Background: While pulmonary nodules can be substantially divided into solid and ground-glass opacity (GGO) ones on CT image, they have different biological natures which could cause false positive or false negative to diagnose malignancy on positron emission tomography with fluorodeoxyglucose (FDG-PET). To determine the effectiveness of PET for small pulmonary nodules, the nodules were classified into solid and GGO ones, of which results were compared with the data of PET scans. The lower limit size of nodules for PET imaging was also evaluated.
Methods: Prospective FDG-PET scans were undertaken for 136 non-calcified nodules less than 3 cm in diameter. CT density histograms were made for each nodule to classify into solid and GGO ones.
Results: Eighty-one nodules were malignant and 55 were benign. All of the 20 nodules less than 1 cm in diameter (n = 8 in malignant, n = 12 in benign), were negative on PET regardless of the histology. In the 116 nodules 1-3 cm in diameter (n = 73 in malignant, n = 43 in benign), there were 15 false negative and 15 false positive nodules, with a sensitivity of 79% and specificity of 65%. CT density histograms showed 101 solid nodules (n = 63 in malignant, n = 38 in benign) and 15 GGO nodules ( n = 10 in malignant, n = 5 in benign). All of the 10 malignant nodules with GGO images were histologically well-differentiated adenocarcinoma and 9 of them (90%) were false negative on PET. Four of the 5 (80%) benign nodules with GGO images were focal pneumonia with well-preserved air spaces, causing false positive on PET. Sensitivity and specificity for nodules with GGO images were 10 and 20%, respectively, which were significantly lower than 90 and 71% for nodules with solid images (P < 0.001).
Conclusion: Pulmonary nodules which are less than 1cm in size or show GGO images on CT cannot be evaluated accurately by PET.
Copyright 2004 Elsevier Ireland Ltd.