Related Experiment Video
Updated: May 28, 2025
A Combined 3D Tissue Engineered In Vitro/In Silico Lung Tumor Model for Predicting Drug Effectiveness in Specific Mutational Backgrounds
Published on: April 6, 2016
A deep-learning model for predicting tyrosine kinase inhibitor response from histology in gastrointestinal stromal
Xue Kong1,2, Jun Shi3, Dongdong Sun4
1Department of Pathology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of PR China, Hefei, PR China.
Abstract:
Over 90% of gastrointestinal stromal tumors (GISTs) harbor mutations in KIT or PDGFRA that can predict response to tyrosine kinase inhibitor (TKI) therapies, as recommended by NCCN (National Comprehensive Cancer Network) guidelines. However, gene sequencing for mutation testing is expensive and time-consuming and is susceptible to a variety of preanalytical factors. To overcome the challenges associated with genetic screening by sequencing, in the current study we developed an artificial intelligence-based deep-learning （DL） model that uses convolutional neural networks (CNN) to analyze digitized hematoxylin and eosin staining in tumor histological sections to predict potential response to imatinib or avapritinib treatment in GIST patients. Assessment with an independent testing set showed that our DL model could predict imatinib sensitivity with an area under the curve (AUC) of 0.902 in case-wise analysis and 0.807 in slide-wise analysis. Case-level AUCs for predicting imatinib-dose-adjustment cases, avapritinib-sensitive cases, and wildtype GISTs were 0.920, 0.958, and 0.776, respectively, while slide-level AUCs for these respective groups were 0.714, 0.922, and 0.886, respectively. Our model showed comparable or better prediction of actual response to TKI than sequencing-based screening (accuracy 0.9286 versus 0.8929; DL model versus sequencing), while predictions of nonresponse to imatinib/avapritinib showed markedly higher accuracy than sequencing (0.7143 versus 0.4286). These results demonstrate the potential of a DL model to improve predictions of treatment response to TKI therapy from histology in GIST patients. © 2025 The Pathological Society of Great Britain and Ireland.
Insights
A new artificial intelligence deep-learning model analyzes histology slides to predict gastrointestinal stromal tumor (GIST) treatment response. This AI approach shows promise for predicting response to tyrosine kinase inhibitor (TKI) therapies, potentially improving patient outcomes.
Area of Science:
- Computational pathology
- Artificial intelligence in oncology
- Gastrointestinal stromal tumors (GIST) research
Background:
- Over 90% of GISTs have KIT or PDGFRA mutations, guiding tyrosine kinase inhibitor (TKI) therapy.
- Current genetic sequencing for mutation testing is costly, time-consuming, and prone to preanalytical errors.
- NCCN guidelines recommend mutation testing for TKI therapy selection in GIST patients.
Purpose of the Study:
- To develop and validate an artificial intelligence (AI)-based deep-learning (DL) model for predicting TKI treatment response in GIST.
- To analyze digitized hematoxylin and eosin (H&E) stained histological sections using convolutional neural networks (CNNs).
- To assess the model's ability to predict response to imatinib and avapritinib in GIST patients.
Main Methods:
- Development of a deep-learning (DL) model utilizing convolutional neural networks (CNNs).
- Analysis of digitized H&E stained histological sections of GIST tumor samples.
- Independent testing set validation for predicting imatinib and avapritinib response, including dose-adjustment and wildtype cases.
Main Results:
- The DL model achieved high predictive performance, with case-level AUCs of 0.902 for imatinib sensitivity and 0.958 for avapritinib sensitivity.
- Slide-level AUCs demonstrated strong predictive capabilities, reaching 0.922 for avapritinib-sensitive cases.
- The DL model showed comparable or superior prediction accuracy to sequencing-based screening for TKI response and significantly higher accuracy for predicting nonresponse.
Conclusions:
- The developed DL model effectively predicts TKI treatment response in GIST patients using histological image analysis.
- This AI-driven approach offers a potentially faster, more cost-effective alternative to genetic sequencing for treatment selection.
- Histology-based DL analysis holds significant potential for improving personalized treatment strategies in GIST management.
More Related Videos
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
07:21Molecular and Immunologic Techniques in a Genetically Engineered Mouse Model of Gastrointestinal Stromal Tumor
Published on: May 2, 2022