A novel scoring system in endobronchial ultrasound-guided transbronchial needle aspiration

<bold>Background:</bold> Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is a minimally invasive diagnostic examination often requiring a deep sedation of the patient, who needs to be assisted by mechanical ventilation with addictional risks to the procedure. In order to reduce procedural time frames, it would be helpful to find a diagnostic tool able to improve the prediction of malignancy. <bold>Objective:</bold> Evaluating all the ultrasound (US) features predictive of malignancy in B-mode, Color-Doppler (CD) and Elastography, we aimed to assess their sensibility, sensitivity, diagnostic accuracy and both predictive values. <bold>Methods:</bold> We enrolled all the patients above 5 months with mediastinal and hilar adenopathy on chest tomography, candidates for diagnostic EBUS-TBNA (n=38) and prospectively recordered the following lymph node (n=99) US characteristics: size, shape, echogenicity, margins, central hilar structure, coagulation necrosis sign (CNS), CD vascularization pattern and elastography. These features were evaluated by 3 raters and then compared with histopathology. Statistical analysis was performed with IBM SPSS 26.0 software. <bold>Results:</bold> At the multivariate analysis 5 US features revealed to be independent predictors for malignancy (p<0,05): short axis>1 cm, round shape, heterogeneous echogenicity, distinct margins and CNS present. They were subjected to ROC analysis, obtaining a 5 items scoring system (scoring from 0 to 5; cut-off point of 3) with a negative predictive value of 91.40%. <bold>Conclusions:</bold> This novel score would eventually allow the ruling out of lymph nodes scoring<3, thus reducing procedural time and risks. Further studies are required to asses its reliability.