Specimen processing techniques for endobronchial ultrasound-guided transbronchial needle aspiration

Abstract

Background: Endobronchial ultrasound is used for sampling thoracic pathologic processes. Histologic examination may provide added diagnostic yield to cytologic preparations owing to superior assessment of architecture and immunohistochemistry. It remains unclear whether specific specimen processing technique impacts diagnostic yield. We hypothesized that diagnostic yield using histologic analysis of core needle biopsies is higher than cytologic preparations alone.

Methods: We evaluated 177 consecutive patients with mediastinal abnormalities. An interventional pulmonologist or thoracic surgeon performed endobronchial ultrasound. We compared diagnostic yields of two specimen processing techniques, fixed slides (cytology) and formalin-fixed core samples (histology). Results were categorized as malignant, benign (infectious, inflammatory), normal nodal tissue, or inadequate sampling (nondiagnostic). Malignancy, a defined benign process, and normal lymph node were considered diagnostic.

Results: The diagnostic yield for benign processes was higher by histologic examination (n = 37) than in cytologic preparations (n = 22; p = 0.0064). The diagnostic yield was comparable in malignancy (p = 0.7530). The combination of both techniques provided a higher overall diagnostic rate: 84% (n = 148) by histology, 82% (n = 146) by cytology, and 89% (n = 158) using both. Using two techniques revealed discordance in 23% (n = 40), demonstrating that the use of one technique alone would have resulted in missed diagnoses.

Conclusions: Adding histologic analysis of tissue cores obtained by endobronchial ultrasound offers higher diagnostic accuracy than only cytologic preparation of needle aspirates. Histologic and cytologic methods offer comparable diagnostic rates for malignancy. However, diagnostic yield for benign conditions is higher using histologic examination. Together, histology and cytology provide fewer missed diagnoses than either individually. When using endobronchial ultrasound, it is ideal to routinely use both needle aspirate cytology and core biopsy histology.