TransientElastographyinAlcoholicLiverDiseaseand … · 2021. 1. 20. · ResearchArticle...

Research ArticleTransient Elastography in Alcoholic Liver Disease andNonalcoholic Fatty Liver Disease: A Systemic Reviewand Meta-Analysis

Changzhou Cai ,1 Xin Song,1 Xueyang Chen,1 Weihua Zhou,1,2 Qi Jin,1 Shenghui Chen,1

and Feng Ji 1

1Department of Gastroenterology, �e First Affiliated Hospital of Zhejiang University School of Medicine,Hangzhou 310003, China2Department of Gastroenterology, Sanmen People’s Hospital, Taizhou, China

Correspondence should be addressed to Feng Ji; [email protected]

Received 22 July 2020; Revised 21 November 2020; Accepted 21 December 2020; Published 20 January 2021

Academic Editor: Alessandro Granito

Copyright © 2021 Changzhou Cai et al. (is is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Background and Aims. Alcoholic liver disease (ALD) and nonalcoholic fatty liver disease (NAFLD) have become common chronicliver diseases. Recent evidence has shown the value of transient elastography (TE) in the context of ALD/NAFLD.(e aim of thisstudy is to investigate the accuracy of TE for diagnosing steatosis and fibrosis in ALD/NAFLD patients. Methods. We retrievedrelevant English studies from the databases of PubMed, Embase, the Web of Science, and the Cochrane Library through March31st 2019. We included studies regarding the diagnosis or staging of steatosis or fibrosis by using controlled attenuation parameter(CAP) or liver stiffness measurement (LSM) measured by TE in patients with ALD or NAFLD. (e reference standard of allincluded studies was liver biopsy. A random-effects model was applied. Statistical analyses were performed using STATA. Results.A total of 62 articles were included and analyzed in our meta-analysis. In patients with ALD/NAFLD, the pooled results revealedthat the sensitivity and specificity of CAP were 0.84, 0.83, and 0.78 and 0.83, 0.71, and 0.62 for steatosis grades ≥S1, ≥S2, and �S3,respectively. (e sensitivity and specificity of LSM for identifying fibrosis grades ≥F1, ≥F2, ≥F3, and �F4 were 0.77, 0.77, 0.83, and0.91 and 0.80, 0.82, 0.84, and 0.86, respectively. Conclusion. In patients with ALD/NAFLD, CAP was feasible for identifying andscreening steatosis, and LSM was accurate for diagnosing fibrosis, especially severe fibrosis and cirrhosis.

1. Introduction

Nonalcoholic fatty liver disease (NAFLD) is one of the mostprevalent chronic liver diseases worldwide, affecting ap-proximately 25% of the adult population [1]. It is expectedthat NAFLD will soon become the leading cause of livertransplantation [2, 3]. (e spectrum of NAFLD ranges fromthe reversible phase of nonalcoholic fatty liver disease(NAFLD) to nonalcoholic steatohepatitis (NASH), whichmay progress to liver fibrosis and hepatocellular carcinoma[4]. Alcoholic liver disease (ALD) is another commonchronic liver disease that includes the steatosis and fibrosispathological classifications [5]. Currently, liver biopsy is still

the gold standard for ALD/NAFLD diagnosis and steatosisand liver fibrosis staging. However, the application of liverbiopsy is not easy because it is an invasive test with po-tentially life-threatening complications after the operation[6]. (erefore, developing noninvasive and accuratemethods for disease diagnosis and staging in ALD/NAFLDpatients is urgently needed.

Recently, controlled attenuation parameter (CAP) andliver stiffness measurement (LSM) measured by transientelastography (TE) with Fibroscan® equipment (Echosens,Paris, France) have been widely used to assess hepaticsteatosis and liver fibrosis, respectively [7, 8]. CAP isevaluated based on the properties of ultrasonic signals

HindawiCanadian Journal of Gastroenterology and HepatologyVolume 2021, Article ID 8859338, 10 pageshttps://doi.org/10.1155/2021/8859338

mailto:[email protected]

https://orcid.org/0000-0003-4602-5046

https://orcid.org/0000-0002-1426-0802

https://creativecommons.org/licenses/by/4.0/

https://creativecommons.org/licenses/by/4.0/

https://doi.org/10.1155/2021/8859338

acquired by the Fibroscan® equipment. Moreover, CAP andLSM shared the same radio-frequency data and the sameregion of interest [9]. Many biopsy-proven ALD/NAFLDstudies have reported excellent performance of CAP indetecting and staging steatosis [10–12]. Additionally, LSMaccurately predicted liver fibrosis and differentiated itsdifferent stages in several biopsy-controlled studies[9, 13, 14].

In this meta-analysis, we aimed to evaluate the efficiencyof CAP and LSMmeasured by TE for diagnosing and stagingsteatosis and fibrosis in patients with ALD/NAFLD usingliver biopsy as the reference standard.

2. Methods and Materials

2.1. Search Strategy. We performed this comprehensivesystematic review and meta-analysis based on the PreferredReporting Items for Systematic Reviews and Meta-Analyses(PRISMA) statement. Studies published in English about thediagnosis and staging of steatosis and fibrosis using CAP andLSM in patients with ALD/NAFLD were retrieved from thePubMed, Embase, Web of Science, and Cochrane Librarydatabases. (e databases were searched through December31st, 2019. (e electronic search strategy included the fol-lowing terms: (“alcoholic liver disease” OR “ALD” OR“alcoholic hepatitis” OR “non-alcoholic fatty liver disease”OR “NAFLD” OR “non-alcoholic steatohepatitis” OR“NASH”) AND (“transient elastography” OR “controlledattenuation parameter” OR “liver stiffness”). (e retrievalstrategies for each database are shown in the supplementarymaterials. Furthermore, we performed a manual search, andwe added two additional suitable articles.

2.2. Paper Selection. We included studies if they met thefollowing criteria: (1) studies regarding the diagnosis orstaging of steatosis or fibrosis by using CAP or LSM mea-sured by TE in patients with ALD or NAFLD; (2) studies inwhich the gold standard diagnosis method for patients withNAFLD/ALD was liver biopsy; and (3) articles that providedspecificity (SPE), sensitivity (SEN), sample size, or enoughinformation to calculate and construct a diagnostic 2 ∗ 2contingency table. (e exclusion criteria included the fol-lowing: (1) cell or animal studies, comments, reviews, andletters; (2) duplicate studies; and (3) studies focused onirrelevant topics or that did not report necessary data.

3. Methodological Quality and Bias Assessment

(e methodological quality of the included studies wasevaluated by the Quality Assessment of Diagnostic AccuracyStudies-2 (QUADAS-2) checklist by Review ManagerVersion 5.3. (is evaluation form consists of four parts:patient selection, index test, reference standard, and flowand timing. We assessed the methodological quality of eacharticle by answering every question using “Yes,” “No,” or“Unclear” or “High concern,” “Low concern,” or “Unclearconcern.”

3.1. Data Extraction. Two researchers (CC and XC) inde-pendently filtered studies by reviewing the titles and ab-stracts and then proceeded with a full-text evaluation. (ethird reviewer resolved the disagreements. We extracted thefollowing data from the literature: (1) author, publicationyear, ethnicity, age, sample size, disease, diagnostic index,probe type, diagnostic thresholds (cut-off values), body massindex (BMI), and study design; (2) diagnostic parameters ofthe diagnostic index, including sensitivity, specificity, andcalculated or extracted numbers of true positive (TP), falsepositive (FP), false negative (FN), and true negative (TN)cases. We then constructed a diagnostic 2 ∗ 2 contingencytable.

3.2. Statistical Analysis. Our meta-analysis was mainly di-vided into two parts. First, we assessed TE-measured CAPfor the diagnosis and grading of steatosis in patients withALD/NAFLD. Subgroup analyses were then conductedbased on BMI and cut-off value when analyzing the effi-ciency of CAP for grade ≥S1, ≥S2, and S3 steatosis. Meta-regression analyses were performed to determine thesource of heterogeneity. Second, we evaluated TE-mea-sured LSM for the diagnosis and grading of fibrosis inpatients with ALD/NAFLD. When analyzing the efficiencyof LSM for grade ≥F1, ≥F2, ≥F3, and F4 fibrosis, weperformed subgroup analyses based on disease status, BMI,and study design. Meta-regression analyses were per-formed to determine the source of heterogeneity. We es-timated SEN, SPE, positive likelihood ratio (PLR), negativelikelihood ratio (NLR), and diagnostic odds ratio (DOR)with their associated 95% confidence intervals (CIs) fromthe diagnostic 2 ∗ 2 contingency tables. To analyze thediagnostic accuracy of TE in patients with ALD/NAFLD,we pooled SEN and SPE graphically using forest plots,constructed summary receiver-operating characteristics(SROC) curves, and calculated areas under the SROC curve(AUC).

We evaluated the heterogeneity between studies usingthe I2 index. I2 values of 25%, 50%, and 75% corresponded tolow, medium, and high heterogeneity, respectively. (erandom-effects model was chosen as the default; if the I2values were less than 50%, we chose the fix-effects model.Furthermore, we constructed the Deeks’ funnel plots, inwhich P< 0.10 indicated that there was publication bias. Weperformed all of these statistical analyses using Stata (version12.0).

4. Results

4.1. Paper Selection and Characteristics. Based on the liter-ature research strategy, we retrieved 3283 articles, of which469 were from PubMed, 1491 were from Embase, 1313 werefrom theWeb of Science, eight were fromCochrane, and twowere identified in the manual search. After removing 1069duplicate studies, 262 reviews or comments, 16 animalstudies, 1741 irrelevant studies, and 33 studies failing toprovide sufficient data, we eventually included 62 articles(supplementary materials) in our study (Figure 1). (e

2 Canadian Journal of Gastroenterology and Hepatology

necessary information and related diagnostic data are shownin Table S1. Among the studies, 17 assessed the efficiency ofCAP for diagnosing and staging steatosis in ALD/NAFLDpatients, and 53 studies evaluated the accuracy of LSM fordiagnosing and grading fibrosis in ALD/NAFLD patients.CAP and LSM were both measured by TE with Fibroscan®equipment. (e methodological quality of each includedarticle assessed by QUADAS is shown in a bar chart(Figure S1).

4.2. Diagnostic Performance of CAP for Steatosis in ALD/NAFLD Patients. 14 studies containing 1936 ALD/NAFLDpatients with pathologically confirmed S1–S3 steatosis and369 healthy controls assessed the diagnostic accuracy of CAPfor steatosis grades ≥S1. (e pooled results indicated highaccuracy with an AUC of 0.90 (95% CI 0.87–0.92)(Figure 2(c)), a sensitivity of 0.84 (95% CI 0.78–0.88)(Figure 2(a)), and a specificity of 0.83 (95% CI 0.77–0.87)(Figure 2(b)) at an average threshold of 272 dB/m. Publi-cation bias was not significant (P � 0.462) (Figure S2).

17 studies containing 1625 pathologically confirmedALD/NAFLD patients with S2–S3 steatosis grades and 1542controls assessed the diagnostic accuracy of CAP for

steatosis grades ≥S2. (e pooled results revealed high ac-curacy with an AUC of 0.83 (95% CI 0.79–0.86)(Figure 2(f)), sensitivity of 0.83 (95% CI 0.77–0.88)(Figure 2(d)), and specificity of 0.71 (95% CI 0.66–0.76)(Figure 2(e)) at an average threshold of 292 dB/m.(ere wassignificant publication bias, with P< 0.05 (Figure S3).

15 studies containing 591 patients with pathologicallyconfirmed ALD/NAFLD with S2–S3 steatosis and 1872controls assessed the diagnostic accuracy of CAP for stea-tosis grade S3. Accuracy was lowest at the S3 thresholdcompared with the ≥S1 threshold and the ≥S2 threshold,with an AUC of 0.79 (95% CI 0.75–0.82) (Figure 2(i)),sensitivity of 0.78 (95% CI 0.72–0.83) (Figure 2(g)), andspecificity of 0.62 (95% CI 0.56–0.69) (Figure 2(h)) at anaverage threshold of 308 dB/m.(e results of PLR, NLR, andDOR from the nomogram are shown in supplementarymaterials (Figures S2–S4).

We then performed subgroup analyses according to BMIand the cut-off values (Table 1). Furthermore, meta-regressionwas performed according to covariates including BMI, cut-offvalue, sample size, ethnicity, disease, and study design, anddisease was shown to relate to the heterogeneity (P � 0.045)(Table S3–S5).

Records identified through database searching (n = 3281)

Scre

enin

gIn

clude

dEl

igib

ility

Iden

tifica

tion Additional records identified through

other sources(n = 2)

Records a�er duplicates removed (n = 2114)

Records screened(n = 95)

Records excluded(n = 2019)

Reviews and comments(n = 262)

Animal studies (n = 16)

Irrelevant researches(n = 1741)

Full-text articles assessed for eligibility (n = 95)

Full-text articles excluded forwithout necessary data for

constructing diagnostic 2 ∗ 2 contingency table.

(n = 33)

Studies included in qualitative synthesis

(n = 62)

Studies included in quantitative synthesis (meta-

analysis)(n = 62)

Figure 1: Flowchart showing the selection of articles included in the meta-analysis.

Canadian Journal of Gastroenterology and Hepatology 3

Study (year) Sensitivity (95% CI)

Naveau 2017 NAFLD M/XL 0.78 [0.69 – 0.86]

Naveau 2017 NAFLD M/XL 0.68 [0.60 – 0.75]

�iele 2018 ALD/ 0.70 [0.63 – 0.76]

Chan 2014 NAFLD M 0.92 [0.85 – 0.96]

Karlas 2014 NAFLD M 0.93 [0.82 – 0.99]

Imajo 2016 NAFLD M 0.82 [0.75 – 0.88]

Lee 2016 NAFLD M/XL 0.88 [0.82 – 0.92]

Chan 2017 NAFLD XL 0.95 [0.85 – 0.99]

Chan 2017 NAFLD M 0.91 [0.80 – 0.97]

Forlano 2017 NAFLD/ 0.90 [0.81 – 0.96]

Garg 2018 NAFLD XL 0.59 [0.46 – 0.70]

Runge 2018 NAFLD M 0.90 [0.78 – 0.97]

Eddowes 2019 NAFLD M/XL 0.80 [0.75 – 0.84]

Siddiqui 2019 NAFLD M/XL 0.80 [0.76 – 0.84]

Combined 0.84 [0.78 – 0.88]

Q = 115.61, df = 13.00, p = 0.00

0.5Sensitivity

1.0

I2 = 88.76 [84.04 – 93.47]

(a)

Study (year) Specificity (95% CI)

Naveau 2017 NAFLD M/XL 0.83 [0.61 – 0.95]

Naveau 2017 NAFLD M/XL 0.69 [0.49 – 0.85]

�iele 2018 ALD/ 0.73 [0.61 – 0.82]

Chan 2014 NAFLD M 0.94 [0.85 – 0.98]

Karlas 2014 NAFLD M 0.87 [0.60 – 0.98]

Imajo 2016 NAFLD M 0.90 [0.55 – 1.00]

Lee 2016 NAFLD M/XL 1.00 [0.66 – 1.00]

Chan 2017 NAFLD XL 0.91 [0.72 – 0.99]

Chan 2017 NAFLD M 0.87 [0.66 – 0.97]

Forlano 2017 NAFLD/ 0.60 [0.36 – 0.81]

Garg 2018 NAFLD XL 0.83 [0.36 – 1.00]

Runge 2018 NAFLD M 0.60 [0.15 – 0.95]

Eddowes 2019 NAFLD M/XL 0.83 [0.69 – 0.92]


Combined 0.83 [0.77 – 0.87]

Q = 29.36, df = 13.00, p = 0.01

0.1Specificity

1.0

I2 = 55.73 [29.18 – 82.27]

(b)

1.0

0.5

0.0

SROC with confidence and predictive ellipses

1.0 0.5Specificity

0.0

Observed dataSummary operating pointSENS = 0.84 [0.78 – 0.88]SPEC = 0.83 [0.77 – 0.87]

95% confidence ellipse95% prediction ellipse

SROC curveAUC = 0.90 [0.87 – 0.92]

Sens

itivi

ty

(c)


Kumar 2013 NAFLD/ 0.78 [0.62 – 0.90]de Ledinghen 2016 NAFLD M 0.79 [0.73 – 0.85]

Naveau 2017 NAFLD M/XL 0.90 [0.81 – 0.96]Naveau 2017 NAFLD M/XL 0.65 [0.55 – 0.74]

�iele 2018 ALD/ 0.64 [0.54 – 0.74]Friedrich-Rust 2012 NAFLD M/XL 0.98 [0.87 – 1.00]

Karlas 2014 NAFLD M 0.97 [0.83 – 1.00]Hiriart 2015 NAFLD M 0.79 [0.73 – 0.85]Imajo 2016 NAFLD M 0.78 [0.68 – 0.87]

Lee 2016 NAFLD M/XL 0.85 [0.76 – 0.91]Chan 2017 NAFLD XL 0.95 [0.84 – 0.99]Chan 2017 NAFLD M 0.91 [0.78 – 0.97]

Ooi 2018 NAFLD XL 0.84 [0.75 – 0.91]

Eddoews 2019 NAFLD M/XL 0.70 [0.63 – 0.75]Siddiqui 2019 NAFLD M/XL 0.77 [0.71 – 0.82]

Runge 2018 NAFLD M 0.92 [0.75 – 0.99]

Garg 2018 NAFLD XL 0.74 [0.52 – 0.90]

Combined 0.83 [0.77 – 0.88]Q = 84.08, df = 16.00, p = 0.00

0.5Sensitivity

1.0

I2 = 80.97 [72.61 – 89.33]

(d)


Kumar 2013 NAFLD/ 0.73 [0.52 – 0.88]de Ledinghen 2016 NAFLD M 0.71 [0.59 – 0.80]

Naveau 2017 NAFLD M/XL 0.69 [0.55 – 0.81]Naveau 2017 NAFLD M/XL 0.79 [0.68 – 0.87]

�iele 2018 ALD/ 0.74 [0.67 – 0.81]Friedrich-Rust 2012 NAFLD M/XL 0.67 [0.38 – 0.88]

Karlas 2014 NAFLD M 0.81 [0.63 – 0.93]Hiriart 2015 NAFLD M 0.71 [0.59 – 0.80]Imajo 2016 NAFLD M 0.81 [0.70 – 0.90]

Lee 2016 NAFLD M/XL 0.80 [0.70 – 0.88]Chan 2017 NAFLD XL 0.61 [0.43 – 0.77]Chan 2017 NAFLD M 0.86 [0.71 – 0.95]

Ooi 2018 NAFLD XL 0.47 [0.37 – 0.58]

Eddoews 2019 NAFLD M/XL 0.76 [0.68 – 0.83]Siddiqui 2019 NAFLD M/XL 0.57 [0.49 – 0.64]

Runge 2018 NAFLD M 0.55 [0.36 – 0.74]

Garg 2018 NAFLD XL 0.75 [0.62 – 0.86]

Combined 0.71 [0.66 – 0.76]Q = 72.80, df = 16.00, p = 0.00

0.4Sensitivity

1.0

I2 = 78.02 [67.98 – 88.07]

(e)

1.0 0.5Specificity

0.0



SROC curveAUC = 0.88 [0.79 – 0.86]

1.0

0.5

Sens

itivi

ty

0.0


(f )


Kumar 2013 NAFLD/ 0.71 [0.29 – 0.96]

Naveau 2017 NAFLD M/XL

de Ledinghen 2016 NAFLD 0.87 [0.78 – 0.93]

�iele 2018 ALD/

0.74 [0.62 – 0.83]

Chan 2014 NAFLD M

0.88 [0.73 – 0.97]

Karlas 2014 NAFLD M

0.75 [0.53 – 0.90]

Imajo 2016 NAFLD M

1.00 [0.77 – 1.00]

Lee 2016 NAFLD M/XL

0.82 [0.48 – 0.98]

Chan 2017 NAFLD XL

0.63 [0.41 – 0.81]

Chan 2017 NAFLD M

0.73 [0.54 – 0.87]

Friedrich-Rust 2012 NAFLD M/XL

0.80 [0.52 – 0.96]

Garg 2018 NAFLD XL

1.00 [0.78 – 1.00]

Runge 2018 NAFLD M

1.00 [0.40 – 1.00]

Eddowes 2019 NAFLD M/XL

0.78 [0.40 – 0.87]

0.72 [0.63 – 0.79]


Combined 0.78 [0.72 – 0.83]

Q = 115.61, df = 13.00, p = 0.00

0.5Sensitivity

1.0

I2 = 88.76 [84.04 – 93.47]

(g)


Kumar 2013 NAFLD/ 0.68 [0.54 – 0.80]

Naveau 2017 NAFLD M/XL

de Ledinghen 2016 NAFLD 0.47 [0.40 – 0.55]

�iele 2018 ALD/

0.74 [0.65 – 0.81]

Chan 2014 NAFLD M

0.67 [0.60 – 0.73]

Karlas 2014 NAFLD M

0.67 [0.48 – 0.82]

Imajo 2016 NAFLD M

0.53 [0.45 – 0.61]

Lee 2016 NAFLD M/XL

0.76 [0.62 – 0.87]

Chan 2017 NAFLD XL

0.73 [0.65 – 0.81]

Chan 2017 NAFLD M

0.61 [0.52 – 0.69]

Friedrich-Rust 2012 NAFLD M/XL

0.55 [0.42 – 0.67]

Garg 2018 NAFLD XL

0.41 [0.29 – 0.54]

Runge 2018 NAFLD M

0.78 [0.66 – 0.87]

Eddowes 2019 NAFLD M/XL

0.76 [0.61 – 0.87]

0.63 [0.56 – 0.69]


Combined 0.78 [0.72 – 0.83]

Q = 115.61, df = 13.00, p = 0.00

0.3Specificity

0.9

I2 = 88.76 [84.04 – 93.47]

(h)

1.0 0.5Specificity

0.0



SROC curveAUC = 0.79 [0.75 – 0.82]

1.0

0.5

0.0


Sens

itivi

ty

(i)

Figure 2: Diagnostic performance of controlled attenuation parameter (CAP) for steatosis in alcoholic liver disease/nonalcoholic fatty liver disease.(a) Sensitivity, (b) specificity, and (c) summary receiver-operating characteristics (SROC) curve of CAP for identifying patients with steatosis grade≥S1. (d) Sensitivity, (e) specificity, and (f) SROC curve of CAP for identifying patients with steatosis grade≥S2. (g) Sensitivity, (h) specificity, and (i)SROC curve of CAP for identifying patients with steatosis grade �S3. CAP: controlled attenuation parameter.


4.3. Diagnostic Performance of LSM for Fibrosis in ALD/NAFLD Patients. 11 studies containing 1097 patients withpathologically confirmed ALD/NAFLD with F1–F4 fibrosisgrades and 484 healthy controls assessed the diagnosticperformance of LSM for fibrosis grades ≥F1. Accuracy waslowest at the ≥F1 threshold, with an AUC of 0.85 (95% CI0.82–0.88) (Figure 3(c)), a sensitivity of 0.77 (95% CI0.68–0.74) (Figure 3(a)), and a specificity of 0.80 (95% CI0.73–0.86) (Figure 3(b)) at an average threshold of 6.3 kPa.Significant publication bias was identified (P � 0.029) in thisanalysis (Figure S5).

40 studies containing 2569 ALD/NAFLD patients withpathologically confirmed F2–F4 fibrosis and 3014 controlsassessed the diagnostic performance of LSM for fibrosisgrades ≥F2.(e pooled results showed high accuracy with anAUC of 0.86 (95% CI 0.83–0.89) (Figure 3(f)), sensitivity of0.77 (95% CI 0.73–0.81) (Figure 3(d)), and specificity of 0.82(95% CI 0.78–0.86) (Figure 3(e)) at an average threshold of8.2 kPa. We also identified significant publication bias, withP< 0.05 (Figure S6).

51 studies containing 2925 patients pathologically con-firmed ALD/NAFLD with F3–F4 fibrosis and 6308 controlsassessed the diagnostic performance of LSM for fibrosisgrades ≥F3. Diagnostic accuracy was high at the ≥F3threshold, with an AUC of 0.90 (95% CI 0.88–0.93)(Figure 4(c)), sensitivity of 0.83 (95% CI 0.79–0.86)(Figure 4(a)), and specificity of 0.84 (95% CI 0.81–0.87)(Figure 4(b)) at an average threshold of 13.4 kPa. Publicationbias was also detected, with P< 0.05 (Figure S7).

34 studies containing 914 ALD/NAFLD patients withpathologically confirmed F4 fibrosis and 4238 controlsassessed the diagnostic performance of LSM for fibrosisgrade F4. Diagnostic accuracy was highest at the F4

threshold, with an AUC of 0.95 (95% CI 0.92–0.96)(Figure 4(f )), sensitivity of 0.91 (95% CI 0.87–0.94)(Figure 4(d)), and specificity of 0.86 (95% CI 0.83–0.89)(Figure 4(e)) at an average threshold of 14.2 kPa. No sig-nificant publication bias was detected (P � 0.054)(Figure S8). (e results of PLR, NLR, and DOR from thenomogram are shown in supplementary materials(Figures S5–S7).

In addition, subgroup analyses were conductedaccording to BMI and study type (Table 2). We then con-ducted meta-regressions according to the sample size,ethnicity, disease, and study design, and we detected thatdisease was the possible source of heterogeneity, with P �

0.042 (Table S6–S9).

5. Discussion

ALD and NAFLD are two common chronic liver diseasesthat pathologically range from steatosis, steatohepatitis, andfibrosis to cirrhosis [7, 15]. (us far, liver biopsy is still thegold standard for diagnosing patients with ALD/NAFLDand staging steatosis and fibrosis. However, the applicationof liver biopsy is difficult due to its invasive characteristics,possible subsequent adverse reactions, and relatively highprice.(erefore, a variety of noninvasive methods, includingserum biomarkers and imaging techniques, have beenevolving and advancing [16, 17]. When compared withtraditional noninvasive diagnostic methods for ALD/NAFLD in clinical practice, such as ultrasound and com-puted tomography scans, TE using Fibroscan® equipmentexhibited high accuracy in diagnosing and staging steatosisand fibrosis for patients with ALD/NAFLD [7, 18–20].Considering the pathological commonalities between ALD

Table 1: Summary diagnostic accuracy parameter estimates and their 95% confidence intervals.

Subgroup Sensitivity(95% CI)

Specificity(95% CI)

Positive LR(95% CI)

Negative LR(95% CI)

DOR(95% CI) AUC

S1–S3BMI ≥28 6 0.76 [0.69–0.81] 0.79 [0.70–0.86] 3.69 [2.41–5.66] 0.30 [0.23–0.41] 12.11 [6.08–24.12] 0.85 [0.81–0.88]BMI <28 7 0.89 [0.83–0.93] 0.88 [0.80–0.93] 7.44 [4.24–13.06] 0.13 [0.07–0.21] 59.05 [22.16–159.76] 0.95 [0.92–0.96]Cut-off ≥270 7 0.77 [0.73–0.81] 0.82 [0.78–0.86] 4.34 [3.45–5.46] 0.28 [0.23–0.33] 15.65 [11.33–21.60] 0.85 [0.83–0.89]Cut-off <270 7 0.89 [0.85–0.92] 0.86 [0.74–0.89] 6.56 [3.20–13.47] 0.13 [0.10–0.17] 50.89 [21.26–121.84] 0.92 [0.89–0.94]

S2–S3BMI ≥28 10 0.79 [0.73–0.84] 0.69 [0.62–0.75] 2.56 [2.09–3.13] 0.30 [0.23–0.38] 8.61 [6.07–2.21] 0.81 [0.77–0.84]BMI <28 7 0.88 [0.78–0.94] 0.88 [0.78–0.90] 3.35 [2.60–4.31] 0.16 [0.09–0.31] 38.28 [25.84–56.70] 0.84 [0.80–0.87]Cut-off ≥290 9 0.76 [0.70–0.82] 0.70 [0.64–0.75] 2.52 [2.15–2.95] 0.34 [0.27–0.42] 7.45 [5.56–9.97] 0.79 [0.75–0.82]Cut-off <290 8 0.89 [0.75–0.81] 0.73 [0.63–0.81] 3.30 [2.33–4.65] 0.15 [0.09–0.25] 31.19 [5.23–17.08] 0.89 [0.86–0.92]

�S3BMI ≥28 7 0.75 [0.69–0.80] 0.63 [0.52–0.73] 2.03 [1.60–2.58] 0.40 [0.33–0.47] 5.13 [3.62–7.26] 0.77 [0.73–0.80]BMI <28 8 0.86 [0.74–0.93] 0.62 [0.54–0.69] 2.25 [1.90–2.67] 0.23 [0.13–0.42] 9.75 [5.05–18.81] 0.78 [0.74–0.81]Cut-off ≥300 11 0.75 [0.70–0.79] 0.64 [0.56–0.71] 2.10 [1.74–2.54] 0.39 [0.33–0.46] 5.40 [3.98–7.32] 0.77 [0.73–0.80]Cut-off <300 4 0.91 [0.85–0.95] 0.83 [0.77–0.87] 2.31 [1.84–2.89] 0.06 [0.01–0.73] 37.38 [3.29–424.09] 0.94 [0.92–0.96]

LR: likelihood ratio, DOR: diagnostic odds ratio, AUC: area under the curve, and CI: confidence interval.


and NAFLD, this meta-analysis focused on the population ofALD and NAFLD patients and consisted of two parts. First,we studied the accuracy of TE-measured CAP in the di-agnosis and grading of steatosis in patients with ALD/NAFLD.We found that the average cut-off values of CAP foridentifying patients with steatosis grades ≥S1, ≥S2, and �S3were 272 dB/m, 292 dB/m, and 308 dB/m, respectively.Previous studies obtained different results regarding CAPmeasuring the steatosis grades of ALD/NAFLD patients.Early research showed the excellent diagnostic performanceof CAP in steatosis, with AUCs of 0.91, 0.95, and 0.89 forsteatosis grades ≥S1, ≥S2, and �S3, respectively [21]. Sub-sequent research failed to repeat such excellent diagnosticefficiency, andmost studies showed high diagnostic accuracyfor steatosis grades ≥S1, while a decline in AUCs was ob-served in the diagnostic performance for steatosis grades ≥S2and �S3 [12, 22–24]. In our meta-analysis, the diagnostic

accuracy of CAP decreased as the steatosis grades increased,with the highest AUC of 0.90 obtained for steatosis grades ≥S1, but the AUCs were 0.83 for steatosis grades ≥S2 and 0.79for steatosis grade S3. (e pooled results revealed that thesensitivity and specificity of CAP were 0.84, 0.83, and 0.78and 0.83, 0.71, and 0.62 for steatosis grades ≥ S1, ≥S2, and�S3, respectively. (e reason for the reduction in the ac-curacy of CAP in grading steatosis as the steatosis deteri-orated is still unclear but might relate to the characteristics ofthe study population, especially the ethnicity, BMI, andproportion of patients with severe steatosis [25]. Furthersubgroup analyses indicated that CAP exhibited higherdiagnostic accuracy in studies whose population BMI was<28 kg/m2 than in studies whose population BMI was≥28 kg/m2. CAP in the studies with thresholds <270 dB/m,<290 dB/m, and <300 dB/m had higher accuracy in thediagnosis of steatosis grades ≥S1, ≥S2, and �S3, respectively,

Sensitivity (95% CI)Study (year)

Nguyen-Khac 2008ALD/

Nobili 2008 NAFLD M

Kumar 2013 NAFLD M

Imajo 2016 NAFLD M

Lee 2016 NAFLD M/XL

Siddiqui 2019 NAFLD M/XL

Chan 2017 NAFLD XL

Garg 2018 NAFLD XL

Chan 2017 NAFLD M

Karlas 2008 NAFLD M/XL

Gaia 2011 NAFLD M/XL

Combined

I2 = 91.62 [87.95 –95.29]

0.4 1.0

Sensitivity

0.83 [0.74 – 0.90]

0.97 [0.87 – 1.00]

0.86 [0.76 – 0.93]

0.57 [0.42 – 0.71]

0.78 [0.71 – 0.84]

0.62 [0.53 – 0.70]

0.66 [0.57 – 0.75]

0.85 [0.69 – 0.95]

0.79 [0.62 – 0.91]

0.80 [0.68 – 0.89]

0.53 [0.47 – 0.59]

0.77 [0.68 – 0.84]

Q = 119.36, df = 10.00, p = 0.00

(a)


Nguyen-Khac 2008ALD/

Nobili 2008 NAFLD M

Kumar 2013 NAFLD M

Imajo 2016 NAFLD M

Lee 2016 NAFLD M/XL


Chan 2017 NAFLD XL

Garg 2018 NAFLD XL

Chan 2017 NAFLD M


Gaia 2011 NAFLD M/XL

Combined

Q = 36.55, df = 10.00, p = 0.00

0.88 [0.47 – 1.00]

0.91 [0.59 – 1.00]

0.89 [0.65 – 0.99]

0.78 [0.56 – 0.93]

0.68 [0.59 – 0.76]

1.00 [0.77 – 1.00]

0.86 [0.76 – 0.93]

0.76 [0.60 – 0.87]

0.80 [0.65 – 0.90]

0.56 [0.30 – 0.80]

0.87 [0.79 – 0.93]

0.80 [0.73 – 0.86]

Specificity0.3 1.0

I2 = 72.64 [56.06 –89.22]

(b)

SROC with confidence and predictive ellipses1.0

0.5

0.0

Sens

itivi

ty

1.0 0.5 0.0Specificity

Observed data

Summary operating pointSENS = 0.77 [0.68 – 0.84]SPEC = 0.80 [0.73 – 0.86]

SROC curveAUC = 0.85 [0.82 – 0.88]


(c)


Nobili 2008 NAFLD M



Gaia 2011 NAFLD M/XLPetta 2011 NAFLD M

CombinedQ = 181.98, df = 39.00, p = 0.00

Chang 2018 NASH M/XLMyers 2010 NAFLD/

Ergelen 2015 NAFLD M/XL


Attia 2016 NAFLD M/XLAttia 2016 NAFLD M/XL

Lee 2017 NAFLD MNguyen-Khac 2008 ALD/

De Ledinghen 2009 NAFLD M/XLLupsor 2010 NASH/

Mahadeva 2010 NAFLD/

Mahadeva 2013 NAFLD M

Wong 2010 NAFLD M/XL

Wong 2012 NAFLD XL

Alkhouri 2013 NAFLD SKumar 2013 NAFLD M

Wong 2012 NAFLD M

Aykut 2014 NAFLD M/XLKarlas 2014 NAFLD M

Naveau 2014 NAFLD M/XLVuppalanchi 2014 NAFLD/

Dincses 2015 NAFLD/Cassinotto 2016 NAFLD MCassinotto 2016 NAFLD M

Imajo 2016 NAFLD MLee 2016 NAFLD M/XL

Thiele 2016 ALD/Chan 2017 NAFLD MChan 2017 NAFLD M

Loong 2017 NAFLD MLoong 2017 NAFLD M

Petta 2017 NAFLD /Garg 2018 NAFLD XL

Ooi 2018 NAFLD XLEddowes 2019 NAFLD M/XL

0.2 1.0

Sensitivity

0.94 [0.73 – 1.00]0.94 [0.80 – 0.99]0.68 [0.49 – 0.83]0.83 [0.52 – 0.98]0.84 [0.64 – 0.95]0.91 [0.75 – 0.98]0.63 [0.48 – 0.77]0.81 [0.70 – 0.89]1.00 [0.74 – 1.00]

0.77 [0.66 – 0.86]0.67 [0.41 – 0.87]0.80 [0.52 – 0.96]0.79 [0.70 – 0.87]0.79 [0.61 – 0.91]0.69 [0.57 – 0.80]0.57 [0.46 – 0.68]0.79 [0.67 – 0.88]1.00 [0.69 – 1.00]0.77 [0.69 – 0.84]0.59 [0.47 – 0.70]0.75 [0.60 – 0.87]1.00 [0.63 – 1.00]0.73 [0.50 – 0.89]0.80 [0.69 – 0.89]0.75 [0.51 – 0.91]0.60 [0.51 – 0.67]0.90 [0.85 – 0.95]0.65 [0.53 – 0.75]0.83 [0.69 – 0.92]0.83 [0.74 – 0.91]0.46 [0.19 – 0.75]0.85 [0.55 – 0.98]0.65 [0.53 – 0.76]0.93 [0.84 – 0.98]0.74 [0.68 – 0.80]0.70 [0.51 – 0.85]1.00 [0.59 – 1.00]0.71 [0.64 – 0.77]0.66 [0.59 – 0.73]

0.77 [0.73 – 0.81]

0.88 [0.76 – 0.96]

I2 = 78.57 [72.29 –84.85]

(d)


CombinedQ = 343.37, df = 39.00, p = 0.00

Nobili 2008 NAFLD M



Gaia 2011 NAFLD M/XLPetta 2011 NAFLD M

Chang 2018 NASH M/XLMyers 2010 NAFLD/




Lee 2017 NAFLD MNguyen-Khac 2008 ALD/


Mahadeva 2010 NAFLD/

Mahadeva 2013 NAFLD M

Wong 2010 NAFLD M/XL

Wong 2012 NAFLD XL

Alkhouri 2013 NAFLD SKumar 2013 NAFLD M

Wong 2012 NAFLD M

Aykut 2014 NAFLD M/XLKarlas 2014 NAFLD M

Naveau 2014 NAFLD M/XLVuppalanchi 2014 NAFLD/

Dincses 2015 NAFLD/Cassinotto 2016 NAFLD MCassinotto 2016 NAFLD M

Imajo 2016 NAFLD MLee 2016 NAFLD M/XL

Thiele 2016 ALD/Chan 2017 NAFLD MChan 2017 NAFLD M

Loong 2017 NAFLD MLoong 2017 NAFLD M

Petta 2017 NAFLD /Garg 2018 NAFLD XL

Ooi 2018 NAFLD XLEddowes 2019 NAFLD M/XL

0.76 [0.58 – 0.89]0.59 [0.33 – 0.82]0.91 [0.79 – 0.97]0.79 [0.49 – 0.95]0.81 [0.64 – 0.92]0.90 [0.74 – 0.98]0.92 [0.80 – 0.98]0.92 [0.75 – 0.99]0.92 [0.79 – 0.98]0.74 [0.59 – 0.86]0.77 [0.68 – 0.84]0.84 [0.71 – 0.93]0.90 [0.55 – 1.00]0.76 [0.68 – 0.83]0.74 [0.58 – 0.87]0.71 [0.59 – 0.80]0.90 [0.83 – 0.95]0.64 [0.54 – 0.74]1.00 [0.94 – 1.00]0.78 [0.71 – 0.84]0.70 [0.56 – 0.81]0.93 [0.81 – 0.99]0.98 [0.87 – 1.00]0.78 [0.67 – 0.87]0.75 [0.55 – 0.89]0.78 [0.60 – 0.91]0.90 [0.80 – 0.96]0.45 [0.33 – 0.57]0.89 [0.79 – 0.96]0.85 [0.78 – 0.90]0.91 [0.85 – 0.96]0.94 [0.85 – 0.98]0.89 [0.79 – 0.96]0.88 [0.81– 0.93]0.42 [0.34 – 0.50]0.73 [0.63 – 0.82]0.59 [0.43 – 0.73]0.74 [0.67 – 0.81]0.70 [0.62 – 0.77]0.80 [0.73 – 0.85]

0.82 [0.78 – 0.86]

0.3 1.0Specificity

I2 = 88.64 [85.86 –91.42]

(e)

SROC with confidence and predictive ellipses1.0

0.5

0.0

Sens

itivi

ty


Observed data


SROC curveAUC = 0.86 [0.83 – 0.89]


(f )

Figure 3: Diagnostic performance of LSM for liver fibrosis in alcoholic liver disease/nonalcoholic fatty liver disease. (a) Sensitivity, (b)specificity, and (c) summary receiver-operating characteristics (SROC) curve of LSM for identifying patients with fibrosis grade ≥F1. (d)Sensitivity, (e) specificity, and (f) SROC curve of LSM for identifying patients with fibrosis grade ≥F2. LSM: liver stiffness measurement;SROC: summary receiver-operating characteristics.


than that in the study with threshold ≥270 dB/m, ≥290 dB/m, and ≥300 dB/m.(is finding might be associated with therelatively higher BMI of the population in studies withhigher CAP cut-off values because higher BMI leads to lesssatisfactory results. (e disease was the only possible factorrelated to the heterogeneity when exploring the diagnosticperformance of CAP for steatosis grades ≥S1 in ALD/NAFLD patients, and separate analyses showed higher di-agnostic accuracy of CAP in NAFLD patients than in ALDpatients (sensitivity 0.85 vs. 0.70, specificity 0.84 vs. 0.73, andAUC 0.91 vs. 0.77).

Second, we analyzed the efficiency of TE-measured LSMin the diagnosis and grading of liver fibrosis in patients withALD/NAFLD. (e average cut-offs for LSM were 6.3 kPa,8.2 kPa, 13.4 kPa, and 14.2 kPa for patients with fibrosisgrades ≥F1, ≥F2, ≥F3, and �F4, respectively. Previous studies

have shown that TE exhibited a high degree of accuracy inthe diagnosis of liver fibrosis by measuring LSM, especiallyin cases of severe fibrosis and cirrhosis [12, 25–27]. Con-sistent with these results, our pooled data demonstrated thatthe AUCs of LSM for diagnosing fibrosis grades ≥F1, ≥F2,≥F3, and �F4 were 0.85, 0.86, 0.90, and 0.95, respectively.(e sensitivity and specificity of LSM for identifying fibrosisgrades ≥ F1, ≥F2, ≥F3, and �F4 were 0.77, 0.77, 0.83, and 0.91and 0.80, 0.82, 0.84, and 0.86, respectively. Subsequentsubgroup analyses revealed that patients with BMI <28 kg/m2 were more sensitive to TE than patients with BMI≥28 kg/m2, with the data showing that TE had higher ac-curacy in the lower BMI population than in the higher BMIpopulation. Meta-regression analyses identified that diseasemight be the reason for the heterogeneity when studying thediagnostic performance of LSM for fibrosis grades ≥F3 in

Kim 2009 ALD MJanssens 2010 ALD/

Myers 2010 NAFLD/Yoneda 2010 NAFLD/Fernandez 2012 ALD/

CombinedQ = 600.90, df = 50.00, p = 0.00

I2 = 91.68 [90.04 – 93.32]

0.97 [0.85 – 1.00]0.72 [0.53 – 0.86]0.69 [0.41 – 0.89]1.00 [0.69 – 1.00]0.91 [0.81 – 0.96]0.84 [0.60 – 0.97]0.90 [0.55 – 1.00]0.79 [0.49 – 0.95]0.81 [0.62 – 0.94]0.81 [0.72 – 0.88]0.87 [0.80 – 0.93]0.87 [0.75 – 0.95]1.00 [0.48 – 1.00]0.85 [0.66 – 0.96]0.84 [0.70 – 0.93]1.00 [0.48 – 1.00]0.91 [0.79 – 0.98]0.84 [0.72 – 0.92]0.82 [0.57 – 0.96]0.76 [0.58 – 0.89]0.78 [0.64 – 0.88]0.83 [0.69 – 0.93]0.85 [0.77 – 0.91]0.69 [0.49 – 0.85]0.96 [0.81 – 1.00]1.00 [0.66 – 1.00]1.00 [0.69 – 1.00]0.91 [0.81 – 0.97]0.89 [0.75 – 0.97]0.68 [0.43 – 0.87]0.86 [0.57 – 0.98]0.72 [0.59 – 0.82]0.88 [0.83 – 0.93]0.90 [0.82 – 0.95]0.87 [0.73 – 0.95]0.96 [0.82 – 1.00]0.95 [0.76 – 1.00]0.88 [0.47 – 1.00]0.88 [0.47 – 1.00]0.84 [0.69 – 0.93]0.75 [0.68 – 0.82]0.72 [0.60 – 0.83]0.74 [0.68 – 0.80]0.77 [0.69 – 0.85]0.90 [0.77 – 0.97]0.92 [0.84 – 0.97]0.64 [0.31 – 0.89]0.71 [0.62 – 0.78]0.80 [0.72 – 0.87]0.27 [0.21 – 0.35]0.34 [0.26 – 0.42]

0.83 [0.79 – 0.86]

Siddiqui 2019 NAFLD M/XLEddowes 2019 NAFLD M/XL

Garg 2018 NAFLD XL

Wong 2019 NAFLD XLWong 2019 NAFLD M

Voican 2017 ALD MSeki 2017 NAFLD M

Petta 2017 NALFD/Petta 2017 NALFD MPetta 2017 NALFD MPetta 2017 NALFD M

Loong 2017 NALFD MChan 2017 NALFD M

Chan 2017 NALFD XLTapper 2016 NALFD MLee 2016 NAFLD M/XL

Imaio 2016 NAFLD MCassinotto 2016 NAFLD M

Boursier 2016 NAFLD MWong 2016 NAFLD/

Petta 2015 NAFLD MPetta 2015 NAFLD M

Pathik 2015 NAFLD MFernandez 2015 ALD/Dincses 2015 NAFLD/

Naveau 2014 NAFLD M/XLAykut 2014 NAFLD M/XL

Mahadeva 2013 NAFLD MKumar 2013 NAFLD MWong 2012 NAFLD M

Wong 2012 NAFLD XLPetta 2011 NAFLD M

Gaia 2011 NAFLD M/XLWong 2010 NAFLD M/XL

Mueller 2010 ALD M



Lee 2017 NAFLD M�iele 2018 ALD/

Nahon 2008 ALD/Nguyen-Khac 2008 ALD/

Nobili 2008 NAFLD MKarlas 2008 NAFLD M/XL



0.2 1.0

Sensitivity

(a)

Kim 2009 ALD MJanssens 2010 ALD/

Myers 2010 NAFLD/Yoneda 2010 NAFLD/Fernandez 2012 ALD/


Garg 2018 NAFLD XL


Voican 2017 ALD MSeki 2017 NAFLD M

Petta 2017 NALFD/Petta 2017 NALFD MPetta 2017 NALFD MPetta 2017 NALFD M

Loong 2017 NALFD MChan 2017 NALFD M

Chan 2017 NALFD XLTapper 2016 NALFD MLee 2016 NAFLD M/XL


Boursier 2016 NAFLD MWong 2016 NAFLD/

Petta 2015 NAFLD MPetta 2015 NAFLD M

Pathik 2015 NAFLD MFernandez 2015 ALD/Dincses 2015 NAFLD/

Naveau 2014 NAFLD M/XLAykut 2014 NAFLD M/XL


Wong 2012 NAFLD XLPetta 2011 NAFLD M


Mueller 2010 ALD M



Lee 2017 NAFLD M�iele 2018 ALD/


Nobili 2008 NAFLD MKarlas 2008 NAFLD M/XL


Study (year)

0.87 [0.81 – 0.92]

0.67 [0.55 – 0.78]

0.97 [0.90 – 1.00]

0.78 [0.72 – 0.84]

0.75 [0.69 – 0.80]

CombinedQ = 513.89, df = 50.00, p = 0.00

I2 = 90.27 [88.26 – 92.28]

0.84 [0.81 – 0.87]

0.96 [0.93 – 0.98]0.98 [0.95 – 0.99]0.74 [0.68 – 0.79]

0.88 [0.77 – 0.95]0.76 [0.67 – 0.83]0.88 [0.81 – 0.93]

0.81 [0.78 – 0.84]0.86 [0.81 – 0.90]0.78 [0.73 – 0.82]0.87 [0.81 – 0.91]

0.97 [0.90 – 1.00]0.77 [0.67 – 0.85]0.86 [0.79 – 0.91]0.84 [0.75 – 0.90]0.61 [0.52 – 0.69]0.63 [0.57 – 0.69]0.86 [0.81 – 0.90]1.00 [0.96 – 1.00]0.86 [0.77 – 0.93]0.81 [0.70 – 0.89]

0.76 [0.61 – 0.88]0.74 [0.63 – 0.82]0.90 [0.80 – 0.96]0.67 [0.57 – 0.76]0.88 [0.83 – 0.92]0.78 [0.69 – 0.85]0.78 [0.70 – 0.85]0.78 [0.69 – 0.85]0.47 [0.34 – 0.61]0.83 [0.77 – 0.88]0.75 [0.62 – 0.86]0.97 [0.89 – 1.00]

0.81 [0.70 – 0.90]1.00 [0.92 – 1.00]0.80 [ 0.66 – 0.90]0.89 [0.75 – 0.97]0.94 [0.89 – 0.97]0.85 [0.74 – 0.93]0.94 [0.82 – 0.99]0.81 [0.54 – 0.96]0.76 [0.65 – 0.86]0.68 [0.55 – 0.79]0.93 [0.81 – 0.99]0.76 [0.59 – 0.89]0.76 [0.50 – 0.93]0.78 [0.40 – 0.97]

Sensitivity (95% CI)

Specificity0.3 1.0

(b)

1.0

0.5

0.0

Sens

itivi

ty


Observed data


SROC curveAUC = 0.95 [0.92 – 0.96]



(c)

Lee 2017 NAFLD MAttia 2016 NAFLD M/XLAttia 2016 NAFLD M/XL

Fernandez 2012 ADL/Yoneda 2010 NAFLD/

Myers 2010 NAFLD/Janssens 2010 ALD/

Kim 2009 ALD MLemoine 2008 NAFLD/

Chang 2018 NASH M/XL




Voican 2017 ALD MChan 2017 NALFD M

Chan 2017 NALFD XL�iele 2016 ALD/

Mota 2016 NASH M/XLLee 2016 NAFLD M/XL


Fernandez 2015 ALD/

CombinedQ = 177.98, df = 33.00, p = 0.00

I2 = 81.46 [75.79 – 87.13]

0.89 [0.78 – 0.96]

0.91 [0.87 – 0.94]

0.54 [0.41 – 0.67]0.47 [0.35 – 0.59]0.89 [0.74 – 0.97]0.85 [0.69 – 0.95]0.86 [0.68 – 0.96]1.00 [0.29 – 1.00]1.00 [0.29 – 1.00]0.97 [0.85 – 1.00]0.83 [0.59 – 0.96]1.00 [0.80 – 1.00]1.00 [0.72 – 1.00]0.92 [0.78 – 0.98]0.90 [0.77 – 0.97]1.00 [0.66 – 1.00]0.88 [0.47 – 1.00]0.91 [0.83 – 0.96]0.81 [0.54 – 0.96]1.00 [0.66 – 1.00]0.92 [0.74 – 0.99]1.00 [0.87 – 1.00]0.95 [0.75 – 1.00]1.00 [0.66 – 1.00]0.85 [0.68 – 0.95]0.84 [0.74 – 0.91]0.93 [0.66 – 1.00]1.00 [0.54 – 1.00]1.00 [0.63 – 1.00]

1.00 [0.54 – 1.00]1.00 [0.54 – 1.00]0.80 [0.56 – 0.94]1.00 [0.88 – 1.00]0.90 [0.76 – 0.97]1.00 [0.54 – 1.00]

Aykut 2014 NAFLD M/XLMahadeva 2013 NAFLD M

Kumar 2013 NAFLD MWong 2012 NAFLD M


Mueller 2010 ALD MDe Ledinghen 2009 NAFLD M/XL



0.3 1.0

Sensitivity

(d)

Lee 2017 NAFLD MAttia 2016 NAFLD M/XLAttia 2016 NAFLD M/XL

Fernandez 2012 ADL/Yoneda 2010 NAFLD/

Myers 2010 NAFLD/Janssens 2010 ALD/

Kim 2009 ALD MLemoine 2008 NAFLD/

Chang 2018 NASH M/XL



Voican 2017 ALD MChan 2017 NALFD M

Chan 2017 NALFD XL�iele 2016 ALD/

Mota 2016 NASH M/XLLee 2016 NAFLD M/XL


Fernandez 2015 ALD/Aykut 2014 NAFLD M/XL



Mueller 2010 ALD MDe Ledinghen 2009 NAFLD M/XL




CombinedQ = 284.28, df = 33.00, p = 0.00

I2 = 88.39 [85.29 – 91.50]

0.81 [0.71 – 0.91]

0.86 [0.83 – 0.89]

0.92 [0.89 – 0.95]0.95 [0.93 – 0.97]0.86 [0.82 – 0.89]0.79 [0.74 – 0.83]0.93 [0.88 – 0.97]0.96 [0.89 – 0.99]0.95 [0.87 – 0.99]

0.66 [0.49 – 0.80]0.66 [0.85 – 0.94]

0.90 [0.84 – 0.94]0.76 [0.67 – 0.83]0.62 [0.55 – 0.69]0.86 [0.78 – 0.92]0.76 [0.65 – 0.85]0.89 [0.83 – 0.94]0.88 [0.83 – 0.92]0.83 [0.76 – 0.89]0.70 [0.57 – 0.81]0.88 [0.83 – 0.92]0.77 [0.66 – 0.86]0.88 [0.82 – 0.92]1.00 [0.96 – 1.00]0.84 [0.74 – 0.92]

0.81 [0.71 – 0.89]0.93 [0.82 – 0.98]0.83 [0.59 – 0.96]0.87 [0.78 – 0.93]0.98 [0.89 – 1.00]0.82 [0.67 – 0.92]0.76 [0.56 – 0.90]0.50 [0.25 – 0.75]0.88 [0.47 – 1.00]0.80 [0.65 – 0.90]

Specificity0.2 1.0

(e)

1.0

0.5

0.0

Sens

itivi

ty


Observed data


SROC curveAUC = 0.95 [0.92 – 0.96]



(f )

Figure 4: Diagnostic performance of liver stiffness measurement (LSM) for liver fibrosis in alcoholic liver disease/nonalcoholic fatty liverdisease. (a) Sensitivity, (b) specificity, and (c) summary receiver-operating characteristics (SROC) curve of LSM for identifying patients withfibrosis grade ≥F3. (d) Sensitivity, (e) specificity, and (f) SROC curve of LSM for identifying patients with fibrosis grade�F4. LSM: liverstiffness measurement; SROC: summary receiver-operating characteristic.


Tabl

e2:

Summarydiagno

stic

accuracy

parameter

estim

ates

andtheir95%

confi

denceintervals.

Subgroup

Sensitivity

(95%

CI)

Specificity

(95%

CI)

PositiveLR

(95%

CI)

NegativeLR

(95%

CI)

DOR(95%

CI)

AUC

F2–F

4BM

I≥2

8kg/m

221

0.74

[0.69–

0.78]

0.78

[0.73–

0.83]

3.42

[2.80–

4.18]

0.33

[0.29–

0.39]

10.24[7.94–13.21]

0.82

[0.79–

0.85]

BMI<2

8kg/m

214

0.80

[0.72–

0.86]

0.85

[0.77–

0.90]

5.26

[3.56–

7.79]

0.24

[0.18–

0.32]

22.06[13.64–3

5.69]

0.89

[0.86–

0.91]

Prospective

370.77

[0.73–

0.81]

0.82

[0.78–

0.86]

4.34

[3.45–

5.46]

0.28

[0.23–

0.33]

15.65[11.33–21.60]

0.86

[0.83–

0.89]

Retrospective

30.74

[0.65–

0.81]

0.82

[0.73–

0.88]

3.51

[1.94–

6.34]

0.31

[0.18–

0.54]

14.74[7.41–

29.32]

0.87

[0.83–

0.90]

F3–F

4NAFL

D42

0.81

[0.77–

0.85]

0.85

[0.82–

0.88]

10.09[2.16–

47.13]

0.14

[0.04–

0.53]

25.19[18.15–3

4.95]

0.90

[0.87–

0.92]

ALD

90.80

[0.72–

0.86]

0.89

[0.85–

0.92]

5.53

[4.44–

6.88]

0.22

[0.18–

0.27]

31.19[5.23–17.08]

0.92

[0.89–

0.94]

BMI≥2

8kg/m

225

0.80

[0.73–

0.85]

0.84

[0.79–

0.88]

5.00

[3.91–

6.40]

0.24

[0.18–

0.32]

20.57[14.59–2

9.00]

0.89

[0.86–

0.91]

BMI<2

8kg/m

223

0.86

[0.83–

0.89]

0.86

[0.82–

0.90]

6.26

[4.67–

8.37]

0.16

[0.13–

0.20]

38.28[25.84–5

6.70]

0.92

[0.89–

0.94]

Prospective

430.82

[0.78–

0.86]

0.84

[0.81–

0.88]

5.30

[4.34–

6.48]

0.21

[0.17–

0.26]

25.20[18.68–3

9.99]

0.90

[0.87–

0.93]

Retrospective

80.80

[0.72–

0.86]

0.89

[0.85–

0.92]

5.53

[4.44–

6.88]

0.22

[0.18–

0.27]

31.19[5.23–17.08]

0.92

[0.89–

0.94]

�F4 N

AFL

D25

0.92

[0.86–

0.96]

0.87

[0.83–

0.90]

7.18

[5.44–

9.47]

0.09

[0.05–

0.17]

78.59[39.17–157.70]

0.95

[0.93–

0.97]

ALD

90.92

[0.84–

0.96]

0.84

[0.76–

0.89]

5.64

[3.83–

8.30]

0.10

[0.05–

0.20]

57.52[26.27–125.92]

0.94

[0.92–

0.96]

BMI≥2

8kg/m

212

0.88

[0.78–

0.94]

0.85

[0.80–

0.89]

5.91

[4.51–

7.75]

0.14

[0.07–

0.26]

43.04[23.05–8

0.35]

0.92

[0.90–

0.94]

BMI<2

8kg/m

217

0.92

[0.86–

0.95]

0.88

[0.82–

0.92]

7.60

[4.97–11.61]

0.09

[0.06–

0.16]

80.28[38.22–168.63]

0.95

[0.93–

0.97]

Prospective

290.92

[0.86–

0.95]

0.87

[0.83–

0.90]

6.90

[5.36–

8.88]

0.10

[0.06–

0.16]

71.57[39.82–128.65]

0.95

[0.93–

0.96]

Retrospective

50.91

[0.85–

0.95]

0.83

[0.770.87]

5.29

[3.90–

7.17]

0.10

[0.06–

0.19]

51.00[24.71–1

05.24]

0.94

[0.92–

0.96]

LR:likelihoo

dratio

,DOR:

diagno

stic

odds

ratio

,AUC:a

reaun

derthecurve,andCI:confi

denceinterval.


ALD/NAFLD patients, but subgroup analyses showed thatthere was no obvious difference in TE diagnostic accuracybetween analyses with ALD patients and NAFLD patients(sensitivity 0.81 vs. 0.80, specificity 0.85 vs. 0.89, and AUC0.90 vs. 0.92).

Although we are trying to comprehensively analyze thevalue of TE in the context of ALD/NAFLD, there are stillmany weaknesses in this meta-analysis. First, the CAP andLSM of all the studies we included had different thresholds.Although we calculated the average cut-offs of CAP for thediagnosis of every steatosis grade and the average cut-offs ofLSM for the determination of every fibrosis grade, we couldnot obtain the best cut-offs for CAP and LSM for diagnosis.Second, we conducted a comprehensive electronic search andmanual search of multiple databases. However, in the sub-sections of our study, significant publication bias was de-tected, which indicated that some failed or poorly performedresearch was missed or not included in SCI journals.(ird, inthe analysis process, many analyses had significant hetero-geneity, and meta-regression could not find all sources ofheterogeneity.(e pooled results were relatively more reliableafter excluding the effects of a single factor by subgroupanalysis. In general, the existing heterogeneity does not affectour interpretation of the results. Despite these shortcomingsin this study, we can still draw some very valuable and strongconclusions from this analysis: (1) this is the first meta-analysis to comprehensively study the diagnostic accuracy ofCAP and LSM for steatosis and fibrosis in ALD/NAFLDpatients; (2) the accuracy of TE used for diagnosing steatosisor fibrosis between ALD and NAFLD patients was equivalent;(3) when assessing the accuracy of CAP to diagnose differentgrades of steatosis in ALD/NAFLD patients, we found thatCAP measured by TE was good at screening for patients withfatty liver but failed to maintain high accuracy in diagnosingpatient with severe steatosis; (4) when evaluating the accuracyof LSM to diagnose different grades of fibrosis in ALD/NAFLD patients, we found that TE was accurate for diag-nosing fibrosis, especially severe fibrosis and cirrhosis; and (5)because of the large number of studies and participants in ourmeta-analysis, the results are convincing and can guideclinical practice.

In conclusion, this study comprehensively analyzed theapplicability of TE for diagnosing steatosis and fibrosis inALD/NAFLD patients. (e pooled results indicated that TEexhibited similar diagnostic efficiency for ALD and NAFLDpatients. In patients with ALD/NAFLD, CAPwas feasible foridentifying steatosis, and LSM was accurate for diagnosingfibrosis, especially severe fibrosis and cirrhosis.

Data Availability

All data of this study are available from the correspondingauthor upon request by email.

Disclosure

(e abstract of this article was presented as poster in theconference of Digestive Disease Week and published in theGastroenterology. (e funders did not play any role in the

study design, data collection, and analysis, decisions re-garding data release, or manuscript preparation

Conflicts of Interest

(e authors declare that they have no conflicts of interest.

Authors’ Contributions

FJ conceived the study; CC and XS performed the study,analyzed the data, and wrote the manuscript. SC acted as thethird reviewer. XS, WZ, and QJ revised the manuscriptcritically. All authors read and approved the final manuscript.

Acknowledgments

(is study was funded by the National Natural ScienceFoundation of China (grant no. 81700485).

Supplementary Materials

Table S1: characteristics of the included studies. Table S2: themeta-regression analysis of CAP detecting patients withsteatosis grade ≥S1. Table S3: the meta-regression analysis ofCAP detecting patients with steatosis grade ≥S2. Table S4:the meta-regression analysis of CAP detecting patients withsteatosis grade �S3. Table S5: the meta-regression analysis ofCAP detecting patients with fibrosis grade ≥F1. Table S6: themeta-regression analysis of CAP detecting patients withfibrosis grade ≥F2. Table S7: the meta-regression analysis ofCAP detecting patients with fibrosis grade ≥F3. Table S8: themeta-regression analysis of CAP detecting patients withfibrosis grade �F4. Figure S1: overall methodology qualityassessment of included articles using the QUADAS-2 cri-teria. Figure S2: diagnostic performance of CAP for steatosisgrade ≥S1 in alcoholic liver disease/nonalcoholic fatty liverdisease. A: positive likelihood; B: negative likelihood; C:diagnostic odds ratio; D: Deeks’ funnel plots; E: Fagan’snomogram. Figure S3: diagnostic performance of CAP forsteatosis grade ≥S2 in alcoholic liver disease/nonalcoholicfatty liver disease. A: positive likelihood; B: negative likeli-hood; C: diagnostic odds ratio; D: Deeks’ funnel plots; E:Fagan’s nomogram. Figure S4: diagnostic performance ofCAP for steatosis grade �S3 in alcoholic liver disease/nonalcoholic fatty liver disease. A: positive likelihood; B:negative likelihood; C: diagnostic odds ratio; D: Deeks’funnel plots; E: Fagan’s nomogram. Figure S5: diagnosticperformance of LSM for fibrosis grade ≥F1 in alcoholic liverdisease/nonalcoholic fatty liver disease. A: positive likeli-hood; B: negative likelihood; C: diagnostic odds ratio; D:Deeks’ funnel plots; E: Fagan’s nomogram. Figure S6: di-agnostic performance of LSM for fibrosis grade ≥F2 in al-coholic liver disease/nonalcoholic fatty liver disease. A:positive likelihood; B: negative likelihood; C: diagnostic oddsratio; D: Deeks’ funnel plots; E: Fagan’s nomogram. FigureS7: diagnostic performance of LSM for fibrosis grade ≥F3 inalcoholic liver disease/nonalcoholic fatty liver disease. A:positive likelihood; B: negative likelihood; C: diagnostic oddsratio; D: Deeks’ funnel plots; E: Fagan’s nomogram. FigureS8: diagnostic performance of LSM for fibrosis grade �F4 in


alcoholic liver disease/nonalcoholic fatty liver disease. A:positive likelihood; B: negative likelihood; C: diagnostic oddsratio; D: Deeks’ funnel plots; E: Fagan’s nomogram. (Sup-plementary Materials)

References

[1] W.-K. Chan, S. Treeprasertsuk, G. B.-B. Goh et al., “Opti-mizing use of nonalcoholic fatty liver disease fibrosis score,fibrosis-4 score, and liver stiffness measurement to identifypatients with advanced fibrosis,” Clinical Gastroenterologyand Hepatology: �e Official Clinical Practice Journal of theAmerican Gastroenterological Association, vol. 17, no. 12,pp. 2570–2580, 2019.

[2] M. R. Charlton, J. M. Burns, R. A. Pedersen, K. D. Watt,J. K. Heimbach, and R. A. Dierkhising, “Frequency andoutcomes of liver transplantation for nonalcoholic steato-hepatitis in the United States,” Gastroenterology, vol. 141,no. 4, pp. 1249–1253, 2011.

[3] E. M. Brunt, D. E. Kleiner, L. A. Wilson, P. Belt, andB. A. Neuschwander-Tetri, “Nonalcoholic fatty liver disease(NAFLD) activity score and the histopathologic diagnosis inNAFLD: distinct clinicopathologic meanings,” Hepatology,vol. 53, no. 3, pp. 810–820, 2011.

[4] L. A. Adams, J. F. Lymp, J. St. Sauver et al., “(e natural historyof nonalcoholic fatty liver disease: a population-based cohortstudy,” Gastroenterology, vol. 129, no. 1, pp. 113–121, 2005.

[5] S. Mueller, G. Millonig, L. Sarovska et al., “Increased liverstiffness in alcoholic liver disease: differentiating fibrosis fromsteatohepatitis,” World Journal of Gastroenterology, vol. 16,no. 8, pp. 966–972, 2010.

[6] J.-F. Cadranel, P. Rufat, and F. Degos, “Practices of liverbiopsy in France: results of a prospective nationwide survey,”Hepatology, vol. 32, no. 3, pp. 477–481, 2000.

[7] M. (iele, V. Rausch, G. Fluhr et al., “Controlled attenuationparameter and alcoholic hepatic steatosis: diagnostic accuracyand role of alcohol detoxification,” Journal of Hepatology,vol. 68, no. 5, pp. 1025–1032, 2018.

[8] H. Garg, S. Aggarwal, Y. R Shalimar et al., “Utility of transientelastography (fibroscan) and impact of bariatric surgery onnonalcoholic fatty liver disease (NAFLD) in morbidly obesepatients,” Surgery for Obesity and Related Diseases, vol. 14,no. 1, pp. 81–91, 2018.

[9] P. J. Eddowes, M. Sasso, M. Allison et al., “Accuracy ofFibroScan controlled attenuation parameter and liver stiffnessmeasurement in assessing steatosis and fibrosis in patientswith non-alcoholic fatty liver disease,” Gastroenterology,vol. 156, no. 6, pp. 1717–1730, 2019.

[10] M. Kumar, A. Rastogi, T. Singh et al., “Controlled attenuationparameter for non-invasive assessment of hepatic steatosis:does etiology affect performance?” Journal of Gastroenterologyand Hepatology, vol. 28, no. 7, pp. 1194–1201, 2013.

[11] S. Naveau, C. S. Voican, A. Lebrun et al., “Controlled at-tenuation parameter for diagnosing steatosis in bariatricsurgery candidates with suspected nonalcoholic fatty liverdisease,” European Journal of Gastroenterology & Hepatology,vol. 29, no. 9, pp. 1022–1030, 2017.

[12] H. W. Lee, S. Y. Park, S. U. Kim et al., “Discrimination ofnonalcoholic steatohepatitis using transient elastography inpatients with nonalcoholic fatty liver disease,” PLoS One,vol. 11, Article ID e0157358, 2016.

[13] V. W.-S. Wong, M. Irles, G. L.-H. Wong et al., “Unifiedinterpretation of liver stiffness measurement by M and XLprobes in non-alcoholic fatty liver disease,” Gut, vol. 68, 2019.

[14] S. Petta, V. W.-S. Wong, C. Camma et al., “Improved non-invasive prediction of liver fibrosis by liver stiffness mea-surement in patients with nonalcoholic fatty liver diseaseaccounting for controlled attenuation parameter values,”Hepatology, vol. 65, no. 4, pp. 1145–1155, 2017.

[15] G. J. Ooi, A. Earnest, W. W. Kemp et al., “Evaluating feasi-bility and accuracy of non-invasive tests for nonalcoholic fattyliver disease in severe and morbid obesity,” InternationalJournal of Obesity, vol. 42, no. 11, pp. 1900–1911, 2018.

[16] C. Moreno, S. Mueller, and G. Szabo, “Non-invasive diagnosisand biomarkers in alcohol-related liver disease,” Journal ofHepatology, vol. 70, no. 2, pp. 273–283, 2019.

[17] U. Agbim and S. K. Asrani, “Non-invasive assessment of liverfibrosis and prognosis: an update on serum and elastographymarkers,” Expert Rev Gastroenterol Hepatol, vol. 13,pp. 361–374, 2019.

[18] R. Vuppalanchi, M. S. Siddiqui, M. L. Van Natta et al.,“Performance characteristics of vibration-controlled transientelastography for evaluation of nonalcoholic fatty liver dis-ease,” Hepatology, vol. 67, no. 1, pp. 134–144, 2018.

[19] J. Trowell, L. Liu, Y. Savva, A. Maheshwari, H. Y. Yoo, andP. J. (uluvath, “Staging of steatosis: a study comparingcontrolled attenuation parameter (CAP) scores with liverhistology in patients with NASH and other chronic liverdiseases,” Hepatology, vol. 68, p. 988A, 2018.

[20] Y. M. Hong, K. T. Yoon, M. Cho et al., “Assessment of thehepatic steatosis using controlled attenuation parameter (CAP)measured with transient elastography in healthy population,”Journal of Hepatology, vol. 66, no. 1, p. S597, 2017.

[21] M. Sasso, M. Beaugrand, V. De Ledinghen et al., “Controlledattenuation parameter (CAP): a novel vcte guided ultrasonicattenuation measurement for the evaluation of hepatic stea-tosis: preliminary study and validation in a cohort of patientswith chronic liver disease from various causes,” Ultrasound inMedicine & Biology, vol. 36, no. 11, pp. 1825–1835, 2010.

[22] V. D. Ledinghen, G. L.-H. Wong, J. Vergniol et al., “Con-trolled attenuation parameter for the diagnosis of steatosis innon-alcoholic fatty liver disease,” Journal of Gastroenterologyand Hepatology, vol. 31, no. 4, pp. 848–855, 2016.

[23] M. S. Siddiqui, R. Vuppalanchi, M. L. Van Natta et al., “Vi-bration-controlled transient elastography to assess fibrosis andsteatosis in patients with nonalcoholic fatty liver disease,”ClinicalGastroenterology and Hepatology, vol. 17, no. 1, p. 156, 2019.

[24] J. H. Runge, L. P. Smits, J. Verheij et al., “MR spectroscopy-derived proton density fat fraction is superior to controlledattenuation parameter for detecting and grading hepaticsteatosis,” Radiology, vol. 286, no. 2, pp. 547–556, 2018.

[25] W.-K. Chan, N. R. Nik Mustapha, G. L.-H. Wong,V. W.-S. Wong, and S. Mahadeva, “Controlled attenuationparameter using the FibroScan XL probe for quantification ofhepatic steatosis for non-alcoholic fatty liver disease in anAsian population,” United European Gastroenterology Jour-nal, vol. 5, no. 1, pp. 76–85, 2017.

[26] C. S. Voican, A. Louvet, J.-B. Trabut et al., “Transient elas-tography alone and in combination with FibroTest for thediagnosis of hepatic fibrosis in alcoholic liver disease,” LiverInternational, vol. 37, no. 11, pp. 1697–1705, 2017.

[27] S. Mahadeva, A. S. Mahfudz, A. Vijayanathan, K. L. Goh,A. Kulenthran, and P. L. Cheah, “Performance of transientelastography (TE) and factors associated with discordance innon-alcoholic fatty liver disease,” Journal of Digestive Diseases,vol. 14, no. 11, pp. 604–610, 2013.


http://downloads.hindawi.com/journals/cjgh/2021/8859338.f1.docx

http://downloads.hindawi.com/journals/cjgh/2021/8859338.f1.docx

TransientElastographyinAlcoholicLiverDiseaseand … · 2021. 1. 20. · ResearchArticle...

Documents

Transcript of TransientElastographyinAlcoholicLiverDiseaseand … · 2021. 1. 20. · ResearchArticle...