BACKGROUND/OBJECTIVES:Improvements in esophageal adenocarcinoma (EAC) treatment have reduced mortality. While chemoradiation before surgery was previously a standard of care, updated guidelines recommend peri-operative chemotherapy without chemoradiation. Continued investigation into optimal non-operative treatment paradigms for patients who defer surgery or are not candidates for surgery and certain chemotherapy regimens is needed. The impact of induction chemotherapy prior to chemoradiation on survival and surgical outcomes remains unclear. This study assessed survival and surgical outcomes in a real-world cohort of EAC patients receiving induction chemotherapy before chemoradiation. METHODS:This single-institution, IRB-approved, retrospective cohort study included patients with newly diagnosed stage II-IVb (oligometastatic for IVb) EAC who received definitive chemoradiation (radiation ≥ 40 Gy and two cycles of chemotherapy) +/- esophagectomy from 2007 to 2022. Patients receiving induction chemotherapy were compared to those who did not. Endpoints included survival and surgical outcomes. RESULTS:A total of 141 EAC patients received definitive chemoradiation; 83 received induction chemotherapy before chemoradiation. Patients receiving induction chemotherapy were younger (p < 0.01) with slightly lower performance status (p = 0.27) and presented at a more advanced stage (p < 0.001). Median OS was 3.5 years in the induction chemotherapy group compared to 2.2 years (p = 0.10). There was no difference in pathologic complete response (p = 0.81), esophagectomy frequency (p = 0.87), or surgical downstaging between treatment groups (p = 0.84). CONCLUSIONS:In this real-world, single-institutional patient cohort investigating induction chemotherapy prior to chemoradiation in EAC, patients receiving induction chemotherapy did well but did not have a statistically significant improvement in survival outcomes or surgical outcomes. This study showed that significant numbers of real-world patients may not receive esophagectomy. Thus, prospective, randomized clinical trials are warranted to better delineate the efficacy and selection of patients for induction chemotherapy when non-operative approaches are favored.

BACKGROUND:Online adaptive radiotherapy (oART) has recently gained traction due to its potential clinical benefits. However, it presents significant challenges in developing effective quality assurance (QA) programs, particularly within multi-centre, multi-platform clinical trials settings. In this study, a Failure Mode and Effects Analysis (FMEA) was conducted to identify high-risk components of oART across different centres using different platforms and workflows to inform the development of risk-based QA processes for multi-centre clinical trials. MATERIALS AND METHODS:The National Radiotherapy Trials Quality Assurance (RTTQA) Group and domain experts from five UK radiotherapy centres using oART conducted an FMEA following AAPM Task Group 100 (TG-100) guidelines. Through collaborative discussions, participants identified a comprehensive list of oART-specific failure modes (FMs) and agreed on scoring guidance for rating their occurrence (O), severity (S), and detectability (D). Finally, this standardised scoring system and guidance were used by multidisciplinary teams from different centres to independently score the FMs and the risk priority numbers were calculated as RPN = O*S*D. Median RPN scores were used to rank the FMs from high to low risk. RESULTS:The two FMs representing the greatest risk in the oART workflow were linked to the daily outlining step; incorrect contouring due to missing or inaccurate clinical information RPN (Median [Range]) = 96 [16 - 120], followed by target or OAR delineation errors RPN = 56 [16 - 140]. These failures were mostly associated with insufficient manual review. Overall, contouring, imaging, dose calculation, and plan appraising FMs were rated as highest risk. These FMs were predominantly operator/workflow dependent. CONCLUSION:FMEA identified critical oART-specific FMs requiring targeted QA within clinical trials. Mitigating variability is fundamental to ensuring the validity of clinical trial endpoints. Workflow variability matters: scoring differences across centres highlight the need for harmonisation or cross-validation via end-to-end testing.

3146 Background: The MTA-cooperative PRMT5 inhibitor BMS-986504 selectively binds to the PRMT5-MTA complex, a synthetic lethal target in MTAP -del cancer cells, while sparing MTAP –wild-type normal cells. In the phase 1 CA240-0007 study, BMS-986504 was well tolerated and had antitumor activity in pts with MTAP -del advanced solid tumors (ORR, 23%; median DOR, 10.5 mo). BMS-986504 also demonstrated dose-dependent reductions in plasma symmetric dimethylarginine (SDMA) levels, a PD biomarker of PRMT5 inhibition. Here, we report expanded PD and genomic analyses and initial results of exploratory proteomic analyses from CA240-0007. Methods: Pts with advanced, unresectable, or metastatic solid tumors with homozygous MTAP -del received BMS-986504 (50–800 mg) in 3-wk cycles. Analyses of SDMA levels and proteomic profiles from paired plasma samples (baseline and C2D1) were performed using mass spectrometry and SomaScan 11K, respectively. Tumor genomic information was extracted from local pathology reports. ORR was assessed per RECIST v1.1. Results: Among 78 pts with paired samples at data cutoff (22 Sept 2025; median f/u, 15.9 mo), dose-dependent reductions were seen in plasma SDMA levels, with the greatest reductions at 400 mg QD (median, 60.3%; n = 31) and 600 mg QD (median, 60.3%; n = 23). Proteomic pathway enrichment analysis demonstrated downregulation of PI3K/Akt/mTOR, Myc target, cell cycle, and DNA damage repair (DDR) pathways with BMS-986504 on C2D1 (n = 87), consistent with the expected effect of PRMT5 inhibition; a trend toward deeper modulation of PRMT5-regulated pathways was observed at higher doses. No proteins at baseline (n = 94) were found to be significantly associated with ORR (false discovery rate > 0.05). Responses in clinically evaluable pts with NSCLC (n = 33) or PDAC (n = 35) were observed regardless of EGFR , KRAS , or TP53 status. In pts with NSCLC, STK11 and SMARCA4 alterations were numerically higher in nonresponders. No gene alterations were associated with response in PDAC, except GABRA6 alterations (n = 2, both responders). A trend toward higher ORRs was seen in pts with alterations in DDR genes ( CHEK2, FANCD2, XRCC4 , and/or MSH2 ), consistent with the role of PRMT5 in DDR via regulation of mRNA splicing. ctDNA analysis will be presented. Conclusions: BMS-986504 demonstrated robust dose-dependent PD effects in pts with advanced solid tumors in plasma SDMA and proteomic analyses. Downregulation of PRMT5-regulated pathways was observed with BMS-986504 treatment. Response was seen regardless of EGFR, KRAS , and TP53 alterations in NSCLC or PDAC, and DDR gene alterations were associated with improved response. These findings further support the MOA of BMS-986504 and its ongoing investigation in the phase 2/3 MountainTAP-9, -29, and -30 studies as a potential treatment option in pts with MTAP -del advanced solid tumors. Clinical trial information: NCT05245500 .

BACKGROUND:The MORPHEUS platform comprised multiple open-label, randomized, phase Ib/II trials to identify early signals with different treatment combinations across multiple cancers. MORPHEUS-PDAC (NCT03193190) evaluated atezolizumab combinations in pancreatic ductal adenocarcinoma (PDAC). We describe outcomes with atezolizumab plus either motixafortide, cobimetinib, or two simlukafusp alfa regimens. METHODS:Eligible patients with advanced, pretreated PDAC were randomized to receive second-line (2 L) atezolizumab plus either motixafortide (BL8040; n = 15), cobimetinib (n = 14), simlukafusp alfa every 2 weeks (q2w; n = 15), or simlukafusp alfa every 3 weeks (q3w; n = 16); or control (mFOLFOX6 [n = 25] or gemcitabine plus nab-paclitaxel [n = 25]). Patients experiencing disease progression or toxicity who met eligibility criteria were enrolled to receive third-line (3 L) atezolizumab plus cobimetinib (n = 14), or atezolizumab plus simlukafusp alfa q2w (n = 1) or q3w (n = 6). Primary endpoints were objective response rates (ORRs) per RECIST 1.1 and safety. RESULTS:ORRs were 7.1% with atezolizumab-simlukafusp alfa q2w, 8.7% with mFOLFOX6 (both 2 L; 0% in other arms), 14.3% with atezolizumab-cobimetinib, and 16.7% with atezolizumab-simlukafusp alfa q3w (both 3 L). Grade 3-5 adverse event rates were 53.3% (2 L atezolizumab-motixafortide), 64.3% (2 L atezolizumab-cobimetinib), 57.1% (2 L atezolizumab-simlukafusp alfa q2w), 53.3% (2 L atezolizumab-simlukafusp alfa q3w), 63.0% (2 L mFOLFOX6 or gemcitabine-nab-paclitaxel), 50.0% (3 L atezolizumab-cobimetinib), and 100% (3 L atezolizumab-simlukafusp alfa q3w). CONCLUSIONS:The overall safety of atezolizumab combinations was manageable and consistent with each agent's known safety profile. This novel trial design enabled rapid evaluations of 3 atezolizumab combinations; all had limited efficacy as 2 L or 3 L treatment for metastatic PDAC. New treatments are needed to improve outcomes in previously treated PDAC.

399 Background: The standard of care for resectable GEA adenocarcinomas is perioperative chemotherapy with the FLOT (fluorouracil, leucovorin, oxaliplatin, docetaxel) regimen. Recent studies investigating the addition of checkpoint inhibitors (CPIs) to perioperative chemotherapy have shown mixed results. An NMA can leverage both direct and indirect comparisons to evaluate the relative efficacies of available regimens in this setting. Methods: MEDLINE, EMBASE, Scopus, Web of Science and CENTRAL were searched until August 27th, 2025. Eligible studies included phase 3 randomized trials evaluating perioperative or neoadjuvant chemotherapy +/- CPIs ± radiation in resectable GEA. The primary outcomes were disease-free survival (DFS) and overall survival (OS). Hazard ratios and 95% confidence intervals (CIs) were estimated using a frequentist NMA framework. Results: Fifteen trials including 8072 patients were analyzed. For DFS, perioperative durvalumab + FLOT (dFLOT) ranked highest (P-score = 0.99), followed by FLOT (P-score = 0.84) (HR for dFLOT vs. FLOT was 0.71, 95% CI 0.58-0.86). dFLOT also ranked higher than perioperative pembrolizumab + cisplatin and fluorouracil/capecitabine (KN-585), though this comparison was not statistically significant (HR 0.65, 95% CI 0.40-1.04). Compared with CROSS, dFLOT showed superior DFS (HR of 0.48, 95% CI 0.36-0.63). For OS, dFLOT ranked the highest (P-score=0.95) and was superior to CROSS (HR 0.57, 95% CI 0.42-0.77). While dFLOT appeared favorable versus FLOT (HR 0.78, 95% CI 0.62-0.98), this finding was primarily driven by MATTERHORN, which did not meet its prespecified OS significance boundary (P = 0.03 vs threshold P < 0.0001), and the proportional hazards assumption was violated. Similarly, no significant OS difference was observed between dFLOT and the KN-585 arms (HR 0.75, 95% CI 0.45-1.23). Conclusions: In this NMA, perioperative FLOT remained the benchmark regimen for resectable GEA. dFLOT has shown promising early results, ranking the highest for DFS and showing favorable, but not statistically definitive OS trends in comparison to FLOT. Importantly, the network enabled indirect comparisons not tested in head-to-head trials, demonstrating that dFLOT ranked above pembrolizumab+CF/CX (KN-585) without reaching statistical significance, and also outperformed CROSS. However, these findings should be interpreted with caution, as they are primarily driven MATTERHORN data in which OS did not cross the prespecified threshold and proportional hazards assumptions were violated. Perioperative immunotherapy combinations appear promising, but final MATTERHORN results and further confirmatory studies are needed before dFLOT can be considered as a new standard of care.

Background: Stromal hyaluronic acid (HA) poses a physical barrier and protects tumor cells from immune surveillance. Stroma targeting with pegylated human recombinant PH20 hyaluronidase (PEGPH20) demonstrated improved infiltration of cytotoxic T-lymphocytes and delivery of chemotherapy and PD1/PD-L1 antibodies in tumor models. This multicenter phase II study of PEGPH20 plus pembrolizumab evaluated the efficacy, safety and immune and stromal biomarkers in patients with HA-high refractory metastatic pancreatic ductal adenocarcinoma (mPDA). Patients and Methods: Patients were treated with PEGPH20 3 µg/kg IV weekly and pembrolizumab 200 mg IV in 3-week cycles. Tumor and blood samples were collected at baseline and on-study for biomarker analyses. Results: Between May and November 2019, 38 patients were screened and 8 treated, with median age 68 years (range 60-73) and median two (range 1-4) prior therapies. The study was closed to accrual early by pharmaceutical sponsor. Treatment was well tolerated, with expected grade 1/2 musculoskeletal toxicities. Best response was stable disease in 2 of 7 evaluable patients (29%). Median overall and progression-free survival were 7.2 months (95% CI 1.2-11.8) and 1.5 months (95% CI 0.9-4.4), respectively. Prolonged survival (range 10.2-27.6 months) occurred in patients treated with subsequent chemotherapy. Higher baseline tumor T cell receptor (TCR) clonality correlated with longer survival. Conclusions: Pembrolizumab with PEGPH20 was safe but did not have significant efficacy in refractory HA-high metastatic PDA.

Magnetic resonance imaging (MRI) offers superior soft tissue contrast compared to computed tomography (CT), making it highly valuable in external beam radiotherapy (EBRT) planning. However, there are a number of barriers that have limited widespread use of MRI for EBRT planning in the UK such as limited access to MRI scanners and lack of training and guidance. Following the 2018 Institute of Physics And Engineering In Medicine (IPEM) survey on MRI use in UK RT centres, data were collected from 68 centres across the UK in 2025 to reassess MRI access, utilisation, and adherence to 2021 IPEM guidance on MRI in EBRT planning. With a 79% (54/68) complete response rate, the survey revealed increased integration of MRI into EBRT planning workflows, particularly for brain, spine, and prostate cancers. However, access remains variable, with only five centres reporting MRI scanners dedicated for RT. Compliance with recommended imaging MRI sequences and quality assurance procedures has improved but remains variable, especially among centres relying on Picture archiving and communication systems-sourced images. Barriers such as capital investment, staffing, and training persist, although clinical engagement and future planning for MRI-only workflows and Artificial Intelligence-based tools are increasing. These findings underscore the need for continued investment, updated guidance, and multidisciplinary collaboration to support the safe and effective expansion of MRI in RT planning.

Therapies targeting the RAF-MEK-ERK pathway are generally considered to have limited efficacy in KRAS-mutant cancers. However, specific KRAS mutants exhibit distinct behaviors. Notably, KRASG12R pancreatic ductal adenocarcinoma (PDAC) tumors have shown sensitivity to MEK inhibitors (MEKi) in combination with autophagy inhibitors, but a better understanding of the underlying mechanisms is needed to optimize this treatment strategy. Using a systems-level approach, we uncovered a mechanistic explanation for this phenomenon. Due to distinct biophysical properties, KRASG12R had an impaired ability to activate wild-type HRAS and NRAS (WT-RAS) compared with other KRAS mutants, such as KRASG12D. This reduced activation stemmed from the weaker interaction between KRASG12R and guanine exchange factors (SOS), as well as the tumor suppressor neurofibromin (NF1), crucial in regulating WT-RAS activity. The impaired ability to activate WT-RAS led to weaker holistic MAPK signaling in KRASG12R-driven tumors, which conferred increased sensitivity to MEKi. To substantiate the preclinical findings, the utility of MEKi in combination with the autophagy inhibitor hydroxychloroquine was analyzed in patients with KRASG12R-mutated metastatic PDAC. Five of the eight (62.5%) patients treated in first- or second-line settings had a progression-free survival exceeding 6 months. Three patients had impressive disease control: two had stable disease of 11 and 22.7 months, and one achieved a partial response with an 83% decrease in tumor size that lasted for 8.9 months. Overall, this work highlights how systems-based approaches in precision medicine can uncover mechanistic insights to guide the identification of patients with PDAC most likely to benefit from tailored therapeutic strategies. SIGNIFICANCE:The unique sensitivity of KRASG12R-mutant cancers to MEK inhibitors offers a critical advancement in understanding MAPK signaling and paving the way for precision-targeted therapies in previously untreatable contexts. See related commentary by Tiriac and Engle, p. 1817 See related article by Burge et al., p. 1854 See related article by Burge et al., p. 1868.

Clinicogenomic characteristics and treatment lines of 8 patients with metastatic PDAC and KRAS G12R alteration treated with MEK inhibitor + hydroxychloroquine

Kaplan-Meier estimates of progression-free survival

Supporting Data for Isogenic cells

Importance:Cancer antigen 19-9 (CA19-9) is used to assess treatment response among patients with pancreatic ductal adenocarcinoma (PDAC); however, nearly 30% of patients with PDAC do not produce elevated CA19-9. Objective:To develop, validate, and apply an electronic tumor marker (e19-9) derived from routine laboratory data available in the electronic health record to assess treatment response and predict outcomes among patients with PDAC who do not produce CA19-9. Design, Setting, and Participants:In this cohort study, an artificial intelligence (AI) model was trained using routinely collected serum laboratory data from patients with PDAC and elevated CA19-9. The model was externally validated and then applied to a separate cohort of CA19-9 nonproducers. Model development and internal testing were conducted at a single institution using patient data from 2010 to 2022. External validation used a deidentified patient network across 58 health care organizations over the same period. The training cohort included 3239 patients with pancreatic cancer and elevated CA19-9. The external validation cohort included 4384 similar patients. The model was applied to 121 patients with resectable or borderline resectable PDAC who did not produce elevated CA19-9 and received neoadjuvant therapy with curative intent. These data were analyzed from November 2021 through March 2025. Main outcomes and measures:Model performance was assessed using root mean square error and R2. Clinical outcomes included completion of all neoadjuvant treatment and surgery, metastatic progression, and overall survival (OS). Results:The final fitted model demonstrated stable performance across both internal and external validation cohorts. Among 121 patients (59 female and 62 male) with localized PDAC who did not produce elevated CA19-9, a 50% or more decline in e19-9 (area under the curve [AUC], 0.79) and e19-9 level of less than 100 (AUC, 0.84) were objectively determined cut points associated with prognosis. A total of 93 patients (77%) completed all planned neoadjuvant therapy and surgery. A 50% or more decline in e19-9 levels and an e19-9 level less than 100 was associated with completion of all intended therapy (odds ratio [OR], 5.00; 95% CI, 1.60-15.66; P = .006 and OR, 19.31; 95% CI, 5.80-64.26; P < .001). An e19-9 level less than 100 was independently associated with OS (hazard ratio, 0.49; 95% CI, 0.25-0.97; P = .04). Conclusions and relevance:In this study, e19-9 was a noninvasive AI-derived marker that may provide accurate and relevant information to assess treatment response for the approximately 30% of patients with PDAC who do not produce CA19-9 at elevated levels. The development and validation of scalable, noninvasive screening methods using machine-learning algorithms may pave the way for early detection, prognostication, and treatment of cancers.

Background and Purpose:Organ at risk (OAR) motion results in increased risk of radiotherapy treatment side effects due to the potential for greater dose than intended being delivered. This is of particular concern in stereotactic ablative body radiotherapy (SABR). The study aim was to assess the role of planning organ at risk volume (PRV) margins applied to OARs of the gastrointestinal (GI) tract over five fraction non-adaptive SABR for pancreatic cancer. Materials and Methods:The clinical adaptive magnetic resonance-guided radiotherapy baseline planning for 10 patients was combined with retrospective replanning respecting constraints to OAR plus a margin of 2-5 mm. Baseline planning/replanning was applied to OAR contouring at treatment fractions without plan adaptation. The impact of PRV margin was assessed in terms of the encompassing of OAR motion, effect on delivered dose, and degree of associated compromise in planned treatment volume (PTV) dose coverage. Results:Whilst variable across individual OARs, a 2-5 mm margin poorly encompassed motion in 40-70% of fractions. However, a 2-5 mm margin did effectively limit dose increases in most cases such that delivered biological effective dose (BED) did not exceed, on average, the baseline value. A margin of 2-3 mm resulted in minor PTV compromise (D70% median PTV BED decrease <10 Gy), whereas a 5 mm margin resulted in moderate compromise (D70% median PTV BED decrease ∼20 Gy). Conclusions:A PRV margin of 2-5 mm for OARs limits the increase in delivered dose. Consideration of such a margin is relevant in optimizing non-adaptive treatment practice.

INTRODUCTION:Pancreatic Ductal Adenocarcinoma (PDAC) is often caused by mutations in multiple genes including KRAS (activating the Ras-Raf-MEK-ERK pathway). This study evaluated the role of MEK inhibitor (MEKi)-based combinatorial targeted therapies in patients with PDAC. Methods. This is a retrospective/prospective observational, single institution study, including 29 patients with metastatic PDAC with KRAS alterations, treated with MEKi therapies between 2022-2024. RESULTS:Ten patients had KRAS G12R (34.5%), ten G12D (34.5%), and nine G12V (31%). Majority of patients received MEKi therapy as third-line and beyond (KRAS G12R/G12D/G12V 60%/50%/78%, respectively). Median overall survival from MEKi initiation for KRAS G12R/G12D/G12V was 8.2/5.1/4.7 months (P = 0.5), respectively, and median progression-free survival was 4.4/2.3/1.4 months (P = 0.11). Six (21%) patients discontinued at least one drug in the treatment combination due to toxicity. CONCLUSIONS:MEKi-based combinatorial therapies had modest disease control in patients with KRAS G12R, and minimal disease control in patients with KRAS G12D/V in the late-line setting.

Pancreatic ductal adenocarcinoma (PDAC) is among the most lethal solid malignancies, characterized by aggressive biology and a paucity of effective treatments. Activating mutations in KRAS occur in more than 90% of cases and are fundamental to tumor initiation, progression, therapeutic resistance, and immune exclusion, establishing KRAS as the dominant oncogenic driver in PDAC. Long considered undruggable, KRAS has recently become a viable therapeutic target with the development of allele-specific inhibitors as well as pan-RAS(ON) agents capable of broadly suppressing mutant RAS signaling. Preclinical models and early-phase clinical trials demonstrate meaningful antitumor activity, with emerging evidence of tumor microenvironment remodeling and delayed resistance. Combination strategies integrating KRAS-directed therapies with chemotherapy, vertical pathway inhibition, immunotherapy, and emerging approaches such as KRAS degradation and RNA-targeted approaches are being explored to improve the depth and durability of response. Together, these advances signal a paradigm shift toward molecularly guided treatment strategies in PDAC and offer a promising path forward in a disease with substantial unmet clinical need.

Patient, tumor and treatment characteristics

Mutant KRAS allele dependent activation of WT-RAS

Purpose: This study aimed to generate a map of local recurrences after neoadjuvant chemotherapy and radiation (total neoadjuvant therapy [TNT]) followed by surgical resection for pancreatic ductal adenocarcinoma (PDAC). Such recurrence patterns will serve to inform radiation treatment planning volumes that should be given in the neoadjuvant setting. Methods and Materials: Locoregional recurrences after TNT followed by surgery treated between 2009 and 2022 were radiologically identified. Recurrences were individually segmented using MIM software and complied in a single base scan. All contour compilations were used to create a threshold contour encompassing 80% of recurrences among all patients, head only, and body/tail only. The distance between organs at risk and the threshold contour were measured to design an optimal clinical target volume contour for patients treated with TNT. Recurrence patterns were also compared with existing adjuvant guidelines to assess coverage. Results: A database of 474 patients managed with TNT for PDAC was queried. While locoregional recurrences were rare in this cohort, we identified 80 patients with either isolated locoregional or simultaneous local and distant recurrences. Patients with diagnostic imaging at the time of recurrence were identified. The majority of recurrences were partially in the field of published contouring guidelines or volumetric expansions off of vessels, and volumetric coverage was low for all. Common areas of recurrence include the aorticodiaphragmatic junction, retropancreatic duodenal nodal basin, and the region to the right of the superior mesenteric artery. A novel set of proposed neoadjuvant contours was designed to cover the central-most 80% of recurrences. Conclusions: This is the largest collection of local/regional PDAC recurrences from a cohort of patients treated exclusively with TNT. Patterns of local/regional recurrence using TNT in PDAC vary significantly from those patients with PDAC treated with a surgery-first approach. Novel contouring guidelines presented in this study can help to ensure optimal coverage of high risk regions and avoid reliance on the current adjuvant guidelines to guide treatment planning. Published by Elsevier Inc. on behalf of American Society for Radiation Oncology.

e13706 Background: Comprehensive clinical trial screening for all oncology patients is challenging due to multiple factors. Traditional approaches rely on manual review, making enterprise-level trial screening infeasible. Centralized Clinical Trials Offices (CTO) prioritize interventional therapeutic trials, further limiting screening for interventional, non-therapeutic trials. Large language model (LLM)-based AI can enhance efficiency by systematically screening all patients against an institution’s full clinical trial portfolio before each visit. This study evaluated whether an LLM-based system could improve trial-to-patient matching using a representative clinical trial. Methods: This retrospective study assessed patients at the Medical College of Wisconsin Cancer Center from December 2024 to January 2025. OncoLLM, a fine-tuned LLM trained on institutional clinical data, cancer guidelines, and oncology textbooks, processed unstructured EHR data for patients scheduled in participating clinics to identify potential trial matches. OncoLLM is deployed using Triomics’ PRISM platform, employing a two-tiered approach: Primary screening: Identify relevant trials using disease site, histology, stage, and biomarker status. Deep screening: Evaluate full inclusion/exclusion (I/E) criteria, transforming each criterion into a structured query with a determination of eligibility. The primary outcome was the number of trial-to-patient matches to ALLIANCE-A222004-ANOREXIA (NCT04939090). Secondary outcomes included measuring LLM accuracy, defined as a recommendation accepted by the study team, and reasons for inaccuracy. Results: A total of 2,277 patients with 38 tumor types were screened, representing 100% of patients seen during the study period. 49 patients were recommended as high-probability trial candidates for further assessment, and all recommended patients met key inclusion criteria, warranting deep screening by PRISM. A total of 22 of the 49 recommendations were accepted by the study team, resulting in patients being watch-listed for the study. Reasons for rejection included not meeting study-specific criteria, medical/health related, and logistical or health communication. Compared to traditional methods, OncoLLM resulted in a 40% increase in identified trial matches. Conclusions: OncoLLM enabled enterprise-scale, AI-driven screening, identifying a greater number of potentially eligible patients than manual methods. OncoLLM demonstrated high criterion-level accuracy and feasibility for broad-scale clinical trial matching. As precision oncology continues to drive complexity in trial eligibility, AI-powered tools like OncoLLM are essential for democratizing trial access, optimizing enrollment, and reducing inefficiencies in oncology research.