Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
May 16, 2026Journal of Pathology InformaticsOpen Access

SCRIPT: Stratified clinical risk prediction from pathology reports using large language models

View Full Paper
Ask AI
Bookmark
Share

Authors

CLChiara M.L. LoefflerNRNic G. ReitsamFWFabian Wolf

Discussion

Loading...

Member takes

Overview

Randomized trial evaluates LLMs extracting prognostic scores from pathology reports, suggesting improved risk stratification in oncology.

Key Points

  • To assess if large language models can extract prognostic information from pathology reports for risk stratification in oncology.
  • Used open-weight LLaMA 3.3 70B model to generate risk scores from pathology reports.
  • Evaluated associations between LLM-generated scores and survival outcomes in gastrointestinal cancers.
  • Conducted multivariate analysis to confirm LLM-generated risk score as an independent prognostic factor.
  • In colorectal cancer, LLM-generated risk scores showed significant prognostic value for overall survival (HR = 2.77, 95% CI = 1.92–3.97, p < 0.001).
  • Progression-free survival was significantly predicted by LLM-generated scores (HR = 2.93, 95% CI = 2.11–4.08, p < 0.001).
  • Disease-specific survival demonstrated strong prognostic value with LLM scores (HR = 5.85, 95% CI = 3.66–9.36, p < 0.001).

Cite This Study

Loeffler et al. (2026) studied this question.

synapsesocial.com/papers/6a080a11a487c87a6a40bdfehttps://doi.org/10.1016/j.jpi.2026.100673
View Full Paper
Ask AI
Bookmark
Share

Also Consider

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

  1. 1Large language models enable prognostic stratification of cancer patients using real-world clinical notes2026
  2. 2Large Language Models Improve Cancer Survival Prediction Using Real-World Clinical Notes2025
  3. 3Prognostic data extraction harnessing a privacy-preserving large language model: a clinician-AI collaborative retrospective evaluation in head and neck oncology2026
  4. 4Toward Automating the Summarization of Cancer Pathology Reports Using Large Language Models to Improve Clinical Usability2026 · 2 citations
  5. 5Validation of large language models for detecting pathologic complete response in breast cancer using population-based pathology reports2024 · 22 citations