OBJECTIVE. The objective of our study was to directly compare the capability of dynamic frst-pass contrast-enhanced (CE) perfusion area-detector CT (ADCT) and FDG PET/CT for differentiation of metastatic from nonmetastatic lymph nodes and assessment of N stage in patients with non-small cell lung carcinoma (NSCLC). SUBJECTS AND METHODS. Seventy-seven consecutive patients, 45 men (mean age ± SD, 70.4 ± 5.9 years) and 32 women (71.2 ± 7.7 years), underwent dynamic frst-pass CE-perfusion ADCT at two or three different positions for covering the entire thorax, FDG PET/CT, surgical treatment, and pathologic examination. From all ADCT data for each of the subjects, a whole-chest perfusion map was computationally generated using the dual- and single-input maximum slope and Patlak plot methods. For quantitative N stage assessment, perfusion parameters and the maximum standardized uptake value (SUVmax) for each lymph node were determined by measuring the relevant ROI. ROC curve analyses were performed for comparing the diagnostic capability of each of the methods on a per-node basis. N stages evaluated by each of the indexes were then statistically compared with the fnal pathologic diagnosis by means of chi-square and kappa statistics. RESULTS. The area under the ROC curve (A z) values of systemic arterial perfusion (Az = 0.89), permeability surface (Az = 0.78), and SUVmax (Az = 0.85) were signifcantly larger than the A z values of total perfusion (Az = 0.70, p < 0.05) and distribution volume (Az = 0.55, p < 0.05). For each of the threshold values, agreement for systemic arterial perfusion calculated using the dual-input maximum slope model was substantial ( = 0.70, p < 0.0001), and agreement for SUV max was moderate ( = 0.60, p < 0.0001). CONCLUSION. Dynamic frst-pass CE-perfusion ADCT is as useful as FDG PET/CT for the differentiation of metastatic from nonmetastatic lymph nodes and assessment of N stage in patients with NSCLC.
All Science Journal Classification (ASJC) codes