Comparison of Diagnostic Accuracy of Aspartate

Aminotransferase to Platelet Ratio Index and

Fibrosis-4 Index for Detecting Liver Fibrosis in
Adult Patients With Chronic Hepatitis B Virus
Infection: A Systemic Review and Meta-analysis
Guangqin Xiao,1 Jiayin Yang,2 and Lunan Yan1

The aspartate aminotransferase-to-platelet ratio index (APRI) and fibrosis index based on
the four factors (Fibrosis 4 index; FIB-4) are the two most widely studied noninvasive tools
for assessing liver fibrosis. Our aims were to systematically review the performance of APRI
and FIB-4 in hepatitis B virus (HBV) infection in adult patients and compare their advan-
tages and disadvantages. We examined the diagnostic accuracy of APRI and FIB-4 for signif-
icant fibrosis, advanced fibrosis, and cirrhosis based on their sensitivity, specificity, positive
predictive value, negative predictive value, and area under the receiver operating characteris-
tic curve (AUROC). Heterogeneity was explored using metaregression. Our systemic review
and meta-analysis included 16 articles of APRI only, 21 articles of APRI and FIB-4 and two
articles of FIB-4 for detecting different levels of liver fibrosis. With an APRI threshold of
0.5, 1.0, and 1.5, the sensitivity and specificity values were 70.0% and 60.0%, 50.0% and
83.0%, and 36.9% and 92.5% for significant fibrosis, advanced fibrosis, and cirrhosis,
respectively. With an FIB-4 threshold of 1.45 and 3.25, the sensitivity and specificity values
were 65.4% and 73.6% and 16.2% and 95.2% for significant fibrosis. The summary
AUROC values using APRI and FIB-4 for the diagnosis of significant fibrosis, advanced
fibrosis, and cirrhosis were 0.7407 (95% confidence interval [CI]: 0.7033-0.7781) and
0.7844 (95% CI: 0.7450-0.8238; (Z 5 1.59, P 5 0.06), 0.7347 (95% CI: 0.6790-0.7904)
and 0.8165 (95% CI: 0.7707-0.8623; Z 5 2.01, P 5 0.02), and 0.7268 (95% CI: 0.6578-
0.7958) and 0.8448 (95% CI: 0.7742-0.9154; (Z 5 2.34, P 5 0.01), respectively. Conclu-
sions: Our meta-analysis suggests that APRI and FIB-4 can identify hepatitis B-related fibro-
sis with a moderate sensitivity and accuracy. (HEPATOLOGY 2015;61:292-302)

C
hronic hepatitis B (CHB) virus (HBV) infection
is a major public health problem. Patients with
CHB infection have a high risk of progressive
liver fibrosis, which can lead to cirrhosis and further
complications, including end-stage liver disease and
hepatocellular carcinoma (HCC).1,2 Therefore, the early
diagnosis of liver fibrosis or cirrhosis plays an important
role not only in the control of disease progression, but
also in therapeutic judgment for HBV infection.
Liver biopsy remains the gold standard for identifying
histological stages because it allows the direct measure-
ment of liver fibrosis. However, this procedure has limita-
tions, including invasiveness, sampling error, and
complications,3 and as an invasive method, it does not

Abbreviations: ALT, alanine aminotransferase; APRI, aspartate aminotransferase to platelet ratio index; AST, aspartate aminotransferase; AUROC, area under
the receiver operating characteristic curve; AUSROC, area under the summary receiver operating characteristic curve; CHB, chronic hepatitis B virus; CI, confidence
interval; DOR, diagnostic odds ratio; FIB-4, Fibrosis 4 index; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HIV, human
immunodeficiency virus; NPV, negative predictive value; PPV, positive predictive value; QUADAS, the Quality Assessment of Diagnostic Accuracy Studies; SD,
standard deviation; SROC, summary receiver operating characteristic.
From the 1Department of Liver Surgery and State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, China; and 2Organ
Transplantation Center, West China Hospital of Sichuan University, Chengdu, China.
Received February 4, 2014; accepted August 14, 2014.
This work was supported by a grant from the National Science and Technology Major Project of China (2012ZX10002-016) and (2012ZX10002017-017).
Additional Supporting Information may be found at onlinelibrary.wiley.com/doi/10.1002/hep.27382/suppinfo.

292
HEPATOLOGY, Vol. 61, No. 1, 2015
allow the dynamic observation of liver fibrosis. All of these
reasons prompted us to explore alternative, noninvasive
methods. Many investigators have attempted to propose
noninvasive methods to assess liver fibrosis. The perfect
noninvasive method should be simple, readily available,
inexpensive, reliable, and accurate in detecting fibrosis.
Currently, there are multiple noninvasive methods based
on inexpensive laboratory tests for predicting liver fibrosis,
including the aspartate aminotransferase (AST) to platelet
index (APRI)4 and the fibrosis index based on the four
factors (Fibrosis-4 index; FIB-4).5 Since these two models
were initially reported, they have been widely investigated
with regard to their diagnostic accuracy for detecting liver
fibrosis resulting from different causes. Recently, numer-
ous studies have attempted to externally validate these
two models in detecting fibrosis in patients with CHB
infection, but the results have been controversial.6-10 Sha-
heen et al.11 and Lin et al.12 have performed meta-
analyses of APRI for detecting liver fibrosis caused by
hepatitis C virus (HCV). The results regarding which
index performs better for the diagnosis of liver fibrosis in
HBV-related patients are inconsistent.
In light of the uncertain clinical utility of APRI and
FIB-4 in patients with CHB infection, we systemati-
cally reviewed the diagnostic accuracy of APRI and
FIB-4 for the prediction of HBV-related fibrosis. Ulti-
mately, we provide a summary of the existing literature
and explore the heterogeneity of the results from dif-
ferent studies using metaregression analysis.
Patients and Methods
Search Strategy. An electronic search was per-
formed on PubMed/Medline, EMBASE, the Cochrane
Library, and the Web of Knowledge (01/2005-12/
2013) including the following search terms: APRI,
AST-to-platelet ratio index, AST, platelet, FIB-4, age,
hepatitis B, and noninvasive models and fibrosis. The
language was limited to English. Studies were included
if they met the following inclusion criteria: (1) the
study evaluated the performance of APRI or FIB-4 for
detecting fibrosis in HBV-infected patients; (2) liver
biopsy was used as the reference standard for assessing
Address reprint requests to: Lunan Yan, Ph.D., Department of Liver Surgery & State Key Laboratory of Biotherapy, West China Hospital of Sichuan University,
No.37, Guoxue Alley, Chengdu City 610041, Sichuan Province, China. E-mail: yanlunan1268@163.com; fax: 186-28-85422867 or Jiayin Yang, Ph.D., Organ
Transplantation Center, West China Hospital of Sichuan University, No.37, Guoxue Alley, Chengdu City 610041, Sichuan Province, China. E-mail: yangjiayin-
doctor@163.com; fax: 186-28-85422867.
Copyright VC 2014 by the American Association for the Study of Liver Diseases.
View this article online at wileyonlinelibrary.com.
DOI 10.1002/hep.27382
Potential conflict of interest: Nothing to report.
XIAO ET AL. 293
fibrosis; (3) data could be extracted to allow the con-
struction of at least one 2 3 2 table of test perform-
ance; and (4) the study included more than
30 patients. Studies including human immunodefi-
ciency virus (HIV)-coinfected patients were excluded.
Data Extraction. Two reviewers independently
evaluated the study eligibility and quality and extracted
the outcome data. Discrepancies were resolved by con-
sensus. Fifteen parameters were collected. The Quality
Assessment of Diagnostic Accuracy Studies (QUA-
DAS) score was used to assess the quality of the stud-
ies included in this review. The outcomes were the
identification of significant fibrosis, advanced fibrosis,
and cirrhosis, defined, respectively, as Metavir, Batts
and Ludwig, or Scheuer stages F2-F4, F3-F4, and F4
or as Ishak stages F3-F6, F4-F6, and F5-F6.
Data Synthesis and Analysis. Stata 12.0, Meta-
Disc software 1.4, Review Manager 5.2, and MedCalc
12.7.0 were used to analyze the reports and tests for
sensitivity, specificity, and area under the summary
receiver operating characteristic curve (AUSROC), in
addition to the meta-regression analysis.
Assessment of Diagnostic Test Accuracy. The data
were extracted and 2 3 2 tables were constructed to cal-
culate sensitivity, specificity, positive predictive value
(PPV), and negative predictive value (NPV) for each
reported test threshold. To provide clinically meaningful
results, three measures of diagnostic test accuracy were
used to examine the accuracy of APRI and FIB-4 for the
diagnosis of significant fibrosis, advanced fibrosis, and
cirrhosis, including the AUSROC, the summary diag-
nostic odds ratios (DORs), and the summary sensitiv-
ities and specificities. The summary receiver operating
characteristic (SROC) curves of APRI and FIB-4 in dif-
ferent studies were simultaneously constructed to com-
pare the SROC values of these two models for detecting
different stages of fibrosis. The Z test was used to com-
pare the AUSROC values of APRI and FIB-4 for pre-
dicting liver fibrosis. Because the SROC curve, which is
generated using linear regression, only represents the
relationship between the sensitivity and specificity values
across studies with different cut-off values of APRI and
FIB-4, we further calculated the summary DORs using
294 XIAO ET AL.
a Der-Simonian and Laird random-effects model with a
corresponding test of heterogeneity. The summary sensi-
tivities and specificities were also calculated employing
the bivariate meta-analytic approach.
Assessment of Heterogeneity and Publication Bias.
A meta-regression technique was used to explore the fac-
tors that may induce heterogeneity in APRI and FIB-4
accuracy between studies using the Q-I2 statistic, which
describes the variability (as a percentage) in estimates aris-
ing from study heterogeneity, rather than sampling error.
Ten factors were selected for the meta-regression analysis.
A value >50% may be considered to represent substantial
heterogeneity. A linear regression test of funnel plot
asymmetry using a Deeks plot was conducted to examine
the possible publication bias of the studies regarding the
performance of APRI and FIB-4 in detecting fibrosis.
Results
Search Results. A total of 319 studies were
retrieved based on the search strategies described.
Sixty-eight full-text articles were eligible for evaluation
and 29 studies were excluded for the similar studies
from one center (n 5 8), failure to use biopsy (n 5 5),
HEPATOLOGY, January 2015
Fig. 1. Flow diagram of study identi-
fication. Abbreviation: SCI, Science
Citation Index.
data incomplete (n 5 8), or small sample size (n 5 8).
Finally, 39 studies were selected for evaluation and
meta-analysis.6-10,13-46 All studies included in this
review were from Europe and Asia (Fig. 1).
Characteristics of the Included Studies. The
main features of the studies included in this review are
shown in Table 1. In total, 9,377 patients (median age:
39.7 years; 71.7% male) were included. The overall
prevalence of significant fibrosis, advanced fibrosis, and
cirrhosis was 56.7% (26.9%-78.9%), 33.6% (9.0%-
77.1%), and 18.7% (3.4%-60.5%). Twenty-six studies
used the Metavir score, six used the Ishak score, five
used the Scheuer score, and one used the Batts and Lud-
wig score. Thirty-three studies included HBV-
monoinfected patients (n 5 8,789)6-10,14-19,22-46; three
studies included HBV-infected patients with other fac-
tors, including alcohol consumption (n 5 512)13,20,34;
and one study included patients with HCC (n 5 76).21
According to the QUADAS score, the methodological
quality of the included studies was satisfactory (Table 1).
Diagnostic Accuracy of APRI and FIB-4 for the
Prediction of Significant Fibrosis APRI. Thirty-
four studies (n 5 8,855) assessed the performance of
 Table 1. Characteristics of the Included Studies
Interval Between Median/Mean Liver Biopsy Liver Significant Fibrosis,
Biopsy and Age, Years Scoring Biopsy Advanced Fibrosis, QUADAS
Author, Year, Region Models Study/Center Description n Blood Test (Male %) Etiology System Blind Length (mm) Cirrhosis (%) Score

Chrysanthos, 2006, Greece13 APRI Retrospective, one center 205 Unclear 51 (74.6) HBV, alcohol Ishak Yes >15 60.5, *, 26.8 12
Wai, 2006, Singapore14 APRI Retrospective, one center 218 Twenty days 35 (82.6) HBV Ishak Yes 16 47.3, *, 18.8 14
Coco, 2007, Italy7 APRI Retrospective, one center 228 Unclear 50.4 (71.5) HBV: 79; Metavir Yes 25 (12-54) 61.9, *, 50.4 13
HCV: 149
Hongbo, 2007, China15 APRI Retrospective, one center 444 Within 1 week 30 (73.0) HBV Metavir Yes >10 30.2, 9.0, 4.5 14
Chang, 2008, Singapore26 APRI Prospective, one center 120 Unclear 49.5 (57.5) HBV: 58; Metavir Yes 15 (2-40) 44.2, 26.7, 10 12
HCV: 62
HEPATOLOGY, Vol. 61, No. 1, 2015

Lin, 2008, Taiwan17 APRI Retrospective, one center 48 Unclear 56.2 (83.3) HBV Metavir Unclear Unclear *, 77.1, 37.5 12
Shin, 2008, Korea9 APRI Retrospective, one center 264 Unclear 27.6 (87.1) HBV Metavir Yes 17 53.4, 29.5, 3.4 14
Zhang, 2008, China18 APRI Retrospective, one center 137 Unclear 35.2 (70.5) HBV Metavir Unclear Unclear 58.4, 24.8, 8.8 12
Mallet, 2009, France19 FIB-4 Retrospective, two centers 138 Unclear 42 (71.0) HBV Metavir Unclear 17.6 6 6.8 49.3, 29.1, * 13
Bonnard, 2010, France20 APRI &FIB-4 Prospective, one center 57 Same day 35 (68.4) HBV, HCV, others Metavir Unclear Unclear 70, 38, 24 12
Hung, 2010, Taiwan21 APRI Retrospective, one center 76 Within 1 week 57 (84.2) HBV-HCC Ishak Yes Unclear 75.0, 67.1, 60.5 13
Kim, 2010, Korea8 APRI &FIB-4 Retrospective, one center 668 Same day 39.2 (66.2) HBV Metavir Yes >15 78.9, 49.4, 34.4 14
Wong, 2010, Hongkong22 APRI &FIB-4 Prospective, one center 156 Same day 45 (76.0) HBV Metavir Yes >15 75, 48, 26 14
Zhou, 2010, China23 APRI Retrospective, multicenter 146 Same day 33 (84.9) HBV Scheuer Yes >15 66.4, 29.5, 7.6 14
Castera, 2011, France24 APRI Retrospective, one center 60 Same day 39 (62.0) HBV Metavir Yes 21.6 6 8.0 73.3, 46.7, 25 13
Cheong, 2011, Korea25 APRI &FIB-4 Prospective, six centers 420 Unclear 41.6 (67.8) HBV Metavir Yes 12.6 6 3.3 78.2, 46.8, 17.3 13
Lesmana, 2011, Indonesia26 APRI Retrospective, one center 117 Unclear 40.6 (53.8) HBV Metavir Yes >15 62.4, 23.9, 3.4 13
Liu, 2011, China27 APRI &FIB-4 Retrospective, three centers 623 Unclear 32.4 (55.5) HBV Metavir Yes >10 34.5, 18.1, 5.6 14
Sebastiani, 2011, France28 APRI Retrospective, multicenter 253 Unclear 43.7 (72.7) HBV Metavir Yes Unclear 57.7, 28.5, 17.8 14
Seto, 2011, Hongkong29 APRI &FIB-4 Retrospective, one center 237 Same day 36.4 (67.5) HBV Ishak Yes >15 32.4, 18.9, 8.4 14
Zhu, 2011, China30 FIB-4 Retrospective, two centers 159 Unclear 42 (71.1) HBV Unclear Unclear >15 65.4, 39.6, 19.5 12
Liu, 2012, China31 APRI &FIB-4 Retrospective, two centers 114 Unclear 38.3 (79.8) HBV Metavir Yes 15-20 50.9, 26.3, 10.5 14
Raftopoulos, 2012, Australia32 APRI Prospective, four centers 179 Within 1 week 41.9 (70.9) HBV Metavir Yes 21 6 9.7 42, 22.4, 8.4 14
Wu, 2012, China33 APRI &FIB-4 Retrospective, three centers 482 Unclear 33.1 (80.9) HBV Metavir Yes Unclear 56.0, 35.5, 19.3 13
Yoon, 2012, Korea34 APRI Prospective, two centers 250 Unclear 46.6 (59.0) HBV: 179; Batts and Yes >10 67, 39, 14 13
HCV: 19; Ludwig
others: 52
Zhu, 2012, China35 APRI &FIB-4 Prospective, one center 175 Same day 36.5 (78.3) HBV Metavir Yes >15 45.4, *, 16.6 14
Basar, 2013, Turkey36 APRI &FIB-4 Prospective, one center 76 Same day 45.2 (44.7) HBV Metavir Yes >10 67, 38.2, 17.1 14
Chen, 2013, China6 APRI &FIB-4 Retrospective, one center 458 Within 1 week 42.1 (61.1) HBV Metavir Yes >15 43.7, 25.8, 12.5 14
de Ledinghen,2013, France37 APRI &FIB-4 Prospective, one center 209 Unclear 42.5 (64.2) HBV Metavir Yes Unclear 65.4, 40.2, 23.4 12
Gumusay, 2013, Turkey38 APRI &FIB-4 Prospective, one center 58 Unclear 41 (56.9) HBV Ishak Unclear >20 *, 17.2, * 12
Ma, 2013, China39 APRI &FIB-4 Retrospective, one center 1168 Unclear 33.3 (84.3) HBV Metavir Unclear Unclear 66.0, 45.6, 17.9 13
Shrivastava, 2013, India40 APRI &FIB-4 Retrospective, one center 52 Unclear 30 (94.2) HBV Ishak Unclear 15-20 26.9, 11.5, * 12
Tseng, 2013, Taiwan41 APRI Retrospective, one center 119 Two weeks 39 (81.0) HBV Metavir Yes 18 6 8 *, 43.5, 32.5 14
Ucar, 2013, Turkey42 APRI &FIB-4 Retrospective, one center 73 Same day 44.9 (64.4) HBV Metavir Yes Unclear 56.0, 24.5, 10.9 13
Wang, 2013, China43 APRI &FIB-4 Retrospective, one center 231 Unclear 34 (67.5) HBV Scheuer Yes >15 29.4, 13.4, 6.9 14
Wang, 2013, China44 APRI &FIB-4 Retrospective, five centers 349 Unclear 37 (87.7) HBV Scheuer Yes >10 58.5, 26.4, 7.16 14
Xu, 2013, China10 APRI &FIB-4 Retrospective, one center 126 Unclear 36 (72.2) HBV Scheuer Yes >10 59.3, 36.1, 17.4 13
Xun, 2013, China45 APRI &FIB-4 Prospective, one center 197 Unclear 31 (75.6) HBV Scheuer Unclear >15 56.9, 33.5, * 13
XIAO ET AL.

Zeng, 2013, China46 APRI &FIB-4 Retrospective, one center 287 Unclear 35.2 (70.5) HBV Metavir Unclear 15-20 74.8, 45.5, 19.2 12
295

*No data in the studies.

296 XIAO ET AL.
Fig. 2. (A) SROC curve of APRI and FIB-4 for detecting significant
fibrosis. (B) SROC curve of APRI and FIB-4 for detecting advanced
fibrosis. (C) SROC curve of the APRI and FIB-4 for detecting cirrhosis.
HEPATOLOGY, January 2015
APRI for detecting significant fibrosis. In these studies,
the prevalence of significant fibrosis ranged from 26.9%
to 78.9% (mean, 56.6%). The mean AUSROC value of
APRI for detecting significant fibrosis was 0.72 (range,
0.61-0.88; Table 4). When the results were combined,
the AUSROC value of the SROC was 0.7407 (standard
deviation [SD]: 0.0191; Fig. 2A). The summary DOR
was 4.55 (95% confidence interval [CI]: 3.53-5.86; Fig.
3A1). Significant heterogeneity was observed in the
analysis of the significant fibrosis stage (Q 5 89.48,
I2 5 74.3%, P < 0.01). Table 2 shows the summary sen-
sitivities and specificities of APRI at various cut-off val-
ues for distinguishing significant fibrosis. At the cut-off
values of 0.5 and 1.5, the summary sensitivities and spe-
cificities were 70.0% (range, 35.0%-97.0%) and 60.0%
(range, 34.0%-86.7%) and 34.1% (range, 14.0%-
75.0%) and 89.5% (range, 81.6%-100.0%), respec-
tively. When we assumed the prevalence rate of
significant fibrosis in HBV patients to be 56.6%, and on
the basis of the data described above, the estimated PPV
and NPV at a cut-off value of 0.5 were 61.7% and
52.9%, respectively. The estimated PPV and NPV at a
cut-off value of 1.5 were 83.2% and 52.0%, respectively.
FIB-4. Twenty-two studies (n 5 6,455) assessed the
performance of FIB-4 for predicting significant fibrosis.
The prevalence of significant fibrosis ranged from
26.9% to 78.9% (mean, 56.4%). The mean AUSROC
value of FIB-4 for detecting significant fibrosis was 0.76
(range, 0.69-0.87; Table 4). Based on the combined
results, the AUSROC value of SROC was 0.7844 (SD,
0.0201; Fig. 2A). The summary DOR was 6.23 (95%
CI: 4.57-8.50; Fig. 3B1). Significant heterogeneity was
observed in the analysis of the significant fibrosis stage
(Q 5 46.93, I2 5 70.2%, P < 0.01). The summary sen-
sitivities and specificities of FIB-4 for distinguishing sig-
nificant fibrosis at different cut-off values are displayed
in Table 3. At cut-off values of 1.45 and 3.25, the sum-
mary sensitivities and specificities were 65.4% (range,
51.9%-87.0%) and 73.6% (range, 65.0%-85.0%) and
16.2% (range, 7.0%-25.1%) and 95.2% (range, 95.0%-
99.4%), respectively. The PPV values were 76.2% and
62.2% at cut-off values of 1.45 and 3.25 when a preva-
lence rate of 56.4% for significant fibrosis in HBV
patients was assumed and given the aforementioned
results. The NPV values were 81.4% and 46.8% at cut-
off values of 1.45 and 3.25.
Diagnostic Accuracy of APRI and FIB-4 for
Prediction of Advanced Fibrosis APRI. Thirty-three
studies in 8,254 patients reported the prevalence of
advanced fibrosis. The mean prevalence of advanced
fibrosis was 33.5% (range, 9.0%-77.1%). The average
AUSROC value of APRI for detecting advanced fibrosis
HEPATOLOGY, Vol. 61, No. 1, 2015 XIAO ET AL. 297

Fig. 3. (A1) Diagnostic accuracy of APRI for significant fibrosis; (A2) diagnostic accuracy of APRI for advanced fibrosis; and (A3) diagnostic
accuracy of APRI for cirrhosis. (B1) Diagnostic accuracy of FIB-4 for significant fibrosis; (B2) diagnostic accuracy of FIB-4 for advanced fibrosis;
and (B3) diagnostic accuracy of FIB-4 for cirrhosis.
298 XIAO ET AL. HEPATOLOGY, January 2015

Table 2. Summary Sensitivities and Specificities of APRI at played in Table 3. The PPV values were 40.2% and
Various Diagnostic Thresholds for Prediction of Significant 80.1% at cut-off values of 1.45 and 3.25 when a prev-
Fibrosis, Advanced Fibrosis, and Cirrhosis alence rate of 32.2% for advanced fibrosis in HBV
Number of Summary patients was assumed and under given the aforemen-
Studies Sensitivity Summary Specificity
(Patients) (Mean, Range) (%) (Mean, Range) (%)
tioned results. The NPV values were 75.9% and
71.3% at cut-off values of 1.45 and 3.25.
Significant fibrosis
0.235-0.425 4 (1,260) 67.2 (62.7-72.3) 71.2 (68.0-74.6)
Diagnostic Accuracy of APRI and FIB-4 for
0.5 10 (2,535) 70.0 (35.0-97.0) 60.0 (34.0-86.7) Prediction of Cirrhosis APRI. Thirty-four studies
0.535-0.85 5 (906) 63.5 (41.4-73.2) 74.1 (59.4-87.5) (n 5 8,773) reported on the prevalence of cirrhosis. In
1.0 2 (109) 40.0 (25.0-55.0) 67.0 (50.0-84.0)
1.5 12 (2,754) 34.1 (14.0-75.0) 89.5 (81.6-100.0)
these studies, the prevalence of cirrhosis ranged from
23.79 1 (126) 77.1 84.3 3.4% to 60.5% (mean, 18.4%). The mean AUSROC
Advanced fibrosis value of APRI for detecting cirrhosis was 0.72 (range,
0.27-0.47 2 (193)* 72.4 71.6 0.50-0.85; Table 4). When the results were combined,
0.5 2 (283) 73.0 (63.0-83.0) 54.5 (37.0-72.0)
0.525-0.8 2 (124) 55.6 (43.2-67.9) 74.2 (66.6-81.8) the AUSROC value of SROC was 0.728 (SD, 0.0352;
1.0 1 (52) 50.0 83.0 Fig. 2C). The summary DOR was 4.41 (95% CI:
1.5 1 (52) 33.0 91.0 2.86-6.80; Fig. 3A3). Significant heterogeneity was
4.719-49.94 2 (1,294) 92.8 (85.0-100.0) 34.9 (2.06-67.7)
Cirrhosis
observed in the analysis of the cirrhosis stage (Q 5
0.5 1 (146) 82.4 38.5 85.64, I2 5 84.8%, P < 0.01). Table 2 shows the sum-
0.529-0.77 3 (309) 74.8 (72.7-76.9) 72.9 (72.8-73.0) mary sensitivities and specificities of APRI at various
1.0 5 (1,247) 66.1 (47.0-85.0) 73.7 (56.0-83.7)
1.2 2 (105) 52.8 (50.0-55.6) 70.5 (51.0-90.0)
cut-off values for distinguishing cirrhosis. When we
1.5 2 (374) 36.9 (25.2-48.5) 92.5 (85.9-99.1) assumed the prevalence rate of cirrhosis in HBV
2.0 6 (1,445) 31.3 (13.0-63.2) 88.8 (81.0-96.0) patients to be 18.4%, and on the basis of the results
2659.57 1 (287) 65.8 78.1 in Table 2, at a cut-off value of 0.5 the estimated PPV
*Hung et al.21 did not report the corresponding sensitivity and specificity and NPV were 23.2% and 90.7%, respectively. The
values. estimated PPV and NPV at a cut-off value of 1.5 were
52.6% and 86.7%, respectively.
was 0.76 (range, 0.68-0.87; Table 4). After combina- FIB-4. Nineteen studies in 6,068 patients reported
tion, the AUSROC value of SROC was 0.7347 (SD, on the prevalence of cirrhosis. The prevalence of cir-
0.0284; Fig. 2B). The summary DOR was 3.70 (95% rhosis ranged from 5.6% to 34.4% (mean, 16.5%).
CI: 2.27-6.03; Fig. 3A2). Significant heterogeneity was The mean AUSROC value of FIB-4 for detecting cir-
not observed in the analysis of the advanced fibrosis rhosis was 0.78 (range, 0.71-0.93; Table 4). After the
stage (Q 5 13.33, I2 5 45.5%, P 5 0.06). The summary results were combined, the AUSROC value of SROC
sensitivities and specificities of APRI at various cut-off was 0.8448 (SD, 0.0360; Fig. 2C). The summary
values for distinguishing advanced fibrosis are shown in DOR was 12.14 (95% CI: 6.87-21.45; Fig. 3B3). Sig-
Table 2. When the prevalence rate of advanced fibrosis nificant heterogeneity was observed in the analysis of
in HBV patients was 33.5%, based on the results in the cirrhosis stage (Q 5 25.52, I2 5 80.4%, P < 0.01).
Table 2, the estimated PPV and NPV were 44.7% and The summary sensitivities and specificities of FIB-4 at
80.0% at a cut-off value of 0.5. The estimated PPV and different cut-off values for distinguishing significant
NPV at a cut-off value of 1.5 were 64.9% and 72.9%. fibrosis are displayed in Table 3.
FIB-4. Twenty-two studies (n 5 6,338) reported Methodological Heterogeneity and Publication
on the prevalence of advanced fibrosis. The prevalence Bias. Because meta-regression analyses are limited by
of significant fibrosis ranged from 11.5% to 49.4% the number of studies, we examined methodological
(mean, 32.2%). The mean AUSROC value of FIB-4 heterogeneity only in groups of more than 10 studies.
for detecting significant fibrosis was 0.80 (range, 0.74- The accuracy of APRI for detecting significant fibrosis
0.91; Table 4). After the results were combined, the and advanced fibrosis was not affected by study- or
AUSROC value of SROC was 0.8165 (SD, 0.0337; patient-related factors. The following factors were not
Fig. 2B). The summary DOR was 8.50 (95% CI: significant in this analysis for detecting significant fibro-
4.96-14.57; Fig. 3B2). Significant heterogeneity was sis and advanced fibrosis: (1) location of the study
observed in the analysis of the significant fibrosis stage (P 5 0.61 and 0.78); (2) study design (P 5 0.78 and
(Q 5 32.07, I2 5 78.2%, P < 0.01). The summary 0.11); (3) sample size (P 5 0.80 and 0.34); (4) median
sensitivities and specificities of FIB-4 at different cut- age (P 5 0.84 and 0.47); (5) percentage of males
off values for distinguishing advanced fibrosis are dis- (P 5 0.32 and 0.24); (6) histopathological scoring
HEPATOLOGY, Vol. 61, No. 1, 2015 XIAO ET AL. 299

Table 3. Summary Sensitivities and Specificities for FIB-4 at The mean AUSROC value of FIB-4 was higher than
Various Diagnostic Thresholds for Prediction of Significant that of APRI (0.76 vs. 0.72) for predicting significant
Fibrosis, Advanced Fibrosis, and Cirrhosis fibrosis in this meta-analysis. When combined, the FIB-
Number of Summary Summary 4 had a SROC AUC of 0.7844 (95% CI: 0.7450-
Studies Sensitivity Specificity
(Patients) (Mean, Range) (%) (Mean, Range) (%)
0.8238), whereas APRI had a value of 0.7407 (95% CI:
0.7033-0.7781; Z 5 1.59, P 5 0.06). The summary
Significant fibrosis
0.8-1.1 5 (1,026) 72.5 (70.7-74.0) 67.3 (60.0-76.0)
DOR of FIB-4 (DOR 5 6.23) was higher than the sum-
1.45 6 (1,809) 65.4 (51.9-87.0) 73.6 (65.0-85.0) mary DOR of APRI (DOR 5 4.55). According to our
1.57-2.5 3 (648) 45.2 (14.0-51.7) 83.2 (69.0-95.0) analysis, at a cut-off value of 0.5, APRI had 70.0% sen-
3.25 5 (1,578) 16.2 (7.0-25.1) 95.2 (95.0-99.4)
4.9-223.7 3 (343) 77.8 (60.0-87.6) 82.5 (81.9-83.3)
sitivity and 60.0% specificity; at a threshold value of
Advanced fibrosis 1.5, the sensitivity and specificity were 34.1% and
0.82-1.175 2 (255) 78.6 64.6 89.5%, respectively. The sensitivity of FIB-4 was lower
1.45 3 (421) 62.7 (50.0-71.0) 55.7 (14.0-80.0)
than the sensitivity of APRI, with values of 65.4% and
1.6-2.711 3 (1,888) 73.7 (33.0-100.0) 65.6 (16.7-96.0)
3.25 1 (52) 17.0 98.0 16.2% at low and high thresholds (1.45 and 3.25).
3.6-808.77 2 (794) 56.8 (30.0-83.6) 81.5 (64.5-98.4) However, FIB-4 had higher specificity, with values of
Cirrhosis 73.6% and 95.2%, at cut-off values of 1.45 and 3.25.
0.84-1.05 4 (1,191) 87.4 (79.0-92.0) 64.7 (51.3-72.8)
1.625-2.65 4 (1,033) 64.3 (38.5-76.9) 85.5 (75.3-97.9)
Assuming a prevalence of 56% for significant fibrosis (as
observed in the included studies), we obtained estimated
PPV values of 61.7% and 76.2% for APRI and FIB-4,
system (P 5 0.63 and 0.24); (7) liver biopsy length
respectively, at low cut-off values and 83.2% and 62.2%
(P 5 0.99 and 0.06); (8) blinded interpretation
for APRI and FIB-4, respectively, at high cut-off values.
(P 5 0.82 and 0.12); (9) prevalence of significant fibro-
The estimated NPV values of APRI and FIB-4 were
sis (P 5 0.45 and 0.19), advanced fibrosis (P 5 0.38 and
52.9% and 81.4% at the low thresholds and 52.0% and
0.11), and cirrhosis (P 5 0.15 and 0.10); and (10)
46.8% at the high thresholds, respectively. Based on
QUADAS score (P 5 0.09 and 0.10). The accuracy of
these analyses, the exclusion strength of FIB-4 for signif-
FIB-4 for detecting significant fibrosis was also not
icant fibrosis is superior to that of APRI at low thresh-
affected by the following factors: country of origin
olds, but at high cut-off values, the exclusion strength of
(P 5 0.40), study design (P 5 0.09); sample size
APRI is higher.
(P 5 0.15); median age (P 5 0.61); and percentage of
The summarized results showed that the SROC values
males (P 5 0.69). FIB-4 accuracy was not affected by
of FIB-4 were higher than that of APRI for advanced
the histopathological classification (P 5 0.13), blinded
fibrosis and cirrhosis, and the results had statistical signifi-
interpretation (P 5 0.60), adequacy of biopsy specimens
cance (advanced fibrosis: Z 5 2.01, P 5 0.02; cirrhosis:
(P 5 0.39), prevalence of significant fibrosis (P 5 0.10),
Z 5 2.34, P 5 0.01). The AUSROC values of APRI for
advanced fibrosis (P 5 0.17), cirrhosis (P 5 0.08), and
predicting significant fibrosis and cirrhosis in the training
methodological quality (P 5 0.07).
and validation cohorts were 0.80/0.89 and 0.88/0.94,
The linear regression test of funnel plot asymmetry
respectively, in the original study of Wei et al.4 Two previ-
demonstrated that there was no publication bias for
ous systematic reviews11,12 regarding the accuracy of
APRI for detecting significant fibrosis, advanced fibrosis,
APRI, mainly in HCV patients, showed that APRI may
and cirrhosis (P 5 0.22, 0.54, and 0.83; Fig 4 A1-A3).
not have great diagnostic worth as initially described in
There was also no evidence of publication bias for FIB-4
HCV or HCV/HIV infection patients. The AUROC
for predicting significant fibrosis, advanced fibrosis, and
cirrhosis (P 5 0.72, 0.39, and 0.68; Fig 4B1-3).
Table 4. AUSROC Values of APRI and FIB-4 for Detecting
Significant Fibrosis, Advanced Fibrosis, and Cirrhosis
Discussion
Significant Advanced
APRI and FIB-4 are the two methods receiving the Fibrosis Fibrosis Cirrhosis

most attention. The diagnostic value of these two APRI

methods is controversial. We summarized the diagnos- Number of studies (patients) 25 (5,565) 10 (3,139) 17 (3,712)
AUSROC value (mean) 0.72 0.76 0.72
tic accuracy of APRI and FIB-4 for predicting AUSROC value (range) 0.61-0.88 0.68-0.87 0.50-0.85
HBV-related fibrosis and weighed their advantages and FIB-4
disadvantages. The summarized results suggest that Number of studies (patients) 14 (3,240) 8 (3,036) 7 (1,726)
APRI and FIB-4 have modest accuracy for detecting AUSROC value (mean) 0.76 0.80 0.78
AUSROC value (range) 0.69-0.87 0.74-0.91 0.71-0.93
significant fibrosis, advanced fibrosis, and cirrhosis.
300 XIAO ET AL. HEPATOLOGY, January 2015

Fig. 4. Linear regression test of funnel plot asymmetry for publication. (A1) APRI detecting significant fibrosis; (A2) APRI detecting advanced
fibrosis; and (A3) APRI detecting cirrhosis. (B1) FIB-4 detecting significant fibrosis; (B2) FIB-4 detecting advanced fibrosis; and (B3) FIB-4
detecting cirrhosis.

values for significant fibrosis and cirrhosis were 0.76-0.77
and 0.82-0.83, respectively. For advanced fibrosis, the
summary AUROC was 0.80. Castera et al.47 reported that
ROC values of FibroScan, Fibrotest, and APRI were 0.83,
0.85, and 0.78 for predicting significant fibrosis, 0.90,
0.90, and 0.84 for detecting advanced fibrosis, and 0.95,
0.87, and 0.83 for diagnosing cirrhosis. However, the vast
majority of the patients in our systemic review and meta-
analysis had HBV infection. The summarized results
showed that APRI had AUROC values of 0.74, 0.73, and
0.73 for detecting significant fibrosis, advanced fibrosis,
and cirrhosis, which suggested that APRI was less accurate
for identification of HBV-related significant fibrosis,
severe fibrosis, and cirrhosis.
HEPATOLOGY, Vol. 61, No. 1, 2015
In the original study of Sterling et al.,5 the AUSROC
of FIB-4 for detecting significant fibrosis and advanced
fibrosis in the training and validation cohorts was
0.711/0.688 and 0.737/0.765, respectively. Surprisingly,
the summarized results showed that the AUROC of
FIB-4 was 0.78, 0.82, and 0.84 for prediction of HBV-
related significant fibrosis, advanced fibrosis, and cirrho-
sis, respectively. FIB-4 tended to present more accurate
results for identification of significant fibrosis and
advanced fibrosis in HBV-monoinfected patients than
in HIV/HCV-coinfected patients. Some studies have
indicated that HIV- or antiretroviral-related thrombocy-
topenia may diminish accuracy,48 which provides a pos-
sible explanation for why the AUROC values were
higher in our systemic review. APRI consists platelet cell
and AST. Besides these two factors, FIB-4 also includes
age and alanine aminotransferase (ALT). Reuber et al.49
displayed that age was associated with cirrhosis. Some
researchers believe that the elevations in AST more than
ALT have been associated with more-advanced fibrosis
and are, in part, related to delayed clearance of AST, rel-
ative to ALT, or to mitochondrial injury associated with
more advanced fibrosis.50
We acknowledge several limitations of our systemic
review and meta-analysis. The first limitation is that
only published manuscripts were identified in this
review. Although the summarized results suggest no
publication- or sample-size-related biases, the exclusion
of potentially eligible abstracts may create the possibility
of publication bias. Second, we focused only on analyz-
ing the utility of APRI and FIB-4 in HBV-infected
patients. It has been reported that the accuracy of APRI
and FIB-4 is lower in HBV-infected patients than in
HCV-infected patients. Though we focus on mono-
HBV infection, there was observed heterogeneity in
diagnostic accuracy of APRI and FIB-4. Our analysis
displayed that the methodological factors included in
our article did not resulted in the heterogeneity, perhaps
the other test characteristics, such as cost-effectiveness,
had an effect on clinical outcomes. Finally, only studies
in English were selected for analysis, and this language
restriction may bias the results to some extent.
In conclusion, this review demonstrates that APRI
and FIB-4 have moderate diagnostic accuracy for pre-
dicting fibrosis in patients with chronic HBV infection.
Although the diagnostic accuracy of FIB-4 and APRI is
not high, they still can be considered as a choice for the
prediction of fibrosis caused by CHB infection in
regions with limited health care resources. These meth-
ods can also be applied simultaneously to mitigate their
respective shortcomings, which may improve their diag-
nostic performance. Novel noninvasive models with the
XIAO ET AL. 301
features of simplicity, low cost, and wide availability
should be proposed in future studies to provide
improved accuracy for detecting fibrosis.
References
1. Kim KH, Shin HJ, Kim K, Choi HM, Rhee SH, Moon HB, et al.
Hepatitis B virus X protein induces hepatic steatosis via transcriptional
activation of SREBP1 and PPARgamma. Gastroenterology 2007;132:
1955-1967.
2. Ganem D, Prince AM. Hepatitis B virus infection—natural history and
clinical consequences. N Engl J Med 2004;350:1118-1129.
3. McGill DB, Rakela J, Zinsmeister AR, Ott BJ. A 21-year experience
with major hemorrhage after percutaneous liver biopsy. Gastroenterol-
ogy 1990;99:1396-1400.
4. Wai CT, Greenson JK, Fontana RJ, Kalbfleisch JD, Marrero JA,
Conjeevaram HS, et al. A simple noninvasive index can predict both
significant fibrosis and cirrhosis in patients with chronic hepatitis C.
HEPATOLOGY 2003;38:518-526.
5. Sterling RK, Lissen E, Clumeck N, Sola R, Correa MC, Montaner J,
et al. Development of a simple noninvasive index to predict significant
fibrosis in patients with HIV/HCV coinfection. HEPATOLOGY 2006;43:
1317-1325.
6. Chen B, Ye B, Zhang J, Ying L, Chen Y. RDW to platelet ratio: a
novel noninvasive index for predicting hepatic fibrosis and cirrhosis in
chronic hepatitis B. PLoS One 2013;8:e68780.
7. Coco B, Oliveri F, Maina AM, Ciccorossi P, Sacco R, Colombatto P,
et al. Transient elastography: a new surrogate marker of liver fibrosis
influenced by major changes of transaminases. J Viral Hepat 2007;14:
360-369.
8. Kim BK, Kim do Y, Park JY, Ahn SH, Chon CY, Kim JK, et al. Vali-
dation of FIB-4 and comparison with other simple noninvasive indices
for predicting liver fibrosis and cirrhosis in hepatitis B virus-infected
patients. Liver Int 2010;30:546-553.
9. Shin WG, Park SH, Jang MK, Hahn TH, Kim JB, Lee MS, et al.
Aspartate aminotransferase to platelet ratio index (APRI) can predict
liver fibrosis in chronic hepatitis B. Dig Liver Dis 2008;40:267-274.
10. Xu MY, Qu Y, Jia XF, Wang ML, Liu H, Wang XP, et al. Serum proteo-
mic MRM identify peptide ions of transferrin as new fibrosis markers in
chronic hepatitis B. Biomed Pharmacother 2013;67:561-567.
11. Shaheen AA, Myers RP. Diagnostic accuracy of the aspartate
aminotransferase-to-platelet ratio index for the prediction of hepatitis
C-related fibrosis: a systematic review. HEPATOLOGY 2007;46:912-921.
12. Lin ZH, Xin YN, Dong QJ, Wang Q, Jiang XJ, Zhan SH, et al. Per-
formance of the aspartate aminotransferase-to-platelet ratio index for
the staging of hepatitis C-related fibrosis: an updated meta-analysis.
HEPATOLOGY 2011;53:726-736.
13. Chrysanthos NV, Papatheodoridis GV, Savvas S, Kafiri G, Petraki K,
Manesis EK, et al. Aspartate aminotransferase to platelet ratio index for
fibrosis evaluation in chronic viral hepatitis. Eur J Gastroenterol Hepa-
tol 2006;18:389-396.
14. Wai CT, Cheng CL, Wee A, Dan YY, Chan E, Chua W, et al. Non-
invasive models for predicting histology in patients with chronic hepati-
tis B. Liver Int 2006;26:666-672.
15. Hongbo L, Xiaohui L, Hong K, Wei W, Yong Z. Assessing routine and
serum markers of liver fibrosis in CHB patients using parallel and serial
interpretation. Clin Biochem 2007;40:562-566.
16. Chang PE, Lui HF, Chau YP, Lim KH, Yap WM, Tan CK, et al. Pro-
spective evaluation of transient elastography for the diagnosis of hepatic
fibrosis in Asians: comparison with liver biopsy and aspartate transami-
nase platelet ratio index. Aliment Pharmacol Ther 2008;28:51-61.
17. Lin CS, Chang CS, Yang SS, Yeh HZ, Lin CW. Retrospective evalua-
tion of serum markers APRI and AST/ALT for assessing liver fibrosis
and cirrhosis in chronic hepatitis B and C patients with hepatocellular
carcinoma. Int Med 2008;47:569-575.
302 XIAO ET AL.
18. Zhang YX, Wu WJ, Zhang YZ, Feng YL, Zhou XX, Pan Q. Noninva-
sive assessment of liver fibrosis with combined serum aminotransferase/
platelet ratio index and hyaluronic acid in patients with chronic hepati-
tis B. World J Gastroenterol 2008;14:7117-7121.
19. Mallet V, Dhalluin-Venier V, Roussin C, Bourliere M, Pettinelli ME,
Giry C, et al. The accuracy of the FIB-4 index for the diagnosis of mild
fibrosis in chronic hepatitis B. Aliment Pharmacol Ther 2009;29:409-
415.
20. Bonnard P, Sombie R, Lescure FX, Bougouma A, Guiard-Schmid JB,
Poynard T, et al. Comparison of elastography, serum marker scores, and
histology for the assessment of liver fibrosis in hepatitis B virus (HBV)-
infected patients in Burkina Faso. Am J Trop Med Hyg 2010;82:454-
458.
21. Hung HH, Su CW, Lai CR, Chau GY, Chan CC, Huang YH, et al.
Fibrosis and AST to platelet ratio index predict post-operative progno-
sis for solitary small hepatitis B-related hepatocellular carcinoma. Hepa-
tol Int 2010;4:691-699.
22. Wong GLH, Wong VWS, Choi PCL, Chan AWH, Chan HLY. Devel-
opment of a non-invasive algorithm with transient elastography (Fibro-
scan) and serum test formula for advanced liver fibrosis in chronic
hepatitis B. Aliment Pharmacol Ther 2010;31:1095-1103.
23. Zhou K, Gao CF, Zhao YP, Liu HL, Zheng RD, Xian JC, et al. Sim-
pler score of routine laboratory tests predicts liver fibrosis in patients
with chronic hepatitis B. J Gastroenterol Hepatol 2010;25:1569-1577.
24. Castera L, Bernard PH, Le Bail B, Foucher J, Trimoulet P, Merrouche
W, et al. Transient elastography and biomarkers for liver fibrosis assess-
ment and follow-up of inactive hepatitis B carriers. Aliment Pharmacol
Ther 2011;33:455-465.
25. Cheong JY, Um SH, Seo YS, Kim DJ, Hwang SG, Lee YJ, et al. Non-
invasive index for predicting significant liver fibrosis: comparison of
diagnostic performances in patients with chronic hepatitis B and C.
Dig Dis Sci 2011;56:555-563.
26. Lesmana CR, Salim S, Hasan I, Sulaiman AS, Gani RA, Pakasi LS,
et al. Diagnostic accuracy of transient elastography (FibroScan) versus
the aspartate transaminase to platelet ratio index in assessing liver fibro-
sis in chronic hepatitis B: the role in primary care setting. J Clin Pathol
2011;64:916-920.
27. Liu HB, Zhou JP, Zhang Y, Lv XH, Wang W. Prediction on liver
fibrosis using different APRI thresholds when patient age is a categori-
cal marker in patients with chronic hepatitis B. Clin Chim Acta 2011;
412:33-37.
28. Sebastiani G, Castera L, Halfon P, Pol S, Mangia A, Di Marco V, et al.
The impact of liver disease aetiology and the stages of hepatic fibrosis on
the performance of non-invasive fibrosis biomarkers: an international
study of 2411 cases. Aliment Pharmacol Ther 2011;34:1202-1216.
29. Seto WK, Lee CF, Lai CL, Ip PP, Fong DY, Fung J, et al. A new
model using routinely available clinical parameters to predict significant
liver fibrosis in chronic hepatitis B. PLoS One 2011;6:e23077.
30. Zhu X, Wang LC, Chen EQ, Chen XB, Chen LY, Liu L, et al. Pro-
spective evaluation of FibroScan for the diagnosis of hepatic fibrosis
compared with liver biopsy/AST platelet ratio index and FIB-4 in
patients with chronic HBV infection. Dig Dis Sci 2011;56:2742-2749.
31. Liu XD, Wu JL, Liang J, Zhang T, Sheng QS. Globulin-platelet model
predicts minimal fibrosis and cirrhosis in chronic hepatitis B virus
infected patients. World J Gastroenterol 2012;18:2784-2792.
32. Raftopoulos SC, George J, Bourliere M, Rossi E, de Boer WB, Jeffrey
GP, et al. Comparison of noninvasive models of fibrosis in chronic hep-
atitis B. Hepatol Int 2012;6:457-467.
33. Wu SD, Ni YJ, Liu LL, Li H, Lu LG, Wang JY. Establishment and vali-
dation of a simple noninvasive model to predict significant liver fibrosis
in patients with chronic hepatitis B. Hepatol Int 2012;6:360-368.
34. Yoon KT, Lim SM, Park JY, Kim DY, Ahn SH, Han KH, et al. Liver
stiffness measurement using acoustic radiation force impulse (ARFI) elas-
tography and effect of necroinflammation. Dig Dis Sci 2012;57:1682-
1691.
HEPATOLOGY, January 2015
35. Zhu CL, Li WT, Li Y, Gao RT. Serum levels of tissue inhibitor of
metalloproteinase-1 are correlated with liver fibrosis in patients with
chronic hepatitis B. J Dig Dis 2012;13:558-563.
36. Basar O, Yimaz B, Ekiz F, Ginis Z, Altinbas A, Aktas B, et al. Non-
invasive tests in prediction of liver fibrosis in chronic hepatitis B and
comparison with post-antiviral treatment results. Clin Res Hepatol
Gastroenterol 2013;37:152-158.
37. de Ledinghen V, Vergniol J, Barthe C, Foucher J, Chermak F, Le Bail
B, et al. Non-invasive tests for fibrosis and liver stiffness predict 5-year
survival of patients chronically infected with hepatitis B virus. Aliment
Pharmacol Ther 2013;37:979-988.
38. Gumusay O, Ozenirler S, Atak A, Sonmez C, Ozkan S, Tuncel AF,
et al. Diagnostic potential of serum direct markers and non-invasive
fibrosis models in patients with chronic hepatitis B. Hepatol Res 2013;
43:228-237.
39. Ma J, Jiang Y, Gong G. Evaluation of seven noninvasive models in
staging liver fibrosis in patients with chronic hepatitis B virus infection.
Eur J Gastroenterol Hepatol 2013;25:428-434.
40. Shrivastava R, Sen S, Banerji D, Praharaj AK, Chopra GS, Gill SS.
Assessment of non-invasive models for liver fibrosis in chronic hepatitis
B virus related liver disease patients in resource limited settings. Indian
J Pathol Microbiol 2013;56:196-199.
41. Tseng PL, Wang JH, Hung CH, Tung HD, Chen TM, Huang WS,
et al. Comparisons of noninvasive indices based on daily practice
parameters for predicting liver cirrhosis in chronic hepatitis B and hep-
atitis C patients in hospital and community populations. Kaohsiung J
Med Sci 2013;29:385-395.
42. Ucar F, Sezer S, Ginis Z, Ozturk G, Albayrak A, Basar O, et al. APRI,
the FIB-4 score, and Forn’s index have noninvasive diagnostic value for
liver fibrosis in patients with chronic hepatitis B. Eur J Gastroenterol
Hepatol 2013;25:1076-1081.
43. Wang H, Xue L, Yan R, Zhou Y, Wang MS, Cheng MJ, et al.
Comparison of FIB-4 and APRI in Chinese HBV-infected patients
with persistently normal ALT and mildly elevated ALT. J Viral Hepat
2013;20.
44. Wang Y, Xu MY, Zheng RD, Xian JC, Xu HT, Shi JP, et al. Prediction
of significant fibrosis and cirrhosis in hepatitis B e-antigen negative
patients with chronic hepatitis B using routine parameters. Hepatol Res
2013;43:441-451.
45. Xun YH, Zang GQ, Guo JC, Yu XL, Liu H, Xiang J, et al. Serum
hepatitis B surface antigen quantification as a useful assessment for sig-
nificant fibrosis in hepatitis B e antigen-positive hepatitis B virus car-
riers. J Gastroenterol Hepatol 2013;28:1746-1755.
46. Zeng DW, Liu YR, Zhang JM, Zhu YY, Lin S, You J, et al. Serum cer-
uloplasmin levels correlate negatively with liver fibrosis in males with
chronic hepatitis B. PloS One 2013;8:e77942.
47. Castera L, Vergniol J, Foucher J, Le Bail B, Chanteloup E, Haaser M,
et al. Prospective comparison of transient elastography, Fibrotest, APRI,
and liver biopsy for the assessment of fibrosis in chronic hepatitis C.
Gastroenterology 2005;128:343-350.
48. Scaradavou A. HIV-related thrombocytopenia. Blood Rev 2002;16:73-76.
49. Reuber MD. Effect of age and testosterone on the induction of hyper-
plastic nodules, carcinomas, and cirrhosis of the liver in rats ingesting
n-2-fluorenyldiacetamide. Eur J Cancer 1976;12:137-141.
50. Pohl A, Behling C, Oliver D, Kilani M, Monson P, Hassanein T.
Serum aminotransferase levels and platelet counts as predictors of
degree of fibrosis in chronic hepatitis C virus infection. Am J Gastroen-
terol 2001;96:3142-3146.
Supporting Information
Additional Supporting Information may be found at
onlinelibrary.wiley.com/doi/10.1002/hep.27382/suppinfo.