Search for a command to run...
This study aimed to evaluate the diagnostic value of four serum biomarkers (CEA, CYFRA21-1, NSE, CA125), both individually and in combination, for identifying clinically significant abnormal chest computed tomography (CT) findings in male miners from Chongqing, China, and to determine their relative contributions within a combined model. In this cross-sectional study (June 2022-December 2023), 110 miners underwent low-dose CT scans and serum biomarker analysis. We employed multivariable logistic regression to assess associations, receiver operating characteristic (ROC) analysis to evaluate diagnostic performance, and Weighted Quantile Sum (WQS) regression to quantify the relative weights of each biomarker within the mixture effect. Logistic regression revealed significant associations for CEA (fully-adjusted OR = 2.01), CYFRA21-1 (OR = 3.36), and NSE (OR = 1.18), but not for CA125. The combined biomarker model demonstrated high diagnostic accuracy (AUC = 0.92), significantly outperforming individual biomarkers (AUC range: 0.58–0.781) and achieving 84.9% sensitivity and 100% specificity. Notably, the specificity suggests its potential utility as a high-specificity “rule-in” tool. WQS regression confirmed a significant positive mixture effect (OR = 4.11) and identified a distinct hierarchy of contribution: CEA was the dominant driver (37.7%), followed by NSE (32.8%) and CYFRA21-1 (29.5%), while CA125’s role was negligible (< 0.1%). A panel combining CEA, CYFRA21-1, and NSE provides promising preliminary performance for detecting CT abnormalities in miners, representing a candidate tool for initial screening in occupational health settings. However, these findings require validation in larger, independent cohorts.