The aim of this study was to identify the perioperative risk factors for postoperative bile leakage after hepatic resection and to propose a treatment strategy for such leakage when it does occur. Between 1992 and 2000 a total of 313 hepatic resections without choledocojejunal anastomosis were performed at our institute. Risk factors related to bile leakage were identified with univariate analysis, and strategies were evaluated in relation to the findings of postoperative fistulography. Postoperative bile leakage developed in 17 patients (5.4%). Univariate analysis identified high risk factors as advanced age, a wide surface area of the incision (bile leakage group versus no bile leakage group: 102.1 vs. 66.4 cm(2), p < 0.05), and exposure of Glisson's sheath at the cut surface (e.g., central bisegmentectomy, S4, S8 subsegmentectomy). Groupings of patients by their postoperative fistulography results showed that patients with involvement of the proximal bile duct were slower to heal than those with no demonstrable bile duct involvement. The one patient whose fistulogram demonstrated peripheral bile duct involvement had uncontrollable leakage and required reoperation. Hepatectomies with a wide surface area and those that expose the major Glisson's sheath present serious risk factors for bile leakage. When the fistulogram shows proximal bile duct involvement, endoscopic nasobiliary tube drainage is necessary; when the fistulogram shows peripheral bile duct involvement, reoperation is needed.

No abstract

Neuroendocrine tumors including carcinoid tumors and pancreatic endocrine tumors are uncommon, and the genetic alterations in these indolent tumors are not well characterized. We studied global hypomethylation by analyzing long interspersed nucleotide elements (LINE)-1 and Alu methylation using pyrosequencing in 35 neuroendocrine tumors and corresponding normal tissue. The tumor samples were less methylated than normal tissue at LINE-1 (P ¼ 0.04) and Alu (P ¼ 0.001). The mean relative tumor hypomethylation (difference in methylation between normal tissue and in tumor) was 11.5710.0 for LINE-1 and 5.876.4 for Alu, and were correlated with each other (correlation coefficient 0.6, P ¼ 0.001). Relative tumor hypomethylation of LINE-1 was higher in ileal carcinoid tumors than in non-ileal carcinoid tumors and pancreatic endocrine tumors (P ¼ 0.047), and tumors with lymph node metastasis (P ¼ 0.02), chromosome 18 loss (P ¼ 0.001) and RAS-association domain family 1, isoform A gene methylation (P ¼ 0.02). Alu methylation in tumors was inversely correlated with methylation of O 6 -methyl-guanine methyltransferase gene (P ¼ 0.02). Our study shows that hypomethylation is more common in carcinoid tumors than in pancreatic endocrine tumors and is associated with clinicopathologic features, and genetic and epigenetic alterations in these tumors, including lymph node metastasis.

Little correlation was observed between imaging response of colorectal cancer liver metastases to chemotherapy and pathologic response. Liver surgery should be undertaken even after a complete response by imaging. Outcome after hepatectomy was favorable in patients showing complete pathologic response of least one metastasis.

No abstract