Editorial

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

NAFLD Fibrosis Score: Is It Ready for Wider Use in Clinical Practice and forClinical Trials?

58

59

60

61

62

63

64

65

See “Simple noninvasive systems predict long-term outcomes of patients with nonalcoholicfatty liver disease,” by Angulo P, Bugianesi E,Bjornsson ES, et al, on page 000.

onalcoholic fatty liver disease (NAFLD) at large is

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

Nrelatively benign, but some individuals with NAFLDdevelop major hepatic and extrahepatic complications(NAFLD-at-risk). Hepatic complications of NAFLD in-clude progressive liver disease leading to cirrhosis and liverfailure and the development of hepatocellular carcinomawhich is largely confined to those with advanced fibrosisand cirrhosis.1,2 It has been suggested that NAFLD is anindependent risk factor for accelerated cardiovasculardisease, and in fact, several recent epidemiologic andcohort studies have shown that cardiovascular disease isthe most common cause of death in individuals withNAFLD.3–5 Currently, liver histology is the most widelyused method to identify individuals with NAFLD who areat risk to develop cirrhosis and subsequent liver failure.Traditionally, the presence of steatohepatitis and advancedfibrosis are considered harbingers of adverse hepatic out-comes in individuals with NAFLD.6 However, performinga liver biopsy on all individuals with NAFLD is not apractical proposition; therefore, there have been numerouscross-sectional investigations to noninvasively identifysteatohepatitis and advanced fibrosis in individuals withNAFLD.7 Some of these noninvasive tests and stratifica-tion methods are reasonably predictive of advanced liverhistology, but to date few studies have investigatednoninvasive tools to predict adverse hepatic outcomes inthis patient population. In fact, the development ofnoninvasive tools that precisely predict long-term hepaticand extrahepatic adverse outcomes remains a major unmetneed in the NAFLD field.

The NAFLD Fibrosis Score (NAFLD-FS), BARD, AST toPlatelet Ratio Index (APRI), and FIB-4 are among the morewidely investigated noninvasive tools to cross-sectionallypredict advanced fibrosis in NAFLD.8–11 These scoringsystems are based on easily available clinical variables andeach exhibited varying degrees of accuracy depending onthe patient population studied. The NAFLD-FS is calcu-lated based on age, body mass index, hyperglycemia ordiabetes, aspartate aminotransferase (AST)/alanine amino-transferase (ALT), platelets, and albumin (www.nafldscore.com).8 The BARD score was developed by Harrison et al9

on a large cohort of patients with biopsy-proven NAFLDand it is based on body mass index, AST/ALT, and thepresence of diabetes. The FIB-4 (calculated based on age,

EDI 5.2.0 DTD � YGAST58620_proof

AST, ALT, and platelets) and the APRI were originallydeveloped to predict advanced fibrosis in patients withchronic hepatitis C infection, but were subsequently inves-tigated for the same task in patients with NAFLD.10–12 Ofthe 4 scoring systems, NAFLD-FS has received the mostextensive validation and it has been recommended forclinical use in the recent US multisociety practice guidelineon the diagnosis and management of NAFLD.6

In this issue of GASTROENTEROLOGY, Angulo et al13 eval-uated the performance of these 4 noninvasive scoringsystems (NAFLD-FS, APRI, FIB-4, and BARD) to predictliver-related complications and death/liver transplantationin patients with NAFLD.14 It was a retrospective, inter-national study that included 320 patients from referralcenters from the United States, Italy, Iceland, Thailand,the United Kingdom, and Australia. The subjects werepredominantly white (92%) and overweight or obese (95%)adults. As it generally is the case with cohorts reportedfrom specialized referral centers, this cohort was enrichedwith nonalcoholic steatohepatitis (NASH) and advancedfibrosis accounting for nearly 50% of the patients. Subjectswere followed for a median duration of 104.8 months(range, 3–317). Per study protocol, 11 patients wereexcluded from the liver-related events analysis becausethose events occurred within first 3 months after liverbiopsy. None of the subjects received vitamin E, thiazoli-dinediones, or had bariatric surgery during the studyperiod. Complete follow-up data were available for 80% ofsubjects. The overall rates of liver-related adverse eventsand death/liver transplantation were 14% and 13%,respectively. Complications of cirrhosis were the thirdmost common cause of death, following cardiovascularevents and nonhepatic malignancies. Interestingly, therewas no greater frequency of liver-related adverse events ordeath/liver transplantation among patients with definiteNASH or severe steatosis. However, there was significantdifference in outcomes between patients with and withoutadvanced fibrosis. The mean score of all 4 noninvasivesystems correlated well with the degree of fibrosis (fibrosisstage 0 vs 1–2 vs 3–4) on baseline liver biopsies. The areaunder the receiver operating characteristic curve forNAFLD-FS, FIB-4, APRI, and BARD for predicting adverseliver-related outcomes were 0.86, 0.81, 0.80, and 0.73, andfor death/liver transplantation were 0.70, 0.67, 0.63, and0.66, respectively. Subsequently, using their establishedcutoff points, the investigators categorized the studycohort into low, intermediate, or high-risk groups. Forliver-related events, the adjusted hazard ratio was signifi-cantly higher for intermediate and high-risk groups forNAFLD-FS, APRI, and BARD, but it was higher only in the

� 22 August 2013 � 7:23 pm � ce

GASTROENTEROLOGY 2013;-:1–3

Table 1. Relationship Between Baseline Fibrosis Scores and Mortality Among National Health and Nutritional Examination Survey IIIParticipants With Nonalcoholic Fatty Liver Disease (NAFLD)

NAFLD-FS APRI FIB-4

Hazard ratio (95% CI)a Hazard ratio (95% CI)a Hazard ratio (95% CI)a

All-cause mortalityLow score Reference Reference ReferenceMedium score 1.26 (0.98-1.64) 1.32 (0.78-2.23) 1.46 (1.16-1.83)High score 1.69 (1.09-2.63) 1.85 (1.02-3.37) 1.66 (0.98-2.82)

Cardiovascular mortalityLow score Reference Reference ReferenceMedium score 2.16 (1.41-3.29) 0.97 (0.40-2.34) 1.75 (1.26-2.43)High score 3.46 (1.91-6.25) 2.53 (1.33-4.83) 2.68 (1.44-4.99)

Liver disease mortalityLow score Reference Reference ReferenceMedium score 0.49 (0.08-2.83) 6.08 (0.77-48.21) 0.68 (0.11-4.05)High score 0.07 (0.00-1.25) 3.01 (0.20-45.62) 1.32 (0.12-14.80)

NOTE. Modified with permission from Kim et al.15 Q3

APRI, AST to Platelet Ratio Index; CI, confidence interval; FIB-4, Fibrosis 4 Score; NAFLD-FS, NAFLD Fibrosis Score.aHazard ratio adjusted for age, gender, race/ethnicity, education, income, diabetes, hypertension, history of cardiovascular disease, lipid loweringtherapy, smoking, waist circumference, alcohol consumption, caffeine intake, total cholesterol, HDL-cholesterol, transferrin saturation, and C-reactiveprotein.

Editorial, continued

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

high-risk category for the FIB-4. The risk of death/livertransplantation steadily increased in the intermediate andhigh-risk groups for NAFLD-FS, only in the high-riskgroups for the APRI and FIB-4 scores, but was not asso-ciated with any of the BARD risk groups. The authorsconcluded that NAFLD-FS was the best indicator ofNAFLD-at-risk, based on hazard ratios. This study onceagain confirmed that it is advanced fibrosis that is asso-ciated with long-term serious outcomes in patients withNASH.14,15

Angulo et al should be congratulated for investigatingthese commonly cited noninvasive tools to predict long-term patient outcomes in NAFLD. As the authors admit,this study is far from perfect and carries several impor-tant limitations as outlined in their discussion. However,it complements a recent population-based study re-ported by Kim et al,16 which consisted of 11,154 par-ticipants of the National Health and NutritionalExamination Survey III conducted from 1988 to 1994who were followed for mortality until December 2006using the National Death Index. At entry, these partici-pants had extensive evaluation, including a gallbladderultrasound which, in turn, was later evaluated for thepresence of hepatic steatosis. From this cohort, theyidentified 4083 individuals (34%) with NAFLD based onclinical, laboratory, and imaging criteria. They investi-gated if baseline NAFLD-FS, FIB-4, and APRI predictedpatient outcomes among these individuals with NAFLDover a median follow-up of 14.5 years (range, 0.3–18.1).There were 779 individuals with NAFLD who died dur-ing follow-up; cardiovascular disease (37%) was the mostcommon cause of death followed by malignancy in 21%,but there were very few liver-related deaths (2.4%). Thisvery low rate of liver-related deaths was not surprisingbecause the prevalence of advanced fibrosis at baseline as

EDI 5.2.0 DTD � YGAST58620_proof

2

characterized by the noninvasive fibrosis markers wasvery low. For example, only 3.2% of the patients withNAFLD had advanced fibrosis using NAFLD-FS of>0.676 as a surrogate. This study showed a progressiveincrease in overall mortality rate with increasing fibrosisscores after adjustment for other known predictors ofmortality (Table 1). These authors have also shown thatthis increased overall mortality was almost entirelyowing to cardiovascular causes. Baseline fibrosis scoresdid not predict liver-related mortality, very likely becauseof very low liver-related mortality observed during thefollow-up period.

To sum up these 2 studies, the NAFLD-FS predictedoverall mortality in 2 separate NAFLD cohorts withvarying underlying characteristics—the National Healthand Nutritional Examination Survey III cohort with lowprevalence (3.2%) and Angulo’s cohort with a greaterprevalence (23.4%) of high NAFLD-FS. These observationsprovide further evidence to the recommendation by therecent multisociety NAFLD practice guidelines that rec-ommended NAFLD-FS may be used to identify NAFLDpatients at high risk for advanced fibrosis.6

Another potentially important finding of the study byAngulo et al is that there was no significant associationbetween histologically evident definite NASH at baselineand overall or liver-related mortality among patients withNAFLD. If confirmed, this observation may have pro-found impact on the design of the clinical trials forpatients with NASH. The improvement in the NAFLDactivity score and the resolution of NASH are widely usedas primary endpoints for clinical trials in NASH. Recently,Younossi et al14 have shown that NAFLD activity scoreof >4 (typical eligibility cutoff for the NASH clinical trials)did not correlate with liver-related mortality (adjustedhazard ratio, 2.92; 95% confidence interval, 0.95–8.95).

� 22 August 2013 � 7:23 pm � ce

Q4

1Q2

Editorial, continued

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

These evolving data call for the upcoming AASLD/FDAworkshop “Trial Designs and Endpoints for Liver DiseaseSecondary to Nonalcoholic Fatty Liver Disease”(September 5–6, 2013 at Silver Springs, Maryland) tocritically discuss (a) the validity of NAFLD activity scoreand steatohepatitis as primary endpoints, (b) the need forenriching clinical trials with individuals with advancedfibrosis, and finally (c) whether NAFLD-FS is ready to beincorporated as one of the primary endpoints for NASHclinical trials.

SAMER GAWRIEHDivision of Gastroenterology and HepatologyMedical College of WisconsinMilwaukee, Wisconsin

NAGA CHALASANIDivision of Gastroenterology and HepatologyIndiana University School of MedicineIndianapolis, Indiana

293

294

295

References

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

1. Vernon G, Baranova A, Younossi ZM. Systematic review: the epide-miology and natural history of non-alcoholic fatty liver disease andnon-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther2011;34:274–285.

2. White DL, Kanwal F, El-Serag HB. Association between nonalco-holic fatty liver disease and risk for hepatocellular cancer, basedon systematic review. Clin Gastroenterol Hepatol 2012;10:1342–1359.

3. Targher G, Day CP, Bonora E. Risk of cardiovascular disease in pa-tients with nonalcoholic fatty liver disease. N Engl J Med 2010;363:1341–1350.

4. Rafiq N, Bai C, Fang Y, et al. Long-term follow-up of patients withnonalcoholic fatty liver. Clin Gastroenterol Hepatol 2009;7:234–238.

5. Ong JP, Pitts A, Younossi ZM. Increased overall mortality and liver-related mortality in non-alcoholic fatty liver disease. J Hepatol2008;49:608–612.

6. Chalasani N, Younossi Z, Lavine JE, et al. The diagnosis and man-agement of non-alcoholic fatty liver disease: practice guideline by theAmerican Gastroenterological Association, American Association forthe Study of Liver Diseases, and American College of Gastroenter-ology. Gastroenterology 2012;142:1592–1609.

7. Machado MV, Cortez-Pinto H. Non-invasive diagnosis of non-alcoholicfatty liver disease. A critical appraisal. J Hepatol 2013;58:1007–1019.

EDI 5.2.0 DTD � YGAST58620_proof

8. Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: anoninvasive system that identifies liver fibrosis in patients withNAFLD. Hepatology 2007;45:846–854.

9. Harrison SA, Oliver D, Arnold HL, et al. Development and validationof a simple NAFLD clinical scoring system for identifying patientswithout advanced disease. Gut 2008;57:1441–1447.

10. Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive indexcan predict both significant fibrosis and cirrhosis in patients withchronic hepatitis C. Hepatology 2003;38:518–526.

11. Sterling RK, Lissen E, Clumeck N, et al. Development of a simplenoninvasive index to predict significant fibrosis in patients with HIV/HCV co-infection. Hepatology 2006;43:1317–1325.

12. Shah AG, Lydecker A, Murray K, et al. Comparison of noninvasivemarkers of fibrosis in patients with nonalcoholic fatty liver disease.Clin Gastroenterol Hepatol 2009;7:1104–1112.

13. Angulo P, Bugianesi E, Bjornsson ES, et al. Simple non-invasivesystems predict long-term outcomes of patients with nonalcoholicfatty liver disease. Gastroenterology 2013;000:000–000. Q

14. Younossi ZM, Stepanova M, Rafiq N, et al. Pathologic criteriafor nonalcoholic steatohepatitis: interprotocol agreement and abilityto predict liver-related mortality. Hepatology 2011;53:1874–1882.

15. Musso G, Gambino R, Cassader M, et al. Meta-analysis: naturalhistory of non-alcoholic fatty liver disease (NAFLD) and diagnosticaccuracy of non-invasive tests for liver disease severity. Ann Med2011;43:617–649.

16. Kim D, Kim WR, Kim HJ, et al. Association between noninvasivefibrosis markers and mortality among adults with nonalcoholicfatty liver disease in the United States. Hepatology 2013;57:1357–1365.

Reprint requestsAddress requests for reprints to: Naga Chalasani, MD, David W. Crabb

Professor of Medicine, Professor of Cellular & Integrative Physiology,Director, Division of Gastroenterology and Hepatology, Indiana UniversitySchool of Medicine, 1050 Wishard Blvd, RG 4100, Indianapolis,IN, 46202. e-mail: nchalasa@iupui.edu.

Conflicts of interestThe authors have made the following disclosures. Dr Chalasani

served as a consultant in past 12 months to Lilly, Salix, Aegerion, Merck,and Abbott in the area of drug hepatotoxicity. He has research grantsupport from Gilead, Takeda, Galectin, Cumberland Pharmaceuticals,Intercept pharmaceuticals and Enterome.

FundingThis work was in part supported by NIH grant K24 DK069290

(Chalasani).© 2013 by the AGA Institute

0016-5085/$36.00http://dx.doi.org/10.1053/j.gastro.2013.08.025

� 22 August 2013 � 7:23 pm � ce

3

316

317

318

319

320

321

322

323

324

325

326

327

328