To the Editor:
We read with great interest the practice guidelines for the diagnosis and management of autoimmune hepatitis recently issued by the American Association for the Study of Liver Diseases (AASLD).1 In particular, we appreciate the new definition of biochemical remission, which now requires not only normal bilirubin and gamma-globulin levels but also normal serum aminotransferases; this is at variance with the 2002 definition,2 which considers aminotransferase levels lower than twice the upper limits of normal to be sufficient. According to the 2002 criteria, nearly 80% of patients with autoimmune hepatitis enter remission within 3 years. The recently coined new definition will result in a tremendous change in the rate of response to immunosuppressive treatment for autoimmune hepatitis.
Here we present our own experience, which has already been published in part,3 and compare the different response rates according to the 2002 and 2010 definitions of remission.
Among 163 consecutive Italian patients with autoimmune hepatitis diagnosed at a single center, 119 (73%) entered remission according to the 2002 AASLD criteria, whereas only 42 patients (26%) of the same cohort fulfilled the 2010 AASLD criteria. Among 89 patients with a follow-up longer than 60 months, 65 (73%) had aminotransferase levels lower than twice the upper limit of normal (2002 criteria), but only 23 (25.8%) consistently maintained normal aminotransferase levels (2010 criteria) with low steroid doses (2-4 mg of methylprednisolone daily or every other day). Interestingly, from a clinical standpoint, after a mean follow-up longer than 100 months, only 1 of the 23 patients (4%) fulfilling the 2010 criteria of remission experienced histological worsening of the disease (mild to severe liver histology), whereas 36 of the 66 patients (54.5%) whose aminotransferase levels did not normalize had histological (14 with severe histology and 9 with cirrhosis) or clinical evidence (11 with end-stage liver disease, 1 with decompensated cirrhosis, and 1 with hepatocellular carcinoma) of uncontrolled and evolving liver disease.
In summary, in our experience, the application of the 2010 criteria flips the previously codified remission rate from 73% to 26%. Complete-response patients have a very good long-term prognosis virtually free of significant clinical events, whereas patients whose serum aminotransferases are unable to be stably normalized are those with the highest probability of developing long-term complications, which not rarely may prove to be lethal. These are the patients most likely to benefit from new pharmacological, cellular, and molecular therapies.4, 5