Liver surgery is safe in compensated liver cirrhosis

Abstract

Background: Liver resection is a standard treatment for primary and secondary livertumours. However, primary liver tumours often arise from cirrhotic livers. The aim of this study was to compare outcome for liver resection in cirrhotic and non-cirrhotic liver. Material and Methods: 264 patients who underwent liver resection between 2008 and 2017 at the university hospital Mainz were analysed. They were categorized into three groups: group 1: HCC in cirrhosis (n=84, 32%), group 2: HCC in non-cirrhotic liver (n=96, 36%) and group 3: CRC-metastases (n=84, 32%) matched to group. Comorbidity following the Charlson Comorbidity Index, pre- and postoperative liver function according to Meld- and Child score as well as encephalopathy and ascites, postoperative complications according the Dindo-Clavien classification and the extent of liver resection were registered. ROC analysis was performed in order to determine cut-off levels for morbidity and mortality. P-values <0.05 were='' considered='' significant='' br=''> Results: Morbidity was not different between HCC in cirrhosis and non-cirrhotic livers, but HCC in non-cirrhotic livers had significantly more major resections (p<0.05). Patients with CRC-Metastases had significantly fewer preoperative liver decompensation, and had fewer postoperative complications grade IIIa-IVb compared to the patients with HCC in cirrhosis (29.8% vs 16.7%, p=0.045). Mortality was 1.2% in group 1, 2.1% and 1.2% in groups 2 and 3, respectively (ns). There was no difference in the postoperative mortality between the groups. ROC-analyses revealed that the preoperative Meld-Score was a poor predictor for postoperative complications as well as for postoperative liver function. Comorbidity had no statistically relevant effect on the outcome, too. Conclusion: Surgery in liver cirrhosis can be safely performed with very low mortality. As neither preoperative Meld-Score nor comorbidity is a reliable predictor for the postoperative outcome each patient should be evaluated individually.