Advances in digital quantification technique enhance discrimination between mild and advanced liver fibrosis in chronic hepatitis C

Liver Int. 2005 Dec;25(6):1142-9. doi: 10.1111/j.1478-3231.2005.01155.x.

Abstract

Background and aims: The necessity of liver biopsy for staging fibrosis and its quantification in patients with chronic hepatitis C (CHC) remains controversial. Semiquantitative scoring of fibrosis is considered more subjective and less objective than digital quantification by image analysis. However, measurement of fibrosis using digital image analysis is thought to be less reliable in determining early stage fibrosis as compared with advanced fibrosis or cirrhosis. Our aims were to correlate all Ishak stages of fibrosis (0-6) with fibrosis percentage (%) using computerized digital image analysis, and thereby seek to improve discrimination between varying levels of liver fibrosis.

Methods: Fibrosis % data were obtained by image analysis on 164 trichrome-stained liver biopsies from untreated patients with CHC, representing all Ishak stages of fibrosis.

Results: Digital analysis of fibrosis % was highly correlated with Ishak scores of fibrosis (Kendall's tau-beta=0.86, P<0.001). Receiver-operator characteristic curves showed reliable discriminative capability of our digital image measurement of fibrosis when compared with semiquantitative assessments of fibrosis. Excellent interobserver reliability was found.

Conclusions: Recent advances in digital quantification of fibrosis have resulted in improved discrimination between the varying stages of liver fibrosis, including mild fibrosis. This method is reproducible, can detect early as well as advanced fibrosis or cirrhosis, may prove to be the best assessment of mild fibrosis, and may be more precise than semiquantitative estimation of changes for monitoring fibrosis progression or regression during clinical therapeutic trials.

MeSH terms

  • Azo Compounds
  • Biopsy, Needle*
  • Coloring Agents
  • Eosine Yellowish-(YS)
  • Fibrosis
  • Hepatitis C, Chronic / complications*
  • Humans
  • Image Processing, Computer-Assisted*
  • Liver / pathology*
  • Liver Cirrhosis / pathology*
  • Methyl Green
  • Microscopy
  • Observer Variation
  • ROC Curve

Substances

  • Azo Compounds
  • Coloring Agents
  • trichrome stain
  • Methyl Green
  • Eosine Yellowish-(YS)