Diagnostic Performance of Tumor-to-Lymph Node SUV Ratio in PET/CT-Guided EBUS-TBNA for NSCLC.

[BACKGROUND] Accurate evaluation of mediastinal and hilar lymph node (LN) is essential in non-small cell lung cancer (NSCLC) management. While PET/CT plays a central role, SUV alone is limited by false positives and false negatives. The tumor-to-lymph node SUV (T/LN SUV) ratio may improve diagnostic accuracy by contextualizing nodal uptake relative to the primary tumor.

[METHODS] This retrospective single-center study analyzed 2469 LNs from 1408 NSCLC patients who underwent both PET/CT and EBUS-TBNA between 2010 and 2021. Sonographic features, including LN size, presence of coagulative necrosis sign (CNS), absence of central hilar structure (CHS), PET/CT findings and cytology results were recorded. Multivariate logistic regression identified independent predictors of malignancy. ROC analysis determined the optimal cut-off value for T/LN SUV ratio.

[RESULTS] Malignant LNs (n = 980, 39.7%) were more likely to be larger size, round in shape, with homogeneous texture, distinct margins, CNS, and lacking CHS (all P < .001). Malignant LNs had significantly lower T/LN SUV ratios than benign LNs (1.74 ± 1.45 vs. 4.12 ± 2.46; P < .001). Lower T/LN SUV ratio, higher SUV, larger size, and absence of CHS were independent predictors of malignancy. A cut-off of 2.10 yielded 79.1% sensitivity, 78.9% specificity, 79% diagnostic accuracy (AUC: 0.838).

[CONCLUSIONS] The T/LN SUV ratio is a simple, and reproducible parameter with high diagnostic performance for predicting LN malignancy in PET/CT-guided EBUS-TBNA for NSCLC. Although this was a single-center retrospective study, the large cohort strengthens the reliability of our findings, and a cut-off value ≤ 2.10 may help clinicians prioritize LN sampling.