Comparison of nine blood tests and transient elastography for liver fibrosis in chronic hepatitis C: The ANRS HCEP-23 study

Journal of Hepatology
Volume 56, Issue 1 , Pages 55-62, January 2012

Jean-Pierre Zarski, Nathalie Sturm, Jérôme Guechot, Adeline Paris, Elie-Serge Zafrani, Tarik Asselah, Renée-Claude Boisson, Jean-Luc Bosson, Dominique Guyader, Jean-Charles Renversez, Jean-Pierre Bronowicki, Marie-Christine Gelineau, Albert Tran, Candice Trocme, Victor De Ledinghen, Elisabeth Lasnier, Armelle Poujol-Robert, Frédéric Ziegler, Marc Bourliere, Hélène Voitot, Dominique Larrey, Maria Alessandra Rosenthal-Allieri, Isabelle Fouchard Hubert, François Bailly, Michel Vaubourdolle, The ANRS HCEP 23 Fibrostar Group

Received 21 October 2010; received in revised form 13 April 2011; accepted 3 May 2011. published online 20 July 2011.

Abstract

Background & Aims

Blood tests and transient elastography (Fibroscan™) have been developed as alternatives to liver biopsy. This ANRS HCEP-23 study compared the diagnostic accuracy of nine blood tests and transient elastography (Fibroscan™) to assess liver fibrosis, vs. liver biopsy, in untreated patients with chronic hepatitis C (CHC).

Methods

This was a multicentre prospective independent study in 19 French University hospitals of consecutive adult patients having simultaneous liver biopsy, biochemical blood tests (performed in a centralized laboratory) and Fibroscan™. Two experienced pathologists independently reviewed the liver biopsies (mean length=25±8.4mm). Performance was assessed using ROC curves corrected by Obuchowski’s method.

Results

Fibroscan™ was not interpretable in 113 (22%) patients. In the 382 patients having both blood tests and interpretable Fibroscan™, Fibroscan™ performed similarly to the best blood tests for the diagnosis of significant fibrosis and cirrhosis. Obuchowski’s measure showed Fibrometer® (0.86), Fibrotest® (0.84), Hepascore® (0.84), and interpretable Fibroscan™ (0.84) to be the most accurate tests. The combination of Fibrotest®, Fibrometer®, or Hepascore® with Fibroscan™ or Apri increases the percentage of well classified patients from 70–73% to 80–83% for significant fibrosis, but for cirrhosis a combination offers no improvement. For the 436 patients having all the blood tests, AUROC’s ranged from 0.82 (Fibrometer®) to 0.75 (Hyaluronate) for significant fibrosis, and from 0.89 (Fibrometer® and Hepascore®) to 0.83 (FIB-4) for cirrhosis.

Conclusions

Contrarily to blood tests, performance of Fibroscan™ was reduced due to uninterpretable results. Fibrotest®, interpretable Fibroscan™, Fibrometer®, and Hepascore® perform best and similarly for diagnosis of significant fibrosis and cirrhosis.

Source