Diagnostic utility of an algorithmic combined use of conventional and EBUS-TBNA in patients with mediastinal and/or hilar lymphadenopathy

Background: A debate is going on whether conventional TBNA(cTBNA) and EBUS-TBNA are truly competitive or complementary techniques. Aims and objectives: We performed a prospective study to investigate the diagnostic efficacy of an algorithm integrating cTBNA and EBUS-TBNA. Methods: cTBNA was performed for nodal stations(ns) 4,7 and 10 with size>1.5cm. EBUS-TBNA was performed for ns 2,11 irrespective of size and for ns 4,7 and 10 with a size≤1.5cm or when cTBNA was inconclusive. Mediastinoscopy, surgery or follow up was performed in non diagnostic cases. Results: 139ns were sampled in 120 patients. cTBNA was performed in 77ns (55.4%) and EBUS-TBNA in 62(44.6%). cTBNA sampled nodes(n) were: 4R(37n-48%), 7(27n-35%), 4L(7n-9.1%), 10R(5n-6.5%) and 10L(1n-1.3%). EBUS-TBNA sampled n were: 2R(4n-6.5%), 4R(20n-32.3%), 4L(6n-9.7%), 7(8n-12.9%), 10R(13n-21%), 10L(3n-4.8%), 11R(4n-6.5%), 11L(4n-6.5%). Cumulative diagnosis was set by both techniques in 115 ns (82.7%). EBUS was performed as a second procedure in 5 patients (6 stations) without setting an alternative diagnosis. Mediastinoscopy or thoracotomy was performed in 17 non-diagnostic stations (15 patients), resulting to alternative diagnosis in 12(70.6%). Follow up of 5 patients did not reveal an alternative diagnosis. Overall sensitivity, diagnostic accuracy and negative predictive value were: 93%, 94% and 68%. Conclusion: Following the proposed algorithm, we used EBUS-TBNA in only 44.6% of cases attaining an overall sensitivity, diagnostic accuracy and negative predictive value of 93%, 94% and 68% respectively. Our algorithm may be further investigated as it can lead to EBUS resources rationalization.