Uncertainties in Completing Conventional Corticosteroid Therapy
Both the AASLD and BSG guidelines recommend that initial corticosteroid therapy be finite ( Table 2 ). Relapse after drug withdrawal occurs in 28–86% of treated patients, but relapse is not inevitable and an attempt should be made to discontinue therapy. Indeed, 21% of patients can achieve a remission that does not require additional therapy during 10.6 ± 2.1 years of observation (median follow-up, 6.3 years), and the prospect that stringent end-point criteria can improve these results has further supported the recommendation to treat with the expectation of drug withdrawal. The AASLD and BSG guidelines are discrepant in the duration of therapy required to achieve an optimal end point, and they are weak and divergent in endorsing liver tissue examination prior to the termination of treatment.
Treatment End Points and Documentation of Histological Resolution
Patients who are treated until they exhibit normal liver tissue prior to corticosteroid withdrawal have the best opportunity to sustain their remission. The frequency of relapse can be reduced from ≥50% to 28% after achieving a normal histological end point. Normalisation of the serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), bilirubin and c-globulin levels indicates the best opportunity to document histological resolution, but it does not establish its occurrence. Inflammatory activity is present in 36–45% of liver tissue specimens from patients with normal liver tests during treatment, and these patients remain at high risk for relapse after drug withdrawal. Histological resolution lags behind laboratory resolution by 3– 8 months, and treatment should be continued for at least this period before histological assessment. The AASLD guidelines recommend that corticosteroid therapy be continued until complete resolution of laboratory tests and liver tissue, and the histological assessment is best performed 3–8 months after laboratory resolution.
Normal liver tests are achieved in 66–91% of treated patients within 2 years, and the average duration of treatment until normalisation of these tests (serum AST, ALT, bilirubin and c-globulin levels) is 19 months. As histological resolution may not occur for an additional 3–8 months, the expected treatment duration to achieve normal liver tests and liver tissue in accordance with the AASLD guidelines is 22–27 months. The guidelines of the BSG support treatment until normal liver tests, but they are not as dogmatic as the AASLD guidelines regarding liver tissue examination prior to drug withdrawal. The BSG recommendation of continuous treatment for at least 2 years, including at least 12 months beyond laboratory resolution, encourages a duration of therapy that is likely to be similar to that required to satisfy the AASLD guidelines. The BSG guidelines assume that the duration of therapy will be sufficient to achieve histological resolution rather than requiring its documentation.
The clinical judgment at this decision point must recognise that the opportunity to induce a sustained long-term remission without further treatment is limited and that this opportunity can be strengthened by the assurance that histological resolution has been achieved ( Table 5 ). Counterbalancing this decision is the acceptance that continuous treatment for at least 2 years with sustained normal liver tests for at least 1 year is usually sufficient time to optimise the histological response without further documentation and that relapse is still possible even after treatment to normal liver tests and liver tissue in 20–28% of patients. The deciding factor regarding the need for liver tissue examination prior to drug withdrawal may relate to the importance of the tissue findings in developing subsequent courses of action.
An exacerbation of the autoimmune hepatitis in a patient with incomplete histological resolution constitutes premature discontinuation of therapy, not relapse of the disease. The medications should not have been withdrawn according to the AASLD guidelines, and an exacerbation under these circumstances justifies resumption of the conventional corticosteroid regimen. An exacerbation of autoimmune hepatitis in a patient with complete histological resolution constitutes an exacerbation despite an optimal histological end point. This patient has truly relapsed, and long-term (indefinite) maintenance therapy with azathioprine (2 mg/kg daily) or low-dose prednisone (≤10 mg daily) is justified. The certainty of histological resolution prior to drug withdrawal directs the immediate and long-term management of autoimmune hepatitis at this decision point, and the performance of a liver tissue examination prior to drug withdrawal should be considered ( Table 5 ).