To determine the life expectancy of patients with primary biliary cirrhosis, we analyzed survival data from 280 patients with either symptomatic (243) or asymptomatic (37) disease. Patients were followed for up to 19 years (mean, 6.9 years). The average length of survival was 11.9 years--nearly twice that reported in other studies. In contrast, over a 12-year period the survival of the asymptomatic patients after diagnosis did not differ from that of a control population matched for age and sex. Jaundice, weight loss, hepatomegaly, splenomegaly, and ascites were each associated with a poor prognosis. Prognosis also correlated with the histologic stages of hepatic fibrosis, cholestasis, and periportal-cell necrosis. A multivariate analysis of clinical features revealed that at the onset of disease, age, hepatomegaly, and elevated levels of serum bilirubin were independent discriminators of a poor prognosis. A histologic finding of fibrosis limited to portal areas improved this discrimination, correlating with prolonged survival. No other factors enchanced the prediction of risk.

No abstract

No abstract

Granulomata occur in the liver not only in patients with systemic granulomatous disease, but also in a variable number with underlying liver disease and in a heterogeneous group of disorders that appear to be neither hepatic nor granulomatous in nature. The hepatic granulomata found in association with liver disease are rarely attributable to complicating systemic granulomatous disease, and probably represent a nonspecific response to the underlying hepatic disease. In the heterogeneous group of diseases that appear to be neither hepatic nor granulomatous in nature, hepatic granulomata may (in some instances) represent a nonspecific response to such conditions as intraabdominal malignancy and ulcerative bowel disease. However, in others, particularly those with unexplained prolonged fever, hepatic granulomata may be attributable to specific agents that are overlooked or escape detection by currently available diagnostic measures. The etiology of hepatic granulomata can seldom be established on histological grounds alone, and usually requires collateral clinical and laboratory evidence for identification.

No abstract