Related Experiment Videos
Reliability and Validity of Imaging-derived Extranodal Extension Grading Using CT and MRI in Nasopharyngeal Carcinoma
Pae Sun Suh1, Jinna Kim1, Dong Eon Kim2
1Department of Radiology, Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea.
Abstract:
Background Imaging-derived extranodal extension (iENE) was added to the N3 criteria for nasopharyngeal carcinoma (NPC) in the ninth version of the tumor-node-metastasis (TNM) staging system. Purpose To evaluate the reliability and prognostic value of a proposed joint task force iENE grading system in NPC. Materials and Methods This retrospective single-center study included patients with NPC treated between January 2007 and March 2025. For patients with cervical lymph node metastases, three radiologists independently graded iENE at pretreatment CT and/or MRI using the proposed four-tier classification system. Inter- and intrarater reliability was assessed using the intraclass correlation coefficient (ICC) and weighted κ. Overall survival (OS), disease-free survival, local-regional recurrence-free survival, and distant metastasis-free survival (DMFS) were evaluated across iENE grades using Kaplan-Meier and multivariable Cox regression analyses adjusted for age, sex, T and M categories, and chemotherapy status. Results A total of 401 patients (median age, 52 years [IQR, 44-60 years]; 296 male patients) were included. According to the consensus-based final iENE grades, 223, 36, 94, and 48 patients were classified as having iENE grade 0, 1, 2, and 3, respectively. Interrater reliability (ICC, 0.87 [95% CI: 0.85, 0.89]) and intrarater reliability (κ = 0.95) for iENE grading were excellent. Compared with patients with iENE grade 0, patients with iENE grade 3 had shorter OS (hazard ratio, 3.30 [95% CI: 1.70, 6.41]; P < .001). In addition, patients with iENE grade 1, 2, and 3 had shorter DMFS than patients with iENE grade 0, with hazard ratios of 2.33 (95% CI: 1.12, 4.87; P = .02), 2.11 (95% CI: 1.24, 3.60; P = .006), and 2.39 (95% CI: 1.28, 4.45; P = .006), respectively. There was no evidence of a difference in prognostic performance between the TNM-8 and TNM-9 systems. Conclusion The joint task force iENE grading system had excellent reliability and good prognostic value in patients with NPC. © RSNA, 2026 Supplemental material is available for this article.