TY - JOUR
T1 - Validation of Diagnostic Thresholds for Compensated Advanced Chronic Liver Disease Using Supersonic Shear Imaging
AU - Lee, Yeri
AU - Kim, Nam Hee
AU - Kang, Ji Hun
AU - Cho, Young Seo
AU - Kim, Yongsoo
AU - An, Jihyun
AU - Sohn, Joo Hyun
N1 - Publisher Copyright:
© RSNA, 2024.
PY - 2024/4
Y1 - 2024/4
N2 - Background: The Society of Radiologists in Ultrasound (SRU) has proposed thresholds for acoustic radiation force impulse techniques to diagnose compensated advanced chronic liver disease (cACLD). However, the diagnostic performance of these thresholds has not been extensively validated. Purpose: To validate the SRU thresholds in patients with chronic liver disease who underwent supersonic shear imaging and, if suboptimal diagnostic performance is observed, to identify optimal values for diagnosing cACLD. Materials and Methods: This retrospective single-center study included high-risk patients with chronic liver disease who had liver stiffness (LS) measurements and had undergone endoscopy or liver biopsy between January 2018 and December 2021. Patients were randomly allocated to test and validation sets. cACLD was defined as varices at endoscopy and/or severe fibrosis or cirrhosis at liver biopsy. The diagnostic performance of the SRU guidelines was evaluated, and optimal threshold values were identified using receiver operating characteristic (ROC) curve analysis. Results: A total of 1180 patients (median age, 57 years [IQR, 50–64 years]; 761 men), of whom 544 (46%) had cACLD, were included. With the SRU recommended thresholds of less than 9 kPa and greater than 13 kPa in the test set (n = 786), the sensitivity and specificity for ruling out and ruling in cACLD were 81% (303 of 374 patients; 95% CI: 77, 85) and 92% (380 of 412 patients; 95% CI: 89, 94), respectively. In ROC curve analysis, the identified optimal threshold values were less than 7 kPa and greater than 12 kPa, showing 91% sensitivity (340 of 374 patients; 95% CI: 88, 93) for ruling out cACLD and 91% specificity (373 of 412 patients; 95% CI: 87, 93) for ruling in cACLD, respectively. In the validation set (n = 394), the optimal thresholds showed 91% sensitivity (155 of 170 patients; 95% CI: 86, 95) and 92% specificity (206 of 224 patients; 95% CI: 88, 95). Conclusion: Compared with the SRU guidelines, the dual LS threshold values of less than 7 kPa and greater than 12 kPa were better for diagnosing cACLD.
AB - Background: The Society of Radiologists in Ultrasound (SRU) has proposed thresholds for acoustic radiation force impulse techniques to diagnose compensated advanced chronic liver disease (cACLD). However, the diagnostic performance of these thresholds has not been extensively validated. Purpose: To validate the SRU thresholds in patients with chronic liver disease who underwent supersonic shear imaging and, if suboptimal diagnostic performance is observed, to identify optimal values for diagnosing cACLD. Materials and Methods: This retrospective single-center study included high-risk patients with chronic liver disease who had liver stiffness (LS) measurements and had undergone endoscopy or liver biopsy between January 2018 and December 2021. Patients were randomly allocated to test and validation sets. cACLD was defined as varices at endoscopy and/or severe fibrosis or cirrhosis at liver biopsy. The diagnostic performance of the SRU guidelines was evaluated, and optimal threshold values were identified using receiver operating characteristic (ROC) curve analysis. Results: A total of 1180 patients (median age, 57 years [IQR, 50–64 years]; 761 men), of whom 544 (46%) had cACLD, were included. With the SRU recommended thresholds of less than 9 kPa and greater than 13 kPa in the test set (n = 786), the sensitivity and specificity for ruling out and ruling in cACLD were 81% (303 of 374 patients; 95% CI: 77, 85) and 92% (380 of 412 patients; 95% CI: 89, 94), respectively. In ROC curve analysis, the identified optimal threshold values were less than 7 kPa and greater than 12 kPa, showing 91% sensitivity (340 of 374 patients; 95% CI: 88, 93) for ruling out cACLD and 91% specificity (373 of 412 patients; 95% CI: 87, 93) for ruling in cACLD, respectively. In the validation set (n = 394), the optimal thresholds showed 91% sensitivity (155 of 170 patients; 95% CI: 86, 95) and 92% specificity (206 of 224 patients; 95% CI: 88, 95). Conclusion: Compared with the SRU guidelines, the dual LS threshold values of less than 7 kPa and greater than 12 kPa were better for diagnosing cACLD.
UR - https://www.scopus.com/pages/publications/85190342116
U2 - 10.1148/radiol.232188
DO - 10.1148/radiol.232188
M3 - Article
C2 - 38591973
AN - SCOPUS:85190342116
SN - 0033-8419
VL - 311
JO - Radiology
JF - Radiology
IS - 1
M1 - e232188
ER -