Article Text

Acute presentation of autoimmune hepatitis: a multicentre study with detailed histological evaluation in a large cohort of patients
  1. Hiep Nguyen Canh1,
  2. Kenichi Harada1,
  3. Hirofumi Ouchi1,
  4. Yasunori Sato1,
  5. Koichi Tsuneyama2,
  6. Masayoshi Kage3,
  7. Masayuki Nakano4,
  8. Kaname Yoshizawa5,
  9. Atsushi Takahashi6,
  10. Masanori Abe7,
  11. Jong-Hon Kang8,
  12. Kazuhiko Koike9,
  13. Ayano Inui10,
  14. Tomoo Fujisawa10,
  15. Akinobu Takaki11,
  16. Teruko Arinaga-Hino12,
  17. Takuji Torimura12,
  18. Yoshiyuki Suzuki13,
  19. Keiichi Fujiwara14,
  20. Mikio Zeniya15,
  21. Hiromasa Ohira6,
  22. Atsushi Tanaka16,
  23. Hajime Takikawa16
  24. Intractable Liver and Biliary Diseases Study Group of Japan
  1. 1Department of Human Pathology, Kanazawa University Graduate School of Medicine, Kanazawa, Japan
  2. 2Department of Pathology Lab. Med., Tokushima University, Graduate School, Tokushima, Japan
  3. 3Kurume University Research Center for Innovative Cancer Therapy, Kurume, Japan
  4. 4Division of Diagnostic Pathology, Shonan Fujisawa Tokushukai Hospital, Fujisawa, Japan
  5. 5Department of Gastroenterology, National Hospital Organization, Shinshu Ueda Medical Center, Ueda, Japan
  6. 6Department of Gastroenterology, Fukushima Medical University School of Medicine, Fukushima, Japan
  7. 7Department of Gastroenterology and Metabiology, Ehime University Graduate School of Medicine, Ehime, Japan
  8. 8Center of Gastroenterology, Teine Keijinkai Hospital, Sapporo, Japan
  9. 9Department of Gastroenterology and Hepatology, The Third Hospital of Jikei University School of Medicine, Tokyo, Japan
  10. 10Department of Pediatric Hepatology and Gastroenterology, Eastern Yokohama Hospital, Yokohama, Japan
  11. 11Department of Gastroenterology and Hepatology, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Okayama, Japan
  12. 12Department of Medicine, Kurume University School of Medicine, Kurume, Japan
  13. 13Department of Hepatology, Toranomon Hospital, Tokyo, Japan
  14. 14Department of Gastroenterology and Nephrology, Graduate School of Medicine, Chiba University, Chiba, Japan
  15. 15Sanno Medical Center, International University of Health and Welfare, Tokyo, Japan
  16. 16Department of Medicine, Teikyo University School of Medicine, Tokyo, Japan
  1. Correspondence to Dr Kenichi Harada, Department of Human Pathology, Kanazawa University Graduate School of Medicine, Kanazawa 920-8640, Japan; kenichih{at}med.kanazawa-u.ac.jp

Abstract

Aims Although liver biopsy is crucial to diagnose and guide treatment decisions, a detailed histological analysis of autoimmune hepatitis (AIH) with clinically acute presentations has not yet been performed. This study aimed to characterise the histological features and explore potential histological hallmarks to diagnose the acute presentation of AIH.

Methods We systematically evaluated liver specimens of 87 adult patients with acute presentation of AIH retrospectively enrolled from Japanese multicentre facilities. Each histological feature was predefined by consensus based on the diagnostic criteria.

Results Key findings were that acute presentation of AIH revealed histological features of both acute hepatitis and chronic hepatitis accompanying various degrees of fibrosis. The prominent features were lobular necrosis/inflammation (97.7%), plasma cell infiltration (96.4%), emperipolesis (89.3%), pigmented macrophages (84.5%), cobblestone appearance of hepatocytes (82.6%) and perivenular necroinflammatory activity, including centrilobular necrosis (81.4%).

Conclusions The acute presentation of AIH represents the entire histological spectrum of acute hepatitis and chronic hepatitis with various activity grades and fibrosis stages that clinically correspond to acute-onset AIH and acute exacerbation of classic AIH, respectively. Although there are no pathognomonic features for the pathological diagnosis, the prominent presence of lobular and perivenular necroinflammatory activity, pigmented macrophages and cobblestone appearance of hepatocytes in addition to the classic AIH features, such as plasma cell infiltration and emperipolesis, are useful for the pathological diagnosis of the acute presentation of AIH.

  • AUTOIMMUNITY
  • HEPATITIS
  • LIVER DISEASE
  • SURGICAL PATHOLOGY

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Autoimmune hepatitis (AIH) is an immune-mediated necroinflammatory liver disorder characterised by autoantibodies, hypergammaglobulinemia, and interface hepatitis with lymphoplasmacytic infiltration on histological examination.1–3 Although defined as a chronic disease,4 AIH can have an acute presentation at its onset that mimics acute hepatitis of other causes (8.7–75%).5–20 In Japanese nationwide surveys conducted in 2009 and 2015, 10.9% and 11.7%, respectively, of patients with AIH had a histological diagnosis of acute hepatitis.21 ,22 AIH may also present as acute severe hepatitis or even acute liver failure.16 ,23–30 Acute presentation of AIH may encompass two different clinical entities:30–34 (1) the acute exacerbation of underlying chronic AIH (acute exacerbation phase of classic AIH) with histological evidence of chronic hepatitis; and (2) the genuine, newly developed, acute-onset AIH (acute hepatitis phase of AIH) without histopathological findings of chronic liver diseases.6 ,7 ,20 ,24 ,28 ,29 ,34–39

Early diagnosis of AIH with an acute presentation remains challenging but urgent. There is no single diagnostic test for AIH. The diagnostic criteria for classic AIH proposed by the International Autoimmune Hepatitis Group have been standardised by consensus and include clinical, serological, biochemical and histological findings, as well as the absence of data suggestive of other diseases.40 ,41 However, some patients with acute presentation of AIH, atypical features such as normal serum immunoglobulin G (IgG) levels and/or negativity of autoantibody usually do not meet the diagnostic criteria.12 ,14 ,24 ,36 ,37 ,40–43 AIH showing an acute presentation may progress to severe hepatitis or acute liver failure with resistance to immunosuppressive therapy and catastrophic outcomes; thus, timely diagnosis is the most important factor for improving the poor prognosis of these patients.7 ,18 ,28 ,44–46 Liver biopsy evaluation is mandatory for precise, early diagnosis, but the histology of AIH with acute presentations is heterogeneous and may not reveal typical histological findings of classic AIH.30 ,41 ,47–49 Centrilobular necrosis (CN) has been repeatedly reported in AIH patients with acute presentations.6 ,16 ,34–36 ,50–58 In particular, CN with plasma cell infiltration is considered to be characteristic for AIH with acute presentations, but its significance remains to be elucidated; the centrilobular lesion with none to mild portal inflammation, without interface hepatitis and fibrosis, may represent characteristics of early AIH.50 ,55–57 However, histological features of AIH with an acute presentation showed wide-ranging divergences among previous studies that were based on case reports, small cohorts of patients or those that mainly focused on clinical characteristics. Some found no significant difference in clinical laboratory features between AIH with an acute presentation and that with a chronic presentation.5 ,42 Further, a detailed histological analysis of acute presentation of AIH has not been performed.

We hypothesised that the acute presentation of AIH should show a wide spectrum of histological features, which should change time dependently, including that of acute hepatitis, convalescent stages from acute hepatitis and acute exacerbation of chronic AIH. Herein, we performed a detailed histological evaluation on liver biopsies from a large cohort of patients with acute presentation of AIH. Our specific aims were to characterise the histological features and explore potential histological hallmarks to diagnose the acute presentation of AIH.

Methods

Patients

The Autoimmune Hepatitis Group, a subgroup of the Intractable Hepato-Biliary Disease Study Group in Japan, conducted a multicentre retrospective survey of patients with AIH who were admitted to eight tertiary hospitals with liver specialists, from 2005 to 2014. Overall, 87 adult patients with acute presentations of AIH, selected from the survey, were eligible for this study. The acute presentation of AIH was diagnosed based on the revised original criteria of the International AIH Group (1999) defining the score for definite or probable AIH40 and/or on an empirical judgment by experienced hepatologists on the basis of combining clinical, laboratory, serological and histological data. All patients showed response to immunosuppressive therapy, with a mean follow-up duration of 41±28.6 (1–115) months. Other aetiologies of liver diseases were strictly ruled out. The acute presentation of AIH was tentatively defined as AIH with (1) acute onset symptoms in conjunction with serum bilirubin levels of more than 5 mg/dL and/or serum alanine aminotransferase (ALT) levels of more than 10-fold the normal limit; and (2) no history of any prior liver disease.7 Patients with an overlap syndrome (overlap of AIH with other diseases, such as primary biliary cholangitis, primary sclerosing cholangitis, autoimmune cholangitis) or a coexistent liver disease were excluded from this analysis.

Clinical and laboratory data

The demographic and biochemical data collected from patients at the time of diagnosis were as follows: age, sex, time from onset to liver biopsy, bilirubin, aspartate aminotransferase (AST), ALT, alkaline phosphatase (ALP), IgG, anti-nuclear antibody (ANA), anti-smooth muscle antibody (ASMA) and anti-mitochondrial antibody (AMA). This study was approved by the Fukushima Medical University Ethics Committee (no. 2099).

Histological evaluation

Liver biopsy was performed before treatment, within 6 months from the onset of symptoms. Tissues from explants or autopsies were excluded. Biopsy specimens were stained with at least H&E and connective tissue stainings such as Masson's trichrome and reticulin stainings. Five experienced pathologists reviewed all biopsy slides, using the same histological scoring sheet as in table 1, in which each histological feature was predefined by consensus based on the diagnostic criteria. Histological features evaluated on each biopsy were divided into two board categories: (1) lobular (parenchymal) changes including CN (collapse, lytic type), centrilobular congestion/haemorrhage, centrilobular endothelial cell injury/central vein endotheliitis, cobblestone appearance of hepatocytes, hepatic rosette formation in areas of cobblestone appearance, lobular pigmented macrophages (in CN, intralobular areas) and centriloblar fibrosis; (2) portal/periportal changes, including portal inflammation, interface hepatitis, portal plasma cell infiltration, hepatic rosette formation in periportal areas, bile duct injury and fibrosis. The features were noted as absent (−), present (+) or scored on a semiquantitative scale based on the severity of the lesion (0=none, 1=mild, 2=moderate and 3=severe) or other characterised as specified. The portal-based fibrosis was scored using the Batts-Ludwig scheme (0, no fibrosis; 1, portal fibrosis; 2, periportal fibrosis; 3, septal fibrosis; and 4, cirrhosis).59

Table 1

Histological features evaluated in this study

Centrilobular fibrosis was recorded as none (no fibrosis), mild (fibrosis limited to perivenular areas, requiring collagen/fibre stain to detect), moderate (fibrosis limited to perivenular areas, visible on H&E staining), or severe (scar-like, board fibrosis). Plasma cell infiltration in lobular or portal/periportal areas was recorded as absent, mild (only occasional cells were identified), moderate (easily recognised plasma cells that were a minority of the inflammatory cell infiltrates) or severe (numerous plasma cells present in groups and sheets in the portal tracts or as groups in the sinusoids).60 Cobblestone appearance of hepatocytes was defined as clusters of small, monomorphic hepatocytes with typically clear hydropic cytoplasm that tend to efface the liver cord arrangement, usually locate around the portal tracts or adjacent to fibrous septa, reflecting increased regenerative activity (figure 1).61 The presence and the extent of cobblestone appearance were assessed as none, mild (≤10% of the parenchyma affected), moderate (10–30% of the parenchyma affected), severe (>30% of parenchyma affected). Bile duct injury was assessed as none (score 0), mild or moderate injury (score 1) and severe injury (score 2) as previously described.62 Cholestasis was evaluated as none (no visible bile), mild (bile identified only after careful search under high magnification), moderate (bile not immediately evident at low magnification but easily found at high magnification, severe (bile easily found at low magnification).63 A modified Ishak scoring system was used to evaluate interface hepatitis, lobular necrosis/inflammation and portal inflammation.64 Interface hepatitis was recorded as absent, mild (focal, involving some or most of portal tracts), moderate (circumferential, involving minority, <50%, of portal tracts or septa) and severe (circumferential, involving majority of portal tracts or septa). Lobular necrosis/inflammation was assessed as absent, mild (few foci of necrosis/apoptosis/inflammation per 10× objective), moderate (5–10 foci) and severe (>10 foci and/or confluent necrosis). Portal inflammation was recorded as absent, mild (few inflammatory cells, involving some or most portal tracts), severe (marked inflammatory cells, involving some or most portal tracts) and moderate (moderate inflammatory cells, between the two entities).

Figure 1

Examples of some histological features of autoimmune hepatitis with an acute presentation. (A) Centrilobular necrosis (CN) without the involvement of portal tracts (P). (B) Higher magnification of the panel (A) reveals a typical CN with confluent hepatocellular necrosis/dropout and mononuclear inflammation including some plasma cells (arrowheads) around a central vein (CV). (C) The right half of the hepatic parenchyma, from the fibrous septa with a hepatic rosette (arrow), represents the cobblestone appearance with small and hydropic hepatocytes. (D) Hepatic rosette formation (arrow). There is also the presence of CN with centrilobular haemorrhage. (E) Perivenular necroinflammatory activity. (F) Emperipolesis (arrow) (H&E (A–F), original magnifications: (A) ×50, (B–E) ×200, (F) ×400).

Findings other than features listed in table 1 were also recorded. After the histological evaluation, cases with disagreement among the five pathologists were discussed. The final histological diagnosis was established based on consensual decision making by all pathologists. CN was defined as confluent hepatocellular necrosis in the centrilobular area or predominantly affecting the centrilobular area (figure 1). Emperipolesis was defined as engulfment of lymphocytes by hepatocytes; the engulfed lymphocytes morphologically remain intact and are surrounded by hepatocyte cytoplasm with a halo (figure 1). Engulfment of plasma cells by hepatocytes was excluded from this analysis.41 ,47 Perivenular necroinflammatory activity was defined as centrilobular injury, other than CN, with prominent hepatocellular necrosis and mononuclear inflammation, including spotty/focal necrosis (figure 1).52 ,65

Statistical analysis

Data are reported as mean±SD or median (range) for continuous variables and proportion/frequency for dichotomous variables. Statistical analysis was conducted with SPSS Statistic for Mac OS V.18.0 (SPSS, Chicago, Illinois, USA). If there were missing values, the analysis was performed with available data.

Results

Patient characteristics

Clinical and laboratory features of AIH with acute presentations are summarised in table 2. Eighty-seven patients with AIH who had acute presentations of the disease were enrolled. Patients were predominantly female (female:male ratio, 6:1), with a mean age of 54.5±15.3 (range, 16–84) years.

Table 2

Clinical and serological features of patients with acute presentation of AIH at initial liver function tests

The median days from the onset of symptoms to biopsy was 21 (range 1–180) days, and 85.1% of the biopsy specimens were taken within 3 months from the onset of symptoms. The mean (±SD) serum IgG level at onset was 1984.7±820 mg/dL. Almost 78% of patients tested positive for ANA and/or ASMA, 74.4% for ANA and 5.5% for AMA.

Histological findings

Histological features of AIH with acute presentations revealed various patterns of hepatic injury and are summarised in table 3.

Table 3

Histological features of autoimmune hepatitis with an acute presentation

The spectrum of histological features ranged from early stages, mild acute lobular hepatitis with virtually unaffected portal and periportal areas (figure 2), through chronic hepatitis with severe activity and negligible fibrosis (figure 3) to severe necroinflammatory lesions in chronic hepatitis with advanced stages of fibrosis (stage 3) (figure 4). However, lobular necrosis/inflammation, plasma cell infiltration in lobular and/or portal areas, emperipolesis, pigmented macrophages, cobblestone appearance of hepatocytes, perivenular necroinflammatory activity and/or CN were the most striking features of AIH with acute presentations. The presence (+) or score ≥1 of each finding was evaluated as a positive one in calculating each value. The frequency of plasma cell infiltration in lobular and/or portal areas was high at 96.4%. Approximately 90% of patients showed emperipolesis, 84.5% pigmented macrophages (regardless of the areas), and 81.4% perivenular necroinflammatory and/or CN. CN was usually observed, but there was no significant difference in the frequency between collapse (39.5%) and lytic (44.0%) types.

Figure 2

A histological spectrum of early-stage autoimmune hepatitis with an acute presentation. (A–B) Case no. 47. A biopsy performed 10 days after presentation reveals lobular changes with negligible perivenular injury (A) and mild portal inflammation (A–B); high-power view of portal inflammation (B) showing mixed inflammatory infiltrates with prominent lymphocytes, plasma cells (arrowheads) and pigmented macrophages. (C, D) Case no. 38. Prominent perivenular necroinflammatory activity with virtually uninvolved portal tracts (P). Areas around the central veins (CV) reveal focal (spotty) necrosis marked by the presence of pigmented macrophages and lymphocytes (D); the nearby portal tract is intact (D). (E, F) A representative case showing histological features of a convalescent stage from acute hepatitis with virtually sparing portal and centrilobular areas. The hepatic parenchyma shows minimal injuries (F) consisting of emperipolesis (arrow) and focal necrosis highlighted by pigmented macrophages (arrowhead) (case no. 35, the biopsy performed 142 days after the onset of symptoms). (H&E (A–F), original magnifications: (A, C, E) ×50, (B, D) ×200, (F) ×400).

Figure 3

Histopathology of autoimmune hepatitis with acute presentations. Low magnification shows severe portal inflammation and marked lobular necroinflammatory changes (A). (B) No apparent portal fibrosis is present (F0). (C) Severe portal inflammatory infiltrates with interface hepatitis contains many plasma cells (arrowheads). (D) The lobular area shows small clusters of mononuclear inflammatory cells, including some plasma cells (arrowheads), indicating sites of hepatocyte necrosis (H&E (A, C, D), Masson's trichrome stain (B), original magnifications: (A) ×50, (B) ×100, (C–D) ×400.)

Figure 4

Histopathology of autoimmune hepatitis with acute presentations. (A) A portal tract is expanded with mononuclear inflammatory cell infiltrates and bridging fibrosis highlighted by Masson's trichrome stain (C). (B) Marked portal enlargement and massive hepatocellular necrosis with severe inflammatory infiltration and bile duct proliferation. (D) Higher magnification shows hydropic degeneration of hepatocytes and lobular plasma cell infiltration (arrowheads). (H&E (A, B, D), original magnifications: (A–C) ×50, (D) ×400.)

Perivenular necroinflammatory activity was more frequent than CN. Portal inflammation was frequently accompanied by plasma cell infiltration (91.6% of all cases). The proportion of interface hepatitis was 75%, and hepatic rosette formation (in cobblestone and/or periportal areas) was 60.5%. Other common findings were centrilobular endothelial cell injury/central vein endotheliitis, fibrosis (stage 1–3, 67.4%), centrilobular fibrosis and centrilobular congestion/haemorrhage. Bile duct injury was commonly seen, but only 7/84 (8.3%) cases showed severe damage. Cholestasis was present in 12 (14%) cases (8 cases of mild cholestasis, 4 cases of moderate cholestasis). No case of cirrhosis was observed.

Discussion

To date, no standardised definition or acceptable diagnostic criteria have been established for the acute presentation of AIH, although this entity has long been recognised.10 ,66 Liver biopsy is crucial to diagnose AIH and guide treatment decisions. However, the morphological spectrum of AIH with an acute presentation has not yet been clarified and remains controversial. This study aimed to address the issue by systematically evaluating liver biopsies from a large cohort of patients with acute presentations of AIH. To our knowledge, this is the largest performed series to date. Key findings were that AIH with acute presentations represented a broad histological spectrum of both acute hepatitis and chronic hepatitis, ranging from mild acute lobular hepatitis to severe inflammatory lesions in chronic hepatitis with advanced stage of fibrosis, but without cirrhosis. The most striking features were lobular necrosis/inflammation, plasma cell infiltration in lobular and/or portal areas, emperipolesis, pigmented macrophage, cobblestone appearance of hepatocytes and perivenular necroinflammatory activity, including CN. The present data revealed various histological patterns of AIH with acute presentations reported in earlier studies, in which histological features of acute hepatitis, including CN, submassive necrosis and massive necrosis, were present in 12.2–88% cases.6 ,7 ,20 ,24 ,28 ,29 ,34–39 However, as mentioned, individual histological features were less specified and their proportions were widely different among previous reports.5–7 ,16 ,24 ,27–29 ,35 ,36 ,38 ,42

Particular attention was paid to CN. In earlier reports,16 ,22 ,34–36 ,52 ,67 CN frequency varied from 21.8% to 100% of AIH patients with acute presentations compared with 57% in the present study. The reason for this divergence is unclear, but several feasible explanations can be proposed, such as patient populations with different diagnostic criteria, differences in disease severities, differences in the histological definition of CN among reports and/or natural heterogeneity of the disease. CN has been more frequently found in the acute hepatitis phase (100%) than in the acute exacerbation phase of AIH (42%).34 Its prevalence also increases with disease severity.37 Thus, patient populations with a higher rate of severe hepatitis or fulminant hepatitis cases may have a higher frequency of CN.34 ,37 Another reason is inconsistency in the diagnostic criteria of CN.52 Some prior studies regarded both confluent necrosis (defined as CN in the present study) and spotty (focal) necrosis in centrilobular areas as CN.16 ,54 Our data actually showed that the rate of CN and/or perivenular necroinflammatory activity was also high (81.4%).

The significance of CN in AIH remains to be clarified, but numerous data support the concept that CN is a histological characteristic of AIH with acute presentations. Indeed, our material revealed some cases with CN and virtually unaffected portal tracts (figure 1). Interestingly, some other cases showed lobular hepatitis with mild portal inflammation without CN or demonstrated lobular hepatitis without an involvement of centrilobular and/or portal areas (figure 2). These cases represented a wide histological spectrum of early-stage AIH with variously combining lesions in lobular, centrilobular and portal areas, including features of the convalescent stage of acute hepatitis. Furthermore, CN was more often observed in AIH with an acute presentation than in that with a chronic presentation (53% vs 17.5–30.7%).22 ,34 ,51 ,53 ,68 Hofer et al51 reported that patients with AIH and CN more frequently have an acute presentation of disease than those without CN. Some CN cases without or with minimal portal involvement, representing typical acute hepatitis, may progress to classic AIH, although other cases may not.35 ,50 ,54–57

Emperipolesis has been widely described in classic AIH and included in the simplified International Autoimmune Hepatitis Group scoring system (2008).41 ,69 The presence of emperipolesis indicates a close immune interaction between lymphocytes and hepatocytes, which may subsequently induce hepatocyte apoptosis and has been proposed as an additional mechanism of autoimmune-mediated liver injury in AIH. The frequency of emperipolesis found in the present study was far higher than that of the 2015 Japanese nationwide survey,22 which was also higher than the rate reported in classic AIH.68–70 Interestingly, emperipolesis was found to be more frequent in AIH patients with acute hepatitis than in those with chronic hepatitis (21.7% vs 12.2%, p=0.02).22 This may be a justification for our higher frequency than that of classic AIH, and emperipolesis may have been underestimated in the 2015 Japanese nationwide survey.

Plasma cell infiltration is a hallmark of AIH, although it is not a specific feature. We frequently found plasma cells, in lobular and/or portal areas, at a rate of 96.4% (lobular 85.7%, portal 91.6%), which was similar to that reported by Fujiwara et al36 ,67 (93–100%), Burgart et al42 (100%), and Takahashi et al (87.6%).22 Besides portal areas, we frequently found plasma cells in hepatic lobules (85.7%), including CN areas that were more likely to be involved in early stages of the disease. Interestingly, the infiltration of plasma cells in central areas reportedly occurred during the acute phase of AIH, but not during acute hepatitis caused by hepatitis virus and drug-induced liver injury (DILI).24 Thus, our data confirm that CN with plasma cell infiltration is useful for supporting the diagnosis of AIH with acute presentations.

Compared with other histological features reported previously, the present data show both similarities and broad differences as follows (our data vs other studies): hepatic rosette formation (60.5% vs 14.6–55.6%),22 ,24 ,27 ,34 ,35 ,58 interface hepatitis (75% vs 66.7–92.3%),5 ,22 ,35 ,36 ,42 ,58 bile duct injury (50% vs 21.1–77.8%),22 ,58 fibrosis (67.4% vs 40–96.1%),5 ,6 ,16 ,22 ,34 ,36 ,42 ,67 end-stage fibrosis (stage 4) (0% vs 0–33.3%),5–7 ,14 ,16 ,42 portal inflammation (92.9% vs 90–100%),16 ,22 ,35 lobular necrosis/inflammation (97.7% vs 73–100%)5 ,16 ,22 ,27 ,42 ,67 and cobblestone appearance of hepatocytes (82.6% vs 44.4%, 8/18 cases).27 It is worth noting that the frequencies of fibrosis and interface hepatitis varied markedly among studies, which may have led to controversial conclusions in the clinicopathological characteristics of AIH with acute presentations compared with classic AIH. Indeed, the rates of fibrosis and interface hepatitis reported previously5 ,42 were higher than those in the present study. In 1994, Nikias et al5 reported 12 cases of acute presentation of AIH. Of those, 75% showed fibrosis, 27% cirrhosis and 91.7% interface hepatitis (piecemeal necrosis), concluding that AIH with an acute presentation is indistinguishable by clinical laboratory features from the chronic presentation. Likewise, in 1995, Burgart et al42 evaluated 26 patients with acute presentation of AIH, of which 25/26 (96.1%) presented portal inflammation and fibrosis, with overt cirrhosis in 4 cases (15%) and interface hepatitis in 24/26 (92.3%). In contrast, Fujiwara et al and Takahashi et al reported lower proportions of fibrosis at 44.4% and 53.9%, respectively.22 ,67 Furthermore, fibrosis, particularly advanced-stage fibrosis (stage 3, 4), has been less frequently observed in AIH with acute presentations compared with that in AIH with chronic presentations.16 ,22 ,34 Causes of this divergence are unclear, but possible explanations are varied in diagnostic criteria as indicated above.

Portal inflammation and lobular necrosis/inflammation showed similarities between our study and other reports with high frequencies. Although patients with histological features of acute liver failure (submassive or massive necrosis) were not included in the present study, cases with these histological patterns have repeatedly been reported.24 ,25 ,27–29 ,42 Portal inflammation was almost always found, accounting for 92.9% of our patients (90–100% in other studies),16 ,22 ,35 including in early-biopsy cases with a mild lobular injury and without any evidence of disease chronicity. Accordingly, AIH with acute presentations can present in a wide histological spectrum of acute hepatitis. Further, portal inflammatory infiltration, like in acute viral hepatitis, may occur in the acute hepatitis phase of AIH and may not be specific for chronic AIH. More importantly, findings that comprise focal (spotty) necrosis in hepatic lobules without any other changes, but with mild portal inflammation, though unspecific, should not preclude a diagnosis of AIH. The appearance of plasma cells in the hepatic lobules and/or portal tracts may be suggestive of AIH in this setting. Besides, in the liver specimens, we often found pigmented macrophages (84.5%) and cobblestone appearance of hepatocytes (82.6%), which indirectly indicate disease activity. The evaluation of these lesions may be helpful for the diagnosis and prognosis of AIH with acute presentations, particularly in the covalence stage when necroinflammatory activity has reduced.71

Histologically, differentiating AIH from viral hepatitis and DILI, particularly DILI with AIH features (drug-induced AIH), can be challenging or indistinguishable.72 ,73 The prominent histological features indicated above are also present in many other acute or chronic liver diseases.60 ,63 ,74 ,75 However, when occurring in combination, they may favour a diagnosis of AIH. Portal and lobular plasma cells, hepatic rosette formation and emperipolesis have been reported to be features that favour AIH over DILI and viral hepatitis.24 ,60 ,70

Our study had a number of strengths and limitations. This multicentre study provided us with the unique opportunity to perform a systematic histological assessment in a large cohort of patients. The histological review was conducted by experienced hepatic pathologists. Additionally, the key histological features were predefined by consensus. The fact that no gold standard exists for the diagnosis of AIH with acute presentations is a limitation of our analyses. However, patients were retrospectively enrolled, and they showed response to immunosuppressive therapy on follow-up; this warranted our diagnosis and results of histological evaluation. Without a control group, we could not indicate whether there are particular histological features that should suggest AIH with acute presentations, as opposed to other diseases, but this was not the aim of the present study.

Herein, we found that AIH with acute presentation is associated with a variety of histological changes, occurring at all stages of the disease, but cirrhosis. The most prominent features were lobular necrosis/inflammation, plasma cell infiltration in lobular and/or portal areas, emperipolesis, pigmented macrophages, cobblestone appearance of hepatocytes, perivenular necroinflammatory activity and/or CN and interface hepatitis. Portal inflammation can occur in early AIH. There are no histological features that are pathognomonic, but lobular necrosis/inflammation, particularly CN, plasma cell infiltration, emperipolesis and pigmented macrophages are characteristics of the acute presentation of AIH. In conclusion, the acute presentation of AIH represents the entire histological spectrum of acute hepatitis and chronic hepatitis with various activity grades and all fibrosis stages. We do encounter AIH that both clinically and histopathologically qualifies as acute hepatitis in our daily practice. We should recognise this wide histological spectrum of AIH to make a timely diagnosis and appropriate treatment decisions.

Take home messages

  • Approximately 9–75% of autoimmune hepatitis (AIH) cases can have an acute clinical presentation that mimics acute hepatitis of other causes.

  • The acute presentation of AIH can represent the entire histological spectrum of acute hepatitis and chronic hepatitis with various activity grades and fibrosis stages.

  • The prominent presence of lobular and perivenular necroinflammatory activity, including centrilobular necrosis, pigmented macrophages and cobblestone appearance of hepatocytes, together with the classic AIH features (plasma cell infiltration and emperipolesis) are useful to support the pathological diagnosis of AIH with acute presentation.

References

Supplementary materials

  • Abstract in Japanese

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Abstract in Vietnamese

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Handling editor Cheok Soon Lee

  • Twitter Follow Hiep Nguyen Canh @Nguyen Canh Hiep

  • Contributors Each of the authors contributed to the manuscript. The concept of this study was created by the Intractable Liver and Biliary Diseases Study Group of Japan. KH, KT, MK and MN developed the protocol. KH, KT, MK, MN, HO, YS and HNC reviewed the slides, collected, analysed and interpreted the data. HNC drafted the manuscript and KH critically revised the manuscript. KY, AT, MA, J-HK, KK, AI, TF, AT, TA-H, TT, YS, KF, MZ, HO, AT and HT identified clinical cases, and collected data. All authors read and approved the final manuscript.

  • Funding This study was conducted and supported by the Health Labour Science Research Grants from Research on Measures for Intractable Diseases, the Intractable Hepato-Biliary Diseases Study Group in Japan and grant No. 26460416 from the Japan Society for the Promotion of Science (KH). HNC received the Japanese Government (Monbukagakusho: MEXT) Scholarship, No. 153345.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Fukushima Medical University Ethics Committee (no.2099).

  • Provenance and peer review Not commissioned; externally peer reviewed.