引用文本：张世纪, 肖君. 结直肠癌临床结局预后评估方法研究进展[J/CD]. 消化肿瘤杂志（电子版）, 2026, 18(2): 283-292.

作者：张世纪，肖君

单位：南京中医药大学附属医院（江苏省中医院）消化内镜中心，江苏 南京210023

Authors：Zhang Shiji, Xiao Jun

Unit：Endoscopy Center, the Affiliated Hospital of Nanjing University of Chinese Medicine, Jiangsu Province Hospital of Chinese Medicine, Nanjing 210023, Jiangsu, China

摘要：

结直肠癌（colorectal cancer, CRC）是全球高发恶性肿瘤。其预后评估主要围绕病理评估、生物标志物和数学预测模型三大领域展开。病理评估以TNM分期和组织学特征为核心，但存在一定的主观性和异质性局限。生物标志物如微卫星不稳定性、循环肿瘤DNA和炎症指标等，为动态监测和个体化预后评估提供了新途径。数学模型尤其是机器学习方法，通过整合多源数据显著提升了预测精度。未来研究应注重多组学整合、液体活检技术的标准化及机器学习模型的综合性，以构建更精准、动态的CRC预后评估体系。

关键词：结直肠癌；预后评估；生物标志物；数学模型；机器学习

Abstract：

Colorectal cancer (CRC) is a highly prevalent malignant tumor worldwide. Prognostic assessment for CRC primarily focuses on three major areas: pathological evaluation, biological markers, and mathematical prediction models. Pathological assessment, centered on TNM staging and histological features, remains limited by subjectivity and heterogeneity. Biological markers such as microsatellite instability (MSI), circulating tumor DNA (ctDNA), and inflammatory indicators offer new avenues for dynamic monitoring and individualized prognostic assessment. Mathematical models, particularly machine learning methods, significantly enhance predictive accuracy by integrating multi-source data. Future research should emphasize multi-omics integration, standardization of liquid biopsy techniques, and comprehensiveness of machine learning models to establish a more precise and dynamic prognostic assessment system for CRC.

Key words：Colorectal cancer; Prognostic assessment; Biomarkers; Mathematical models; Machine learning