How to assess liver fibrosis and for what purpose?

Abstract

“When the liver is stiff the prognosis is bad”.—HippocratesIt is undisputable that most of the fatalities in patients with chronic liver diseases occur in patients with cirrhosis and that the evolution to cirrhosis is characterized by a progressive increase in liver fibrosis. To assess fibrosis and fibrogenesis in the liver is therefore of tremendous importance but how to do it is increasingly a subject of debate. Furthermore, the answer obviously depends on the purpose of this assessment and three main questions have to be answered. How to recognize among patients with chronic liver disease those with ‘significant’ fibrosis? How to recognize patients with cirrhosis? How to quantify and monitor the amount of fibrotic tissue in the liver? This last question may gain importance in the near future when antifibrotic drugs will be available and tested.Up to now, we have had three mains tools to answer these questions at our disposal: liver biopsy, blood tests and Fibroscan (Echosens, Paris, France), all of them having limitations and pitfalls.Liver biopsy came first historically and is still considered the gold standard despite limitations [[1]Afdhal N.H. Diagnosing fibrosis in hepatitis C: is the pendulum swinging from biopsy to blood tests?.Hepatology. 2003; 37: 972-974Crossref PubMed Scopus (114) Google Scholar]. These limitations have been emphasized repeatedly during recent years. Liver biopsy is often reluctantly accepted by the patients and even by some doctors, particularly when it has to be repeated. It carries important costs, some morbidity and most of all it is not entirely reliable. The most important point is sampling error. The mean length of liver samples in a French multicentric study comparing liver biopsy to Fibroscan was 18 mm [[2]Ziol M. Handra-Luca A. Kettaneh A. Christidis C. Mal F. Kazemi F. et al.Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C.Hepatology. 2005; 41: 48-54Crossref PubMed Scopus (1246) Google Scholar]. This is less than the minimum suitable length of 25 mm suggested for assessing liver fibrosis reliably [[3]Bedossa P. Dargere D. Paradis V. Sampling variability of liver fibrosis in chronic hepatitis C.Hepatology. 2003; 38: 1449-1457Crossref PubMed Scopus (1950) Google Scholar]. Additionally, a proper interpretation needs an experienced pathologist, a well validated scoring system and, even in this favourable situation, high rates of either intraobserver or interobserver discrepancies have been reported. These discrepancies are more frequent in patients with nonextensive fibrosis (for example in making a distinction between F1 and F2 in the Metavir classification) but other causes of error are observed in patients with cirrhosis particularly due to underscoring of patients with macronodular cirrhosis and it is generally considered that around 20% of patients are misdiagnosed.Many years ago these limitations led to the testing of blood parameters to assess liver fibrosis. In the first step, attention was paid to the extracellular matrix components and to the serum concentration of these compounds or their by products [[4]Trinchet J.C. Clinical use of serum markers of fibrosis in chronic hepatitis.J Hepatol. 1995; 22: 89-95PubMed Google Scholar]. Unfortunately, this approach was partly a failure due to several reasons: (a) there is no specific type of collagen or extracellular matrix components in the liver; even production of collagen III, if particularly dominant in chronic active liver diseases, is encountered in other organs and could be increased in the course of lung or bone marrow diseases for example; (b) serum concentration of these compounds depends on their degradation rate which could be impaired in various conditions such as renal failure or cholestasis; (c) serum concentration reflects their active metabolism more than their static amount in the liver, that is to say they reflect fibrogenesis and fibrolysis more than fibrosis itself. Furthermore collagen I, which is the hallmark of established liver fibrosis is too ubiquitous to be considered a marker of liver diseases. Despite these restrictions, at least two serum markers emerged from this period which are still in use by clinicians: PIIINP, the N-terminal propeptide of collagen III, a marker of fibrogenesis [[4]Trinchet J.C. Clinical use of serum markers of fibrosis in chronic hepatitis.J Hepatol. 1995; 22: 89-95PubMed Google Scholar], and hyaluronic acid, a marker of perisinusoidal fibrosis and cirrhosis [[5]Halfon P. Bourliere M. Penaranda G. Deydier R. Renou C. Botta-Fridlund D. et al.Accuracy of hyaluronic acid level for predicting liver fibrosis stages in patients with hepatitis C virus.Comp Hepatol. 2005; 4: 6Crossref PubMed Scopus (137) Google Scholar].In the second period, indexes considered as ‘serum markers of fibrosis’ were constructed using the association of blood tests selected on the basis of a purely statistical evaluation. These indexes could be considered as surrogate markers as most of the items tested have no direct link with liver fibrosis. Such an approach was innovative and was justified by the fact the stage of fibrosis does not take into account the amount of fibrotic tissue in the liver but the extension of liver fibrosis in the parenchyma and the resulting distortion of liver architecture. This distortion is responsible for vascular changes leading to portal hypertension and liver dysfunction. It is noteworthy that platelet count or prothrombin time for example, both predictive of the stage of liver fibrosis, reflect these later phenomena more than the amount of fibrosis in the liver. Obviously there are limitations to the use of these surrogate markers: (a) some of the parameters tested such as gamma-glutamyl transpeptidase or total bilirubin have a genetic heterogeneity which could explain elevated values in normal subjects; (b) almost all of them could be influenced by extrahepatic diseases or conditions such as hemolysis; (c) to what extent their variation with the time reflects variation in liver fibrosis is unknown but a close relationship seems improbable as they have no direct link. Despite pitfalls, these indexes seemed to have a fair predictive value for the stage of fibrosis at least in the patients tested who were mostly naive and devoid of extrahepatic diseases.The third and more recent step is the assessment of liver fibrosis through the measurement of liver stiffness. A new device called Fibroscan is able to record a simple parameter, the velocity of a shear wave inside the liver, and to translate it in liver elasticity (or stiffness). The volume of liver parenchyma explored is around 2 cm3. The technique is too recent to have been properly evaluated in most clinical situations [[6]Sandrin L. Fourquet B. Hasquenoph J.M. Yon S. Fournier C. Mal F. et al.Transient elastography: a new noninvasive method for assessment of hepatic fibrosis.Ultrasound Med Biol. 2003; 29: 1705-1713Abstract Full Text Full Text PDF PubMed Scopus (2107) Google Scholar]. But the present interest surrounding Fibroscan leads us to hope that larger and more diverse data will be available soon. What has already been suggested is the following: (a) Fibroscan achieves a proper measurement or liver elasticity in more than 90% of patients [2Ziol M. Handra-Luca A. Kettaneh A. Christidis C. Mal F. Kazemi F. et al.Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C.Hepatology. 2005; 41: 48-54Crossref PubMed Scopus (1246) Google Scholar, 7Castera L. Vergniol J. Foucher J. Le Bail B. Chanteloup E. Haaser M. et al.Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C.Gastroenterology. 2005; 128: 343-350Abstract Full Text Full Text PDF PubMed Scopus (2026) Google Scholar], ascites being a contra-indication of the method and obesity (or more precisely the fatness of the chest wall) being the most important cause of failure; (b) Fibroscan values depend on the extent of fibrosis and are highly correlated to the area of fibrosis measured by morphometry; the liver architecture seems not to influence the measurement; (c) Fibroscan is more accurate than most of the blood indexes for the diagnosis of cirrhosis [[7]Castera L. Vergniol J. Foucher J. Le Bail B. Chanteloup E. Haaser M. et al.Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C.Gastroenterology. 2005; 128: 343-350Abstract Full Text Full Text PDF PubMed Scopus (2026) Google Scholar]; (d) Its performance for selecting patients with chronic hepatitis C with Metavir score upper or equal to F2 is in the range of those serum indexes.These data, although rapidly being generated, are not sufficient to give a definite answer to the questions asked and the place for these different tools in the future remains uncertain. Nevertheless, although speculative, some statements could be proposed concerning the present and the near future. Firstly, to recognize among patients with chronic liver diseases those with significant fibrosis is not an easy task; as far as it has important implications for the patient (such as the decision to initiate or not long term treatment), it may require either a properly sized liver biopsy interpreted by an experienced pathologist or the association of two methods as recently suggested [[7]Castera L. Vergniol J. Foucher J. Le Bail B. Chanteloup E. Haaser M. et al.Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C.Gastroenterology. 2005; 128: 343-350Abstract Full Text Full Text PDF PubMed Scopus (2026) Google Scholar]. Secondly, the diagnosis of cirrhosis could be more easily assessed either by biopsy or Fibroscan or both. Blood indexes are probably not specific enough. A careful ultrasonographic examination could bring very specific findings not discussed in this editorial, the limitation of the technique being its rather low sensitivity and the need for a well trained observer [[8]Aube C. Winkfield B. Oberti F. Vuillemin E. Rousselet M.C. Caron C. et al.New doppler ultrasound signs improve the non-invasive diagnosis of cirrhosis or severe liver fibrosis.Eur J Gastroenterol Hepatol. 2004; 16: 743-751Crossref PubMed Scopus (56) Google Scholar]. Thirdly, the most uncertain issue is probably the monitoring of fibrosis in treated and untreated patients. In this setting, the role of liver biopsy although frequently advocated is questionable. Repeated liver biopsies are poorly accepted and the sampling error is such that they are in most of the cases inappropriate for individual follow-up. If confirmed, the results obtained by Fibroscan could be of major interest in this situation.Let's now turn to the future: new ways for assessing fibrosis such as MRI will be developed, proteomics could revive blood testing, new data will be available concerning Fibroscan and, to a certain extent, this technique could be improved by the designation of probes according to the morphology of the patients. Important future developments will not only concern evaluation of fibrosis but of fibrogenesis. For this purpose liver biopsy could regain interest [[9]Asselah T, Bièche I, Laurendeau I, Paradis V, Vidaud D, Degott C, et al. Liver gene expression signature of mild fibrosis in patients with chronic hepatitis C. Gastroenterology, 2005;129:2064–2075.Google Scholar]. “When the liver is stiff the prognosis is bad”. —Hippocrates It is undisputable that most of the fatalities in patients with chronic liver diseases occur in patients with cirrhosis and that the evolution to cirrhosis is characterized by a progressive increase in liver fibrosis. To assess fibrosis and fibrogenesis in the liver is therefore of tremendous importance but how to do it is increasingly a subject of debate. Furthermore, the answer obviously depends on the purpose of this assessment and three main questions have to be answered. How to recognize among patients with chronic liver disease those with ‘significant’ fibrosis? How to recognize patients with cirrhosis? How to quantify and monitor the amount of fibrotic tissue in the liver? This last question may gain importance in the near future when antifibrotic drugs will be available and tested. Up to now, we have had three mains tools to answer these questions at our disposal: liver biopsy, blood tests and Fibroscan (Echosens, Paris, France), all of them having limitations and pitfalls. Liver biopsy came first historically and is still considered the gold standard despite limitations [[1]Afdhal N.H. Diagnosing fibrosis in hepatitis C: is the pendulum swinging from biopsy to blood tests?.Hepatology. 2003; 37: 972-974Crossref PubMed Scopus (114) Google Scholar]. These limitations have been emphasized repeatedly during recent years. Liver biopsy is often reluctantly accepted by the patients and even by some doctors, particularly when it has to be repeated. It carries important costs, some morbidity and most of all it is not entirely reliable. The most important point is sampling error. The mean length of liver samples in a French multicentric study comparing liver biopsy to Fibroscan was 18 mm [[2]Ziol M. Handra-Luca A. Kettaneh A. Christidis C. Mal F. Kazemi F. et al.Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C.Hepatology. 2005; 41: 48-54Crossref PubMed Scopus (1246) Google Scholar]. This is less than the minimum suitable length of 25 mm suggested for assessing liver fibrosis reliably [[3]Bedossa P. Dargere D. Paradis V. Sampling variability of liver fibrosis in chronic hepatitis C.Hepatology. 2003; 38: 1449-1457Crossref PubMed Scopus (1950) Google Scholar]. Additionally, a proper interpretation needs an experienced pathologist, a well validated scoring system and, even in this favourable situation, high rates of either intraobserver or interobserver discrepancies have been reported. These discrepancies are more frequent in patients with nonextensive fibrosis (for example in making a distinction between F1 and F2 in the Metavir classification) but other causes of error are observed in patients with cirrhosis particularly due to underscoring of patients with macronodular cirrhosis and it is generally considered that around 20% of patients are misdiagnosed. Many years ago these limitations led to the testing of blood parameters to assess liver fibrosis. In the first step, attention was paid to the extracellular matrix components and to the serum concentration of these compounds or their by products [[4]Trinchet J.C. Clinical use of serum markers of fibrosis in chronic hepatitis.J Hepatol. 1995; 22: 89-95PubMed Google Scholar]. Unfortunately, this approach was partly a failure due to several reasons: (a) there is no specific type of collagen or extracellular matrix components in the liver; even production of collagen III, if particularly dominant in chronic active liver diseases, is encountered in other organs and could be increased in the course of lung or bone marrow diseases for example; (b) serum concentration of these compounds depends on their degradation rate which could be impaired in various conditions such as renal failure or cholestasis; (c) serum concentration reflects their active metabolism more than their static amount in the liver, that is to say they reflect fibrogenesis and fibrolysis more than fibrosis itself. Furthermore collagen I, which is the hallmark of established liver fibrosis is too ubiquitous to be considered a marker of liver diseases. Despite these restrictions, at least two serum markers emerged from this period which are still in use by clinicians: PIIINP, the N-terminal propeptide of collagen III, a marker of fibrogenesis [[4]Trinchet J.C. Clinical use of serum markers of fibrosis in chronic hepatitis.J Hepatol. 1995; 22: 89-95PubMed Google Scholar], and hyaluronic acid, a marker of perisinusoidal fibrosis and cirrhosis [[5]Halfon P. Bourliere M. Penaranda G. Deydier R. Renou C. Botta-Fridlund D. et al.Accuracy of hyaluronic acid level for predicting liver fibrosis stages in patients with hepatitis C virus.Comp Hepatol. 2005; 4: 6Crossref PubMed Scopus (137) Google Scholar]. In the second period, indexes considered as ‘serum markers of fibrosis’ were constructed using the association of blood tests selected on the basis of a purely statistical evaluation. These indexes could be considered as surrogate markers as most of the items tested have no direct link with liver fibrosis. Such an approach was innovative and was justified by the fact the stage of fibrosis does not take into account the amount of fibrotic tissue in the liver but the extension of liver fibrosis in the parenchyma and the resulting distortion of liver architecture. This distortion is responsible for vascular changes leading to portal hypertension and liver dysfunction. It is noteworthy that platelet count or prothrombin time for example, both predictive of the stage of liver fibrosis, reflect these later phenomena more than the amount of fibrosis in the liver. Obviously there are limitations to the use of these surrogate markers: (a) some of the parameters tested such as gamma-glutamyl transpeptidase or total bilirubin have a genetic heterogeneity which could explain elevated values in normal subjects; (b) almost all of them could be influenced by extrahepatic diseases or conditions such as hemolysis; (c) to what extent their variation with the time reflects variation in liver fibrosis is unknown but a close relationship seems improbable as they have no direct link. Despite pitfalls, these indexes seemed to have a fair predictive value for the stage of fibrosis at least in the patients tested who were mostly naive and devoid of extrahepatic diseases. The third and more recent step is the assessment of liver fibrosis through the measurement of liver stiffness. A new device called Fibroscan is able to record a simple parameter, the velocity of a shear wave inside the liver, and to translate it in liver elasticity (or stiffness). The volume of liver parenchyma explored is around 2 cm3. The technique is too recent to have been properly evaluated in most clinical situations [[6]Sandrin L. Fourquet B. Hasquenoph J.M. Yon S. Fournier C. Mal F. et al.Transient elastography: a new noninvasive method for assessment of hepatic fibrosis.Ultrasound Med Biol. 2003; 29: 1705-1713Abstract Full Text Full Text PDF PubMed Scopus (2107) Google Scholar]. But the present interest surrounding Fibroscan leads us to hope that larger and more diverse data will be available soon. What has already been suggested is the following: (a) Fibroscan achieves a proper measurement or liver elasticity in more than 90% of patients [2Ziol M. Handra-Luca A. Kettaneh A. Christidis C. Mal F. Kazemi F. et al.Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C.Hepatology. 2005; 41: 48-54Crossref PubMed Scopus (1246) Google Scholar, 7Castera L. Vergniol J. Foucher J. Le Bail B. Chanteloup E. Haaser M. et al.Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C.Gastroenterology. 2005; 128: 343-350Abstract Full Text Full Text PDF PubMed Scopus (2026) Google Scholar], ascites being a contra-indication of the method and obesity (or more precisely the fatness of the chest wall) being the most important cause of failure; (b) Fibroscan values depend on the extent of fibrosis and are highly correlated to the area of fibrosis measured by morphometry; the liver architecture seems not to influence the measurement; (c) Fibroscan is more accurate than most of the blood indexes for the diagnosis of cirrhosis [[7]Castera L. Vergniol J. Foucher J. Le Bail B. Chanteloup E. Haaser M. et al.Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C.Gastroenterology. 2005; 128: 343-350Abstract Full Text Full Text PDF PubMed Scopus (2026) Google Scholar]; (d) Its performance for selecting patients with chronic hepatitis C with Metavir score upper or equal to F2 is in the range of those serum indexes. These data, although rapidly being generated, are not sufficient to give a definite answer to the questions asked and the place for these different tools in the future remains uncertain. Nevertheless, although speculative, some statements could be proposed concerning the present and the near future. Firstly, to recognize among patients with chronic liver diseases those with significant fibrosis is not an easy task; as far as it has important implications for the patient (such as the decision to initiate or not long term treatment), it may require either a properly sized liver biopsy interpreted by an experienced pathologist or the association of two methods as recently suggested [[7]Castera L. Vergniol J. Foucher J. Le Bail B. Chanteloup E. Haaser M. et al.Prospective comparison of transient elastography, fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C.Gastroenterology. 2005; 128: 343-350Abstract Full Text Full Text PDF PubMed Scopus (2026) Google Scholar]. Secondly, the diagnosis of cirrhosis could be more easily assessed either by biopsy or Fibroscan or both. Blood indexes are probably not specific enough. A careful ultrasonographic examination could bring very specific findings not discussed in this editorial, the limitation of the technique being its rather low sensitivity and the need for a well trained observer [[8]Aube C. Winkfield B. Oberti F. Vuillemin E. Rousselet M.C. Caron C. et al.New doppler ultrasound signs improve the non-invasive diagnosis of cirrhosis or severe liver fibrosis.Eur J Gastroenterol Hepatol. 2004; 16: 743-751Crossref PubMed Scopus (56) Google Scholar]. Thirdly, the most uncertain issue is probably the monitoring of fibrosis in treated and untreated patients. In this setting, the role of liver biopsy although frequently advocated is questionable. Repeated liver biopsies are poorly accepted and the sampling error is such that they are in most of the cases inappropriate for individual follow-up. If confirmed, the results obtained by Fibroscan could be of major interest in this situation. Let's now turn to the future: new ways for assessing fibrosis such as MRI will be developed, proteomics could revive blood testing, new data will be available concerning Fibroscan and, to a certain extent, this technique could be improved by the designation of probes according to the morphology of the patients. Important future developments will not only concern evaluation of fibrosis but of fibrogenesis. For this purpose liver biopsy could regain interest [[9]Asselah T, Bièche I, Laurendeau I, Paradis V, Vidaud D, Degott C, et al. Liver gene expression signature of mild fibrosis in patients with chronic hepatitis C. Gastroenterology, 2005;129:2064–2075.Google Scholar].