Evaluation of diagnostic accuracy of APRI for prediction of fibrosis in hepatitis C patients.

BACKGROUND Several non-invasive markers are being used to assess the structural liver damage in patients with chronic hepatitis C (CHC). We evaluated Aspartate aminotransferase (AST) to platelet ratio index (APRI) in comparison with Metavir scoring for assessing the severity of hepatic fibrosis in the CHC patients in district Rawalpindi. METHODS One hundred twenty CHC patients, naive for HCV treatment, underwent liver biopsy in tertiary care hospitals of district Rawalpindi, participated in the study. Liver biopsies were reviewed by Metavir scoring system. Serum AST was analyzed by IFCC method. Platelets were measured on a haematology Analyzer. Patients with mild fibrosis (F0, F1) were differentiated from significant fibrosis (F2, F3, F4) and those with mild/moderate fibrosis (F0, F1, F2) from advanced fibrosis (F3, F4) based on APRI score as compared to liver biopsy. RESULTS Liver biopsies examination revealed that out of 120 patients 10 (8.3%) had no fibrosis (F0), 46 (38%) portal fibrosis (F1), 34 (28%) septal fibrosis (F2), 21 (18%) bridging fibrosis (F3) and 9 (8%) cirrhosis (F4). APRI correctly classified 58 (48%) patients of significant fibrosis with AUC = 0.82 (95% CI, 0.73-0.88) at cut-off 0.5 and 1.5 with negative predictive value (NPV), Positive predictive value (PPV), sensitivity and specificity of 78%, 72%, 66%, 83% and 58%, 90%, 41% and 90% respectively. Eighty-seven (66%) CHC patients were correctly classified for advanced fibrosis with AUC = 0.87 (95% CI 0.79-0.94) at cutoffs 0.90 and 1.75 with a 95%NPV at 0.90 and 78% PPV at 1.75. CONCLUSION APRI could correctly identify significant fibrosis in 48% and advanced fibrosis in 66% cases with acceptable degree of diagnostic accuracy in CHC patients in our clinical practice.