EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Large language model processing capabilities of ChatGPT 4.0 to generate molecular tumor board recommendations—a critical evaluation on real world data

Maximilian Schmutz iD, Sebastian Sommer, Julia Sander, David Graumann, Johannes Raffler iD, Iñaki Soto-Rey, Seyedmostafa Sheikhalishahi, Lisa Schmidt, Leonhard Paul Unkelbach, Levent Ortak, Tina Schaller, Sebastian Dintner und 6 weitere

The Oncologist · 01.10.2025 · 30 (10)

DOI: 10.1093/oncolo/oyaf293 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Oncologist, online erschienen am 18.09.2025. Verfasst von Maximilian Schmutz und 17 weiteren Autor:innen. Beteiligte Einrichtung: Department of Hematology and Oncology, Faculty of Medicine, University of Augsburg, Augsburg 86156, Germany. Hauptthemen laut MEDLINE: Neoplasms, Precision Medicine. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Large language models (LLMs) like ChatGPT 4.0 hold promise for enhancing clinical decision-making in precision oncology, particularly within molecular tumor boards (MTBs). This study assesses ChatGPT 4.0’s performance in generating therapy recommendations for complex real-world cancer cases compared to expert human MTB (hMTB) teams.

Methods

We retrospectively analyzed 20 anonymized MTB cases from the Comprehensive Cancer Center Augsburg (CCCA), covering breast cancer (n = 3), glioblastoma (n = 3), colorectal cancer (n = 2), and rare tumors. ChatGPT 4.0 recommendations were evaluated against hMTB outputs using metrics including recommendation type (therapeutic/diagnostic), information density (IDM), consistency, quality (level of evidence [LoE]), and efficiency. Each case was prompted thrice to evaluate variability (Fleiss’ Kappa).

Results

ChatGPT 4.0 generated more therapeutic recommendations per case than hMTB (median 3 vs 1, P = .005), with comparable diagnostic suggestions (median 1 vs 2, P = .501). Therapeutic scope from ChatGPT 4.0 included off-label and clinical trial options. IDM scores indicated similar content depth between ChatGPT 4.0 (median 0.67) and hMTB (median 0.75; P = .084). Moderate consistency was observed across replicate runs (median Fleiss’ Kappa = 0.51). ChatGPT 4.0 occasionally utilized lower-level or preclinical evidence more frequently (P = .0019). Efficiency favored ChatGPT 4.0 significantly (median 15.2 vs 34.7 minutes; P < .001).

Conclusion

Incorporating ChatGPT 4.0 into MTB workflows enhances efficiency and provides relevant recommendations, especially in guideline-supported cases. However, variability in evidence prioritization highlights the need for ongoing human oversight. A hybrid approach, integrating human expertise with LLM support, may optimize precision oncology decision-making.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

NeoplasmsPrecision MedicineHumansRetrospective StudiesFemaleClinical Decision-MakingLarge Language ModelsGenerative Artificial Intelligence

Stichwörter der Autor:innen

ChatGPT 4.0artificial intelligencelarge language modelsmolecular tumor boardprecision oncologyvariant annotation

Autor:innen und Institutionen

  • Department of Hematology and Oncology, Faculty of Medicine, University of Augsburg, Augsburg 86156, Germany.
  • Department of Data Management and Clinical Decision Support, Faculty of Medicine, University of Augsburg, Augsburg 86356, Germany.
  • Bavarian Cancer Research Center (BZKF), Erlangen 91054, Germany.
  • Faculty of Medicine, University of Augsburg, Augsburg 86195, Germany.
  • General Pathology and Molecular Diagnostics, Faculty of Medicine, University of Augsburg, Augsburg 86156, Germany.
  • Pediatrics and Adolescent Medicine, Faculty of Medicine, University of Augsburg, Augsburg 86156, Germany.

ORCID-Profile: Maximilian Schmutz, Johannes Raffler, Michaela Kuhlen, Rainer Claus

Zitieren

APA

Schmutz, M., Sommer, S., Sander, J., Graumann, D., Raffler, J., Soto-Rey, I., Sheikhalishahi, S., Schmidt, L., Unkelbach, L. P., Ortak, L., Schaller, T., Dintner, S., Hildebrand, K., Kuhlen, M., Jordan, F., Trepel, M., Hinske, C., & Claus, R. (2025). Large language model processing capabilities of ChatGPT 4.0 to generate molecular tumor board recommendations—a critical evaluation on real world data. The Oncologist, 30(10). https://doi.org/10.1093/oncolo/oyaf293

Vancouver

Schmutz M, Sommer S, Sander J, Graumann D, Raffler J, Soto-Rey I, et al. Large language model processing capabilities of ChatGPT 4.0 to generate molecular tumor board recommendations—a critical evaluation on real world data. The Oncologist. 2025;30(10). doi:10.1093/oncolo/oyaf293

BibTeX
@article{Schmutz2025,
  author = {Schmutz, Maximilian and Sommer, Sebastian and Sander, Julia and Graumann, David and Raffler, Johannes and Soto-Rey, Iñaki and Sheikhalishahi, Seyedmostafa and Schmidt, Lisa and Unkelbach, Leonhard Paul and Ortak, Levent and Schaller, Tina and Dintner, Sebastian and Hildebrand, Kathrin and Kuhlen, Michaela and Jordan, Frank and Trepel, Martin and Hinske, Christian and Claus, Rainer},
  title = {Large language model processing capabilities of ChatGPT 4.0 to generate molecular tumor board recommendations—a critical evaluation on real world data},
  journal = {The Oncologist},
  year = {2025},
  volume = {30},
  number = {10},
  doi = {10.1093/oncolo/oyaf293}
}
Kostenlos starten →