Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning

Abstract Background Tumor mutation burden (TMB) and VHL mutation play a crucial role in the management of patients with clear cell renal cell carcinoma (ccRCC), such as guiding adjuvant chemotherapy and improving clinical outcomes. However, the time‐consuming and expensive high‐throughput sequencing...

Full description

Saved in:
Bibliographic Details
Main Authors: Qingyuan Zheng, Xinyu Wang, Rui Yang, Junjie Fan, Jingping Yuan, Xiuheng Liu, Lei Wang, Zhuoni Xiao, Zhiyuan Chen
Format: Article
Language:English
Published: Wiley 2024-08-01
Series:Cancer Medicine
Subjects:
Online Access:https://doi.org/10.1002/cam4.70112
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846156193313062912
author Qingyuan Zheng
Xinyu Wang
Rui Yang
Junjie Fan
Jingping Yuan
Xiuheng Liu
Lei Wang
Zhuoni Xiao
Zhiyuan Chen
author_facet Qingyuan Zheng
Xinyu Wang
Rui Yang
Junjie Fan
Jingping Yuan
Xiuheng Liu
Lei Wang
Zhuoni Xiao
Zhiyuan Chen
author_sort Qingyuan Zheng
collection DOAJ
description Abstract Background Tumor mutation burden (TMB) and VHL mutation play a crucial role in the management of patients with clear cell renal cell carcinoma (ccRCC), such as guiding adjuvant chemotherapy and improving clinical outcomes. However, the time‐consuming and expensive high‐throughput sequencing methods severely limit their clinical applicability. Predicting intratumoral heterogeneity poses significant challenges in biology and clinical settings. Our aimed to develop a self‐supervised attention‐based multiple instance learning (SSL‐ABMIL) model to predict TMB and VHL mutation status from hematoxylin and eosin‐stained histopathological images. Methods We obtained whole slide images (WSIs) and somatic mutation data of 350 ccRCC patients from The Cancer Genome Atlas for developing SSL‐ABMIL model. In parallel, 163 ccRCC patients from Clinical Proteomic Tumor Analysis Consortium cohort was used as independent external validation set. We systematically compared three different models (Wang‐ABMIL, Ciga‐ABMIL, and ImageNet‐MIL) for their ability to predict TMB and VHL alterations. Results We first identified two groups of populations with high‐ and low‐TMB (cut‐off point = 0.9). In two independent cohorts, the Wang‐ABMIL model achieved the highest performance with decent generalization performance (AUROC = 0.83 ± 0.02 and 0.8 ± 0.04 in predicting TMB and VHL, respectively). Attention heatmaps revealed that the Wang‐ABMIL model paid the highest attention to tumor regions in high‐TMB patients, while in VHL mutation prediction, non‐tumor regions were also assigned high attention, particularly the stromal regions infiltrated by lymphocytes. Conclusions Our results indicated that SSL‐ABMIL can effectively extract histological features for predicting TMB and VHL mutation, demonstrating promising results in linking tumor morphology and molecular biology.
format Article
id doaj-art-958a9f80ac0d44b48c3f379119d3a381
institution Kabale University
issn 2045-7634
language English
publishDate 2024-08-01
publisher Wiley
record_format Article
series Cancer Medicine
spelling doaj-art-958a9f80ac0d44b48c3f379119d3a3812024-11-26T10:12:30ZengWileyCancer Medicine2045-76342024-08-011316n/an/a10.1002/cam4.70112Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learningQingyuan Zheng0Xinyu Wang1Rui Yang2Junjie Fan3Jingping Yuan4Xiuheng Liu5Lei Wang6Zhuoni Xiao7Zhiyuan Chen8Department of Urology Renmin Hospital of Wuhan University Wuhan Hubei ChinaCentre for Reproductive Science Renmin Hospital of Wuhan University Wuhan Hubei ChinaDepartment of Urology Renmin Hospital of Wuhan University Wuhan Hubei ChinaUniversity of Chinese Academy of Sciences Beijing ChinaDepartment of Pathology Renmin Hospital of Wuhan University Wuhan Hubei ChinaDepartment of Urology Renmin Hospital of Wuhan University Wuhan Hubei ChinaDepartment of Urology Renmin Hospital of Wuhan University Wuhan Hubei ChinaCentre for Reproductive Science Renmin Hospital of Wuhan University Wuhan Hubei ChinaDepartment of Urology Renmin Hospital of Wuhan University Wuhan Hubei ChinaAbstract Background Tumor mutation burden (TMB) and VHL mutation play a crucial role in the management of patients with clear cell renal cell carcinoma (ccRCC), such as guiding adjuvant chemotherapy and improving clinical outcomes. However, the time‐consuming and expensive high‐throughput sequencing methods severely limit their clinical applicability. Predicting intratumoral heterogeneity poses significant challenges in biology and clinical settings. Our aimed to develop a self‐supervised attention‐based multiple instance learning (SSL‐ABMIL) model to predict TMB and VHL mutation status from hematoxylin and eosin‐stained histopathological images. Methods We obtained whole slide images (WSIs) and somatic mutation data of 350 ccRCC patients from The Cancer Genome Atlas for developing SSL‐ABMIL model. In parallel, 163 ccRCC patients from Clinical Proteomic Tumor Analysis Consortium cohort was used as independent external validation set. We systematically compared three different models (Wang‐ABMIL, Ciga‐ABMIL, and ImageNet‐MIL) for their ability to predict TMB and VHL alterations. Results We first identified two groups of populations with high‐ and low‐TMB (cut‐off point = 0.9). In two independent cohorts, the Wang‐ABMIL model achieved the highest performance with decent generalization performance (AUROC = 0.83 ± 0.02 and 0.8 ± 0.04 in predicting TMB and VHL, respectively). Attention heatmaps revealed that the Wang‐ABMIL model paid the highest attention to tumor regions in high‐TMB patients, while in VHL mutation prediction, non‐tumor regions were also assigned high attention, particularly the stromal regions infiltrated by lymphocytes. Conclusions Our results indicated that SSL‐ABMIL can effectively extract histological features for predicting TMB and VHL mutation, demonstrating promising results in linking tumor morphology and molecular biology.https://doi.org/10.1002/cam4.70112artificial intelligenceattention‐based multiple instance learningclear cell renal cell carcinomaself‐supervised learningtumor mutation burdenVHL mutation
spellingShingle Qingyuan Zheng
Xinyu Wang
Rui Yang
Junjie Fan
Jingping Yuan
Xiuheng Liu
Lei Wang
Zhuoni Xiao
Zhiyuan Chen
Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning
Cancer Medicine
artificial intelligence
attention‐based multiple instance learning
clear cell renal cell carcinoma
self‐supervised learning
tumor mutation burden
VHL mutation
title Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning
title_full Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning
title_fullStr Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning
title_full_unstemmed Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning
title_short Predicting tumor mutation burden and VHL mutation from renal cancer pathology slides with self‐supervised deep learning
title_sort predicting tumor mutation burden and vhl mutation from renal cancer pathology slides with self supervised deep learning
topic artificial intelligence
attention‐based multiple instance learning
clear cell renal cell carcinoma
self‐supervised learning
tumor mutation burden
VHL mutation
url https://doi.org/10.1002/cam4.70112
work_keys_str_mv AT qingyuanzheng predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT xinyuwang predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT ruiyang predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT junjiefan predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT jingpingyuan predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT xiuhengliu predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT leiwang predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT zhuonixiao predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning
AT zhiyuanchen predictingtumormutationburdenandvhlmutationfromrenalcancerpathologyslideswithselfsuperviseddeeplearning