Related Experiment Video
Updated: Jan 9, 2026
Author Spotlight: Advancements in Molecular Biomarker Testing for Non-Squamous Non-Small Cell Lung Cancer
Published on: September 8, 2023
50-Year age threshold for early-onset NSCLC: A SEER-TCGA retrospective analysis reveals a prognostic paradox based on
Zheng Lu1, Peilan Zheng2, Yi Liao3
1The School of Basic Medical Sciences, Fujian Medical University, Fuzhou, China.
Objective:
To determine the optimal age threshold for distinguishing early-onset non-small cell lung cancer (NSCLC) and explore the mechanisms underlying its unique prognosis.
Methods:
Cases with NSCLC from 2004 to 2015 were identified from Surveillance, Epidemiology, and End Results (SEER) database. X-tile survival analysis and quadratic polynomial fitting identified age thresholds. Random survival forests (RSFs) and restricted cubic splines evaluated age-treatment efficacy interactions. Propensity score matching controlled confounding factors. Molecular analysis used The Cancer Genome Atlas, GEO, cBioPortal database to identify differentially expressed genes and mutations.
Results:
A total of 161 410 cases with NSCLC were extracted from the SEER database, 6396 cases were defined as early-onset lung cancer (EOLC), and 155 014 cases as later-onset lung cancer (LOLC). The optimal cutoff for defining EOLC is 50 years. Patients with EOLC had worse baseline features (EOLC vs. LOLC: 37.3% Stage IV vs. 26.4%, P < 0.001) but superior survival [LOLC: hazard ratio: 1.54, P < 0.001]. Surgery's protective effect declined with age, while chemotherapy and radiotherapy benefits increased. RSF ranked surgery (importance = 0.2584) and age (0.0616) as top prognostic factors. Female and lower lobe tumor location conferred survival advantages. Molecular analysis revealed five intersecting genes upregulated in EOLC and lower-frequency KRAS/TTN mutations in EOLC tumors.
Conclusion:
The age threshold for optimal differentiation between patients with EOLC and LOLC is 50 years, Patients with EOLC benefit from aggressive surgery and age-dependent treatment, supported by treatment efficacy and molecular heterogeneity.
