Gastroenterologìa (Nov 2017)
Development of a non-invasive model to improve the accuracy of determining liver fibrosis stage in nonalcoholic fatty liver disease
Abstract
Background. The differentiation of mild (F1-F2) and advanced fibrosis (F3-F4), as well as the exclusion of fibrosis in patients with nonalcoholic fatty liver disease (NAFLD), are extremely important for prediction of the disease course. Integrative analyses of serum markers have been proposed as promising alternatives to biopsy method. Our study was targeted to develop a new model for determining the stage of fibrosis based on a more efficient combination of serological markers and to compare it with well-established algorithms. Materials and methods. Sixty patients with biopsy-proven NAFLD, including 26 (43 %) men and 34 (57 %) women, with average age of 37.10 ± 12.4 and 44.30 ± 7.25 years, respectively, were recruited for the study. Particularly, advanced fibrosis was diagnosed in 8 patients, 28 had mild fibrosis and 24 didn’t have any fibrosis according to morphological study. The following fibrosis markers were calculated: aspartate aminotransferase and alanine aminotransferase ratio (AAR), aspartate aminotransferase to platelet ratio index (APRI), fibrosis index based on the 4 factor (FIB-4). Among many variables, hyaluronic acid, α2-macroglobulin, apolipoprotein A1, fibronectin, and haptoglobin were included in comprehensive study. Integrative model have been built up to determine the stage of fibrosis. The models were compared with the area under the receiver operating characteristic (AUROC) curves. Results. The ROC analysis showed that the FIB-4 demonstrated the largest AUROC, for the F2 — 0.72, F3 — 0.8, F4 — 0.82, respectively. Obtained results of the APRI were significantly higher for mild and advanced fibrosis (F2 — 0.74, F3 — 0.82). The AAR values were reliable only for liver cirrhosis (AUROC 0.89). A strong direct correlation was determined between the stage of fibrosis and the level of hyaluronic acid, α2-macroglobulin and fibronectin (r = 0.72, 0.93 and 0.71, p < 0.05, respectively). Whereas, we observed a moderate negative linear correlation between fibrosis stage and the indices of both apolipoprotein A1 and haptoglobin (r = –0.61; r = –0.35, respectively, p < 0.05). The positive correlation was determined between activity of the inflammatory process and the content of hyaluronic acid, α2-macroglobulin and fibronectin (r = 0.54, 0.67 and 0.55 at p < 0.05), while the reverse moderate relation observed for apolipoprotein A1 and hapthoglobin (r = –0.56 and –0.33, p < 0.05). Conclusions. The analysis of obtained results showed that α2-macroglobulin, apolipoprotein A1, hyaluronic acid, and fibronectin had the greatest diagnostic validity among non-invasive markers of fibrosis. Every of them get the AUROC level higher than 0.75 for minimal fibrosis and, moreover, for moderate, significant fibrosis and cirrhosis they had an area more than 0.9.
Keywords
