Aims
Endoscopic ultrasound–guided shear wave elastography (EUS-SWE) allows quantitative liver stiffness assessment from both hepatic lobes during standard EUS. Its diagnostic accuracy for staging liver fibrosis remains incompletely defined. We performed systematic review and meta-analysis of cumulative data to evaluate the accuracy of EUS-SWE for detecting advanced fibrosis (F3–F4) compared with no/mild fibrosis (F0–F2).
Methods
We systematically searched PubMed, Embase, and Scopus from inception. 490records were identified, 161 underwent title/abstract screening and 156 were excluded;5 studies met inclusion criteria.
5 prospective studies of patients undergoing EUS-SWE with liver biopsy and/or validated non-invasive reference standards and reported diagnostic performance for fibrosis staging were included. EUS-SWE was performed, using 10 valid measurements per lobe and a reliability index (VsN) ≥60%. The primary outcome was the area under the ROC curve (AUROC) for advanced fibrosis (F3–F4 vs F0–F2). Where multiple lobes were reported, we used the best-performing lobe for the main analysis and summarized lobe-specific data separately. Because standard errors were not consistently reported, we calculated a pooled AUROC as the simple mean of study-level AUROCs and described ranges of sensitivity and specificity at authors’ optimal cut-offs.
Results
Five prospective studies included 323 patients (range 42–108 per study; mean age 48–63 years; BMI 26.8–40.7 kg/m²), predominantly MASLD, with smaller proportions of viral, autoimmune, or alcohol-related liver disease. EUS-SWE was technically successful in >95% of cases and rescued patients with unreliable or failed VCTE in two series. Four studies, (264 patients) reported AUROCs for advanced fibrosis (F3–F4 vs F0–F2). The reported AUROC for left lobe stiffness was 0.93 with sensitivity 0.90 and specificity 0.83 at a cut-off of 8.48 kPa. [1] Diehl et al. found AUROCs of 0.98 (right) and 0.95 (left) versus biopsy, with right-lobe sensitivity 0.60 and specificity 0.98 at 25.5 kPa [2]. Kohli et al. reported AUROCs of 0.80 (left) and 0.78 (right) with sensitivities 0.58–0.75 and specificities 0.86–0.93 at 18–24 kPa [3]. AbiMansour et al. observed AUROCs of 0.80 (right) and 0.73 (left) for advanced liver disease (stage ≥3) defined by biopsy, MRE, or transient elastography[4]. The pooled AUROC for advanced fibrosis across these studies was 0.87 indicating good overall diagnostic accuracy. Sensitivities ranged from 0.58 to 0.96 and specificities from 0.83 to 0.98 at study-specific thresholds. Cirrhosis-focused data from Del Valle et al. (n=59) showed AUROCs of 0.97 (right) and 0.99 (left) for biopsy-proven cirrhosis, with sensitivity 0.97 and specificity 0.90–0.97 at cut-offs of 10.7 and 14 kPa. [5] In Kohli and Wang, cirrhosis AUROCs ranged from 0.90 to 0.96 with sensitivities 0.89–1.00 and specificities 0.87–0.92.[1, 3]. Right- versus left-lobe analyses showed broadly comparable performance.[2, 4] Diehl and AbiMansour reported slightly higher AUROCs or sensitivity for the right lobe (0.98 vs 0.95; 0.80 vs 0.73, respectively), whereas Kohli and Del Valle observed marginally higher AUROCs for the left lobe (0.80 vs 0.78; 0.99 vs 0.97). [3, 5] Wang assessed only the left lobe [1]
Conclusions
EUS-SWE demonstrates good overall diagnostic accuracy for advanced fibrosis (pooled AUROC ~0.87) and excellent accuracy for cirrhosis (AUROCs often >0.95). Its performance is comparable or superior to VCTE, particularly in patients with obesity or unreliable transcutaneous measurements. EUS-SWE be integrated into routine EUS examinations as a comprehensive tool for non-invasive staging of liver fibrosis, especially when transabdominal elastography are impractical or inconclusive.