Computed Tomography-Based Radiomic Nomogram to Predict Occult Pleural Metastasis in Lung Cancer.

Area of Science:

• Medical Imaging
• Oncology
• Radiology

Background:

• Preoperative identification of occult pleural metastasis (OPM) in lung cancer is a significant clinical challenge.
• Accurate OPM detection is vital for appropriate treatment planning and patient management.

Purpose of the Study:

• To develop and validate a predictive model for preoperative OPM identification in lung cancer.
• The model integrates clinical data with radiomic features from chest CT scans.

Main Methods:

• A training cohort of 50 OPM-positive and 50 non-metastatic lung cancer patients was used.
• Least absolute shrinkage and selection operator (LASSO) logistic regression identified key radiomic features.
• A predictive nomogram was built using clinical factors and radiomic scores, then validated on 545 patients.

Main Results:

• Significant clinical predictors included carcinoembryonic antigen (CEA), neutrophil-to-lymphocyte ratio (NLR), clinical T stage, and tumor-pleural relationship.
• The integrated model combining radiomic scores (VOI) with CEA and NLR achieved high predictive performance (AUCs of 0.890 training, 0.855 validation).

Conclusions:

• CT-derived radiomic features show promise for identifying lung cancer patients at risk of OPM.
• The developed nomogram, integrating CEA, NLR, and radiomic scores, improves preoperative OPM prediction accuracy for clinical decision-making.