Background and study aims: Transesophageal endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) of mediastinal lymph nodes is increasingly used to detect noncaseating granulomas in patients with suspected sarcoidosis. The optimal needle size and tissue processing method for detecting noncaseating granulomas are debated. We assessed the value of cell-block analysis when added to conventional cytological evaluation of EUS aspirates obtained by 22-gauge needles in patients with stage I and II sarcoidosis.
Patients and methods: Data from 101 consecutive patients (55 % of whom had previously had a nondiagnostic bronchoscopy) with suspected pulmonary sarcoidosis (stage I and II), who underwent EUS-FNA of mediastinal lymph nodes with 22-gauge needles were retrospectively analyzed.
Results: The sensitivity of EUS in detecting granulomas was 87 % (cytology and cell-block analysis together) (stage I, 92 %; stage II, 77 %). In 33 % of cytology negative patients (n = 6), granulomas were present in the cell block. The optimal yield for granuloma detection was reached with four needle passes. One patient developed mediastinitis after EUS-FNA.
Conclusions: Cell-block analysis added to conventional cytological evaluation of 22-gauge EUS aspirates, results in a high yield in detecting granulomas in patients with suspected sarcoidosis and reduces the false-negative rate substantially. EUS has a considerably higher yield in stage I compared with stage II sarcoidosis. For an optimal yield, four needle passes are required.
Georg Thieme Verlag KG Stuttgart. New York.