Citations

ABSTRACT Purpose We studied histologic Ewing sarcoma (ES) tumor response at local control surgery to explore associations between histologic response and chemotherapy practice changes over time and confirm its prognostic significance in the modern era. Methods We reviewed clinical data for 126 patients with ES treated at Mayo Clinic from 1977 to 2022, analyzing the percentage necrosis as a prognostic factor across chemotherapy regimens (including modern interval‐compressed VDC‐IE). Results In the historical cohort, 5‐year event‐free survival (EFS) and overall survival (OS) for patients with ≤5% versus greater than 5% viable tumor were 76% versus 59% ( p = 0.01) and 83% versus 62% ( p = 0.01). Review of 104 specimens by two pathologists identified the percentage remaining viable tumor as more reproducible than the percentage necrosis and/or fibrosis. On multivariable analysis of patients with localized disease ( n = 87), pelvic tumors and greater than 5% viable tumor remained independent risk factors for poor survival (EFS: hazard ratio [HR] 3.1, p < 0.01; OS: HR 2.1, p = 0.048). The modern cohort consisted of 22 patients. No patients with tumor size less than 8 cm at diagnosis and 100% necrosis after surgery had events; four of seven with tumors greater than 8 cm had events despite 100% necrosis. Conclusions Our results validate the historical threshold of greater than 5% viable tumor with older chemotherapy regimens but not with modern interval‐compressed VDC‐IE. In addition, inter‐rater reliability for the percentage viable is more reproducible than the percentage necrosis. Although numbers are small, tumors less than 8 cm with no viable tumor appear to be a favorable subgroup and suggest risk stratification using a combination of tumor size and any viable tumor should be evaluated in ES.

Abstract Purpose: Despite advances in molecular diagnostics, glioblastoma (GBM) remains incurable, with recurrence driven by residual invasive tumor cells that escape surgical resection and therapy. Current genomic profiling to guide therapy for recurrent disease typically relies on a single-site clinical biopsy (SSCB) obtained from the contrasting-enhancing (CE) tumor core - the region targeted for maximal resection and most visible on Magnetic Resonance Imaging (MRI). However, this region poorly represents the non-enhancing (NE) invasive tumor margins, which harbor treatment-resistant and genetically distinct cell populations responsible for disease recurrence. GBM exhibits extensive spatial heterogeneity, raising concerns that a single tissue may underrepresent the true actionable landscape of the tumor. Experimental Procedures: We analyzed 302 biopsies from 66 GBM patients who underwent multiregional whole exome and RNA sequencing across CE and NE tumor regions. Of these, 17 patients also had standard-of-care SSCB molecular profiling utilizing clinically available genomic sequencing panels outside of the multiregional protocol. Variants that have been targeted in ongoing or successful human clinical trials were annotated as actionable. We compared actionability profiles between SSCB and multiregional sampling, to determine the gain of actionable variant detection across the two sampling strategies. We performed an in-silico biopsy simulation to model the coverage of these alterations across the tumor landscape with increasing sampling frequency in the multiregional protocol, versus the coverage from clinical sequencing panels on SSCBs. Results: Clinical sampling identified a median of one actionable mutation per patient (range 0-3). In contrast, multiregional profiling revealed additional potentially actionable variants in over 70% of patients, particularly within NE infiltrative regions that are not captured by conventional resection. Notably, nine potentially actionable copy-number variants, including PDGFRA amplification, were consistently absent from SSCB results and underrepresented in most clinical panels. In silico biopsy simulations demonstrated that actionable variant detection continues to rise even after sampling five additional regions, reflecting spatial segregation of therapeutic targets. Conclusions: These findings demonstrate that a single sample from the CE tumor core fails to capture the biologically diverse invasive tumor populations responsible for GBM recurrence. Multiregional sequencing uncovers clinically meaningful alterations that would otherwise remain undetected with conventional sampling, highlighting the risk of incomplete molecular characterization in current clinical biopsy workflows. Citation Format: Enmar Harfoush, Taylor Weiskittel, Mylan Blomquist, Dustin Grief, Aadel A. Chaudhuri, Leland S. Hu, Nhan L. Tran, Shannon Fortin-Ensign. Beyond a single core biopsy: Spatial genomic profiling improves therapeutic target detection in glioblastoma [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 706.

Abstract BACKGROUND Intraductal papillary mucinous neoplasm (IPMN) is one of the two most common precursor lesions leading to the development of pancreatic ductal adenocarcinoma (PDA). IPMNs comprise a heterogeneous group of tumors with a wide range of grades and histotypes, and the emergence of single-cell RNA sequencing (RNA-seq) and multiplex digital spatial profiling have characterized unique cell populations, including dysplastic epithelial and immune cells, within the heterogeneous tumor microenvironment that carry signature gene expressions, which could be used as markers for disease progression. However, it is not clear on the heterogeneity of the epithelial and immune response of the evolution during progression of IPMN to PDA. OBJECTIVE Combining nuclear GSK-3β with oncogenic KRas mice (referred to as KNGC – KRas, nuclear GSK-3β, Cre) resulted in the development of IPMN. We have found that 4-week-old KNGC mice show progressive desmoplasia and ductal lesion development with increased M2-like macrophage infiltration. Interestingly, CX3CL1 (fractalkine), a chemokine involved in the recruitment of CX3CR1-expressing monocytes and M2-like macrophage conversion in various disease models is expressed by IPMN progenitor ductal cells. We aim to investigate expression and contribution of CX3CL1-CX3CL1 axis in the development and progression of IPMN. METHODS We utilized a newly generated CX3CL1-mCherry and well-established CX3CR1-GFP knockin/knockout mouse for crossbreeding with KNGC and KGC (KRas, GNas, Cre) for lineage tracing and function analysis by immunofluorescent staining, flow cytometry and CyTOF. RESULTS Both KNGC- and KGC-3LmCherry/+ mice harbor a large population of CX3CL1-mCherry/EPCAM+ epithelial cells as compared to 3LmCherry/+ and wildtype (WT) control mice. In KNGC and KGC mice, we show that there was a unique population of macrophages (CD45+ CD11b+ Ly6G- F4/80+ Ly6c- ) with significantly higher expression of CX3CR1 (CX3CR1Hi) that was not seen in WT mice. Significantly, Knockout of CX3CR1 in KNGC-3RGFP/GFP animals had a paucity of M2-like macrophages, reduced T regulatory cells (Tregs), substantially reduced desmoplasia and impaired development of IPMN. CONCLUSIONS The study provides important knowledge regarding the role of CX3CL1-CX3CR1 axis in recruitment and generation of M2-like macrophages, and their impact on the initiation of the immune/stroma microenvironment facilitating IPMN development. Further investigations using these models have the potential to identify new biomarkers for risk stratification and therapeutic intervention. Citation Format: Li Ding, Kaely Roeck, Easton Maeder, Cheng Zhang, Taylor Weiskittel, Esther Rodman, Mark Maynes, Bibo Zhu, Aaron Johnson, Jie Sun, Li Hu, Daniel Billadeau. Uncoupling CX3CL1-CX3CR1 Axis in IPMN Development Through Transgenic Mouse Models [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research—Emerging Science Driving Transformative Solutions; Boston, MA; 2025 Sep 28-Oct 1; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_3):Abstract nr A064.

4185 Background: Elraglusib is a first-in-class inhibitor of GSK-3ß, a well-credentialed target in cancer implicated in intrinsic oncologic processes and tumor immune response. Preliminary results of the 1801 Part 3B trial (NCT03678883) showed statistically significant benefits for elraglusib+GnP versus GnP for 1-year survival and mOS in mPDAC (companion abstract: Mahalingam et al.). We investigated plasma levels of cytokines/chemokines/soluble cell receptors/growth factors (CCSG) as potential biomarkers of favorable outcomes on elraglusib. Methods: Forty CCSGs were evaluated in pre-dose plasma from patients with previously untreated mPDAC enrolled in 1801 Part 3B treated with GnP (n = 78) or elraglusib+GnP (n = 155) using a Luminex immunoassay. Using Kaplan-Meier statistics and cutpoint determination, CCSGs were assessed for OS predictive ability in the elraglusib+GnP arm. Multivariate models were constructed to predict binarized survival at 12 months using machine learning (ML). Fivefold cross-validation was used in both analyses, and all methods were applied to the GnP arm to identify elraglusib+GnP specific predictors. Results: Data shown as of November 15, 2024. Multiple CCSGs significantly stratified elraglusib+GnP patients. The most extreme prognosticators were IFN-ß (average HR = 2.34), and PD-L1 (average HR = 0.52) (HR shown as high CCSG vs low CCSG). Screening of different ML approaches ranked logistic regression at the top, and hyperparameter grid search identified stochastic gradient solver with ridge regularization as the optimal method. The model had an accuracy of 88% (SD: 3.9%) and a balanced accuracy of 80% (SD: 8.5%). IFN-beta had the most substantial effect size (odds ratio (OR) = 72.28), followed by IL-18 (OR = 23.38) and PD-L1 (OR = 15.32). All other CCSGs had an OR between 0.03 and 6.71. When this model was applied to patients in the GnP arm, accuracy was 68.4%, and balanced accuracy was 43.1%. Conclusions: Many CCSG biomarkers were identified as promising predictors of survival benefit in mPDAC patients treated with elraglusib+GnP. Both univariate statistical and multivariate ML approaches show predictive significance with high interpretability. The initial ML model is specific to elraglusib treatment, not GnP alone. These results indicate that patients’ initial immune state plays a role in response to elraglusib+GnP. However, single CCSGs for enrichment currently exclude too many patients ( > 75%), and thus, panels of biomarkers are being investigated with ML to overcome this limitation. The 1801 Part 3B clinical trial recruitment has been completed, and updated biomarker models reflective of topline OS data, available by April 2025, will be presented.

Abstract Reprogramming of aged donor tissue cells into induced pluripotent stem cells (A-iPSC) preserved the epigenetic memory of aged-donor tissue, defined as genomic instability and poor tissue differentiation in our previous study. The unbalanced expression of RNA exosome subunits affects the RNA degradation complex function and is associated with geriatric diseases including premature aging and cancer progression. We hypothesized that the age-dependent progressive subtle dysregulation of EXOSC2 (exosome component 2) causes the aging traits (abnormal cell cycle and poor tissue differentiation). We used embryonic stem cells as a tool to study EXOSC2 function as the aging trait epigenetic memory determined in A-iPSC because these aging traits could not be studied in senesced aged cells or immortalized cancer cells. We found that the regulatory subunit of PP2A phosphatase, PPP2R5E, is a key target of EXOSC2 and this regulation is preserved in stem cells and cancer.

424 Background: Advances in prostate cancer (PC) care have led to a 98% 10-year survival rate. Prostate Specific Antigen (PSA) is the primary method of follow-up after definitive radiotherapy (RT), with benign PSA ‘bounce’ occurring in up to 30% of patients, causing uncertainty and anxiety. Distinguishing between PSA bounce and biochemical recurrence (BCR) remains challenging. The Phoenix Criteria is the standard definition of BCR since 2006, defined as PSA nadir + 2 ng/dl, but does not consider patient specific information. Our study aims to explore factors associated with bounce and BCR and develop a predictive tool to differentiate between these two events. Methods: PC patients who received RT at our institution between 2006 and 2022 were identified, including only patients with a PSA increase during follow-up. Exclusion criteria included palliative RT, prostatectomy, or insufficient follow-up information. Patients’ clinical data was collected, and PSA increase was classified as bounce or BCR. Variables were compared using Kruskal Wallis or Chi-square. Multivariate logistic regression was performed in the subset of patients with absolute PSA value >2 (n= 581). Results: Of 1783 patients reviewed due to PSA rise after treatment, 694 met the inclusion criteria. There were 213 (30.7%) with PSA bounce and 481 (69.3%) with recurrence. Prior to meeting the Phoenix Criteria, 20 (4.2%) had biopsy-proven recurrence and 18 (3.7%) imaging-proven recurrence. The median delta PSA was 0.8 ng/mL for bounce vs 5 ng/mL for recurrence (p<0.001). The median time from treatment to event was 14.4 months for bounce vs 45.3 months for BCR (p<0.001). Recurrence was significantly associated (p<0.001) with older age at diagnosis, N1 stage, higher risk category, Gleason score, and higher baseline PSA. In the subgroup of patients without ADT, recurrence was associated with a lower median first PSA post-RT (1.9 vs. 2.4, p=0.046) and a lower median nadir before event (0.6 vs. 1.8, p<0.001). Multivariate logistic regression achieved an accuracy of 0.94, with a sensitivity of 0.83, specificity of 0.98, and an area under the curve (AUC) of 0.97. Conclusions: Our results highlight significant differences in clinical factors between bounce and recurrence in post-RT patients. The multivariate logistic regression differentiated the events with 94% accuracy for patients with a rising PSA value >2 subgroup. We highlight the possible improvement that could be accomplished by personalizing a PSA threshold for recurrence versus the Phoenix definition. This tool offers patients and providers a highly specific tool to differentiate between PSA recurrence and bounce, and may reduce stress and unnecessary testing, or help diagnosis recurrence faster. Next steps include prospective validation of the model to assess the impact on clinical practice costs, patient experience, and cancer outcomes.

ABSTRACT Symmetry refers to properties that remain invariant upon mathematical transformations. The principles of symmetry have guided numerous important discoveries in physics and chemistry but not in biology and medicine. Here, we aim to explore the presence of symmetry relationships at the gene expression level as a mean to distinguish between healthy and disease states. We deployed Learning-Based Invariant Feature Engineering - LIFE, a hybrid machine learning approach implemented with two symmetric invariant feature functions (IFFs) to identify Invariant Feature Genes (IFGs), which are gene pairs whose IFF single-value outputs remain invariant across individual samples in a given biological phenotype. Our multiclass classification results across the transcriptomes of 25 normal organs, 25 cancer types, and blood samples obtained from 4 different types of neurodegenerative diseases revealed the presence of unique phenotype-specific IFGs. We constructed networks using these IFGs (IF-Nets) and intriguingly, we demonstrated that the hubs could serve as information encoders, capable of reconstructing sample-wise expression values in relation to their counterpart genes. More importantly, we found that hubs of cancer IF-Nets were enriched with both approved and clinical trial drugs, highlighting “symmetry breaking” as a novel approach for treating diseases.

Abstract Clinical and preclinical research has demonstrated that iPSC-derived NK (iNK) cells have a high therapeutic potential, yet poor understanding of the detailed process of their differentiation in vitro and their counterpart cell development in vivo has hindered therapeutic iNK cell production and engineering. Here we dissect the crucial differentiation of both fetal liver NK cells and iNK cells to enable the rational design of advanced iNK production protocols. We use a comparative analysis of single-cell RNA-seq (scRNA-seq) to pinpoint key factors lacking in the induced setting which we hypothesized would hinder iNK differentiation and/ or functionality. By analyzing key transcription factor regulatory networks, we discovered the importance of TBX21, EOMES, and STAT5A in the differentiation timeline. This analysis provides a blueprint for further engineering new iPSC lines to obtain iNK cells with enhanced functions. We validated this approach by creating a new line of STAT5A-iPSCs which can be differentiated to STAT5A-expressing macrophages with both NK cell and macrophage features such as perforin production, phagocytosis, and anti-tumor functions.

e16331 Background: Elraglusib (9-ING-41), a potent and selective GSK-3 inhibitor, is a targeted anticancer agent with clinical activity in phase 1 trial and undergoing clinical development for the treatment of metastatic pancreatic ductal adenocarcinoma (mPDAC). Elraglusib (9-ING-41) displayed immune-modulatory activity in preclinical models and modulation of serum cytokine profile in clinical specimens. Here, we investigated the correlation between plasma levels of cytokines/chemokines/soluble cell receptors/growth factors (CCSG) and clinical outcomes of patients with mPDAC treated with elraglusib and gemcitabine/nab-paclitaxel (GnP) in an exploratory study. Methods: CCSGs were evaluated in the pre-dose plasma samples of patients with newly diagnosed mPDAC enrolled in the 1801 phase 2 study (NCT03678883) treated with GnP (n = 13) or elraglusib/GnP (n = 32). Using a Luminex immunoassay, 40 CCSGs were screened as predictive biomarkers of median overall survival (mOS) in the elraglusib/GnP arm using 5-fold cross-validation and maximally selected rank statistic cutpoint determination. A subset of CCSGs (n = 10) was remeasured using an MSD immunoassay with greater sensitivity. Maximally selected rank statistic cutpoint determinations were conducted using MSD results for elraglusib/GnP samples with mOS and median progression free survival (mPFS). P values < 0.05 were considered statistically significant. Results: Luminex screening revealed several CCSGs significantly stratified elraglusib/GnP patients by mOS. IL-6 and IL-8 were identified as top candidates with stable and significant results on cross-validation. Using the MSD assay, both were confirmed to stratify elraglusib/GnP patients by OS and extended to analyze PFS. Low level of IL-8 conferred a better OS prognosis (HR: 0.0919, 95%, CI: 0.0205-0.411, P < 0.001) than low IL-6 (HR: 0.205, 95%, CI: 0.063-0.665, P = 0.005) in the elraglusib/GnP treated patients. Correlations with mPFS were similar in the elraglusib/GnP patients for IL-6 (HR: 0.248, 95% CI: 0.072-0.848, P = 0.02) and IL-8 (HR: 0.297, 95%, CI: 0.111-0.743, P = 0.01). Conclusions: Preliminary results of our exploratory study identified low plasma levels of IL-6 and IL-8 prior to elraglusib/GnP treatment as potential predictive biomarkers associated with improved outcomes in previously untreated patients with mPDAC. The 1801 Part 3 Arm B clinical trial recruitment has been completed. We plan to analyze additional pre-dose plasma samples obtained from approximately 200 patients by April 2024. These will be analyzed using interim OS data available by April 2024 and included in the final presentation. Clinical trial information: NCT03678883 .

e16306 Background: Glycogen Synthase Kinase-3beta has been well-credentialed as a therapeutic target in patients with advanced cancer including pancreatic cancer. Elraglusib (9-ING-41), a novel best-in-class GSK-3 inhibitor, is being developed in combination with gemcitabine/nab-paclitaxel (GnP) as first-line therapy for patients with metastatic pancreatic cancer (mPDAC). Using liquid biopsy samples from an ongoing randomized phase 2 trial (1801 Part 3B; NCT03678883), we examined whether gene mutations in cfDNA correlated with clinical outcomes in mPDAC patients treated with GnP or elraglusib/GnP. Methods: Pre-dose plasma samples were obtained from 37 (GnP arm) and 89 (elraglusib/GnP arm) mPDAC patients previously untreated for advanced disease. Mutational analysis of cfDNA extracted from plasma samples was performed by Tempus (Chicago, IL) using the xF+ platform. The Tempus xF+ assay is an NGS cfDNA liquid biopsy tumor profiling assay for identifying genomic alterations in 523 genes. Tempus xF+ reports were used for analysis of gene mutations. Results: In 126 pancreatic cancer patients, frequently mutated genes ( > 10%) consisted of KRAS (93/126, 74%), TP53 (79/126, 63%), CDKN2A (26/126, 21%), DNMT3A (25/126, 20%), NOTCH1/2 (25/126, 20%), and MLL3 (23/126, 18%). Gain-of-function mutations in KRAS, and loss-of-function mutations of TP53, and CDKN2A genes were associated with worse overall survival (OS) in patients treated with elraglusib/GnP (P < 0.05) but not in GnP-treated patients. Multivariate analysis of frequently mutated genes showed that patients with double (KRAS+TP53) and triple (KRAS+TP53+CDKN2A) mutations had significantly worse OS only in the elraglusib/GnP arm. The presence of DNMT3A, NOTCH1/2 or MLL3 gene mutations did not correlate with any clinical outcomes. Conclusions: Our preliminary analysis of gene mutations in liquid biopsy samples obtained from patients in the ongoing 1801 Part 3B phase 2 trial identified potential predictive biomarkers of clinical outcome in mPDAC patients treated with combination of elraglusib/GnP. Clinical trial information: NCT03678883 .