HLA typing and autoantibodies in hepatitis B surface antigen-negative chronic active hepatitis

An unusual form of autoimmune hepatitis in young Somalian men

BACKGROUND

Significant diversity in disease severity has been identified for autoimmune disorders among different ethnic groups. Current knowledge of both the natural history and management of autoimmune hepatitis (AIH) has been derived from European or Japanese patients, and there is limited information about the disease in patients from other ethnic groups.

AIMS

To assess the clinical, histological and immunological features of AIH in patients from Somalia and to determine their response to therapy.

METHODS

Retrospective review of a cohort of young Somalian men with atypical AIH compared with a control group of European patients.

RESULTS

The six Somalian men were younger at presentation (median age 37 (range 24-59) years) than the seven female and three male European controls (55 (34-54) years, P = 0.06). The Somalians had slightly more severe disease at presentation-median modified Ishak stage of 2.5 compared with 2 in Europeans (P = 0.61) and four (66%) had features of cholestasis compared with only one (10%) European patient (P = 0.04). Therapy with prednisolone and azathioprine was completely effective for eight of 10 Europeans but only one of seven Somalians (P = 0.04). Analysis of human leucocyte antigen types revealed differences between the Somalian and European patients, although these differences did not reach statistical significance.

CONCLUSIONS

Somalian men with AIH present with cholestatic features and respond poorly to standard immunosuppressive regimes.

Susceptibility to autoimmune chronic active hepatitis: human leukocyte antigens DR4 and A1-B8-DR3 are independent risk factors

After nearly 18 years of research, the association between human leukocyte antigens A1-B8-DR3 and autoimmune chronic active hepatitis still provokes debate. The principal reasons for this are disease heterogeneity and racial variation in the distribution of human leukocyte antigens between populations. The aim of the present study was to reexamine the relationship between these antigens and autoimmune chronic active hepatitis in a well-characterized series of patients. Ninety-six outpatients with autoimmune chronic active hepatitis and an additional 14 referred for liver transplantation with end-stage autoimmune chronic active hepatitis were studied. Human leukocyte antigen frequencies were compared with those of 100 racially and geographically matched controls. The A1-B8-DR3 haplotype was present in 38% of patients compared with 11% of controls (chi 2 = 20.6, p less than 0.0005). When all the DR3-positive patients were eliminated, there was a striking secondary association with DR4; 35 (80%) of 44 remaining patients were DR4 positive compared with 31 (39%) of 79 DR3-negative controls (Fisher's exact probability test p = 0.000031, pc = 0.0013). In addition patients with A1-B8-DR3 are seen at a significantly younger age than those without (39.75 yr vs. 48.21 yr, p less than 0.025), relapse more frequently (52% of patients with A1-B8-DR3 relapsed on one or more occasions compared with 34% of patients without this haplotype) and are more frequently referred for liver transplantation. These data indicate for the first time that two genes within the major histocompatibility complex closely linked to the DR3 and DR4 genes independently confer susceptibility to autoimmune chronic active hepatitis.

Clinical and prognostic implications of HLA B8 in corticosteroid-treated severe autoimmune chronic active hepatitis

To assess the clinical and prognostic implications of human leukocyte antigen B8 in corticosteroid-treated severe autoimmune chronic active hepatitis, 81 consecutive patients were tested for histocompatibility antigens on the A and B loci, treated with corticosteroids, and followed prospectively for 111 +/- 8 mo. The 47 patients with HLA-B8 were younger (38 +/- 2 yr vs. 48 +/- 2 yr; p less than 0.01), had higher serum levels of aspartate aminotransferase (658 +/- 60 U/L vs. 465 +/- 49 U/L; p = 0.02) and bilirubin (7 +/- 1 mg/dl vs. 2.8 +/- 0.4 mg/dl; p = 0.003), and more commonly had histologic features of bridging necrosis, multilobular necrosis, and cirrhosis (85% vs. 56%; p less than 0.01) at presentation than the 34 patients without HLA-B8. Remission (79% vs. 71%), relapse after drug withdrawal (76% vs. 71%), treatment failure (13% vs. 6%), progression to cirrhosis (46% vs. 32%), and death from liver failure (6% vs. 3%) occurred as frequently in patients with and without HLA-B8. Importantly, HLA-B8-negative patients with HLA-A1 relapsed less frequently than HLA-B8-positive patients with and without HLA-A1- and HLA-B8-negative counterparts without HLA-A1. It is concluded that HLA-B8-positive patients are younger and have more severe disease at presentation than HLA-B8-negative patients. The HLA-B8 phenotype does not influence the response to corticosteroid therapy. HLA-B8-negative patients with HLA-A1 relapse less frequently than other phenotypes.

Clinical features and prognostic implications of severe corticosteroid-treated cryptogenic chronic active hepatitis

To assess the nature and prognosis of severe chronic active hepatitis of unknown cause, we compared 26 patients who had been fully screened for etiologic factors with 112 patients who had autoimmune chronic active hepatitis after similar durations of corticosteroid therapy (17(+)/- 2 versus 23 (+)/- 2 months), and follow-up versus 103 +/- 7 months). Patients with cryptogenic disease could not be distinguished from those with autoimmune disease on the basis of age, sex distribution, duration of illness, immunoglobulin levels, frequency of concurrent immunologic disorders, or histologic findings. Serum gamma-globulin levels were higher (3.4 +/- 0.1 versus 2.5 +/- 0.2 g/dl, P = 0.007) and albumin levels were lower (2.9 +/- 0.1 versus 3.3 +/- 0.1 g/dl, P = 0.003) in patients with autoimmune disease than in those with cryptogenic disease, but individual findings did not differentiate the patients. Remission (69 versus 75%), treatment failure (23 versus 13%), relapse after drug withdrawal (67 versus 68%), progression to cirrhosis (57 versus 36%), and death from hepatic failure (12 versus 11%) occurred as commonly in patients with cryptogenic as in those with autoimmune disease. Patients with different constellations of immunoserologic findings were similar clinically. We conclude that patients with severe chronic active hepatitis who have been fully screened for etiologic factors cannot be distinguished from patients with autoimmune disease of comparable severity. These two groups of patients have a similar prognosis after corticosteroid therapy, and such treatment should be considered in these highly selected patients.

The nature of unexplained chronic aminotransferase elevations of a mild to moderate degree in asymptomatic patients

To determine the nature of unexplained chronic serum aspartate aminotransferase elevations of a mild to moderate degree in asymptomatic patients, we performed systematic clinical, biochemical and histologic examinations in 47 individuals who had been screened for virus-, alcohol- or drug-related disease. Serum aspartate aminotransferase levels ranged from 3- to 8-fold normal (mean: 156 +/- 7 units per liter) for at least 6 months (mean: 30 +/- 6 months). Serum alanine aminotransferase levels were also increased but to a lesser degree in most patients. Thirty-four patients (72%) had histologic features of chronic active hepatitis, including 16 with cirrhosis. Ten patients (21%) had steatohepatitis and three (6%) had miscellaneous disorders. Patients with chronic active hepatitis and cirrhosis could not be distinguished from counterparts without cirrhosis by individual clinical or laboratory findings. Antinuclear or smooth muscle antibodies were detected in 18 of the patients with chronic active hepatitis (53%). All patients with steatohepatitis were women, and they had laboratory changes at presentation, including seropositivity for antinuclear antibodies, that overlapped with those of patients with chronic active hepatitis. We conclude that asymptomatic patients with unexplained chronic aspartate aminotransferase elevations of a mild to moderate degree frequently have chronic active hepatitis and that many have cirrhosis. Immunoserologic findings compatible with autoimmune hepatitis are commonly present. Steatohepatitis is the most frequent alternative diagnosis, especially in women, and it is not excluded by the presence of antinuclear antibodies. Differentiation of the disorders is possible only by histologic examination.