Background: Improved prediction of prognosis among lung cancer patients could facilitate better clinical management. We aimed to study the prognostic significance of circulating proteins at the time of lung cancer diagnosis, among patients with and without smoking history.
Methods: We measured 91 proteins using the Olink Immune-Oncology panel in plasma samples that were collected at diagnosis from 244 never smoking and 742 ever smoking patients with stage I-IIIA non-small cell lung cancer (NSCLC). Patients were recruited from nine centres in Russian Federation, Poland, Serbia, Czechia, and Romania, between 2007-2016 and were prospectively followed through 2020. We used multivariable Survey-weighted Cox models to assess the relationship between overall survival and levels of proteins by adjusting for smoking, age at diagnosis, sex, education, alcohol intake, histology, and stage.
Results: The 5-year survival rate was higher among never than ever smoking patients (63.1% vs. 46.6%, P<0.001). In age- and sex-adjusted survival analysis, 23 proteins were nominally associated with overall survival, but after adjustment for potential confounders and correcting for multiple testing, none of the proteins showed a significant association with overall survival. In stratified analysis by smoking status, IL8 [hazard ratio (HR) per standard deviation (SD): 1.40, 95% confidence interval (CI): 1.18-1.65, P=1×10-4] and hepatocyte growth factor (HGF) (HR: 1.45, 95% CI: 1.18-1.79, P=5×10-4) were associated with survival among never smokers, but no protein was found associated with survival among ever smokers. Integrating proteins into the models with clinical risk factors did not improve the predictive performance of NSCLC prognosis [C-index of 0.63 (clinical) vs. 0.64 (clinical + proteins) for ever smokers, P=0.20; C-index of 0.68 (clinical) vs. 0.72 (clinical + proteins) for never smokers, P=0.28].
Conclusions: We found limited evidence of a potential for circulating immune- and cancer-related protein markers in lung cancer prognosis. Whereas some specific proteins appear to be uniquely associated with lung cancer survival in never smokers.
Keywords: Lung cancer; prognosis; proteomics; smoking.
2024 AME Publishing Company. All rights reserved.