Introduction: Liver stiffness measurement (LSM) is used for the assessment of liver fibrosis. However, there is limited data in Indian patients.
Aims and objective: The aim of this study was to find the correlation of LSM, aspartate transaminase to platelet ratio index (APRI) with fibrosis as assessed by liver biopsy (LB), and predictors of discordance between LB and LSM.
Methods: One hundred and eighty-five consecutive patients who underwent liver biopsy and transient elastography (TE) were enrolled. Fibrosis was graded by two independent pathologists using the METAVIR classification. Area under receiver operating curves (AUROC) was used to evaluate the accuracy of transient elastography and APRI in diagnosing significant fibrosis (F>2) and cirrhosis (F4).
Results: Predominant etiologies were hepatitis B (46 %) and hepatitis C (26 %). LSM was unsuccessful in ten patients (5 %) because of small intercostal space (n = 3) and obesity (n = 7). Fibrosis is significantly correlated with LSM (r = 0.901, p = 0.001) and APRI (r = 0.736, p = 0.001). There was a significant difference in median LSM value in patients with no fibrosis (F0) in comparison to patients having mild fibrosis [mild portal fibrosis (F1) + fibrosis with few septa (F2)] (4.5 vs. 7.5 kPa, p = 0.001) and advanced fibrosis [bridging fibrosis that is spreading and connecting to other areas that contain fibrosis (F3) + cirrhosis or advanced scarring of the liver (F4)] (4.5 vs. 19.4 kPa, p = 0.001). Similarly, there was a significant difference in mean APRI value in patients with F0 in comparison to patients having mild fibrosis (F1 + F2) (0.55 ± 0.31 vs. 1.09 ± 0.81, p = 0.001) and advanced fibrosis (F3 + F4) (2.3 ± 1.3, p = 0.001). AUROC for diagnosis of significant fibrosis was 0.98 (95 % confidence interval (CI) 0.963-0.999) for TE and 0.865 (95 % CI 0.810-0.920) for APRI. Optimal TE value was 10.0 kPa for diagnosis of significant fibrosis and 14.7 kPa for cirrhosis with specificity and sensitivity of 89 %, 98 % and 96 %, and 97 %, respectively. On multivariate analysis, total bilirubin and histological activity index (HAI) were identified as an independent predictor of TE inaccuracy.
Conclusion: LSM is a reliable predictor of hepatic fibrosis in Indian patients. LSM is superior to APRI for noninvasive diagnosis of hepatic fibrosis and cirrhosis, and high bilirubin (10.5 mg/dL) and Ishak HAI grade (>11) were independent predictors of discordance between LB and LSM.