Understanding confounding factors allows for accurate interpretation of liver stiffness measurements by ElastQ, a novel 2D shear wave elastography technique

David Jm Bauer,Annalisa De Silvestri,Laura Maiocchi,Ambra Raimondi,Ruxandra Mare,Mattias Mandorfer,Ioan Sporea,Theresa Müllner-Bucsics,Giovanna Ferraioli,Thomas Reiberger
DOI: https://doi.org/10.1055/a-2329-2801
2024-08-08
Abstract:Purpose: Liver stiffness measurement (LSM) using vibration-controlled transient elastography (VCTE) or two-dimensional shear wave elastography (2D-SWE) is recommended to assess the risk of liver fibrosis and advanced chronic liver disease. Even though both techniques measure liver stiffness, their numerical results often diverge. Confounders and reliability criteria for 2D-SWE have not been systematically investigated. Materials and methods: We prospectively recruited participants with paired LSM by VCTE and the novel 2D-SWE technique ElastQ (Philips) in three European tertiary centers. The following parameters were recorded: sex, age, body mass index (BMI), etiology, laboratory markers of liver damage and function, as well as cholestasis, LSM by VCTE and controlled attenuation parameter (CAP), interquartile range (IQR)/median for VCTE-LSM and ElastQ-LSM, and the skin-to-liver capsule distance. Results: We included 840 participants: 447 (53.2%) males; median age 57.0 [IQR:19.0] years; median BMI 25.4 [6.0] kg/m2; median VCTE-LSM 7.25 [9.2] kPa; median ElastQ-LSM 6.7 [5.4] kPa. On uni- and multivariable modeling (adjusted for LSM), we found that the discrepancy increased with liver stiffness and markers of disease severity. Skin-to-liver capsule distance and BMI affected VCTE-LSM more compared to ElastQ-LSM and significantly increased the discordance between the two measurements. Conclusion: The discrepancy of ElastQ-LSM to VCTE-LSM increases with liver stiffness and disease severity. BMI and skin-to-liver capsule distance increase the discrepancy between VCTE- and ElastQ-LSM but affect ElastQ-LSM less. The quality criterion IQR/median ≤ 30% indicates reliable ElastQ-LSM.
What problem does this paper attempt to address?