Role of transesophageal endosonography-guided fine-needle aspiration in the diagnosis of lung cancer.

Study Objective: Bronchoscopic methods fail to diagnose lung cancer in up to 30% of patients. We studied the role of transesophageal endosonography (EUS)-guided fine-needle aspiration (FNA; EUS-FNA) in such patients.

Design: Prospective study. The final diagnosis was confirmed by cytology, histology, or clinical follow-up.

Setting: University hospital.

Patients: Thirty-five patients (30 male and 5 female; mean age, 60.9 years; range, 34 to 88 years) with suspected lung cancer in whom bronchoscopic methods failed. Patients with a known diagnosis, recurrence of lung cancer, or mediastinal metastasis from an extrathoracic primary were excluded.

Interventions: EUS and guided FNA of mediastinal lymph nodes.

Results: The procedure was uneventful, and material was adequate in all. The final diagnosis by EUS-FNA was malignancy in 25 patients (11 adenocarcinoma, 10 small cell, 3 squamous cell, and 1 lymphoma) and benign disease in 9 patients (5 inflammatory, 2 sarcoidosis, and 2 anthracosis). Another patient with a benign result had signet-ring cell carcinoma diagnosed on pleural fluid cytology (probably false-negative in EUS-FNA). The sensitivity, specificity, accuracy, and positive and negative predictive values were 96, 100, 97, 100, and 90%, respectively. There were no complications. Reviewing the EUS morphology, the nodes were predominantly located in levels 7 and 8 of American Thoracic Society mediastinal lymph node mapping (subcarinal and paraesophageal region). In seven patients, the punctured nodes were < 1 cm (four malignant and three benign), which are difficult to sample by other methods. The malignant nodes had a hypoechoic, homogenous echotexture.

Conclusions: EUS-FNA is a safe, reliable, and accurate method to establish the diagnosis of suspected lung cancer when bronchoscopic methods fail, especially in the presence of small nodes.