Thiazolidinediones and Advanced Liver Fibrosis in Nonalcoholic Steatohepatitis: A Meta-analysis

doi:10.1001/JAMAINTERNMED.2016.9607

Thiazolidinediones and Advanced Liver Fibrosis

in Nonalcoholic Steatohepatitis

A Meta-analysis

Giovanni Musso, MD; Maurizio Cassader, PhD; Elena Paschetta, MD; Roberto Gambino, PhD

IMPORTANCE

Nonalcoholic steatohepatitis (NASH) is projected to be the leading cause of

liver transplantation by 2020. Advanced fibrosis (stage F3-F4) on liver biopsy independently

predicts all-cause and liver-related mortality in NASH. There are no known efficacious

treatments for advanced fibrosis related to NASH. Thiazolidinedione therapy has been

extensively evaluated in NASH, and new randomized clinical trials (RCTs) of its efficacy have

been completed.

OBJECTIVE To synthesize the evidence about the association of thiazolidinedione therapy

with advanced liver fibrosis in NASH.

DATA SOURCES MEDLINE, Ovid MEDLINE In-Process, Cochrane Library, EMBASE,

clinicaltrials.gov, PubMed, and Scopus databases (without language restrictions), as well as

other registries and scientific meeting presentations, from database inception through

August 15, 2016.

STUDY SELECTION Randomized clinical trials evaluating the effect of thiazolidinedione

therapy on histologic features of the liver in biopsy-proven NASH.

DATA EXTRACTION AND SYNTHESIS Two investigators extracted study data independently and

in duplicate and rated the risk of bias using the Cochrane Risk of Bias Tool.

MAIN OUTCOMES AND MEASURES The primary outcome was a dichotomous improvement in

advanced fibrosis on liver biopsy, defined as an improvement in fibrosis stage from F3-F4 to

F0-F2. Secondary outcomes were at least a 1-point improvement in fibrosis of any stage and

NASH resolution. This meta-analysis also evaluated adverse effects of thiazolidinedione

therapy, including weight gain, lower limb edema, congestive heart failure, bone fractures,

cancer, and anemia. With the use of random-effects models, dichotomous variables are

presented as odds ratios (ORs) with 95% CIs, and continuous variables are presented as

weighted mean differences with 95% CIs.

RESULTS This study analyzed 8 RCTs (5 evaluating pioglitazone use and 3 evaluating

rosiglitazone maleate use) enrolling 516 patients with biopsy-proven NASH for a duration of 6

to 24 months. Among all studies combined, thiazolidinedione therapy was associated with

improved advanced fibrosis (OR, 3.15; 95% CI, 1.25-7.93; P =.01;I

2

= 0%), fibrosis of any

stage (OR, 1.66; 95% CI, 1.12-2.47; P =.01;I

2

= 0%), and NASH resolution (OR, 3.22; 95% CI,

2.17-4.79; P < .001; I

2

= 0%). Analyses restricted to RCTs enrolling patients without diabetes

yielded similar results for improvement in advanced fibrosis (OR, 2.95; 95% CI, 1.04-10.90;

P =.02;I

2

= 0%), improvement in fibrosis of any stage (OR, 1.76; 95% CI, 1.02-3.03; P =.02;

I

2

= 0%), and NASH resolution (OR, 3.40; 95% CI, 1.95-5.93; P < .001; I

2

= 0%). All effects

were accounted for by pioglitazone use. Weight gain and lower limb edema occurred more

frequently with thiazolidinedione therapy (initial body weight +2.70%; 95% CI, 1.96%-4.34%;

P = .001). The small sample size of included RCTs prevented evaluation of more serious

adverse effects of thiazolidinedione therapy.

CONCLUSIONS AND RELEVANCE Pioglitazone use improves advanced fibrosis in NASH, even in

patients without diabetes. Whether this finding translates to improvement in risk for clinical

outcomes requires further study.

JAMA Intern Med. 2017;177(5):633-640. doi:10.1001/jamainternmed.2016.9607

Published online February 27, 2017. Last corrected on April 4, 2017.

Invited Commentary

page 640

Supplemental content

Author Affiliations: Emergency

Department, Humanitas Gradenigo

Hospital, Turin, Italy (Musso,

Paschetta); Department of Medical

Sciences, University of Turin, Turin,

Italy (Cassader, Gambino).

Corresponding Author: Giovanni

Musso, MD, Emergency Department,

Humanitas Gradenigo Hospital, Gorso

Regina Margherita 8, 10132 Turin,

Italy (giovanni_musso@yahoo.it).

Research

JAMA Internal Medicine | Original Investigation

(Reprinted) 633

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/intemed/936202/ by a Universita Torino User on 05/02/2017

N

onalcoholic fatty liver disease (NAFLD) is the most

common chronic liver disease in the world, encom-

passing a histological spectrum ranging from simple

steatosis to steatosis plus necroinflammation, known as non-

alcoholic steatohepatitis (NASH), with variable stages of fibro-

sis. Both fibrosis stage and NASH can only be assessed by liver

biopsy.

1

There is no established treatment for NASH, which is the

second leading cause of liver disease among adults awaiting

liver transplant and is projected by 2020 to be the leading in-

dication for liver transplant.

2,3

Extensive experimental and epi-

demiological evidence suggests that the presence of ad-

vanced fibrosis (stage F3-F4) (ie, bridging fibrosis or cirrhosis)

on liver biopsy is the only independent predictor of poor out-

comes in NAFLD; overall and liver-related mortality, liver trans-

plant, and liver-related complications are increased in ad-

vanced fibrosis but not in patients with NASH or milder fibrosis

(stage F0-F2), whose prognosis is similar to that of the gen-

eral population.

4,5

Recent guidelines highlight the need to iden-

tify patients with NAFLD with advanced fibrosis to target them

for more intensive monitoring for the onset of complications.

6,7

However, although reversal of advanced fibrosis has been gen-

erally associated with improved clinical outcomes in other

causes of chronic liver disease,

8,9

this stage of disease was not

improved by any of the treatments evaluated to date in ran-

domized clinical trials (RCTs) of NASH.

10,11

The thiazolidinedione antidiabetic agents have been ex-

tensively evaluated in NASH. While access to rosiglitazone ma-

leate has been restricted by the US Food and Drug Administra-

tion, pioglitazone hydrochloride continues to be recommended

in current diabetes guidelines, and novel data evaluating this

drug in NASH have been recently published.

12

The results of a

previous meta-analysis

10

suggested that thiazolidinedione

therapy improved histological features of NASH but not ad-

vanced fibrosis. Therefore, we analyzed the evidence on thia-

zolidinedione therapy in NASH, focusing on their effect in ad-

vanced fibrosis.

Methods

Data Sources and Searches

The study protocol was approved by the Humanitas Grad-

enigo Review Board. We searched English-language and non–

English-language publications in MEDLINE, Ovid MEDLINE

In-Process, Cochrane Library, EMBASE, clinicaltrials.gov,

PubMed, and Scopus databases from database inception

through August 15, 2016. We also reviewed abstracts from an-

nual meetings of the American Association for the Study of

Liver Disease, American Gastroenterological Association, Eu-

ropean Association for the Study of Liver, American Diabetes

Association, European Association for the Study of Diabetes,

and Digestive Disease Week. All included references were sub-

jected to the same quality assessment.

Search terms were nonalcoholic steatohepatitis (NASH),

nonalcoholic fatty liver disease (NAFLD), fatty liver, liver fat,

steatosis, liver enzymes, transaminase, alanine aminotransfer-

ase (ALT), aspartate aminotransferase (AST), γ-glutamyl trans-

ferase (GGT), severity of liver disease, fibrosis, advanced fibro-

sis, fibrosis stage F3, fibrosis stage F4,bridging fibrosis, cirrhosis,

treatment, therapy, efficacy, trial, thiazolidinedione, rosiglita-

zone, pioglitazone, troglitazone, glitazone, and peroxisome pro-

liferator–activated receptor γ agonist (PPAR-γ agonist). An ex-

ample of the full electronic search strategy is included in the

eMethods in the Supplement.

Study Selection

Inclusion criteria were English-language and non–English

-language articles reporting RCTs enrolling participants of any

sex or racial/ethnic origin with NAFLD or NASH, diagnosed

on the basis of radiological or histological evidence of steato-

sis according to accepted criteria.

1

Relevant meta-analyses

were also included if they followed the Preferred Reporting

Items for Systematic Reviews and Meta-analyses (PRISMA)

guidelines.

13

Excluded from the meta-analysis were nonhuman stud-

ies, nonrandomized trials, letters, and case reports. Also ex-

cluded were studies enrolling fewer than 10 participants, ar-

ticles not reporting outcomes of interest or primary data

(editorials and review articles), and investigations using inad-

equate case definitions or enrolling patients with secondary

steatosis (eg, drug-induced steatosis and total parenteral nu-

trition–induced steatosis).

Data Extraction and Quality Assessment

Data were extracted from each study by 2 of us (G.M. and M.C.)

independently and in duplicate. Agreement between the 2 re-

viewers on study selection and quality assessment ofstudieswas

evaluated by κ statistics, and disagreement was resolved by mu-

tual discussion. Authors were contacted to obtain further data

and to verify methodological quality. Data were then extracted

from each study independently and in duplicate by 2 of us (G.M.

and R.G.) using a predefined protocol and a data extraction sheet.

Discrepancies were resolvedby mutual discussion. Methodologi-

cal quality of RCTswas assessed using each item specified by the

Cochrane Risk of Bias Tool(score range, 0-8)

14

(Table). Random-

ized clinical trials scoring higher than 6 were arbitrarily consid-

ered as having a low risk of bias. The analysis was performed in

Key Points

Question What is the association of thiazolidinedione therapy

with advanced liver fibrosis in nonalcoholic steatohepatitis?

Findings In this meta-analysis of 8 randomized clinical trials

enrolling 516 patients with biopsy-proven nonalcoholic

steatohepatitis, thiazolidinedione therapy was associated with

reversed advanced fibrosis, improved overall fibrosis stages, and

resolution of nonalcoholic steatohepatitis. Pioglitazone

hydrochloride use accounted for all of the effects of

thiazolidinedione therapy in nonalcoholic steatohepatitis, and

these benefits were observed in patients without diabetes as well.

Meaning Pioglitazone use improves advanced fibrosis in

nonalcoholic steatohepatitis, even in patients without diabetes,

and may thus halt disease progression to end-stage liver disease in

this patient population.

Research Original Investigation Thiazolidinedione Therapy and Advanced Liver Fibrosis

634 JAMA Internal Medicine May 2017 Volume 177, Number 5 (Reprinted) jamainternalmedicine.com

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/intemed/936202/ by a Universita Torino User on 05/02/2017

accord with the Cochrane Handbook for Systematic Reviews of

Interventions

14

using a statistical software program (RevMan, ver-

sion 5.3.5; The Cochrane Collaboration), and data were reported

according to PRISMA guidelines.

13

Data Synthesis and Analysis

The primary outcome variable was a dichotomous improve-

ment in advanced fibrosis (stage F3-F4) on liver biopsy, de-

fined as a 2-point improvement in fibrosis stage from F3-F4

to F0-F2 on the NASH Clinical Research Network Scale. An im-

provement in advanced fibrosis was defined in the following

2 ways: (1) the number of individuals among all patients with

NASH included in the RCT whose fibrosis stage had changed

from F3-F4 to F0-F2 at the end of treatment and (2) the num-

ber of individuals among patients with NASH with advanced

fibrosis (stage F3-F4) at baseline whose fibrosis stage had

changed from F3-F4 to F0-F2 at the end of treatment.

Secondary dichotomous outcome variables were at least

a 1-point improvement in fibrosis of any stage on the NASH

Clinical Research Network Scale in patients with NASH and

NASH resolution, with fibrosis stage and NASH defined ac-

cording to current guidelines.

1

We also evaluated adverse ef-

fects of thiazolidinedione therapy, including weight gain, lower

limb edema, congestive heart failure, bone fractures, cancer,

and anemia.

Dichotomous variables are presented as odds ratios (ORs)

with 95% CIs, and continuous variables are presented as weighted

mean differences with 95% CIs. We conservativelyused random-

effects models, with significance set at P = .05.Statistical hetero-

geneity was assessed with the I

2

statistic. Using I

2

of 50% or

higher, we planned to explore individual study characteristics

and those of subgroups in the main body of evidence.

Sensitivity analysis was performed by removing 1 study at

a time and repeating the meta-analysis to assess whether any

single study substantially affected pooled estimates. In addi-

tion, we planned a priori subgroup analysis according to the

following criteria: RCTs evaluating rosiglitazone use vs RCTs

evaluating pioglitazone use, RCTs enrolling exclusively pa-

tients without diabetes vs RCTs also enrolling patients with dia-

betes, and treatment duration of 1 year or less vs longer than 1

year and different dosages, as well as for each item of the

Cochrane Risk of Bias Tool.

When at least 8 comparisons were available, the effect of

age, changes in insulin resistance (as estimated by the homeo-

stasis model of insulin resistance index), and treatment dura-

tion on assessed outcomes was evaluated by meta-regression

analysis (using random-effects models, with within-study vari-

ance estimated with the unrestricted maximum likelihood

method). Publication bias was examined using funnel plots and

the Egger test.

Management of Missing Data

Missing data were managed by contacting the corresponding

authors of the RCTs. If this contact was unsuccessful, missing

Table. Randomized Clinical Trials (RCTs) With Posttreatment Histological Features of the Liver Assessing Thiazolidinedione Therapy

in Nonalcoholic Steatohepatitis (NASH) Included in the Meta-analysis

Source

No. of

Patients

Mean

Age, y

Male,

%

Mean

BMI

Diabetes,

%

Agent

(Daily Dosage)

Trial

Duration, mo Comparator

BMI

Change From

Baseline, %

Quality

Score

a

Ratziu

et al,

15

2008

63 54 59 31 31 Rosiglitazone

maleate (8 mg)

12 Placebo +1 7 (H)

Sanyal

et al,

16

2004

20 46 50 32 0 Pioglitazone

hydrochloride

(30 mg)

6 Vitamin E 0 7 (E)

Belfort

et al,

17

2006

55 51 45 34 48 Pioglitazone

hydrochloride

(45 mg)

6 Placebo +2.7 7 (E)

Aithal

et al,

18

2008

74 54 61 31 0 Pioglitazone

hydrochloride

(30 mg)

12 Placebo +3 7 (E)

Idilman

et al,

19

2008

74 47 59 32 0 Rosiglitazone

maleate (8 mg)

12 Metformin

hydrochloride,

placebo

−2.6 4 (B, C, D, E)

Omer

et al,

20

2010

64 49 55 31 70 Rosiglitazone

maleate (4 mg)

12 Metformin

hydrochloride,

metformin

hydrochloride

plus

rosiglitazone

maleate

0 4 (B, C, D, E)

Sanyal

et al,

21

2010

247 46 40 34 0 Pioglitazone

hydrochloride

(30 mg)

24 Vitamin E,

placebo

+4.8 8

Cusi

et al,

12

2016

101 51 70 34 51 Pioglitazone

hydrochloride

(45 mg)

18 Placebo +1 8

Abbreviation: BMI, body mass index (calculated as weight in kilograms divided

by height in meters squared).

a

The Cochrane Risk of Bias Tool (score range, 0-8) score for RCTs is reported,

with failing items in parentheses. Quality items of RCTs according to the

Cochrane Risk of Bias Tool are as follows: A (adequate method of sequence

generation), B (masking of participants performed), C (masking of personnel

performed), D (masking of assessors performed), E (randomization

concealment adequate), F (adequate assessment of each outcome),

G (selective outcome reporting avoided), and H (intent-to-treat analysis of the

results).

Thiazolidinedione Therapy and Advanced Liver Fibrosis Original Investigation Research

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine May 2017 Volume 177, Number 5 635

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/intemed/936202/ by a Universita Torino User on 05/02/2017

histological scores were calculated from the raw numbers given

in tables or estimated from bar charts. For missing SDs of the

mean change in scores and where the P value was provided

for a comparison between treated and control groups, the SD

was calculated by converting the P value to a t statistic with

appropriate df and then calculating SEs and SDs. If neither the

SD nor the P values were supplied, imputation of an SD from

studies with similar measurement methods, trial duration, and

measurement error was used if available, tested in a sensitiv-

ity analysis, and reported if the estimate differed meaning-

fully from previous estimates.

Results

The agreement between reviewers was good to excellent. The

κ statistics were 0.88 for study selection and 0.92 for quality

assessment.

A flow diagram of study selection is shown in Figure 1.

We identified 8 RCTs (5 evaluating pioglitazone use and 3

evaluating rosiglitazone use, with posttreatment histological

features of the liver) enrolling 516 patients. Trial durations

were 6 to 24 months, with daily dosages ranging from 4 to 8

mg for rosiglitazone maleate and from 30 to 45 mg for piogli-

tazone hydrochloride (Table).

12,15-21

For the included RCTs,

the histopathological scoring system proposed by the NASH

Clinical Research Network was used to score the severity of

histological features of the liver, as recommended by current

guidelines.

1,6,7

Fibrosis was assessed by Masson trichrome

stain, and the pathologist intraobserver and interobserver

agreement for fibrosis staging was good to excellent (κ statis-

tic, ≥0.82).

The agreement between reviewers for quality assess-

ment was good (κ statistic, 0.84). Overall, 6 RCTs had a low risk

of bias in key domains, while 2 RCTs (both evaluating rosi-

glitazone use) demonstrated a higher risk of bias because of

unclear blinding and randomization concealment.

19,20

Pooled results of RCTs showed that thiazolidinedione

therapy was associated with improved advanced fibrosis

(Figure 2). The effect size was significant when considering all

patients with NASH (Figure 2A) and only patients with NASH

with advanced fibrosis at baseline (Figure 2B). In addition, thia-

zolidinedione therapy was associated with improved fibrosis

of any stage and induced NASH resolution (Figure 3). Statis-

tical heterogeneity was low for all evaluated outcomes, sug-

gesting a consistent effect size across studies.

After the analysis to RCTs enrolling exclusively to

patients without diabetes was restricted, pooled ORs

remained similar in magnitude and direction to the overall

effect. Among the 4 studies, thiazolidinedione therapy was

associated with improvement in advanced fibrosis (OR, 2.95;

95% CI, 1.04-10.90; P = .02; I

2

= 0%), improvement in fibro-

sis of any stage (OR, 1.76; 95% CI, 1.02-3.03; P =.02;

I

2

= 0%), and NASH resolution (OR, 3.40; 95% CI, 1.95-5.93;

P < .001; I

2

= 0%).

Trial duration, dosage, and exclusion of the 2 RCTs

16,18

with

a high risk of bias (both evaluating rosiglitazone use) did not

affect the magnitude and direction of the overall effect. Meta-

regression analysis found no association between assessed out-

comes and age, homeostasis model of insulin resistance in-

dex, and treatment duration.

The separate analyses of rosiglitazone and pioglitazone

demonstrated that the observed effects of thiazolidinedione

therapy were accounted for by pioglitazone use. Rosiglita-

zone use did not reach statistical significance for any histo-

logical outcome (Figure 2).

The Egger test and funnel plot analysis found no strong evi-

dence for publication bias. These results are shown in eFig-

ure 1 in the Supplement.

Thiazolidinedione therapy was associated with a mean

2.7% weight gain compared with controls (eFigure 2 in the

Supplement). It was also associated with a higher OR for lower

limb edema (2.36; 95% CI, 1.15-4.84; P = .02; I

2

= 0%) (6 stud-

ies), without any significant difference in agents, RCTs, or trial

duration.

Reporting of other adverse events was variable. Recog-

nized adverse effects, such as congestive heart failure, were

reported in fewer than half of the RCTs.

Figure 1. Evidence Acquisition Flow Diagram

12

337 Excluded by specific criteria

Nonhuman studies

Letters or case reports

Did not report outcome of interest

Did not report primary data

(editorials, reviews)

Pediatric population only

Inadequate case definition

<10 Patients included

200 RCTs excluded

21 Inadequate case definition

106 Not assessing treatment of interest

64 Nonrandomized trials

13 Controlled nonrandomized

51 Uncontrolled

4 Retrospective studies

5 Meta-analyses

8 RCTs without posttreatment histological

features of the liver excluded

12

553 Publications identified

16 RCTs assessing thiazolidinediones

therapy in NAFLD

8 RCTs with posttreatment histological

features of the liver assessing

thiazolidinediones therapy in NASH

included

216 Unique publications meeting

inclusion criteria

143 RCTs

13 Controlled nonrandomized

trials

51 Uncontrolled trials

4 Retrospective studies

5 Meta-analyses

NAFLD indicates nonalcoholic fatty liver disease; NASH, nonalcoholic

steatohepatitis; and RCTs, randomized clinical trials.

Research Original Investigation Thiazolidinedione Therapy and Advanced Liver Fibrosis

636 JAMA Internal Medicine May 2017 Volume 177, Number 5 (Reprinted) jamainternalmedicine.com

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/intemed/936202/ by a Universita Torino User on 05/02/2017

Discussion

In this meta-analysis of 8 RCTs of thiazolidinedione therapy,

we found that treatment for up to 24 months was associated

with improved advanced fibrosis and fibrosis of any stage and

NASH resolution. These effects were mainly accounted for by

pioglitazone use. Benefits were also observed in patients with

NASH without diabetes. Aside from weight gain and lower limb

edema, no major adverse events were reported during the trial

durations, with recognizable power limitations of our analy-

sis because of the few included RCTs.

Figure 2. Thiazolidinedione Therapy (TZD) and Improvement in Advanced Fibrosis, Improved Fibrosis of Any Stage, and Nonalcoholic Steatohepatitis

(NASH) Resolution

All patients with NASH

A

Weight, %

Favors

Controls

Favors

TZD

0.01 101.0 1000.1

OR (95% CI)

TZD

No. of

Events

No. of

Patients

Control

No. of

Events

No. of

PatientsSource

Rosiglitazone maleate

Odds Ratio (95% CI)

7.71110 8Idilman et al,

19

2008 2.43 (0.09-67.57)

10.51201 22Omer et al,

21

2010 1.11 (0.06-18.93)

10.71321 31Ratziu et al,

18

2008 0.97 (0.06-16.19)

Heterogeneity: τ

2

=

0.00; χ

2

=

0.19; P

=

.91; I

2

=

0%

Overall effect: z

=

0.30; P

=

.77

29.03632 61Total (95% CI) 1.30 (0.23-7.20)

Pioglitazone hydrochloride

9.43310 30Aithal et al,

17

2008 7.49 (0.37-151.50)

9.97260 21Belfort et al,

16

2006 16.54 (0.89-308.98)

9.84500 51Cusi et al,

12

2016 9.97 (0.52-190.16)

10.01101 10Sanyal et al,

15

2004 1.00 (0.05-18.57)

32.06802 83Sanyal et al,

20

2010 3.28 (0.64-16.78)

Heterogeneity: τ

2

=

0.00; χ

2

=

2.39; P

=

.66; I

2

=

0%

Overall effect: z

=

2.71; P

=

.007

71.021 197 3 195Total (95% CI) 4.53 (1.52-13.52)

Heterogeneity: τ

2

=

0.00; χ

2

=

4.12; P

=

.77; I

2

=

0%

Overall effect: z

=

2.44; P

=

.01

10024 260 5 256Total (95% CI) 3.15 (1.25-7.93)

Patients with NASH with advanced fibrosis at baseline

B

Weight, %

Favors

Controls

Favors

TZD

0.01 101.0 1000.1

OR (95% CI)

TZD

No. of

Events

No. of

Patients

Control

No. of

Events

No. of

Patients

Source

Rosiglitazone maleate

Odds Ratio (95% CI)

8.6130 3Idilman et al,

19

2008 4.20 (0.12-151.97)

11.5171 4Omer et al,

21

2010 0.50 (0.02-11.09)

12.4151 15Ratziu et al,

18

2008 3.50 (0.18-69.34)

Heterogeneity: τ

2

=

0.00; χ

2

=

1.06; P

=

.59; I

2

=

0%

Overall effect: z

=

0.65; P

=

.52

32.53152 22Total (95% CI) 1.84 (0.29-11.66)

Pioglitazone hydrochloride

11.1370 11Aithal et al,

17

2008 17.89 (0.76-420.49)

6.3770 2Belfort et al,

16

2006 75.00 (1.16-4868.64)

10.7470 5Cusi et al,

12

2016 14.14 (0.57-352.00)

7.2121 2Sanyal et al,

15

2004 1.00 (0.02-50.40)

32.26122 19Sanyal et al,

20

2010 8.50 (1.33-54.13)

Heterogeneity: τ

2

=

0.00; χ

2

=

2.43; P

=

.66; I

2

=

0%

Overall effect: z

=

3.55; P

<.001

67.521 35 3 39Total (95% CI) 10.17 (2.83-36.54)

Heterogeneity: τ

2

=

0.00; χ

2

=

5.71; P

=

.57; I

2

=

0%

Overall effect: z

=

3.29; P

=

.001

10024 50 5 61Total (95% CI) 5.84 (2.04-16.71)

A, Improvement in advanced fibrosis (stage F3-F4) in patients with

biopsy-proven NASH, defined as the number of patients with NASH whose

fibrosis stage had changed from F3-F4 to F0-F2 at the end of treatment.

B, Improvement in advanced fibrosis (stage F3-F4) in patients with NASH with

advanced fibrosis, defined as the number of patients with NASH with advanced

(F3-F4) fibrosis at baseline whose fibrosis stage had changed from F3-F4 to

F0-F2 at the end of treatment. In contrast to A, only patients with NASH and

advanced fibrosis were included as the denominator in B.

Thiazolidinedione Therapy and Advanced Liver Fibrosis Original Investigation Research

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine May 2017 Volume 177, Number 5 637

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/intemed/936202/ by a Universita Torino User on 05/02/2017

Thiazolidinediones and Advanced Liver Fibrosis in Nonalcoholic Steatohepatitis: A Meta-analysis

Citations

Review of recent research on hepatitis C therapy for 54th annual meeting of the American association for the study of liver diseases

Non-alcoholic fatty liver disease: causes, diagnosis, cardiometabolic consequences, and treatment strategies.

非アルコール性脂肪肝炎の治療薬およびバイオマーカーの開発における課題と機会 : 米国肝臓病学会(American Association for the Study of Liver Diseases)と米国食品医薬品局の合同ワークショップの知見と推奨

Asia-Pacific Working Party on Non-alcoholic Fatty Liver Disease guidelines 2017-Part 1: Definition, risk factors and assessment.

NAFLD and increased risk of cardiovascular disease: clinical associations, pathophysiological mechanisms and pharmacological implications

References

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement.

Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement

Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0. The Cochrane Collaboration

Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement (Chinese edition)

Related Papers (5)

Pioglitazone, Vitamin E, or Placebo for Nonalcoholic Steatohepatitis

Liraglutide safety and efficacy in patients with non-alcoholic steatohepatitis (LEAN): a multicentre, double-blind, randomised, placebo-controlled phase 2 study

The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases.

EASL-EASD-EASO Clinical Practice Guidelines for the Management of Non-Alcoholic Fatty Liver Disease

Elafibranor, an Agonist of the Peroxisome Proliferator−Activated Receptor−α and −δ, Induces Resolution of Nonalcoholic Steatohepatitis Without Fibrosis Worsening