PET-FDG scan enhances but does not replace preoperative surgical staging in non-small cell lung carcinoma

Objective: To assess the effectiveness of positron emission tomography with radiolabeled [F-18]-2-fluoro-deaxy-D-glucose (PET-FDG) imaging in mediastinal lymph node (LN) staging for non-small cell lung carcinoma (NSCLC) and to compare it to conventional clinical and surgical staging. Methods: From June 1998 to February 2000, we enrolled 64 potentially resectable NSCLC patients in a prospective study of PET-FDG imaging of the mediastinum to assess LN involvement. Results of this technique were compared to conventional clinical and surgical staging. Diagnostic efficacy was determined by calculating sensitivity, specificity, overall accuracy, and positive and negative predictive values for each method. Results: PET-FDG imaging correctly identified nodal stage (N0-N1 vs. N2) in 50 out of 61 patients (82%), overstaging occurred in eight patients (13%), and understaging in three patients (4.9%). The sensitivity, specificity, accuracy, and positive and negative predictive values for PET-FDG scan imaging were 67, 85, 82, 43, and 93.6%, respectively. Conventional staging correctly identified nodal stage (N0-N1 vs. N2) in 51 out of 62 patients (82%), overstaging occurred in five patients (8.1%), and understaging in six patients (9.7%). The sensitivity, specificity, accuracy, and positive and negative predictive values for conventional staging were 33, 90.6, 82, 37, and 89%, respectively. With regard to N2 disease, conventional staging showed a poor sensitivity (33%). Indeed, six out of 64 patients were understaged for mediastinal LN involvement. Even though the improvement was not statistically significant (McNemar P = 0.08), the combined use of PET-FDG scan and computerized tomography (CT) scan allowed a two-fold increase in the sensitivity of our clinical preoperative staging. Moreover, relying on the PET-scan high negative predictive value might have contributed to a three-fold decrease in the number of required surgical staging procedures. Conclusions: Our study shows that the PET-FDG imaging strength lies in its very high negative predictive value and increased sensitivity. In this study, the overall accuracy of PET-FDG scan (82%) was lower than previously reported. Combined with chest CT-scan preoperatively, it may alleviate the need for surgical staging when PET-FDG studies of the mediastinum are negative. However, with a positive PET-FDG scan result, further diagnostic procedures should be pursued in order to avoid overstaging and allow better surgical patient selection. (C) 2001 Elsevier Science B.V. All rights reserved.