Chen H, Shen Y, Wu SD, Zhu Q, Weng CZ, Zhang J, Wang MX, Jiang W. Diagnostic role of transient elastography in patients with autoimmune liver diseases: A systematic review and meta-analysis. World J Gastroenterol 2023; 29(39): 5503-5525 [PMID: 37900994 DOI: 10.3748/wjg.v29.i39.5503]
Corresponding Author of This Article
Wei Jiang, MD, PhD, Academic Research, Chief Doctor, Professor, Department of Gastroenterology and Hepatology, Zhongshan Hospital (Xiamen), Fudan University, No. 666 Jinhu Road, Huli District, Xiamen 361015, Fujian Province, China. jiang.wei@zs-hospital.sh.cn
Research Domain of This Article
Gastroenterology & Hepatology
Article-Type of This Article
Systematic Reviews
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Hong Chen, Sheng-Di Wu, Cheng-Zhao Weng, Jun Zhang, Mei-Xia Wang, Wei Jiang, Department of Gastroenterology and Hepatology, Zhongshan Hospital (Xiamen), Fudan University, Xiamen 361015, Fujian Province, China
Hong Chen, Yue Shen, Sheng-Di Wu, Qin Zhu, Wei Jiang, Department of Gastroenterology and Hepatology, Zhongshan Hospital of Fudan University, Shanghai 200032, China
Hong Chen, Yue Shen, Sheng-Di Wu, Qin Zhu, Wei Jiang, Shanghai Institute of Liver Diseases, Fudan University Shanghai Medical College, Shanghai 200032, China
Author contributions: Jiang W conceived and designed the study; Chen H and Shen Y contributed to acquiring the data; Wang MX and Weng CZ contributed statistical analysis support; Chen H, Shen Y, Wu SD and Zhu Q contributed to analysis and interpretation of the data; Chen H wrote the manuscript; Wu SD and Zhang J contributed to manuscript revision; All authors read and approved the submitted version and are accountable for all aspects of the work.
Supported byNatural Science and Technology Major Project of Fujian Province, No. 2021D033; Natural Science Foundation of Shanghai, No. 20ZR1410900; Medical Innovation Project of Fujian Province, No. 2022CXB020; National Science and Technology Major Project, No. 2017ZX 10203202-003-002.
Conflict-of-interest statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
PRISMA 2009 Checklist statement: The authors have read the PRISMA 2009 Checklist, and the manuscript was prepared and revised according to the PRISMA 2009 Checklist.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Wei Jiang, MD, PhD, Academic Research, Chief Doctor, Professor, Department of Gastroenterology and Hepatology, Zhongshan Hospital (Xiamen), Fudan University, No. 666 Jinhu Road, Huli District, Xiamen 361015, Fujian Province, China. jiang.wei@zs-hospital.sh.cn
Received: July 2, 2023 Peer-review started: July 2, 2023 First decision: August 30, 2023 Revised: September 9, 2023 Accepted: October 11, 2023 Article in press: October 11, 2023 Published online: October 21, 2023 Processing time: 109 Days and 3.6 Hours
Abstract
BACKGROUND
Noninvasive methods have been developed to detect fibrosis in many liver diseases due to the limits of liver biopsy. However, previous studies have focused primarily on chronic viral hepatitis and nonalcoholic fatty liver disease. The diagnostic value of transient elastography for autoimmune liver diseases (AILDs) is worth studying.
AIM
To compare the diagnostic accuracy of imaging techniques with serum biomarkers of fibrosis in AILD.
METHODS
The PubMed, Cochrane Library and EMBASE databases were searched. Studies evaluating the efficacy of noninvasive methods in the diagnosis of AILDs [autoimmune hepatitis (AIH), primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC)] were included. The summary area under the receiver operating characteristic curve (AUROC), diagnostic odds ratio, sensitivity and specificity were used to assess the accuracy of these noninvasive methods for staging fibrosis.
RESULTS
A total of 60 articles were included in this study, and the number of patients with AIH, PBC and PSC was 1594, 3126 and 501, respectively. The summary AUROC of transient elastography in the diagnosis of significant fibrosis, advanced fibrosis and cirrhosis in patients with AIH were 0.84, 0.88 and 0.90, respectively, while those in patients with PBC were 0.93, 0.93 and 0.91, respectively. The AUROC of cirrhosis for patients with PSC was 0.95. However, other noninvasive indices (aspartate aminotransferase to platelet ratio index, aspartate aminotransferase/alanine aminotransferase ratio, fibrosis-4 index) had corresponding AUROCs less than 0.80.
CONCLUSION
Transient elastography exerts better diagnostic accuracy in AILD patients, especially in PBC patients. The appropriate cutoff values for staging advanced fibrosis and cirrhosis ranged from 9.6 to 10.7 and 14.4 to 16.9 KPa for PBC patients.
Core Tip: Onset of autoimmune liver diseases is frequently insidious, and immune cell infiltration and continuous inflammation drive hepatic fibrosis, which gradually progresses to cirrhosis, causing poorer long-term outcomes. Liver biopsy as the reference standard is an invasive procedure. Thus, repeated biopsies are difficult to implement. Consequently, appropriate noninvasive methods are essential to dynamically monitor the degree of liver fibrosis. Our meta-analysis compared the diagnostic accuracy of imaging techniques with serum biomarkers of fibrosis in autoimmune liver diseases.
Citation: Chen H, Shen Y, Wu SD, Zhu Q, Weng CZ, Zhang J, Wang MX, Jiang W. Diagnostic role of transient elastography in patients with autoimmune liver diseases: A systematic review and meta-analysis. World J Gastroenterol 2023; 29(39): 5503-5525
The incidence of autoimmune liver diseases (AILDs), including autoimmune hepatitis (AIH), primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), and multiple overlap syndromes, a group of autoimmune diseases associated with the liver and bile duct is increasing[1,2]. Onset is frequently insidious, with nonspecific symptoms. Immune cell infiltration and continuous inflammation drive hepatic fibrosis, which gradually progresses to cirrhosis, causing poorer long-term outcomes in patients[3-5]. Accordingly, accurate identification of high-risk patients for such conditions is essential in clinical care to guide timely treatment and delay disease progression.
For years, liver biopsy has been recognized as the reference standard for the assessment of liver fibrosis. However, biopsy area restrictions, sampling errors, and interobserver variability may affect the diagnostic accuracy[6,7]. Moreover, because biopsy is an invasive procedure with potentially hazardous complications ranging from pain to more severe events and even death, many patients are reluctant to undergo repeat biopsies[8,9]. Consequently, an increasing number of studies have focused on noninvasive methods to identify the ideal approach for dynamically monitoring the degree of liver fibrosis[10].
In recent years, some noninvasive methods, including biochemical tests and imaging techniques, have been widely developed, including the aspartate aminotransferase to platelet ratio index (APRI), aspartate aminotransferase/alanine aminotransferase (ALT) ratio (AAR), fibrosis-4 index (FIB-4), red cell distribution width to platelet ratio (RPR), Mac-2 binding protein (M2BP), platelet count to spleen diameter (PC/SD) ratio, transient elastography (TE), shear wave elastography (SWE), acoustic radiation force impulse (ARFI), magnetic resonance spectroscopy (MRS) and magnetic resonance elastography (MRE). Previous studies have validated that elastography is a reliable method with a diagnostic accuracy higher than that of blood tests for staging liver fibrosis in chronic viral hepatitis[11-13], nonalcoholic fatty liver disease[14] and AIH[15]; however, no studies have explored the diagnostic accuracy of noninvasive methods for the other two types of AILDs (PBC and PSC).
Therefore, the present meta-analysis aimed to compare the diagnostic accuracy of biochemical tests and imaging techniques for detecting liver fibrosis in patients with AILD, determine whether the same noninvasive methods show different diagnostic values in the three types of AILDs and recommend appropriate cutoff values for different fibrosis stages.
MATERIALS AND METHODS
Literature search strategy
Studies on the diagnosis of AILD published between January 2006 and December 2022 were searched in PubMed, Cochrane Library and EMBASE databases using the following keywords: AIH, PSC, PBC, liver fibrosis, TE, SWE, MRE, APRI, FIB-4 and AAR. The detailed search strategy is presented in Supplementary Table 1.
Study selection criteria
Original studies that fulfilled the following criteria were enrolled: (1) Studies with patient populations with AIH, PBC or PSC with discrete data that could be separately extracted from the mixed liver disease study cohort; (2) Studies in which liver biopsy was used as the gold standard to assess fibrosis based on the Metavir score or another score that could be converted to the Metavir score; (3) Studies assessing the performance and utility of APRI, AAR, FIB-4, RPR, M2BP, ARFI, PC/SD ratio, TE, SWE, MRE or MRS for staging liver fibrosis; and (4) Studies directly reporting the true-positive, false-positive, false-negative and true-negative values or provided data by which they could be calculated to allow the construction of a 2 × 2 table for each test.
The following studies were excluded: (1) Studies exploring the prognostic value of liver stiffness measurement (LSM) for patients with AILD; (2) Animal experiments, reviews, protocols, guidelines, case reports or meta-analyses; (3) Studies on liver fibrosis due to other etiologies, including nonalcoholic fatty liver disease, chronic hepatitis B, or chronic hepatitis C; and (4) Studies without sufficient data for further analysis or with the same or overlapping group of participants.
Data extraction and quality assessment
Two investigators (Chen H and Shen Y) independently evaluated the eligibility and quality of the included studies and extracted the data. Any disagreements were resolved by a senior researcher (Wu SD). We collated the following parameters in Microsoft Excel 2010: authors; year of publication; country; study period and design; pathological type; diagnostic methods; sample size; patient characteristics [age, sex, body mass index (BMI), ALT level, treatment condition]; quality of liver biopsy; and performance of the index test, including cutoff values, sensitivity, specificity, positive predictive value, negative predictive value, and area under the receiver operating characteristic curve (AUROC). Two reviewers (Chen H and Shen Y) independently assessed the risk of bias using the Quality Assessment of Diagnostic Accuracy Studies-2 tool[16].
Data synthesis and statistical analysis
According to the Metavir, Batts-Ludwig and Scheuer scores, liver fibrosis was classified into five stages (F0, F1, F2, F3 and F4), whereas there were seven stages according to the Ishak score. Given that Shiha et al[17] proposed that an Ishak score of 3 corresponds to METAVIR score of F2, significant fibrosis (SF), advanced fibrosis (AF) and cirrhosis were defined as stages F2-F4, F3-F4, and F4, respectively. For the data analysis, we constructed 2 × 2 contingency tables with true-positive, false-positive, false-negative and true-negative values based on data directly extracted from the original studies or calculated from indirect variables (sensitivity, specificity and sample size). A bivariate random effects model was subsequently applied to calculate the diagnostic test accuracy variables, including summary sensitivity, summary specificity, positive likelihood ratio, negative likelihood ratio and diagnostic odds ratio (DOR) with their associated 95% confidence intervals (95%CIs). We also performed meta-analyses using hierarchical models to produce summary ROC curves, from which we obtained summary AUROC values to evaluate the diagnostic accuracy of the different noninvasive methods. The method was considered to have excellent accuracy if the summary AUROC value was above 0.90, moderate accuracy if it was greater than 0.80, and poor accuracy if it was less than 0.80[18].
The heterogeneity was assessed using multiple methods. Spearman’s correlation coefficient was calculated to evaluate the threshold heterogeneity of the included studies, whereas Cochran’s Q and I2 values were used to assess nonthreshold heterogeneity. If an I2 value > 50% or P < 0.05 indicated distinct statistical heterogeneity, a random effects model was used to combine the data. A fixed effect model was chosen when the I2 value ≤ 50% or P ≤ 0.05. However, the number of original studies was sufficient to perform a meta-regression to explore the potential heterogeneity of certain index tests. In addition, we conducted a subgroup analysis according to the sample size, treatment conditions and cutoff value. Deeks’ funnel plots were used to evaluate the possible publication bias. The meta-analysis was performed using Stata 12.0, Reviewer Manager Version 5.3 and Meta-Disc Version 1.4.
RESULTS
Characteristics of the included studies and patients
The study selection process is illustrated in Figure 1. A total of 1386 studies were retrieved through our search strategy, of which 427 were excluded as duplicates and 602 were removed following the screening of titles, abstracts and reviews. The remaining 355 potentially eligible studies were selected for further evaluation. Of these, 60 articles were included in the evaluation and analysis. Among them, 22, 29 and 6 studies were regarding AIH, PBC and PSC, respectively (2 studies focused on both AIH and PBC[19,20], while 1 study focused on both PBC and PSC[21]). In total, they included 11 noninvasive index tests. The basic characteristics of the included studies are presented in Table 1. We selected articles published between 2006 and 2022, of which 46 (76.7%) were published between 2016 and 2022. There were 31 (51.7%) retrospective studies, 17 (28.3%) prospective studies, 10 (16.7%) unknown studies and 2 studies with both designs. Most included studies were conducted in Asia (28 studies) or Europe (24 studies). A total of 27 studies utilized the Metavir score, 8 studies used the Batts-Ludwig score, 17 studies used the Scheuer score, 5 studies used the Ishak score, and 3 studies used the Ludwig-Scheuer score.
A total of 1594, 3126 and 501 patients with AIH, PBC, and PSC, respectively, were included to analyze the diagnostic performance of noninvasive methods in staging liver fibrosis. Most patients with AIH and PBC were female (72.4% and 87.6%, respectively). In contrast, patients with PSC were predominantly male (73.7%). The average ages of patients with AIH, PBC and PSC were 47.0, 55.2 and 41.5 years, respectively. Patients with AIH (160.29 IU/mL) had higher ALT levels than patients with PBC (69.81 IU/mL).
Quality assessment of the included studies
The results of the quality assessment based on the Quality Assessment of Diagnostic Accuracy Studies-2 scale for all 60 eligible studies are shown in Figure 2 and Supplementary Figure 1. Regarding patient selection, eight studies had an unclear risk of bias owing to the lack of information on whether patients were enrolled randomly or consecutively. Regarding the index test, four studies were determined to have an unclear risk of bias because the results of the index test were interpreted without blinded information on the results of the reference standard. Likewise, 22 studies were regarded as having an unclear risk of bias because the results of the reference standard were interpreted without blinded information regarding the results of the index test. In terms of flow and timing, two studies were considered high-risk because not every subject received a reference standard, while 19 studies were considered unclear risk because of an unknown time interval between the index and reference tests.
Figure 2 Quality assessment of included studies by Quality Assessment of Diagnostic Accuracy Studies-2.
Risk of bias and applicability concerns graph.
Performance of noninvasive methods in diagnosing SF (F ≥ 2)
Diagnosis of SF for AIH: Fifteen studies (n = 1001) evaluated eight noninvasive methods for detecting SF in patients with AIH. Of these, five (n = 459), two (n = 129), five (n = 459), nine (n = 523) and three (n = 234) studies focused on APRI, AAR, FIB-4, TE and SWE separately, whereas only one study each utilized the ARFI, PC/SD ratio and RPR.
The APRI had moderate summary sensitivity (exceeding 70%) with poor summary specificity (less than 50%), whereas the FIB-4 had the opposite result (Table 2). Interestingly, TE had a relatively greater diagnostic performance than the other laboratory tests, with summary sensitivity and specificity values of 0.82 and 0.73, respectively, and cutoff values ranging from 5.8–7.0 KPa. The summary sensitivity of SWE (0.89; 95%CI: 0.83–0.93) was significantly higher than that of the other six noninvasive methods and slightly greater than that of TE (0.83; 95%CI: 0.78–0.87).
Table 2 Summary sensitivities, specificities, positive likelihood ratio and negative likelihood ratio of noninvasive methods at various diagnostic thresholds for prediction of significant fibrosis, advanced fibrosis and cirrhosis in autoimmune liver diseases patients.
Disease
Diagnostic model/Stage
Cutoff values
No. of patients (n)
Summary sensitivity
Summary specificity
Summary PLR
Summary NLR
AIH
APRI
SF
0.27-0.70
2 (195)
0.80 (0.72-0.86)
0.35 (0.23-0.48)
1.46 (0.55-3.89)
0.36 (0.061-2.09)
0.88-1.55
3 (264)
0.74 (0.68-0.80)
0.51 (0.38-0.63)
1.52 (1.18-1.96)
0.50 (0.36-0.69)
AF
0.38-0.90
4 (379)
0.86 (0.81-0.90)
0.48 (0.41-0.56)
1.60 (1.18-2.15)
0.33 (0.24-0.47)
1.12-3.40
6 (538)
0.80 (0.72-0.86)
0.35 (0.23-0.48)
0.57 (0.51-0.64)
0.68 (0.62-0.73)
Cirrhosis
0.55-1.81
3 (330)
0.65 (0.56-0.74)
0.47 (0.40-0.54)
1.49 (0.93-2.39)
0.62 (0.46-0.83)
1.85-2.00
3 (213)
0.70 (0.57-0.81)
0.73 (0.65-0.79)
2.48 (1.75-3.52)
0.42 (0.28-0.62)
AAR
SF
0.72-0.93
2 (129)
0.67 (0.57-0.77)
0.68 (0.49-0.83)
2.22 (1.32-3.72)
0.46 (0.23-0.90)
AF
0.76-1.18
7 (532)
0.61 (0.54-0.68)
0.72 (0.66-0.77)
2.01 (1.59-2.53)
0.58 (0.46-0.73)
Cirrhosis
0.94-1.40
3 (213)
0.61 (0.47-0.74)
0.83 (0.76-0.88)
3.31 (1.96-5.59)
0.49 (0.36-0.69)
FIB-4
SF
1.95-2.90
3 (303)
0.64 (0.58-0.71)
0.71 (0.61-0.81)
2.20 (1.48-3.27)
0.50 (0.38-0.65)
3.20-5.07
2 (156)
0.60 (0.51-0.69)
0.77 (0.60-0.90)
2.66 (1.44-4.92)
0.52 (0.39-0.69)
AF
1.75-2.37
6 (476)
0.66 (0.59-0.72)
0.63 (0.57-0.69)
2.05 (1.56-2.70)
0.40 (0.20-0.81)
3.21-5.60
5 (486)
0.47 (0.41-0.54)
0.73 (0.67-0.79)
1.83 (1.24-2.71)
0.68 (0.50-0.92)
Cirrhosis
2.21-3.40
5 (440)
0.75 (0.67-0.82)
0.56 (0.50-0.62)
2.06 (1.53-2.77)
0.37 (0.19-0.71)
6.44
1 (103)
0.68
0.64
1.88
0.51
TE
SF
5.80-7.00
7 (423)
0.83 (0.78-0.87)
0.73 (0.65-0.80)
2.89 (2.23-3.76)
0.23 (0.12-0.42)
9.10-10.05
2 (100)
0.77 (0.67-0.86)
0.94 (0.70-1.00)
7.65 (1.66-35.32)
0.18 (0.02-1.47)
AF
8.18-9.00
3 (286)
0.80 (0.72-0.87)
0.80 (0.73-0.86)
4.09 (2.64-6.33)
0.24 (0.17-0.35)
10.40-12.10
4 (174)
0.73 (0.60-0.83)
0.93 (0.86-0.97)
7.67 (2.89-20.31)
0.27 (0.12-0.61)
Cirrhosis
11.00-12.67
4 (213)
0.89 (0.82-0.94)
0.88 (0.81-0.93)
6.89 (4.38-10.85)
0.14 (0.09-0.23)
16.00-19.00
3 (147)
0.88 (0.74-0.96)
0.97 (0.92-0.99)
22.08 (5.35-91.22)
0.16 (0.08-0.33)
2D-SWE
SF
8.20-10.00
3 (234)
0.89 (0.83-0.93)
0.72 (0.59-0.83)
3.25 (1.67-6.32)
0.17 (0.11-0.28)
AF
12.20-15.80
3 (234)
0.82 (0.73-0.89)
0.79 (0.72-0.86)
3.92 (2.79-5.52)
0.24 (0.13-0.44)
Cirrhosis
14.30-19.30
4 (297)
0.83 (0.74-0.90)
0.86 (0.81-0.91)
5.85 (4.09-8.37)
0.21 (0.13-0.34)
PBC
APRI
SF
0.26-1.20
4 (584)
0.84 (0.80-0.87)
0.63 (0.56-0.70)
1.98 (1.54-2.55)
0.34 (0.23-0.51)
AF
0.3.0-0.75
8 (858)
0.62 (0.57-0.68)
0.54 (0.50-0.58)
1.39 (1.09-1.79)
0.68 (0.48-0.98)
0.93-2.00
7 (731)
0.73 (0.68-0.78)
0.68 (0.64-0.72)
2.68 (1.80-3.97)
0.46 (0.36-0.58)
Cirrhosis
0.65-1.39
6 (852)
0.75 (0.67-0.83)
0.51 (0.48-0.55)
2.19 (1.38-3.50)
0.31 (0.10-0.99)
AAR
SF
0.92-1.00
3 (323)
0.69 (0.61-0.76)
0.56 (0.48-0.63)
1.61 (1.33-1.95)
0.52 (0.31-0.87)
AF
0.81-1.01
5 (559)
0.54 (0.47-0.62)
0.73 (0.68-0.77)
2.15 (1.52-3.03)
0.63 (0.44-0.91)
Cirrhosis
1.00-1.10
4 (407)
0.81 (0.70-0.90)
0.77 (0.72-0.82)
4.55 (1.98-10.49)
0.28 (0.10-0.79)
FIB-4
SF
1.39-3.90
3 (462)
0.85 (0.81-0.89)
0.77 (0.69-0.83)
2.89 (2.10-3.98)
0.26 (0.10-0.66)
AF
2.05-2.63
7 (865)
0.77 (0.72-0.81)
0.57 (0.53-0.61)
1.95 (1.51-2.52)
0.31 (0.16-0.61)
2.81-4.60
6 (431)
0.63 (0.55-0.71)
0.80 (0.75-0.85)
3.25 (1.78-5.94)
0.49 (0.30-0.81)
Cirrhosis
2.05-4.60
6 (852)
0.87 (0.80-0.93)
0.61 (0.58-0.65)
2.79 (1.92-4.07)
0.16 (0.05-0.52)
TE
SF
5.90-8.80
4 (402)
0.81 (0.76-0.85)
0.95 (0.89-0.98)
10.51 (2.03-54.36)
0.23 (0.12-0.44)
16.00
1 (44)
0.94
0.81
4.90
0.07
AF
6.75-7.60
4 (377)
0.80 (0.73-0.86)
0.81 (0.76-0.86)
4.19 (2.35-7.46)
0.19 (0.05-0.79)
9.60-10.70
3 (317)
0.91 (0.84-0.95)
0.82 (0.77-0.87)
5.68 (2.55-12.69)
0.12 (0.07-0.21)
11.90-17.90
3 (180)
0.75 (0.60-0.86)
0.94 (0.88-0.97)
11.76 (2.29-60.48)
0.22 (0.06-0.80)
Cirrhosis
11.40-14.40
3 (256)
0.84 (0.69-0.93)
0.94 (0.90-0.97)
13.46 (7.66-23.65)
0.19 (0.10-0.38)
15.60-25.10
3 (227)
0.90 (0.74-0.98)
0.93 (0.89-0.96)
22.8 (0.81-639.69)
0.12 (0.04-0.34)
RPR
AF
0.10-0.14
4 (362)
0.49 (0.40-0.58)
0.89 (0.84-0.92)
4.27 (2.22-8.22)
0.59 (0.47-0.74)
M2BP
AF
1.00-1.40
4 (370)
0.68 (0.59-0.77)
0.80 (0.75-0.85)
4.26 (1.82-9.96)
0.32 (0.14-0.75)
PSC
TE
SF
8.80
2 (121)
0.76 (0.62-0.87)
0.88 (0.79-0.95)
6.34 (3.25-12.37)
0.29 (0.18-0.46)
AF
9.60
3 (177)
0.82 (0.70-0.91)
0.83 (0.75-0.89)
4.75 (2.21-10.19)
0.15 (0.02-1.04)
Cirrhosis
13.70-14.40
4 (207)
0.82 (0.68-0.91)
0.89 (0.83-0.94)
7.46 (3.74-14.88)
0.25 (0.15-0.43)
Diagnosis of SF for PBC: Thirteen studies (n = 1389) evaluated nine noninvasive methods for diagnosing SF in patients with PBC. Among them, four (n = 584), three (n = 323), three (n = 462), two (n = 87), five (n = 446) and two (n = 210) studies focused on APRI, AAR, FIB-4, ARFI, TE and SWE, respectively; however, only one study each utilized the PC/SD ratio, MRE and M2BP.
As shown in Table 2, the APRI and FIB-4 index had relatively good summary sensitivities of 0.84 and 0.85, respectively, with mild summary specificities of 0.63 and 0.77, respectively. The corresponding values for sensitivity and specificity of the AAR were poor (0.69, 0.56). In contrast, both the summary sensitivity (0.81; 95%CI: 0.76–0.85) and specificity (0.95; 95%CI: 0.89–0.98) of TE were significantly higher than those of the other five noninvasive methods for predicting SF with cutoff values ranging from 5.9–8.8 KPa.
Furthermore, Table 3 shows that the summary DORs of APRI, FIB-4 and TE were 3.9 (95%CI: 2.1–7.3), 5.1 (95%CI: 3.1–8.5) and 16.8 (95%CI: 8.8–32.2), respectively, in patients with AIH, while the summary DORs of APRI and TE were 6.3 (95%CI: 3.5–11.2) and 74.5 (95%CI: 12.2–455.5), respectively, in patients with PBC. Additionally, the summary AUROC value of TE in patients with PBC (0.93, 95%CI: 0.91–0.95) was relatively higher than that of TE in patients with AIH (0.84, 95%CI: 0.80–0.87) but significantly higher than that of FIB-4 (0.74) and APRI (0.67) in patients with AIH and APRI (0.77) in patients with PBC (Table 3 and Figure 3).
Figure 3 The summary receiver operating characteristic curve plots of transient elastography in autoimmune liver disease patients.
A and B: Transient elastography (TE) for detecting significant fibrosis in autoimmune hepatitis (AIH, A) and primary biliary cholangitis (PBC, B) patients; C and D: TE for detecting advanced fibrosis in AIH (C) and PBC (D) patients; E-G: TE for detecting cirrhosis in AIH (E), PBC (F) and primary sclerosing cholangitis (G) patients; H: Shear wave elastography for detecting cirrhosis in AIH patients. AUC: Area under the curve; SENS: Sensitivity; SPEC: Specificity; SROC: Summary receiver operating characteristic.
Table 3 Summary area under the receiver operator curve and diagnostic odds ratio of noninvasive methods for prediction of significant fibrosis, advanced fibrosis and cirrhosis in autoimmune liver diseases patients.
Disease
Diagnostic model/Stage
No. of studies (patients)
AUROC (95%CI)
DOR (95%CI)
AIH
APRI
SF
4 (383)
0.67 (0.63-0.71)
3.87 (2.1-7.3)
AF
10 (917)
0.71 (0.67-0.75)
3.85 (2.8-5.3)
Cirrhosis
6 (543)
0.71 (0.67-0.75)
3.77 (2.2-6.4)
FIB-4
SF
5 (459)
0.74 (0.70-0.78)
5.11 (3.1-8.5)
AF
11 (962)
0.73 (0.69-0.76)
4.04 (2.4-6.8)
Cirrhosis
6 (543)
0.72 (0.68-0.76)
5.48 (2.4-12.6)
TE
SF
9 (523)
0.84 (0.80-0.87)
16.83 (8.8-32.2)
AF
7 (460)
0.88 (0.85-0.90)
25.14 (9.7-65.3)
Cirrhosis
7 (415)
0.90 (0.87-0.92)
91.77 (40.1-201.2)
AAR
AF
6 (410)
0.73 (0.69-0.77)
4.94 (3.2-7.8)
2D-SWE
Cirrhosis
4 (297)
0.91 (0.89-0.94)
30.68 (15.7-59.9)
PBC
APRI
SF
4 (584)
0.77 (0.73-0.80)
6.27 (3.5-11.2)
AF
15 (1589)
0.70 (0.66-0.74)
3.67 (2.3-5.9)
Cirrhosis
6 (852)
0.83 (0.79-0.86)
14.55 (1.9-113.8)
FIB-4
AF
13 (1296)
0.79 (0.75-0.82)
7.13 (4.0-12.8)
Cirrhosis
6 (852)
0.88 (0.85-0.91)
29.79 (5.9-150.3)
TE
SF
5 (446)
0.93 (0.91-0.95)
74.45 (12.2-455.5)
AF
10 (880)
0.93 (0.90-0.95)
41.84 (19.3-91.0)
Cirrhosis
6 (483)
0.91 (0.88-0.93)
134.83 (33.0-551.8)
AAR
AF
6 (559)
0.74 (0.70-0.78)
4.13 (2.0-8.6)
Cirrhosis
4 (407)
0.88 (0.84-0.90)
25.29 (9.0-70.9)
RPR
AF
4 (362)
0.53 (0.49-0.58)
7.98 (4.0-15.8)
M2BP
AF
4 (370)
0.86 (0.82-0.88)
13.17 (4.1-42.4)
PSC
TE
Cirrhosis
4 (207)
0.95 (0.93-0.97)
70.59 (15.4-322.7)
Performance of noninvasive methods in diagnosing AF (F ≥ 3)
Diagnosis of AF for AIH: Twenty studies (n = 1435) evaluated 11 noninvasive methods for detecting AF in patients with AIH. Among them, 10 (n = 917), 7 (n = 532), 11 (n = 962), 7 (n = 460), 3 (n = 234) and 2 (n = 191) studies focused on APRI, AAR, FIB-4, TE, SWE and RPR, respectively; however, only one study each utilized the ARFI, MRS, PC/SD ratio, MRE and M2BP methods.
As shown in Table 2, with a cutoff value of 8.2–9.0 KPa, both the summary sensitivity and specificity exceeded 80% when TE was used for predicting AF, whereas with a cutoff value of 10.4–12.1 KPa, there was a better summary specificity (0.93; 95%CI: 0.86–0.97) with a mild summary sensitivity (0.73; 95%CI: 0.60–0.83). Regarding SWE, MRE and the PC/SD ratio, the summary sensitivity and specificity also exceeded 80%. The specificity of MRE was 1.00, but only one study assessed it[22]. For AAR and FIB-4 index, there was a relatively modest summary specificity (< 0.80) and poor summary sensitivity (< 0.60).
Diagnosis of AF for PBC: Twenty-eight studies (n = 2737) evaluated 11 noninvasive methods for detecting AF in patients with PBC. Of these, 15 (n = 1589), 6 (n = 559), 13 (n = 1296), 2 (n = 97), 10 (n = 874), 5 (n = 362), 4 (n = 370) and 2 (n = 210) studies were focused on APRI, AAR, FIB-4, ARFI, TE, RPR, M2BP and SWE, respectively. Only one study each utilized the methods of MRE and RLR.
As shown in Table 2, TE had a good summary sensitivity and specificity (0.91, 0.82) with a cutoff value of 9.6–10.7 KPa, while RPR and M2BP had good summary specificity (0.89 and 0.80, respectively) with poor summary sensitivity (0.49 and 0.68, respectively). Regardless of the cutoff values, the summary sensitivities and specificities of the AAR, APRI and FIB-4 were less than 0.80.
Furthermore, Table 3 shows that the summary DORs of AAR, APRI, FIB-4 and TE were 4.9 (95%CI: 3.2–7.8), 3.9 (95%CI: 2.8–5.3), 4.0 (95%CI: 2.4–6.8) and 25.1 (95%CI: 9.7–65.3), respectively, in patients with AIH and 4.1 (95%CI: 2.0–8.6), 3.7 (95%CI: 2.3–6.0), 7.1 (95%CI: 4.0–12.8) and 41.8 (95%CI: 19.3–91.0), respectively, in patients with PBC. Moreover, the summary DORs of RPR and M2BP in patients with PBC were 8.0 (95%CI: 4.0–15.8) and 13.2 (95%CI: 4.1–42.4), respectively. As shown in Table 3 and Figure 3, the summary AUROC value of TE for detecting AF was 0.88 (95%CI: 0.85–0.90) and 0.93 (95%CI: 0.90–0.95) in patients with AIH and PBC, respectively. The value of M2BP was 0.86 (95%CI: 0.82–0.88) in patients with PBC, whereas the summary AUROC values for AAR, APRI and FIB-4 were less than 0.80 in both patients with AIH and PBC, and the value for RPR in patients with PBC was less than 0.60.
Performance of noninvasive methods in diagnosing cirrhosis (F = 4)
Diagnosis of cirrhosis for AIH: Sixteen studies (n = 1076) evaluated ten noninvasive methods for detecting cirrhosis in patients with AIH. Of these, six (n = 543), three (n = 213), six (n = 543), two (n = 82), seven (n = 415) and four (n = 297) studies focused on APRI, AAR, FIB-4, ARFI, TE and SWE, respectively. Only one study each utilized the PC/SD ratio, MRE, RPR and M2BP.
As shown in Table 2, the summary sensitivities and specificities of APRI and FIB-4 were less than 75%, and those of AAR were 0.61 and 0.83, respectively. Moreover, the summary sensitivity and specificity of TE (cutoff value ranging from 11.0–12.7 KPa) were significantly higher for predicting cirrhosis, with 0.89 (95%CI: 0.82–0.94) and 0.88 (95%CI: 0.81–0.93), respectively, while the summary sensitivity and specificity of SWE (0.83, 0.86) were close to those of TE. Surprisingly, the summary specificity dramatically rose to 0.97 (95%CI 0.92–0.99) with a cutoff value ranging from 16.0–19.0 KPa.
Diagnosis of cirrhosis for PBC: Sixteen studies (n = 1568) evaluated nine noninvasive methods for detecting cirrhosis in patients with PBC. Among them, six (n = 852), four (n = 407), six (n = 852), six (n = 483), two (n = 210) and two (n = 194) studies focused on APRI, AAR, FIB-4, TE, SWE and M2BP, respectively. However, only one study utilized the ARFI and MRE methods.
As listed in Table 2, the summary sensitivities of APRI, AAR and FIB-4 for predicting cirrhosis were 0.75, 0.81 and 0.87, respectively, and their corresponding summary specificities were 0.51, 0.77 and 0.61, respectively. In contrast, TE had higher summary sensitivity (0.90; 95%CI: 0.74–0.98) and specificity (0.93; 95%CI: 0.89–0.96) with a cutoff value ranging from 15.6–25.1 KPa.
Diagnosis of cirrhosis for PSC: Four studies (n = 207) evaluated TE as a predictor of cirrhosis in patients with PSC. Because the diagnosis of PSC does not rely on liver biopsy, few related studies have been conducted. As listed in Table 2, the summary sensitivity and specificity of TE were 0.82 (95%CI: 0.68–0.91) and 0.89 (95%CI: 0.83–0.94), respectively.
Furthermore, Table 3 shows that the summary DORs of APRI and FIB-4 were 3.8 (95%CI: 2.2–6.4) and 5.5 (95%CI: 2.4–12.6) in patients with AIH and 14.6 (95%CI: 1.9–113.8) and 29.8 (95%CI: 5.9–150.3) in patients with PBC. In addition, the summary DOR of TE was highest in patients with PBC, with values of 91.8 (95%CI: 40.1–201.2), 134.8 (95%CI: 33.0–551.8) and 70.6 (95%CI: 15.4–322.7) in patients with AIH, PBC and PSC, respectively. The summary AUROC values of TE for detecting cirrhosis in patients with AIH, PBC and PSC were 0.90 (95%CI: 0.87–0.92), 0.91 (95%CI: 0.88–0.93) and 0.95 (95%CI: 0.93–0.97), respectively, while the summary AUROC values for APRI and FIB-4 were less than 0.80 in patients with AIH and 0.90 in patients with PBC (Table 3 and Figure 3).
Methodological heterogeneity, subgroup analysis and publication bias
As shown in Table 4, threshold heterogeneity was observed only in APRI F2 in both patients with AIH and PBC, whereas nonthreshold heterogeneity was observed in most groups (Figure 4, Supplementary Figures 2 and 3). Because meta-regression to explore the source of heterogeneity requires the number of original studies to exceed 10, we only conducted a meta-regression for AF in patients with AIH and PBC. The heterogeneity of APRI, FIB-4 and TE accuracy was mainly affected by the cutoff value with regard to specificity, whereas FIB-4 and TE were affected by sample size with regard to sensitivity, according to the meta-regression analysis (Supplementary Figure 4).
Figure 4 Forest plots of diagnostic odds ratio of transient elastography in autoimmune liver disease patients.
A and B: Transient elastography (TE) for detecting significant fibrosis in autoimmune hepatitis (AIH, A) and primary biliary cholangitis (PBC, B) patients, respectively; C and D: TE for detecting advanced fibrosis in AIH (C) and PBC (D) patients; E-G: TE for detecting cirrhosis in AIH (E), PBC (F) and primary sclerosing cholangitis (G) patients. 95%CI: 95% confidence interval.
Table 4 Heterogeneity of all the included studies.
Fibrosis stage
Threshold heterogeneity
Non-threshold heterogeneity
r
P value
I2 (%)
P value
AIH
TE
SF
0.176
0.651
82.9
0
AF
-0.429
0.337
93.8
0
Cirrhosis
0.321
0.482
56.59
0.03
APRI
SF
1.0
0
62.34
0.05
AF
0.717
0.02
71.94
0
Cirrhosis
0.714
0.111
90.93
0
FIB-4
SF
0.70
0.188
46.81
0.11
AF
0.627
0.039
98.29
0
Cirrhosis
-0.029
0.957
88.98
0
AAR
AF
0.857
0.014
36.46
0.16
SWE
Cirrhosis
0
1.0
61.97
0.05
PBC
TE
SF
-0.10
0.873
99.99
0
AF
0.195
0.590
100
0
Cirrhosis
-0.726
0.027
93.77
0
APRI
SF
1.0
0
99.9
0
AF
0.209
0.454
100
0
Cirrhosis
-0.657
0.156
100
0
FIB-4
AF
0.418
0.156
100
0
Cirrhosis
0.029
0.957
100
0
AAR
AF
0.60
0.208
96.05
0
Cirrhosis
0.40
0.60
95.37
0
M2BP
AF
0
1.0
99.24
0
RPR
AF
0
1.0
94.49
0
PSC
TE
Cirrhosis
0.80
0.20
50.07
0.11
Subgroup analyses of TE according to sample size, cutoff value and treatment status are shown in Table 5. Because of the limited data, we only conducted an analysis for posttreatment combined with original data for pretreatment (Supplementary Table 2).
Table 5 Subgroup analysis of sample size and treatment status in prediction of significant fibrosis, advanced fibrosis and cirrhosis in autoimmune hepatitis and primary biliary cholangitis patients.
Disease
Parameter
Stage
Subgroup
Sensitivity (95%CI)
Specificity (95%CI)
AUROC (95%CI)
AIH
Sample size
SF
n < 50
0.83 (0.55-0.95)
0.82 (0.65-0.92)
0.85 (0.82-0.88)
n > 50
0.84 (0.73-0.91)
0.77 (0.63-0.87)
0.87 (0.84-0.90)
AF
n < 50
0.78 (0.54-0.91)
0.91 (0.78-0.96)
0.92 (0.89-0.94)
n > 50
0.78 (0.68-0.86)
0.87 (0.73-0.94)
0.88 (0.84-0.90)
Cirrhosis
n < 50
0.90 (0.65-0.98)
0.92 (0.74-0.98)
0.96 (0.94-0.97)
n > 50
0.88 (0.82-0.93)
0.93 (0.86-0.96)
0.92 (0.89-0.94)
Treatment status
SF
Post
0.78 (0.44-0.94)
0.76 (0.60-0.86)
0.79 (0.75-0.82)
AF
Post
0.83 (0.66-0.93)
0.96 (0.84-0.99)
0.93 (0.91-0.95)
Cirrhosis
Post
0.91(0.77-0.97)
0.97 (0.73-1.00)
0.94 (0.91-0.95)
Cutoff value
SF
5.80-6.27
0.87 (0.81-0.92)
0.69 (0.60-0.77)
0.86 (0.83-0.89)
6.40-9.10
0.82 (0.75-0.88)
0.89 (0.74-0.96)
0.92 (0.89-0.94)
AF
9.00-11.00
0.83 (0.69-0.91)
0.92 (0.73-0.98)
0.88 (0.85-0.91)
8.18-12.10
0.77 (0.71-0.83)
0.85 (0.80-0.89)
0.88 (0.85-0.90)
Cirrhosis
11.00-12.67
0.89 (0.82-0.94)
0.88 (0.81-0.93)
0.92 (0.94-0.96)
11.00-19.00
0.88 (0.82-0.93)
0.92 (0.88-0.95)
0.90 (0.87-0.92)
PBC
Sample size
SF
n < 100
0.81 (0.48-0.95)
0.78 (0.60-0.89)
0.82 (0.79-0.86)
n > 100
0.83 (0.68-0.92)
0.98 (0.74-1.00)
0.97 (0.95-0.98)
AF
n < 100
0.90 (0.85-0.94)
0.88 (0.78-0.94)
0.91 (0.88-0.93)
n > 100
0.81 (0.64-0.91)
0.88 (0.75-0.94)
0.91 (0.88-0.93)
Cirrhosis
n < 100
0.82 (0.67-0.91)
0.94 (0.80-0.99)
0.86 (0.82-0.89)
n > 100
0.91 (0.76-0.97)
0.97 (0.90-0.99)
0.94 (0.92-0.96)
Treatment status
SF
Post
0.89 (0.70-0.97)
0.98 (0.41-1.00)
0.97 (0.95-0.98)
AF
Post
0.85 (0.68-0.94)
0.92 (0.63-0.99)
0.93 (0.91-0.95)
Cirrhosis
Post
0.90 (0.71-0.97)
0.96 (0.74-1.00)
0.94 (0.92-0.96)
Cutoff value
AF
6.75-7.60
0.80 (0.73-0.86)
0.81 (0.76-0.86)
0.88 (0.85-0.91)
9.60-10.70
0.91 (0.84-0.95)
0.82 (0.77-0.87)
0.92 (0.89-0.94)
11.90-17.90
0.75 (0.60-0.86)
0.94 (0.88-0.97)
0.93 (0.91-0.95)
Cirrhosis
11.40-14.40
0.84 (0.69-0.93)
0.94 (0.90-0.97)
0.96 (0.94-0.97)
14.40-16.90
0.88 (0.72-0.97)
0.99 (0.96-1.00)
0.99 (0.98-1.00)
Deeks’ funnel plot of these noninvasive methods was generated to assess publication bias. There was a publication bias for APRI in detecting SF (P = 0.06) and cirrhosis (P = 0.08) in patients with PBC but not in other methods for detecting SF, AF and cirrhosis (Figure 5, Supplementary Figure 5). Moreover, no publication bias was observed for any noninvasive method in patients with AIH (Figure 5, Supplementary Figure 6).
Figure 5 Deeks’ funnel plot asymmetry test for publication bias of transient elastography in autoimmune liver disease patients.
A and B: Transient elastography (TE) for detecting significant fibrosis in autoimmune hepatitis (AIH, A) and primary biliary cholangitis (PBC, B) patients; C and D: TE for detecting advanced fibrosis in AIH (C) and PBC (D) patients; E-G: TE for detecting cirrhosis in AIH (E), PBC (F) and primary sclerosing cholangitis (G) patients.
DISCUSSION
In our review, a total of 60 studies (including 1594, 3126 and 501 patients with AIH, PBC and PSC, respectively) were included to evaluate the diagnostic accuracy of noninvasive methods for predicting SF, AF and cirrhosis in patients with AILDs. TE had excellent accuracy with summary AUROC values of 0.93, 0.93 and 0.91 for SF, AF and cirrhosis, respectively, in patients with PBC, while TE had a moderate to excellent accuracy of 0.84, 0.88 and 0.90, respectively, in patients with AIH. Moreover, the summary AUROC was 0.95 for cirrhosis in patients with PSC. In contrast, other noninvasive methods, such as AAR, APRI, FIB-4 and RPR, had poor accuracy, with summary AUROC values of < 0.80. In addition, the pooled sensitivity and specificity of TE were higher than those of the other noninvasive methods. Our results indicated that LSM using TE had a better diagnostic performance for staging hepatic fibrosis in AILDs, especially in patients with PBC. Moreover, our results showed that TE had mostly higher specificity and relatively low sensitivity for the diagnosis of AILDs. Koizumi et al[23] found that TE had high sensitivity and relatively low specificity for the diagnosis of PBC. However, the optimal cutoff values were higher and the range was wider than those in other studies, indicating that different optimal cutoff values may have an effect on diagnostic accuracy.
Meta-regression analysis is a reliable method for screening heterogeneity. In our study, the sample size, cutoff values, prevalence of SF and scoring system provided heterogeneity in summarizing the test results, consistent with previous studies[24,25]. We conducted subgroup analyses based on the sample size and cutoff values. Our results revealed that TE had a better predictive effect in a larger sample of patients with PBC. LSM by TE is the best surrogate marker for staging in SF and AF with a cutoff ranging from 6.4–9.1 KPa and 9.0–11.0 KPa, respectively, in patients with AIH and staging in AF and cirrhosis with a cutoff ranging from 9.6–10.7 KPa and 14.4–16.9 KPa, respectively, in patients with PBC.
Several previous studies have demonstrated that inflammation in the liver (reflected by elevated ALT levels)[26] and extrahepatic cholestasis (reflected by total bilirubin)[27] may increase the stiffness value, causing a decrease in the diagnostic accuracy of TE, whereas ALT and bilirubin levels decline after treatment. Because a limited number of studies have reported results for the ALT subgroup, we only conducted a subgroup analysis of treatment conditions, which showed that the diagnostic accuracy for staging liver fibrosis was comparable between pretreatment and posttreatment in patients with both PBC and AIH. In other words, this may indicate that ALT levels have no significant effect on diagnostic accuracy. Meanwhile, two scoring systems [International Autoimmune Hepatitis Group (IAIHG) 1999 and IAIHG 2008] proposed by Granito et al[28] for the diagnosis of AIH are not interchangeable. According to our subgroup analysis regarding diagnostic criteria, the IAIHG 2008 showed diagnostic accuracy comparable to that of the IAIHG 1999 in distinguishing patients with AIH (Supplementary Tables 3 and 4). However, due to the limited number of studies, further investigation is required to confirm these results.
In addition, some studies have shown that other imaging technologies, including two-dimensional-SWE (2D-SWE)[29-32], MRE[22,33] and ARFI[19,34], also perform well in staging liver fibrosis (Supplementary Table 5). Further, 2D-SWE had excellent accuracy, with a summary AUROC of 0.91 for cirrhosis in patients with AIH (Table 3). In comparison, our findings indicated that 2D-SWE and ARFI had good accuracy with higher sensitivity, specificity and AUROC for AF and cirrhosis in patients with PBC, while the AUROC of MRE was higher in patients with AIH. Interestingly, compared with TE, 2D-SWE produces a two-dimensional grayscale image so that interference from the gallbladder, ascites and large tubular structures in the liver can be effectively avoided. However, the number of studies on 2D-SWE, MRE and ARFI included in our analysis was small. Indeed, the diagnostic accuracies of 2D-SWE and MRE require further studies with larger sample sizes to determine the best method for staging fibrosis in patients with AILDs.
However, the overlap syndrome, one of the AILDs, also deserves attention because it exhibits significantly higher rates of various complications, progresses to cirrhosis more rapidly and has a poor treatment response to ursodeoxycholic acid[35,36]. Hence, the development of noninvasive methods is beneficial for this disease. Wu et al[37] reported that the AUROCs of TE for SF, AF and cirrhosis were 0.837, 0.910 and 0.996, respectively. Yan et al[38] reported that the AUROCs of SWE were 0.91, 0.97 and 0.96, respectively. These results show that noninvasive imaging techniques have excellent accuracy for overlap syndrome, although more studies are required for further validation.
Our study had some limitations. First, we only included studies published in the English language; therefore, a language bias may have influenced the results. Second, we did not consider the confounding factors such as obesity, whereas a previous study proposed that a high BMI may reduce the efficiency of ultrasound-based elastography techniques in detecting fibrosis[39]. However, only a limited number of studies have provided sufficient data to conduct subgroup analyses to explore the potential impact of BMI on the diagnostic effects. Third, it is unknown whether ALT level is responsible for the difference in the diagnosis of TE between patients with AIH and PBC due to a lack of sufficient data. Moreover, the treatment conditions before inclusion in the study were unknown, and the lack of pretreatment studies made it impossible to compare the effects of treatment on outcomes. Finally, the number of studies on SWE, MRE and ARFI was inadequate to compare the effects of these imaging technologies and TE.
CONCLUSION
In conclusion, LSM using TE had better diagnostic performance for staging hepatic fibrosis in patients with AILDs compared to other serum biomarkers, especially in patients with PBC. The appropriate cutoff value for staging in AF and cirrhosis ranged from 9.6 to 10.7 KPa and 14.4 to 16.9 KPa, respectively, for patients with PBC. Additional recommended optimal cutoff values warrant further investigation to provide a better reference for clinical applications.
ARTICLE HIGHLIGHTS
Research background
Noninvasive criteria are needed for autoimmune liver diseases (AILDs) to assess liver fibrosis stage for prognosis and treatment decisions.
Research motivation
Results of individual diagnostic test accuracy studies assessing the diagnostic accuracy of transient elastography (TE) for the diagnosis of AILD appear promising. However, previous systematic review and meta-analyses have focused primarily on other liver diseases, which is still lacking in AILD.
Research objectives
To compare the diagnostic accuracy of imaging techniques with serum biomarkers of fibrosis in AILD.
Research methods
The PubMed, Cochrane and EMBASE databases were searched for literature. The Quality Assessment of Diagnostic Accuracy Studies-2 tool was used to evaluate the quality. Meta-Disc 1.4 and STATA 12.0 software were used to analyze the combined statistics: sensitivity; specificity; positive likelihood ratio; negative likelihood ratio; diagnostic odds ratio; and area under the curve fitted to the total receiver operating characteristic curve (AUROC).
Research results
A total of 60 studies were included in the meta-analysis. The AUROC curve values were 0.93, 0.93 and 0.91 for significant fibrosis, advanced fibrosis and cirrhosis, respectively, in primary biliary cholangitis patients, while the AUROC curve values were 0.84, 0.88 and 0.90, respectively, in autoimmune hepatitis patients.
Research conclusions
TE is a reliable method for diagnosis in AILD patients, especially in primary biliary cholangitis patients. The appropriate cutoff value for staging advanced fibrosis and cirrhosis ranged from 9.6 to 10.7 KPa and 14.4 to 16.9 KPa, respectively.
Research perspectives
We propose a suitable diagnostic threshold for TE in PBC patients. However, future prospective multicenter studies with TE and histopathology protocol are required to validate our results.
Footnotes
Provenance and peer review: Invited article; Externally peer reviewed.
Tanaka A, Ma X, Yokosuka O, Weltman M, You H, Amarapurkar DN, Kim YJ, Abbas Z, Payawal DA, Chang ML, Efe C, Ozaslan E, Abe M, Mitchell-Thain R, Zeniya M, Han KH, Vierling JM, Takikawa H. Autoimmune liver diseases in the Asia-Pacific region: Proceedings of APASL symposium on AIH and PBC 2016.Hepatol Int. 2016;10:909-915.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 28][Cited by in F6Publishing: 32][Article Influence: 4.0][Reference Citation Analysis (0)]
Poynard T, Halfon P, Castera L, Charlotte F, Le Bail B, Munteanu M, Messous D, Ratziu V, Benhamou Y, Bourlière M, De Ledinghen V; FibroPaca Group. Variability of the area under the receiver operating characteristic curves in the diagnostic evaluation of liver fibrosis markers: impact of biopsy length and fragmentation.Aliment Pharmacol Ther. 2007;25:733-739.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 70][Cited by in F6Publishing: 71][Article Influence: 4.2][Reference Citation Analysis (0)]
Johannessen A, Stockdale AJ, Henrion MYR, Okeke E, Seydi M, Wandeler G, Sonderup M, Spearman CW, Vinikoor M, Sinkala E, Desalegn H, Fall F, Riches N, Davwar P, Duguru M, Maponga T, Taljaard J, Matthews PC, Andersson M, Mboup S, Sombie R, Shimakawa Y, Lemoine M. Systematic review and individual-patient-data meta-analysis of non-invasive fibrosis markers for chronic hepatitis B in Africa.Nat Commun. 2023;14:45.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 11][Cited by in F6Publishing: 15][Article Influence: 15.0][Reference Citation Analysis (0)]
Shiha G, Zalata K. Ishak versus METAVIR: Terminology, Convertibility and Correlation with Laboratory Changes in Chronic Hepatitis C.Liver Biopsy. 2011;10:155-170.
[PubMed] [DOI][Cited in This Article: ]
Park DW, Lee YJ, Chang W, Park JH, Lee KH, Kim YH, Kang NK, Chung JW, Jang HY, Ahn S, Kim H, Jeong SH, Kim JW, Jang ES. Diagnostic performance of a point shear wave elastography (pSWE) for hepatic fibrosis in patients with autoimmune liver disease.PLoS One. 2019;14:e0212771.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 15][Cited by in F6Publishing: 13][Article Influence: 2.6][Reference Citation Analysis (0)]
Koizumi Y, Hirooka M, Abe M, Tokumoto Y, Yoshida O, Watanabe T, Nakamura Y, Imai Y, Yukimoto A, Kumagi T, Takeshita E, Ikeda Y, Hiasa Y. Comparison between real-time tissue elastography and vibration-controlled transient elastography for the assessment of liver fibrosis and disease progression in patients with primary biliary cholangitis.Hepatol Res. 2017;47:1252-1259.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 12][Cited by in F6Publishing: 16][Article Influence: 2.3][Reference Citation Analysis (0)]
Xu XY, Kong H, Song RX, Zhai YH, Wu XF, Ai WS, Liu HB. The effectiveness of noninvasive biomarkers to predict hepatitis B-related significant fibrosis and cirrhosis: a systematic review and meta-analysis of diagnostic test accuracy.PLoS One. 2014;9:e100182.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 33][Cited by in F6Publishing: 34][Article Influence: 3.4][Reference Citation Analysis (0)]
Janik MK, Kruk B, Szczepankiewicz B, Kostrzewa K, Raszeja-Wyszomirska J, Górnicka B, Lammert F, Milkiewicz P, Krawczyk M. Measurement of liver and spleen stiffness as complementary methods for assessment of liver fibrosis in autoimmune hepatitis.Liver Int. 2021;41:348-356.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 8][Cited by in F6Publishing: 7][Article Influence: 2.3][Reference Citation Analysis (0)]
Osman KT, Maselli DB, Idilman IS, Rowan DJ, Viehman JK, Harmsen WS, Harnois DM, Carey EJ, Gossard AA, LaRusso NF, Lindor KD, Venkatesh SK, Eaton JE. Liver Stiffness Measured by Either Magnetic Resonance or Transient Elastography Is Associated With Liver Fibrosis and Is an Independent Predictor of Outcomes Among Patients With Primary Biliary Cholangitis.J Clin Gastroenterol. 2021;55:449-457.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 31][Cited by in F6Publishing: 33][Article Influence: 11.0][Reference Citation Analysis (0)]
Youssef A, Abdel Ghaffar TY, Esmat G, Wanis AAA. Non invasive assessment of hepatic fibrosis in children: Performance of liver stiffness measurement and aspartate transaminase to platelet ratio.Hepatology. 2013;58:815A.
[PubMed] [DOI][Cited in This Article: ]
Kim JK, Kim HW, Lee JI, Lee KS. Analysis of liver stiffness measured by transient elastography in autoimmune hepatitis.Hepatology. 2014;60:364A-365A.
[PubMed] [DOI][Cited in This Article: ]
Harrison L, McFarlane E, Dube A, Gleeson D. Accuracy of transient elastography in predicting histological fibrosis severity in treated autoimmune hepatitis?Gut. 2016;65:A268-A269.
[PubMed] [DOI][Cited in This Article: ]
Hartl J, Denzer U, Ehlken H, Zenouzi R, Peiseler M, Sebode M, Hübener S, Pannicke N, Weiler-Normann C, Quaas A, Lohse AW, Schramm C. Transient elastography in autoimmune hepatitis: Timing determines the impact of inflammation and fibrosis.J Hepatol. 2016;65:769-775.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 96][Cited by in F6Publishing: 104][Article Influence: 13.0][Reference Citation Analysis (0)]
Nishikawa H, Enomoto H, Iwata Y, Hasegawa K, Nakano C, Takata R, Nishimura T, Yoh K, Aizawa N, Sakai Y, Ikeda N, Takashima T, Iijima H, Nishiguchi S. Clinical significance of serum Wisteria floribunda agglutinin positive Mac-2-binding protein level and high-sensitivity C-reactive protein concentration in autoimmune hepatitis.Hepatol Res. 2016;46:613-621.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 61][Cited by in F6Publishing: 66][Article Influence: 8.3][Reference Citation Analysis (0)]
E Anastasiou O, Büchter M, A Baba H, Korth J, Canbay A, Gerken G, Kahraman A. Performance and Utility of Transient Elastography and Non-Invasive Markers of Liver Fiibrosis in Patients with Autoimmune Hepatitis: A Single Centre Experience.Hepat Mon. 2016;16:e40737.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 14][Cited by in F6Publishing: 19][Article Influence: 2.4][Reference Citation Analysis (0)]
Piwczyńska K, Marek W, Maciej D, Małgorzata W. APRI as a fibrosis marker in children with autoimmune hepatitis(AIH). In: ESPGHAN 49th ANNUAL MEETING of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition.Journal of Pediatric Gastroenterology and Nutrition. 2016;62:1-890.
[PubMed] [DOI][Cited in This Article: ]
Paranagua-Vezozzo D, Terrabuio DR, Moutinho RD, Ono S, Salas V, Carrilho F, Alves VF, Cancado EL. Transient elastography (TE) and acoustic radiation force impulse imaging (ARFI) can predict degree of advanced fibrosis for autoimmune hepatitis in biochemical remission.Hepatology. 2017;66:187A.
[PubMed] [DOI][Cited in This Article: ]
Li X, Xu H, Gao P. Red Blood Cell Distribution Width-to-Platelet Ratio and Other Laboratory Indices Associated with Severity of Histological Hepatic Fibrosis in Patients with Autoimmune Hepatitis: A Retrospective Study at a Single Center.Med Sci Monit. 2020;26:e927946.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 4][Cited by in F6Publishing: 4][Article Influence: 1.0][Reference Citation Analysis (0)]
Ferronato M, Lenzi M, Muratori P, Muratori L. Blood-Based Non-Invasive Tests of Hepatic Fibrosis in Autoimmune Hepatitis: Application among Selected Patients Leads to Higher Accuracy.Gastroenterology Insights. 2022;13:286-295.
[PubMed] [DOI][Cited in This Article: ][Reference Citation Analysis (0)]
Soh EG, Lee YH, Kim YR, Yoon KH, Choi KH. Usefulness of 2D shear wave elastography for the evaluation of hepatic fibrosis and treatment response in patients with autoimmune hepatitis.Ultrasonography. 2022;41:740-749.
[PubMed] [DOI][Cited in This Article: ][Reference Citation Analysis (0)]
Ferrara F, Caroli D, Antoniazzi S, Variola A, Cazzagon N, Baldo V, Floreani A. Performance of surrogate markers of hepatic fibrosis in primary biliary cirrhosis.J Hepatol. 2009;50:S243.
[PubMed] [DOI][Cited in This Article: ]
Su CW, Chan CC, Hung HH, Huo TI, Huang YH, Li CP, Lin HC, Tsay SH, Lee PC, Lee SD, Wu JC. Predictive value of aspartate aminotransferase to alanine aminotransferase ratio for hepatic fibrosis and clinical adverse outcomes in patients with primary biliary cirrhosis.J Clin Gastroenterol. 2009;43:876-883.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 19][Cited by in F6Publishing: 15][Article Influence: 1.0][Reference Citation Analysis (0)]
Umemura T, Joshita S, Sekiguchi T, Usami Y, Shibata S, Kimura T, Komatsu M, Matsumoto A, Ota M, Tanaka E. Serum Wisteria floribunda Agglutinin-Positive Mac-2-Binding Protein Level Predicts Liver Fibrosis and Prognosis in Primary Biliary Cirrhosis.Am J Gastroenterol. 2015;110:857-864.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 101][Cited by in F6Publishing: 106][Article Influence: 11.8][Reference Citation Analysis (0)]
Nishikawa H, Enomoto H, Iwata Y, Hasegawa K, Nakano C, Takata R, Nishimura T, Yoh K, Aizawa N, Sakai Y, Ikeda N, Takashima T, Ishii A, Iijima H, Nishiguchi S. Impact of serum Wisteria floribunda agglutinin positive Mac-2-binding protein and serum interferon-γ-inducible protein-10 in primary biliary cirrhosis.Hepatol Res. 2016;46:575-583.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 29][Cited by in F6Publishing: 32][Article Influence: 4.0][Reference Citation Analysis (0)]
Kamal N, Surana P, Noureddin M, Kleiner D, Hoofnagle J, Heller T, Koh C. Bile Duct Damage and Cirrhosis in Primary Biliary Cholangitis: An Investigation of Non-Invasive Biomarkers.Gastroenterology. 2018;154:S1209.
[PubMed] [DOI][Cited in This Article: ]
Milovanović T, Copertino A, Boričić I, Miličić B, Marković AP, Krstić M, Matović V, Popović D. Transient elastography for noninvasive assessment of liver fibrosis in patients with primary biliary cirrhosis.Vojnosanitetski Pregled. 2018;75:374-379.
[PubMed] [DOI][Cited in This Article: ][Reference Citation Analysis (0)]
Rossi M, Alessi N, Cabibi A, Fichera A, Zarcone A, Bianco AL, Marco VD, Cammà C, Craxì A. High rate of misclassification of fibrosis stage using Transient Elastography in patients with Primary Biliary Cholangitis.Digestive and Liver Disease. 2020;52:e34.
[PubMed] [DOI][Cited in This Article: ]
Cristoferi L, Calvaruso V, Overi D, Viganò M, Rigamonti C, Degasperi E, Cardinale V, Labanca S, Zucchini N, Fichera A, Di Marco V, Leutner M, Venere R, Picciotto A, Lucà M, Mulinacci G, Palermo A, Gerussi A, D'Amato D, Elisabeth O'Donnell S, Cerini F, De Benedittis C, Malinverno F, Ronca V, Mancuso C, Cazzagon N, Ciaccio A, Barisani D, Marzioni M, Floreani A, Alvaro D, Gaudio E, Invernizzi P, Carpino G, Nardi A, Carbone M; Italian PBC Registry. Accuracy of Transient Elastography in Assessing Fibrosis at Diagnosis in Naïve Patients With Primary Biliary Cholangitis: A Dual Cut-Off Approach.Hepatology. 2021;74:1496-1508.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 16][Cited by in F6Publishing: 25][Article Influence: 8.3][Reference Citation Analysis (0)]
Fujinaga Y, Namisaki T, Takaya H, Tsuji Y, Suzuki J, Shibamoto A, Kubo T, Iwai S, Tomooka F, Takeda S, Fujimoto Y, Enomoto M, Murata K, Ishida K, Ogawa H, Takagi H, Ozutsumi T, Furukawa M, Nishimura N, Sawada Y, Kitagawa K, Sato S, Kaji K, Kawaratani H, Moriya K, Noguchi R, Akahane T, Mitoro A, Yoshiji H. Enhanced liver fibrosis score as a surrogate of liver-related complications and mortality in primary biliary cholangitis.Medicine (Baltimore). 2021;100:e27403.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 3][Cited by in F6Publishing: 10][Article Influence: 3.3][Reference Citation Analysis (0)]
Manesis EK, Schina M, Vafiadis I, Gatos I, Theotokas J, Zoumpoulis P, Drazinos P, Ketikoglou J, Delladetsima IK, Tiniakos DG. Liver stiffness measurements by 2-dimensional shear wave elastography compared to histological and ultrasound parameters in primary biliary cholangitis.Scand J Gastroenterol. 2021;56:1187-1193.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 2][Cited by in F6Publishing: 6][Article Influence: 2.0][Reference Citation Analysis (0)]
Corpechot C, Gaouar F, El Naggar A, Kemgang A, Wendum D, Poupon R, Carrat F, Chazouillères O. Baseline values and changes in liver stiffness measured by transient elastography are associated with severity of fibrosis and outcomes of patients with primary sclerosing cholangitis.Gastroenterology. 2014;146:970-9; quiz e15.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 175][Cited by in F6Publishing: 179][Article Influence: 17.9][Reference Citation Analysis (0)]
Bowlus CL, Montano-Loza AJ, Invernizzi P, Chazouillères O, Hirschfield G, Metselaar HJ, Goodman Z, Myers RP, Aguilar R, Subramanian GM, McHutchison JG, Chapman R, Muir AJ, Eksteen B, Levy C. Liver stiffness measurement by transient elastography for the prediction of fibrosis in patients with primary sclerosing cholangitis in a randomized trial of simtuzumab.J Hepatol. 2016;64:S434.
[PubMed] [DOI][Cited in This Article: ]
Ehlken H, Wroblewski R, Corpechot C, Arrivé L, Rieger T, Hartl J, Lezius S, Hübener P, Schulze K, Zenouzi R, Sebode M, Peiseler M, Denzer UW, Quaas A, Weiler-Normann C, Lohse AW, Chazouilleres O, Schramm C. Validation of Transient Elastography and Comparison with Spleen Length Measurement for Staging of Fibrosis and Clinical Prognosis in Primary Sclerosing Cholangitis.PLoS One. 2016;11:e0164224.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 32][Cited by in F6Publishing: 33][Article Influence: 4.1][Reference Citation Analysis (0)]
Krawczyk M, Ligocka J, Ligocki M, Raszeja-Wyszomirska J, Milkiewicz M, Szparecki G, Ilczuk T, Górnicka B, Zieniewicz K, Krawczyk M, Lammert F, Milkiewicz P. Does transient elastography correlate with liver fibrosis in patients with PSC? Laennec score-based analysis of explanted livers.Scand J Gastroenterol. 2017;52:1407-1412.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 12][Cited by in F6Publishing: 12][Article Influence: 1.7][Reference Citation Analysis (0)]
