Endoscopic Ultrasound Elastography (Strain Ratio) and Fractal-based Quantitative Analysis for the Mediastinal and Abdominal Lymphadenopathies

Endoscopic ultrasound (EUS) and EUS-guided tissue acquisition (EUS-TA) are minimally invasive techniques to establish the benign or malignant nature of lymph nodes (LNs) with a very high accuracy. In particular, for mediastinal LNs, a meta-analysis underlined that EUS features alone are not enough to reach a good sensitivity and specificity, while EUS-TA showed a slightly high sensitivity (88% vs. 85%) and a significantly high specificity (96% vs. 85%) for diagnosing the cause of LN enlargement. Considering all sites, according to the meta-analysis of Xu and colleagues, the assessment of LNs by EUS-elastography is a useful tool in differentiating benign and malignant LNs, with a sensitivity of 88% and a specificity of 85%. For lymphadenopathy of unknown origin, ESGE recommends performing EUS-TA (or alternatively endobronchial ultrasound \[EBUS\]-guided) if the patient management requires a pathological result and no superficial lymphadenopathy is easily accessible. Unfortunately, EUS-TA may sometimes be challenging or inconclusive, due to blood contamination or insufficient material. EUS elastography (EUS-E) is a non-invasive ultrasound technique that measures the hardness of tissues. The level of hardness of region of interest (ROI) can be evaluated using qualitative scores and/or quantitative methods (strain ratio; SR). To date, most of the studies on EUS-E have been carried out using the Hitachi ultrasound machine. In particular, Paterson performed a quantitative analysis considering as a pathological value cut-off, the SR ≥7.5 for EUS-E in the nodal staging of esophageal cancer. The study involved 53 LNs using cytology as gold standard: their data showed that EUS-E had a sensitivity of 83%, specificity of 96%, PPV of 95% and NPV of 86%. Recently, a new compact ultrasound processor, EU-ME2 (Olympus SE \& CO. KG, Hamburg, Germany) that includes an elastography software was developed although data regarding its application in LNs evaluation are still not availa