Diagnostic accuracy of 2D-SWE ultrasound for liver fibrosis assessment in MASLD: A multilevel random effects model meta-analysis

Background & Aims: Metabolic dysfunction-associated steatotic liver disease (MASLD) imposes significant healthcare burdens. Early detection of advanced fibrosis and cirrhosis in MASLD is essential due to their unfavourable outcomes. This multi-level random-effects meta-analysis aimed to provide the best evidence for the diagnostic accuracy of two-dimensional shear wave elastography (2D-SWE) in detecting liver fibrosis in biopsy-proven MASLD. Approach & Results: Systematic search in PubMed/MEDLINE, Embase, Scopus, Web of Science, LILACS, and Cochrane Library electronic databases for full-text articles published in any language up to the 26th of February 2024. Included studies reported liver stiffness measurement (LSM) by 2D-SWE and used histological diagnosis as gold standard. A linear mixed-effects multiple thresholds model was employed, and summary estimates for sensitivity (Se), specificity (Sp), and summary area under the curve (sAUC) were computed. 20 observational studies (SuperSonic Imagine, General Electric Healthcare, Canon Medical Systems) fulfilled the inclusion criteria, comprising 2223 participants with biopsy-proven MASLD. The prevalence of mild fibrosis (F1), significant fibrosis (F2), advanced fibrosis (F3), and cirrhosis (F4) was 30.0%, 18.5%, 17.9%, and 10.9%, respectively. The sAUCs [95%CI] in detecting ≥F1, ≥F2, ≥F3, and F4 for all ultrasound machines considered together were 0.82 [0.16-0.98], 0.82 [0.76-0.88], 0.86 [0.77-0.93], and 0.89 [0.80-0.95], respectively. The optimal cut-off values were 6.432 kPa for ≥F1, 8.174 kPa for ≥F2, 9.418 kPa for ≥F3, and 11.548 kPa for F4, respectively. Conclusions: Our meta-analysis identified optimised cut-offs for fibrosis staging by 2D-SWE in etiology-specific chronic liver diseases (MASLD), with excellent diagnostic performance, underscoring the potential for standardising cut-off values.