Single-centre experience of laparoscopic pancreatic surgery.

BACKGROUND: Laparoscopic resection is regarded as safe and feasible in selected patients with benign pancreatic tumours. Few data exist on laparoscopic surgery for malignant lesions and larger neoplasms in unselected patients. METHODS: The study included all patients admitted to Oslo University Hospital, Rikshospitalet, from March 1997 to March 2009 for surgery of lesions in the body and tail of the pancreas, and selected patients with lesions in the pancreatic head, who underwent surgery by a laparoscopic approach with curative intent. RESULTS: A total of 166 patients had 170 operations, including 138 pancreatic resections, 18 explorations, nine resections of peripancreatic tissue and five other therapeutic procedures. Four patients had repeat procedures. There were 53 endocrine tumours (31.0 per cent), 28 pancreatic carcinomas (16.4 per cent), five cases of metastases (2.9 per cent), 48 cystic tumours (28.1 per cent) and 37 other lesions (21.6 per cent). The total morbidity rate was 16.5 per cent. Fistula was the most common complication (10.0 per cent). Three patients needed reoperation for complications. There were three hospital deaths (1.8 per cent). Median hospital stay following surgery was 4 days. CONCLUSION: Laparoscopic resection of lesions in the body and tail of the pancreas in an unselected patient series was safe and feasible, and should be the method of choice for this patient group in specialized centres.

Hepatic and splanchnic metabolism of plasma proteolytic enzymes before, during, and after clamping of the hepatic artery and portal vein.

Parameters of the kallikrein-kinin, fibrinolytic, and coagulation systems before (control), during, and after occlusion of the hepatic artery and the portal vein were studied in anesthetized pigs. Hepatic synthesis was observed for kallikrein, plasmin, and plasminogen. The other factors, prekallikrein, kallikrein inhibitor, alpha 2-antiplasmin, prothrombin, and antithrombin-III were all released by the liver and underwent partial hepatic inactivation. The splanchnic area showed a significant synthesis of antithrombin-III. After 90 min of hepatic arterial occlusion, the hepatic release of all factors declined, and synthesis of kallikrein, plasmin, and plasminogen fell an average of 54%. The inactivation of prekallikrein and alpha 2-antiplasmin was increased and the splanchnic synthesis of antithrombin-III fell. Portal vein blood flow fell by 38%. After 15 min of portal vein occlusion (combined occlusion of the hepatic artery and portal vein) the splanchnic area started to synthetize kallikrein and the hepatic inflow of kallikrein increased by more than 50%. Splanchnic metabolism of other factors was also altered, but after revascularization of the hepatic artery changes were mostly insignificant. Hepatic synthesis of kallikrein and plasmin remained depressed after reperfusion, and that of plasminogen fell to zero. After this short period of occlusion of the hepatic artery and portal vein, plasma concentrations of prekallikrein, plasmin, and alpha 2-antiplasmin were decreased, causing a state of fibrinolysis. A role for kallikrein/bradykinin in hepatic blood flow regulation is proposed and implications of these observations for liver transplantation are discussed.

Activation of the plasma contact system and hemodynamic changes after graft revascularization in liver transplantation.

In this study, the relation between activation of the plasma contact system and hemodynamic changes during orthotopic liver transplantation was evaluated. Nineteen consecutive courses of OLT in 17 adult patients were investigated. Veno-venous bypass was used in all patients. Blood samples were drawn through all phases of the procedure, and analyzed for the following parameters using functional techniques (chromogenic peptide substrate assays): plasma kallikrein (KK), prekallikrein, functional plasma kallikrein inhibition, C1 inhibitor, and alpha 2-macroglobulin. Plasma high molecular weight kininogen (HK) degradation was evaluated using the immunoblotting technique. An abrupt rise in KK activities occurred within 1 min after portal reperfusion of the liver graft (7-16 U/L, P < 0.05). Simultaneously, proteolytic breakdown of HK was seen. The elevated KK activities were maintained the next 1 1/2 hr. Ten min after graft reperfusion, a significant increase in cardiac output compared with the anhepatic phase (7.2-12.4 L/min, P < 0.05) was found. At the same time, systemic vascular resistance fell significantly (817-408 dynes x sec/cm-5, P < 0.05). The increase in plasma KK activities accompanied by simultaneous degradation of HK seen immediately after reperfusion of the liver graft may be due to contact activation as recipient blood contacts with the underlying basement membrane of injured sinusoidal endothelium in the transplanted liver. We suggest that hemodynamic changes associated with the postreperfusion syndrome seen after revascularization of the liver in OLT could at least be caused in part by bradykinin release due to contact activation.

A comparative study of the short-term outcome following open and laparoscopic liver resection of colorectal metastases.

BACKGROUND: Laparoscopic resection of liver tumors is feasible, but few studies have compared short-term outcome of the laparoscopic approach to that of a conventional technique. METHODS: Eighteen tumor resections performed during 14 procedures (14 patients) by conventional surgery were compared to 21 similar resections performed laparoscopically during 15 procedures (13 patients). All patients had colorectal liver metastases. RESULTS: No perioperative mortality occurred. Surgical time, peroperative bleeding and blood transfusion requirement were similar in the two groups. The resection margin was involved by tumor tissue in one specimen laparoscopically resected and in two specimens conventionally resected (p = 0.58). Patients operated laparoscopically remained in hospital for median 4 days, while patients operated conventionally stayed median 8.5 days (p <0.001). Patients operated laparoscopically required less opioid medication than patients having conventional surgery (median 1 vs 5 days; p = 0.001). CONCLUSIONS: Short-term outcome of laparoscopic liver resection compares to that of conventional surgery, with the additional benefits derived from minimal invasive therapy.

Laparoscopic resection of the pancreas: a feasibility study of the short-term outcome.

BACKGROUND: Laparoscopic resection is not an established treatment for tumors of the pancreas. We report our preliminary experience with this innovative approach to pancreatic disease. METHODS: Thirty two patients with pancreatic disease were included in the study on an intention-to-treat basis. The preoperative indications for surgery were as follows: neuroendocrine tumors ( n=13), unspecified tumors ( n=11), cysts ( n=2), idiopathic thrombocytopenic purpura with ectopic spleen ( n=2), annular pancreas ( n=1), trauma ( n=1), aneurysm of the splenic artery ( n=1), and adenocarcinoma ( n=1). RESULTS: Enucleations ( n=7) and distal pancreatectomy with ( n=12) and without splenectomy ( n=5) were performed. Three patients underwent laparoscopic exploration only. Four procedures (13%) were converted to an open technique. One resection was converted to a hand-assisted procedure. The mortality rate for patients undergoing laparoscopic resection was 8.3% (two of 24). Complications occurred after resection in nine of 24 procedures (38%). The median hospital stay was 5.5 days (range, 2-22). Postoperatively, opioid medication was given for a median of 2 days (range, 0-13). CONCLUSION: Resection of the pancreas can be performed safely via the laparoscopic approach with all the potential benefits to the patients of minimally invasive surgery.

Mechanism of ursodeoxycholic acid- and canrenoate-induced biliary bicarbonate secretion and the effect on glucose- and amino acid-induced cholestasis.

The mechanism of ursodeoxycholic acid (UDCA)- and canrenoate-induced bicarbonate choleresis was studied before and during the administration of glucose or amino acids in anaesthetized pigs. Previous studies have shown that the canalicular secretion has, on a molar basis, a relationship among the secretion of chloride, bicarbonate, and bile acids of 0.9, 0.3, and 1, respectively. Ductular secretion is associated with the transport of 0.25 mol chloride per 1 mol bicarbonate. In control experiments UDCA was associated with a biliary secretion of about 1.3 mol chloride and 0.5 mol bicarbonate per 1 mol bile acid, and canrenoate caused the secretion of 1.2 mol chloride per 1 mol bicarbonate. Intravenous infusion of glucose or amino acids increased these relationships, and after administration of UDCA or canrenoate, these relationships were still increased by at least 70% on average when compared with the control experiments. A reduction in bile secretion after glucose or amino acid infusion is opposed by UDCA or canrenoate. The effect of UDCA or canrenoate on bile secretion is not disturbed by glucose or amino acids. Both substances stimulate canalicular bicarbonate secretion and could be of importance in improving cholestatic conditions.

Effect of insulin on hepatic bile secretion during normoglycaemia and hyperglycaemia.

The roles of hyperosmolality, hyperglycaemia, and insulin in glucose-induced reduction of bile secretion were examined in anaesthetized pigs. Compared with normoglycaemia, intravenous infusion of isotonic glucose reduced bile acid-dependent bile secretion at a plasma glucose concentration of 18 mmol/l, with 34 +/- 4%. Lowering of plasma glucose concentration to normoglycaemia after administration of insulin (10 U/kg body wt. low dose) increased bicarbonate-dependent bile secretion by 23 +/- 3%. Induction of hyperglycaemia (plasma glucose concentration, 16 mmol/l) and the combined infusion of isotonic glucose and the low insulin dose decreased bile secretion by 22 +/- 3%. During hyperglycaemia (plasma glucose concentration, 16 mmol/l) the combined infusion of isotonic glucose and a high dose of insulin (60 U/kg body wt) increased bile acid-dependent bile secretion by 26 +/- 3%. Hyperglycaemia reduces bile secretion without altering plasma osmolality. Endogenous production (or too rapid degradation) of insulin may be too small during intravenous glucose infusion to cope with the metabolic demands of hepatocellular glucose conversion. This may be overcome by administration of insulin in a large dose, which stimulates bile acid secretion.

Role of sodium, bicarbonate, and plasma osmolality in biliary secretion.

To examine the effect of changes in biliary sodium and bicarbonate secretion on bile formation, experiments were performed on fasted, pentobarbital-anesthetized pigs. During continuous intravenous secretin infusion (2.7 CU X kg body wt-1 X h-1) sodium secretion was altered by increasing or reducing plasma sodium concentration. Bicarbonate secretion was altered by varying arterial plasma pH. At increased biliary sodium secretion, bile formation was depressed, but changes in bicarbonate secretion were accompanied by parallel alterations in bile formation. Bile acid secretion was increased during elevated plasma sodium concentration, whereas reduced plasma sodium concentration depressed bile acid secretion. To distinguish between the effect of changes in plasma osmolality and sodium concentration, bile formation was also studied during intravenous sucrose infusion at normal plasma sodium concentration. About 50% of the effect on bile formation of changing plasma sodium concentration is solely caused by the changes in plasma osmolality. During secretin stimulation bile formation is mainly determined by bicarbonate. Changes in plasma osmolality affect bile secretion through alterations in the net osmotic force across the hepatocellular membrane. Sodium has an impact on the bile-acid-dependent fraction, whereas bicarbonate is the mediator of the bile-acid-independent fraction of bile secretion.

Abolished relationship between pancreatic HCO-3 secretion and arterial pH during carbonic anhydrase inhibition.

After acetazolamide administration, CO2 hydration in pancreatic cells would be slow and might become a rate-limiting factor to pancreatic HCO-3 secretion. Correspondingly, pancreatic HCO-3 secretion-normally pH dependent-would become slow and pH-independent. However, acetazolamide would not be expected to interfere with the capacity of the secretory mechanism to generate a proton potential gradient between pancreatic cells and interstitial fluid. These predictions were examined in 5 anesthetized, secretion infused (2.7 C. U./kg b.wt. h-1) pigs. Pancreatic juice was collected from a catheter in the pancreatic duct. Arterial pH was varied through i.v. HCl and NaHCO3 infusions and CO2 addition to inspired air. Before acetazolamide, HCO-3 secretion varied with plasma pH and averaged 298 +/- 30 mumol/min at control arterial pH. Acetazolamide (150 mg/kg, i.v.) reduced HCO3 secretion to 84 +/- 12 mumol/min and rendered secretion independent of arterial pH between pH 7.6 and pH 7.0. It is concluded that acetazolamide imposes a pH-independent transport maximum on pancreatic HCO-3 secretion, but does not reduce the capacity of the secretory mechanism to sustain a proton potential gradient between cells and interstitial fluid.

Mechanism of hepatic bicarbonate secretion and bile acid independent bile secretion.

To examine hepatic bicarbonate transport and bile acid independent bile secretion, bile was sampled via a T-tube inserted into the common bile duct of anaesthetized pigs. Secretin was infused intravenously at a rate of 2.7 C.U./kg body weight h-1 (large dose) or 0.45 C.U./kg body weight h-1 (small dose). Hepatic water and electrolyte secretion were studied during systemic acid-base disturbances while secretin was continuously administered. Systemic acidosis reduced the rate of NaHCO3 secretion which fell in proportion to changes in plasma pH, by 9% and 2% per 0.1 pH unit for the large and small dose of secretin, respectively. Plasma pCO2 and bicarbonate concentration had little influence on NaHCO3 secretion. Consequently, plasma pH appeared to be the main determinant of hepatic NaHCO3 secretion during acid-base changes. Secretion of 1 mol NaHCO3 was accompanied by an isotonic solution containing water and 0.25 mol NaCl. After secretin infusion, 14C-erythritol clearance increased in proportion to bile flow. Bicarbonate secretion is determined by a gradient limited H+-pump at the contraluminal cell. During secretin stimulation bile acid independent bile secretion is osmotically driven by bile NaHCO3 flux.

Effect of intravenous infusion of hypertonic glucose solutions on pancreatic HCO3(-) secretion.

To examine why intravenous infusion of hypertonic non-electrolyte solutions inhibit pancreatic HCO3(-) secretion, the relationship between pancreatic HCO3(-) secretion and plasma pH was examined before and following intravenous infusion of hypertonic glucose to 5 anesthetized, secretin infused (2.7 C.U./kg b.wt.h-1) pigs. Hyperglycemia (plasma glucose 103 +/- 6 mmol/l) did not significantly change plasma pH, Na+, K+, Cl- and HCO3(-) concentrations. Hyperglycemia reduced pancreatic water flux by 48 +/- 5% and raised pancreatic juice HCO3(-) concentration by 43 +/- 4 mmol/l. Concurrently, HCO3(-) secretion fell by 34 +/- 5%. Acidosis, produced through intravenous HCl infusion and CO2 addition to inspired air, reduced HCO3(-) secretion by 40 +/- 6 mumol/min and 30 +/- 5 mumol/min per 0.1 pH unit reduction in plasma pH before and during hyperglycemia, respectively, and abolished HCO3(-) secretion at an estimated plasma pH of 6.51 +/- 0.06 before and a pH of 6.63 +/- 0.05 during hyperglycemia. We conclude that hypertonic glucose infusions inhibit pancreatic water flux and cause an increase in pancreatic juice HCO3(-) concentration which may inhibit HCO3(-) secretion through an effect on acid-base balance in secretory cells.

Plasma Na+-iron concentration or pH as regulator of pancreatic HCO-3 secretion.

Pancreatic HCO-3 secretion is caused by proton flux from pancreatic cells to interstitial fluid which, hypothetically, may be derived by a Na-pump or a proton pump. A Na-pump would reabsorb protons from pancreatic duct in proportion with plasma Na+ concentration (PNa+). A proton pump would cause passive HCO-3 flux into pancreatic ducts that would vary with pancreatic jice HCO-3 concentration (CHCO-3). Because pancreatic ducts are water-permeable, CHCO-3 varies with plasma osmolarity. This phenomenon allows testing of the two hypotheses. Intravenous infusion of hypotonic salt solution to 5 anesthetized, secretin infused pigs (2.7 C.U./kg b. wt. H-1) lowered PNa+ by 20 +/- 2 mmol/l and CHCO-3 by 20 +/- 4 mmol/l and increased pancreatic HCO-3 secretion by 71 +/- 5 mumol/min. Intravenous infusion of hypertonic salt solutions to 5 other pigs raised PNa+ by 52 +/- 3 mmol/l and CHCO-3 by 54 +/- 3 mmol/l and reduced HCO-3 secretion by 86 +/0 26 mumol/min. Isotonic glucose infusion lowered PNa+ by 27 +/- 2 mmol/l and did not change CHCO-3 nor HCO-3 secretion rate in 5 pigs. These findings comply with the proton pump hypothesis and are at variance with the Na-pump hypothesis of pancreatic HCO-3 secretion.

Dose-dependent effects of glycine, alanine, and glucose on hepatic bile secretion, oxygen consumption, and hemodynamics.

To assess the effects of a small (0.5%) and a large dose (5%) of glycine and alanine and of hypertonic glucose on hepatic bile secretion, oxygen consumption, and hemodynamics, experiments were performed on anesthetized pigs. Only the large dose of amino acids exerted significant changes. Glycine, alanine, and glucose reduced bile acid-dependent bile secretion gradually, which was nearly halved from a control value of 0.32 +/- 0.04 ml/min. Oxygen consumption was thereby continuously stimulated during amino acid and glucose infusion and increased from 448 +/- 132 mumol/min before to 995 +/- 226 mumol/min after the infusion of glycine, alanine, and glucose. Hepatic arterial blood flow increased from 214 +/- 14 ml/min to 238 +/- 14 ml/min after glycine infusion, whereas portal venous blood flow decreased from 542 +/- 50 ml/min to 481 +/- 47 ml/min. Total hepatic blood flow remained unchanged. Alanine and glucose provoked no further changes in hepatic blood flow. Bile secretion is a sensitive marker of hepatic metabolism, whereas hepatic blood flow is not a dominant regulator of bile secretion. Stimulation of hepatic metabolism is not followed by changes in total hepatic blood flow.

Amino acid- and glucose-induced cholestasis before and during secretin stimulation.

To identify the mechanisms of reduced bile flow after hypertonic amino acid and glucose infusion, acute experiments were performed on anesthetized pigs. When secretin was not administered, amino acids or glucose reduced bile acid-dependent bile secretion to 65 +/- 3% of control. During secretin stimulation amino acids or glucose diminished bile acid-independent bile secretion to 78 +/- 2% of control. No changes in serum bilirubin, alanine aminotransferase, and aspartate aminotransferase were observed. Amino acids and glucose attack different mechanisms responsible for bile formation, but the result is that when secretin is not administered, biliary secretion of bile acids is reduced, and, accordingly, bile acid-dependent bile flow diminished. During secretin stimulation biliary NaHCO3 secretion is depressed, accounting for a fall in bile acid-independent bile flow. Amino acids exert no effect on bile acid secretion or, as a result, on bile acid-dependent bile flow after secretin infusion.

Laparoscopic cholecystectomy: bile duct and vascular injuries: management and outcome.

BACKGROUND: This is a retrospective study of 32 consecutive patients referred in the period 1992-2000 for management of serious bile duct injuries caused by elective laparoscopic cholecystectomy. METHODS: The patients were referred on median 29 days (0 days to 34 months). Only 7 patients were referred immediately after discovery of the injury. At the local hospital, 25 patients underwent various procedures in attempts at repair. Ten of the patients were treated for bile duct strictures after previous repairs in other hospitals. RESULTS: At referral, 23 patients (72%) had complete transection of the bile duct, while 9 had bile leakage injuries. Additional complications were occlusion of the right hepatic artery in 8 patients (24%) and occlusion of the mesenteric superior artery in 1 patient. Infectious complications were prominent in 21 patients (70%), 6 of whom had septicaemia. Operative management with hepaticojejunostomy Roux-Y was employed in 22 patients. Various non-operative strategies were chosen, including endoscopically or transhepatic stenting of the bile duct and embolization of the right hepatic artery. There was no difference in hospital stay between operative and non-operative procedures which on median was 16 days ( range 7-69 days). Three patients died: one had thrombosis of the superior mesenteric artery, while the other two died of complications to bile peritonitis. Median observation period is 5 years (5 months to 8 years). Two patients have cholangitis; both had injury to the right hepatic artery. The other patients all had normal ultrasonograms of the liver and normal/almost normal liver function tests. CONCLUSIONS: Bile duct injuries continue to occur, are serious and may result in death. Injury to the right hepatic artery is present in many cases. Patients are referred late to a competent center, resulting in serious infection in 70%.

Perivascular low attenuation zone at CT of liver transplants. A follow-up study.

A follow-up study of 30 liver transplant patients confirmed that periportal and pericaval rims of low attenuation found at CT reflect hepatic lymph stasis. The perivascular rims are found coexistent with acute allograft rejection, but are not correlated with acute rejection, immunosuppressive agents (cyclosporin A), or intercurrent infection. Disappearance of the perivascular low attenuation zone in our material coincided with normalization of liver biochemistry probably associated with reestablishment of lymphatic drainage. Like previous reports on experimental hepatic lymph stasis, our study thus indicates that lymph stasis may have harmful effects on human liver allografts.

Effect of phlorizin on hepatic bile production before and during glucose infusion.

During intravenous infusion of glucose, bile secretion is reduced (cholestasis), indicating that hepatocellular metabolism of glucose could have harmful effects on the liver. Phlorizin has been identified as a compound capable of impeding glucose uptake of liver cells. To examine whether phlorizin had any effect on glucose-associated cholestasis, three groups of experiments were performed on anaesthetized pigs. In group I phlorizin (100 mg/kg body wt) during normoglycaemia stimulated bicarbonate-dependent bile secretion by 56 +/- 4%. After phlorizin, hyperglycaemia decreased both bile acid- and bicarbonate-dependent bile secretion by 37 +/- 4%. But after the glucose load normalization of plasma glucose concentration increased the bicarbonate-dependent fraction by 38 +/- 4%. In group II phlorizin (100 mg/kg body wt, infused intravenously) during hyperglycaemia stimulated bicarbonate-dependent bile secretion by 35 +/- 5%. In group III bile secretion was continuously stimulated by infusion of Na-taurocholate. Hyperglycaemia reduced bicarbonate-dependent bile secretion by 33 +/- 4%, but after phlorizin both bile acid- and bicarbonate-dependent bile secretion increased on average by 121 +/- 8%. The osmotic effect of hyperglycaemia cannot be blocked by phlorizin, but judged by the effect on bile secretion, phlorizin may decrease the cholestatic effect induced by metabolism of glucose. Phlorizin could be an interesting compound for use in solutions for organ preservation.

Risk factors associated with mortality and morbidity after elective splenectomy.

OBJECTIVE: To assess the effect of certain risk factors on the outcome of elective splenectomy. DESIGN: Retrospective cohort study. SETTING: University hospital, Norway. SUBJECTS: 135 patients who underwent elective splenectomy from 1978 to 1992 for autoimmune diseases (n = 58), myeloproliferative disorders (n = 58), and other diagnoses (n = 19). MAIN OUTCOME MEASURES: Morbidity and mortality, and the effect on outcome of age, malignancy, platelet count, treatment with steroids, and blood transfusion during the operative period. RESULTS: Three patients died, one from each diagnostic group; all had been treated with steroids. The overall complication rate was 31% (42/135). Most of the complications (31, 74%) were infective (pneumonia n = 23, abscess or sepsis n = 8). Nine patients required reoperation (haemorrhage, n = 6, and wound dehiscence, n = 2, and bowel obstruction, n = 1). Neither treatment with steroids nor severe thrombocytopaenia were risk factors for any complication, but age (p = 0.05) and a diagnosis of myeloproliferative disease (p = 0.08) had an important though not significant role. Blood transfusion during the operative period was the only independent risk factor (p = 0.001), and further analysis showed that transfusion of blood had a dose dependent effect on the incidence of infective complications (chi-square for linear trend 14.21 p < 0.001). CONCLUSION: To reduce the complication rate of elective splenectomy blood transfusion during the operative period should be avoided.