Long-term prognosis of patients with alcoholic liver cirrhosis: a 15-year follow-up study of 100 Norwegian patients admitted to one unit.

BACKGROUND

Most follow-up studies in patients with alcoholic liver cirrhosis have been for a 5-year period or less. The aim of this study was to assess the long-term mortality and causes of death among patients with alcoholic liver cirrhosis and to identify predictors of mortality.

METHODS

One hundred patients with alcoholic liver cirrhosis, consecutively admitted to one medical department, were included in the study from May 1984 until December 1988. All patients had a history of alcohol abuse of at least 100 g ethanol daily for several years. The study comprised 65 men and 35 women with a median age of 58 years (range 34-82). Percutaneous liver biopsies and/or autopsies were obtained on 89 patients. Sixty-seven had ascites at admission and 34% had bleeding oesophageal varices. All patients were followed prospectively until death or until October 2000.

RESULTS

During the follow-up period 90% of the patients died, 68 of whom (76 %) had been autopsied. The cumulative actuarial mortality after 1, 3, 6 and 12 months was 18%, 28%, 36% and 49%, respectively and after 5, 10 and 15 years 71%, 84% and 90%, respectively. None of the patients underwent liver transplantation during the study. The causes of death were bleeding, liver failure or a combination of these two conditions in 52 of 90 patients (58%), while 9 (11%) died of hepatocellular carcinoma 0.5 to 73 months after inclusion in the study. Using the Cox regression analysis, age, alcohol abuse and alkaline phosphatase were independent and significant predictors of mortality, but Child-Pugh class was not.

CONCLUSIONS

The mortality in a group of patients with advanced alcoholic cirrhosis was extremely high with 5 and 15 years' mortality in 71% and 90%, respectively. Independent predictors of a poor prognosis were high age, continuous alcohol consumption of more than 10 g ethanol per day and high levels of alkaline phosphatase.