Pelster, Meredith S.

TPS3184 Background: Replication stress (RS) creates cancer cell vulnerability to which the critical DNA damage response kinase, ATR (Ataxia-Telangiectasia and Rad3-related) responds. ATR senses the stress, halts the cell cycle, and promotes DNA repair to allow proliferation. Alnodesertib is an ATR inhibitor developed to prevent cancer cells from successfully repairing RS and proliferating further. In earlier stages of alnodesertib development, preclinical and translational research demonstrated that tumor cells experience three combined biological insults: irinotecan-induced RS, endogenous RS via ataxia-telangiectasia mutated (ATM) loss, and RS rescue prevention by ATR inhibition. Data from the dose escalation and initial dose expansion cohorts of this trial in a tumor agnostic population showed encouraging results in ATM-negative cancers (confirmed Overall Response Rate of 50% in 20 patients with ATM negative cancers and responses across 8 cancer types including complete responses, Ulahannan et al, AACR 2025). Tumor-specific dose expansion cohorts are now enrolling, including colorectal cancer (CRC) and pancreas ductal adenocarcinoma (PDAC). Methods: Two tumor-specific cohorts are now open, each enrolling approximately 50 patients (pts): ATM-negative CRC and ATM-negative PDAC. Pts will receive the recommended dose of alnodesertib 200mg orally on days 1-3 and 8-10 combined with low-dose irinotecan 60mg/m 2 IV on days 1 and 8, of each 21-day cycle. Tumor tissue is prescreened by immunohistochemistry for ATM protein loss (H score 0) for enrollment eligibility. Pts with CRC must have received a maximum of 2 prior chemotherapies excluding prior trifluridine/tipiracil, fruquintinib, or regorafenib, and pts with PDAC must have received a maximum of 1 prior chemotherapy for advanced disease. Prior targeted agents are allowed and do not count as prior lines of therapy. Key objectives include safety, tolerability, and preliminary efficacy. Clinical trial information: NCT04657068 .

e15509 Background: Large language models (LLMs) are increasingly explored for oncology decision support, yet comparative evidence evaluating medical-specific versus general-purpose LLMs using expert-reviewed clinical cases is limited. We compared a medical-specific LLM (Google MedGemma 27B) with a general-purpose LLM (ChatGPT-5) using gastrointestinal (GI) and genitourinary (GU) oncology vignettes. Methods: This cross-sectional study evaluated 32 de-identified oncology vignettes (26 GI, 6 GU) created by a U.S. board-certified medical oncologist and informed by real-world cases. Vignettes included text-only information on the clinical context, description of pathology, tumor staging, biomarkers, and relevant comorbidities. Both LLMs were prompted to provide National Comprehensive Cancer Network(NCCN) guideline-based treatment recommendations, monitoring strategies, maintenance therapy considerations, and references, along with a self-reported confidence score (0–100). Board-certified U.S. medical oncologists independently reviewed outputs using a structured instrument; two disease specific expert oncologists evaluated each case. Outcomes included correctness and comprehensiveness (1–10 scales), hallucinations (factually fabricated or unsupported statements), NCCN traceability, potential patient harm, and clinician trust. Associations between accuracy and comprehensiveness were assessed using correlation coefficients. Inter-rater reliability was assessed using a two-way random-effects intraclass correlation coefficient (ICC). Results: ChatGPT-5 demonstrated higher mean correctness (8.65 vs 4.68) and comprehensiveness (8.68 vs 5.22) scores compared with MedGemma 27B. Accuracy and comprehensiveness were strongly correlated overall (Pearson r = 0.93, p < 0.001), with significant model-specific correlations (ChatGPT-5 r = 0.86, p < 0.001; MedGemma r = 0.77, p < 0.001). Hallucinations occurred less frequently with ChatGPT-5 (12.3% vs 65%), and NCCN-traceable recommendations were more common (75.4% vs 36.7%). Responses judged as potentially harmful occurred in 18.5% of ChatGPT-5 outputs versus 70% of MedGemma outputs. Clinician trust favored ChatGPT-5 (84.6% vs 15%). Inter-rater reliability improved with rater averaging (ICC(A,k): correctness 0.72; comprehensiveness 0.64). Conclusions: In expert-reviewed GI and GU oncology cases, ChatGPT-5 demonstrated higher clinical accuracy, comprehensiveness, and guideline alignment than a medical-specific LLM; however, a substantial rate of hallucinations and potentially harmful recommendations persisted across models. These findings underscore that, despite performance differences, current LLMs require rigorous clinician oversight and should not be used as standalone decision-support tools in oncology.

4013 Background: Pancreatic ductal adenocarcinoma (PDAC) is driven predominantly by oncogenic KRAS signaling, yet prior attempts to target the MAPK pathway have been limited by toxicity and emergence of resistance. Atebimetinib is a next-generation, deep cyclic inhibitor (DCI) of MEK designed to achieve pulsatile pathway inhibition with improved tolerability. We report a Phase 2a cohort evaluating atebimetinib in combination with modified gemcitabine/nab-paclitaxel (mGnP) in the first-line treatment of advanced or metastatic PDAC (NCT05585320). Methods: In this multicenter, open-label, nonrandomized Phase 2a study, patients with previously untreated advanced or metastatic PDAC were eligible for participation and were treated with orally administered atebimetinib daily in combination with mGnP (modification: every other week dosing). The primary endpoint was objective response rate (ORR) per investigator assessment using RECIST 1.1 criteria in response-evaluable patients. The cohort was powered (~80%, one sided alpha 0.05) using an optimal Simon’s 2-stage design. Results: Fifty-five patients were enrolled and treated with 320 mg atebimetinib daily + mGnP. ECOG performance status was 0-1 (100%); median age was 68 years (range 42-85 years), 62% were ≥65 years, and 55% were male. No grade 5 treatment-related adverse events (AE) were observed. No AE related to atebimetinib were higher than grade 3. Grade 3 AE related to atebimetinib occurred in 29% of participants, with the most frequent being rash (5%), ALT increased (5%), and AST increased (5%). Among the 50 response-evaluable patients, best overall response (BOR) achieved were 21 PR, 21 SD, 7 PD, 1 NE. The ORR was 42%, and DCR was 84%. With a median follow-up of 10.4 months (data cutoff Jan 7, 2026), 6- and 9-month OS rates were 88% and 79%, respectively; median PFS was 8.3 months and median OS was not reached. Conclusions: Atebimetinib in combination with mGnP demonstrated favorable safety and promising efficacy compared to historic standard-of-care (SoC) GnP. 1 A phase 3 registrational trial is planned to start in mid-2026 for evaluation of atebimetinib + mGnP versus SoC GnP. 1. Von Hoff DD, Ervin T, Arena FP, et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med . 2013;369(18):1691-1703. doi:10.1056/NEJMoa1304369 (PMID: 24131140). Clinical trial information: NCT05585320 .

ABSTRACT Background A randomized phase II screening trial of gemcitabine, nab‐paclitaxel, and cisplatin with a medically supervised ketogenic diet (MSKD) versus usual diet (non‐MSKD) was conducted in patients with treatment‐naive metastatic pancreatic ductal adenocarcinoma (PDAC). Methods Patients with untreated metastatic PDAC were randomized 1:1 to MSKD or non‐MSKD while receiving gemcitabine, nab‐paclitaxel, and cisplatin on days 1 and 8 of a 21‐day cycle. The MSKD was guided by a remote health care team and daily ketone (beta‐hydroxybutyrate) levels, with goal beta‐hydroxybutyrate of 0.5 to 3.0 mM. The primary endpoint was progression‐free survival (PFS) using a one‐sided alpha level of 0.20. Secondary endpoints included overall survival (OS), safety, and quality of life (QOL). Changes in microbiome were an exploratory endpoint. Findings Overall, there were 32 evaluable patients. In the MSKD arm, 15 of 16 patients achieved nutritional ketosis; the median proportion of days in ketosis was 39.4%. The median PFS was 8.5 months in MSKD patients and 6.2 months in non‐MSKD patients: hazard ratio, 0.53 (95% CI, 0.21–1.37); one‐sided p = .096. The median OS was 13.7 months with MSKD and 10.2 months in the non‐MSKD arm: hazard ratio, 0.58 (95% CI, 0.25–1.37); one‐sided p = .107). All MSKD‐related adverse events were grade 1‐2. There were no significant differences in grade ≥3 chemotherapy‐related adverse events between the arms. MSKD patients had no decline in QOL and had significant enrichment of beneficial taxa in the microbiome ( p < .05, log‐fold change ≥2). Conclusions The MSKD is feasible in patients with PDAC and, although not powered for definitive outcomes, shows trends in improved PFS and OS when combined with gemcitabine, nab‐paclitaxel, and cisplatin, without added toxicity or detriment to QOL. Larger studies are required to confirm these findings and establish the value of the MSKD in pancreatic cancer treatment.

608 Background: Tinengotinib, a spectrum-selective multi-kinase inhibitor with unique binding properties to FGFR , potently inhibited FGFR2 fusion/rearrangement and acquired resistant/gatekeeper mutations in pre-clinical models and exhibited antitumor activity in cholangiocarcinoma (CCA) patients (pts) in phase 1/2 trials. Here we present the overall survival (OS) data and biomarker correlative analysis from a phase 2 clinical trial of tinengotinib in CCA. Methods: Eligible pts with advanced/metastatic CCA who had received ≥ 1 prior systemic chemotherapy with ECOG PS 0-1 were treated with tinengotinib 10 mg QD across four Cohorts: A1: FGFR2 fusion(s) with primary progression on previous FGFR inhibitor ( FGFR i); A2: FGFR2 fusion(s) with progression after prior response to FGFR i; B: FGFR alteration(s) without FGFR2 fusion(s); C: FGFR wild-type. Efficacy was evaluated per RECIST v1.1 and safety was assessed by CTCAE v5.0. Genomic alterations in circulating tumor DNA (ctDNA) were assessed by Foundation Medicine NGS panel at baseline, C3D1 and EOT. Planned correlative analyses of response (PR+SD>6 months) and PFS in pt subgroups based on ctDNA mutational status were performed. Results: 55 eligible pts were enrolled (18 in A1, 11 in A2, 13 in B, 13 in C) with median age 61.0 [range 24-81] years, ECOG PS 0 in 50.9% pts, 56.4% female, and 60.0% with ≥ 3 lines of prior therapy. Among 42 pts with FGFR alterations, 78.4% had ≥ 1 prior FGFR i, and 97.6% had prior chemotherapy. Median follow up time was 8.6 months (0.4-30.5). The median OS (months) was 17.1 (95%CI 7.5-19.5) in A1; not reached (95%CI 9.6, -) in A2; 18.0 (95%CI 9.6, -) in A1+A2; not reached (95%CI 8.0, -) in B; 6.5 (95%CI 4.8-16.4) in C. The OS rate in A, B and C was 100%, 91.7% and 75% at month 3; 83.7%, 83.3% and 66.7% at month 6, and 65.8%, 55.6% and 23.8% at month 12. Partial response was observed in 26% (10/39) of evaluable FGFR-altered pts (A1+A2+B). The safety profile was consistent with previous reports. The most common any-grade treatment-related AEs were hypertension (63.6%) and diarrhea (47.3%). Baseline ctDNA results were available in 46 pts, 35 in cohorts A+B and 11 in C. A positive correlation was shown in MED12 mutation vs response (P=0.03476). BCOR and ARID1A mutation were found to be negatively associated with PFS (P=0.01366 and P=0.0437). Acquired/secondary mutations at baseline were observed in cohort A, including V564F/I (3/24 and 5/24) and N549K/H (5/24 and 5/24); favorable anti-tumor activity was noted despite these resistant mutations, consistent with preclinical data. Conclusions: Tinengotinib has shown promising anti-tumor efficacy in CCA pts with FGFR fusion after prior FGFR i and in those with primary FGFR mutations. An ongoing randomized phase III study will investigate tinengotinib vs. chemotherapy in FGFR inhibitor refractory CCA (NCT05948475). Clinical trial information: NCT04919642 .

Abstract Background: Pancreatic ductal adenocarcinoma (PDAC) is characterized by stromal fibrosis, hypoxia, and nutritional deprivation. Nutritional metabolic interventions may fundamentally change the tumor microenvironment and improve outcomes. A ketogenic diet (KD) (lower carbohydrate, moderate protein, higher fat) can significantly reduce glucose and insulin and stimulate ketogenesis. Triplet chemotherapy with gemcitabine/cisplatin/nab-paclitaxel has been associated with a median OS 16.4 months, PFS 10.1 months in patients with metastatic PDAC in the first-line setting. Recently, KD plus triplet chemotherapy was shown to inhibit murine pancreatic KPC tumor growth and prolong animal survival over chemotherapy alone (Yang et al, Med, 2022). Tumor growth inhibition was associated with glucose depletion, altered TCA substrate use and NADH elevation. Methods: In this Phase II trial, patients with treatment-naive metastatic PDAC were randomized to triplet chemotherapy with a medically supervised ketogenic diet (MSKD) or continuing their usual diet (non-MSKD). The MSKD is supported by Virta Health, who offers tracking of daily ketone (beta-hydroxybutyrate, BHB) and glucose levels, a web-based application, education, and communication with a remote care team to ensure sustained nutritional ketosis (BHB 0.5–3.0mM). Patients with BMI &lt; 18, type 1 DM or history of DKA were excluded. Primary endpoint was PFS. Secondary endpoints: DCR (PR+CR+SD ≥ 9 weeks), change in CA 19-9 (CA125, CEA if not CA 19-9 expressers), average fasting insulin levels, HbA1c, body weight, gut microbiome, serum metabolites, QOL (QLQ- C30). Results: A total of 36 patients with untreated metastatic PDAC were enrolled. Among 32 evaluable patients (median age 65.9 years; 53% male), 16 were randomized to each arm. Both groups had similar baseline weight, HgbA1c, insulin and BHB levels (all p&gt;0.05). Nine patients had type 2 DM; 6 required insulin. In the MSKD group, 15/16 achieved nutritional ketosis at any point during the study, with mean BHB of 0.57 mM (95% CI 0.40–0.73) and median proportion of days in ketosis of 39.4% (range 0-95.8%) by daily ketone tracking. Lab data showed significantly higher mean ± SD maximum change in BHB in the MSKD (0.5 ± 0.9 mmol/L) versus non-MSKD (–0.2 ± 0.3 mol/L) group (p = 0.018). There were no significant differences between groups in maximum change in weight, HbA1C, or insulin (all p &gt; 0.05). The most common MSKD-related adverse events were Gr 1-2 fatigue (n=4), constipation (n=3), weight loss (n=3), and none stopped MSKD due to adverse events.Conclusion: A medically supervised ketogenic diet is feasible in patients with metastatic PDAC receiving triplet chemotherapy, with median proportion of days in ketosis of 39.4% and an acceptable safety profile. The MSKD is associated with comparable changes in insulin and HgbA1c levels compared to the non-MSKD diet, and sustained ketosis does not result in significantly greater weight loss. MSKD warrants further prospective exploration and analyses of the primary endpoint and correlative studies are ongoing. Citation Format: Gayle S Jameson, Erkut Borazanci, Diana L Hanna, Caroline GP Roberts, Meredith S Pelster, Richard C Frank, Angela T Alistar, Julia E Wiedmeier-Nutor, Sandra D Algaze, Brandon Fell, Sarah J Hallberg, Denise J Roe, Betsy C Wertheim, Derek Cridebring, Joshua D Rabinowitz, Stephen Gately, Daniel D Von Hoff, Drew W Rasco. Randomized phase II trial of two different nutritional approaches for patients receiving treatment for their advanced pancreatic cancer: initial results of safety and feasibility of the medically supervised ketogenic diet [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research; 2024 Sep 15-18; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl_2):Abstract nr B012.

Abstract Adagrasib, an irreversible, selective KRASG12C inhibitor, may be an effective treatment in KRASG12C-mutated colorectal cancer, particularly when combined with an anti-EGFR antibody. In this analysis of the KRYSTAL-1 trial, patients with previously treated KRASG12C-mutated unresectable or metastatic colorectal cancer received adagrasib (600 mg twice daily) plus cetuximab. The primary endpoint was objective response rate (ORR) by blinded independent central review. Ninety-four patients received adagrasib plus cetuximab. With a median follow-up of 11.9 months, ORR was 34.0%, disease control rate was 85.1%, and median duration of response was 5.8 months (95% confidence interval [CI], 4.2–7.6). Median progression-free survival was 6.9 months (95% CI, 5.7–7.4) and median overall survival was 15.9 months (95% CI, 11.8–18.8). Treatment-related adverse events (TRAEs) occurred in all patients; grade 3–4 in 27.7% and no grade 5. No TRAEs led to adagrasib discontinuation. Exploratory analyses suggest circulating tumor DNA may identify features of response and acquired resistance. Significance: Adagrasib plus cetuximab demonstrates promising clinical activity and tolerable safety in heavily pretreated patients with unresectable or metastatic KRASG12C-mutated colorectal cancer. These data support a potential new standard of care and highlight the significance of testing and identification of KRASG12C mutations in patients with colorectal cancer. This article is featured in Selected Articles from This Issue, p. 897 See co-corresponding author Rona Yaeger discuss this research article, published simultaneously at the AACR Annual Meeting 2024: https://vimeo.com/932606282/f27a6e46f4

434 Background: Tinengotinib is a spectrum-selective multi-kinase inhibitor with unique binding properties to FGFR, potently inhibited FGFR2 fusion/rearrangement and acquired resistant mutations in pre-clinical models and in phase I trials that included cholangiocarcinoma (CCA) patients (pts). Here we present the efficacy and safety of tinengotinib in a phase II clinical trial. Methods: Eligible pts with advanced/metastatic CCA who had received ≥ 1 prior systemic chemotherapy therapy and ECOG PS 0 or 1 were treated with tinengotinib 10 mg QD. Four cohorts included: Cohort A1: FGFR2 fusion(s) with primary progression on previous FGFR inhibitor ( FGFRi), A2: FGFR2 fusion(s) with progression after prior response to FGFRi (acquired resistance); B: non-fusion FGFR alteration(s): C: FGFR wild-type ( FGFRwt). Primary endpoint was objective response rate (ORR) per RECIST v1.1. CTCAE V5.0 was used for safety assessments. Results: As of 7Aug2023, 48 pts with CCA were enrolled, 13 in Cohort A1, 10 in A2, 12 in B, 13 in C. Median age 61.5 [range 25-81] years old, 41.7% male, 58.3% had ≥ 3 lines of prior therapy. ECOG PS 0 in 47.9% pts. Among 35 pts with FGFR alterations, 80.0% had ≥ 1 prior FGFRi therapy, and 97.1% had prior chemotherapy. Forty (40) pts were efficacy evaluable. In A1, 1 out of 11 pts (9.1%) achieved PR with tumor reduction of 31.8%. In A2, 3 out of 8 pts (37.5%) achieved PR with tumor reduction of 40.7%, 47.0% and 54.6%. In B, 3 out of 9 pts (33.3%) achieved PR with tumor reduction of 36.5%, 48.6%, and 60.6%. No PR was observed in C. Overall DCR was 94.7% (18/19) in FGFR2 fusion/rearrangement pts (A1+A2), 88.9% (8/9) in other FGFR alterations pts (B), and 75% (9/12) in FGFRwt pts (C). Median progression-free survival (mPFS) was 5.26 months (95%CI, 2.86-9.10) in A1+A2, 5.98 months (95%CI, 1.87-NA) in B and 3.84 months (95% CI, 1.84-4.80) in C. Among 48 treated pts, treatment-related AEs (TRAEs) occurred in 45 (93.8%) pts, 14 (29.2%) in G1-2, 29 (60.4%) in G3 and 2 (4.2%) in G4. The most common G3 TRAEs were hypertension in 12 (25%), palmar-plantar erythrodysesthesia syndrome in 3 (6.3%), diarrhea in 3 (6.3%) and stomatitis in 3 (6.3%). One subject had G4 increased TSH and G4 increased lipase, and another subject had G4 posterior reversible encephalopathy syndrome. No G5 TRAE was observed. Conclusions: Tinengotinib has promising clinical benefit for FGFR2 fusion CCA after prior FGFRi and for non-fusion FGFR alterations. Tinengotinib-related toxicities were manageable. An ongoing randomized, controlled phase III study will evaluate the clinical efficacy, safety, and pharmacodynamic effect of Tinengotinib vs Physicians’ choice in subjects with FGFR2-altered refractory/relapsed CCA after prior chemotherapy and FGFRi therapy. Clinical trial information: NCT04919642 .