Agreement Between 2-Dimensional Shear Wave and Transient Elastography Values for Diagnosis of Advanced Chronic Liver...

Inter-platform variation in liver stiffness evaluation (LSE) could hinder dissemination and clinical implementation of new ultrasound methods. We aimed to determine whether measurements of liver stiffness by bi-dimensional shear wave elastography (2D-SWE) with a Supersonic Imagine apparatus are comparable to those made by vibration-controlled transient elastography (VCTE).

We collected data from 1219 consecutive patients with chronic liver disease who underwent LSE by VCTE and 2D-SWE (performed by blinded operators), on the same day, at a single center in France from September 2011 through June 2019. We assessed the ability of liver stiffness value distributions and 2D-SWE performances to identify patients with compensated advanced chronic liver disease (cACLD) according to the Baveno VI criteria, based on VCTE cut-off values.

VCTE and 2D-SWE values correlated (Pearson's correlation coefficient, 0.882; P < .0001; Lin concordance coefficient, 0.846; P < .0001). The median stiffness values were 6.7 kPa with VCTE (interquartile range, 4.8-11.6 kPa) and 7.1 kPa with 2D-SWE (interquartile range, 5.4-11.1 kPa) (P = .736). 2D-SWE values were slightly higher in the low percentiles and lower in the high percentiles; the best match with VCTE values were at approximately 7-9 kPa. The area under the curve of 2D-SWE for identifying of VCTE values below 10 was 0.964 (95% CI, 0.952-0.976) and for VCTE values above 15 kPa was 0.976 (95% CI, 0.963-0.988), with Youden index-associated cut-off values of 9.5 and 13kPa and best accuracy cut-off values of 10 kPa and 14 kPa, respectively. A 2D-SWE cut-off value of 10 kPa detected VCTE values below 10k Pa with 92% sensitivity, 87% specificity, and 91% accuracy.

Measurement of liver stiffness by VCTE or 2D-SWE produces comparable results. 2D-SWE accurately identifies patients with cACLD according to the Baveno VI criteria based on VCTE cut-off values. A 10 kPa 2D-SWE cut-off value can be used to rule out cACLD.