A. Bosma

Prolonged continuous or intermittent vascular inflow occlusion during hemihepatectomy in pigs.

OBJECTIVE To assess ischemia and reperfusion (I/R) injury in a hemihepatectomy model in pigs after prolonged continuous or intermittent vascular inflow occlusion in the liver. SUMMARY BACKGROUND DATA Massive intraoperative blood loss during liver resections can be prevented by temporary vascular inflow occlusion, consequently leading to ischemia and reperfusion injury in the remnant liver. Previously, in a pig liver resection model in which only limited I/R injury was induced during brief (90 min) vascular inflow occlusion, the authors demonstrated reduced I/R injury after continuous (CNT) occlusion, compared to intermittent (INT). This liver resection study on pigs was undertaken to assess I/R injury after prolonged (120 min) CNT or INT occlusion. METHODS In pigs (37.0 +/- 1.5 kg), liver ischemia during 2 hours was CNT (n = 6) or INT (n = 6) (eight subsequent periods of 12 min ischemia and 3 min recirculation), followed by 6 hours of reperfusion. A left hemihepatectomy (45.5% +/- 1.4%) was performed within the first 12 minutes of ischemia. No hepatic pedicle clamping or liver resection was performed in control experiments (n = 6). Microvascular damage was assessed by hyaluronic acid (HA) uptake capacity of the liver (parameter of early sinusoidal endothelial cell damage) and restoration of intrahepatic tissue pO2 during reperfusion. Hepatocellular damage was tested by plasma concentrations of aspartate aminotransferase (AST), alanine aminotransferase, and lactate dehydrogenase (LDH). RESULTS Hyaluronic acid uptake after 6 hours of reperfusion, compared to preischemic uptake, was unaltered in the control group, but was significantly reduced in both resection groups. However, more HA was taken up after INT occlusion, compared to CNT (60.4% +/- 5.6% and 39.5% +/- 3.7%, respectively; ANOVA: p = 0.001). Intrahepatic tissue pO2 distribution after 6 hours of reperfusion more closely returned to preischemic configuration in the INT group than in the CNT group, indicating reduced microcirculatory disturbances after INT occlusion. Release of AST and LDH after 6 hours of reperfusion was significantly increased in both CNT and INT groups. Lower AST levels, however, were found after INT occlusion than after CNT occlusion (267.0 +/- 74.7 U/l and 603.3 +/- 132.4 U/l, respectively; p = 0.06). CONCLUSIONS Intermittent hepatic vascular inflow occlusion during prolonged liver ischemia in pigs resulted in less microcirculatory and hepatocellular injury, compared to continuous occlusion. Intermittent clamping is preferable when prolonged periods of vascular inflow occlusion are applied during liver resections.

Long-term survival after resection of proximal bile duct carcinoma (Klatskin tumors)

This retrospective study in 79 surgically treated patients with a proximal bile duct carcinoma, revealed 12 patients [median age (range): 59.5 years (21-73)] who survived for more than 5 years. These 12 patients were analyzed in order to identify specific patient characteristics for long-term survival. Fifteen patients died due to postoperative complications and were excluded from this survival analysis. In relation with preoperative Bismuth classification, there were 3/15 (20%) longterm survivors with type I tumors and 9/26 (35%) long-term survivors with type II tumors. In the group of type III and IV tumors, there were no long-term survivors. Concerning type of resection, 9/51 (18%) patients had long-term survival after local resection and 3/13 (23%) patients after local resection combined with hemi-hepatectomy. Complete tumor free surgical specimen margins were found in only 4/64 cases (6%), of which only one survived for more than 5 years. Negative proximal bile duct margins, absence of multifocality and diploid tumors showed a significant correlation with long-term survival. There was no significant correlation between long-term survival and postoperative radiotherapy Of the 12 long-term survivors, five died after 5 years, of which 2 had developed metastases and one a local recurrence. The other two died of a metastasis of an ovarian adenocarcinoma and of cachexia, respectively. The mean survival of the 64 patients analyzed in this study (in which hospital mortality was excluded) was 33.7 months, with a median survival of 18.8 months. In conclusion, preoperative Bismuth classification of the tumor, absence of multifocality, diploid type tumors, and negative proximal bile duct margins at histopathological examination, were found to be the only significant prognostic factors for long-term survival.