Putting flesh and polish on autoimmune hepatitis and moving the disease of exclusion to inclusion


  • Albert J. Czaja

    Corresponding author
    1. Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine, Rochester, MN
    • Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905
    Search for more papers by this author
    • fax: 507-284-0538

  • Potential conflict of interest: Nothing to report.


Autoimmune hepatitis emerged during an era when concepts of neonatal immune tolerance, clonal selection of lymphocytes, and “forbidden clones” of activated immune cells were forming. The diagnosis had to be deduced from circumstantial evidence and by exclusion of other conditions. The goals of this review are to demonstrate how a clinician nonscientist can contribute to the maturation of autoimmune hepatitis and to illustrate the principles of clinical investigation that can be applied broadly to other projects. Autoimmune hepatitis initially had to be distinguished from other diseases, and improvements in the tests for viral and immune markers were instrumental in this regard. Diversification of the clinical phenotype to accommodate acute severe, asymptomatic, elderly, and variant forms enhanced the pertinence of the disease, and the formation of the International Autoimmune Hepatitis Group standardized the diagnosis, interconnected investigators, and promoted global acceptance of the condition. Subsequent studies refined current corticosteroid-based therapies, identified prognostic markers, assessed genetic predispositions, explored new pharmacological agents, and forecast the emergence of cellular and molecular interventions. Good fortune, stimulating mentors, career dedication, practical goal selection, protocol compliance, compulsive record keeping, personal resilience, and strong collaborations were the bases for progress. Autoimmune hepatitis exemplifies an evolutionary process in the science of autoimmunity and the people committed to its study. Lessons derived from this experience can be far-reaching. (HEPATOLOGY 2010;52:1177-1184)

When I graduated from Harvard Medical School in 1968, the designation “autoimmune hepatitis” did not exist, and the clinical phenotype of “lupoid hepatitis” connoted cirrhosis in young amenorrheic women with hirsutism, acne, and cushingoid features.1-4 The remarkable early observations associating the lupus erythematosus cell phenomenon with chronic hepatitis5,6 were followed by the recognition that antinuclear antibodies7 and smooth muscle antibodies8 frequently accompanied the condition. These features raised the possibility of an immune-mediated chronic aggressive liver disease.9 This novel consideration was fitting since Frank Macfarlane Burnet and Peter Medawar had recently won the Nobel Prize in Physiology or Medicine (1960) for describing the clonal selection of lymphocytes and explaining neonatal immune tolerance.10 Autoimmunity was an emerging and exciting frontier.

The concept of Burnet that autoreactive cells could escape into the peripheral circulation as “forbidden clones”11-13 heralded an era of disease discovery and understanding, and autoimmune hepatitis was a product of this surge. Autoimmunity, however, was still a vague pathogenic mechanism; it was not an etiologic agent like a virus or a drug; and it could not be measured in the clinic. The evolving requisites for autoimmunity, especially the requirement for the transfer of disease by antibodies or lymphocytes, were restrictive,14,15 and “autoimmune” vied with “idiopathic” as an apt descriptor for the fledgling condition. The wobbly legs of autoimmune hepatitis would persist for at least 2 decades. Systemic lupus erythematosus almost swallowed it16 and drug-induced17 and virus-related18,19 conditions repeatedly threatened its legitimacy.

The goals of this review are to illustrate the dynamics of successful clinical investigation in liver disease and to underscore the vital role of the clinician nonscientist in starting and completing the circle of care from bedside-to-bench-to-bedside. Autoimmune hepatitis will be the “illustrative model” by which to accomplish these goals, and I will be the typical “clinician nonscientist.” The script can be applied broadly and accommodate any substitute model or actor. The principal components of this tutorial are indicated below, and they rely heavily of good fortune, good mentoring, appropriate goal identification, adherence to protocol, compulsive record keeping, personal resilience, and strong collaborations.


CALD, chronic active liver disease; HBsAg, hepatitis B surface antigen; HLA, human leukocyte antigen; IAIHG, International Autoimmune Hepatitis Group; MELD, Model for End-Stage Liver Disease.

Career Awakening and Good Fortune

From 1969 to 1972, I had the good fortune to interact with academic clinicians who had a keen interest in the study of liver disease (Table 1). At the Philadelphia General Hospital, Geobel Marin advocated the principles of controlled clinical trial and “double-blinded” investigation as the bases for new knowledge in clinical medicine, and my first article comparing peritoneoscopy with unguided needle biopsy of the liver illustrated some of these principles.20 At the University of Pennsylvania, Roger Soloway had just returned from a fellowship at the Mayo Clinic, and he presented wonderful data derived from a now classic controlled clinical trial that described the natural history and treatment of “chronic active liver disease”.21 My commitment to the study of liver disease was established through these contacts in Philadelphia as was my desire to train at the Mayo Clinic. Fortunately, Bill Summerskill agreed to accommodate this desire.

Table 1. Requisites for Successful Clinical Investigation and Personal Examples
RequisitesPersonal Examples
  1. Numbers in parenthesis are references.

Keen interest in field of studyEarly exposure to academic clinician investigators (193, 194)
Introduction to methods of clinical trial (20)
Training in study design, cross-discipline collaboration, and data analysisStudies of Curling's ulcer (22, 23, 195, 196)
Description of acute liver injury after burns (24)
Identification of key clinical problems, resources and talent Total personnel commitmentInteractions with Solko Schalm (34), Archie Baggenstoss (51, 83, 108) and Jurgen Ludwig (57, 75)
Direction of W.H.J. Summerskill (56, 197, 198)
Rigid adherence to protocol and compulsive long-term follow-upAll the studies on autoimmune hepatitis at the Mayo Clinic
SerendipidityAssignment to U.S. Army Institute of Surgical Research
Acceptance to Mayo Clinic Hepatology Research Fellowship
Death of W.H.J. Summerskill and Mayo opportunity
Timely discovery of hepatitis C virus (87)
Interaction with cadre of like-minded investigatorsInternational Autoimmune Hepatitis Group (91)
Development and modification diagnostic criteria (94)
Evaluation of scoring systems (92, 95, 96)
Identification of treatment challenges and opportunities (97)
Acceptance of failed hypotheses and resiliency of effortFailed trials of prednisone bolus therapy (176), ursodeoxycholic acid (177), and budesonide (178)
Inability to complete trial on mild autoimmune hepatitis (89)
Lack of value for certain serological markers (199-201)
Strong collaborationsMultiple studies within (140) and outside (62, 68, 71, 81, 119, 133, 161, 202, 203) the United States
Development of new skillsLearning restriction fragment length polymorphism (204-206)
Self-perpetuating hypothesesProgress toward new therapeutic interventions (180-182)
FundingSporadic, unpredictable, creative

The military draft interrupted my transition to Mayo, but my assignment to the U.S. Army Institute of Surgical Research (“Burn Unit”) at the Brooke Army Medical Center in San Antonio, Texas was fortuitous and formative (Table 1). From 1972-1975, I was steeped in clinical investigation, collaborative study, protocol development, critical study review, and data analysis. I was also surrounded by superb clinical investigators in all subspecialties. Doug Wilmore, who later assumed the Francis Moore Chair of Surgery at Harvard, and Basil Pruitt, who was the quintessential trauma surgeon, clinical investigator and unit commander, kept my compass fixed on pertinent areas of clinical study. Joseph McAlhany, who later joined the surgical faculty at the University of South Carolina, taught me the value of collaboration across disciplines. Together we learned much about Curling's ulcers22 and their prevention,23 and I was still able to pursue my interest in liver disease.24 By this time, I had learned that successful clinical investigation required a contagious excitement about the topic, accurate identification of the key clinical problems, appropriate resources, talented individuals, and total personal commitment. I also realized that clinical problems were abundantly evident in routine medical practice and that most clinical environments could accommodate and benefit from their study (Table 1).

Mentoring and Opportunity

In 1975, Bill Summerskill headed a research unit that was enriched by the studies of Alan Hofmann, Sydney Phillips, Juan Malagelada, Bill Go, and Gene DiMagno and energized by trainees in liver disease such as Nick LaRusso, Solko Schalm, Misael Uribe, Arnold Vogten, and Gerry van Berge Henegouwen. Collaboration, critical interactive review, and the importance of high quality data were evident daily. Chronic active liver disease (CALD) was a term that had been developed by Bill Summerskill. It included all patients with the same clinical, laboratory and histological features regardless of etiology, and it was the generic name for the liver disease that we all studied.25-27

Misael Uribe defined the bases for corticosteroid-induced complications in treated CALD28-30; Arnold Vogten was the first person to recognize that human leukocyte antigen (HLA) DR3 was associated with a poor prognosis31; and Solko Schalm demonstrated that reduced conversion of prednisone to prednisolone in advanced CALD was insufficient to affect treatment outcome.32,33 Solko Schalm also demonstrated with Archie Baggenstoss that initial histological patterns of CALD had different prognoses and that they could undergo transitions during corticosteroid treatment.34 Etiologic distinctions within CALD were just being recognized,35 and Solko Schalm started the dissection of CALD into subcategories by demonstrating differences between patients with and without hepatitis B surface antigen (HBsAg).36 Autoimmune hepatitis was hidden within the rubric of “HBsAg-negative chronic active liver disease,” and its existence was uncertain.37,38

Bill Summerskill taught discipline in the acquisition, organization, analysis and understanding of data, and he believed that the Mayo studies in liver disease were distinguished not by their innovation but their “tenacity.” Systematic “protocol” evaluations at each visit, liver tissue examinations at 6 month intervals, uniform treatment schedules, predefined responses to disease behavior, regular surveillance schedules using mailed serum specimens, serum and liver tissue banks, and commitment to indefinite patient follow-up were the manifestations of this “tenacity.” I remain convinced that rigid adherence to protocol and compulsive follow-up are essential components of successful clinical investigation (Table 1).

Bill Summerskill died suddenly in March 1977, and I was abruptly launched solo into the realm of CALD. Life-saving therapy had now been established by three controlled clinical trials; the disease was rare; funding sources were limited or uninterested, and my principal initial concern was that there was nothing more to study.21,39,40 The practical clinical problems that required answers became obvious quickly through routine patient care, and they generated a compelling urgency for further clinical studies. Remarkable work from Meyer zum Buschenfelde's group in Mainz, Germany,41-44 and Roger Williams' group in London, England,45-49 invigorated the concept of autoimmune hepatitis, and I suddenly realized anew that I was in an exciting place at an exciting time.

Fleshing Out the Disease: Phase I

The first objective was to describe the clinical phenotype of autoimmune hepatitis and to distinguish it from other diseases. This was done by describing its autoantibodies,50 histological manifestations,51,52 clinical presentations,53-55 and response to corticosteroid treatment.56-58 The disease had to be distinguished from systemic lupus erythematosus59 and chronic viral hepatitis60; it had to be released from the early restrictive requirement for 6 months of disease activity61; subtypes based on mutually exclusive serological markers had to be explored62-64; and it had to accommodate patients with nonclassical manifestations.65 This was the era of serological exploration, and collaborations with Mikio Nishioka,66-70 Francesco Bianchi,71,72 Michael Manns62-64 and their coworkers were essential to understand the nature and clinical significance of antinuclear reactivities, including antibodies to ribonucleoproteins, histones, single-stranded DNA and doubled-stranded DNA, antibodies to actin, antibodies to liver kidney microsome type 1 (anti-LKM1), antibodies to soluble liver antigen (anti-SLA), and antimitochondrial antibodies in autoimmune hepatitis.

These efforts complemented studies performed elsewhere, and they supported concepts of an acute autoimmune hepatitis,54 two serologically distinct forms of the disease,73 variant syndromes characterized by antimitochondrial antibodies,74 bile duct changes,75-79 or concurrent viral infection,80-82 and a seronegative state frequently misclassified as cryptogenic chronic hepatitis.83-85 Michael Manns of Hannover, Germany was a critical collaborator during this early period, and it was through his generosity that the assay for anti-LKM1 was established at Mayo.62 CALD was splitting apart and assuming a more complex phenotype65 (Fig. 1). Terms such as idiopathic chronic active hepatitis, HBsAg-negative chronic active hepatitis, chronic persistent hepatitis, and plasma cell hepatitis were no longer appropriate, and a codification of diagnostic criteria and nomenclature was sorely needed.86

Figure 1.

Division of chronic active liver disease (CALD) into subgroups based on etiologic distinctions or atypical features. Patients with chronic viral hepatitis could have concurrent immune features, and patients with cryptogenic chronic hepatitis could be indistinguishable from autoimmune hepatitis except for the absence of autoantibodies. Antimitochondrial antibodies (AMA) and histological findings of bile duct injury could suggest primary biliary cirrhosis (PBC), and the presence of chronic ulcerative colitis (CUC) and biliary changes by endoscopic retrograde cholangiography (ERC) could suggest primary sclerosing cholangitis (PSC). Cholestatic clinical and laboratory changes in the absence of AMA or cholangiographic abnormalities characterize a cholestatic variant (“autoimmune cholangitis”). The percentages indicate the estimated frequencies of these occurrences.

The discovery of the hepatitis C virus in 198987 initially challenged the concept of autoimmune hepatitis, but later validated its existence by demarking a subset of liver disease that was not virus-related.60,81,88,89 The clinical, laboratory and histological trappings of autoimmunity, the confident exclusion of viral infection in most cases, and the responsiveness of the disease to corticosteroid therapy ultimately justified its formal designation as autoimmune hepatitis.90

In 1992, Philip Johnson and Ian McFarlane organized an international panel in Brighton, United Kingdom to address the issues of nomenclature, diagnosis and treatment, and the International Autoimmune Hepatitis Group (IAIHG) was formed.91 Diagnostic criteria were codified, reviewed, and updated. The diagnostic scoring system, initially sketched on a tabletop napkin by Drs Johnson and McFarlane, was tested,92,93 refined,94 and simplified.95,96 The IAIHG continues to identify key clinical and investigational issues,97,98 develop strategies for the standardization of serological testing,99 modify diagnostic criteria,94,95 and explore new therapies. It has repeatedly demonstrated the importance of dialogue within a cadre of like-minded clinical investigators to maintain research direction and motivation (Table 1).

Polishing Treatment: Phase I

The next objective was to refine the treatment of autoimmune hepatitis and identify prognostic factors. The indications for therapy had been established for only the most severe life-threatening forms of the disease.100 An appropriate endpoint of treatment was uncertain21; relapse after drug withdrawal was common56,101; side effects frequently prompted premature discontinuation of medication57,102; disease progression was possible58,103; and improvements were often incomplete despite protracted therapy104 (Fig. 2).

Figure 2.

Outcomes of corticosteroid therapy. Drug toxicity, disease progression (treatment failure), and incomplete response are the adverse outcomes during treatment, and relapse is the most common problem after drug withdrawal. Reproduced from: Czaja AJ. Autoimmune hepatitis: treatment. (figure 19-1, page 135) In: McNally PR, ed. GI/Liver Secrets Plus. Philadelphia: Mosby Elsevier; 2010, with permission from Elsevier.

A controlled clinical trial was developed in the 1980s to randomize patients with mild autoimmune hepatitis to prednisone or placebo, but this project was abandoned when asymptomatic patients with mild disease frequently refused liver biopsy assessment or the possibility of taking prednisone. Furthermore, the recently available test for hepatitis C virus disclosed that 37% of asymptomatic patients were virus-infected.89 We learned that mild autoimmune hepatitis was not “a submerged iceberg” and that controlled clinical trials were not always doable or appropriate (Table 1). The management of asymptomatic patients with mild disease continued to be controversial for almost 2 decades105,106 until retrospective analyses finally justified their routine treatment.107

Correlations between the laboratory indices of inflammatory activity (serum aspartate aminotransferase and γ-globulin levels) and histological features of liver injury reduced the need for liver tissue examinations during and after therapy.108 Complete histological resolution,57 especially disappearance of plasma cells from the liver specimen,109 was recognized as important in decreasing the frequency of relapse, and the ideal treatment endpoint was defined as normal liver tests and tissue.110 Unfortunately, an ideal laboratory and histological response was not achievable in all patients nor did it prevent relapse.110,111 Repeated relapse and re-treatment increased the cumulative probabilities of progression to cirrhosis and liver failure,112 and alternative strategies were warranted after the first relapse, including indefinite azathioprine therapy or maintenance treatment with low dose prednisone.113,114

Multiple analyses failed to disclose a critical feature at presentation that predicted outcome, and yet the early recognition of problematic patients was clearly an essential requirement to improve results. The response to corticosteroid therapy was the only determinant of immediate prognosis.115 This response could be assessed within 2 weeks, but recognition by treatment trial was still too slow and arbitrary. Earlier studies by Arnold Vogten had demonstrated an association between HLA DR3 and poor outcome,31 and the sophisticated analyses of Peter Donaldson and his colleagues now energized this research direction.116,117

My collaborations with Peter Donaldson and his group were some of my most informative and personally rewarding interactions in clinical research. The clinician nonscientist had to learn new skills in order to study the genetic bases for autoimmune hepatitis in white North American and northern European patients (Table 1). We required a normal control population of the same ethnic background as our patients. Under the direction of Drs. Paula Santrach and S. Breanndan Moore of the Mayo Tissue Typing Laboratory, I learned to perform HLA typing by restriction fragment length polymorphism and I personally secured the necessary control group. This normal control material constituted the bases for our subsequent studies of susceptibility alleles and polymorphisms in autoimmune hepatitis and fostered collaborations with other Mayo investigators whose genetic studies also required a normal control population.118

The alleles, DRB1*0301 and DRB1*0401, and the polymorphisms of the tumor necrosis factor-α gene (TNFA*2), cytotoxic T lymphocyte antigen 4 gene (CTLA-4), and Fas gene (TNFRSF6) influenced the clinical phenotype and behavior of the disease,119-125 but the low sensitivity and specificity of these markers for treatment failure did not justify their expense. The model of end-stage liver disease (MELD) was also applied to autoimmune hepatitis, and a MELD score of at least 12 points at presentation had a sensitivity of 97% and specificity of 68% for treatment failure.126 The MELD score did encompass the problematic patients, but it also included many others.

Antibodies to asialoglycoprotein receptor had been associated with histological activity and a propensity for relapse after drug withdrawal,127 and these original observations by Ian McFarlane were substantiated.128 Antibodies to soluble liver antigen had been discovered in two independent laboratories,129,130 and the target antigen had been characterized.131-134 These antibodies were associated with DRB1*0301 and severe disease.135-137 Antibodies to chromatin,138 antibodies to cyclic citrullinated peptide (anti-CCP),139 antibodies to double-stranded DNA,67 and antibodies to actin71 were all shown to have prognostic implications (depending on the assay applied), whereas antibodies to nuclear antigens (histones, centromere) 66,68,70 and antibodies to neutrophilic cytoplasm140 did not.

The serological markers of autoimmune hepatitis were studied, either singly or in constellation,141 because of their perceived importance as imprints of pathogenic mechanisms that affected disease severity. Their pursuit also demonstrated the frustration of searching for a “needle in the haystack”.142 Successful clinical investigation can be fraught with failed hypotheses and attempts to find the “needle.” Importantly, these efforts are often manifestations of the vigor and resiliency of the research program. They almost always teach something, and they can enhance the precision of the next search (Table 1).

Fleshing Out the Disease: Phase II

The emerging worldwide experiences with autoimmune hepatitis indicated the diversity of its clinical phenotypes143-148 and genetic predispositions,149-152 and they compelled comparative studies of the disease in different geographical regions and age groups. The rarity of anti-LKM1 in North American patients with autoimmune hepatitis62; the association of autoimmune hepatitis in South American children with HLA DRB1*1301150,152,153; and the late onset of mild disease in Japan154 were observations that generated new questions about the nature of autoimmune hepatitis and its causes.

The “shared motif hypothesis,” which had been espoused in rheumatoid arthritis, was developed as a basis for autoimmune hepatitis in white northern European and North American patients by collaborations with Peter Donaldson and his colleagues.155 The alleles, DRB1*0301 and DRB1*0401, encoded a six-amino-acid motif, L (leucine) L (leucine) E (glutamic acid) Q (glutamine) K (lysine) R (arginine), at positions 67-72 on the DRβ polypeptide chain of the class II MHC molecule that characterized the disease.119 A lysine (K) at position DRβ71 was the key susceptibility factor (Fig. 3). Alleles with similar encoding properties typified the disease in Mexican,156 Japanese,157 and Chinese158 patients in whom the shared motif differed only by the substitution of an arginine (R) for lysine (K) at DRβ71.159

Figure 3.

Susceptibility site on antigen binding groove of class II molecule of the major histocompatibility complex. The critical motif that is associated with susceptibility to autoimmune hepatitis in white northern European and North American patients is between positions 67 and 72 on the DRβ chain. It is encoded as L (leucine), L (leucine), E (glutamic acid), Q (glutamine) K (lysine), and R (arginine). K (lysine) at position DRβ71 is the principal susceptibility factor. The autoimmune hepatitis in Japan, mainland China, and Mexico is associated with R (arginine) instead of K (lysine) at position DRβ71 (“shared motif”). Reproduced with kind permission of Springer Science and Business Media from Czaja AJ. Autoimmune hepatitis. (figure 3-39, page 73) In: Maddrey WC, Feldman M, eds. Atlas of the Liver. ed. 4. Philadelphia: Springer; 2007.

The association of autoimmune hepatitis in South America with HLA DRB1*1301 countered the “shared motif hypothesis,” but the observation that the hepatitis A virus was slowly cleared in individuals with HLA DRB1*1301160 suggested another basis for autoimmune hepatitis in certain regions. Susceptibility alleles could differ because of region-specific indigenous etiological agents. HLA DRB1*1301 was infrequently associated with autoimmune hepatitis in North America,161,162 but in other regions infection with certain viruses or sensitization to particular environmental agents might be favored by this phenotype.160 The association of anti-LKM1 with HLA DRB1*0764,149,163 and the lower frequency of this genetic marker in the normal population of the United States compared to those of Germany64 and Italy164 further supported concepts that the occurrence, manifestations and behavior of autoimmune hepatitis were strongly influenced by genetic predispositions that were region-specific and ethnic-specific.165

Age also influenced the clinical manifestations of autoimmune hepatitis as well as its outcome, and these latter aspects were in turn associated with HLA phenotype166-168 (Fig. 4). Elderly patients (aged ≥60 years) had cirrhosis more commonly at presentation than young adults, and they responded more rapidly and completely to corticosteroid therapy.166,168 Patients aged ≥60 years also had HLA DRB1*0401 more commonly and HLA DRB1*0301 less frequently than young adults.166 These distinctions suggested that elderly patients were exposed to a different battery of triggering antigens than young patients or had a weaker autoimmune response.

Figure 4.

Age-related frequencies of treatment failure, HLA DR3, and HLA DR4 in the autoimmune hepatitis of white North American patients. The frequency of treatment failure decreases with age as the frequency of the human leukocyte antigen (HLA) DR4 increases. The highest frequency of treatment failure occurs in young adults aged 18-30 years, and it is accompanied by the highest frequency of HLA DR3.

The rapidity of the treatment response was identified as important in preventing cirrhosis and liver failure, and the factors influencing the speed of improvement were age and HLA status168,169 (Fig. 5). A total of 94% of elderly patients who responded to standard corticosteroid therapy did so within 24 months, whereas only 64% of adults aged <40 years did so.168 These different rates of response to the same treatment indicated the need for individualized treatment schedules. An ideal response interval of 6-12 months was defined, and the outer limit for response that reduced progression to cirrhosis and frequency of liver failure was 24 months.168 Additional studies are now necessary to determine how to achieve a rapid result in all patients, and they logically must explore individualized dosing schedules and new therapeutic interventions. Clearly, successful clinical investigations are self-perpetuating and potentially endless (Table 1).

Figure 5.

Rapidity of treatment response according to age. Elderly patients respond more quickly to standard corticosteroid therapy than young adults. Treatment response within 12 months is associated with a lower frequency of progression to cirrhosis and liver failure. Reprinted from Czaja AJ. Rapidity of treatment response and outcome in type 1 autoimmune hepatitis. (figure 2, page 164) J Hepatol 2009;51:161-167, with permission from Elsevier.

Polishing Treatment: Phase II

Corticosteroid therapy had been shown to resolve symptoms,21 improve liver tissue,57 normalize 10-year survival,170 and prevent or improve hepatic fibrosis.171 The indications for therapy had been extended to the asymptomatic107 and the elderly166; the ideal treatment response168 and the optimal endpoints of treatment110 had been described; the consequences and management of relapse after drug withdrawal had been clarified112,114; the clinical, serological, and genetic factors affecting outcome had been expanded 115,119,121,123,128,172; and long-term surveillance strategies to detect late relapse111 and hepatocellular cancer173 had been formulated. The next objective was to understand the role of liver transplantation and other immunosuppressive agents as salvage therapies.

Liver transplantation proved to be remarkably effective for the decompensated patient with autoimmune hepatitis. Five year graft and patient survivals exceeded 90% in the early Mayo experience, and the trappings of autoimmunity, including autoantibodies and hypergammaglobulinemia, disappeared in all patients within 2 years.104 Recurrent autoimmune hepatitis was recognized in 17% of transplanted patients, but it was typically managed easily by adjustments in the doses of the immunosuppressive medication.174

Liver transplantation also introduced new medications, such as the calcineurin inhibitors (cyclosporine and tacrolimus) and the next-generation purine antagonist (mycophenolate mofetil). Furthermore, it stimulated interest in expanding site-specific drug and molecular interventions.175 Salvage therapies, including high-dose bolus prednisone,176 ursodeoxycholic acid,177 budesonide,178 and mycophenolate mofetil,179 were evaluated, and their ineffectiveness strengthened the commitment to refresh current corticosteroid-based therapies and improve the timing for liver transplantation.110,126

Autoimmune hepatitis is now poised to enter the next phase of investigation.180,181 Murine models based on DNA vaccination or infection with viral vectors promise to enhance the opportunity to develop new therapies based on site-specific cellular and molecular therapies.182 The immunization of female mice using plasmids of cytomegalovirus containing the antigenic region of human CYP2D6 and human formiminotransferase cyclodeaminase183,184 and the infection of mice with adenovirus expressing human CYP2D6 have produced animal models that closely resemble the human disease.185 The rodent model based on viral infection produces human autoantibodies specific for autoimmune hepatitis, exhibits liver-infiltrating CD4+ T lymphocytes, induces typical histological changes, and progresses to hepatic fibrosis.185 Its self-perpetuating and aggressive nature will allow the study of novel therapeutic manipulations prior to clinical trial.

Studies have already demonstrated the importance and feasibility of the adoptive transfer of regulatory CD4+CD25+ T cells (T-reg cells) in the management of autoimmune hepatitis.186-189 The soluble dimeric recombinant human fusion protein of cytotoxic T lymphocyte antigen 4 (abatacept), which has been approved by the United States Food and Drug Administration and the European Medicines Agency for the treatment of rheumatoid arthritis,190 has a strong rationale as an immune modulator in autoimmune hepatitis,181 and humanized monoclonal antibodies against CD3, which target antigen-specific activated T cells and have been used successfully in the treatment of human type 1 diabetes mellitus,191,192 represent a genre of evolving interventions that have potential application in autoimmune hepatitis.180

Moving the Disease of Exclusion to Inclusion

Autoimmune hepatitis has been a product of evolution both in the science of autoimmunity and the people committed to its study. Its emergence on the global scene reflects this process. The improved tests for viral infection, the diverse battery of available serological assays with diagnostic specificity, and the general awareness of its different presentations have ensured the consideration of autoimmune hepatitis in all patients with acute or chronic liver disease regardless of age, gender, ethnic background, duration or severity of illness, and transplantation status. The simplified diagnostic scoring system of the IAIHG underscores the diagnostic specificity of a few key features, and the litany of exclusion factors that had characterized its early description have been largely bypassed.95 Autoimmune hepatitis has acquired an identity that can be recognized immediately, secured by codified diagnostic criteria, quantified by scoring systems, found throughout the world in all age groups, and treated effectively. The disease that had existed previously only by the exclusion of others can now stand almost fully on its own.

Autoimmune hepatitis still lacks an etiology, but this deficiency in fact ensures its differentiation from the hereditary, metabolic, virus-related, and drug-induced conditions that resemble it. Its responsiveness to corticosteroid therapy remains its salient feature and the principal reason for its consideration. Few other liver diseases have a therapy that so consistently benefits the patient and rewards the physician. Like any successful clinical investigation, progress depends on a multitude of contributors who nurture each other and expand global recognition of the disease. Progress is an endless process that must be pursued by relays of newly committed investigators that reflect global rather than institutional alliances. Future advances in autoimmune hepatitis will depend on the strength of this multi-national network. For the clinician nonscientist, collaborators are the true bases for research success, the principal sources of instruction, and the appropriate recipients of undying gratitude (Table 2).

Table 2. Collaborative Network in the Study of Autoimmune Hepatitis
Country of OriginCollaborators and Contributions
  1. Numbers in parenthesis are references.

BrazilPL Bittencourt (161), ELR Cancado (203), AC Goldberg (161), G Porta (95, 161)
GermanyEM Hennes (95), M Kruger (64), AW Lohse (95, 135), MP Manns (62-64)
GreeceGN Dalekos, C Liaskos, EI Rigopoulou (207)
IrelandDJ Doherty (155)
ItalyFB Bianchi (71, 72), F Cassani (71, 72), S Magrin (81), P Muratori (164, 208, 209), U Volta (202)
JapanM Chen (70), SA Morshed (66-68), M Nishioka (67-69), S Parveen (66), M Zeniya (94, 95, 97)
PortugalRM Dos Santos (210), A Porto (210)
SpainM Costa (133), C Gelpi (133, 172), JL Rodrigues-Sanchez (133), AX Torres-Collado (172)
United KingdomK Agarwal (124, 125), DP Bogdanos (207), S Cookson (122, 123), PT Donaldson (120, 123, 135, 211), IG McFarlane (97, 98), MDJ Strettell (119, 120)
United States
 • Hepatologists and CliniciansEL Krawitt (95, 97), KD Lindor (177, 178, 212), MK Porayko (174), JJ Poterucha (213), J Rakela (85, 89, 115, 214), RD Soloway (215), JA Talwalkar (216), SR Targan (140), J Vierling (97, 217), RH Wiesner (104, 174), N Zein (218-220)
 • Laboratory ScientistsRH Decker (128), HA Homburger (60, 62, 84, 221), SB Moore (121, 206, 222), GL Norman (136, 138, 141, 211), P Santrach (121, 206, 222), Z Shums (136, 138, 141, 211), Taswell HF (57, 60, 83)
 • TraineesGL Davis (53, 57, 58, 83), RP Kenny (74), A Gonzalez-Koch (174), JE Hay (85, 223), AJ McCullough (199-201), KVN Menon (114), AJ Montano-Loza (110, 112, 139, 173, 224, 225), GA Nikias (54, 55), SZ Park (226), R Perdigoto (78), SK Roberts (170, 174), L Sanchez-Urdazpal (104), MT Shiels (102, 176, 200), KK Wang (176, 227-229)
 • PathologistsAH Baggenstoss (51, 83, 108, 230), KP Batts (54, 55), LJ Burgart (54, 55), HA Carpenter (52, 76, 77, 84, 88, 92, 109, 169, 171, 218, 231), J Ludwig (57, 75, 83, 223)


See online Supporting Information for reference cites.