Nomogram based on quantitative lung CT features to identify cardiovascular disease in chronic obstructive pulmonary disease and predict prognosis

To explore the performance of a CT-derived lung nomogram for identifying cardiovascular disease (CVD) among individuals with chronic obstructive pulmonary disease (COPD) and assessing the relationship with COPD prognosis. This retrospective analysis enrolled hospitalized COPD patients between September 2015 and April 2023, with clinical data and visually assessed coronary artery calcium scores (CACS) collected for all participants. A quantitative model was constructed by extracting features from lung CT images and employing the least absolute shrinkage and selection operator algorithm for feature selection. The quantitative features and clinical factors were merged to formulate a nomogram. Area under the ROC curve (AUC) and decision curve analysis were used to study the ability of the nomogram to identify prevalent CVD. In Kaplan-Meier analysis, the predictive value of the nomogram for COPD re-hospitalization and all-cause mortality as endpoints was studied. Of 643 COPD patients (mean age, 68 years ± 10SD; 110 female), 159 had a history of CVD. The derived nomogram had better ability to identify CVD (AUC: 0.87; 95%CI 0.78, 0.94) than the clinical factors alone (AUC: 0.75; 95%CI 0.63, 0.86) and visual CACS (AUC: 0.68; 95%CI 0.56, 0.79) in internal validation, and achieved good performance in external validation with highest AUC (0.77; 95%CI 0.71, 0.84). The nomogram demonstrated a strong association with events (P < 0.001). A nomogram based on quantitative CT features and clinical factors could effectively identify CVD in COPD patients, with bette discriminatory capacity than visual CACS or clinical factors alone. The nomogram also showed association with COPD re-hospitalization and all-cause mortality. In accordance with the Declaration of Helsinki, this retrospective study was approved (Ethical Approval: Second Affiliated Hospital of Naval Medical University Ethics Committee, 2022SL068, December 6, 2022; Trial Registration: Chinese Clinical Trial Registry, ChiCTR2300069929, March 29, 2023), with a waiver for individual patient consent requirements.