Retrospective analysis of radial EBUS outcome for the diagnosis of peripheral pulmonary lesion: sensitivity and complications

BACKGROUND: The purpose of the current study was to clarify the sensitivity and complication rate of the radial (endobronchial ultrasound, EBUS) without the use of guide-sheath (GS) and fluoroscopy for lung cancer (LC), by measuring the distance from the orifice of the bronchus to the pulmonary lesion, as well as to analyze factors that can predict the diagnostic outcome. MATERIALS AND METHODS: A total of 147 patients with peripheral pulmonary lesions (PPL) underwent radial EBUS-guided transbronchial biopsy (TBB) in between August 1, 2013, and August 31, 2014. We analyzed retrospectively radiological data, diagnostic work-up in everyday clinical settings, final diagnosis and complication rates, as well as factors influencing the diagnostic outcome. RESULTS: Around 63.9% of PPLs were visualized by ultrasound. A definitive malignant diagnosis was established in 39 patients (26.5%) using radial EBUS. In the remaining 108 patients, additional procedures were performed. We missed LC diagnosis in 40 cases that results in a sensitivity of 49%. For malignant lesions visualized by radial EBUS, the sensitivity was 60%, compared with 24% for not visualized lesions. For malignant lesions, logistic regression was performed to identify the factors that had significant influence on visualization of the lesion and on diagnostic yield. Logistic regression analysis showed significant odds ratios (OR) for visualization depending on location of the lesion; upper lobe lesions were identified more frequent with OR of 3.85 (95% CI 1.42 – 10.98, p=0.009). Size above 30 mm had a non-significant OR of 2.11 (95% CI 0.80−5.73, p=0.134) for visualization. Diagnostic yield was only significantly influenced by visualization with the radial EBUS, OR 3.70 (95% CI 1.35−11.02, p=0.014). Location (p=0.745) and size above 30 mm (p=0.308) showed no significant increase in diagnostic yield. Other lesion characteristics defined on computed tomography, such as distance to carina and pleura, did not show any significant influence on the diagnostic yield. The complications rate was low with three cases of pneumothorax. CONCLUSION: Radial EBUS has definitely its place in the diagnostic work-up of PPL, especially for the lesions that can be visualized by radial ultrasound. However, prospective randomized controlled studies are necessary to raise the diagnostic yield and to define factors that can predict the outcome, which will consequently enable selection of the ‘right’ patients for this diagnostic procedure.