Abstract Background We externally validated the proposed 9th edition of the TNM staging classification with our institution’s prospectively collected data and compared it to the 8th edition for overall survival (OS) and freedom from recurrence (FFR). Methods A retrospective analysis was conducted of 4029 cases of stage I-III non-small cell lung cancer that underwent surgical treatment from January 2004 to December 2020. Survival was compared using Kaplan-Meier curves and multivariable Cox regression analysis. The concordance index (C-index), Alkaike information criterion (AIC), and R2 were used to assess the discriminatory ability. Results In the 9th edition, the N2 category (n = 352) was subdivided into N2a (n = 256, 72.7%) and N2b (n = 96, 27.3%). The TNM stage changes were as follows: (1) IIB to IIA, 151 cases (26.0%); (2) IIIA to IIB, 52 cases (11.5%); (3) IIIB to IIIA, 57 cases (61.3%); (4) IIIA to IIIB, 56 cases (12.4%). The survival curves of the proposed 9th edition demonstrated similar patterns to those of the 8th edition, but with a greater discriminative ability for OS and FFR. Subdividing N2 into N2a and N2b refined prognosis prediction. The C-index, AIC, and R2 demonstrated improved values in the proposed 9th edition compared to the 8th edition. Conclusions The proposed 9th edition of the TNM staging classification for lung cancer showed favorable prognostic validity and better discrimination ability than the 8th edition.
