Predictive Biomarkers Of Response Research Articles

Abstract 1178: Tissue nanomechanics as a novel, clinically translatable predictive early biomarker of response in combination immunotherapy

Abstract Background: Recently, the successful integration of immunotherapy with other treatment modalities has demonstrated its efficacy in enhancing immune activation and suppressing tumors in solid cancers. The most promising of such efforts include the use of low dose radiation (LD-XRT) to prime the tumor microenvironment and enhance the T cell infiltration and activation [Barsoumian HB, et al., 2020; Patel RR, et al., 2021; He K, et al., 2023]. With these advancements, there is an urgent need to develop clinical biomarkers to predict response and optimize treatment decisions. Recent evidence underscores the importance of nanomechanical alterations within the tumor microenvironment as mechanistic indicators of aggressiveness [Plodinec M, et al., 2012]. This study aims to demonstrate the value of ARTIDIS nanomechanical signature as early predictor of elucidate of response to combination LD-XRT/immune checkpoint inhibitors (ICI). Our measurement protocol in mice follows the same settings for biopsies acquired in routine biopsy collection protocol, thus supporting prompt clinical translation and integration. Methods: We established a mouse model of 344SQ lung adenocarcinoma tumors that exhibited resistance to anti-PD1 treatment in 129Sv/Ev mice. These mice underwent treatment protocols involving a combination of anti-PD1 and anti-CTLA4 antibodies, both with and without LD-XRT pre-treatment. Throughout the study, we closely monitored survival rates and tumor growth across diverse experimental groups. Employing the ARTIDIS platform, a cutting-edge technology that integrates atomic force microscopy with proprietary artificial intelligence algorithms, we extracted a multiparametric nanomechanical signature and we evaluated its outcome prediction value. This signature was then complemented by histopathology, multiplex immunofluorescence, and Nanostring assays to elucidate mechanisms of stroma remodulation and immune infiltration. Results: This study identifies the unique nanomechanical signature predictive of response to immunotherapy alone or in combination with radiation. Using the tissue nanomechanical signature as a predictive biomarker, we achieved precise differentiation between responders and non-responders, boasting a 90% sensitivity, 99.1% specificity, and a 96% AUC. To our knowledge, this study is the first to demonstrate the response-predicting power of a tissue derived nanomechanical signature for immune checkpoint inhibitors administered in combination with LD-XRT. These findings support clinical translation of the ARTIDIS nanomechanical signature as a pivotal tool for clinically predicting and evaluating responses to integrated radiation and immunotherapy approaches. Citation Format: Gitika Srivastava, Sara Nizzero, Mark Wasley, Mingee Kim, Kathleen Graham, Papa Diogop Ndiaye, Mariam Gachechiladze, Nahum Puebla-Osorio, Philipp Oertle, Tobias Appenzeller, Vittorio Cristini, Marko Loparic, Marija Plodinec, James W. Welsh. Tissue nanomechanics as a novel, clinically translatable predictive early biomarker of response in combination immunotherapy [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 1178.

Abstract 2523: Predictive biomarkers for neoadjuvant FLOT4 chemotherapy response in gastroesophageal cancer: Results of multi-omics approach

Abstract Background: The FLOT4 chemotherapy regimen is a standard preoperative treatment for gastroesophageal junction (GEJ) and gastric cancer (GC), yet the response to this treatment is variable. Identification of predictive biomarkers is crucial for optimizing therapeutic strategies. This study aims to identify predictive biomarkers for FLOT4 treatment efficacy using a multi-omics approach. Methods: We conducted a preliminary differential gene expression analysis in diagnostic tissue pre-treatment and surgical resection post-treatment in six patients (two GEJ and four GC) who all exhibited a partial pathological response (G2) to neoadjuvant FLOT4 chemotherapy. Utilizing DESeq2 for differential gene expression and David pathway analysis for functional insights, we have begun to unravel the complex molecular response to treatment. Upcoming assays include blood gDNA and tumor tissue whole exome sequencing (WES), comprehensive methylation profiling, miRNA characterization, and proteomics via LC-MSMS. Results: Our preliminary data suggest alterations in gene expression associated with cell adhesion, extracellular matrix interaction, and immune response pathways. Three common genes (CXCL2, CXCL3, CXCL1) were identified among upregulated pathways in GEJ patients, while one common gene (IL1A) was observed in GC patients. These findings suggest a complex network of biological processes influenced by FLOT4. In-depth results, including whole-exome sequencing, methylation patterns, and miRNA profiles, will be completed and presented at the conference. Conclusion: These initial findings propose that changes in gene expression related to cellular architecture and immune response are indicative of a partial response to FLOT4 in both GEJ and GC patients. The identified genes and pathways may serve as potential biomarkers for predicting treatment outcomes. Future Directions: By the time of the conference, we anticipate including a more extensive patient cohort with complete response or no response to FLOT4, and comparative analyses with a FOLFOX treatment group. The integration of forthcoming WES, methylation, miRNA, and proteomics data will provide a more comprehensive molecular predictive model. Keywords: FLOT4 chemotherapy, gastroesophageal junction cancer, gastric cancer, predictive biomarkers, differential gene expression. Citation Format: Amira Al-Fahdi, Shoaib Al Zadjali, Ibrahim Al Haddabi, Anoopa Pullanhi, Muna Al Jabri, Aida Al Yayahee, Nada Al Shuaili, Nahad Al Mahrouqi, Mansour Al Moundhri. Predictive biomarkers for neoadjuvant FLOT4 chemotherapy response in gastroesophageal cancer: Results of multi-omics approach [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 2523.

Abstract 7610: Predictive biomarkers for immuno-chemotherapy response in biliary tract malignancies

Abstract Background: Biliary tract cancers (BTCs) refer to invasive malignancies of the biliary tree. Lack of specific symptoms in early-stages frequently leads to late-stage diagnosis, reducing therapeutic options. Surgical resection remains the only curative option for early-stage disease; instead, patients with advanced disease usually receive chemotherapy (cisplatin plus gemcitabine) in combination with durvalumab as first-line treatment and mFOLFOX-6 as second-line, but effects are limited and only a small proportion of patients respond to treatment. Recent advances in molecular characterization of BTCs allowed the identification of several alterations (e.g. IDH1/2 mutations, FGFR2 rearrangement) for which targeted agents are now available in second line setting but only for a small subset of patients. However, prognosis for patients with advanced stage disease remains poor. Treatment failure depends on several factors such as intra- and inter-tumor heterogeneity and the complex immunological landscape. Therefore, the identification of predictive biomarkers able to individuate patients that will benefit from immuno-chemotherapy will be crucial. Methods: BTC patients at any stage were enrolled in two arms: A) BTCs undergoing surgery; B) advanced/relapsed BTCs. For this study, patients belonging to arm B were selected, in particular those who received first-line immuno-chemotherapy. Blood samples were collected before and after three months of therapy. Ascites samples were collected for patients with advanced disease who underwent paracentesis during therapy. DEPArray NxT (Menarini Silicon Biosystems) was used for phenotypic analysis and single ascites-derived cell isolation. The Ampli1 Lowpass kit was used for copy number aberration (CNA) profiling. Extracellular vesicles (EVs) were isolated from plasma at baseline and after three months of immuno-chemotherapy by size-exclusion chromatography and analyzed for size distribution and concentration using Nanosight NS300. Surface EV markers were analyzed by flow cytometry using the MACSPlex Exosome Kit human (MiltenyiBiotec). Results: We enrolled 43 patients starting from August 2022: 14 in arm A and 29 in B. 23 patients from arm B received immuno-chemotherapy and 3 underwent paracentesis during therapy. Patient-derived organoids (PDOs) were successfully established from tumor cells recovered from ascites of these patients. Epithelial cells recovered from ascites displayed aberrant CNA profiles with gains in chromosomes 1q, 8q, 10, 12p, 20, and losses in 1q, 3q, 4, 8q, 9 and X. EV size distribution and concentration were similar before and after treatment as well as surface markers exposure. Further analyses in correlation with clinical features of patients will be performed. Conclusions: Our approach will contribute to identify potential predictive factors associated with response/resistance to immuno-chemotherapy in BTCs. Citation Format: Michele Zanoni, Tania Rossi, Martina Valgiusti, Giuliano La Barba, Michela Cortesi, Sara Bandini, Giulia Bartolini, Alessandra Dubini, Paolo Di Gioia, Chiara Gallio, Luca Esposito, Alessandra Virga, Giorgia Gurioli, Sara Bravaccini, Giovanni Martinelli, Paola Ulivi, Giovanni Luca Frassineti, Giorgio Ercolani, Ilario Giovanni Rapposelli. Predictive biomarkers for immuno-chemotherapy response in biliary tract malignancies [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 7610.

Abstract 2689: Mass cytometry analysis of CXCR3 expression in tumor-infiltrated myeloid cells as a predictive biomarker For immunotherapy response

Abstract Monocytes have emerged as important regulators of cancer progression. We have recently found that CXCR3+ circulating monocytes in humans are associated with better response to immunotherapy. Here, we aimed to study the functions and heterogeneity of CXCR3+ monocytes in two different triple negative breast cancer (TNBC) tumors in mice, using either AT3 or EO771 tumor cell lines. AT-3 tumors exhibit higher resistance to anti-PD-1 and anti-CTLA-4 therapy in C57BL/6 breast cancer mouse models compared to EO771 tumors. We evaluated 36 surface protein markers on tumor-infiltrated myeloid cells in tumors isolated from each model using CyTOF mass cytometry. Analysis of monocytes, macrophages and dendritic cells in the tumors using unsupervised clustering identified 13 cell subsets. Significant differences were discovered between the two tumor types in frequencies of nine of 13 identified subsets. However, 3 of these subsets displayed relatively high expression of CXCR3 in EO771 tumors when compared to AT-3 tumors. The significant CXCR3 positive subsets were identified as CXCR3+ early macrophages, CXCR3+ Ly6C+CD11blo classical monocytes and Trem2+ CXCR3+ dendritic cells. We observed higher expression of CXCR4, Trem2, and CCR1 in CXCR3+ myeloid cells when compared to their CXCR3− counterparts in both tumor models. To elucidate the role of these CXCR3+ cells, we performed low-input bulk RNA sequencing of CD115+/CD11b+ cells from blood and spleen of EO771 tumor bearing mice with a wild type or CXCR3KO genotype. Ingenuity Pathway Analysis (IPA) of the differentially expressed genes suggests a significant inhibition of the IL-10 signaling pathway of CXCR3KO cells. These differentially expressed markers and genes suggest a functional importance in this subset in migration to the tumor and antigen presentation, both of which are anti-tumoral functions. Ongoing studies are aimed towards understanding the role of CXCR3+ myeloid cells in regulating response to immunotherapies. Citation Format: Gabriel A. Valentin-Guillama, Nandini Chatterjee, Ahmad Alimadadi, Catherine C. Hedrick. Mass cytometry analysis of CXCR3 expression in tumor-infiltrated myeloid cells as a predictive biomarker For immunotherapy response [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2689.

Abstract 5149: Extracellular matrix remodeling in metastatic mismatch repair deficient endometrial cancer:Implications for immune checkpoint inhibitor response prediction

Abstract Mismatch repair deficient (MMRd) status is a predictive biomarker for Immune checkpoint inhibitors (ICI) in endometrial cancer (EC), however, half of these patients (pts) do not respond. Most studies exploring biomarkers of response to ICI have focused on tumor or immune cell factors. We aimed to describe the extracellular matrix components of the tumor microenvironment (TME) in ICI-Responder (R) versus Non-Responder (NR) MMRd EC pts to identify new predictive biomarkers of response. Clinical data and outcomes of metastatic MMRd EC pts, treated with ICI (2016-2021), were retrospectively collected. Pts were classified as Rs (CR, PR, or SD ≥12 months) or NRs (PD or SD <12 months). Pre-ICI FFPE tumor samples were subjected to Biognosys UltraDeep TrueDiscovery™ Mass Spectrometry (MS) for identification of differentially regulated protein expression between R and NR. Criteria for protein candidate selection were p-value < 0.01 and average absolute log2 fold change (AVG Log ratio) > 0.58. Collagen was visualized on whole slides using Masson’s trichrome and quantified by Qupath. Collagen density was expressed as % positive surface in the intratumoral area. Second-harmonic generation (SHG) microscopy plus polarimetry (p-SHG) was used to estimate collagen content and describe the spatial orientation of fibrillar collagen in the intratumoral area. intraepithelial (ie) and stromal (s) CD3+ and CD8+ cells were quantified by Qupath. Non-parametric t-tests and Spearman’s rank coefficient (r) for linear correlation were used. Twenty-five tumor samples were included in the analysis: 17 from Rs and 8 from NRs. MS identified collagen type V alpha-2 (AVR Log ratio 2.9, p 0.008) and type XXII alpha-1 chains (AVR Log ratio 1.14, p 0.006) as significantly upregulated in NRs. Using spatial imaging, collagen density in the intratumoral area was significantly higher in NRs (median: 14.2% vs 4.6%, p 0.01). Intratumoral collagen density was positively correlated with collagen type V expression (r 0,54, p 0.016), and negatively correlated with ie and s CD3+ cell (r -0.39, p 0.054; r -0.4, p 0.048, respectively) and ie and s CD8+ cell (r -0.39, p 0.053; and r -0.29, p 0.166) infiltration. SHG confirmed that fibrillar collagen content in the tumor area was significantly greater in NRs (p 0.02). p-SHG revealed that NR tumors tended to show a highly disorganized collagen matrix. In contrast, a linear collagen pattern was associated with response to ICI: 89% of tumors displaying this pattern were Rs. Both MS and two different spatial approaches consistently demonstrated that increased collagen expression or density may serve as a potential predictive biomarker of resistance to ICI in MMRd metastatic EC, contributing to create an immunosuppressive TME. The spatial orientation pattern of collagen fibers may also be crucial in shaping immune TME and influencing the response to ICI. Citation Format: Juan Francisco Grau, Carole Aimé, Martin Mehnert, Elisa Yaniz-Galende, Catherine Genestie, Audrey Le Formal, Elodie Edmond, Antoine Amaury Lachaud, Patricia Pautier, Judith Michels, Sébastien Gouy, Amandine Maulard, Emeline Colomba-Blameble, Felix Blanc-Durand, Marie-Claire Schanne-Klein, Alexandra Leary. Extracellular matrix remodeling in metastatic mismatch repair deficient endometrial cancer:Implications for immune checkpoint inhibitor response prediction [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 5149.

Abstract 5943: Development of a SMART theranostic agent for precision pancreatic cancer therapeutics

Abstract Pancreatic cancer is predicted to emerge as the second leading cause of cancer-related deaths in the United States by 2030. Pancreatic ductal adenocarcinoma (PDAC) - the most common form of pancreatic cancer - is often diagnosed as metastatic disease and has demonstrated the least improvement in overall survival over the last three decades with incidence rising by 0.5% to 1.0% per year due in part to lack of effective and tumor-selective treatment strategies. One approach to reduce host toxicity is the development of novel molecules and innovative therapeutic approaches that are nontoxic to normal cells but can be activated only under tumor-specific conditions for lethality, which we coined as SMART (Selective Method of Activation for Response in Tumors) therapeutic agents. We utilized explicit properties and biomarkers present in PDAC tumors but not in normal tissues to develop a SMART DNA-damaging anti-cancer agent with both therapeutic and diagnostic (theranostic) activities. We evaluated the tumoricidal effect of our SMART agent using long-term survival assays in PDAC cell lines and 3D co-culture spheroids, as well as in tumor xenograft model studies in mice. We also tested the ability of our SMART agent to treat, detect/diagnose (via click chemistry), and identify critical resistance mechanisms (via proteomics) as a single agent or as a chemosensitizer with predictive tumor biomarkers for tumor-selective therapeutic response. We found that our SMART small molecule induces tumor-selective lethal effects as a monotherapy or in combination with DNA repair inhibitors or small molecules known to generate hydrogen peroxide (H2O2) in a tumor-selective manner for activation. Using “click” chemistry, we detected and quantified the incorporation of our SMART small molecule as a DNA-damaging agent via flow cytometry that correlated with therapeutic response, which is a feature that could facilitate the stratification, treatment monitoring, and management of patients with pancreatic cancers. We also found that our SMART agent is less toxic and more effective as a monotherapy than cisplatin in PDAC cell line-derived xenograft studies in mice. Overall, our novel SMART theranostic agent could provide a significant advancement as they are preferentially activated under tumor-specific conditions to induce tumor-selective therapeutic response. Furthermore, the diagnostic component our SMART agent could provide oncologists with essential information to make rapid clinical decisions on treatment strategies. Our SMART agent is also remarkably versatile as it can be utilized as a chemosensitizer or as a chemical probe to identify targetable vulnerabilities to potentiate its therapeutic effects at lower doses. Citation Format: Edward A. Motea, Jarrett J. Smith, Oluwasijibomi Ketiku. Development of a SMART theranostic agent for precision pancreatic cancer therapeutics [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 5943.

Abstract 4527: TNG348 is synergistic with PARP inhibitors in tumor models with elevated replication stress

Abstract Defective maintenance of genomic integrity is a hallmark of cancer cells that can result from oncogene-induced replication stress and by loss of DNA repair mechanisms. DNA repair deficiencies and elevated replication stress present targetable vulnerabilities for cancer treatment. Notably, BRCA1/2 mutant and homologous recombination deficient (HRD) tumors cannot repair double-strand breaks by homologous recombination and rely on alternative pathways of DNA repair. PARP inhibitors (PARPi), which are a standard of care in many BRCA1/2 mutant tumors, cause synthetic lethality with BRCA1/2 mutation by inhibiting the DNA base excision repair pathway. Despite the clinical benefit of PARPi, they are not effective in every HRD tumor and the acquisition of PARPi resistance limits long-term response. TNG348, a selective allosteric inhibitor of the deubiquitinating enzyme USP1, was specifically designed to target HRD vulnerabilities through an alternative mechanism. We previously showed that the anti-tumor activity of USP1 inhibition results from disruption of the translesion synthesis DNA damage tolerance pathway, a mechanism of action that is functionally distinct from base excision repair targeted by PARPi. Our preclinical studies show that TNG348 is active in HRD models and strongly synergizes with PARP inhibitors to drive strong anti-tumor responses. We have identified replication stress as a predictive biomarker of TNG348 response using cell line profiling and genome-wide CRISPR screens. For example, overexpression of oncogenes known to induce replication stress sensitized to the TNG348 and PARPi combination both in vitro and in vivo. These data indicate that cancer-specific elevated DNA replication stress could contribute to tumor sensitivity to TNG348 and provide additional patient stratification strategies and opportunities for indication expansion. Citation Format: Antoine Simoneau, Hsin-Jung Wu, Charlotte Pratt, Grace Comer, Shangtao Liu, Samuel Meier, Tenzing Khendu, Ashley Choi, Hongxiang Zhang, Binzhang Shen, Douglas Whittington, Sirimas Sudsakorn, Wenhai Zhang, Yi Yu, Yingnan Chen, Brian Haines, Adam Crystal, Jannik N. Andersen, John Maxwell, Scott Throner. TNG348 is synergistic with PARP inhibitors in tumor models with elevated replication stress [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 4527.

Sarcopenia is associated with leukopenia in urothelial carcinoma patients who receive tislelizumab combined with gemcitabine and cisplatin therapy

BackgroundIn the era of combination therapy, there has been limited research on body composition. Specific body composition, such as sarcopenia, possesses the potential to serve as a predictive biomarker for toxic effects and clinical response in patients with urothelial carcinoma (UC) undergoing tislelizumab combined with gemcitabine and cisplatin (T + GC).Materials and MethodsA total of 112 UC patients who received T + GC were selected at the Affiliated Hospital of Xuzhou Medical University from April 2020 to January 2023. Baseline patient characteristics and detailed hematological parameters were collected using the electronic medical system and laboratory examinations. The computed tomography images of patients were analyzed to calculate psoas muscle mass index (PMI). We evaluated the association between sarcopenia (PMI < 4.5 cm2/m2 in men; PMI < 3.3 cm2/m2 in women) and both hematological toxicity and tumor response.ResultsOverall, of the 112 patients (65.2% male, median age 56 years), 43 (38.4%) were defined as sarcopenia. Patients with sarcopenia were notably older (p = 0.037), more likely to have hypertension (p = 0.009), and had poorer ECOG-PS (p = 0.027). Patients with sarcopenia were more likely to develop leukopenia (OR 2.969, 95% CI 1.028–8.575, p = 0.044) after receiving at least two cycles of T + GC. However, these significant differences were not observed in thrombocytopenia and anemia. There were no significant differences in the tumor response and grade 3–4 hematological toxicity between patients with sarcopenia and those without sarcopenia.ConclusionsPatients with sarcopenia were more likely to develop leukopenia after receiving T + GC. There were no notable alterations observed in relation to anemia or thrombocytopenia. No significant difference was found between the sarcopenia group and non-sarcopenia group in terms of tumor response and grade 3–4 hematological toxicity.

Abstract 2799: Intestinal Akkermansia muciniphila complements the therapeutic efficacy of PD1 therapy by reshaping the immunosuppressive microenvironment in nonalcoholic fatty liver disease (NAFLD)-induced hepatocellular carcinoma

Abstract Objectives: Nonalcoholic fatty liver disease (NAFLD)-induced hepatocellular carcinoma (HCC) is an emerging malignancy in the developed world. However, immune checkpoint inhibitors (ICIs) are ineffective for patients with this disease. Advances in research have shown that gut-liver communication contributes to NAFLD progression. Therefore, identifying a key gut microbiota for predicting ICIs response and an effective combinatorial strategy is urgently needed. Design: 16S rRNA sequencing was employed to identify key gut microbiota crucial in developing the mouse NAFLD-induced HCC model. The in vivo roles of Akkermansia muciniphila (Akk) were evaluated by immunofluorescence staining, qPCR, mass spectrometry analysis, ELISA assay, and single-cell RNA sequencing (scRNA-seq) coupled with immune profiling analysis. The clinical and therapeutic value of Akk alone in combination with PD1 treatment was investigated. Results: Akk was decreased from healthy to HCC tissues during HCC tumor development, and daily administration of Akk not only ameliorates liver steatosis and cholesterol biosynthesis but could effectively attenuate the development of NAFLD-induced HCC. Akk repairs the intestinal lining, with a concurrent decrease in the serum concentration of lipopolysaccharide (LPS) and bile acid metabolites. Using scRNA-seq coupled with immune profiling analyses, we found that Akk suppressed the populations of immunosuppressive cells, including monocytic myeloid-derived suppressor cells (m-MDSCs) and M2 macrophages, by suppressing the differentiation of these cells from monocytes, which leads to T cell proliferation and activation. Consistent with this finding, Akk administration, combined with PD1 treatment, exerted a maximal growth-suppressive effect, accompanied by increased infiltration and activation of T cells. Clinically, the fecal level of Akk in HCC patients was decreased in NAFLD-induced HCC patients and serves as a potential biomarker for predicting PD1 response in HCC patients. Conclusions: Akk regulates the immune tumor environment by regulating two key immune suppressive cells via regulation of the LPS flux to develop NALFD-induced HCC. This finding provides a solid therapeutic basis for a rational combination with PD1 for the treatment of this deadly disease. Interestingly, Akk may serve as a predictive biomarker for PD1 response in HCC patients. Citation Format: Xueqian Wu, Fan Ying, Katherine Po Sin Chung, Carmen Oi Ning Leung, Terence Kin Wah Lee. Intestinal Akkermansia muciniphila complements the therapeutic efficacy of PD1 therapy by reshaping the immunosuppressive microenvironment in nonalcoholic fatty liver disease (NAFLD)-induced hepatocellular 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 2799.

Abstract 5165: Development and validation of an assay for the characterization of SSTR2 expression on CTCs from liquid biopsies

Abstract Background: Expression of Somatostatin Receptor subtype 2 (SSTR2) is observed in multiple cancers, including breast and colorectal cancer. Its role as a prognostic or predictive marker is cancer-type dependent. Recently, there has been renewed interest in assessing this biomarker since it plays a critical role as a target for radiotherapeutics (peptide receptor radionuclide therapy - PRRT) in patients with neuroendocrine tumors (NETs). Blood-based liquid biopsies are a non-invasive tool with proven effectiveness in characterizing cancer biomarker expression in patients. In this study, we describe the development and validation of a novel assay to analyze circulating tumor cells (CTCs) for expression of SSTR2 using the RareCyte platform. Experimental Procedures: Using AccuCyte®, an unbiased density-based method for collecting nucleated cells from whole blood, we transferred nucleated cells to slides and stained for SSTR2, CD45, and CK/EpCAM to identify CTCs (CK+/EpCAM+, CD45-), and evaluate SSTR2 expression. Slides were imaged with CyteFinder®, an automated multiparameter immunofluorescent (IF) microscopy system that applies machine learning algorithms for rare cell identification. Both clinical and spike-in samples were used for assay validation. Single-cell SSTR2 mean fluorescence intensities (MFI) were analyzed to determine protein expression levels. Colorectal cancer with neuroendocrine features and Merkel cell cancer (neuroendocrine tumor of the skin) patient samples were evaluated for CTC enumeration and SSTR2 expression. Results: The assay’s performance metrics for SSTR2 expression were 90+% accuracy, 85% sensitivity, 100% specificity with repeatability and intermediate precision CVs of 14.4% and 16.4%, respectively. A small cohort of colorectal cancer (1) and Merkel cell cancer (2) patient blood samples were tested using the assay. All 3 patients had high CTC levels (300+ per 7.5 mL). The percentage of SSTR2-positive CTCs ranged from 20-60%, reflective of tumor heterogeneity. Conclusions: SSTR2 is an important target now in the advent of theranostics - PRRT. Prognostic and therapeutic implications of SSTR2 expression is an area of active clinical research. SSTR2 targeted therapeutics like 177Lu-dotatate illustrate the possible utility of SSTR2 as a predictive biomarker of treatment response. To date, most of the work has been focused on assessing SSTR2 on tissue. Herein, we show how a CTC platform can be adapted to assess a clinically relevant cell surface protein biomarker. This is now included as an important correlative in a clinical trial for Merkel Cell Cancer patients called iPRRT that is actively accruing (NCT05583708). Citation Format: Erin G. Bayer, Rachel N. Ponting, Ciara M. Kelly, Areeb Lutfi, Maaz K. Afghan, Arturo B. Ramirez, Pashtoon M. Kasi. Development and validation of an assay for the characterization of SSTR2 expression on CTCs from liquid biopsies [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 5165.

Abstract 5028: Detection of non-small cell lung and bladder cancer signatures in peripheral blood using a novel antibody-free chromatin capture assay

Abstract Immune checkpoint inhibition (ICI) via anti-PD-1/PD-L1 has improved clinical outcomes for multiple cancers, including bladder and non-small cell lung cancer (NSCLC). Yet, the response rates remain low highlighting a significant knowledge gap in the mechanisms driving response and resistance. While the expression of PD-L1 has been established as a predictive biomarker of ICI response, other features, including molecular tumor burden, tumor microenvironment (TME), and the host’s immune mechanisms may contribute to overall response. The complex interactions between a tumor and the immune system underscore the need for a more comprehensive approach to explore novel biological and immunological insights of cancer progression. Circulating tumor DNA assays, while integral to contemporary decision-making in oncology, fall short in delineating the transcriptional programs that dictate cancer phenotypes and their evolution over the disease trajectory. To bridge this gap, we have developed a plasma-based active chromatin capture method that enables the comprehensive epigenetic and transcriptomic interrogation of both tumor and immune components. In this study, we quantified epigenetic profiles from plasma samples (n=100) of patients with bladder cancer (n=20) and NSCLC (n=24) undergoing ICI therapy, along with samples (n=70) from healthy individuals (n=5). As expected, the total cfDNA yield (ng) of lung (1.7x, p=2.3e-08) and bladder (1.98x, p=1.6e-11) cancer samples were higher than those of healthy individuals. More Interestingly, the longer cfDNA fragments derived from regulatory-active chromatin, also showed a significantly higher abundance in plasma samples from both lung (2.0x, p=2.5e-09) and bladder (3.2x, p=3.2e-14) cancer patients. Furthermore, the patterns of epigenetic regulations across cancer and healthy conditions showed distinct and stable clusters, consistent between gene bodies and transcription start sites, highlighting the epigenetic changes that may be driving cancer progression. We also observed a common enrichment of immune epigenetic signatures in both lung and bladder cancer samples, likely attributed to shared mechanisms associated with the modulation of the tumor microenvironment and immune interactions. Understanding these changes at an epigenetic level could provide insights into novel therapeutic approaches that target the tumor microenvironment and the immune interactions. Citation Format: Kevin Lai, Katharine Dilger, Rachael Cunningham, Timothy Barnes, Khiet Truong, Kathy Lam, Rhea Boquiren, Maggie C. Louie, Diana Abdueva. Detection of non-small cell lung and bladder cancer signatures in peripheral blood using a novel antibody-free chromatin capture assay [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 5028.

Abstract 5182: Exploiting tumor RNA-sequencing data for prediction of immune checkpoint inhibition response

Abstract Immune checkpoint inhibitors (ICI) have become the standard of care as a first- or second-line systemic treatment for patients with various cancer types. Unfortunately, many patients do not respond to this treatment and the occurrence of immune-related adverse events varies widely. There is an urgent need for robust predictive biomarkers for ICI response to identify patients with likely clinical benefit from this costly treatment. Several biomarkers, aimed at predicting response to ICI, have been proposed, including PD-L1 expression, tumor mutation burden (TMB), infiltration of cytotoxic T-cells in the tumor, Microsatellite Instability (MSI) and expression of various immune gene signatures. However, these individual biomarkers have suboptimal performance and multiple complementary omics technologies are required to properly quantify them. Integration of these multi-omic biomarkers has proven valuable in increasing the accuracy and robustness of ICI response prediction.We have developed a computational pipeline that determines the expressed mutation burden (eTMB), MSI status, fraction of infiltrating immune cells and various immune gene expression signatures directly from the RNA-sequencing profile of the tumor. Each biomarker is quantified using a dedicated algorithm that applies machine learning (eTMB and MSI) or computational deconvolution (infiltrating immune cells) and integrates external data sources to maximize performance. Algorithm performance has been validated on large cohorts of tumor RNA-sequencing data with matching gold standard quantification of said biomarkers. As a proof of concept, we applied our pipeline to tumor RNA-sequencing data of 45 gastric cancer patients treated with pembrolizumab. Responders showed significantly higher eTMB scores and were enriched among the MSI-H patients. Fractions of CD8 T-cells and M1 Macrophages, together with interferon gamma and cytotoxic T-cell gene expression signatures, were significantly increased in responders. Notably, integrating these biomarkers into a single prediction model improved prediction of response to checkpoint inhibition therapy compared to predictions based on eTMB or MSI alone. Taken together, our approach enables the quantification of various ICI response biomarkers from a single omics layer (RNA-sequencing) and can contribute to ICI response prediction. Citation Format: Pieter Mestdagh, Pieter-Jan Van Dam, Frederick De Baene, Carolina Fierro, Elise Van Hoof, Emmanuel Rivière, Hanne Dangreau, Reindert Van Cauwenberge, Jo Vandesompele. Exploiting tumor RNA-sequencing data for prediction of immune checkpoint inhibition response [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5182.

Abstract 68: Imaging Endothelial PD-L1 expression with ultrasound molecular imaging to stratify immune checkpoint inhibitor responders

Abstract Immune checkpoint blockade therapies (ICTs) have revolutionized oncology through substantial and lasting treatment outcomes, but only in 15-20% of patients. Non-invasive diagnostic tools to select patients that are most likely to respond to ICT and guide treatment regimens, and thus minimize toxicities to non-responder patients, are in high demand. Program death-1 ligand (PD-L1) marker evaluation in the tumor vasculature and tumor microenvironment is an essential criterion for identifying a suitable patient for PD-L1 based ICT, and it also serves as a prognostic marker in tumor response evaluation. Ultrasound molecular imaging (USMI) is an up-and-coming non-invasive imaging modality that enables longitudinal characterization of endothelial-specific markers. We are developing USMI methods to image immune markers expressed specifically on endothelial cells (EC) of tumors as a decision-support tool for guiding ICT strategies. In this work, we demonstrate for the first time the potential of imaging EC-specific immunosuppressive markers such as PD-L1 and investigate whether it can act as a marker for predicting ICT response. Target-ready microbubble (MB) was used (Trust Bio-sonics, Taiwan) to generate PD-L1 antibody (Ab) linked MBs (TMBs). DSPE-PEG was functionalized to anti-PD-L1 Ab or isotype-Ab (control) to generate stable TMBs (~5 × 107 MBs/mL) to target PD-L1 expressing ECs in vivo. We studied the treatment effect of anti-PD-L1-Ab and isotype-Ab using a syngeneic CT-26 tumor in mice and demonstrated the targeting potency of TMB or iso-Ab-MB (NMB). US imaging was performed using Vevo 2100 high-frequency US system with M250 transducer to evaluate and quantify the differentially targeted enhancement parameter (dTE). We observed that TMBs bound PD-L1 expressing ECs in vivo, with high and low enhancement for TMB and NMB, respectively, in tumors validated by dTE quantification through characterization of the slower washout kinetics. Furthermore, TMBs exhibited increased tumor retention and prolonged binding with average differential targeted enhancement values (dTE) (+/- SD) of 5.2 ± 7.4 for TMBs compared to 2.4 ± 3.9 for NMBs (p=0.14). Preliminary results of USMI (dTE values of TMBs) indicate that PD-L1-treated mice showed a higher response (4.5 ± 5.2; n=9) to the treatment group than control mice that did not receive Ab (1.4 ± 0.8; n=4). The tumor growth rate for the treated mice group was 5.4 -fold retarded compared to the control group which correlated with USMI dTE values. The USMI signals correlated with IHC performed on tumor tissue ex vivo from the respective mouse used for US imaging. Overall, our pre-clinical study suggests that USMI using TMBs may contribute to non-invasive assessment of IC expression on tumor ECs, and that it may serve as a predictive biomarker of ICT response. This novel TMB agent could be a game changer for CE USMI-based cancer prognosis in ICB therapy. Citation Format: Arutselvan Natarajan, Farbod Tabesh, Mahsa Bataghva, Ramasamy Paulmurugan, Ahmed Nagy El Kaffas. Imaging Endothelial PD-L1 expression with ultrasound molecular imaging to stratify immune checkpoint inhibitor responders [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 68.

Abstract 3799: Spatial biomarker prediction in papillary renal cell carcinoma using artificial intelligence-based immunofluorescent inference

Abstract Papillary renal cell carcinoma (pRCC) accounts for ~15-20% of the global burden for kidney cancers but still lacks effective mechanisms to predict patient responses to therapeutic intervention. Existing therapies targeting kidney cancers (anti-angiogenesis and immunotherapy) intimately involve the tumor microenvironment in their mechanism of action. However, the role of the physical organization of the tumor in determining therapeutic efficacy remains unknown. Previous efforts have utilized transcriptional profiling to stratify patients into treatment categories but here we propose to leverage artificial neural networks (“AI”) to infer the expression of marker genes for particular cell types of interest based on ubiquitously expressed immunofluorescence channels, unstained brightfield images, and hematoxylin and eosin (H&amp;E) stained images. We have developed AI models that are capable of the prediction of a range of relevant cell types based on the expression profile of DAPI, type IV collagen, and integrin beta-1, including endothelial cells, T cells, macrophages and other cell types. This work applies an AI model class known as generative adversarial networks (GANs) to produce photo-realistic simulated immunofluorescence images based on the input channels. This approach may allow us to accurately predict complex multiplex immunofluorescence (mIF) images from simpler, more routine histological methods, greatly reducing the cost burden required to profile histopathology images for cell types of interest. Our previous work identified a number of interesting cell-cell association clusters correlated with patient prognosis and this extension of that previous work aims to allow for the prediction of these interactions while obviating the need to perform the original, costly multiplex immunofluorescence assay or even a more reduced immunofluorescence staining panel which may still be out of reach for many clinical pathology labs and the interpretation of which may be challenging. By simplifying and increasing the availability of predictive AI models for digital pathology, we hope to further develop and apply predictive biomarkers for patient prognosis and treatment response in papillary renal cell carcinoma. Citation Format: Jared Brewer, Joseph Vento, Scott Haake, Anupama Reddy. Spatial biomarker prediction in papillary renal cell carcinoma using artificial intelligence-based immunofluorescent inference [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 3799.

Abstract 619: Comprehensive cell line profiling of monomer- and dimer-selective small molecule RAF inhibitors uncovers determinants of cellular responses

Abstract Dysregulation of the Mitogen-Activated Protein Kinase (MAPK) pathway is a frequent event in many cancers. The RAF family of kinases, which includes ARAF, BRAF, and CRAF (RAF1), plays a key role in MAPK signaling. Of the three isoforms, only BRAF is currently targeted by approved RAF inhibitors. These inhibitors were designed to target BRAF-mutant variants that act as oncogenic monomers. However, BRAF-monomer inhibitors can induce activation of wild-type RAF by the formation of dimers, resulting in hyperproliferative lesions of BRAF wild-type cells. Additionally, BRAF-monomer inhibitors do not target RAF-dimers, although this is a potential vulnerability of many MAPK-driven cancers. Novel inhibitors are in development to circumvent these limitations. Further insights into the precise targeting and mechanism of action of RAF inhibitors can help in determining the most promising avenues for drug development. We performed extensive cellular profiling experiments with three approved and five clinical stage inhibitors of RAF to gain more insight into the differences and cellular targeting of these agents. The inhibitors were tested in cell viability assays on a panel of 130 human cancer cell lines from diverse tissue origins. Cell lines were exposed to dose ranges of each inhibitor spanning four logarithmic orders, and viability was determined by measurement of intracellular ATP. Effects of compound on cell viability were related to genomic, transcriptomic, and proteomic status using bioinformatics, with the aim to identify predictive biomarkers for drug response, and to determine selective targeting of MAPK-driven cell lines. Additionally, the primary mechanism of cellular inhibitor activity was determined by relating cell line responses to gene dependency data from large CRISPR knockout screens. Our results showed striking similarity in the overall inhibitory profile of BRAF-monomer inhibitors and the vemurafenib-derivative and paradox breaker plixorafenib, which all selectively targeted BRAF-mutant cell lines. However, whereas vemurafenib increased cell viability in RAS-mutant cell lines at high drug concentrations, this effect was greatly reduced for plixorafenib. Western blot analysis of phosphorylated MEK and ERK levels in RAS-mutant cell lines treated with BRAF-monomer inhibitors confirmed MAPK pathway activation as cause of the increased cell viability. In addition, RAF-dimer inhibitors potently targeted BRAF-mutant cell lines, as well as several KRAS- and NRAS-mutant cell lines. Lastly, correlation analyses of cell line profiling data with CRISPR knockout data revealed the preferential targeting of RAF1-dependent cell lines by the dual RAF/MEK inhibitor avutometinib. The results of our analyses shed light on the intricacies of cellular targeting by RAF inhibitors and provide insights to guide the development of new RAF inhibitors. Citation Format: Jeffrey J. Kooijman, Awan Al Koerdi, Sjoerd B. van der Leeuw, Daphne J. Kluitmans, Esmee van den Bossche, Jelle Dylus, Jeroen A. de Roos, Janneke J. Melis, Nicole Willemsen-Seegers, Guido J. Zaman. Comprehensive cell line profiling of monomer- and dimer-selective small molecule RAF inhibitors uncovers determinants of cellular responses [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 619.

Abstract 3514: A deep learning model for detection and characterization of tertiary lymphoid structures in H&amp;E-stained images from pancreatic ductal adenocarcinoma

Abstract Tertiary lymphoid structures (TLSs) are ectopic aggregates of a number of immune cells in nonlymphoid tissues under chronically inflamed environments and cancer. Emergence evidence suggested that TLSs were found in diverse cancers and TLSs have been referred as an independently predictive biomarker for immunotherapy response, especially immune checkpoint inhibitors. Moreover, the density and mature status of TLSs often positively associated with the favorable outcomes in cancers. Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related mortality with low survival rate in the whole world. Here, we developed a multi-resolution machine learning model for detection of TLSs from hema-toxylin and eosin (H&amp;E) stained digital pathology slides as a potential clinical biomarker in PDAC. We selected 120 H&amp;E slides (patients) with a total of 1,896 TLSs manually annotated as immature or mature TLSs by two pathologists. We also used immunohistochemistry (IHC) for CD20 and CD23 markers as a reference standard for H&amp;E-based manual annotations. Our model mainly included detection and segmentation of individual TLSs. First, we used a transformer convolutional neural network to build a classification task to detect the presence of immature or mature TLSs. Then the instance segmentation model was applied to quantify the number of lymphocytes per unit square to validate the presence and maturation of TLSs. F1-score, calculated by the harmonic mean of precision (positive predictive value) and recall (sensitivity), was finally used to quantitatively evaluate the performance of model for each H&amp;E slide. In validation dataset with 250 H&amp;E slides (patients), we selected 50 slides to calculate F1-score, suggesting that our machine learning model achieved high F1-scores for detection and classification of immature TLSs (mean = 0.87, range from 0.79 to 0.95), and mature TLSs (mean = 0.92, range from 0.86 to 0.97). We further stratified 370 PDAC patients into three groups with mature TLSs (23%), immature TLSs (16%), and no TLSs (61%), respectively. Combined the clinical information, survival analysis revealed that patients with mature TLS had better survival outcomes than those without TLS (HR = 0.63; 95% CI, 0.44 - 0.90; P = 0.010). We developed a machine learning model that can accurately detect and classify TLSs in PDAC, and also demonstrated the prognosis value of predictive TLSs in H&amp;E slides by automatedly machine learning model. These data highlight the promise of machine learning model for automated identification and quantification of the TLS on H&amp;E slides in PDAC pathology. Citation Format: Chaoxian Zhao, Jidong Jia, Mingxi Lv, Jiayi Feng, Yijun Wang, Yingbin Liu. A deep learning model for detection and characterization of tertiary lymphoid structures in H&amp;E-stained images from pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3514.

Abstract 2521: New predictive markers for dasatinib responsiveness in triple-negative breast cancer: A biomarker discovery study

Abstract Triple negative breast cancer is the most heterogeneous and aggressive type of breast tumor type and is characterized by the lack of expression of clinical biomarkers (ER, PR, and HER2) and targeted therapies. In this regard, different clinical trials have failed, unable to stratify these patients to find an effective response to specific therapy. This study aimed to identify biomarkers exclusively expressed in the basal mammary compartment that can help us to subclassify the triple-negative breast cancer subtype. Based on computational analysis from single-cell RNA sequencing data, we have identified a list of basal identity-associated genes (BC-markers). Subsequent histological validation confirmed the expression pattern of these markers in a cohort of 50 samples, which has allowed us to stratify triple-negative breast cancer patients into BC-TNBC and luminal-like subtypes. Importantly, the expression of these markers correlated with poorer prognosis in TNBC patients. Functional analysis revealed that TAGLN played a significant role in cell proliferation and migration, potentially impacting tumor aggressiveness. Moreover, the expression of these BC-markers is associated with a mesenchymal phenotype which corresponds to the previously described Lehmannn’s mesenchymal-like signature. In this study, TAGLN appeared to influence dasatinib sensitivity, suggesting its potential as a predictive biomarker for dasatinib response in TNBC patients. Citation Format: Daniel Ortega-Álvarez, David Tebar, Marta Casado, Elena Castillo, Cristina Guardia, David Olivares-Osuna, Eva Musulén, Elisabetta Mereu, Ginés Luengo, Eva M. Galán-Moya, Veronica Rodilla. New predictive markers for dasatinib responsiveness in triple-negative breast cancer: A biomarker discovery study [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 2521.

Abstract 5162: Tumor transcriptional identity as a key predictor of clinical outcome in EGFR-Mutant non-small cell lung cancer

Abstract Background: EGFR tyrosine-kinase inhibitors (TKIs) produce significant clinical benefit in patients (pts) harboring EGFR-activating mutations, however, patient outcomes remain heterogeneous, highlighting the need to identify novel biomarkers for treatment response prediction. We aim to investigate the association of tumor transcriptional identity with clinical outcomes in EGFR-mutant NSCLC. Methods: We used consensus classification to derive unsupervised transcriptionally distinct clusters in EGFR-mutant NSCLC from the TCGA (N= 44). Then, we built a predictive model using extreme-gradient boosting to (1) classify an additional 88 surgically resected tumors from the LCCS study and analyzed for overall survival (OS) prediction; (2) classify metastatic EGFR-mutant NSCLC from the RELAY trial with available RNAseq data (N = 106) and analyzed for progression-free survival (PFS) with erlotinib treatment. Results: Consensus clustering revealed two distinct clusters, one associated with more lineage-preserving features and one with more lineage-independent features that could transcriptionally distinguish EGFR-mutant lung cancer. No differences in primary EGFR mutations were detected across the two clusters. Ingenuity Pathway Analysis revealed TGFB1 as the most significantly enriched pathway in the lineage-independent cluster which was also significantly more mesenchymal (p = 0.0028). For surgically resectioned EGFR-mutant NSCLC, OS was consistently shorter in the lineage-independent cluster for TCGA (23.6 vs 58.3 mo; p = 0.049; HR = 0.4) and the LCCS cohort (55.7 vs not-reached; p = 0.0073; HR = 0.27). In the metastatic setting with erlotinib treatment, there was a trend for PFS to be shorter in the lineage-independent cluster (15.2 vs 19.4 mo; p = 0.12; HR = 0.61). Conclusions: Transcriptional identity in EGFR-mutant lung cancer are prognostic of clinical outcome. The lineage-independent tumors have inferior benefit to EGFR TKI as well as inferior OS, indicating that transcriptional biomarkers can be independent of oncogene biomarkers. These transcriptional biomarkers may help identify pts with poor outcome for future intensification of therapies. Citation Format: Simon Heeke, Monique Nilsson, Don Gibbons, Jianjun Zhang, Ignatio I. Wistuba, Keunchil Park, John V. Heymach, Xiuning Le. Tumor transcriptional identity as a key predictor of clinical outcome in EGFR-Mutant non-small cell lung 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 5162.

Abstract 81: High-resolution monitoring of B cell isotype switching for biomarker analysis of adenosine pathway inhibition immunotherapy

Abstract Background B cell isotype switching gives rise to IgG-, IgA-, and IgE-expressing B cells and is an essential step in generating effective humoral responses to antigen challenge. Within the tumor microenvironment, extracellular adenosine may contribute to cancer immune evasion by inhibiting isotype switching through interaction with A2aR, a possibility that has driven investigation of adenosine pathway inhibition as an immunotherapeutic modality. In this context, methods to quantify isotype switching may provide insight into the pharmacodynamics of candidate agent activity and serve as a source of predictive biomarkers of response. We recently clinically validated the Oncomine IGH-LR assay for the measurement of CLL somatic hypermutation (SHM). Given that this assay also detects isotype switching events, we retrospectively analyzed our validation dataset to determine the suitability of the assay as a potential tool for biomarker analysis of adenosine pathway inhibition immunotherapy. Methods The Oncomine IGH-LR assay determines B cell SHM and isotype through targeted next-generation sequencing of IGH chains from RNA. The assay software organizes detected IGH chains into B cell clonal lineages, of which a subset may comprise multi-isotype lineages indicative of isotype switching events. The software further reports secondary repertoire features such as clonality/evenness. Total RNA was extracted from peripheral blood mononuclear cells or bone marrow aspirates from patients identified as having &amp;gt;10% monoclonal B-cell population by flow cytometry. Intra-assay precision was evaluated by processing RNA from 7 samples in triplicate and sequencing in the same run. Inter-assay precision was evaluated by having different operators process 10 samples on different days with different sequencing barcodes. Dominant clone isotype and clonal lineage assignment, total isotype abundance, and repertoire clonality were evaluated across replicates. Results Overall, 100% concordance was achieved for dominant clone isotype and clonal lineage calls for both intra- and inter-assay studies. Intra-assay and inter-assay median coefficient of variation (CV) for reported isotype frequencies ranged from 0.10-0.16 and 0.15-0.28, respectively. Intra-assay and inter-assay median CV for clonality (1 - Normalized Shannon Entropy) was 0.06 and 0.04, respectively. Conclusions We observe high reproducibility and repeatability of the IGH-LR assay for B cell isotyping, clonal lineage analysis, and clonality assessment. Multi-isotype clonal lineages are automatically reported by the software and provide a means to directly monitor isotype switching events in the context of antigen challenge or immunomodulation. Taken together, we anticipate this assay to be a useful tool for biomarker analysis of adenosine pathway inhibition immunotherapy. Citation Format: Ramadevi Prathapam, Kristen Champion, Anne Burkhardt, Chelsey L. Smith, Rajesh Singh, Timothy J. Looney. High-resolution monitoring of B cell isotype switching for biomarker analysis of adenosine pathway inhibition immunotherapy [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 81.

Abstract 5186: Spatial single-cell deconvolution of the bone marrow microenvironment to define predictive biomarker for response to immunotherapy in patients treated with a check point inhibitor in smoldering multiple myeloma, phase II trial of nivolumab in combination with lenalidomide and low dose dexamethasone

Abstract Background:The immune system is critical for response and resistance in all cancer types and therapies. Patients with Smoldering Multiple Myeloma (SMM) have altered immune cell composition in the bone marrow (BM). However, little is known about the role of BM spatial immune profiling in patients treated with checkpoint inhibitors in MM. Methods: We performed spatial profiling with imaging mass cytometry of trephine biopsies of the BM (n=15; pre-treatment n=7 and post-treatment n=8) from patients with high-risk SMM enrolled in a phase II trial of nivolumab in combination with lenalidomide and low dose dexamethasone (NCT02903381). Patients received 6 cycles of induction therapy of nivolumab (240mg) at days 1 and 15, in combination with lenalidomide (25mg) at days 1-21 and dexamethasone at days 1, 8, and 15. The induction phase was followed by nivolumab (240mg) and lenalidomide (25mg) maintenance for another 6 cycles. A treatment cycle was defined as 28 consecutive days for a total of 12 months period. Event monitoring was performed up to 3 years. The primary efficacy outcome was progression-free survival 2 years after treatment. Results: Eight patients were enrolled on this study from January 2017 to June 2017. The study was put on hold due to the toxicity of checkpoint inhibitors observed in other trials of MM. The median age for all patients at enrolment was 60 years (range 49 to 81), with 5 males (62%) and 3 females (38%). All patients met criteria for high risk SMM disease. Median follow-up for all 8 patients was 38.7 months. Median overall survival was not reached; median progression-free survival was 33.8 months. Seven total patients had a response of Minimal Response (MR) or better, and 4 patients who initially responded eventually progressed. The total number of patients who were followed for at least two years and remained progression-free was 4 of 8 evaluable patients (50%; 90% CI: 19 - 81%).Spatial profiling with a panel of 37 markers on the pre-treatment samples allowed us to assess the states and abundances of multiple immune subpopulations, including rare regulatory NK cells. To assess the accuracy of our cell type annotation, we correlated the proportion of malignant plasma cells identified per patient with the value determined by pathologists for clinical diagnosis. These measures were significantly positively correlated (R=0.92, p&amp;lt;0.01). Spatial neighborhood analysis revealed statistically significant interactions enriched in the responders, even with our small sample size (t-test; p &amp;lt; 0.05). Conclusions and future directions: This pilot study suggests that BM spatial immune profiling before receiving immune check point inhibitors can be a biomarker for response in patients with SMM. Citation Format: Yoshinobu Konishi, Giuseppe Tarantino, Yiwen He, Romanos Sklavenitis-Pistofidis, Sujal I. Shah, Katherine Towle, Christian J. Cea-Curry, Oksana Zavidij, Ruben D. Carrasco, Scott J. Rodig, Jon C. Aster, David Liu, Gad Getz, Irene M. Ghobrial. Spatial single-cell deconvolution of the bone marrow microenvironment to define predictive biomarker for response to immunotherapy in patients treated with a check point inhibitor in smoldering multiple myeloma, phase II trial of nivolumab in combination with lenalidomide and low dose dexamethasone [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 5186.