Advertisement

Development and prognostic relevance of a histologic grading and staging system for alcohol-related liver disease

  • Author Footnotes
    § C. Lackner and R. E. Stauber share first authorship.
    ,
    Author Footnotes
    & Members of the SALVE Histopathology Group.
    Carolin Lackner
    Correspondence
    Corresponding author. Address: Institute of Pathology, Medical University of Graz, Neue Stiftingtalstraße 6, 8010 Graz, Austria; Tel.: +43 316 385 71739, fax: +43 316 385 79000.
    Footnotes
    § C. Lackner and R. E. Stauber share first authorship.
    & Members of the SALVE Histopathology Group.
    Affiliations
    Institute of Pathology, Medical University of Graz, Austria
    Search for articles by this author
  • Author Footnotes
    § C. Lackner and R. E. Stauber share first authorship.
    Rudolf E. Stauber
    Footnotes
    § C. Lackner and R. E. Stauber share first authorship.
    Affiliations
    Division of Gastroenterology and Hepatology, Department of Internal Medicine, Medical University of Graz, Graz, Austria
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Susan Davies
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Department of Histopathology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, United Kingdom
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Helmut Denk
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Institute of Pathology, Medical University of Graz, Austria
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Hans Peter Dienes
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Department of Pathology, Medical University of Vienna, Vienna, Austria
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Viviane Gnemmi
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Université Lille, Canther, Inserm, UMR-S 1277, CHU Lille, Service de Pathologie, Lille, France
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Maria Guido
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Department of Medicine - DIMED, University of Padova, Padova, Italy
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Rosa Miquel
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Liver Histopathology Laboratory, Institute of Liver Studies, King's College Hospital, London, United Kingdom
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Valerie Paradis
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Assistance Publique-Hôpitaux de Paris, Service d'Anatomie et de Cytologie Pathologiques, Hôpital Universitaire Beaujon, France

    Université Paris Diderot, CNRS, Centre de Recherche sur l'Inflammation (CRI), Paris, France

    Département Hospitalo-Universitaire (DHU) UNITY, Clichy, France
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Peter Schirmacher
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Luigi Terracciano
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Anatomic Pathology Institute, Humanitas University Research Hospital, Rozzano, Milano, Italy
    Search for articles by this author
  • Andrea Berghold
    Affiliations
    Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Graz, Austria
    Search for articles by this author
  • Gudrun Pregartner
    Affiliations
    Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Graz, Austria
    Search for articles by this author
  • Lukas Binder
    Affiliations
    Division of Gastroenterology and Hepatology, Department of Internal Medicine, Medical University of Graz, Graz, Austria
    Search for articles by this author
  • Philipp Douschan
    Affiliations
    Division of Gastroenterology and Hepatology, Department of Internal Medicine, Medical University of Graz, Graz, Austria
    Search for articles by this author
  • Florian Rainer
    Affiliations
    Division of Gastroenterology and Hepatology, Department of Internal Medicine, Medical University of Graz, Graz, Austria
    Search for articles by this author
  • Stephan Sygulla
    Affiliations
    Institute of Pathology, Medical University of Graz, Austria
    Search for articles by this author
  • Marion Jager
    Affiliations
    Université de Paris, AP-HP, Hôpital Beaujon, Service d'Hépatologie, DMU DIGEST, Centre de Référence des Maladies Vasculaires du Foie, FILFOIE, ERN RARE-LIVER, Centre de recherche sur l'inflammation, Inserm, UMR 1149, Paris, France
    Search for articles by this author
  • Pierre-Emmanuel Rautou
    Affiliations
    Université de Paris, AP-HP, Hôpital Beaujon, Service d'Hépatologie, DMU DIGEST, Centre de Référence des Maladies Vasculaires du Foie, FILFOIE, ERN RARE-LIVER, Centre de recherche sur l'inflammation, Inserm, UMR 1149, Paris, France
    Search for articles by this author
  • Andreea Bumbu
    Affiliations
    Liver Unit, Regional Institute of Gastroenterology and Hepatology, Cluj-Napoca, Romania

    Liver Research Club, Cluj-Napoca, Romania
    Search for articles by this author
  • Adelina Horhat
    Affiliations
    Liver Unit, Regional Institute of Gastroenterology and Hepatology, Cluj-Napoca, Romania

    Liver Research Club, Cluj-Napoca, Romania
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Ioana Rusu
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    University of Medicine and Pharmacy "Iuliu Hatieganu", Cluj-Napoca, Romania
    Search for articles by this author
  • Horia Stefanescu
    Affiliations
    Liver Unit, Regional Institute of Gastroenterology and Hepatology, Cluj-Napoca, Romania

    Liver Research Club, Cluj-Napoca, Romania
    Search for articles by this author
  • Author Footnotes
    & Members of the SALVE Histopathology Group.
    Sönke Detlefsen
    Footnotes
    & Members of the SALVE Histopathology Group.
    Affiliations
    Department of Pathology, Odense University Hospital, Odense C, Denmark

    Department of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense C, Denmark
    Search for articles by this author
  • Aleksander Krag
    Affiliations
    Department of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense C, Denmark

    Department of Gastroenterology and Hepatology and OPEN, Odense Patient data Explorative Network, Odense University Hospital, Odense C, Denmark
    Search for articles by this author
  • Maja Thiele
    Affiliations
    Department of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense C, Denmark

    Department of Gastroenterology and Hepatology and OPEN, Odense Patient data Explorative Network, Odense University Hospital, Odense C, Denmark
    Search for articles by this author
  • Helena Cortez-Pinto
    Affiliations
    Clínica Universitária de Gastrenterologia, Laboratório de Nutrição, Faculdade de Medicina, Universidade de Lisboa, Portugal
    Search for articles by this author
  • Christophe Moreno
    Affiliations
    Department of Gastroenterology, Hepatopancreatology and Digestive Oncology, CUB Hôpital Erasme, Université Libre de Bruxelles, Bruxelles, Belgium
    Search for articles by this author
  • Author Footnotes
    $ A. S. H. Gouw and D. G. Tiniakos share senior authorship.
    ,
    Author Footnotes
    & Members of the SALVE Histopathology Group.
    Annette S.H. Gouw
    Footnotes
    $ A. S. H. Gouw and D. G. Tiniakos share senior authorship.
    & Members of the SALVE Histopathology Group.
    Affiliations
    Department of Pathology and Medical Biology, University Medical Center Groningen, Groningen, Netherlands
    Search for articles by this author
  • Author Footnotes
    $ A. S. H. Gouw and D. G. Tiniakos share senior authorship.
    ,
    Author Footnotes
    & Members of the SALVE Histopathology Group.
    Dina G. Tiniakos
    Footnotes
    $ A. S. H. Gouw and D. G. Tiniakos share senior authorship.
    & Members of the SALVE Histopathology Group.
    Affiliations
    Transitional and Clinical Research Institute, Newcastle University, Newcastle upon Tyne NE2 4HH, UK

    Department of Pathology, Aretaieio Hospital, Medical School, National & Kapodistrian University of Athens, Athens 11528, Greece
    Search for articles by this author
  • Author Footnotes
    § C. Lackner and R. E. Stauber share first authorship.
    $ A. S. H. Gouw and D. G. Tiniakos share senior authorship.
    & Members of the SALVE Histopathology Group.
Open AccessPublished:June 11, 2021DOI:https://doi.org/10.1016/j.jhep.2021.05.029

      Highlights

      • ALD is the most frequent cause of cirrhosis in Europe, but an ALD-specific histologic grading and staging system is lacking.
      • The SALVE Histopathology Group developed a reproducible grading and staging system for ALD.
      • Prognostic value of the new grading and staging system was evaluated in a large cohort of patients with ALD enrolled at 4 European centers.
      • ASH and severe cirrhosis on histology were independent predictors of liver-related events/deaths in compensated/decompensated ALD.

      Background & Aims

      The SALVE Histopathology Group (SHG) developed and validated a grading and staging system for the clinical and full histological spectrum of alcohol-related liver disease (ALD) and evaluated its prognostic utility in a multinational cohort of 445 patients.

      Methods

      SALVE grade was described by semiquantitative scores for steatosis, activity (hepatocellular injury and lobular neutrophils) and cholestasis. The histological diagnosis of steatohepatitis due to ALD (histological ASH, hASH) was based on the presence of hepatocellular ballooning and lobular neutrophils. Fibrosis staging was adapted from the Clinical Research Network staging system for non-alcoholic fatty liver disease and the Laennec staging system and reflects the pattern and extent of ALD fibrosis. There are 7 SALVE fibrosis stages (SFS) ranging from no fibrosis to severe cirrhosis.

      Results

      Interobserver κ-value for each grading and staging parameter was >0.6. In the whole study cohort, long-term outcome was associated with activity grade and cholestasis, as well as cirrhosis with very broad septa (severe cirrhosis) (p <0.001 for all parameters). In decompensated ALD, adverse short-term outcome was associated with activity grade, hASH and cholestasis (p = 0.038, 0.012 and 0.001, respectively), whereas in compensated ALD, hASH and severe fibrosis/cirrhosis were associated with decompensation-free survival (p = 0.011 and 0.001, respectively). On multivariable analysis, severe cirrhosis emerged as an independent histological predictor of long-term survival in the whole study cohort. Severe cirrhosis and hASH were identified as independent predictors of short-term survival in decompensated ALD, and also as independent predictors of decompensation-free survival in compensated ALD.

      Conclusion

      The SALVE grading and staging system is a reproducible and prognostically relevant method for the histological assessment of disease activity and fibrosis in ALD.

      Lay summary

      Patients with alcohol-related liver disease (ALD) may undergo liver biopsy to assess disease severity. We developed a system to classify ALD under the microscope by grading ALD activity and staging the extent of liver scarring. We validated the prognostic performance of this system in 445 patients from 4 European centers.

      Graphical abstract

      Keywords

      Introduction

      Alcohol abuse is a major global health concern, being a frequent cause of chronic liver disease, cirrhosis, hepatocellular carcinoma and indication for liver transplantation. Alcohol-related liver disease (ALD) shows a spectrum of liver pathology ranging from steatosis to steatohepatitis and fibrosis.
      • Yip W.W.
      • Burt A.D.
      Alcoholic liver disease.
      For the sake of clarity, steatohepatitis related to ALD
      European Association for the Study of the Liver
      EASL Clinical Practice Guidelines: Management of Alcohol-Related Liver Disease.
      is referred to as histological alcoholic steatohepatitis (hASH) in this manuscript, to differentiate it from non-alcoholic steatohepatitis and it is not synonymous with the clinical scenario of alcoholic hepatitis.
      Clinically, steatosis is associated with few, if any, symptoms and has a low risk of progression, whereas hASH is a major driver of fibrogenesis and disease progression. In turn, progression can be associated with clinical abnormalities, development of cirrhosis and hepatocellular carcinoma. Severe symptoms may be due to decompensation of cirrhosis and/or clinical alcoholic hepatitis the latter being associated with 3-month mortality rates of 20–50%.
      European Association for the Study of the Liver
      EASL Clinical Practice Guidelines: Management of Alcohol-Related Liver Disease.
      ,
      • Lucey M.R.
      • Mathurin P.
      • Morgan T.R.
      Alcoholic hepatitis.
      As clinical ALD classifications may correspond poorly with histology,
      • Shen N.T.
      • Salajegheh A.
      • Brown Jr., R.S.
      A call to standardize definitions, data collection, and outcome assessment to improve care in alcohol-related liver disease.
      EASL Clinical Practice Guidelines for the Management of ALD recommend liver biopsy in cases where the diagnosis of ALD is uncertain in both clinical practice and clinical trials.
      European Association for the Study of the Liver
      EASL Clinical Practice Guidelines: Management of Alcohol-Related Liver Disease.
      Standardized and reproducible assessment of disease activity and fibrosis is a prerequisite for histology-based patient stratification, prognosis and monitoring of treatment effects. Several histological grading and staging systems have been developed for use in chronic liver disease including non-alcoholic fatty liver disease (NAFLD)
      • Kleiner D.E.
      • Brunt E.M.
      • Van Natta M.
      • Behling C.
      • Contos M.J.
      • Cummings O.W.
      • et al.
      Design and validation of a histological scoring system for nonalcoholic fatty liver disease.
      ,
      • Bedossa P.
      FLIP Pathology Consortium
      Utility and appropriateness of the fatty liver inhibition of progression (FLIP) algorithm and steatosis, activity, and fibrosis (SAF) score in the evaluation of biopsies of nonalcoholic fatty liver disease.
      and viral hepatitis.
      • Ishak K.
      • Baptista A.
      • Bianchi L.
      • Callea F.
      • De Groote J.
      • Gudat F.
      • et al.
      Histological grading and staging of chronic hepatitis.
      ,
      • Bedossa P.
      • Poynard T.
      An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group.
      Although ALD is among the most frequent of liver diseases and its morphological features are well described, few proposals for specific grading and staging systems have been made
      • Yip W.W.
      • Burt A.D.
      Alcoholic liver disease.
      and a universally accepted system for the full clinical spectrum of ALD is currently lacking.
      As ALD and NAFLD show histological overlap, some have proposed applying NAFLD grading and staging systems for ALD,
      • Lefkowitch J.H.
      Morphology of alcoholic liver disease.
      but several prognostically relevant ALD features, like cholestasis, Mallory-Denk bodies (MDB) and megamitochondria, are not considered in NAFLD grading.
      • Brunt E.M.
      • Neuschwander-Tetri B.A.
      • Burt A.D.
      Fatty liver disease: alcoholic and non-alcoholic.
      Further, the vast majority of patients with ALD have cirrhosis at first presentation
      • Lackner C.
      • Spindelboeck W.
      • Haybaeck J.
      • Douschan P.
      • Rainer F.
      • Terracciano L.
      • et al.
      Histological parameters and alcohol abstinence determine long-term prognosis in patients with alcoholic liver disease.
      in contrast to patients with NAFLD in whom it is infrequent.
      • Goh G.B.
      • Pagadala M.R.
      • Dasarathy J.
      • Unalp-Arida A.
      • Sargent R.
      • Hawkins C.
      • et al.
      Clinical spectrum of non-alcoholic fatty liver disease in diabetic and non-diabetic patients.
      Histological substages of cirrhosis
      • Tsochatzis E.
      • Bruno S.
      • Isgro G.
      • Hall A.
      • Theocharidou E.
      • Manousou P.
      • et al.
      Collagen proportionate area is superior to other histological methods for sub-classifying cirrhosis and determining prognosis.
      ,
      • Kim S.U.
      • Oh H.J.
      • Wanless I.R.
      • Lee S.
      • Han K.H.
      • Park Y.N.
      The Laennec staging system for histological sub-classification of cirrhosis is useful for stratification of prognosis in patients with liver cirrhosis.
      and the extent of pericellular fibrosis (PCF)
      • Altamirano J.
      • Miquel R.
      • Katoonizadeh A.
      • Abraldes J.G.
      • Duarte-Rojo A.
      • Louvet A.
      • et al.
      A histologic scoring system for prognosis of patients with alcoholic hepatitis.
      are prognostically relevant in ALD but are not reflected in current NAFLD staging systems. Finally, ALD is associated with fibro-obliteration of hepatic veins, perivenular fibrosis and sclerosing hyaline necrosis, all predictors of progression and adverse prognosis which are rare in NAFLD.
      • Burt A.D.
      • MacSween R.N.
      Hepatic vein lesions in alcoholic liver disease: retrospective biopsy and necropsy study.
      ,
      • Nakano M.
      • Worner T.M.
      • Lieber C.S.
      Perivenular fibrosis in alcoholic liver injury: ultrastructure and histologic progression.
      A group of European liver pathologists, members of the EASL-endorsed consortium for the Study of Alcohol-related LiVer disease in Europe (SALVE), comprising the SALVE Histopathology Group (SHG), convened to design a morphological grading and staging system valid for the whole clinical spectrum of the disease and evaluated its prognostic utility.

      Patients and methods

       Study cohort

      A previously described retrospective cohort with clinical and histologically confirmed ALD, the Graz cohort,
      • Lackner C.
      • Spindelboeck W.
      • Haybaeck J.
      • Douschan P.
      • Rainer F.
      • Terracciano L.
      • et al.
      Histological parameters and alcohol abstinence determine long-term prognosis in patients with alcoholic liver disease.
      was used to design the morphological grading and staging system and to assess interobserver variation. The prognostic utility of the grading and staging system was then evaluated in the Graz cohort and 3 additional cohorts from SALVE centres in Odense (Odense University Hospital, Denmark), Paris (Hôpital Beaujon, Clichy, France), and Cluj-Napoca (Regional Institute of Gastroenterology and Hepatology, Cluj-Napoca, Romania). Patients in the Graz cohort underwent liver biopsy for diagnosis and/or staging of liver disease. The Paris and Cluj cohorts included consecutive patients undergoing liver biopsy for suspicion of clinical alcoholic hepatitis. The Odense cohort included patients from a prospective diagnostic study on patients with compensated ALD in Southern Denmark.
      • Thiele M.
      • Detlefsen S.
      • Sevelsted Moller L.
      • Madsen B.S.
      • Fuglsang Hansen J.
      • Fialla A.D.
      • et al.
      Transient and 2-dimensional shear-wave elastography provide comparable assessment of alcoholic liver fibrosis and cirrhosis.
      ,
      • Thiele M.
      • Madsen B.S.
      • Hansen J.F.
      • Detlefsen S.
      • Antonsen S.
      • Krag A.
      Accuracy of the enhanced liver fibrosis test vs FibroTest, elastography, and indirect markers in detection of advanced fibrosis in patients with alcoholic liver disease.
      Patients received standard of care or, if needed, intensive care support.
      Follow-up data on survival and liver transplantation were available for all patients. Length of survival and cause of death were documented based on data from hospitals, family practitioners and national death registries. Data on abstinence during follow-up were available in 323 out of 445 patients. In addition, data on liver-related events were collected in the subgroup of patients with compensated ALD. All studies received approval by the local Ethics Committees of all centres. Informed consent was obtained in accordance with the Declaration of Helsinki.

       Design of the SALVE grading and staging system

      The SALVE grading and staging system was based on (i) characteristic morphological features of ALD, (ii) previously reported independent prognostic parameters of grade
      • Altamirano J.
      • Miquel R.
      • Katoonizadeh A.
      • Abraldes J.G.
      • Duarte-Rojo A.
      • Louvet A.
      • et al.
      A histologic scoring system for prognosis of patients with alcoholic hepatitis.
      ,
      • Katoonizadeh A.
      • Laleman W.
      • Verslype C.
      • Wilmer A.
      • Maleux G.
      • Roskams T.
      • et al.
      Early features of acute-on-chronic alcoholic liver failure: a prospective cohort study.
      ,
      • Spahr L.
      • Rubbia-Brandt L.
      • Genevay M.
      • Hadengue A.
      • Giostra E.
      Early liver biopsy, intraparenchymal cholestasis, and prognosis in patients with alcoholic steatohepatitis.
      and stage,
      • Lackner C.
      • Spindelboeck W.
      • Haybaeck J.
      • Douschan P.
      • Rainer F.
      • Terracciano L.
      • et al.
      Histological parameters and alcohol abstinence determine long-term prognosis in patients with alcoholic liver disease.
      ,
      • Kim S.U.
      • Oh H.J.
      • Wanless I.R.
      • Lee S.
      • Han K.H.
      • Park Y.N.
      The Laennec staging system for histological sub-classification of cirrhosis is useful for stratification of prognosis in patients with liver cirrhosis.
      ,
      • Altamirano J.
      • Bataller R.
      Alcoholic liver disease: pathogenesis and new targets for therapy.
      and (iii) at least substantial interobserver agreement as documented in the literature or according to the results of studies by the SHG (described below). For the Graz cohort, morphological evaluations were carried out by members of the SHG in consensus using a multiheaded microscope. Scanned liver biopsy slides of the Cluj-Napoca, Paris and Odense patients were scored in consensus by groups of at least 3 SHG pathologists using the “share screen” option on a digital platform. One SHG pathologist (CL) attended all virtual scoring sessions to ensure homogeneity of histological evaluation. The observers were unaware of the clinical data.
      All studied samples were routinely stained with hematoxylin-eosin and either chromotrope aniline blue or Sirius red.

       SALVE grading

      A semiquantitative evaluation method was defined using numerical scores for macrovesicular steatosis (0-3), hepatocellular ballooning (0-2), MDB (0-2), and lobular neutrophils (0-2). Ballooning was defined as hepatocellular enlargement (at least 2x the size of normal hepatocytes), rounded cellular shape and rarefied cytoplasm (cytoplasmic clarification). Cholestasis was specified as hepatocellular, canalicular, or ductular cholestasis and scored as absent (0) or present (1). The SHG evaluated 30 cases on digitized slides for the assessment of interobserver variation.
      Parameters with substantial interobserver agreement (κ >0.6) namely steatosis, ballooning, MDB, lobular neutrophils, canalicular and ductular cholestasis were selected as descriptors of SALVE Grade defined by scores for steatosis (0-3), activity (0-4), canalicular (0-1) and ductular cholestasis (0-1) (Fig. S1). Because ballooning and MDB scores showed a strong correlation (Spearman´s rho = 0.9), the higher score of either feature rather than their sum was considered, to avoid overestimation of hepatocellular injury. The activity range of 0-4 was based on cellular injury and inflammation as the sum of scores for ballooning or MDB and lobular neutrophils. The definition of ASH
      • Yip W.W.
      • Burt A.D.
      Alcoholic liver disease.
      was based on ballooning and neutrophil scores of ≥1 each and cases with activity scores of 3 and 4, or in some cases 2, were diagnosed with hASH. The SALVE grading system is shown in Box 1.
      SALVE grading.
      SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe.
      Lipid vacuoles in the cytoplasm of hepatocytes larger than the hepatocellular nucleus.
      ∗∗If scores for ballooning and Mallory-Denk bodies are unequal the higher score is applied.
      §Feature is appreciated after a reasonable search and is present in few microscopic fields.
      §§Feature is frequent and easy to find without searching and present in many microscopic fields.
      %Neutrophils surrounding ballooned hepatocytes.

       SALVE staging

      Fibrosis stages were described based on a combination of the NASH Clinical Research Network (CRN)
      • Kleiner D.E.
      • Brunt E.M.
      • Van Natta M.
      • Behling C.
      • Contos M.J.
      • Cummings O.W.
      • et al.
      Design and validation of a histological scoring system for nonalcoholic fatty liver disease.
      and the Laennec systems
      • Kim S.U.
      • Oh H.J.
      • Wanless I.R.
      • Lee S.
      • Han K.H.
      • Park Y.N.
      The Laennec staging system for histological sub-classification of cirrhosis is useful for stratification of prognosis in patients with liver cirrhosis.
      with some modifications. The SALVE staging system details 7 SALVE fibrosis stages (SFS) comprising 4 pre-cirrhotic (SFS 0, 1, 2, and 3), similar to the CRN, and 3 cirrhotic stages (SFS 4A, 4B, and 4C), similar to Laennec. For some clinical settings, and also within trials, the presence of severe pericellular fibrosis (PCF) may be included, as described below and in Table 1. Morphological aspects of staging are also illustrated in Table 1 and Figs S2, S3 and S4.
      Table 1SALVE staging.
      SFSDescriptionMorphological changes
      Description of the full range of topographical abnormal fibrosis including the degree of both dense septal and pericellular fibrosis.
      Examples
      0No fibrosisFibrosis is absent
      1Mild fibrosisPeriportal fibrosis only or PCF
      Pericellular fibrosis: Collagen fibers surrounding single or small groups of hepatocytes.
      in zone(s) 3 ± 2
      SFS 1P
      OPTIONAL, the presence of pericellular fibrosis as a dominant fibrosis type may be classified as “P” substage.
      : PCF in zone(s) 3 ± 2 only
      2Moderate fibrosisPeriportal fibrosis and PCF in zone(s) 3 ± 2
      3Severe fibrosis≥1 complete septum
      Complete septum: Fibrous band consisting mainly of collagen fibers resembling septa in viral hepatitis or septal PCF crossing biopsy diameter and linking portal tracts, portal tracts and central veins, or central veins.
      bridging portal tracts and/or central veins, ±PCF
      SFS 3P
      OPTIONAL, the presence of pericellular fibrosis as a dominant fibrosis type may be classified as “P” substage.
      : Panlobular PCF and/or complete septal PCF ± few dense septa ± venous lesions
      4ACirrhosis thin septa≥1 parenchymal nodule
      Parenchymal nodule without evidence of portal-central relations surrounded by dense septa.
      , thin septa
      Thin septum: Dense septum, <50% of diameter of smallest parenchymal nodule.
      , ± 1 broad septum
      Broad septum: Dense septum, ≥50% of the diameter of smallest parenchymal nodule but not thicker.
      ,

      ±PCF

      SFS 4AP
      OPTIONAL, the presence of pericellular fibrosis as a dominant fibrosis type may be classified as “P” substage.
      : Severe PCF
      PCF evaluated at LOW magnification (20x or 40x total magnification).
      in >50% of parenchyma, indistinct parenchymal nodules
      Parenchymal areas of indistinct nodular shape dissected by severe PCF.
      4BCirrhosis broad septaParenchymal nodules, >1 broad septum, ± 1 very broad septum
      Very broad septum: Dense septum, wider than the diameter of smallest parenchymal nodule.
      , ±PCF
      SFS 4BP
      OPTIONAL, the presence of pericellular fibrosis as a dominant fibrosis type may be classified as “P” substage.
      : Severe PCF in >50% of parenchyma
      4CCirrhosis very broad septaParenchymal nodules, >1 very broad septum, ±PCF

      SFS 4CP
      OPTIONAL, the presence of pericellular fibrosis as a dominant fibrosis type may be classified as “P” substage.
      : Severe PCF in >50% of parenchyma
      PCF, pericellular fibrosis; SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe; SFS, SALVE Fibrosis Stage.
      a Description of the full range of topographical abnormal fibrosis including the degree of both dense septal and pericellular fibrosis.
      b Pericellular fibrosis: Collagen fibers surrounding single or small groups of hepatocytes.
      c OPTIONAL, the presence of pericellular fibrosis as a dominant fibrosis type may be classified as “P” substage.
      d Complete septum: Fibrous band consisting mainly of collagen fibers resembling septa in viral hepatitis or septal PCF crossing biopsy diameter and linking portal tracts, portal tracts and central veins, or central veins.
      e Parenchymal nodule without evidence of portal-central relations surrounded by dense septa.
      f Thin septum: Dense septum, <50% of diameter of smallest parenchymal nodule.
      g Broad septum: Dense septum, ≥50% of the diameter of smallest parenchymal nodule but not thicker.
      h PCF evaluated at LOW magnification (20x or 40x total magnification).
      i Parenchymal areas of indistinct nodular shape dissected by severe PCF.
      j Very broad septum: Dense septum, wider than the diameter of smallest parenchymal nodule.
      SALVE fibrosis stage 1 (SFS 1) comprises 2 distinct morphological patterns: Typically, centrilobular regions are affected by PCF occasionally extending to intermediate lobular areas (Fig. S2B). Alternatively, there may be predominantly portal-based fibrosis with periportal extension (Fig. S2C).
      • Brunt E.M.
      • Neuschwander-Tetri B.A.
      • Burt A.D.
      Fatty liver disease: alcoholic and non-alcoholic.
      SFS 2 is defined by coexisting centrilobular PCF and periportal fibrosis (Fig. S2D). In SFS 3 portal-based dense fibrous septa develop, linking portal tracts, portal tracts and central veins as well as central veins. This may be accompanied by variable degrees of PCF (see below for when PCF predominates) (SFS 3) (Fig. S2E,F). The cirrhosis stage SFS 4 is characterized by destruction of lobular architecture and development of parenchymal nodules surrounded by septa, the thickness of which are used for sub-classification. Septa may be dense and thin (SFS 4A; Fig. S3A,B), broad (SFS 4B; Fig. S3C,D) or very broad (SFS 4C; Fig. S3E,F). The assessment of septal thickness was based on the dimension of the smallest distinct parenchymal nodule as detailed in Table 1.

       Expanded SALVE fibrosis staging in consideration of severe forms of pericellular fibrosis

      In ALD, PCF can be the predominant pattern of fibrosis, present at all stages and therefore SFS 1, 3, and 4A-C can be further classified as outlined below and in Table 1.
      In SFS1 cases, the designation SFS 1P may be used to indicate the presence of centrilobular PCF only (Fig. S2B). In some cases of SFS 3, SFS 3P denotes that severe PCF may assume a septum-like configuration (centro-central septal PCF) (Fig. S4A) or involve entire hepatic lobules but with preservation of porto-central relations (Figs S2E and S4B). In this setting obliterative venous lesions are frequently noted, and a few dense septa may be seen. In SFS 4AP severe PCF is present, destroying portal-central relations resulting in indistinct parenchymal nodules (Fig. S3B). SFS 4B and 4C with severe PCF and indistinct parenchymal nodules may be referred to as SFS 4BP (Fig. S3D) or 4CP (Fig. S3F), respectively.
      An algorithm was designed to facilitate and standardize the staging procedure (Fig. 1). Two groups of observers of the SHG independently assessed SFS of the first consecutive 140 cases of the Graz cohort using the SALVE Staging Algorithm for interobserver studies.
      Figure thumbnail gr1
      Fig. 1SALVE staging algorithm.
      ALD, alcohol-related liver disease; PCF, pericellular fibrosis; SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe; SFS, SALVE fibrosis stage; cOptional stages to indicate presence of PCF as the predominant fibrosis type. For further details regarding definitions please see footnote.

       Assessment of venous lesions of ALD

      Perivenular fibrosis, sclerosing hyaline necrosis and fibro-obliteration of hepatic veins are presumed to be of prognostic relevance. Therefore, their presence was evaluated by 5 SHG observers in 140 cases of the Graz cohort.

       Statistical analyses

      Continuous variables were reported as median (Q1, Q3), whereas categorical data are presented as relative frequencies. Liver-related mortality at short-term (90 days) or long-term (end of follow-up) was defined as death due to liver failure, complications of cirrhosis or HCC. Patients with non-liver-related death and those undergoing liver transplantation during follow-up were censored and counted as non-event. Decompensation-free survival was defined as absence of liver-related events (new-onset jaundice, ascites, portal hypertensive bleeding, hepatic encephalopathy) or liver-related death during follow-up. The effect of prognostic variables on survival was analysed by the Kaplan-Meier method and compared by log-rank tests performing Bonferroni correction for pairwise comparisons. The association of clinical, biochemical, and histological variables with survival was analysed by univariable and multivariable Cox regression. Multicollinearity was assessed with variance inflation factors (VIF). Statistical analyses were performed using SPSS Statistics Version 26 (SPSS Inc., Chicago, IL, USA) or R version 3.6.1. A p value of <0.05 was considered significant.

      Results

       Clinical, biochemical, demographical, and histological characteristics of the study cohort

      The whole study cohort consisted of 445 patients representing the entire clinical spectrum of ALD. Subgroup analysis was performed in patients with (i) compensated ALD defined by lack of clinical symptoms and no evidence of cirrhosis on ultrasonography or biochemistry (n = 159), (ii) decompensated ALD characterized by bilirubin levels >3 mg/dl and/or signs of decompensation (new-onset jaundice, ascites, hepatic encephalopathy, portal hypertensive gastrointestinal bleeding) (n = 286), and (iii) decompensated ALD and hASH (n = 181). Twenty-eight patients died from non-liver-related causes. Twenty-two patients underwent liver transplantation during follow-up (6 patients within 90 days from liver biopsy). Clinical, biochemical, and demographic characteristics of the study cohorts are shown in Table 2.
      Table 2Clinical, biochemical and demographic characteristics of the study cohorts.
      CentreGraz (n = 172)Cluj-Napoca (n = 92)Paris (n = 75)Odense (n = 106)
      Period of enrolment1995-20092016-20192011-20192013-2016
      Decompensated ALD, %691001000
      Age, years49 (41–57)51 (43–57)54 (48–59)56 (49–62)
      BMI25 (22–29)26 (22–30)26 (22–30)26 (23–28)
      Sex female, %31272026
      AST, U/L45 (26–77)140 (99–188)146 (96–197)39 (27–55)
      ALT, U/L28 (16–54)44 (27–57)41 (31–67)30 (21–43)
      GGT, U/L148 (62–327)332 (207–642)272 (134–667)103 (47–238)
      Alkaline phosphatase, U/L160 (109–226)463 (322–573)164 (122–234)96 (76–125)
      Bilirubin, mg/dl2.6 (1.1–8.5)8.1 (4.0–19.5)11.1 (6.7–19.4)0.6 (0.4–0.9)
      INR1.22 (1.03–1.57)1.93 (1.65–2.29)1.82 (1.58–2.36)1.0 (0.9–1.1)
      Creatinine, mg/dl0.9 (0.8–1.1)0.7 (0.6–0.8)0.7 (0.6–1.0)0.8 (0.7–0.9)
      Albumin, g/dl3.5 (2.9–4.3)2.8 (2.5–3.1)2.0 (1.8–2.4)4.1 (3.8–4.3)
      Platelet count, G/L138 (94–231)110 (78–162)127 (73–173)227 (163–297)
      Leucocyte count, G/L6.8 (5.3–10.8)9.8 (7.2–13.0)9.9 (7.0–15.2)6.7 (5.4–9.2)
      MCV, fl100 (94–105)102 (97–108)not reported95 (91–101)
      Sodium, mmol/L138 (135–141)136 (133–139)132 (128–136)139 (138–141)
      Variceal bleeding, %1516not reported0
      Hepatic encephalopathy, %2137310
      Ascites, %4484880
      MELD14 (9–20)22 (19–26)23 (19–27)6 (6–8)
      Child-Pugh score8 (6–10)10 (8–12)12 (11–13)5 (5–5)
      Route of biopsy, n (percutaneous/transjugular)164/80/920/75106/0
      Length of biopsy core, mm18 (12–27)21 (16–28)37 (28–50)33 (27–40)
      Survival, years4.1 (0.9–8.8)1.6 (0.4–3.5)0.5 (0.1–1.4)4.0 (3.5–4.9)
      90-day mortality, %923270
      5-year mortality, %3449445
      ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamyltransferase; INR, international normalized ratio; MCV, mean corpuscular volume; MELD, model for end-stage liver disease.
      Data are given as median (IQR).
      Histological characteristics of the study cohort are compiled in Table S1. Median biopsy length was 25 mm in transcutaneous and 27 mm in transjugular biopsies. Decompensated patients more often had severe cirrhosis (SFS 4C). In addition, all cirrhotic stages with severe PCF (SFS 4AP, 4BP and 4CP) were more frequent in decompensated than in compensated patients. Individuals with decompensated disease had higher activity grade and presence of hASH than patients with compensated ALD. Canalicular cholestasis was infrequent and ductular cholestasis was nearly absent in compensated patients, but they were present in approximately 57% and 27% of those with decompensation, respectively.

       Parameters of SALVE grading and staging in interobserver studies

      Substantial interobserver agreement was found for steatosis, ballooning, MDBs, lobular neutrophils, canalicular and ductular cholestasis, whereas agreement was moderate for the interpretation of hepatocellular cholestasis. Interobserver agreement for interpretation of the 7 SFS as well as the SFS with severe PCF was substantial (Table 3).
      Table 3Interobserver variation in scoring of histological features of SALVE grade and stage (Graz cohort).
      ItemScoring systemKappa value
      SteatosisScore 0: <5%; 1:5-33%; 2:34-66%; 3: >66%0.88
      Kendall’s W, 10 raters, 27-29 observations.
      Hepatocellular ballooningScore 0: none, 1: few; 2: many0.66
      Kendall’s W, 10 raters, 27-29 observations.
      Mallory-Denk bodiesScore 0: none, 1: few; 2: many0.78
      Kendall’s W, 10 raters, 27-29 observations.
      Lobular neutrophilsScore 0: none, 1: few, 2: many and/or satellitosis0.67
      Kendall’s W, 10 raters, 27-29 observations.
      Hepatocellular cholestasis0: none, 1: present0.33
      Fleiss‘ Kappa, 10 raters, 28-29 observations.
      Canalicular cholestasis0: none, 1: present0.65
      Fleiss‘ Kappa, 10 raters, 28-29 observations.
      Ductular cholestasis0: none, 1: present0.66
      Fleiss‘ Kappa, 10 raters, 28-29 observations.
      SFS (all substages)0, 1/1P, 2, 3/3P, 4A/4AP, 4B/4BP, 4C/4CP0.69
      Cohen’s Kappa, 2 rater groups, 138 observations.
      SFS (main stages)0, 1, 2, 3, 4A, 4B, 4C0.80
      Cohen’s Kappa, 2 rater groups, 138 observations.
      Pericellular fibrosis0, 10.69
      Cohen’s Kappa, 2 rater groups, 138 observations.
      SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe; SFS, SALVE fibrosis stage.
      a Kendall’s W, 10 raters, 27-29 observations.
      b Fleiss‘ Kappa, 10 raters, 28-29 observations.
      c Cohen’s Kappa, 2 rater groups, 138 observations.

       Association of SALVE grading with survival

      Kaplan-Meier analysis in the whole study cohort revealed no prognostic utility for steatosis grade (data not shown). The association of SALVE activity grade with survival is shown in Fig. 2A. High activity (grade 2-4) was associated with significantly shorter survival (p <0.001 vs. grade 0-1). An association with shorter survival was also seen for patients with hASH compared to those without hASH (p <0.001; Fig. S5). Survival of patients with canalicular or ductular cholestasis was significantly shorter than that of individuals without cholestasis (both p <0.001, significance level p = 0.016) (Fig. 2B).
      Figure thumbnail gr2
      Fig. 2Kaplan-Meier plots of long-term survival by SALVE grade (whole cohort, n = 445).
      (A) Effect of activity grade on survival: 0-1, no or mild activity; 2-4, high activity; p <0.001 (log-rank test). (B) Effect of CC and DC on survival; p <0.001 (log-rank test). CC, canalicular cholestasis; DC, ductular cholestasis; SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe.
      Patients with decompensated ALD and activity grade 2-4, hASH or canalicular and/or ductular cholestasis had significantly higher 90-day mortality than patients with activity grade 0-1, no hASH or no cholestasis (p = 0.038, 0.012 or 0.001, respectively) (Fig. S6A-C). In the subgroup of decompensated patients with hASH, canalicular and/or ductular cholestasis was associated with lower 90-day survival compared to those without cholestasis (p = 0.029) (Fig. S7).
      In patients with compensated ALD, activity grade 2-4 or hASH was associated with a higher incidence of liver-related events during follow-up than activity grade 0-1 or no hASH (p = 0.011, respectively) (Fig. S8A-B).

       Association of SALVE staging with survival

      SFS staging was aggregated in a 5-tiered system based on relation to survival: no fibrosis (SFS 0), mild and moderate fibrosis (SFS 1 and 2), severe fibrosis (SFS 3), cirrhosis with thin or broad septa (SFS 4A and 4B), and cirrhosis with very broad septa (severe cirrhosis; SFS 4C). On Kaplan-Meier analysis the respective 10-year survival probabilities of patients were 100%, 89%, 65%, 43% and 32% (Fig. 3A).
      Figure thumbnail gr3
      Fig. 3Kaplan-Meier plots of long-term outcome by SFS.
      (A) Effect of SFS tiers on long-term survival (whole cohort, n = 445): no fibrosis, SFS 0; mild fibrosis, SFS 1-2; severe fibrosis, SFS 3; cirrhosis, SFS 4A-4B; severe cirrhosis, SFS 4C; p <0.001 (log-rank test). (B) Effect of SFS on decompensation-free survival during follow-up (compensated ALD, n = 159): no/mild fibrosis, SFS 0-2; severe fibrosis/cirrhosis, SFS 3-4; p <0.001 (log-rank test). ALD, alcohol-related liver disease; SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe; SFS, SALVE fibrosis stage.
      Long-term mortality was significantly higher for patients with severe cirrhosis vs. pre-cirrhotic stages and vs. lesser cirrhosis (p <0.001 and p = 0.003, significance level p = 0.016). In contrast, severe cirrhosis showed only a trend in relation to short-term outcome (p = 0.162) in patients with decompensated disease and also in the subgroup of decompensated patients with hASH (p = 0.070) (Fig. S9). In compensated ALD, severe fibrosis/cirrhosis was related to the development of liver-related events on long-term follow-up (p <0.001) (Fig. 3B).

       Association of the pericellular fibrosis type with outcome, inflammation, and venous lesions

      Stages with severe PCF (SFS 3P, 4AP, 4BP, 4CP) had significantly worse long-term outcome than the respective SFS stages without severe PCF (SFS 3, 4A, 4B, 4C) (p = 0.042) (Fig. S10).
      Severe PCF was associated with ballooning/MDB, cholestasis and lobular neutrophils (Chi-square test, p <0.001 for all parameters). In the Graz cohort, venous lesions were assessed and associated with severe PCF although no association of any of the venous lesions with long- or short-term prognosis was found.

       Independent predictors of survival and liver-related events

      Clinical, biochemical and histological variables associated with long- or short-term survival on univariable Cox regression are detailed in Table S2. On multivariable Cox regression, sex, model for end-stage liver disease (MELD), platelet count, hepatic encephalopathy, and severe cirrhosis emerged as independent predictors of long-term liver-related mortality (Table 4). In a subgroup of patients in whom follow-up data on abstinence were available (n = 323), sex, MELD, hepatic encephalopathy, abstinence, and severe cirrhosis were independent predictors of long-term liver-related mortality.
      Table 4Clinical, biochemical and histological predictors of outcome in patients with ALD.
      VariableHazard ratio (95% CI)p value
      Predictors of long-term liver-related mortality in the whole cohort (n = 445)
      Severe cirrhosis (SFS 4C)1.88 (1.32-2.68)<0.001
      Male sex0.68 (0.47–0.97)0.036
      MELD1.11 (1.08-1.13)<0.001
      Platelet count0.997 (0.995–-0.999)0.014
      Hepatic encephalopathy1.58 (1.09–2.27)0.015
      Entered variables: severe cirrhosis (SFS 4C), hASH, ductular cholestasis, sex, MELD, WBC, platelet count, hepatic encephalopathy. Maximal VIF = 2.17.
      Predictors of 90-day liver-related mortality in patients with decompensated ALD (n = 286)
      Severe cirrhosis (SFS 4C)2.21 (1.24-3.96)0.008
      hASH present1.98 (1.02–3.85)0.043
      MELD1.19 (1.14–1.24)<0.001
      Hepatic encephalopathy2.44 (1.38–4.30)0.002
      Entered variables: severe cirrhosis (SFS 4C), hASH, ductular cholestasis, sex, MELD, WBC, hepatic encephalopathy. Maximal VIF = 1.64.
      Predictors of 90-day liver-related mortality in decompensated patients with hASH (n = 181)
      Severe cirrhosis (SFS 4C)2.16 (1.11–4.18)0.023
      MELD1.19 (1.13-1.26)<0.001
      Hepatic encephalopathy2.28 (1.20-4.35)0.012
      Entered variables: severe cirrhosis (SFS 4C), canalicular cholestasis, sex, MELD, WBC, platelet count, hepatic encephalopathy. Maximal VIF = 1.51.
      Predictors of decompensation-free survival in patients with compensated ALD (n = 159)
      Severe cirrhosis (SFS 4C)3.26 (1.38–7.69)0.007
      hASH present2.80 (1.32–5.96)0.008
      MELD1.22 (1.05-1.42)0.011
      Albumin0.44 (0.22–0.91)0.026
      Abstinence during follow-up0.33 (0.12–0.91)0.032
      Entered variables: severe cirrhosis (SFS 4C), hASH, age, MELD, albumin, abstinence during follow-up. Maximal VIF = 1.32.
      Multivariable Cox regression.
      ALD, alcohol-related liver disease; hASH, histological steatohepatitis due to ALD; MELD, model for end-stage liver disease; SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe; SFS, SALVE fibrosis stage; VIF, variance inflation factor; WBC, white blood cell count.
      In decompensated ALD, MELD, hepatic encephalopathy, hASH, and severe cirrhosis were independent predictors of short-term (90-day) liver-related mortality. In the subgroup of decompensated patients with hASH, MELD, hepatic encephalopathy, and severe cirrhosis independently predicted 90-day liver-related death.
      In patients with compensated ALD, decompensation-free survival was independently predicted by MELD, albumin, hASH, severe cirrhosis, and abstinence during follow-up.

      Discussion

      The aim of our study was to design and validate an ALD-specific histological grading and staging system that has hitherto been lacking. This study presents the SALVE grading and staging system, a robust histological method with substantial interobserver agreement and clear associations to clinical outcomes in ALD.
      While none of the grading features have been identified as independent prognostic factors for long-term outcome in previous studies,
      • Lackner C.
      • Spindelboeck W.
      • Haybaeck J.
      • Douschan P.
      • Rainer F.
      • Terracciano L.
      • et al.
      Histological parameters and alcohol abstinence determine long-term prognosis in patients with alcoholic liver disease.
      ,
      • Masson S.
      • Emmerson I.
      • Henderson E.
      • Fletcher E.H.
      • Burt A.D.
      • Day C.P.
      • et al.
      Clinical but not histological factors predict long-term prognosis in patients with histologically advanced non-decompensated alcoholic liver disease.
      ballooning, MDBs or lobular neutrophils have been described as independent predictors of short-term mortality in patients with decompensated ALD.
      • Altamirano J.
      • Miquel R.
      • Katoonizadeh A.
      • Abraldes J.G.
      • Duarte-Rojo A.
      • Louvet A.
      • et al.
      A histologic scoring system for prognosis of patients with alcoholic hepatitis.
      ,
      • Katoonizadeh A.
      • Laleman W.
      • Verslype C.
      • Wilmer A.
      • Maleux G.
      • Roskams T.
      • et al.
      Early features of acute-on-chronic alcoholic liver failure: a prospective cohort study.
      ,
      • Mookerjee R.P.
      • Lackner C.
      • Stauber R.
      • Stadlbauer V.
      • Deheragoda M.
      • Aigelsreiter A.
      • et al.
      The role of liver biopsy in the diagnosis and prognosis of patients with acute deterioration of alcoholic cirrhosis.
      The association of hASH with short-term outcome in decompensated patients in our study confirms these results. Histological diagnosis of hASH in patients with decompensated ALD is clinically important to identify those with worse prognosis. Furthermore, in those with decompensation and hASH, short-term prognosis is predicted by clinical factors along with severe cirrhosis. The results emphasize the clinical utility of liver biopsy in these high-risk situations.
      Cholestasis is a distinct feature of severe ALD, which is not described in NAFLD.
      • Yip W.W.
      • Burt A.D.
      Alcoholic liver disease.
      ,
      • Tiniakos D.G.
      Liver biopsy in alcoholic and non-alcoholic steatohepatitis patients.
      Ballooning-associated obstruction of bile radicles,
      • McGill D.B.
      Steatosis, cholestasis, and alkaline phosphatase in alcoholic liver disease.
      as well as impaired bile formation and transport in hepatocytes,
      • Jones A.
      • Selby P.J.
      • Viner C.
      • Hobbs S.
      • Gore M.E.
      • McElwain T.J.
      Tumour necrosis factor, cholestatic jaundice, and chronic liver disease.
      may be involved in canalicular cholestasis. Defective bile secretion via canalicular transporters and/or decreased bile flow have been implicated in sepsis-associated cholestasis.
      • Geier A.
      • Fickert P.
      • Trauner M.
      Mechanisms of disease: mechanisms and clinical implications of cholestasis in sepsis.
      ,
      • Lefkowitch J.H.
      Bile ductular cholestasis: an ominous histopathologic sign related to sepsis and "cholangitis lenta.
      Bacterial infection and sepsis-associated immune paralysis are frequent in advanced ALD and often fatal complications triggering acute-on-chronic liver failure.
      • Wasmuth H.E.
      • Kunz D.
      • Yagmur E.
      • Timmer-Stranghoner A.
      • Vidacek D.
      • Siewert E.
      • et al.
      Patients with acute on chronic liver failure display "sepsis-like" immune paralysis.
      Ductular cholestasis has been associated with evolving (subclinical) sepsis and thus may indicate infection at an early stage.
      • Altamirano J.
      • Miquel R.
      • Katoonizadeh A.
      • Abraldes J.G.
      • Duarte-Rojo A.
      • Louvet A.
      • et al.
      A histologic scoring system for prognosis of patients with alcoholic hepatitis.
      ,
      • Katoonizadeh A.
      • Laleman W.
      • Verslype C.
      • Wilmer A.
      • Maleux G.
      • Roskams T.
      • et al.
      Early features of acute-on-chronic alcoholic liver failure: a prospective cohort study.
      ,
      • Mookerjee R.P.
      • Lackner C.
      • Stauber R.
      • Stadlbauer V.
      • Deheragoda M.
      • Aigelsreiter A.
      • et al.
      The role of liver biopsy in the diagnosis and prognosis of patients with acute deterioration of alcoholic cirrhosis.
      Both canalicular and ductular cholestasis convey prognostic information in the whole study cohort, in the decompensated subgroup as well as in decompensated patients with hASH. This supports similar results of others
      • Lefkowitch J.H.
      Bile ductular cholestasis: an ominous histopathologic sign related to sepsis and "cholangitis lenta.
      ,
      • Nissenbaum M.
      • Chedid A.
      • Mendenhall C.
      • Gartside P.
      Prognostic significance of cholestatic alcoholic hepatitis. VA Cooperative Study Group #119.
      and underscores the prognostic utility of morphological cholestasis as an integral factor of SALVE grading.
      Since most patients with ALD exhibit severe fibrosis or cirrhosis at first diagnosis, any stepwise staging method should allow sub-classification into prognostically meaningful categories. Based on the SALVE staging system and Kaplan-Meier analyses, 2 cirrhosis substages with different mortality risk could be defined. Patients with severe cirrhosis (SFS 4C) had worse outcome than patients with lesser grades of cirrhosis (SFS 4A and 4B). Moreover, severe cirrhosis emerged as an independent histological predictor for both long- and short-term survival in the whole study cohort, in subgroups of patients with symptomatic/decompensated disease as well as in patients with hASH. Our data are thus in line with results from earlier studies, indicating the prognostic utility of cirrhosis substaging in chronic liver disease.
      • Kim S.U.
      • Oh H.J.
      • Wanless I.R.
      • Lee S.
      • Han K.H.
      • Park Y.N.
      The Laennec staging system for histological sub-classification of cirrhosis is useful for stratification of prognosis in patients with liver cirrhosis.
      ,
      • Kim M.Y.
      • Cho M.Y.
      • Baik S.K.
      • Park H.J.
      • Jeon H.K.
      • Im C.K.
      • et al.
      Histological subclassification of cirrhosis using the Laennec fibrosis scoring system correlates with clinical stage and grade of portal hypertension.
      PCF can be a striking feature in ALD. Interestingly, in contrast to SFS with predominant septal fibrosis, (like SFS 3 or 4A/B/C), severe PCF (SFS 3P or 4AP/BP/CP) was associated with morphological features of liver injury, inflammation, fibro-obliterative venous lesions, cholestasis, and significantly shorter long-term outcome. The PCF pattern may indicate active disease and ongoing fibrogenesis. It may represent an immature type of fibrosis because it lacks elastic fibres and clusterin, a potent inhibitor of matrix-degrading metalloproteinases
      • Jeong S.
      • Ledee D.R.
      • Gordon G.M.
      • Itakura T.
      • Patel N.
      • Martin A.
      • et al.
      Interaction of clusterin and matrix metalloproteinase-9 and its implication for epithelial homeostasis and inflammation.
      present in mature scar tissue.
      • Aigelsreiter A.
      • Janig E.
      • Sostaric J.
      • Pichler M.
      • Unterthor D.
      • Halasz J.
      • et al.
      Clusterin expression in cholestasis, hepatocellular carcinoma and liver fibrosis.
      Data from rodent models of cirrhosis suggest that recent fibrous septa are readily degraded whereas older septa rich in elastin are more resistant.
      • Pellicoro A.
      • Ramachandran P.
      • Iredale J.P.
      Reversibility of liver fibrosis.
      Therefore, it could be speculated that PCF is more sensitive to degradation than dense fibrotic septa. Regression of PCF observed in a paired biopsy on follow-up could indicate decreased or resolved liver injury in phases of abstinence or medical intervention and could be useful to trace early antifibrotic treatment effects to monitor the evolution of fibrosis. Specific stages – SFS 3P, 4AP, 4BP and 4CP – were introduced in the SALVE staging system as an option to identify cases in which PCF is the predominant fibrosis type and to define these stages, which are not currently represented in other staging systems. If applied with the help of the staging algorithm, interrater agreement is substantial.
      Our study has some limitations related to its retrospective design. Although a large panel of expert hepatopathologists reached substantial interobserver agreement, the proposed system should also be validated in a general pathology setting.
      In conclusion, the SALVE histopathology group developed and validated an ALD-specific grading and staging system based on a large cohort of patients representing the whole clinical spectrum of ALD. This histological system integrates features of disease activity and fibrosis in a prognostic context. The large patient number has enabled us to evaluate the applicability and prognostic utility of SALVE grading and staging in clinically important subgroups, with compensated and decompensated disease, as well as in patients with histological ASH. Activity scores can be used to define the severity of injury and inflammation and to diagnose steatohepatitis and may, along with SALVE stage, help guide patient management.

       Abbreviations

      ALD, alcohol-related liver disease; hASH, histological steatohepatitis due to ALD; MDB, Mallory-Denk body; PCF, pericellular fibrosis; SALVE, Consortium for the Study of Alcohol-related LiVer disease in Europe; SFS, SALVE fibrosis stage; SHG, SALVE Histopathology Group.

      Financial support

      The authors received no financial support to produce this manuscript.

      Authors’ contributions

      CL, RES, HCP, CM, ASHG, DT: study design, data evaluation, manuscript writing and final review; CL, SD, HD, HPD, VG, MG, RM, VP, PS, LT, IR, SD, SS, ASHG, DT: review of histologic slides and final review; CL, RES, AB, GP: data management, statistical analysis and final review; RES, LB, PD, FR, PER, MJ, HS, AB, AH, MT, AK: patient recruitment, data collection and final review.

      Data availability statement

      The data that support the findings of this study are available from the corresponding author, upon reasonable request.

      Conflicts of interest

      The authors declare no conflicts of interest that pertain to this work.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Acknowledgement

      The authors thank Drs. Elizabeth M. Brunt and Ian R. Wanless for fruitful discussion of the ALD-specific staging system.

      Supplementary data

      The following are the supplementary data to this article:

      References

        • Yip W.W.
        • Burt A.D.
        Alcoholic liver disease.
        Semin Diagn Pathol. 2006; 23: 149-160
        • European Association for the Study of the Liver
        EASL Clinical Practice Guidelines: Management of Alcohol-Related Liver Disease.
        J Hepatol. 2018; 69: 154-181
        • Lucey M.R.
        • Mathurin P.
        • Morgan T.R.
        Alcoholic hepatitis.
        N Engl J Med. 2009; 360: 2758-2769
        • Shen N.T.
        • Salajegheh A.
        • Brown Jr., R.S.
        A call to standardize definitions, data collection, and outcome assessment to improve care in alcohol-related liver disease.
        Hepatology. 2019; 70: 1038-1044
        • Kleiner D.E.
        • Brunt E.M.
        • Van Natta M.
        • Behling C.
        • Contos M.J.
        • Cummings O.W.
        • et al.
        Design and validation of a histological scoring system for nonalcoholic fatty liver disease.
        Hepatology. 2005; 41: 1313-1321
        • Bedossa P.
        • FLIP Pathology Consortium
        Utility and appropriateness of the fatty liver inhibition of progression (FLIP) algorithm and steatosis, activity, and fibrosis (SAF) score in the evaluation of biopsies of nonalcoholic fatty liver disease.
        Hepatology. 2014; 60: 565-575
        • Ishak K.
        • Baptista A.
        • Bianchi L.
        • Callea F.
        • De Groote J.
        • Gudat F.
        • et al.
        Histological grading and staging of chronic hepatitis.
        J Hepatol. 1995; 22: 696-699
        • Bedossa P.
        • Poynard T.
        An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group.
        Hepatology. 1996; 24: 289-293
        • Lefkowitch J.H.
        Morphology of alcoholic liver disease.
        Clin Liver Dis. 2005; 9: 37-53
        • Brunt E.M.
        • Neuschwander-Tetri B.A.
        • Burt A.D.
        Fatty liver disease: alcoholic and non-alcoholic.
        in: Burt A.D. Portmann B. Ferrel L. MacSween´s Pathology of the Liver. Churchill Livingstone, Edinbourgh2012: 293-359
        • Lackner C.
        • Spindelboeck W.
        • Haybaeck J.
        • Douschan P.
        • Rainer F.
        • Terracciano L.
        • et al.
        Histological parameters and alcohol abstinence determine long-term prognosis in patients with alcoholic liver disease.
        J Hepatol. 2017; 66: 610-618
        • Goh G.B.
        • Pagadala M.R.
        • Dasarathy J.
        • Unalp-Arida A.
        • Sargent R.
        • Hawkins C.
        • et al.
        Clinical spectrum of non-alcoholic fatty liver disease in diabetic and non-diabetic patients.
        BBA Clin. 2014; 3: 141-145
        • Tsochatzis E.
        • Bruno S.
        • Isgro G.
        • Hall A.
        • Theocharidou E.
        • Manousou P.
        • et al.
        Collagen proportionate area is superior to other histological methods for sub-classifying cirrhosis and determining prognosis.
        J Hepatol. 2014; 60: 948-954
        • Kim S.U.
        • Oh H.J.
        • Wanless I.R.
        • Lee S.
        • Han K.H.
        • Park Y.N.
        The Laennec staging system for histological sub-classification of cirrhosis is useful for stratification of prognosis in patients with liver cirrhosis.
        J Hepatol. 2012; 57: 556-563
        • Altamirano J.
        • Miquel R.
        • Katoonizadeh A.
        • Abraldes J.G.
        • Duarte-Rojo A.
        • Louvet A.
        • et al.
        A histologic scoring system for prognosis of patients with alcoholic hepatitis.
        Gastroenterology. 2014; 146 (1231-1239.e1-6)
        • Burt A.D.
        • MacSween R.N.
        Hepatic vein lesions in alcoholic liver disease: retrospective biopsy and necropsy study.
        J Clin Pathol. 1986; 39: 63-67
        • Nakano M.
        • Worner T.M.
        • Lieber C.S.
        Perivenular fibrosis in alcoholic liver injury: ultrastructure and histologic progression.
        Gastroenterology. 1982; 83: 777-785
        • Thiele M.
        • Detlefsen S.
        • Sevelsted Moller L.
        • Madsen B.S.
        • Fuglsang Hansen J.
        • Fialla A.D.
        • et al.
        Transient and 2-dimensional shear-wave elastography provide comparable assessment of alcoholic liver fibrosis and cirrhosis.
        Gastroenterology. 2016; 150: 123-133
        • Thiele M.
        • Madsen B.S.
        • Hansen J.F.
        • Detlefsen S.
        • Antonsen S.
        • Krag A.
        Accuracy of the enhanced liver fibrosis test vs FibroTest, elastography, and indirect markers in detection of advanced fibrosis in patients with alcoholic liver disease.
        Gastroenterology. 2018; 154: 1369-1379
        • Katoonizadeh A.
        • Laleman W.
        • Verslype C.
        • Wilmer A.
        • Maleux G.
        • Roskams T.
        • et al.
        Early features of acute-on-chronic alcoholic liver failure: a prospective cohort study.
        Gut. 2010; 59: 1561-1569
        • Spahr L.
        • Rubbia-Brandt L.
        • Genevay M.
        • Hadengue A.
        • Giostra E.
        Early liver biopsy, intraparenchymal cholestasis, and prognosis in patients with alcoholic steatohepatitis.
        BMC Gastroenterol. 2011; 11: 115
        • Altamirano J.
        • Bataller R.
        Alcoholic liver disease: pathogenesis and new targets for therapy.
        Nat Rev Gastroenterol Hepatol. 2011; 8: 491-501
        • Masson S.
        • Emmerson I.
        • Henderson E.
        • Fletcher E.H.
        • Burt A.D.
        • Day C.P.
        • et al.
        Clinical but not histological factors predict long-term prognosis in patients with histologically advanced non-decompensated alcoholic liver disease.
        Liver Int. 2014; 34: 235-242
        • Mookerjee R.P.
        • Lackner C.
        • Stauber R.
        • Stadlbauer V.
        • Deheragoda M.
        • Aigelsreiter A.
        • et al.
        The role of liver biopsy in the diagnosis and prognosis of patients with acute deterioration of alcoholic cirrhosis.
        J Hepatol. 2011; 55: 1103-1111
        • Tiniakos D.G.
        Liver biopsy in alcoholic and non-alcoholic steatohepatitis patients.
        Gastroenterol Clin Biol. 2009; 33: 930-939
        • McGill D.B.
        Steatosis, cholestasis, and alkaline phosphatase in alcoholic liver disease.
        Am J Dig Dis. 1978; 23: 1057-1060
        • Jones A.
        • Selby P.J.
        • Viner C.
        • Hobbs S.
        • Gore M.E.
        • McElwain T.J.
        Tumour necrosis factor, cholestatic jaundice, and chronic liver disease.
        Gut. 1990; 31: 938-939
        • Geier A.
        • Fickert P.
        • Trauner M.
        Mechanisms of disease: mechanisms and clinical implications of cholestasis in sepsis.
        Nat Clin Pract Gastroenterol Hepatol. 2006; 3: 574-585
        • Lefkowitch J.H.
        Bile ductular cholestasis: an ominous histopathologic sign related to sepsis and "cholangitis lenta.
        Hum Pathol. 1982; 13: 19-24
        • Wasmuth H.E.
        • Kunz D.
        • Yagmur E.
        • Timmer-Stranghoner A.
        • Vidacek D.
        • Siewert E.
        • et al.
        Patients with acute on chronic liver failure display "sepsis-like" immune paralysis.
        J Hepatol. 2005; 42: 195-201
        • Nissenbaum M.
        • Chedid A.
        • Mendenhall C.
        • Gartside P.
        Prognostic significance of cholestatic alcoholic hepatitis. VA Cooperative Study Group #119.
        Dig Dis Sci. 1990; 35: 891-896
        • Kim M.Y.
        • Cho M.Y.
        • Baik S.K.
        • Park H.J.
        • Jeon H.K.
        • Im C.K.
        • et al.
        Histological subclassification of cirrhosis using the Laennec fibrosis scoring system correlates with clinical stage and grade of portal hypertension.
        J Hepatol. 2011; 55: 1004-1009
        • Jeong S.
        • Ledee D.R.
        • Gordon G.M.
        • Itakura T.
        • Patel N.
        • Martin A.
        • et al.
        Interaction of clusterin and matrix metalloproteinase-9 and its implication for epithelial homeostasis and inflammation.
        Am J Pathol. 2012; 180: 2028-2039
        • Aigelsreiter A.
        • Janig E.
        • Sostaric J.
        • Pichler M.
        • Unterthor D.
        • Halasz J.
        • et al.
        Clusterin expression in cholestasis, hepatocellular carcinoma and liver fibrosis.
        Histopathology. 2009; 54: 561-570
        • Pellicoro A.
        • Ramachandran P.
        • Iredale J.P.
        Reversibility of liver fibrosis.
        Fibrogenesis Tissue Repair. 2012; 5: S26

      Linked Article