Performance of Endobronchial Ultrasound Elastography in the Differentiation of Malignant and Benign Mediastinal Lymph Nodes: Results in Real-life Practice

Abstract

Background: Little data exists regarding the performance of elastography in EBUS-TBNA. The aim of the study was to evaluate the elastography score proposed and previously published by Izumo, in particular its capacity to perfectly identify benign lymph node, and to discriminate malignant ones.

Methods: This study included patients undergoing EBUS-TBNA for mediastinal lymph nodes (LN). Before LN needle aspiration, an elastography was performed which allowed a color elastogram to be superimposed on the ultrasound image. Three blinded assessors classified these elastograms into 3 types using the score published by Izumo: type 1 (predominantly not blue), type 2 (partially blue, partially not blue), or type 3 (predominantly blue). These types were then compared with pathology results.

Results: A total of 217 LN (114 patients) were analyzed: histologic findings identified 97 benign LN (44.7% of the lymph nodes) and 120 malignant LN (55.3%). There were 44 elastographies (20.2%) that were classified as type 1, 90 elastographies (41.5%) classified as type 2, and 83 elastographies (38.3%) classified as type 3. Considering type 1 as benign and type 3 as malignant, sensitivity, specificity, positive predictive value, and negative predictive value were respectively 87.0%, 68.0% , 80.0% , and 77.0%. Ten (23%) of the 44 lymph nodes with a type 1 elastogram were malignant.

Conclusion: Elastography does not preclude performing TBNA of the lymph nodes. It does not preclude EBUS-TBNA when a type 1 elastogram pattern is found. All lymph nodes visualized should be sampled by EBUS-TBNA, regardless of elastography pattern.