Integrating lymphovascular and perineural invasion into TNM staging: a novel ITNM system for enhanced prognostic stratification in colorectal cancer

Read full text
Authors: Jiaqi Hu, Minghao Zhang, Yang Qi, Y. Wang, Dalin Xu, Kejin Zhu, Jun Bu
Cited by: 3
References: 27

Abstract

Background The standard TNM staging system for colorectal cancer (CRC) fails to reflect true tumor biology because it ignores key histopathologic features such as lymphovascular and perineural invasion. We therefore developed and validated an “ITNM” classification that folds these two factors into the N stage, yielding more accurate prognoses and greater clinical value. Methods A retrospective cohort of 442 stage I–III CRC patients underwent radical resection (2015–2021). Propensity score matching (PSM) created balanced exposure (LVI/PNI-positive, n = 185) and control (LVI/PNI-negative, n = 257) groups. The ITNM system was constructed by upstaging the N category based on LVI/PNI status. Predictive performance was evaluated using C-index, ROC-AUC, calibration curves, decision curve analysis (DCA), net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Results Multivariate analysis confirmed LVI (HR = 2.10, p = 0.018) and PNI (HR = 2.28, p = 0.025) as independent predictors of overall survival (OS). The LVI/PNI-positive group had significantly lower 5-year OS (69.5% vs . 88.2%, p < 0.001). The ITNM system demonstrated superior discriminative ability for 5-year OS (C-index = 0.715; AUC = 0.735, 95% CI: 0.691–0.776), excellent calibration (p = 0.489), and higher net benefit on DCA. Significant reclassification improvement was confirmed (NRI = 0.306, p < 0.001; IDI = 0.061, p < 0.001). Conclusion The ITNM system significantly enhances prognostic accuracy by integrating LVI and PNI into TNM staging, enabling risk-adapted therapeutic decision-making and representing a paradigm shift in CRC stratification.

Fields of Study & Subjects

Colorectal Cancer Surgical Treatments Radiomics and Machine Learning in Medical Imaging AI in cancer detection

Publication details

Journal / source
Frontiers in Oncology
Publisher
Frontiers Media SA
Volume
16
Pages
1745324–1745324
Publication date
28 January 2026
Language
EN