Repeatability of vibration-controlled transient elastography versus magnetic resonance elastography in patients with cirrhosis: A prospective study.

Abstract

Magnetic resonance elastography (MRE) and vibration-controlled transient elastography (VCTE) have the potential to assess disease progression; however, repeatability data in people with cirrhosis are lacking. We aimed to assess the effect of disease severity on measurement variability and contribute to the evidentiary basis for the qualification of repeating liver stiffness measurements (LSM) in practice and research. This prospective study included 49 adult participants (58.3% female) with cirrhosis who underwent same-day repeat LSM examinations. The primary outcome was the same-day, same-operator repeatability coefficient% (RC%) and the within-case coefficient of variation (wCV) for each modality. Secondary outcomes include the intra-class correlation coefficient (ICC). The relationship between measurement variability (interquartile for VCTE, standard deviation for MRE) and disease severity (mean liver stiffness) was evaluated by linear regression with the coefficient of determination R2 reported. Same-day repeat MRE and VCTE exams were prospectively conducted in 33 and 45 participants, respectively. The RC% appeared 82% higher for VCTE versus MRE (38% vs. 21%), with consistent findings in head-to-head analyses. The wCV for VCTE and MRE was 14% and 8% respectively, indicating VCTE has 75% higher within-subject measurement variation than MRE. ICC was excellent for LSM by VCTE (0.92) and MRE (0.96). Measurement variability increased with mean liver stiffness for VCTE (R2 = 0.78) and MRE (R2 = 0.93). Both VCTE and MRE demonstrated increased measurement variability with disease severity. However, MRE outperformed VCTE in terms of technical repeatability in patients with cirrhosis. These repeatability estimates may improve the qualification of NITs in practice.