Combined fluorine-18-FDG and carbon-11-methionine PET for diagnosis of tumors in lung and mediastinum.

UNLABELLED We evaluated the value of PET using 18F-fluorodeoxyglucose (FDG) and 11C-methionine, individually or in combination, to distinguish malignant from benign tumors and to identify or exclude mediastinal metastases. METHODS Seventeen patients with a tumor in the lung or mediastinum were evaluated with 18F-FDG and 11C-methionine PET. For morphological comparison, we used CT, and all findings were confirmed by histology of surgical resection specimens (n = 16) or by cytology (n = 1). RESULTS All tumors were visualized equally well with both tracers, and there were no false-positive results. In 2 patients with a malignant tumor, coexisting pneumonia was correctly diagnosed as an inflammatory lesion because of its wedge-like shape. PET correctly excluded hilar invasion and mediastinal lymph node metastases in 10 of 14 patients with primary lung tumor. PET identified mediastinal metastases in 4 of 4 patients. CT failed to detect mediastinal tumor spread in 2 patients and gave a false-positive reading in 2 others. Significantly higher uptake (SUV) and transport rate (slope) values were obtained from malignant than benign lesions with both tracers. No major differences were seen in either the levels of significance or accuracy when the two tracers were compared. Slope values did not add further information to what was obtained with SUV. Density correction of SUV and slope values, to avoid the influence of surrounding air as well as tumor heterogeneity, increased these differences somewhat. Both tracers distinguished malignant from benign lesions with a 93% sensitivity and an accuracy of 89%-95%, but sensitivity improved to 100% when values from both tracers were combined. CONCLUSION Fluorine-18-FDG and 11C-methionine PET visualized all tumors equally well and detected mediastinal spread better than CT. For differentiation purposes, the problems of false-positive and false-negative PET findings could not be safely overcome in a limited number of cases either by the use of both tracers, by the additional use of slope values or by lesion density correction.