Invasive staging of non-small cell lung cancer: a review of the current evidence.

STUDY OBJECTIVES To determine the test performance characteristics of transbronchial needle aspiration (TBNA), transthoracic needle aspiration (TTNA), endoscopic ultrasound-guided needle aspiration (EUS-NA), and mediastinoscopy in staging non-small cell lung cancer (NSCLC). DESIGN, SETTING, AND PARTICIPANTS Systematic search of MEDLINE, HealthStar, and Cochrane Library databases to July 2001 and print bibliographies. Included were studies comparing staging results of TBNA, TTNA, EUS-NA, or mediastinoscopy against either tissue histologic confirmation or long-term clinical follow-up (> or = 1 year). Patients included were those with NSCLC or small cell lung cancer. MEASUREMENT AND RESULTS For patients with lung cancer, the pooled sensitivity for TBNA was 0.76, the pooled specificity was 0.96, and the negative predictive value (NPV) was 0.71. For TTNA, the pooled sensitivity was 0.91, with an NPV of 0.78. EUS-NA had a pooled sensitivity of 0.88, a pooled specificity of 0.91, and an NPV of 0.77. For standard cervical mediastinoscopy, the pooled sensitivity was 0.81, with an NPV of 0.91. The addition of either extended cervical mediastinoscopy or anterior mediastinotomy to standard cervical mediastinoscopy appeared to improve the sensitivity of any of the procedures alone. CONCLUSIONS Invasive clinical staging of NSCLC can be performed effectively by TBNA, TTNA, EUS-NA, or mediastinoscopy. Selection of the appropriate study is dependent on the degree of suspicion for metastatic disease, the patient's comorbid illnesses, and the availability and performance characteristics of procedural options.