Esophageal endoscopic ultrasound with fine-needle aspiration with an on-site cytopathologist : high accuracy for the diagnosis of mediastinal lymphadenopathy
Faculty of Medicine and Health Sciences
Chest. - Chicago, Ill.
, p. 3004-3009
Study objectives: To analyze the accuracy of esophageal endoscopic ultrasound (EUS) with real-time, guided fine-needle aspiration (EUS-FNA) with an on-site cytopathologist in patients with (presumed) lung cancer presenting with mediastinal lymphadenopathy (ML) or a suspect left adrenal gland (LAG). Design: A single-center prospective study. Patients: Sixty-seven outpatients with (presumed) lung cancer with ML or a suspect LAG on either CT and/or positron emission tomography with F-18-fluorodeoxyglucose (FDG-PET) scan. Interventions: All patients underwent EUS-FNA under conscious sedation. A cytopathologist was present during all procedures. Measurements: EUS with and without fine-needle aspiration (FNA) as compared to FDG-PET was evaluated for accuracy in diagnosing cancer, safety, and rate of avoidance for further surgery. Results: Of 67 consecutive patients (56 men; median age, 64 years), malignant AIL or LAG were found in 47 patients (70.1%). In 20 patients (29.9%) without EUS-FNA proof of malignancy, confirmation was obtained by surgical procedure in 13 patients (sarcoidosis [n = 5], infection [n = 1], lung cancer [n = 7]) or by clinical follow-up in 5 patients suggesting benign disease. Sixty-five patients were included in the calculation of test characteristics. With malignancy as an end point, the accuracy for EUS-FNA was 100%. This was better than EUS without FNA (accuracy, 75.4%; p < 0.001) or FDG-PET (accuracy, 75.0% [n = 28]; p = 0.0011). When using final histopathologic diagnosis as an end point, the accuracy of EUS-FNA was 92.3%, since EUS-FNA was unable to show noncaseating granulomas in those patients with sarcoidosis diagnosed after mediastinoscopy. Related to the presence of the in situ cytopathologist, there were no inconclusive samples. No adverse events were recorded, and 67.7% of surgical interventions were avoided following EUS-FNA. Conclusions: The accuracy in this series of EUS-FNA with cytopathologist-assisted rapid on-site evaluation is high. The technique is safe and greatly reduces the number of surgical interventions.