FDG-PET parameters predicting mediastinal malignancy in lung cancer.

Area of Science:

• Oncology
• Nuclear Medicine
• Radiology

Background:

• Mediastinal lymph node staging is crucial for non-small cell lung cancer (NSCLC).
• F-18 fluorodeoxyglucose positron emission tomography (FDG-PET) offers non-invasive staging, but its accuracy varies between centers.
• Identifying reliable FDG-PET predictors is essential to minimize variability and optimize staging procedures.

Purpose of the Study:

• To identify FDG-PET predictors of mediastinal malignancy in NSCLC.
• To minimize intercenter variability in FDG-PET staging measures.
• To improve the selection of subsequent staging procedures for NSCLC patients.

Main Methods:

• A multicenter study involving NSCLC patients staged with FDG-PET and endobronchial ultrasonography with needle aspiration (EBUS-NA).
• Systematic nodal dissection after therapeutic surgery served as the gold standard.
• Assessment of intercenter variability and predictive power of various FDG-PET measures, including SUVmax mediastinum/tumor ratio.

Main Results:

• The SUVmax mediastinum/tumor ratio significantly reduced intercenter variability (p=0.083).
• This ratio demonstrated high accuracy (AUC 0.77) and predictive power for mediastinal malignancy at a 0.4 cutoff (OR 6.62).
• FDG-PET combined with EBUS-NA improved staging sensitivity and negative predictive values, particularly for larger tumors or high SUVmax mediastinum/tumor ratios.

Conclusions:

• The SUVmax mediastinum/tumor ratio is a robust predictor of regional tumor spread in NSCLC.
• This standardized FDG-PET measure is not affected by intercenter variability.
• An SUVmax mediastinum/tumor ratio cutoff of 0.4 provides over 70% accuracy for identifying mediastinal malignancy.