Oral Presentations

Introduction:A70yearold ladypresentedwitha3 dayhistoryofbilateral knee and ankle pain and a 1 day history of right knee swelling. Plain X-ray showed chondrocalcinosis and calcium pyrophosphate crystals were seen on polarised light microscopy of the joint aspirate. Despite intraarticular and oral steroids she went on to develop polyarticular joint inflammation involving the right shoulder, right sternoclavicular joint, knee, ankle and left first MTP. Active inflammation was confirmed on imaging despite treatment with steroids and she was commenced on methotrexate10mgonceweekly. Casedescription:A70yearoldladypresentedwitha3 dayhistoryofpolyarthralgia affecting the right shoulder, both knees and ankles, with a 1 dayhistoryofanacutelyswollenrightknee.Therehadbeenahistoryofa vomiting and diarrhoea illness 1week previously.Therewas apast medical history of hypothyroidism, hypertension, fibromyalgia and previous Helicobacter pylori infection. She was afebrile. Blood tests showed normal fullbloodcountwith raised inflammatorymarkers (C-reactiveprotein 192, erythrocyte sediment rate 63). Uric acid was 257. Blood cultures were negative.Plain X-rayof the rightknee showed chondrocalcinosis. A joint aspiration of the right knee was performed and calcium pyrophosphate crystals were seen under polarised light microscopy. She was managedwithintra-articularjoint injectionattherightkneeandareducing regime of oral prednisolone, starting at 15mg and reducing over 28 days. Inflammatory markers were improving and she was discharged with a diagnosis of acute calcium pyrophosphate crystal arthritis. On review in rheumatology several weeks later she had developed polyarticular joint inflammation at the right shoulder, right sternoclavicular joint, knee, left ankleandgreattoe,requiringhertorecommenceonlowdoseoralprednisolone 5mg daily. Blood tests showed normal full blood count, renal and liver function.CRP was 46.6 and ESR was 32. Uric acid and CK were normal. Calcium, magnesium, phosphate, alkaline phosphatase, ferritin, iron, transferrin, glucose and parathyroid hormone wereall within normal limits. Autoantibody screening showed positive rheumatoid factor (23), negative anti-cyclic-citrunillated peptide antibodies and negative antinuclear antibody. Plain x-ray of the right shoulder showed calcific tendonitisofthedistalsupraspinatustendoninkeepingwithcalciumpyrophosphate arthropathy. Magnetic resonance imaging of the right sternoclavicular jointshowedactive inflammation,despiteongoingtreatment with low dose oral steroids. She was commenced on oral methotrexate 10mg once weekly with folic acid, with a diagnosis of chronic calcium pyrophosphate crystal inflammatory arthritis. It is currently too early to assess her response; she experienced significant nausea from oral methotrexate and therefore the dose cannot be escalated until her switchtosubcutaneousmethotrexate. Discussion: There is a spectrum of disease in calcium pyrophosphate crystal deposition ranging from asymptomatic patients with incidental plain x-ray findings, up to joint destruction and even spinal involvement. Most common is an acute and often short-lived arthritis, commonly affecting the knee, wrist or shoulder. In only approximately 5% of cases does it result in chronic calcium pyrophosphate crystal inflammatory arthritis. In this case there were no associated metabolic or endocrine conditions. Incasesofchroniccalciumpyrophosphatecrystal inflammatory arthritis, knees and the jointsof the upper arm suchas wrists, elbows andshouldersarecommonly involved. Involvementof thesternoclavicular joint, however, is rare. In the absence of any aspirate from the sternoclavicular joint, or specific radiological findings, the inflammation within thisparticular jointcannotbeconfirmedasbeinga resultofcalciumpyrophosphate deposition, therefore experiences from other clinicians regarding further investigation and management would be useful. If symptoms are not controlled with non-steroidal anti-inflammatory drugs (NSAIDs) or colchicine then low dose oral steroids may be of benefit. In cases suchas this where these medications fail to control joint inflammation there may be a role for disease-modifying anti-rheumatic drugs (DMARDs). Current European League Against Rheumatism (EULAR) recommendations in the management of calcium pyrophosphate depositionsuggest theremaybeabenefit fromhxdroxychloroquineormethotrexate. This is based upon several smaller trials, and larger studies are ongoing. The patient is in the early stage of treatment with methotrexate, and this is the first case of a patient in this trust being treated with a DMARD for chronic calcium pyrophosphate crystal inflammatory arthritis. Understanding the experiences of others who have managed similar patientsonmethotrexatecouldhelp toguidehowfuturepatientswith this conditionaremanagedwithinthetrust. Key learning points: Chronic calcium pyrophosphate crystal inflammatory arthritis should be considered as a differential in any polyarticular inflammatory arthritis. Joint aspiration can provide confirmation if calcium pyrophosphate crystals are detected however an aspirate which is negativeforcrystalsdoesnotexcludethisasadiagnosis.Acarefulhistoryshould betakentoelicitanypotentialtriggers,inthiscasetheprecedingillnesswith diarrhoea is likely toberelevant.Associatedendocrineandmetabolicconditions including haemochromatosis, parathyroid disease and electrolyte abnormalitiesshouldbeexcluded.Otherfeaturestosuggestcalciumpyrophosphatecrystaldepositionasacause include radiological findingssuch as chondrocalcinosis or calcification of tendons or ligaments. Calcium pyrophosphate deposition more commonly affects older patients. Often there are contraindications to the use of NSAIDs in this patient group, for example in renal failureorcardiacdisease.Thesepatientsarealsoatriskof the consequences ofprolonged steroid use. Additionally somecasesmay be unresponsive to NSAIDs and steroids. There is a developing role of DMARDssuchashydroxychloroquineandmethotrexatetocontrol inflammationandreducejointdamageinsuchcases. i42 11–12 October 2017 ORAL PRESENTATIONS

Hospital, Villejuif, France In order to establish a treatment strategy based on the perspective of liver transplantation, we reviewed 1000 consecutive liver transplants over a 10 year period with special emphasis on prior history of portal hypertension with variceal hemorrhage. The impact of bleeding and subsequent therapy was analyzed to integrate the proper timing of liver transplant in the multimodality therapy of portal hypertension. Transplantation for fulminant hepatitis (n=148) and retransplantation (n=122) were excluded. Of 730 primary transplanted patients with chronic liver disease, 544 (74%) had no, prior history of variceal hemorrhage. There were 186 (26%) patients with variceal bleed prior to transplantation of which 130 (70%) required interventional therapy to palliate the bleeding. In the two groups of patients with and without bleed prior to transplantation there were no significant differences in regards to age, sex, etiology of the liver disease or donor liver morphology. Sclerotherapy was Ferformed in 93 (50%), surgical portal diversion in 27 (15%) and TIPS in 10 (5%) patients. Moderate to severe liver dysfunction (grade B and C of a modified Child classification) accounted for 91 percent of the patients with bleeding complications. The overall survival for all patients was 76 percent at five years. Previous history of variceal bleeding alone or treatment by initial sclerotherapy demonstrated no significant difference with either graft or patient survival. The patients who had TIPS to control hemorrhage had lower, but insignificant, graft and patient survival. The group of patients with variceal hemorrhage who had prior surgical shunt did, however, demonstrate a significant increased survival at five years when compared to the non-shunted group (96% versus 73%, p<0.007). In conclusion, the impact of variceal bleeding does not seem to be critical to subsequent liver transplantation. In contrast to sclerotherapy and TIPS, portocaval shunt demonstrated an improved outcome following liver transplantation. The perspective of liver transplantation should not be a contraindication to perform portocaval shunting in properly selected patients. The effect ofnitric oxid in acute liver injury at different time intervals was evaluated in an acute liver injury model induced by D-galactosamine (1.1 gm/kg body wt.) intraperitoneally. Rats were divided into 4 groups: normal control, acute liver injury, acute liver injury + N-nitroL-arginine methyl ester (L-NAME) and acute liver injury + L-NAME + L-arginine. After 6 hours ofthe liver injury, the Alkaline Phosphatase (ALP), Bilirubin (B/L), Aspartate Aminotransferase (ASAT) and Alanine Aminotransferase (ALAT) increase in the acute liver injury + L-NAME group compared to the acute liver injury control group with significant difference in ALP (P-<0.01), BIL (P<0.05) and ASAT (P<0.05). The acute liver injury + L-NAME + L-arginine group show reduced levels of ALP, B1L, ASAT and ALAT compared to acute liver injury + L-NAME group, with a significant difference in ALP (P<0.05). After 12 hours the inhibition ofnitric oxid increase the level ofliver enzymes and translocated bacteria but without significant difference. After 24 hours in the acute liver injury + L-NAME group there is a significant increase in B/L (P<0.05) compared to acute liver injury group. The acute liver injury + L-NAME + L-arginine show significant reduction in the level of ALP (P<0.05) compared to acute liver injury + L-NAME group. The number ofthe transloeated bacteria to arterial blood, portal blood, liver and mesenteric lymph nodes at all time intervals increased in acute liver injury + L-NAME groups compared to acute liver injury group with a significant difference in the arterial blood after 24 hours (P<0.05) and a decreased number in acute liver injury + L-NAME+ L-arginine groups compared to acute liver injury+ L-NAME groups with a significant difference in the arterial blood after 24 hours (P<0.05). These show that inhibition ofnitric oxid increase the number of the translocated bacteria and potentiate the liver injury.

lstanbul, TURKEY
The treatment of hydatid cysts of the liver is still primarily surgical. The surgica treatment of this disease, however, is far from ideal. Intraoperative spillage and postoperative recurrence around 10 % are yet unresolved problems. Although in experimental models the efficacy of albendazole has been demonstrated, clinical data are still lacking. In addition to the gross appearance of the cyst, the intracystic pressure (ICP) has also been found to be a reliable guide for the assessment of viability. High intracystic pressures are found in viable cysts.
In this study a three week course of preoperative albendazole (10 mg/kg) was given to patients with liver hydatid cysts and the intraoperative viability ofthe cyst assessed.
The study consisted of two groups and the first group had 15 patients (5male, 10 female)with a median age of 31 (21-75). All cysts were located in the liver. In two patients the cysts were grossly degenerated and the ICP was 0. In two others the cysts were partly degenerated and the ICPs were 4 and 5.5 era. In the remaining patients the mean ICP was 27 em H20 (range  cm H20). Direct microscopy with eosin exclusion test revealed viable scolices in four patients in whom the test was performed.
The second group consisted of 40 patients with liver hydatid cysts without any preoperative treatment. In this control group, there were 9 non-viable (mean ICP 0era H20) and 31 viable cysts (mean ICP 35cm H20). The differences among the groups were not significant (p< 0.05).
lhysioloical ar%erial lqase of live vasculaisation was X+/-SD=8+/-lsec, vahile in CT (n=5) and it was prolored (p O. Oi) and lasted 12+/sec and O+/-2sec rspectively, witl)ut tl3e dif-es between the last two groups (p > O. 05). Portal. lase in C was 12+/-sec, in Cr (n:5) wasn't registered (p <0. 01), vahile in CPV it was prolonged (21+/-II sec) in ccoa'ison to C (p O. 0) and CT (p O. 01). Relative liver portal per/kion (HPI, (mean, 42 years). Twentyseven cases had only one cyst and 3 cases had more than one. Half volume of the cysts we're aspirated under the guidance of US or CT and 0.5 % silver nitrate solution was injected into the cyst as well as. aspirated fluid. Five minutes later, cyst cavity was aspirated totally and the catheter was withdrawn. Three patients had allergic reactions (mild cyanosis in two and bronchospasm in one patient) due to silver nitrate injection, not necessarily to quit the procedure. The patients were administered Albendazole (10 mg/kg/day) orally, for 13 days, 3 days before the application and 10 days following the aspiration. Totally 5 secondary aspiration was necessary in three patients in the early post drainage period because of insufficient evacuation of the cysts. The mean follow-up period was 27 months. Recurrence was seen only in one patient with abdominal cyst which was also treated percutaneously. We concluded that, percutaneous aspiration of the hepatic hydatid cysts was an effective and an alternative method to the surgery in convenierit cases. The good results of liver transplantation (LT) have allowed its larger use. However this increased use of LT has uncovered a relative organ shortage. The split liver is one of the means to increase the number of available grafts. We report here our recent experience with this technique. From January to December 1995, a systematic proposal to perform a split was made to the French organ sharing network as often as possible. Ninety LT were performed using 61 whole grafts, 27 split liver grafts, 2 reduced size grafts. Twenty livers were splitted at our center generating 40 split liver grafts: 23 transplanted at our center in 23 patients and 17 shipped to other centers. We received 4 split liver grafts from 4 livers splitted in other centers. Our 27 patients were transplanted for cirrhosis in 19 cases, amyloid polyneuropathy in 6 cases and fulminant hepatitis in 2 cases. Operative mortality (< day 60 post-LT) occured in one case of fulminant hepatitis and long term mortality (> day 60 post LT) occured in case of cirrhosis. One patient was retransplanted at day 6 for primary non function due to a too small graft (the lowest liver to recipient weight ratio 0.87).Patient and graft actuarial survival are respectively91.4 + 5.8% and 87.5 + 6.8%. Twelve technical complications occured in 10 patients: 3 arterial complications (2 thrombosis and dissection)of which 2 were successfully treated by urgent desobstruction; biliary fistula" 4 cases, biliary stenosis: 2 cases; hemopedtoneum: 2 cases, segment 4 necrosis: case. Eight of these complications needed surgery to be controlled. Conclusion 111 LT were performed with 87 livers realising an economy of 24/111 grafts (22%). During the same period, 16 proposals of split were refused in France. The (LT). Patients were classified Child class A in 5 cases, B in 11 cases, C in 18 cases. Indication for TIPS was sclerotherapy failure in 23 cases and intractable ascites in 11 cases.Two patients were excluded because of technical failures which were treated by OLT in one case and open calibrated porta-caval shunt in one case. The follow-up with LT as end point was to 34 months (7.6+1.6 M). Results: Early thrombosis (< 3 months) occurred in 8 cases: 6 were desobstructed via the internal jugular vein and 2 were desobstructed surgically together with calibrated porta-caval shunt. Late thrombosis occurred in case with portal vein thrombosis and was treated by mesenterico-caval shunt followed by LT 6 months later. Recurrence of hemorrhage occurred in 2/22 patients who underwent TIPS for sclerotherapy failure (one rupture of varices, one duodenal ulcus).Ascites disappeared in 7/10 patients who underwent TIPS for intractable ascites and was controlled together with diuretics in 2 patients. Ascites remained unchanged in patient. 21 patients were transplanted following TIPS with a mean delay 6.4 + 1.6 (range:   Portal vein embolization (PVE) is useful with liver tumors when extended hepatic resection is technically feasible, but, the potential for postoperative liver failure prohibits primary surgical treatment. Achieving a "functional hepatectomy" percutaneously, while inducing controlateral hypertrophy, in anticipation of major hepatic resection is the goal of this modality. This technique was applied in 20 patients who were already enrolled in different protocols with neoadjuvant chemotherapy including arterial chemoembolization for primary tumor, and chronomodulated and/or hepatic artery infusion chemotherapy for hepatic metastatic cancer. Although most patients had either primary or metastatic hepatic tumor, one patient had cholangiocarcinoma and another had a neuroendocrine tumor. Liver cirrhosis was associated with tumor in 5 (23%) patients. Final decision to perform hepatic resection was based upon the degree of liver hypertrophy of the future remaining liver by clinical, biologic, volumetric computed tomographic (CT) scan and tumor response to chemotherapy. There were no deaths and one complication. Exploration was performed after PVE in 18 (90%) patients, while 2 (10%) are still awaiting decision to operate. At exploration 5/18 (28%) had disseminated disease and were considered incurable, while the remaining 13 (72%) had hepatic resection with curative intent. Specimens at the time of resection revealed tumor necrosis in 8 (62%) patients (3 with 100% necrosis) and in 5 (38%) patients the resection was less extensive than anticipated. Tumor margins were negative in all patients resected (68%), while the remainder were either inoperable (23%) or are still awaiting surgery (9%). There was a significant increase in serum bilirubin after PVE (p<0.02) and a decrease in lactate dehydrogenase (p=0.05) prior to surgery. There was no significant change in other biochemical liver function tests or with indocyanine gree (ICGR15). When primary resection of liver malignancy is not feasible for various reasons, in both cirrhotic and non cirrhotic livers, neoadjuvant PVE which was performed in conjunction with chemotherapy, permitted resection with intent to cure, when otherwise was prohibitive.
The main causes of death following liver resection (LR} for hepatocellular carcinoma (HCC) in cirrhotic patients are tumour recurrence and liver failure. The aim of this study was to analyze patientc who underwent liver transplantation (LTX) after LR of their HCC between Jul-87 and Feb-95.
LR was performed in 57 cirrhotic patients with HCC.
Recurrence of HCC was detected in 28 of 50 patients who survived surgery (56%). Seven patients underwent LTX, the indications being tumour recurrence in 4 and liver failure in 3. Those patients transplanted for tumour recurrence, only one patient is alive 9 months later with no evidence of tumour recurrence; one died from carcinomatosis 5 months post-LTX and the other two died from cryptococcal meningitis and from upper digestive bleeding with dyseminated aspergillosis respectively and with severe recurrence of viral hepatitis C in both. Liver failure was the indication for LTX in one patient during the early postoperative LR period and is still alive 38 months post-LTX. Other two patients underwent LTX 47 and 5 months after LR, but both died from septicaemia in early and late postoperative periods, respectively. The actuarial survival rates at 1 and 4 years for patients who underwent LR and afterwards a LTX were 86% and 43% respectively. (Actuarial survival of patients with LR alone were 60% and 32%, respectively). patients were operated in emergency and in this situation we used 10 mm. prosthesis in order to obtain better decompression of the portal system. Postoperative complications were: neuropathy prob]ably related to criog]obulinemia, bleeding fron erosive gastritis, bleeding duodenal u|cers and duodenal perforation. In pt. (5.3%) an HCC was showed year after shunt procedure. pts. died after surgery so the overall operative mortality has been 21% of these pts. underwent shunt procedure in emergency setting and they were Child C (50%) and Child (20%). Cumulative shunt patency in survived pts. was 100% and have not observed neither episodes of reb]eeding from oesophageal varices nor graft thrombosis during follow up. out 15 (26.6%) survived pts. had an episode of acute encepha]opathy while only patient (6.6%) developped chronic encephalopathy. Our results show that partial portacaval shunt with small diameter P.T.F.E. prosthesis (8 or 10 mm.) is an effective procedufor the treatment of varicea] rebleeding with low rate of chronic encephalopathy. We conclude that this technique, which does not compromise liver transplantation, should be used in Child Apts. when sc]erotherapy has failed, in presence of P.H.G., when the pt. does not comply with sclerotherapy or when he or she lives in non-urban area. Pts. with poor liver function showed high morbidity and mortality so they are probably better served with T.I.P.S. procedure. A clinical trial comparing costs, complications and results of variceal injection-]igation and partial portacaval shunt in Child Apts. might be indicated. In order to identify risk factors for mortality at'mr elective hepatic resection a retrospective tmivariate analysis of risk factors was performed. Between 1988 and 1995 315 hepatic resections were performed in 161 males and 154 females (median age 58 yrs., range 14-90 yrs.). 187 Resections were performed for metastases, 64 for primary liver tumors, and 64 for benign lesions such as hemangioma or FNH. Resection methods included 121 anatomical lobeetomies, 56 extended lobectomies and 138 so-called minor resections with removal of solitary ofmultiple segments or atypical subsegmenteetomies.Tbe 30-day mortality in the whole series was 3.8% (12/315). The following risk factors were significantly different between patients whodied within 30 d and those that survived >30 days: bilirubin (Bilir),AST, alkaline phosphatase (AP), albumin (Alb), pseudo-eholinesterase (pCHE), prothrombin ace. to Quick, and duration of surgery (OP-time (25 vs.26 rain), and median number ofunits red packed cells (1 vs. 0) did not differ significantly between both groups. There were significantly more early deaths in patients with cirrhosis (5/42) compared to those without cirrhosis (7/273) and in those with primary liver tumors (6/64) compared to those with metastases (6/187) or those with benign lesions (0/64) ( Beaujon, University Paris VII, Clichy, France. Auxiliary liver transplantation (ALT) theoretically, bridges the period of acute liver failure until the native liver (NL) rccovcrs and immunosuppression can be discontinued. However, this attractive concept is burdened by technical problems and by the selection of candidates. We report our experience of ALT wilh special references to early and long term graft function in prospective study including all patients who undcrwent emergency liver transplantation for acute liver failure from April 1.993 to October 1995. Patients: Thirty patients aged from 16 to 62 years with acute liver failure wcrc candidates for emergency liver transplantation according to Clichy criteria. Causal disease was drug toxicity (n=10) including paracetamol in 3; hepatitis B (n=6); hepatitis A (n=2) and other (n=12). We decided to perform OLT in 18 because of age>60 years (n=3), pre-existing chronic liver discasc (n=4), hacmodynamic instability (n=4), poor liver graft (n=2) and poor neurological status with immediate risk .of cerebral herniation (n=5). Seven patients died postoperatively including 5 after ALT; in the latter group mortality was due to vascular thrombosis (n=3) graft compression (n=l) and sepsis (n=l). With a follow up ranging from 3 to 31 months among the 7 surviving patients, graft was removed in 2 respectively after and 7 months, immunosuppression was stoped in 2 respectively after 9 and 27 months. Liver biopsy demonstrated the presence of mild fibrosis in 3 respectively after 6 and 9 months. Conclusion: After auxiliary liver transplantation for fulminant hepatitis, there is no predictive value of the extent nor the delay of sufficient regeneration of the native liver. The higher operative risk associated with ALT suggests that this procedure should: (a) not be indicated earlier than standard OLT; (b) be restricted to patients < 50 years without haemodynamic instability and (c) be performed using good quality ABO compatible graft. MATERIAL Between 0ct-88 and Dec-94, 172 liver transplants were performed in 158 patients with end-stage liver disease. The median age was 51 years (r:16-66). Child's-classification prior to LT was: A(IO%); B(37%); and C(53%). Ninety percent were UNOS and II status. Renal insufficiency prior to LT attained to 18% of patients. Donor and recipient factors studied in relationship to ARF were: preoperative parameters (age, sex, indication to LT, Child-Pugh score, UNOS status, laboratory data, renal function prior to LT), intraoperative parameters (le6gth of surgery, type and length of anhepatic phase, preservation time, periods of hypotension, blood product requirements, diuretic doses, balance, etc), and postoperative parameters (prymary liver disfunction, blood products transfused.ere).
Univariate analysis of donor and recipient factors showed that: urgent retransplantation, advanced Child-Pugh score, renal function prior to LT, preservation time, blood product requirements, and primary liver disfunction, were the variables observed statistically significant in relationship to ARF. However, multivariate analysis revealed that among 38 variables investigated in our serie, only two had independent prognostic value preoperative SCr 1.5 mg/dl and graft dysfunction grades III and IV. CONCI,USIOIi Early ARF is common and severe complication in LT, with high morbi-mortality, that can be predicted particularly in relation to some wellknown factors, mentioned aboved. RISK Barcelona, Spain. To meet the demand for grafts to transplant patients with end-stage liver diseases acute liver failure, it is necessary to accept the called non-optimal donors, which in adittion to other negative factors lead to important graft dysfunction primary nonfunction. The aim of this study is to analyse the risk factors that could influence in the appearence of IL, originated either in the donor (prepreservation injury), during hypothermic storage (preservation injury), during reperfusion/reimplantation period. MATERIAL AND METHODS From October 1988to December 1994 performed in 158 patients. The mean age was 51 years. Sex distribution 65% males and 35% females.
Ischemic hepatic lesion classified regarding liver function test during the first days Mild (GPT < I000, and prothrombin time (PT) 60%); Moderate (GPT 1000-5000 and PT 30%-60%; and Severe (GPT > 5000 and PT < 30%). All the events ocurring during the immediate postoperative period during their stay in ICU respectively, with differences statistically significant compared to the rest of transplants. Patients with ischemic lesion showed higher morbidity in terms of greater need of dialysis, rate of severe infections, need of respirator, and ICU stay. CONCLUSION Incidence of severe ischemic injury was 10.4% with 2.3% of PNF. Risk factors for severe ischemic lesion recipient age, Child-C, and donor ICU stay longer than 5 days. As  GV bleeding presents high risck of mortality. Emergency surgery has high mortality. ES og GV using Bucrylate (B) material, may be a valid alternative particularly in emergency. The aim of this paper is to presents our experience in Polidocanol Vs Bucrylate ES of GV bleeding. MATERIALS: 52 Pts mean age 48+/-16 range 43-68 were trataed in our Istitution. 50% had alcoholic cirrhosis, Ace.to Child pug risk 6% were A,61% B, 33% C. Ace.to NIEC Criteria 61% were of type I, 35% of Type II,4% of type III (type I+II 17% ). 88% and 12% were respectively emergency and electively trated. 16pts were treated with Polydocanol + Saline ES respectively 30+/-6 range 38-465 and 40+/-7 ml range 36-48 36 Pts were treated with B -ES (with or without LIPIODOL 1"1 with a mean of 2.9+/-0.9 ml. RESULTS" Pol Hepatocellular carcinoma (HCC) is an established but still debated indication of liver transplantation (LT). The high risk of recurrence and five yearsuival.rates significantly lower (0-5.0% in different series) than those of benign diseases have questionned the place of LT for HCC in the current period of organ shortage. We report in this study the consequences of a new selection of patients adapted from prognostic indicators established in the first phase of a same series. From November 1985 to March 1994, 109 patients with cirrhosis were transplanted for HCC. Of these 109 patients, only 95 patients with HCC diagnosed betbre LT were included in the study. The presence of extrahepatic deposits on pretransplant staging or peroperative exploration was considered as a contraindication to LT. In the first period of our experience (November 85-December 91), the selection criteria only included the absence of any extrahepatic tumour (60 patients). After assessment of prognostic factors in this first period (mainly tumor size > 30 mm, number of nodules > 3 and presence of a portal thrombosis), we proceeded to a more restrictive selection of those patients at very high risk of recurrence. Results in terms of patient selection and 3 year-survival were as. follows 1st period (85-91) n=60 2sd period ( APOLT has been proposed in the treatment of FH to provide temporary hepatic support until the native liver regeneration. We report herein our experience with APOLT for FH in 5 patients during the past 12 months. During the same period X patients were treated at our center for FH needing liver transplantation. Conventional orthotopic liver transplantation (OLT) was chosen in X most severe cases to shorten as mudh as possible the waiting time for a liver graft. APOLT was chosen as a potential bridge in 2 cases because the graft was ABe incompatible, in case because of a too small graft from a living related donor, and in 3 cases because the potential for native liver regeneration was judged reasonnable in stable patients. The main caracteristics of the 5 patients are sumarized in table 1. In conclusion, Our experience with APOLT confirms the technical feasibility of this approach in FH. ABe incompatible graft may be a good indication of APOLT in FH. If, due to rejection or biliary complications, the graft should be removed, there is a chance of regeneration of the native liver allowing to avoid retransplantation. The objective of this work was to verify and analyse, in Wistar rats, the alterations caused by 90-minutes partial hepatic ischemia, followed by immediate, 15 and 60 minutes reperfusion, using a PAF-antagonist (WEB2086). A total of 48 male Wistar rats were used, with mean weight of 310g, previously submitted to a 6-hour fast with water ad libitum. Twelve rats were separated for hemodynamic and electrolytes analysis of: MAP, CVP, pH, pCO2, PO2, BE, HCO3, Na, K, Ca++, GLY, Hb, htc. Thirty-six rats were separated for analysis of mitochondrial function, RCR, liver function tests (AST, ALT, LDH, LDHs) and GLY. In the first part of the study the twelve rats were divided in two sub-groups of 6 rats each. The first sub-group was submitted to partial hepatic ischemia of the median and left lobes; the second sub-group received a specific PAF-antagonist (WEB2086) minutes before the ischemia. In the second part of the study the rats were also divided in two sub-groups, one sub-group submitted to partial hepatic ischemia, the other receiving WEB 2086 5 minutes before (20mg kg). The rats were submitted to partial hepatic ischemia for 90 minutes and studied in reperfusion times of zero, 15 and 60 minutes (R0, R15, R60).The results of the several groups were analysed with ANOVA and Friedman test. We observed statistically significant difference in the time R0 of the following MAP, CVP, Na, Hb, htc, ADPO and OPR. In the time R15: E III, MAP, HCO3, BE. In the time R60: CVP, BE, HCO3,GLY, E IV and LDH. Although, the use of WEB 2086 (PAF-antagonist) had inhibited the hypotension proprierties of PAF that occurred during the reperfusion, we did not observed improvement in that liver function and the mitochondrial function was maintained. Non controlled studies have reported better results for the association of TACE with PEI than for PEI alone in the treatment unifocal HCC. Patients and methods. We report here the preliminary results of a prospective multicentric controlled study comparing these two treatment strategies in 150 cases of small (<5 cm) unifocal HCC in liver cirrhosis in Child Pugh class A (n=89) and B (n=61) in whom liver resection was not feasible (age >65 years; Child B class; esophageal varices at risk; location of HCC in central segments). Patients were enrolled in 12 different italian centers since February 1992 to May 1994 and 84 of them were randomly assigned to PEI and 66 to TACE + PEI. The two groups were matched for sex, age, Child-Pugh class, size of HCC and AFP level at entry. Diagnosis of HCC was established by echo-guided liver biopsy and unifocality at US examination was confirmed in all cases by contrast-enhanced CT. In the group TACE + PEI, TACE was performed prior to PEI. All patients were followed at four months intervals by US and AFP. PEI was repeated in case of recurrence. The mean follow up was 18 months in the group TACE + PEI and 19 months in the group PEI. The treatment outcome was evaluated as follows: 1) successful when the lesion remained stable or decreased over time; 2) local progression; 3) diffuse progression. Survival curves in the two groups were then assessed by the Kaplan-Meier method and compared by the Mantel-Cox method.
Results. A decrease of the AFP levels was observed in both groups of treatment at 12 months interval, but any significant difference was not found between the two groups. A non significant (p<0.1>0.05) trend towards a better outcome (80% stable disease vs 59%) at 16 months was found in the group TACE + PEI. The analysis of survival curves showed a non significant trend for an improved survival in the group TACE + PEI in comparison with PEI alone (estimated survival at 24 months 65% vs 45%). The severity of liver cirrhosis was not worsened over time in the group TACE + PEI in comparison with the group PEI. Conclusions (p=NS). Long-term HE was significantly lower (p < .05) in group (13%) vs group 2 (39%). No patient of group and 2 patients of group 2 had severe chronic HE. Long-term evaluation of hepatic score was 0.59'-0.11 in group and 0.67:L-0.11 in group 2 (p < .02). Actuarial survival rate at 1,3 and 5 years was 100%0, 77% and 67% in group and 86%, 63% and 52% in group 2 respectively (p .04 Gehan's Wilcoxon Test). In conclusion: partial shunt, mantaining a good hepatic flow, significantly improves long-term survival and reduces postoperative encephalopathy and postoperative hepatic deterioration vs. total shunt.

F026
Advantages and disadvantages ofinferior vena cava preservation in ortothopie liver transplantation D. Casanova, M G Fleitas, E Martino, L Herrera, F Hemanz, JM Rabanal, G Solares Department of Surgery, University HospitalValdecilla, University of Cantabria, Santander, Spain Inferior vana cava preservation (IVCP) has been proposed an alternative procedure in ortothopic liver transplantation (OLT). This techniques provides continuous caval flow to the heart, minimizing the hemodynamic alterations and increasing the renal perfusion. Aim: The objective is prospective study of ortothopic liver transplantation using IVCP. Patients andmethods: Between November 1990 andNovember 1995, 130 OLT were performed in 118 adult patients (12 were retransplants). Donors and recipients were matched for size (weight and height) and ABO blood groups. During the surgery, in all cases, the liver was removed preserving the inferior cava vena. Cross clamping the IVC, was placed laterally preserving blood flow throughout the anhepatic phase. Results Prototype MRS machine and phosphorus spectra collected at 26 MHz. The liver was reperfused with the respective oxygenated buffer solution and the regeneration of ATP monitored following the acquisition of 2 min time resolved spectra. Concurrently, changes in inorganic phosphate, phosphomonoesters and phosphodiesters could also be resolved in real time. Spectra analysed using dedicated program designed specifically from the biochemical composition profiles of resonant peaks. Initial rates of ATP regeneration in the UW group was 9.7 x 103s whereas groups with added adenosine (Ad) or the prostacylin derivative (PD) had rates of 9.5 x 103s and 14.2 x 10"3s , respectively. The maximal attained amounts of ATP with respect to total phosphate in these same groups were 4.49 + 0.52% (UW), 5.82 :k 0.27% (UW + Ad) and 6.79 + 0.40 % (UW + PD). These changes represent an increase from the UW solution of 30% with added adenosine (p < 0.05) and 51% with added prostacyclin (p<0.02). This study suggests that improvements can be made in ATP regeneration using prostacyclin derivatives and adenosine to buffer the adenylate pool. The two deaths were due sepsis and myocardal infarct and occurred in one patient with the highest serum bilirubin concentration and hypoalbuminemia and in one patient with renal failure. These results suggest that most patients with obstructive jaundice can undergo major liver resection without preoperative biliary drainage. A long and complex procedure is necessary and the incidence biliary fistula is high. Increased operative bleeding was observed but tolerance to ischemia and regeneration were not affected by jaundice in this series. Preoperative drainage may be indicated in selected cases (bilirubin > 300, prolonged jaundice, hypoalbuminemia, renal failure). Korea, and the mortality rate from HCC is highest in this country all over the world, probably due to high hepatitis B virus infection rate and extremely poor prognosis. Established prognostic factors in HCC are presence of venous invasion, multiplicity, curative resection, but we do nok know all about prognosis in curatively resected cases. So, we studied i53 protein expression in Korea HCC patients, which is independent prognostic factor in breast cancer and lung cancer. We performed immunohistochemistry using Pab 1801, which is monoclonal antibody for protein in 39 HCC patients, who underwent curative liver resection in Korea Cancer Center Hospital. Positive expression rate of ix53 protein was 2.6%. p53 protein expression seemed to be higher in HCC patients with poor prognostic factors such as venous invasion, encapsulation, and multiplicity, but they did not reach the stastistical significance.

EFFECT OF INOTROPES AND STEATOSIS
And slightly better survival rates were shown in p53 negative protein group, but no significant difference was detected(p=0.593).
In the other hand, presence of veflous invasion, multiplicity and no capsule formation were related with bad prognosis statistically(p<0.05) In conclusion, we could not detect the prognostic significance of p53 protein expression in curatively resected HCC patients, but we think that more extensive study will be needed in more patierrts using diverse antibodies to mutant p53 protein. Microchimerism after organ transplantation has been demonstrated repeatedly in clinical and experimental transplantation. Little is known about the functional relevance of the transferred immune cells. This study analyses the immunological activity of the chimeric donor cells in the recipient. Male Lewis rats were either transplanted with a liver from a previously ACI-skin-sensitized BN-donor or a nontreated BN-rat and tested 2 weeks after liver transplantation with an ACI-test-skin graft for the transfer ofthis. sensitization. Recipient sensitization before liver transplantation was included as a control. Transplantation of a liver graft from previously skin sensitized BNliver donors to Lew-recipients led to accelerated rejection of ACI-test skin grafts (median of 10 days (n=11) compared to median of 13 days in the control group (n=6); p-value (Mann-Whitney U test) 0.004729); indicating a transfer of donor sensitization to the recipient. Similar results were achieved after recipient sensitization (median rejection time of test skin graft 9.5 days, (n=2)). BN-skin grafts used as marker for the tolerogenicity of the BN-liver graft were prolonged to the same extent, including permanent acceptance, in all three groups. Preliminary experiments using ACI-hearts as test grafts are pointing in the same direction. This is the first systematic approach to demonstrate donor specific immune functions in a liver graft recipient, most likely explained by transfer of donor derived lymphocytes. In addition the recipient specific development of liver graft induced tolerance was not impaired; leading to observation of both, donor and recipient specific immune properties in the liver graft recipient, thus pointing towards merging of donor and recipients immune system.  Oct. 1995. According to this rationale we began a phase II thai of schedule-oriented biochemical modulation of FUra bolus by MTX and I Interferon, and FUra continous infusion by Leucovorin. In particular, the treatment schedule provided a hybrid regimen of two biweekly administrations of 600 mg/sqm of FUra bolus, which had to be preceded the day before by 200 mg/sqm of methotrexate, and had to be followed, the same day and the next day, by two administrations of 3 x 106 l-Interferon i.m.; after an interval of two weeks, the cycle arried on with three weeks of continuous infusion of 200 mg/sqm per day of FUra, which was preceded every first day of the week by an administration of 20mg/sqm of Leucovorin bolus. The entire cycle was repeated after a week of rest, having first evaluated the lesions through US/CT scans and plotted the percent change of total measured tumour mass and dosed tumor markers. All the 14 patients, with no prechemotherapy, had the primary colorectal tumor mass resected for necessity and their livers were considered unresectable for the characteristics of the hepatic metastases: their number, dimensions, contiguity/continuity with important vascular structures didn't allow a radical operation. 13patients (average age 63) have completed at least one cycle of the treatment and have been reevaluated, while one patient has just been included in the study; we have obtained two complete responses, after six months of chemotherapy, and, at the moment, also nine partial responses (75% of all); two patients died after 8 months because of the advancement of their illness. We have had one death due to toxicity after the first administration of FUra, MTX and Interferon in the first cycle, and two cases of III grade toxicity (diarrhoea and mucositis). The two patients that had a complete response to the chemotherapy were operated, two hepatic metastasis, in each one, that had reached dimensions of 2 cm in diameter, were resected, after a previous US Antibodies directed against hepatocellular carcinoma (HCC) antigens have been used in animal models as well as in humans. However lack of specific antibodies and the high background level when radiolabelled antibodies are administered, have hampered a wide use of immunoscintigraphy. Recently a new method, employing biotinilated monoclonal antibodie (MoAb), not directly radiolabelled, and avidin has been used, for diagnosis of CEA producing tumours, with good results. Aim of this work is to evaluate the sensitivity of anti-Ferritin MoAb immunoscintigraphy with the avidin biotin system, in identifying HCC in cirrhotic patients. Between Jenuary and December 1995, 9 patients with proved HCC were studied. After performing abdominal ultrasound, NMR(6 pts), Lipiodol Tc (6 pts) and biopsy, the patients were injected with mg anti-Ferritin MoAb; 24-36 hours later 10 mg of avidin were administered and after further 24 hours In-111 labelled biotin was injected. Two hours after the radioactivity administration planar and tomographic imaging were obtained. Eight patients subsequently underwent surgery during which 10 lesions were diagnosed. The lesions ranged in size between 1.5 and 13 cm. Seven patients underwent surgical resection. Ultrasound, NMR and TC correctly diagnosed HCC in 8/9,6/8 and 6/6 cases respectively. Immunoscintigraphy identified 6/9 HCC; no false positive radioactivity localization was observed. The lower resolutive power of immunoscintigraphy was 1.5 cm. Our results suggest that anti-Ferritin immunoscintigraphy is effective in localizing HCC nodes in cirrhotic patients. INTRODUCTION. The hepatitis C virus (HCV) is one of the major etiological agents of end-stage liver diseases requiring liver transplantation (OLT). However, recent studies demonstrate the high rate of reinfection after transplantation in HCV carriers. The purpose of the study was to evaluate: 1) the evolution of post-transplant viraemia in patients anti-HCV positive pre-OLT; 2) the clinical behavior of transplanted patients reinfected with the HCV; 3) the relation between HCV viraemia and ALT levels. PATIENTS AND METHODS. Sixty-one patients transplanted between 1987 and 1994, who were anti-HCV positive after OLT, with a post-OLT follow-up >3 mos. entered the study. Twelve (19,7%) became anti-HCV positive after OLT and 49 (80,3%) were anti-HCV positive pre-OLT. Their follow-up is 50,4+30,7 mos. Fifteen (24,6%) were also HBsAg positive pre-OLT. Antibodies against HCV were assayed using an ELISA assay and a RIBA assay of 2nd and/or 3rd generation. For HCV-RNA detection, a PCR was employed. RESULTS. Ten patients were HBsAg positive after OLT. Among 51 HBsAg negative patients, persistent or intermittent elevations of ALT levels were >2xN in 37 and <2xN in 7; five patients always showed normal ALT levels (two other patients were excluded). Seriated post-OLT determiantions of the HCV-RNA were available in 51 patients out of 61 (6+4 determinations per patient): 5 (9,8%) were HCV-RNA negative (3 were HBsAg positive after OLT and 2 had a follow-up <5 mos.), while 46 (90,2%) resulted HCV-RNA positive at least once. In this group of patients, a total of 307 post-OLT determinations were avail.dole: 103 (33,7%) resulted HCV-RNA negative. In 16 patients, one or more determinations of the HCV-RNA resulted negative after demonstration of viremia. No correlation was evident between viraemia and ALT levels: the HCV genome was usually found also in patients with persistingly normal ALT values, while HCV-RNA fluctuations were observed in patients with persistent ALT alterations. CONCLUSIONS. In our experience, patients carrying the hepatitis C virus were at high risk of reinfection after OLT. The HCV-RNA reappearead eady in the serum in most cases, although without any apparent correlation with ALT levels. Signs of hepatitis of the transplanted liver, although usually mild and subjectively asymptomatic, were frequent. In these patients, repeated PCR assays may be useful for confirmation of HCV active replication, especially in case of abnormal liver function tests and histology, when a differential diagnosis with rejection is required. Introduction. Hepatic suicide gene therapy may provide approach to treat irreseetable hepatic malignancies, either primary metastatic. This strategy involves the infection of tumor and liver cells with replication-defective adenoviral vectors carrying suicide genes. Subsequently, systemically administered cytotoxie substrates selectively destruct replicating tumorcells, infected with suicide genes, whereas infected liver parenehymal cells unaffected because of their minimal proliferative activity. Systems have been developed to deliver genes in-vivo with replication-defective adenoviral vectors. A major in this approach is that suicide genes could theoretically enter non-target organs with high mitotic activity and, in this way, cause their destruction. The present study has focused the development of strategies to achieve targeted and  (1,8%); wedge resections 10 (18,1%). Colonic procedures consisted of: right colectomy 13 (23.6%); left colectomy 4 (7,2%); total colectomy 3 (5,4 %) subtotal .colectomy (1,8%); Anterior rectal resection procedures 19 (34,5%); low anterior resection 10 (18,1%); low anterior resection and hysterectomy 2 (3,6%); Miles .procedures 3 (5,4%). Average surgical time was 4,3hs (2-6hs). Average ventilatory support time was 12.8hs ; average intensive care stay was 49.4 (12-96hs). Mean hospital stay was 9 days (6-24). Postoperative complications were: intralxIominal abscesses 7 cases (12,7%); hepatic failure 3 (5,4%); sepsis 3 (5,4%); atelectasis 2 (3,6%); bile leakage (1.85); pleural effusion (1.8%) and anastomotic leakage (1.8%); 34 patients received post-operative systemic chemotherapy; operative mortality rate was 0%. The follow up was possible in 80% of patients, with an average time of 25,6 (3-108). Free of disease survival was 53,5% at year; 17% at 3 yr. and 7% at 5 yr. Actuarial survival rate was 87% at yr.; 37 % at 3 yr. and 30% at 5 yr. Conclusion: Simultaneous surgery can be performed with similar mortality and morbility rate than sucesive colorectal and hepatic surgeries. HPB Surgery Section, General Surgery Service, Hospital Italiano, Buenos Aires, ARGENTINA Scarcity of cadavedc donors, especially pediatric donors, has prompted the implementation of the living related donors liver transplantation program around the world.Despite the use of the reduced-size liver technique, mortality rate of patients On waiting list was 39.2%, in 1992 and above 50% in recipients with weight less than 10 kg. Since then, the living related donors technique has been an option for patients with low weight and end-stage liver disease. 24 donors whose mean age was 30 years (13 fathers, 9 mothers, uncle, grandfather) were studied. Preoperative evaluation included liver volume via CT scan, hepatic vasculature and a complete medical and psychiatric evaluation. 54% of potential living related donors were refused due to liver conditions 6, psychological reasons 4, asthma 1, hepatitis c. and pregnancy 1. Mean age of the 11 donors was 29.3 years (22-37) and weight was 65.5 Kg (47-76).

F031
Postoperative complications included case of duodenal ulcer and case of wound infection. Hospital mean stay was 5.6 days. All donors are currently well and alive. Recipients included 10 children with biliary atresia and child with ATT deficiency. The group's mean age was 2.2 years (0.9-10) and weight was 0.7Kg (7.4-26). Recipients urgency status was elective in 10 cases and of high urgency in case (primary non-function of cadavedc liver). The technique employed was similar to the described by Broelsch and others. Postoperative complications in pediatric transplants were: Biliary (1 biliary leakage and 2 strictures) 3 cases (27%); infectious (2 peritonitis and intrabdominai abscess) 3 cases (27%); portal thrombosis, case (9%); pulmonary haematoma, case (9%) and atelectasia case (9%). The patient in high urgency status died of a stroke while in the other group patient received a retransplant due to portal thrombosis and patient died due to multiorganic failure.Currently, mortality rate of children on waiting fist has decreased by 16.4%. Conclusion: The living related donor technique has allowed a decrease of mortality rate of children in waiting list. In this randomized controlled study, the effects of diuretics on ascitie fluid interleukin-1 (IL-11), opsonie activity (OA)and complement C3 concentrations were compared with those of therapeutic paracentesis in patients with mixed cirrhosis (i.e. mixed schistosomal hepatic fibrosis and cirrhosis) and tense ascites. Twenty four patients were randomly allocated to two groups: group A included twelve patients treated with spironolactone (200-400 rag/day) and group B included twelve patients treated with twice weekly 3-4 liters paracentesis for two weeks. Ascitic fluid samples from all patients were analyzed for total protein and albumin concentrations, C3 and lJ_,-l13 levels and OA at the beginning and 2 weeks after treatment. IL-1[, an immunoregulatory cytokine that stimulates a variety of cells that function as effector ofimmune response towards antigens, was significantly decreased following, paracentesis while it remained almost stable among the diuretic treated patients. The ascitic fluid OA and C3 concentrationsincreased significantly in diuretic treated patients (P<0.05) while patients treated with paracentesis had significantly decreased C3 concentration, andtheir ascitic fluid OA remained stable. It can be concluded that diuretics, may have the potential advantage over therapeutic paracentesis of providing better protection from spontaneous bacterial peritonitis. In the present study, systemic, portal and renal hemodynamics were assessed using Doppler flowmetry in 24 patients with schistosomal hepatic fibrosis (SHF); 12 of them had ascites due to portal hypertension and 12 were non ascitic and in 12 healthy subjects.
Also, renal function tests, fractional sodium excretion (FENa) and plasma level of glucagon, an endogenous vasodilator were determined in all patients and healthy subjects. Hemodynamic assessment ofsystemic and portal circulations showed a significant decrease in systemic vascular resistance and significant increases in cardiac index, and portal vein blood flow volume and congestion index in patients with SHF regardless ofthe presence of ascites as compared with healthy subjects. These circulatory changes were associated with significant increases in plasma glucagon levels in the ascitic and non-ascitic patients but without" significant correlations between these parameters. Moreover, these hemodynamic alterations become more marked with development of ascites. Renal hemodynamics showed a significant decrease in hilar renal blood flow and significant increases in renovascular resistance indices in ascitic and non-ascitic patients as compared with healthy subjects. It can be concluded that peripheral arterial vasodilatation particularly in the splanchnic area is the trigerring signal for sodium retention and ascites formation in patients with SHF. Hyperglucagonemia might contribute to the occurrence of these hemodynamic abnormalities. Considering these hemodynamic changes, the use ofvasoactive drugs in the management of ascites should be carefully studied in the future. The objective was to determine retrospectively the efciency of selective preoperative portal vein embolization (PPVE) in inducing hypertrophy of the future remaining liver before major hepatectomy resecting a substantial portion of functional liver parenchyma. Twenty eight patients with initially unresectable liver tumors had PPVE between sept 87 and sept'95. Twenty-three of them were curatively hepatectomized after PPVE. They represent 6% of the 371 patients hepatectomized for malignant lesions in the same period. PPVE was done under sonographic guidance, using a simple 5-F catheter, by the transparenchymatous approach of a branch of the portal vein, controlateral to the embolized liver. Different types of PPVE were performed, according to the site of the tumors: right liver, right liver and segment IV, right liver and left lobe, left liver and bisegment V-VIII, and central liver (segment IV-V-VIII). In 23 cases, isobutyl-2-cyancrylate glue was injected for free flow distal and proximal embolization, to block the portal tree definitively. 3-D volumetric assessments of the liver were done before PPVE and one month later, before hepatectomy. Remits Induced hypeophy of the non-embolized liver was successful in 25 cases (89%). The mean percent increase in volume was 70% (SD:50) for the 28 patients. The mean ratio between the remaining liver and the whole functional liver was 21.5% before PPVE, and 33.9% after PPVE. Three patients were not hetectomized because liver hypertrophy could not be induced and two patients for cancer-related reasons. Two patients died during the postoperative course with no symptoms of liver failure during the three first weeks following hepatectmy.
Conclusion PPVE was an efficient means of inducing hyperthrophy of the future remaining liver in 89% of the cases and permitted a 12.4% mean enhancement in the ratio between this remaining liver and the whole liver. These very good results were due to the distal and proximal free flow embolization technique with a non absorbable material, and an interval of one month between PPVE and hepatectomy. Portocaval shunt surgery for end-stage Budd Chiari Syndrome (BCS) is often unsucessful. We report the long term results in 8 BCS patients who were treated by hepatic transplantation over a period of 5 years (age range: 27-58 years). The indication for transplantation in 6 patients was chronic end stage decompensated liver disease and 2 patients was fulminant liver failure secondary to acute BCS. All patients had hepatic outflow obstruction confirmed by venography which also revealed significant caval obstruction in three patients (retrohepatic gradient> 12mm Hg). Standard orthotopic transplantation was performed with veno-venous bypass being used in 5 patients (mean ascitic volume 6 litres). Recipient hepatectomy was noted to be difficult in all patients with chronic disease because of severe peri-caval fibrosis. Coagulation profiles were continually monitored and the underlying hypercoagulable state treated appropriately by full early postoperative anticoagulation. Using this method, no arterial or portal thrombotic episodes have been encountered. Both patients from the fulminant group died. The first patient with acute BCS died of cardiac arrythmia after unclamping of the graft at the time of surgery. The second patient with acute BCS presented very acutely and initially underwent portocaval shunt surgerybut because of gross ascitic fluid production and secondary dilutional coagulopathy required urgent transplantation. Subsequent to this she developed multiorgan failure secondary to sepsis and died month after transplantation. All patients in whom OLT was performed for chronic disease (6)  2L' Hopital de la Pitie Salpetriere, Paris, France.
The integration and over expression of the herpes simplex virus type thymidine kinase (HSV1-TK) gene in localised tumours, results in turnout regression following the administration of the specific nucleoside analogue ganciclovir (GCV). Although only 10-20 per cent of the tumour cells take up the HSV1-TK the neighbouring cells die as a consequence of what has been termed the "bystander effect". Subcutaneous tumours were created following the injection of x 106 cells of the mouse colon adenocarcinoma cell line MC26. All control mice were co-injected with x 106 cells of the NB16 packaging cell line expressing the nls-lacZ gene and all test mice were co-injected with x 106 cells of the PLJ-TK packaging cell line expressing the HSV1-TK gene. The mice were divided into four groups: nude Balb/C mice into a control and a test group (Groups and 2 respectively) and normal Balb/C mice into a control and test group (Groups and 4 respectively). Seven days were allowed for retroviral gene transduction and tumour growth prior to treatment with GCV twice daily for five days. At the end of this time the animals were sacrificed and the tumour volume in each group was assessed. A significant tumour regression was observed in the test groups versus the control groups. The mean tumour volume was 42.1mm in the control groups (Groups and 3) compared with 3.3nun in the test group 4 (p<0.01). The test group for nude mice did not respond with the same efficacy only reaching a reduced tumour volume of 20.5ram (p<0.05) These data demonstrate a near. complete regression of established subcutaneous tumour in normal Balb/C mice following the successful transduction of the HSV1-TK suicide gene followed by treatment with GCV. The same was not true for the Balb/C nude mice suggesting a strong cell mediated immune component to the 'bystander effect'. It is therefore possible that suicide gene therapy may trigger a more general antineoplastic action by facilitating a specific anti-tumour immune response. Further experiments to determine if there is a generalised systemic antitumour immune response are in progress. We have also developed an animal model for the treatment of multiple hepatic metastases in the rat, with the packaging cell line delivered by hepatic artery cannulation. In this way it may be possible to treat patients with otherwise inoperable hepatic metastases by suicide gene therapy. 12 In order to capture hepatocellular carcinoma (HCC) cells in circulating peripheral blood, we ade analysis to see if a-fetoprotein {AFP) mPaNA exists in the peripheral blood obtained from patients with HCC and also, as a control, from hepatitis-viral-aarker-pasitive patients without HCC and a healthy volunteer. As the number of HCC cells in lcc of peripheral whole blood and the quantity of AFP alLNA are expected to be very stall, the analysis was performed by the reverse transcription followed by an original three-step polymerase chain reaction. By this highly-sensitive method, 5 of 7 HCC patients were positive for AFP mRNA. These 5 positive patients consisted of 3 with clinically apparent recurrence, one preoperative patient with tumor thrombus in the portal vein and one recurrence-free patient who developed clinically detectable recurrence 3 months after this analysis. On the other hand, one follow-up patient without any clinical recurrence and negative for AFP eflNA showed no recurrence even after 3 months. One preoperative patient, whose serum AFP protein level was as high as 67213 ng/ml was negative for AFP mRNA Neither 4 patients with positive viral markers nor a healthy volunteer was positive. The results suggest that detection of AFP mRNA from HCC patients' peripheral blood by our highly-sensitive RT-PCR may be a practical and powerful tool to diagnose the preoperative spreading of ItCC and to monitor its recurrence. The reoxygenation phase that follows period of prolonged anoxia i, characterized by the massive formation of oxygen free radicals (OFR), highly reaetiw, substances responsible of damage to cell membranes. Ageing is associated with reduction of antioxidant status. Aim of this study was to determine the effects o:f ageing on the sensitivity of liver cells to a period of anoxia/reoxygenation. OFIt formation and cell injury were evaluated in hepatoeytes isolated from rats of different ages.
METHODS: Sprague ,Dawley male fed rats of 2 or 8 months of age were utilized.
After isolation, liver cells were cast in agarose gel threads and perfused with oxygenated Krebs Henseleit bicarbonate buffer (KHB). A 2 h period of anoxia wa., obtained shifting the gas phase of the perfusate to 95%N2-5%CO2. Successively, tht cells were reperfused for h with oxygenated KHB. OFR were evaluated by enhanced chemiluminescence: anion superoxide (02-) by adding mM lueigenin to the solution, hydrogen peroxide (H202) by adding 10 mM luminol. Cell damage was assessed by measuring LDH release in the effluent. RESULTS: in basal conditions, no differences were observed in the levels o:f lucigenin or luminol-enlumced chemiluminescence between the two groups (13:t:1 vs 144-4 nA and 234-3 vs 244-5 rtA, respectively). During anoxia, lucigenin ox luminol-enhanced chemiluminescence, expression of 02-and H202 formation, respectively, decrease to background values, while LDH release was signifieamtly greater in cells isolated by older animals (750% vs 490% after 2 h of anoxia; p<0.05). During reoxigenation, OFR formation increased in both groups; such a rise, however, was markedly greater in cells obtained from older rats: 02-reached a peak after 15 rnin (1404-10 vs 1004-11 hA; p<0.05), while H202 increased progressively during the hour of reoxygenation (1214-9 vs 834-12 rib.; p<0.05). LDH release rise markedly during reoxygenation in both groups; newly, however, higher values were observed cell obtained from older rats (1300% 750%; p<0.05). The peak of lueigenin.enhanced chemiluminescence, expression of 02production, was observed 10-15 min before the peak release ofLDH. CONCLUSIONS: our results show that: 1) liver cells produce high level of anion superoxide and hydrogen peroxide during the reoxygenation phase that follows period of prolonged anoxia; 2) the age of the rats influences the sensitivity of liver cells to anoxia/reoxygenation injury. Maintenance of high portal perfusion is the most important feature of DSRS. When spleno-pancreatic disconnection (SPD) is added survival and quality of living are increased. 91 Patients that previously bleeded from oesophageal varices underwent DSRS in the first 26 a traditional, limited, porto-splenic disconnection was performed in the other 65 a splenopancreatic, spleno-colic and gastro-colic disconnection (SPD) was added. Peri-operative mortality was 14%, related to Child's risk (Child A-B 4%).
No mortality due to rebleeding from esophageal or gastric varices was reported; loss of long-term PP was observed in 23%. Persistence of disease and incomplete SPD were the two prognostic factors related to survival (38% vs 11%-p<0.001 and 43% vs 14%-p<0.05 respectively). A retrospective study was carried out using our liver resection data base to determine the relationship between Central Venous Pressure (CVP) and blood loss during liver resection. The data base contained ii0 patients who underwent liver resection between 1980 and 1995 and who had their CVP recorded. Since January I, 1991 an attempt was made to keep the CVP as low as possible especially during the liver transection phase. Prior to January i, 1991 no attempt was made to maintain a low CVP. Patients were divided into 2 groups: major liver resections (n=66) and minor liver resections (n=44). A liver resection of 1-2 Couinaud segments was considered a minor liver resection and a resection of 3 or more segments was considered a major liver resection.
Mean CVP, Estimated Blood Loss (EBL), and Number of Units of. Blood transfused (NUB) were compared in major and minor liver resections before (n=49) and after (n=71) January I, 1991. There were 31 major and 18 minor liver resections before January i, 1991 and 35 major and 26 minor liver resections after that date. Analysis of variance confirmed that after the introduction of our low CVP policy there was a decrease in both EBL and NUB; and as expected the reduction was greater in patients undergoing major liver surgery (P=O.OI). A low CVP is associated with less blood loss during liver resection. It is recommended that liver surgeons communicate this information to the anaesthetists involved in the care of their patients in order to achieve the goal of keeping the CVP at or below 5 cm H20 to minimiz blood loss during liver surgery. splenorenal shunt (DSRS) from 10/92 to 10/95. Selection for DSRS was based on: variceal bleeding refractory to endoscopic and pharmacologic therapy; adequate liver function (Childs A-31, B-10, C-1); etiologies were alcoholic-14, cryptogenic-9, PBC-6, post hepatic-6, portal vein thrombosis-3, others-4. Five patients were shunted emergently, and 37 had elective DSRS, with standard DSRS in 37 patients and modified selective shunts in 5. There was no hospital mortality. The median hospital stay was 9 days, with median hospital charges of $23,000. There was one shunt thrombosis in a patient with a myeloproliferative disorder who required splenectomy/devascularization. Early rebleeding occurred in 3 patients (7%): all had patent shunts, and bleeding resolved with nonoperative management. Late rebleeding occurred in one patient (2%) in the presence of a patent shunt. Mild/moderate encephalopathy was found in 4 patients (10%) at median follow-up of 18 months. There have been 2 late mortalities, one from liver failure and one with a progressive myeloproliferative disorder. Two patients have required liver transplant to date, one with sclerosing cholangitis, and one with a cryptogenic cirrhosis 16 months after emergent DSRS. We conclude that DSRS plays a role in managing patients with variceal bleeding in the 1990's: good risk patients can have variceal bleeding controlled by DSRS with low morbidity. PORTAL VEINTHROMBOSINADULT CANDIDATES TO LIVER TRANSPLANTATION.
Sidalqo,E.,Bilbao,I.,Lzaro,JL.,Murio,JE., Balsells,J.,Charco,R.,Gifre,E.,Ruiz,C., Margarit Eleven non cirrhotic patients that underwent major resectional liver surgery under TVE were studied. The duration of TVE (warm liver ischaemia) ranged between 16 and 48 minutes. Biopsies were taken at three stages: pre-ischaemia, before the application of vascular clamps; ischaemic, just at the end of TVE; and post-ischaemia, 12 to 120 minutes following reperfusion. Light-electron microscopy study was performed to study the hepatic architecture, hepatocyte morphology, bile canaliculi and sinusoids. All specimens were coded and assessed blindly for tissue (light microscope level) and cellular (electron microscope level) changes. Prior to TVE hepatocytes and hepatic architecture were as normal as could be expected. During ischaemia the hepatic architecture was distorted due to widespread collapse of sinusoidal spaces. Hepatocytes showed focal chromatin condensation at the nuclear margins, distented nuclear envelope and swelling of both mitochondria and endoplastic reticulum. After reperfusion these changes reversed.
It can be concluded that TVE over a period of 48 minutes has no irreversable deleterious effects on the ultrastructure of the non-cirrhotic liver. Liver resection in patients with liver cirrhosis (even in the absence of overt liver insufficiency) is associated with greater risk than in patients without underlying liver disease. Because the incidence of HCV-cirrhosis related HCC is anticipated to increase rapidly in the near future we have assessed, by multivariate analysis, parameters associated with in-hospital mortality and morbidity in a consecutive series (1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994) of 108 Childs' grade A cirrhotic patients undergoing liver resection of an HCC (1 or less liver segment, 2 segments or or more segments in 42, 23 and 43 patients respectively).
Parameters entered for analysis included age, aetiology of cirrhosis, preoperative serum bilirubin, AST, ALT, GGT, albumin, creatinine levels as well as prothrombin time, presence or absence of pathological features of superimposed active hepatitis, extent of resection, type and duration of vascular clamping and amount of intraoperative blood loss. Overall incidence of in-hospital death and major postoperative complications were 8.3% and 48.1% respectively. By univ,'uiate analysis, preoperative serum ALT levels (p=0.001) and intraoperative transfusions (p=0.01) were the only parameters significantly associated with inhospital death. However, only serum ALT concentrations was an independent risk factor. In-hospital mortality in patients whose preoperative serum ALT was below 2N (n 77), comprised between 2 and 4N (n 23) and greater than 4N (n 8) was 3.9 %, 13.0 % and 37.5 % respectively. Increased ALT levels (> 2N) was also associated with an increased incidence of postoperative ascites ( Membrane fluidity of tumor cells has been suggested to influence tumor invasion or metastases. The present study was conducted to clarify whether or not membrane fluidity (MF-value) of hepatocellular carcinoma (HCC) is related to postoperative recurrence. Membranes of reseCted HCC tumors and non-tumorou liver tissues from the same patients were prepared by gradient ultracentrifugation and fluidity was determined by 'pectrofluorophotometer equipped with polarizer using 1,6-diphenyl-1, -hexatriene as a probe dye.
MF-values of both tumor and non-tumor_ tissues (0.232 and 0.231, n=55, individually) were significantly higher than that of normal liver controls (0.190, n=12, p<0.001). However, no significant difTerence was observed between tumor and non-tumor, as a whole.
Although MF-values of HCC's did not concern tumor size, number, capsular invasion or intrahepatic metastasis, they significantly related to tumor invasion into the portal vein (p<0.05) and serum a-fetoprotein value (p<0.05}. Recurrence group (0.212, n=23) had a significantly lower F-value than that of recurrence-free group (0.242, n=32, p<0.01). From the comparison of MF-value between tumor and non-tumorous liver tissue in individual case, HCC patients were classified into three groups. Group MF-value of tumor is kigher than that of non-tumor and the difference is more than 0.01.
(n=28) Group 11: MF-value of tumor is lower than that of non-tumor and the difference is more than 0.01.
(n=19) Group I: Difference in MF-value between tumor and non-tumor is less than 0.01.

(n=9)
Three-year disease-free survival rates were 54 in group and 8 in groupH and p=O.021 by og-rank analysis.
In conclusion, membrane fluidity intimately concerns postoperative recurrence of HCC after hepatic resection. A follow-up programme was carried out for 240 patients that were operated in our center due to hydatid disease, but only 131 patients could be contacted and clinically evaluated. A complete physical examination, routine chest x-ray, abdominal ultrasound (US), abdominopelvic CT and indirect hemaglutination for hydatid disease (IHA) were performed on all patients. In this study, the data of 45 patients who were either found to be recurrent after this investigation or were operated in our center initially with the diagnosis of secondary hydatidozis, were analized. The effects of misdiagnosis, mistreatment and inappropriate follow up methods on the reccurrence rate have been evaluated.
Of the 45 cases, 6 had recurred twice, 2 had recurred three times. Altogether there were 55 recurrences. 30% dextrose solution was used as a scolosidal agent at the first operation in 51.6% of the case. While in 71.1% of the 45 cases multipl hydatid cysts (HC) were found, 48.8% of these had recurred in the liver as uniloculer cysts. 24.4% of the first recurrences were free intraabdominal cysts (FIAC). 4 of these "were FIAC only and 7 of them were accompaind by liver hydatitozis. Only of these 11 FIAC were diagnosed as FIAC in the first operation. Totally, 13 of the 55 recurrences (23.6%) had an extrahepatic intraabdominal localization. When 13 iHD cases were considered, the reccurrence rate in solitary cysts was 18.2% and the recurrence rate in multipl cysts was 31.6%. In FIAC this recurrence rate was even higher, being 56.25 %. 9 cases (16.4) were diagnosed in the early 2-6 months period following the first operation. Those were probably not real recurrences and should be accepted as preoperative or intraoperative insuffient evaluation. It was found that US was 20% less efficient than CT in determining the recurrences. In four cases who were not included in these 45, CT and/or US suggested recurrence but diagnostic aspiration excluded this possibility. In tiffs study IHA was also used as a parameter to define recurrence: When titrations which were higher or equal to previous titrations were considered as an index of recurrence, IHA had a 92.3% sensitivity and 46.8% specitity.
In this series, total cystectomy was performed in a very limited number of cases and there was not a cese in which pericystectomy was performed. Therefore, the effects of the application of radical procedure on recurrence could not be evaluated.
In conclusion, careful preoperative and intraoperative evaluation and meticulous surgery will decrease recurrences in hydatid diseases. COMPLIANCE of EXTENDED LIVER RESECTIONS Kupcsulik P.K.,Pinkola K.Antony P.,Harsanyi L,Flautner L.
1st Dept of Surgery,Semmelweis Univ.of Medicine,Budapest Between 1989 and 1995 443 patients underwent surgery for liver tumor. 244 interventions were performed for liver malignancy. Preoperative chemoembolisation was applied in 47 cases of liver carcinoma. No direct intraoperative advantage of arterial embolisation was detected comPared to the nonembolised patients. In contrast to this 2 year survival of chemoembolised patients is superior to the controls ,however. Intraoperative portal clamping was used in 26 cases. Noninvasive circulatory monitoring by impedance cardiography showed transient, but significant diminution of CO and decrease of PVR. 3 minutes reperfusion after 30 minutes liver ischaemia resulted significant elevation of CO, but complete restitution occured only in the early postoperative phase. Intraoperative blood loss interferes significantly with CO recovery. In 1/26 pts with portal clamping splenic rupture occured. No long lasting splanchnic stasis or any other form of circulatory consequences after portal clamping was observed.
Despite to the risk of transient portal devascularization, its combination with ultrasonic dissector significantly diminishes blood requirements during extended liver resections.186 surgical interventions resulted in removal of tumorous mass. Lethality of this group was 3.1. percent respectively.
The use of ultrasonic dissector, its combination with partial vascular liver exclusion diminishes intraoperative blood loss, and promotes extended resection of the liver. The best method of biliary reconstruction is controversial and much debated technical aspect of orthotopic liver transplantation (OLT). The aim of this study to compare end to end (E-E) and side (S-S) duct to duct biliary anastomosis after OLT.
Patients undergoing OLT were randomized to receive E-E (n=50) or S-S (n=50) duct to duct biliary anastomosis without T tube. Patients age, sex, graft preservatior time and indication for transplantation were similar in both groups. Within 30 days of OLT all patients were scheduled to have endoscopic retrograde cholangiography (ERC). Cholangiographic findings were classified normal, leak or stricture. Biliary complications were defined as leak or stricture which required endoscopic or surgical treatment. Data were analyzed according to intention to treat. 60 patients received E-E and 40 S-S anastomosis. 10 patients randomized to have S-S had E-E anastomosis done due to shortness of recipient or donor duct. Number of biliary abnormalities on ERC were similar in both groups (E-E 32 % vs S-S 30%, NS), leaks (18% vs 16%) and biliary strictures (14% vs 14%).
There was difference in .number of biliary complications (E-E 22 % vs S-S 22 %). Patients and anastomosis survival was similar (median months E-E 11 vs S-S 13.5,NS, E-E 7.5vs S-S 8.5,NS). This is the first randomized study of E-E versus S-S biliary reconstruction following OLT and it showed no short term benefit of either technique. We would like to conclude that either of anastomoses could be used according to surgeon preference. Between May 1993 and October 1994, 34 patients with suspected colorectal liver metastases were evaluated. There were 18 males and 16 females with an average age of 62 years, 5 month. Staging by conventional radiology consisted of abdominal CT (n=34), chest Xrays (n=15), and chest CT scans (n=19) to evaluate extrahepatic disease i:t all patients, and magnetic resonance imaging (n=24) and CTantiography (n=3) of the liver to determine anatomical resectability in 27 patients.
PET scan was performed in all patients within 8 weeks of conventional radiology.
PET consequently affected the clinical management in i0 patients (29%). However, PET did not provide any additional information compared to conventional radiology in the assessment of hepatic metastases per se.
In conclusion, PET improves the preoperative evaluation and selection of patients with isolated colorectal liver metastases for hepatic resection. Liver resection still remains the best treatment for the majority of hepatocellular carcinomas (HCC) in cirrhotic patients. The purpose of this work was to analyze the results of resection in a series of 200 consecutive patients with HCC and cirrhosis operated on from 1992 to 1994. Mean age was 61 years (range 62 to 78 years). Cirrhosis was of alcoholic origin in 40% of patients. HBV and HCV markers were positive in respectively 23% and 59% of patients. Liver function was good in 79% of patients and most of them (74%) had a single HCC nodule. Nineteen p.cent had major (right or left) and 81% had limited liver resection. Operative mortality was 7.5%.
Cumulative 5-year survival was 30% and cumulative 5-year recurrence rate was 69%. The survival rate was significantly higher in Pugh's A (34%) than Pugh's B-C patients (12%n p<0.001) and in patients with preoperative serum oFP concentration less than 500 ng/ml (36%) than in those with higher levels of aFP (10%, p<0.01).
The survival rate was significantly higher in patients without pathological predictive markers of recurrence. Significant pathological markers of recurrence were fhe absence of free margin of non tumorous parenchyma (p<0.01), a tumor above 5 cm (p<0.02), the absence of a capsule around the tumor (p<0.01) and the presence of satellite nodules and distal portal invasion (p<0.001). Postoperative adjuvant arterial chemotherapy slightly decreased the recurrence rate but did not increase survival. This study confirms that liver resection is a good treatment of HCC in cirrhotic patients with a good liver function and a small unique nodule with no biological nor pathological signs of invasiveness. Our previous studies have shown that radioactive yttrium-90 microspheres, when given in the hepatic artery, concentrated preferentially in liver tumours, giving an average tumour normal internal radiation dose ratio of 4:1. The need for intraoperative dosimetry by laparotomy has been replaced by the use of simulation technetium-99m macroaggregated albumin scan which have been shown to accurately predict the distribution of the therapeutic yttrium-90 microspheres.
Over a period of 3 years, 89 patients with inoperable hepatocellular carcinoma were treated with hepatic intra-arterial yttrium-90 microspheres given through an angiographic catheter placed percutaneously using the Seldinger technique. The dosage given was calculated using a partition model taking into account of extrahepatic shunting, the tumour normal ratio and the liver tumour non tumour volumes measured on computed tomography. There were 76 males and 13 females. The median age was 54 years (range 24-85). Sixty-eight patients had primary inoperable tumours, 20 had post-operative recurrences, and had recurrence after lipiodol-I TM therapy. The turnout dose given was > 12,000 rads.
All patients had a drop in alpha-fetoprotein of over 80% of the pre-treatment level. Patients with pre-treatment normal alpha-fetoprotein were monitored with ferritin levels which showed a similar drop after therapy. The median survival was 8 months (median survival for systemic chemotherapy in previous studies was 10 weeks). The patient with the longest survival was 25.9 months after treatment. He is still alive.
This study suggests that radioactive yttrium-90 microspheres may be of use in patients with inoperable hepatocellular carcinoma. A randomised study to further evaluate this treatment is indicated. Obstructive jaundice as the main presenting feature of hepatocellular carcinoma is uncommon.
In a period of 111/4 years, 48 of 2,037 patients with hepatocellular carcinoma seen in our hospital presented with obstructive jaundice (2.4%). There were 39 men and 9 women. The mean age was 54 years (s.d. 12.9). Obstructive jaundice was diagnosed by blood tests, ultrasound and cholangiography (ERCP or PTC). Hepatocellular carcinoma was diagnosed by either histological proof or in patients with space occupying lesions in the liver with serum alpha fetoprotein of over 500 ng/ml. Ultrasound showed dilated intrahepatic biliary system in all 48 patients. The pathologies causing obstructive jaundice included tumour casts (n 10), free tumour fragments in the extrahepatic biliary systems (n 7), diffuse tumour involvement of intrahepatic ducts (n 28), extrahepatic bile duct obstruction by enlarged porta hepatis lymph nodes (n 3) and haemobilia (n 8). Some patients had more than one pathology. Patients with potentially operable liver tumours were further evaluated with computed tomography and hepatic angiography. Preoperative investigations revealed 37 patients to have inoperable tumours and they were treated with endoscopic stent (n 24), percutaneous stent (n 6), stent put in by combined endoscopic percutaneous approach (n 2) or supportive treatment only (n 5). Eleven patients underwent laparotomy and "curative resection" was possible in 9, while the other 2 had surgical intubation only. With proper management the survival of patients with hepatocellular carcinoma with obstructive jaundice was similar to those without jaundice. Four patients who had "curative" liver resection were still alive, had no evidence of disease at 11.6, 13.7 and 99.2 months after operation and one had recurrence and is still alive 20.5 months after surgery.
Patients with hepatocellular carcinoma with obstructive jaundice should be treated actively. Good palliation, and occasional cure, are possible. Between 1972 and 1995, 612 patients underwent laparotomy in view of liver resection. There were 400 males and 212 females with a mean age of 56.4 +/-14.3 years. Indication for surgery included hepatocellular carcinoma (n=207), other primary liver tumors (n=47), liver metastases (n=221), benign liver tumors (n=77), liver trauma (n=21). and others (n=39). Variables were compared using X test or Fisher's exact test. Survival was calculated according to the Kaplan-Meier method and survival curves were compared using the Logrank test. RESULTS 205 patients underwent major liver resections comprising at least segments. 128 patients underwent right hepatectomy, 63 patients underwent left hepatectomy and 14 patients had or 4 non-contiguous segments removed. 199 patients had or 2 segments resected and 68 had non-anatomical resections. 131 exploratory laparotomies were performed for unresectable tumors. Overall resectability rate was 78.6 %. Operative mortality and morbidity was 6% and 21% respectively. 5-year survival following resection of colorectal metastases, hepatocellular carcinoma developed on cirrhotic and non-cirrhotic livers was 24.6%, 21.7% and 28.6% respectively (p= NS) When results of our early experience were compared to those obtained after 1990 there was a significant decrease in operative mortality and morbidity. Five year survival was also improved. Indications of OLT for NEM are still a matter of controversy, as far as only limited data are available. This retrospective study involves 31 cases of OLT collected among 11 centers from 1989 to 1994. There were 17 males and 14 females, age 45 + 9 years (M + SD). Primary tumor site was pancreatic in 17 cases, ileal (7), bronchial (3), gastric (2) colonic, rectal or lymphatic (1 each). Tumor varieties were carcinoid (15 cases), gastrinoma (7), glucagonoma (1), nonfunctioning (8). Hormonal-related symptoms were present in 17 patients. Seventeen patients (54 %) underwent excision of the primary tumor and 23 (74 %) received chemotherapy before OLT. There was a 30 + 32 months interval (median 19, range 2-120) between the diagnosis of NEM and OLT it was 31 + 34 months for carcinoid and 29 + 30 months for others-NEM. Fourteen patients had not undergone resection of primary tumor at the time of OLT simultaneous removal of the primary tumor and NEM was performed in 11, one by ileal resection, 3 by Whipple resection and 7 by upper abdominal exenteration (with synchronous pancreatic transplant in 3 cases). In the 3 remaining cases, the site of the primary tumor was discovered only after OLT.

LIVER RESECTIONS
Six patients (19 %) died post-operatively 4 out of 7 after "cluster 'esection" and all 3 after composite liver and pancreatic graft. Major surgical complications occured in half of the patients. Among the 25 survivors, 4 died from late complications of the procedure, without evidence of recurrent disease (4, 5, 8 and 10 months after OLT), and 8 patients died of recurence (2 to 41 months after OLT). Thirteen patients are still alive. Overall actuarial survival rates were 58 % at one year and 36 % at 5 years (5 patients). Theses rates were 80 %, 80 % and 69 % at 1, 3 and 5 years respectively for the 15 patients with carcinoid tumors, versus 38 %, 15 % and 0 % respectively for the 16 patients with others NEM. Theses results suggest that OLT can achieve benefit in patients with carcinoid metastases, but not in patients with other type of NEM.
The efficacy of hepatic resection for more than 4 liver metastases from colorectal carcinoma?
During the last decade, surgeons have taken a more aggressive attitude in the treatment of primary or secondary liver tumors, partieulary those from colorectal primaries. Several authors have demodstrated, that resection of a single metastastie lesion is benifieial for the patient. From jan. 1986 til dee. 1993, we underwent by 142 patients liverreseetions for hepatic metastases of colorectal cancer. The patients devided in three groups (group 1: one metastase n 61, group 2:2-4 metastases n=48, group3: more than 4 metastases n=23), which were analysed retrospectivly. The overall hospital mortality was 3,52%. The results are illustrated in table 1. Tab.l: Our results represent, that the most benifit have patients with RO resection. The number of metastases in the liver is a deeeiding factor for the prognoses and the survival benifit patients will have from the resection. The group 3 had the highest hospital mortality and only Ro reseeted patients will have a benifit from the liverresection. Normal hepatic parenchyma is supplied by portal and arterial blood. In contrast, overt liver tumours have a predominant arterial supply, which is the basis for treating hepatic tumour with arterial ischemia. This report reviews our total experience for hepatic non-endocrine tumours during 1971-1995. The material consists of 80 patients (aged 59 (37-77) years) with primary liver cancer (n=15), colorectal liver metastases (n=61) or other hepatic malignancies (n=4) and ineludes all consecutive developments: complete (permanent) dearterialization (n=12), intermittent (16 h) dearterialization (n=26) and repeat transient dearterialization (1-2 h once or twice daily using an implantable occluder) (n=42 A retrospective study was conducted to determine the incidence, risk factors and consequence of biliary complications after hepatic resection and to evaluate the management. Between January 1989 and October 1995, 347 hepatic resections were performed at the Department of Surgery, The University of Hong Kong at Queen Mary Hospital. There were 251 male and 96 female patients with a mean age of 53.2 years (range 2-86). Major hepatic resection (defined as resection of or more segments according to Couinaud's description) was performed in 229 (66%) cases"and the liver was not cirrhotic in 169 (49%) patients. Twenty-eight (8.1%) patients developed biliary complications. The biliary complication rate was 6.5% from 1993 to 1995 and 9.5% from 1989 to 1992 (p=0.31). The hospital mortality rates for the 2 periods were 4.8% and 12.3% respectively (p=0.01). The clinical presentation included biliocutaneous fistula (n 18), peritonitis (n 3), intraabdominal abscess (n=6) or intraoperative bile duct injury (n=l). The site of leakage was from the raw surface (n =2), right/left hepatic duct stump (n=6), common hepatic duct (n=3), bilioenteric anastomosis (n=4), Ttube insertion site (n=l) or unidentified (n= 12). Logistic regression analysis revealed that the risk of biliary complication was increased by older age, higher preoperative white cell count, longer operation time and a higher serum total bilirubin days after operation. Ten of 28 (35.7%) patients with biliary complications died in hospital, most often due to intraabdominal sepsis. Initial treatment consisted of surgery (n 9; 6 died), percutaneous drainage (n =9; died), endoscopic papillotomy 4stent (n =4; died) and conservative (n =6; nil died). The development of biliary complication was associated with a higher incidence of liver failure and prolonged hospital stay. We conclude that biliary complication is a frequent cause of morbidity and mortality after hepatic resection and may be avoided by shortening of operation time. A high bilirubin level days after hepatic resection should arouse suspicion and facilitate early detection. Reoperation carries a high mortality and non-operative management is effective in selected patients. Liver transplantation (OLTx) is most commonly performed with vena caval replacement.Veno-venous bypass (VVB) during the anhepatic phase is used routinely, selectively, or not at all. Without VVB, significant volume loading is required; there maybe hemodynamic instability, and bowel edema maybe problematic. By preserving the vena cava, and anastomosing the donor vena cava to the recipient's hepatic veins ("piggy-back"), the complications associated with vena caval clamping and replacement are obviated. Portal vein (PV) decompression is accomplished by creating a temporary portacaval shunt (PCS) during the native hepatectomy, which is then ligated and taken down during the graft implantation.
Results: We have used this technique in 58 consecutive adult OLTx recipients. The etiology for liver disease included: Hepatitis C (n=14), ETOH (n=ll), Cryptogenic (n=6), PBC (n=7), PSC (n=6), and other (n=14). 2 patients had previously been transplanted. 10 patients had previous TIPSS procedures. 2 patients required PV extension grafts because the TIPSS was extrahepatic. patient had a large spontaneous porta-systemic shunt and did not require a PCS. All  Jejuno-ileal (JI) bypass was developed as a therapy for morbid obesity in the late 1960' s, but has since been abandoned because of a high rate of complications, including cirrhosis (incidence 7-10%). The need for liver transplantation (OLTx) after JI bypass has been infrequent, with only 4 previous patients reported in the literature. However, as the time to the development of symptomatic end-stage liver disease (ESLD) after JI bypass maybe quite long (25 years or more), the incidence of patients who will require OLTx may now only be increasing. We retrospectively reviewed our experience with JI bypass and OLTx in 280 adult patients since 1985] Results: 3 patients (2 females) have undergone OLTx for decompensated ESLD due to steatohepatitis after JI bypass, all within the last 48 months. One patient had concomitant alcoholic liver disease. The mean age of this group was 51 years (range 46-60). The mean duration from time to JI bypass to OLTx was 24 years (range 23-35). Two of three patients had other complications related to the JI bypass (renal and biliary stones). One patient had the JI bypass taken-down prior to OLTx, which precipitated acute liver and renal failure, necessitating urgent OLTx. One patient had the JI bypass taken-down at the time of OLTx. The third patient continues to have the JI bypass and is followed closely with monthly liver function studies and yearly liver biopsies,.and to date (2 years post-OLTx) has not demonstrated any complications attributable to the JI bypass. The patient who underwent concomitant OLTx and JI bypass takedown, has developed recurrent obesity, while the other two patients have not..Conclusions: The incidence of patients who require OLTx after JI bypass maybe on theincrease. Takedown of the JI bypass may precipitate acute liver failure in the cirrhotic patient, and lead to the need for urgent OLTx. JI bypass reversal should be accomplished either at the time of OLTx, or if signs of liver dysfunction occur after OLTx. OLTx recipients maybe at risk for recurrent obesity after takedown of the JI bypass. Ist. di Clinica Chirugica and *Radiologia, University of Bad, Italy Changes in liver blood flow were observed in presence of overt metastases by invasive and non invasive methods. Doppler Ultrasound is a non invasive method to quantify liver blood flow by measuring the mean velocity Vm and the crosssection area A of the hepatic artery and the portal vein. To assess the relationship between blood liver perfusion and the presence and extent of liver metastases, the Hepatic Perfusion Index (HPI), defined as the Arterial/Total blood flow ratio, was evaluated by Colour Doppler US in 20 patients with US detected hepatic colorectal cancer metastasis (MET), in 33 patients with confined colorectal cancer (CA) and in 24 control subjects (CON) Nearly 40% of patients with colorectal malignancies will develop liver metastases in the course of their illness. The aim of this retrospective study was to determine predictive factors of long-term survival in patients treated surgically.
Between 1976 and.1995, 166 patients underwent surgery for colorectal liver metastases. There were 98 males and 68 females with a mean age of 59.8 years (range: 31-80). 123 metastases were metachronous and 43 were synchronous of the primary tumor. Prognostic variables were analyzed using the 2 test and Fisher's exact test. Survival was calculated according to the Kaplan-Meier method and survival curves were compared using the Logrank test.
100 patients underwent major liver resections comprising at .least 2 liver segments: right hepatectomy (n=48), left hepatectomy (n=17), resection of 2,3 or 4 non-contiguous segments (n=35). Minor resections and non-anatomical resections were performed in 22 patients. The resectability rate was 86.5%. Operative mortality and morbidity was 6.2% and 28% respectively. Mean hospital stay for survivors was 16 days. 1,3 and year survival rate was 75%, 32.3% and 24.6% respectively. Predictive factors of long term survival included size of metastases, number of metastases, staging of the primary tumor and resective margins. Worldwide hepatocellular carcinoma is among the most common malignancies. The aim of the present study was-to identify prognostic variables having an influence on long-term survival.
Between 1974 and 1995, 207 patients underwent surgery for hepatocellular carcinoma. There were 185 males and 22 females with a mean age of 60.1 +/-11.7 years. The liver was cirrhotic in 149 patients and non-cirrhotic in 58 patients. Prognostic variables were analyzed using the Z test and Fisher's exact test. Survival was calculated according to the Kaplan-Meier method and survival curves were compared using the Logrank test. 37 patients (18%) underwent major liver resection comprising or more segments. There were 25 right hepatectomies, 12 left hepatectomies, 25 left lobectomies, 6 right lateral segmentectomies, 57 of or 2 segments and 11 non anatomical resections. 71 patients had exploratory laparotomy only. The resectability rate was 65.7% Operative mortality for patients resected was 8.4%, 9.5% for those with cirrhosis versus 4.8% for patients with non-cirrhotic livers (p= NS.). Morbidity was 34.7%. 1, and year survival for cirrhotic patients was 70%, 43.1% and 21.7% respectively. 1, and 5 year survival for non-cirrhotic patients was 81%, 50% and 28.6% respectively (p= NS). Prognostic  Since October 1979, we have laparotomized 1391 patients for hepatectomy. Hepatic resection was undergone 1274 (91.6%). among them, 829 had hepatocellular carcinoma and hepatectomy was carried out 777 (93.7%). Five hundred ninety five patients had the largest tumor no more than 5 cm in diameter. Hepatic resection was camed out in 572 (96.1%).
Resection area of the liver was decided depending on the indocyanine green retention rate at 15 min. After laparotomy, intraoperative ultrasonography was performed for .evaluation of extent of HCC in the liver and indication of hepatic resection and operative procedure were finally decided. Vascular occlusion techniques such as selective vascular occlusion or Pringle maneuver were routinely applied. Division of the liver was proceeded under the guide of ultrasound.
Survival rate of all the series at five years was 40 %. Survival rate in patients received systematic subsegmentectomy was significantly better than that in limited resection. The most significant prognostic factor was year trends. After 1985, 5-year survival reached to.65 %. Repeated resections were one of the important factor to improve long term survival.
Since 1990, percutaneous ethanol injection therapy (PELT) has been generally performed. However our recent group study clearly suggested that first choice of treatment in small HCC was surgery, not PELT. Even in early HCC, hepatic resection showed much better survival than PELT when pathological diagnosis was made by 6 pathologists specialized in liver tumors. Therefore, indication of PElT in Japan may be limited in near future. Bortolasi*, C. lacono*, A. Ace. rbi*, G. Serio* Departments of Surgery and Radiology , University of Verona, Verona, Italy A certain preoperative diagnosis is mandatory in the management of benign tumors of the liver as hemangioma (HMG), focal nodular hyperplasia (FNH) and hepatocellular adenoma (HCA); periodic follow-up is the best therapeutic choice. In order to evaluate diagnostic and therapeutic work-up we review retrospectively the results of biochemical data and diagnostic imaging procedures as US, CT, abdominal angiography (ANG), magnetic resonance (MRI), tecnetium colloid scintigraphy (TCS), DISIDA scintigraphy (DISIDAS) of 90 pts with HMG, 50 pts with FNH and 12 pts with HCA observed from 1975 to 1995. Sensitivity and specificity of each test for benign lesions were determined through the ratio of the true .and false positive rate and the true and false negative rate at radiologic findings of the total number of hepatic neoplasms, considering not only benign tumors but also malignant lesions as hepatocellular carcinoma, colangiocellular carcinoma and metastases. All the results were confirmed by hystologic diagnosis. In the pts with HMG, MRI and ANG reached the best sensitivity and specificity (respectively 96% and 100%), while specificity of US and CT were respectively 66% and 99.3%. In the pts with FNH US had a low sensitivity (66%) but a very high specificity (98%), while CT scan showed sensitivity 88% and specificity 99%; in this group of pts the best results were reached by TCS (sensitivity 94% and specificity 100%) and DISIDAS (sensitivity 91%, specificity 100%). In the pts with HCA the best results were achieved with ANG (sensitivity 75%, specificity 98%), while US showed the worst sensitivity (only 7%); in this case the most common differential diagnosis was with malignant form. The present study was done to assess the efficacy of vmious therapeutic modslities in the management of gastric vmiceal bleeding. Dining the last I0 yesrs, 60 patients presented with bleeding from astric vsrices (GV). The cause ofportsl hertemion was cirrhosis in 23, extrshepatic portal venous obstnon (EHPVO) in 22, non cirrhotic portal fibrosis in 13 and/solsted splenic vein thrombosis in 2 patients. SRe ofbleeding was lesser eue GV (LCGV)in 16 und fundnl GV (FG in 44 patients (isolated FO-'V-13 snd FGV + esophageal varices in 31). Endoscopic sdrotherapy (EST) was initially successful in controlling bleeding from LCGV in all 16 patients, 15 continued on chronic EST and one required ekctive mrgety. Of 44 patients with bleed from F, itj/tiai control could be achieved with balloon tamponade in 12 of25 (48%) patients, with EST in 2 of 15 patients (13/6) and with injection of glue in 3 of 4 patients (75%). Fwe ofthese 17 responders were continued on chronic EST, 2 patients wre lost to follow up and 10 patients were subjected to elective modified Sugiura's mrgety. Emergency urgety was performed in 27 patients (modified Sugiura's operation-25 and porto-caval shunt -2) alter other therapies had failed or in those who had rebled a/ter initial control Remits'. Operative mortality was nt for the elective group, 360/6 for emergency devmculaation and 50% for shunt surgety.Mortty ws 63% for crrhotics (all Childs score C), 25% for VO and 25% for NCPF. There were 2 late deaths due to liver cell failure in the cirrhotic patients. Twenty five patients are alive and free from rebleed. In conclmion, while LCGV can be successfully managed with EST, Fundal varices respond poorly to conservative therapy. Surgery in the form of gastroesophageal devacularbation with stpling i highly successful in controBing active bleeding as well as for long term prevention ofrebleed from fundal variees and should be the therapy of choice for these patients. detected preoperatively are thought to exist in resected specimens. This study was performed to evaluate the detectability of hepatic nodules on computed tomography during angiography (angio CT) compared other modalities including digital subtraction angiography (DSA), computed tomography with other methods (conventional CT), magnetic resonance imaging (MRI) and ultrasonography (US). Forty-seven patients with HCC underwent hepatic resection or liver biopsy after angio CT. In these 47 patients, seventy-three nodules were detected in resected specimens or in needle biopsy specimens. These nodules included 54 nodules of HCC, 3 of early hepatocellular carcinoma (eHCC), 5 of adenomatous hyperplasia (AH), 5 of hemangioma, and 6 of other benign lesions. The rate of detection of all hepatic nodules was 91.8% (67 nodllles out of 73; 67/73} with angio CT, 58.9% (43/73) with DSA, 74% (54/73) ith conventional CT, 72.5% (37/51) with MRI and 80.8% (59/73) with US. The rate of detection of eHCC and AH was 50% (4/8) with angio CT, 12.5% (i/8) with DSA and conventional CT, and 25% (2/8) with MRI and US. In conclusion, in this study angio CT could detect hepatic nodules more frequently than other imaging modalities. However, thirteen nodules in 9 cases detected preoperatively with anglo CT were not detected with intraoperative US. This suggests that it is important to follow these nodules carefully. One cause of high incidence of intrahepatic recurrence after curative operation for hepatocellular carcinoma (HCC) might be the multicentric occurrence (MC) of HCC different from the primary tumor. In this study we tried to clarify the characteristics of MC determined by histopathological and molecular biological analysis. <Patients and methods> We analyzed the recurrence pattern of 106 patients who underwent curative primary resection between October, 1983 and December, 1994 in our hospital. While the cumulative survival rates 2 and 5 years after the operation were 82 and 64%, respectively, the corresponding cumulative disease-free surviVal rates after the operation were 56 and 31%, respectively. <Results> Recurrence was seen in 45 patients. Clinicopathologic variables in the recurrence group were cx}mpared with those in the disease-ee group. The sex of the patients, HBs Ag, HCV-Ab, tumor staging, tumor size, extent of hepatic resection did not affect the incidence of recurrence, but recurrence occurred less frequently in the patients with noncirrhotic liver (p=0.004) and in the patients with multiple HCC (p=0.0008). Among 41 patients with intrahepatic recurrence, one to three nodules in the same lobe as the primary tumor were seen in 9, those in different lobe in 12, multiple nodules in 18 and marginal recurrences in 2. In 29 patients whose recurrent HCC were analyzed for donality, 14 showed MC of HCC and 10 metastatic HCC. Repeat resection was performed in 10 patients with the form of one to three nodular recurrence and in 8 the recurrences were MC. The survival rate after the re-operation was the same as that after the first resection. <Conclusion> It seems important to detect MC, especially in cirrhotic liver, in the early stage after the first operation and perform radi"cal treatments in order to prolong the survival. INTRODUCTION. The most used technique to control intraoperatory bleeding during hepatoresection is the "Pringle manouvre" (temporary clamping of the hepatic pedicle). In spite of advantages, reperfusion injury represents an important problem in case of prolonged clamping. Intermittent clamping has been recently proposed as a method to reduce organ injury. In this study, we compared survival, hepatic adenosine triphosphate (ATP) level and lipid peroxidation (LP) in rats exposed to intermittent vs continuous ischemia of the liver. MATERIALS AND METHODS. Liver ischemia was induced in male Wistar rats (weight range 200-250 g), by clam-ping the appropriate branches of the portal vein and hepatic artery of the left lateral and medial lobes of the liver with noncrushing microvascular clamps. The times of ischemia were 60, 90 and 120 min intermittently (periods of 30 min of ischemia followed by 10 min of reperfusion) or continuously. Partial ischemia was performed to avoid splanchnic congestion. At the end of the ischemia, the right lateral and caudate lobes of the liver were removed. Survival of the animals was assessed at days. ATP and LP lev.els were evaluated at the end of the ischemic time and hour after reperfusion by bioluminescence analysis and malondialdehyde (MDA) tissue concentration. RESULTS. Intermittent clamping markedly improved the animal survival at all the ischemic time considered when compared to the continuous group (60 min: 100% vs 25%; 90 min: 75% vs 0%; 120 min: 50% vs 0%). ATP and MDA concentrations were significantly reduced after 2 hours of intermittent or continuous ischemia; however, the decrease was markedly greater in liver exposed to continuous ischemia (0.66:-0.05 vs 0.910.07 pmol/mg; p<0.01). After reperfusion, ATP and MDA levels increased in both groups; however after hour of reperfusion, MDA concentration was significantly lower in liver exposed to intermittent ischemia (2.1:-0.6 vs 4.6:0.7; 13<0.05). CONCLUSIONS. Intermittent oxigen deprivation of the liver is followed by an higher rat survival, if compared to continuous ischemia.
Moreover, it reduces liver energy depletion caused by anoxia and decreases tissue lipid peroxidation that occurs during reoxygenation. These results suggest that intermittent rather than continuous "Pringle manouvre" may reduces liver injury during resections. A Prospective, double-blinded study was undertaken to assess the sensitivity of ultrasound in determining patency, stenosls or occlusion of transjugular intrahepatic portosystemic shunts on routine followup examination. Thirty-six evaluations were preformed on twenty-seven patients over a four month period.
The ultrasonographer and the interventional radiologist were blinded to each others findings. The data was collected and analyzed by the gastroenterologist.
Of the-thirty-six examinations preformed, ultrasound was inaccurate in assessing.the status of the shunt 42% of the time.
Our study contraindicates the current openion that ultrasound is highly sensitive in detecting early stenosis of intrahepatic portosystemic shunts. With the high restenosis rate seen with transjugular intrahepatic portosystemic shunts and the high morbidity associated with recurrent variceal bleed, accurate assessment of shunt patency and early detection and correction of shunt stenosis is paramount in the management of patients with portal hypertension and intrahepatic shunts. Dept of Surgery, Medicine, HEINZ KALK-Hospitai (HKH), D-97688 Bad Kissingen, Germany Endoscopic scierothezapy (ES) and recentiy Iigation (EL) has been estabiished as therapy of first choice in cirrhotics with bIeeding esophageai varices. However, eariy or.iater scIerotherapy faiiures (SF} occur in a frequency between 20 to 50%. In concurrence to ES and in spite of TIPS shunt procedures are stiii the best prophyiaxis for recurrent variceai hemorrhage. Shouid this modaiitiy not offered to SF with good iiver reserve? From Jan. I, 1983 untx Jan. I, 1995 922 pat. with bieeding esophageaI varices were admitted to HKH; 507 (55%) beIonged to the CHILD-PUGH (CP) ciassification A and B. 162 (32%) were SF; 72 of them were seiected for distai spienorenai shunt (DSRS) on the basis of the foiiowing criteria: iiver voIume between 1000 and 2500mi, portai perfusion index over 30%, exciusion'of activity and progression of iiver cirrhosis (LC) by biopsy. LC was mainiy of aicohoiic origin (51/75%). 48 were maIe and 24 femaie with a median age of 52. 3 (16-71) years. 38 were CP A and 34 B. In ten cases DSRS was technicaiiy to risky; a narrow-iumen mesocavai interposition shunt was performed. Thus, 62 pat. (6.7%, 12.2%) received DSRS. -Hospitai mortaiity was 4.8%. In two cases (3.2%) recurrence of variceai hemorrhage occured because of shunt thrombosis (I) and a shunt iumen of 7mm (I). VariceaI hemorrhage couid be stopped by emergency or eiective ES. AII pat. couid be foiiowed up to Juiy I, 1995; there was no case of. shunt encephaiopathy and 8 further deaths. Cumuiative survivai time according to KAPLAN-MEIER is 80% after five and 70% after ten years. Thus, eiective DSRS in a highiy seiected patients popuiation offers the current best decompressive method for recurrent variceai hemorrhage after SF. Histologic finding of the primary failure suggest that at least the final pathway of the process leads to ischemic-type injury and many observations document altered coagulation and fibrinolysis during the postoperative period of OLT.
PAI-1 is the main modulator of fibrinolysis, is a component of the acute phase response to injury and is synthesized by endothelial cells and hepatocytes. Aim of this work is to study .the effect of ischemia-riperfusion injury on PAI-1 mRNA synthesis in human liver. In 10 subjects undergoing partial hepatectomy for localized lesions the unaffected portion of the liver was biopsed before the clamp of the hepatic-vessels and at the end of the surgical procedure, after an average period of recovery from ischemia of 76+/-3 9 rain. Total RNA was isolated by the guanidine isothiocyanate method. The amount of transcripts encoding for PAI-1 and T-actin mRNAs was determined by the Norhtern technique, with 32p-labeled eDNA probes. After an average 30 min. of ischemia a four fold increase of PAl-1 mRNA (2p<0.01).is detectable in the hepatic tissue. The increase of PAl-1 mRNA is evident even after a very short time of ischemia (7 min.) and does not seem to be time dependent. The mRNA de novo synthesis seems also very rapidly induced, since its increase is detectable already after 20 rain. of recovery from the vascular occlusion. It seems also relevant the great individual variability of PAI-I synthesis both in basal (1 to 5 in an arbitrary scale) and in the post ischemia-riperfusion state (4 to 20). A decrease of fibrinolytic activity due to the ischemia-riperfusion induced PAl-1 overproduction may play a key role in the pathogenesis of both the primary failure and thrombotic complications.

COMPARISON OF DIFFERENT METHODS FOR THE TREATMENT OF NONPARASITIC LIVER CYSTS
A. Petri, 1. Makula*, S. Karesonyi Department of Surgery, Institute of Radiology, Albert Sz(C)nt-Gyfrgyi Medical University, Szeged, Hungary Nonparasitic liver cysts (NC) belong to the most common benign focal diseases. Usually they are accidentally detected by different radiological imagine methods when examining patients with atypical complaints. Only a small part of them cause characteristic complaints to the patients. Between I982 and 1995 eightytwo patients underwent operations because of NC of the liver, 18 of them were operated on by laparoseopie way. The were 16 male (mean age: 51.9 years) and 66 female (mean age: 54,6 years) patients. The most typical complaint was pain in the right hypoehondrial region (n 80). The mean size of the cysts were 57.4 mm in diameter (20 140 ram). Seventyfour were solitary and eight were multiple. In thirtyone eases we performed an enueleation, in sixty eases a fenestration and in one ease a punction of the cysts. A liver resection had to be carded out in two eases. The gall-bladder was removed simultaneously at eleven patients.
Except for the laparoseopie way the operations have been carried out in temporary occlusion of the hepatoduodenal ligament according to Pringle.
Postoperative morbidity rate was low in both groups. Fever could be observed in eight eases, leueoeytosis in one ease, jaundice in three eases, pleuritis and hydrothorax in one ease, suppuration of the wound in one ease and a sterile abdominal wall disruption also in one case. There was no mortality. Another twentysix patients with NC have been treated with ultrasound guided puntion of the liver cysts. There were four male and twentytwo female patients (mean age: 58.27 years). The cysts were solitary in twentythree cases and multiple in three other cases. The number of the cysts were thirtytwo. The cysts have been sclerotized with aetoxysclerol in three cases and by means of absolut alcohole in four cases. In one case an ultrasound guided drainage of the liver cyst was carded out. One case had to be operated on besause of recurrence of the cyst. Mortality could not be observed.
In the laparotomised group the everage time of the hospitalisation was 14.3 days, and among the patients who were operated on by laparoscopic way 7.0 days. In the third group the patients did not require any hospitalisation. Cell-matrix interaction is an essential parameter in the process of liver metastasis formation. This interaction is via the binding of extracellular matrix and cellular integrin and activation of downstream signals including the focal adhesion kinase (FAK) and paxillin. This study examined the role of hepatocyte growth factor/scatter factor (HGF/SF), a tumour invasion promoter, in tumour-matrix interaction.
Human colon cancer cell HT 115 was used. The basement mmbrane (matrigel) was used to determine cell-matrix interaction. Cell numbers after adhesion were determined by Hoescht 33258 assay. Tyrosine phosphorylation of focal adhesion kinase and paxillin was detected by immunoprecipitation and Western blotting. The presence ofHGF/SF in the assay system significantly increased cell-matrix binding and this was seen in a concentration dependent manner. Pretreatment of cell with HGF/SF and subsequently depletion of HGF/SF also increased the binding, suggesting an activated adhesion mechanism. HGF/SF increased the tyrosine phosphorylation of both focal adhesion kinase and paxillin shortly after treatment. This was seen together with the activation of HGF/SF receptor (cMET proto-oncogene).
We conclude that HGF/SF enhances tumour-matrix interaction by stimulation ofthe phosphorylation of focal adhesion kinase and paxillin. Satisfactory graft re-perfusion is fundamental to successful liver transplantation. Laser Doppler flowmetry is a technique which can measure hepatic blood flow 'continuously without altering the floff in the microcireulation. Here we report its first application to clinical liver transplantation.
Hepatic tissue microcirculation was measured intraoperatively in 22 liver grafts. Data was examined with respect to (a) effects of portal vein (PV) inflow, and (b) effect of hepatic artery (HA) inflow, both up to 30 min after revascularisation. Laser Doppler flowmeter (LDF) readings from the surface of the left lobe were validated against total liver .blood flow using an electromagnetic flowmeter (EMF) on PV after revascular.isation of PV flow.
There was a significant correlation (r 0.96; p < 0.001, n=8) between surface left lobe liver perfusion using the LDF and total liver blood flow measured by EMF. The LDF was reliable and robust under clinical conditions and provided reproducible measurements of perfusion with a coefficient of variation of approximately 4%. Re-perfusion of the transplanted liver with venous blood was accompanied by an immediate increase in liver blood flow perfusion. Over the subsequent 10-30 minutes there was no significant increase in flow and reperfusion of the graft with arterial blood did not increase liver blood flow peffusion. Hepatic tissue blood flow showed a negative correlation with cold isehaemic time (r -0.48, p < 0.025, n=22) on graft perfusion at the early phase of revascularisation of portal vein blood, but this correlation disappeared by the end of the operation (r -0.11, p 0.611).
One CCA patient(NED 4yrs) is alive. Two died of reccurence; 8/12 and 2 yrs, confirming poor prognosis of OLT in CCA reported by others. 2/17 HCC have reccurenee, 9/17 have diedmetastasis(l), reccurent Hepatitis B with no tumor-4(5/12 to 17/12 post OLT), other causes (4/17). 7/17 are alive NED (4/12 to 5yrs). 3. 14/42 had incidental HCC diagnosed postoperatively(7/14) or suspected preoperatively(7/14) in OLT for ESLD. Lesions were < 1.Ocm (1), lto eros (6), > 3.0cms(5);unifocal (8), and multifocal(4). 5/14 died with NED, and 9/14 are alive (5/12 to 6.4 yrs) with recurrence in 1. OLT has a role in the treatment of HCC. Careful preoperative staging is required,including exploratory laparotomy, backup recipient, and improved imaging techniques for tumors in cirrhoties. Study protocols are required to evaluate adjuvant therapy in HCC. Incidental tumors have better prognosis but pathological staging determines the outcome. The hepatic hemangioma is the most frequent benign tumor of the liver. It is generally an ordinary and incidental finding in imaging routine examinations being mainly a problem for oncologic patients. They are more frequently found in women in the dght lobe of the liver. Many authors have pointed out the differences among the imaging methods in order to establish the diagnoses of this lesion. The purpose of this paper is to introduce the US, CT, and MRI features of the hepatic hemangioma in a great number of patients (791)  We investigated the protective effect of triiodothyronine (T3) on a liver ischemia reperfusion model on rats. Portal and left lateral lobes of the liver clamped and 70% liver ischemia was made. Following the ischemia clamp was opened and reperfusion was established.
Study was carried out on T3 treatment and control groups. L triiodothyronine in a dose of 300 mcg/kg/day per os for 10 days was given to the treatment group and hyperthyroidism constituted. Each treatment and control groups has 7 subgroups: sham operation, ischemia only, 15 minutes, 2 hours, 24 hours, 48 hours and 7 days following ischemia reperfusion injury. Animals are sacrificed after obtaining blood and tissue samples, repeated reoperation is avoided.
Serum AST ALT LDH values and tissue lipid peroxidation were studied, and histopathologic examinations of the ischemic liver were done.
AST, ALT, LDH and lipid peroxidation of the ischemic tissue were significantly lower in the treatment group when compared controls. Histopathologic damage was also less severe in the treatment groups.
These findings suggests that L triiodothyronine pretreatment has protective effects on ischemia reperfusion injury of the liver. Bichat, 75018 Paris, France.
In case of fulminant hepatitis, auxiliary liver transplantion (ALT) is proposed to supply for the naiive liver (NL) until it regenerates. Technical factors of the NL regeneration are not yet known. The aim of this experimental study was to assess the NL regeneration according to the graft (G) weight and the location of portal anastomosis in ALT. Material and Methods: 24 syngenic ALT using an arterialized G were performed in a rat model after a 80% reduction of the NL and were allocated into 4 groups A 50% reduced-sized G anastomosed to the superior mesenteric vein (SMV) and NL vascularized by the pancreas, the spleen and the stomach (n=6), B full-sized G on SMV (n=6), C 50% reduced-sizexl G anastomosed to the portal vein (PV) and NL vascularized by the stomach (n=5), et D full-sized G on PV (n7). NL and G regeneration were assessed at day 2 or day 4 by in vivo incorporation of tritiated thymidine, then by weighting at day 30 after sacrifice.
Weight variation of the NL (AW (W30-W0)/W0) was calculated as well as the ratio NL/W NL weight/rat weight and L/W NL weight + G weight/rat weight at ALT and at sacrifice (respectively NL/W0, NL/W30, L/W0, et L/W30). Resllts: At day 0, the ratio L/W 0 was 2.88+0.25% (group A), 4.58+0.28% (group B), 2.68+0.15% (group C), and 4.73+0.16% (group D) respectively (normal value 3.88+0.24%). At day 2 and day 4, tritiated thymidine incorporation in NL was more important when G was anastomosed to PV but did not change according to the G size. At day 30, AW of the NL was more important in case of a full-sized G (groups B+D vs A+C) and G anastomosis to the PV (groups A+B vs C+D). The AP of NL was + 212+123% in group B and + 96+81% in group C (p=0.07, Mann-Whitney's U-test). The ratio NL/W30 was 2.32+0.68% in group B and 1.21+0.63% in group C (p=0.02). The ratio L/W30 was normalized in groups A (3.87+0.37%) and C (3.82+0.41%), and was still increased in groups B (4.65+0.60%) and D (4.56+0.63%). Conclusions: In this model of ALT 1) A better NL regeneration at day 30 was observed using a full-sized G anastomosed to the SMV; 2) This result is not fully determined by the regeneration occuring until day 4 3)  performed. In cases with rupture ofthe cyst in billim7 tract the explbration otthe billiary tract was performed, with the remora ofthe twin cyst and closing ofcystobiliary fistula with omentum. The secondary cysts in the peritoneal cavity were completely removed. There were no intra or postoperative complications in any ofthe cases mentioned.
Postoperatily all the eases were examined by ultrasound after 3fi and 12 months.
There was no complication evident in controls performed by ultrasound examination in the cases, whatsoever.
We have specifically examined the remaining cavity.in the liver at..er the rein.oval ofthe cysts and came to a conslusion that in cases where omentoplicasion the cavity within three months while on cases without the omentoplication it took over six months. As an outcome ofour research we consider that the partial pericystectomia with omentoplication gives satisfactory results with a low postoperative morbidity. We sugest ultrasound controls in evaluation ofthe remaining cavity in the liver and appearance ofeventual complications. Hepatosplenic Schistosomiasis (HS) is the leading cause of Portal Hypertension (PH) in Brazil. Variceal bleeding is the major complication of the disease and carries high mortality. Although the studies comparing efficacy of Endoscopic Sclerotherapy (ES) and Variceal Ligation (EVL) in the trealanent of esophageal varices in cirrhotic patients showed similar efficacy and lower incidence of complication with EVL, there are no studies using only patients with pre sinusoidal PH. We performed a prospective, double blind, randomized study to compare efficacy, complications and recurrence of esophageal varices treated by ES or EVL in patients with HS and Esophageal Varices (EV). Thirty six patients were included (18 in each group). They were treated with repeated endoscopy sessions until eradication of EV. Six months after eradication, endoscopy was performed to look for recurrence and local complications. Less, sessions were necessary to eradicate VE in EVL group (2,9 vs. 3,9, p=0,04). Efficacy was similar in both groups (100% in EVL and 88,8% in ES). Treatment failure occurred in 2 (12,2%) patients in the ES group, who developed uncontrollable gastric variceal bleeding and were referred to surgery. There were not significant differences in the incidence of complications. However, 50% of the patients in the ES group needed additional sedation during or inunediately after the procedures and it did not happen in any patient ofthe other group.
There was only one case of recurrence in ES group: We concluded that both endoscopic methods of treatment of EV are efficient, with similar incidence of complications and recurrence in this study with schistosomotic patients. Eradication occurred faster and with less pain in EVL group. In unselected clrrhotlcs the 2 yr. FVB rate is 25/, and 40"1, of the episodes occur within the first 6 mo (Hepatology 1994;20:66). Pts with HRV should have at least a 10"/. higher FVB rate respect the baseline risk. To minimise heterogeneity we grouped RCTs with similar: a)FVB rate in the placebo groups; b)length of f-u. RCTs whose placebo groups had a 24 mo FVB rate >35% were defined as RC,Ts of HRV. We Included RCTs: a)with an average f-u of 24 mo; b)with a yr. FVB rate >35% in the placebo group; c)whose results were published in life lables or data communicated by the Authors when requested. Eight of 9 RCTs met these criteria. A total number of 749 pts was included :347 treated and 372 controls (CTR). The FVB rate in the pooled control groups was 48.4% (range 34%-80*/.), 42*/0 of the episodes occurred within 6 mo. The sample size of 749 pts was required to significantly decrease by >10"/, the risk of a FVB (a=5%; 1-1]=80%). The Messod's method for censored data was used (CompuL Progr.Meth.Biom.1993;40:261). The table show the results: among treated pts them was a trend toward a reduction of the FVB rate dudng the first 6 mo, then the FVB rate significantly decreased within and after the first six months.  Purpose: To assess SPIO-(superparamagnetic iron oxide, a reticuloendothelial system-RES-contrast agent) MR imaging (MRI) in detecting and characterizing liver lesions before surgery. Methods: 50 patients with known or suspected liver lesions were examined with SPIO-enhanced MRI. Qualitative and quantitative analyses and correlation with surgical and pathological findings were performed.
Results: SPIO enhanced T2-weighted MRI images demonstrated marked decrease in signal intensity of liver (96%). This sequence provided improved liver lesion detection and diagnostic confidence compared to plain MR (81% and 86% resp.). Cases of FNH (n=16) were all depicted and characterized in SPIO-enhanced MRI and the central scar was better seen in post contrast images. Hemangiomas (n=12) revealed different signal behavior than FNH and all were correct identified in enhanced MRI. Adenomas (n=4) were found to have different RES activity. Hypervascular metastases (n=18), because of lack of RES cells were all depicted in SPIO-MRI.
Conclusion: SPIO-enhanced MRI has the potential to be the non-invasive method of choice in the preoperative evaluation of focal liver lesions due to its efficacy to detect and potentially characterize surgical from nonsurgical liver lesions. This study investigated the value of intense follow-up after curative surgery of cancer in the colon or rectum. 107 patients were randomized to no follow-up (control group, n=54) or intense follow-up (follow-up group, n=53) after surgery and early postoperative colonoscopy. Patients in the follow-up group were followed at frequent intervals with clinical examination, rigid proctosigmoidoscopy, colonoscopy, computed tomography of the pelvis (in patients operated with abdominoperineal resection), pulmonary x-ray, liver function tests, and determinations of carcinoembryonic antigen (CEA) and faecal hemoglobin. Follow-up ranged from 5.5 to 8.8 years after primary surgery. Tumour recurred in 18 patients (33 %) in the control group and in 17 patients (32 %) in the follow-ul group. Reresection with curative intent was performed in three patients in the control group and in five patients (four of whom were asymptomatic) in the follow-up group. In the follow-up group two asymptomatic patients with elevated CEA levels were disease-free three and five and a half years after reresection and were the only patients apparently cured by reresection. No patient underwent surgery for metastatic disease in the liver or lungs. Symptomatic metachronous carcinoma was detected in one patient (control group) after three years. Five-year survival rate was 67 % in the control group and 75 % in the follow-up group (p> 0.05); the corresponding cancer-specific survival rates were 71% and 78 %, respectively. It is concluded that intense follow-up after resection of colorectal cancer did not prolong survival in this study.  (39 %), gave no conelusive information 28 patients (13 %) and incorrect information in 24 patients (12 %). Cytological diagnosis was. valuable for planning further investigation and treatment in 58 patients (28 %) but was of no or doubtful importance in 139 patients (67 %). In 11 patients (5 %) it led to repeat aspiration biopies without definite diagnosis, delay of treatment or unnecessary investigations or operations. Implantation metastasis was recorded in 7 patients (3 %). It is coneluded that fine needle biopsy is of limited value in the management of liver tumours and that the benefits and risks involved do not justify its use in candidates for curative resection. Liver transplantation LT has been widely accepted in of irreseetable HCC because of the severity of the underlying cirrhosis, the location of the tumour within the liver hilus, or the extent of the liver involvement, but the benefit of pretransplantation tumor treatment with ehemoembolization (TAE and/or percutaneous alcohol injection PEI is still under investigation. We report our results after preoperative treatment with TAE and PEI, regarding the clinical outcome and the recurrence rate after LT for HCC. Material and methods: 17 patients underwent total hepatectomy and orthotopic liver transplantation for irresectable HCC on cirrhosis. There were 15 males and 2 females, whose ages ranged from 39 to 63 years (median age +/-sd 54 +/-5.7 ). All the patients were eirrhotics, with prevalence of HCV infection in all but two, one affected by alcoholic cirrhosis and the other by cholestatic liver disease. The tumour nodules were solitary in 9 cases, multiple in 8, with mean size of 28 +/-15 mm, ranging from 10 to 70 mm. TAE and PEI were performed in all but two patients, and repeated every 2-3 months (TAE until time and every 2-3 weeks PEI until 6 time ). Results: the serum a-fetoprotein level was useful for evaluating the therapeutic effect, showing important decrease in patients with level higher than 100 ng/ml. Mild to severe local pain occurred immediately after PEI and high fever >38 *C developed after TAE in half of the patients without any complications. Extensive tumour necrosis 90% was seen in 9 patients, but tumour could not be found in despite positive preoperative liver biopsy (in two) and most suggestive preoperative radiographic imaging in the other one. We lost four patients: two because of tumour recurrence at 8 and 37 months after LT, one because fulminant recurrence of HBV infection after retransplantation and the other one because of cerebral bleeding, respectively at 16 and months after LT. The median follow-up is 20 +/-15 months range 6-48 ), with total recurrence rate of 11%. The overall actuarial survival and tumour free survival is 76% at four years. In a previous study, we have shown a beneficial effect of intermittent warm (20C) rinsing on microcirculatory injury in cold stored rat livers. Aim of this study was to evaluate the temperature effect (4C vs 20C) of a single flush preceding reperfusion in a rat liver preservation model. Preservation injury was assessed by measuring AST, LDH, hyaluronic acid (HA) and purine nucleoside phosphorylase (PNP). HA is metabolized by the sinusoidal endothelial cells (SEC) of the liver; uptake or release therefore reflects SEC function or death, resp. Release of PNP is associated with SEC death. Methods.
.Rat livers were washed out in situ via the portal vein with UW solution (4C) and after hepatectomy, were stored at 4C. Immediately after hepatectomy (tO, control livers) and after 8hr, 16hr and 24 hr cold ischemic time (CIT), resp, the livers were reperfused for 90 mins via the portal vein with oxygenated Krebs Henseleit buffer (37C, without albumin) containing HA (34-55 pgr/L). At the end of CIT, prior to reperfusion, the livers were flushed with 10 ml UW solution, either at 4C or 20C. During reperfusion, bile production was monitored and HA (radiometric assay), PNP (fluorescence spectroscopy), LDH and AST were measured in the reperfusion medium. Results. Mean bile production was highest in control livers (1.88

Madrid. Spain
Liver transplantation is now widely accepted as a worthwhile treatment for selected patients with primary liver cancer. At Hospital Ramtn y Cajal and Clfnica Puerta de Hierro, we performed 340 liver transplantation between March 1986 and December 1995. 31 patients underwent transplantation for primary malignant disease. 23 patients had hepatocellular carcinoma, including of the fibrolamellar type, 7 were diagnosed of cholangiocarcinoma and of hepatoblastoma. Of the 31 patients, 6 were female and 25 were males. The ages of the patients ranged from 12 to 63 years. 4 of whom were less than 20 years old. patients (9%) died in the first 3 months from complication of transplantation. Of the remaining patients who survived more than months, 4 (14%) developed recurrent disease, diagnosed 7 months to 9 months after liver replacement. The only patient treated for fibrolamellar hepatoma is alive without tumor at more than 8 years after transplantation.
CARCINOMA HEPATOCELLULAR 20 of 22 patients with carcinoma hepatocellular had associated underlying liver disease. These patients were stratified according to the TNM classification: Stage I: 3 patients, Stage II: 10 patients, Stage III: 6 patients and Stage IVa: 3 patients. Actuarial survivals for all 22 patients with carcinoma hepatocellular at year, 2 years and 3 years were 57%, 50% and 41%, respectively. Stage and II (5 years actuarial survival: 56 %) and incidental hepatoma (4 years actuarial survival: 77 %) have a good prognosis after transplantation.. CHOLANGIOCARCINOMA of the 7 patients who underwent liver transplantation for cholangiocarcinoma had nodal involvement and/or perineural and vascular invasion with infiltration of tissue around biliary tract. The 7 year actuarial survival was 66 %. 2 patients survived tumor free for more than seven years. In patient, recurrence occurred in the trasplanted liver, seven months after transplantation.
We conclude that liver replacement for malignant hepatic neoplasms should be considered for a selected group of patients with unresectable lessions and without evidence of extrahepatic disease. age, sex, Iocalisation and stage of primary tumor, size, number and Iocalisation of metastases and extent and radicality of the liver resection.
The results showed, that a significant influence regarding the long term survival rate could only be archieved in patients, in which a R0-resection with a histollogically proven tumor free margin of more than cm was performed (p= 0,0087). For patients in this group (n=46) a 5-year-survival rate of 40% was observed: In patients with tumor free resection margins less than cm (n=38) the 5-year-survival rate was only 10% and comparible to,patients with Rl-resections (n=7). None of the patients with R2-resections (n=15) survived 5 years. Other significant prognostic factors were the number of metastases and their size. Patients with more than 3 metastases or metastases > 10 cm had a significant shorter survival (p= 0,0467/p= 0,0472) than patients with 1-3 metastases or metastases < 10 cm.
Our data support the importance of a sufficient safty margin in liver resections for colorectal metastases. Gastroenterologia, * Clinica Chirurgica II, Univ Bologna, ITALY Introdoction: liver cirrhosis is characterized by increased intestinal blood flow with reduced arterial resistances. On the contrary hepatic, splenic and renal artery resistances are reported to be increased. Aim: to assess the effect of orthotopic liver transplantation (OLT) on splanchnic hemodynamic changes induced by liver cirrhosis. Materials nd methods: 20 patients submitted to OLT for end stage liver cirrhosis (m=14; with ascites=12; 14 HCV+ and/or HBV, 3 alcohol, PSC, PBC, Byler's disease), when first in the waiting list, underwent .an abdominal Doppler US exam to assess resistance indexes (RI) (an indirect measurement of resistance) in the superior mesenteric, intrahepatic, intrasplenic and renal interlobular arteries and portal blood flow velocity in the right portal branch. The same measurements were repeated at six months in 17 patients (none with ascites, two suffering from rejection, two retransplantated). The 3 other patients had died. Student t-test for paired data was used for statistical analysis. Results: at six months portal blood flow velocity increased from 14.4 to 28.6 crn/sec (p<0.01) and splenic artery RI decreased from 0.601 to 0.525 (p<0.025). RI changed from 0.678 to 0.644 in the intrahepatic artery, from 0.651 to 0.618 in the renal artery and from 0.810 to 0.841 in the superior mesenteric artery: these changes didn't reach, however statistical significance. C0nclu.si0ns: OLT reverts the splenic artery RI increase. Persistance of a mild overflow in the splanchnic vascular bed after OLT is suggested by the marked increase of portal flow velocity and the incomplete reversal of superior mesenteric artery dilation (normal RI values in our experience=0.88). The contribution of the persistance of collateral pathways to this event should be investigated. Morevoer the high portal flow, probably stimulating the arterial buffer response with arterial costriction, could hide significant reduction of hepatic artery RI. Renal artery RI tends to decrease after OLT, probably not reaching statistical significance, only because in this series of patients the preOLT levels were already in the normal range (<0.70). Objective: We analyze, using a cerulein-induced acute pancreatitis model, the role of the activated complement system (ACS) in PMNs priming and lung sequestration, event related to the acute lung injury and ARDS occurring in this disease. Participation of the ACS at an early stage is evaluated using a specific receptor of the complement system activated components. Material and metlods: Supramaximal doses of the secretagogue cerulein were infused for 4h to produce a mild edematous pancreatitis in Lewis rats. Soluble complement receptor, sCR! (BRL55730-Smithklein Beecham Pharmaceuticals), was used to block the complement cascade. Rat lungs were resected at the end of the experiment for measurement of myeloperoxidase (MPO), a marker of PMN accumulation. Mac-1 expression on surface of leukocytes ("in vitro" and "in vivo' flowcytometric studies) was used to access activation status. Bronchoalveolar lavage (BAL) and wet:dry weight ratio indicated the degree of lung injury, Histological studies were performed in HE-stained specimens. R..sults: High MPO levels were found in rat lungs 4h after induction of pancreatitis, indicating a high concentration of PMNs at this time (OD460=1.54+/-0.13, n=8). This data was confirmed by histological examination. Treatment with sCRI before pancreatitis induction significantly reduced the pulmonary concentrations of PMN (OD460=1.00+/-0.16, n=8). Also Macupregulation (activated pattern) was observed in circulating PMNs of rats with pancreatitis and in neutrophils incubated with serum of pancreatitis group rats as early as 2h after induction. This could be reversed by addition of sCR-1. During the time observed it could be seen no evidence of lung injury as shown by no change in wet:dry weight ratio and BAL fluid microscopy..Conclu sion: l)Cornplement system is responsible for neutrophils priming and consequent lung sequestratio erly in the course of pancreatitis in this model, and this could be effici,zntly reversed by the use of sCR-1.2)No lung injury could be demnstrated in the period observed, suggesting thai the function enhancement of the primed neutrophils occurs at a later stage probably under int.uence of another inflammatory mediator. Alterations in pancreatic endothelial barrier integrity results in leakage el blood components, tissue edema, hemorrhage and necrosis. Over-activation of macrophages might be one ofthe responsible factors initiating pancreatic endothelial barrier compromise. In the first experiment, three kinds of macrophage stimulators (zymosan, concanavalin A (CCV) and thioglycolate medium (TM)), were chosen to evaluate the influence on pancreatic endothelial barrier integrity 24 hours after intraperitoneal challenge with 0.25 and 0.50 mg of the various substances per g bodyweight (BW). Zymosan induced a significant increase in pancreatic tissue water content, interstitial fluid volume and extravascular protein volume and a decrease in intravascular plasma volume as compared to controls. The administration of TM and CCV had none or minor effects on the endothelial barrier and induced an increase in reticuloendothelial system (RES) function, while zymosan resulted in a compromise of RES. In the second experiment, the dynamics of pancreatic endothelial alterations induced by various doses of zymosan by determining endothelial permeability to 24 hours after challenge. Pancreatic endothelial injury was evident from one hour increasing by time. In the third experiment, possible mechanisms by which zymosan induced endothelial injury was investigated. Pretreatment with dimethyl sulphoxide (a scavenger), indomethacin (a cyclooxygenase inhibitor), and verapamil (a calcium channel blocker) partly prevented the increase in endothelial barrier permeability. Pretreatment with N-acetyl-Lcysteine, a scavenger of multiple oxygen free radicals completely protected from endothelial injury. Pretreatment with allopurinol (xanthine oxidase inhibitor), L-NNA and L-NAME (nitric oxide inhibitors), and pargyline (monoamine oxidase inhibitor) had no significant effects. Thus, zymosan resulted in a compromised pancreatic endothelial barrier. Over-activation of macrophages might play an important role in the etiology of pancreatic dysfunction. Multiple factors including a variety of oxygen free radicals are probably involved. We have rewieved our experience with pancreatic resection for cxocrinc pancreatic cancer in order to evaluate short-and Ions-term survival. A number of variables were evaluated to identify factors predictive of longterm survival. From January 1982 to ber 1994 we performed 92 pancreatic resections for adenocarcinoma of cxocrine pancreas. Sixty-one were males and 31 females, mean age was 62+11 years (range:34-82). The operative procedures consisted of 66 pancreaticodumics (PD), 7 distal pancreatectomies and 19 total pancreatcctomies. Fortythree of the PD included a distal 8astrectomy and 23 were pylorus preserving; patients had a pancrcatogastrostomy. Twenty-two patients (23.9%) had associated vascular resoctions. Survival was analyzed by the method of Kaplan-Meier. Differences in survival among variables were compare with the log-rank test. Operative mortality rate (60 days) was 6.5% (6 patients). Major morbidity related to operative procedure was seen in 18.4% of patients(17 patients). Actuarial survival rate at 1, 5 and 10 years was 51.7%, 10.5% and 5.6% respectively. The factor significantly influencing a poor prognosis were: neoplastic invasionof preaortic lamina (j0.006) and metastatic lymph node involvemm (pffi0.002). Vascular invasion, age >70 years and perioperative blood transfusions were not associated with a worse prognosis. Open laparotomy has traditionally been required to stage hepatobiliary and pancreatic (HBP) cancers accurately. For unresectable patients, costs and morbidity have been high. Today, laparoscopy alone or combined with laparoscopic ultrasonography (LUS) is being examined for its value in defining the extent of malignancy. We have analyzed our routine implementation of this new staging technique in our HBP Center. METHODS: Staging laparoscopy (SL) with LUS was performed in 50 consecutive patients with HBP malignancies. All patients were considered to have resectable tumors as determined by traditional preoperative staging modalities. Primary tumors were located in the liver (n=7), biliary tract (n=ll), or pancreas (n=32). Preoperative staging studies included computed tomography (96%) alone or combined with antegrade/ retrograde cholangiography (72%), arterioportography (22%) and magnetic resonance imaging (14%). An average of 2.7 preoperative studies per patient were required to initially determine resectability. SL-LUS was performed to detect occult hepatic, lymphatic or peritoneal metastases, and to detect local tumor invasion rendering the tumor unresectable. SL-LUS was performed under the same anesthetic as the subsequent laparotomy (70%) or as a separate staging procedure (30%). RESULTS: SL-LUS predicted a resectable tumor in 28 patients (56%). At laparotomy, 26 of 28 were actually resectable indicating a false-negative rate of 4%. SL-LUS indicated unresectability in 22 patients (44%). SL alone demonstrated previously unrecognized occult metastases in 11 patients (22%). For an additional 11 patients (22%) in whom SL alone was negative, LUS determined unresectability as a result of vascular invasion (n=5), lymph node metastases (n=5), or intraparenchymal hepatic tumor (n=l). All cases of unresectability due to vascular invasion were validated by laparotomy. 5 of 6 lymph node or hepatic metastases were proven histologically by LUS-guided needle biopsy rather than laparotomy. CONCLUSIONS: SL with LUS optimizes patient selection for curative resection of HBP malignancies. Unnecessary laparotomy can be safely avoided in unresectable patients reducing costs and morbidity. Malignant ampullomas can be curatively treated by pancreaticoduodenectomy (PD). A local excision can be proposed in case of benign ampulloma. The aim of this study was to evaluate the value of side-viewing duodenoscopy (SVD) with biopsies, endoscopic sphincterotomy (ES), and endoscopic ultrasonography (EUS) for preoperative diagnosis of benign and malignant ampullomas.

ORAL PRESENTATIONS
M..aterial et methods From October 1989 to September 1995, 26 patients with ampulloma were explored preoperatively by SVD including ES in 9 cases and EUS. The papilla of Vater was always explored at SVD and tbrceps biopsies were performed in all patients except one with a typical malignant tumour. EUS evaluated the T stage of the TNM classification in all patients except six because of a previous ES (n=5) and difficulty to localize the papilla (n=l). The N stage of the TNM classification was evaluated by EUS in all cases. A curative resection was always performed 2 local excisions of the ampulla (one adenoma with low-grade dysplasia and one xanthoma), and 24 PD for 20 carcinomas (including 9 (45%) N1) and 4 benign lesions. Results At SVD, papilla was. both ulcerated and protruding into the duodenal lumen in 10 cases, prominent and smooth in 15 cases, and normal in one case. Histologic examination of the 25 biopsies revealed malignancy in 1.0 cases (always confirmed by pathological examination of the resected specimen) and a benign lesion in 15 cases (resected specimen 6 benign lesions and 9 carcinomas); among the 9 biopsies performed after ES, two only revealed carcinoma (resected specimen 4 benign lesions and carcinomas), with an accuracy rate of 64% globally and 66% after ES. At EUS, 13 lesions were presumed limited to ampulla (benign lesion or stages T in situ and T1) but 4 out of these 13 were classified T2 or more histologically; among the 7 tumours classified T2 or more by EUS, one was histologically limited to ampulla. The accuracy rate of EUS for the T stage was 75%. At EUS, lymph nodes were presumed benign (NO) in 21 cases (histology NO in 15 cases and NI in 6 cases) and metastatic (N1) in cases (histology NO in 2 cases and N1 in cases), with an accuracy rate of 69%. All endoscopic explorations were compatible with benign lesion in 11 patients histologically, 6 patient out of these 11 had a carcinoma including one with a T1N1 and two with a T2N0 tumour.
Conclusions In patients with ampulloma tumour, SVD with biopsies even after ES, and EUS are not accurate enough to make sure preoperatively that the tumour is benign. Therefore a local resection cannot be safely indicated. Tissue factor (TF), the transmembrane cellular receptor which is the physiological initiator .of blood coagulation is not expressed in normal pancreatic tissue, but in the adenocarcinoma of the pancreas its expression increases with progressive dedifferentiation. We have studied the effect of TF expression on in vitro invasion by the human pancreatic carcinoma, cell line MIA PaCa-2. The full length TF gene (1360 base pairs) was cloned into the plasmid DNA vector pcDNA3 in sense and antisense orientations. Diagnostic DNA digests confirmed the correct orientation of the cloned genes. These vectors were used to transfect the MIA PaCa-2 cell line in vitro using the lipofectin method. Expression of TF sense gene resulted in a five fold increase in cell surface procoagulant activity frooa 7.8 thromboplastin units per 10 cells for wild type to 40 units (p=0.001). There was a small reduction in procoagulant activity for the TF/antisense clones to 7.6 units per 10 cells. The total antigenic content of TF gene product (sense or antisense) markedly increased from 160 pg/mg for wild type to 7733 pg/mg for sense and 856 pg/mg for antisense, h vitro invasion was assessed in a standard matrigel assay by counting the number of cells per high powered field, after haematoxylin staining. Pseudocysts and abscesses of pancreas in abdominal surgery, particularly at the peak of the acute inflammation, are one of the most often and serious complications ofpancreatitis.
From 1986 to 1995 years 128 patients with pancreatic pseudocysts were hospitalised in oar clinic. Traditional surgical treatment was performed in 17 patients with calculous pancreariris complicated by formarion ofpancreatic pseudocysts and in 3 patients with cystadenoma.
The other patients were treated with ultrasound-guided percutaneous punction and drainage ofpancrearic pseudocysts, including 13 cases of internal percutaneous cystogastro-and cystoduodenostomy. Simultaneously a complex set of conservative therapy including high doses of sandostatin (600-800 mg/day) was carried out for preventative purposes (3 days) or treatment of pancreatiris. Percutaneous fulfdling of cystogastroanastomosis was performed in 10 patients with pseudocysts of pancreas body and tail by means of the Hancke percutaneous pancreatic cyst drainage set/"COOK", Denmark/. Obtained material was sent for cytological, biochemical and bacteriologic investigations.
For the fwst time in our clinic we performed ultrasoundand endoscope.guided percutaneous formation of internal cystoduodenoanastomosis using, catheter of our own modification in patients with pseudocysts of pancreatic head. Both patients had pancreonecrosis and acute destructive pancreariris complicated by obstructivejaundice. On the 3-rd 4-th day after cystoduodenostomy the level of direct bilirubin decreased to normal and on the 7-th 8-th day the patients were discharged.
In one prient during the earliest stage of our work the at,en to form cystogastroanastomosis was failed, which demanded sursical treatment. Long-term results (14-46 months) didn't show disease recurrence. Therefore, we think that percutaneous cystogastro-and cystoduodenostomy is less traumatic and highly effective and should be the method of choice in treatment of such patients. Pancreatic pseudocysts have been successfully treated by endoscopic drainage (cystogastrostomy or duodenostomy, and transpapillary drainage). We report our experience with endoscopic therapy of pancreatic pseudocysts.'From July/94 to August/95, 16 patients with pancreatic pseudocysts were referred to ERCP because of persistent pain and/or jaundice. In 4/16 (25%) endoscopic therapy was not performed because we were not able to place a guide wire beyond a pancreatic stenosis and there was no indentation of gastric or duodenal wall. In the remaining 12 patients (9/3 mal, e/fernale ratio), mean age 38.2 years (range 24 64 years), 15 pseudocysts were treated with cystogastrost0my (5), cystoduodenostomy (2) and transpapillary drainage (8). Etiology was: alcoholic chronic pancreatitis (8), blunt abdominal trauma (2), surgical trauma (2). Pseudocysts mean size was 7.83 cm (range 3,5 18 cm) and were located in the head (4), body (8) and tail (3).
Complications were present in 7/16 patients: 2 early stent occlusion, fever, pneumoperitoneum, bleeding, proximal migration and perforation. Except for the perforation that required surgery, all complications were minor and medically and/or endoscopically managed. There were no deaths. Mean follow up was 150 days (range 15 360 days) and mean stent period was 134 days (range 30 210 days). Clinical improvement was noted in 11/12 (91%): all but one pseudocyst resolved, 8 patients are asymptomatic and 3 are taking small doses of analgesics. In the only patient that persisted, with pain, stent was removed and she was sent to surgery because of chronic pancreatitis.

F140 F141
The expression of growth associated protein 43  We used GAP-43 as a marker of. neuronal plasticity in chronic pancreatitis and correlated histological findings with clinical data. Methods. The innervation ofthe normal pancreas from 14 organ donors was determined immunohistochemically with the panneuronal marker protein-gene product 9.5 (PGP 9.5) and the expression of GAP-43 was also characterized. The findings were compared with results obtained from 29 patients with chronic pancreatitis. The density of PGP 9.5 and GAP-43 was quantified by image analysis. In addition, patients responded .to a symptom questionnaire for the evaluation of clinical findings. Results. In chronic pancreatitis a marked increase of PGP 9.5 and GAP-43 immunostaining was observed: digitized morphometry revealed that 70% of PGP-9.5-immunoreactive nerves were also GAP-43-immunoreactive whereas in the controls the relative area of GAP-43 was less than 3% (p<0.01). GAP-43-immunoreactivity significantly correlated with the intensity of pain (p<0.01) and duration of disease (p<0.05). Conclusions. GAP-43 re-expression at high levels both in nerves and intrinsic neurons indicates an axonal sprouting mechanism in chronic panereatitis; the correlation between clinical symptoms and GAP-43-immunostaining suggest a link between plasticity of peripheral nervous system and pain generation in chronic pancreatitis. From March 1990 we adopted a combined multirnodal treatment of resectable pancreatic head carcinomas including subtotal (or total, when needed) pancreatectomy, intraoperative irradiation (IORT)and external beam radiation therapy (EBRT). Twenty-two patients were treated following the above mentiorted protocol: 20 subtotal and 2 total panereatectornies were performed. IORT was delivered to the tumor bed (including celiac axis, portal vein and origin of superior mesenteric vessels) with a dose of 10 G-y. Mean time spent to cma3' out IORT (patient's transfer to the bunker, irradiation and return to the operating room) was 54 minutes. Postoperative major complications were observed in 5 patients (23 %) and mortality was 9 % (2 cases). Postoperative EBRT of the tumoral and lymphatic bed (3 beams or box technique 50 Cry) was performed in 16 patients. In 6 cases (27 %) postoperative EBRT was not started because of postoperative death (2 cases) or unsuitable clinical status. In the more recent period of our experience, 12 patients underwent preoperative "flash" radiotherapy of the liver and the pancreas (opposite beam technique 5 Gy) to treat possible liver micrometastases and to reduce tumor spread during surgery. Median survival was 11.8 months. Statistical analysis showed a significantly better survival in women (19.9 months vs. 9 months: p<0.04) and in patients undergoing preoperative "flash" EBRT (25.6 months vs.

ROLE OF ADJUVANT RADIOTHERAPY IN TREATMENT
9.5 months: p<0.03).
Our experience suggests that combination of EBRT and IORT with surgical resection of pancreatic cancer is safe, tolerable and offers a good local control. High frequency of liver metastases suggests the need to test new therapeutic combinations: in this respect, neoadjuvant radiochemotherapy seems to be the most promising. and eonfLrmed in all eases by both intraoperative and bacteriological findings. Conservative treatment based fluids and electrolytes replacement, cardiac and emodynamie monitoring, wide spectrum antibiotics (in ease of proven sepsis) and total parenteral nutrition; when necessary patients were traaferred in ICU for supportive treatment of renal and respiratory failure. The indications for surgical treatment were: the development of secondary infection, acute abdomen, MOF syndrome and eardioeireolatory shock. The surgical procedure consisted of abdominal entrance by bilateral subcostal incision, exposure of the pancreas by dividing the gastroeolie omentum, debridment and neerosectomy; in the presence of pancreatic abscess, the purulent 17.5 %: 8% (n.5) for sterile necrosis and 31.1% (n.14) for secondary infections (38.7% for infected necrosis and 14% for pancreatic abscess). DISCUSSION. Controversy still surrounds the optimal management of ANP particularly regarding the indications, the timing and the methods of surgical intervention. For sterile necrosis in the absence of severe systemic eomplieatious conservative treatment seems to be justified beth for the natural history of the illness (high incidence of resolution) and the poor results of early surgery. In presence of secondary infection surgical treatment is mandatory. Puestow procedure, 20 distal pancreatectomy without anastomosis, 5 total pancreatectomy). In two series were found no significant differences as regards sex, age, etiology, calcifications, time of appearance and severity of symptoms; main pancreatic duct was significantly more enlarged in derivated than resected patients (p<0.0001).
Length of operation, units of transfused blood, postoperative mortality (2 versus 5) and morbidity, hospital staying time were significantly lesser in derivated group (p<0.0001).
A similar number of patients in two series (77.2% vs. 71.9%) were painfree after a 6.1 years median follow-up. Given comparable results in reducing pain, we suggest that derivative surgery has however to be preferred owing to the less biological load; pancreatic resection instead has to be performed when pancreatic duct is not enlarged and in case of complications related to pseudocysts and suspect of pancreatic cancer. The biological behaviour and slow evolution of endocrine pancreatic tumors, even in presence of liver metastases, have prompted to most radical surgical approach and to research of any effective adjuvant treatment. Since 1991, hepatic arterial chemoembolization (TACE) has been proposed in the treatment of malignant endocrine pancreatic tumors, on the experience of primary hepatic malignancy.
From 1985 to June 1995, we observed 63 patients (25 males, 38 females; mean age 53.8 years, range 17-78 ys) suffering from histologically proven endocrine turuour of the pancreas, in 28 cases (44.5%) of functioning type, non functioning in 35 (55.5%). In 23 patients (36.5%) the turnouts were malignant with presence at diagnosis. (17 pts) or onset during follow-up (6 pts) of liver metastases. Of 17 patients (74%) presenting synchronous metastases, one patient underwent radical resection with the removal of the only metastasis detected; 8 patients were not submitted to radical surgery because of either locally advanced disease or not available TACE.
Eight patients (3 ruales, 5 females; mean age 43.8 years, range 49-67) underwent one or more treatment of TACE prior and/or after the palliative resection of pancreatic malignancy. After cannulation of proper hepatic artery, according to Seldinger technique, whole liver parenchyma was microembolized with solution of Lipioidol FU (10 ml) and dacarbazine (Deticene) 500 mg; afferent artery was then embolized with fragments of Spongostan or contour ruicrospheres. The response to the treatment was assessed by CT-scan within 30 days. TACE was repeated, if possible, within 90 days and in case of relapsing disease and the patients were followed up by CT-scans every 3 months. Of the 8 patients undergoing TACE, 2 died respectively after 11 and 18 months from surgery, while 6 patients are still alive after mean follow-up of 26.3 months (range 12-55), 2 with stable and 4 with progressive disease. The response to the first course of TACE was positive in 5 of these cases, with 14 months mean time of stabilization. In one recently treated patient the response cannot yet be assessed. The median actuarial survival of these 8 patients is more than 55 months.
These results suggest that TACE, combined with surgical resection of pancreatic turuor, seems to be able of stabilizing the disease for certain time and increasing patient survival. Further studies are necessary to settle the choice of better chemoterapeutic, dosage and time intervals between treatments. Infected pancreatic necrosis and sepsis are the leading causes of mortality in necrotizing pancreatitis. Sine 1986, 136 patients with infected pancreatic necrosis have been treated. The mean number of APACHE II score was 17.5 (range 11-31). In all cases, the infected necrosis was combined with retroperitoneal abscesses. 93 were situated in the right or left retrocolic area, 13 in the subphrenic region, and 30 in the retroduodenal or subhepatic area. The surgical treatmcr' was performed on average 18.5 days (range 8-25 days) after the ,,,set of aerate pancreatitis. The operative management consisted of wide-ranging necrosectomy in the total affected area, combined with widespread lavage and suction drainage. In 62 of the 136 cases (45.5%), some other surgical intervention (distal pancreatic resection, splenectomy, cholecystectomy, sphincteroplasty or colon resection) was also performed. Continuous lavage and suction drainage was applied for an average of 39.5 days (range 21-90 days), with an average of 8.5 (range 5-15) litres of saline per day. The bacteriologic findings revealed mainly enteral bacteria, but Candida infection was detected frequently. The incidence of fungal infection was 21%. Twenty-three patients (17%) had to undergo reoperation. The cytokine production capacity (TNF, IL-1 and IL-6) was shown to correlate with the prognosis. The overall hospital mortality was 6.6% (9 patients died). In our experience, infected pancreatic necrosis responds well to aggressive surgical treatment, continuous, long-standing lavage and suction drainage, together with supportive therapy consisting of immunonutrition and blockade of cytokine production, combined with adequate antibiotic and antifungal medication. Some reports suggest that a low dose of eholeeystokinin oetapeptide (CCK-8) may induce regeneration of the pancreas. This regeneration can be detected in ever increasing degree up to the end of the first month; it then becomes stable.
We wanted to investigate whether changes occur in the serum TGF-[31 and IL-6 levels during regeneration, and whether there is a eormeetion between their levels and the rate of regeneration in rats.
Distal pancreas resection (75%) was performed. CCK-8 was administered subcutaneously in a 300 ng/kg dose 3 times per day to the investigated group, while the control animals received the same amount of saline. The rats were examined 3, 7, 14 and 28 days after the first injection. Serum TGF-[31 levels were determined by ELISA, IL-6 levels by bioassay, DNA content by Giles & Meyers method. The weights of the residual pancreas.were increased in both groups on day 3. At subsequent times the weights decreased in the controls but increased continuously in the CCK-8-treated group. There was significant difference on day 14, andon day 28 the pancreas weight almost doubled in the CCK-8 group, whereas it decreased to the normal level in the controls. The DNA content of the pancreas was continuously higher in the treated than in the control group. It reached the maximum level on day 28, with a significant different (1800 + 350 vs. 780 + 240 y/pancreas). The protein content reached its highest level on day 28 in the CCK-8-treated group (32.435 + 7.88 rag/pancreas). A significantly higher level of IL-6 was measured on day 7 ,s. the control (250 + 70 vs. 50 + 30 pg/ml). It later decreased, but remained above the control level. Significantly different TGF-[3 levels were measured on days 7 and 14 (290 + 40 vs. 155 + 70, and 275 + 10 vs. 180+ 65 ng/ml, respectively). There was no difference between the TGF-[31 and IL-6 levels in the two groups by day 28. No significant changes in the amylase levels were observed; they remained at a normal level (4.8 + 0.8 U/ml). This indicates that the increase in the pancreas weight was not caused by panereatitis. Our results reveal that regular low dose of CCK'8 injections resulted in pancreas regeneration following 75 % distal resection. This was indicated by increases in the pancreas weight, and in the DNA and protein contents of the pancreas.
Significantly elevated serum TGF-I and IL-6 levels were also detected up to day 14. These data suggests that TGF-[31 and IL-6 might play a stimulatory effect in the early stage of regeneration of the pancreas. Forty patients underwent total pancreato-duodenectomy for end-stage chronic pancreatitis. There were 34 men and 6 women, with a median age of 39 years (range 21-66 years). Alcoholism was the major aetiological agent (30 patients), and five patients had had previous acute idiopathic pancreatitis. The overwhelming indication for operation was severe abdominal pain, complicated by failing exoefine and endocrine function. Resection was performed in one (n=17) or two stages (n=23), following previous proximal (7) or distal (16) panereateetomy; progression from partial to total pancreateetomy occurred over an interval of 8-96 months (median 15 months). Another 6 patients had undergone previous pseudoeyst or duet drainage procedures. The pylorus was preserved in 28 patients and the spleen in 10. Median operation time was 6 hours (range 2.5-8.5 hours) and median blood loss 2000 ml (range 500-16000 ml). There were two hospital deaths and three patients required reoperation. Of 38 survivors, 30 obtained complete or substantial relief of pain. There were 15 late deaths at 2.5-120 months postoperatively, 13 in the alcohol group and 11 disease-related.
Total panereateetomy can relieve the intractable pain of chronic panereafifis at the cost of possible premature death from continuing alcohol abuse. The aiin of this study was to analyse comparatively results of controlled open pancreatic drainage in noninfected and infected pancreatic necrosis. Twenty four patients with necrotizing pancreatitis were treated surgically by open lesser-omental-sac drainage (celiostomy) in the last five years. They were classified into two groups according to date and indications of surgical drainage: a group of 12 patients with early celiostomy (in the first week after the onset of pancreatitis) for noninfected necrosis, and the second group of 12 patients with late celiostomy (2 to 12 weeks after pancreatitis) for infected necrosis and pancreatic abscess. Age, sex and etiology of pancreatitis were similar between surgical groups. The diagnosis has been based on clinical syndrome, laboratory tests including bacteriologic cultures, ultrasound and CT scan findings. Drainage procedure consisted in marsupialization of lesser-omental sac by suturing open gastrocolic ligament to anterior peritoneum. Abdominal wall was left largely open in this site and the drains were inserted via celiostomy. Only 6 patients with early operation underwent necrosectomy at date of laparotomy and repeated necrosectomy was possible in 4 patients. The necrosectomy was carried out during laparotomy in all patients of the second group and three to nine reexploration via celiostomy were needed for removal large residual necrosis and purulent material. In group with noninfected necrosis, eight patients died postoperatively (66.6%) because of multiple-organ-system failure. In the second group with pancreatic abscess, only one patient died early after surgery (8.3%) of septic shock. Two patients developed postoperatively distant submesocolie abscesses resolved by relaparotomy. Other postoperative complications were pancreaticocutaneous fistulas resolved spontaneously (2 patients) and external bleeding (1 patient [Aim] To investigate the validity of K-ras codon 12 point mutation, values of CEA, CA19-9, and cytology of pure pancreatic juice and bile for a diagnosis of pancreaticobiliary carcinoma. [Materials and Methods] Sixty four cases suspected ofpancreaticobiliary diseases, were examined. Pure pancreaticjuice could be obtained and measured in 33 cases-malignant :14 cases (42%), neoplasitc :19 cases (58%), and benign 14 cases (42%). Bile could be measured in33 cases--malignant: 20 cases (61%) and other 13 cases were benign non-neoplastic diseases. Pure pancreatic juice was taken from ERP catheter for 5 min without stimulation of secretin. Bile was also taken without stiulation. K-ras codon 12 point mutation, values of both CEA and CA19-9, and cytology-were examined. K-ras mutation was detected by two-step PCR-RP method. Serous CEA, CA19-9 were also examined.
[Results] 1. CEA values of pancreatic juice in cases with carcinoma were significantly greater than those in benign group (P<0.01). When cut-off line ofCEA values was set up to 50 ng/ml, predictive values were described bellow. 2. K-ras mutation of pure pancreatic juice was more frequent in carcinoma group (P<0.02). 3. When discrimination analysis was applied (Formula 1), accuracy was 100%. 4. CA19-9 in sera was most useful for diagnosis of bile duct carcinoma. In a pilot phase I/II study we have tested synthetic ras peptides used as a cancer vaccine in five patients with advanced pancreatic carcinoma. The treatment principle used was based on loading professional antigen-presenting cells (APCs) from peripheral blood with a synthetic ras peptide corresponding to the K-RAS mutation found in tumour tissue from the patient. Peptide loading was performed ex vivo and the next day APCs were re-injected into the patients after washing to remove unbound peptide. The patients were vaccinated in the first and second week and thereafter every 4-6 weeks. In 2 of the 5 patients treated, an immune response against the immunising ras peptide could be induced. None of the patients showed evidence of a T-cell response against any of the ras peptides before vaccination. The treatment was well tolerated, and could be repeated multiple times in the same patient. Side effects were not observed even if an immunological response against the ras peptide was evident. We conclude that ras peptide vaccination according to the present protocol is safe and can result in an immune response even in patients with advanced malignant disease. In a clinical setting where tumour is surgically removed, such Tcell responses may be of potential benefit. The purpose of this report is to define the outcome of acute pancreatitis and to characterize and compare the disease and treatment patterns in cardiac and renal allograft recipients. Medical records of cardiac and renal allograft recipients from 1980 to 1994 whose ICD-9 codes included acute pancreatitis reviewed. Pancreatitis defined abdominal pain associated with either hyperamylasemia (E200) and/or classic radiographic findings. Severity of pancreatitis classified mild (hospitalization ranging 1-7 days), moderate (>7 days), (ICU admission There is a broad spectrum of duodenal complications after bladder-drained PA TX using the duodenal segment technique; little is known of the effect of duodenal complications on the long-term prognosis of patients and PA grafts. MATERIALS AND METHODS: We studied incidence and outcome of duodenal complications after 373 bladder-drained (stapled duodenoscystostomy in 95%) whole organ pancreatico-duodenal transplants (7/85 thru 1/95)-.
Complications were defined as early if they occurred within the first postoperative month, and late otherwise. Mean follow-up was 5.5 months (range, to 108 months). RESULTS: 1. There were 42 duodenal leaks (11.3%); 12 early with a mean of 11.5 days (range, 1-28 days) and 30 late with a mean of 6.9 months (range, 1-36 months). The site of the leak was at the duodeno-cystostomy site (tne bladder anastomotic leaks) in 15 cases, at the stapled proximal duodenal stump in 8, and at the stapled distal duodenal stump in 4. In the other 15 cases it was impossible to identify the exact site of the leakage because the patients were treated conservatively and the studies (CystogranffCT Scan) were unable to identify the site. In 23 (55%) patients the leakage xvas oversexvn; but 6 (14%). 4 patients had a recurrent leak requiring enteric conversion 2 to 12 months after lhe first leak. 12 (28%) patients with small leaks were treated conservatively with an indwelling catheter for to 2 months with resolntion of the leak. 2. Gross hematuria (defined as severe enough to requi.re cystoscopy) occurred in 26 patients (7%), 10 early with a mean of 14 days (range, 7-21 days) and 16 later with a mean of 11.5 months (range, 1.5-60 months); 2 (8%) patients had an enteric conversion and 2 (8%) had a graft pancreatectomy. 3 In patients with periampullary carcinomas partial pancreaticoduodenectomy with only regional lymphadenectomy seems to be superior to partial pancreaticoduodenectomy with extended radical retroperitoneal lymphadenectomy, but might be explained by selection of early stages of Group B and small numbers of cases.  (111)in 24 pts. Four cases were colloid carcinoma and the degree of differentiation was not assessed. The 3 small cell neuroendocrine carcinoma presented with regional lymph nodes metastases.
The three pts with small-cell neuroendocrine carcinoma died at 6, 7 and 17 months because of disseminated metastases. The actuarial survival rate at 5 and 10 years was calculated in 51 pts with adenocarcinoma and was 48.8% and 16.9% respectively (Kaplan-Meier method). The anastomosis between the remaining pancreas and the intestinal tract after pancreatoduodenectomy(PD) has been the site of complications responsible for considerable morbidity and mortality. Pancreatogastrostomy(PG) has been recommened by a few surgeons for some theoretical and practical advantages over pancreatojejunostomy (PJ). The purpose of this study is to determine whether PG can be a safe alternative to PJ. The indications of surgery and the operation of chronic pancreatitis with duct abnormalities is a contentious subject. The main indication for operation is the intractable pain. There are three main types of operation: partial or total resection of the gland duct drainage and denervation. There are two main drainage method in painful chronic pancreatitis: the longitudinal pancreaticogastrostomy -which has not spread widely-and the conventional pancreaticojejunostomy. The aim of our prospective study to measure whether the pacreatic juice released into the stomach had any effect on gastric acidity compared both form of surgery and evaluate the results of surgical treatment. Between 1992 and Oct. 95 43 patients with chronic pancreatitis were selected to this study, investigated and operated in our Surgical Department. 22 patients underwent pancreaticogastrostomy, 7 cases had pancreaticocysto-gastrostomy and in 14 patients pancreaticojejunostomy was applied when a two layer wide anastomosis was created between the pancreatic duct and the antrum or the jejunum loop to relieve the symptoms. A 24 hours gastric monitoring was taken on every patient before and 6 weeks after the operation. The patients have received netiimicin prophylaxis before surgery. Following a complete postoperative check up was found that both types of operations are effective for pain relief (74%), the median pain scores reduced from 12o (30-220) to 40 . 83 % of the patients had no digestive problems due to pancreatic enzyme substitution. The average postoperative weightgain was 3.8 kg. The number of diabetic patients increased from 9 to 12. There was not perioperative death. Three patients needed reoperations.
According to our statistical evaluation of 24 hours gastric pH monitoring test no alteration was detected in gastric pH level in both groups pre-and postoperatively. It might be the lack of activation of the pancreatic enzymes at the anastomotic site by the low gastric pH level and the decreased pancreatic exocrin function. On the basis of pH measuring and evaluated data we can consider the pancreaticogastrostomy -which is a simple and quick procedureis also a good operation choice to releive intractable pain in selected patients with chronic pancreatitis associated with duct dilatation. Severe acute pancreatitis is associated with infectious complications and multiple organ failure. The present study was carried out to determine the incidence of bacterial translocation in severe experimental pancreatitis in the rat and the influence of an antioxidant (Nacetyl-L-cystein).
Acute pancreatitis was induced by the intraductal infusion of taurodeoxycholate in the rat. At induction of acute pancreatitis or sham operation the animals received either sterile saline or N-acetyl-Lcystein intravenously. Sampling for bacterial cultures from portal and caval vein blood, mesentdric lymph nodes, spleen, pancreas, liver, lungs, peritoneal fluid, distal small intestine and colon was performed 12 h after induction of acutepancreatitis. The bacterial overgrowth in the distal small intestine noted 12 h after induction of acute pancreatitis was prevented by treatment with Nacetyl-L-cystein. The frequent bacterial translocation to especially mesenteric lymph nodes, but also to peritoneal fluid, liver, pancreas, spleen and lungs, was completely prevented by pretreatment with Nacetyl-L-cystein. Severe acute pancreatitits in the rat thus results in changes in the bacterial ecology in the intestine with bacterial overgrowth and bacterial translocation, both to mesenteric lymph nodes and also systemic dissemination to blood and various organs. Survival time in patients with adenocarcinoma of the pancreas is short, also when radical surgery is attempted. Various biological properties may be responsible for the rapid tumor progression. In the present study the expression of proliferation associated nuclear antigen Ki-67 and p53 gene mutations were studied as prognostic factors.
All original specimens were re-evaluated and paraffin embedded material was used for immunohistochemical analyses. Ist. Radiologia Univ. Messina-*IV Div. Radiologia Osp. 5. Manhao, Genova Our goal is to assess the usefulness of CT i.n the classification of pancreatic trauma and tbr determining a prompt and correct diagnosis. 6 patients with pancreatic trauma were examined with CT. Scans were obtained after administration of both oral and intravenous urographic contrast agents. In all cases we repeated sequences at 4 mm sfice thickness in the pancreatic region. Pancreatic trauma classified in contusion, parenchymal lacerations with or without duct distruption and complete rupture. 2/6 patients showed, at CT scamfing, a pancreatic contusion with enlargement of the pancreas and peri pancreatic edema in the region of the body of pancreas around the superior mcscntcfic artery.  (n=2), and splenectomy (n=l). Venous extension grafts on the kidney and/or pancreas were needed in 3 patients. Mean HLA-ABDr match was 2.3 with a mean kidney cold ischemic time of 16.7 hours. Mean operative blood replacement was 2.0 units, and the mean operating time was 6.2 hours. All but one of the patients received quadruple immunosuppression. The mean length of initial hospital stay was 20 days with a.mean of 1.5 readmissions. There were 4 acute rejection episodes (1 pancreas, 3 kidney), all successfully treated with steroids. Seven major infections occurred (3 CMV, 2 line sepsis, 2 peritonitis). Patient survival is 100% with a mean follow-up of 37 months (range 2-65). Kidney and pancreas allograft survival are 91% and 73%, respectively. Three grafts (2 pancreas, kidney) were lost early from thrombosis, with two (1 kidney, pancreas) successfully retransplanted.
No patient required dialysis after transplant, and all patients are currently dialysis-independent with a mean serum creatinine of 1.9 mg/dl. Conclusion: PKT after previous transplant is a challenging but safe procedure that often requires concomitant procedures, the use of vascular extension grafts, and atypical placement of the allografts. However, the excellent results justifies an aggressive policy of retransplantation in the diabetic patient with a failed or failing allograft(s). Cystic neoplasms are an uncommon group among pancreatic tumors. These lesions are seen more frequently in recent surgical practice, probably because of advances in diagnostic and surgical techniques. The aim of this study is to describe the diagnostic features and therapeutical options in the management of cystic tumor of the pancreas. We report here our experience with 72 patients with pancreatic cystic tumors over a ten-year period. Sixty-two four patients were women and ten were men. The .mean age of patients was 55.2 years (range, 21 to 81 years).
Mild abdominal pain was the main symptom in 70 % of patients. The lesion were incidental finding in 10% of patientsl CT scan provided the diagnosis of cystic tumor in 94% of patients while ultrasonography provided the same diagnosis in 78% of patients. All patients underwent surgical treatment. The pathological diagnosis was: thirty patients with mutinous cystadenoma (41.7%), thirty-two patients with serous cystadenoma .(44.4%), ten patients with mutinous cystadenocarcinoma (13.9%). There was no operative mortality. Seven of ten patients with cystadenocarcinoma ultimately died of the disease. One patient with extended resection is still alive years after surgery without recurrence of the tumor. The survival rate was 20.5% at years. All patients gath cystadenomas (mucinous or serous type) that underwent complete resection are alive or died from other causes.
Only complete resection of the cystic tumors of the pancreas provides certain pathological diagnosis, the best chance of cure and may remove the risk of malign transformation of the cystadenomas, particularly of the mucinous type, with minimum operative risk. Cystic neoplasms represent up to 15% of cystic lesions of the pancreas. The aim of this study is to analyze retrospectively 60 cases of cystic tumors of the pancreas operated on between 1990 and 1995. There were 53 women and 7 men with mean age of 55.9 years. All patients were operated on. 27 patients presented a mucinous cystadenoma (45%), 24 presented a serous cystadenoma (40%) and 9 patients presenteda mucinous cystadenocarcinoma. There were 12 pancreaticoduodenectomies, 2 total pancreatectomies, 28 distal pancreatectomies, 8

enucleation and 8 biopsies.
There was no operative mortality. A preoperative diagnosis of serous cystadenoma was made in 15 patients, confirmed in 12 of them (80%). A preoperative diagnosis of mucinous cystadenoma was made in 19 patients, confirmed in 12 of them (63.2%) while 2 presented a cystadenocarcinoma. Undetermined cystic lesion was the diagnosis proposed for 16 patients. The final diagnosis was serous cystadenoma (n=10), mucinous cystadenoma (n=2) and cystadenocarcinoma (n=4). Ten patients received preoperative diagnosis of pancreatic pseudocyst, but in eight of them the peroperative diagnosis was correct. Unfortunately twocases were mistaken by pseudocysts and a drainage into a viscous organ was performed. Both patients were reoperated .on after diagnosis of cystic neoplasm. The final diagnosis was mucinous cystadenoma (n=6) and cystadenocarcinoma (n=4) ( Table 1).  INTRODUCTION We have often experienced delayed gastric emptying after pylorus preserving pancreaticoduodenectomy (PPPD), but the mechanism of the delayed gastric emptying is still under. Therefore we have investigated the cause of that by dogs. MATERIAL AND METHODS PPPD performed on the nine mongrel dogs. Reconstruction was carried out by modified Cattd's procedure. The right gastric artery and the pyloric branch of the vagus nerve were preserved. Straingage force transducers (S.G.T) were placed the stomach body, antrum and the jejunum. On the 7th postoperative day we gave five dogs barium and meal, and confirmed the patency of the pyloms ring by X-ray. On the 7th, 14th and 28th postoperative days we studied the contraction of the stomach and the jejunum for 24hrs and measured plasma motilin concentration in four dogs. Five normal dogs were implanted S.G.T. on the stomach body, antrum and the duodenum to obtain control measurements of gastrointestinal motility and measured plasma motilin concentration. RESULTS Pylorus ring patency In all dogs given barium and meal could it passed through the pylorus ring and confirmed the patency of that. Motor activity of the stomach On the 7th postoperative days only irregular and short contractions were observed. The interdigestive myoelectrical complex (IMC) of the stomach did not appea. On the 14th postoperative day bands of strong contractions like the IMC appeared. But the duration and the cycle were different from the IMC significantly. Plasma motilin concentration Plasma motilin concentration when the bands of strong contractions appeared did not differ from those of other phases. CONCLUSION In the dogs undergone PPPD the pyloms was open and the disorder of the gastric motility was the cause of the delayed gtric emptying after PPPD. On the 14 postoperative day the bands of strong contractions like the IMC appearext It was supposed that appearing those contractions showed the process of recovering from the delayed gastric emptying. It was speculated the low motilin concentration gave rise to the disorder of the gastric motility. In 38% of the cases an acute exacerbation of the previously known chronic pancreatitis have been detected, whereas in the remaining 62% the pseudocysts constantly persisted without new onset of acute inflammatory changes. The surgical procedures of choice were drainage in 88%, resection in 12%.-The majority (73%) of drainage procedures were internal, directed towards the gastrointestinal tract. Based on the favourable results the authors preferred anastomoses performed between the pseudocyst and stomach (posterior cystogastrostomy), as well as the pseudocyst and duodenum (blunt forced cystoduodenostomy).
Additionally at the same time efforts have been made to resolve the morphological alterations of the duct of Wirsung caused by chronic pancreatitis. This has been achieved by pancreatic duct decompression procedures such as cysto-Wirsungo-gastrostomy, and sphincteroplasty of the papilla of Vater. (53% of the operations were such combined interventions.) The early postoperative morbidity was 12.9%, mortality was 1.46%. The early mortality was significantly increased by complications requiring reoperations, such as insufficiency of the anastomosis, haemorrhage, and abscess formation. There were no statistically significant differences regarding complications and early death rate between the combined and non-combined procedures.
In the patient group (n=700) that have undergone operations in the first 5 years a long term follow up using standardised questionnaires was completed. 87% of the answers were satisfactory for evaluation with an average of 44 months follow up period. The late recovery results were found to be excellent in 23%, good in 36%, satisfactory in 30% and poor in 11%. The late mortality after 4 years follow up was found to be 15.5%. (2) tail resection and left pancreatectomy; (3) resection of head of pancreas, with or withdut pancreatojejunostomy; (4) drainage operation: and (5) traumatic lesions.
1. Fibrin sealing allows seamless sealing of the tissue defect.
2. Following tail resection or left pancreatectomy, the resection surface is trimmed in the form of a fish mouth: the gaping wound edges are approximated and sealed, and the "lips" are covered with an additional coat of the fibrin sealant. 3. Following Whipple's operation for chronic, pancreatitis both anastomoses hepaticoand pancreaticointestinal are sealed with fibrin sealant. In cancer surgery the remaining pancreas is occluded with fibrin sealant via the pancreatic duct. The pancreatic duct is ligated and the resection surface sealed with an additional layer of fibrin sealant, eventually in combination with collagen fleece. In 92 patients there was no lethal case. As to local complications, there was one case of a clinically necrosing remaining pancreatitis and 9 cases of postoperative pancreas fistulae. Cystic pancreatic tumours are uncommon. In the ist Surgical Clinic of the University of Bologna, from 1975 to 1995, 0 cases have been treated: they represent 7.5% of all pancreatic neoplasms (532 cases) and 24.4% of all cystic pathologies seen at our Institute dring this period. Different histopathological types of cystic tun mours were observed of which 19 (47.5%) were malignant (19 mucinous cystadenocarcinomas), 16 (40.0%) benign (7 serous cystadenomas, 6 mucinous cystadenomas, 2 lymphangiomas, cystic insulinoma) and 5 (12.5%) of unce tain malignancy (3 intraductal papillary neoplasms, 2 solid and cystic tumours). Of the 40 cases, 35 (87.5%) underwent surgical resection, of which 19 left pancreatectomy, 8 pancreatico-duodenectomy (4 according Child, 4 Traverso-Longmire), 4 exeresis of neoplasia, 3 subtotal pancreatectomy and, finally, 2 total pancreatectomy. Resection of benign and uncertain malignancy cystic tumours was always possible (resecability index 100%); in mucinous cystadenocarcinomas, instead, pancreatectomy was performed in 14 out of 19 cases (resecability index 73.6%). In these patients two-year and five year survival was respectively 46.2% and 36.9%. Of the 16 p tients with benign cystic tumours, died of postoperative complication and 15 were still alive and well at a mean follow-up time of 45 months (range 4-132 months).
Three patients with intraductal papillary neoplasms and two with solid and cystic tumours of the pancreas were still alive and well at a mean follow-up time of 68months (range 1-147 months). For all cases two-year and five-.year actuarial survival was respectively 71.1% and 62.2%. These results indicate that cystic tumours of the pancreas have a high Xndex of resecability (87.5%) and a relatively good prognosis with respect to solid pancre tic neoplasms. For these reasons careful differential diagnosis is essential and, in the case of positivity for cystic tumours, aggressivesurgicaltreatmentmust be done. F170 F171 ENDOCRINE PANCREATIC TUMOURS: OUR EXPERICEWITH59 CASES Marrano D., Greco V.M., Casadei R. Taffurelli M. ist Surgical Clinic, University of Bologna, BolognItaly Endocrine pancreatic tumours are rare. In the ist Surgical Clinic of the University of Bologna from 1975 to 1995, 59 cases were observed: they represent 11.1% of all pancreatic neoplasms (532 cases). Two groups of endocrine tumours were recognized: functional and non func. tional. Functional endocrine tumours were distinguished in 21 insulinomas, 15 Zollinger-Ellison syndrome, 2 Vipomas, glucagonoma, somatostatinoma and PPoma. Insulinoma were benign in 17 case, malignant in 2 cases and, finally., in 2 cases on iperplasia of B cells was shown. All were resected (14 left pancreatectomy with splenectomy, 6 enucleation and left pancreatectomy with hepatic resection of metastasis). Zollinger-Ellison syndrome were supported by a gastrinomas in 9 (64.2%)c_a ses; in 5 (35.8%) gastrinomas were occult; in case iperplasia of G antral cells was shown. All identified gastrinomas were resected. The other very rare functional endocrine tumours were all resected; in case (somatostatinoma) with hepatic resection of some small metastasis and in another one (vipoma) with resection of retroperitoneal lymph nodes. Non functional endocrine tumours were resected in 15 out of 18 cases (index of resecabili ty=83.3%) (8 left pancreatectomy, 3 enucleation, 2 pancrea_ toduodenectomy and 2 intermediate pancreatectomy). Of the 59 endocrine pancreatic tumours 32 (54.2%) were benign,27 (45.8%) malignant. Resection of benign endocrine tumours was always possible while a pancreatectomy was performed in 24 out of 27 malignant tumours (index of resecability 88.8%). In 3 cases pancreatic resection was associated to hepatic metastatic resection. In all cases postoperative cours were uneventful with relief of the endocrine symptoms. Five-year survival were 100% for benign tumours, 60.2% for malignant. These data indicate that endocrine pancreatic tumours have an high index of resecability(95%). Moreover it is important to resect alo hepate etasta sis when it is possible because of relief of endocrine symptoms and relatively good prognosis due to slow growing of these tumours.
EFFECT OF E,.ARLY JEJUNAL FEEDING ON THE SEPTIC
Purpose of study: The necrotized tissue of the pancreas is colonized by bacteria r,ainly from the colon which may lead to abscess or infectet necrosis. Our purpose was to prove, that early jejunal feeding decreasing the paralytic condition and distension of colonic wall, starting the passage and preserving .the normal colonic bacterial flora can reduce the rate of septic complicati.om. Patients and method: The study included 38 patients with acute, biliary panereatitis, randomizing into two groups. In group A (n=lS)jejunal feeding was started in the fh'st 24 hours (second jejunal loop; Survimed OPD, Freseaius). In group B (n=20) total parenteral nutrition was applied. Between the two groups neither in the male:female ratio 13:5 and I8:2"1, nor in the average age (48.2 and 44.7 years) nor in the etiolomy 13 and 16 alcoholic; 5 and 4 idiopathic) were significant differencefound. Results: Necrosis developed altogether in 14 patients (36.8 %). In group A two infected and three sterile necrosis were detected. In group B four infected, one sterile necrosis and four abscess were found. Septic complication due to bacterial contamination developed in cases in group A (11%) and in 8 cases in group B (40 %). Statistical difference is significant (p=0.047; Fisher-test). In group A four patients, in group B eight patients underwent operating procedure. Mortality rate was 5.5 % in the group ofjejunal feeding and 10 % in the control group.
Conclusion: Our results suggest that early jejunal feeding in the treatment of acute panereatitis reduce the rate of bacterial contamination of necrotized pancreatic tissue mainly in the later phase of disease, after the first week. Seven patients required reoperation. In 73.3% of the laparotomies a tumor was felt or seen. In 3 cases, additional undetected adenomas were resected by chance. Overall negative surgical exploration rate was 27.7% (20/72 cases) and in 14/20 the preoperative imaging was wrong or negative. Operative U.S. correctly localized the tumor in 22128 (78.6%) patients detecting also 2 "occult" insulinomas, but had 2 false positive (7.1%). Only 51/65 patients (78.5%) had typical single adenoma. Five (7.6%) had multiple adenomas and 5 had hyperplasia or nesidioblastosis. In 26.1% of patients the tumor size was cm.
Two cases had liver metastases and in 2 patients the tumor was not found. Excluding the cases with metastases, the outcome of the patients was as follows: 60163 (95.2%) were cured after surgery (1 of them recurred 3 years later); 2/63 (3.2%) were unchanged, and 2/63 developed diabetes after a near-total pancreatectomy (including the case who recurred In patients with acute pancreatitis, evaluation of the morphology and contents of the biliary tract is essential for diagnosis and optimal management of the disease. The aim of this study was fo compare the morphological findings pertinent to assessment of the common bile duct by ultrasonography (US) and endoscopic retrograde cholangiopancreatography (ERCP) in patients with biliary acute pancreatitis. Thirty-one patients were studied (10 males, 21 females, mean age 67 years, range 40-90); the diagnosis of acute pancreatitis was made on the basis of characteristic abdminal pain associated with an elevation of serum amylase and lipase. US was performed in all patients on admission to the hospital and all subsequently underwent urgent ERCP and endoscopic sphincterotomy within 24 hours. US was .perfo.rmed with an Ansaldo AU 450 apparatus, using a 3.5 MHz transaucer; ERCP was performed using a Fuji ED7-XU2 duodenoscope. For the purpose of the study, the endoscopist was kept unaware of the morphological details reported by the sonographer. US showed choledocolithiasis in 2 patients and microlithiasis of the common bile duct in 12. ERCP showed choledocolithiasis in 4 patients and microlithiasis of the common bile duct was detected in 20. The mean of the common bile duct diameters determined by US was 7.6 mm (range4-12 mm), which was significantly smaller (P<0.001) than the value obtained with ERCP (mean value 10.2 mm, range 4-17.5 mm). For this.. Preoperative differentiation of pancreatic cysts is important for appropriate treatment. Cystic neoplasms of the pancreas are observed with increasing frequency, often as incidental finding. CT-scan is the most common radiologic investigation for preoperative detection of pancreatic lesions, but misdiagnosis of pancreatic cysts is relatively frequent. We reviewed our experience of CT-scan in 63 patients with 69 cystic lesions of the pancreatic area, observed from May 1988 to March 1995. We evaluated the accuracy of CT-scan to distinguish pancreatic or extrapancreatic lesions, to differentiate between benign and malignant cysts, and to identify different type of cysts. The final diagnosis included: 19 pseudocysts (28%), 6 serous cystadenomas, 20 mucinous neoplasms (29%; 5 adenomas, 12 carcinomas,3 IHMN), 9 ductal carcinomas, 3 endocrine tumors, papillary cystic tumor, lymphoma, lleiomyosarcoma, 2 solitary true cysts, 7 extrapancreatic lesions (10%).
Among different type of cyst CT correctly identified 15119 pseudocysts (79%), 14/20 mucinous neoplasms (70%), 316 serous cystadenomas, 7/9 ductal carcinomas (78%). None of the other rare types of tumor was correctly identified. In 8 % of cases the radiologic diagnosis was changed when the clinical history of the patient was known. In 25% of patients, with a cyst of the pancreatic area CT-scan in not able to differentiate the type and the malignant nature of the cyst. In our series 12% of case's were rare types of pancreatic lesion and in these patients there is no chance to recognize the nature and behaviour of the cyst. Popiela T., Matyia A., Thor P., Herman R.M., Plebankiewicz S., Lorens K., Kawiorski W., Ist Dept. of General and GI Surgery Jagiellonian University, Cracow, Poland.
Pylorus preserving pancreatoduodenectomy (PPP) has been proposed as an alternative procedure to conventional Whipple procedure (CWP).
The aim of this study was to evaluate the gastric motility and Ph using modern methods of studies.
Patients: Two groups of patients, each consisted of 20 persons were examined before and after surgery group patients following PPP and group II patients after CWP.
Methods: Transcutaneous electrogastrographic examination and 24-hour Ph monitoring was performed in each patients before and up to one year after surgery.The following EGG patterns were evaluated in fasting and postprandial state: slow waves time and space distribution, percentage of normal slow waves, percentage of gastric dysarrhytmias. During the 24hr pH monitoring the number of reflux of jejunal contents episodes and its duration were evaluated.
Results: In PPP the frequency of gastric dysarrhytmias is lower than following CWP. The dominant frequency of SW increased up to 10 + 2 cpm which occupied 56% of the examination time. In CWP the significant decrease in SW frequency occurred, although in symptomatic patients the aboral propagation of jejunal pacemakers with frequency of 12 cpm were observed.

Conclusion: EGG evaluation and continuous 24-h
.monitoring appeared to be useful methods in diagnosing and monitoring of motility sequels following surgery. The frequency of motility disturbances following PPP was significantly lower than after CWP. The surgical strategy used in treating symptomatic chronic pancreatitis, namely drainage or resection, has generated much controversy. The Fray modification of the Puestow procedure combines the principle of wide drainage of the pancreatic duct with in situ resection of the head of the pancreas. The aim of this study was to prospectively assess the results of the Frey procedure in our patient population, this being the procedure of choice since early 1992.
Between March 1992 and November 1995, 24 patients underwent the Frey procedure for symptomatic chronic pancreatitis. There were 20 males and 4 females, mean age was 38.6 years (29-48). The aetiology was alcohol in 21, and idiopathic in 3. Pain alone was the presenting symptom in 17 (71%) patients, 6 had pain and jaundice, and one had jaundice alone. Six patients (25%) had previous pancreatic surgery, 3 having a distal pancreatectomy, and 3 a Puestow procedure. Exocdne insufficiency was present in 25% preoperatively, and 37% were diabetic. The Frey procedure was performed alone in 12 patients (50%), 11 were combined with biliary drainage (46%), and with distal pancreatectomy. Significant morbidity.requiring reoperation occurred in two patients, one an eady adhesive obstruction, and in another major bleeding from the pancreatic bed occurred two weeks following surgery. Overall morbidity was 42% and mostly minor, respiratory problems being the commonest. One patient died of intra-abdominal sepsis following leakage of an entero-enterostomy, an avoidable complication (mortality 4%). Of the 23 survivors, 4 have been lost to follow-up. In the remaining 19 patients, mean follow-up time is 13.2 months (range 1-41). Relief of pain has been good or excellent in 18 patients (95%), and poor in 1. No recurrence of jaundice has occurred in any of the patients. One of 12 patients seen at follow-up had developed new symptomatic exocdne insufficiency. No new cases of diabetes were detected following the surgery.
In conclusion, the eady results of the Frey procedure have been promising. Sedous morbidity is low and the one death in the sedes was avoidable. Although follow-up is short, very good relief of symptoms has been achieved. No significant deterioration in exocdne or endocrine function was detected following this operation. Introduction: The cystic tumours of the pancreas are not very common. Must be done a differential diagnosis with pseudocysts and the treatment is always surgical.

Material and Methods:
We reported on a retrospective trial of 9 patients with CTP operated on during the last 10 years. Two were serosous cystoadet,rnas, 5 mucinous cystoadenomas and 2 cystoadenocarcinomas. Eight were females and was male, with a median age of47,2 years (28-72). The cystic character of the ,,-T was preoperatively diagnosed (laparoscopy, ultrasonography CAT scanner) but not.even the fine needle aspiration can establish a certain diagnosis.

Results:
Depending on the tumour localizationthe surgical treatment is performed, carrying out 6 distal pancreatectomies (in 2 cases with splenic preservation) and cephalic duodenopancreatectomies(CDP). During the postoperative period one patient died, a 72 years old male with CDP due to cystadenocarcinoma, of caute renal insufficiency. There were no complications in the postoperative period with the remaining patients.
After a follow-up, varying from 4 months and 10 years (3 years), there was.recurrence in one case, the patient subsequently died. One patient died due to urdmown causes. The remaining (6 cases) still alive. Conclusiqns: 1) The cytological and radiological criteria are not enough to achieve a preoperative diagnosis of the CPT.
2) These tumours must be treated surgically. It is curative in benign forms and offers good results in malignant cases regarding non-cystic types of adenocarcinomas. Pancreatic injuries from blunt trauma are infrequent, and their diagnosis and management can be difficult. Over the last 5 years we treated 13 patients with major pancreatic injuries from blunt trauma. 12 were involved in MVA, none of whom were wearing seat-belts.
assault. Only 5 patients had physical findings suggesting intraabdominal injury. Serum amylase were elevated in 7 of the 9 patients tested. CT demonstrated injury in 3 of 4 patients scanned.
diagnosed by u/s. 5 patients who had DPL suffered other concomitant injuries that produced haemoperitoneum. Injuries were equally distributed throughout the pancreas.All patients underwent celiotomy. 8  Preoperative. differential diagnosis of cystic lesions of the pancreas may be difficult because there are no reliable clinical or radiological cdteria to assist in making the differentiation. The crucial point is to recognize mucinous and/or malignant tumors, in which a surgical resection is mandatory;Recently, analysis of aspirated cyst fluid for enzymes (amylase, lipase), tu,mor markers and cytology has been used to provide a preoperative diagnosis of pancreatic cysts. Aim of the study was to evaluate the utility of serum and cyst fluid analysis. for enzymes (amylase and lipase) and tumor markers (CEA, CA 19-9, CA 125, CA 72-4) in the differential diagnosis of pancreatic cystic lesions. Serum and cyst fluid were obtained from 48 patients with pancreatic cyst (21 pseudocysts, 14 mucinous cystic neoplasms, 6 ductal carcinomas, 7 serous cystadenomas), observed from 1089 to 1994. Preoperatively, a basal sample of blood was collected and the serum was kept frozen until assay. The cyst fluid was collected by percutaneous fine needle aspiration in 20 cases, and by intraoperative aspiration in 28. The tumor markers were measured using commercially available double antibody immunoassaies (CEA: n.v. < 5 ng/ml; CA 19-9: n.v. < 37 U/ml; CA 125 n.v. < 35 U/ml; CA 72-4, n.v. < 4 U/ml). Serum CA 19-9 levels were significantly (p < 0.05) higher in ductal carcinomas (all > 100 U/ml) and mucinous cystic neoplasms. CA 72-4 cyst fluid levels were significantly higher in mucinous cystic tumors (p < 0.005) with 95% specificity and 80 % sensitivity in detecting mucinous or malignant cysts. Combined assay of serum CA 19-9 and cyst fluid Ca 72-4 correctly identified 19/20 (95 %) of (pre-)malignant lesions, with only one (3.6 %) false positive. Cytology correctly classified all 21 pseudocysts as inflammatory lesions. Mucin-containing epithelium were found in 8114 MCN and malignant cells in 316 ductal carcinomas. Cytology showed a sensitivity of 48% and specificity of 100%.
Any pancreatic cyst with high serum CA 19-9 values, or positive cytology, or high CA 72-4 in the fluid should be considered for resection. Study supported by C.N.R., contract nr. 94.01179.PF39. The purpose of this retrospective study was to find a new scoring system for early differentiation of the two most common acute pancreatitis etiologies: biliary and alcoholic, because biliary pancreatitis can be treated early by endoscopic sphyncterotomy, whereas such treatment is unnecessary in alcoholics.
A hundred and forty-five patients (57 males and 88 females) with diagnosis of acute pancreatitis based on a combination of clinical features, a typical case history, elevation of serum pancreatic enzymes and confirmation with imaging studies (ultrasonography or contrast enhanced computed tomography), satisfied require ments for participation in the study. Seven parameters (serum amylase, aspartate aminotransferase /AST/, alanine aminotransferase /ALT/, alkaline phosphatase /ALP/, urine amylase, lipase/amylase ratio /L/A/and erythrocyte mean corpuscular volume /MCV/ that differ (statistically significant, p<0.001) between patients with biliary and alcoholic pancreatitis were included in the scoring system. Each parameter according to its values was counted as 0 or i, so the patients reached scores from 0 to 7. Score >4 differs biliary pancreatitis from alcoholic with snsitivity of 92,92%, specificity of 93,75%, positive predictive value of 98,11% and negative predictive value of 76,92%. We conclude that our new scoring system calculated from routine laboratory parameters could be important support in early differentiation of acute pancreatitis etiology because it is non-invasive, fast and inexpensive. EXPERIENCE WITH 59 CONSECUTIVE SOLITARY PANCREAS TRANSPLANTS (PTX) R,J. Stratta, R.J. Taylor, I.S. Gill, L.G. Weide, D. Sudan, R. Sindhi, P. Castaldo, J.T. Jerius, K. Cushing, K. Frisbie, M.T. Grune, S.J. Radio Department of Surgery, University of Nebraska Medical Center, Omaha, Nebraska, USA From 3/91 through 11/95, we performed 59 solitary PTXs (17 sequential pancreas after kidney, 42 PTXs alone) in 54 adult Type diabetic patients. Indications for solitary PTX were: 1) the presence of 2 or more overt diabetic complications; and/or 2) glucose hyperlability with hypoglycemic unawareness and impaired quality of life. The recipient group consisted of 28 men and 26 women with a mean age of 38 years (range 25-62) and a mean duration of diabetes of 27 years (range 14-52). The recipient evaluation emphasized the documentation of diabetic complications, as well as adequate cardiac and renal reserve. Organ acceptance was restricted to ideal donors and mandated a minimum of a 2 antigen match (mean ABDR match 2.5). The mean cold ischemia time was 16.6 hours. Whole organ PTX was performed with bladder drainage. Ten patients (17%) received pancreas retransplants. All patients received systemic anti-coagulation intra-operatively, anti-platelet therapy post-operatively, and were managed with either triple or quadruple immunosuppression. Monitoring included prospective urine cytology as well as protocol and clinically indicated cystoscopic transduodenal needle biopsies. Twenty-six cases were managed with either FK506 induction (11) or rescue (15) therapy. The mean initial length of hospital stay was 16.7 days, and mean hospital charges were $100,664. The incidence of rejection, surgical complications, and major infections was 71%, 47%, and 59%, respectively. Actuarial year patient and graft survival is 90% and 62%, respectively. All patients with functioning grafts have excellent metabolic control (mean glycohemoglobin level 5.1%) and have achieved good rehabilitation without major cardiovascular, renal, o( progressive diabetic problems. Conclusion: Despite morbidity, solitary PTX can be performed with improving success, enhances quality of life, and offers an opportunity to arrest secondary diabetic complications. Pancreatic cancer is a disease with a poor therapeutic outcome. The frequency oflocal recurrence and liver metastasis is high even when extended resection (D2) is performed. In this study we assessed the results of performing IOR to prevent local reeunence and PC means of preventing liver metastasis. The subjects of this study were 65 patients with cancer of the head of the pancreas who underwent resection between September 1985 and February 1995. The overall degree of progression was stage in 7 patients, stage II in 6, stage HI in 42, stage IV in 10 (pTNM, UICC). The surgical procedure performed panereatoduodenectomy (PD) in 30 patients, PD + portal vein resection in 20, total pancreatectomy (TP) in 2, and TP + portal vein resection in 13.
We divided the patients into four groups, group surgery (D2) alone in 23 patients, group 2 surgery + IOR in 14, group 3 surgery + PC in 10, group 4 surgery + IOR + PC in 18, and assessed the sites of recurrence and survival time retrospectively. Intraoperative radiation consisted of administering 20-30 Gy of electron beam radiation retroperitoneally after performing resection, while portal vein catheterization was used to continuously infuse 5FU, 250 mg/day, for two weeks through the recanalized umbilical vein.
There was no significant difference in the stage distribution in these groups. Investigation of sites of recurrence revealed the following, group local 87% (20/23), liver 74%(17/23), peritoneal dissemination 4%(1/23), group 2 79%(11/14), 57%(8/14), 29% (4/14), group 3 50%(5/10), 40%(4/10), 0%(0/10), group 4 61%(11/18), 22%(4/18), 33%(6/18). Comparison of the rate of local recurrence in groupl and group 2, and in group and group 4, failed to show a significant difference between either pair of groups (P=-0.50, P=0.06, respectively), and no preventive effect of intraoperative radiation on local recunence was detected. Comparison of the rate of liver metastasis in groupl and group 3, failed to show a significant difference (P=0.06), but comparison of group and group 4 revealed a significantly lower liver metastasis rate in group 4 (P=0.001). Mean survival time was 11.8 months in group 1, 9.7 months in group 2, 23.6 months in group 3 and 18.4 months in group 4, and group 4 had significantly longer survival than group and 2. There was local recurrence preventing effect and liver metastasis preventing effect in the surgery (D2) + IOR + PC "group. Increasing evidence suggests that motility disorders of the sphincter of Oddi may lead to episodes of recurrent pancreatitis in a small percentage of patients who are given the diagnosis of idiopathic recurrent pancreatitis. Over the course of 10 years, 35 patients have been identified and treated for this condition. The aim of this study was to assess symptomatic outcome in these patients. Following the exclusion of common causes of pancreatitis the patients underwent sphincter of Oddi manometry. Patients with manometric abnormalities and 3 patients with normal manometry underwent treatment. Initially, patients were treated conservatively and in ten an endoscopic sphincterotomy of the biliary part of the sphincter was performed. Twenty six patients with persistent symptoms underwent total division of the sphincter via an open sphincteroplasty and septectomy. Patients were followed up according to symptoms and classed as cured, mild symptoms or no change. On a median follow up of 24 months (9 to 105), 15 of the 26 patients (58%) were cured. Eight (31%) had only mild symptoms which did not require medical treatment; 3 patients remained unchanged. In the majority of patients having a good clinical outcome, manometry had demonstrated sphincter of Oddi stenosis.
Total sphincter of Oddi division is associated with a good symptomatic outcome in patients with recurrent episodes of pancreatitis and documented sphincter of Oddi stenosis.  Our recent experience, as noted above, reveals a 0% perioperative mortality rate in the 1990's of the 16 patients who underwent curative resection with median survival of 17.4 mos. and 5-yr survival of 16.7%. This is consistent with previous recent reports that the Whipple operation is safe in the 1990% with perioperative mortality of less than 1%. The resection of pancreatic cancer does improve these patients' five-year survival. Operative treatment of necrotizing pancreatitis at our institution has evolved from open packing to repeated operative necrosectomy/debridement every other day with delayed primary wound closure over drains. We prospectively analyzed our results. with operative management of 72 patients (mean age 61 yr; range 20-93) with NP from 1983-1995. RESULTS Overall mortality was 25%. Univariate analysis showed that mortality was increased 1) in patients over age 59 (p=0.06), 2) when the preop APACHE II score was > 13 (p<0.005), 3) with pancreatic parenchymal versus extrapancreatic necrosis (p=0.05), and 4) when the perioperative course was complicated (13 patients) by intraabdominal hemorrhage (p<0.01). In contrast, the number of operative debridements (mean 3, range 0-21) or the development of pancreatic and/or gastrointestinal fistulas (35%; 25 patients) and recurrent intraabdominal abscesses (12 %; 9 patients, 5 ofwhom were treated percutaneously) were not associated with increased mortality. Multivariate analysis showed that APACHE II score and perioperative hemorrhage maintained significance (p=0.05 and 0.03, respectively) as prognostic factors for mortality. CONCLUSIONS: NP still carries very significant morbidity and mortality. Planned reoperative necrosectomy with delayed primary closure is associated with a very low incidence of recurrent intraabdominal abscess. Peripancreatic tissue necrosis with preservation ofthe viability ofthe pancreatic parenchyma, younger age and especially APACHE-2 score <13 and absence of hemorrhage are associated with more favorable outcome. Gouma. Department of Surgery, Academical Medical Center, Amsterdam, the Netherlands.
Surgical palliation of unresectable pancreatic head carcinoma is commonly associated with high morbidity and mortality, thus favoring the use of nonsurgical (endoscopical) procedures for palliation. More recent reports, however, have shown improved results of surgical palliative treatment. This prompted us to review the results of palliative biliary and gastric bypass surgery in 126 patients who were found to have an unresectable tumor. Patients: Between 1983 and1994, 126 patients (m:f 64:62) with a median age of 64 years (range: 39 to 90 years) underwent a biliary and gastric bypass (n=118), a biliary bypass alone (n=6) or a gastric bypass alone (n=2).
Biliary bypass surgery in most cases consisted of Roux-and-Y hepaticojejunostomy (79%), and gastric bypass of gastro-jejunostomy (GJ). Surgical palliation was undertaken when laparotomy revealed an unresectable tumor (n=44), when endoscopical treatment failed (n=l 9), when gastric outlet obstruction had occurred (GO0, n=28) or for miscellaneous reasons (n=35 Pancreatic cancer is characterised by frequent mutations in the p53 gene, leading to p53 overexpression. Detection of p53 overexpression strongly correlates with neoplasia in many cytological specimens. In order to test the usefullness of p53 assessment in detecting early stages of pancreatic cancer we analysed cytological specimens of pancreatic juice samples for p53 expression. 16 out of 27 cytological specimens from patients with pancreatitis (59%) and 10 out of 15 specimens from patients with pancreatic carcinoma (67%) were positive for p53 expression. The wildtype (wt) p53 specific monoclonal antibody PAb1620 was positive in 11 out of 27 specimens from patients with pancreatitis and 10 out of 13 specimens from patients with pancreatic carcinoma. The results clearly indicate, that p53 overexpression is observed in cytological specimens of patients with pancreatic cancer as well as patients with pancreatitis, whereas the expression of p53-protein could never be observed in normal pancreatic tissue. The expression of heat shock protein 70/72 was analysed in parallel: HSP 70 in those cytological specimens. In vitro, wt p53 expression was inducible in pancreatic adenocarcinoma cells by TNF-( treatment, followed by apoptotic cell death as revealed by in situ "Terminal Transferase Test". A corresponding result was achieved by analyzing cytological specimens from patients with pancreatitis. Our findings suggest, that TNF-( (increased levels during inflammation and accompanied by a Orelease) is a factor that stimultates p53 expression. This might be one reason for wt p53 overexpression in cytological specimens of pancreatic juice samples from patients with pancreatitis. Chronic inflammation might therefore exhaust the function of wt p 53 as the "guardian of the genome" The pancreatic-digestive anastomosis is then highly endangered when a normal lienat pancreatic tissue after pancreatic head dissection has to be anastomosed with the jejunum. In the literature many procedmes for the permanent as well as the temporary elimination of the exocrine pancreatic secretion are known which in most cases is responsible for the anastomosis-related complications. In animal experiments we could show the efficiency of PDO with FS for the protection of this anastomosis, in the meantime we could gain great experience with this method in clinical trials. Since 1987 87 patients underwent partial duodenopancreatectomies (Whipple's procedure) due to pancreatic head or peri-ampullary carcinomas. After typical pancreatic head resection the duct of the remaining lienal pancreas was intubated with a thin catheter and then occluded with an average.of 2 (+/-0.4) ml FS. A high concentration of aprotinin (anthqbrinolytic substance) of 20.000 I.U./ml added to the sealant is required to prevent a premature dissolution prior to the 5th postoperative day. Then the remaining pancreas was anastomosed in single-layer fashion. Results: The postoperative lethality rate amounted to 2.4% (n 2) and was not method-related. In cases bile fistulas occurred, in one a partial liver necrosis and in another one a colonic necrosis, which required relaparotomies. No pancreatic fistulas occurred. A remarkable fact is that postoperatively only 5 of the patients suffered from exocrine and only 3 from endocrine, insulin-dependent insufficiencies.
Conclusion: The PDO with FS represents a safe and effective method for the protection of the pahcreatic-digestive anastomosis and can therefore be recommended for Whipple's procedure. The patients with exocrine pancreatic insufficiency of maldigestive state is sometimes severe and causes to death. To clarify the mechanism of fat maldigestion in patients with exocrine pancreatic insufficiency, we analysed upper small intestinal contents and investigated the effect of pancreatic enzyme on maldigestion. Lipase activities, pH, and micellar lipids in the upper small intestinal contents were studied after intragastric infusion of test meal in 9 Japanese patients with exocrine pancreatic insufficiency and 8 healthy subjects. As the same. way, we infused the patients test meal with pancreatic enzyme. The upper small intestinal pH was slightly less in patients with exocrine pancreatic insufficiency than healthy controls. Lipase activities and micellar lipids were significantly decreased in patients with exocrine pancreatic insufficiency compared with those in healthy controls. There was a significant correlation between serum cholesterol and micellar cholesterol concentrations.
In micellar lipids, the rate of monoglyceride was decreased and triglyceride was increased. Lipase activities and micellar lipids, especially monoglyceride were increased after the administration of pancreatic enzyme to the test meal.
The results suggest 1)that in patients with exocrine pancreatic insufficiency there is maldigestive state due to decrease in lipase secretion, 2)that insufficiency of lipase secretion disturbs hydrolysis of triglyceride, prevents micelle formation and leads to decrease in uptake of cholesterol into micellar phase, and 3)that such decrease is  always performed. Advanced aged (>75 years), poor general conditions, regional metastases, bilateral involvement of the hepatic ducts beyond the secondary branches, involvement of the main trunk of the hepatic artery, bilateral involvement of the portal vein branches combination of vascular involvement to one side of liver with the extensive chelangiographic involvement on the other, were considered contraindications to resection. RESULTS On the basis of the above criteria, 10 patients out of 78 considered suitable for resection of the tumour. At laparotomy, resectability confn'med in patients. Local excision of hilar turnout was performed in cases; local excision plus liver resection was performed in other left hepatectomy required in 2 Type mb cases and patient underwent right hepatectomy plus excision of segment because there local disease spread suspected to this segment. In all all loco-regional lymphatics and areolar tissue excised. The morbidity and mortality rate 62% and 12% respectively. Macroscopic and microscopic tumour clearance was obtained in patients. The overall survival was 21 months (range 3-73); 4 patients still alive, disease free, 3,16,21 and 39 months respectively. For the remaining 70 patients considered unresectable for cure stenting by endoscopic mean the only palliative treatment performed but in 8 cases PTBD was associated to complete endoscopic intrahepatie drainage. Most patients had placement of plastic endoprosthesis; in 41% of patients drainage obtained with endoprosthesis, 54% of required 2 endoprosthesis while only in 4% of cases stents placed.The morbidity rate and the direct procedure related mortality of 14% and 1% respectively with survival of 10 months (rangel-24). CONCLUSION: from the present study is justifiable to conclude that long term survival and potential cure can be achieved, in selected eases, by radical surgical resection. Endoscopic insertion of single multiple endoprostheses for patients unsuitable to surgery is safe procedure and provides good palliation. Gastric surgery increases the risk of cholelythiasisis due to impaired gallbladder motility secondary to denervation and/or loss of the coordination with gastro-duodenum motility. Erythromycin, a macrolide antibiotic with powerfull prokinetic effect on the GI tract has been shown to promote gallbladder emptying in normal subjects and in those with gallstones. We investigated the effect of oral Erythromycin on gallbladder emptying in 15 patients (7M, 8F, median age 63y) subjected to total (8) or subtotal (7) gastrectomy for cancer with a median follow-up of 18 months and free of disease. Five healthy subject, were considered as control.
After an overnight fasting, the gallbladder volume (GV) was measured before and h after 500mg of Erythromycin and before and h after a standardized meal. The GV was calculated ultrasonographically using the 3 Dodds' formula.
A wide variation in GVat fast was seen, particularly in totally gastrectomized patients. After meal, the GV decreased significantly in both groups of patients: from 36.4+21ml to 23.5+22ml (p=0.025) in patients with total gastrectomy and from 43.4+_12ml to 19.1+13ml (p=0.018) in patients with subtotal gastrectomy. In the controls the GV decreased from 25.2ml +_ to 6.5+_2ml. After Erythromycin the GV decreased clearly in controls(from 23.3+/-10ml to 16+/-5.6ml), while in patients with total gastrectomy did not change significantly from 41.8+/-31ml to 38.8+23ml, and in patients with subtotal gastrectomy increased from 31+/-7ml to45+17ml (p=O.06). Two patients in the group of total gastrectomy and one in the group of subtotal gastrectomy had developed gallbladder stones after gastric surgery. Gallbladder motility is only partially impaired by the parasimpathetic denervation after total or subtotal gastric removal for cancer since the gallbladder is still able to empty in response to physiological stimuli. However it does seem to become insensitive to the effects of oral administration of Erythromycin within one hour.  The routine of intraoperative cholansiogram during laparoscopic surgery is still controversial. Its in laparoscopic surgery at institution has been useful in identifying bile duct stones in patients who had dinical, laboratory ultrasonographic evidence of choledocholithlasts. We have performed this retrospective study in effort to identify the incidence of "silent bile.duct stones" Methods: We have reviewed hundred and twenty patients that underwent laparoscopic cholecystectomy since August 1990. The inclusion criteria for "silent bile duct stones" l)the absence of clinical features(jaundice, cholangitis) 2) Normal laboratory (liver function test) 3) and suggestion of choledocholithiasis (stone dilation) by ultrasound. Patients having any of the above indicators recorded in the category "probable bile duct stones" and excluded for the purpose of this study.

MG-Brazil
Cholecystectomy through laparoscopy has become the first-choice for procedures of gallbladder and bile duet because of the rapid recovering of patients, which may be associated to less anesthetic-surgical trauma. We compared endocrine response in 30 patients undergoing operative procedures for eholecystectomy using standard technique with subcostal incision (n 16) and videolaparoseopy (n 14) through plasmatic dosages of e'ortisol and adrenocorticotrophic hormone (ACTH) during surgery and at early postoperative period. Results showed that plasma concentrations of cortisol and ACTH showed no significant differences between the two techniques during surgery and at the first 6 hours postoperatively; only after 12 hours postoperatively there was a trend to significance (0.05 < P < 0.10) relating to cortisol, being greater in the standard group. Analysis of different periods in each group showed significant increasing of cortisol postoperatively compared to initial values for both groups; and ACTH showed significant difference after extubation and 12 hours postextubation for standard group; for videolaparoscopy group it was evident 12 hours postoperatively.
Because these results shows that plasma concentrations of ACTH and cortisol were not significantly different between both groups we concluded that neuroendoerine stimulation induced for both procedures is similar. Some elements are -likely to be involved and may be responsible for the rapid recovering of patients undergoing cholecystectomy through videolaparoscopy. Also, pneumoperitoneum and videosurgery were not different from standard procedure relating to the endocrine changes studied. Surgery and endoscopic otomy can both be proposed as detlnitive treatmeat for patients with common bile duct stones. floweret, the choice between the two procedures has not yet been well establislg4. The aim of this randozed trial was to compare sursery and endoscopic sphintcrotomy for common bile duct stones, with Weaal ra' to opeve and short term results, From 1989 to 19o.4, 204 patients (64 men and 140 wollll mes/l 67 *_ iS yer', raa 2S.90) were indud in the stu. One and -ven patients were opetttd on and 97 undawa endoscopic sphincterotomy, Before treatment, th two Stoups ofpatients were not silFdflca17 different as resards for rnc al sex ratio, ASA score, 'cvious ebole, cystectomy, jauntily, choiis, or Vancrtitis. In the surgew group, ston(C)s were found in 72 % of csses, 10.: % of the psfieats had a negative exploration of emnmon bile duct, In the spincterotomy lpoup, llla was een/n all instra. Common bil(C) duct etsm was pore'hie in 9 %, anti sphinctotomy n 82 % of cas. Four post operative dths were reportl: on (1%) in the surgery group and 3 (3 %) in the phkerotomy tFoup. Medi hospital st,xy was 6 days al 12 day. respectively (n$). P, nidual stones were diagnosed early after treatme.t in 8 patiet of surgery p'Otlp (7 %) and in 19 patients of spb/ncterotomy roup (20 %) p<0.02,. Early opemio, was nec in :Z patient of ,,8ery 8ro,p (2 %) and in IS pents of sphincterototw IFoup (20 %) -p<0.000} Although operative rnorudiry was not sisnificantly differem between the two procedures, surggry allowed a sijptificant lower ro0pe--mtion and residual stone rates than endoscopic sphincteotomy. This study xhowed that surgery should be prdered to endoscopic sphincterotomy, which should be restricted to patients with a sibmificant operative Introduction. We have previously observed that: I) multiple black gallstones (GS) are frequently associated with an increase in number and depth of the Rokitansky. Aschoff (R-A) sinuses of the gallbladder (GB) (1): II) black mieronuclei can act as initial nuclei for the formation of multiple cholesterol GS (2). The aim of this study is to analyse by scanning electron microscopy (SEM) black pigment microstones found within R-A sinuses and the pigmented center of the multiple cholesterol GS which ,ere concomitantly pres( in the same gallbladder looking at microstructural differences or analogies. Material and Methods. During the prospective study of 168 consutive patients who had systematic stone and bile analysis and histologic examination of GB wall, 32 patients were found with adenomyomatosis (ADM), i. e., an increase in number and depth of R-A sinuses, and black stones alone or in association with single cholesterol (n 7) or multiple mixed (n= 4) stones. In 4 patients with intraparietal black microstones and multiple mixed Cholesterol GS,. SEM analysis of stones and GB specimens were performed. In addition to x-ray diffractometry, x-ray fluorescence elemental analysis of stones also was performed in 2 of these cases. Results. SEM analysis of GB specimens has demonstrated that: 1) both black pigment microstones within the R-A sinuses and mixed cholesterol GS in the main lumen of GB can grovth concomitantly in the same GB; 2) black pigment microstones are mostly built up from granules of calcium bilirubinate rather than glassy masses of bilirubinate. SEM analysis of the pigmented center of multiple cholesterol GS showed the presence of a nidus of granules of calcium bilirubinate alone or in association with calcium phosphate in all the observed stones. The presence of calcium phosphate and other calcium salts in the pigmented center of mixed cholesterol stones was also demonstrated by x-ray fluorescence elemental analysis. Conclusions. It is suggested that: i) both black microstones and mixed cholesterol GS can groWth together in the same GB; ii) microstructures of black With the advent of, laparoscopic cholecystectomy (LC), the optimal management ofcommon bile duct stones remains controversial. AIM_.._d We report our experience of selective. ERCP and intraoperativc cholangiography (IOC) in the management of common bile duct stones in a large series of LC from a single center. METHODS 1847 consecutive LC pefform&l .from 1990-1995 were analyzed in terms of ERCP and IOC involvemznt. A high likelyhood for risk of CBD stones was considered an indicatiot for pro-operative ERCP and was defined as either presence of bilirub >2mg%, alkaline phosphatase (ALP) >lSOU/I, present/recent jaundicedpancreatitis or dilated CBD/stone on ultrasound or CT-scan. Selective IOC was performed for intermediate risk based on either bilirubin 1.5-2, ALP 110-150, ALT/AST greater than twice normal or remote history of jaundice/pancreatitis. Post-operative ERCP was performl in patients with suspected retained stones or bile duct injury. RESULTS Pro-operative ERCP was performed in 143 (7.7%) of patients; was successful in 141 (98.6%) and demonstrated CBD stones in 43 (30%) which were successfully extracted. Of 36 patients with mild gallstone pancreatitis, stones were found only in 6 (17%). Selective IOC was performed in 87 (5%) and stones were found in 4 (5%). Postoperative ERCP was performed in 66(3.6%). Bile leaks were found in 21 (32%), stones in 20 (30%), ductal ihjury in 3 (4%), papillary stenosis in 3(4%) and pancreatic duct stricture in 2 (3%). Bile leaks, injuries and stones wore all managed endoscopically. Complications were pancreatitis in 6 (4.2%), bleeding in 2 (1.4%), fever in (0.7%) and all settled with conservative treatment. CONCLUSIONS Even in selected patients considered likely to have CBD stones, the positive diagnostic yield of preoperative ERCP is low. Mild gallstone pancreatitis is associated with a low incidence of CBD stones. There is a higher incidence of positive findings with post-operative ERCP following selective IOC. Objective: In order to identify intrahepatic bile ducts and their anatomical modalities and to comprehend accurately the extent of bile duct cancer in each segmental duct includ!ng caudate branches, we attempted to make 3 dimensional images of intrabepatic bile ducts which were reconstructed from MR-cholangiography using maximum intensity projection (MIP). Materials and Methods: 8 patients with hilar bile duct carcinoma underwent imaging with a MR imager (Magneton H15 Siemens, Erlanger, Germany) and a surface coil. A turbo spin echo pulse sequence (8000 msec/ 91 msec TR TE) was used for data acquisition, with 35 seconds breath holding, 280 mm field of view, and a 120 256 matrix. These images sections were processed by using a standard MIP algorithm to obtain views of the entire hepatic biliary tree. Result: Subsegmental bile ducts and their modalities could be visualized on MR-cholangiography. Some caudate branches which could not be visualized on conventional cholangiography because they were located behind the hepatic hilum could be visualized on MR-cholangiography.
Other caudate branches which could not be filled of contrast medium because of the invasion of cancer also could be visualized. increase (at least x2) oftotal bilirubin, ALT, AST, GGT and alk. phosp.; US evidence of CBD or intrahepatic stone/s and/or a CBD size >7mm; Endoscopic cannulation of the Bile Duct was successful in every patient. Fifty-two patients (i.e. 54% of those undergoing ERCP and 12.4% of the whole group) underwent ES and extraction of stones and/or of biliary sludge was executed in all patient but one. There was no mortality and the morbidity rate was 0.2% (one case of mild acute panereatitis). All these patients subsequently underwent LC at a mean interval time of 11.2 days (min. 0 days, max 150 days), lntra Operative Cholangiography was never performed. CBD injury, intra/postoperative major morbidity and conversion rate were in 0.2%, 0.7% and 5.8% respectively. More than 95% of all the patients have been followed up by a self-administered postal questionnaire or telephonic interview. 0.4% (1 case of CBD and case of intrahepatic stone) is the residual stone rate after a median follow-up of 21.4 months (range 1-44 months). These results support the safety and the efficacy of the sequential treatment. The high number of negative preoperative ERCP is counterbalanced by a very low rate of residual CBD Up to date it's unclear laparoscopic surgery determine less immunosuppressive effects than traditional laparotomic procedures. We determined in siries of 38 patients affected by symptomatic gallstone disease and operated either by laparoscopic (Group 1) and by traditional open surgery (Group 2), the changes of lymphocyte subpopulations in the postoperative, compared to preoperative period, different time point .up to day 30 after surgery. We collected 15 ml of blood from all patients in both group and 2, day-1 (1 day before surgery) and postoperative day 1,7,15 and 30. A single blood sample from control group (Group 3) formed by 56 healty volunteers obtained. In the patients submitt to open cholecystectomy we observed significant fall in total lymphocyte count in postoperative day 1. Basal levels of lymphocyte subpopulations did not show any statistical significant difference (Wilcoxon test) between study and control groups (results with P<O,O1 considered significant). No differences were found in preoperative lymphocyte cell count between the two groups submitted to cholecystectomy. Pan cell (CD3) showed marked statistically significant reduction throughout the observation period. The count of helper (CD4), suppressor (CD8) and NK (CD16) T cells reduced postoperative day 1, NK cell (CD16) count remaining low until posfoperative day 30. B lymphocytes group showed postoperative reduction. In patients submitted to laparoscopic cholecystectomy significant postoperative fall of total lymphocytic count, CD3, CD4 and CD8 subpopulations observed day only. No reduction of CD16 and CD19 subpopulations was noted. A comparative statistical analysis between lymphocyte subpopulations inthe two groups was carried out: in the open cholecystectomy group compared to laparoscopic group, CD3, CD4, CD8 and CD16 lymphocyte subpopulations showed marked reduction different time points. In particular, statistically significant differences were found in CD3 levels from postoperative day 30, in CD4 from day through day 7 and in CD8 and CD 16 only day 1. Material and Methods: We performed a revision of 37 bile duct carcinomas surgically resected in the Department of Surgery, Bologna University, from 1982. The 37 carcinomas were classified as middle (5 cases, 13,5%), lower (11 cases, 29,7%) and hilar (21 cases, 56,7%) bile duct carcinomas. In each case we evaluated the following clinico-pathologic variables: age, sex, location of primary tumor, serosal invasion, peritoneal dissemination, hepatic metastasis, lymph-node metastasis, pancreatic invasion, duodenal invasion, microscopic vessels involvement, perineural invasion, resected proximal and distal margin involvement, histologic type depth of cancer invasion and survival. Results and follow-up: Lower bile duct carcinomas: All but two pts died for neoplastic progression (mean survival 19,6 months), two pts are alive and free of disease ms= 90 months). Middle bile duct carcinomas: 3 pts died for neoplastic disease (ms= 15,3 months) and 2 pts are alive (ms= 96 months). Hilar bile duct carcinomas: 2 pts died after surgery, 10 pts died for neoplastic disease (ms= 11,8 months), 4 pts are alive at 13 months, 2 pts died at 51,5 months (ms), and 3 pts are free of disease at 56 months (ms). The main clinico-pathologic factors that seem to correlate with .prognosis are location of primary tumor, size of tumor, depth of cancer invasion, histologic type, lymphatic or perineural invasion, lymph-node metastasis, hepatic or duodenal diffusion, and pancreatic metastasis.
Conclusions: On the basis of this retrospective revision morphological features seem to be closely related to the clinical behaviour of bile duct carcinoma, and they can be evaluated only by the mean of an accurate intraoperative staging. The ,'tim of this study is to elucidate the role of extended lymph node dissection for advanced gallbladder cancer. Forty-one consecutive patients with advanced gallbladder cancer(tumor extended more than subserosal layer) were underwent extended lymph node dissection combined with hepatectomy and resection of bile duct. The range of lymph node dissection covered N1, N2 of UICC and para-aortic region. Lymph node metastasis found in 28 out of 41 patients(%). Metastatic rate of each regional lymph node were 29%(cystic node), 45.2%(pericholedochal node), 45.2% (posterosuperior pancreaticoduodenal node), 25.8%(retroportal node), 19.4% (around the common hepatic artery). 8 out of 28 patients underwent paraaortic lymph node dissection. Jumping metastasis(means N1 negative and N2 positive) was observed in 2 cases. Cumulative survival rates of patients underwent curative resection were significantly higher than non-curative cases. Overall survival rates at 1, 3 and 5 years were 87.1, 62.4 and 32.8% respectively. There was statistical significant defferrence between the survival rates of patients without metastasis and with N2 metastasis. There which recurred in para-aortic reNgn in N1 positive patients. On the other hand 7 out of 11 patients with N2 and para-aortic nodes metastasis recurred in para-aortic region. Recurrence sites were out of the whitch dissected previously such as retrocaval area and hilus of left kidney. Six out of 11 patients with N2 and para-aortic nodes metastases survived than 2 years. We conclude that extended lymph node dissection is benefitial for precise estimation of tumor extention and prolonged survival periods. Methods: MRCP using a respiratory-triggered multi-slab threedimensional fast spin echo sequence was performed in 37 consecutive patients referred during an 18 month period because of an unsuccessful ERCP (n=20), presence of post-surgical biliary-entedc anatomy (n=10) or evidence of complete pancreatic duct obstruction on ERCP (n=7). MR examinations were acquired in the coronal plane on a 1.5T system (Philips ACS Gyroscan II) using the following parameters: TR 2500-5000 msec, TE 240 mSec, field of view 240 mm, or 2 signal averages, matrix 186 x 256, echo train length 31, slice thickness 2 mm. Eight or 10 slabs were obtained as part of the multislab acquisition with 5 or 6 slices per slab (total slices 40 to 50). Subsequent course and impact on clinical management was determined.
Multiple findings were present in 7 patients. Based on MRCP results, patients subsequently underwent laparotomy (n=l 1), therapeutic ERCP with precut papillotomy (n=3), therapeutic PTC (n=2), diagnostic PTC (n=l) or ultrasound-guided biopsy (n=l). The 11 patients with normal findings on MRCP required no intervention. The remaining 8 patients had abnormalities detected on MRCP but were followed clinically. Conclusion: MRCP plays an important role in the management of patients when ERCP is unsuccessful or incomplete and in patients in whom technical difficulties can be anticipated. Failed ERCP represents one of the main clinical indications for performing MRCP. Very low calorie diets (VLCD's) have increasingly been used in the treatment of morbid obesity; however, an increased risk of gallstone formation has been reported during rapid weight loss. Impaired gallbladder motility is considered important pathogenetic factor for gallstone development. Since diet composition modulates gallbladder contraction and most VLCD's are characterized by low fat content, this study aimed at evaluating whether V1LT)'s with low and higher fat content might influence gallbladder motility and consequence modify the risk of gallstone formation during weight loss. Sixteen obese gallstone-free subjects (4 males, 12  After transhepatic catheterization of a segment II or III bile duct, the left lobe of the liver and the lesser curvature of. the stomach were perforated under fluoroscopic and laparoscopic guidance using three trocars. Anastomosis between the biliary tree and the stomach was maintained with a gastrostomy tube placed across the tract. After 2 weeks, the tube was removed and patency of the tract was preserved with a metallic stent. Two-thirds of the patients had a hilar level of obstruction, and 65% of patients had an unresectable cholangiocarcinoma or pancreatic adenocarcinoma during laparoscopic staging. Onefourth of patients palliated also had biliary obstruction due to metastatic, colon adenocarcinoma, or gastric adenocarcinoma. The total bilirubin fell from 271 to 32 (p<0.001) in less than four weeks. The mean hospital stay was 17 days. After a follow-up period of 47 months, the mean survival was 7-months with 35% of patients surviving more than 12 months. Two patients died of septicemia and pneumonia in the hospital. Early complications were cholangitis (3), subcapsular hematoma (2), and gastric outlet obstruction (1) The magnitude of metabolic response to injury has been shown to be proportional to the degree of magical trauma. For this reason many investigatior have henceforth tried to find ways of reducing the metabolic response to surgery. Laparoscopic surgery is a minimal invasive procedure becoming the operation of choice in many field of surgery. In this study of patients undergoing open cholecystectomy (OC) and laparoscopic cholecystectomy (LC) we aimed to determine, whether or not there are any indications of postoperative inflammatory reaction by measuring serum concentration of Creaktive protein (CRP) and polymorphonuclear leukocytes (PMN) elastase as the most sensitive biochemical markers of the granulocyta stimulation. An unselected number of patients (pts) were investigated and divided into two groups. Group I. contained 15 pts operated by OC; Group II. involved 15 pts with LC. The indication for surgery was chronic cholecystitis in all cases. Anaesthetic and operative procedures were standardised in both groups. Venous blood samples were taken hours before surgical intervention and every day for 4 days following the operation. CRP level was measured by immonological assay (CRP-latex a88utination slide test), the PMNelastase vas measured by heterogeneous enzyme immunoassay (Reagent kit, MERCK) to specific determine its complex formatione with cl-proteinase inhibitor.The postoperative CRP concentration rose markedly following OC, and reached peak of 84,1+!4,32 ng/ml on the second day. This elevation was hishly significant (p<0,001), while in the LC group only a moderate increase to 52,7+17,70 ng/l was shown, which was not statistically significant. The level of CRP in the OC group remained higher during the whole observation period, while in the LC group it decreased to the starting value 19+4,93 n8/l by the 4th day. Our results owed significant PMN-elastase elevation in both groups on the fn-st postoperative day (135,1:L-9,1 n8/l, p<0,001; 150,5:E24,8 ng/l p<0,05 respectively), ttowever there was marked difference between the two groups by the 3rd postoperative day. While the value of PMN-elastase in the LC group decreased considerably (! 16,5:E20,9 ng/l) it maintained a high level in OC group (164,6+18,9 ng/l; p<0,01). Following the 4th postoperative day we still found an elevated enzyme level in the OC group (! 52,8-9,3 ng/lg p<0,0 ), while in the LC group a continuous fall towards the baseline value was noted (87,2+!0,5 ng/l).These results provide us with biochemical evidence supporting the clinical observation that LC is far less traumatic to the patients than OC. 100 surgical repairs without stenting were performed for benign bile duct strictures (Bismuth 3-5 types) from 1987 through 1994. There were 78 women and 22 men with ages ranging from 23 to 69 years. The most common cause ofthe stricture (85,7% ofthe patients) was the iatrogenic trauma during cholecystectomy or -resection ofthe stomach. Manycf our patients had undergone previously 3, 4 and 5 operations on the bile ducts, and suffered from long-term jaundice and cholangitis (69% ofthe patients). 25 patients had the stricture below the bifurcation ofthe hepatic duct( less then cm), 49 patients had the stricture ofthe confluence, 19 patients had biductal stricture, and 7 patients had monoductal stricture. Performing high mucosa-to-mucosa biliary enteral anastomosis, we tried to free the hepatic ducts above the stricture.

EXPERIENCE OF THE
In our report we shall present some van'ation ofthe "platform of the duct branches" for biliary enteral anastomosis. No fatal outcome was noted and the 17 complications were: subphrenic abscess case, temporary biliary fistulae 6 cases, intraabdominal bleeding 3 cases, gastrointestinal.bleeding 6 cases, intestinal liens case. Two patients had been. operated over the long-term follow-up period from 6 months to 9 years): recurrent stricture case, intrahepatic cholangiolitias -1 case. Our experience shows, that in benign biliary (type 3 5) stricture thorough dissection of the bile ducts above the stricture permits precise anastomosis to be constructed without stenting. A retrospective anonymous survey was conducted in Belgium in a multicentric group of general surgeons about 9959 LC. The incidence of BTI was 0.5 %.
The prevalence is 1.3 % below 50 cases of experience compared to 0.35 % (p < 0.0001) over 50 cases of experience. 36% of these BTI have been performed by a surgeon with more than 100 cases of experience. Predisposing local risk factors were present in 61%. Severe BTI occured in 42 %. Intraoperative detection of BTI occured in 44 %, mainly related to the performance of intraoperative cholangiography. In patients with biliary peritonitis, the overall mortality and late biliary stenosis were 19 % and 38 % respectively, compared to 4 % and 12 % in patients with intraoperative detection of BTI. Hospital mortality was related to the occurrence .of biliary peritonitis, while late biliary stenosis was related to the occurrence of postoperative biliary complications after initial biliary repair.
In conclusion, mortality and morbidity of BTI during LC are high. Intraoperative detection mainly with the use of IOC is an essential factor to improve outcome. Non-advanced gallbladder carcinoma (GC) usually remains unsuspected at operation and diagnosis is made postoperatively by the pathologist. However, the entire organ is not included for pathological examination in a macroseopically normal gallbladder.
Aim: To. analyze the value of bile cytology in the diagnosis of gallbladder carcinoma. Methods: Bile samples obtained intraoperatively by fine needle aspiration (FNA) from 311 patients were studied (99 men, 215 women; mean age 65 years, range 27-95). All the patients were operated either by laparotomic or laparoscopic approach for gallbladder disease (mainly cholelithiasis). Ultrasonography was suspicious of GC in 4 patients. Sample A was obtained through aspiration of 5 mL of gallbladder bile. Sample B was obtained after infusion of 5 mL of physiologic saline in the gallbladder lumen and subsequent aspiration. Early cytologic diagnosis was done intraoperatively (Dif-quiek stain), but definitive diagnosis was delayed until a Papanieolau technique was performed. Routine pathologic examination of the gallbladder was carried out by the pathologist without previous knowledge of the cytologic result. Results: Cytologic examination was positive for malignant cells (GC) in 11 patients. The 4 cases of preoperative suspicion of carcinoma were-confirmed., There was a false positLve ina case-of chronic reactive dysplasia. The remaining 6 cases of GC were deemed chronic or acute cholecystitis by surgeons at operation. No differences were observed between bile samples A and B, neither between Dif-quik and Papanicolau methods. Pathologists were unable to identify 2 cases of GC in situ in the first microscopic examination. The diagnosis was made in a second evaluation after the cytologic results were disclosed. Conclusions: 1) Intraoperative bile cytology obtained by FNA of the gallbladder allows an early diagnosis of GC; 2) Routine pathologic examination of the gallbladder can miss,a small number of GC cases that can be diagnosed by bile cytology; and 3) lntraoperative bile cytology is indicated when macroscopic appearance of the gallbladder suggests GC. Between January 1985 and June 1995, 16 patients with hilar cholangiocarcinoma were treated by external or/with intraluminal radiation therapy (ILRT). The mode of radiation therapy were: external radiation only in 6 patients, external + ILRT in 9 patients and another one ILRT only. The clinical status of 16 patients with hilar carcinoma are: unresectable or unsuitable resection in 9 patients, post-resection residual tumor in 3 and pos,toperative recurrent cancer in another 4. Radiation effects were assessed on the basis of clinical respone, cholangiographic change, choledochoscopic finding and survival times. Clinical response after radiation are: subsided ofmucinous substance from bile duct in 3 patients, general condition deteriorated in two patients and other eleven patients with stationary condition after radiation. Post-treatment. cholangiogram revealed patency of bile duct, decrease size or diminished the filling defect ofbile duct in six patients (37.5%). Post-treatment choledochoscopic finding revealed nodular or papillary tumor necrosis with decreased tumor size in six patients (6/13;. 46.1%), tumor necrosis only in 4 patients and dutal fibrosis in another 3. The survival times of 16 patients with hilar carcinoma after complete radiative therapy are ranged from 3 to 112 months (mean 29.8M, median 15.5 M) with l-y, 3-y and 5-y survival rates are 56.2%, 37.5% and 25% respectively. These data suggest that patients with multiple cholesterol gallstones have significant increases in gallbladder bile transferrin, percent transferfin, total iron ,and percent iron saturation. We conclude that an alteration in iron metabolism may be linked to the pathogenesis of multiple cholesterol gallstones. Stones in the gallbladder with inflammation or carcinoma make bilio-biliary fistula (BBF) or confluence stone (CS) in some cases. Although the differentiation of malignant and benign lesions in the gallbladder is important for a rational treatment, it is not easy to achieve a definite diagnosis due to interference of the stone in or around the fistula. To assess the effects of bile endotoxin on the pathophysiologyof acute cholangitis (AC), we investigated changes in the levels of bile endotoxin and serum cytokines after biliary drainage for obstructive jaundice with or without AC. Patients who underwent percutaneous transhepatic cholangiodrainage (PTCD) for obstructive jaundice were classified as those with AC (group A; n 5); those with a history of AC (group B; n 4); and those without a history of.AC (group C; n i0). The concentrations of bile and serum endotoxin, the serum inflammatory cytokines interleukin 6 (IL-6), and IL-8, and IL-I receptor antagonist (IL-Ira) were measured before PTCD, 5 h after, and i, 3, 5, 7, 10, and 14 days after. Endotoxin was assayed by the Endospecy method. The levels of IL-6, IL-8, and IL-ira were measured with kits for enzyme-linked immunosorbent assays. Results for serum endotoxin were positive (at above the cut-off value of pg/ml) in patients in group A, one patient in group B, and 2 patients in group C. The median concentration of bile endotoxin was higher in groups A and B than in group C. Serum endotoxin levels, if high, decreased to below the cut-off value after PTCD. Bile endotoxin levels in group C, if high, had decreased by 5 hours. This concentration ingroup A, if high, decreased slowly or stayed high after PTCD. In patients with poor drainage, bile endotoxin decreased only transiently. Intestinal flora including Escherichia coli and Enterococcus were isolated from the bile in all patiehts in groups A and B. The median serum levels of IL-6 and IL-8 before PTCD were higher in group A than groups B and C. These levels in group A increased after PTCD, peaking before 1 day, and then decreased. The serum levels of IL-6 were significantly correlated to the concentration of bile endotoxin at 5 hours. The median serumlevels of IL-ira before PTCD were higher in group A than in groups B or C. Increases in bile endotoxin and serum cytokines may be important in the pathophysiologyof AC. Biliary drainage eventually decreases the concentrations of these Background: Prolonged biliary obstruction results in bile duct proliferation and disruption of the hepatocyte tight-junction at the level of the bile canaliculi. Aim: To evaluate the respective roles of biliary obstruction and intrabiliary pressure in passage of bacteria from the biliary tract to the bloodstream and lymphatic system. M&M: 37 male Wistar rats underwent distal common bile duct (= CBD) ligation or a sham operation. After 2 weeks a laparotomy was performed and the CBD, the caval vein and the thoracic duct were canulated. Next, a broth containing 108 bacteria/ml of a specific pathogenous E.coli strain was retrogradely infused in the CBD at 5 or 20 cm H20 above baseline biliary pressure. After 5 minutes perfusion, blood and lymph samples were collected for quantitative culture analysis. Subsequently the liver was fixed for light microscopy. The infused E.coli could be visualized in the liver sections by immunohistological staining with specific monoclonal antibodies. Results: A higher biliary infusion pressure resulted in more colony forming units E.coli per ml blood in both the sham operated rats (5 cm H20, n 10, mean 1.99 x104 vs 20 cm H20, n 9, mean 11400 xl 04 ;p 0.0015, MannWhitney(MW)) and the bile duct ligated rats (5 cm HO, n=9, mean 3.5 xl05 vs 20 cm HO, n=9, mean 330 105; p=0.034, MW). Bile duct obstructed rats showed, at infusion pressures of 5 cm HO, more bacterial reflux to the bloodstream as compared to the sham operated rats (n 10, mean 1.99 104 vs n =9, mean 35 x104 ;p=0.0092, MW). However, at 20 cm HO infusion pressure, there was no significant difference between the two groups (n 9, mean 1.14 108 vs n 9, mean 33.2 108 ;p 0.7, MW). At 20 cm H20 infusion pressure, 2 of 9 lymph cultures were positive in the sham operated group and of 9 in the CBD ligated group. None of the lymph cultures showed growth at 5 cm H20 infusion pressure in both groups. Conclusion: This study confirms the increase of translocation of bacteria from biliary tract to bloodstream with higher intrabiliary pressures. Increased translocation, however, is also present at low intrabiliary pressures after a period of biliary obstruction. Bacterial migration to the lymphatic system is of no major significance in the early phase of bacterial infusion in the biliary tract. We have adopted the radical resection for the patients with gallbladder carcinoma according to strategy, corresponded to modes of spread of the carcinoma since 1985. We present results of prospective study. A total of 101 patients who underwent the resection from 1965 to 1995 divided into non-protocol group (A) (n=60) and protocol group (B) (n-41). The age of the patients 61 years-old in both group. In group B, intraoperatlvE. ultrasonography adopted for the diagnosis of the tumor spreid in depth of the wall and the invasion into the liver according to TNM system. In the protocol, wedge resection of the li.er with reglonal lymphadenectomy for Tlb, wedge resection and duct resection with regional and paraaortic lymphadenectomy for T2 adopted. A segmentectomy of the liver including $5 and S4b added to the with the direct invasion into the within 20mm. A pancreatoduodenectomy (PD) performed the with regional nodal involvement diagnosed by the froze] biopsy.
In pT2 patients, 10 of group A died from recurrence (8 local and 2 distant hepatic metastases), whereas only died form distant hepatic metastases in 14 of group @. Th 5year survival rate of pT2 tumr 21% in group A, and 88% in group B (P=0.006). In pT3 tumor, the 3 and 5-year survival rates 23% and 15% in group A, and 38% and 28% in group B, respectively. Six of 8 patients with recurrence in group B died from the blood-borne disease; distant hepatic (4) and lung (2) metastases without any recurrences at the hepatic resection margin. stents. RESULTS:Stent placement was technically successful in all patients. None of procedure -related complication was confirmed. Thirty day mortality was identical ior both groups(0). 25 weeks survival rates were 60 of polyurethane -covered and uncovered metallic stents. Stent obstruction prior to death or last follow up occurred in 4/10(40Z) of uncovered metallic stents with a median time to obstruction of 24 weeks and 0/11(0) of polyurethane-covered metallic stents. Purpose: To evaluate the mode of lymphatic spread of the advanced gallbladder carcinoma and to determine standard lymph node dissection are the purpose of this study.
Method s: The resected specimen, intraoperative findings and prognosis in 107 patients who underwent radical surgery were studied clinicopathologically using modification of the pTNM stage of AJCC. Primary tumor was classified into threestage of pTl(N=16), pT2(N=46)and pT3-4(N=45). Paraaortic lymph node belonged to pN2 group. Results: The frequency of nodal involvement (0% in pT1, 48% in pT2, and 73 % in pT3-4) was significantly different among three groups based on the stage of primary tumor. Distribution of the lymph node metasases was more widely in pT3-4. The paraao rtic node metastasis was recogn ized as a 12% in patients with pT2 and a 23% in patients with pT3-4 and other clinicopathlogical findings were positive in most patients with pT3-4 and some patients with pT2. The patients with pN2 metastases had significantly poorer survival (16 % of 5 year) than those with pN0 (65 %) or pN1 metastases (55 %). However, 5 patients with pN2 metastases survived more than 36 months. Conclusions: Understanding the characteristics of the mode of lymphatic spread is very helpful for achieving the approp date dissection of the lymph nodes. Dissection of regional lymph nodes including the paraaortic lymph node is recommended for T2-4 carcinoma ofthe gallbladder. for hilar cholangiocarcinoma is difficult due to invasion of perihilar soft tissues and adjacent main vascular structures and extensive metastasis to regional lymphnodes. But high postoperative mortality and morbidity rates have been reported with such an extensive operation, and the operator should select patient and determine reasonable extent of resection. In this study, we analyzed the short-term results of hepatic resection for hilar cholangiocarcinoma and tried to establish rational preoperative assessment to determine resectability and extent of resection. From October 1993 to October 1995, 15 patients had undergone hepatic resection for hilar cholangiocarcinoma. CT and cholangiography were done as preoperative assessment of tumors.
In some patients, angiography and/or choledochoscopy were added. In 13 patients, jaundice had relieved preoperatively with percutaneous transhepatic biliary drainage. In 4 cases of Bismuth type IIIA, 3 extended right lobectomies and rigfit lobectomy were performed and in 8 cases of Bismuth type IIIB, left lobectomies were done. In 3 cases of-Bismuth type IV, 2 left lobectomies with portal vein resections and left lobectomy were carried out. Among these 15 cases, 13 caudate lobectomies were combined. Early postoperative complications developed in 7 cases including case with arterial bleeding and case with portal vein thrombosis, which were managed operatively. Others were controlled with conservative managements and all the complicated cases had improved without any sequalae. To summarize, hepatic resections were performed in 15 consecutive patients with Bismuth type III and IV hilar cholangiocarcinomas without operative mortality. In 2 cases, proximal resection margins were postive microscopically and in another 2 cases, disease recurred at 10 and 14 postoperative months, respectively. In case, anastomotic site obstruction developed as a late complication and reanastomosis was performed without evidence of recurrence. All the patients are alive(mean follow-up of 13 months) and disease free survivors(13/15) have good quality of life.
In conclusion, lobectomy or extended lobectomy with caudate lobectomy is safe for most hilar cholangiocarcinomas with Bismuth type III and even Bismuth type IV in selected cases, and more extensive surgery for advanced hilar cholangiocarcinoma should be considered whenever resection is feasible. Surgery in patients with obstructive jaundice (OJ) is associated with higher morbidity than in non-jaundiced .patients, due to increased susceptibility to endotoxin (LPS) resulting in the inflammatory cascade. Different interventions have been studied to reduce endotoxemia and cytokine induction and the resulting complications. Bactericidal/permeability increasing protein (BPI) is a naturally occurring endotoxin binding and neutralizing protein, released from the primary granules of neutrophils. It binds endotoxin, neutralizing the activity and therefore inhibiting cytokine production by mononuclear cells. In animal studies and in healthy human volunteers BPI has a protective effect in experimental endotoxemia.
The aim of this study is to determine if BPI can protect against the increased endotoxin sensitivity in rats with OJ and, by that, reduce mortality.
Male Wistar rats were used, weighing approximately 250g. A dose of 2.0 mg/kg intraperitoneal E-coli 0111 B4 LPS was chosen, given week after Sham operation or bile duct ligation (BDL). Three groups of rats were studied Sham, BDL with saline, and BDL with recombinant BPI21 (recombinant 21 kD protein).
Conclusions: Intraperitoneal recombinant BPI21 treatment in BDL rats reduced the endotoxin induced mortality from 75% to 8%, a mortality rate comparable to that in non-jaundiced rats. BPI could be an interesting perioperative treatment possibility in olinical OJ. A retrospective study was carried out to define definitive criteria for choosing the most appropriate treatment for each type of polypoid lesion of the gallbladder (PLG). The shapes and sizes of PLGs were evaluated using ultrasound in 82 patients who had undergone surgery. Histologic examinations showed cholesterol polyps in 55 patients, adenomas in 9, cancers in 16, an inflammatory polyp in and a hyperplastic polyp in 1. The diameters of 58% of the benign PLGs were less than 10 mm, whereas those of 88% of the cancers were more than 10 mm; 78% of the former were pedunculated and 56% of the latter were sessile. 7 of 8 early-stage cancers had diameters less than 18 mm, whereas those .of all 8 advanced cancers were greater than 18 mm. 5 of the 8 early-stage cancers were pedunculated, and 6 of the more advanced cancers were sessile. Cholecystectomies with or without full-thickness dissection (removal of the entire connective tissue layers of the gallbladder bed to expose the liver surface) were main surgical procedures used to resect benign PLGs and early-stage cancers, whereas cholecystectomy with partial liver resection was used for more advanced cancers. Laparoscopic cholecystectomy was performed in the recent 42 patients, 4 of whom had early-stage cancers. In 17 of the 42 patients, the procedure with full-thickness dissection was performed. 8 patients with early-stage cancer and 5 with more advanced cancers were alive with no signs of recurrence after respective observation periods of 1.8 to 17.5 years and 1.8 to 16.5 years. In conclusion, a PLG with a diameter of less than 18 mm is a potential early-stage cancer and therefore can be resected by laparoscopic cholecystectomy with full-thickness dissection. .However, when cancer invades the subserosal layer or beyond, a second-look operation is necessary. A PLG with a diameter of greater than 18 mm may be an advanced cancer and should be removed by cholecystectomy with partial liver resection or a more extended procedure with lymph node dissection. Cytology despite becoming standard method for diagnosing the malignant nature of biliary strictures, has never been correlated with tumour type and differentiation. The aim of the study was to compare cytology results with tumour type and differentiation.
The study included 79 patients with biliary strictures (50 M, 29 F, median age 65yrs, range 19-85), who had both biliary cytology (92 samples taken at ERCP) and tissue biopsy for histology. Cytology ,eported as positive or negative for malignant cells. Tumour type and differentiation obtained by histology of resected specimen (n 30), percutaneous intraoperative biopsy (45) or post mortem examination (4). 23 patients had pancreatic, 29 bile duct, 20 ampullary and 6 gallbladder cancer. In case histology of resected bile duct showed no malignant, but cytological examination was positive for malignant cells. Cancers were graded as well (20), moderately (27) and poorly differentiated (1 carcinoma in situ, 9 differentiation not known). Overall sensitivity of cytology was 55 % (43/78) and positive predictive value 98 %. There was no association between positive biliary cytology and the degree of tumour differentiation. Sensitivity of cytology for well, moderately and poorly differentiated tumours was 65% (13/20), 52% (14/27) and 48% (10/21) respectively (chi square test, p>0.5). However, there was an association between tumour type and positive cytology (p > 0.02). Sensitivity for bile duct and ampullary cancer was 59 % and 80% for pancreatic and gallbladder cancer 30% and 50%. Sensitivity of biliary cytology depends on tumour type and is highest for ampullary and bile duct cancer, but unexpectedly not on degree of tumour differentiation.

60
Ifoperative eholangiography (OC) was a widely accepted and leasable method, there would be no need for a preoperative exam to aeertain the presence or absence of a common bile duet (CBD.) stone. In the era of taparoseopic choleeystetomy, as this exam Is not 100 % leasable and/or interpretable, preoperative endoscopic ultrasonograpy (EU) has been proposed to detect CBD stones. Because of its drawbacks (cost and need of general anesthesia), use of EU is still controversial.The aim of this multieentrie prospective study was determine the place of EU in patients for whom laparoscopie eholecysteetomy is planned. Patients operated on for eholelithiasis were selected on the basis of a preoperative score to be at risk of having CBD stones. Preoperative EU was performed within 10 days before the operation and compared with OC.-Presence of CBD stones on one of these exams was systematically confirmed (or infirmed) by operative exploration.
Two hundred and fifty patients were included in the study. EU and OC were feasab|e in 99 % and 91% of cases, respectively. Of 225 eases available for analysis, 206 were concordant (92 %) and .19 were. discordant (8 %). When both exams were in favor of CBD stones, the presence of stones was confirmed operatively in each case (n=45).
When both exams were not in favor of CBD stones, follow up.
confirmed the absence (n=161). In 12 instances, EU was in favor of CBD stones and OC was not 10 were EU false positive for CBD stbne diagnosis and 2 were OC false negative. In 7 instances, OC was in favor ofCBD stones and EU was not 6 were EU false negative for CBD stone diagnosis and was OC false positive. In our study t/EU s more otten feasable than OC 2/in pntients at risk of having CBD stones, performance of EU was very similar to OC 3/if any, its use should decrease with the feasabitity of OC in laparoseopic eholeeysteetomy. The choice treatment of malignant ampulloma is pancreaticoduodenectomy. Before engaging in such major surgery, some clinicians prefer to have histological confirmation of malignancy. In case of negative biopsies, surveillance, ampullectomy and endoscopic sphincterotomy have been advocated. Patients and methods 35 patients with adenocarcinomas of the Ampulla of Vater confirmed by pathological analysis of the surgical specimens (3 ampullectomies, 31 pancreaticoduodenectomies, total pancreatectomy) underwent preoperative endoscopic biopsy. There were 22 males and 13 females with a mean age of 62.5+/-9.5 years. Jaundice, abdominal pain, poor general health status and gastrointestinal bleeding were present in 19, 21, 22 and 8 patients respectively. Results Specimens showed protuberant and hemorrhagic papillary tumor (n 17), a pseudovillous tumor (n 2), an enlarged papilla (n 5), a common bile duct dilatation (n 4), a common bile duct nodule (n 4) and a papillary obstruction (n 1). Data was not available for 2 patients. 17 of 35 biopsies showed infiltrating adenocarcinoma, biopsy having been obtained only after sphincterotomy. Biopsies showed an adenovillous tumor suspect of malignant transformation, an unspecified suspect lesion and a positive smear in case each. Other biopsies showed mild (n 1), medium (n 2), or severe dysplasia (n 4), benign tumors (n 4), inflammation (n 1) and hyperplasia (n 1). No anomalies were noted in 2 patients. Conclusion In some patients preoperative endoscopic biopsies are capable of ascertaining malignancy for tumors of Vater's ampulla. However the possibility of malignancy should not be discarded in the presence of a negative biopsy and patients should be denied the benefits of resective surgery solely on the basis of a negative biopsy. The .place of liver resection in the management of Klatskin tumors remains controversial with some authors suggesting that liver resection should be the rule. We assess this attitude in the light of our personal experience with Klatskin tumors.
Patients and methods Between 1974 and 1993, 40 patients underwent resections for Klatskin tumors." The group comprised 23 males and 17 females with a mean age of 60 years (range 34-81 years). The majority of tumors were stage T3 (n 24) and 25 presented with type III biliary extension. The resectability rate was 42.5%. Surgical procedures included 11 tumor resections, and 27 combined tumor and liver resections. The latter included 7 extended right hepatectomies and in 8 cases resections were supplemented by regional vascular resections. There were also 4 liver transplantations of which 2 were preceded by organ cluster-type resections.
Results: There was no operative mortality among patients having undergone tumor resections, combined tumor, liver and vascular resections, or transplantation. There were 4 hospital deaths in the group having had combined tumor and liver resections. Following resections considered to be curative, the median time of survival was 23 months. When the site of the tumor was considered, mean survival was greatest for type II lesions (52 months). When TNM staging was considered, mean survival was 5 years for Tis and T1 lesions. Survival was as high as 26.7 months for type III lesions. When surgical procedures were considered, 5-year survival was excellent following tumor resection (27%) but was poor following liver resection (7%) although the latter increased resectability. Conclusions: Resectability of Klatskin tumors is increased by adding liver resection. Median survival time for "curative" procedures is 23 months. Mean survival for Tis and T1 lesions is excellent (beyond 5 years) and remains good for T3 and type II lesions. However, longterm results of liver resection are disappointing when compared to those achieved by simple tumor resections. There were 1200 patients of periampullary carcinoma who had laparotomy from 1962 to 1994 in our department. 320 cases of them underwent PD, the resection rate was 25.8%, hospital mortality were 23 cases,and the mortality rate was 7.19%, hospital morbidity were 69 cases, the morbidity rate was 21.9%, Of the 320 PD, 214 were male, 106 female, The ages of the patients ranged from 26 to 73 years .with a median of 52.9 years. The lesions of the 320 cases were that carcinoma in the head of pancreas were 81 cases, in common bile duct were 85, in ampulla of Vater were 104, in duodenum were 50. The method of resection we preferred is to follow the order of gallbladder, bile duct stomach proximal jejunum and" duodenum initially and leave the pancreas at last, this method provides excellent exposure of uncinate process and controls bleeding easily .Gastrojejunual anastomosis was retrocolic procedure The end of the jejunum is brought into upper abdomen in a retrocolic position, but anterior to the mesenteric vessels. Pancreatic fistula is a common and severous complication following PD. From 1962to 1970 PD with end-to-side anastomosis without pancreatic drainage were performed, the fistula occurred in 10 patients 6 deaths after that. So we change the method to end-to-end anastomosis between the pancreas and jejunum 14 patients with internal drainage by using of short tube ,fistula occurred in 2 cases, then we changed to a long catheter external drainage for 237 cases, fistula only occurred in 3 cases. PPPD were performed in 28 cases, gastric stasis occurred in 5 patients, it was cured by nasogastric drainage and drugs within a week. From our data shows that postoperative survival rate is poor for pancreatic carcinoma, one year survival rate is no more than half, and the three years is only 13.58%, but for carcinoma of ampulla of Vater and common bile duct, the result is better, five years survival rate for carcinoma of ampulla of Vater is 41.54 % which is the best. Renal dysfunction frequently occurs in patients with obstructive jaundice (OJ) and changes of the total volume and distribution of body water are probably involved in its pathogenesis and were detectedby invasive methodsz. Bio-impedence analysis BIA was reported as a reliable and non invasive method to assess the total bodywater TBW ), the intracellular (ICW) and the extracellular (ECW) distribution in normal and pathologic conditions, by measuring the body resistance (R) and reactance (Xc) 3,4.
This study was undertaken to determine the total body water and its distribution in patients with OJ Bilirubin > 200 mol/I for > week with no evidence of renal dysfunction Creatinine < 20 lmol/I ). Three repeated measurements were taken at steady state in 5 cancer OJ patients before treatment and in 5 control subjects matched for sex, age and weight. In conclusion, the mesurement of body water in OJ patients by BIA gave results similar to those obtained with invasive methods and confirmed the alterations of water volume and distribution in these patients.
Though laparoscoplc cholecystectomy has become the operation of choice for uncomplicated gallstones, minilaparotomy is shorter, without the disadvantages of the carbon dioxide pneumoperltoneum, and with a more rapid postoperative recovery compared with use of a standard incision (Surg.1994;1151533-9 and Br.J.Surg. 1992;79: 1061-4). The aim of this study is to compare the two methods evaluating respiratory and metabolic functions in patients (pts) undergoing electlve laparoscopic (group I =30) in the Giaveno Hospital and mlnilaparotomy (group II=30) cholecystectomy in the Giaveno and Asti Hospitals for symptomatic cholelithiasis. Hinilaparotomy was performed by use of the smallest feasible transverse subcostal incision (7-12 cm.) depending on the habitus of the patient. Arterlal blood gases" full blood count; serum cortisol" urinary vanillylmandelic acid (VHA), epinephrine (EN), norepinephrine (NEN) and cateeholamine (CCA); serum C-reactive protein (CRP), fibrinogen, erythrocyte sedimentation rate (ESR) and serum electrolytes (Na+, K+ and Ca++) were assessed in the preoperative and the immediate postoperative time and 24 hs afterwards. The data were analyzed by "t"student test. Blood gas data demonstrated a more significant decrease in arterial oxlgen pressure in pts of group II compared with those of group I, 24 hs later (P(0.000), reflecting poorer respiratory performance. Serum fibrinogen, electrolytes and cortsol dosages showed no significant dfferences in both groups, while urinary VA, EN NEN CCA were sJgnlficantly less for pts of group I P(O.4; P(0.04; P(O.01 and P(O.OOg,respect.). Also acute-phase responses were greatest in patients undergoing minilaparotomy as determined by ESR and CRP levels (P(O.01 and P(O.01, respect.). Horeover, the-decrease in hospltal stay after surgery, could be observed in the group I. These findings suggest that laparoscopic cholecystectomy may result in a reduced risk of postoperative complications. BACKGROUND: Cicatricial biliary strictures are usually associated with high morbidity and mortality rates, frequently related to technical difficulties of the surgical repair, mainly in the hilar lesions and those complicated with portal hypertension. Extended follow-up is needed to adequately evaluate results achieved with appropriate surgical repair techniques.
METHODS: The medical records of 45 patients surgically treated for cicatdcial biliary strictures between January 1984 and July 1992 were reviewed and the immediate and long term results retrospectively analyzed RESULTS: There were 34 females and 11 males. The average age was 42.7 years (11-72). The cause of the biliary lesion was: cholecystectomy in 18; cholecystectomy with duct exploration or reoperations for biliodigestive anastomosis in 25 and trauma in two. Thirty-seven patients. (82.2%) presented episodes of jaundice after the lesion and 31 (68.8%) presented cholangitis. Plasma bilirrubin levels were high in 31 patients. (68.8%) and alkaline phosphatase was elevated in 37 (82.2%). Diagnosis was possible by ecography in 25/33 cases with a sensitivity (S) of 75.7%, by ERCP in 10/11 (S=90.9%) and by transhepatic cholangiogram in 21/22 (S=95.4%). In 24 cases (53.3%) the stricture was located at the upper third of the bile duct, in 20 (44.4%) at the middle third and in one case (2.2%) it was low. All patients were submitted to Roux-en-Y hepaticojejunostomy with mucosa apposition. No transanastomotic stents were used. Six patients (13.3%) presented eight postoperative complications: biliary fistula (3); duodenal fistula (1) and wound infection (4). Average hospital stay was 10.8 days and there was no mortality. Only three patients developed secondary biliary cirrhosis, one with ascites, after a follow-up period of two to 10.5 years (average three years). There were no episodes of cholangitis in the late postoperative pedod.
CONCLUSION: Roux-en-Y hepaticojejunostomy with mucosa apposition without transanastomotic stent is a safe and efficient method for the surgical treatment of cicatdcial biliary strictures. Laparoscopic cholecystectomy (LC) is actually considered the treatment of choice for gallbladder lithiasis. Since 1991 many cases of inapparent gallbladder carcinoma (GBC) discovered after LC have been reported.

MANAGEMENT OF EARLY GALLBLADDER
Abdominal wall metastases, expecially at the port sites, and peritoneal metastatic diffusion have been described as complications of this new technique. From March 1990 to October 1995 11 cases of inapparent GBC have been observed: out of 1492 LC personally performed for cholelithiasis (0.2%) and 8 referred to our institution by other surgeons. At the time of the first operation patients were classified stade I, stade II and stade III. The median interval between LC and re-exploration was 211 days at this time 6 patients (2 stade 1,3 stade II and stade III) were submitted to a 4th-5th bisegmentectomy with radical pedicle lymphadenectomy and umbilical resection. The others patients (1 stade I, 2 stade II and 2 stade III) had unresectable tumor and received a palliative treatment: 2 had diffused liver metastasis, 2 had an invasion of the common bile duct, had a peritoneal carcinosis. Among the 6 resected patients one (stade I) developed a peritoneal carcinosis and died of recurrence 6 months after the hepatic resection. A second patient (stade II) developed a cutaneous metastases on the surgical limb. After the removal of this seeding he's alive without tumor recurrence 40 months after LC. 4 patients are alive 20,9,7 and 6 months after hepatectomy without recurrence. In the group with unresectable tumor median survival time was 8 months. We suppose that LC may be involved in the early appearance of abdominal and cutaneous recurrences: to confirm this hypothesis we have organized a multicentric group to study gallbbladder carcinoma and the consequence of laparoscopic procedure. At present, according to the literature and to our personal experience: 1. we avoid to perform LC if a GBC is suspected preoperatively; 2. we abandon laparoscopy if a GBC is diagnosed intraoperatively; 3. we perform a 4th-5th bisegmentectomy, a radical pedicle lymphadenectomy and an umbilical resection whenever a GBC is postoperatively In the hamster diisopropanolnitrosamine (DIPN) induces papillary hyperplasia in the biliary epithelium which progresses to eholangioearcinoma. In rodent models of colon cancer diets high in cholie acid,have been shown to promote carcinogenesis. However, the effect of dietary cholic acid in the pathogenesis of cholangiocarcinoma has not been investigated. Therefore, we tested the hypothesis that dietary eholic acid would increase the incidence ofbile duct hyperplasia in DIPN treated hamsters. Eight week old male Syrian Golden hamsters were fed either a control chow (CHOW) or a 0.5% cholic acid enriched (CA) diet. In each group, ammals underwent weekly subcutaneous injection with either normal saline (NS), or DIPN (500 mg/kg) for 10 weeks. At thirty weeks, livers were harvested, hepatic bile and serum were collected. The incidence of bile duet papillary hyperplasia, the liver function tests, and the percent ofglychocholic acid in hepatic bile were: Diagnostic laparoseopy and laparoscopic ultrasonography is increasingly used for staging of gastrointestinal malignancies. Recently many reports on port site metastases after laparoscopic colonreseetions and laparoscopic eholecystectomies fox occult galbladder carcinoma were published and induction ofport site metastases after diagnostic laparoscopy is also suggested. Therefore the prevelance ofport site metastases after diagnostic laparoscopy was assessed in this study.
Patients and Methods: All records ofpatients, who underwent diagnostic laparoscopy between January 1992 and July 1995 for staging of a gastrointestinal malignancy, were retrospectively analysed for the appearence ofport site metastases in the trocar-scars. Included were 250 patients; patients with an esophageal tumor (n=66), periampullairy tumor (n= 121), proximal bile duct tumor (n=26), a livertttmor (n=24) or other intra-abdominal malignancies (n=l 3). Results: Seven patients (2,8%) were lost from follow up. Four patients developed port site metastases, one with neuro-endocrine tumor, one with proximal bile duct tumor and two with pancreatic head tumors, respectively 2, 3, 5 and 10 months after diagnostic laparoscopy. Two patients had atypical, but no malignant cells in the peritoneal lavage fluid ofthe diagnostic laparoscopy and during the procedure biopsies were taken in 2 patients to prove irresectable disease. None ofthe 4 patients underwent tumor resection, the 2 patients with pancreatic head malignancies both underwent laparotomy with palliative bypass. These patients did not develop metastases in their laparotomy scars. All 4 patients were in an end stage of their disease and underwent only palliative treatment. One patient is still alive one month after detection ofhis port site metastases, the other patients died within months after development ofport site metastases. Conclusion: Port site metastases occured in 1,6% (4/250 patients) as late complication ofdiagnostic laparoscopy for gastrointestinal malignancies. Remarkable is the fact that no metastases were found in the laparotomy scars, indicating that not only dissiminated intra-abdominal disease but also laparoscopy related-factor, for instance the pneumoperitoneum, must be responsible for this phenomenon. Although the precise mechanism for this complication is unclear, the occurence in potentially curable disease should be prevented. Therefore during laparoscopy biopsies should only be taken to prove incurable disease and resectable tumors should not be biopsied.

F243 F244
Laparoscopic treatment of residual stones after cholecystectomy. Cholecystectomy carried out with the video-laparoscopic technique, today considered tested and codified operation, not only has the merit of having introduced method in the field of surgery, but, above all, ofhaving proposed way of interpreting and dealing with surgery: mini-invasive surgery or, it has been also called, "respectful" surgery, could in the next few years be increasingly performed and have wder and important application When bile stone is encountered after laparoscopic cholecystectomy, two options available to the surgeon: to perform ERCP-PTSE surgical treatment. In the period between February 1992 and October 1995, 1041 consecutive laparoscopic cholecystectomies performed. In four patients were found choledochal calculosis after surgical treatment and underwent ERCP choice that still consider optimal when is sufficient.
In two patients not possible to the stones by endoscopic tecnique and underwent the patients to laparoscopic choledochotomy. Direct laparoscopic choledochotomy, after laparoscopic cholecystectomy, requires very carefully dissection to define the anterior common duct wall, usually fibrotic tissue be found. Then, longitudinal incision is made in the duct for distance of about cm, and exploration proceed. At the conclusion of the exploration, T-tube is placed and the duct is securely closed with intracorporeal suturing techniques. It is racommended that small drain be placed in the region of Morrison's pouch to control any liquid collection. The patients had regular post-operative course; intracholedochal drainage removed 13 days aRer the operation, aRer x-ray check-up; this patient dismissed eighteen days aRer the operation. At present believe that the laparoscopic cholecystectomy be considered the