Abstract

The clinical staging of lung cancer, especially stage IIIA, provides important information for clinical analysis and treatment method to surgeon ar radiation oncologist. The aim of this analysis is to evaluate of difference between clinical staging and pathologic staging and determine to accuracy of preoperative pathologic diagnosis and clinical staging for clinically staged IIIA NSCLC. Analysis was performed on 138 patients who has been staged IIIA NSCLC. Pathologic diagnosis was performed using sputum cytology, bronchoscopy or percutaneous needle aspiration. All patients were evaluated by CT and staged according to American Joint Commit- tee on Cancer staging classification. All patients received the thoracotomy. The pathologic diagnosis was changed after surgery in 9 patients(6.5%). The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CT in dis- tinguishing T3 from Tl-T2 were 62.5%, 89.7%, 47.6%, 87.5%, and 75.3%, respectively. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CT in distinguishing N2 from NO-Nl were 91.8%, 24.3%, 42.7%, 78.3%, and 54.8% respec- tively. According to pathologic staging, there were 27(19.6%) patients in stage I, 24(17.4%) in stage II, 64(46.4%) in stage IIIA, 21(15.2%) in stage IIIB, and 2(1.5%) in stage IV. Comparing clincal staging with pathologic staging, concordant results were found in 46.4%, 37.0% were clinically overestimated and 16.7% underestimated. Preoperative pathologic diagnosis was well correlate postoperative pathologic finding. But CT accuracy was low in both primary tumor and mediastinal lymph node. Especially to diagnose malignant involvement of mediastinal nodes, CT is not a valid means. Patients would not be denied the opportunity for curative surgery on the basis of CT signs.

• Keywords: Non-small cell lung cancer, Staging, Accuracy