Portal de Periódicos da CAPES

Liver biopsy: The best standard…when everything else fails

2009; Elsevier BV; Volume: 50; Issue: 6 Linguagem: Inglês

10.1016/j.jhep.2009.02.010

1600-0641

Thierry Poynard, Yves Benhamou, Dominique Thabut, Vlad Ratziu,

Hepatitis B Virus Studies

We fully agree with the basic statement by Mehta et al. [[1]Mehta S.H. Lau B. Afdhal N.H. Thomas D.L. Exceeding the limits of liver histology markers.J Hepatol. 2009; 50: 36-41Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar] that the conventional method used so far, for validation of surrogate markers has an intrinsic limitation, because the reference standard, in this case liver biopsy, is imperfect in regard to the prediction of the disease. In chronic hepatitis C core fragments of 40 mm have a 25% rate of misclassification of METAVIR stages [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. Ironically, striving for the highest AUROC will only result in a test having the same high rate of misclassification vs. the presence of the disease. Depending on the prevalence of the disease and the sensitivity and specificity of liver biopsy, a surrogate with a lower AUROC might provide a better prediction of the disease than one with a higher AUROC. Another rational consequence is that studies observing 100% AUROC for any biomarker validated with biopsy are either underpowered or faked. Consequently Mehta et al. correctly suggest that new methodology is needed given the lack of a perfect reference standard. Clinical discussion of discordance cases and exclusion of spectrum bias are examples of such new methods [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. Spectrum bias is a significant confounder of AUROC and that differences in prevalence of stages defining advanced or non-advanced fibrosis (or defining non-cirrhosis versus cirrhosis) can explain variability of AUROCs for the same test when performed in different populations [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. We believe this is the reason why “changing the definition of diseased liver from F4 to F2F3F4 is associated with a decrease in the AUROC for a biomarker” and not because the biomarker is “not adequate” as stated in the accompanying editorial [[3]Bedossa P. Carrat F. Liver biopsy: the best, not the gold standard.J Hepatol. 2009; 50: 1-3Abstract Full Text Full Text PDF PubMed Scopus (246) Google Scholar]. We have shown that an AUROC of 0.90 for F4, an AUROC of 0.80 for F2F3F4 and an AUROC of 0.65 for the distinction between F1 and F2 are all similarly “adequate” [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. Remarkably, this is true for any diagnostic test including liver biopsy. When compared to the real gold standard, which is the liver in its entirety, liver biopsy displays similarly reduced AUROCs for the distinction of adjacent fibrosis stages (e.g. F1 vs. F2) than for distant/end of spectrum stages [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. Biomarkers or biopsy do not perform any differently in that regard. Bedossa and Carrat state that serum markers have been developed to provide a binary assessment of advanced vs. non-advanced fibrosis which they perceive as limited and insufficiently informative [[3]Bedossa P. Carrat F. Liver biopsy: the best, not the gold standard.J Hepatol. 2009; 50: 1-3Abstract Full Text Full Text PDF PubMed Scopus (246) Google Scholar]. It is worth pointing out that this pragmatic separation has been widely adopted by the FDA, the EMEA and multiple medical society guidelines. However we do agree that with surrogate markers, a further step needs to be taken, both in terms of information provided by these and validation vs. hard clinical end points. For instance, in an attempt to validate FibroTest independent of the imperfect reference standard (liver biopsy), we tested it using repeated measurements, the impact of effective treatment and prognostic value [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. In line with what Mehta et al. proposed, three prognostic validations of FibroTest in viral hepatitis C, B and alcoholic liver disease have been performed [2Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar, 4Ngo Y. Benhamou Y. Thibault V. Ingiliz P. Munteanu M. Lebray P. et al.An accurate definition of the status of inactive hepatitis B virus carrier by a combination of biomarkers (Fibrotest-Actitest) and viral load.PlosOne. 2008; 3: e2573Crossref PubMed Scopus (71) Google Scholar, 5Naveau S. Gaudé G. Asnacios A. Agostini H. Abella A. Barri-Ova N. et al.Diagnostic and prognostic values of non-invasive biomarkers of fibrosis in patients with alcoholic liver disease.Hepatology. 2009; 49: 9-11Crossref PubMed Scopus (223) Google Scholar] and similar data with other biomarkers will certainly follow. Therefore we strongly refute the editorial’s claim that “because of the conditional relationship with liver biopsy, serum markers might represent a dead end” [[3]Bedossa P. Carrat F. Liver biopsy: the best, not the gold standard.J Hepatol. 2009; 50: 1-3Abstract Full Text Full Text PDF PubMed Scopus (246) Google Scholar]. We rather side with Mehta et al.’s quite opposing view that attempts to validate markers are limited by imperfections of liver biopsy when taken as the reference standard. Contrarily to the editorialist’s statement, the other frequent liver injury such as activity, steatosis, NASH and alcoholic hepatitis can be diagnosed with available biomarkers [[2]Poynard T, Muntenau M, Morra R, Ngo Y, Imbert-Bismut F, Thabut D, et al. Methodological aspects for the interpretation of liver fibrosis non-invasive biomarkers: a 2008 update. Gastroenterol Clin Biol 2008;32:8–21.Google Scholar]. We fully agree with Mehta et al. that new methodological developments are needed, specifically designed for comparing estimates such as biopsy or biomarkers, without using any gold standard. For instance, factors associated with false-positive or false-negative of fibrosis biomarkers can be identified without using biopsy [[6]Poynard T. Ingiliz P. Elkrief L. Munteanu M. Lebray P. Morra R. et al.Concordance in a world without a gold standard: a new non-invasive methodology for improving accuracy of fibrosis markers.PlosOne. 2008; 3: e3857Crossref PubMed Scopus (78) Google Scholar]. But beyond these methodological considerations, there are some basic issues when discussing the use of liver biopsy versus that of surrogate non-invasive markers. The editorial totally overlooked the risk-benefit ratio of the different diagnostic tests. In 2007, two deaths attributable to liver biopsy were reported to a nationwide malpractice insurance company: one in chronic hepatitis C and another in primary biliary cirrhosis [[7]Sicot C. Rapport du conseil médical du Sou Médical Responsabilité. Novembre 2008.Google Scholar]. We developed research on biomarkers in order to prevent these and other serious side effects. Given the increasing incidence and mortality of some liver diseases, screening for advanced fibrosis in the general population is a major public health challenge. This can only be achieved through a moratorium on liver biopsy as a first-line estimate of injury in chronic liver diseases and the validation of non-invasive markers [[8]Poynard T. Ratziu V. Benhamou Y. Thabut D. Moussalli J. Biomarkers as a first-line estimate of injury in chronic liver diseases: time for a moratorium on liver biopsy?.Gastroenterology. 2005; 128: 1146-1148Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar]. While we agree that liver biopsy (of 40 mm) remains the best reference standard, it should only be used when all available non-invasive methods have failed to convince the clinician.