcriptome analyses revealed that 16 genes were statistically upregulated in high outflow regions and 57 genes were statistically downregulated in high outflow regions when compared to low outflow regions. Gene ontology enrichment analysis indicated that the top three biological categories of these differentially expressed genes were ECM/cell adhesion, signal transduction, and transcription. The ECM/cell adhesion genes that showed the largest differential expression (Log2FC ±1.5) were ADAM15, BGN, LDB3, and CRKL. ADAM15, which is a metalloproteinase that can bind integrins, was upregulated in high outflow regions, while the proteoglycan BGN and two genes associated with integrin signaling (LDB3, and CRKL) were downregulated. Immunolabeling studies supported the differential expression of ADAM15 and showed that it was specifically upregulated in high outflow regions along the inner wall of Schlemm's canal and in the juxtacanalicular (JCT) region of the TM. In addition to these genes, the studies showed that genes for decorin, a small leucine-rich proteoglycan, and the α8 integrin subunit were enriched in high outflow regions. These studies identify several novel genes that could be involved in segmental outflow, thus demonstrating that digital spatial profiling could be a useful approach for understanding segmental flow through the TM. Furthermore, this study suggests that changes in the expression of genes involved in regulating the activity and/or organization of the ECM and integrins in the TM are likely to be key players in segmental outflow.

Abstract WP304: Investigating Thrombus Architecture Using Spatial Transcriptomics

Introduction: Spatial transcriptomic (ST) platforms provide high-resolution localization of gene expression profiles within intact tissue, offering insight to cellular phenotype and ultimately tissue function. While cerebral thrombus morphology has been well-characterized, their functional architecture has not. Although this study is insufficiently powered to infer specific pathophysiologic conclusions, our findings offer preliminary insight into the functional segmentation of cerebral thrombi using ST. Methods: Four thrombi retrieved from cerebral arteries of ischemic stroke patients during standard of care thrombectomy were formalin-fixed and paraffin-embedded. Morphology marker antibodies (vimentin, CD45) and RNA probes coupled to photocleavable oligonucleotide tags were applied to thrombus specimens. Regions of interest (ROI) were randomly selected based on visualized cellular clustering within each specimen. ST analysis was performed using NanoString's GeoMx Digital Spatial Profiling platform. Results: Of the 24 ROIs, 0 samples were below the 50% warning, and 18,676 targets were analyzed. 1,127 genes were expressed in 50% of ROIs. 8,793 genes normalized by 3rd quartile, expressed above the Limit of Quantitation in at least 1% of ROIs. Differential expression analysis using a false discovery rate of &lt; 0.05 revealed 54 differentially expressed genes (DEG), all of which were protein-coding using edgeR’s exact test. Conclusions: Although ST does not offer visibility into the presence of fully formed ribosomes, the identification of intact cytoplasmic RNA suggests there is significant transcriptional activity within thrombi. We cautiously present pilot data to support feasibility for spatially-resolved transcriptomic analysis of cerebral thrombi which, upon further technical refinement, may facilitate discovery of new targets for stroke prevention or treatment.

Abstract B067: Identification of Gemcitabine-resistant Populations using scRNA-Sequencing in Triple Negative Breast Cancer Patient-Derived Xenografts

Abstract Despite major advances in the treatment of breast cancer (BC), it remains the most diagnosed and second deadliest cancer among American women. BC consists of distinct clinical subtypes, including estrogen receptor or HER2 receptor-positive and triple-negative breast cancer (TNBC) that lacks these receptors preventing the use of established precision therapies and the standard of care remains neoadjuvant chemotherapy (NACT), followed by surgery, radiation and emerging combinations with immunotherapy. While NACT is effective in some patients, up to 50% of patients are intrinsically unresponsive or acquire resistance. These patients demonstrate the highest rates of recurrence and distant metastasis. It is known that chemotherapies can effectively eliminate proliferating tumor cells, but a small population of drug-tolerant persister (DTPs) cells can remain and are thought to play a key role in relapse and resistance. We hypothesized that chemotherapy influences TNBC tumor plasticity by exerting selective pressures leading to the outgrowth of resistant subpopulations with the greatest survival advantage. To evaluate this hypothesis, we generated in vivo models of disease progression and chemotherapy resistance and investigated the plasticity of tumor cell subpopulations by single-cell RNA sequencing. We developed PDX models from a multi-drug resistant primary TNBC treated with standard-of-care chemotherapy and its patient-matched lung metastasis. To model response and resistance at the metastatic stage, we challenged this metastasis PDX with several cycles of Gemcitabine. We obtained residual disease that relapsed and eventually developed resistance, mirroring the patient response. We performed single-cell RNA sequencing (scRNAseq) of these models using a droplet-based technology from 10X Genomics. This scRNAseq data was used to compare the changes in the proportions of cellular subpopulations in each model. Our data show that the rebound model presents greater similarity to the untreated metastasis, while the resistant model significantly differs in cell population expression profiles. Interestingly, the residual disease demonstrates an intermediated state with similarities to both the untreated metastasis and the resistant model. We identified populations with hypoxic signatures in the primary tumor, its matched metastasis and the residual, rebound and resistant models. Immunohistochemistry and Nanostring GeoMx Digital Spatial Profiler validated these populations in these models. In addition, we have identified other varying populations and are currently investigating their cellular mechanisms and gene expression patterns. Using scRNA-sequencing to understand the clonal expansion of resistant subpopulations following chemotherapy reveals distinctive resistant cell features enabling the identification of the vulnerabilities of these tumors. Citation Format: Sandrine Busque, Constanza Martinez Ramirez, Ariel Madrigal Aguirre, Paul Savage, Anie Monast, Anne-Marie Fortier, Gerardo Zapata, Atilla Omeroglu, Jamil Asselah, Nathaniel Bouganim, Sarkis Meterissian, Francine Tremblay, Ari Meguerditchian, Yasser Riaz Alhosseini, Hamed S. Najafabadi, Hellen Kuasne, Morag Park. Identification of Gemcitabine-resistant Populations using scRNA-Sequencing in Triple Negative Breast Cancer Patient-Derived Xenografts [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr B067.

Placenta accreta spectrum disorder at single-cell resolution: a loss of boundary limits in the decidua and endothelium

Placenta accreta spectrum disorders are associated with severe maternal morbidity and mortality. Placenta accreta spectrum disorders involve excessive adherence of the placenta preventing separation at birth. Traditionally, this condition has been attributed to excessive trophoblast invasion; however, an alternative view is a fundamental defect in decidual biology. This study aimed to gain insights into the understanding of placenta accreta spectrum disorder by using single-cell and spatially resolved transcriptomics to characterize cellular heterogeneity at the maternal-fetal interface in placenta accreta spectrum disorders. To assess cellular heterogeneity and the function of cell types, single-cell RNA sequencing and spatially resolved transcriptomics were used. A total of 12 placentas were included, 6 placentas with placenta accreta spectrum disorder and 6 controls. For each placenta with placenta accreta spectrum disorder, multiple biopsies were taken at the following sites: placenta accreta spectrum adherent and nonadherent sites in the same placenta. Of note, 2 platforms were used to generate libraries: the 10× Chromium and NanoString GeoMX Digital Spatial Profiler for single-cell and spatially resolved transcriptomes, respectively. Differential gene expression analysis was performed using a suite of bioinformatic tools (Seurat and GeoMxTools R packages). Correction for multiple testing was performed using Clipper. In situ hybridization was performed with RNAscope, and immunohistochemistry was used to assess protein expression. In creating a placenta accreta cell atlas, there were dramatic difference in the transcriptional profile by site of biopsy between placenta accreta spectrum and controls. Most of the differences were noted at the site of adherence; however, differences existed within the placenta between the adherent and nonadherent site of the same placenta in placenta accreta. Among all cell types, the endothelial-stromal populations exhibited the greatest difference in gene expression, driven by changes in collagen genes, namely collagen type III alpha 1 chain (COL3A1), growth factors, epidermal growth factor-like protein 6 (EGFL6), and hepatocyte growth factor (HGF), and angiogenesis-related genes, namely delta-like noncanonical Notch ligand 1 (DLK1) and platelet endothelial cell adhesion molecule-1 (PECAM1). Intraplacental tropism (adherent versus non-adherent sites in the same placenta) was driven by differences in endothelial-stromal cells with notable differences in bone morphogenic protein 5 (BMP5) and osteopontin (SPP1) in the adherent vs nonadherent site of placenta accreta spectrum. Placenta accreta spectrum disorders were characterized at single-cell resolution to gain insight into the pathophysiology of the disease. An atlas of the placenta at single cell resolution in accreta allows for understanding in the biology of the intimate maternal and fetal interaction. The contributions of stromal and endothelial cells were demonstrated through alterations in the extracellular matrix, growth factors, and angiogenesis. Transcriptional and protein changes in the stroma of placenta accreta spectrum shift the etiologic explanation away from "invasive trophoblast" to "loss of boundary limits" in the decidua. Gene targets identified in this study may be used to refine diagnostic assays in early pregnancy, track disease progression over time, and inform therapeutic discoveries.

Glomerular spatial transcriptomics of IgA nephropathy according to the presence of mesangial proliferation

Mesangial proliferation is a diagnostic feature and a prognostic predictor of immunoglobulin A nephropathy (IgAN). We aimed to investigate the gene expression profiles of IgAN glomerulus according to the presence of mesangial proliferation. We performed spatial-specific transcriptomic profiling on kidney biopsy tissues using the GeoMx Digital Spatial Profiler. Twelve cases with three glomeruli for each case were profiled using direct pathologic classification (4 M1-IgAN, 4 M0-IgAN, and 4 donor controls). The results of enriched glom-specific genes demonstrated that M1-IgAN could be distinguished from controls (77 upregulated and 55 downregulated DEGs), while some DEGs were identified between M1-IgAN and M0-IgAN cases (24 upregulated and 8 downregulated DEGs) or between M0 and controls (1 upregulated and 16 downregulated DEGs). TCF21, an early podocyte damage marker, was the only differentially expressed gene (DEG) consistently upregulated in both M1-IgAN and M0-IgAN patients, whereas ATF3, EGR1, DUSP1, FOS, JUNB, KLF2, NR4A1, RHOB, and ZFP36 were consistently downregulated in IgAN cases. Glomeruli from M1-IgAN cases were significantly enriched for cell surface/adhesion molecules and gene expressions associated with vascular development or the extracellular matrix. Spatial transcriptomic analysis may contribute to dissecting structure-specific pathophysiology and molecular changes in IgAN.

Abstract A108: Spatial transcriptomic analysis of venous invasion in human pancreatic adenocarcinoma reveals widespread attenuation in features associated with aggression and mortality

Abstract In pancreatic ductal adenocarcinoma (PDAC), venous invasion (VI) is a necessary precursor to hematogenous dissemination, but little is known about the molecular features of VI. To investigate these features, we performed spatial transcriptomic analysis of foci of VI, PDAC in stroma (PDAC-S), and normal pancreatic duct in surgically resected human PDAC tissue samples from eight treatment naïve patients. We performed GeoMx Digital Spatial Profiler analysis of 35 regions of VI, 39 regions of PDAC-S, and 14 regions of normal duct. Through unbiased hierarchical clustering, we observe that normal duct foci cluster more closely to VI than to PDAC-S. When comparing differentially expressed genes (DEGs) between VI and PDAC-S, we observe that VI shows high expression of genes which are highly expressed by normal duct relative to PDAC-S. These genes have well characterized roles in normal pancreatic exocrine function and development. Within the upregulated VI DEGs, we observe another group of genes which are elevated in VI relative to both PDAC-S and normal duct, which includes mucins, cell adhesion molecules such as CEACAM5 and TSPAN8, and transcription factors such as FOS and ZFPM1. Pathway analysis of the differentially expressed genes reveals enrichment for MYC signaling and oxidative phosphorylation. When compared to PDAC-S, VI demonstrates downregulation of many genes involved in epithelial-mesenchymal-transition (EMT), such as collagens, laminins, S100 proteins and other EMT-associated genes such as FN1, VIM, MMP11, TGFBI, TGM2, and POSTN. When compared to PDAC-S, VI shows increased expression of genes from the classical phenotype of the Moffitt classification, while PDAC-S has increased expression of the basal-like signature. We performed morphologic subtyping of the VI foci, with 11 samples classified as VI-Destructive (disruption of the venous lumen and associated fibrosis) and 22 samples classified as VI intraepithelial neoplasia-like or VI-IN-Like (complete circumferential replacement of the venous lumen). Compared to VI-IN-Like, VI-Destructive shows increased expression of EMT-associated genes, increased expression of the basal-like gene signature, and reduced expression of the classical signature. Compared to VI-IN-Like and PDAC-S, VI-Destructive showed intermediate expression of EMT-associated genes. Transfer learning of our data to publicly available RNA-Seq data sets reveals that metastatic pancreatic cancer bears greater similarity to PDAC-S than to VI, and circulating tumor cells express features of both VI and PDAC-S. Overall, our data suggest that VI regains transcriptional features of normal ductal epithelium, and demonstrates widespread downregulation of features associated with aggression and mortality, such as genes associated with EMT and the basal-like phenotype. We observe that this transcriptional shift is associated predominantly with the VI-IN-Like morphology, and that VI-Destructive is a transcriptional intermediate between VI-IN-Like and PDAC-S. Citation Format: Alexander T.F. Bell, Peter Chianchiano, Katsuya Hirose, Doreen Gisder, Alexander I. Damanakis, Luciane T. Kagohara, Elana J. Fertig, Laura D. Wood. Spatial transcriptomic analysis of venous invasion in human pancreatic adenocarcinoma reveals widespread attenuation in features associated with aggression and mortality [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr A108.

Spatial Transcriptomic Profiling of Tetraspanins in Stage 4 Colon Cancer from Primary Tumor and Liver Metastasis.

Stage 4 colon cancer (CC) presents a significant global health challenge due to its poor prognosis and limited treatment options. Tetraspanins, the transmembrane proteins involved in crucial cancer processes, have recently gained attention as diagnostic markers and therapeutic targets. However, their spatial expression and potential roles in stage 4 CC tissues remain unknown. Using the GeoMx digital spatial profiler, we profiled all 33 human tetraspanin genes in 48 areas within stage 4 CC tissues, segmented into immune, fibroblast, and tumor compartments. Our results unveiled diverse gene expression patterns across different primary tumor sub-regions. CD53 exhibited distinct overexpression in the immune compartment, hinting at a potential role in immune modulation. TSPAN9 was specifically overexpressed in the fibroblast compartment, suggesting involvement in tumor invasion and metastasis. CD9, CD151, TSPAN1, TSPAN3, TSPAN8, and TSPAN13 displayed specific overexpression in the tumor compartment, indicating potential roles in tumor growth. Furthermore, our differential analysis revealed significant spatial changes in tetraspanin expression between patient-matched stage 4 primary CC and metastatic liver tissues. These findings provide spatially resolved insights into the expression and potential roles of tetraspanins in stage 4 CC progression, proposing their utility as diagnostic markers and therapeutic targets. Understanding this landscape is beneficial for tailoring therapeutic strategies to specific sub-tumor regions in the context of stage 4 CC and liver metastasis.