The diagnostic yield and safety of ultrasound-assisted transthoracic fine-needle aspiration of drowned lung.

BACKGROUND Proximal lung tumors, though not discernable by means of transthoracic ultrasound (US), may cause varying degrees of pulmonary collapse and postobstructive pneumonitis which may give rise to a 'drowned lung' appearance on chest computed tomography (CT) and US. The diagnostic yield for malignancy of US-assisted transthoracic fine-needle aspiration (FNA) of these areas of drowned lung is unknown. OBJECTIVES We aimed to explore the feasibility of US-assisted FNA in this setting by prospectively investigating its diagnostic yield and safety. METHODS We enrolled 31 patients (aged 59.4 ± 9.7 years, 17 males) with central tumors and secondary drowned lung on CT scan. A respiratory physician performed transthoracic US to identify the target drowned lung tissue. Three US-assisted superficial FNA passes (≤20 mm from the pleura) were followed by 3 deeper FNA passes (>20 mm) aimed in the direction of a visible or approximated central mass. Rapid on-site evaluation of specimens was used. RESULTS Superficial FNA was diagnostic in 11 patients (35.5%), whereas deeper FNA was diagnostic in 23 patients (74.2%, p = 0.002). Deeper FNA confirmed malignancy in all cases with diagnostic superficial FNA. We observed no pneumothoraces or major hemorrhage. All patients were ultimately diagnosed with malignancy (bronchogenic carcinoma, n = 30; lymphoma, n = 1). CONCLUSIONS US-assisted FNA of drowned lung has an acceptable diagnostic yield and is safe.