Citations

Abstract Glioblastoma (GBM) is a high grade glioma that accounts for the majority of primary malignant brain tumors. The standard of care for GBM consists of maximal surgical resection, concurrent radiation and temozolomide (TMZ), and adjuvant TMZ. However, this has only led to a modest increase in median patient survival. TROY (TNFRSF19) is an orphan member of the TNF receptor superfamily that is differentially upregulated in GBM with low expression in normal brain tissue. Previously, we have reported that TROY is overexpressed in GBM and signals to mediate cell invasion and therapeutic resistance. Propentofylline (PPF) is a well-studied synthetic methylxanthine derivative that has strong BBB penetrance. We have shown that PPF decreases TROY expression in gliomas, inhibits invasion, and sensitizes GBM to TMZ in vitro. Here we examined the efficacy of PPF in combination with TMZ using in vivo studies of intracranial models of GBM patient-derived xenograft (PDX) models in murine. We tested 3 models of GBM PDX cells with various TROY expression levels (GBM8-high TROY; GBM43-moderate TROY; GBM22-low/no TROY). Our intracranial PDX experiments showed that PPF treatment alone did not increase survival in all PDX cells. The combination of PPF and TMZ treatment did provide a survival benefit (p < 0.039) relative to TMZ alone (only in GBM8-high TROY and GBM43-moderate TROY) that was partially lost when treatment was discontinued. However, continuous PPF administration following a concomitant PPF and TMZ treatment course led to extended survival (p < 0.0015) compared to TMZ alone resulting in increased DNA damage (γH2AX) and apoptotic markers (Cleaved Caspase-3). Immunohistochemical analyses showed that TROY expression was decreased in tumors treated with PPF. In addition, PPF-treated tumors showed a differential myeloid population with well circumscribed tumors as compared to control-treated or TMZ-treated tumors. Thus, our data demonstrates that PPF may provide a pharmacological approach to targeting TROY concomitant with standard-of-care to increase patient survival. Citation Format: Ryan Eghlimi, David Nascari, Angad Beniwal, Drake Alton, Dustin Grief, Joseph Loftus, Melanie Haluska, David Coleman, Nhan Tran. Propentofylline enhances GBM sensitivity to temozolomide via targeting TROY signaling [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 5905.

Abstract Background: Cholangiocarcinoma (CCA) is a lethal and heterogenous malignancy of the biliary tree characterized by perineural invasion. About 85% of CCA show activation of YAP signaling, which promotes proliferation, anti-apoptosis, and therapeutic resistance. YAP signaling is mediated by Src family kinases (SFK) by phosphorylation of YAP (pYAPy357). NXP900 is a first-in-class, highly selective, SFK inhibitor with a novel mechanism of action which locks SFK in its closed and inhibited conformation, in contrast to other Src inhibitors such as dasatinib. A phase 1a study of NXP900 in patients with advanced solid tumors was recently initiated. Here, we examined treatment responses of CCA to NXP900 in vitro and in vivo. Methods: Cell viability was determined in seven CCA cell lines by CellTiter-Glo. Cell death and apoptosis were assayed by PI/Hoechst and Caspase-Glo 3/7 assay. CalcuSyn software was used to determine synergistic drug effects. Immunoblot analysis, RT-PCR, and IF staining of YAP localization were used to evaluate YAP knockdown. Clones of human CCA cell lines resistant to NXP900 were generated through escalating exposure to NXP900 over 6 months. Five patient-derived xenograft (PDX) tumors were expanded into flanks of NOD/SCID mice. Tumor bearing mice were randomized 1:1 with 5 mice per arm and treated with vehicle or NXP900 (40 mg/kg) once daily for up to 4 weeks. To determine drivers of sensitivity and resistance, predicted sensitivity scores based on tumor growth were linked to multi-omics (RNA seq and [phospho-]proteomics) of our PDX models. Results: All cell lines were sensitive to NXP900 with IC50 values between 7nM-15µM; IC50’s of resistant clones were 1000 times higher. NXP900 induced more cell death compared to vehicle, which appeared to be through apoptosis. NXP900 inhibited pSrc, decreased pYAPy357, and upregulated inactive pYAPs127. Correspondingly, decreased YAP target gene (Cyr61, NUAK, and CTGF) levels and a nuclear-to-cytoplasmic translocation were observed. NXP900/GemCis combination therapy increased cell death demonstrated a synergistic effect (Combination Indices <1) at all concentrations. In our PDX models, treatment was associated with a significant decrease in tumor growth in 3 models (mean fold change 3.1 vs. 14.2; 1.0 vs. 7.6; 1.4 vs. 2.6). The major resistant signature was the TRK signaling network, involved in nerve growth factor binding activity, while drivers of sensitivity included the SRC network. Conclusion: NXP900 demonstrated therapeutic activity in vitro and in human PDX models. We are currently performing multi-omic approaches in NXP900 sensitive and resistant cell lines to unravel determinants of activity and resistance. Additional in vivo studies will be performed to determine effects of NXP900/GemCis and NXP900/anti-PDL-1 combination therapy. Citation Format: Hendrien Kuipers, Jennifer L. Tomlinson, Danielle M. Carlson, Amro M. Abdelrahman, Erik Jessen, Jack W. Sample, Nathan W. Werneburg, Hannah E. Stumpf, Mark J. Truty, Sumera I. Ilyas, Gregory J. Gores, Rory L. Smoot. Src family kinase inhibition demonstrates antitumor activity in vitro and in patient-derived xenograft models of human cholangiocarcinoma [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 614.

Abstract Background: Nutritional and inflammatory factors collected in routine cancer labs, neutrophil percentage to albumin ratio (NPAR), neutrophil lymphocyte ratio (NLR), monocyte lymphocyte ratio (MLR), and platelet lymphocyte ratio (PLR), have been shown feasible in identification of patients at risk of increased treatment complications, decreased cognition, anxiety and depression, and increased mortality. These have proven sensitive to multiple cancer types and stages, though little is known about causal pathways. The Vitamin/Mineral and Inflammatory Correlations To Understand Chronic Stress in the U.S. (The VICTUS Study) sought to elucidate a relationship between diet and food insecurity within these panels. Methods: Using NHANES 2017/2018 surveys, adults > 18 years without renal or hepatic disorders, pregnancy, alcohol, or supplement use were included. NPAR, NLR, MLR, and PLR were categorized as quartiles and compared to demographic, food security, plasma micronutrient statuses [vitamins A, B9, C, D, E, calcium (Ca), ferritin (Fe), phosphorus (P), potassium (K), sodium (Na)], and smoking, activity level, and sun exposure, using chi-square and ANOVA. ANOVA models compared food security and poverty to NPAR, NLR, MLR, and PLR means. Results: From 3,317 participants, 50.2% were male, 54.3% non-Hispanic White, 33.7%, food insecure, 51.1% food stamp recipients, 27.7% smokers, with a mean age of 42.5 (SD 0.44), poverty to income ratio 2.79 (SD 0.08), CRP 3.90 (SD 0.18), NPAR 14.1 (SD 0.14), NLR 2.07 (SD 0.04), MLR 0.28 (SD 0.003), and PLR 118.7 (SD 1.61). CRP was found to be statistically significantly associated with all biomarker ratios (p < 0.018). Significant micronutrient biomarker-inflammatory biomarker associations were identified in Fe [NPAR, MLR, PLR (p < 0.001)], P [NPAR, NLR, MLR, PLR (p < 0.033)], C [NPAR, NLR (p < 0.008)], E [MLR, PLR (p < 0.016)], K [NPAR, NLR, MLR, PLR (p < 0.015)], Ca [NPAR, PLR (p < 0.04)], A [NPAR (p < 0.023)] and D [NLR (p < 0.035)]. Significant micronutrient intake-inflammatory biomarker associations were detected in vitamin A [NPAR, PLR (p < 0.013)], lycopene [NLR, PLR (p < 0.035)], and B12 [NLR, PLR (p < 0.031)]. Age [NPAR, NLR, MLR, PLR (p < 0.001)], gender [NPAR, MLR, PLR (p < 0.001)], ethnicity [NPAR, NLR, MLR (p < 0.006)], and sedentariness [MLR, PLR (p < 0.039)] were also associated with inflammatory markers. Poverty was significantly associated with NPAR, NLR, MLR, and CRP (p < 0.018). Food insecurity was significantly associated with NPAR (p < 0.001) and CRP (p < 0.001). Discussion: Food insecurity and poverty were significantly associated with NPAR, a ratio implicated in overall poorer cancer health outcomes. Multiple inadequate micronutrient intakes and plasma biomarkers were associated with inflammatory panels, though neither nutritional biomarker nor food insecurity statuses and their contributions to inflammatory processes are routinely surveilled within cancer disease and treatment. Citation Format: Jennifer Crook, Eunkyung Lee, Rui Xie, Opeyemi Bolajoko, Victoria Loerzel. Nutritional mediators of inflammation: A cause for increased surveillance [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 2171.

Abstract Background: Advanced prostate cancer is defined as metastatic hormone-sensitive prostate cancer (mHSPC) or castration-resistant prostate cancer (CRPC). Newer and aggressive treatments are leading to increasing lifespans in this population. As such, studying quality of life could have a large impact on the survivorship experience. An important component of quality of life is financial toxicity, defined as the financial burden and distress that can arise for patients and their families because of cancer treatment. Objectives: To understand the prevalence and predictors of financial toxicity in advanced prostate cancer. Methods: IRONMAN (International Registry of Men with Advanced Prostate Cancer) is a prospective cohort of patients with newly diagnosed mHSPC or CRPC enrolling in 15 countries. Financial toxicity was collected using the EORTC QLQ-C30 (“has your physical condition or medical treatment caused you financial difficulties?”) at enrollment and every three months thereafter. For this analysis, we studied 997 patients cared for at 37 US sites enrolled from 7/2017 to 7/2023. We examined baseline and longitudinal reports of financial toxicity across clinical and demographic factors: age at enrollment, race and ethnicity, disease state, marital status, education, employment, military status, and type of health center. We conducted adjusted risk difference analysis and multivariable regression modeling to identify predictors of financial toxicity using a generalized estimating equations approach to calculate odds ratios (OR and 95% confidence intervals, CI). Results: At enrollment, 125 (13%) patients reported experiencing “quite a bit” or “very much” financial toxicity. 21% of Black patients experienced toxicity compared to 10% of White patients. Even after incorporating all other predictors, Black patients experienced 5% (95% CI: 0, 11) greater toxicity than White patients. Independently of other predictors, patients >75 experienced 11% (CI: -17, -4) lower toxicity compared to those <65 years (while patients in the 65-74 group experienced 6% less, CI: -12, 1). Toxicity was 6% greater in patients with mHSPC (CI: 2, 10) than in patients with CRPC, and disabled or unemployed patients were at substantially higher risk. In the multivariable longitudinal analyses, older age (> 75) remained associated with significantly lower odds for financial toxicity (OR: 0.30, 0.15-0.61) while mHSPC (OR: 1.91, 1.14-3.20) and being unemployed (OR: 2.69, 1.03-6.99) or disabled (OR: 5.14, 2.29-11.6) were associated with higher odds. Conclusions: Financial toxicity is prevalent in advanced prostate cancer, especially in patients who are Black, younger than 75, diagnosed with mHSPC, or unemployed/disabled. These results may inform identification of patients at high risk for financial toxicity to support delivery of patient-situation-informed care by clinicians and targeted aid for patients at highest risk. Citation Format: Isani Singh, Bailey Vaselkiv, Emily Rencsok, Karen Autio, Deborah Enting, Alicia Morgans, Joel Nowak, Kevin Kensler, David Nanus, Robert Dreicer, Brian Lewis, Oliver Sartor, Elisabeth Heath, Phillip Kantoff, Deborah Kaye, Daniel George, Lorelei Mucci. Financial toxicity among patients with advanced prostate cancer in the United States [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 3410.

Abstract Introduction: Cholangiocarcinoma (CCA) is an aggressive and heterogenous biliary malignancy. Previous therapeutic shortcomings have encouraged the investigation of new strategies, such as targeted therapeutics. YAP/TAZ, transcriptional coactivators of the hippo pathway, have been implicated in the tumorigenesis of CCA along with the Src familiar kinase, LCK. Milk-derived nanovesicles (MNVs) have emerged as a promising biologic delivery apparatus for hepatobiliary malignancy due to their predilection for hepatic uptake. We sought to investigate the utility of aptamer guided MNVs loaded with short interfering RNA (siRNA) targeting YAP, TAZ, and LCK. Methods: MNVs were prepared and loaded with species-specific siRNA prior to decoration with an epithelial cellular adhesion molecule (EpCAM) targeting aptamer. In vitro studies were performed using murine and human CCA cell lines, SB1 and HuCCT-1, respectively. In vivo studies were carried out using C57BL/6 mice following Institutional Animal Care and Use Committee protocols. Results: We previously validated EpCAM as an effective aptamer demonstrated by increased levels of expression in CCA cell lines using immunoblot, RT-PCR and immunofluorescence. Additionally, in vivo experiments demonstrated excellent biodistribution and affinity for tumor-selective uptake. To evaluate knockdown efficacy, CCA cell lines were incubated with EpCAM aptamer guided MNVs individually targeting YAP, TAZ, and LCK which resulted in downregulation of these targets on immunoblot and RT-PCR. MNVs were then loaded with equal concentrations of siRNA targeting YAP, TAZ, and LCK in combination which resulted in similar degrees of gene downregulation compared with single transcriptome targeting MNVs. Due to the role of YAP/TAZ activation in CCA systemic therapy resistance, we investigated the effect of MNVs loaded with either single target or combinatory target siRNA alongside and in combination with gemcitabine and cisplatin (GemCis). Incubation with both combinatory siRNA and GemCis significantly intensified cell death on Pi/Hoechst staining compared with single target knockdown alone. Similar findings were found using separate assays of cell viability, including ATP luminescent quantitation and caspase 3/7 activity. These data suggest MNV transcriptome targeting may induce a state of cell stress which allows for amplification of chemotherapy treatment effect. Conclusion: Aptamer guided, nanovesicle mediated transcriptome targeting is a promising therapeutic strategy in preclinical CCA models that may potentiate chemotherapy response. Our future direction involves further investigating the role of CCA target downregulation and its effect on the tumor immune microenvironment utilizing aptamer guided nanovesicle delivery and inducible knockdown cell lines. Citation Format: Jack W. Sample, Mincheng Yu, Hendrien Kuipers, Danielle M. Carlson, Nathan W. Werneberg, Jennifer L. Tomlinson, Gregory J. Gores, Rory L. Smoot. Milk-derived extracellular nanovesicles: A novel delivery strategy for transcriptome targeting therapies in cholangiocarcinoma [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 5753.

Abstract Lung cancer is the leading cause of cancer deaths in the US. Non-small cell lung cancer (NSCLC) accounts for ~80% of lung cancer cases and is further classified into adenocarcinoma (LUAD), squamous cell carcinoma (LUSC), and large cell carcinoma. New approaches targeting pathways in NSCLC have emerged, resulting in encouraging new treatments that benefit LUAD patients, but very few advances have been made in the treatment options for LUSC which accounts for ~30% of lung cancer cases. NSCLCs, including LUSCs, contain a sub-population of cells that exhibit properties of cancer stem cells (CSCs) and are responsible for tumor initiation, maintenance, therapeutic resistance, and relapse. Therefore, CSCs must be effectively targeted to elicit long lasting therapeutic results in LUSC patients. Our studies have shown that: 1) PKCι is an oncogene and prognostic marker in LUSC; 2) copy number gains (CNGs) of the PKCι gene (PRKCI) drives overexpression of PKCι in 90% of LUSCs; 3) PKCι phosphorylates SOX2, and controls SOX2-mediated transcriptional activation of Hedgehog Acyl Transferase, resulting in Hedgehog (Hh) pathway activation that drives a CSC phenotype in LUSC cells; and 4) the FDA approved compound Auranofin (ANF) is a potent inhibitor of oncogenic PKCι. In this study, we assessed the effect of ANF in combination with the Hh pathway inhibitor, LDE225, on LUSC CSCs and tumor growth. Cell viability, sphere formation and soft agar assays were used to assess the effects of ANF and LDE225 alone or in combination on the growth of LUSC CSCs with PRKCI CNGs or LUAD and LUSC CSCs without PRKCI CNG to determine if PRKCI CNG can be used to predict response to ANF and LDE225 and test the selectivity of response for LUSC. The effects of ANF and LDE225 on the levels of PKCι-Hh signaling components was assessed by qPCR and western blotting. Finally, the efficacy of ANF and LDE225 in combination was evaluated in human primary LUSC patient-derived xenograft (PDX) models that harbor PRKCI CNGs and the ability of PKCι-Hh signaling intermediates to serve as biomarkers of PKCι-Hh pathway inhibition was validated in vivo. We observed a dose-dependent decrease in cell proliferation, clonal expansion and transformed growth in CSCs treated with ANF and LDE225 alone. ANF and LDE225 combined treatment of LUSC CSCs with PRKCI CNGs (but not LUAD or LUSC CSCs without PRKCI CNG) synergistically inhibited CSC phenotypes based on results of combination index analysis. Furthermore, treatment of LUSC PDX models with ANF and LDE225 in combination had a synergistic effect on tumor growth inhibition. Lastly, ANF and LDE225 treatment led to inhibition of PKCι-Hh signaling intermediates demonstrating the on-target effects of treatment. Our data indicate that a PKCι-Hh vertical blockade strategy of combined ANF and LDE225 effectively targets LUSC CSCs and blocks tumor growth in vivo and provides compelling rationale to develop this novel combination strategy further in clinical trials. Citation Format: Dania Alqasrawi, Ryan Argo, Christine Ratliff Rang, Prita Pandya, Skyeler Klinge, Jennifer Lindemann, Nyla Searl, Alan Fields, Verline Justilien. A novel combination therapy for lung squamous cell 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 5434.

Abstract Glioblastoma (GBM) tumors contain a subpopulation of glioma stem cells (GSC) which can escape therapeutic effects through self-renewal and invasion. Epidermal growth factor receptor (EGFR) mutations are the most common alteration in GBM (50%). One example includes extracellular domain missense mutations like G598V, A289V, and R108K (10-15%). These extracellular domain missense mutations increase ligand binding affinity, inducing a highly active receptor. Earlier reports have shown these EGFR missense mutations exhibit therapeutic resistance; however, the molecular mechanism of these missense mutations in GBM remains unknown. In this study, we show high oligodendrocyte transcription factor 2 (OLIG2) expression in GSCs with EGFR-G598V and EGFR-A289V missense mutations. OLIG2 is universally expressed in glioma and essential for reprogramming GBM cells into stem-like tumor cells. Here, we report GSCs with these point mutations display increased self-renewal, growth, invasion, and therapeutic resistance. In addition, EGFR-G598V and EGFR-A289V trigger distinct downstream signaling pathways in GSCs to regulate OLIG2 function. These missense mutation(s) also represent different pathway-based GBM subclassifications, where A289V induces a Neuronal subtype and G598V induces a Glycolytic/Plurimetabolic subtype. We hypothesize that GBM tumors harboring EGFR missense mutations (G598V, A289V) influence stem cell properties via multiple pathways to evade therapeutic effects and inhibiting these different signaling pathways will decrease therapeutic resistance in GBM. Understanding the mechanism of these EGFR mutations may lead to improved therapeutic strategies for GBM patients. Citation Format: Sidney Weaver, Aunay Miller, Mylan Blomquist, James McNamara, Keely Orndorff, Chris Sereduk, Jean Kloss, Joseph Loftus, Shwetal Mehta, Nhan L. Tran. EGFR missense mutation(s) induce OLIG2 expression to regulate glioma stem cells maintenance and therapeutic resistance [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 5444.

Abstract Glioblastoma (GBM) is the most common and lethal primary brain tumor in adults, despite current treatment options. One of the major contributing factors to this dismal outcome is a subset of cells called glioma stem cells (GSCs) that are highly tumorigenic, invasive, and resistant to conventional treatments. Thus, studies that focus on the properties of GSC biology are needed to develop novel therapies to improve overall survival. Here we report a novel role of STAT5 in GSC stemness and survival. In GBM tumors, STAT5 has been shown to be active in the infiltrative, residual population of cells remaining post-surgery. Knockdown of STAT5 results in the decrease in GSC stemness and therapeutic resistance. Interestingly, we find that expression of OLIG2, a key cell fate factor important for the tumorigenic potential of GSCs, affects STAT5 activation with a positive correlation between OLIG2 and STAT5 mRNA expression in GBM tumors. Furthermore, we find that in GSCs expressing EGFRvIII, STAT5 interacts with OLIG2 and pharmacological inhibition of STAT5 activation or knockdown of STAT5 expression results in decrease of OLIG2-mediated GSC invasion. Together, these data suggest a novel transcriptional complex, STAT5-OLIG2, in the potential role for GSC survival, maintenance, and invasion. Citation Format: Aunay Miller, Costanza Lo Cascio, Mylan Blomquist, James B. McNamara, Keely Orndorff, Matthew Dufault, Christopher Sereduk, Joseph Loftus, Shwetal Mehta, Nhan L. Tran. A novel transcriptional complex of OLIG2-STAT5 mediates glioma stem cell self renewal and invasion [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 256.

Abstract Background: Neuropilin 2 (NRP2) is a co-receptor that enhances signaling of growth factor and cytokine receptors and is expressed in tumor and microenvironment (TME) cells in melanoma and other tumors. Neuropilin 2 expression in tumor cells is associated with more aggressive tumors. Expression in tumor associated macrophages enhances efferocytosis and immunologically silent tumor cell death. Methods: Whole excisional lymph node biopsies were obtained from 10 treatment naïve patients with metastatic cutaneous melanoma who subsequently received anti-PD1 treatment. A single 5mm formalin-fixed paraffin embedded tissue section was used to assess a panel of 39 analytes by cyclic multiplex immunofluorescence (MxIF). Fields of view (~1mm2, n=288) were selected from pathologist-annotated regions of the tumor core, tumor-immune interface and adjacent lymphoid tissue. Pixel-based single cell segmentation and a supervised classifier approach was applied to resolve 12 distinct tumor, stromal and immune cell phenotypes and functional states (NRP2 positive or negative) in 1.5 million cells. Results: Tumor recurrence was observed in 6 patients (relapsed), whereas 4 patients were recurrence-free (non-relapsed) during follow up. Among the different cell types, NRP2 expression was more frequent in macrophages (44%), tumor cells (40%), dendritic cells (40%), followed by Tregs (31%) and T helper cells (29%), and less frequent in neutrophils (24%), cytotoxic T cells (23%) and B cells (15%). In colocalization permutation analysis, NRP2 expressing macrophages clustered away from tumor cells in non-relapsed patients while a random spatial distribution of NRP2 expressing macrophages with respect to tumor cells was observed in relapsed patients. In comparing the odds of tumor cell interactions with NRP2-expressing macrophages to those with NRP2 non-expressing macrophages, we found in the tumor:stromal interface (10% tumor cells) that cells from both non-relapsed and relapsed patients were more likely to interact with NRP2 non-expressing macrophages (OR = 0.73, 95% CI = [0.71, 0.75] and OR = 0.56, 95% CI = [0.55, 0.58] for non-relapsed and relapsed patients, respectively). However, in the tumor core (90% tumor cells), tumor cells from relapsed patients had significantly higher odds of interacting with NRP2 expressing macrophages, (OR = 1.15, 95% CI = [1.12, 1.18]), and this was not seen for non-relapsed patients (OR = 0.68, 95% CI = [0.66, 0.71]). In FOVs dominated by tumor cells (75% tumor), tumor cells from relapsed patients had 4.97 times the odds of expressing NRP2 compared to tumor cells from non-relapsed patients (95% CI = [2.23, 11.09], p = 0.0001). Conclusions: These spatial analytics suggest complex interactions among NRP2 expression in tumor and TME of melanoma determining risk for relapse with anti-PD1 treatment. Citation Format: Anastasios Dimou, Raymond M. Moore, Caitlin Ward, Alexey Leontovich, Ruifeng Guo, Chen Wang, Shankar Suman, Betty Dicke, Jill M. Schimke, Noah A. Stueven, Chathu A. Atherton, Wendy Nevala, Svetomir N. Markovic. Spatial analysis of neuropilin 2 expression in the microenvironment of 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 2905.

Abstract Background: Recent studies indicate a correlation between increased stromal composition in tumors and resistance to immune checkpoint inhibitors (ICIs) across various cancer types. Specifically, the extracellular matrix (ECM) within the stromal compartment significantly influence the tumor microenvironment, affecting cancer cell proliferation and immune cell recruitment. However, identifying and quantifying the ECM at the collagen fiber level in whole slide images (WSIs) is challenging due to its inherent heterogeneity and complexity. In this work, we utilized statistical analysis and a deep learning (DL)-based approach to identify and assess collagen in the ECM, examining its characteristics in relation to the ICIs treatment response in gastric cancer. Methods: We analyzed 116 WSIs of hematoxylin and eosin (H&E) stained slides before ICIs treatment from three institutions. Responders were defined as those with a complete response or partial response (N=18); those with stable disease (SD) (N=34) or progressive disease (PD) (N=64) were considered non-responders (N=98). For non-responders, post-treatment H&E slides were available for 21 patients; SD (N=5) and PD (N=16). We applied a DL-based second harmonic generation (SHG) synthesis algorithm to tumor regions identified by a tumor detection model. We then calculated various collagen properties, including volume, fiber orientation and alignment, waviness, straightness, and fiber thickness, based on collagen fiber centerlines extracted from the generated SHG images. An attention-based multiple instance learning (ABMIL) model was trained to automatically predict the treatment response solely based on collagen features. The dataset of 116 pre-treatment slides was divided into training and validation subsets using stratified 3-fold cross-validation. The area under the curve (AUC) was used to assess the performance of the trained models. Results: Collagen fiber features were significantly associated with ICIs response. Notably, ICIs responders have a decrease in collagen waviness and fiber orientation variation (p-value<0.05, by Mann-Whitney U test). Among non-responders, the PD group showed a statistical decrease in waviness after treatment (p-value<0.05, by Mann-Whitney U test). The ABMIL model achieved a mean AUC score of 0.823, with a standard deviation of 0.023. Conclusion: This study provides a comprehensive analysis of collagen in the ECM, and its properties in relation to gastric cancer ICIs response within Tumor Immune Microenvironment. We also developed a model using collagen features and demonstrated high accuracy in predicting treatment response. Citation Format: Minji Kim, Ekaterina Redekop, Yoonho Choi, Sung Hak Lee, Jae-Ho Cheong, Yu Aoki, Kohei Shitara, Tae Hyun Hwang. AI-driven analysis of collagen characteristics in the tumor immune microenvironment predicts immune checkpoint inhibitors treatment responsiveness in gastric cancer [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 4262.