Abstract
Chronic inflammation has been associated with the development of lung cancer. In this study, we examined the association between C-reactive protein (CRP) and lung cancer in a prospective cohort study and used Mendelian randomization (MR) to clarify the causality. We included 420 977 participants from the UK Biobank (UKB) in the analyses; 1892 thereof were diagnosed with lung cancer during the follow-up. Hazards ratios (HRs) of CRP concentrations were estimated by Cox proportional hazard models and two approaches of MR analysis were performed. Besides, we added CRP concentrations to epidemiological model of lung cancer to evaluate its prediagnostic role through time-dependent receiver operating characteristic curve analysis. Elevated CRP levels were associated with a 22% increased lung cancer risk per 1 SD increase (HR = 1.22, 95% confidence interval [CI] = 1.18-1.26). Positive associations were observed in small cell lung cancer (HR = 1.21, 95% CI = 1.10-1.33), lung adenocarcinoma (HR = 1.17, 95% CI = 1.11-1.23) and lung squamous cell carcinoma (HR = 1.22, 95% CI = 1.14-1.31). No genetical association of circulating CRP levels and lung cancer risk was observed in MR analysis. When added to a risk model of lung cancer, CRP improved the performance of model as long as 8 years among current smokers (basic model: C-statistic = 0.78 [95% CI = 0.75-0.80]; CRP model: C-statistic = 0.79 [95% CI = 0.76-0.81]; Pnonadjusted = .003, Padjusted = .014). Our results did not support the causal association of circulating CRP with lung cancer risk. However, circulating CRP could be a prediagnostic marker of lung cancer as long as 8 years in advance for current smokers.