Tumor Etiology and Screening Department of Cancer Institute and General Surgery, The First Hospital of China Medical University, Shenyang, China; Key Laboratory of Cancer Etiology and Prevention, Liaoning Education Department, The First Hospital of China Medical University, Shenyang, China; Key Laboratory of GI Cancer Etiology and Prevention, Liaoning Province, The First Hospital of China Medical University, Shenyang, China.
Mathematical Computer Teaching and Research Office, Liaoning Vocational College of Medicine, Shenyang, China.
Am J Pathol. 2023 Oct;193(10):1517-1527. doi: 10.1016/j.ajpath.2023.06.004. Epub 2023 Jun 24.
Determining the molecular characteristics of cancer patients is crucial for optimal immunotherapy decisions. The aim of this study was to screen immunotherapy beneficiaries by predicting key molecular features from hematoxylin and eosin-stained images based on deep learning models. An independent data set from Asian gastric cancer patients was included for external validation. In addition, a segmentation model (Horizontal-Vertical Network) was used to quantify the cellular composition of tumor stroma. The model performance was evaluated by measuring the area under the curve (AUC). The tumor extraction model achieved an AUC of 0.9386 and 0.9062 in the internal and external test sets, respectively. The stratification model could predict the immunotherapy-sensitive subtypes (AUC range, 0.8685 to 0.9461), the genetic mutations (AUC range, 0.8283 to 0.9225), and the pathway activity (AUC range, 0.7568 to 0.8612) fairly accurately. In external validation, the prediction performance of Epstein-Barr virus and programmed cell death ligand 1 expression status achieved AUCs of 0.7906 and 0.6384, respectively. The segmentation model identified a relatively high proportion of inflammatory cells and connective cells in some immunotherapy-sensitive subtypes. The deep learning-based models potentially may serve as a valuable tool to screen for the beneficiaries of immunotherapy in gastric cancer patients.
确定癌症患者的分子特征对于制定最佳免疫治疗决策至关重要。本研究旨在通过基于深度学习模型从苏木精和伊红染色图像预测关键分子特征来筛选免疫治疗受益人群。一项来自亚洲胃癌患者的独立数据集用于外部验证。此外，还使用分割模型（水平-垂直网络）来量化肿瘤基质的细胞组成。通过测量曲线下面积（AUC）来评估模型性能。肿瘤提取模型在内部和外部测试集中的 AUC 分别为 0.9386 和 0.9062。分层模型可以相当准确地预测免疫治疗敏感亚型（AUC 范围为 0.8685 至 0.9461）、遗传突变（AUC 范围为 0.8283 至 0.9225）和通路活性（AUC 范围为 0.7568 至 0.8612）。在外部验证中，EB 病毒和程序性细胞死亡配体 1 表达状态的预测性能的 AUC 分别为 0.7906 和 0.6384。分割模型在一些免疫治疗敏感亚型中识别出了相对较高比例的炎症细胞和结缔组织细胞。基于深度学习的模型可能成为筛选胃癌患者免疫治疗受益人群的有价值工具。
