The cellular response to hypoxia is driven by hypoxia-inducible factors (HIFs), which regulate genes involved in glycolysis, angiogenesis, and cell proliferation, as well as inflammation and tumor progression. HIF activation is well-characterized and is primarily regulated by oxygen-dependent prolyl hydroxylation and subsequent degradation. SET1B, a histone H3 lysine 4 (H3K4) methyltransferase, has recently emerged as a key modulator of HIF target gene transcription, but evidence suggests that it plays a broader role in modulating HIF transcriptional activity beyond histone methylation. Here, we revealed that SET1B interacts with RNA polymerase II to coordinate sustained HIF-mediated transcriptional activity through multiple functional domains. In clear cell renal cell carcinoma (ccRCC), SET1B was critical for sustained HIF activity, and SET1B expression correlated with disease progression and metastasis in patient samples. Moreover, SET1B depletion enhanced the efficacy of HIF-2 inhibitors. These findings establish SET1B as a driver of tumor progression and potential therapeutic target in ccRCC.

Maintaining sustained deoxyribonucleotide triphosphate (dNTP) pools is essential for DNA replication fidelity and genome stability. In EGFR-mutant non-small cell lung carcinoma (NSCLC), we found that disruption of dNTP homeostasis plays a critical role in determining sensitivity to the EGFR inhibitor osimertinib and in shaping mechanisms of acquired resistance. Transcriptomic and biochemical analyses revealed that osimertinib suppresses RRM2 expression, a key regulator of dNTP synthesis, through downregulation of the transcription factor MYBL2. In response to osimertinib-mediated replication stress and dNTP depletion, cells activated a compensatory pathway involving the stress-inducible ribonucleotide reductase subunit RRM2B via a transcriptional regulator, TNNT3. CHK2 signaling was essential for TNNT3 nuclear translocation and RRM2B transcriptional activation. Inhibition of CHK2 or combined CHK1/2 blockade impaired RRM2B induction, exacerbated replication stress, and delayed the development of osimertinib resistance both in cell lines and in xenograft models. Collectively, these findings reveal that EGFR-mutant NSCLC cells rely on dynamic signaling through EGFR-MYBL2-RRM2 and CHK2-TNNT3-RRM2B regulatory pathways to maintain dNTP pool balance under therapeutic pressure. Disruption of this signaling network sensitizes tumors to osimertinib and impairs the acquisition of resistance, linking metabolic regulation to therapeutic resistance and disease progression.

Abstract Introduction: Replication repair deficiency (RRD) is a pan-cancer mechanism caused by germline and/or somatically acquired mutations in the replication repair machinery - DNA polymerase proofreading and the mismatch repair (MMR) system. Germline monoallelic (Lynch Syndrome, LS) or biallelic (Constitutional Mismatch Repair Deficiency, CMMRD) mutations in MMR genes are present in 5-10% of glioblastomas in children, adolescents, and young adults. RRD gliomas are lethal, chemoradiation-resistant cancers, characterized by universal hypermutation and variable susceptibility to immune-checkpoint inhibition (ICI). These tumors exhibit variability in patient age of onset, type, location, and response to ICI. Methods: To understand the clinical and biological differences associated with RRD central nervous system (CNS) tumors, we used germline mutations and brain development-specific Cre-drivers to generate murine models that recapitulate the phenotypic and genomic characteristics of each human RRD subgroup: 1) MMRD+PPD (Nestin- and Olig2-Cre+/ Msh2LoxP/LoxP/PoleS459F/+ and LSL-PoleP286R/+): MMR-deficiency (MMRD) in combination with polymerase proofreading deficiency (PPD). 2) MMRD-only (Nestin-Cre+/Trp53LoxP/LoxP and Msh2LoxP/LoxP or Mlh1-/-): MMRD lacking PPD associated with TP53 mutations. Results: Using trans-species comparative approach, we elucidated a mechanistic model of RRD-driven brain tumorigenesis. We revealed that the cell-of-origin significantly contributes to determining brain tumor type, location, and age of tumor onset, suggesting a strong impact of early- or late-RRD mutational onset in shaping tumor biology (p<0.0001). Importantly, using murine neural stem cells, we discovered that germline mutagenesis onset directly influences timeline of brain tumor formation and survival between CMMRD and LS patients (p<0.05). We further demonstrate the interplay between POLE mutations and MMRD status in modulating the likelihood of brain tumorigenesis in both species. To understand the interaction between hypermutation and the immune system, we characterized the tumor immune microenvironment in spontaneously forming tumors. We uncovered subgroup-specific immune landscapes, with CD8+ T cell activity emerging as a key modulator in controlling brain tumor growth (p<0.0001), suggesting an underlying mechanism that may inform therapeutic strategies in RRD patients. Significance: Altogether, our models accurately mimic the human condition, providing a mechanistic framework of RRD-driven brain tumorigenesis, optimization of subgroup-tailored immunotherapy approaches, and putative surveillance protocols. Citation Format: Zoya Aamir, Melissa A. Galati, Emma Gattoni, Owen Crump, Nemanja Ilic, Anirban Das, Nicholas R. Fernandez, Angel K. Wong, Lucie Stengs, Jose R. Dimayacyac, Yuan Chang, Vanessa Bianchi, Melissa Edwards, David Malkin, Cynthia Hawkins, Nuno M. Nunes, Uri Tabori. Trans-species analysis of central nervous system developmental-specific replication repair deficiency reveals differential patterns of gliomagenesis and response to immunotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 631.

Abstract Background: Spatial imaging outputs continue to grow in scale and complexity. While brightfield IHC and H&E remain the qualitative gold standard for antibody-based assessment, mIF offers quantitative protein measurement on a single slide. However, challenges such as non-specific binding, imaging artifacts, and variability across sites and operators limit confidence in mIF reproducibility. A quantitative, robust method is needed to assess concordance between IHC and mIF stains. Methods: Using a pan-cancer dataset with a 4-plex mIF panel and matched IHC sections from consecutive slides, we first co-registered images into a shared coordinate space with Valis, applying global rigid and non-rigid transformations from feature matches. IHC stain channels were isolated via stain-matrix-based deconvolution. A tissue mask was generated on the mIF image using Otsu thresholding and morphology operations and then projected onto the IHC slide.Tissue was divided into tiles whose size accounted for section-to-section distance, registration error, and biological variability. Within each tile, random windows were sampled to perform two tests: (1) identify whether the tile contains high stain intensity and (2) determine whether the corresponding IHC and mIF tiles exhibit statistically concordant staining. This approach yields both a DICE score for high-stain region overlap and a stain concordance metric capturing agreement across high- and low-stain regions. Tile-level results and heatmaps are visualized in SpaceIQ™. Results: Concordance between mIF and IHC varied substantially across markers, with CD8 showing the highest and FoxP3 the lowest agreement, a trend consistent across samples. Concordance heatmaps also revealed strong spatial effects, with some tissue regions highly concordant and others clearly discordant. Expert visual review matched these quantitative findings. Conclusions: This segmentation-free framework identifies substantial marker- and region-specific variation in concordance between mIF and IHC staining. Because the method is marker-agnostic and compensates for registration error and inter-section biological differences, it provides a generalized, quantitative approach for evaluating agreement between paired mIF and IHC slides across platforms. Citation Format: Brian Falkenstein, Raymond Yan, A. Burak Tosun, S. Chakra Chennubhotla, Filippo Pullara. Robust segmentation-free stain quality concordance metrics in the SpaceIQ™ multi-omic analysis platform [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 715.

Abstract Introduction: Chronic lymphocytic leukaemia (CLL) cells abnormally express lipoprotein lipase (LPL), an enzyme typically restricted to adipocytes and myocytes for lipid-mediated energy utilisation. This enables CLL cells to store and utilise lipids, potentially competing with or diverting resources from healthy tissues. In vitro studies suggest that reducing fatty acid availability may limit CLL proliferation; however, little is known about how patients can modulate this process in vivo. Exercise training offers a systemic, non-pharmacological approach to counter metabolic dysregulation, with potential benefits for tumour control and overall health. Methods: We conducted a 12-week exercise trial involving five treatment-naive (TN-CLL) and five previously treated (Td-CLL) patients. We assessed the metabolic fate of ingested lipids before (Baseline) and after (Post-Intervention) the program. Patients consumed a meal containing 200mg palmitic acid tracer (13CPA), and blood samples were collected hourly for 3 hours (T0h-T3h). We assessed 13CPA enrichment in plasma triacylglycerol (TAG) and non-esterified fatty acids (NEFA), and total fatty acids in immune cells (PBMC) using mass spectrometry, and data were analysed using RM-ANOVA. Results: Post-meal ingestion, 13CPA enrichment in plasma TAG and NEFA increased steadily from T1h-T3h (p<0.001). At Baseline T3h, TN-CLL exhibited higher plasma 13CPA-TAG and unlabelled PA-TAG incorporation than Td-CLL (p<0.001). Post-Intervention T3h, TN-CLL 13CPA-TAG levels decreased (p<0.05) and were no longer significantly different than Td-CLL. TN-CLL 13CPA-NEFA enrichment increased post-Intervention compared to Td-CLL (p<0.05), suggesting enhanced 13CPA-TAG hydrolysis. Similarly, 13CPA uptake into PBMCs, which was higher in TN-CLL at Baseline T3h (p<0.05), reduced Post-intervention. Conclusion: This pilot study demonstrates the feasibility of stable isotope tracing to assess in vivo lipid uptake in CLL. Exercise training in TN-CLL patients reduced lipid uptake, suggesting a shift towards a more balanced and healthier metabolic profile. Further research is needed to determine whether exercise can disrupt the lipid dependence of CLL cells. Citation Format: Uzma Zaheer, Ellie Miles, Angela Avramovska, Vithushan Srikumaran, Andrew Hulton, Long Li, Caitlin Jeary, Andrea Sitlinger, Renata Walewska, Barbara Fielding, David Bartlett. Rebalancing systemic and cellular energy dysmetabolism in Chronic Lymphocytic Leukemia through exercise training [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3269.

Abstract Background: ABI1 (Abelson interactor-1) is classically recognized as a multifunctional adaptor protein with homeostatic roles in cancer biology. It functions as a tumor suppressor in some cancer such as prostate cancer, yet exhibits oncogenic activity in other cancers such as for example breast cancer. Historically, ABI1 has been studied for its actin-cytoskeleton-associated functions—including cell-cell adhesion, cell motility, and lamellipodia formation—as well as its role in regulating major signaling hubs such as c-Abl, PI3K, and Src. Our recent findings reveal an unanticipated function of ABI1: direct DNA binding mediated by a conserved homeodomain homology region (HHR). This discovery led us to hypothesize that ABI1 may act as a previously unrecognized transcriptional regulator. Here, we sought to define the molecular mechanisms through which ABI1 contributes to transcriptional control. Methods: To determine sequence specificity and genomic occupancy, we performed ChIP using HHR-intact and HHR-mutant ABI1 constructs, complemented by in vitro DNA binding assays using purified proteins. Subcellular fractionation and chromatin enrichment assays assessed ABI1 nuclear localization and association with chromatin. ABI1-interacting transcriptional machinery was identified through co-immunoprecipitation (co-IP). RNA-seq comparing cells expressing wild-type ABI1 versus an HHR-defective DNA-binding mutant defined ABI1-dependent transcriptional outputs. Results: ABI1 binds DNA both in vitro and in vivo and displays reproducible sequence motifs from integrated ChIP and in vitro binding analyses. ABI1 variants containing an intact HHR domain localize preferentially to the nucleus and chromatin fractions. Co-IP studies identify ABI1 as a component of a defined transcriptional complex. RNA-seq analyses reveal that HHR-mediated DNA binding is required for a discrete ABI1-dependent transcriptional program. Conclusions: We identify ABI1 as a novel DNA-binding protein with sequence preference and transcriptional regulatory capacity mediated through its HHR domain. These findings expand the functional repertoire of ABI1 beyond actin regulation and kinase signaling, providing the first mechanistic framework for ABI1-driven transcriptional control. Citation Format: Kate Livingston, XIANG Li, Kevin M. Lin, Leszek Kotula. Mechanistic dissection of ABI1 as DNA-binding transcriptional regulator in cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7245.

Abstract Cancer genomes accumulate mutations through diverse biological processes, including defective DNA repair, exposure to carcinogens, and normal aging. These processes leave characteristic patterns known as mutational signatures. Identifying these signatures in patient tumors enables researchers to understand disease mechanisms and infer past exposures.The current standard resource, COSMIC v3.5, catalogs over 90 single base substitution (SBS) signatures using the SBS-96 mutational context, which characterizes each mutation by the substituted base and one flanking nucleotide on each side. This reference is widely used across cancer genomics studies to decode the mutational processes active in tumors. However, challenges have emerged where signatures overlap and can be misassigned, leading to incorrect interpretations of the mutational processes driving individual cancers. While we showed that higher-resolution methods using extended sequence contexts can differentiate these overlapping signatures and reveal novel processes, these findings were limited to specific cancer types such as colorectal, esophageal, renal, and head and neck. Additionally, no comprehensive high-resolution reference currently exists.To address this, we developed a high-definition (HD) mutational signature SBS reference set using over 40,000 whole-genome sequences from 14 cohorts spanning diverse cancer types and non-cancer tissues, including both primary and metastatic tumors. We analyzed mutations at SBS-4608 resolution, combining extended pentanucleotide sequence contexts with strand orientation information. This represents the practical limit of current sequencing technology. Our analysis clarifies several ambiguous signatures from the existing reference, improving accuracy when identifying mutational processes in individual tumors. We also discovered novel signatures that were previously undetectable at standard resolution, revealing new insights into the mutational landscape of human cancers.This HD reference set addresses critical limitations in current mutational signature analysis, enabling more accurate interpretation of mutational processes in cancer genomes and improving the reliability of downstream biological and clinical insights. By providing a comprehensive resource that resolves signature overlaps and reveals previously hidden mutational processes, this work will enhance precision in understanding cancer biology and tumor mutagenesis. The reference will be made publicly available to the research community upon publication. Citation Format: Jessica N. Au, Marcos Diaz-Gay, Raviteja Vangara, Pilar Gallego-Garcia, Mousumy Kundu, S.M Ashiqul Islam, Maria Zhivagui, Zichen Jiang, Christopher Steele, Sarah Moody, Michael R. Stratton, Paul J. Brennan, Ludmil B. Alexandrov. High-definition signatures of single-base substitutions in human cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 46.

Abstract Background: Although three generations of EGFR tyrosine kinase inhibitors (TKIs) have achieved remarkable clinical success, resistance driven by secondary EGFR mutations remains a major therapeutic challenge in non-small cell lung cancer (NSCLC). To address this unmet need, we developed HJ-004, a potent, selective, and orally bioavailable pan-EGFR mutation degrader that targets classical, rare, and exon 20 insertion mutations. Methods and Results: HJ-004 induces potent and selective degradation of mutant EGFR proteins, but not wild-type EGFR, in Ba/F3 and human cell lines harboring diverse EGFR alterations, with DC50 values below 5 nM, and exhibits strong antiproliferative activity. Mechanistic studies confirmed E3 ligase-mediated ubiquitination and proteasomal degradation. In vivo, HJ-004 produces dose-dependent tumor regression in both murine and human-derived cell-derived xenograft (CDX) and patient-derived xenograft (PDX) models carrying various resistant EGFR genotypes. Pharmacokinetic evaluation demonstrated favorable PK properties, including 30-40% oral bioavailability across multiple species and sustained exposure supporting once-daily dosing. Importantly, HJ-004 exhibits an approximately 10-fold therapeutic window in both mice and monkeys, without any notable skin rashes or diarrhea, indicating excellent selectivity and safety margin. Conclusions: HJ-004 represents a first-in-class pan-EGFR mutation targeted protein degrader that effectively overcomes osimertinib resistance across classical, rare, and exon 20 insertion variants. Its potent degradation activity, broad antitumor efficacy, and favorable pharmacokinetic and safety profiles support advancement into clinical development as a next-generation therapy for EGFR-mutated NSCLC. HJ-004 was independently developed by Jing Medicine Technology (Shanghai) Ltd., and has received IND clearance from both the U.S. FDA and China NMPA. Citation Format: Li Zeng, Yao Liu, Ya Geng, Yue Zhu. HJ-004, a potent and selective pan-EGFR mutation targeted protein degrader (TPD) that effectively overcomes osimertinib resistance in EGFR-mutant NSCLC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5161.

Abstract A major challenge in drug development is the gap between early-phase signals and late-stage success. Despite promising early-stage clinical trial readouts, fewer than 10% of drug candidates ultimately progress to regulatory approval, highlighting substantial gaps in predictive fidelity along the development pathway. This gap complicates decision-making, leading to increased development time, costs and risks. Key contributors to this challenge include the small sample sizes inherent to early-phase cohorts and the limited use of patient stratification, shown to be a contributing factor for higher success rates. Recent advances in artificial intelligence (AI) have introduced powerful tools that have the potential to support clinical trial progress evaluation. By integrating multimodal models with real-world data (RWD) within an AI-driven framework, Imagene AI is developing approaches to address the discrepancies between early and late-phase outcomes, and to support better evaluation of late-phase readouts, such as survival outcomes and biomarkers identification, based on early phase cohorts. In this study, we adopted a multimodal foundation-model strategy, built on a diverse set of foundation models. Among them, our digital pathology foundation model, CanvOI, played a central role in enabling prediction of large-cohort outcomes from small-cohort data. Models were trained on a limited sample set of Trastuzumab (Herceptin)-treated breast cancer patients with outcome data. We then generated Kaplan-Meier survival curves for this small cohort with and without an AI-augmented workflow and compared the results with published outcomes from a Phase III trial. Our findings show that the AI-augmented predictions better align with the Phase III clinical trial outcomes, suggesting that this approach has the potential to support more informed decisions using early-phase data. *ChatGPT was used for editing this abstract Confidentiality Notice: This document is confidential and contains proprietary information and intellectual property of Imagene AI LTD. Neither this document nor any of the information contained herein may be reproduced or disclosed under any circumstances without the express written permission of Imagene AI LTD. Please be aware that disclosure, copying, distribution or use of this document and the information contained therein is strictly prohibited. Citation Format: Inbal Gazy, Assaf Avinoam, Reva Basho, Jonathan Zalach. An AI-driven multimodal workflow for enhancing late-phase clinical trial outcome prediction [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4170.

Abstract Background: PD-1/PD-L1 antibody-based therapies have demonstrated tremendous success in the treatment of a variety of cancers. However, these antibody drugs are associated with several disadvantages, such as weak tumor penetration, immune-related adverse events and emergence of anti-drug antibodies. Here, we report the discovery and characterization of ALG-094295 as a highly potent and orally bioavailable small molecule PD-1/PD-L1 inhibitor that binds to PD-L1 and promotes PD-L1 dimerization, internalization and degradation, offering a different mechanism of action with potential advantages over PD-1/PD-L1 antibody therapeutics. Methods: The interaction of PD-1/PD-L1 and PD-L1 dimerization were assessed by AlphaLISA®. Cellular activity was measured using PD-1 expressing Jurkat NFAT luciferase T cells and CHO-hPD-L1 cells. In vivo PD-L1 target engagement, tumor growth inhibition and tumor infiltration of T-cells were assessed in a humanized-PD-L1 MC38 subcutaneous tumor mouse model. In vitro ADME tox profile was established using standard assays. Pharmacokinetic (PK) studies were performed with rat, dog and cynomolgus monkey. Results: ALG-094295 demonstrated inhibition of PD-1/PD-L1 interaction at sub-nanomolar concentrations and induced PD-L1 dimerization. In vitro studies showed that ALG-094295 activated T cells with approximately ten times greater potency compared to INCB086550, an orally administered small molecule PD-L1 inhibitor that has demonstrated clinical responses in a phase I trial. Furthermore, treatment of CHO-hPD-L1 cells with ALG-094295 resulted in internalization and reduction of PD-L1 protein levels. In ex vivo human PBMC assays, ALG-094295 demonstrated PD-L1 target engagement, T cell activation and immune cell mediated tumor cell killing. In a humanized PD-L1 MC38 mouse model, a single oral dose of ALG-094295 (5 mg/kg) achieved PD-L1 target engagement comparable to INCB086550 (150 mg/kg PO). Daily oral dosing of ALG-094295 (50 or 150 mg/kg) in humanized PD-L1 MC38 mice over 21 days resulted in tumor growth inhibition equivalent to twice-weekly administration of durvalumab (10 mg/kg IV), with tumor size correlating with increased CD8+ T-cell infiltration. ALG-094295 demonstrated no in vitro liabilities for CYP450 inhibition or induction mediated drug-drug interactions, off target toxicity, cardiovascular safety, or genotoxicity. Preclinical in vivo PK data suggests once-daily oral dosing is feasible in humans. Conclusion: ALG-094295 is a highly potent and orally bioavailable small molecule PD-1/PD-L1 inhibitor that promotes PD-L1 dimerization, internalization and degradation. ALG-094295 has the potential to overcome some limitations of antibody-based therapies due to potent PD-L1 blockade, oral delivery and novel mechanism of action. Citation Format: Heleen Roose, Kristina Rekstyte-Matiene, Sarah Stevens, Kusum Gupta, Sandra Chang, Nour Fayyad, Cheng Liu, Vladimir Serebryany, Lillian Adame, Kha Le, Antitsa Stoycheva, Dinah Misner, Lawrence M. Blatt, Sushmita Chanda, David B. Smith, Julian A. Symons, Andreas Jekle, Tongfei Wu. Preclinical Characterization of ALG-094295, a highly potent and orally bioavailable small molecule PD-1/PD-L1 inhibitor targeting dimerization, internalization and degradation of PD-L1 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2818.