Mediastinal lymph node staging in suspected lung cancer: comparison of positron emission tomography with F-18-fluorodeoxyglucose and mediastinoscopy.

Background: In patients with bronchogenic carcinoma, mediastinal lymph node staging is essential for determining treatment options. In this retrospective analysis we compared the results of positron emission tomography (PET) using F-18 fluorodeoxyglucose with those of mediastinoscopy in nodal staging for suspected bronchogenic carcinoma.
Methods: From March 1997 to June 2001, 102 patients (86 male,16 female, age 62 +/- 9 years) underwent both PET and mediastinoscopy for radiologically suspected mediastinal lymph node disease in bronchogenic carcinoma. Total body emission scans were acquired 90 to 150 minutes after injection of 230 MBq of F-18 fluorodeoxyglucose. Mediastinoscopic evaluation of lymph node stations was performed according to the method of Mountain and Dresler (1R, 1L, 2L, 2R, 4L, 4R,7). Patients were eligible if surgical staging was performed within 6 weeks after the PET scan. RESULTS. Of the 102 patients, benign lesions were diagnosed in 15. In 87 patients malignant disease was proven by histology, and bronchogenic carcinoma was found in 82. Of 469 nodal stations analyzed, malignancy was documented by histology in 84. In PET analysis 79 true-positive and 304 true-negative samples were found. Five lymph node stations were false negative, and 81 samples were false positive. False-positive findings in PET frequently were seen in inflammatory lung disease. The sensitivity of PET was 94.1%, specificity was 79% with a diagnostic accuracy of 81.6%. The positive predictive value of PET was 49.3%, and the negative predictive value was 98.4%.
Conclusions: In patients with positive PET scan results histologic verification appears necessary for exact lymph node staging. In view of the negative predictive value mediastinoscopy can be omitted in patients with bronchogenic carcinoma whose PET scan results were negative.