Prediction of significant fibrosis in chronic hepatitis C patients with normal ALT

Jae Jun Park, Jun Yong Park, Do Young Kim, Young Nyun Park, Sang Hoon Ahn, Chae Yoon Chon, Kwang Hyub Han

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)


Background/Aims: Prediction of significant fibrosis (F≥2) using non-invasive methods for chronic hepatitis C (CHC) patients with persistently normal alanine aminotransferase (PNALT) levels remains a challenging problem. We aimed to develop a novel non-invasive model for predicting the presence of significant fibrosis in CHC patients with PNALT. Methodology: We prospectively enrolled 40 treatment-naive CHC patients with PNALT who underwent liver biopsy and liver stiffness measurements (LSM). Age-platelet index (API), aspartate aminotransferase to platelet ratio index (APRI), LSM and LSM to platelet ratio index (LPRI) were compared with liver histology results. Results: Significant fibrosis was diagnosed in 17 patients (42.5%). The diagnostic accuracy of LPRI was the highest for the prediction of significant fibrosis (AUROC=0.859) when compared to that of APRI (0.770), LSM (0.769) and API (0.703). Using a cutoff value of LPRI ≥37, the significant fibrosis could be correctly identified with high accuracy (100% PPV) in 6 (15.0%) patients. While, using an LPRI cutoff value <20, it could be excluded with 94.1% NPV in 17 (42.5%) patients. Consequently, 57.5% of the CHC patients with PNALT levels could avoid liver biopsy. Conclusions: The LPRI was useful in predicting significant fibrosis. Screening CHC patients PNALT using LPRI has the potential to reduce the number of liver biopsies and might help designing appropriate management plans.

Original languageEnglish
Pages (from-to)1321-1327
Number of pages7
Issue number109
Publication statusPublished - 2011 Jul

All Science Journal Classification (ASJC) codes

  • Hepatology
  • Gastroenterology


Dive into the research topics of 'Prediction of significant fibrosis in chronic hepatitis C patients with normal ALT'. Together they form a unique fingerprint.

Cite this