18F-FDG PET/CT in Differential Diagnosis of Solitary Pulmonary Nodule with Different Size

郑建国,屈婉莹,姚稚明,王梅,罗志福,管一晖,王全师,陈秋松,张永学,黄庆娟,尹吉林,李亚明,刘庆伟,郭万华
DOI: https://doi.org/10.3760/cma.j.issn.2095-2848.2008.03.009
2008-01-01
Abstract:Objective The differential diagnosis of solitary pulmonary nodule (SPN) is not an easy task for the thoracic imaging. The aim of this study was to assess the value of 18F-fluorodeoxyglucose (FDG) PET/CT in helping the differential diagnosis of benign and malignant SPN. Methods SPN patients with known pathological results underwent PET/CT imaging in 10 different hospitals. The results of CT, PET/CT, pathology, quantitative data [standardized uptake value (SUV), size of nodules] were collected. The characteristic wordings on the CT and PET/CT reports that were considered to be positive included lung cancer, highly possible lung cancer and possible lung cancer. Those of the negative findings included benign lesion, highly possible benign lesion and possible benign lesion. The results, which were difficult to be judged as positive or negative, were considered to be indeterminate. The analysis of the data was carried out using SPSS 10.0. Results There were 120 SPNs included for analysis: 28 benign and 92 malignant nodules. When the lesion size was≤ 1.0cm, the percentage of malignant lesions was 52.94%(9/17). When the nodule size ranged from 1.1 to 2.0cm and from 2.1 to 3.0cm, the percentage rose to 75.44%(43/57) and 86.96%(40/46), respectively. The average of individual lesion SUVmax was 4.5±4.9 and SUVmean was 2.9±3.6 in 28 benign nodules. The corresponding SUVmax and SUVmean for the 92 malignant lesions were 6.7±4.2 and 4.9±3.6, respectively. The SUV of benign nodules was significantly lower than that of malignant nodules. The SUV increased as the size of nodules increased regardless of benignity or malignancy. The sensitivity, specificity and accuracy of CT and PET/CT were 78.26%,3/8,72.22% and 94.51%,64.00%,87.93% respectively. Conclusions 18F-FDG PET/CT is superior to CT with higher sensitivity, specificity and accuracy for differentiating benign from malignant SPNs. The size of SPN is an important parameter for differentiation. The possibility of malignancy was about 87% when the nodule is greater than 2.0cm. SUV can be helpful but the conventional threshold value of 2.5 was only a rough reference, because the SUVmean statistics of the benign nodules (2.9±3.6) were highly variable, particularly with the subcentimeter nodules.
What problem does this paper attempt to address?