Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou, 510060, PR China; Department of Colorectal Surgery, Sun Yat-sen University Cancer Center, Guangzhou, PR China.
Bio-totem Pte Ltd, Foshan, PR China; Department of General Surgery, Nanfang Hospital, Southern Medical University, Guangzhou, PR China.
EBioMedicine. 2022 Jul;81:104120. doi: 10.1016/j.ebiom.2022.104120. Epub 2022 Jun 23.
We aimed to develop a deep learning (DL) model to predict DNA mismatch repair (MMR) status in colorectal cancers (CRC) based on hematoxylin and eosin-stained whole-slide images (WSIs) and assess its clinical applicability.
The DL model was developed and validated through three-fold cross validation using 441 WSIs from the Cancer Genome Atlas (TCGA) and externally validated using 78 WSIs from the Pathology AI Platform (PAIP), and 355 WSIs from surgical specimens and 341 WSIs from biopsy specimens of the Sun Yet-sun University Cancer Center (SYSUCC). Domain adaption and multiple instance learning (MIL) techniques were adopted for model development. The performance of the models was evaluated using the area under the receiver operating characteristic curve (AUROC). A dual-threshold strategy was also built from the surgical cohorts and validated in the biopsy cohort. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), F1-score, and the percentage of patients avoiding IHC testing were evaluated.
The MIL model achieved an AUROC of 0·8888±0·0357 in the TCGA-validation cohort, 0·8806±0·0232 in the PAIP cohort, 0·8457±0·0233 in the SYSUCC-surgical cohort, and 0·7679±0·0342 in the SYSUCC-biopsy cohort. A dual-threshold triage strategy was used to rule-in and rule-out dMMR patients with remaining uncertain patients recommended for further IHC testing, which kept sensitivity higher than 90% and specificity higher than 95% on deficient MMR patient triage from both the surgical and biopsy specimens, result in more than half of patients avoiding IHC based MMR testing.
A DL-based method that could directly predict CRC MMR status from WSIs was successfully developed, and a dual-threshold triage strategy was established to minimize the number of patients for further IHC testing.
The study was funded by the National Natural Science Foundation of China (82073159, 81871971 and 81700576), the Natural Science Foundation of Guangdong Province (No. 2021A1515011792 and No.2022A1515012403) and Medical Scientific Research Foundation of Guangdong Province of China (No. A2020392).
本研究旨在开发一种基于苏木精和伊红染色全切片图像（WSI）的深度学习（DL）模型，用于预测结直肠癌（CRC）的错配修复（MMR）状态，并评估其临床适用性。
采用三折交叉验证方法，利用癌症基因组图谱（TCGA）中的 441 张 WSI 数据和病理人工智能平台（PAIP）中的 78 张 WSI 数据对 DL 模型进行开发和验证，并利用中山大学肿瘤防治中心（SYSUCC）的 355 张手术标本 WSI 和 341 张活检标本 WSI 进行外部验证。采用域自适应和多实例学习（MIL）技术进行模型开发。采用受试者工作特征曲线（ROC）下面积（AUROC）评估模型性能。还从手术队列中构建了双阈值策略，并在活检队列中进行了验证。评估了敏感性、特异性、阳性预测值（PPV）、阴性预测值（NPV）、F1 评分和避免免疫组织化学（IHC）检测的患者比例。
MIL 模型在 TCGA 验证队列中的 AUROC 为 0.8888±0.0357，在 PAIP 队列中的 AUROC 为 0.8806±0.0232，在 SYSUCC 手术队列中的 AUROC 为 0.8457±0.0233，在 SYSUCC 活检队列中的 AUROC 为 0.7679±0.0342。采用双阈值分诊策略，对可疑的 dMMR 患者进行分类，将其分为确诊和排除，对仍存在不确定性的患者建议进一步进行 IHC 检测，该策略在手术和活检标本中均能保持较高的敏感性（均大于 90%）和特异性（均大于 95%），使超过一半的患者避免了基于 IHC 的 MMR 检测。
成功开发了一种可直接从 WSI 预测 CRC MMR 状态的基于 DL 的方法，并建立了双阈值分诊策略，以减少进一步进行 IHC 检测的患者数量。
本研究得到了国家自然科学基金（82073159、81871971 和 81700576）、广东省自然科学基金（No. 2021A1515011792 和 No.2022A1515012403）和广东省医学科学技术研究基金（No. A2020392）的资助。
