Towards Precision Therapy in Hepatocellular Carcinoma: A Clinical-Reasoning LLM for Risk Stratification and Treatment Guidance

2026-07-09

AI Summary

Current hepatocellular carcinoma (HCC) staging systems are overly coarse-grained, failing to capture patient heterogeneity and the nuanced clinical context embedded in electronic medical records (EMRs), thereby limiting precision oncology. To address this, this work proposes HCC-STAR, a novel framework that deeply integrates clinical guideline knowledge into large language model reasoning to establish a knowledge-aligned clinical decision-support system. HCC-STAR jointly outputs risk stratification, evidence-based treatment recommendations, and individualized survival predictions. The model leverages prompt-augmented generation to synthesize 30,000 EMR-style training samples and is validated on 6,668 multicenter real-world patients. Results demonstrate that HCC-STAR significantly outperforms established guidelines (BCLC and CNLC) as well as state-of-the-art models including GPT-5 and Gemini-2.5 Pro, with simulated adherence to its recommendations yielding a median survival of 51 months. The model garners high trust from hepatobiliary specialists and effectively enhances diagnostic accuracy and clinical workflow efficiency.

Abstract

Hepatocellular carcinoma (HCC) is a common malignancy and a leading cause of cancer-related mortality. Current guidelines and staging systems provide coarse categories, but often miss within-stage heterogeneity and the clinical context in electronic medical records (EMRs). We present HCC-STAR (Hepatocellular Carcinoma Staging, Treatment And pRognosis), a clinically aligned large language model that reads routine EMR narratives and jointly outputs risk score-based staging, ranked guideline-consistent treatments with evidence-based rationales, and individualized survival estimates. We curated about 30,000 HCC cases from SEER and expanded them into EMR-style narrative training data using a clinician-validated, prompt-based augmentation workflow. On this corpus, we developed a knowledge-aligned reasoning framework optimized with a step-verifiable composite reward, moving beyond text-level memorization of clinical guidelines. In a multi-center cohort of 6,668 patients from 12 hospitals in China, HCC-STAR achieved state-of-the-art performance in treatment recommendation and risk stratification compared with clinical guidelines and competitive models, including GPT-5 and Gemini-2.5 Pro. Hypothetical overall-survival analysis showed a median survival of 51 months under adherence to HCC-STAR recommendations, compared with 29 and 32 months under BCLC and CNLC. In clinician-centric evaluations, blinded hepatobiliary specialists rated HCC-STAR's reasoning and evidence-based justifications as trustworthy. The model surpassed resident and attending physicians in treatment accuracy and helped physicians make more accurate decisions faster when used as an assistant. These findings support HCC-STAR as a reliable and verifiable decision-support system for risk stratification and precision therapy in HCC.

Problem

• hepatocellular carcinoma
• risk stratification
• precision therapy
• clinical guidelines
• electronic medical records

Innovation

• clinical reasoning LLM
• risk stratification
• treatment recommendation
• electronic medical records
• precision therapy