Deep learning with whole slide images can improve the prognostic risk stratification with stage III colorectal cancer

• Published in: Computer methods and programs in biomedicine Vol. 221; p. 106914
• Main Authors: Sun, Caixia, Li, Bingbing, Wei, Genxia, Qiu, Weihao, Li, Danyi, Li, Xiangzhao, Liu, Xiangyu, Wei, Wei, Wang, Shuo, Liu, Zhenyu, Tian, Jie, Liang, Li
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.06.2022
• Subjects: Chemotherapy duration; Colorectal cancer; Deep learning; Prognosis; Whole slide images; Deep learning; Whole slide images; Prognosis; Chemotherapy duration; Colorectal cancer
• ISSN: 0169-2607; 1872-7565
• DOI: 10.1016/j.cmpb.2022.106914

•An innovative whole slide image signature is developed to facilitate the prognostic risk stratification for stage III colorectal cancer patients.•An unsupervised multiple instance learning model is proposed to mine the prognostic information from pathological images.•The feasibility of this signature was initially verified in this study, using a retrospectively collected dataset of 210 stage III colorectal cancer patients. Adjuvant chemotherapy is recommended as standard treatment for colorectal cancer (CRC) with stage III according to TNM stage. However, outcomes are varied even among patients receiving similar treatments. We aimed to develop a prognostic signature to stratify outcomes and benefit from different chemotherapy regimens by analyzing whole slide images (WSI) using deep learning. We proposed an unsupervised deep learning network (variational autoencoder and generative adversarial network) in 180,819 image tiles from the training set (147 patients) to develop a WSI signature for predicting the disease-free survival (DFS) and overall survival (OS) of patients, and tested in validation set of 63 patients. An integrated nomogram was constructed to investigate the incremental value of deep learning signature (DLS) to TNM stage for individualized outcomes prediction. The DLS was associated with DFS and OS in both training and validation sets and proved to be an independent prognostic factor. Integrating the DLS and clinicopathologic factors showed better performance (C-index: DFS, 0.748; OS, 0.794; in the validation set) than TNM stage. In patients whose DLS and clinical risk levels were inconsistent, their risk of relapse was reclassified. In the subgroup of patients treated with 3 months, high-DLS was associated with worse DFS (hazard ratio: 3.622–7.728). The proposed based-WSI DLS improved risk stratification and could help identify patients with stage III CRC who may benefit from the prolonged duration of chemotherapy.