RESUMO
BACKGROUND: Non invasive approaches will likely be increasing utilized to assess liver fibrosis. This work provides a new non invasive index to predict liver fibrosis induced in mice. METHODS: Fibrosis was generated by thioacetamide (TAA), chronic intake of ethanol, or infection with S. mansoni in 240 mice. Both progression and regression of fibrosis (after treatment with silymarin and/or praziquantel) were monitored. The following methods were employed: (i) The METAVIR system was utilized to grade and stage liver inflammation and fibosis; (ii) Determination of hepatic hydroxyproline and collagen; and (iii) Derivation of a new hepatic fibrosis index from the induced changes, and its prospective validation in a group of 70 mice. RESULTS: The index is composed of 4 serum variable including total proteins, gamma-GT, bilirubin and reduced glutathione (GSH), measured in diseased, treated and normal mice. These parameters were highly correlated with both the histological stage and the grade. They were combined in a logarithmic formula, which non-invasively scores the severity of liver fibrosis through a range (0 to 2), starting with healthy liver (corresponding to stage 0) to advanced fibrosis (corresponding stage 3).Receiver operating characteristic curves (ROC) for the accuracy of the index to predict the histological stages demonstrated that the areas under the curve (AUC) were 0.954, 0.979 and 0.99 for index values corresponding to histological stages 1, 2 and 3, respectively. Also, the index was correlated with stage and grade, (0.947 and 0.859, respectively). The cut off values that cover the range between stages 0-1, 1-2 and 2-3 are 0.4, 1.12 and 1.79, respectively. The results in the validation group confirmed the accuracy of the test. The AUROC was 0.869 and there was good correlation with the stage of fibrosis and grade of inflammation. CONCLUSION: The index fulfils the basic criteria of non-invasive marker of liver fibrosis since it is liver-specific, easy to implement, reliable, and inexpensive. It proved to be accurate in discriminating precirrhotic stages.
