Borad, Mitesh J.

ABSTRACT The limited efficacy of current therapies against cholangiocarcinoma (CCA) necessitates the development of novel treatment strategies. Src family kinases (SFKs) contribute significantly to tumor progression and resistance in CCA. Therefore, we investigated the novel, first-in-class SFK ‘OFF’ inhibitor NXP900 in diverse preclinical CCA models, including those with acquired resistance. This study evaluated the therapeutic effects of NXP900 and detailed adaptive molecular responses to SFK inhibitor therapy. We also aimed to identify biomarkers predictive of drug sensitivity using integrated multiomic profiling and develop strategies to overcome resistance. NXP900 inhibited YAP activity through direct inhibition of tyrosine phosphorylation and indirect activation of the Hippo pathway via LATS. These effects were associated with decreased tumor cell viability in CCA cell lines and several in vivo models. Notably, IDH -mutant patient-derived xenograft CCA models were particularly sensitive to NXP900. NXP900 also synergized with gemcitabine/cisplatin chemotherapy, enhancing antitumor efficacy in both in vitro and in vivo models. Multiomic analyses combining transcriptomics, global proteomics, and phosphoproteomics identified molecular features associated with primary response and acquired resistance. IL13RA-AKT signaling was upregulated in resistant models; NXP900 sensitivity could be restored with AKT or IL13RA2 inhibition. Together, these findings demonstrate the therapeutic potential of NXP900 as a novel YAP inhibitor in CCA and support further investigation in a clinical trial. Graphical Abstract

837 Background: A recent randomized trial suggested that CR may influence ICI activity in patients with lung cancer. However, the same has not been established in GI cancers, where reports have shown mixed results with variable definitions of 'early' vs 'late' IT. We evaluated multiple approaches to assess the impact of CR on first-line ICI in GI malignancies. Methods: This retrospective study included patients from three Mayo Clinic sites (N=455). Both previously reported IT cutoffs and a new cutoff derived from ROC analysis were evaluated. Patients were categorized by the median of all ITs, the median of the first four ITs, and whether ≥20% of ITs occurred before the identified cutoff. Univariate and multivariate Cox regression (adjusting for age, MSI status, type of cancer, metastasis status, number of cycles, and receipt of chemotherapy with ICI) assessed associations with OS. KM analyses were performed to illustrate survival differences. The primary endpoint was OS. Results: When patients were categorized based on whether ≥20% of infusions occurred before 14:00 (ROC-derived cutoff), a significant association with OS was observed. On univariate analysis, early infusion timing was associated with improved survival (HR 2.30, 95% CI 1.58–3.32, p <0.001). This remained significant after multivariable adjustment (HR 2.11, 95% CI 1.41–3.16, p < 0.001). Median OS was 14.2 months in the early group vs 3.8 months in the late group (log-rank p < 0.001). Using previously defined cutoffs, categorization by the median IT of all infusions or the median of the first four infusions did not demonstrate a significant association with OS in either univariate or multivariate Cox regression analyses. Conclusions: These findings highlight the potential role of infusion timing as a chronotherapy factor in GI cancers. Given the retrospective design and heterogeneous population, prospective studies are warranted to validate this signal and clarify its clinical relevance. Baseline demographics of patients by infusion timing (early vs late). Variables Early N=412 (%) LateN=43(%) Total N=455(%) p-value Age <70 249(60.4) 21(48.8) 270 (59.3) 0.1 >70 163(39.6) 22(51.2) 185 (40.7) Gender F 118(28.6) 16(37.2) 134 (29.5) 0.3 M 294(71.4) 27(62.8) 321 (70.5) Cancer Type Gastroesophageal 109 (26.5) 8 (18.6) 117 (25.7) 0.1 Hepatocellular 159 (38.6) 24 (55.8) 183 (40.2) Biliary 86 (20.9) 4 (9.3) 90 (19.8) MSI-H 58 (14.1) 7 (16.3) 65 (14.3) Metastasis No 150 (36.4) 22 (51.2) 172 (37.8) 0.1 Yes 262 (63.6) 21 (48.8) 283 (62.2) Immunotherapy Cycle Number <6 189 (45.9) 33 (76.7) 222 (48.8) <0.001 >6 223 (54.1) 10 (23.3) 233 (51.2)

830 Background: Immune checkpoint inhibitors (ICIs) are standard therapy for MSI-H GI cancers. Lynch syndrome, caused by germline mutations in mismatch repair (MMR) genes, leads to MSI-H tumors and robust ICI responses, but its effect on irAEs remains poorly defined. Clarifying this risk may inform treatment selection and toxicity monitoring. Methods: We retrospectively reviewed MSI-H or MMR-deficient (dMMR) GI cancer patients (pts) treated with ICIs across Mayo Clinic (AZ, MN, FL) from January 2017 to June 2025. Eligible pts were identified using Mayo Data Explorer. Demographics, treatment, and irAEs (graded per CTCAE v5.0) were collected. Lynch status was determined by germline testing when available. Logistic regression estimated odds ratios (ORs) for any grade and grade ≥3 irAEs, adjusting for age, cancer type, prior chemotherapy, and treatment regimen. Results: A total of 209 pts with dMMR/MSI-H GI cancer were included: 61 with Lynch syndrome and 148 without. Lynch pts were younger (median 58 vs 76 years, p < 0.001) and more likely to have received prior chemotherapy (62% vs 43%, p = 0.009). Colorectal cancer was the predominant cancer type (74%), but Lynch pts had proportionally more non-colorectal GI cancers (15% vs 2.7%, p = 0.007). Overall, 29% of pts experienced an irAE, with rates significantly higher in Lynch vs non-Lynch pts (43% vs 24%, p = 0.005). The most frequent irAEs were gastrointestinal, endocrine, and dermatologic, followed by hepatic and pulmonary events. Grade ≥3 irAEs occurred in 16% overall, with no significant difference between groups (p = 0.124). In adjusted models, Lynch syndrome was independently associated with a higher risk of any grade irAE (OR 2.14, 95% CI 1.09 - 4.19, p = 0.026), while risk of severe irAEs was not increased (OR 1.55, 95% CI 0.64 - 3.59, p = 0.316). Lynch syndrome pts were more likely to have combination therapy (17% vs 6.8%, p = 0.029). However, there were no differences observed in any grade irAE between monotherapy and combination immunotherapy (OR 1.36, 95% CI 0.49 - 3.54, p = 0.535). Conclusions: Lynch syndrome is associated with a higher prevalence of any-grade irAEs among dMMR/MSI-H GI cancer pts treated with ICIs, though severe events were not increased. Incorporating Lynch status into immunotherapy planning may improve risk stratification and guide regimen selection.

776 Background: Surgery is the only potentially curative option for PC. However, only a minority of patients (pts) undergo resection with current perioperative (periop)-chemotherapy (CT). In the absence of phase III trials, selection between mFOLFIRINOX and gemcitabine/nab-paclitaxel (gem-nab) is based on limited evidence. We assessed whether NGS–based tumor profiling can guide tailoring of CT. Methods: pts with resected PC (resectable, borderline resectable, and downstaged locally advanced) that underwent Tempus xT, xF (DNA), and/or xR (RNA) testing were selected (89% tissue, 11% blood). Most samples (73%) were collected post-CT. Samples were stratified by KRAS status into KRAS -mutated ( KRAS mut) and KRAS wild-type ( KRAS wt). PD-L1 IHC was reported as tumor proportion score (TPS). Primary endpoint was median overall survival (mOS) defined from CT initiation to death with censoring of pts alive at the last follow-up or at a study cutoff of 5 years. Secondary endpoint was molecular profiling by KRAS status. Pearson chi-square and Wilcoxon rank-sum tests were used for descriptive comparisons; univariate Cox regression and log-rank tests (p<0.05) for survival. Results: We included 1,325 pts (median age 66 years [IQR 59–72]) with resected PC [stage II (15%), stage III (12%) and stage I (10%), unknown (62%)]. 95% had adenocarcinoma. Pts received neoadjuvant (50%), neoadjuvant plus adjuvant (26%), neoadjuvant with indeterminate adjuvant allocation (13%) or adjuvant CT (11%). First-line periop-CT was mFOLFIRINOX in 73% and gem-nab in 27% (median duration 4.3 months (mo) [range, 2.6–6.7]). KRAS mutations were present in 75% (G12X 91%: G12D 44%, G12V 34%, G12R 21%, G12C <1%). mOS was 26.7 mo (95%CI, 22.4-35.7) in the total cohort; KRAS mut pts showed a clinically meaningful worse mOS compared to KRAS wt [24.1 vs 35.7 mo; HR 1.61 (95%CI, 0.98 - 2.63); p=0.06]. In the KRAS G12D pts, mFOLFIRINOX showed a trend towards longer mOS compared to gem-nab [17.79 vs 14.53 mo; HR 0.75, (95% CI, 0.56 - 1); p=0.050]; no significant differences were observed in other KRAS subgroups. Compared to KRAS wt, KRAS mut pts were enriched for TP53 (76% vs 22%; p<0.001), SMAD4 (28% vs 5.8%; p<0.001), CDKN2A (33% vs 8.5%; p<0.001), MTAP deletions (7.7% vs 1.2%; p<0.001), and chromatin-regulator alterations ( ARID1A 7.8% vs 1.2%; p<0.001; KMT2C 2.5% vs 0.6%; p=0.034). KRAS wt pts were enriched for BRAF V600E (1.5% vs 0%; p<0.001). PD-L1 TPS ≥1% was more frequent in KRAS mut compared to KRAS wt (11.0% vs 8.1%; p=0.033). Conclusions: KRAS status does not predict benefit from mFOLFIRINOX vs gem-nab in resected PC. Interestingly, resected PC showed a higher prevalence of KRAS wt compared to the metastatic setting. KRAS status is associated with distinct profiles of potentially targetable co-alterations. These findings may suggest the integration of genomic profiling in clinical trials to develop biomarker-driven tailored strategies in the early stage.

679 Background: DNA-damaging agents improve outcomes in metastatic pancreatic cancer (mPC) with homologous recombination repair (HRR) deficiency, albeit use is limited to 6-9% of patients (pts) with BRCA1/2 or PALB2 mutations. We hypothesized that BRCA/PALB2 wildtype mPC may exhibit platinum sensitivity driven by altered HRR gene expression. We correlated HRR gene RNA profiles with real-world overall survival (rwOS) following first-line (1L) platinum therapy (PT). Methods: Tempus Lens, a platform used to query multimodal data from millions of de-identified patient records in the Tempus Database was used to identify mPC pts with wildtype somatic BRCA1/2 and PALB2 who had Tempus xT DNA and xR RNA testing. RNA-seq data were normalized to correct for assay/batch effects, quantified as transcripts per million (TPM) and reported as log2(TPM+1) for 17 HRR genes and GATA6 . Pts were classified as high and low expressors based on the top and bottom quartile gene expression. Pts were grouped as PT-treated (received 1L PT) or PT-naïve (never received PT). Real world overall survival (rwOS) was calculated from 1L start to death or loss to follow up. Risk set adjustment was applied to mitigate immortal time bias. Median rwOS was estimated using Kaplan-Meier and univariate Cox models. A multivariate Cox Proportional (CoxPh) model included PT use, RAS status, and expression of genes (as continuous variables). Results: Among 1,068 pts, the median age was 66 years and 609 (57%) received 1L PT. In low RAD51 expressors, PT-treated pts (n=136) had significantly longer rwOS than PT-naïve (n=112): 11.4 months (95% CI, 7.8–13.8) vs 8 months (95% CI, 5.5–9.0); p=0.028. In contrast, in high RAD51 expressors, rwOS was similar between PT-treated (n=150) and PT-naïve (n=97): 7.9 vs 6.9 months (p=0.236). High GATA6 expression was also associated with improved rwOS in PT-treated pts vs PT-naïve: 13.5 vs. 9.3 months; p=0.011. In multivariate analysis, increased GATA6 expression remained a positive predictor of rwOS (HR 0.82; 95% CI, 0.76–0.88; p<0.001), and high RAD51 expression was associated with shorter rwOS with PT (HR 1.18; 95% CI, 1.01–1.39; p=0.043). Conclusions: In BRCA/PALB2 wt mPC, transcriptomic profiling identified low RAD51 and high GATA6 expression as predictors for improved rwOS when treated with PT. Integrating these biomarkers may improve development of DNA-damaging therapies beyond canonically defined HRD. Hazard ratios of mPC mortality by selected HRR genes estimated by multivariate CoxPh models. HR 95% CI p value 1L PT (vs PT-naïve) 0.87 0.75, 1.00 0.054 BRCA1 1.15 0.98, 1.35 0.079 BRCA2 0.89 0.77, 1.04 0.14 GATA6 0.82 0.76, 0.88 <0.001 NBN 0.97 0.81, 1.15 0.7 RAD51 1.18 1.01, 1.39 0.043 RECQL4 0.94 0.84, 1.04 0.2 Any RAS mutation 1.43 1.14, 1.81 0.002

599 Background: Biliary tract cancers (BTC) remain aggressive malignancies with limited treatment options. Alterations in PIK3CA are implicated in tumor progression across solid tumors, but their prognostic impact in BTC remains unclear. We investigated the association of PIK3CA alterations(alt), alone and in combination with TP53 and ARID1A , with survival outcomes in advanced BTC. Methods: We conducted a retrospective analysis of 1,079 patients with advanced BTC treated with first-line chemotherapy (CT) or chemoimmunotherapy (CIO) across three sites in the USA, Japan, and Brazil. Genomic profiling was performed using next-generation sequencing on tissue and/or liquid biopsy samples. Survival outcomes were estimated by Kaplan-Meier analysis, with comparisons by log-rank testing and hazard ratios (HR) derived from Cox regression models. Results: PIK3CA alterations were identified in 11% (123/1,079) of patients. The median age at diagnosis was 68 yrs (range, 20–79), with 50% (62/123) female. The most common primary site was intrahepatic in 43.1% (53/123) while majority of patients received CT only 92.5% (111/123). When analyzed with TP53 , 68 pts had PIK3CA/TP53 co-alterations(co-alt), 501 had TP53 alt, 55 had PIK3CA alt alone, and 455 had neither PIK3CA nor TP53 alt (WT/WT). Overall survival (OS) was significantly worse in patients with PIK3CA alt, particularly in the setting of TP53 co-alt. Median OS (mOS) was 22.6 months(m) (95% CI, 20.4–24.8) for PIK3CA/TP53 WT, 15.9 m (95% CI, 11.6–18.5) for PIK3CA alt /TP53 WT, 14.6 m (95% CI, 13.1–19.5) for PIK3CA alt/ TP53 alt, and 17.1 m (95% CI, 15.5–18.8) for TP53alt/PIK3CA WT (p<0.001). In analyses with ARID1A , 32 pts had PIK3CA/ARID1A co-alt, 91 had ARID1A alt only, 120 had PIK3CA alt only, and 836 were WT/WT. PIK3CA/ARID1A co-alt tumors had a mOS of 12.8 m (95% CI, 9.1–NE) compared with 19.3 m (95% CI, 18.1–20.9) in PIK3CA WT /ARID1A WT (p=0.009). Multivariate analysis identified PIK3CA alt (hazard ratio [HR] 1.37, 95%CI 1.09–1.72; p=0.008), PIK3CA alt/ TP53 alt (HR 1.52 95% CI, 1.11–2.07; p=0.008), ARID1A alt/ PIK3CA WT (HR 1.40, 95% CI 1.08-1.82, p=0.011), and metastatic status (HR 1.82, 95% CI 1.52-2.19, P<0.001) as independent prognostic factors for shorter OS. Conclusions: In advanced BTC, PIK3CA alt were associated with significantly worse overall survival, particularly when co-occurring with TP53 or ARID1A alt. These findings establish PIK3CA as a prognostic biomarker of poor outcome and highlight the need to explore PI3K-targeted therapeutic strategies in BTC.

593 Background: TP53 and CDKN2A are the most frequently mutated tumor suppressors in BTC, each linked to poor outcomes. Their co-alteration remains unexplored due to the current single-gene classification. Methods: This is an international multicenter study involving 35 institutions across 3 continents (North America, Asia, South America). Genomic profiling was performed on tissue or blood samples using NGS. Patients (pts) were categorized as wild-type (wt) TP53/ wt CDKN2A (WT/WT) (42%), mutated (mut) TP53 /wt CDKN2A (45%), wt TP53 /mut CDKN2A (5%), mut TP53 /mut CDKN2A (8%). Associations between TP53 / CDKN2A status , clinical and genomic features were evaluated using Chi-squared/Kruskal-Wallis tests adjusted for multiple comparisons. Overall survival (OS) from start of first-line systemic therapy was analyzed by Kaplan–Meier and multivariate Cox regression (covariates: age, sex, stage, primary site, first-line treatment, TP53/CDKN2A status), p or q<0.05. Results: We included 1079 pts with BTC. Median age 66 years with intrahepatic BTC (46%), gallbladder cancer (25%), extrahepatic BTC (22%), undefined (6%), ampullary cancer (< 1%). Most had de novo metastatic disease (54%), followed by recurrent (38%) and locally advanced disease (8%). Treatment regimens included chemotherapy (87%), chemo-immunotherapy (12%) and immunotherapy alone (<1%). The most frequently actionable genes were HER2 (8%), IDH1 (7%), BRAF (6%), FGFR2 fusions (4%), MSI-High (H) (3%). HER2 alterations were prevalent in mut TP53 /wt CDKN2A vs WT/WT (11 vs 4%, q = 0.002). IDH1 alterations were more frequent in WT/WT vs TP53 mut/wt CDKN2A (10 vs 4%, q < 0.001). FGFR2 fusions were more common in wt TP53 /mut CDKN2A vs WT/WT (17 vs 6%, q < 0.001), mut TP53 /mut CDKN2A (17 vs 3%, q = 0.001) and mut TP53 /wt CDKN2A (17 vs 1%, q < 0.001). MSI-H was enriched in mut TP53 /mut CDKN2A vs wt TP53 /mut CDKN2A (11 vs 2%, q = 0.035), mut TP53 /wt CDKN2A (11 vs 3%, q = 0.002) and WT/WT (11 vs 2%, q < 0.001). TMB-H was more prevalent in mut TP53 /mut CDKN2A vs mut TP53 /wt CDKN2A (16 vs 8%, q < 0.001) and WT/WT (16 vs 3%, q < 0.024). There was no significant difference in ORR, DCR and PFS among the four groups. mut TP53 /mut CDKN2A had the shortest median OS, while WT/WT had the longest (Table). Only 14% of mut TP53 /mut CDKN2A received regimens including immunotherapy. Conclusions: mut TP53 /mut CDKN2A represents a high-risk subset enriched with MSI-H and TMB-H pts, underscoring the need to prioritize the development of immunotherapy strategies. TP53 / CDKN2A status is prognostic, stratifying patients into distinct survival groups and supporting NGS use in the first-line setting. mOS (months) (95% CI) HR (95% CI) P-value WT/WT 21.4 (19.0-24.4) reference wt TP53 /mut CDKN2A 17.3 (12.4-25.9) 1.16 (0.79-1.71) 0.442 mut TP53 /wt CDKN2A 17.4 (16.0-19.2) 1.21 (1.01-1.45) 0.040 mut TP53 /mut CDKN2A 13.8 (11.5-18.1) 1.54 (1.13-2.09) <0.001

561 Background: For patients with aHCC, second-line treatment options are not well established. The primary aim of this study was to evaluate the efficacy of Futi, an FGFR 1-4 inhibitor, in combination with Pem as a second line therapy in patients with FGF19-expressing aHCC. Methods: Patients aged ≥ 18 years, with histological/cytological diagnosis or radiological diagnosis of HCC, Child-Pugh (CP) class A-B7, Barcelona Clinic Liver Cancer (BCLC) A, B or C, ECOG performance status of 0-2, and FGF19+, by in-situ hybridization or immunohistochemistry, received Futi 20 mg once daily plus i.v. Pem 200mg every 3 weeks. The primary endpoint was progression-free survival at 6 months (PFS6). Secondary endpoints include overall survival (OS). The study was designed to enroll 22 evaluable patients to detect PFS6 from 25% (historical control) to 47%, assuming a 1-sided alpha of 0.1 with 80% power. All patients provided written consent. Eligible patients with consent who started treatment are considered evaluable. Results: The study was closed to enrollment early at the request of funding agency due to poor accrual. Between 8/2021 and 3/2024, 14 patients enrolled with 1 patient withdrawing consent before beginning protocol treatment. Among 13 evaluable patients, median age was 72 (range: 60-87) years, 7 (53.8%) were male, 12 CP A (92.3%), 7 BCLC C (53.8%), with 7 (53.8%) having received prior liver directed therapy. Median PFS was 15 weeks (95% confidence interval [CI]: 9.0-Not Estimable) and PFS6 was 23.1% (95% CI: 8.6%, 62.3%). Objective response rate was 0% (95% CI: 0%, 24.71%) and stable disease occurred in 7 (53.9%) of patients. Median OS was 55.7 weeks (95% CI: 47.3-NE). Nine (69.2%) patients experienced at least 1 grade 3+ adverse event (AE), and 2 (15.4%) patients experienced at least a grade 4+ AE. One patient had a grade 5 AE unrelated to study treatment (acute kidney injury). Most commonly reported AEs (occurring in more than 31% of patients) were hyperphosphatemia (76.9%), fatigue (69.3%), diarrhea (61.6%), AST (53.9%) and ALT (46.2%), dry eye (38.5%), nail changes (38.5%), rash-maculo-papular (38.5%), and mucositis oral (38.5%). Conclusions: This prospective study evaluated the use of a targeted treatment in combination with immunotherapy following first line immune-based therapy. This combination had modest activity as a second line therapy for patients with aHCC but did not meet the primary endpoint. The safety profile was acceptable and consistent with prior reports. An optimal second-line regimen remains a critical need. Clinical trial information: NCT04828486 .

476 Background: Outcomes remain poor for patients with advanced/metastatic CCA treated with first-line gemcitabine-cisplatin +/- anti-PD(L)1. Lirafugratinib is the first highly selective, irreversible FGFR2 inhibitor designed to target oncogenic FGFR2 driver alterations and resistance mutations. ReFocus (NCT04526106) is an open-label, multicenter phase 1/2 study in patients with advanced/metastatic CCA and other solid tumors with FGFRok2 alterations. Methods: Pivotal cohort of patients (n=116) with advanced/metastatic CCA harboring FGFR2 f/r previously treated with ≥1 systemic therapy and FGFR i-naive were treated with oral lirafugratinib 70 mg once daily until disease progression or unacceptable toxicity. The primary endpoint was confirmed ORR per RECIST v1.1 by independent review committee (IRC). Key secondary endpoint was duration of response (DOR); other secondary endpoints: disease control rate (DCR), progression-free survival (PFS), overall survival (OS), safety, and quality of life per EORTC QLC-C30. Results: As of 27SEP2024, the primary efficacy analysis of IRC-assessed data (n=114) and the secondary analysis of investigator-assessed data (n=116) set was done. The primary efficacy analysis excluded 2 patients as not available. Median age 57 years with 61.4% female. IRC-assessed ORR was 47%, and median DOR was 11.8 months (mos) (95% CI, 7.5-13.0), where 76.2% of responses lasted >6 mos. Median PFS was 11.3 mos (95% CI, 9.2, 14.8), with 12-month rate of 49.2%. Median OS was 22.8 mos (95% CI, 17.3-27.2), with 12-month rate of 74.6%. DCR was 96.5%. Common grade ≥3 on-target, off-tumor treatment-related adverse events (TRAEs) included palmar-plantar erythrodysesthesia (32.8%) and stomatitis (12.1%). TRAEs led to dose reductions of 75.9%, dose interruption of 82.8%, and treatment discontinuation of 4.3%. QoL was maintained. Conclusions: Lirafugratinib demonstrated clinically meaningful anti-tumor activity (ORR, DOR, and PFS), manageable and tolerable safety in CCA FGFR2 f/r patients. Clinical trial information: NCT04526106 . Best Overall Response, n (%) IRC c Investigator d Complete Response Partial Response Stable Disease Progression Disease Not Evaluable 3 (2.6)50 (43.9)57 (50.0)3 (2.6)1 (0.9) 061 (52.6)51 (44.0)3 (2.6)1 (0.9) ORR a , n (%)95% CI 53 (46.5)37.1, 56.1 61 (52.6)43.1, 61.9 DCR b , n (%)95% CI 110 (96.5)91.3, 99.0 112 (96.6)91.4, 99.1 a ORR defined as proportion achieving confirmed response of complete response or partial response per RECIST v1.1. b DCR defined as proportion with confirmed response of complete response, partial response, or stable disease per RECIST v1.1. c Primary Efficacy Analysis Set. d Secondary Efficacy Analysis Set. Percentages based on number of patients in the analysis set for each treatment arm. Exact Clopper-Pearson used for 95% CI. IRC and investigator assessments are based on different populations and cannot compare directly.

103 Background: Clinicians spend significant time extracting and synthesizing information to risk-stratify and make adjuvant chemotherapy (AC) recommendations for CRC. By automating parts of this process, LLMs could help reduce documentation burden and variability in decision-making. Methods: We selected 144 stage III, pMMR CRC patients treated at Mayo Clinic (2017–2021). HIPAA-compliant Gemini-2.0-Flash-001 and GPT-4o extracted clinical variables (age, neuropathy grade, pathological stage) from EHRs and generated AC recommendations. Two approaches were compared: (1) a rule-based algorithm with nested if/else statements translating variables into NCCN-based recommendations, and (2) a dynamic LLM agent generating recommendations using a structured prompt informed by NCCN guidelines. Concordance with treating physicians was assessed, with disagreements reviewed by a blinded GI oncologist. Results: The dataset included 144 patients with median age 67 (IQR 66–68), 48.6% female. The LLM pipeline extracted neuropathy grade with 100% accuracy and pathological T/N stages with 99.3% accuracy. Concordance with physician recommendations was 93.1% for the rule-based algorithm and 62.5% for the LLM agent. In discordant cases, blinded adjudication by a GI oncologist favored the algorithm and LLM agent over physicians in most cases (Table 1). Conclusions: This exploratory work shows LLM agents can reliably extract variables and generate evidence-based treatment recommendations for stage III CRC. The high rate of expert preference for LLM recommendations in discordant cases is intriguing but limited by small sample size and possible adjudicator bias. Discordant physician recommendations reflect the complexity of decision-making, including patient preferences, performance status, comorbidities, and subtle findings not fully captured in EHRs. Rather than replacing judgment, we envision LLMs as decision support tools—analogous to an advanced trainee reviewing charts and presenting evidence-based recommendations to the attending physician with ultimate authority. Larger multi-institutional studies with multiple adjudicators are needed to validate these findings and clarify AI’s role in oncology. Patient characteristics, LLM extraction accuracy, and concordance with physician recommendations. Metric Result Total patients 144 Median age (IQR) 67 (66-68) Female (%) 48.6% LLM extraction accuracy – Neuropathy grade 100% LLM extraction accuracy – Pathological T and N stages 99.3% Primary Concordance – Rule-based algorithm vs. physician 93.1% Adjudicator preference of rule-based algorithm over treating physician for discordant cases 100% Concordance – LLM agent vs. physician 62.5% Adjudicator preference of LLM agent over treating physician for discordant cases 82.0%