Pulmonary embolism in lung cancer: a retrospective exploratory study

Abstract

Lung cancer is associated with a high risk of venous thromboembolism, and pulmonary embolism (PE) is a critical complication. The synchronous diagnosis of PE at the initial presentation of lung cancer presents diagnostic challenges and affects management. The predictive factors for this dual diagnosis remain poorly defined. A retrospective comparative study was conducted on 251 newly diagnosed lung cancer patients, including 79 with concurrent PE (observation group) and 172 without PE (control group). Demographic, clinical, laboratory, and tumor - related data were collected. Statistical analysis included univariate and multivariate logistic regression to seek independent risk factors and ROC curve analysis to evaluate their potential predictive performance. Univariate analysis revealed significant differences in multiple variables, including pericardial/pleural effusion, smoking index, hematologic parameters, coagulation markers, D - dimer, liver enzymes, albumin, and tumor markers (CEA, CA125, CA199) (P < 0.05). Multivariate logistic regression analysis indicated that smoking index, hemoglobin, PT, direct bilirubin, albumin, and CEA may be independent risk factors for synchronous pulmonary embolism in lung cancer patients. The combination of the six independent risk factors also demonstrated significant predictive potential (AUC = 0.857, sensitivity 79.1%, specificity 79.7%). Lung cancer patients presenting with synchronous PE exhibit a distinct clinical phenotype. The smoking index, hemoglobin, PT, direct bilirubin, albumin, and CEA may be independent predictive factors. D - dimer is a well diagnostic marker. These findings suggest a practical framework to help guide clinical suspicion and encourage targeted diagnostic evaluation for PE at the initial diagnosis of lung cancer, contributing to improved early intervention and patient outcomes.