SEARCH

SEARCH BY CITATION

To the Editor:

We read with interest the article by Kotronen et al.,1 reporting on liver fat score and liver fat equation—new noninvasive, easy-to-calculate indexes that estimate the presence and severity of hepatic fat accumulation. We assessed the ability of these two indexes and of plasma cytokeratin-18 fragments (CK-18) to predict the presence of nonalcoholic fatty liver disease (NAFLD) and of nonalcoholic steatohepatitis (NASH),1, 2 respectively, and their relations to validated predictors of incident cardiovascular disease and diabetes.3, 4

To this purpose, 125 subjects (40 nondiabetic patients with biopsy-proven NAFLD and 85 healthy controls) underwent an oral fat tolerance test,5 with measurement of postprandial plasma lipid responses, and a standard oral glucose tolerance test (OGTT), whose results were elaborated by Minimal Model analysis to assess whole-body, hepatic, and muscle insulin sensitivity and indexes of pancreatic β-cell function (namely, CP-genic index [CGI] and Adaptation Index [AI]), as previously described.5–7 Finally, circulating markers of inflammation (C-reactive protein), endothelial dysfunction (E-selectin and intercellular adhesion molecule-1 [ICAM-1]) and oxidative stress (nitrotyrosine and oxidized low-density lipoproteins) were measured.

Results are shown in Table 1. NASH group showed higher postprandial lipemia and oxidative stress than either steatosis or controls. Patients with NASH had also more severe whole-body insulin resistance, hepatic insulin resistance, and pancreatic β-cell dysfunction and higher plasma C-reactive protein, E-selectin, ICAM-1, and nitrotyrosine levels than steatosis and control groups.

Table 1. Baseline Characteristics of Patients with NAFLD and Controls
CharacteristicControls (n = 85)Steatosis (n = 17)NASH (n = 23)P Value
  • Data are presented as mean ± standard error of the mean. Differences between groups were analyzed by analysis of variance for normal variables; otherwise, the Mann-Whitney test was used for nonparametric variables. Normality was evaluated by Shapiro-Wilk test. Fisher or chi-squared test were used to compare categorical variables, as appropriate. Differences were considered statistically significant at P < 0.05.

  • BP, blood pressure; FFA, free fatty acids; HDL-C, high-density lipoprotein cholesterol; IAUC, incremental area under the curve; ICAM-1, intercellular adhesion molecule-1; LDL-C, low-density lipoprotein cholesterol; met sy, metabolic syndrome; OGIS, oral glucse insulin sensitivity index; Tg, triglyceride; total C, total cholesterol.

  • CGI (Cpgenic index) = ΔC– peptide30′/Δglucose30′ during the OGTT; Adaptation Index was computed by multiplying CGI × OGIS. Hepatic extraction is the percent secreted insulin extracted by the liver.

  • This is a corrected version of Table 1 first published online on 09 October 2009 — the NASH value for CK-18 has been changed to 253.

  • *

    P < 0.05 versus controls;

  • P < 0.01 versus controls.

Age (years)49 ± 448 ± 346 ± 40.712
Sex (%males)6770710.549
BMI (kg/m2)25.1 ± 1.724.9 ± 1.625.3 ± 1.50.656
Systolic BP (mmHg)129 ± 6130 ± 7131 ± 90.978
Diastolic BP (mmHg)81 ± 484 ± 688 ± 6*0.034
Waist (cm)88 ± 290 ± 491 ± 40.478
Tg (mg/dL)89 ± 2198 ± 35111 ± 39*0.112
LDL-C (mg/dL)99 ± 7124 ± 8150 ± 7*0.056
HDL-C (mg/dL)59 ± 257 ± 249 ± 2*0.013
Total C (mg/dL)156 ± 9205 ± 9241 ± 100.058
Glucose (mg/dL)92 ± 396 ± 397 ± 40.541
Insulin (μU/mL)4.2 ± 1.614.3 ± 2.922.7 ± 5.60.009
AST (U/L)13 ± 356 ± 342 ± 30.123
ALT (U/L)16 ± 5127 ± 8116 ± 80.245
ICAM-1 (mg/mL)190.5 ± 7.1220.4 ± 7.2*264.4 ± 8.70.031
E-selectin (mg/mL)18.9 ± 2.023.1 ± 2.338.2 ± 2.50.001
C-reactive protein1.1 ± 0.71.6 ± 1.0*2.4 ± 1.10.028
CK-18 (IU/L) (IU/L)fragments98 ± 6136 ± 11253 ± 120.030
Nitrotyrosine3.1 ± 3.79.9 ± 4.1*27.4 ± 7.30.002
Met sy (%)2136470.052
Histological steatosis (% hepatocytes)16±733±90.032
Necroinflammatory grade2.1±0.3
Fibrosis stage1.9±0.4
Liver fat score2.428 ± 0.1121.563 ± 0.2905.114 ± 0.9310.039
Liver fat (%)1.5 ± 113 ± 330 ± 60.025
OGTT-derived indexes of glucose homeostasis
OGIS (mL · minute−1 · m−2)459.1 ± 16.5400 ± 11.9*369 ± 10.20.032
Hepatic extraction (%)78 ± 481 ± 570 ± 4*0.030
Cpgenic Index (CGI) (ngC-pep · ginline image)614 ± 36.1540 ± 29.5402 ± 19.30.009
Adaptation Index (ngC-pep · ginline image · mL−1 · m−2) (ngCpep · ginline image · mL−1 · m−2)281826 ± 15391216703 ± 12678148401 ± 106810.007
Muscle insulin sensitivity0.032 ± 0.0090.016 ± 0.008*0.015 ± 0.006*0.267
Hepatic insulin resistance738 ± 961109 ± 1121812 ± 1830.011
Oral Fat Load
IAUC Tg (mg/dL × hour)124 ± 28344 ± 49398 ± 560.034
Fasting FFA (μmol/L)0.47 ± 0.131.09 ± 0.221.96 ± 0.290.112
IAUC FFA (μmol/L × hour)0.8 ± 0.32.5 ± 0.54.9 ± 0.90.009
Fasting oxidizedLDL (uA 234 nm/uA 200 nm × 100)6.42 ± 1.636.44 ± 1.237.94 ± 2.590.356
IAUC oxidized LDL (uA 234 nm/uA 200 nm × 100 × hour)1.29 ± 0.514.29 ± 1.258.02 ± 1.930.001

Liver fat equation correlated with the degree of histological steatosis in both NASH and steatosis groups (in both groups: rs > 0.66, P < 0.003). The area under the receiver operating characteristic curve (AUROC) of liver fat score for predicting NAFLD was 0.86 (95% confidence interval [CI]: 0.82–0.91). A cutoff of −0.640 individuated NAFLD with a sensitivity, specificity, and positive and negative likelihood ratio of 0.93, 0.80, 4.63, and 0.09, respectively.

The AUROC of CK-18 for NASH was 0.83 (95% CI: 0.80–0.90). A cutoff of 246 IU/L for CK-18 individuated NASH with a sensitivity, specificity, and positive and negative likelihood ratio of 0.78, 0.88, 6.65, and 0.25, respectively.

On multiple regression analysis, liver fat equation independently correlated with hepatic insulin resistance (β = 0.52; 95% CI: 0.48–0.56, P = 0.005) and with indexes of pancreatic β-cell function (for CGI: β = −0.43; 95% CI: 0.48–0.56, P = 0.01; for AI: β = −0.46; 95% CI: 0.42–0.51, P = 0.009) in the whole sample. Liver fat equation also independently predicted plasma C-reactive protein (β = 0.40; 95% CI: 0.37–0.44, P = 0.02), nitrotyrosine (β = 0.41; 95% CI: 0.38–0.46, P = 0.02), E-selectin (β = 0.49; 95% CI: 0.45–0.54, P = 0.006), and postprandial triglyceride (β = 0.42; 95% CI: 0.39–0.46, P = 0.02) and oxidized low-density lipoprotein (β = 0.40; 95% CI: 0.38–0.45, P = 0.03) responses to the fat load.

Histological steatosis, inflammatory grade, and fibrosis stage were not independently related to the abovementioned parameters.

In conclusion, liver fat equation and CK-18 accurately individuated the presence and severity of liver fat infiltration and the presence of NASH in our cohort of nondiabetic subjects. Most importantly, liver fat equation was tightly related to validated predictors of increased cardiometabolic risk in both healthy and NAFLD subjects. The clinical significance of liver fat equation may thus go beyond hepatic fat content estimation and this easy-to-calculate index may aid in predicting individual cardio-metabolic risk of patients with NAFLD. Our cross-sectional findings warrant prospective confirmation in independent large cohorts.

References

  1. Top of page
  • 1
    Kotronen A, Peltonen M, Hakkarainen A, Sevastianova K, Bergholm R, Johansson LM, Lundbom N, Rissanen A, Ridderstråle M, Groop L, Orho-Melander M, Yki-Järvinen H. Prediction of non-alcoholic fatty liver disease and liver fat using metabolic and genetic factors. Gastroenterology 2009; 137: 865872.
  • 2
    Feldstein AE, Wieckowska A, Lopez AR, Liu YC, Zein NN, McCullough AJ. Cytokeratin-18 fragment levels as noninvasive biomarker for nonalcoholic steatohepatitis: A multicenter validation study. HEPATOLOGY 2009;doi:10.1002/hep.23050.
  • 3
    Abdul-Ghani MA, Williams K, DeFronzo RA, Stern M. What is the best predictor of future type 2 diabetes? Diabetes Care 2007; 30: 15441548.
  • 4
    Adiels M, Olofsson SO, Taskinen MR, Boren J. Diabetic dyslipidaemia. Curr Opin Lipidol 2006; 17: 238246.
  • 5
    Musso G, Gambino R, Pacini G, Pagano G, Durazzo M, Cassader M. Transcription factor 7-like 2 polymorphism modulates glucose and lipid homeostasis, adipokine profile, and hepatocyte apoptosis in NASH. HEPATOLOGY 2009; 49: 426435.
  • 6
    Tura A, Kautzky-Willer A, Pacini G. Insulinogenic indices from insulin and C-peptide: comparison of beta-cell function from OGTT and IVGTT. Diabetes Res Clin Pract 2006; 72: 298301.
  • 7
    Ahrén B, Pacini G. Importance of quantifying insulin secretion in relation to insulin sensitivity to accurately assess beta cell function in clinical studies. Eur J Endocrinol 2004; 150: 97104.

Giovanni Musso M.D.*, Roberto Gambino Ph.D.†, Marilena Durazzo Ph.D.†, Maurizio Cassader Ph.D.†, * Gradenigo Hospital, Turin, Italy, † Department of Internal Medicine, University of Turin, Italy.