Abstract:Background: The components of the tumor invasive margin (IM) site in colorectal cancer (CRC), such as tumor infiltration pattern and immune cell infiltration, play a crucial role in controlling CRC invasion and metastasis, however, further exploration of their relevance and prognostic value is still needed.Methods: A total of 479 patients with stage I–III CRC from two centers were included in this retrospective study, divided into a discovery cohort (N = 347) and a validation cohort (N = 132). Types of tumor growth pattern (TGP) were assessed in hematoxylin and eosin (HE) stained whole-slide images (WSIs). A deep learning method was used to segment immunohistochemical (IHC) stained WSIs and fully automated to quantify the density of CD3+ T-cells at the tumor infiltration margin. Based on the TGP and the level of immune cell infiltration at the IM, we proposed the 'TGP-I score'. Survival analysis was performed using the Cox proportional hazard model. Harrell’s C-statistics was used to assess the performance of the model.Findings: Patients with a higher TGP-I score3 had a worse prognosis. In the discovery cohort, the 5-year survival rates for high and low TGP-I score3 were 61·7% and 86·8%, respectively (unadjusted hazard ratio [HR] for high vs. low 3·62, 95% confidence interval [CI] 2·22–5·90; P＜0·0001). The results were further confirmed in the validation cohort, with 5-year survival rates of 67·7% and 96·5% for high and low TGP-I score3, respectively (unadjusted HR for high vs. low 5·79, 95% CI 1·84–18·2; P = 0·003). After adjusting for other clinicopathological factors, the prognostic association of the TGP-I score3 with OS was maintained in multivariate analysis. The relative contribution of each parameter to predicting survival was analyzed, and the TGP-I score3 had similar importance compared to tumor-node-metastasis (TNM) staging (31·2% vs. 32·9%) and was stronger than other clinical parameters.Interpretation: This automated workflow and the proposed TGP-I score could further provide accurate prognostic stratification and have potential value for supporting the clinical decision-making of stage I–III CRC patients.Funding Information: This work was supported by Key-Area Research and Development Program of Guangdong Province (2021B0101420006), the National Science Fund for Distinguished Young Scholars (81925023), the National Key R&D Program of China (2021YFF1201003), Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application (2022B1212010011), NSFC Incubation Project of Guangdong Provincial People's Hospital (KY0120220037), and High-level Hospital Construction Project (DFJH201914 and DFJHBF202105).Declaration of Interests: The authors have declared that no competing interest exists.Ethics Approval Statement: The informed consent was waived for this retrospective analysis after it received approval from the Research Ethics Committees of both hospitals. (The Sixth Affiliated Hospital of Sun Yat-sen University (SYSU6), the Guangdong Provincial People's Hospital (GDPH)).

Prognostic Significance of Tumor-Infiltrating Lymphocytes Using Deep Learning on Pathology Images in Colorectal Cancers

The Prognostic Value of Tumor-Infiltrating Lymphocytes in Colorectal Cancer Differs by Anatomical Subsite: a Systematic Review and Meta-Analysis

Spatial analysis of tumor‐infiltrating lymphocytes in histological sections using deep learning techniques predicts survival in colorectal carcinoma

Exploring prognostic biomarkers in pathological images of colorectal cancer patients via deep learning

Diagnosis and Survival Prediction of Colorectal Cancer Using Deep Learning from the Histopathological HE Images

Deep-Learning-Based Characterization of Tumor-Infiltrating Lymphocytes in Breast Cancers From Histopathology Images and Multiomics Data

Deep Learning-Based Mapping of Tumor Infiltrating Lymphocytes in Whole Slide Images of 23 Types of Cancer

Assessing The Prognostic Significance Of Tumor-Infiltrating Lymphocytes In Patients With Melanoma Using Pathologic Features Identified By Natural Language Processing

Preoperative Prediction of Lymph Node Metastasis in Colorectal Cancer with Deep Learning

A systematic review of machine learning-based tumor-infiltrating lymphocytes analysis in colorectal cancer: Overview of techniques, performance metrics, and clinical outcomes

Abstract LB384: Artificial intelligence measures of tumor infiltrating lymphocytes predict colorectal cancer-specific and overall survival

Deep learning-based quantitative assessment of tumor-infiltrating lymphocytes from hematoxylin and eosin-stained slides in triple-negative breast cancer: A prognostic study.

Prediction of Tumor Mutation Load in Colorectal Cancer Histopathological Images Based on Deep Learning

Artificial intelligence-based analysis of tumor-infiltrating lymphocyte spatial distribution for colorectal cancer prognosis

Artificial Intelligence-Assisted Analysis of Tumor Growth Pattern, Tumor-Infiltrating Lymphocytes, and Survival from Colorectal Cancer

Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes for prediction of prognosis in resected colon cancer

Tumor-infiltrating lymphocyte: features and prognosis of lymphocytes infiltration on colorectal cancer

Tumor-Infiltrating Lymphocyte Recognition in Primary Melanoma by Deep Learning Convolutional Neural Network

Automated Analysis of Lymphocytic Infiltration, Tumor Budding, and Their Spatial Relationship Improves Prognostic Accuracy in Colorectal Cancer

Deep learning analysis of histopathological images predicts immunotherapy prognosis and reveals tumour microenvironment features in non-small cell lung cancer

Deep learning-based tumor microenvironment analysis in colon adenocarcinoma histopathological whole-slide images