Slow-pull capillary technique versus suction technique in endobronchial ultrasound-guided transbronchial needle aspiration for diagnosing diseases involving hilar and mediastinal lymph node enlargement
Xin He,Yanjun Wu,Haoyan Wang,Ganggang Yu,Bo Xu,Nan Jia,Zhigang Yao
DOI: https://doi.org/10.1177/1753466620907037
2020-01-01
Therapeutic Advances in Respiratory Disease
Abstract:Background: Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is a widely used, safe, and accurate technique for obtaining pathological specimens to be used in the diagnosis of diseases involving lung hilar and mediastinal lymph node (LN) enlargement. However, application of the suction technique during EBUS-TBNA remains controversial. In addition, the effectiveness of the slow-pull capillary technique for the diagnosis of pancreatic masses was recently reported. The aim of this study was to compare the diagnostic accuracy of EBUS-TBNA using these two techniques. Methods: The accuracy, sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and availability of tissue cores of the suction and slow-pull capillary techniques were studied retrospectively in patients who underwent EBUS-TBNA for the diagnosis of diseases involving lung hilar and mediastinal LN enlargement. Results: A total of 97 patients with hilar and mediastinal LN enlargement underwent EBUS-TBNA; 30 patients underwent the suction technique, 56 patients underwent the slow-pull capillary technique, 5 patients underwent both techniques, and 6 patients had failed operations. The accuracy, sensitivity, specificity, NPV, PPV, and the number of tissue cores obtained with the suction and slow-pull capillary techniques were 66.67% versus 85.71% ( p = 0.039), 43.75% versus 85.42% ( p 0.05), 59.09% versus 50% ( p > 0.05), 87.5% versus 97.62% ( p > 0.05), and 19 versus 50 ( p = 0.004), respectively. In both univariate and multivariate analyses, the acquisition of tissue core was significantly associated with the diagnostic accuracy of EBUS-TBNA. Moreover, the slow-pull capillary technique was significantly associated with the acquisition of tissue core in EBUS-TBNA. There were no significant differences between the two groups in the blood contamination of samples. Conclusions: Use of the slow-pull capillary technique in EBUS-TBNA can significantly increase the accuracy related to the diagnosis of diseases involving hilar and mediastinal LN enlargement by improving the acquisition of tissue core. The reviews of this paper are available via the supplemental material section.
respiratory system