Chronic Hepatitis with Combined Features of Autoimmune Chronic Hepatitis and Chronic Hepatitis C: Favorable Response to Prednisone and Azathioprine

Case Reports

Patient 1

A 74-year-old white woman had an 18-month history of right-sided abdominal pain, progressive fatigue, and persistently elevated serum alanine aminotransferase levels ranging from 1.18 to 1.55 µkat/L. Other medical problems included insulin-dependent diabetes mellitus, hypertension, and hypothyroid disease. She reported no history of blood transfusion, intravenous drug abuse, tattoos, or excessive alcohol use. Her daily medications included 25 mg of hydrochlorothiazide, 16 units of humulin U-100 insulin, 10 mg of enalapril, 0.01 mg of synthroid, and 1 g of acetaminophen. She had no family history of liver disease. Physical examination showed no cutaneous stigmata of chronic liver disease, hepatosplenomegaly, or ascites. Laboratory data are shown in Table 1. Results of abdominal ultrasound and computed tomographic scans and upper gastrointestinal series were unremarkable. Examination of a liver biopsy specimen showed features of moderately severe chronic hepatitis with a Knodell hepatitis activity index score [13] of 16/22 (Figure 1, top).

Given the patient's immune serologic results and hypergammaglobulinemia, she was treated with 30 mg of prednisone daily and 50 mg of azathioprine daily despite positive results for anti-HCV antibody (by recombinant immunoblot assay) and HCV RNA [as shown by results of a polymerase chain reaction]. She responded clinically and biochemically, with alanine aminotransferase levels returning to normal 8 months after therapy began. The prednisone dose was adjusted accordingly, and she has been well for 3 years while receiving 5 mg of prednisone daily and 50 mg of azathioprine daily. Examination of a liver biopsy specimen obtained 3 years after treatment showed marked improvement (Figure 1, bottom); the Knodell hepatitis activity index score was 6/22. This was accompanied by resolution of hypergammaglobulinemia. The patient remains positive for HCV RNA, with an increased quantitative level of HCV RNA as determined by the semi-quantitative method described by Ulrich and colleagues [14].

Patient 2

A 38-year-old white woman had a 19-month history of persistently elevated alanine aminotransferase levels. She had had rheumatoid arthritis since the age of 18 years and had received various nonsteroidal anti-inflammatory drugs, most recently piroxicam. Therapy with this drug had been discontinued for 6 months, with no substantial change in serum alanine aminotransferase levels, which ranged from 1.16 to 3.5 µkat/L. She had no history of hepatitis or blood transfusion. She reported intravenous drug abuse between the ages of 18 and 20 years. Other medical problems included gastroesophageal reflux disease and hyperprolactinemia. She reported previous excessive use of alcohol. Her medications included 150 mg of ranitidine twice daily, 10 mg of metoclopramide three times daily, 400 mg of hydroxychloroquine sulfate daily, 2.5 mg of indapamide daily, 0.5 mg of alprazolam daily, and 5 mg of bromocriptine mesylate daily. Physical examination showed a firm, nontender liver edge 1 inch below the right costal margin, without splenomegaly, ascites, or cutaneous stigmata of chronic liver disease. Joint changes of inactive rheumatoid arthritis were observed. Laboratory data are shown in Table 1. Examination of a liver biopsy specimen showed moderately severe chronic hepatitis; the Knodell hepatitis activity index score [13] was 12/22. Results of anti-HCV tests by first-generation enzyme-linked immunosorbent assay were negative.

Because of a positive antinuclear antibody titer of 1:640 and hypergammaglobulinemia, she was given 40 mg of prednisone daily and 50 mg of azathioprine daily as a therapeutic trial. She responded clinically and biochemically, with serum alanine aminotransferase levels returning to normal 5 months after the initiation of therapy. However, results of a repeat anti-HCV antibody test by second-generation enzyme-linked immunosorbent assay were positive, and HCV RNA was detected by polymerase chain reaction. Because cushingoid features developed, she discontinued prednisone therapy on her own, which resulted in a flare-up of hepatitis and an elevation of alanine aminotransferase levels to 3.3 µkat/L. Therapy with prednisone and azathioprine was reinstituted, and serum alanine aminotransferase levels rapidly returned to normal. Since then, she has remained well while receiving 10 mg of prednisone daily and 50 mg of azathioprine daily. Examination of a liver biopsy specimen obtained 2 years after the initiation of therapy showed marked reduction of inflammation; the Knodell hepatitis activity index score was 7/22. She continues to be well and remains positive for HCV RNA, with no apparent change in the level of viremia as determined by the semi-quantitative methods described by Ulrich and colleagues [14].

Discussion

The association between autoimmune chronic hepatitis and chronic hepatitis C remains controversial. In this context, the two cases we report provide interesting insights. Both patients had serologic and histologic features of autoimmune chronic hepatitis and chronic hepatitis C, and both had an excellent therapeutic response to immunosuppressive therapy. Although anti-liver-kidney microsomal antibody has been reported with HCV-associated autoimmune chronic hepatitis, neither of our patients had this antibody. Both patients were positive for HCV RNA by the polymerase chain reaction, which excludes the possibility of a false-positive result on the anti-HCV antibody test. Thus, HCV could have acted as a trigger for autoimmune chronic hepatitis, or the patients could have had underlying autoimmune chronic hepatitis and incidentally became infected with HCV. Determining the sequence of events is difficult. In any case, the combined features of autoimmune chronic hepatitis and chronic hepatitis C raise an important therapeutic dilemma because the therapies for each disease are believed to be mutually exclusive.

Nishiguchi and associates [15] described 21 patients with steroid-responsive autoimmune hepatitis, 2 of whom tested positive for HCV RNA. In contrast, Magrin and colleagues [16] described positive recombinant immunoblot assay test results and serum HCV RNA in 15 patients whom they considered to have “autoimmune” hepatitis. None of their patients responded to prednisone therapy, whereas the few who were treated with interferon responded. The difficulty in interpreting their results stems from the broad criteria used to define autoimmune hepatitis. Autoantibodies are frequently found in low titers in patients with chronic hepatitis C, and most such patients who test positive for HCV RNA by the polymerase chain reaction have chronic hepatitis C and do not have autoimmune hepatitis. However, more recently the same investigators [12] described histologic improvement in 4 of 12 patients with autoimmune hepatitis treated with prednisone who tested positive for HCV RNA, and the improvement occurred despite increased serum HCV RNA titers. Two of the six patients responded to subsequent interferon treatment after not responding to prednisone therapy. Shindo and coworkers [17] reported exacerbation of liver disease during interferon-α therapy for chronic hepatitis C in a patient who tested positive for both HCV RNA and antinuclear antibody but who responded partially to subsequent prednisone therapy. Both of our patients met the narrow criteria for autoimmune hepatitis; that is, they had high titers of autoantibodies and steroid-responsive chronic hepatitis. Our concern with prednisone therapy was the risk for increasing HCV replication under immunosuppression and, thus, worsening the liver disease. However, in patient 2, the degree of viremia remained stable despite immunosuppression; in addition, patient 1 showed biochemical and histologic improvement despite an apparent increase in the degree of viremia. This suggests that the risk for exacerbation of liver disease caused by HCV by immunosuppressive therapy may be substantially less than the risk for exacerbation of autoimmune chronic hepatitis by interferon-α therapy. An unequivocal steroid response (that is, return of aminotransferase levels to normal and resolution of hypergammaglobulinemia) may be the only way to differentiate a disease that is primarily immune-mediated from chronic hepatitis C, in which injury reflects viral effects on the liver. This is different from the modest decreases in aminotransferase levels that are sometimes seen with steroids in chronic hepatitis C, especially in those patients with immunologic manifestations. Whether these patients should be further treated with interferon while they are in remission and receiving steroids is debatable. We recommend that patients with combined features of autoimmune chronic hepatitis and chronic hepatitis C be treated with prednisone and azathioprine first; only if they do not respond should interferon therapy be considered.