ABSTRACT
In patients with chronic hepatitis C, the precise stage of hepatic fibrosis is the most important predictor of disease progression and it determines the need for antiviral therapy. Although liver biopsy is acknowledged as the gold standard for evaluating fibrosis, it is occasionally prone to sampling error and complications. We aimed to correlate an index of biochemical markers with histological features of fibrosis to predict hepatic fibrosis in patients with chronic hepatitis C virus, patients with combined hepatitis C virus and non-alcoholic steatohepatitis and those with non-alcoholic steatohepatitis, aiming to reduce the use of the liver biopsy. Out of those attending our out patient clinic for clinical, haematological, biochemical, virological, histological and ultrasonographic assessment prior to interferon therapy for hepatitis C virus, we enrolled 41 patients and grouped them according to histopathological examination of their liver biopsies into: Group I: 21 chronic hepatitis C virus patients as defined by positive 3rd generation ELISA; Group II: 20 patients with combined hepatitis C virus and NASH. We added a third group [Group III] of 15 patients having non alcoholic steatohepatitis as defined clinically, biochemically and through diagnostic percutanous liver biopsy. There were 33 male 23 female patients; 35 [62.5%] of them were from rural areas and 21 [37.5%] were from urban areas; the mean ages were 40.5 +/- 9, 46.6 +/- 7.7 and 42.13 +/- 11.06 in Group I, II and III respectively. Twenty apparently healthy individuals served as the control group. All the patients and the control group were submitted to full clinical history and examination, abdominal ultrasonography, CBC, liver biochemical profile and fibrosis biomarkers [apolipoprotein A1, haptoglobin, alpha2 marcoglobulin, GGT]. Liver biopsy was done for suitable patients after taking a consent and the results of fibrosis seromarkers were compared with the results of liver biopsy using the Metavir scoring system, We also estimated patients' body mass index, fasting and post prandial blood glucose. We excluded patients with other causes of chronic liver disease and co-morbidities that could confound the results of the non-invasive markers adopted, including schistosomiasis which was excluded by serological test. 43% of Group I and 40% of Group II had advanced fibrosis. None of Group III had advanced fibrosis; mild fibrosis was detected in 80% of them. gamma-GT was found positively correlated to the degree of hepatic fibrosis in Groups I, II and III [r = 0.667, 0.656 and 0.121, respectively] with P values of 0.001, 0.002, 0.668, respectively. alpha2 macroglobulin was found to be a reliable predictor of fibrosis [r = 0.30, P = 0.02] with ROC curve [area under the curve = 0.70] best cutoff value 2.55 g/L with sensitivity of 0.80 and specificity of 0.50. The results of haptoglobin were negatively related to the degree of hepatic fibrosis in Group I and II with ROC curve area under the curve of 0.33 and P value of 0.04. Significant direct correlation was seen in Group III [r = 0.55, P = 0.03], so by regression analysis, haptoglobin can be used as a good predictor for fibrosis in Group III [r = 0.54, P=0.04]. Apolipoprotein A1 has negative correlation to the stage of fibrosis in Groups I and II although the results were statistically insignificant. APRI index was found significantly directly correlated to the fibrosis stage and the grade of inflammation of all studied groups [r= 0.57, P< 0.01 and r = 0.36, P< 0.01, respectively] with a best cutoff value of 0.62, with sensitivity of 0.86 and specificity of 0.57. In patients with advanced fibrosis the best cutoff value was found to be 0.72 with sensitivity of 0.94 and specificity of 0.67. Mordified APRI test showed AUC of 0.79 [P<0.01] with a best cutoff value of 0.067 at which sensitivity and specificity were 0.82 and 0.61, respectively. Alpha macroglobulin, haptoglobin, apolipoprotein A1, APRI index and a modified APRI index, were found be significant predictors of hepatic fibrosis and were reprocessed by stepwise logistic regression