Posters

There is now a large data base incriminating the oxygen free radical system as an important mechanism of ischemic liver injury. The purpose of this study was to examine the effectiveness of mannitol over wellknown free radical scavengers in preventing liver damage from controlled hepatic ischemia and reperfusion.Regional hepatic ischemia was induced by occluding the vessels to the left and median lobes of the liver for one hour.The reperfusion was established by removing the occlusion at the end of the hour and the animals were sacrificed after 24hours of reperfusion. There were two control groups;sham-op. controls(n:7)and non-treated ischemic controls(n:8),and other four study groups;mannitol pretreated, superoxide dismutase ( "SOD" )treated, verapamil treated and SOD/Verapamil treated(n-9, 10, 10, 7).The liver ATP level was significantly higher(p 0.01) than ischemic controls, however it was indifferent compared to sham operated rats or rats protected by control scavengers. The plasma lipid peroxidase level was significantly lower in the mannitol pretreated group compared to ischemic controls(p O.01),.but similar to verapamil and SOD/Verapamil treated groups. We conclude that mannitol serves compatibly to other known oxygen radical scavengers in preventing the liver cell from toxic damage after ischemia and reperfusion injury of the liver.

There is now a large data base incriminating the oxygen free radical system as an important mechanism of ischemic liver injury. The purpose of this study was to examine the effectiveness of mannitol over wellknown free radical scavengers in preventing liver damage from controlled hepatic ischemia and reperfusion.Regional hepatic ischemia was induced by occluding the vessels to the left and median lobes of the liver for one hour.The reperfusion was established by removing the occlusion at the end of the hour and the animals were sacrificed after 24hours of reperfusion. There were two control groups;sham-op. controls(n:7)and non-treated ischemic controls(n:8),and other four study groups;mannitol pretreated, superoxide dismutase ( "SOD" )treated, verapamil treated and SOD/Verapamil treated(n-9, 10, 10, 7).The liver ATP level was significantly higher(p 0.01) than ischemic controls, however it was indifferent compared to sham operated rats or rats protected by control scavengers. The plasma lipid peroxidase level was significantly lower in the mannitol pretreated group compared to ischemic controls(p O.01),.but similar to verapamil and SOD/Verapamil treated groups. We conclude that mannitol serves compatibly to other known oxygen radical scavengers in preventing the liver cell from toxic damage after ischemia and reperfusion injury of the liver. The paper presented herein is to study characteristics of ultrastructural changes in the regenerating hepatic cells following hepatectomy. Healthy Wistar rats of either sex weighing 200-270g were used. Rats underwent partial hepatectomy were performed under anesthesia with sodium pen tobarbital and an amount of 40% of the liver tissue were removed. At time intervals of 10,15,20,25, 30 and 35 days post resection liver tissues were removed and prepared for electromicroscopy. The treatment group animals received ATP and 654-2 at the 3rd and 7th postoperative day. In another group animals were rendered cirrhotic by hypodermic injection of CC14 and feed with 5% alcoholic solution. Serial studies of ultrastructural changes of liver cells after partial hepatectomy revealed evidence of enhaced cellular metabolism namely increase and aggregation of roughened endoplasmic reticulum; swelling of mitochondria with accumulation of glycogen particles; and nuclear enlargement. But in the cirrhotic group ultrastructural pertrophy of collengenous f ibri liver cells. It also revealed t operatively can promote liver c therefore suggested that fracti and treatment with 654-2 and AT treating begign hepatic lesions paid in handling cirrotic patie changes were mainly hyals with little functionl hat 654-2 and ATP postell regeneration. It is onal resection of liver P may be beneficial in but cautions must be nts.
are argyrophilic (Ag-NOR). The results show an increase in the hepatocytic nuclei in the caudate and middle lobes in portacaval shunt rats in relation to control rats; the middle lobe increase is statistically significant (p<O.05). The Ag-NOR area is greater in the right lateral and caudate lobe in the control rats and this parameter increases in the middle and caudate lobes in portacaval shunt rats than in the control rats. The Ag-NOR area per nucleus is superior in the right lateral and caudate lobes in both the control and portacaval shunt rats. This parameter increases in the left lateral, middle and caudate lobes in portacaval shunt rats in relation to the control rats. The number of Ag-NORs per nucleus increases in all the hepatic, lobes after the portacaval shunt and this increase is statistically significant (p<O.05) in the middle lobe. In the parameters studied, the differences between the hepatic lobes of the control rats as well as their variations after a portacaval shunt make it possible to hypothesize the existence of a hepatic lobular functional heterogeneity in relation to the protein synthesis, ,hich changes after the deprivation of port.l flow produced by an end-to-side por(:a.caval shunt.
A temporary increase in arterial pH occurred 2 h after 70 % hepatectomy. Intramucosal pH of the GIT decreased from 2 h and on following 90 % hepatectomy, reaching statistical significance at 4 h. Intraluminal pH of the GIT increased 2 and 4 h after 70 % and 90 % hepatectomy, as compared with animals with sham operation. Systemic oxygen extraction increased immediately following hepatectomy. GIT oxygen extraction gradually increased in 70 % hepatectomized animals, while a decrease was seen following 90 % hepatectomy.
We conclude that major liver resection induces alterations in systemic and GIT homeostasis, which may contribute to explain disturbances noted in enteric bacterial ecology, the increase in bacterial adherence onto the intestinal surface and the increased translocation of enteric bacteria from gut following major liver resection in the rat. 90 % hepatectomy, while only 20 % to the portal vein. An increase in bacterial adherence onto the intestinal surface, damage of the permeability of the CMS and BTB, and pathological alterations in the ileum and colon developed, correlating with the extent of liver removed and the time that had passed following subtotal hepatectomy. Most translocating bacteria appeared in morphologically intact entemcytes, with increased membrane permeability, in the presence of antigen-preseming cells, and in the submucosal lymphatics, but also some bacteria were seen within damaged enterocytes from the 90 % hepatectomy-4 h group.
Our results indicate that an altered permeability of the CMS is one of the earliest changes in challenged entemcytes, and enteric bacteria translocate through both morphologically normal and abnormal entemcytes, mainly into the lymphatics, either being "carried" by antigen-presenting cells or actively invading by themselves. Intestinal transit time is an important indicator of gastrointestinal tract function. The relationship between enteric bacterial overgrowth and intestinal transit time following subtotal liver resection has not been elucidated.
In the present study, intestinal morphology, immunocytochemistry of the enteric nervous system, enteric bacterial growth in the small intestine and colon as well as intestinal transit time were determined in rats subjected to sham operation, 90 % hepatectomy or portal vein obstruction.
Histo-pathological alterations and E.coli overgrowth in the intestine were observed 2 h after 90 % hepatectomy, while a delayed intestinal transit was noted already 1 h following 90 % hepatectomy. Intestinal transit time was significantly delayed in rats subjected to 90 % hepatectomy as compared with both animals with sham operation and portal vein obstruction. No difference in the intestinal transit time was noted between rats with portal vein obstruction and sham operation.
It is concluded that the delayed intestinal transit following 90 % hepatectomy may contribute to enteric bacterial overgrowth and concomitant bacterial translocation from the gut.
EHEC was administered 1 and 12 h prior to 90 % hepatectomy in the rat. 90 % hepatectomy resulted in 80-100 % enteric bacterial translocation to MLNs or blood 2 and 4 h after operation while translocation did not occur in rats undergoing sham operation or in animals with 90 % hepatectomy and EHEC pre-treatment (p < 0.01). Bacterial overgrowth, increased bacterial adherence on the intestinal surface, as well as diminished intestinal and mucosal mass were observed in animals with subtotal liver resection alone, but not in those with enteral administration of EHEC. A delay in intestinal 2 h transit time was present in both groups receiving subtotal liver resection, with or without EHEC. EHEC inhibited bacterial growth and DNA synthesis, and altered bacterial surface properties following 1 h interaction with bacteria.
In conclusion, EHEC seems to alter enterobacterial capacities of metabolism, proliferation and invasion by direct effects on bacterial surface characteristics, and possesses atrophic action on the intestine rather than enhancing intestinal motility following subtotal experimental liver resection.
A colonic adenocarcinoma cell line (WB2054M) syngeneic to F1 hybrid rats was used to induce liver tumour. 4x10 6 cells were injected into the portal vein of 36 F1 rats. 12 rats were treated with octreotide (2/g bd) for 4 weeks and 12 treated with saline as a control. At 4 weeks the rats were killed and the Percentage hepatic replacement by tumour was calculated.
In the remaining group of 12 rats with liver tumour, tumour blood flow was determined before and after an intravenous infusion of octreotide (0.05 ug\min for 10 minutes) using a dual radio labelled microsphere technique.
The results of this study indicate that octreotide inhibits the growth of hepatic tumour derived from a colonic adenocarcinoma. Furthermore, octreotide reduced tumour blood flow, suggesting that this may, at least in part, be its mechanism of action in inhibiting growth.
It has been suggested that stimulation of hepatic reticuloendothelial system (RES) activity may be the mechanism whereby octreotide inhibits the growth of hepatic metastases. In order to investigate this hypothesis, further we have assessed the effects of octreotide and gadolinium chloride (GAD), a known inhibitor of the RES, on the growth and development of heptic tumour.
Two group of 12 rats (BDIX) received intravenous GAD (5mg\Kg) and 2 groups saline (controls). All animals then received an intraportal injection of lxl0 7 K12\Tr cells. Rats from each group received either octreotide 2/g, or saline s.c.b.d, for 4 weeks and the percentage hepatic replacement by tumour determined.
These results indicate that RES blockade with GAD significantly (Mann Whitney U p<0.01) increases tumour growth compared to controls.
Octreotide significantly inhibited tumour growth, but is more effective in the absence of GAD. It appears that RES activity is important in the growth of liver metastases. The efficacy of octreotide may be partly dependant on a functioning RES system, but other mechanisms are also operating. Departments of Surgery and Bioengineering, Royal Liverpool University Hospital, Liverpool, U.K.
Targeting of 5-fluorouracil (5FU) to the liver of patients with hepatic metastases may maximise tumour kill, minimise systemic side effects and possibly improve survival. The aim of this study was to compare the uptake of 5FU and a novel 5-FU disaccharide (galactose and fructose) adduct by normal liver tissue and hepatic tumour in rats.
Hepatic tumour was induced in BD1XZ rats by an intraportal injection of 1 x 107 K12/TR adenocarcinoma cells. Groups of rats with hepatic tumour received 5mg C-FU or C-5FU adduct (22mg) containing the same concentration of 5FU and blood samples removed at 10 min intervals. One hour after administration of the 5FU or adduct the rats were killed, the visceral organs removed, weighed and the radioactivity measured in a scintillation counter.
The hepatic uptake of 5FU and the adduct was not significantly different between the two groups of animals (p > 0.05 Student's t test). However, the uptake of adduct by tumour tissue was four times greater than 5FU (4226 + 190 v 1003 + 190 cpm/g; p < 0.001). Conversely the maximal blood level of the adduct (57.9 + 0.65 cpm/ml) was significantly less than 5FU (212.7 +__2.9 cpm/ml).
The results of this study indicate that coupling of 5FU to a simple carbohydrate containing a galactose moiety selectively targets the cytotoxic to hepatic adenocarcinoma and may therefore potentiate tumour kill and improve survival. tumors are still missed in early stages" efforts are made to improve the detection rate of HCC and thus its resectability.
MATERIA AND METHODS Over a 2 years period, 178 pts. were tested for AFP, serum ferritin (SF), tissue polipeptide antigen (TPA), carcinoembrionic antigen (CEA) and des-gamma-carboxy prothrombin (DCP) (E-1023, Eitest Mono P-II, Eisai Co., Ltd., Tokyo). Ninety two (51.7%) had HCC (of these 75 were cirrhotics) and the remaining 86 (48.3%) liver cirrhosis (24 pts.), liver metastases (19 pts.), benign liver lesions (19 pts.), non-HCC tumors (14 pts.) or others miscellaneous diseases (i0 pts.). Effectiveness tests were assessed for each marker and crosstabulated with clinical data to evaluate differences between groups. RESULTS AFP and DCP levels were higher in the HCC group (p<.05). TPA, CEA and SF were not helpful in discriminate HCC. The combined use of AFP and DCP improved the sensibility rate of almost 20% (Tab. I). University of Verona, Italy BCA and BCAC are rare,but interesting neoplasms of the liver for the possibillty of the former to evolve into mallgnant lesion.BCA is a slngle round multilocular tumor:the loouls llned by a mucin-secreting epithelium with papillary projections are filled by a mutinous fluid.The presence of a cellular mesenchymal layer supporting the eplthelium differentiates BCA in 2 groups:BCA wlth-BCMS-or wlthout-BCWMS-mesenchymal stroma.BCMS seems to more prone to undergo malignant change than BCWMS. BCAC shows thick papillary projections with invasion of the papillary stroma by neoplastic cells,ln a fashion that closely resembles degenerated pollps of the colon. The possibility of BCA/BCAC should be always ruled out in every cystic lesion of the liver:in every duobt leslon surgical exploration is indicated and at least a biopsy should be performed.
The complete excision of the lesion is the best therapeutical choice.Neoplasms can be excised with enucleation,but a correct oncologlcal procedure require wide resection in normal parenchyma since no BCA should be considered benign before thorough hystological sampllng.Only in older and poor risk patients not radlcal procedure,as in our case n.5,could be Justified. During hepatic resection in the cirrhotic liver a prolonged interruption of liver blood flow may produce different degrees of parenchymal damage; the safe limit of duration ofliver ischemia in cirrhosis has not yet been defined conclusively.
To minimize both intraoperative bleeding and biochemical disturbances after interruption of blood flow to the liver, the technique of selective hemihepatic vascular occlusion (HVO) has been developed.
AIM OF THE STUDY. To assess the variations of cirrhotic liver function occurring during and after HVO we recorded sequentially the biochemical parameters of hepatic function, including enzymes, serum bilirubin and acute phase proteins in five patients with liver cirrhosis who underwent hepatic resection for hepatocarcinoma with the HVO technique and in five patients with Pdngle manouver (PM).
RESULTS. Average duration of liver ischemia was 45 +_ 5 minutes for HVO technique and 32 _/ 8 minutes for PM. The biochemical alterations observed after HVO included an early postoperative increase of enzymes serum levels, which peaked on postop, day 2 and reversed by postop, day 6. A similar transient postoperative increase of serum acute phase proteins was detected, associated with a modest increase of serum bilirubin. In patients operated with PM, AST levels were significantly higher from the 1st to the 7th day; a similar pattern was recorded for bilirubin.
CONCLUSIONS. The data observed indicate that in patients with HCC in liver cirrhosis the technique of liver resection with HVO is well tolerated and causes modest and reversible changes of biochemical indices of hepatic function. In 39 pts (1986-88) we used Mitomycin C microcapsules while in 49 pts  the treatment was performed with fluid Mitomycin C in Lipiodol suspension with or without Gelfoam. 65 were males and 23 females with a median age of 61 years (range 27-79); 72 pts were Child A and 16 Child B; 59 and 29 were Okuda I and II respectively. The median diameter of the HCC was 6 cm (range 1-13 cm). 30 pts had a single neoplastic lesion, 51 multiple and 7 had diffuse HCC.
No related treatment mortality or major toxicity was observed: only minor complications were detected (ascitis, pain, temperature). Overall objective response was 30%: complete response in 8 pts (9%) and partial response in 19 pts (21%) with a median duration of response of 26 and 9 months respectively. Actuarial survival was calculated: overall 1-year and 2-year survival were 58% and 44% respectively. 2-year survival of pts Child class A was 51% while 2-year survival of pts Okuda stage I was 47%. Chemoembolization with Mitomycin C can be considered a safe and useful treatment for unresectable HCC in cirrhosis.
158 PO021 SURGICAL RFECTION FOR SMALL HEPATOCELLULAR CARCINOMA I Nagashima, S Inoune, T Nagao, N Kawano, T Muto The First Department of Surgery, University of Tokyo, Tokyo, Japan Surgical resection could be curative in selected patients with a single small hepatocellular carcinoma (sHCC), but the cirrhotic patients often have the risk of tumour recurrence or death from underlying liver dysfunction. Orthotopic liver transplantation (OLT) might be a rational treatment for those patients. We retrospectively reviewed twenty three patients with a single sHCC undergoing hepatic resection at our department between 1979 and 1991. 17 patients of 23(74 %) were cirrhotic and the other 6 had chronic hepatitis. Three (13 %) died from hepatic failure in hospital after surgery. One, 3 and 5 year survival rates of the other 20 patients were 90,79 and 61% respectively. As for prognostic factors influencing long-term survival, the presence of vascular invasion of the distance of free surgical margin (more than 1 cm or not) was not significant. The only significant factor was the severity of co-existing liver disease. One, We reviewed the pre-operative data on 20 cases of FNH (out of a sample of 41)and 8 cases of HCA (out of a sample of 16), investigations in all cases included US CT To-colloid and Tc-DISIDA scintigraphy, andblood-chemistry tests ( AST, ALT, ALP, BIL, CHE, GGT, AFP, CEA) The definitive diagnosis was confirmed by histolog of surgical specimens or,in the case of FNH only,by wide-wedge surgical biopsy.US proved particularly reliable in differentiating between the two formswhen the central searthe echogenicity similar to that of normal parenchyma, and a hypertrophic supply artery to the mass were clearly discernible; these elements however, were detectable only in 3 eases. Ct diagnostic reliability corresponded to 16/20 cases of FNH as against only 2/8 cases of HCA (with a misdiagnosis of FNH in I case and of possible malignancy in 5). In each case of FNH, To-toll revealed no changes in uptake of the lesion compared to the surrounding liver whereas Tc-DISIDA failed to show a hyperconeentration in the elimination phase in one case only. As regards HCA, Te-coll revealed a cold lesional area in 6 cases, a low-uptake area in one case and a normal uptake area in another. Tc-DISIDA showed increased accumulation of contrast medium in the elimination phase in one case only. The perfusional phases of the scintiraphic procedures were hypervascularized in all HCA eases and in 16/20 FNH cases.Laboratory test results were normal in all easesapart from increased GGT in 12/20 and increased ALP in 8/20 FNH cases.We observed no increases in serum tumor markers in patients with FNH and HCA.A correct pre-operative diagnosis was achievable in approximately 95% of FNH cases using imaging techniques alone,eliminating the need for surgery in the case of asymptomatic lesions.Any doubtful or suspect cases of HCA call for laparotomic diagnostic investigations and, in the case of evidence of HCA,surgical removal both on account of the haemorragic potential of the lesion and in order to obtain correct histological differentiation from a well differentiated HCC. showed an improvement in subjective symptoms and 14 (64%) had an objective improvement. Karnofsky score was improved in 8 (36%) patients while 9 (41%) remained unchanged. The survival rate was 64% at 6 months and 14% at 12 months. The median survival was 7.14 months (1 to 14 months). There was equal survival of cirrhotic and non-cirrhotic patients. From 1986 to 1992, a prospective study was done on 22 patients who were shown to have inoperable hepatocellular carcinoma on preoperative investigations. These patients were assessed to have good surgical risks, good liver function and no extra-hepatic spread of the disease. The tumours were inoperable because of extensive bilobar involvement of the liver. One patient was found to have an operable tumour intraoperatively and resection with curative intent was done instead.
Twenty one patients underwent cytoreductive surgery with liver resection, cryosurgery, microwave tissue coagulation and/or absolute alcohol injection. In-hospital mortality was 9.5%. One patient died of multi-organ failure 1 day after surgery because of reactionary haemorrhage which required a second operation to stop the bleeding, and another patient died 8 days after surgery because of liver failure. Other morbidity included intra-peritoneal abscess (5 patients), hyperbilirubinaemia which improved gradually (2 patients), bleeding from stress ulcers (2 patients), reactionary haemorrhage (1 patient) and chest infection (1 patient).
The symptomatic relief and quality of survival were excellent. The median survival of patients after cytoreductive surgery was 12 months and the survival was much better than those of 234 patients who received chemotherapy during the same period of the study (log rank test, p = 0.0003). There was no statistical difference between the survival curves of those patients who received (12 patients) and those who did not receive (7 patients) adjuvant chemotherapy 'or radiation therapy after cytoreduction.
We believe cytoreductive surgery contributed to the improved quality and quantity of survival in our patients. Hepatic Hemangiomas (HHe) are the most frequent benign lesions of the liver. Less than 1/3 of the patients with HHe are symptomatic, usually with the largest lesions referred to as "giant", (GHHe) (max diameter > 4 cm). From 12/01/87 to 12/1/92 fifteen patients, thirteen females (86,7%) and two males (13,3%), with GHHe underwent surgery: maximum diameter of the lesion was 9.3 + 2.3 cm. Mean age of patients was 50.6 years (range 31-69). Symptomatic patients were 86.7% (13/15 pts), their major complaint being pain in fight hypocondrium or epigastrium associated in 56.3 % of cases with non specific gastrointestinal symptoms. In 1/13 patient the lesion became symptomatic three years after diagnosis (initial max diameter: 6 cm preoperative diameter: 12 cm). Two/15pts with asymptomatic GHHE underwent surgery because of rapidly growing lesions. Forty percent of cases (six females) showed multiple HHe besides GHHe, their diameter being < 2 cm. Another patient showed four HHe in the fight lobe, two of which were GHHe (max diameter 5 and 6 cm). All patients underwent Ultrasonography associated with AngioCT (9 pts), with Angiography (2 pts), with CT + Angiography (3 pts), with MRI (1 pt). None of the patients showed abnormality of blood tests. In 66.7% of cases simple enucleation of the HHe was performed (10/15 pts), associated in three cases with colecystectomy and in one case with colecystectomy + right adrenalectomy (pheocromocytoma). In 33.3% of cases (5/15 pts) liver resection was performed: right hepatectomy (3/15 pts), left lateral segmentectomy (1/15 pt), fight lateral segmentectomy (1/15 pt). Mean blood loss (MBL) was 403.8 cc (15 pts). In the last nine cases the Pringle maneuvre was adopted, mantaining clamping for a mean time of 11 minutes (range 6-15). In this group seven enucleations and two liver resections (one fight lateral segmentectomy and one right hepatectomy) were performed. MBL were 155 ml and 400 ml for enucleation plus the Pringle maneuvre ( mean diam of GHHe 9.5 cm) and resection plus the Pringle maneuvre mean diam of GHHe 12.7 cm), respectively. Total MBL (8pts) was 231.2 ml. MBL in 3 pts undergoing simple enucleation without the Pringle maneuvre was 441.5 mlMBL in 3 pts undergoing liver resection without the Pringle maneuvre was 1080 ml. Mean hospitalization of all patients after surgery was 9.8 days. Postoperative complications included one case of biliary fistula after fight hepatectomy, which resolved spontaneously one month after surgery, and only minor problems (two fight pleural effusions) after enucleation (2 cases). No death has occured up to date.Surgical treatment of GHHe should be limited to symptomatic and rapidly growing lesions with high risk of rupture.Benefit of resection must be outweighed against risk of operative morbility and mortality. Surgical therapy must be therefore the simplest and safest one. Enucleation is performed by digitoclasia along the natural pseudocapsular plane, clipping and sectioning afferent vessels. No biliary duct is encountered in the right surgical plane. In our experience enucleation is the first choice operation. This procedure, especially when combined with the Pringle maneuvre, is very quick, allows minimum blood loss, and is associated with a very low incidence of postoperative complications.   have had no treatment or their metastatic liver disease except or symptomatic therapy. In the sameoperiod o time in the case o an other 18 patients I.V.600 mB/m/d Fluorouracil /5-FU/ monochemotherapy as estabilished or 5 days and repeated on every 28 days. This regiment as continued at least or six month or until sings o toxicity or death. No severe side eect as observed and reduction o dose ith 50% ere indicated only in to cases. From the 18 patients o the chemotherapy Broup one underent let sided hemihepatectomy and an another excision o to small liver metastases. In this retrospective non-randomized study the authors analysed the data o survival in the 5-FU and in the untreated group o patients, respectively. There has been no significantly differences between the age and sex distributions. From those cases having had no chemotherapy only two survived or more than 12 month, and their median survival as poor, 5 (range 2-15) month. In contrast in the chemotherapy group six patients were alive in the end o the study 6-24 (median 12) month ater the tumour spred to liver had been recoBnised. Moreover a prolonged median survival time 9 (range 2-24) month was observed ater 5-FU treatment. The difference in survival proved to be statistically significant (Student's t test, p<O.O1).
The authors emphasise the need o systhemic Fluorouracil chemotherapy instead o negligence to improve the survival time with better quality o lie or patients suering rom liver metastases o colorectal cancer. They also consider to conduct a prospective randomized trial on larger number o cases to analyse the superiority o Ca-olinate modulated 5-FU over 5-FU alone in this metastatic disease. CT-scan during arterial portography (CTAP) seems to be the best preoperative imaging technique for detection of liver metastases. Intraoperative ultrasonography (US) has been advocated to enhance the accuracy of preoperative evaluation. In our experience, intraoperative finger palpation of the liver when liver vascular exclusion (LVE) is performed detects with excellent accuracy liver metastases.
The aim of this study was to compare preoperative CTAP, intraoperative US, and finger palpation under LVE in patients operated on for liver metastases of colorectal cancer.

Patients and methods
From october 1990 to july 1992, ten patients were operated for liver metastases from colorectal adenocarcinoma; there were 7 males and 3 females, with a mean age of 59,3 _.+ 7,4 years (range: 43 to 70 years). Metastases were synchrone for four patients, and occurred for the others six at a mean of 27,5 + 13,4 months after initial surgery. All patients had a simple CT-Scan and a CTAP, 8 had a preoperative ultrasonography (US) and 7 had a magnetic resonance imaging (MRI). Intraoperative US was routinely performed all livem were palpated before liver vascular exclusion (LVE) and 5 were palpated once again after LVE.

Results
Intraoperative US when compared with CTAP detected :the same metastases in 8 cases, more metastases in 2 cases, three new metastases in 2 cases. Finger palpation before LVE was able to detect these three metastases. However, 2 metastases detected by CTAP and intraoperative US were not by finger palpation before LVE. Nevertheless, these metastases were found after LVE.

Conclusion
The CTAP shows a greater number of metastases than other preoperative explorations. In this study, there were no false positives, and 2 false negatives when compared to intraoperative US. We showed that CTAP and finger palpation before and after LVE, gave the same results than intraoperative US. Further evaluations are however needed. The specific role of hepatocytes and or endothelial cells on the metastatic cells adhesiveness and growth remain unknown.
After passages in vivo, in the rodent model a colonic adenocarcinoma line (DHD/K 12-PROb Prob-), two variants of cells were isolated The third selection tumour was chosen as the metastatic variant (Prob mp1); cells derived from spontaneous lung metastases bearing primary tumour isografts were reimplanted subcutaneously ("seed selection"). The second selection tumour was chosen as the liver affinity variant (Prob h2); cells derived from hepatic tumour obtained after repeated intraportal vein injections ("soil selection").
After intracaecal injections, only the see selection tumour permitted to obtain a local tumour with liver and pulmonary metastases. The adhesion and the growth of different cells on contact with endothelial or hepatocyte cell monolayer cultures were compared; it resulted that: 1 o) the growth of the tumoural cells were not dependent on the cell supports.
2) The adhesiveness of the seexl selection tumours decreases on contact with endothelial cells and the adhesiveness of the soil selection increases on contact with hepatocyte cells.
Comparison with data in the literature, our results obtained suggest that metastasis is a dynamic phenomenon and the cell adhesiveness varies with cell position to give a metastasis or not. However, the hepatocyte influence is important because the process of cell maturation along the acinus results in a matrix of varying composition. The liver is the most frequent target organ of metastases from carcinomas of the digestive tract and other organs. Possibilities of early detection followed by effective treatment improve the prospects for patients with liver metastases.
Tumours were synchronous in 4 and metachronous in 11 patients. Metastases were colorectal in 13 and non-colorectal (hypemephroma) in 2 patients. Of all the patients, 2 died in one month and another 2 in 3 months after operation. Theextentofsurgery was: hemihepatectomy, bisegmentectomy, segmentectomy, and extra-anatomical resection in 2,3,4 and 6 patients, respectively. In all of them extra-hepatic propagation of the tumour was excluded. Indications for surgery were: (i) absolute (peritonitis or hemoperitoneum after rupture of metastasis); (ii) relative (metastases f digestive tract carcinoma, endrine tumurs or hypernephroma); (iii) casional (sarcoma, seminma).
In future the following algorhythm of treatment in hepatic metastases will be used" 1) pre-operation cytostatic treatment to assess tumour sensitivity to cytostatics in vivo; 2) surgery (resection); 3) specific immunotherapy (production of autologous xenogeneic vaccine from tumour cells): 4) application of cytostatics to the arteda lienalis; 5) application of Interleukin 2 to the arteda lienalis. Mean longest single inflow period was 44.5 minutes, range 20-80. Mean operative blood loss was 1200 ml, range 150-1700 ml. 21 patients did not require a blood transfusion. In 15 patients receiving a transfusion, mean blood replacement was 2200 ml, range 500-7000ml. In 17 resections of <4 segments, mean blood loss was 692 ml of whom 2 required a transfusion. In 21 resections of 4 or more segments, mean blood loss was 1563 ml and 13 required a (1) and urinary tract infection (1). One patient died post-operatively of multi-organ failure and coagulopathy. Mean hospital stay was 13.5 days (range 5-34 days). Significantly more complications occurred in patients receiving a transfusion (4/22 vs 8/14, x 2 =4.2, p =0.03). No significant difference in the incidence of complications was found with regard to extent of resection (>4 segments 10/21 vs <4 2/15. x2=3.2, p=0.07) age or duration of ischaemia. Inflow occlusion is well tolerated and facilitates liver resection especially in nonanatomical resections. HOSPITAL "GREGORIO MARAON" MADRID, SPAIN The most common vascular problems after liver transplant occur in arterial anastomoses. However, portal and caval anastomotic stenoses have been also described, especially in children. Starlz el al published in 1984 a technical modification in small vascular anastomosis, the so called "grow factor technique" in wich a continuous monofilament suture is tied to "a considerable distance" from the vessel wall in order to permit a maximal distension in the anastomotic line (Starlz TE, Iwatsuki S, Shaw BW Jr: A grow factor in fine vascular anastomoses. Surg Gynecol Obstet 1984, 159:164-165). The distance to which the knot is to be tied has been difficult to determine in our hands. If the suture is tied too far away from the vessel, additional stiches have to be placed to create a watertight anastomosis. We describe here a technical variation to avoid pursstring effect and anastomotic stenosis. After a countinous monofilament polypropylene suture is placed, a second clamp is applied to the donor vessels. The proximal clamp is then released allowing the distension of the anastomotic line to its maximal diameter. If the suture is then. tied gently over the vessel, the blood pressure will avoid a pursestrig effect. From april 1990 to december 1992 92 adult liver transplant were done in our Liver Transplant Unit. The vascular patency was checked by sistematic echo-doppler or by angiography in cases of class III or IV malfunction, postoperative ascitis, gastroesophageal bleeding, or inconclusive echo-doppler studies. In the first 22 cases in wich the "growth-factor effect" was used we found two portal stenosis and an arterial thrombosis.
In the 70 following transplants, the two-clamps technique was applied to arterial and portal anastomosis and no vascular stenosis or thrombosis were discovered to date. As the study is not prospective nor ramdomized no statistical conclusion must be drawn, but we believe that this technique is a useful tool in termino-terminal anastomosis between vessels with elastic wall.
When orthotopic liver transplantation (OLT) was first considered for patients with hepatocellular carcinoma (HCC) the indications were those tumours deemed unresectable by conventional menas. These were large tumours associated with decompensating cirrhosis.
Thus it was not surprising that until recently OLT for HCC had been beset with high recurrence rates and poor results.
From past experience at King's we made a determined effort to transplant only cirrhotic patients with (HCC) <6 cm in size.
Between October 1989 to December 1991 20 such patients were transplanted. They were followed until December 1992, a median follow up of 30 months and a minimal follow up of 12 months.
The aetlology were HBV in 8, HCV in 4, Alcohol in 4 and one each of alpha-l-antitrypsin deficiency and cryptogenlc cirrhosis.
Seventeen patients were in Child-Pugh grade B or C, only 3 were grade A. the size of the tumours were <4cm in 12 patients, 4-6cm in 6 and >6cm in 2. Methods of bile duct anastomosis were reviewed in a consecutive series of 123 liver transplants performed in 118 patients. The 30 day mortality was 12/118 (10%). Mean survival was 13 months with a range from 1 day to 36 months. Methods of bile duct repair included gall bladder conduit (N 2), primary roux loop repair (N 12), direct end to end biliary anastomosis (Nffi 106). Indications were standard but PBC, Hepatitis B and C, alcoholic cirrhosis, cryptogenic cirrhosis and primary liver cell canceraccounted for 75 % of cases. 16/106 duct to duct anastomoses were splinted with a T-tube whilst 90/106 were not. Biliary leak occured in 1/14 (7 %) roux loop repairs, 4/16 (25 ) splinted duct to duct anastomoses and 10 of 90 (11 ) unsplinted duct to duct anastomoses. Strictures developed in 11% of unsplinted duct to duct anastomosis and no other group. Strictures occurred early and were successfully dilated ( 8 E.R.C.P., 2 PTHC). None have had to be dilated on more than one occasion and none have recurred. There is no significant difference in leak rate or stricture formation between any of the methods of repair. Direct duct to duct anastomosis is a satisfactory method of biliary repair following liver transplantation. Routine use of T-tubes to protect this anastomosis does not result in a significantly lower instance of either leakage or stricture formation. Retransplantation and biliary leak is best treated by roux loop repair. An independent factor affecting HBV recurrence was perioperative anti-viral prophylaxis. In fact, in pts with high anti-HBs post-OLT titer, the HBV recurrence rate was significantly lower than in pts without prophylaxis (28% vs 65%, p=0.0005). Furthermore, at a preliminary analysis the HBV recurrence in the transplanted liver seems to be primarily affected by pre-OLT viral replication.
, INCT  We present our experience with 4 patients who had OLT for SBC due to traumatic bile duct strictures. The aetiology of the primary bile duct injury in the 4 patients were; post-operative in 2, blunt abdominal trauma in 1 and shrapnel injury in 1. All four patients had failed biliary reconstruction and three had 4 or more attempts at stricture repair. All four suffered from persistent jaundice and had portal hypertension. One patient had recurrent haematemesis due to oesophageal varices and one suffered from chronic encephalopathy (Grade 1-2). Liver biopsy in all four patients confirmed the presence of SBC.
Surgery was technically difficult due to the presence of extensive vascular adhesions and lack of tissue planes due to the previous operations. Three patients are alive and well with normal liver function tests. One patient's postoperative course was complicated by multiple duodenal perforations and she died 3 weeks post-transplant due to multiorgan failure secondary to sepsis.
OLT is a useful option in the management of patients with established SBC secondary to recurrent bile duct strictures. Conventional surgery in these patients is associated with high incidence of postoperative morbidity and mortality. High potassium "intracellular" liver preservation solution is used to prevent ionic shifts during cold ischaemic storage. High sodium lactobionate solution (with raffmose) suppresses hypothermia-induced cell swelling and offers potential advantages in preventing potassium.
induced vasconstriction and hyperkalaemia on reperfusion. No prospective double blind, randomised clinical studies comparing these two solutions have been reported.
The aim of this study was to evaluate the efficacy of a high sodium liver preservation solution compared with the high potassium preservation solution used widely in clinical practice.
Thirty-six patients randomised into two groups were included in the study. Eighteen patients (nine male, nine female, mean age 44, range 22-62) were transplanted with a liver preserved with high sodium solution (Group I) and eighteen (eleven male, seven female, mean age 44, range 22-28) with a liver preserved using the standard high potassium solution (Group II). The indications for OLT were similar in both groups. The quality of liver preservation was assessed by post reperfusion liver biopsy, maximum serum concentration of bilirubin, AST, ALT and minimum platelet count within the first 48 hours post transplantation. In addition peal-operative blood loss and potassium requirements were assessed. Number of histologically confirmed acute rejection episodes were recorded and current graR survival was calculated. Mann Whitney U test was used for statistical, analysis and median values are shown.
Cold ischaemic times (705 mins vs 684 mins) were similar in both groups (NS, p >0.05) as were the time required to fashion the three venous anastomoses before reperfusion of the liver (50 rains vs 50 mins, NS p >0.05). Histological evidence of damage related to preservation was present in both groups. Maximum levels of bilirubin (109 tmol/l vs 112 pmol/l), AST (944 U/I vs 596 U/I), ALT (543 U/I vs 429 U/I) within 48 hours since transplantation were similar in both groups (NS, p > 0.05). There was no significant difference in minimum platelet count (94 x 10-9/1 vs 58 x 10-*/1, NS p >0.05). Number of acute rejection episodes was 2.5 (range 0-4) in group I and 1 (range 1-3) in group II (NS, p >0.05). Current graft survival was 16 in group I and 15 in group II.
We conclude that this study has not shown any significant differences between high sodium and high potassium solutions in preservation related graR damage and long term outcome of OLT. From December 1988 to July 1992, 123 liver transplants were performed in 118 patients at a single centre with a thirty day mortality ofl0 .T he preferred biliary anastomotic technique was end to end duct anastomosis without T-tube. This paper documents our experience with E.R.C.P. in post-transplant patients to investigate possible bile duct problems.
25 patients had E.R.C.P.s post liver transplant with 4 patients having two studies and 21 patients having a single study. All were performed on patients with an end to end bile duct anastomosis. Time post transplant was 5 560 days (median 30). The indications for liver transplantation in this group were similar to those for the whole series. 6 studies were performed as an emergency with the rest performed semi-electively on routine lists. The indications included clinical evidence of leakage or stricture, abnormal LFTs, evidence of cholangitis or ultrasound evidence of dilated ducts.
The results included failure (2); stricture (8); biliary leak (4); combined stricture and biliary leak (2) and a normal study in 9. 5 patients who had biliary leaks were converted to a roux loop anastomosis. 6 patients who had post liver transplant E.R.C.P. have died between 18 days and 16 months post-operatively and there were no complications causing significant morbidity or mortality following E.R.C.P.
E.R.C.P. is a reliable and effective way of diagnosing and dealing with biliary problems post liver transplant. The biliary tract anastomosis has been considered the "Achiile's heel" of the liver transplantation, with morbility rate between $-86% in the literature. The Choledocho-choledochostomy (CD-CD) with T tube is more frequently used; however, the complications T tube related are higher. At present, some patients may be benefit of the anastomosis without T tube. Between April 1986 to November 1992, 293 liver transplants (239 adults  The CD-CD diminishes the operative time, avoid T tube complications, had not morbi-moality advantages vs T tube use, may be delay radiological diagnosis of complications. The choledocho-cho|edochostomy is our first choice method of biliary reconstruction; the use of T tube in our hands is a more secure type of anastomosis. CAH.-Chronic Active Hepatitis (HCV) PBC Primary Biliary Chirrhosis Four patients had portal hypertension with variceal bleeding before the transplant. Tw patient had important ascitis (Tx i, Tx 5). Donor recipient matching was by size and ABO status. Standard operative techniques were used, including venous-venous bypass in 1 case (Tx I). Average lengths and ranges of donor liver ischemia, operating time and blood replact were 13:50 hours (range 6:18 24: 30 hours), 16: 36 hours (range 9: 00 25: 00 hours), and 24 units packed cells (range 6-45 units). Arterial reconstruction was termino-terminal in all cases. Biliary reconstruction was choledochocholedochostomy with T tube stent in 4 cases and Roux-en-Y choledochojejunostomy with straight tube stent in 1 case (Tx 5). Imunossuppression consisted of triple therapy with Azathioprine, Prednisona, and Ciclosporin. The overal biliary complications rate was 20% (i T tube migration Tx 2). One patient (Tx 2) with previous abdominal surgery had a ilium fistula 16 days posttransplant and required reoperation (ileostomy) and dialysis for persistent renal failure during the postoperative period. 57 days posttransplant (Tx 2) this patient presented hepatic necrosis in segment VII and VIII with rtant hepatic disfunction without hepatic artery thrombosis. The treat consisted in clinical support and antibiotic. One patient died (Tx 3) due to primary graft failure 2 days after the transplantation. The mean duration of Hospital stay was 32 days. Currently 4 patients are alive between 30 days and 15 months post transplant. The treatment of alcoholic cirrhosis by hepatic transplantation remains controversial. A balance must be found between the number of donor livers, the financial resources and the long-term benefit in survival quality of life and relapse of alcoholism. We report our experience in 16 patients (12 males, 4 females, mean age 43.5 years, range 35-58) with alcoholic cirrhosis who underwent liver transplantation between 1987 and 1991. There was a median period of follow-up of 8.3 months (range 0-34 months). All patients had a psychiatric assessment prior to transplantation and were considered preoperatively to be abstaining and suitable for surgery. Eight patients were Child A, 2 Child B and 6 Child C. There was one intraoperative death but no other hospital mortality.
At time of followup 13 patients remained alive, the three deaths all occurring in Child C patients. No patient had required a second graft. Two of the group had reverted to alcoholism on clinical and biochemical grounds. We conclude from these early results that in carefully selected patients with alcoholic liver disease hepatic transplantation is an effective treatment with reversion to alcoholism in less than 20 % of survivors.  The aim ofthis study was to analyze ;he significance and the mechanisms of ascites following orthotopic liver transplantation (OLT) in children. Patients and ,methods From 86 to 91,148 liver allografs were performed in 131 children, 15 (25 gratis) of whom were excluded because of early death or regraiting. Ascites was defined when ascitic fluid output was greater than 25 ml/kg/day (or >500 ml/day) and persisted at least 72 hours alter removal of the abdominal drains. Group I included 31 (25.2%) allograffs with post-operative ascites. Group II was the control group of 92 liver grafts without ascites.
(2) Post-transplant ascites is an exsudate with a predominantly lymphocytic cellular content and a low lipid concentration, originating from the liver allograff. (3) Severe pre-transplant hepatic insufficiency, supra-hepatic outflow block and early grat injury (primary graPt dysfunction, acute rejection) are the main mechanisms of post-transplant ascites. Since the description by Starzl et al., in 1987, of the rapid technique for multiple organ harvesting, it is generally admitted that minimal preliminary dissection of intraabdominal viscus is preferable for two reasons (a) it is associated with an improvement in graft function, and (b) it is better accepted by the hospital personnel hosting the procedure. We report here the results of a simplified technique of liver harvesting without in situ cannulation of the portal system. From 1989From to 1992 OLTs were performed in 88 patients. Nine grafts were excluded (2 cluster grafts, 3 reduced-sized grafts from a split procedure, 1 graft harvested using cardiopulmonary bypass and 3 imported grafts). University of Wisconsin (UW) solution was used for preservation in all cases. Two groups were compared. Group I included 20 grafts retrieved according to the Starzl rapid technique with cannulation of the aorta and the inferior mesenteric vein. Group 2 included 60 grafts retrieved with a simplified technique. The only step of the procedure specifically performed by the liver team required 5 minutes and included inspection of the graft for quality and arterial supply, and ligation of the cystic duct. The organs were only perfused by aortic route, with 3-4 liters of UW solution after cannulation of the distal aorta and clamping of the supraceliac aorta. Immediately after harvesting, the graft was perfused on the back Hepatic abscess-amoebic or pyogenic-can be diagnosed with great accuracy by either ultrasonography or computed tomographie (CT) scanning. Ultrasound is the modality of choice and will detect almost 100% of abscesses. Confirmation of a diagnosis of amoebic liver abscess is made by indirect haemagglutination test that should be positive in almost 100% of cases. Cultures of pus from the abscess and from the blood must be obtained in cases of pyogenic liver abscess. A positive culture of pus from the abscess has been achieved in 90% of cases. Ultrasound or CT guidance is utilised in aspiration of a hepatic abscess.
In the treatment of an amoebic liver abscess, Metronidazole is the amoebicide of choice. Open drainage is contra-indicated. For cases that fail to respond to therapy with amoebicides, closed drainage guided by CT or ultrasound is performed. Secondary bacterial infection of an amoebic liver abscess is an extremely rare event.
The identification and determination of the antibiotic sensitivity of organisms responsible for pyogenic liver abscess is a crucially imlrtant step. Unless a coeliotomy is necessary to correct an intra-abdominal process or the abscess is extremely large, the initial treatment of pyogenie liver abscess is a 3 week course of appropriate antibiotics followed by a 1 month course of oral antibiotics. The majority of pygenic liver abscesses will reslnd t such treatment. If drainage of a pyogenic abscess is required, the preferable technique is wi a rcutaneus CT-or ultrasound directed caeter. Otn surgical drainage should be reserved for those eases in which a eoeliotomy is required for oer purlses or for the patient who has failed a curse of appropriate antibiotic therapy and closed percutaneous drainage is not feasible.
We have experience with 18 patients. In all, 354 patients with liver hydatidosis were treated by surgical operation in our hospital between 1988 and 1991. One hundred and ten hydatid disease of these patients underwent CT scanning: 102 patients had cystic hydatid disease and 8 had alveolar hydatid disease.
The diagnosis of hydatid disease is based on: (1) a history of contact with dogs or sheep in prevailing areas; (2) absence of subjective symptoms in early stage of infection and the gradual occurrence of pressure symptoms with the slow growth of hydatid cyst: (3) the formation of liver abscess as a result of complicated infection of hydatid cyst and the development of acute abdominal pain and anaphylactic shock as a consequence of rupture of hydatid cyst; (4) the typical sign on palpation of hydatid cyst projecting under the liver; (5) the acoustic images displayed by ultrasonic detection: and (6) immunological diagnosis.
These procures, however, are difficult to find early pathological changes.
The use of computerised X-ray tomography in the diagnosis of hepatic echinococcosis not only makes it possible to diagnose an asymptomatic parasitecarrier but also detects accurately images of various specific pathological pictures such as solitary cyst, multiple cyst, daughter cyst, calcification of cyst, complicated infection of cyst, complicated rupture of cyst. Degeneration of cyst was revealed on the basis of bservations of morphological changes of clinical pathophysiology and of evolutional dynamics of various complications.
The pre-operative diagnostic accuracy rate in this series was 99.1%. Removal of the cysts left in the liver residual ectocyst cavities as large as the cysts and a bile-ntaining bluely effusion frming retention cyst, usually resulted in secondary infection, causing liver abscess.
The clinical observations and experimental studies of this series revealed the mechanism of intrahepatic biliary fistula formation and the natural rule of cavity closure, allowing rational cavity management and modification of operative methods: 2. 3. 4.

5.
closure of ectocyst cavity by suture partial resection of ectocyst with cavity open closure of ectocyst filling greater omentum closure of ectocyst inserting catheter drainage "Marsupialization" and "Roux-en-y" drainage were abandoned This reduced the frequency of post-operative biliary fistula and of secondary infection, and also shortened the healing period.
Endoscopy identified fresh blood in the second part of the duodenum in 7 of 10 occasions. A liver lesion was identified in 6 of 10 patients who underwent either CT scanning or liver ultrasound. Selective hepatic angiography demonstrated an intrahepatic bleeding source in 13 patients. An arteriobiliary fistula in the gallbladder in 1 patient was not identified by angiography. Selective hepatic arterial embolization using either gelfoam pledgers or Gianturco coils controlled bleeding in 10 of 12 patients. Embolization failed in 2 patients (1 with segmental liver necrosis required a right hepatic Iobectomy and a second patient underwent surgery and ligation of the left hepatic artery). Bleeding from the gallbladder in 1 patient was treated by cholecystectomy. Selective hepatic artery embolization was not attempted in 1 patient who underwent a left hepatic Iobectomy.
Selective hepatic artery embolization was successful in 10 of 12 patients (83%) of whom 1 patient developed subsequent complications.
Selective hepatic artery embolization provides definitive control of liver bleeding with a low incidence of complications and should be considered the primary treatment of choice for intrahepatic haemobilia. For many years hydatidosis has been considered a benign disease and therefore treated by conservative surgery nor alays followed by good results.
The biliary fistula is often a complication of the conservative surgery and involve an inadequate treatment because implies removal of the contents leaving peicystium in situ.
PATIENTS AND METHODS. Three females are presented in this poster (62, 6} and 7} year-old). They had been treated out of our Department, 12 years, 4 and 6 moth ago. The first surgery was conservative in all cases (2 tube drainage and I partial cystopeicystectomy) and the cyst was located in the right lobe.
Patients arrived to the emergency with fewer, acute colangitis and external biliary fistula. One also showed extracellular fluid volume deficit, metabolic acidosis and hipokalemia.
The diagnosis was completed by ECO, TAC and fistulography.
In all patients was necessary to complete the cystopericystectomy and the biliay fistuIa was isolated and ligated on hepatic tissue. Intaoperative studies with x-ray were done outineIy. In a patient a sphintero-pIasty was associate.
One patient died with sepsis and two emains asynto- The use of artificial conduits (Gorotex, Dacron) as interposition or bypass grafts in the portal venous system is associated with high rates of early thrombosis. The use of homologous veins (e.g. the internal jugular vein) adds a further operation to what is already a long procedure. Preserved iliac veins, harvested from organ donors, are used infrequently in graft recipients and so provide a ready supply of venous allografts for use in other circumstances. We have used such allografts in 6 patients, four for replacement of a portal vein resected en bloc with an invading cancer (pancreatic cancer 2, bile duct cancer 2, all with partial liver resection) and two for complicated cases of portal hypertension. The length of preservation time for the vein graft prior to insertion ranged from 1 29 days with a median of 6 days. No patient was placed on immunosuppression and all patients survived operation and left hospital. In two patients, the graft apparently functioned initially but later thrombosed. In the remaining patients the clinical course suggests that the grafts have remained patent. The longest follow up at present is 13 months. If venous allografts used without immunosuppression prove to retain patency over a long term then the implications for the use of such veins are many and varied. sphincterotomy, the patient remained asymptomatic for 5 years. In August 1990, he developed acute angiocholitis with ensuing septic shock and acute respiratory distress syndrome. At laparotomy, all superficial liver cysts were fenestrated in order to remove intrahepatic bile stones. Bile duct anomalies in the left liver lobe were minimal, but the idea of performing the fight hepatectomy was abandoned because of the small size of the left lobe and the existence of associated congenital hepatic fibrosis. Resection would almost certainly have led to acute hepatic failure and encephalopathy. The patient is alive and well 8 months post-operatively, but remains at risk of developing lifethreatening angiocholitis.
Caroli's disease, a rare condition with an incidence of 1 per million, is characterised by congenital dilatation of the intrahepatic bile ducts. The bile duct anomaly may, in 20% of cases, be limited to a liver segment or lobe, but in it's usual form, it affects the entire intrahepatic biliary tree. In the latter form, biliary drainage techniques have often proved ineffective in preventing recurring bouts of angiocholitis. Moreover, resection of the most severely affected liver parenchyma is seldom feasible because of associated congenital hepatic fibrosis or secondary biliary cirrhosis. In such cases, OLTX (orthotopic liver transplantation) may represent the only treatment, but optimal timing for this procedure has yet to be determined. In these typically young and otherwise healthy individuals, OLTX may be perceived by the patient and his practitioner, as an unacceptable burden. Later however, if acute angiocholitis resistant to antibiotics develops, the patient may no longer be suitable for liver transplantation. The improved technique for bloodless hepatic resection using the total hepatic vascular isolation under the normothermic or hypothermic condition were reported to deal with the large hepatoma and the severe liver trauma involved in the liver hilum, the main hepatic veins and retrohepatic inferior vena cava.
The original Heaney's or Fortner's methods were modified so that the improved techniques could be simpler and more practicable to perform otherwise hazardous liver resection.
Over 4 years, major hepatic surgery was successfully performed on 20 occasions with the normothermic or hythermic total vascular exclusion techniques in our department for patients with malignant and benign liver tumour or trauma.
Among 20 cases, 17 with the normothermic selective total vascular exclusion (extended fight lobectomy: 5 cases; extended left lobectomy: 2 cases; fight lobectomy: 5 cases; central segmentectomy: 3 cases; repair of the ruptures of the hepatic veins and the retrohepatic vena cava: 2 cases); 3 with the ttal vascular isolation on the in situ cold perfused liver (left trisegmentectomy 1 case, right trisegmentectomy'l case; extended left lobectomy 1 case). The indications and the clinical experience of the application of the techniques were discussed. They may increase the repertoire of the surgeon in the management of a variety of hepatic lesions. From 1986 to 1992, 31 benign tumors of the liver were diagnosed in 29 patients, 25 women and 4 men. These tumors were 15 hemangiomas, 12 focal nodular hyperplasias, and 6 adenomas.
Six of the hemangiomas were symptomatic, as well as 3 of the focal nodular hyperplasias and 1 of the adenomas. The diagnosis was assessed on behalf of imaging techniques: ultrasound (23), C.T. scan (26), M.R.I. (13), arteriography (7). A liver biopsy was performed in 9 patients. In 11 patients, such diagnosis data were sufficient to avoid surgery in asymptomatic hemangiomas and focal nodular hyperplasias.
The remaining 18 patients underwent surgery. Preoperative diagnosis was confirmed in only seven patients (39%): 4 hemangiomas out of 6, 1 focal nodular hyperplasia out of 8 and 2 adenomas out of 4. Surgical procedures were 5 major hepatectomies, 3 bisegmentectomies, 7 segmentectomies and 2 tumorectomies. This retrospective study enhances the difficulties encountered to assess an exact pre operative diagnosis in benign tumors of the liver. The present study is an account of our experience in the management of our cases. The methods of present-day diagnosis and treatment a re al so reviewed.
Fourteen patients (4 males and 0 females) were hospitalized for nonparacytic cysts of the liver during the years 986-99. The age rangeo!from 36-76 years (mean 58.4 years). Almost all patients complained of moderate heaviness in the epigastrium or the right hypochondrium. In 5 patients (35.7%) indigestion and a history of duodenal ulcer co-existed; patient had mild jaundice without fever. The blood tests, microbiological and biochemical examinations, mainly those concerninn liver function, were within normal limits exceptonepatient who had moderate diabetes mellitus. The laboratory tests for hydatidosis were negative. U1 trasonography, computerized tomography, retrograde cholan.oiography and liver scannin. were an essential diagnostic tool in 0 of our patients. In the remaining 4 patients, the cysts were an incidental finding. Localization of the cysts in 8 patients (57.2%) was in the left lobe and in 4 (28.6%) the cysts were found bilaterally.
In 9 patients (64.2%) the cysts were solitary and in 5 patients (35.8%) multiple; their diameter ranged from 0.5-7.5 cm. Radical cystectomy was carried out in 8 cases (57.3%), marsipulization in (7.1%), atypical left hepatectomy in (7.1%), capitonage in 2 (14.2%) and partial Lin-Chen-Wang cystectomy in 2 (14.2%). Cytology of the cystic fluid revealed a small number of RBC's, histiocytes, some elements of bile and rare hepatic cells. Histology revealed features of nonparacytic cysts. Mean hospitalization was 11.4 days (7-22) and mortality was nil. In concluding it can be stressed that the treatment of choice is radical cystectomy with or without hepatectomy, while in the cases in which the liver is polycystic, the Lin-Chen-Wang procedure is sun.Qested. to this purpose we preferably use H202. In the postoperative course, we observed a higher rate of early morbidity in conservative methods than in radical ones, being complications represented by local infection, biliary fistula and cholangitis. The average hospitalization time was 13 days for radical operations and 17 days for conservative ones. However, the follow up has not shown significant differences in relapses between radical and conservative techniques. In pulmonary hydatidosis we carried out 3 wedge resections and 3 lobectomies. One patient presented multiple location, hepatic, pulmonary and splenic, treated in three different times: 1 pericystoresection of Bourgeon, 2 inf. left lobectomy and splenectomy, 3 pericystectomy for cystic relapse to the left lung. At last, we report one case of splenic hydatidosis as the only location treated with splenectomy. In conclusion, our results indicate that the operative choice for echinococcal cysts must be balanced by the aim at radicality on one hand, and by saving parenchyma on the other. In hepatic hydatidosis, pericystectomy is the first choice operation, but when multiple cysts are present, partial cystic resection and drainage, previous sterilization of cystic cavity could represent a good alternative to more aggressive hepatic resection. Recently was described a technique (i) for approaching the hepatic pedicle structures and their branches by an intrahepatic posterior approach that allows early delineation of the segments of the liver without the need for extrahepatic pedicle structures dissected separately.
We used this technique during the second semester of 1992, in 6 patients (table). This technique allows the surgeon to dissect and clamp the required sheath early in an operation, eliminating any guesswork involved in defining the boundaries of the segment (S) to be removed. This approach can be used in partial hepatectomy and for resections of tumurs of the proximal bile duct.
We believe also that this technique is one of the factors that leads to diminished blood loss in hepatic resections. Certanly, it is a useful addition to the atarium of operations upon the A new technique of insulin delivery using the greater omentum was designed to exert greater metabolic effects for the patients with liver dysfunction. This new method was easy, simple, and safety as follows The catheter using for insulin administration was inserted into the abdomen and the catheter tip was attached the greater omentum under laparotomy. A continuous infusion of insulin using syringe pump was underwent during operation. Since it might be questioned whether the omental route is suitable for insulin administration, a clinical examination was performed in 4 cases who were received colectomy as follows The examination was a voulus administration of 20 U human regular insulin using our method.
Peak levels of insulin in portal blood were achieved 5 to 15 minutes, and those were twice levels of peripheral blood.
A prospective randomized trial of insulin administration in 15 patients received hepatectomy.
The patients were divided into two groups In group A, 2 to 4 units per hour of insulin were administrated using our method, and group B was not administrated. During the operation, average minimal value of arterial blood keton body ratio (AKBR) in group A was more than 0.7, but that in group B was less than 0.7. Conclusion 1.A new method of insulin delivery using the greater omentum was simple, safety, and effective.
2.In hepatectomy to the patients with liver dysfunction, our method was useful at the aim of inhibitory effect on the suppression of AKBR. The author believes that a new technique of insulin delivery using the omental route is widespread and has proposed ways of application to the treatment of diabetes mellitus. An experimental model of extrahepatic cholestasis in the rat, using a microsurgical technique, is described. Sixteen days postoperatively all of the animals (n=lO) were alive and had hepatomegaly, splenomegaly, splenorenal collateral circulation, jaundice and hyperbilirubinemia. An increase in the size of the portal spaces, with intense proliferation and a slight infiltrate of inflammatory cells, was observed. The proliferation of the bile ducts projected out of the portal spaces, invading the hepatic trabecula, and connected the spaces together. The use of this technique prevents the development of hepatic cysts and other complications inherent in the surgical techniques of cholestasis, such as hepatopneumonic abscesses. Because of the lack of postoperative mortality and the non-existence of hilus biliary cysts, the microsurgical technique described for the obstruction of the extrahepatic biliary route in the rat could be a useful experimental model for the study of cholestasis in the early stage postoperatively and for secondary cirrhosis. Choledochal cysts presenting after childhood are uncommon. Clinical features and results of treatment in 14 adults who presented between 1979 and 1992 were evaluated. The 10 women and 4 men had a median age of 26 years (14-62 years). Diagnosis was delayed (> 1year) in 7 patients and missed in 2 who had biliary surgery elsewhere. In most cases symptoms were nonspecific. One patient presented with cyst perforation secondary to worm obstruction of the common bile duct (CBD) and another had pancreatitis. Of 13 patients who underwent cholangiography, 8 had the characteristic common-channel between CBD and pancreatic duct. Two had associated cholangiocarcinoma. Nine of 10 patients with Todani type-I cysts underwent cyst excision (leaving the posterior adventitial wall intact Lilly technique) and Roux-Y hepatico-jejunostomy. One with malignancy was treated non-operatively. Three patients with type-IVa cysts had extrahepatic cyst excision with hepaticocystjejunostomy. Simple removal was performed in 1 case of a type-II cyst. There was no surgical mortality.
Median follow-up was 5.5years (6 months-9 years). Three patients developed hepatico-jejunostomy strictures 2, 4 and 40 months after definitive surgery: 1 had a left hepatic Iobectomy after failed percutaneous removal of intrahepatic stones and another underwent revision hepatico-jejunostomy. A third patient developed sclerosing cholangitis with secondary biliary cirrhosis and in spite of revisionary surgery, died 8 years later from hepatic failure. Both patients with cholangiocarcinoma have died. Of the 11 survivors, 10 are currently well and 1 was lost to follow-up after 3yrs. One of the main complications in the use of biliary stents is the blockage formation due to bacterial adherence. The present study aimed at modifying the surface of biliary drainage in order to reduce bacterial adherence.
T-tubes, used as a representative of biliar drain materials, was cut longitudinally into slices with 1 cm 2 of square urface. The slices was treated with phosphatidylcholine (PC) and phosphatidylinositol (PI) by soaking them in phospholipid-chloroform solution for 4 minutes and drying at 60C for 24 hours at Karlshamns LipidTeknik AB (Stockholm, Sweden) and sterilised by gas at Grambro Lundia AB (Lund, Sweden. Control slices were sterilised at 120C for 30 minutes. The adherence of cells of seven E. coli strains to the native T-tube slices, T-tube slices with PC or PI were studied in vitro, and the adherence of cells of E. coli strain NG7C to implanted PCor PI-treated slices from the common bile duct in the rat. The T-tube slices were incubated with 1 x 107 cfu radiolabeled E. coli cells/ml at 37C for 60 rain and then drained and washed thrice in 2 ml phosphate buffered saline (PBS). Adherent E. coli cells were quantified by radioactivity counting.
Both PC and PI adsorbed on the surface of slices prevented against adherence of E. coli cells of all seven strains, whereas, after implanted into the common duct in rats between 6 and 12 days, phospholipid-treated slices partially prevented bacterial adherence. Additional studies showed that phospholipidtreated slices inhibited fibronectin to deposit on the surface.
In conclusion, the results suggest that the phospholipid-treated T-tube slices reduce the adherence of E. coli in at least two ways, i.e. by changing surface properties in vitro and by reducing deposition of host-derived molecules on the phospholipid-ra-eated surface in vivo. The results may be of benefit for the modification of biomaterial surfaces in the clinical situation. T-tubes, used as a representative of biliary drain materials, were cut longitudinally into the slices with 1 cm 2 of square surface each. The T-tube slices were sterilised at 120C for 30 min before used. After 5-day period of common bile duct occlusion by the use of a minioccluder surrounding the common duct followed by 7-day relief of the occlusion, male Sprague-Dawley rats were divided into three groups. Group A (n =27) was used as controls. Group B (n = 36) underwent T-tube slice implantation into the dilated but patent common duct, while the animals in group C (n =27) underwent sham implantation. Between one and four weeks after implantation, the animals were challenged with an inoculum between 102 105 cfu of E. coli(clinical isolate) in 100 ttl NaC1 into the biliary tract. Spontaneous bacterobilia were identified by bile sampling before inoculation and routine bacterial cultures and animals with positive cultures were excluded from the study. Complement-mediated opsonic activity in bile and sera was analysed by use of a phagocytic bactericidal assay.
Neither positive blood culture nor spontaneous bacterobilia was found. Bile was sampled and bacterial culture were performed within 24 hours after inoculation.
2 10 cfu injected into the common duct produced biliary infection in 34/36 animals in group B, but only in 6 of 27 animals in Groups A and C ( p < 0.01, Fisher's exact test). 105 cfu resulted in 100% bacterobilia in groups A and C. Bacterial counts in bile in Group B was 2 x 108 cfu/ml(n =34) 24 hours after inoculation of 102 cfu, whereas 6.3 x 104 cfu in infected animals in Group A and 4 x 104 cfu in Group C. Culture 7_.9. hours after inoculation showed, that the bacterial number in bile in groups A and C significantly decreased, whereas it remained high in Group B. Complement-mediated opsonic activity in bile in Group B decreased gradually with time(24+2.8 before implantation vs. 6_+2 four weeks after implantation, P < 0.01) whereas opsonic activity in serum in all groups, and in bile in Groups A and C remained the same level as initially.
Our results suggest that, impaired local defence in bile may at least partly play an inportant role in bacterial infection associated with implants in the biliary tract. based on a retrospective study. These authors advocated a score using five objective, easily collected variables that are independent in multivariate analysis.
The following formula has been propose: 0.04 x A + 3.1 x B + 1.2 x C + 0.7 x E with A age, B = common bile duct diameter > 12 mm (no = 0, yes = 1), C = gallstone diameter < 10 mm (no = 0, yes = 1), D = biliary colic (no = 0, yes = 1), and E = cholecysfitis (no ffi 0, yes 1). According to whether the patient has a score of less or more than 3.5, two groups of patients can be selected with a low (2%) or high risk (32%) of having choledocholithiasis. In an 11 month period, we prospectively compared the preoperative scores of 175 patients who then underwent traditional cholecystectomy with routine intra-operative cholangiography. We were able to select a low risk group (3.4% if score < 3.5) and a high risk group (25% if score > 3.5) of having choledocholithiasis. Laparoscopic cholecystetomy may be performed without any further investigations in the low risk group (with a risk of 3.5 % of choledocholithiasis). In the high risk group, either laparoscopic cholangiography, endoscopic ultrasonography or cholangiography should be entertained. The association of cholelithiasis and peptic ulcer in this study was established by investigating gallbladder abnormalities in 330 patients with endoscopic evidence ofduodenal ulcer. Ultrasonography revealed the presence ofgallstones in 15 patients (3.3%). 4 patients were diagnosed as having non-functioning gallbladder. 2 of them were subjected to surgery for acute abdomen and both had gallstones. 35 patients out of 50 with gallbladder abnormalities showed a picture of cholecystitis on ultrasonography. Peptic ulcer and diseases of the biliary system are possible in association because of the direct anatomical and physiological neighbourhood and the biochemica interaction between the biliary system, stomach, and duodenum. METHOD If the biliary tract was free of stone, and 6 or more weeks had passed since cholecystostomy, a microwave antenna for medical use was inserted into the cystic duct under the guidance of the cholecystoscope. The duct wall and the tissue around the orifice were coagulated by microwave heating for 10 seconds at several points and repeted to 4 times at each point. The current used to generate microwave was 50 mA. Within 20 to 36 hours after coagulation, temporary occlusion of the cystic duct was confirmed by cholecystography, and the volume of the gallbladder lumen was measured. The same volume of 95 % ethanol alcohol and 5 % Tetracycline were injected into the lumen and were kept there for 30 minutes successively. The injection was repeated 4 to 6 hours later. When the drainage was free of bile and less than 10 ml per day, and the cavity was substituted by a narrow fistula on cholecystography, the drainage tube was withdrawn.
RESULT AND DISCUSSION In all cases, the fistula closed quickly, and ultrasonography revealed that the gallbladder cavity was obliterated in 4 to 5 weeks after coagulation. One of them had an attack of fight subcostal pain. The gallbladder cavity was detected by ultrasonography again in the 9th month after treatment. The occlusion of the cystic duct induced by microwave was temporary, it would be patent again in 3 days after coagulation when the edema subsided. Therefore, we performed chemical ablation within 20 to 36 hours after coagulation, and the volume of sclerosing agent was restricted to the measured gallbladder volume. Following this principle, the leakage of contrast medium into common bile duct and t.he evidence of bile duct strcture did not occur in any case. A pediatric 5 Fr. feeding tube is advanced through the cystic duct until it reaches the duodenum through the papilla, and serves as a leading track for sphincterotomy. The duodenoscope is advanced while the patient is in supine position. Alternatively, the papilla is canulated with the sphincterotome in a standard manner, and sterile methylene blue is injected through the sphincterotome and reflux observed at the cystic duct level.
Of 30 consecutive patients (9 males and 21 females, mean age 58) who had their stones identified by routine intraoperative cholangiography, 28 had their stones extracted after sphincterotomywith a dormia basket or flushed through the cystic duct with saline. Three patients had postoperative asymptomatic transient high serum amylase. One patient underwent laparotomyfor an impacted stone in the lower common bile duct. One patient had an esophageal stenosis preventing endoscope insertion and one underwent direct CBD laparoscopic exploration. Two patients had transcystic electrohydraulic lithotripsy for impacted stones. 24 patients had multiple stones and 6 patients had a single stone. Median postopoperative stay was 2 days (range 2-6) and average endoscopic time was 22 minutes (range 15-45).
Intraoperative ERCP does not increase the risk of pancreatitis. It permits clearance of the common duct and prevents to rely on a postoperative ERCP.These results are in favor of intraoperative endoscopic sphincterotomy as an alternative method to extract choledocholithiasis during laparoscopic cholecystectomy. Cholecystectomy on these patients is usually difficult and, because the marked inflammation around Calot's triangle, the dissection is potentially dangerous in this area. To avoid injury of the common bile duct, the Surgeon must be aware of the condition and the diagnosis has to be considered pre-opcratively. To minimise dissection around Calot's triangle we dissect the gallbladder from the liver bed, to its neck, then open it by a parallel cholecystotomy, remove the impacted stone and perform cholecystectomy by leaving its neck as a cuff.
Then we proce to suture the wall of the abandoned neck, avoiding potential damage to the common bile duct and the fight hepatic artery. In our 91 patients no iatrogenic damage was done pre-operatively and the post-operative mortality was 0%.
We conclude that the essentials to the management of this syndrome are: Preoperative diagnosis of the condition by ultrasonography and ER. Preoperative classification of the case.
In type II syndrome, partial cholecystectmy, removal of the impacted stone and suturing of Hartmann's pouch, protects iatrogenic injury of the fight and common hepatic ducts. The aim of this study is to correlate the pre-operative clinical and laboratory findings of acute cholangitis with the operative findings in defining the severity and the prognosis of the disease.
During the last 15 years at our Institution, we operated on 3,300 patients with benign diseases of the biliary tract. Among them 602 patients had acute cholecystitis (18%), and 180 were operate on for acute cholangitis (ACHO) (5.5 %). According to their clinical presentation defined by the Charcot's triad (fever, jaundice, pain or chills) or Reynold's fivefold signs (fever, jaundice, pain or chills, mental confusion and phenomena of shock), leucocytes, prothrombin time, lymphocytes and blood urea nitrogen in comparing with the operative findings of bile peritonitis, edema around the extra.epatic bile ducts, edematous wall of the bile ducts, green bile, turbid bile or pus in the main ducts, redness of the mucosa of the common hepatic duct and multiple abscesses of the liver, we calculated these as risk factors. According to the number of risk factors present in every patient with ACHO we divided our patients into 3 groups. Group A with 1-5 risk factors (52 patients). Group B with 5-10 risk factors (59 patients) and Group C with 10-14 risk factors (69 patients). To Group A belong the patients with mild ACHO. To Group B with moderate and to Group C with severe or suppurative ACHO.
We conclude that there was absolute correlation between the presence of clinical signs pre-operatively, the laboratory parameters and the operative findings as to the diagnosis, the severity and the prognosis of acute cholangitis in our cases.
By calculating all the above parameters, we feel that we can define ACHO patients into the 3 categories of mild, moderate and severe form of the disease, the 3 types of parameters are absolutely correlated and also define prognostically the morbidity and the mortality of patients with ACHO. The pre-operative parameters as calculated in our cases, compared with the operative findings, are worthy of credit for the diagnosis of the disease. There was an increasing morbidity and moity rate according to the increasing number of risk factors present in each one group of patients. The mortality in Group A was 0%. In Group B was 0% and in Group C 12%. All the patients who died had present more than 12 risk factors. Calcic bile (CB) syndrome refers to gallbladder mudding with CaCO3. CaCO3 concentrations higher than 50% in bile lead t Sl:ntaneus radio-opacity (so that radio-opaque dyes are no longer necessary for preparing the gallbladder for radiology).
We have followed 9 cases with CB, 6 of them having been confirmed by surgery, 1 during autopsy and the other 2 by radiography.
In 6 cases (women, 46-61 years old) we performed chemical analysis of the CB. This revealed the presence of CaCO3 in 41-93% of the cases. In 3 cases, cholesterol calculi were mixed into the CB, 2 cases had radio-opaque bile (containing 83% and 89% CaCO3) and in 1 case the CB was inconclusively radio-opaque (41% CaCO3).
In 2 eases (women, 47, 52 years old) we found CB in the gallbladder basinet, while the gallbladder body was occupied by cholesterinic calculi (CaCO3 72-93 %). 1 case woman, 76 years old died of lung cancer; autopsy revealed CB in a crescent form (Rg) at the bottom of the gallbladder and brownish bile at the top. In 1 case (woman, 53 years old) the pasty content of the gallbladder, was drained into the duodenum during surgery for para-esophageal hiatus hernia. In another case (woman, 62 years old) the gallbladder content was spontaneously evacuated into the duodenum, the phenomenon being accompanied by severe biliary colic with subicterus. In the last case (unoperated, man, 78 years old), the radiologie image of the gallbladder varied according to the position of the patient. All our patients had free cystic ducts. Gallbladders showed aspects of chronic inflammation. Echography revealed an echodense structure of the gallbladder with posterior cone of shadow.
Spontaneous radio-opacity without previous administration of radio-opaque dyes is a sign of diagnostic certainty.
CaCO3 appears to the bile by the combination of Ca2 + ions and HCO3 ions when pH is 8 or above. Gallbladder evacuation disorders lead to CaCO3 storage, its density being higher than the density of bile.  (b). RESULTS On group a in 32 pts. operative cholangiography detected choledocal lithiasis but in 118 cases it was not possible to perform any type of this investigation. In the 172 complicated forms occurred 3 jatrogenic lesions (1.7 %) and 5 deaths (2.9 %); in elective surgery 1 residual lithiasis (0.I %) and 2 deaths (0.3 %). The mean hospital stay in elective and emergency surgery was respectively 7 and 14 days. On group b cholangiography pre-or operative has been always done and the LC, always elected, was useful without main complications or deaths. The mean hospital stay was 2.7 days. CONCLUSIONS a) LC is preferable in elective surgery of cholelithiasis, b) pre-or operative cholangiography to prevent residual stones or detect jatrogenic lesions in LC is mandatory and the omission is hazardous because the visual perspective of the operating surgeon as been changed. Athen's General Hospital Athen's Greece This is a retrospective interpretation of our experience in biliary tract surgery in patients with cholelithiasis. During the last 12 years 2344 pts were operated for cholelithiasis and 563 pts of them (24,0%) underwent exploration of the biliary tract (12 pts were explored by choledochoscopy though the dilated cystic duct). of the patients were females. The indications for bile duct exploration were: dilatation of the extrahepatic biliary tract (25,4%), multiple small gallstones (24,7%), x-ray diagnosis of choledocholithiasis (20,2%), obstructive jaundice (17,4%), acute gallstone pancreatitis (7,6%), acute cholangitis (3,2%), palpable choledocholithiasis (1,4%). In 288 pts (51,2%) choledocholithiasis accompanied cholelithiasis, in 6,2% of the pts choledocholithiasis wasn't accompanied by gallbladder lithiasis and in 61 pts (10,8%) choledocholithiasis was present in pts subjected to cholecystectomy in the past. In 32% of our pts the bile duct exploration proved negative for stones despite tholelithiasis. A T-tube was placed in 57,2% of the pts and 35,3% underwent a choledochoduodenostomy, 2% had a Roux-en-Y choledochojejunostomy, 2,2% sphincterotomy-sphincteroplasty and 2,3% had primary closure of the common bile duct. In 6 pts the choledochoscopy though the wide cystic duct didn't prove lithiasis and the operation was limited to cholecystectomy. The immediate postoperative complications associated to the operation, were 12,1%. Most important of them were T-tube related complications (22 pts), retained stones (21 pts), bile leakage (13 pts), postoperative mild pancreatitis (9 pts) and subhepatic collection (3 pts). 10 pts (1,8%) were reoperated for postoperative problems. The total postoperative mortality was 2,3% (13 pts). It is concluded: (a) Bile duct exploration needs special knowledge and attention because of possible complications related to the procedure (b) T-tube intubation is associated with multible problems and whenever is used, care must be undertaken to exclude retained stones with cholangiography and/or choledochoscopy. (c) Choledochopeptic anastomoses is a useful and safe procedure to treat elderly patients with common bile duct stones. PO080 TRANSDUODENAL SPHINCTEROPLASTY.

IS STILL INDICATED?
Kr Daskalakis, E Mouloudaki, P Vernio$, G Diamantopoulos 3d Surgical Department, Evangelismos Hospital, Athens, Greece After the wide use of endoscopic sphincterotomy, transduodenal sphincteroplasty became less popular. Despite this, transduodenal sphincteroplasty is a very good procedure to produce free common bile duct drainage, if clear indications are present. In the 3d Surgical Department of "EVANGELISMOS" Hospital, were performed 31 transduodenal sphincteroplastles in the last 8 years (1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992). Eighteen patients were male and 13 female. Their age ranged between 22 to 78 years. In 17 patients of this series transduodenal sphincteroplasty was the initial treatment and in 14 it was reoperation after cholecystectomy with or without common bile duct exploration in the past. The indications for this procedure were: Common bile duct stones because of hemolytic disease in 2 patients, stenosis of the papilla in 3, acute suppurative cholangitis with impacted stones in 5, obstructive jaundice with impacted stones in 12, residual or recurrent ductal calculi in 7 and rupture of hydatid cysts into the biliary tract in 2. One patient died in the immediate postoperative period because of acute necrotizing pancreatitis (Mortality: 3,2%). This patient reoperated on, necrosectomy and bursa lavage was performed, but he died in the 20th postoperative day from multiple organ failure. The immediate postoperative complications were acute necrotizing pancreatitis in one and respiratory infection in another one patient (Morbidity: 6,4%). The mean hospital stay was 9,1 (range: 8-20) days. The long-term results were good. In conclusion, transduodenal sphincteroplasty is a very good but a little difficult procedure to produce free common bile duct drainage. Complete gallstone dissolution after bile .acid.ther.apy in eligible patients (radiolucent stones -15mm an saze an a functioning gallbladder) takes place in about 1/3 of cases, indicating that such stones are composed of pure cholesterol. A relationship between sonographic gallstone patterns and chemical stone composition has been described.
Therefore, it as interesting to learn whether sonographic stone images can identify pure cholesterol gallstones and improve the efficacy of oral bile acid treatment. Aim. To investigate the relationship between the sonographic pattern and the chemical composition of gallstones.
Patients an mehos. 35 patients with symptomatic gallstones who were eligible for oral bile acid therapy and underwent cholecistectomy were studied. An abdominal sonographic examination was performed before surgery, and at the operation the stones were retrieved and their chemical composition was analized by infrared spectroscopy. Patients were divided into 3 groups: I, 15 with pure cholesterol stones (>90% cholesterol). II, 10 with pigment stones (<20% cholesterol). III, 10 with mixed cholesterol stones (>10% pigment or calcium carbonate ). Sonograms were independently reported by 2 radiologists who were unaware of the chemical composition of the stones. The images were assigned to one of two patterns: 1) One or more hyperechogenic amages with distal acoustic shadow. 2) One or more hiperechogenic images without distal acoustic shadow. Though 120 cases only have been reported in the literature the surgeons have to think about this rare entity because of its worse prognosis than that of acute cholecystitis with a mortality rate 15% (in acute cholecystitis 4,1%) and its higher frequency of perforation of the gallbladder (4 times more often) than that of acute cholecystitis. During the last 10 year-periqd, three cases of emphysematous cholecystitis were treated in 2 n Department of Propedeutic Surgery of University of Athens. The age of the patients ranged from 55 to 67 years and there were two men and one woman. The general condition of the patients at their admittance was severely affected, presenting acute pain in the right subcostal region (space), vomiting and fever ranged from 38,2oc to 39,5oc with main laboratory finding the leucocytosis. All of the three patients had known cholelithiasis and two of them diabetes mellitus. The diagnosis was established upon the plain roentgenogram of the abdomen and the u ltrasonography and in one case the CT scan of the abdomen. One of the patients was operated urgently because of his bad and aggravated general condition and the rest two patients in 48hrs period after the admission. All of the patients underwent cholecystectomy. The cultures of the gall revealed. E.coli and Cl Welchi in two cases. The postoperative recovery was uncomplicated and the patients left the hospital doing well. Conclusively, emphysematous cholecystitis requires timely diagnosis and surgical treatment promptly under antibiotic cover, in order to reduce morbitity and mortal i ty rates. In 3 patients stones of up to 5 mm diameter were found in the gallbladder intraoperatively. All patients survived the operation.
The challenge in securing a preoperative diagnosis arises from the presence of other organ system injuries and the need for analgesia and sedation. Fever and leucocytosis, while commonly present, were too nonspecific to be of any great diagnostic benefit. Abdominal tenderness and liver function abnormalities were frequently absent. We conclude that ultrasound is an important modality at the patient's bedside in the prompt diagnosis and treatment of AAC.

Athens, Greece
The management of any patient with the diagnosis of an acute biliary condition, must be argued for the individual case, guided by certain general indications. Indications for surgery during an acute attack of cholecystitis are" a sustained fever beyond 48 hours, signs of peritoneal irritation, features of cholangitis, general peritonitis, presence of heart disease. 520 cholecystectomies were performed over a 6 year period in our surgical unit of which 52 were early cholecystectomies for acute cholecystitis, cholecystostomies were carried out in 4,2%. The overall hospital stay was 16,1 days. Ultrasonography was carried out in all cases. Empyema and gangrene were found in 36,1% and calculi of the common bile duct in 16,2% of the cases. The mortality rate for early cholecystectomy was 0,4%. There were no significant differences in postoperative complications between early and ellective surgery. But if the structures cannot be recognized because of the presence of oedema it is safer to abandon cholecystectomy in favour of cholecystostomy. Cholecystectomy is clearly ideal provided it is safe. Experts in gallbladder surgery regard cholecystostomy with condescension because they always find it possible to remove the gallbladder during an acute attack. Thus no shame should be experienced in occasionally carrying out cholecystostomy in an emergency. Peri-ampullary duodenal diverticula are known to be associated with an increased incidence of biliary tract and pancreatic diseases. The nature of the association remains uncertain. An analysis of 1336 ERCP had been carried out within a period of five years. The result was 89 (7,5 %) duodenal diverticula, found in association with papilla Vateri. Of these 40 were in male patients of an average age of 65, and 49 in women averaging 70 years of age. The diverticula were divided into six types, so that the possibility of cannulation of the papilla opening existed in 62 (70%) of the patients.
The most frequent indication showing the need for examination was the obstruction of gallbladder drainage in 35, obstructive icterus in 13, and postcholecystectomy stage in 5 patients, and chronic pancreatitis in 10 patients. In 52% of the patients thole or choledholithiasis was fund, in 16% chronic pancreatitis, and in 13 % stenosis of the distal part of the choledochus. In 3 % of the patients a turnout in assiation with diverticulum was fund, and in 16% of the patients the findings were normal. On the basis of the results the nclusion is drawn that ere is close assiation between biliary and pancreatic diseases and duodenal diverticula associated with papilla Vateri.  (8), liver (8), lung or bone (5) and para-aortic node (1). Most important contributing factor to hilar recurrence was cancer positive margin in cases with direct hilar invasion. In patients with definite hilar involvement, negative margin was obtained in only 2 of 16 cases even by exteded surgery, l4ode of recurrence after curative resection was mainly hematogenouso Conclusion: Even though results of curative resection of GBC is satisfactory, hilar invasion is still the barrier for surgical cure. For early detection before jaundice, routine use of ultrasonography for abdominal symptoms and macroscopic examination of the gallbladder removed for presumed benign diseases were quite useful. The management of benign intrahepatic strictures and stones, especially in patients with bilateral disease, is complex with a high incidence of recurrence. The efficacy of operative stricture dilatation, stone extraction and creation of a Roux-en-Y hepaticojejunostomy with an afferent access loop for subsequent radiological percutaneous transjejunal biliary intervention, was evaluated.
Twenty two symptomatic adult patients (14 women; 8 men mean age 31 years; range 22-64 years) with benign intrahepatic strictures and stones underwent 72 postoperative percutaneous transjejunal biliary balloon dilatations between 1986 and 1992. The intrahepatic strictures and stones were secondary to iatrogenic hepatic duct injury (n=5) Caroli's disease (n=2) choledocal cyst (n=2) and primary intrahepatic stones (n= 13). All patients underwent intraoperative stricture dilatation, stone extraction and construction of a Roux-en-Y side-to-side hepaticojejunostomy with a 12 cm jejunal access loop marked with silver clips and attached to the anterior abdominal wall. Subsequent postoperative percutaneous transjejunal biliary dilatation of residual or recurrent intrahepatic strictures and stone extraction was performed using a biliary guide-wire, co-axial catheter and 7 Fr Gruntzig angioplasty balloon catheter. Three patients with intrahepatic stones had in addition resection of the left lateral segment during the primary procedure.
Five patients with iatrogenic strictures underwent 22 dilatations (mean 3, range 1-6) and are asymptomatic without residual stones. Seventeen patients who had Caroli's, choledocal cysts or primary intrahepatic stones underwent 50 dilatations (mean 2.9, range 1-9). Two patients have residual intrahepatic stones and one has required a left lateral segmentectomy for removal of stones. Minor local or ductal complications related to percutaneous dilatation and stone extraction occurred in three patients without long-term sequelae.
The combined radiological and surgical approach using a Roux-en-Y hepaticojejunostomy and biliary access loop with post-operative percutaneous transjejunal biliary dilatation and stone extraction provides an effective method of treating symptomatic patients with complex residual or recurrent intrahepatic strictures and stones. A surgical procedure was possible in 35 of the cases. Cholecystectomies for the gallbladder carcinoma were performed in 18 patients (two associated with wedge hepatic resection), hepatobilia-T resections were performed for the tumours of the hilum in 5 patients, biliary resections were performed for tumours localised in the choledochus or the common bile duct in 4 patients, duodenocephalopanereatectomy in 2 patients for carcinoma of the papilla, side to side choledochoduodenostomy in 4 elderly patients for carcinoma of the papilla and two tumours of the common bile duct. No surgical treatment was performed in two elderly patients with severe diabetes and multiple metastases. Complications appeared in 8 patients. The post-operative mortality was 9.2 %. Long-term survival was good for the gallbladder carcinoma (<2.5 years), 6-18 months for the hilar carcinoma, 18-28 months for the common bile duct and 10-22 months for the papilloma tumours.
In conclusion, the results of our study seem to show that a curative resection is rarely lssible fr treatment of biliary tract carcinoma, although the palliative techniques are able to improve the extension of survival and the quality of life of these patients. The so-called Mirizzi syndrome is a rare condition in the heterogeneous field of benign biliary diseases that can complicate the natural history of gallstones. In this study we report our experience in the surgical treatment of type 1 Mirizzi syndrome.
Patients and method: Between 1982 and 1991 in our department we operated on 2420 patients for benign biliary tract diseases. Among these we have encountered 19 cases of the Mirizzi syndrome, representing a prevalence of 0,78%. The mean age was 48 years range 24-62 nearly all of them (79%) females. The most common signs and symptoms were pain (100%), fever (89%), jaundice (74%) and pruritus (63%). The diagnosis was made preoperatively by intravenous cholangiography in 2 cases (10,5 %), by ERCP in 8 (42%), by ultrasonography in 7 (37%) and in 2 the syndrome was discovered intra-operatively.
All patients underwent cholecystectomy and bile duct exploration; in 13 cases (68%) a T-tube was placed, in 4 (21%) a Roux-en-Y biliojejunostomy was performed and 2 patients required the repair of superficial erosion of the hepatic duct wall. There was no operative mortality. The morbidity was 10,5 % 1 subphreni abscess and 1 biliary lge.
Discussion: There are nowadays two surgical and cholangiographic variants of the syndrome: tyle 1 obstruction of the cmmon hepatic duet causexl by a solitary stone impacted in the cystic duct or in Hartmann's pouch of the gallbladder and type 2 cholecystocholedochol fistula due to a calculus that has eroded partly or completely into the common bile duct. As recently suggested by Csendes and others (1989), we believe that the original Mirizzi syndrome is only the first but probably the most important step of a complicated process that can lead to the development of a true cholecystobiliary fistula. So e diagnosis at this first stage must be very careful in order to prevent the progression of the inflammatory disease. In every way the Mirizzi syndrome still remains a difficult challenge for biliary surgeons. Parma Italy The appearance of extra-hepatic obstructive jaundice in patients who undergone cholecystectomy represents a complex pathogenetic problem;the causes of obstruction beeing almost exclusively represented by residual or recurrent stones,papillitis,and iatrogenic injuries overlooked during surgery. Apart from these conditions,generally the appearance of obstructive jaundice due to incomplete stricture of the common bile duct,in absence of liver and[or bile duct lesions present at surgery,has been repeatedly reported. Ischemic pathogenesis of late strictures of the common bile duct following cholecystectomy have been for a long time hypotized and anatomo-surgical studies on blood supply to the extrahepatic bile ducts have undubtedly contributed to support this point of view. Two clinical cases,in which ischemic pathogenesis of late (11 months and 5 years after surgery respectively) bile duct stricture could be suggested,are reported.
Beside an anao-surgtcal appraisal on blood supply to the comnon bile duct,clinical features of these lesions and their menagement are discussed,pointing out that bilio-digestive anastomosis results as the procedure of choice. The use of intrahepatic biliary anastomoses is sometimes necessary in patients undergoing resective or palliative surgery at the hepatic hilus. Over a 23 year period we have performed intrahepatic bilio-enteric anastomoses in 54 patients (15 palliative, 35 curative, 4 benign) with a median age of 55 years (range 15-82). There was a hospital mortality of 4 patients in both the "palliative" and "curative" group. A significant fall in both the serum bilirubin and alcaline phosphatase occurred after surgery (p < 0.05). Long term follow up was possible in 42 patients. Recurrent stenosis due to recurrent disease occurred in 89 % of the palliative group and 69 % of the "curative" group. There were 2 stenosis in 4 patients with benign disease. Thirtyfive patients developed recurrent cholangitis, 5 without apparent stenosis. The median survival for the palliative group was 4.5 months (range 2 days-32 months) and in the curative group it was 21.7 months (range 0.5-148 months) while all patients in the benign group remain alive. Despite the advent of modern endoscopic and percutaneous intubation techniques, intra-hepatic anastomoses after tumour resection offer the only chance of cure for obstructing hilar malignancy. Center, Zaloska 7, 61000 Ljubljana, Slovenia The incidence of proximal bile duct carcinoma in our country is increasing. During the last decade our approach to the treatment of these patients has changed and aggresive surgical treatment has been accepted. Sceletonization-resection is justified as long as it is technically feasible and the patients do not present any signs of tumor dissemination.
During the 5-year period from Jan.lst 1987 to Dec.31st 1991, 34 patients with carcinoma of the proximal bile duct were treated at our Department (13 males, 21 females).
Supposed radical or only palliative resection was possible in 27 patients (resectability rate 78%). Thus presumably radical procedure was performed in 17 patients, and palliative in I0 patients. In 7 patients resection was not possible any more: two of them had biliodigestive by-pass performed and in another two intraoperative endoprosthesis was inserted, while only exploration and biopsy were done in the remaining three patients.
We had no intraoperative deaths but postoperatively death occured in four patients (mortality rate 11,7%)" two patients died after palliative resection, one after intraoperative endoprosthesis insertion and one after exploration. Mortality rate for 27 resected patients was 7,4% (2 patients). (1) Earlier operation. The mortality of AOSC increases as clinical features such as pain, jaundice, fever and shaking chills become heavier, but the manifestations of hepatocholangitis (121/431 cases-28.07%) might be lacking in jaundice even with more serious lesions. Improper delayed operation would be associated with a very high mortality. (2)  Laparoscopic cholecystectomy has emerged as the treatment of choice for symptomatic cholelithiasis, and the number of contraindications has sharply decreased in the last months. Nevertheless, gallbladder carcinoma (GC) remains as a formal contraindication for LC. The incidence of gallbladder carcinoma in patients operated for cholelithiasis is 1,5-2%, and 80% of early GC are incidental findings during cholecystectomy. Diagnostic of GC during LC may be impaired by decreased tactil feeling. In recent years, has been described 5 cases of inddental GC found during LC, and all patients developed parietal seeding few months after surgery. We present a case of G C that not was recognised during LC. CASE REPORT: A 67 y. female diagnosed 25 years before of colelithiasis, presented with an history of biliary colic pain. 2 ultrasonographic explorations performed in the last 5 months revealed the existence of 2 stones of 2 cm. with a normal gallbladder wall. The bile duct was dilated ( 10 mm). Blood hepatic test were normal. During LC, a distended and immflammatory gallbladder was found, and the diagnosis o acute cholecystitis was stablished. An intraoperative colangiography was normal. The patient was discharged at fourth day without incidences. The pathological study reported a moderatelly differentiated adenocarcinoma afecting the whole wall and perineural invasion. 5 months after, the patient developed two lumps in the umbilical and right subcostal scar, and the biopsy was compatible with adenocarcinoma deposits. CONCLUSION: The incidence of unsuspected GC during LC is low, but the misdiagnosis of this tumour may preclude a correct treatment or may induce the dissemination of the tumours trough the trocars tracks. Laparoscopic surgeons must to keep this possibility in mind, and if GC is suspected during LC, a frozen pathological study should be performed and the cholecystectomy converted to an open procedure.

MATERIAL AND METHODS
Between April '90 and August '91, 25 biliary prostheses of Wallsten were implanted in 20 patients with biliary obstruction: 11 men and 9 women. Age distribution ranged 33-84 years old (68.2). Nineteen were malignant, 10 cholangiocarcinoma, 5 pancreatic cancer and 4 nodes from metastases in the area of ampulla of Vater. One was benign in a patient with two stenoses of biliodigestive bypass. Early complications (< 72 hours) in 15 % cases: biliary sepsis, pancreatitis and subcapsular haematoma.
Late complications found" cholangitis without obstruction of the prosthesis (20%), prosthesis obstruction (30%). Post operative mortality represented 48 % (11 patients) 2 due to cholangitis, one due to digestive haemorrhage and 4 due to evolution of carcinoma.

CONCLUSIONS
The advantage of this new prosthesis is its large internal diameter and the fact that it can be inserted percutaneusly. In this report, Olympus mother-scope TJF type M20 and baby-scope CHF type B20 were used. Three cases underwent the procedures successfully in a half year period. The results of PC compared with the imaging of endoscopic retrograde cholangiopancreatography (ERCP) were reviewed as follows: Case I ERCP confirmed the presence of left hepatic duct stones and proximal stricture which were treated by PC after endoscopic sphincterotomy (EST). The proximal stricture was dilated by baby-scope and calculi in situ were removed through the channel of the baby-scope by Dormia basket; Case II was proved to suffer with residual choledocholithiases after cholecystectomy. Following preliminary extraction of stones through duodenoscope channel after EST, PC was used and all residual calculi were removed under direct view; Case III had a history of biliary tract exploration with left hepatic duct stones suspected by ERCP and B-ultrasound imaging. An attempt to directly insert the baby-scope failed prior to EST even under the guidewire aid. The following procedure carried out immediately after EST was quite easy. Under the direct view of the baby-scope, no calculi could be found and unnecessary laparotomy was therefore avoided.
The preliminary experiences showed that: 1. PC has a definite role in diagnostic and therapeutic biliary endoscopy; 2. EST is a necessary step and should not be omitted even though the insertion of the baby-scope could succeeM in some cases; 3. The stricture decreases the effects of PC if the dilation of the stricture cannot be effected. 4. As for the shorter hospitalisation, the total expenditure is lower compared with the surgical interventions. Development of laparoscopic cholecystectomy has firstly reduced the proportion of routine operative cholangiography because of technical problems. The necessity to obtain a cholangiogram has not to be influenced by the procedure of cholecystectomy. The aim of this study was to appreciate the feasability and performance of laparoscopic cholangiography (LC) during the treatment of symptomatic gallstones. Between January and December 1992, 100 patients with symptomatic gallstones, eligible for laparoscopic cholecystectomy were prospectively studied. Of these, acute cholecystitis was present in twenty six patients. All patients were operated on by laparoscopic procedure, except 2 conversions to laparotomy. Routine LC was tried in 87 of 100 cases, and was not carried out in 13, because of iodine allergy or a previous endoscopic sphincterotomy. Cholangiography was sucessfull in 72 patients (83%) and failed in 15 patients, because of narrowness of cystic duct in 10 and cystic valves in 5. According to preoperative clinical, biological, ultrasonographic and cholangiographic findings, 68 of 72 cholangiograms were normal. Four were abnormal because of an abnormal biliary anatomy (no cystic duct) in one case, a common bile duct stone in one case and cystic stones in two cases. In the two former cases, a laparotomy was decided for the treatment of anatomic biliary abnormality and common bile duct stone. In the two later cases, laparoscopic cholecystectomy was performed after removal of cystic stones. In conclusion, LC permits to detect "1) anatomic biliary abnormalities, usually unsuspected on routine preoperative investigations, avoiding severe bile duct injuries.
2) cystic stones, whose laparoscopic treatment is generally easy.3) possibly common bile duct stones, usually identified preoperatively. LC can be carried out simply and effectively in many cases, but failures are unpredictible. Others preoperative bile duct explorations (endoscopic retrograde cholangiopancreatography, endoscopic ultrasonography) have to be evaluated. POLO5 LAPAROSCOPIC CHOLECYSTECTOMY (The first thirty five cases) .M.. VeleQrakis, G. Perpirakis,Th.Kokkinakis, G.Fasseas, A.Apalakis. First Department of Surgery, Venizelion General Hospital, Iraklion Crete GREECE.
From the begining of 1992, laparoscopic cholecystectomy was undertaken on 36 chololethiasic patients, which was successful in 35 of those (97.2%). Criteria for selection of this method were the absence of obstructive jaundice in patient's past history and previous operations in upper abdomen. In one female patient the laparoscopic method was converted to open operation due do the thickness of the gallbladder wall and to the solid adhesions in Calot's triangle. Hasson's procedure was used in 8 patients because of previous sub-umbilical laparotomies and laparoscopic cholangiography was performed in 3 cases. Retrograde cholecystectomy was carried out in two cases due to difficulties in preparation of cystic duct and artery, which were ligated after the gall bladder was mobilized. Serious intraoperative complications were not observed and the mean operative time was 2 hours.
One patient required prolonged exploration due to bile leak after a presumed diathermy injury of the C.B.D. Two more complications were observed: D.V.T. in one case and sub-cutaneous emphysema in an other one. In all the patients antibiotics were administrated perioperatively, anti-thrombotic agents were also given and subhepatic drainage was established in 22 patients for 24 hours. In the rest of patients, the post-operative recovery was uneventful, the average time of hospitalization was two days and the patients returned to their normal activities after one week. The past years have been a time of great excitement in laparoscopic surgery and laparoscopic cholccystcctomy in particular. Sixty nine patients suffering from acute calcular cholecystitis subjected to laparoscopic procedures were included in this study. Patients were collected at random, 41 patients were women and 28 wcrc men, patients aged 26-84 years with a mean age 58.5 years, laparoscopic cholecystectomy could be done in 54 cases (78.26%). Cholecystcctomy was conducted safely in 42 cases (60.87%). In 18 cases laparoscopic cholangiography through cannulation of gallbladder or cystic duct was indicated and it was successful in 14 cases and laparoscopic cholecystcctomy was proceeded in 12 cases (17.39%). Minor complications occurred in 9 cases and included bile leakage in 4 cases, umbilical sepsis in 2 cases and 3 cases of abdominal wall hacmatoma. Major complication in the form of duodenal perforation occurred in one case. In fifteen patients (21.73 %) the operation was converted into open procedures because of difficult dissection or unclear anatomy in 8 cases, failed cholangiography in 4 cases, cholangiographic cvidcncexl of common bile duct stones in 2 cases, cystic artery blceing in one case and perforated duodenum in one case. No mortality occurred in this group of patients. Dept of Surgery, Ag. Anargiri Hospital Kifisia, Athens, GREECE.
The aim of this study is to evaluate the efficacy and safety of the combined endoscopic and laparoscopic treatment of patients with gallbladder and common bile duct stones. For the last 18 months until August 1992, 355 laparoscopic cholecystectomes (LC) were performed. Twenty two patients who had a history of jaundice and abnormal LFTs or dilated CBD on U/S underwent an ERCP before the LC. Fifteen patients had CBD stones which were removed in 14 (93%) patients after a successful endoscopic sphincterotomy. LC was done 1-3 days later. One (7%) patient wth large CBD stones (>2.5 cm) required laparotomy with CBD exploration.
Intraoperative cholangography was attempted in 114/355 (32) patients and was successful in 112/114 (98%). Unsuspected CBD stones were found in 16 (14) patients. Stones were removed intraoperatively using a small choledochoscope through the cystic duct in 1 (6) patient. ERCP and EST were performed in 15 (94%) patients wth successful stones removal in all patients. CBD access was possible in 3 (20) patients after a needle-knife papillotomy. Nasobliary drain was used in 7 patients after the ERCP. One patient (3%) developed mId pancreattis which was treated conservatively. There were no deaths.
Preor post-operatve ERCP, EST and CBD clearance combined wth Laparoscopc Cholecystectomy is a safe and effective treatment in patients with gallbladder and CBD stones. ] the operation ws begun with lposoopio technique end later on was converted to open cholecystectomy. The average operation time was 95 rain. No complications of pneumoperitoneum were observed. We did not noticed any complications of introducing the trooars. In 15 [ 6,7 the gallbladsr was opened intraoperativsly, including 10 osos 4,5 1 where the stones failed out to peritoneal cmvity. The intraoperative bleeding rsquireing additional clip plication was noticed in 4 patients 1,8 |. In a single 0,4 ] few hours lsting bile leakage ws observed through the thin drain left in the gallblader bed. It stopped spontaneously. In the other single ease the cystio duct in the place of the connection with common hepatic duet was punotered with the hook. The conversion to open cholecystectomy was done and Ttube was inserted in the place of injury The suppuration o| the umbilioal wound was diagnosed in 2 oases [ 0,R % ]. None of the patients required reoperation we did not observed any lethal omplication. Conclusion The small number of intraoperative and postoperative complications let us suggest, that LCI should replace the open cholecystectomy in the majority of oases. Pancreas being relatively protected in the retroperitoneum presents a low incidence of injury after blunt abdominal trauma. The vast majority of pancreatic traumata are associated with injuries to major vessels, solid and hollow viscera. Isolated pancreatic damage associated or not to extraperitoneal injuries is extremely rare.
Three such cases from a total of 34 pancreatic injuries (8.8%) in a lO-year period are presented. The general pancreatic trauma index in the same period was relatively low (2.8%).
In conclusion isolated intraperitoneal pancreatic injuries are very rare and there are usually diagnosed with delay. Due to this fact fulminant course of acute inflammation or other post-traumatic complications are responsible for the high mortality rates. It is obvious that the mechanism of injury may leed to the suspicion of a potential pancreatic injury which must be followed by an aggressive approach to diagnosis and early operative intervention. ERCP showed normal CBD with external compression on common hepatic duct and both hepatic ducts.
Cook's pig tail (10.2F) percutaneous catheter drain under ultrasound guidance was fixed. Mixed blood and bile was initially drained. Later on it became clear bile.
Repeat ERCP after 2 weeks revealed a normal biliary tree with a minute communication between fight hepatic duct radicle and the cavity. The catheter was removed after 1 month. Repeat CT scan showed complete collapse of the cavity. Percutaneous catheter drainage was a simple successful procedure that obviated surgery. Rethimnon, Gr liver trauma is not frequent but its management is usually very difficult and its outcome is combined with high mortality.
Over a 7 year period from 1986, twenty-five patients with hepatic trauma were managed operatively in our department. Twenty patients had blunt and 5 penetrating liver trauma, including 2 gunshot wounds. 17 of 20 blunt traumas were caused by road traffic accidents, with direct hepatic damage and 3 were due to a fall. Three cases were pure hepatic ruptures and 22 were related to multiple injures. The hepatic injures were classified according to severity and regarded as being grade I in 3 cases, grade II in 14, grade III in 5 and grade IV in 3 cases according to the classification of Calne.
All the above patients required urgent laparotomy. The 3 cases of grade I,the 14 of grade II and 2 of 5 cases of grade III were treated with lavage of hepatic parenchyma with normal saline by suturing profound ruptures and by stable pressure of the liver to control haemorrhage. 2 cases of grade III required excision of the damaged part of liver and ligation of bleeding vessels. One patient of grade III and in 3 of grade IV required pedhepatie packing for life saving bleeding control. Of those 4 cases, 1 died during the operation as a result of severe co-existing injuries, 3 were referred to a specialist centre. Of the 21 cases re-operation was required in 2 .for haemorrhage and 2 for sepsis.
Of the 25 patients, 5 died. Only 5 out of 25 (20%) died. Grade IV 2 patients, grade III 2, and 1 patient of grade II. The cause of death in 3 cases was uncontrollable haemorrhage, coagulopathy and in the other 2 cases the cause of death was severe co-existing injuries.
The successful outcome of these injuries depends on the resuscitation and the urgent laparotomy, in which conservative surgical procedures must be used.
We conclude that those simple surgical procures such as direct sutures, ligating bleexling vessels, excision of damaged hepatic parenchyma if necessary, perihepatic packing and selective hepaticartery ligation. However the morbidity and mortality from these injuries is still high in all centres specialise or not. A thirty-one year old patient was admitted to hospital after a car accident. The patient had multiple bone injuries and signs of head injury. During the first hours of hospitalisation he developed an abdominal pain in the supra-umbilical region. Ultrasound showed minimal free liquid in the abdominal cavity. There was no sign of an acute abdomen. Two days later ultrasound was normal. The patient was treated because of bone injury.
Seven days after admission the patient becamejaundice and liver function tests showed.the evidence of a bile duct obstruction. On ultrasound we saw dilated bile ducts and ERCP showed amputation of the cholexiochal duct in the intrapancreati portion and dilated pancreatic duct. CT scan showed enlarged head of the pancreas with necrotic changes and dilated bile ducts. Body and tail of the pancreas were normal and the pancreatic duct was dilated.
During the surgical procure intra-opcrative cholangiography showed "stop" in the head of the pancreas and swelling and hardness of the head of the pancreas with multiple adhesions. Choledochojejunostomy with "Roux" loop was performed with gxt results. Six months after the operation the patient is without trouble and at this time we planned to perform a second operation (pancreaticojejunostomy) if the stenosis of the pancreatic duct in the head of the pancreas develol dilatation of Wirsung's duct, and chronic pancreatitis.
Two years after the operation the condition of the patient is satisfactory, and liver function tests are normal. On ultrasound there is no evidence of dilatation of the pancreatic duct. Hepatic injuries, relativery rare in peacetime in the past,are becoming more common nowadays because of the enormous increase of traffic accidents.Injuries of the liver represent a very severe clinical condition,both from the point of view of iiate management of liver injuries which requires the taking of rapid,sound and correct decision as to the appropiate operative procedure to be followed in each specific case. Open lesions of the liver are less difficult to handle,with the exception of those cases where we have associated lesions of a large vessel,e.g.the inferior vena cava of the aorta. In contrast,several problems arise in the management of closed hepatic injuries which are still associated with high mortality and complication rates. There were reviewed 48 cases with traumatic injuries of the liver. The over all mortality rate was 29.16%.Eleven from the patients who died during the operation or the postoperative period suffered from minor or moderate liver injuries treated by simple suture.
Deaths were mainly due to multiple organ inguries or oligemic shock than to the hepatic injuru itself.The majority of liver injuries may be safely managed with present knowledges on hepatic surgery with exception injuries involving hepatic veins and retrohepatic vena caval area.

Bulgaria
In Bulgaria trauma of the abdomen for the last 10 years rates 4th place as a reason for death. 50% of traumatic injuries are caused by transport vehicles and firearms. According to the data injuries of the liver are from 5 to 11% of abdominal trauma. Trauma of the extrahepatic bile ducts is comparatively rare, which is due to their anatomical location. Penetrating injuries of the pancreas are less frequent than blunt injuries.
For the period 1985-1992 in our surgical clinic we treated 164 patients with trauma of the abdominal system. 132 were blunt injuries and 42 were penetrating injuries. The male patients predominated in numberie. 101. The age of the patients varied between 16 and 70 years. Trauma of the HPB system was mainly multiple. In 13 cases we observed single liver trauma with 15 cases of post traumatic pancreatitis. Single trauma of the extrahepatic bile ducts was found in 3 patients and injury of the gallbladder in 1 case. Diagnosis, reanimation and operation are done at the same time in our practice. Treatment is done on the basis of 2 principles to obtain control of the intra-abdominal bleeding and abdominal sepsis.
Our data shows that trauma of the HPB system is 15, 1% of the patients with abdominal trauma with a mortality rate of 20, 3 %. We retrospectively classified all 172 patients with liver trauma admitted to Sunnybrook Regional Trauma Unit from Jan I, 1987 June 30, 1992. 90% were due to blunt trauma and 87% underwent aparotomy. 34.5% had liver related complications and 6.4% had liver related death. We compared liver related mortality and morbidity (M & M). Morbidity was evaluated by measuring blood tranfusion requirement, liver complications and liver treatment using the 3 different classification systems. Statistical analysis revealed that each of the 3 systems was internally consistent and had a linear relationship to each variable in predictng M & M.
The Moore was the best predictor of liver related death and blood requirement, the OIS and the Buechter were able to predict major liver treatment better and the Buechter predicted liver complications well. Subjectively the Buechter is easier to use due to its simplicity; but it tends to cluster most of the patients into grade 2 and the choice of treatment affects the grade of injury. The OIS, although more complex, leads to a better separation of the different grades of Iiver trauma; thus al owing comparison between different institutions.
In conclusion, all 3 systems have merit and none proved superior in all aspects. The major advantage of Buechter is its simplicity and ease of recall. The Moore is closely related to the OIS which is more accurate and allows precise comparisons between different institutions. The high frequency of traffic accidents has contributed to an increasing incidence of pancreatic trauma. The surgical adage "the pancreas is not your friend" is apparent considering mortality and morbidity rate over than the one half of pancreatic trauma patients.
As associated major vascular or solid organ injuries lead to early mortality due to exsanguination, mortality rate of pancreatic trauma due mostly to complications, is easily understood.
During the last decade 34 cases of pancreatic trauma after abdominal injuries were treated. The spectrum of injuries comprised contusion with haematomas, laceration, fracture and complete disruption of the gland, but the main point to determine trauma severity was the integrity of the pancreatic duct. Associated intraabdominal trauma rate exceeded 2.5 per patient. Operative techniques used were wide closed suction drainage, debridement, suture, distal or hemipancreatectomy and pancreatoduodenectomy.
Hypovolemic shock was the cause of death in 10 patients due to associated injuries. In another 14 patients secondary haemorrhage, fistula, pseudocyst or intraabdominal abscess formation, and anastomotic leakage obliged to reoperations, but finally 8 patients died because of sepsis or multiple organ failure.
It is concluded that the final outcome after pancreatic trauma is in close relation with exact intraoperative evaluation of the injury, especially of the integrity or not of the major duct and is also depended on the degree of parenchymal damage and the anatomic location of the injury. Bacterial translocation from the gut is thought to be responsible for infectious complications in surgery and in different disease states. We studied the translocation of bacteria in galactosamine induced liver injury and compared the results to liver resection and control rats.
Material and methods: Liver injury was induced by administration of Dgalactosamine. Results were compared to 70 % liver resection and control rats. Animals were studied 24 and 48 h after injury. The extent of liver injury was assessed by liver function tests and histology. Translocation to medal and portal blood as well as solid organs was studied after administration of labeled bacteria. Intestinal permeability was also measured by sodium fluoroscein administration. Bacterial cultures of intestinal mucosa and content was performed.
Results: The incidence of bacterial translocation to arterial and portal blood was found to be 50% and 83% resp. after galactosamine administraion at 24h and remained elevated to a similar extent at 48h. The rate of translocation after liver resection was 100% to medal and portal blood at 24h and dropped to 50% at 48h. There was no translocation to blood in control group. Bacterial translocation to liver, spleen and mesenteric lymph nodes was 100% after both galactosamine administration and liver resection at 24 and 48h. Intestinal permeability was increased similarly in both groups compared to control animals. Bacterial cultures did not show any differences between the groups. Conclusion: Bacterial translocation and intestinal permeability are increased after severe hepatocellular damage to a similar extent as seen after liver resection. They may play a pathogenetic role in the infectious complications seen in these conditions. Although there is a great deal of enthusiasm for the consumption of fish oil enriched with omega-3 (o-3) fatty acids, it is known that prolonged provision of marine oils to subjects with type II (adult-onset) diabetes mellitus may raise their daily insulin requirements. Also, it has been shown that acute exposure of pancreatic islets to high concentrations of polyunsaturated fatty acids will stimulate insulin secretion whereas chronic exposure to these fatty acids will desensitize the islets to glucose via a mechanism linked to fatty acid oxidation. However, it is not clear whether omega-3 (0-3) fatty acids present in fish oil have a direct effect on the release of insulin from pancreatic beta cells. AIM: The purpose of the present study was to examine the effects on insulin release of a dietary precursor of 0-3 fatty acids, linolenate (18:3,o3) and the main constituent 0-3 fatty acid of fish oil, eicosapentaenoic acid (EPA, 18:5 0-3), and to compare these effects with those of omega-6 fatty acids. METHOD: In each experiment, a total of six islets microdissected from the pancreata of three female CD-1 albino mice were placed in a flow-through perifusion chamber and preperifused for I hour, at the rate of lml/min with a Krebs-Ringer bicarbonate buffer pH 7.4, containing 2% bovine albumin, 5.5mM (basal) glucose which was continuously gassed with 95%/5% Oz/CO z. After basal samples were taken, the perifusion was continued in 20 mins random cycles of different fatty acids separated by 20 min "washout" periods of basal glucose perifusion. Solutions were changed using a stopcock and effluent perifusate samples collected on ice at 2 min intervals were stored frozen until radioimmunoassay for insulin. .RESULTS: Insulin secretion in response to the different fatty acids was assessed as the mean integrated area under the curve (AUC/20 mins) above basal. Perifusion of islets with 5mM of each of linoleate (18:2, o-6) and linolenate (18:3 0-3) showed that the latter was more potent insulin secretagogue as the AUC's respectively were 2842__.417 and 10506+_1490 pg/20 mins, p<0.002, n=5). Also, l mM EPA (20:5, -3) was significantly more potent than lmM arachidonic acid (20:4#0-6) in insulin stimulation (2168+_452 vs 597+_251 pg/20 min, respectively, p<0.003, n =5). CONCLUSION: These data demonstrate a direct effect of omega-3 fatty acids on insulin secretion by perifused islets and suggest that -3 fatty acids are more potent insulin secretagogues than the corresponding 0 6 fatty acids. Aneurysms of the splenic hilum often require splectomy, which has been reported to be associated with a high incidence of septic complications (1-4%). Utilizing microsurgical techniques for vascular reconstruction, splenic preservation may be technically feasible even in case of hilar aneury.sm. CASE REPORT: B.P. 37-year-old-female with no past history of vascular disease, 2 regular pregnancies, came to our observation for an asyntomatic aneurysm of the splenic hilum (diameter 1.5 cm) incidentally detected at ecotomography. Because of the increasing volume of the aneurysm (diameter 2.1 cm) after one year of clinical follow-up the operation was decided.
Because of the overlapping of vessel images selective angiography was not diagnostic; ecotomography with pulse doppler confirmed the diagnosis of hilar aneurysm involving the bifurcation of the splenic artery (figure A). Through a midline incision with self retractor the spleen was anteriorly displaced sectioning the posterior legament. The aneurysm was completely mobilized and resected working from the posterior aspect of the spleen. Using an operating microscope (Olympus.ll@ the two splenic a.fferent vessels were anastomosed end to end (7/0 prolene) and the afferent artery anastomosed to the loop end to side (figureB). The post-operative course was uneventful. A post-op, digital angiography showed the excellent splenic circulation with a normal vascular anatomy.
Microsurgical technique can be used also in general surgery for difficult vascular reconstructions.  Survival after an acute pancreatitis attack, has improved in the early stage of the disease as a result of aggressive organ support, over the past two decades. But patients continue to die at a later stage from necrotic and septic complications. The natural history of necrotizing pancreatitis has been better elucidated, with appreciation of a spectrum of pancreatic necrosis and pancreatic abscess; infected pseudocyst and peripancreatic phlegmon are excluded from the latter term.
Over a 7 year period from 1986 to 1992, 196 patients with an acute pancreatitis attack were treated in our units, including 24 cases (12, 25 %) with acute necrotising pancreatitis (18 men and 6 women). Pre-operatively the diagnosis in 22 cases (91,5 ) was documented by combined estimation of the clinical course, laboratory evaluation and by findings on C/T. In all those patients high levels of serum CRP (> 200mg/lt) leucocytosis and positive limulus amoebocyte lysate test for detecting endotoxinemia, were found. Intra-operatively findings were: 6(25 ) cases with pancreatic necrosis, 10(41.6) cases with pancreatic abscess 3 on the pancreas head and 7 on the pancreas body and tail and 8 infected.
All patients with documented necrotising pancreatitis underwent operative management via a midline incision and trangastrocolic approach to the lesser sac. Our surgical method is based on atraumatic necrosectomy and lesser sac drainage using 4 large diameter (20 mm) silastic tubes, for continuous post-operative lavage. The average amount of fluid used was approximately 4 litres per day, for 12+2,3 days. Four of 24 patients (16,66) developed a recurrent intra-abdominal abscess, requiring re-operation. Total mortality was 8 of 24 patients (33,33 ), including 4 cases with pancreatic abscess, 3 with infected pancreatic necrosis. In 7 cases the death was the multiple organ failure syndrome and in another one it was massive pulmonary embolisation. Later complications were observed in 4 of 16 (25 ) survivals, which analytically were: 2 pancreatocutaneous fistulae, 1 pseudocyst formation and 2 exocrine pancreatic insufficiency.
We conclude that the surgical management in patients with acute necrotising pancreatitis is a difficult decision and requires clinical judgment with a thorough consideration of the clinical situation and other non-operative means of supportive care. The proposed method is necrosectomy accompanied by post operative lavage to avoid severe operative complications, such as severe intra-abdominal bleeding and bowel necrosis, which increases morbidity and mortality in these patients. pancreas may be post-inflammatory or post-traumatic. Elective operation is indicated if the pseudocyst is bigger than 4 cm in diameter and after unsuccessful conservative treatment for 6 weeks. The choice of operation depends on the size, nature, situation and type of the cyst.
In the 3d Surgical Department of "EVANGELISMOS" Hospital, 32 patients with pancreatic pseudocysts were operated on the last 10 years (1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992). Twenty of these were female and 12 male. Their age ranged between 25 to 65 years. The cause of the pseudocyst was acute pancreatitis in 24 patients, trauma in 2 and unknown in 5. Eight patients with spontaneous regression of the pseudocyst are not included in this study. Internal drainage was performed in 27 patients: in 4 patients a cystogastrostomy, in 22 a cysto-jejunostomy using a Roux-en-Y jejunal loop and in 1 a cysto-jejunostomy with a simple jejunal loop with jejuno-jejunal anastomosis (Braun) was performed. External drainage of the pseudocyst was done surgically in 3 patients because of infection. Two of them reoperated on after a few months using a Roux-en-Y jejunal loop and the other one has had spontaneous healing of the remaining cyst and the pancreatocutaneous fistulla. Two distal pancreatectomies without splenectomy were performed. There were no postoperative deaths in this series. The early and late results are good. It is concluded that in our opinion, if the removal of the pseudocyst is impossible, internal drainage using a Roux-en-Y jejunal loop is the treatment of choice. In conclusion, i) MOSF score is of similar value as the APACHE II score and the multiple-factor scoring systems in early identification of high risk patients with acute pancreatitis. 2) Only MOSF and APACHE II scores are available soon after admission and they allow close monitoring in these patients and may therefore help in deciding on therapy. In this work, the results of necrotizing pancreatitis surgery in 30 consecutive patients admitted to our Department of Surgery between January 1988 and October 1992 are presented. Twenty patients were males and i0 females, with an average age of 58.2+/-14.6 years (range 28-87). The mean hospitalization period was 36.7+/-31.4 days. Most frecuent among the etiologies presented was alcohol with 14 cases, followed by biliary lithiasis with 9, idiopathic 4, postoperative 2 and hyperlipidemia i. Twenty two patients presented pancreatic and/or peripancreatic infection either on admission or during hospitalization (13 PA, 7 IPN, and 2 IP). Surgical interventions (i to 5 per patient) included operations over the pancreatic area in 39 cases (28 drainages and/or lavages, 9 necrosectomies, 2 distal resections with splenectomy), 15 on gallbladder and biliary tract (4 cholecystectomies, 7 cholecystostomies, 4 drainages of the common bile duct) and 6 on the digestive tract (3 colostomies, 2 jejunostomies and 1 gastrostomy). Seventeen patients died.
Clinical, bacteriological, radiological findings and the results of treatment are analyzed. It is concluded that pancreatic infection secondary to necrotizing pancreatitis is associated with a high morbidity and mortality related to its clinical presentation (localized vs. diffuse, p=0.04). The APACHE II severity of illness index is considered to be an accurate predictor of mortality (p=0.0005). Semeiotica Chirurgica, University of Padova, Italy. Fistulas are a frequent complication in parcreatic surgery. We think that their treatment must include an adequate drainage, the functio nal suppression of the pancreatic gland, a careful evaluation of all the nutritional parameters and surgical treatment in selected cases. We performed all the above conservative techniques in order to achie vea good healing of the fistulas we observed. In addition we used a human fibrin sealant to fill their tracts. Our overall experience in the treatment of fistulas with fibrin sealant includes 9 enteric, 2 biliary, 1 vaginal and 13 pancreatic fistulas. Since 1984, 13 pancreatic fistulas underwent fibrin sealing: 6 followed a pancreaticoduodenectomy (2 for cancer, 2 for papillary carcinoma, 2 for endocri ne tumors); 2 after left pancreatectomy (i for chronic pancreatitis, 1 for cystoadenoma); 1 followed a pancreaticoeunostomy for chronic pancreatitis; 1 after excision of an insulinoma in the pancreatic head; 3 after surgery for acute pancreatitis (i necrosectomy and drainage, 1 percutaneous drainage of pseudocyst, 1 cysto-eunostomy). All our patients received an adequate nutritional support and had their pancreatic secretions reduced by pharmacological treat ment. Moreover, they were all submitted to repeated X-ray controls in order to position an accurate and proper drainage. As soon as a regular tract and a low outflow were achieved, the patient underwent the sealing treatment. We used a double lumen catheter under X-ray control which permitted a selective inection of the sealant at the origin of the fistula up to the skin. The tract was thereby completely filled. In ii cases we obtained a good healing with a single inection; 2 patients required 2 treatments. The sealant is selfshaping and its pressure prevents the out-flow of secretions through the fistula, diverting them intotheir natural channels. years with acute necrotizing pancreatitis were studied in a prospective clinical and laboratory trial. The treatment which had been proposed for the management of patients with pancreonecrosis included a wide variety of medical and surgical measures i.e. biliary drainage, early operative drainage of the pancreas, thoracic duct drainage and drainage of the peritoneal cavity if pefipancreatic infection was recognised. In 12 patients the gland was oedematous, usually with scattered areas of fat necrosis. In this group mortality was 16 per cent. In 14 patients there was gross retropefitoneal and peripancreatic haemorrhage, and in this group mortality was 50 per cent. In 8 patients the findings were classified as a pancreatic phlegmon, mortality was 32 per cent. Examination of lymphocyte function showed reduced cell percentages in peripheral blood and T cell percentage, a serum inhibitor of sheep red blood cell rosette formation by T cells.
Our results show significantly less T helper cell a day or three after operation. Proliferative reactions by T cell to PHA mitogen are decreased as compared to controls. Patients produce less antibody than controls. No statistical difference can be identified in IgA reslnse between 2 groups of patients with pyogenic complications or without them. Neutrophil chemotaxis and bactericidal function are decreased as sn as pyogenic complications are developed.
We have examined the levels of IL 1 and receptors on the surface on the sensitized lymphocytes in patients with pancreonecrosis, interactions between helper T cells, B cells and cytokines. These findings correlate statistically with an increased risk of death or purulent complications. The inclusion of immunocorrection (tactivin, thymogen, thymalin, thymoptin) in the complex of therapeutic measures after operation for acute panereatitis led to a fall of pancreonecrosis mortality from 33-50 to 29 per cent. We will present a clinical prospective study of one hundred cases of acute pancrcatitis, admitted to our service during a period of 35 months, trying to evaluate, above all, the criteria of diagnosis and the therapeutic options according to the various clinical presentations.
The diagnosis on clinical and/or ultrasonographic bases was not very precise, having a high number of diagnostic laparotomies as a consequence. A better diagnostic accuracy was achieved with the paracentesis or peritoneal lavage.
The evaluation by computed tomography was not possible because it was infrequently available at our emergency service.
Biliary surgery was the most common surgical treatment in acute pancrcatitis and direct pancreatic surgery was performed only in cases of necrotizing pancreatitis or in the treatment of pancreatic complications.
Morbidity and mortality rates depended, above all, on the initial severity of the pancreatitis, but intensive medical care of the "severe pancreatitis" and early surgical treatment of the local complications, had a favourable influence on the final outcome.
The overall mortality rate was 4 % in all cases as a result of pancreatic sepsis. In group A complete recovery was obtained without any adjunctive therapeutic procedure in 11 cases (1,6%),8 of wich developed pancreatic pseudocyst during the follow-up (6 months), 1 case was operated on for hemorrha gic and necroziting pancreatitis and died in 14th post operatory day.0f the cases of pancreatic pseudocyst 2 have already been successfully operated on;the third is still waiting for operation. in group B complete recovery was obtained in 10 cases (8,%, n.s. in camparison with group A) ;the other two cases,in wich clinical evaluation showed an necrosing severity,theraphy was changed and i.v. continuous perfu sion of somastatine was used respectively in 4th and th day.these two cases had complete recovery,but devepo_ led pancreatic pseudocyst during the follow-up (one has already been successfully operated). In conclusion, the treatment of acute edematous pancrea titis by means of i.v. somatostatine or s.c. longastatine had similar results.the incidence of late complications (psudocyst) wassimilar too. the preference accrded to longastatine in comparison with somastatine is based on an easier execution of the therapy and a higher cost containment. We retrospectively studied 18 patients with chronic pancreatitis associated with an enlarged fibrotic pancreatic head, who undet longitudinal pancreaticojejunostomy. All patients were alcohol abusers presenting with abdominal pain and loss of weight. Jaundice ws present in 10 patients, although alkaline phosphatase was elevated in 17 patients. DuodeDml obstruction was present in 2 cases. During surgical procedure the pancreatic head was 7.0cm or more, with retention cysts in 14 cases.
All patients had common bile duct obstruction. The head of the pancreas was cored out and the cysts were drained as related by Frey. In 13 cases a carefull and meticulous dissection of .the common bile duct in its intrapancreatic course was perfornd. However, in 2 patients with marked dilatation of choledocus and intense fibrosis w perfo a separate biliary bypass. In the follow-up the patients were submited to a control E.R.C.P. which demonstrated a complete drainage of the pancreatic duct. The pain relief was considered excelent in all but two patients who persisted in alcohol consume.  (1985 1992). Twenty-two (22 %) of these had some complication because of the necrotic process. Nine patients haemorrhaged (intra-abdominal in 5, upper gastro-intestinal tract bleexling in 4); necrosis of colon was seen in 5 patients, necrosis of the medial wall of the duodenum in 3 and isolated bile duct necrosis was seen in 2 patients and in one patient the necrosis involved the spleen.
A total of 34 surgical procedures had to be done in these 22 patients. Five (22 %) patients died, 3 because of haemorrhage and one each with duodenal and colonic necrosis. In contrast, in the 76 patients who had no complication, 4 (5 %) patients died.
Pancreatic necrosis is a serious disease with a potential for causing varied complications each of which needs a planned surgical approach. Of all the complications of pancreatic necrosis haemorrhage seems to be the most lethal and difficult to treat.

Mantova-Italy
The film presented the surgical treatment with pancreatico-jejunostomy in a 68 years old male patient suffering from chronic nonalcoholic pancreatitis with intractable pain,an alteration in his general condition and insulin-dependent diabetes for the last 8 years. The patient presents an obliterant arteriopathy affecting the lower limbs on an arteriosclerotic base. The CAT scan and preoperative ERCP revealed a calcified chronic pancreatitis with dilatation of Wirsung's duct,occupied by stones situated,above all,in the head. The most interesting points of the technique qre to locate the Wirsung duct by puncture and to perform a direct Wirsunggraphy with exact morphological definition, extraction with Fogarty balloon catheter of the pancreatic calcifications found within the duct and finally, fashioning of a one-layer, wide side-to-side transmesocolic wirsung-jejunostomy on a Roux-en-Y loop, that guarantees adeguate pancreatic drainage, in the presence of a firm stenosis of the duct in its cephalic trajectory. The patient was discharged on the 12th postoperative day, after ? days on total parenteral nutrition.
In the long-term out-patient controls, the patient referred to complete relief of pain and weight gain.

Athens, Greece
If the malignant tumor proves to irremovable, the majority of surgeons perform one of the many short-circuiting or decompressive operations.
The choise of these palliative procedures continues to be cholecystojejunostomy using the Roux-en-Y loop or the simple jejunal loop technique combined with side-to-side jejunojejunostomy or end-toside hepaticojejunostomy is carried out to circumvent the possibility of subsequent duodenal obstruction. Authors report a retrospective study of patients with primary pancreatic tumors who underwent palliative procedures in our department. They were 32 male and 22 female who ranged in age from 48 to 82 years. Surgical treatment included laparotomy with biopsy alone in 8 patients, choledochojejunostomy in 31 and cholecystojejunostomy in 15 patients. The mortality rate was 9,2%. In 38 patients a side-toside jejunojejunostomy was carried out. A number of surgeons consider that cholecystogastrostomy is the operation of choise for anatomical reasons. This procedure, however, has disadvantages. The results of the palliative operations are disappointing, as the average mortality rate is about 10% and within 5-9 months of the operation nearly all such patients are dead. When survival exceeds 1 year the question always arises as to whether or not the primary lesion in the pancreas was in fact cancerous. A novel method of pancreatic anastomosis following proximal Whippletype resection (classical PD or pylorus-preserving PPPD) has been evaluated over a five year period, 1987-92 in 52 patients.
Indicatations for resection included chronic pancreatitis (n = 9) and neoplasms (n 43). Reconstruction involved a cephalad end to end duodeno-/gastro-jejunal anastomosis with a biliary anastomosis 6-8cm downstream. A separate isolated, defunctioned Roux loop was used to construct a duct-to-mucosa pancreaticojejunostomy.
Median postoperative stay was 18.0 days (11-32). Three deaths (operative mortality 5.8%) occurred due to sepsis (subhepatic abscess), profound hypoglycaemia and necrotising pancreatitis respectively. These deaths were not related to pancreatic fistula. There were no pancreatic leaks (defined as > 50ml amylase-rich fluid for more than 7 days) Twenty patients considered to have normal pancreatic remnants underwent a PABA excretion test at 3-18 months after operation, median PABA excretion index as 48% (24-100%). Only 4 of 43 (9.3%) patients resected for neoplasm required pancreatic supplements after operation.
Isolated defunctioned duct-to-mucosa pancreaticojejunostomy is a safe procedure offering good functional results after Whipple's PD or PPPD resection. The extended excision of the ampulla of Vater was first performed by author in 1989. It was published as a new operative technique in Hepato-Gastroenterology in 1992.
The procedure comprises a pylorus-preserving resection of the descending segment of duodenum, excision of a large part of pancreatic head herewith concerning the anatomic pattern of common bile duct and pancreatic duct, which are running together up to the level of subpyloric region. Regional lymph nodes are removed and reconstruction with Roux-en-Y jejunal loop on the subpyloric part of the duodenum and both ducts sutured together, performed. Our first six cases are presented, patients for the operation being strongly selected. Their basic data, pathology, operative variations, postoperative course, morbidity as well as short term survival time are described.
There were four patients with ampullary tumors, one with periampullary tumor and one with duodenal carcinoma. Four of them are alive and well, the first operated patient, who had had recidivant ampullary carcinoma died because of the progress of the disease two years after the extended excision, and one patient died 34 days after the operation from heart attack.
According to our first experience the advantages of the described procedure are" it is more radical operation than simple excision, it is less mutilant than Whipple operation, with few possibility for complications, and resection as well as reconstruction can be simply performed.
Disantvantages may be: questionable radicality for some cases, problems with reconstruction when pancreatic duct is not dilated, possibility of common poscibial difficulties after pylorus preserving resections.
We conclude that more operations should be done for better evaluation of the value of this operative procedure. In the last year 3 young females were diagnosed pre-operatively as SPCENP due to better awareness of the condition. A retrospective analysis of pancreatic resection for cystic turnouts of the pancreas over a five year period (1987)(1988)(1989)(1990)(1991) was undertaken to identify SPCENP. Specific information was obtained regarding pre-operative diagnosis, imaging and the surgical approach. Only 4 more cases could be identified. This indicates referral pattern, as a large lesion in pancreas is generally associated with poor prognosis. In this series of seven female patients the age range was 16-40 years with two elderly patients aged 35 and 40, an unusual feature. Pre-operative diagnosis in earlier periods was adenocarcinoma of pancreas/cystic turnout ofpancreas. Imaging errors included diagnosis of liver abscess, pseudocyst. One patient underwent cystogastrostomy elsewhere due to mistaken diagnosis of pancreatic pseudocyst. Pre-operative histology was obtained in recent three cases by fine needle aspiration cytology. 6/7 lesions were resected completely by distal pancreatectomy. One patient underwent Whipple's resection for SPCENP of head of pancreas. There was no detectable recurrence in this series with follow up ranging from 3-60 months.
It is important to identify SPCENP as a distinct entity as resectability is good, with better prognosis. Endocrine tumours of the periampullary area and the head of the pancreas (PET) may require an extended surgical procexture for cure. Since 1968 we observed 37 patients with proven PET: 19 cases of insulinomas, 9 Zollinger-Ellison Syndrome and 9 "non functioning" tumours. 18/19 insulinomas were typical single adenomas of the head of the pancreas ranging in size from 1.0 to 4.0 cm. 14 were treated by local excision 2 with pancreato-duodenectomy (PD) and two with subtotal head resection; all patients were cured by surgery. One patient had a multiple endocrine neoplasia (MEN 1) and 4 adenomas were excised from the head of the pancreas. The patient is still normoglycaemic 8 years later. Five/9 patients with ZES had multiple tumours (2 had MEN 1) and 3 patients had microadenomas (size < 0.5 era). Five cases were treated with a PD, 4 had an excision of duodenal adenomas. In two cases ZES was not cured by surgery and one recurred 14 years after surgery.
Five patients had normal gastrinaemia after 5 16 years follow-up. Six/8 patients with non-functioning tumours underwent resective surgery. Two PD, 1 partial head and body resection for double adenoma, 2 local excisions and 1 partial dutenal resection. One of them died 6 months later for tumour progression, and 4 are disease free after 1 to 11 years follow-up. One/3 patients with liver metastases was still living 5 years after diagnosis.
While preservation of the disease free pancreas is the rule in insulinomas and extended predures are usually due to technical problems, in our series 4/19 (21%) had a pancreatic head resection at least subtotal.
Since long term survival in non beta PET is a frequent event these tumours, even if advanced, require an aggressive surgical approach, as a curative resection or a long term palliation can be frequently achieved. The septic risk after splenectomy, although a matter of controversy, must become a surgeon's concern in left pancreas resections for benign or traumatic diseases.
Twelve patients with benign tumors of the tail of the pancreas and one with a traumatic rupture of the neck of the pancreas underwent a left pancreatectomy with preservation of the spleen. The splenic pedicle was always resected along with the left pancreas. The pancreas is exposed by separating the omentum from the transverse colon taking care not to harm the short gastric vessels. The tail of the pancreas is mobilized and the splenic vessels are ligated and divided proximal to the splenic hilum. Left pancreatectomy is completed by division of the neck of the pancreas and of the medial segment of the splenic pedicle. Residual blood flow to the spleen originates from short gastric, gastroepiploic and left gastric arteries. This technique was described for pancreas donation from a living relative and is anatomically possible in most cases. It must become the standard in patients undergoing left pancreatectomy for a benign disease. Hypoglycemic disorders, due to an absolute or relative increase of insulin secretion, may represent a complex diagnostic and therapeutic problem in pediatric patients by presenting possible cerebral les ion due to the lack of glucose and ketons. Our statistic includes 7 patients aging from 3 months to 14 years, with stable or recurrent non ketogenetic hypoglycemia with convulsive crises beginning sometimes in the first hours of life. The operative sequence was: i-Assess the presence of hyperinsulinemic state (insulin plasma lev el> 150nu/ml, or lack of insulin secretion'suppression during hypoglycemiaor inadeguate secretion) by means of insulin/glucose index and also by looking at the metabolic effects of hyperinsulinism ( inibition of lipolisis, increase of cellular incorporation of aminoacids) or using stimulation testwith leucine or glucagone. We don't use suppression test.
2-Medical theraphy-We use parenteral administration of glucose to mantein blood glycemia over 40 mg/dl. The use of diazoxide doesn't give very good results. We don't use chlorpromazine or phenytoin or streptozocinand somatostatinfor the inibitor effect on other hormones. 3-Differential diagnosis between nesidioblastosis and B-cell adenoma: the study of proinsulin as biochemical markerwas an inconclusive diagnosis. Only in two cases (28,5%) we localized pre-operatively a B-cell adenoma( one by CAT and the other by arteriography). 4-Surgical theraphy: it was the ultimate theraphyin all cases. We found 3 -cell adenoma ( one in the tail, one in the head and one in the mesentery). In the other patients we performed a subtotal pancreatectomy for nesidioblastosis. The surgical procedure was done without removing the spleen. In all patients the followup shows a good metabolic control. Survival rates have been reported to be lower after orthotopic liver retransplantation (re-OLT) than primary liver transplantation (OLT). The aim of this retrospective study was to determine whether mortality rate was similar in cases of re-OLT compared to elective re-OLT. Between January 1988 and October 1991, 100 liver transplantations were performed in 85 children (mean age" 44.4 months). Fourteen of these patients (16%) underwent re-OLT. Five patients were retransplanted in elective conditions (group 1)" four patients had a second graft and one had two new grafts. Indications for retransplantation were four hepatic artery thrombosis (HAT) and two vanishing bile duct syndromes. In this group, all patients but one who had an elective re-OLT for HAT had had an attempt for immediate surgical revision. In group 2, nine children had a second graft in emergency for primary liver non function (n=5), HAT (n--3), portal vein thrombosis (n--l). In this group, no patient retransplanted for HAT had had an attempt for revascularization. The overall survival of patients retransplanted was 71%. In group 1, there was no death, while in group 2, 5 deaths (55%) occured. Our results suggest that, in paediatric liver transplantation, retransplantation might be performed in elective conditions when possible, as emergency retransplantation yields higher mortality rate. As HAT was the main cause for retransplantation, this emphasizes the interest of immediate surgical revision which might delay retransplantation until recipient state allows retransplantation in better conditions. P0169 SMALL PORTAL VEINS AND PAEDIATRIC ORTHOTOPIC LIVER TRANSPLANTATION. T. Yandza, F. Gauthier, J. Valayer. Surgical unit, Department of Pediatrics, Hopital Bicetre, Kremlin Bicetre, France.
Considered previously to be a contraindication for orthotopic liver transplantation (OLT), small portal veins are no longer a reason to reject paediatric patients who are candidates for OLT. The aim of this study was to determine the frequency of small portal veins and their effect on the outcome of paediatric OLT. Of 85 chidren (mean age" 44.4 months; range" 3 to 156 months) receiving 100 liver transplants, two groups were studied according to portal vein diameter" in group 1 (n=12), portal vein diameter was <4mm, and in group 2 (n=73), portal vein diameter was >4mm. The main indication for OLT was biliary atresia in both groups with a higher percentage in group 1 (9/12=75%) compared to group 2 (39/73=53%) (NS).
There was no significant difference between the two groups in terms of age and weight. In group 1, the mean portal vein diameter was 3 mm versus 6 mm in group 2 (p<0.001). Three children (25%) presented with portal vein thrombosis in group 1, leading to variceal bleeding episodes in two cases, and to death in one. There was no portal vein thrombosis in group 2 (p<0.002). In conclusion, small portal veins are encountered in 14% of paediatric OLT and are associated with a higher risk of portal vein thrombosis. Progressive decrease in portal vein diameter observed in children awaiting OLT should be an indication of rapid transplantation. 64 patients were Child A, 47 were Child B, 13 were Child C and 1 was undetermined. The types of shunts performed were 112 endto-side portal-caval anatomoses (89.6 %), 1 side-to-side portal caval anastomosis, 3 meso-caval anastomoses, 6 proximal splenorenal anastomoses, and 3 distal spleno-renal anastomoses. The end to side portal-caval anastomosis was the standard procedure, with the proximal spleno-renal shunt being preferred in the presence of hypersplenism, and the meso-caval shunt being performed prior to liver transplantation. Operative mortality was 8 % (n IO) and was correlated with the Child classification. Postoperative complications included ascites, variceal bleeding and encephalopathy in 5, 4 and ii patients respectively. Survival (Kaplan-Meier curve) at 1,5 and i0 years was 77 %, 40 %, and 18 % (operative mortality included). Survival was not related to the Child status but improved significantly in patients abstaining from alcohol. Late complications (13 %) included chronic ascites (n 3) and encephalopathy (n 13 ).
Despite success in the early follow-up period, the results of portal-systemic shunting are poor in the longer term, with high rates of late comlications. These are to be compared to the stable results obtained with liver transplantation. A randomized prospective trial could be considered.