A PROSPECTIVE PILOT STUDY ON THE RELIABILITY AND UTILITY OF EUS SHEAR WAVE ELASTOGRAPHY FOR LIVER AND SPLEEN STIFFNESS MEASUREMENT IN PREDICTING GASTROESOPHAGEAL VARICES IN PATIENTS WITH CHRONIC LIVER DISEASES

Background/aims:
Liver (LSM) and spleen stiffness measurements (SSM) are representative surrogate markers for cirrhosis and clinically significant portal hypertension, which can lead to development of gastroesophageal varices and life-threatening gastrointestinal bleeding. Endoscopic ultrasound shear wave elastography (EUS-SWE) has been newly developed as a real-time quantitative tissue stiffness assessment. Although liver and spleen are more readily accessible by EUS and less affected by patients’ body habitus as in transient elastography (TE), the reliability and clinical application of EUS-SWE LSM and SSM in chronic liver diseases has not been well investigated.

Methods:
This was a prospective pilot study using EUS-SWE in adults with liver cirrhosis. Patients had TE LSM and SSM by FibroScan®. Three endoscopists then independently performed EUS-SWE for 10 valid (defined as Vsn ≥ 70%) measurements each, after esophagogastroduodenoscopy (OGD) for variceal screening in the same session. The inter-observer agreements for LSM and SSM (in terms of Vs, m/s and Elastography, kPa) among the endoscopists were analyzed by the intraclass correlation coefficient (ICC). The area under receiving operating characteristics (AUROC) curve analysis was performed to assess the association of EUS-SWE SSM with presence of varices.

Results:
40 patients with chronic liver diseases of Child's Pugh Class A (chronic hepatitis B: 27, chronic hepatitis C: 2, non-alcoholic fatty liver diseases [NAFLD]: 6, alcoholic liver disease: 1, chronic hepatitis B and NAFLD: 4) and a mean body-mass-index of 25.6 kg/m² were studied. 72.5% of them had central obesity. The mean LSM by TE was 9.9 kPa. The mean procedure times for EUS-SWM LSM were 218 seconds (operator 1), 243 seconds (operator 2) and 192 seconds (operator 3), and that for EUS-SWM SSM were 208 seconds (operator 1), 261 seconds (operator 2) and 258 seconds (operator 3). The ICCs among the three endoscopists of EUS-SWE LSM were 0.71 (95% CI 0.51 – 0.83, p<0.001) for Vs and 0.72 (95% CI 0.53 – 0.84, p<0.001) for Elastography measurements, while that of EUS-SWE SSM for both were 0.82 (95% CI 0.68 – 0.90, p<0.001). After performing 12 cases, the ICC of EUS-SWM SSM among the three operators could reach over 0.8 while 24 cases were needed to increase the ICC of EUS-SWM LSM over 0.6 (Figure 1). The AUROCs of EUS-SSM to predict the presence of varices by one of the operators was up to 0.72 (95% CI 0.50 – 0.95, p =0.55) in terms of Elastography and 0.71 (95% CI 0.48 – 0.93, p = 0.07) for Vs. (Figure 2) Taking 3.16 m/s and 29.8 kPa as the cut-off Vs and Elastography values respectively, EUS-SWE SSM had 80% sensitivity and 61.3% specificity.

Conclusions:
Both EUS-SWE LSM and SSM had good inter-operator reliability. EUS-SWE SSM might be a potential marker that correlates with gastroesophageal varices. Further prospective data is warranted.