Scholarlink

Background <p>Liver stiffness measurement (LSM) is a non-invasive surrogate for hepatic fibrosis that informs risk stratification and management. Vibration-controlled transient elastography (VCTE; FibroScan®) is well established. Endoscopic ultrasound–guided shear wave elastography (EUS-SWE) offers the potential to assess LSM during routine EUS; however, its diagnostic performance remains uncertain.</p> Methods <p>In this prospective, single-center diagnostic accuracy study, consecutive adults without biliary obstruction who were&#xa0;undergoing EUS additionally underwent LSM measurement from&#xa0;the right hepatic lobe using EUS-SWE (Olympus ME3) and same-day VCTE. VCTE thresholds defined advanced chronic liver disease (ACLD; &gt; 15 kPa), cirrhosis (&gt; 12 kPa), and advanced fibrosis (&gt; 8 kPa). EUS-SWE performance was evaluated using correlation, Bland–Altman agreement, and area under the&#xa0;receiver-operating characteristic&#xa0;curve (AUROC).</p> Results <p>Among 146 patients (median VCTE-LSM 6.05 kPa, range 3–53.3), 57, 41, and 31 had LSM ≥ 8, ≥ 12, and ≥ 15&#xa0;kPa, respectively. EUS-SWE showed moderate correlation with VCTE (Spearman&#xa0;ρ = 0.65, p &lt; 0.001) and reported higher values by 34.1% (95% limits of&#xa0;agreement:-64.8%&#xa0;to&#xa0;133.1%) without proportional bias on Bland–Altman analysis. EUS–SWE showed excellent accuracy for screening relevant VCTE-LSM thresholds with AUROC (95% CI) 0.89 (0.81–0.97) for ACLD, 0.89 (0.82–0.96) for cirrhosis, and 0.85 (0.79–0.92) for advanced fibrosis. Optimal EUS-SWE threshold for VCTE-defined ACLD was 12.9 kPa (sensitivity 93%, specificity 77%), representing the optimal screening cutoff; a regression-derived threshold of 21.4 kPa (sensitivity 67%, specificity 93%) represents the mathematical equivalent EUS-SWE value to VCTE ≥ 15 kPa, appropriate for direct modality comparison. Accuracy&#xa0;for ACLD declined monotonously with increasing variability (IQR/LSM%): AUROC&#xa0;0.97&#xa0;(for&#xa0;&lt; 30%), 0.86 (30–45%), 0.83 (45–60%), and 0.78 (≥ 60%).</p> Conclusion <p>EUS-SWE yields LSM values approximately 34% higher than VCTE-LSM; the two modalities are not interchangeable and require modality-specific thresholds. EUS-SWE provides accurate prediction of VCTE-defined fibrosis thresholds, particularly when IQR/LSM% ≤ 45%, and represents a promising adjunct for opportunistic liver stiffness assessment during routine EUS.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

EUS-Guided Shear Wave Elastography in a Real-World Cohort: Diagnostic Accuracy and the Impact of Measurement Variability

  • 期刊论文

  • 发表日期：

    2026年7月27日

查看全文
  • Umang Arora,
  • Shivam Khare,
  • Shrihari Anikhindi,
  • Ashiesh Khandelwal,
  • Randeep Rana,
  • Ashish Kumar,
  • Praveen Sharma,
  • Naresh Bansal,
  • Munish Sachdeva,
  • Piyush Ranjan,
  • Mandhir Kumar,
  • Anil Arora

摘要

Background

Liver stiffness measurement (LSM) is a non-invasive surrogate for hepatic fibrosis that informs risk stratification and management. Vibration-controlled transient elastography (VCTE; FibroScan®) is well established. Endoscopic ultrasound–guided shear wave elastography (EUS-SWE) offers the potential to assess LSM during routine EUS; however, its diagnostic performance remains uncertain.

Methods

In this prospective, single-center diagnostic accuracy study, consecutive adults without biliary obstruction who were undergoing EUS additionally underwent LSM measurement from the right hepatic lobe using EUS-SWE (Olympus ME3) and same-day VCTE. VCTE thresholds defined advanced chronic liver disease (ACLD; > 15 kPa), cirrhosis (> 12 kPa), and advanced fibrosis (> 8 kPa). EUS-SWE performance was evaluated using correlation, Bland–Altman agreement, and area under the receiver-operating characteristic curve (AUROC).

Results

Among 146 patients (median VCTE-LSM 6.05 kPa, range 3–53.3), 57, 41, and 31 had LSM ≥ 8, ≥ 12, and ≥ 15 kPa, respectively. EUS-SWE showed moderate correlation with VCTE (Spearman ρ = 0.65, p < 0.001) and reported higher values by 34.1% (95% limits of agreement:-64.8% to 133.1%) without proportional bias on Bland–Altman analysis. EUS–SWE showed excellent accuracy for screening relevant VCTE-LSM thresholds with AUROC (95% CI) 0.89 (0.81–0.97) for ACLD, 0.89 (0.82–0.96) for cirrhosis, and 0.85 (0.79–0.92) for advanced fibrosis. Optimal EUS-SWE threshold for VCTE-defined ACLD was 12.9 kPa (sensitivity 93%, specificity 77%), representing the optimal screening cutoff; a regression-derived threshold of 21.4 kPa (sensitivity 67%, specificity 93%) represents the mathematical equivalent EUS-SWE value to VCTE ≥ 15 kPa, appropriate for direct modality comparison. Accuracy for ACLD declined monotonously with increasing variability (IQR/LSM%): AUROC 0.97 (for < 30%), 0.86 (30–45%), 0.83 (45–60%), and 0.78 (≥ 60%).

Conclusion

EUS-SWE yields LSM values approximately 34% higher than VCTE-LSM; the two modalities are not interchangeable and require modality-specific thresholds. EUS-SWE provides accurate prediction of VCTE-defined fibrosis thresholds, particularly when IQR/LSM% ≤ 45%, and represents a promising adjunct for opportunistic liver stiffness assessment during routine EUS.