• Potential conflict of interest: Nothing to report.


We acknowledge Drs. Fraquelli and Rigamondi for their comments that add new data to the debate. We strongly agree that it is necessary to investigate Fibroscan performances in etiologically defined groups of patients and according to the clinical background. Even if in our paper we had examined the performances of Fibroscan for the diagnosis of cirrhosis in a large group of patients with various chronic liver diseases,1 we have realized that liver stiffness measurement (LSM) values are not related to liver architectural distortion but to the amount of liver fibrosis. This fact may explain some discrepancies, such as low values recorded in patients with macronodular inactive cirrhosis, a condition often associated with low liver collagen content. On the contrary, we disagree with them in regards to the influence of histological activity on LSM values. The French multicenter study has enrolled more than 1,000 patients with chronic hepatitis C or B. In patients with hepatitis C, LSM values were weakly correlated to activity in univariate analysis. This activity was also correlated with fibrosis but such relationship completely disappeared in multivariate analysis. This outcome has been shown in an initial group of 251 patients with HCV chronic hepatitis2 and confirmed in a validation group of 494.3 Similar results were observed in patients with hepatitis B virus, including those with high activity grades.4 We can therefore confidently conclude in these cases that there was an absence of relationship. However, we also observed patients with subacute hepatitis and severe necroinflammatory lesions with very high LSM values without cirrhosis.1 The meaning of this reversible phenomenon is unknown. Is inflammation, edema, or necrosis responsible? Looking at the picrosirius staining on biopsy samples and comparing theses cases with patients who had cirrhosis, one can suggest that the collapse of sinusoids resulting from hepatocyte loss could lead to an increased density of collagen in the liver comparable to the increase observed in cirrhotic livers. Similar findings have been observed in patients with acute alcoholic hepatitis and extensive perisinusoidal fibrosis.

Finally, if Fibroscan is becoming a validated tool for assessing liver fibrosis, we must be cautious to consider it as an automat providing instant diagnosis. As for other tests, the results of LSM must be interpreted according to the clinical background.

Nathalie Ganne-Carrié*, Marianne Ziol† **, Victor De Ledinghen‡, Catherine Douvin? ††, Patrick Marcellin?, Laurent Castera‡, Daniel Dhumeaux? ††, Jean-Claude Trinchet*, Michel Beaugrand*, * Hépato-Gastroenterologie, Université Paris 13, Bobigny, France, † Anatomopathologie, AP-HP, Hôpital Jean Verdier, Bondy, France, ‡ Hépato-Gastroenterologie, Centre d'investigation de la Fibrose, hépatique, Hôpital Haut Lévêque, CHU Bordeaux, Pessac, France, ? Hépato-Gastroenterologie, AP-HP, Hôpital Henri Mondor, Créteil, France, ? Hépatologie, AP-HP, Hôpital Beaujon, Clichy, France. INSERM U481, ** UPRES EA3409, Université Paris 13, Bobigny, France, †† INSERM E362, Université Victor Segalen, Bordeaux, France.