Can stage-IIB lung cancer be divided into subgroups in terms of prognosis? A modelling study.

Abstract

Introduction: Whether changes should be made to the TNM classification of non-small cell lung cancer (NSCLC) according to the newly proposed nodal classification is unclear. We aim to compare the survival between stage-IIB subsets using a modelling study performed using the newly proposed nodal classification.

Patients and methods: A total of 682 patients with stage-IIB NSCLC based on the 8th TNM classification were analysed. Hazard ratio (HR) values calculated from survival comparisons between stage-IIB subgroups were used to create a model for patients with stage-IIB NSCLC, and modelling was performed according to the HR values that were close to each other.

Results: Patients with T1N1a cancer had the best survival rate (58.2%), whereas the worst prognosis was observed in those with T2bN1b cancer (39.2%). The models were created using the following HR results: Model A (T1N1a, n = 85; 12.4%), Model B (T2a/T2bN1a and T3N0, n = 438; 64.2%), and Model C (T1/T2a/T2bN1b, n = 159; 23.4%). There was a significant difference between the models in terms of overall survival (p = 0.03). The median survival time was 69 months in Model A, 56 months in Model B, and 47 months in Model C (Model A vs. Model B, p = 0.224; Model A vs. Model C, p = 0.01; and Model B vs. Model C, p = 0.04). Multivariate analysis showed that age (p < 0.001), pleural invasion (p < 0.001), and the developed modelling system (p = 0.02) were independently negative prognostic factors.

Conclusion: There was a prognostic difference between stage-IIB subsets in NSCLC patients. The model created for stage-IIB lung cancer showed a high discriminatory power for prognosis.