Back to Annual Meeting Program

Patients with cirrhosis have an increased risk of postoperative mortality. In addition, patients with cirrhosis per se have a reduced life expectancy. Little is known about the combined effect of these reduced outcomes after surgery. We thus evaluated early and long-term survival after surgery in patients with cirrhosis.
Methods: For prediction of long-term outcome we evaluated survival after general surgical procedures performed during the last decade by two different approaches: I) Actuarial survival was estimated in 180 patients after surgery (thus including postoperative deaths) and II) Survival was analyzed separately in 143 patients who were discharged after postoperative treatment (i. e. patients without in-hospital mortality) and with survival information available. Survival was analyzed by actuarial methods, risk factors were assessed univariately (log rank) and multivariately (Cox regression, multiple models).
Results: I) Survival in all 180 operated patients was 54% after one and 25% after 5 years (median survival 1.24 years). In univariate analysis the CHILD-score (p<0.001), MELD-score (p<0.001), ASA-score (p=0.05), emergency procedures (p<0.001), viral hepatitis (p<0.01 vs alcoholic/other origin), hyponatremia (p<0.01) and major procedures (p<0.03 vs minor) were associated with decreased survival. In multivariate analysis CHILD- (p<0.02) and MELD-score (p<0.001), ASA-class (p<0.01), preoperative hyponatremia and thrombocytopenia were independently associated with poor prognosis. II) Survival in 143 patients discharged after surgery was 68% after one and 32% after 5 years (median survival 2.8 years). Long-term survival (univariately) correlated with CHILD- and MELD-scores (p<0.01/<0.001), (preoperative) hyponatremia (p<0.01) and ASA class (p<0.05). In multivariate analysis the MELD-score (p<0.001) and hyponatremia (p<0.01) but not significantly the CHILD-score (p=0.06) or ASA-class independently predicted the outcome after hospital discharge. Neither an underlying malignant disease nor age nor emergency operations independently correlated with long-term survival.
Conclusions: Long-term survival in patients with liver cirrhosis requiring general surgery is poor. In the entire patient group poor liver function and, in part, factors influencing postoperative mortality like comorbidity (ASA) or thrombocytopenia are prognostic factors. In patients surviving the early postoperative period the long-term outcome is determined mainly by the natural course and severity of liver disease (MELD better than Child).