Abstract
BACKGROUND: Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is an innovative tool for diagnosing mediastinal diseases. We investigated the factors affecting the diagnostic yield of EBUS-TBNA and evaluated whether the effects of these factors (number of biopsies, core tissue acquisition rate, and diameter and volume of tissue) vary depending on computed tomography (CT) and/or positron emission tomography (PET)/CT results. METHODS: We retrospectively analyzed lung cancer patients who underwent EBUS-TBNA at Korea University Ansan Hospital (January 2019-December 2022). Patients in whom EBUS-TBNA failed and those with missing diameter or volume data and no imaging data interpretation were excluded. Subgroup analysis was performed by dividing the patients into None (no cancer detected on CT or PET/CT), Either (cancer detected on either CT or PET/CT), and Both (cancer detected on both CT and PET/CT) groups. RESULTS: In all, 228 patients were enrolled; 351 lymph node stations were analyzed. The median age of the patients was 69 years (male, 76.8%). Adenocarcinoma (28.5%) was the most common diagnosis. EBUS-TBNA was predominantly performed at station #4R (30.5%). Each examination involved two stations with a total procedure time of 30 minutes. An increased number of passes led to a higher diagnostic yield for EBUS-TBNA (P<0.001). Additionally, successful tissue sampling was associated with a large diameter (P=0.016) and volume (P=0.002) of the tissue. The effect of these factors was modified by imaging results. In the None and Either groups, an increase in the pass number was correlated with an increased diagnostic yield (adjusted P=0.003 and 0.007, respectively). However, in the Both group, it was not significant and remained at a suggestive level (P=0.304). The diameter and volume did not differ significantly across subgroups (adjusted P>0.05). CONCLUSIONS: Increasing the number of passes during EBUS-TBNA can maximize the diagnostic yield, especially when CT and/or PET/CT results are inconclusive.