Related Experiment Video
Updated: May 22, 2026

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
Prognostic Value and Therapeutic Implications of MRI-Based Negative Lymph Nodes in Nasopharyngeal Carcinoma
Ze-Qi Lin1, Chu-Yu He1, Shui-Qing He1
1Department of Radiation Oncology, Sun Yat-Sen University Cancer Center, State Key Laboratory of Oncology in South China, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Guangdong Provincial Clinical Research Center for Cancer, Guangzhou, Guangdong, P. R. China.
Background:
This study aimed to evaluate the prognostic significance of negative lymph nodes (LNnegs) identified by magnetic resonance imaging (MRI) and treatment implications in patients with N0 locoregionally advanced NPC.
Methods:
This retrospective cohort study included 753 patients with T3-4N0M0 NPC receiving concurrent chemoradiotherapy (CCRT) with or without induction chemotherapy (IC) at two independent hospitals. Pre-treatment and post-radiotherapy MRI scans were performed. Recursive partitioning analysis (RPA) stratified patients by radiological LNneg features. Cox regression and Kaplan-Meier analysis were performed with overall survival (OS) as the primary endpoint.
Results:
Neck LNneg levels and the short-axial diameter (SD) of the largest cervical LNneg were identified as significant factors associated with OS through Cox analysis (all p ≤ 0.009). Incorporating these two factors, RPA categorized patients into RPA-I (n = 177; ≥ 4 LNneg levels and SD decreased by > 40%) and RPA-II (n = 576; < 4 LNneg levels, or ≥ 4 levels and SD decreased by ≤ 40%). The 5-year OS in RPA-I was significantly better than in RPA-II (98.3% vs. 89.5%; hazard ratio [HR]: 0.16; 95% confidence interval [CI]: 0.05-0.50; p < 0.001), validated after propensity score matching (n = 177 vs. 177; HR: 0.20; 95% CI: 0.06-0.70; p = 0.005). The OS advantage of RPA-I was significantly influenced by the cumulative cisplatin dose during CCRT (≥ 200 mg/m2, p < 0.001; < 200 mg/m2, p = 0.660), but was independent of IC (all p ≥ 0.040).
Conclusions:
Dispersed LNnegs with a considerable decrease in the SD of the largest cervical LNneg in T3-4N0M0 NPC indicated a low OS risk and potential benefit from a cumulative cisplatin dose of ≥ 200 mg/m2 in CCRT.