Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer

Abstract Homologous recombination deficiency (HRD) has been recognized as a key biomarker for poly-ADP ribose polymerase inhibitors (PARPi) and platinum-based chemotherapy in breast cancer (BC). HRD prediction typically relies on molecular biology assays, which have a high turnaround time, and cost....

Full description

Saved in:
Bibliographic Details
Main Authors: Haijing Luan, Taiyuan Hu, Jifang Hu, Weier Liu, Kaixing Yang, Yue Pei, Ruilin Li, Jiayin He, Yajun Gao, Dawei Sun, Xiaohong Duan, Rui Yan, S. Kevin Zhou, Beifang Niu
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:npj Precision Oncology
Online Access:https://doi.org/10.1038/s41698-025-00950-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849688433021157376
author Haijing Luan
Taiyuan Hu
Jifang Hu
Weier Liu
Kaixing Yang
Yue Pei
Ruilin Li
Jiayin He
Yajun Gao
Dawei Sun
Xiaohong Duan
Rui Yan
S. Kevin Zhou
Beifang Niu
author_facet Haijing Luan
Taiyuan Hu
Jifang Hu
Weier Liu
Kaixing Yang
Yue Pei
Ruilin Li
Jiayin He
Yajun Gao
Dawei Sun
Xiaohong Duan
Rui Yan
S. Kevin Zhou
Beifang Niu
author_sort Haijing Luan
collection DOAJ
description Abstract Homologous recombination deficiency (HRD) has been recognized as a key biomarker for poly-ADP ribose polymerase inhibitors (PARPi) and platinum-based chemotherapy in breast cancer (BC). HRD prediction typically relies on molecular biology assays, which have a high turnaround time, and cost. In contrast, tissue sections stained with hematoxylin and eosin (H&E) are ubiquitously available. However, current HRD prediction methods that utilize pathological images are usually based on attention-based multiple instance learning, which is ineffective for modeling the global context of whole slide images (WSIs). To address this challenge, we propose a Sufficient and Representative Transformer (SuRe-Transformer) for WSI-based prediction of HRD. Experimental results demonstrate the superior performance of SuRe-Transformer in predicting HRD status compared to state-of-the-art methods, achieving an AUROC of 0.887 ± 0.034. Furthermore, SuRe-Transformer demonstrates generalizability across multiple external patient cohorts and achieves state-of-the-art performance in predicting several gene mutation biomarkers from BC WSIs.
format Article
id doaj-art-2b5a0ff6a97d458f8a2dcc5d79bf642e
institution DOAJ
issn 2397-768X
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series npj Precision Oncology
spelling doaj-art-2b5a0ff6a97d458f8a2dcc5d79bf642e2025-08-20T03:22:00ZengNature Portfolionpj Precision Oncology2397-768X2025-05-019111410.1038/s41698-025-00950-5Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative TransformerHaijing Luan0Taiyuan Hu1Jifang Hu2Weier Liu3Kaixing Yang4Yue Pei5Ruilin Li6Jiayin He7Yajun Gao8Dawei Sun9Xiaohong Duan10Rui Yan11S. Kevin Zhou12Beifang Niu13Computer Network Information Center, Chinese Academy of SciencesComputer Network Information Center, Chinese Academy of SciencesComputer Network Information Center, Chinese Academy of SciencesComputer Network Information Center, Chinese Academy of SciencesUniversity of Chinese Academy of SciencesComputer Network Information Center, Chinese Academy of SciencesComputer Network Information Center, Chinese Academy of SciencesComputer Network Information Center, Chinese Academy of SciencesBeijing ChosenMed Clinical Laboratory Co. Ltd.Beijing ChosenMed Clinical Laboratory Co. Ltd.School of Disaster and Emergency Medicine, Faculty of Medicine, Tianjin UniversitySchool of Biomedical Engineering, University of Science and Technology of ChinaSchool of Biomedical Engineering, University of Science and Technology of ChinaComputer Network Information Center, Chinese Academy of SciencesAbstract Homologous recombination deficiency (HRD) has been recognized as a key biomarker for poly-ADP ribose polymerase inhibitors (PARPi) and platinum-based chemotherapy in breast cancer (BC). HRD prediction typically relies on molecular biology assays, which have a high turnaround time, and cost. In contrast, tissue sections stained with hematoxylin and eosin (H&E) are ubiquitously available. However, current HRD prediction methods that utilize pathological images are usually based on attention-based multiple instance learning, which is ineffective for modeling the global context of whole slide images (WSIs). To address this challenge, we propose a Sufficient and Representative Transformer (SuRe-Transformer) for WSI-based prediction of HRD. Experimental results demonstrate the superior performance of SuRe-Transformer in predicting HRD status compared to state-of-the-art methods, achieving an AUROC of 0.887 ± 0.034. Furthermore, SuRe-Transformer demonstrates generalizability across multiple external patient cohorts and achieves state-of-the-art performance in predicting several gene mutation biomarkers from BC WSIs.https://doi.org/10.1038/s41698-025-00950-5
spellingShingle Haijing Luan
Taiyuan Hu
Jifang Hu
Weier Liu
Kaixing Yang
Yue Pei
Ruilin Li
Jiayin He
Yajun Gao
Dawei Sun
Xiaohong Duan
Rui Yan
S. Kevin Zhou
Beifang Niu
Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer
npj Precision Oncology
title Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer
title_full Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer
title_fullStr Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer
title_full_unstemmed Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer
title_short Breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative Transformer
title_sort breast cancer homologous recombination deficiency prediction from pathological images with a sufficient and representative transformer
url https://doi.org/10.1038/s41698-025-00950-5
work_keys_str_mv AT haijingluan breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT taiyuanhu breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT jifanghu breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT weierliu breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT kaixingyang breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT yuepei breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT ruilinli breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT jiayinhe breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT yajungao breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT daweisun breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT xiaohongduan breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT ruiyan breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT skevinzhou breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer
AT beifangniu breastcancerhomologousrecombinationdeficiencypredictionfrompathologicalimageswithasufficientandrepresentativetransformer