2. Academic Validation
  3. Intelligent identification and targeted intervention of GRP75-caused drug resistant hepatocellular carcinoma, a study based on radiomics, machine learning, and molecular pharmacology

Intelligent identification and targeted intervention of GRP75-caused drug resistant hepatocellular carcinoma, a study based on radiomics, machine learning, and molecular pharmacology

  • Int J Surg. 2026 Feb 19. doi: 10.1097/JS9.0000000000004913.
Zhendong Wang 1 2,  Jianguo Zhu 3,  Ziyi Wang 1 2,  Qinliang Mo 4 5,  Xinru Du 1 2,  Yue Ma 1 2,  Hongquan Zhang 4 5,  Yancheng Gao 1 2,  Sha Liu 1 2,  Chengwu Tang 4 5,  Yan Wang 6,  Chunxiao Zhou 7,  Yuan Li 1 2 7
Affiliations

Affiliations

  • 1 School of Public Health, Key Laboratory of Public Health Safety and Emergency Prevention and Control Technology of Higher Education Institutions in Jiangsu Province, Nanjing Medical University, Nanjing, China.
  • 2 The Key Laboratory of Modern Toxicology, Ministry of Education, School of Public Health, Nanjing Medical University, Nanjing, China.
  • 3 Department of Radiology, The Second Affiliated Hospital of Nanjing Medical University, Nanjing, China.
  • 4 Department of General Surgery, The First Affiliated Hospital of Huzhou University, Huzhou, China.
  • 5 Zhejiang Key Laboratory of Digital Technology in Medical Diagnostics, The First affiliated Hospital of Huzhou University, Huzhou, Zhejiang Province, China.
  • 6 Department of Pathology, The Second Affiliated Hospital of Nanjing Medical University, Nanjing, China.
  • 7 Department of Gastroenterology, The Affiliated Suzhou Hospital of Nanjing Medical University, Suzhou, Jiangsu Province, China.
PMID: 41711200 DOI: 10.1097/JS9.0000000000004913
Abstract

Background: Tumor heterogeneity mediated drug resistance was a core clinical challenge to improve the prognosis of patients with advanced hepatocellular carcinoma (HCC). Glucose-regulated protein 75 (GRP75) was a key molecular target driving this heterogeneity. Therefore, accurate identification of HCC patients with GRP75-related poor prognosis and targeted intervention was a key issue to be solved.

Materials and methods: This study integrated radiomics and artificial intelligence (AI) algorithms to construct a preoperative imaging-based risk stratification model, aiming to accurately identify patients with prognosis associated with GRP75. The traditional Chinese medicine (TCM) databases were combined with molecular docking and dynamics simulations techniques to screen for GRP75-targeting TCM monomers. The underlying mechanisms of action were further explored, and functional validation was performed using in vitro and in vivo models.

Results: Based on preoperative imaging, we successfully developed a system that could accurately identify HCC patients with poor prognosis associated with high expression of GRP75. Furthermore, the TCM monomer baicalin was identified and validated as a GRP75-targeting agent. Mechanistic studies revealed that baicalin effectively reversed drug resistance in HCC by specifically targeting GRP75 and disrupting the mitochondria-associated endoplasmic reticulum membranes, thereby modulating the calcium/Autophagy regulatory axis. This process restored the sensitivity of drug-resistant cells to cisplatin and sorafenib.

Conclusions: This study established an AI-based radiomics system for identifying HCC patients with poor prognosis associated with high expression of GRP75, and revealed a novel therapeutic approach utilizing the TCM monomer baicalin to reverse multi-drug resistance through GRP75 targeting, further clarifying its mechanism.

Keywords

artificial intelligence; early identification and personalized intervention; hepatocellular carcinoma; heterogeneity of drug resistance; molecular pharmacology.

Figures
Products