Getting started as a hepatobiliary surgeon: lessons learned from the first 100 hepatectomies as a consultant

Jonathan Koea

2005-02-25

Abstract:Aim: Liver resection has historically been regarded as difficult and dangerous surgery associated with significant perioperative mortality and morbidity rates. Partly as a result of this, adequate training in hepatic surgery has been difficult to obtain with most surgical trainees exposed only to hepatic trauma and damage control scenarios. This report describes the first 100 liver resections undertaken as a surgical consultant. Method: Clinical, diagnostic, pathological, and follow-up data were collected prospectively on 100 patients undergoing liver resection, and was stored in a computerised database. Factors associated with morbidity and trends in operative and perioperative variables over the period of the study were then analysed. Results: Malignant tumours were the most common indication for hepatic resection (88 cases) with 40 resections being undertaken for metastatic colorectal cancer. A further 34 resections were performed for cholangiocarcinoma, gallbladder cancer, or hepatocellular carcinoma. Fifty-six patients underwent a lobectomy (right lobectomy 38, left lobectomy 18)--while a further 13 patients underwent extended resections and 31 patients underwent segmental hepatic resections. The median blood loss for all patients was 375 ml (range 100-5800 ml), and 48 patients required red cell transfusion at any time during their hospital admission. The median hospital stay was 7 days (range 4-38 days), with 52% of the patients developing complications of which 7 patients experienced a major complication, and a 3% mortality rate. The risk of complications was directly related to the magnitude of resection, with 11 of 31 patients undergoing segmental resections developing complications. In comparison, 28 of 56 patients undergoing lobectomy (and all 11 patients undergoing extended resections) developed perioperative complications. Over the study period there was an increase in the use of segmental resections and extended resections with a decrease in blood loss, in-flow occlusion, and mortality. However, hospital stay and morbidity rates remained constant. On multivariate analysis operative blood loss greater than and equal to 1000 ml, resection of greater than and equal to 4 hepatic segments, and the presence of abnormal parenchyma (cirrhosis, steatosis or fibrosis) were independent predictors of perioperative morbidity and mortality. Conclusions: Hepatic resection can be undertaken safely, and increasing experience as a hepatic surgeon is associated with greater utilisation of parenchymal sparing and extended resections (without the routine use of in-flow occlusion).