Diagnostic value of two liver imaging techniques combined with serological models for hepatitis B virus-related liver fibrosis
10.3760/cma.j.issn.1000-6680.2018.05.003
- VernacularTitle: 两种肝脏成像技术联合血清型模型对乙型肝炎相关肝纤维化的诊断价值
- Author:
Liu WU
1
;
Dongying XUE
;
Jie ZHANG
;
Weizheng LI
;
Lin SUN
;
Ji LI
;
Wenhong ZHANG
;
Lingyun SHAO
Author Information
1. Department of Infectious Diseases, Putuo Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai 200062, China
- Publication Type:Journal Article
- Keywords:
FibroScan;
Acoustic radiation force impulse imaging;
APRI;
FIB-4;
Liver fibrosis
- From:
Chinese Journal of Infectious Diseases
2018;36(5):270-276
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To compare the diagnostic efficacy of transient elastography (TE) FibroScan and acoustic radiation force impulse imaging (ARFI) combined with serological models including aspartate aminotransferase-to-platelet ratio (APRI) and fibrosis-4 (FIB-4) in hepatitis B virus-related fibrosis.
Methods:Sixty-seven patients with chronic HBV infection from October 2014 to May 2017 in Department of Infectious Diseases, Putuo Hospital were enrolled. Both FibroScan and ARFI were conducted in all patients together with serological tests. According to the golden standard of pathology results, the diagnosis values of FibroScan, ARFI combined with APRI or FIB-4 were compared as noninvasive assessment for liver fibrosis. Data with homogeneity of variance were tested by t test, and data with heterogeneity of variance were tested by Mann-Whitney U test.
Results:Based on the pathology results, the receiver operating characteristic (ROC) areas under the curve (AUC) of APRI, FIB-4, FibroScan and ARFI in diagnosis of hepatic fibrosis ≥S2 were 0.752, 0.612, 0.885, and 0.850, respectively. The AUC of ROC curve in diagnosis of hepatic fibrosis ≥S3 were 0.746, 0.733, 0.851, and 0.863, respectively. The AUC of ROC curve in diagnosis of hepatic fibrosis ≥S4 were 0.782, 0.705, 0.962 and 0.981, respectively. Combined liver imaging technique and serological tests, such as APRI with FibroScan, APRI with ARFI, FIB-4 with FibroScan or FIB-4 with ARFI, the AUC of ROC curve in the 4 groups in diagnosis of hepatic fibrosis ≥S2 were 0.887, 0.861, 0.893, and 0.853, respectively; in the diagnosis of hepatic fibrosis ≥S3 were 0.873, 0.871, 0.900, and 0.875, respectively; and in diagnosis of hepatic fibrosis ≥S4 were 0.952, 0.981, 0.969, and 0.981, respectively. FibroScan and ARFI were positively correlated with liver inflammation (r=0.467, P=0.000; r=0.371, P=0.002) and jaundice (r=0.424, P=0.000; r=0.0.312, P=0.01), while negatively correlated with platelet (r=-0.331, P=0.006; r=-0.312, P=0.01). The AUC of ROC curve of FibroScan, ARFI and their combination with serological model were significantly increased compared with the single serological model (all P<0.05).
Conclusions:Serological models such as APRI and FIB-4 as well as liver imaging techniques such as FibroScan and ARFI are all valuable in assessment of hepatic fibrosis, while FibroScan and ARFI have better diagnostic value. ARFI is convenient to application for its integration with the ordinary ultrasound system. The sensitivity and specificity for diagnosis of hepatic fibrosis could be improved by combining serological model with FibroScan or ARFI. Combination of APRI and ARFI show the highest accuracy in diagnosis of hepatic fibrosis. Combination of serological models and transient elastic liver imaging is recommended for assessment and follow-up of HBV-related fibrosis.
