Impact of Smoking on Overall Survival in Lung Cancer: Insights from the TNM Classification

Article Impact Level: HIGH
Data Quality: STRONG
Summary of Journal of Thoracic Oncology, S155608642500824X. https://doi.org/10.1016/j.jtho.2025.07.002 
Dr. Lawson Eng  et al.

Points

The study analyzed the impact of smoking status on lung cancer staging using the IASLC database.
Smoking at diagnosis is known to reduce overall survival and increase treatment-related toxicity risks.
The research evaluated associations between smoking status and stage-stratified overall survival in 48,531 NSCLC patients.
Smoking status was categorized as never, former, or current, with significant differences observed across groups.
Findings suggest incorporating smoking status into the TNM classification could improve prognostic value and treatment guidance.

Summary

A retrospective analysis from the International Association for the Study of Lung Cancer (IASLC) database investigated the impact of smoking status on overall survival (OS) across the ninth edition TNM classification for lung cancer. The study cohort comprised 48,531 patients with histologically proven non-small cell lung cancer (NSCLC) from 78 institutions across 25 countries, diagnosed between 2011 and 2019, with complete demographic, smoking status, disease stage, and survival data. Smoking status was categorized as never, former (quit 1 year before diagnosis), or current (smoked at diagnosis or quit <1 year before). OS was the primary endpoint, assessed from diagnosis to death or last follow-up.

Baseline characteristics of the cohort showed 56% were aged ≥65 years, 51% male, and 66% from Asia. Adenocarcinoma accounted for 75% of cases. Approximately half (51%) presented with stage IA or IB disease, and 69% underwent curative-intent surgical resection. Regarding smoking status, 26% (n=12,439) were current smokers, 41% (n=19,677) were former smokers, and 34% (n=16,415) had never smoked. Significant geographic variations were noted; 85% of never smokers were from Asia, compared to 60% of current smokers and 53% of former smokers (p < 0.01).

Multivariable Cox proportional hazards regression models, stratified by data source and adjusted for age, sex, histologic type, region, ECOG performance status, and anatomical stage, were used to quantify the prognostic effect of smoking status. Subgroup analyses by histologic type, geographic region, sex, and individual ninth edition stage groups (IA-IVB) were performed. Recursive partitioning and amalgamation (RPA)-generated survival trees further evaluated the importance of smoking status relative to anatomical descriptors. The findings aim to inform potential improvements in the prognostic value of lung cancer staging by incorporating non-anatomical factors like smoking.

Link to the article: https://www.jto.org/article/S1556-0864(25)00824-X/fulltext

References

Eng, L., Evans, W. K., Eisele, M., Nishimura, K. K., Land, S. R., Rami-Porta, R., Dresler, C. M., Asamura, H., Fathi, J. T., Lechtenberg, K., Rusch, V., Sheikh, M., Wynes, M. W., Zius, K., Osarogiagbon, R. U., & Warren, G. W. (2025). The international association for the study of lung cancer staging project: The impact of smoking status on lung cancer staging in the ninth edition of the tnm classification. Journal of Thoracic Oncology, S155608642500824X. https://doi.org/10.1016/j.jtho.2025.07.002