Value of FDG PET/CT in the differential diagnosis of benign and malignant hilar mediastinal lymph nodes in patients with non-small cell lung cancer

Objective: To investigate the diagnostic value of ¹⁸F-fluorodeoxyglucose (FDG) PET/CT in detecting N₁ or N₂ metastasis of lymph node (LN) with different densities in patients with non-small cell lung cancer (NSCLC). Methods: A total of 118 patients (68 males, 50 females, age range: 27-87 (65.4 ± 10.8) years) with N₀-N₂ M₀ NSCLC in Beijing Hospital between October 2007 and December 2017 were included in this study. All patients underwent ¹⁸F-FDG PET/CT, followed by surgery within 1 month. The pathological findings of the resected hilar mediastinal LN were taken as the gold standard, and ratios of LN metastasis were calculated for different density groups (calcification, partial calcification, hyper density, hypodensity/isodensity). The cut-off values of LN size (short diameter) and the maximum standardized uptake value (SUV_max) in the detection of N₁ and N₂ metastases was determined by the receiver operating characteristic (ROC) curve, and the diagnostic efficiencies were calculated. Independent-sample t test, Mann-Whitney U test and X² test (partition of X²) were used for data analysis. Results: A total of 433 LN has the histopathologic results: 365 LN was in stage N₀, 68 LN was in stage N₁-N₂. There were no metastases in calcification group (0/8). The metastatic LN proportions in partial calcification group (28.6%, 8/28), hypodensity/isodensity group (20.3%, 44/217) were significantly higher than that in the hyper density group (8.9%, 16/180 X² values: 7.369, 9.945, both P < 0.017 (threshold for partition of X²)). There was no significant difference between the partial group and hypodensity/isodensity group (X² = 1.021, P > 0.017). The SUV_max of N₁+ N₂ group was significantly higher than that in N₀ group (6.94 (4.51, 11.36) vs 2.45 (1.93, 3.42); z = -10.388, P < 0.01). According to the ROC curve, the cut-off value of SUV_max in detecting hilar and mediastinal LN was 3.66. The diagnostic sensitivity, specificity and accuracy for N₁ + N₂ metastasis was 85.3%, 78.9%, 80.0% respectively. The cut-off values of SUV_max for hypodensity/isodensity group and hyper density group were 3.66 and 2.79 respectively, and the corresponding sensitivities, specificities, accuracies and positive predictive values (PPV) were 93.2%, 86.7%, 88.0%, 64.1% vs 93.8%, 57.9% (X² = 10.724), 61.1% (X² = 7.326, P < 0.01), 17.9% (X² = 32.971, P < 0.01). The specificity of LN size (1.0 cm) + SUV_max was significantly higher that of LN size or SUV_max alone (94.2% vs 80.9%, 86.7%; X² values: 14.048, 5.661, both P < 0.05) in hypodensity/isodensity group. The specificity and accuracy of LN size (1.0 cm) + SUV_max were significantly higher those of SUV_max alone (X² values: 58.043, 37.037, both P < 0.01) in hyper density group. Conclusions: ¹⁸F-FDG PET/CT is useful in diagnosing the N₁ + N₂ metastases in hypodensity/isodensity LN, but has limitation in diagnosing the partial calcification LN. Combination of lymph node short diameter and SUV_max cut-off value can improve the diagnostic specificity or accuracy of ¹⁸F-FDG PET/CT for hypodensity/isodensity and high density LN. (authors)