2D shear wave elastography combined with MELD improved prognostic accuracy in patients with acute-on-chronic hepatitis B liver failure

Objectives: To evaluate accuracy of two-dimensional shear wave elastography (2D SWE) and develop and validate a new prognostic score in predicting prognosis of acute-on-chronic liver failure (ACLF) patients. Methods: From 1 October 2013 to 30 September 2015, we consecutively enrolled 290 patients, sequentially collected data (including 2D SWE, ultrasound parameters, laboratory data and prognostic scores) and recorded patients’ outcome (recovering/steady or worsening) during a 90-day follow-up period. We evaluated ability of 2D SWE to predict outcomes of acute-on-chronic hepatitis B liver failure (ACLF-HBV) patients. We developed a new score (MELD-SWE, combining MELD and SWE values) for predicting mortality risk of ACLF-HBV in 179 patients in a derivation group, and validated in 111 patients. Results: 2D SWE values were higher in worsening patients than recovering/steady ones (p < 0.001). Accuracy of 2D SWE in predicting outcomes of ACLF-HBV was comparable to that of the MELD score (p = 0.441). MELD-SWE showed a significantly higher prognostic value than MELD in both derivation (AUROC, 0.80 vs. 0.76, p = 0.040) and validation (AUROC, 0.87 vs. 0.82, p = 0.018) group. Conclusions: The MELD-SWE score, combining MELD and SWE values, was superior to MELD alone for outcoming prediction in patients with ACLF-HBV. Key Points: • 2D SWE is a simple prognostic evaluation tool in patients with ACLF-HBV. • MELD-SWE was created in this study: 1.3×MELD + 0.3×2D SWE (kPa). • MELD-SWE score was superior to MELD alone for outcoming prediction in ACLF-HBV. • In this study, 46.8 was the optimal cut-off value of MELD-SWE score.