BACKGROUND: Cryptogenic chronic active hepatitis may be an autoimmune or viral disease. Our aims were to determine the clinical features, human leukocyte antigen phenotype, and response to corticosteroid therapy of severe cryptogenic chronic active hepatitis and to compare it with these other diseases. METHODS: Twelve patients with cryptogenic hepatitis were compared with 94 patients with autoimmune hepatitis and 30 patients with chronic viral hepatitis. RESULTS: Patients with cryptogenic hepatitis were indistinguishable from those with autoimmune hepatitis by age, gender, and individual laboratory and histological findings. HLA B8 (75% vs. 49%, P = 0.2), DR3 (71% vs. 51%, P = 0.5), and A1-B8-DR3 (57% vs. 38%, P = 0.6) occurred as commonly in each group. Patients with cryptogenic hepatitis entered remission (83% vs. 78%, P > 0.9) and failed treatment (9% vs. 11%, P > 0.8) as frequently as those with autoimmune hepatitis during corticosteroid therapy. In contrast, patients with chronic viral hepatitis had lower biochemical abnormalities, less frequent multilobular necrosis at presentation, and different human leukocyte phenotypes than those with cryptogenic or autoimmune disease. CONCLUSIONS: Severe cryptogenic hepatitis has a clinical expression, genetic phenotype, and corticosteroid responsiveness that is similar to autoimmune hepatitis. It may be an autoimmune disorder that has escaped detection by conventional immunoserological markers.
BACKGROUND: Cryptogenic chronic active hepatitis may be an autoimmune or viral disease. Our aims were to determine the clinical features, human leukocyte antigen phenotype, and response to corticosteroid therapy of severe cryptogenic chronic active hepatitis and to compare it with these other diseases. METHODS: Twelve patients with cryptogenic hepatitis were compared with 94 patients with autoimmune hepatitis and 30 patients with chronic viral hepatitis. RESULTS:Patients with cryptogenic hepatitis were indistinguishable from those with autoimmune hepatitis by age, gender, and individual laboratory and histological findings. HLA B8 (75% vs. 49%, P = 0.2), DR3 (71% vs. 51%, P = 0.5), and A1-B8-DR3 (57% vs. 38%, P = 0.6) occurred as commonly in each group. Patients with cryptogenic hepatitis entered remission (83% vs. 78%, P > 0.9) and failed treatment (9% vs. 11%, P > 0.8) as frequently as those with autoimmune hepatitis during corticosteroid therapy. In contrast, patients with chronic viral hepatitis had lower biochemical abnormalities, less frequent multilobular necrosis at presentation, and different human leukocyte phenotypes than those with cryptogenic or autoimmune disease. CONCLUSIONS: Severe cryptogenic hepatitis has a clinical expression, genetic phenotype, and corticosteroid responsiveness that is similar to autoimmune hepatitis. It may be an autoimmune disorder that has escaped detection by conventional immunoserological markers.