Background: The significance of diffusion-weighted imaging (DWI) is uncertain for the diagnosis of nodal involvement. The purpose of this study was to examine diagnostic capability of DWI compared with PET-CT for nodal involvement of lung cancer.
Methods: A total of 160 lung cancers (114 adenocarcinomas, 36 squamous cell carcinomas, and 10 other cell types) were analyzed in this study. DWI and PET-CT were performed preoperatively.
Results: The optimal cutoff values to diagnose metastatic lymph nodes were 1.70 × 10(-3) mm(2)/s for ADC value and 4.45 for SUVmax. DWI correctly diagnosed N staging in 144 carcinomas (90 %) but incorrectly diagnosed N staging in 16 (10 %) [3 (1.9 %) had overstaging, 13 (8.1 %) had understaging]. PET-CT correctly diagnosed N staging in 133 carcinomas (83.1 %) but incorrectly diagnosed N staging in 27 (16.8 %) [4 (2.5 %) had overstaging, 23 (14.4 %) had understaging]. Sensitivity, accuracy, and negative predictive value for N staging by DWI were significantly higher than those by PET-CT. Of the 705 lymph node stations examined, 61 had metastases, and 644 did not. The maximum diameter of metastatic lesions in lymph nodes were 3.0 ± 0.9 mm in 21 lymph node stations not detected by either DWI or PET-CT: 7.2 ± 4.1 mm in 39 detected by DWI, and 11.9 ± 4.1 mm in 24 detected by PET-CT. There were significant differences among them. The sensitivity (63.9 %) for metastatic lymph node stations by DWI was significantly higher than that (39.3 %) by PET-CT. The accuracy (96.2 %) for all lymph node stations by DWI was significantly higher than that (94.3 %) by PET-CT.
Conclusions: DWI has advantages over PET-CT in diagnosing malignant from benign lymph nodes of lung cancers.