In autoimmune hepatitis, patients who are intolerant or with toxicity experience, non-responders, relapsers or refractory are challenging. Non-standard drugs are being tried to preemptively avoid corticosteroid-related side effects. Prognosis and quality of life of life rely on treatment optimization. Recently, emergence of powerful immunosuppressive agents, mainly from liver transplantation, challenged the supremacy of the corticosteroid regime and promise greater immunosuppression than conventional medications, offer site-specific actions and satisfactory patient tolerance. Successes in experimental models of related diseases have primed these molecular interventions. We performed a literature review on alternative treatments. Azatioprine intolerance is the principal indication for mycophenolate use but it can be used as a front-line therapy. Cyclosporine A and tacrolimus have been tested for non-responders or relapsers. Rituximab may be used as salvage therapy. Anti-tumor necrosis factor-alpha agents may be used for incomplete responses or non-responders. Methotrexate is possibly an alternative for induction of remission and maintenance in refractory patients. Cyclophosphamide has been included in the induction regimen with corticosteroids. Ursodeoxycholic acid action is mainly immunomodulatory. Non-standard treatments are coming slowly to the attention, but its use should be cautious performed by experienced centers.

Liver chronic inflammation, interface hepatitis (on histology), hypergammaglobulinemia, and autoantibodies presence are landmarks of autoimmune hepatitis (AIH)[1,2]. AIH is an immune-mediated liver disease, its etiology is unknown[3]. A loss of tolerance seems to be the principal immunologic explanation[3,4]. Women are more affected than men and it occurs across all ages. Men are diagnosed at 40 years of age and women at 50 years of age in median[5]. Prevalence and incidence data on AIH are still limited. In Western Europe and North America Caucasian people the estimated prevalence ranges from 50 to 200 cases per million[6]; the annual incidence in Northern Europeans is 1.9 cases per 100000 persons per year[1,2,7]. An acute presentation occurs in 25% of patients, fulminant presentation is rare but AIH should be considered as etiology in the study of acute liver failure[4]. In addition, the prognosis of disease is influenced by age (young patients having an increased risk), presence of cirrhosis, treatment response (as opposed to activity) and relapses.

The clinical manifestations are heterogeneous. Unspecific symptoms like fatigue, lethargy, jaundice and right upper quadrant pain, are the most frequent clinical presentation. The complications of portal hypertension, i.e., ascites, esophageal varices, hypersplenism and encephalopathy, may ensue on the natural course of the disease. About 25% of patients present extrahepatic immune-mediated symptoms and diseases, arthralgia is the most frequent[4]. Clinical criteria were developed in 1993 and they help to establish the diagnosis when there isn’t a single clinical or biochemical test to affirm it[8,9]. These diagnostic criteria include hypergammaglobulinaemia; positivity for autoantibodies: anti-nuclear antibody (ANA), smooth muscle antibody (SMA) or anti-LKM1; typical histology; other causes of hepatitis (viral or toxic) should be excluded as well as other diseases with similar presentation of AIH[8,10]. The autoantibody profile helps to classify AIH: in type 1, SMA and ANA are present; on type 2 anti-LKM1 antibodies are present. Type 1 AIH affects adults and children, while type 2 AIH is mainly a disease of children and adolescents[11]. The scoring system for AIH has a sensitivity of 97% to 100%. In the presence of chronic hepatitis C, the specificity for excluding AIH relies between 66% to 92%[7,10].

Inflammatory activity at the onset of disease and cirrhosis are the main determinants of natural history and prognosis of AIH. Without treatment, mortality of 90% in 10 years is expected when a 5- to 10-times elevation of aspartate aminotransferase and a twofold increase of γ-globulins are present. Cirrhosis occurs in 17% within 5 years of the diagnosis in patients periportal hepatitis and in 82% of patients with bridging necrosis or necrosis of multiple lobules[2]. At diagnosis, 58% of mortality is expected within 5 years of diagnosis[2,4].

The diagnostic criteria may be too strict when applied to diverse ethnic groups because heterogeneous clinical phenotypes and outcomes may be present[12-14] which may be determined by antigenic exposure, variations in immune response, genetic predisposition and cultural, social and economic factors[14]. The diagnosis may be delayed as the institution of corticosteroid treatment[14].

The human leukocyte antigen (HLA) profile also determines the clinical outcome of AIH: HLA DR3 is associated with more severe disease; HLA DR4 is associated with onset at a later age and a more benign outcome of AIH[4]. The HLA-DRB1 locus, specially the alleles HLA-DR3 (DRB1*0301) and DR4 (DRB1*0401) are related to AIH type 1 susceptibility in European populations and North Americans and the strongest genetic associations contributing to the diagnosis of AIH and are included in the IAIHG revised diagnostic scoring system[15]. Genetic profile determines the response to treatment: those who do not respond to corticosteroid treatment have usually DRB1*0301 alleles[15]. Also, clinical manifestations and prognosis may be determined by genetic profile.

AIH, if left untreated, may lead to cirrhosis, liver failure and even death[16]. Survival is increased when immunosuppressive therapy is used. Initially, induction of remission is the main goal[4]. Corticosteroid regimens are effective[2,17]. Prednisolone, alone or in association to azathioprine leads to symptom improvement, laboratory and histologic manifestations of liver inflammation within 6-12 mo in the majority of patients[4,18]. Standard therapy leads to complete biochemical response in 77% in 6 mo[19], improves hepatic fibrosis[18] and 20-year life expectancy is increased in 80%[20].

Early recognition and treatment of the disease, treatment until complete resolution of inflammation, prevention of complications of treatment and early identification and treatment of problematic patients may improve the outcomes of current therapy[21]. Main prognostic determinant is the response to corticosteroid therapy: rapid disease progression is expected when the treatment is delayed or deferred[17].

Between those 23% that do not respond, 5% are intolerant or present toxicity, 7% are non-responders or have refractory disease and the remain 10% have incomplete responses[2]. The relapses after drug withdrawal are frequent (50%-86%)[2,22]. Other efficient treatments are needed. Complete biochemical remission determines the outcome[23] and, therefore, optimization of treatment has implications on prognosis and quality of life[2]. Liver transplantation supersedes empirical drug therapy in decompensated patients[21].

The recognition of the response of AIH to immunosuppression changed its prognosis[24,25]. Immunosuppressive treatments should be established immediately, especially in the presence of severe disease[24].

The goal of AIH treatment includes: induction of remission; maintenance of remission; prevention of the establishment of cirrhosis and complications using the lowest possible dose of medication[11,15]. AIH has a good response to immunosuppressive treatment with 80% of remission rate[15,26].

According to the American Association for Study of Liver Diseases guidelines treatment is indicated for patients with established diagnosis of AIH, elevation of aminotransferase activities [≥ 5 times upper normal limit-(ULN)], rises of immunoglobulin G ( ≥ 2 times upper normal value) and presence of interface hepatitis or necroinflammatory activity (Ishak score 4-6)[2,16]. If untreated, high mortality of 60% at 6 mo is expected when serum aspartate aminotransferase (AST) levels of 10 times the ULN or more than 5 times the ULN especially if associated with serum γ-globulin level more than twice the ULN. Also, in 82% there is progression from bridging necrosis or multilobular necrosis at presentation to cirrhosis, associated with 45% mortality within 5 years[2,16]. Corticosteroid treatment is indicated in the presence of these findings[2,16]. Treatment should also be started in the presence of incapacitating, such as fatigue and arthralgia[2].

Standard therapy may be not an option if coticosteroids, azathioprine or other immunosuppressive therapies are contraindicated by itself or by patient risk factors. The treatment doesn’t alter the outcome in patients with decompensated liver cirrhosis on waiting list for liver transplantation or in those with cirrhosis without inflammatory activity[6].

The outcomes of therapy include: remission, relapse, treatment failure, and stabilization[11]. Normal inflammatory parameters and histology is necessary to assume remission. Histological remission should be differentiated from biochemical remission (complete normalization of aminotransferase levels including IgG). Treatment should definitely be considered in any patient with proven AIH, histological activity and a more than marginal elevation of aminotransferase levels, not only in patients with levels greater 5 × ULN. In 65% to 75% of patients after 24 mo on standard therapy remission is achieved[11,27]. Relapse is defined as a flare in aminotransferase levels with symptoms under treatment, following the minimum dose of maintenance therapy, or after withdrawal. Relapse occurs in about 50%; loss of remission in 42% within 6 mo of treatment withdrawal and in 80% after 3 years; progression to cirrhosis occurs in 38% and liver failure in 14%[11,17]. Retrospective analysis indicates that loss of remission or relapse occurs in virtually all patients with AIH in long-term remission when immunosuppressive therapy is discontinued[28]. Treatment failure should be assumed when there is progression of symptoms, non-improvement of histological parameters and deterioration of serologic features during standard therapy. In case of treatment failure, diagnosis should be reconsidered to exclude an overlap syndrome with primary sclerosing colangitis or primary biliar cirrhosis or different etiologies[11]. Partial remission corresponds to stabilization of the disease[11].

The standard initial treatment of AIH includes the corticosteroids only or combined with azathioprine. Combination therapy is the first choice and low-dose of prednisolone (30 mg/d) with 1 mg/kg azathioprine are used in the induction phase[11]. In the United States, 50 mg is used for azathioprine, but in Europe a dose of 1-2 mg/kg bodyweight is used[2]. Alternatively, monotherapy may be used, with 60 mg of steroid and reductions of 10 mg/wk to maintenance dose of 20 mg for at least 6 mo, and further reduction until lowest dose in 2.5 mg decrements. Maybe the initial prednisolone dose in combination therapy should be considered since the percentage of response is higher. There are no differences in the remission induction. Combined treatment is preferred because it allows to decrease the dose of the prednisone dose to below 10 mg and reduces the steroid side effects[6].

Standard therapy is the best option unless contra-indicated and may be especially useful by reducing corticosteroids side-effects in older patients, in patients with osteoporosis, metabolic syndrome or psychiatric lability[6]. Monotherapy with steroids is the best treatment option in patients with hematological abnormalities or a proven homozygous deficiency of thiopurine methyltransferase because azathioprine causes hematological side effects such as leukopenia or anemia[6]. Thiopurine methyltransferase (TPMT) is an enzyme responsible for the conversion in one of azathioprine to 6-mercaptopurine (active metabolite) and in 6-methyl mercaptopurine or 6-thiouric acid (inactive metabolites)[12]. In patients with azathioprine intolerance, lower TPMT activity is documented but measurements of TPMT activity cannot be used to identify those patients[29]. Pre-treatment TPMT testing provides some certain of the presence of risk for azathioprine toxicity and strengths physician confidence in the treatment regimens[12]. Allopurinol may safely and effectively optimize thiopurine therapy in patients with intolerance and/or nonresponse due to an unfavourable thiopurine metabolism and this is another option in order to maintain the standard treatment[30].

Corticosteroids are the first option of treatment in all populations, but its use should be individualized in the presence of cholestatic features[14].

However, as the combination treatment fails, other drugs have been tried although its use requires further validation[24].

In fulminant hepatic failure and in en-stage liver disease, transplantation is the treatment of choice. Post-transplantation AIH recurrence may occur[24].

Budesonide is glucocorticoid from the next-generation, more than 90% has first pass hepatic clearance and metabolites don’t have glucocorticoid activity[31]. These pharmacological properties seem to predict less secondary effects. In non-cirrhotic patients, treatment combination between budesonide and azathioprine may be an alternative in uncomplicated AIH with mild disease[32,33] or with conditions that may be worsened by prednisone treatment like hypertension, osteopenia, diabetes and obesity[34].

In corticosteroids refractory or dependent AIH patients may not be used as a rescue treatment. The budesonide regimen normalized serum AST and alanine aminotransferase (ALT). Budesonide histological resolution and persistent response is unknown[34].

There are some difficult-to-treat patients for whom newer immunosuppressive agents, usually employed as anti-rejection drugs, have been tried with variable success. Immunosuppression with non-standard drugs is being tried to avoid corticosteroid side effects (13%) but are being used specially as superior regimens to corticosteroid treatment[21]. The use of such regimens has to be weighed and data available comes only from few small studies or case reports[15]. Other treatments considered are: mycophenolate mofetil (MMF)[35-43], cyclosporine A (CyA)[44,45], tacrolimus (FK506)[46-49], ritximab[50], anti-tumor necrosis factor-alpha (TNF-α) agents[51], methotrexate[52], cyclophosphamide and ursodeoxycholic acid (UDCA)[53] (Table 1). These non-standard treatments application is not widespread and they are not included into any standard management algorithm[15].

Only recently has the emergence of powerful immunosuppressive agents, mainly from liver transplantation, challenged the supremacy of the corticosteroid regimens[22,54]. Drugs outside of the standard repertoire now promise greater immune suppression than conventional medications, offer site-specific actions and satisfactory patient tolerance[22,54]. Site-specific molecular treatments are also possible because of improved understanding of the central pathogenic disease pathways and technological advances that now enable modulation of these pathways[22,54]. Furthermore, successes in experimental models and in other autoimmune diseases have primed these molecular interventions for study in AIH[22,54].

Importantly, publication bias may be considered since there is, probably, underreport of studies with negative results. Also, target populations, dosing schedules, safety profiles and monitoring strategies are not yet clear; adjunctive therapy with corticosteroids is still required; the standard algorithms do not include already the risks and expense of these drugs[55]. Newer agents are much more expensive than the standard treatment irrespectively of the generic use (as recently available generic MMF may attenuate this problem) and this may be a limitation to the accessibility to these treatments.

In AIH, identification of efficient salvage treatment options is urgently needed for the difficult-to-treat patients: those who experience intolerance or toxicity, non-responders, relapsers or with refractory disease. Also, non-standard drugs are being tried as superior drugs to corticosteroid regimens and to minimize its side effects. Optimization of treatment plays a major role in long-term prognosis and quality of life for patients with AIH. Recently, the emergence of powerful immunosuppressive agents, mainly from liver transplantation, challenged the supremacy of the corticosteroid regime and promise greater immunosuppression than conventional medications, offer site-specific actions and satisfactory patient tolerance. Successes in experimental models and in other autoimmune diseases have pointed these molecular interventions for study in AIH. Some encouraging results were described, but the establishment of these non-standard drugs as alternative treatments has evolved slowly and they weren’t already included into a standard management algorithm. Therefore, those treatments should be used with caution and only in experienced centers.