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

BACKGROUND AND AIMS: Metabolic dysfunction-associated steatotic liver disease (MASLD) imposes significant health care burdens. Early detection of advanced fibrosis and cirrhosis in MASLD is essential due to their unfavorable outcomes. This multilevel random-effects meta-analysis aimed to provide the best evidence for the diagnostic accuracy of 2-dimensional shear wave elastography in detecting liver fibrosis in biopsy-proven MASLD. APPROACH AND RESULTS: This study involves 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 February 26, 2024. Included studies reported liver stiffness measurement by 2-dimensional shear wave elastography and used histological diagnosis as the gold standard. A linear mixed-effects multiple thresholds model was employed, and summary estimates for sensitivity, specificity (Sp), and summary area under the receiver operator characteristic curve were computed. Twenty observational studies (SuperSonic Imagine, General Electric Healthcare, and 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 summary area under the receiver operator characteristic curve 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 cutoff 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 optimized cutoffs for fibrosis staging by 2-dimensional shear wave elastography in etiology-specific chronic liver diseases (MASLD), with excellent diagnostic performance, underscoring the potential for standardizing cutoff values.