GeoMx Digital Spatial Profiling Research Articles

824 Spatial transcriptomic analysis of HS skin lesions reveals that tunnels are immunologically active and activity correlates with disease severity

Hidradenitis suppurativa (HS) is an inflammatory skin disease characterized by neutrophil-rich inflammatory nodules, abscesses or tunnels (sinus tracts). Tunnels, unique to HS, are deep dermal cavities lined with squamous epithelium that are immunologically active, contributing to disease pathology. The role of tunnels in driving ongoing disease is poorly understood. We used the GeoMX Digital Spatial Profiler to analyze a moderate (Hurley stage II) compared to a severe (Hurley stage III) HS skin lesion to explore biological differences between and within lesions. Four distinct regions per sample were analyzed: tunnel epithelium, immune infiltrate adjacent to a tunnel, immune infiltrate not associated with a tunnel, and sub-epidermis. We observed higher expression of keratinization genes (e.g. K16, S100A7) in the more differentiated tunnel epithelium of the severe lesion compared to the moderate lesion tunnel. Genes for proinflammatory cytokines (e.g. IL36, IL1B), chemokines (e.g. CXCL1, CCL20), neutrophil activation (e.g. ELANE, MPO) and B cells (CD79A, SDC1) had higher expression in the severe compared to moderate HS lesion. We found a spatial gradient in gene expression for IL36 and neutrophil-attracting chemokines (CXCL1, CXCL8), from high expression in the tunnel epithelium to lower expression in the sub-epidermis. Both lesions displayed these gradients, but those in the severe HS lesion were more pronounced. Conversely, spatial gradients of fibroblast-related genes (e.g. PDGFR, FAP) had higher expression in the sub-epidermis and lower expression adjacent to tunnel epithelium. These data suggest a marked increase in dermal inflammation in more severe HS lesions. The spatial gradients of cytokine and chemokine gene expression support the idea that more differentiated tunnels are more immunologically active, potentially contributing to ongoing severe disease, rather than representing remnants of prior disease.

Abstract 5676: Digital spatial profiling for determination of pancreatic stromal components and correlation with survival

Abstract Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related death in the United States. Recently, intensive research has revealed the remarkable heterogeneity and complexity of the different components in the pancreatic cancer tumor microenvironment. However, our knowledge regarding active stroma (dominated with activated cancer-associated fibroblasts), infiltrated immune cells and actual tumor cells interactions in the tumor microenvironment, and how these interactions impact patient outcomes, remains limited. Digital spatial profiling (DSP) technology now allows us to quantitatively analyze transcript and protein expression on intact tissue, offering an opportunity to study the intra-tumor cell-to-cell communication. Formalin-fixed paraffin-embedded (FFPE) pancreatic tumors from patients with shorter survival (DFS < 12 months, n=2), and longer survival (DFS > 36 months, n=2) were recruited in this study. Immunofluorescent histochemistry staining of tissue morphology identification uses antibodies to pan-cytokeratin (panCK+) for PDAC cells, alpha-smooth muscle actin (aSMA+) for cancer-associated fibroblasts (CAFs). Six replicated regions of interest (ROI) of tumor, CAFs adjacent and distal to tumor area were annotated from each tissue sample and a panel of 78 transcripts (73 target genes and 5 reference genes) were profiled using GeoMx DSP (Nanostring). Person’s correlation (R2) was computed for regression of GeoMx DSP data. All statistical analyses were performed using the GraphPad Prism software. p-values were calculated on two-sided t-tests and p < 0.05 was considered statistically significant. As expected, the morphologic selection successfully separates panCK+ PDAC cells from CAFs demonstrated by the correspondent expression of epithelial genes of EpCam and KRT, allowing single cell profiling. Within the panel of 78 genes, 9 genes (approximal 12%) were detected with significantly higher expression levels in PDAC cells than CAFs. Five genes (BATF3, IL12b, ITGB8, CD4 and IFNAR1) were found increased expression levels in CAFs-adj comparing to CAFs-dis, regulating molecular functions of Organismal Death (p value = 3.88E-03), Maturation of Cells (p value = 2.94E-06), and Migration of Cells (p value = 2.37E-03). When comparing the differential transcriptomes in PDAC cells between the longer and shorter survival groups, higher expression of antigen presentation gene HLA-E and integrin alpha- and beta- family genes (ITGAV, ITGAX, ITGB2, and ITGB8) was seen in longer-lived patients. Proliferation genes (CCND1 and MKI67) were also found to be increased in CAFs in patients with longer survival. Our results suggest that further exploration of gene and protein expression and interactions within the pancreatic tumor microenvironment may improve our understanding of tumor and stromal communication and how it impacts on patient outcomes. Citation Format: Gerik W Tushoski, Dongtao Ann Fu, Lingsong Meng, Kelly Herremans, Andrea N Riner, Christopher E Forsmark, Zhiguang Huo, Steven J Hughes, Song Han. Digital spatial profiling for determination of pancreatic stromal components and correlation with survival [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5676.

Abstract 989: Spatial profiling of androgen receptor splice variant 7 transcriptional activity in prostate cancer metastases

Abstract Introduction: Androgen receptor splice variant 7 (AR-V7) is expressed in metastases from patients with castration resistant prostate cancer (CRPC) and shows a high level of inter- and intra-tumoral variability. However, the downstream activity generated by AR-V7 in tissue has not been shown. Whether AR-V7 is active in tissue or whether AR-V7 is a non-functioning biomarker with full-length AR is not known. To address this question, we constructed a tissue microarray (TMA) of 56 metastases from 27 patients with CRPC to analyze spatial gene expression using the GeoMx Digital Spatial Profiler. Immunostaining was performed to define epithelial, vascular, and stromal compartments. The stained tissues were then hybridized with barcoded tagged oligonucleotides targeting 2093 unique genes, which included those representing AR, AR-V cryptic exons, AR and neuroendocrine activity, and immune cell markers. One 500 um region of interest (ROI) was assessed per tissue core (approximately 1200 cells). A sequential section from the same TMA was then stained with AR-V7 and AR-C-terminal specific antibodies. ROIs for RNA and protein were selected to be similar between slides. In addition to DSP, each metastasis was assessed by RNA-seq on the bulk tissue. Results: The most differentially expressed genes (FDR<0.05) based on association with AR-V7 staining were known downstream AR regulated genes including KLK2 and 3, FKBP5, NKX3.1, TMPRSS2, FASN, and TARP. Additionally, genes associated with proliferation and stemness, e.g., POLB, KRT1, SOX2, were significantly expressed. Since 93% of patients were on ADT at time of tissue collection and over 80% also had been treated with either abiraterone or enzalutamide, the increase in AR downstream genes would not be expected to occur from ligand activation of AR-FL. We also have previously shown that knock-down of AR-V7 in LNCaP95 cells results in loss of AR binding to AREs. In these metastases, then, activation of AR downstream genes would be a result of AR-V7 nuclear transport of AR-FL through AR-V7/AR-FL heterodimers or transcriptional activation by AR-V7 homodimers. Of further note, AR cryptic exons 1, 2, and 5 were also significantly expressed in AR-V7 positive ROIs (p< 0.0001). RNA-seq intron/exon junction reads were used to demonstrate that additional AR-Vs, most commonly AR-V9, were also expressed in tissues positive for AR-V7, suggesting that AR splicing is a common event in CRPC. Finally, expression of neuroendocrine (NE) genes INSM1 and TUBB2 were not expressed in AR-V7 positive ROIs (negatively correlated, p<0.001), indicating that AR-V7 and NE phenotypes cannot co-exist in the same cell. Conclusion: AR-V7 continues to drive prostate cancer through activation of the AR-cistrome. Its expression is heterogenous in metastases along with NE cells, suggesting that in the presence of AR-V7 and NE markers, therapy needs to be directed at the N-terminus of AR and NE components. Citation Format: Cynthia C. Sprenger, Ilsa Coleman, Michelle Kriner, Lauren Brady, Margaret Hoang, Martine Roudier, Mamatha Damodarasamy, Radhika Patel, Lawrence True, Peter Nelson, Michael Haffner, Stephen Plymate. Spatial profiling of androgen receptor splice variant 7 transcriptional activity in prostate cancer metastases [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 989.

Abstract 1715: Phenotypic characterization of spatial immune infiltration niches in non-small cell lung cancer

Abstract Introduction: Prolonged response to checkpoint inhibitors is still rare in patients with non-small cell lung cancer (NSCLC) and the performance of current predictive biomarkers is limited. Thus, more accurate biomarkers and alternative treatment targets are needed to optimize immunotherapy. The objective of this study was to advance the understanding of cancer immunity by characterization of spatially resolved compartments of immune infiltration. Material and methods: Using the GeoMx Digital Spatial Profiler (NanoString), 49 proteins were measured in 183 regions of interest (ROIs) from FFPE tissue sections of 33 NSCLC patients. The pan-leukocyte marker CD45 was used to define region of interests (ROIs) of three spatial phenotypes: (1) tumor infiltrating leukocytes (n=39), (2) tertiary lymphoid structures (TLS, n=19) and (3) tumor stroma without specific leukocyte aggregation (n=125). Differential protein expression (FDR<0.05) was determined using generalized linear mixed models with patient as random effect, to account for multiple samples per patient. Results: Normalized levels of lymphocyte phenotypic markers (CD3, CD8, CD4, and CD20) did not differ between stroma and tumor infiltrates, however the infiltrates had higher expression of markers of immune suppression (PD-L1, B7-H3, CTLA-4, IDO-1, PD-L2, 4-1BB, LAG3), regulatory T-cells (FOXP3), macrophages (CD68), and antigen presentation (CD40, OX40L, CD11c). The stromal ROIs were characterized by markers of fibroblast activation (Fibronectin, SMA, FAP alpha), and higher levels of VISTA and CD27 compared to infiltrating regions, and VISTA, ARG1 and CD66b compared to TLS. In comparison, TLS had higher levels of lymphocyte phenotypic markers (CD20, CD3, CD4, CD8), markers of activation (CD45RO, CD40, ICOS), proliferation (Ki67), and antigen presentation (HLA-DR, B2M, CD11c). In general, markers for macrophages (CD68), neutrophils (CD66b), regulatory T-cells (FOXP3) and immune suppression (B7-H3, IDO1, PD-L1, CTLA4) were more abundant in the vicinity of tumor cells, while CD27, ICOS, SMA and CD3 increased with longer distance to tumor cells. PD-L1 and IDO1 protein expression was higher in tumors that were immunohistochemical positive for PD-L1. Conclusions: Spatially resolved protein profiling showed that NSCLC TLS are characterized by activated, proliferating lymphocytes and antigen presentation. We identified a general immunosuppressive microenvironment and indication of T cell exhaustion in the tumor cell compartment and its close vicinity. The clear discrimination of different immune infiltration niches independent of the overall level of leukocytes in the sampled regions illustrates the value of spatial profiling for deconvoluting how immune cells can dictate tumors, and to identify predictive immune signatures in the context of immunomodulatory treatment of NSCLC patients. Citation Format: Anna Sandström Gerdtsson, Artur Mezheyeuski, Patrick Micke, Sara Ek. Phenotypic characterization of spatial immune infiltration niches in non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1715.

Abstract CT154: B-cell infiltration in lung cancer predicts response to neoadjuvant pembrolizumab

Abstract Background Neoadjuvant immune checkpoint inhibitor treatment is a promising approach for resectable cancer, including non-small cell lung cancer (NSCLC). The characteristics of potential responders to such treatments and the molecular underlying events are not known. Methods We have conducted a phase I, investigator-initiated single-center study (MK3457-223), to examine the safety of neoadjuvant pembrolizumab for stage I-II resectable NSCLC and to determine the recommended phase II dose/schedule (RP2D/S). FFPE biopsies and surgical specimens were subjected to correlative studies. NanoString’s GeoMx Digital Spatial Profiler (DSP) analysis was conducted on pre-treatment samples and post-treatment responder samples. Protein (72 proteins) and mRNA expression data (73 genes) analysis was conducted on regions of interest (ROIs), defined as mostly CD8 positive or mostly pan-cytokeratin positive (presumed cancer cells). Pathology assessment was done on the surgical specimen to identify major pathologic response (MPR; ≤10% remaining viable cells). Statistical analysis was done to compare responders (MPR+) to non-responders (MPR-) by Mann Whitney with false discovery rate correction. Immunohistochemistry (IHC) was conducted on post-treatment samples. Results Twenty-six patients initiated treatment on the study. Two patients (8%, 95% C.I 0-18%) had adverse events that precluded surgery, 1 patient refused surgery after treatment. 7 patients (27%, 95% C.I 10-44%) achieved a major pathologic response (MPR; responders), 3 patients (12%, 95% C.I 0-24%) achieved complete pathologic response. Responders had a longer interval from treatment to surgery (43 days vs. 36 days, univariate analysis, p-value 0.043). RP2D/S was determined as 2 treatments of 200mg pembrolizumab at 3 week interval, followed by surgery at least 2 weeks later. The expression of several proteins and genes differed between responders and non-responders. Pre-treatment, CD20 protein was the most differentially expressed protein both in in CD8+ (4.7 fold, p=0.002) and in cancer cells (4.8 fold, p=0.001) ROIs, in both cases higher in the responders compared to the non-responders. Comparing pre to post-treatment expression in responding tumors, the protein found to be upregulated to the highest extent following pembrolizumab treatment was CD20 protein (6.2-fold, p=0.001), as was its encoding gene, MS4A1 (2.4-fold, p=0.006). CD20 IHC of post-treatment samples demonstrated tertiary lymphoid structures (TLS) to be more prevalent in responders compared to non-responders (3.2-fold, p<0.05). Conclusions Longer interval from treatment to surgery was associated with higher rate of MPR. Presence of tumor-infiltrating B-cells and evolvement of TLSs was strongly correlated with pathologic response to neoadjuvant pembrolizumab in early stage NSCLC. Citation Format: Jair Bar, Iris Kamer, Oranit Zadok, Damien Urban, Marina Perelman, Ilanit Redinsky, Aliza Ackerstein, Sameh Daher, Efrat Ofek, Amir Onn, Nona Zeitlin, Alon Ben-Nun, Ran Kremer, Inbal Daniel, Yossef Glantzspiegel, Irit Gat-Viks. B-cell infiltration in lung cancer predicts response to neoadjuvant pembrolizumab [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr CT154.

Abstract 2578: High regulatory T cells infiltrate within tertiary lymphoid structure restricts response to immune checkpoint blockers in sarcomas

Abstract Background: While composition of the tumor microenvironment (TME) is a prerequisite for an effective antitumor immunity, infiltration of organized B- and T-cells aggregates called tertiary lymphoid structures (TLS) has been recently demonstrated to predict response to immune checkpoint blockers (ICB) in sarcomas. However, only a minority of patient derive benefit, suggesting the implication of additional key determinants of ICB-mediated response in TLS-positive sarcomas, such as TLS composition. Using high-throughput spatial transcriptomics and multiplex immunofluorescence (IHF), we aimed at investigating the association between TLS composition and clinical outcome to ICB. Methods: In an exploratory cohort, we spatially profiled the expression of more than 18000-protein encoding genes from responders (R) and non-responders (NR) using Nanostring’s GeoMx Digital Spatial Profiler (DSP) Whole Transcriptome Atlas (WTA) assay. A first set of regions of interest (ROI) was selected in the TLS and further segmented in “B-cells” vs “no B-cells” areas according to CD20+ staining; a second set of ROI was selected in the tumor tissue and further segmented into “tumor” vs “stroma” areas according to CD45+ staining. Deconvolution of data was performed using SpatialDecon algorithm to estimate cell population within TLS. We then evaluated the association between immune cell composition and response to ICB in each segment. In a validation cohort, we performed multiplexed-IHF assay enabling detection of T cells (CD8/GzmA/CD4/FoxP3/CD56) and B cells. These panel was applied to whole sections baseline sarcoma samples. We investigated the association between immune cell composition and clinical benefit in term of progression-free survival (PFS) and overall survival (OS). Results: Six patients were selected for the exploratory cohort, including 3 R and 3 NR. Among the top immune cell infiltrate within TLS segment, NR demonstrated higher Treg infiltrate versus R in “no B-cells” compartment (3.4% vs 2.0%, respectively; p=0.010), whereas no association was observed between Treg infiltration and response to ICB in both stromal (p=0.67) or tumor cells (p=0.36) compartments from tumor area. In the validation cohort (N=16), we observed that Treg density within TLS was higher in NR versus R (p=0.0059). Patients with Treg-enriched TLS had shorter PFS (2.6 vs 11.1 months, p=0.042) and OS (9.0 vs 18.3 months, p=0.12) compared to those with Treg-low TLS infiltration. Concordantly, the CTLA-4 key Treg regulator gene was upregulated in the TLS regions from NR. Conclusions: Altogether, our findings suggest that the presence of Treg within TLS may exert a negative influence on the capacity of TLS to generate an effective antitumor immune response in sarcoma patients treated with ICB, providing new insights in understanding role of TLS in antitumor immunotherapy. Citation Format: Florent Peyraud, Jean-Philippe Guegan, Christophe Rey, Marina Pulido, Emmanuelle Bompas, Sophie Piperno-Neumann, Christine Chevreau, Nicolas Penel, François Bertucci, Maud Toulmonde, Carine Bellera, Catherine Sautès-Fridman, Antoine Bougoüin, Coralie Cantarel, Michèle Kind, Mariella Spalato-Ceruso, Bérengère Dadone-Montaudie, Jean-Yves Blay, Wolf Herman Fridman, François Le Loarer, Alban Bessede, Antoine Italiano. High regulatory T cells infiltrate within tertiary lymphoid structure restricts response to immune checkpoint blockers in sarcomas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2578.

Abstract 2028: Spatially resolved proteomic profiling identifies tumor cell CD44 as a novel indicative biomarker of sensitivity to PD-1 axis blockade in advanced non-small cell lung cancer (NSCLC)

Abstract Introduction: Most patients with advanced NSCLC fail to develop durable responses from PD-1 axis blockade, and more robust predictive biomarkers are needed. In this study, we aimed to discover and validate spatially resolved protein markers associated with sensitivity to PD-1 axis inhibition across multiple NSCLC cohorts. Methods: We first used the GeoMx Digital Spatial Profiling (DSP) system in a discovery cohort of 56 NSCLC patients treated with PD-1 axis inhibitors at Yale (USA). Pre-treatment tumors were represented in a tissue microarray (YTMA471) and analyzed using a 71-plex primary antibody panel. Proteins were measured from three molecularly defined tissue compartments: tumor (CK+), leukocytes (CD45+/CD68-), and macrophages (CD68+). We used quantitative immunofluorescence (QIF) to orthogonally validate candidate biomarkers. For external validation of identified predictors, we assessed pre-treatment whole tissue sections from a cohort of 128 NSCLC patients treated with single-agent PD-1 axis inhibitors at the Hospital 12 de Octubre (Madrid, Spain) using DSP (39-plex protein panel, measured from CK+ and CD45+ compartments). In addition, we analyzed two immunotherapy untreated cohorts to address prognostic significance: YTMA423 (Yale, USA; n = 252) and CIMA-CUN (UNAV, Spain; n = 124), using QIF and DSP respectively. Results: Using continuous log-scaled data, we found 5 markers independently associated with progression-free survival (PFS) in the tumor compartment (including PD-L1, p = 0.002). Among the novel candidate predictors, tumor cell CD44, a marker of pluripotency and stemness, was associated with longer PFS (multivariate HR = 0.63, p= 0.002). Using QIF, we orthogonally validated that CD44 expression in the tumor compartment was associated with longer PFS (p <0.001) and overall survival (OS) (p = 0.03). In the external validation cohort, CD44 expressed in the tumor compartment, but not in the immune compartment, was predictive of clinical benefit (OR 1.22, p = 0.018), and was significantly associated with longer PFS (first tertile cutpoint: HR 0.62, p = 0.03). In contrast, high CD44 expression was not associated with survival in the two untreated cohorts. Using DSP data from two cohorts, regions of interest with elevated expression of CD44 in tumor cells consistently showed prominent (Fold Change>1.5, p<0.05) upregulation of TIM-3, and PD-L1 and multiple co-stimulatory molecules (including CD40, ICOS, CD27). Conclusion: This work highlights CD44 as a novel indicative biomarker of sensitivity to PD-1 axis blockade in NSCLC. NSCLCs with high CD44 expression in epithelial cells may be associated with an immune contexture primed for higher response to immune checkpoint blockade. Further studies are needed to understand the interplay between CD44+ cancer stem cell phenotype and mechanisms of immune evasion. Citation Format: Myrto Moutafi, Magdalena Molero, Sandra Martinez-Morilla, Javier Baena, Ioannis Vathiotis, Niki Gavrielatou, Laura Castro, Gorka Ruiz de Garibay, Vera Adradas, Daniel Orive, Karmele Valencia, Alfonso Galvo, Luis M. Montuenga, Santiago Ponce, Kurt Schalper, Luis Paz-Ares, David L. Rimm, Jon Zugazagoitia. Spatially resolved proteomic profiling identifies tumor cell CD44 as a novel indicative biomarker of sensitivity to PD-1 axis blockade in advanced non-small cell lung cancer (NSCLC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2028.

Abstract CT139: Intratumoral (IT) microdosing of the investigational SUMOylation Inhibitor TAK-981 in a phase 0 CIVO trial demonstrates the reactivation of type I Interferon (IFN1) signaling in head and neck squamous cell carcinoma (HNSCC)

Abstract Purpose/Objectives: The human tumor microenvironment (TME) has a dramatic impact on cancer prognosis and therapeutic response, but accurate models of the native TME do not exist. The Comparative In Vivo Oncology (CIVO) platform was developed as a means to assess the effect of investigational agents on the native TME in a Phase 0 microdosing study. CIVO was clinically validated using approved agents and is used for the first time here to assess the impact of an investigational agent - the SUMOylation inhibitor TAK-981 - on the native human TME in HNSCC. Materials/Methods: Eligible subjects have a confirmed HNSCC diagnosis, ECOG 0-2, and planned surgical resection. Injectable tumors were at the primary site or within cervical lymph nodes but had to be surface-accessible and ≥ 2cm. TAK-981 or control microdoses were simultaneously administered via a CIVO device and co-injected with a fluorescent tracking marker for injection site identification and visualization. Tumors were resected 24 or 72 hours after injection, processed, and then analyzed at a central site. Multiplexed biomarker staining and molecular profiling via GeoMx Digital Spatial Profiling were performed to capture pharmacodynamic responses in the native TME. Results: As of January 2022, 8 subjects provided informed consent and were enrolled, and no adverse events associated with the injection procedure or microdoses have been reported. Biomarker analysis demonstrated TAK-981 distribution around the injection site accompanied by reduction of SUMOylation. Dose-dependent elevation of IFN1 signaling was also observed in TAK-981-exposed areas within the TME. Elevated IFN1 signaling was accompanied by TME reconfiguration, with increased macrophage M1 polarization and activation of dendritic cells, NK cells, and CD8+ T cells. TAK-981 exposure was also associated with upregulation of CXCL10, PD-L1, and an IFNγ gene expression signature predictive of response to immune checkpoint blockade. Conclusion: IT microdosing with CIVO provided early insights into complex functional responses induced by the investigational agent TAK-981 that can only be accurately evaluated in the intact, native TME of a patient’s tumor. SUMO pathway inhibition in HNSCC tumors following TAK-981 exposure led to functional activation of multiple immune cell types, effectively shifting the local TME toward an inflamed “hot” state, highlighting TAK-981’s potential as an immune stimulating agent for treating patients with solid tumors. These data were generated while TAK-981 was still in Phase I dose escalation trials (via IV administration), highlighting CIVO’s ability to safely study investigational agents. Further evaluation of TAK-981 alone and in combination with other agents is ongoing in this Phase 0 CIVO microdosing trial. Citation Format: Jeffrey Houlton, Harrison Cash, Haodong Xu, Paul L. Swiecicki, Keith Casper, Steven B. Chinn, Daniel R. Clayburgh, Ryan J. Li, Robert J. Christian, Aaron Halfpenny, Annemieke van Zante, Beryl A. Hatton, Kimberly Sottero, Marc O. Grenley, Connor Burns, Jason Frazier, Jonathan Derry, Gloria Kung, Emily Beirne, Nathan J. Schauer, Atticus Turner, Wendy Jenkins, Kirsten Anderson, Richard A. Klinghoffer, Dennis Huszar, Allison Berger, Karuppiah Kannan. Intratumoral (IT) microdosing of the investigational SUMOylation Inhibitor TAK-981 in a phase 0 CIVO trial demonstrates the reactivation of type I Interferon (IFN1) signaling in head and neck squamous cell carcinoma (HNSCC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr CT139.

Abstract 2027: Spatial omics and multiplexed imaging to discover new biomarkers of response or resistance to immune checkpoint inhibitors (ICI) in advanced non-small cell lung cancer (NSCLC)

Abstract Introduction: Despite the impressive outcomes with ICI in NSCLC, only a minority of the patients show long-term benefit from ICI. While PD-L1 immunohistochemistry is an approved companion diagnostic test, it is neither sensitive nor specific. Here, we describe the use of spatial transcriptomics using the GeoMx Digital Spatial Profiler (DSP) as a discovery platform to find biomarkers for ICI response or resistance. Methods: Pre-treatment samples from 56 patients with NSCLC treated with ICI were collected and analyzed in a tissue microarray (TMA) format in two-fold redundancy. The human whole transcriptome, represented by 18000 genes assessed by oligonucleotide-tagged in situ hybridization, was sequenced on the NovaSeq platform to quantify the mRNAs present in each region of interest. Three molecularly defined tissue compartments, defined by fluorescence co-localization (tumor [panCK+], leukocytes [CD45+/CD68-], macrophages [CD68+]) were generated to assess mRNA. All statistical testing was performed using a two-sided significance level of α=0.05. Result: 54000 gene variables were generated per case, from them about 27000 were analyzed after removing targets with expression lower than a prespecified frequency. Expression distribution of EPCAM in tumor, PTPRC in leukocytes and CD68 in macrophages confirmed that DSP successfully profiled the molecular compartments. Differential gene expression (DEG) analysis was performed for 6-month clinical benefit across two distinct tumor TMA spots per patient. This resulted in a DEG list of about 250, 125 and 122 genes in CK, CD45 and CD68 compartment respectively. When we used the intersection of significantly associated with benefit genes the candidate list was reduced dramatically to 5, 10 and 10 distinct genes in CK (CYBA, ESD, SH3BGRL3, TOP2B, UFM1), CD45 (ATRN, DGAT1, DNAAF9, GMPS, IFT20, MPHOSPH6, PKP4, SOCS2, TFAP4, TGFB3) and CD68 (ANKS3, ARFRP1, CTSD, DPYD, GTF3C5, HS2ST1, KIAA0895L, PIGR, SFTPA2, YEATS4) compartment respectively. From them, using the Cox Proportional-Hazards Model we further identified CYBA, ESD and UFM1 in CK, DGAT1 and IFT20 in CD45 and ANKS3 and DPYD in CD68 to predict survival as well. Conclusions: Using DSP technology allows rapid, patient-specific assessment of the transcriptome in TMAs by in situ hybridization in spatially defined molecular compartments. Here we show a small set of candidate genes that are associated with outcome in this ICI-treated cohort. By intersecting two non-adjacent cores of the same patient tissue sample, we have begun to dissect tumor heterogeneity and pilot biomarker candidate information with unique molecularly defined compartments for tumor cells, lymphocytes, and macrophages. Results from this cohort may lead to a novel, spatially defined, transcriptomic approach for developing new biomarkers for immunotherapy. Citation Format: Myrto Moutafi, Sandra Martinez-Morilla, Rolando Garcia-Milian, Thazin Nwe Aung, Ioannis Vathiotis, Niki Gavrielatou, Vasiliki Xirou, Leonidas Salichos, David L. Rimm. Spatial omics and multiplexed imaging to discover new biomarkers of response or resistance to immune checkpoint inhibitors (ICI) in advanced non-small cell lung cancer (NSCLC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2027.

Abstract 6327: Profiling T-cell subpopulations based on spatial proteome in mantle cell lymphoma

Abstract Background: In precision medicine era, we need to develop tools that better predict treatment response with, for e.g. immuno-oncology drugs. One hypothesis is that diverse treatment response is governed by differences in the tumor microenvironment (TME) both within the tumor and surrounding areas. Thus, novel ways to measure the heterogeneity within defined immune-cell populations is the key and now possible by using spatially-resolved multiplexed platforms. Here, we have investigated the phenotype of four T-cell subtypes both within and excluded from the tumor-rich regions in a large cohort of primary mantle cell lymphoma (MCL) patients. We propose that profiling different cell populations could potentially provide insight into patient prognosis and identify clinically actionable targets, but also in understanding whether tumor cells locally re-educate T-cells through cell-to-cell interaction or by soluble factors Method: We analyzed archival FFPE presented as tissue microarrays of 196 cores from 104 MCL patients, collected from the Biobank of Lymphoma of Southern Sweden, by using the, GeoMX-digital spatial profiler (DSP, Nanostring Inc.). Cellular regions of interest were selected by immunofluorescense staining of CD8, CD3 and CD57 to identify four subpopulations of T cells - early cytotoxic (CD3+ CD8+ CD57-), early non-cytotoxic (CD3+ CD8- CD57-), late cytotoxic (CD3+ CD8+ CD57+) and late non-cytotoxic (CD3+ CD8- CD57+). 601 regions distributed among the four T-cell subtypes were collected from both within the tumor-rich and tumor deficient zones and probed for 69 proteins for sub-characterization. Result: Our results show that the distribution and expression profile of all subpopulations differ considerably based on their position relative to the tumor. Using linear mixed effects model to account for effect of multiple sampling per patient, we observed higher cytolytic activity (upregulation of granzyme A and B) (FDR cut-off < 0.05) in early cytotoxic and late non-cytotoxic cells, within the tumor-rich zone compared to the tumor-deficient area. In contrast, proteins such as STING, VISTA, neurofibromin and fibronectin were consecutively upregulated in all subtypes in tumor-deficient zones. We also observe differential expression profiles for proteins involved in for example cell death regulation and apoptotic signaling pathways. Our analysis provide insight into effector functions important for response to treatment in MCL. Conclusion: This study reveals the phenotypic differences of T-cell subsets in MCL with a higher resolution compared to previous studies. The spatially-resolved analysis shows that the proximity to tumor cells highly affect the profile of different T-cell subsets and that further analysis of re-education mechanisms should be pursued. This study advance the understanding of the MCL tissue microenvironment and for the basis for connecting such differences to clinical outcome. Citation Format: Lavanya Lokhande, Joana de Matos Rodrigues, Anna Sandstrom Gerdtsson, Anna Porwit, Mats Jerkeman, Sara Ek. Profiling T-cell subpopulations based on spatial proteome in mantle cell lymphoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6327.

Abstract 2031: High plex proteomic prognostic marker discovery for patients with pancreatic adenocarcinoma using digital spatial profiling

Abstract Pancreatic Ductal Adenocarcinoma (PDAC) long-term outcomes following surgical resection remain poor, with only 20% of patients surviving 5 years post pancreatectomy. With its immune-privileged nature, starting from the early pre-neoplastic state, it appears to easily evade the antitumor immune response. Although there is rationale for targeting PDAC immune pathways, little benefit has been observed to date. The aim of the study was to interrogate the immune landscape of PDAC utilising the Nanostring GeoMx Digital Spatial Profiler (DSP), enabling high-plex proteomic characterisation whilst maintaining tumor microenvironment (TME) topographical features. We assessed tumor samples from 28 treatment-naive PDAC cases represented in a multiregional FFPE tissue microarray (TMA) for which extensive IHC, molecular, genomic characterisation and clinicopathological follow-up data is available. Multiplex immunofluorescent staining for nuclei, panCK, aSMA and CD3 was used to guide region selection using the GeoMx DSP system in multiple TMA cores. 60 immune markers were quantified simultaneously in different tissue compartments defined by immunofluorescence co-localization including tumor (panCK+), immune stromal (PanCK-) regions. The spatially informed assessment by DSP was validated by both regression and variable prognostication compared with IHC for stromal CD3, CD8 and CD68 in near serial TMA PDAC sections. Unsupervised analysis of proteome data (DSP) in the panCK-negative regions identified an immune poor group associated with shorter OS (13.0 versus 31.1mths, P=0.005). When transcriptomic subtype was considered, the checkpoint inhibitor B7-H3 was significantly upregulated in the squamous subtype tumours versus the classical group (Log2: 1.63, P = 0.001). Patients with high B7-H3 expression, using a median expression cut-off, were associated with shorter OS on multivariate analysis (HR: 4.16, P=0.01) a finding that was validated in an external cohort at the bulk transcriptome level. This pilot scale discovery study shows the potential of the Nanostring DSP technology to identify spatially-informed biomarkers with prognostic relevance in biopsy sized samples from treatment-naive PDAC. A number of relevant candidate immune predictors were identified in a spatial context, which are currently undergoing validation in a larger independent cohort and within the neoadjuvant setting. Future studies will apply this technology to pre and post treatment biopsy samples. Citation Format: Assya Legrini, Holly Leslie, Stephan Dreyer, Joanne Edwards, Andrew Biankin, David Chang, Nigel Jamieson. High plex proteomic prognostic marker discovery for patients with pancreatic adenocarcinoma using digital spatial profiling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2031.

Abstract 66: A multi-institution examination of concordance in spatial transcriptomics using the GeoMx Cancer Transcriptome Atlas

Abstract Multiplexed spatial profiling can enable biological insights by characterizing gene expression within discrete physical locations of a tissue. However, these advanced techniques can be confounded by variability of sample collection, storage, or profiling protocols, making it difficult to accurately compare data generated by different laboratories. Further, as more spatial profiling datasets become publicly available, having methods to enable meta-analysis of samples collected on different studies will maximize the learnings to support the advancement of treatments or identifying patient segments. To quantify the variability of spatial profiling data at different laboratories and to advance data normalization methodologies, 4 independent laboratories used the NanoString® GeoMx® Digital Spatial Profiling (DSP) to profile serial 5 µm sections of tissue and cell pellet arrays (CPAs). GeoMx DSP enables high throughput, spatially resolved analysis of gene or protein expression from fresh or archival human tissues. In this study, the GeoMx Cancer Transcriptome Atlas was used to profile >1800 genes simultaneously. We examined the concordance of GeoMx data generated in the different laboratories when controlling for methodical variation (e.g., reagents, tissue source) and experimentally varying region of interest (ROI) size, collection site, and sample preservation methods. Sections of tonsil, colon, and 2 CPAs were profiled separately at the 4 laboratories. Each analyzed fresh cut (FC) tissues and two sites examined sample stability by analyzing the impact of storing slides at -80°C for 1 month prior to spatial profiling. Concordance analysis was performed using the Horn-Morisita Index on raw data comparing paired and unpaired ROIs across each set of slides. In CPA samples where each pellet was a different tumor type (e.g., NSCLC, melanoma), we observed strong clustering by cell line. While data initially showed varying degrees of clustering by slide, factoring out this variable removed the association of slide, allowing integration of the data across profiling locations without affecting concordance within slides. In tonsil, ROIs with increasing area were profiled. Comparing expression between pairs of samples for a given area, concordance increased with ROI size (R = -0.40, p<6e-06). Finally, we observed little impact of preservation method (FC vs -80°C) in these data. In this study, we quantify slide-specific variation observed in high-plex RNA profiling by the DSP platform and detail methods for accounting for this variation. We note that many downstream analyses (e.g., differential expression) already model slide effects during the analysis, but modeling it explicitly allows for direct comparison of concordance with other approaches (e.g., clustering, PCA). These methods support the use of multi-institution studies leveraging the GeoMx platform. Citation Format: Tyler Hether, Tim Howes, David Scoville, Charlie Glaser, Yanyun Li, Rami Vanguri, Neeman Mohibullah, Wan-Jung Chang, Todd Yoder, Minnal Gupta, Kathy Ton, Yan Liang, Ying Huang, Zach Herbert, Jason Reeves, Elizabeth Mittendorf, Simon Lacey, Travis Hollmann, Sarah Warren, Theresa LaVallee. A multi-institution examination of concordance in spatial transcriptomics using the GeoMx Cancer Transcriptome Atlas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 66.

Abstract 2029: Multiomic analysis of whole transcriptome and high plex protein assays on a single FFPE slide

Abstract The GeoMx࣪ Digital Spatial Profiler (DSP) enables high-plex, high-throughput spatial profiling and quantification from a single slide for either protein or RNA. To fully understand the interplay between RNA and protein in tissue, we have developed a novel multi analyte assay that allows for the profiling of both analytes from the area of interest (AOI) on a single slide. Here we describe the development and performance of the multi analyte assay on cell pellet array (CPA) and various tissues including tonsil and colorectal cancer (CRC) using a high plex GeoMx Protein Panel and the GeoMx Human Whole Transcriptome Atlas (WTA). We successfully identified conditions to detect RNA and protein simultaneously, with comparable sensitivity and specificity to the single analyte conditions. The data accurately matched cell lines in an unbiased correlation analysis between GeoMx WTA and the entire CCLE RNAseq database. Furthermore, the assay performance was maintained in tissue samples. Tonsil and colorectal cancer biopsies were profiled for RNA (whole transcriptome) and protein from the same AOI, allowing simultaneous measurements of transcriptional and translational regulation of key biological pathways. This multi-analyte profiling enables deeper characterization of precious biological samples that are available in limited quantities. Citation Format: Shilah Bonnett, Nicholas Confuorto, Kit Fuhrman, Giang Ong, Alyssa Rosenbloom, John Lyssand, Gary Geiss, Joseph Beechem. Multiomic analysis of whole transcriptome and high plex protein assays on a single FFPE slide [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2029.

Abstract 2023: Spatial transcriptomic evidence for an inflammatory, pro tumorigenic microenvironment in normal human dense breast tissue

Abstract Breast cancer affects over 2,000,000 women worldwide, with more than 680,000 deaths per year. Increased breast density is one of the risk factors for breast cancer, albeit less than the risk associated with increased age and genetic mutation. It has been suggested that high breast density accounts for 15% of the breast cancers diagnosed. On radiographic examination, breast density appears as opaque regions associated with increased cellularity and matrix, which in some circumstances contribute to difficulty observing small tumors. Dense regions in the breast have been associated with both morphological changes as well as molecular changes in the tissue. How these changes may play a role in increased risk for tumorigenesis are not well described. In this study, we examine the partial spatial transcriptomes of specific regions in normal breast tissue containing dense regions, including regions of normal density, the interface of dense and non-dense regions and the dense regions themselves using the GeoMx Digital Spatial Profiling system (DSP). Multi-label immunofluorescence was used to distinguish tissue morphology in FFPE samples from two health patients. We selected 24 regions of interest (ROIs) from each tissue section included dense, interface and non-dense breast. Using barcoded DNA oligos attached in situ hybridization probes (for RNA) via ultraviolet (UV)- photocleavable linkers, we screened over 1800 genes from Cancer Transcriptome Atlas panel. Our results indicated an elevated expression of CD68 and CD33, markers for macrophages and myeloid-derived suppressor cells (MDSCs) in dense breast regions as compared to the non-dense areas. Identifying markers for myeloid lineage cells was followed by an increased expression of the genes EOMES, TIGIT and RORA, which are engaged in the immunosuppressive phenotype. This could be hypothesized to support a pro-tumorigenesis microenvironment preventing T cells from recognizing newly arisen tumor cells. In addition, we also observed a significant expression of cytokines (IL12A and IL26) and chemokines (CCL19 and CXCL10) in the dense breast, which are involved in the regulation of inflammatory response and chemotaxis of monocytes and T-lymphocytes cells. Both CCL19 and IL26 have been demonstrated to play a role in cell proliferation. Thus, these initial studies suggest that the dense breast microenvironment is potentially a pro-tumorigenic environment that, with sufficient factors, gives rise to tumorigenesis. These data also suggest potential nodes of therapeutic intervention that may lower such risk. Citation Format: Ana Karina de Oliveira, Patcharin Pramoonjago, Christopher A. Moskaluk, Jay W. Fox. Spatial transcriptomic evidence for an inflammatory, pro tumorigenic microenvironment in normal human dense breast tissue [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2023.

Abstract 3806: High dimensional spatial gene expression analysis of tumor micro-environment in head and neck squamous cell carcinomas

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 for a 770 gene expression panel was previously performed utilizing the NanoString PanCancer IO 360 Gene Expression Panel. Specific tumor microenvironment (TME) region gene expression was analyzed using two different gene expression panels with the NanoString GeoMx digital spatial profiling technology, the GeoMx Immune Pathways Panel (84 genes) and the GeoMx Human Whole Transcriptome Atlas (WTA) panel (18,000 genes). For both panels, 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. Results: The IO360 nCounter readout was high-throughput but lacked the spatial definition that can be crucial to understanding the heterogeneity of the TME. The GeoMx Immune Pathways Panel provided counts for 84 targets known to be crucial in the immuno-oncology field, including the PD-1/PD-L1 inhibitor pathway blockade. These counts were successfully mapped to the 12 ROIs on the tissue. The WTA panel provided an unbiased view of 18,000 protein-coding genes with the same ROI spatial resolution as the Immune Pathways Panel along with increased insight into the TME. Conclusions: The technologies described above enable the investigation of the TME for use in biomarker discovery in the field of immunotherapy but can also facilitate the discovery of prognostic markers or mechanism of action studies because of the high dimensional data with spatial context. The detailed analysis of the 18,000 genes provides exhaustive information on the biology of the tumor microenvironment that was formerly limited in spatial resolution. Citation Format: Carlee Hemphill. High dimensional spatial gene expression analysis of tumor micro-environment in head and neck squamous cell carcinomas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3806.

Abstract 1228: Spatial transcriptomics: Data processing revisited to address noise interference

Abstract Background: Spatially resolved transcriptomics is a novel and already highly recognized method that allows RNA sequencing results to be annotated with local tissue phenotypes. The NanoString GeoMx Digital Spatial Profiling (DSP) Platform allows users to collect RNA expression data from manually selected Regions of Interest (ROIs) on FFPE tissue sections. Here, we extensively evaluated data from the DSP platform with its associated pipeline and identify significant background noise interference issues which compromise data interpretation. Alternative and more suitable workflows are presented for correct data analysis. Methods: In this study, 12 paired tumor samples were collected from six glioma patients who underwent two separate resections. For all patients, the first resection was a low grade astrocytoma (WHO grade II or III) and the second resection was a high grade astrocytoma (WHO grade IV). The DSP platform was used to collect expression data of 1,800 genes from 72 ROIs (i.e. 6 per sample). Biological replicates were made of eight tumors from four patients. Gene expression data was normalized with both standard NanoString methods and several alternative methods (e.g. DeSeq2, gamma fit correction and quantile normalization). Weighted Gene Co-expression Network analysis (WGCNA) was used for biological validation. In addition to our own study, six publicly available NanoString DSP datasets were evaluated. Results: Data distributions of all glioma samples, when exposed to standard data processing, were burdened with significant background noise interference. Notably, differences in noise interference were largest between biologically distinct tumor subgroups (i.e. between first and second glioma resections), which was confirmed in replicate experiments. The noise interference patterns were also present in all six publicly available NanoString DSP datasets which will invariably lead to incorrect interpretation of the underlying biology. To correct for noise interference, we tested several normalization methods. The relatively crude quantile normalization method provided the least biased result and showed the highest concordance with bulk RNA sequencing data. To evaluate the biological validity of our alternative approach, we used T cell counts from our tissue regions as an independent parameter, that were quantified using immune fluorescence. Unsupervised WGCNA identified gene clusters enriched for lymphocyte genes that highly correlated with T cell quantities in ROIs, confirming that alternative normalization can extract a biological signal from the DSP platform. Conclusion: The DSP Platform platform suffers from significant noise interference when using standard analysis tools that obscure its results. Here, we revised the workflow and provide an alternative normalization that adequately addresses noise interference and enables correct interpretation of gene expression data. Citation Format: Levi van Hijfte, Marjolein Geurts, Wies R. Vallentgoed, Paul H. Eilers, Peter A. Sillevis Smitt, Reno Debets, Pim J. French. Spatial transcriptomics: Data processing revisited to address noise interference [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1228.

Abstract 3424: Deep mechanistic profiling of immune oncology (I-O) drug combinations in cancer patients with CIVO® intratumoral microdosing and NanoString GeoMx DSP

Abstract Despite over four thousand clinical trials investigating combinations of various anti-cancer agents with immune checkpoint inhibitors (CPI), few have demonstrated significant clinical improvement over CPI alone. This indicates a need for approaches that can guide development of complex (≥ 3 drug) I-O combination treatments by enabling far greater mechanistic understanding of response to drug exposure by components of the authentic tumor microenvironment (TME). Here we show how an approach based on two technologies, the Presage Comparative In Vivo Oncology (CIVO) Platform and the NanoString GeoMx Digital Spatial Profiler (GeoMx DSP), highlights drug synergies as well as mechanisms of resistance to drug exposure in the native and intact TME of cancer patients, providing a path to rapid identification of effective drug combinations. The CIVO platform enables trackable multiplexed intratumoral delivery of microdosed drugs, either as single agents or in combinations, allowing evaluation of the localized tumor response to drug candidates in the TME while capturing tumor heterogeneity and patient diversity. GeoMx DSP is a method for high-plex spatial profiling of mRNAs with rare-cell sensitivity. Combining these technologies in Phase 0 clinical studies in human soft tissue sarcoma patients we evaluated distinct tumor sites microdosed with either nivolumab (anti-PD1), aldesleukin (recombinant IL-2), the combination of nivolumab and aldesleukin, vehicle (no drug), or chemotherapy controls. Using the Nanostring Cancer Transcriptome Atlas we evaluated the effect of drug exposure on expression of over 1,800 genes simultaneously with spatial resolution across individual patient tumors. Synergistic elevation of multiple transcripts was observed at tumor sites exposed to the combination of aldesleukin and nivolumab. This included T-cell specific elevation of granzyme B, a classic biomarker of T-cell activation. Importantly, we show how events such as T-cell activation may be dampened by feedback loops involving immune suppressive responses including, but not limited to, elevation of IDO-1 in non-T-cell components of the TME. Results were observed across multiple patient tumors and were verified by conventional immuno-histochemistry or in situ hybridization. Performing these studies across different patients allows us to explore the inter-individual variation in drug responses. We believe this is an efficient and biologically relevant approach to better understanding feedback loops activated within the authentic TME and ultimately prioritizing drug combinations for clinical development. Citation Format: Jason P. Frazier, Kenneth R. Gundle, Marc Grenley, Gloria Kung, Kimberly Sottero, Kirsten Anderson, Richard Klinghoffer, Jonathan M. Derry. Deep mechanistic profiling of immune oncology (I-O) drug combinations in cancer patients with CIVO® intratumoral microdosing and NanoString GeoMx DSP [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3424.

Abstract 2026: Spatially resolved multiplexed analysisreveals how macrophages adapt to the mantle cell lymphoma microenvironment

Abstract Introduction: Mantle cell lymphoma (MCL) is an incurable disease and we have previously shown that presence of CD163+ cells in the tumor microenvironment in MCL is associated with an adverse outcome. We therefore profiled CD163+ cells to better understand their impact on this tumor and identify possible targetable markers. We hypothesize that, not only the presence of CD163+ cells, a marker of M2 macrophages, in the tumor tissue, but also the spatial context of CD163+ cells might be relevant for MCL development and treatment response. Material and Methods: Sixty-nine protein targets were measured by the Nanostring GeoMx Digital Spatial Profiler. This method allowed for preserved spatial context, in tissue microarrays from a population-based cohort of de novo MCL patients while investigating different markers. A total of 235 cores from 131 different patients were sampled. Regions of interest for CD20+, CD3+ and CD163+ cells were selected, with more than 600 areas of illumination (AOI) collected. From those, 85 AOI, belonging to 61 patients, were from CD163+ cells. These were subsequently classified according to their spatial profile: CD163+ cells within the tumor area and CD163+ cells excluded from the tumor area. Differentially expressed proteins between different tissue types were also explored. Results: Among the highest expressed proteins in the CD163 AOI were CD45, HLA-DR and CD68. The higher expression of both HLA-DR and CD68 suggests that macrophages in MCL microenvironment may display mixed phenotypes between pro- and anti-inflammatory stages, the called M1 and M2 polarization types. Interestingly, p53 and Ki67 were highly correlated among the CD163 AOI collected, which might suggest higher aggressiveness of CD163+ cells. We explored the differently expressed proteins between CD163 AOI sampled located within the tumor and excluded from the tumor/tumor barrier. We identified significant differences in the expressed proteins between the two groups, such as p53, VISTA, ARG1, LAG3 and HLA-DR. p53 was higher in the CD163 AOI within the tumor area, whereas VISTA was highest expressed in CD163 AOI excluded from the tumor cells/tumor barrier. Among the 85 CD163+ AOIs, 47 were sampled from lymph node biopsies, 14 from bone marrow, 10 from tonsil samples, four from the gastrointestinal tract, six from other tissue types and four from unknown origin. Principal component analysis and k-means clustering showed that most of bone marrow and gastrointestinal tract biopsies clustered apart from the remaining samples, with a few differentially expressed proteins, such as NF1 and CD66b. Conclusions: Our study demonstrates that spatial localization within the MCL tumor affects the expression of CD163+ macrophages, adding to the premise that macrophages in tumor are not easily characterized by the concept of M1 and M2 types. Our results shed light on targetable features of the MCL tumor microenvironment. Citation Format: Joana Matos Rodrigues, Lavanya Lokhande, Anna Gerdtsson, Anna Nikkarinen, Peter Hollander, Anna Porwit, Ingrid Glimelius, Mats Jerkeman, Sara Ek. Spatially resolved multiplexed analysisreveals how macrophages adapt to the mantle cell lymphoma microenvironment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2026.

Abstract 1960: Invasive potential of colorectal cancer through tumor budding in 3D

Abstract Background and Objectives: Tumor budding (TB) is defined as isolated single cancer cells or clusters of up to four cells located at the invasive tumor front of colorectal cancer (CRC) and other solid cancers. Despite the growing body of evidence establishing TB as a strong independent prognostic marker in CRC, further investigation is needed to elucidate the biological mechanisms behind TB and its invasive potential. Pseudobudding (PsB), in contrast to TB, is defined as single cancer cells or clusters of up to four cells that occur due to gland disruption and external influences such as inflammation. We hypothesize that the invasive potential of a tumor is therefore better reflected by the active process of TB rather than through the reactive process of PsB. By improved visualization of the invasive tumor front, we can characterize TB and PsB, elucidate their underlying biological processes, and improve future TB scoring accuracy in routine diagnostic practice. Methods: Cases containing TB and PsB were selected from the pathology archives. Thick sections (80-90µm) of formalin-fixed paraffin-embedded CRC tissue were taken using a microtome and stained using a free-floating immunostaining protocol for Pan-cytokeratin (epithelium) and DAPI (nuclei). Thick sections were imaged via confocal microscopy. Three-dimensional z-stacks were processed and analyzed using FIJI software. Tissue microarrays (TMAs) of regions with TB and PsB were created and RNA spatial transcriptomics was conducted on thin TMA sections (5µm) using the Nanostring GeoMx Digital Spatial Profiling Platform. Results: For sections where the invasive margin contained TB, both single cells and cell clusters appear isolated from the core tumor in 3D. These single cells and cell clusters exhibit migratory and invasive features associated with epithelial-mesenchymal transition (EMT), as evident by complete loss of cell-to-cell adhesions and shape shifting through elongation of the entire cell. In contrast, PsB is characterized by cellular debris surrounding a discontinuous epithelial lining. Undivided cells and cell clusters within the damaged area are observed as involuntarily disassociated with lagging remnants of cell-to-cell adhesions. This may be due to inflammatory cells disrupting the extracellular matrix and pulling the cells apart from one another. The identification of differing morphologies allows for further study of the RNA spatial transcriptome (of which the data will be presented during the meeting). Conclusion: Our findings show that TB and PsB are independent features within the invasive tumor margin which represent two separate biological processes. Considering the strong predictive capacity of TB as a marker of poor prognosis in CRC, TB scoring within routine diagnostic practice benefits from avoiding areas with a strong inflammatory reaction and PsB. Citation Format: Tariq Haddad, Luuk van den Dobbelsteen, Alessandro Lugli, Shannon van Lent-van Vliet, Femke Simmer, Inti Zlobec, Iris Nagtegaal. Invasive potential of colorectal cancer through tumor budding in 3D [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1960.

Abstract 3484: Development of a custom high-plex GeoMx digital spatial profiler breast cancer protein biomarker assay

Abstract The goal of this study was to develop a high-plex assay to simultaneously quantitate 27 established and novel breast cancer (BC)-related, immune protein and phosphoprotein biomarkers using the GeoMx® Digital Spatial Profiler (DSP). The custom assay performance was compared to standard clinical BC biomarker assays (e.g.ER, PR, HER2, Ki67) across the spectrum of BC subtypes and in multiple laboratories. Commercially available antibodies to 27 BC-related protein biomarkers, including ER, PR, HER2, Ki-67, AR, immune-related (e.g. PD-L1) and several cell cycle/proliferation markers were oligonucleotide-tagged, and verified by immunohistochemistry for performance against untagged antibodies. The tagged antibodies were combined with 3 isotype controls and 2 housekeeping proteins into a custom BC high-plex assay for DSP. Confirmation of target specificity was done on a custom tissue microarray (TMA) (Run control) composed of (un)treated cancer cell lines and normal tissues. For clinical BC samples, each biopsy had 4x600 µm regions of interest segmented into pan-Cytokeratin+ tumoral epithelium and pan-Cytokeratin- adjacent stroma segments. With targeted UV light, oligonucleotides were collected from each segment sequentially and quantitated with nCounter. Raw counts were geomean normalized for analysis. Intra-site and inter-site assay reproducibility was assessed with the TMA and serial FFPE sections from BC biopsies. Quantitation of proteins showed high reproducibility within sites (e.g. Run 1 v. Run 2, p>0.05 for all protein targets) and the high-plex custom assay showed good concordance between 3 participating laboratories. GeoMx DSP-derived protein data correlated well with orthogonal methodologies performed on a sample subset including immunohistochemistry, reverse phase protein array, DNA and RNA-sequencing. For example, DSP protein counts for p53 protein were higher in tumors with TP53 point mutation than those with truncating TP53 alterations. Expanded sample testing is underway with 120 well-characterized breast cancer biopsies of various histologies, immune subtypes (e.g. TIL scores) and clinical biomarker phenotypes (e.g. ER+/HER2-/Ki-67 high BC; ER+/HER2-/Ki-67 low BC; HER2+ BC; ER-/PR-/HER2- BC).Our preliminary data demonstrate that this custom high-plex BC assay can quantitate protein biomarkers across a wide dynamic range with high intra-lab and inter-lab reproducibility. The assay requirement of a single 5-µm tissue section facilitates complex biomarker profiling in biopsies with limited material. The custom assay alone or in combination with other targeted DSP protein modules can simultaneously interrogate the biomarker profile, immune-based and other drug target profile, and immune microenvironment of BC specimens, providing a novel approach for actionable tumor subtyping. For research use only. Not for use in diagnostic procedures. Citation Format: Christopher Corless, Amber Bridgeman, Jinho Lee, Gary Geiss, Sarah Warren, Sarah Church, Joseph M. Beechem, E Aubrey Thompson, Jodi Carter. Development of a custom high-plex GeoMx digital spatial profiler breast cancer protein biomarker assay [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3484.