Review
Copyright ©The Author(s) 2023.
World J Hepatol. May 27, 2023; 15(5): 609-640
Published online May 27, 2023. doi: 10.4254/wjh.v15.i5.609
Table 1 Cellular location(s) of synthesis of circulating biomarkers in nonalcoholic fatty liver disease
Serological biomarker
Cellular location(s) of synthesis
AdiponectinAdipocytes
Adipo R2Hepatocytes, skeletal muscle
ALTHepatocytes
Ang-2Liver sinusoidal endothelial cells
CatDLysosomes
ChemerinAdipocytes
CK18Hepatocytes
FGF21Hepatocytes
HAHepatic stellate cells
IL-1βAdipocytes
IL-6Adipocytes
IL-17T helper 17 cells
IL-18Macrophages
LeptinAdipocytes
PIIINPReleased during procollagen processing
PRO-C3Extracellular matrix
RBP4Adipocytes, Hepatocytes
ResistinAdipocytes
VisfatinAdipocytes, Hepatocytes
IL: Interleukin; Adipo R2: Soluble adiponectin receptor 2; FGF: Fibroblast growth factor; RBP4: Retinol binding protein 4; ALT: Alanine aminotransferase; Ang-2: Angiopoietin-2; CK18: Cytokeratin 18; CatD: Cathepsin D; HA: Hyaluronic acid; PRO-C3: N-terminal type III collagen propeptide; PIIINP: Amino-terminal propeptide of type III procollagen. Serological biomarkers are listed in the order they appear in the text.
Table 2 Serological biomarkers for the detection of nonalcoholic fatty liver disease
Ref.
Marker
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
P value
Boyraz et al[24], 2013Adipo-nectin [µg/mL]TurkeyNAFLD vs no-NAFLDObese with NAFLD (n = 63), obese non-NAFLD (n = 85)US3.20.948 (0.924-0.972)10083.5< 0.001
Boyraz et al[24], 2013Adipo-nectin [µg/mL]TurkeyGrade 3 vs Grade 1-2 steatosisObese with NAFLD (n = 63), obese non-NAFLD (n = 85)US2.560.809 (0.751-0.867)84.263.6< 0.001
Mohamed et al[25], 2017Adipo-nectin [µg/mL]EgyptNAFLD vs no-NAFLDNAFLD (n = 101), non-NAFLD controls (n = 57)Biopsy2.40.921374.396.5< 0.001
Flisiak-Jackiewicz et al[19], 2018ALTPolandNAFLD vs no-NAFLDObese with steatosis (n = 72), obese without steatosis (n = 20)MRS0.668 (0.514-0.822)0.0325
Flisiak-Jackiewicz et al[19], 2018ASTPolandNAFLD vs no-NAFLDObese with steatosis (n = 72), obese without steatosis (n = 20)MRS0.683 (0.532-0.834)0.0173
Mohamed et al[25], 2017Chemerin [ng/mL]EgyptNAFLD vs no-NAFLDNAFLD (n = 101), non-NAFLD controls (n = 57)Biopsy186.70.783656.487.788.952.6< 0.001
Kłusek-Oksiuta et al[46], 2014Chemerin [ng/mL]PolandNAFLD vs no-NAFLDSteatosis (n = 33 via MRS)MRS1900.775580.04
Hua et al[38], 2019FGF-21 [pg/mL]TaiwanPredicting high grade steatosisObese (n = 31), obese with liver steatosis (n = 83), controls (n = 89)US106.10.781 (0.687–0.874)86.560< 0.001
Hua et al[38], 2019FGF-21 + GGTTaiwanPredicting high grade steatosisObese (n = 31), obese with liver steatosis (n = 83), controls (n = 89)US3.3180.861 (0.786–0.937)89.274.6< 0.001
Hua et al[38], 2019FGF-21 + GGT + TGTaiwanPredicting high grade steatosisObese (n = 31), obese with liver steatosis (n = 83), controls (n = 89)US5.4030.871 (0.801–0.942)83.882.5< 0.001
Hua et al[38], 2019FGF-21 + GGT + TGTaiwanPredicting high grade steatosisObese (n = 31), obese with liver steatosis (n = 83), controls (n = 89)US6.6610.873 (0.801–0.945)94.672.9< 0.001
Flisiak-Jackiewicz et al[19], 2018GGTPolandNAFLD vs. no-NAFLDObese with steatosis (n = 72), obese without steatosis (n = 20)MRS0.677 (0.521-0.832)0.0257
Hua et al[38], 2019GGT [U/L]TaiwanPredicting high grade steatosisObese (n = 31), obese with liver steatosis (n = 83), controls (n = 89)US21.50.840 (0.765–0.915)82.570.5< 0.001
Duan et al[18], 2022IL-17 [pg/mL]ChinaObese with NAFLD vs obeseObese with NAFLD (n = 176), obese non-NAFLD (n = 91)US40.030.97 (0.96-0.99)8993.8< 0.001
Flisiak-Jackiewicz et al[19], 2018IL-18 [pg/mL]PolandNAFLD vs no-NAFLDObese with steatosis (n = 72), obese without steatosis (n = 20)MRS326.80.680 (0.552-0.808)607534600.0058
Flisiak-Jackiewicz et al[19], 2018IL18 + ALT + AST + GGT + TGPolandNAFLD vs no-NAFLDObese with steatosis (n = 72), obese without steatosis (n = 20)MRS0.782 (0.678-0.887)61859438< 0.001
Duan et al[18], 2022IL-1β [pg/mL]ChinaObese with NAFLD vs obeseObese with NAFLD (n = 176), obese non-NAFLD (n = 91)US11.740.94 (0.91-0.97)84.685.2< 0.001
Duan et al[18], 2022IL-6 [pg/mL]ChinaObese with NAFLD vs obeseObese with NAFLD (n = 176), obese non-NAFLD (n = 91)US8.10.94 (0.91-0.96)91.280.1< 0.001
Boyraz et al[24], 2013RBP4 [µg/mL]TurkeyNAFLD vs no-NAFLDObese with NAFLD (n = 63), obese non-NAFLD (n = 85)US260.974 (0.960-0.988)10092.9< 0.001
Boyraz et al[24], 2013RBP4 [µg/mL]TurkeyGrade 3 vs Grade 1-2 steatosisObese with NAFLD (n = 63), obese non-NAFLD (n = 85)US350.782 (0.726-0.838)84.268.2< 0.001
Boyraz et al[24], 2013Resistin [ng/mL]TurkeyNAFLD vs no-NAFLDObese with NAFLD (n = 63), obese non-NAFLD (n = 85)US120.884 (0.849-0.919)10077.7< 0.001
Boyraz et al[24], 2013Resistin [ng/mL]TurkeyGrade 3 vs Grade 1-2 steatosisObese with NAFLD (n = 63), obese non-NAFLD (n = 85)US5.20.661 (0.586-0.736)36.895.5< 0.05
Flisiak-Jackiewicz et al[19], 2018TGPolandNAFLD vs no-NAFLDObese with steatosis (n = 72), obese without steatosis (n = 20)MRS0.694 (0.574-0.815)0.0015
Hua et al[38], 2019TG [mg/dL]TaiwanPredicting high grade steatosisObese (n = 31), obese with liver steatosis (n = 83), controls (n = 89)US770.732 (0.639–0.824)90.250< 0.001
Elkabany et al[47], 2020Visfatin [ng/mL]EgyptNAFLD vs no-NAFLDObese with NAFLD (n = 31), obese (n = 49), nonobese controls (n = 40)US1883.981.4
ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; AUROC: Area under the receiving operating characteristic; CI: Confidence interval; Dx: Diagnosis; IL, Interleukin; FGF: Fibroblast growth factor; GGT: Gamma-glutamyl transferase; MRS: Magnetic spectroscopy; NAFLD: Nonalcoholic fatty liver disease; NPV: Negative predictive value; PPV: Positive predictive value; RBP4: Retinol binding protein 4; Sens: Sensitivity; Spec: Specificity; TG: Triglycerides; US: Ultrasound.
Table 3 Serological biomarkers and composite scores for the detection of nonalcoholic steatohepatitis
Ref.
Marker
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
P value
Feldstein et al[66], 2013ALTItalyDiagnosing NASHNASH (n = 140), non-NASH (n = 61)Biopsy0.635 (0.556, 0.715)< 0.001
Walenbergh et al[70], 2015ALTItalyBorderline NASH vs definite NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.570.0011 (CatD vs ALT)
Walenbergh et al[70], 2015ALTItalySteatosis + Borderline NASH vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.53< 0.001 (CatD vs ALT)
Walenbergh et al[70], 2015ALTItalySteatosis vs borderline NASH + NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.660.103 (CatD vs ALT)
Walenbergh et al[70], 2015ALT [U/L]ItalySteatosis vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy> 64.50.5961.568.472.756.50.0004 (CatD vs ALT)
Manco et al[57], 2022Ang-2 [ng/mL]ItalyDiagnosing NASHNAFLD (n = 76), controls (n = 28, by ultrasound)Biopsy135.40.911 (0.844–0.979)85.785.38387.5< 0.001
Feldstein et al[66], 2013ASTItalyDiagnosing NASHNASH (n = 140), non-NASH (n = 61)Biopsy0.651 (0.573, 0.728)< 0.001
Walenbergh et al[70], 2015CatDItalyBorderline NASH vs definite NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.85
Walenbergh et al[70], 2015CatDItalySteatosis + Borderline NASH vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.88
Walenbergh et al[70], 2015CatDItalySteatosis vs borderline NASH + NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.81
Walenbergh et al[70], 2015CatD [pg/mL]ItalySteatosis vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy< 184450.9410089.592.9100
Vuppalanchi et al[68], 2014Change in ALTUnited StatesOverall histologic improvementNAFLD (n = 117)Biopsy0.79 (0.70-0.87)
Vuppalanchi et al[68], 2014Change in ALTUnited StatesResolution of NASHNAFLD (n = 117)Biopsy0.84 (0.76-0.93)
Vuppalanchi et al[68], 2014Change in ALT + CK18United StatesOverall histologic improvementNAFLD (n = 117)Biopsy0.79 (0.71-0.87)0.08 (CK18+ALT vs ALT)
Vuppalanchi et al[68], 2014Change in ALT + CK18United StatesResolution of NASHNAFLD (n = 117)Biopsy0.83 (0.75-0.92)0.92 (CK18+ALT vs ALT)
Vuppalanchi et al[68], 2014Change in CK18United States Overall histologic improvementNAFLD (n = 117)Biopsy0.72 (0.63-0.81)0.42 (CK18 vs ALT)
Vuppa-lanchi et al[68], 2014Change in CK18United StatesResolution of NASHNAFLD (n = 117)Biopsy0.69 (0.58-0.79)0.005 (CK18 vs ALT)
Walenbergh et al[70], 2015CK18ItalyBorderline NASH vs definite NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.570.0003 (CatD vs CK18)
Walenbergh et al[70], 2015CK18ItalySteatosis + Borderline NASH vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.52< 0.0001 (CatD vs CK18)
Walenbergh et al[70], 2015CK18ItalySteatosis vs borderline NASH + NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.740.4299 (CatD vs CK18)
Manco et al[57], 2022CK18 [U/L]ItalyDiagnosing NASHNAFLD (n = 76), controls (n = 28, by ultrasound)Biopsy3520.827 (0.735–0.919)77.173.27178.9< 0.001
Fitzpatrick et al[67], 2010CK18 [U/L]United KingdomPredicting NASHNAFLD (n = 45), controls (n = 13)Biopsy2070.85 (0.73–0.96)84889080
Feldstein et al[66], 2013CK18 [U/L]ItalyDiagnosing NASHNASH (n = 140), non-NASH (n = 61)Biopsy2330.93348586.993.771.6< 0.001
Walenbergh et al[70], 2015CK18 [U/L]ItalySteatosis vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy> 327.50.727263.272.862.20.0225 (CatD vs CK18)
Feldstein et al[66], 2013GGTItalyDiagnosing NASHNASH (n = 140), non-NASH (n = 61)Biopsy0.672 (0.594-0.750)< 0.001
Manco et al[95], 2007Leptin [ng/mL]ItalyPredicting NAFLD Activity ScoreNAFLD (n = 72), F0 (n = 31), F1 (n = 41)Biopsy≤ 14.90.833936547
Manco et al[95], 2007Leptin [ng/mL]ItalyPredicting NAFLD Activity ScoreNAFLD (n = 72), F0 (n = 31), F1 (n = 41)Biopsy≥ 20.454765079
Mosca et al[101], 2019PIIINP [ng/mL]ItalyDefinite NASH vs No/Borderline NASHNo/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 7.600.737 (0.66-0.81)62918575
Manco et al[95], 2007TNF-α [pg/mL]ItalyPredicting NAFLD Activity ScoreNAFLD (n = 72), F0 (n = 31), F1 (n = 41)Biopsy≤ 5.90.9111836115
Manco et al[95], 2007TNF-α [pg/mL]ItalyPredicting NAFLD Activity ScoreNAFLD (n = 72), F0 (n = 31), F1 (n = 41)Biopsy≥ 7.982969096
AUROC: Area under the receiver operating characteristic curve; ALT: Alanine aminotransferase; Ang-2: Angiopoietin-2; AST: Aspartate aminotransferase; CatD: Cathepsin D; CK18: Cytokeratin 18; CI: Confidence interval; Dx: Diagnosis; GGT: Gamma-glutamyl transferase; NAFLD: Nonalcoholic fatty liver disease; NASH: Nonalcoholic steatohepatitis; NPV: Negative predictive value; PPV: Positive predictive value; Sens: Sensitivity; Spec: Specificity; TNF-a: Tumor necrosis factor alpha.
Table 4 Serological biomarkers for the detection of fibrosis in nonalcoholic fatty liver disease
Ref.
Marker
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
P value
Mandelia et al[84], 2016CK18ItalyF1-F3 vs F0NAFLD (n = 201), F0 (n = 65), F1–F3 (n = 136)Biopsy0.75 (0.68-0.81)
Mandelia et al[84], 2016CK18ItalyF2-F2 vs F0NAFLD (n = 201), F0 (n = 65), F1–F3 (n = 136)Biopsy0.67 (0.54-0.80)
Mandelia et al[84], 2016CK18ItalyF3 vs F0NAFLD (n = 201), F0 (n = 65), F1–F3 (n = 136)Biopsy0.77 (0.56-0.97)
Fitzpatrick et al[67], 2010CK18 [U/L]United KingdomSignificant Fibrosis (≥ F2)NAFLD (n = 45), healthy controls (n = 13)Biopsy2000.66 (0.5-0.82)8340
Lebensztejn et al[80], 2011CK18 [U/L]PolandFibrosis (F1-F3) vs F0NAFLD (n = 52), NAFLD with obesity/overweight (n = 42), healthy non-obese controls (n = 25)Biopsy2100.666796056820.05
Nobili et al[79], 2010HA [ng/mL]ItalyF1 and F2+ vs F0NAFLD (n = 100), F0 (n = 35), ≥ F1 (n = 65)Biopsy≥ 12000.88 (0.81–0.96)9050
Nobili et al[79], 2010HA [ng/mL]ItalyF2+ vs F0 and F1NAFLD (n = 100), F0 (n = 35), ≥ F1 (n = 65)Biopsy21000.95 (0.91–0.99)4090
Lebensztejn et al[80], 2011HA [ng/mL]PolandFibrosis (F1-F3) vs F0NAFLD (n = 52), NAFLD with obesity/overweight (n = 42), healthy non-obese controls (n = 25)Biopsy19.10.672845552860.04
Lebensztejn et al[80], 2011HA + CK18PolandFibrosis (F1-F3) vs F0NAFLD (n = 52), NAFLD with obesity/overweight (n = 42), healthy non-obese controls (n = 25)Biopsy0.73747956630.002
Hamza et al[99], 2016PIIINP [ng/mL]EgyptPresence of steatosis in obese childrenObese with NAFLD (n = 50), obese without NAFLD (n = 5), nonobese healthy controls (n = 30)US8.57433
Mosca et al[101], 2019PIIINP [ng/mL]ItalyPresence of ≥ F2No/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 8.890.921 (0.87-0.97)84949579
Mosca et al[101], 2019PIIINP [ng/mL]ItalyPresence of F3No/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 13.20.993 (0.98-1.0)1009878100
AUROC: Area under the receiver operating characteristic curve; CK18: Cytokeratin 18; CI: Confidence interval; Dx: Diagnosis; F: Fibrosis stage; HA: Hyaluronic acid; NAFLD: Nonalcoholic fatty liver disease; NASH: Nonalcoholic steatohepatitis; NPV: Negative predictive value; PIIINP: Amino-terminal propeptide of type III procollagen; PPV: Positive predictive value; Sens: Sensitivity; Spec: Specificity; US: Ultrasound.
Table 5 Imaging biomarkers for the detection of nonalcoholic fatty liver disease
Ref.
Marker
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
P value
Yang et al[106], 2022CAP [dB/m]ChinaPredicting NAFL in with obesityNAFLD (n = 61), Non-NAFLD (n = 59), NAFL (n = 44), NASH (n = 17)US> 262.50.659 (0.561-0.758)59.160.50.0037
Yang et al[106], 2022CAP [dB/m]ChinaPredicting NAFL in with obesityNAFLD (n = 61), Non-NAFLD (n = 59), NAFL (n = 44), NASH (n = 17)US> 2580.757 (0.668-0.845)67.267.2<0.001
Chaidez et al[105], 2022CAP [dB/m]United StatesS1-S3 vs S0Total (n = 206), NAFLD (n = 116), Non-NAFLD (n = 90)Biopsy≥ 259 0.98 (0.96-0.99)94919791
Yang et al[106], 2022LSM [kPa]ChinaPredicting NAFL in with obesityNAFLD (n = 61), Non-NAFLD (n = 59), NAFL (n = 44), NASH (n = 17)US> 4.950.674 (0.577-0.771)61.464.50.0015
Yang et al[106], 2022LSM [kPa]ChinaPredicting NAFL in with obesityNAFLD (n = 61), Non-NAFLD (n = 59), NAFL (n = 44), NASH (n = 17)US> 4.650.768 (0.684-0.852)70.570.7< 0.001
Trout et al[134], 2018MRE [kPa]United StatesNAFLD stage 0-1 vs ≥ stage 2 fibrosis in patients with steatosisTotal (n = 86), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy2.280.53 (0.35-0.71)52.271.4
Trout et al[134], 2018MRE [kPa]United StatesNAFLD stage 0-1 vs ≥ stage 2 fibrosis in patients with steatosisTotal (n = 86), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy0.9413100
Middleton et al[121], 2018MRI-PDFF [%]United StatesGrade 1 steatosis vs grade 2-3Baseline MRI (n = 110), no baseline MRI (n = 59)Biopsy17.50.87 (0.80-0.94)74909741
Middleton et al[121], 2018MRI-PDFF [%]United StatesGrade 1-2 steatosis vs grade 3Baseline MRI (n = 110), no baseline MRI (n = 59)Biopsy23.30.79 (0.70-0.87)60908865
Middleton et al[121], 2018MRI-PDFF [%]United StatesDecrease in steatosis gradeBaseline MRI (n = 110), no baseline MRI (n = 59)Biopsy-110.76 (0.66-0.87)31907860
Middleton et al[121], 2018MRI-PDFF [%]United StatesIncrease in steatosis gradeBaseline MRI (n = 110), no baseline MRI (n = 59)Biopsy5.50.83 (0.73-0.92)40903392
Zhao et al[123], 2019MRI-PDFF [%]ChinaDetecting ≥ S1Total (n = 86), Obese/overweight (n = 65), healthy nonobese controls (n = 21)MRS5.10.991 (0.977-1.00)95100
Di Martino et al[122], 2016MRI-PDFF [%]United StatesPresence of steatosisNASH (n = 27), healthy controls (n = 27)Biopsy3.58988
Di Martino et al[122], 2016MRS [%]United StatesPresence of steatosisNASH (n = 27), healthy controls (n = 27)Biopsy692.695.7
AUROC: Area under the receiving operating characteristic; CAP: Controlled attenuation parameter; CI: Confidence interval; Dx: Diagnosis; LSM: Liver stiffness measurement; MRE: Magnetic resonance elastography; MRI: Magnetic resonance imaging; MRS: Magnetic spectroscopy; NAFL: Nonalcoholic fatty liver; NAFLD: Nonalcoholic fatty liver disease; NPV: Negative predictive value; PDFF: Proton density fat fraction; PPV: Positive predictive value; Sens: Sensitivity; Spec: Specificity; US: Ultrasound.
Table 6 Imaging biomarkers for the detection of nonalcoholic steatohepatitis and the detection of fibrosis in nonalcoholic fatty liver disease
Ref.
Marker
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
P value
Yang et al[106], 2022CAP [dB/m]ChinaPredicting NASH in children with obesityNAFLD (n = 61), Non-NAFLD (n = 59), NAFL (n = 44), NASH (n = 17)US> 2760.722 (0.602-0.843)70.672.80.0058
Chaidez et al[105], 2022LSMUnited StatesF0-F2 vs F3-F6 (Ishak)Total (n = 206)Biopsy0.73 (0.64-0.81)
Chaidez et al[105], 2022LSMUnited StatesF0-F2 vs F3-F6 (Ishak)NAFLD (n = 116)Biopsy0.77 (0.67-0.88)
Chaidez et al[105], 2022LSMUnited StatesF0-F2 vs F3-F6 (Ishak)Non-NAFLD (n = 90)Biopsy0.70 (0.56-0.83)
Yang et al[106], 2022LSM [kPa]ChinaPredicting NASH in children with obesityNAFLD (n = 61), Non-NAFLD (n = 59), NAFL (n = 44), NASH (n = 17)US> 5.150.725 (0.611-0.839)64.7650.0048
Xanthakos et al[132], 2014MRE [kPa]United StatesF2-F4 vs F0-F1Chronic liver disease (n = 35; 27 with NAFLD); F0-F1 (n = 27), F2-F4 (n = 8)Biopsy2.710.92 (0.79-1.00)88850.02
Schwimmer et al[133], 2017MRE [kPa]United StatesAny Fibrosis (F0 vs F1-4)F0 (n = 54), F1 (n = 24), F2 (n = 6), F3 (n = 5), F4 (n = 1)Biopsy≥ 2.770.7744.490.776.271
Schwimmer et al[133], 2017MRE [kPa]United StatesAny Fibrosis (F0 vs F1-4)F0 (n = 54), F1 (n = 24), F2 (n = 6), F3 (n = 5), F4 (n = 1)Biopsy≥ 2.690.7947.288.973.971.6
Schwimmer et al[133], 2017MRE [kPa]United StatesAny Fibrosis (F0 vs F1-4)F0 (n = 54), F1 (n = 24), F2 (n = 6), F3 (n = 5), F4 (n = 1)Biopsy≥ 2.780.77244.490.776.271
Schwimmer et al[133], 2017MRE [kPa]United StatesAdvanced Fibrosis (F0-2 vs F3-4)F0 (n = 54), F1 (n = 24), F2 (n = 6), F3 (n = 5), F4 (n = 1)Biopsy≥ 3.050.925 (0.539-0.989)5091.73096.2
Schwimmer et al[133], 2017MRE [kPa]United StatesAdvanced Fibrosis (F0-2 vs F3-4)F0 (n = 54), F1 (n = 24), F2 (n = 6), F3 (n = 5), F4 (n = 1)Biopsy≥ 3.030.879 (0.539-0.898)33.39428.695.2
Schwimmer et al[133], 2017MRE [kPa]United StatesAdvanced Fibrosis (F0-2 vs F3-4)F0 (n = 54), F1 (n = 24), F2 (n = 6), F3 (n = 5), F4 (n = 1)Biopsy≥ 3.330.894 (0.682-0.959)33.390.52095
Trout et al[134], 2018MRE [kPa]United StatesLudwig stage 0-1 vs ≥ stage 2 fibrosis in total cohortTotal (n = 86; 48 with NAFLD), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy2.270.70 (0.59-0.81)68.674.3
Trout et al[134], 2018MRE [kPa]United StatesLudwig stage 0-1 vs ≥ stage 2 fibrosis in total cohortTotal (n = 86; 48 with NAFLD), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy1.6735.391.4
Trout et al[134], 2018MRE [kPa]United StatesLudwig stage 0-2 from ≥ stage 3 fibrosisTotal (n = 86; 48 with NAFLD), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy6.550.90 (0.83-0.97)85.777.8
Trout et al[134], 2018MRE [kPa]United StatesLudwig stage 0-2 from ≥ stage 3 fibrosisTotal (n = 86; 48 with NAFLD), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy5.4164.393.1
Trout et al[134], 2018MRE [kPa]United StatesLudwig stage 0-1 vs ≥ stage 2 fibrosis in patients with steatosis (n = 41)Total (n = 86; 48 with NAFLD), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy0.53 (0.35-0.71)
Trout et al[134], 2018MRE [kPa]United StatesLudwig stage 0-1 vs ≥ stage 2 fibrosis in patients without steatosis (n = 45)Total (n = 86; 48 with NAFLD), Ludwig ≥ stage 2 (n = 51), steatosis (n = 44)Biopsy0.82 (0.67-0.96)
Alkhouri et al[108], 2012PNFIItaly≥ F2F0-F1 (n = 57), F2-F3 (n = 10)Biopsy8.20.747 (0.632-0.820)0.005
Nobili et al[107], 2008TE [kPa]Italy≥ F1F0 (n = 11), F1 (n = 27), F2 (n = 7), F3-4 (n = 5)Biopsy5.10.97 (0.90-0.99)97919791
Nobili et al[107], 2008TE [kPa]Italy≥ F2F0 (n = 11), F1 (n = 27), F2 (n = 7), F3-4 (n = 5)Biopsy7.40.99 (0.92-0.99)1009280100
Nobili et al[107], 2008TE [kPa]Italy≥ F3F0 (n = 11), F1 (n = 27), F2 (n = 7), F3-4 (n = 5)Biopsy10.21.00 (0.94-1.00)100100100100
Alkhouri et al[108], 2012TE [kPa]Italy≥ F2F0-F1 (n = 57), F2-F3 (n = 10)Biopsy8.61.00 (0.981-1.00)
AUROC: Area under the receiving operating characteristic; CAP: Controlled attenuation parameter; CI: Confidence interval; DX: Diagnosis; F: Fibrosis stage; LSM: Liver stiffness measurement; MRE: Magnetic resonance elastography; NAFL: Nonalcoholic fatty liver; NAFLD: Nonalcoholic fatty liver disease; NASH: Nonalcoholic steatohepatitis; NPV: Negative predictive value; PNFI: Proton density fat fraction index; PPV: Positive predictive value; Sens: Sensitivity; Spec: Specificity; TE: Transient elastography.
Table 7 Composite scores for the detection of nonalcoholic steatohepatitis
Ref.
Scores
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
Manco et al[57], 2022Ang-2 + CK18ItalyDiagnosing NASHNAFLD (n = 76), healthy controls (n = 28, by ultrasound)Biopsy71.410010080.4
Mosca et al[101], 2019APRIItalyDefinite NASH vs No/Borderline NASHNo/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 0.240.682658726962
Walenbergh et al[70], 2015CatD + CK18ItalySteatosis from NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.998
Walenbergh et al[70], 2015CatD + CK18ItalyBorderline NASH vs definite NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.858
Walenbergh et al[70], 2015CatD + CK18ItalySteatosis + Borderline NASH vs NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.892
Walenbergh et al[70], 2015CatD + CK18ItalySteatosis vs borderline NASH + NASHNASH (n = 26), borderline NASH (n = 51), steatosis (n = 19), obese (n = 96)Biopsy0.85
Mosca et al[101], 2019FIB-4ItalyDefinite NASH vs No/Borderline NASHNo/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 0.220.636948736558
Arsik et al[54], 2018Mean ALT over 96 wkUnited StatesNASHFibrosis (n = 128), NASH (n = 131)Biopsy81.8480.583
Arsik et al[54], 2018Mean ALT over 96 wkUnited StatesNASH + FibrosisFibrosis (n = 128), NASH (n = 131)Biopsy77.7871.880.8
Kwon et al[137], 2022P1NP/ALP ratioKoreaPresence of steatohepatitisNAFLD (n = 60)US1.460.788 (0.658-0.918)78.881.3
Kwon et al[137], 2022P1NP/ALP ratio × ALTKoreaPresence of steatohepatitisNAFLD (n = 60)US119.080.894 (0.812-0.977)82.692.9
Kwon et al[137], 2022P1NP/osteocalcin ratioKoreaPresence of steatohepatitisNAFLD (n = 60)US3.540.782 (0.647-0.918)80.976.9
Kwon et al[137], 2022P1NP/Osteocalcin ratio × ALTKoreaPresence of steatohepatitisNAFLD (n = 60)US305.380.939 (0.88-0.999)8392.3
Manco et al[95], 2007Risk ScoreItalyPredicting NAFLD Activity ScoreNAFLD (n = 72), F0 (n = 31), F1 (n = 41)Biopsy≤ 12.90.98592433
Manco et al[95], 2007Risk ScoreItalyPredicting NAFLD Activity ScoreNAFLD (n = 72), F0 (n = 31), F1 (n = 41)Biopsy≥ 13.581928292
ALP: Alkaline phosphatase; ALT: Alanine transaminase; Ang-2: Angiopoietin-2; APRI: AST to platelet ratio index; AUROC: Area under the receiving operating characteristic; CatD: Cathepsin D; CK18: Cytokeratin 18; CI: Confidence interval; Dx: Diagnosis; F: Fibrosis stage; FIB-4: Fibrosis-4; NAFLD: Nonalcoholic fatty liver disease; NASH: Nonalcoholic steatohepatitis; P1NP: Procollagen type 1 amino-terminal propeptide; PPV: Positive predictive value; NPV: Negative predictive value; Risk Score: 0.440 + (1.454 × ln leptin) + (4.617 × ln TNF-α); Sens: Sensitivity; Spec: Specificity; TNF-α: Tumor necrosis factor-alpha; US: Ultrasound.
Table 8 Composite scores for the detection of fibrosis in nonalcoholic fatty liver disease
Ref.
Scores
Country
Categories Tested
Sample size (n)
Dx
Cutoff
AUROC (95%CI)
Sens (%)
Spec (%)
PPV (%)
NPV (%)
P value
Mosca et al[101], 2019APRIItalyPresence of ≥ F2No/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 0.240.765980709243
Mosca et al[101], 2019APRIItalyPresence of F3No/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 0.260.853510049100100
Mosca et al[142], 2022APRIItaly> F1NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.61962.852
Mosca et al[142], 2022APRIItaly> F2NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.748678.1
Mansoor et al[140], 2015APRIUnited StatesPresence of F1-F4NAFLD (n = 92)Biopsy0.800 (0.695-0.904)
Mansoor et al[140], 2015APRIUnited StatesPresence of F2-F4NAFLD (n = 92)Biopsy0.666 (0.553-0.778)
Mansoor et al[140], 2015APRIUnited StatesPresence of F3-F4NAFLD (n = 92)Biopsy0.628 (0.478-0.778)
Mansoor et al[140], 2015AST/ALT ratioUnited StatesPresence of F1-F4NAFLD (n = 92)Biopsy0.572 (0.350, 0.793)
Mansoor et al[140], 2015AST/ALT ratioUnited StatesPresence of F2-F4NAFLD (n = 92)Biopsy0.585 (0.466-0.703)
Mansoor et al[140], 2015AST/ALT ratioUnited StatesPresence of F3 - F4NAFLD (n = 92)Biopsy0.441 (0.316-0.565)
Mandelia et al[84], 2016CK18 + WC per-centileItalyPresence of ≥ F1NAFLD (n = 201), F0 (n = 65), F1–F3 (n = 136)Biopsy≥ 350.84 (0.79-0.90)97387686
Mandelia et al[84], 2016CK18 + WC per-centileItalyPresence of ≥ F1NAFLD (n = 201), F0 (n = 65), F1–F3 (n = 136)Biopsy> 8259889151
Gawrieh et al[150], 2021ELFUnited StatesAny fibrosis (≥ F1)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.60 (0.50–0.70)0.11
Gawrieh et al[150], 2021ELFUnited StatesClinically significant (≥ F2)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.70 (0.60–0.80)< 0.001
Gawrieh et al[150], 2021ELFUnited StatesAdvanced fibrosis (≥ F3)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.79 (0.69–0.89)< 0.001
Chaidez et al[105], 2022FAST scoreUnited StatesSignificant liver disease (NAS ≥ 4 and Ishak ≥ 3) vs NAS < 4 / Ishak < 3)Chronic liver disease (n = 206; 116 with NAFLD)Biopsy≥ 0.670.75 (0.56-0.94)8962
Mosca et al[101], 2019FIB-4ItalyPresence of ≥ F2No/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 0.220.741264728839
Mosca et al[101], 2019FIB-4ItalyPresence of F3No/borderline NASH (n = 115), definite NASH (n = 89)Biopsy> 0.240.76878671999
Arsik et al[54], 2018Mean ALT over 96 weeksUnited StatesFibrosisFibrosis (n = 128), NASH (n = 131)Biopsy58.5656.564.6
Mosca et al[142], 2022FIB-4Italy> F1NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.5456252
Mosca et al[142], 2022FIB-4Italy> F2NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.588
Mansoor et al[140], 2015FIB-4United StatesPresence of F1-F4NAFLD (n = 92)Biopsy0.547 (0.375-0.719)
Mansoor et al[140], 2015FIB-4United StatesPresence of F2-F4NAFLD (n = 92)Biopsy0.686 (0.576-0.797)
Mansoor et al[140], 2015FIB-4United StatesPresence of F3-F4NAFLD (n = 92)Biopsy0.367 (0.231-0.503)
Gawrieh et al[150], 2021HAUnited StatesAny fibrosis (≥ F1)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.57 (0.47–0.67)0.32
Gawrieh et al[150], 2021HAUnited StatesSignificant fibrosis (≥ F2)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.64 (0.54–0.74)0.002
Gawrieh et al[150], 2021HAUnited StatesAdvanced fibrosis (≥ F3)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.77 (0.66–0.88)0.001
Mosca et al[142], 2022HepametItaly≥ F2NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.7388.876.6
Mosca et al[142], 2022HepametItaly> F1NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.77863.261.3
Mansoor et al[140], 2015NFSUnited StatesPresence of F1-F4NAFLD (n = 92)Biopsy0.470 (0.259-0.681)
Mansoor et al[140], 2015NFSUnited StatesPresence of F2-F4NAFLD (n = 92)Biopsy0.554 (0.435-0.673)
Mansoor et al[140], 2015NFSUnited StatesPresence of F3-F4NAFLD (n = 92)Biopsy0.521 (0.385-0.657)
Mosca et al[142], 2022NFSItaly> F1NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.5376252
Mosca et al[142], 2022NFSItaly> F2NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.6
Gawrieh et al[150], 2021PIIINPUnited StatesAny fibrosis (≥ F1)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.55 (0.45–0.65)0.18
Gawrieh et al[150], 2021PIIINPUnited StatesClinically significant (≥ F2)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.66 (0.57–0.75)0.002
Gawrieh et al[150], 2021PIIINPUnited StatesAdvanced fibrosis (≥ F3)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.65 (0.53–0.76)0.06
Mosca et al[142], 2022PNFIItaly> F1NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.8190.375.4
Nobili et al[148], 2009PNFIItaly≥ F1NAFLD (n = 203), Fibrosis (n = 141), no fibrosis (n = 62), stage 1 fibrosis (n = 115), stage 2 fibrosis (n = 9), stage 3 fibrosis (n = 17)Biopsy≥ 90.85 (0.80-0.90)98.5
Mosca et al[142], 2022PNFIItaly> F2NAFLD (n = 286), F0 (n = 105), F1 (n = 140), F2 (n = 31), F3 (n = 2)Biopsy0.8497.572.6
Gawrieh et al[150], 2021TIMP-1United StatesAny fibrosis (≥ F1)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.63 (0.54–0.72)0.02
Gawrieh et al[150], 2021TIMP-1United StatesClinically significant (≥ F2)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.63 (0.53–0.72)0.01
Gawrieh et al[150], 2021TIMP-1United StatesAdvanced fibrosis (≥ F3)NAFLD (n = 173), borderline/suspicious NASH (n = 73), definite NASH (n = 71)Biopsy0.76 (0.64–0.88)< 0.001
ALT: Alanine aminotransferase; AUROC: Area under the receiver operating characteristic curve; AST: Aspartate aminotransferase; APRI: AST to platelet ratio index; CI: Confidence interval; Dx: Diagnosis; ELF: Enhanced liver fibrosis; F: Fibrosis stage; FAST: FibroScan-aspartate aminotransferase; FIB-4: Fibrosis-4, Fibrosis-4; HA: Hyaluronic acid; NAS: NAFLD activity score; NAFLD: Nonalcoholic fatty liver disease; NFS: NAFLD Fibrosis Score; NPV: Negative predictive value; PIIINP: Amino-terminal propeptide of type III procollagen; PNFI: Pediatric NAFLD Fibrosis Index; PPV: Positive predictive value; Sens: Sensitivity; Spec: Specificity; TIMP-1: Tissue inhibitor of metalloproteinase-1; WC, waist circumference.