The 2018 release of the 8th edition of the AJCC staging system for medullary thyroid cancer (MTC) aimed to standardize prognosis assessment. However, its effectiveness in predicting patient outcomes is still debated. Patient records were extracted from both the SEER database and multicenter cohorts. The study’s main endpoint was overall survival (OS). Prognostic performance of different staging models was evaluated using the concordance index (C-index). A total of 1450 MTC cases were included from SEER, along with 349 cases from the multicenter cohort. No significant OS difference was observed between T4a and T4b groups according to the AJCC criteria (P = .299). Therefore, the T4 group was redefined based on tumor size into T4a’ (≤3.5 cm) and T4b’ (>3.5 cm), which provided better discrimination of survival outcomes (P = .003). Additional analyses revealed that the T stage was closely linked to both lymph node (LN) location and total number (P < .001). Based on these findings, the N stage was revised to integrate LN location with LN count. Finally, the modified T and N categories were used to update the 8th AJCC system via recursive partitioning analysis, resulting in improved prognostic accuracy (C-index 0.811 vs. 0.792). Refining the 8th AJCC staging system by considering the relationship between T stage, LN location, and LN count enhances its prognostic performance, supporting better clinical decision-making and follow-up strategies.
