Currently submitted to: Journal of Medical Internet Research
Date Submitted: Mar 16, 2026
Open Peer Review Period: Mar 17, 2026 - May 12, 2026
(currently open for review)
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Large Language Model vs. Multidisciplinary Team: A Feasibility Study on Pancreatic Cancer Management Recommendations
ABSTRACT
Pancreatic cancer (PC) is a highly lethal malignancy requiring multidisciplinary team (MDT) management for optimal care, yet MDT is constrained by resource limitations. This single-center retrospective feasibility study enrolled 125 treatment-naive PC patients to assess concordance between ChatGPT-5.2-generated treatment recommendations and real MDT consensus decisions. Results demonstrated high alignment of large language model (LLM) suggestions with MDT conclusions: 80% for resectable, 100% for borderline resectable, 85.7% for locally advanced, and 100% for metastatic disease, with full concordance in biomarker-guided therapy for BRCA1/2-mutant and MSI-H/dMMR patients. Expert scoring showed mean 3.85 for concordance, 3.97 for rationality, and 3.21 for comprehensiveness, with moderate-to-near perfect inter-rater reliability (κw=0.70–0.83). The LLM’s main shortcoming was insufficient details in perioperative and surveillance management. In conclusion, ChatGPT-5.2 presents high feasibility as an auxiliary tool for PC MDTs, matching guideline-consistent and personalized decisions, though multimodal data integration and large-scale prospective validation are needed to improve comprehensiveness and clinical utility.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.