Accuracy and Predictors of Success of EUS-B-FNA in the Diagnosis of Pulmonary Malignant Lesions: A Prospective Multicenter Italian Study

Background: The role of endoscopic ultrasound with bronchoscope fine-needle aspiration (EUS-B-FNA) in the diagnosis of suspected malignant pulmonary lesions adjacent to the esophagus has been poorly investigated. The aim of the present study was to assess the accuracy of EUS-B-FNA for the diagnosis and molecular profiling of paraesophageal pulmonary lesions, as well as its predictors of success. Materials and methods: Patients who underwent EUS-B-FNA for the diagnosis of paraesophageal lesions were consecutively enrolled in four Italian centers. Demographic, clinical, procedural, pathological, and molecular characteristics of the malignant samples were collected. The primary outcome was the diagnostic accuracy for pulmonary malignancies. Secondary outcomes were diagnostic yield and predictors of success for diagnosis and molecular profiling. Results: 107 adult patients (60 [56.1%] males; median (interquartile range) age: 69 [60-70] years) were enrolled. The diagnostic accuracy of EUS-B-FNA was 95.3% in the overall cohort and 95.2% in the 99 patients with a final diagnosis of malignancy. Neither clinical nor procedural variables significantly affected the diagnostic accuracy, whereas rapid on-site evaluation (ROSE), performed by pathologists or trained pulmonologists, was a strong predictor for a complete molecular profiling (OR [95% CI]: 12.9 [1.2-137.4]; p value: 0.03). Conclusion: EUS-B-FNA is a safe and accurate method for the diagnosis of paraesophageal pulmonary lesions. The presence of ROSE is relevant for a complete molecular profiling in this selected cohort of patients with advanced lung cancer.