Novel non-invasive biological predictive index for liver fibrosis in hepatitis C virus genotype 4 patients

Abstract

Aim: To investigate the diagnostic ability of a non-invasive biological marker to predict liver fibrosis in hepatitis C genotype 4 patients with high accuracy.

Methods: A cohort of 332 patients infected with hepatitis C genotype 4 was included in this cross-sectional study. Fasting plasma glucose, insulin, C-peptide, and angiotensin-converting enzyme serum levels were measured. Insulin resistance was mathematically calculated using the homeostasis model of insulin resistance (HOMA-IR).

Results: Fibrosis stages were distributed based on Metavir score as follows: F0 = 43, F1 = 136, F2 = 64, F3 = 45 and F4 = 44. Statistical analysis relied upon reclassification of fibrosis stages into mild fibrosis (F0-F) = 179, moderate fibrosis (F2) = 64, and advanced fibrosis (F3-F4) = 89. Univariate analysis indicated that age, log aspartate amino transaminase, log HOMA-IR and log platelet count were independent predictors of liver fibrosis stage (P < 0.0001). A stepwise multivariate discriminant functional analysis was used to drive a discriminative model for liver fibrosis. Our index used cut-off values of ≥ 0.86 and ≤ -0.31 to diagnose advanced and mild fibrosis, respectively, with receiving operating characteristics of 0.91 and 0.88, respectively. The sensitivity, specificity, positive predictive value, negative predictive value and positive likelihood ratio were: 73%, 91%, 75%, 90% and 8.0 respectively for advanced fibrosis, and 67%, 88%, 84%, 70% and 4.9, respectively, for mild fibrosis.

Conclusion: Our predictive model is easily available and reproducible, and predicted liver fibrosis with acceptable accuracy.

Keywords: Age; Aspartate amino transaminase; Insulin resistance; Liver fibrosis; Platelets.

Figure 1

Receiving operating characteristic curve for discriminating mild fibrosis. At cut off value: -0.31 or more negative: AUC 0.88, 95%CI: 0.84 -0.91, sensitivity 67.2%, specificity 6.3%, PPV 83.6%, NPV 69.5%, PLR 4.9 and NLR 0.38. AUC: Area under the curve; PPV: Positive predictive value; NPV: Negative predictive value; PLR: Positive likelihood ratio; NLR: Negative likelihood ratio; ROC: Receiving operating characteristic.

Figure 2

Receiving operating characteristic curve for discriminating advanced fibrosis. At cut off value > 0.86: AUC 0.91, 95%CI: 0.88-0.94, sensitivity 73%, specificity 90.9%, PPV 74.4%, NPV 90.1%, PLR 8.0 and NLR 0.3. PPV: Positive predictive value; NPV: Negative predictive value; PLR: Positive likelihood ratio; NLR: Negative likelihood ratio; ROC: Receiving operating characteristic.

Figure 3

Receiving operating characteristic curve for discriminating moderate fibrosis. At cut off value > -0.31 up to +0.86: AUC 0.64, 95%CI: 0.61-0.74, sensitivity 53.1%, specificity 74.1%, PPV 33%, NPV 86.8%, PLR 2.0 and NLR 0.63. AUC: Area under the curve; PPV: Positive predictive value; NPV: Negative predictive value; PLR: Positive likelihood ratio; NLR: Negative likelihood ratio; ROC: Receiving operating characteristic.