What question did this study set out to answer?

The study aims to explore the use of large language models to extract prognostic information from unstructured clinical notes for personalized cancer risk assessment.

July 10, 2026PLOS Digital HealthOpen Access

Large language models enable prognostic stratification of cancer patients using real-world clinical notes

View Full Paper

Ask AI

Bookmark

Authors

NKNiklas Kiermeyer TLTim Lenfers ADAmin Dada

Discussion

Member takes

Overview

Randomized trial demonstrates improved survival prediction in cancer patients using language models, suggesting enhanced clinical decision-making.

Key Points

The study aims to explore the use of large language models to extract prognostic information from unstructured clinical notes for personalized cancer risk assessment.
Collected clinical notes from 2,708 non-small cell lung cancer and 814 colon cancer patients.
Utilized self-hosted large language models to extract prognostic indicators without prior training.
Integrated extracted features into machine learning models and compared prediction accuracy to TNM staging.
Overall survival prediction improved with LLM integration (C-Index: NSCLC 0.72 vs 0.64, colon cancer 0.70 vs 0.59).
Reclassification of patients occurred for 61.4% of NSCLC and 68.3% of colon cancer patients.
LLM-derived factors significantly modified the prognostic impact of TNM staging.

Cite This Study

Kiermeyer et al. (2026) studied this question.

synapsesocial.com/papers/6a508c766eeac72a437a08eb https://doi.org/10.1371/journal.pdig.0001546

View Full Paper

Ask AI

Bookmark

July 10, 2026PLOS Digital HealthOpen Access

Large language models enable prognostic stratification of cancer patients using real-world clinical notes

View Full Paper

Ask AI

Bookmark

Authors

NKNiklas Kiermeyer TLTim Lenfers ADAmin Dada

Discussion

Member takes

Overview

Randomized trial demonstrates improved survival prediction in cancer patients using language models, suggesting enhanced clinical decision-making.

Key Points

The study aims to explore the use of large language models to extract prognostic information from unstructured clinical notes for personalized cancer risk assessment.
Collected clinical notes from 2,708 non-small cell lung cancer and 814 colon cancer patients.
Utilized self-hosted large language models to extract prognostic indicators without prior training.
Integrated extracted features into machine learning models and compared prediction accuracy to TNM staging.
Overall survival prediction improved with LLM integration (C-Index: NSCLC 0.72 vs 0.64, colon cancer 0.70 vs 0.59).
Reclassification of patients occurred for 61.4% of NSCLC and 68.3% of colon cancer patients.
LLM-derived factors significantly modified the prognostic impact of TNM staging.

Cite This Study

Kiermeyer et al. (2026) studied this question.

synapsesocial.com/papers/6a508c766eeac72a437a08eb https://doi.org/10.1371/journal.pdig.0001546

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark