Related Experiment Video
Updated: Jun 17, 2026
"Sun's Seven-Step Technique" for Endoscopic En-Bloc Resection of Thyroid Cancer via the Chest-Breast Approach
Published on: November 28, 2025
Recursive partitioning analysis improves risk stratification in node-negative differentiated thyroid cancer: a
Xin Hua1, Hai-Ping You1, Yu-Ling Zhang2
1Department of Radiation Oncology, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou, China.
Background:
The applicability of the Tumor-Node-Metastasis (TNM) staging system in node-negative differentiated thyroid cancer (DTC) remains uncertain. While recursive partitioning analysis (RPA) offers a powerful alternative for staging refinement, its utility for prognostic prediction in node-negative DTC remains unexplored. This study aimed to develop and validate a novel RPA-based staging system for node-negative DTC.
Methods:
This retrospective cohort study included 8,816 eligible node-negative DTC patients (15-year follow-up) from the Surveillance, Epidemiology, and End Results (SEER) database, which were randomly assigned to training (n=6,172) and validation sets (n=2,644). Overall survival (OS) was the primary outcome. Independent prognostic factors were identified using Cox proportional hazards models and were used to develop an RPA-based staging system. Prognostic performance was compared with the American Joint Committee on Cancer (AJCC) TNM system using Kaplan-Meier analysis, time-dependent receiver operating characteristic (tROC) curves, and Harrell's concordance index (C-index).
Results:
In the training set, 15-year OS rates for AJCC TNM stages T1-T4 were 94.4%, 89.9%, 85.4%, and 54.1%. Multivariate analysis (MVA) identified older age, higher tumor grade, advanced T-stage, etc., as independent OS predictors (all P<0.05). RPA categorized patients into stages I-III with 15-year OS rates of 93.6%, 5.6%, and 0% (P<0.05), showing a significantly higher prognostic accuracy (Harrell's C-index: 0.947 vs. 0.750, P<0.001) and larger area under the tROC curves. This superior performance was confirmed in validation set.
Conclusions:
The developed and validated RPA-based model demonstrates superior prognostic performance it holds the potential to improve the staging of lymph node-negative DTC.
