key: cord-0006547-esnsa4u5
authors: nan
title: Abstracts 5(th) Tripartite Meeting Salzburg/Austria, September 9–11,1982
date: 1982
journal: Langenbecks Arch Chir
DOI: 10.1007/bf01279099
sha: 6c9ba3ab92c01158f7d600e535d46d78ced7cbf8
doc_id: 6547
cord_uid: esnsa4u5

nan

The gastric secretion capacity increases during the first year after all types of vagotomy which is assumed to be important for the development of recurrent ulceration. However, the cause of this reestablishment of the preoperative gastric function is not adequately explained. The purpose of this study was to investigate the vagal innervation of the parietal cell mass in dogs 11/2 year after truncal vagotomy (TV), selective gastric vagotomy (SGV) and parietal cell vagotomy (PCV). Material and methods: Mongrel dogs supplied with a gastric fistula were used. The groups comprised 5 dogs with TV, 6 dogs with SGV and 6 dogs with PCV.

The acid secretion was measured before and one month and one year after vagotomy following insulin and pentagastrin stimulations. Operative technique: A gastroduodenostomy was performed in addition to the gastric fistula operation in the dogs randomised for either TV or SGV. The vagotomy operations were performed after the prevagotomy secretion experiments. TV by a 4 cm resection of the main truncs through a left side thoracotomy and SGV as described by Amdrup [1] . The PCV dogs had no gastroduodenostomy added but the technique at the lower esophagus was identical with the SGV and the dissection of the lesser curvature was extended 1 cm distal of the sharp physiological and histologically determined borderline between the fundus and the antrum.

About 11/2 year after the vagotomy the dogs were sacrificed by an acute experiment. The persistent or the regenerated vagal innervation of the parietal cell area was mapped out by neutral red excretion elicited by electrical stimulation of either the thoracic or the cervical vagal nerves [2] . Results: One year after vagotomy no increase in insulin response could be demonstrated in the TV and SGV groups, but a gradual increase to about 50 % of prevagotomy output occured in all PCV dogs. Reliable neutral red experiments were performed in 2 TV, 6 SGV and 4 PCV dogs. The stimmulated neutral red excretion was scanty to the same extent at the cardia region in all three groups, but a distinct antralfundic border was outlined in all 4 PCV dogs. The border was, however, only suggested in one of the SGV dogs. Conclusion: PCV including denervation of 1 cm distal of the antral-fundic border do not prevent a functional important reinnervation of the distal part of the parietal cell area. No important reinnervation seems to occur at the cardia region after any of the three vagotomy procedures. 1 and insulin (0.2 ugkg -1) i.v. bolus; after uni-and then bilateral truncal vagotomy. Each study was 120 minutes and blood was taken at 1, 3, 5, 7 and then 10 min intervals. Gastric acid was measured by autobiuret titration. Plasma was stored at -20°C until assayed for PP by RIA using an antibody sensitive to 3 fmol ml-1. Results are expressed as mean total increment _+ SE. Significance: p < 0.01". Bombesin-stimulated gastric acid secretion was not significantly altered by vagotomy (p < 0.05), whereas that stimulated by insulin was significantly inhibited by bilateral truncal vagotomy (p < 0.01). Bilateral and right hemi-vagotomy significantly inhibited PP release by bombesin (17 < 0.01), however; only bilateral truncal vagotomy significantly inhibited PP release by insulin (p < 0.01). These results suggest that the measurement of PP release by insulin or bombesin is a sensitive index of vagal integrity and that bombesin-released PP may specifically delineate the integrity of the right vagus. Since the measurement of gastric acid secretion after operation is both uncomfortable and often difficult to interpret, the value of a simple blood test to determine vagal integrity may be of considerable clinical relevance. Glucose (G) or oleate (O) empty more slowly from the stomach than 0.15M NaC1 (S). This chemoregulation may be achieved by resistance beyond the proximal stomach [1] . We sougth to define the site of this resistance. Expt. 1: Gastric emptying of S and 600 mM G was measured in 5 dogs with intestinal cannulas 45 cm distal to the pylorus while gastric pressure was maintained at 10, 18 or 26 cm H20 with a barostat. The dogs were studied either with an uninterrupted intestinal stream or with duodenal effluent diverted through a second barostat prior to its return downstream. Since the latter ensured a constant pressure at the ligament of Treitz, any observed effects on emptying could not be due to resistance further distally. Expt. 2: Emptying of S, G, and 40 mM oleate emulsion was measurd in 5 dogs without cannulas after control antroduodenal transection and after antrectomy while gastric pressure was controlled at 11, 17 and 23 cm H20. Emptying rates expressed as the average of the volumes emptied at the 3 pressures.

Under both conditions in Expt. 1 emptying rose with pressure yet S emptied faster than G. Thus a major site of chemoselective resistance to emptying of liquids lies proximal to the ligament of Treitz. After antrectomy S and O emptied faster than before, however, the differences between saline and nutrients were maintained. Resistance, therefore, does not reside in the antrum or pylorus and clearly the duodenum is implicated. Gastric emptying of liquids may be abnormally rapid in DU. This abnormality my be due to a selective loss of inhibition in response to acid since previous studies [1] have demonstrated rapid emptying of acid solutions but normal emptying of glucose and fat. Previous investigations, however, used meals of only one concentration, so differences between DUs and normals (N) may have been minimized by selection of a supra-maximal inhibitory dose. Furthermore intragastric pH was not maintained constant so emptying could have been slowed by higher rates of acidification in DU. We therefore studied emptying of graded concentrations of citric acid, glucose and oleate in 500 ml meals. Gastric volumes were measured by the George technique in DUs and N at 5 rain intervals for 30 min after ingestion. Gastric pH was maintained constant by intragastric titration (3.0 for acid; 7.0 for nutrients). Subjects received one type of meal on separate days and the doses were randomly administered. Dose dependent inhibition of emptying in response to acid and nutrients occured in N and DU, but emptying was faster in DU (p<0.05). With glucose and acid the difference occured within 0-5 min whereas with fat the difference occurred between 5-30 min. Thus (1) DUs do not have a selective loss of inhibition of emptying in response to acid, and (2) the mechanisms which control emptying of fat and which are disturbed in DU differ from those which control emptying of glucose and acid. 1 Gross RA, Isenberg JI, Hogan D, Samloff IM (1978) The effect of fat on meal-stimulated duodenal acid load, duodenal pepsin load and serum gastrin in duodenal ulcer and normal subjects. A means of reducing the requirement for postoperative nasogastric intubation would be advantageous. We report the effect of cimetidine on postoperative nasogastric aspirates. Thirty patients undergoing elective abdominal aortic surgery or colonic resection were entered into a double blind randomised study, receiving cimetidine 200 mg 4 hourly or placebo (saline) by continuous intravenous infusion from the morning of the first postoperative day. Volumes of 4 hourly aspirates were recorded and pH and (H ÷) of samples measured. Nasogastric tube removal accorded to standard clinical practice. It has been shown, that the duodenum has a better ability to resist acid than jejunum or ileum. This resistance was attributed to alkaline secretion (a.s.) of the mucosa, because pancreatic and bile secretion were excluded. Recent in vitro studies demonstrated an energy dependent HCO3-transport (Flemstroem 1980 Am J Physiol G 239: 198). We were interested in studying the role of blood flow (bf) in the production of alkali in an in vivo preparation. min. Three groups of animals were studied: I) Controis with normovolemia for 120 min. II) Normovolemia for 60 min, thereafter 60 min of shock induced by bleeding to 40 mmHg mean arterial blood pressure. III) 60 min normovolemia and 60 min vasopressin (0.2 I. U./kg/min) under normovolemia.

In controls a small decrease in bfand a.s. as well was observed. Both shock and vasopressin induced a significant reduction in a.s. and bf. The relation ofbf to a.s. was found to be exponential (y = 2.71 + 0.21x-0.003x 2 r= 0.71; p<0.05). The reduction in a.s. during shock was much less, when shock induced acidosis was compensated by i.v.-infusion of HCO 3 (0.3 ,uEqu/kg/min). Conclusion: The alkaline secretion of the proximal duodenum is dependent on blood flow and the arterial [HCO~. There has been considerable controversy as to the nature of the offending agent in the etiology in reflux oesophagitis in man. In rat studies, surgically induced reflux oesophagitis was shown to be correlated with the presence of active trypsin in the reflucting juice. Reflux of bile and gastric juice with a pH of 4 did not induce oesophagitis. Even short periods of oesophagitis (1-6 weeks) showed an increasing amount ofpanmural fibrosis which even further increased after a reflux abolishing Roux-Y operation. The pattern of the collagen content of the full thikkness oesophageal wall was similar to that found in human reflux oesophagitis.

The hypothesis that active trypsin is also in man responsible for reflux oesophagitis prompted to the following investigation.

24 patients with typical gastrooesophageal reflux symptoms and 17 control patients were endoscopically examined. On entering the stomach with the endoscope samples of gastric juice were taken for pH and active trypsin determinations. Trypsin activity was determined with a kinetic method using S-2160 (Kabi Vitrum B.V. Amsterdam) as substrate. Results: In all patients some trypsin could be detected in the gastric juice. In those patients with endoscopically erosive and/or ulcerative changes (n = 14) the trypsin content was significantly elevated compared to the controls (/7<0.05 Wilcoxon). Conclusions: The composition of gastric juice is determined by gastric secretion and duodenogastric reflux. Duodenogastric reflux is increased in patients with symptoms of reflux oesophagitis [1] . Trypsin in gastric juice (pH 3.5-7) will stable and partly active over prolonged periods [2] .

So from our study in man one may conclude that the high concentration of trypsin in the gastric juice in patients with reflux oesophagitis, may well be the etiological factor of the oesophagitis. disinfection, using artificial contamination of hands with Escherichia coil and 'surgical' disinfection directed against the resident microflora of the hands.

The authors who developed the procedure have reported unexpectedly high potency for n-propanol compared to iso-propanol, and, in turn, for iso-propanol compared to povidone-iodine and chlorhexidine preparations [1] . However, using the same procedure we have been unable to find any significant differences in activity between these agents. We have identified several potential sources of error including the method of applying E. coliand its spontaneous loss of viability, the use of neutralizers before disinfection, the differing surfactant effects of the agents, and the absence both of a control untreated area and of a cross-over of disinfectants studied sequentially.

In our parallel tests using an excision-sample technique [2] which is considerably more sensitive than the DGHM procedure, we have observed the following mean reductions in the counts of accessible bacteria: iodine in ethanol, 96%; povidone-iodine, 89%; chlorhexidine in ethanol, 88%; iso-propanol, The purpose of this study was to compare radiation injury in Guinea Pig small bowel (1) devoid of contents (2) containing bile (3) containing pancreatic juice. One group was intact control animals not radiated. Another was intact animals radiated. In Group 3 a Roux Y cholecystojejunostomy was constructed and the bile duct ligated. Thus one limb contained bile, the other pancreatic juice. In Group 4 a blind Roux Y was constructed such that one limb was empty of contents and the other contained both bile and pancreatic juice. Each animal was subjected to a single dose of 1600 rads via an abdominal port and sacrificed four days later. The severity of injury was judged by counting the number of surviving crypts per circumference. Damage was further evaluated by histologic criteria -mucosal loss, edema, inflammation, hypervascularity and loss of mucus. The maximum histologlc grading score on this scale was 20. The microscopist was ,,blinded" as to the origin of each tissue section. For histologic grading all radiated groups were significantly different from control (p<0.02) and from one another (/r<0.05) except pancreatic juice vs. empty. Intestine devoid of contents sustains a mild, but significant radiation injury. The presence of pancreatic juice enhances the damage. Pure bile makes for yet more severe injury but still significantly less than normal whole intestinal contents which contain both bile and pancreatic juice. The main problem facing patients with ulcerative colitis after mucosal proctectomy and ileo-anal anastomosis is severe frequency of bowel action. Our hypothesis was that either an artificial valve [1] or reversed ileal loop might improve intestinal absorption and slow transit. We tested it in dogs after colectomy and low ileo-rectal anastomosis (IRA).

Body weight, xylose absorption, serum albumin, folate, vitamin B12, calcium, urea and electrolytes, full blood count, faecal weight and mouth to anus transit time [2] were measured before operation. The dogs then randomly underwent either IRA alone (control, C, n = 7), IRA with a 10 cm reversed loop (IRA + RL, n = 6) or IRA with an artifical valve (IRA + v, n = 6). Body weight, haematological estimations and faecal chemistry were measured weekly for 3 months post operatively. Xylose absorption and transit time were measured a minimum of 2 months after operation.

Body weight decreased significantly after each type of operation. There was no difference however, between (IRA + RL) and C or (IRA + v) and C. Likewise, haematological and faecal measurements after both test operations did not differ significantly from C or from pre-op, measurements. The Grosfeld valve prevented the reduction in transit time that occurred after the control operation, whereas the reversed loop did not.

Use of such a valve in combination with mucosal proctectomy might slow transit, but is unlikely to improve absorption. It seems worthy of further study.

Histologic grading___ SEM Crypt count-+-SEM We have examined the effects of changes in intestinal blood flow induced by hypovolaemia, the mesenteric vasoconstrictors somatostatin and vasopressin and the mesenteric vasodilator prostaglandin E1 (PGE1), on intestinal absorption, electrical activity and volume. In each of 7 dogs a 75 cm jejunal segment was isolated and serosal electrodes attached. In absorption experiments the segments were perfused at 2.9 ml/ rain with a solution of 90 retool/1 sodium and 100 retool/1 glucose. To measure segment volume, a solution of 50.7 g/1 mannitol was perfused. This solution shows zero net absorption but does not alter electrical activity. Intestinal motility and absolute volume were monitored by gamma camera. Experiments were performed after a steady state was reached. Absorption and transit time were measured using non-absorbable markers.

Electrical fast wave activity was reduced by (a) intravenous somatostatin 2.5 mcg/kg/h (28+1.4 to 11_+3.4; mean_+SEM/5 rain) and abolished by (b) intravenous vasopressin 1.2 U/kg/h (p<0.001). Motility was inhibited and mean transit time was prolonged (a) 7.4 to 15.3 rain and (b) 4.4 to 19.9 rain (/r<0.001). Intestinal volume rose by (a) 12.3___ 3.7 ml and (b) 15.4-+4.7 ml (p<0,001). Absorption was unchanged by somatostatin and fell with vasopressin (5.58---0.7 to 4.35-+0.68 ml/5 min). Acute blood loss sufficient to lower the central venous pressure by 5 cm of water produced identical changes to somatostatin. PGE1 produced no alteration in electrical activity or absorption.

Intestinal absorption is resistant to changes in intestinal blood flow whilst electrical activity is depressed and intestinal volume increased by mesentric vasoconstriction. The accepted views on the pathogenesis of amoebic lesions in complicated amoebic colitis are that amoebae produce a toxin resulting in cytolysis and necrosis. This concept may adversely affect the management of patients with complicated amoebic colitis by implying that once the amoebae are killed, the disease process is arrested and colonic perforation is unlikely.

We present an alternative hypothesis that 'complicated amoebic colitis is the result of vascular compromise'. Transmural disease is caused by amoebic invasion of vessels supplying a segment of the colon with subsequent thrombosis and ischaemic necrosis of the affected area. The ischaemic nature of the necrosis is suggested by its shape, and the demonstration of vascular thrombosis on dissection ofresected colons which have perforated. Amoebic invasion of blood vessels can be demonstrated histologically. The ischaemic nature of the lesions can be confirmed by angiographic examination of the resected colon.

Vascular occlusion can be demonstrated pre-operatively with the use of selective mesenteric angiography, which clearly delineates the ischaemic segment of the colon. Selective rnesenteric angiography in 27 patients with fulminating amoebic colitis aided the preoperative diagnosis of colonic infarction before signs of visceral perforation had developed and permitted life saving surgical intervention. There is accurate correlation between angiographic and operative findings.

In postdysenteric colitits associated with amoebic strictures of the colon, mesenteric angiography will demonstrate the ischaemic nature of the stricture and accurately define the extent of resection. after CA. The mean age at operation was 33 and 61 years respectively.

Anal canal length was 3.5___0.1 cm (mean + s.e.m.) after IA and 3.7___0.1 cm after CA. A resting tone of 66___ 3 cm water after IA and 53 -----10 cm water after CAwas recorded. Squeeze pressure was 123___9 cm water after IA and 79-----3 cm water after CA. The recto-anal reflex was present in 54% of the IA patients and in 75 % of the CA patients. A mean daily bowel frequency of 3.5___0.2 followed IA and 4+_1 after CA.

These findings show that after IA anal pressures are within the normal reference range for our laboratory [2] . Lower pressures were found after CA, probably due to the older age of the patients in this group [3] . Normal anal canal length, the integritiy of the striated sphincter and in most cases preservation of the recto-anal reflex contribute to satisfactory continence following peranal anastomoses. Perineal descent is found in patients with idiopathic faecal incontinence (IFI) and patients presenting with symptoms of the descending perineum syndrome (DPS), who have no incontinence. Manometric, radiological and neurophysiological studies were performed in 17 patients with IFI, all of whom leaked during rectal infusion of 1500 ml saline, 18 patients with DPS, who were continent of rectally infused saline, and 14 matched controls. Both patient groups exhibited similar degrees of perineal descent below the pubococcygeal line of straining (IFI, -5.1___2.1 cm; DPS, -5.9 + 1.7 cm) compared with controis (-2.4__+ 1.5 cm;p<0.005 in both cases), and similar prolongation of both the latency of the cutaneoanal reflex (IFI, 13___5 ms; DPS, 16+6 ms; controls, 9+4 ms;/,<0.005 in both cases), and motor unit potential duration (IFI, 9.1___2.6 ms; DPS, 8.4___2.0 ms, control, 6.7+1.5 ms; p<0.01 in both cases). Moreover, both groups had an abnormal ano-rectal angle (/,<0.005), though this was more obtuse in IFI than DPS (120___21 ° vs 107___ 15°;p<0.05). However, while patients with IFI had lower sphincter pressures than normal (basal; 39___17 vs 81___29 cm water; /7<0.005; squeeze, 96___18 vs 208___84 cm water; /7<0.001), and required a lower rectal volume to inhibit sphincter tone for more than one minute (32 + 18 vs 76___29 ml; p<0.00l), these values were normal in patients with DPS (basal; 87___ 30 cm water; squeeze, 198_+90 cm water; rectal volume, 52___46 ml). These findings suggest that perineal descent and neuropathy are not necessarily associated with incontinence if sphincter pressures remain normal. Aminoacids administered either enterally or intravenously stimulate gastric secretion in both dog and man. It has been suggested that absorption of amino acids from the gut into the circulation might contribute to the intestinal phase of gastric secretion. The mechanism of this stimulation by amino acids remains uncertain but in an earlier communication we reported the direct action on the parietal cell of phenylalanine, glutamine and alanine [1] . In the present study using (14C) aminopyrine uptake as an index of dispersed parietal cell secretory function, the interaction between histamine and individual amino acids and the effect of the histamine H2-receptor blocker ranitidine (10-4 molar) on these interactions have been examined. Results (percentage of maximal stimulation produced by 10-4 molar histamine)

Results (percentage of maximal stimulation produced by 10-4molar histamine)

Methionine Glutamine Alanine

Amino acids only (10-2molar) 26 32 29

Amino acids and histamine (10-4molar) 175 -137

Amino acids and ranitidine (10-4molar) 10 17 7

Histamine and ranitidine -complete inhibition -25 125 1. The response to the combination of amino acids and histamine was greater than the sum of the maximal responses to each of these agents alone.

2. The histamine H2-receptor antagonist ranitidine completely inhibited the histamine stimulated response but only partially inhibited secretion stimulated by amino acids.

We conclude that amino acids act directly on the parietal cell by a mechanism which is not solely dependent on histamine. 1 Sex related differences in gastric acid secretion have been documented (Lilja et al, 1967) . It has been also reported that male rats have a significantly higher serum gastrin concentration than females and that oophorectomy increases serum gastrin to male levels (Lichtenberger et al., 1976) . Estimation of serum gastrin levels after administration of androgens has not been reported.

Aim. The aim of this study was to investigate the serum gastrin levels before and after oophorectomy and administration of estrogens and testosterone in female guinea pigs.

Method. 4 Groups of guinea pigs were used for this study, 8 animals as controls, 6 were given estrogens for 2 weeks, 8 were given testosterone for 2 weeks and 5 underwent oophorectomy.

Results. The mean serum (-+SEM) gastrin value in controls was found 35+ 1.63 pg/ml, after estrogens 13.33-+1.22 pg/ml, after testosterone 55_+1.89 pg/ ml, after oophorectomy 46-+2.45 pg/ml. Conclusion. It is concluded that estrogens decrease the serum gastrin, and that the increase of serum gastrin after administration of androgens and oophorectomy is statistically significant (/r<0.01). The mechanism by which the ovarian hormones influence gastrin levels may be a sex-dependent change in the synthesis or secretion of gastrin. The reported inhibitory influence of estrogen on food consumption may be also the predominant factor in serum gastrin alterations. Mean caloric intake was 2336--+427 kcal/day. Moderate to severe dumping occurred in one patient. Faecal fat was normal in 11 while 7 patients had greater than 10 grams fat loss per day.

Weight loss was 14.6_+8.6% of recall weight. Muscle mass measured by arm muscle circumference (25.7_+3.9 cm) and creatinine heigth index (99___26%) was normal for our patients but triceps skin fold was 30% less than normal (8.6_+2.2 mm) indicating that in these patients fat stores were reduced. In six patients haemoglobin was less than 13 grams per cent but serum iron, iron binding capacity and serum folate were normal. Red cell folate was depressed in six patients whose haemoglobin values were normal.

It is concluded that with this reconstruction total gastrectomy produces satisfactory digestive and "nutritional results. Examination of factors affecting faecal fat loss and folate metabolism may help improve the nutritional status of these patients. In order to assess the optimal conditions for segmental pancreatic graft viability the following experiment was conducted. The duct obliterated splenic lobe of the canine pancreas was autotransplanted by end to side fashion to the femoral vessels (n = 10). The graft was lodged in a subcutaneous pocket. A month later total pancreatectomy was completed. The dogs were supplemented with pancreatic enzymes (Viokase). Mean survival was 42___ 12 days. Deaths were due to hyperglycemia when grafts lysed from local infection and thrombosis. In the second experimental group (n= 10) the autografts were anastomosed to the iliac vessels and placed intraperitoneally followed by total pancreatectomy one month later. Construction of a distal splenic arterio-venous fistula increased blood flow from 31.4 cc/min to 108 cc/min. The pancreatic duct was obliterated with neoprene (n=4) silastic (n=3) or left open (n=3). Three dogs died one month to four months due to graft fibrosis and failure. Seven dogs had been followed from four months up to seven months when sacrificed. All were normoglycemic. Mean K values for IVGTT were 1.5. The normoglycemic dogs had mean plasma concentrations of amino acids similar to healthy controls. A two to threefold elevation was observed in pancreatectomized diabetic dogs for plasma leucine, isoleucine and valine. Immunoreactivity of pancreatic polypeptide was measured by radioimmunoassay with an antibody raised against the human hormone by RE. Chance, Lilly Research Laboratories. Mean plasma levels rose form 256--+28 S.E. pg/ml to over 1000 pg/ml following protein meal (20g ground lean beef/kg) in all normal control dogs (n=4), and to levels in the range of 390 to over 1000 pg/ml in the same dogs following infusion ofbombesin (1 pg/kg/ h) for 15 min. Mean basal levels were lower (42___ 2.2 S.E. pg/ml) in dogs with autotransplants and did not increase significantly during any stimulatory test. The levels of pancreatic polypeptide did not distinguish between viable and failing grafts. In successful grafts histology showed fibrosis of the exocrine component. Transmission electron microscopy revealed the presence of clusters of functional islet cells in the six-month implants. A and B cells were common. D cells were much less frequent. Release of secretion granules into the perivascular connective tissue space was observed. The results indicate that vascularized segmental pancreatic graft comprising 25 % to 30 % of the pancreatic mass maintains normal carbohydrate and amino acid metabolism. Graft survival was extended by intraperitoneal location and creation od distal splenic A-V fistula, but chronic graft fibrosis occurred regardless of the method of pancreatic duct treatment. The results further suggest the necessity of an intact vagal innervation for the physiological or pharmacological stimulation of pancreatic polypeptide release from thendocrine pancreas which is otherwise apparently functional. Besides, these considerations rule out pancreatic polypeptide levels as a useful marker for evaluation of pancreas graft. Polyisoprene ductal occlusion allows segmental pancreatic transplantation in man with satisfactory early islet cell function. However acute rejection remains a problem as the diagnosis, based on hyperglycaemia and glycosuria, represents a late manifestation of rejection. Although llqndiumlabelled platelets have been used in the diagnosis of renal rejection [1] , their use in pancreas transplants has not been studied. ~ In 6 patients, autologous llqn-labelled platelets were injected on the second, sixth and tenth days following combined renal and pancreatic transplantation. Graft radioactivity was expressed as the ratio of counts from the pancreas over a reference area on daily gamma images for 14 days.

One patient developed hyperglycaemia coinciding with renal rejection on day 5. Over the previous 24 h pancreas radioactivity had increased by 94 %. In a further patient whose pancreas continued to function, a perigraft haematoma was recognised on the gamma image. The remaining 4 patients had good pancreatic function and no 11XIn-platelet accumulation: Mean (_s.c. mean) pancreatic radioactivity was 1.76+0.190 on the third postoperative day and 1.75_+0.135 at the end of study.

These results demonstrate that ductal occlusion with polyisoprene does not cause significant platelet accumulation. Hence mIn-platelets may potentially be used for the earlier diagnosis of pancreatic rejection. Isograft models of pancreatic transplantation, methods involving closure of the duct system result in severe inflammatory changes and eventual fibrosis [1] . These changes can be avoided by formal drainage of the duct into the bowel of urinary tract. Inflammatory changes initiated by duct closure may contribute to and enhance allograft rejection. Pancreas transplants were performed in rats using Lewis (RT11) donors and streptozotocin-induced dia-

betic DA(RT1 a) recipients, using duct-ligation, open duct and ureteric duct drainage.

Cyclophosphamide produced significant prolongation ofnormoglycaemia in all groups and although there was a tendency towards longer function in the unligated groups there was not statistical difference (Wilcoxon's unpaired test) between the methods of duct management.

Management of the pancreatic duct did not seem to have any immunological consequences for graft survival but septic complications, associated with the normoglycaemic deaths, were more common in immunosuppressed animals with draining ducts. Between July 1, 1980 and March 30, 1982 combined pancreatic and kidney transplantation was performed in 9 patients with type I diabetes who had all been on insulin therapy for at least 16 years. Age at the time of transplantation was 28 to 45 years. The pancreatic segment used for transplantation consisted of body and tail of the organ based on a vascular pedicle of the celiac axis and the splenic vein. Imediately prior to intraperitoneal transplantation to the iliac vessels the ductal system was filled with 4 to 6 ml of prolamine, a rapidly solidifying alcoholic protein solution. Simultaneous kidney transplantation was performed through a separate incision. Five pancreas transplants were lost for non-immunological reasons. Four of them never had any useful endocrine function, one graft was lost of venous thrombosis after 48 hours of excellent function. In none of these patients did failure of the graft lead to any substantial complication.

In the 4 remaining patients initial graft function was excellent with plasma glucose normalizing within 12 hours and subsequent normoglycemia on a regular diet without exogenous insulin administration. Plasma glucose did not rise spontaneously during rejection episodes of the kidney and eleva-tions due antirejection treatment were promptly; reversed as high dose prednisolone was discontinued. In oral glucose tolerance test (OGTT) performed at 4 to 6 weeks in 3 patients median glucose rose from a basal 5.1 mmol/l to 8.6 mmol/1 and was back in normal range (4.7 mmol/1) at 3 h. Basal insulin was 43-237 pmol/1 and also returned to the normal range after 3 h after a peak of 201-531 pmol/1 at 50-210 rain. C-peptide showed a significant 2 peaked rise over basal values (230, 233 pmol/1) in 2 patients while all values were above 4500 pmol/l in the third. Two of these patients have since rejected both organs. Two tranplants still function very well after 12 and 21 months, respectively. In one of these patients the OGTT after 13 months is even slightly better than the above -mentioned first test, and he shows a norreal insulin and C-peptide response. In the fourth patient an IVGTT performed at 7 months and an OGTT at sixteen months were normal with insulin peaking at 251 pmol/1 and a C-peptide peak of 1267 pmol/1 at 60 min in the latter. -Results at 26 and 17 months, respectively, will be presented.

Three problems persist in clinical organ preservation. These are failure of current systems to replenish the ischemically injured organ, to reliably extend the period of organ preservation and to definitely determine organ viability. Previous studies have documented the value of electrochemical redox control in rejuvenation of the ischemically injured kidney during perfusion preservation. This study was undertaken to develop the cost effective technique applicable to current organ preservation systems, to test the reliability of redox measurement in prediction of organ viability and to determine the ability ofredox maintenance to safely extend the period of organ preservation. A disposable cell has been developed using Reticulated Vitreous Carbon as an electrode which is driven by a potentiostat powered by a nine volt transistor battery. Short term preservation studies using ischemically injured dog kidneys (60 rain in-situ warm ischemia) were auto transplanted after 24 h of pulsatile preservation to determine the optimum redox level of the hypothermic kidney. This was determined to be a -20 mv vs the standard calomel electrode. Twenty-one adult mongrel dogs were equally divided into three groups. Pulsatile perfusion preservation was extended to four days in Group A with redox level monitored only. Group B was treated with electrochemical reduction for four days and Group C for six days. Three of seven dogs survived in Group A following auto transplantation and immediate contralateral nephrectomy. The kidneys of all survivors were able to bring perfusate redox potential under control and maintain this level throughout the preservation period. Six of seven dogs in Group B survived with a mean posttransplant peak Serum Creatinine of 4.5 rag/100 ml. In Group C four of seven kidneys supported life immediately. All redox controlled kidneys made copious amounts of urine. Our data indicate that perfusate potential may be either monitored as a reliable index of organ viability or controlled to allow extended safe preservation. Antithymocyte globulin (ATG) has been shown to be an effective agent in combination with increased doses of steroids in reversing renal allograft rejection. Since it is frequently undesirable to employ raised doses of steroids, the following study evaluated the effect of 15-21 i.v. daily doses of horse ATG alone without additional steroids, on 2nd-5th rejection episodes in 10 recipients of cadaveric renal allografts (28 rejections: 25 biopsy-proven) detected within 3-18 mos. of transplantation. Of 28 rejection episodes, 25 were successfully reversed, with return of serum creatinine to pre-rejection levels in 18 episodes (7 patients). Three patients had primarily humoral rejections and returned to dialysis 2-4 mos. after treatment of the last rejection. Levels of circulating horse immunoglobulins were obtained in all 10 patients during and follwing administration of ATG. Recurrent rejections (3rd-5th) following last treatment with ATG (3-11 mos.) were seen in 5/10 patients. All 5 patients had rapid immune eliminiation of ATG (1/2 life 2-4 days) as compared to the 5 patients who had no more than 2 rejection episodes (1/2 life 6-14 days). ATG without additional steroids is an effective agent for reversal of multiple renal allograft rejections which by biopsy are primarily cell-mediated. To be effective, heterologous ATG must be given in adequate total doses and/or from appropriate heterologous source, to prevent rapid immune elimination by the recipient. The use of ATG alone for treatment of recurrent allograft rejections is particularly recommended for its steroidsparing effect in treatment of multiple rejections and for those patients at high risk from steroid side effects.

The Significance of the Monocyte Crossmatch in LRD Recipients of HLA Identical Kidney Grafts J. Cerilli, L. Brasile and S. Rogers Ohio State University Hospitals, Columbus, Ohio, USA In a preliminary study from our Transplant Center, the presence of pre-formed antibody in recipient sera directed against monocytes from their respective living-related donors correlated with a poor clinical course. A poor prognosis for graft survival was found regardless of the HLA match grade. To minimize the role of the HLA system, only those living-related recipient/donor pairs who were HLA identical at the A, B, C, D and DR antigen loci, and who exhibited severe immunological types of rejection were evaluated. Due to the small numbers found in this category at any one center, this abstract represents an international study from 12 different transplant centers.

26 patients who met the criteria were studied for the presence of antibody directed against their respective donor's monocytes both pre-and posttransplant. In eighteen of these patients, cytotoxic antibody against their donor's monocytes was found in their pre-transplant sera. There was no detectable cytotoxic activity against their donor's T or B lymphocytes. Two additional transplant recipients exhibited this antibody in post-transplant sera. Again, no T or B lymphocyte cytotoxicity was detected. A control group of HLA identically matched siblings who incurred no or minimal rejection demonstrated no anti-donor monocyte antibody.

The results of this international study points towards a correlation between a high incidence of graft rejection and the presence of antibody directed against their respective donor's monocytes. Therefore, in our view, the presence of anti-monocyte antibody to the prospective donor pre-transplant is a contraindication to transplantation. In patients with different liver diseases the reduced concentration of the peripheral blood T-cells and altered immune response were observed. The purpose of our studies was to investigate the mechanism of the immunological modification of the lymphoid system in the hepatic injury.

Studies were carried out in 3 groups: 1) LEW rates were treated with dimethylnitrosamine (DMNA), 35 mg/kg b.w., i.v. for acute liver necrosis induction, 2) CC14 (0.1 ml/100g b.w., i.v.) twice weekly, over 4 weeks for chronic liver damage, 3) CCl 4 over 10 weeks for liver cirrhosis (histologically examined). Serum and thymus, spleen, mesenteric lymph nodes (LN) were removed 2 and 32 days after DMNA treatment and 5 and 35 days after last CC14 injection. The total number of thymocytes and spleen cells was counted and the reactivity to PHA and ConA was measured. Normal liver perfusate (LP) was prepared 6 h after removing (20°C) by 5 times perfusion in 30 rain intervals (flow rate 1 ml/g/min with 2 ml/g of Ringer). The effect ofLP, sera, DMNA and CC14 on the viability and PHA response of normal thymocytes was tested in vitro. Liver enzymes were evaluated,

We found the decreased total number of thymocytes immediately after the stopping of medication (group 1: 9,2___ 1,8 %, group 2: 4,6___ 1%, group 3: less than 1% of control). Proliferative response of the remaining cells in thymus to PHA was much higher than normal thymocytes (group 1:325 ± 30 % group 2: 608___164%, group 3: ND). The total number of spleen cells was not changed but their response to ConA was altered in the all groups and to PHA only in group 3. PHA response of LN-cells was decreased in the all groups.

Liver perfusate was cytotoxic (53___3 % of viable cells) and suppressive for PHA response of thymocytes (6_+5% of control). The sera of rats with the hepatic damage showed an enhanced suppressive activity. CC14 and DMNA did not have any effect on thymocyte in vitro. One month after last medication we observed the recovery of thymus involution (total i in the acute and partial in the chronic hepatic damage and cirrhosis).

Our results suggest that the liver origin cytotoxic and immunosuppressiv e factor(s) can be released from the damaged liver into the circulation (like to LP) and can destroy the thymus leading to the secondary changes in the other lymphoid organs. The grade of thymus alteration is dependent on the degree and the duration of hepatic damage and is reversable. The hepatic factor (s) showed the similar effect on the cortical population of thymocytes like the steroid immunosuppressants. Liver grafts in the rat are in certain strain combinations not rejected and in this situation there is evidence for spontaneous donor specific tolerance [1] .

We have developed a model of auxiliary liver transplantation which would allow us to study the immunosuppressive properties apparently produced by a liver allograft.

The portal vein is anastomosed to the left renal artery, the I.V.C. to the renal vein and the bile duct to the ureter. Simultaneous kidney or heart allografts were performed. Conclusions:Auxiliary liver grafts are rejected. Survival of heart or kidney grafts is not influenced by a simultaneous kidney or heart allograft. Heart and kidney grafts are prolonged by simultaneous auxiliary liver grafts. [1] . The role of suppressor cells in the initiation and propagation of malignant tumours in man, is less clearly defined. The present study, using in vitro mitogen assays (PHA, CONA, PWM) and various rosetting assays [2] with specific monoclonal antibodies to lymphocytic helper (OKT4) and suppressor (OKT8) cells, has revealed the presence of such suppressor lymphocytes in women with clinically localised (breast and axilla) mammary carcinoma.

Lymplaocyte hyporeactivity to mitogens was found in 30% of lymphocyte preparations from blood and axillary lymph nodes of patients with breast cancer. In 10% of patients nodal lymphocytes were totally anergic. The most profound hyporeactivity, however, was detected in the lymphocyte subsets (<75 %) of specimens isolated from the breast carcinomas by collagenase digestion and sephadex G-10 column passage [3] . The lymphocyte preparative techniques were not responsible for the low levels of responses detected. Also, in situ prostaglandin synthesis and release did not appear to be involved in depressing lymphocyte reactivity [4] .

Comparable percentages of suppressor cells (OKT8+) were detected within these different lymphocyte preparations. Suppressor cells were not found in the lymphocyte preparations from the blood and lymph nodes of appropriate controls. From clinical and experimental studies it is known that blood transfusions may have immunosupressive as well as immunostimulating consequences. The effect of transfusions on graft survival has been extensively studied by our group in the BN to WAG rat model. In this donor-host combination it was found that a donor specific pre-transplant blood transfusion could lead to a marked prolongation of heart and kidney graft survival, whereas the similar pretreatment resulted in accelerated rejection of BN skin allografts. This specific model was used to investigate the influence of a single BN transfusion on the growth of 2 different syngeneic transplantable tumors in WAG rats. The first tumor was a radiation induced basal cell carcinoma of the Skin (T 1), the second tumor was a chemically induced adenocarcinoma of the duodenum (T 2). The antigenicity of both tumors was assessed in vivo, using classical challenge-protection experiments. It was observed that T I exhibited strong immunogenetic properties, whereas T 2 was only weakly immunogenic. The doubling time ofT I was 2.5 days, the doubling-time of the adenocarcinoma was 14 days. Intravenous inoculation of isolated T 1 cells led to development of lung nodules which could be counted after 14 days. WAG rats were injected i.e. with 1 ml of BN blood or syngeneic blood (controls) at 7-14 days before tumor challenge. T l was given in two different ways: a) sc. implantation of+2x2 mm pieces, b) i.e. injection of 106 isolated tumor cells. T 2 was implanted sc. only. Each experimental group consisted of 8 animals. For T 1 it was found that allogeneic blood transfusions caused a slight (but not significant) inhibition of subcutaneous tumor growth. However, in the T 1 lung-metastasis model it was observed that a single BN blood transfusion led to a 50 % reduction of nodules, counted at 3 weeks after inoculation. This reduction in number and size of nodules was highly significant (p<0.01). For tumor T 2, the BN blood transfusions evoked a strong inhibitory effect on the growth of the sc. implanted tumor. At 8 weeks after implantation all tumors in the control group had grown to a diameter of 10-16 rnm (average diameter 12.2 ram). In the group pretreated with BN blood, only 3 of 8 tumors were palpable at that time (average diamter 4 mm). For T 1 it was further investigated whether a single BN transfusion, given one week after i.e. tumor cell inoculation, would have any influence on tumor growth. No significant effect on number or size of the lung nodules could be noticed, if anything, the transfusion appeared to have a stimulatory effect.

The results indicate that allogeneic transfusions can lead to a substantial modification of tumor growth, depending on tumor type and site of implantation. This observation may have important clinical implications. We report here the serial study of circulating immune complexes (CIC) in two human tumor systems, colorectal cancer and gestational trophoblastic neoplasia (GTN). CIC were assayed by antigen nonspecific insolubilization induced by 3.75 % polyethylene glycol (PEG) and monitored as A OD45o changes by spectrophotometry. All 11 of the serially studied colorectal cancer patients presented with elevated CIC levels (mean = 610 + 396 zt OD450) as compared to our standard CIC level for pooled normal human sera (202--+_4 AOD450, p<0.05). Initial values in these patients range from 304 to 1407 A OD450 with no correlation to tumor load, site of presentation, or subsequent clinical course. In 6/7 patients who underwent ,,curative" resection of primary or metastatic colorectal cancers, serial CIC elevations occured only when antigen excess (measured by simultaneous carci-noembryonic antigen [CEA] assay) decreased. Immunoglobulin components of fractionated CIC showed predominantly IgA subclass. In 30 GTN patients followed with serial CIC and simultaneous human chorionic gonadotropin (hCG) assay, only those patients documented to enter hCG remission after molar evacuation showed significant elevation of CIC. Chromatographic fractionation of peak CIC in one such patient defined three irnmunoglobulin containing fractions showing immunoreactivity to one of four paternal HLA haplotypes (AW32). One ,,antigen" fraction (<25.000 MW) from this complex completely inhibited reference anti-AW32 binding. As in the colorectal cancer patients, these data show that CIC rise only when antigen excess decreases (reflected in the GTN patients by hCG normalization). In addition, some GTN patients may react to immunogenic paternal HLA haplotypes as part of their response to molar pregnancy. DFMO, an enzyme-activated irreversible inhibitor of ornithine decarboxylase (ODC), reduces tumor polyamine levels, inhibits growth of EMT6 sarcomas and hepatomas in experimental animals, and induces remission in human leukemia. A renal adenocarcinoma (RA) cell suspension, or a 1 mm segment ofWilms' (WM) tumor was transplanted intrarenally into Balb/C mice (n--105) or subcutaneously into Wistar-Furth rats (n=80) respectively. DFMO (2 %) in drinking water was administered to half the animals in each group throughout the experiment. At 28 days RA tumors in DFMO-fed mice weighed 72 % less than tumors in control animals (p<0.001); WM tumor weight at 28 days was not affected by DFMO feeding. The mean number of lung metastases in DFMO-fed R.A-bearing mice was 0.1 and in RA-bearing control mice was 1.4 (p<0.001). DFMO caused 72-75 % inactivation of tumor ODC, reduced RA putrescine levels by 710/o (p<0.001), reduced WM putrescine levels by 65% (d7<0.01), reduced WM spermidine levels by 58% (p<0.001) and increased WM spermine levels by 37% (p<0.01). DFMO feeding did not alter DNA content of RA or WM tumors. Although final carcass weight was similar in all animals, DFMO feeding progressively reduced total body weight (TBW) of mice, but not rats, until at day 20 the TBW of DFMO-fed mice was 10.5% less than tumor-bearing control mice (p<0.01). DFMO-fed mice bearing RA tumors survived 3.0-t-0.8 days longer than control mice (p<0.05).

Reduction of polyamine levels in wilms' tumors does not affect tumor growth. Lowering of renal ade-nocarcinoma putrescine levels by continuous feeding of DFMO to tumor-bearing animals decreases tumor growth, reduces lung metastases, and increases host survival. We have previously demonstrated that vagal nerve stimulation releases 5-HT into the lumen of the feline gut. This study was initiated to: 1. determine if substance P (SP) and motilin (MT), other enterochromaffin cell products, are released simultaneously, 2. to evaluate if this release is under cholinergic or adrenergic control. In 6 cats, 15cm isolated in situ segments of proximal jejunum were perfused with saline at 1 ml/min (37 ~C). Perfusate samples were evaluated at 5 rain intervals and concentrations of 5-HT, SP, motilin were measured by RIA's developed in our laboratory. After two 5-min basal periods, the supradiaphragmatic sectioned vagus nerves were stimulated electrically (10V, 5m s, 10Hz). The output from the loop in ng/5 min (5-HT) and pg/5 rain (SP and MT) was: introduced into the jejunum of conscious dogs through an external small bowel fistula. The gut was perfused at 5 ml min-1 with a physiological electrolyte solution containing the non-absorbable marker polyethylene glycol (mol. wt. 4000; 5 g 1-1); water and electrolyte absorption and transit time (TT) were measured during intravenous (I.V.) administration of each peptide, and during preceding and succeeding I.V. control infusions of 0.15M NaCI. Separate studies showed jejunal absorption and TT to be constant over prolonged periods during I.V. NaC1 administration. Bombesin (10 pmol kg-lmin-1) and neurotensin (1 pmol kg-~min 1) significantly reduced jejunal water absorption; bombesin and enkephalin (0.5 nmol kag-1rain-1) significantly prolonged TT; and enkephalin encreased water absorption (/7<0.05 in all cases). Measurement of plasma neurotensin during I.V. infusion indicated that physiological blood levels were not exceeded during these studies. Conclusion: A number of peptides may be involved in the regulation of small bowel function. The effect of neurotensin on jejunal water transport provides a possible mechanism linking raised blood neurotensin levels with intestinal intraluminal fluid accumulation in the dumping syndrome. In order to establish Whether this release was under adrenergic control, 4 cats had cervical ganglionectomies. Using the same electrical paramenters, stimulation of the cut cervical vagus nerves resulted in identical 5-HT responses as above. In 4 additional cats atropine administration (lmg/kg IV) totally abolished the 5-HT responses to vagal nerve stimulation.

The paralled release of 5-HT, SP, and MT following vagal nerve stimulation, strongly suggest that the EC cell is the source of these luminal hormones. This release appears to be under cholinergic control. Since diabetes mellitus is markedly improved immediately after jejunoileal bypass before significant weight loss, but only gradually and often incompletely changed after gastric bypass, it seemed appropriate to investigate the effects of intestinal exclusion on experimental diabetes. Studies were performed on alloxan diabetic Sprague-Dawley rats. Two days after jejunal exclusion (JE) in 8 rats (resection of proximal 1/3 of small intestine), fasting blood sugars (FBS) decreased 580, from 344 to 146mg/dl, and averaged 164 mg/dl at 3 weeks. After ileal exclusion (IE) in 7 rats (resection of distal 1/3 of small intestine), FBS fell 39% in 2 days, from 250 to 153 mg/dl, but increased 40% above preoperative levels to 350 mg/ dl at 3 weeks. Sham operated rats responded similarly to IE rats. After alloxan and operations, all groups lost weight, but only JE rats began to increase weight at a normal rate. Increased water intake, polyuria (140 ml/24 h), and glycosuria (2756 rag/all), were present in IE rats; JE rats were normal (urinary output <20 ml/24 h, and urinary glucose 36 mg/dl). Oral glucose tolerance tests (GTT) were extremely abnormal in IE rats, similar to those in non-operated alloxan rats, while GTT curves in JE rats were similar to normal animals, with some elevation at 30, 60 and 120 min. Serum insulin levels remained low in all alloxan treated rats after jejunal exclusion. Possible mechanisms, currently under study, relate to carbohydrate malabsorption and changes in enteric chemical mediators. The purpose of the present study was to investigate the changes in somatostatin release and somatostatin-containing cells of the pancreas and stomach of the streptozotocin (STZ) -induced diabetic rat after the amelioration of diabetes by whole pancreatic transplantation. Highly inbred Lewis rats were divided into three groups: (1) normal rats, (2) STZinduced diabetic rats and (3) transplanted rats. Diabetes was induced by the administration of STZ (60 mg/kg). On the seventh day after STZ treatment, pancreatic transplantation was performed. Four weeks after the transplantation, in vivo and in vitro, studies were performed. Pancreatic D cells and gastric somatostatin-containing cells were stained with antibody enzyme method. Studies in vivo showed marked improvement of the impaired arginine-induced insulin release by the transplantation. Studies in vitro employing isolated perfused rat pancreas and stomach revealed following results: Mean basal pancreatic somatostatin release in normal, diabetic and transplanted rats were 12___3, 24-t-7, and 17__+4 pg/ml, respectively. Total amount of pancreatic somatostatin release in each group during arginine stimulation (19.2. mM) were 946--+200, 2321___266, and 1013-+126 pg/15 min, respectively. Significantly higher somatostatin release was obtained from the diabetic pancreas, which was, however, reduced to normal after the whole pancreatic transplantation. On the other hand, insulin release from the diabetic pancreas was severly impaired and pancreatic transplantation had not effect on insulin release from the host pancreas in the transplanted rats. As to the glucagon release, there was not significant difference among them. Mean basal gastric somatostatin release in normal, diabetic and transplanted rats were 179-t-5, 173___5, and 135 + 5 pg/ml, respectively. There was no significant difference between normal and diabetic rats, though the significant decreased value was obtained in the transplanted rats. (vs. normal;/7<0.01, and diabetes; /7<0.01). On the other hand, glucagon-stimulated peak values in these groups were 498_ 36, 662___47, and 412+25pg/ml, respectively. Glucagon-stimulated gastric somatostatin release in diabetic rats was significantly increased, but reduced to normal value by pancreatic transplantation. Also, a number of pancreatic D cells and gastric somatostatin-containing cells were markedly increased on the diabetic rats. On the other hand, a number of these cells in the transplanted rats were descreased to normal levels.

In summary, enhanced pancreatic and gastric somatostatin release and cells in the diabetic rats were both normalized after the amelioration of diabetes by the whole pancreatic transplantation. From these results, it is suggested that pancreatic and gastric somatostation are regulated by circulation and/or metablic of nutrients. Doppler velocity recordings are widely used for the non-invasive diagnosis of carotid arterial disease. Although detailed analysis of carotid Doppler spectral information has been suggested as a method for improving diagnostic sensitivity, the accuracy of the relationship between the Doppler recording and the true instantaneous velocity profile has not been established. The Purpose of this study is to determine if a CW Doppler velocitymeter can accurately transduce the true instantaneous blood flow velocity information. Methods: A pulsatile flow model has been constructed in which it is possible to record the instantaneous Doppler spectrum and simultaneously photograph and measure the true velocity profile. A computer controlled pump generates a carotid waveform in tubes without stenoses and with, symmetrical stenoses. Flow is visualized using the photochromic dye tracer technique. A short burst of UV light from a laser is passed across the tube. A narrow band of the fluid turns blue, its movement is photographed, and the instantaneous velocity profile determined every 10 msec throughout the pulse cycle. At the same time, the instantaneous Doppler spectral information is recorded by a frequency analyzer. The Results follow. For pulsatile laminar flow, the Doppler spectrum correctly recorded the true velocity spectrum, including the instant/~neous maximum velocity and mean velocity. For disturbed flow, it was not possible to show the same direct relationship between the Doppler spectral recordings and the blood flow velocity. In Conclusion Doppler velocitymeters accurately transduce velocity information when flow is laminar but when flow is disturbed there is not a direct relationship between Doppler recordings and the true velocity profile. Consequently, one should be cautious in attempting to relate Doppler measurements of disturbed flow directly to the true changes in the velocity pattern. Early failure of arterial reconstruction may originate in poor patient selection. In aorto-iliac stenosis (AIS) selection for operation relies upon clinical examination of the femoral pulse and radiology. Since single plane arteriography is inadequate for accurate definition ofiliac stenosis [1, 2] , this paper compares clinical examination, Doppler ankle systolic pressure (ASPI) and femoral signal analysis (Laplace Transform Damping, LTD, and Pulsatility Index, PI) [3] with biplanar contrast studies. i At biplanar angiography 66 of 102 ischaemic lower limbs had AIS with diameter reduction from 25-84 %, Of the remainder, 25 were normal (<25 % stenosis) and 11 were ,,occluded" ('_--850/0). Nearly two thirds (61%) of limbs with a clinically normal femoral pulse had identifiable arteriographic stenosis (~-__.250/0), Upstream abnormality was predicted incorrectly in 15 % and the overall accuracy of clinical examination was 67 %, both for detecting stenosis and predicting its severtiy. ASPI (median 0.61; 95% Confidence ,Limits 0.3-1.0) and PI (5.4; 2.3-0.05) although correlated with stenosis (ASPIr = 0.65; p = 0.05, PIr = 0.62; p=0.05 Variance analysis for linear regression), did not aid the clinician further (accuracy 62 %). However LTD (0.72; 0.37-1.0) was well correlated (r=0.76, p=0.001) and did improve assessment ofiliac stenosis (accuracy 84 %).

The need for biplanar arteriography is reiterated and its use with Doppler signal analysis should improve the evaluation of aorto-iliac disease. In the absence ofa non-invasive method for estimating volume flow in an individual artery, local blood pressure measurement has proved, with certain limitations, useful in assessing the cardiovascular system. Now ultrasound technology has progressed to enable blood flow in an artery to be measured noninvasively.

We report the results o four evaluation ofa 5 mHz, computerized, 30 channel, pulsed Doppler vessel imaging and flow measuring instrument in in-vivo experiments. Computed blood flow was compared to actual blood flow (calculated by timed collection) in 17 anaesthetised dogs.

Correlation between computed and actual blood flow was stronger in the larger abdominal aorta than in the smaller common carotid arteries. From the regression plot, the coefficients of determination, P, were: 0.927 (exposed aorta scans); 0.857 (transcutaneous carotid scans); and 0.784 (exposed carotid scans). Stepwise regression analysis showed the computed flow values to be independent of probevessel angle, depth and lumen diameter for vessels greater than 2.5 mm in diameter. These results suggest that this pulsed Doppler instrument has the versatility and accuracy essential for diagnostic flow measurements in the main conducting arteries of the neck and limbs and in vascular bypass grafts. In the assessment of patients undergoing carotid artery surgery, many laboratory methods are available in addition to angiography. In a series of 210 patients experience has been gained with the use of EEG, TC 99m isotope scanning, OPG, CT scanning and Doppler. In a 10 year follow up over 85% of patients had a satisfactory outcome. An early mortality of 2% in the beginning of the series has been eliminated due to improved selection. In this report the application of multi-gated pulsed Doppler techniques is reported. This allows a non invasive measurement of mean volume flow in the common carotid artery with a method reproducibility of + 5 %. Mean volume flow in 50 undiseased arteries (25 subjects mean age 46 years) was found to be 536± 104 (S.D.) ml/min. From this a lower range for normal flow of 300 ml per minute (2 x S.D.) was selected. 30 patients were investigated before surgery and a follow up examination was performed at a mean interval of 4V2 months post operatively. 3 Groups were defined. Group A >300 ml/min; Group B 200-300 ml/min; Group C <200 ml/min, Of 14 arteries in Group A before surgery, 11 remained in the Group and 3 dropped to Group B. In Group B, of 8 arteries, 5 go to Group A, 2 remain and 1 dropped to Group C. In Group C, 7 out of 8 arteries moved to Group A and 1 to Group B. Thus of 16 Arteries with below normal volume flow before surgery, 12 were returned to normal range and further 1 improved. 2 remain unchanged and 1 disimproved. Of the 30 arteries examined 4;/2 months after surgery, 23 are in the normal range and a further 1 improved. 2 remain unchanged and 1 disimproved. Of the 30 arteries examined 41/2 months after surgery, 23 are in the normal range in flow values and a further 2 remain unchanged.

The non-invasive and isotope techniques have a valuable and practical application in assessment of carotid artery surgery. Timing is influenced by the finding fo infarction on CT or isotope scanning. Doppler techniques are useful not only in defining severity of diseas and sub-radiological plaques, but valuable flow information can be obtained by pulsed Doppler pre and post operatively. This may help in identifying patients who need further medical or surgical treatment.

Stepwise Logistic Regression-AModel for Predicing Success of FemoraI-Popliteal Bypass Grafts The objectives of this study were to identify the preoperative factors that influenced postoperative patency of femoral-popliteal grafts and to develop a model that could be used prospectively to determine the probability of successful outcome. Data base material consisting of history, physical examination, laboratory data, angiography, and operative findings in 199 patients undergoing femoral-popliteal bypass grafting was entered into a computer programmed for stepwise logistic regression analysis. The computer identified and ranked 24 factors that influenced outcome. The top five factors (other than technical problems) included quantity of runoff, previous ipsilateral femoral-popliteal bypass, preoperative prediction of potential amputation level, concurrent proximal vascular reconstruction, and the location for distal graft anastomosis. Having established the computer data base, it is now possible to enter information from new patients into the computer which will weigh all factors and indicate the likelihood of surgical success. In addition, tables can be generated which will look at simple combinations of variables to predict patency. For example, in a patient about to undergo a primary femoral-popliteal bypass with no anticipated technical problems, the likelihood of success as a function of runoff and preoperative amputation level is as follows: Irreversibility of shock and ischemic injury is generally considered a consequence of extensive cellular injury. To study the role of intravascular coagulation in shock, 56 rats were bled to a mean arterial pressure of 50 mm Hg for 3 hrs or 25% uptake of shed blood, whichever occured first. Return of shed blood with These data provide the patient and the surgeon with a quantitative prediction for success and permit an informed decision when considering therapeutic alternatives.

Potential cytoprotection by heparin was studied by similarly bleeding 52 additional rats; controls were only cannulated. Twenty pairs were heparinized (H); 32 were not (NH). Paired-bled and control rats were sacrificed following hemorrhage, liver (L) and kidney (K) mitochondria were isolated, and the isolates were studied by the polarographic technic with glutamate and succinate to determine the Respiratory Control Index (RCI) as a measure of cellular injury. Results were: an equal volume of isotonic saline resulted in a 43 % survival; % uptake of blood during shock allowed prediction of survival. An additional 55 rats were then randomized (coin toss) to heparinization versus no heparin prior to shock, and were similarly bled and resuscitated. Significantly improved survival (p<0.025) was seen in heparinized (17/23; 740/0) versus nonheparinized rats (13/32; 41%).

Uncoupling and inhibition of mitochondrail function were noted in both H and NH rats with RCI being significantly reduced from control. However, there was no difference between H and NH compared to each other. Heparin does not provide cytoprotection during shock; improved survival with heparin may rather be a consequence of improved reperfusion of tissues following the shock episode. fl-endorphin (B-END) has been postulated to play a role in the pathogenesis of shock because the opiate "antagonist naloxone improves the macrohemodynamics in various shock models [1] . However, plasma levels ofopioid peptides have not been determined as yet. The aime of our study was to measure the plasma concentrations of various peptides and to evaluate the influence of naloxone particularly on the plasma concentration of r-END.

In 16 anesthetized foxhounds, the adrenolumbar vein was cannulated and hemorrhagic shock (MAP = 4 mm Hg for 3 h) was induced according to Wiggerstechnique. The plasma levels offl-END, methioninenkephalin (M-ENK), and leucine-enkephaline (L-ENK) were simultaneously determined in c.v. and/ or adrenal venous blood by a specific RIA. Crossreactivity of r-END withfl-lipotropin was about 4 %. The ENK-antibodies crossreacted with less than 5 %. Five dogs received an i.v. bolus of naloxone (2 mg/ kg) and a subsequent naloxone infusion of 2 mg/kg/ h after 2 h of hypovolemia. Eleven dogs served as control and received equivalent volumes (1 mg/kg per h) of Ringer solution.

Hemorrhage resulted in a sharp rise of central venous plasma levels particularly of M-ENK and L-ENIC This effect was even more pronounced in the adrenal's effluent system.fl-endorphin levels remain elevated whereas the ENK secretion began to decrease 1 h after hemorrhage. Naloxone treatment inhibited any spontaneous fall of adrenal enkephalin release during the shock phase and the values remained elevated 10-15 fold. Volume substitution with autologous blood resulted in a normalization of all peptide levels.

These data demonstrate that hemorrhagic shock will cause stimulation of endogenous opioid peptides. The high levels of enkephalins in the adrenolumbar vein indicate that the adrenal gland is the main source of these peptides in the circulation. In addition toil-END, the ENK seem to play a role in the pathogenesis of shock as well. At our present state of knowledge, however, it is difficult to design a coherent concept of mechanisms involved. This shows that CP treated cells bound nearly as much [125t]-ACTH analog as control cells but there was very little specific binding to SP treated cells. Low concentrations of ACTH effectively displaced the ACTH analog whereas exposure of adrenocortical cells to SP resulted in a significant decrease in ACTH receptors. This suggests that SP has a factor(s) that binds to ACTH receptors of adrenocortical ceils which may adversely affect the stress response of shock. F-43 has been proposed as superior to other asanguinous fluids due to increased oxygen carrying capacity. Evaluation to date has been largely uncontrolled and at extremes of hemodilution (Hct. <2 %) rarely seen clinically. Near infrared spectrophotometric monitoring of brain cytochrome a, a3 redox state, a sensitive indicator ofintramitochondrial oxygen availability, offers a unique opportunity to contrast F-43 with balanced salt-albumin (BSA) and whole blood saline (WBS) as a resuscitative regimen in a clinically relevant model. Fifteen rats were subjected to 30 minutes of hypoxia (FIO2 = 7,5%) and hemorrhagic hypotension (MAP = 30 mmHC), then randomly allocated to one of three groups and resuscitated by FIO2 = 100% and infusion of either F-43, BSA, or WBS. Cytochrome a, a3 redox state was monitored continuously at 813 run. Thirty additional rats were sacrificed at baseline, end shock and 20 and 120 minutes post resuscitation for cerebral cortical ATP and lactate assay. Despite hematocrits as low as 15 % in the BSA and F-43 groups, there were no significant differences /7<0.05 between groups in the parameters of oxygen sufficiency; ATP, lactate, and cytochrome a, a3 redox state. We assume differences in cardiac output compensated for differences in arterial oxygen content. On this basis we suggest perfluorochemical utilization should be limited to situations in which hematocrits are <15 % and when cardiac reserve is limited. Metabolites of the prostaglandin endoperoxide H2 (PGH2) affect both vascular tone and platelet aggregation and thereby may influence blood flow. We, therefore, determined the metabolites formed from PGH2 by microsomes isolated from human saphenous vein used for aortocoronary bypass surgery. In the absence of reduced glutathione (GSH), the enzymatic metabolism of 14C-PGH2 produced only prostacyclin (PGI2) as measured by the formation of its stable breakdown product 6-keto-PGFla. The amount of PGI2 formed varied from 10-50% of the substrate depending upon the microsomal protein and PGH2 concentrations. In addition, the nonenzymatic breakdown of PGH2 resulted in the formation of PGF2a, PGE2, PGD2 and heptadecatrienoic acid (HHT). There was no formation of thromboxane A2 (TXA2) as measured by the absence of its stable breakdown product TXB2. In the presence ofGSH, a required cofactor for microsomal PGE2 isomerase activity, the formation of PGE2 was augmented 2 fold (up to 35 % of the substrate) indicating enzymatic for-mation of PGE2. The GSH either did not alter or augmented (less than 1 fold) the formation of PGI2. The increased formation of PGE2 in the presence of GSH was at the expense of decreased nonenzymatic breakdown of PGH2 to PGF2a, PGD2 and HHT. These data suggest that prostacyclin synthetase activity may serve to protect the vessel graft from platelet aggregation and/or vessel spasms and may possibly serve as an indicator of graft viability. Thrombosis is a frequent cause of early arterial bypass graft failure and platelets are known to be major determinants ofthrombus formation on arterial surfaces. PGI2 and flbriolytic activators from the vascular wall counteract intravascular thrombosis. The aim of this work was to study the effect of arterial grafting on the aforementioned mechanisms.

6 cm lengths of tanned human umbilical vein grafts (HUVG) with an internal diameter of 5 mm were interposed end-to-end in the carotid arteries (c.a.) and jugular veins (j.v.) of sheep. Placed in the c.a.'s, 12 grafts with restricted flow (50 cc/min) were removed on the 10th postoperative day (group I); 18 grafts with unrestricted flow (140___20 cc/min after placement) were taken out 90 days later (group II) and 4 grafts placed in the j.v.'s were removed 10 days after surgery (group III). Upon removal, the grafts were checked for patency and sections from the proximal and distal anastomoses and midgraft were obtained for determination of PGI2 production (RIA) fibrinolysis activators activity (histochemical method) and for light and scanning electron microscopy. Sections from the femoral arteries were also obtained. The results of PGI2 generation are expressed in ng/ml/cm% All grafts showed fibrinolytic activity in the adventitia but 4 grafts in group II also showed fibrinolytic activity in the intima. Early neointimal fibrous hyperplasia (NFH) characterized by proliferation of smooth muscle cells was present in group II. The, occluded grafts showed organizing thrombus material and inflammatory cells and the patent ones showed fibrin and scattered inflammatory cells. In groups I and III, SEM revealed numerous platelets and RBC's incorporated into a proteinaceous material overlying the anastomoses and in some specimens obvious thrombus material was present. In group II, the anastomotic areas were covered with large endothelial cells, nonetheless, some areas were denuded and small thrombi were occasionally noticed. In conclusion: 1. Anastomotic sites create a strong stimulus for thrombus formation despite a high production of PGI2. This suggests that antithrombotic therapy may be necessary to prevent early failures. 2. HUVG develop the capacity to produce PGI2 and fibrinolytic activators and 3. although the etiology of NFH remains obscure, the decreased levels of PGI2 in group II suggest that exhaustion of PGI2 generation from the endothelium might occur leading to proliferation of smooth muscle cells (NFH). These cells will in turn supply PGI2 if a persistent stimulus exists.

Permeability of Intestinal Capillaries to Fibrinolytic Products D. Manwaring and P. William Curreri Department of Surgery, University of South Alabama College of Medicine, USA Fibrin/fibrinogen degradation product D (FDP-D) is significantly elevated in the serum of patients after trauma or sepsis. Purified FDP-D infused into nontraumatized rabbits precipitates thrombocytopenia, complement depletion, pulmonary dysfunction and increased permeability of lung capillaries to I12S-albu -Composite fibrin plate assay Size oflytic zones (mmx) after 17 h incubation at 37 °C rain. In order to determin of products of fibrinolysis alter fluid filtration or permeability, either purified FDP-D or FDP-E were tested in an isolated, autoperfused cat ileum preparation. Steady-state lymphatic: plasma protein concentration ratio (CL/Cp) and lymph flows (QL) were measured at a venous outflow pressure of 5 mmHg. Data was analyzed for each animal group by the paired student t test for QL, CL/Cp and protein clearance (QL x CL/Cp). In cats which received FDP-D (n=7), QL and clearance increased five-fold (P<0.002), but CL/Cpwas not altered, which suggests a permeability change. Ileal mucosal biopsies prepared for histology had villi that were de-epithelialized and platelet clots in blood vessels. FDP-E (n=7) provoked a slight increase in QL (P<0.01), but not in CL/Cp or clearance. Histology was normal. (FDP-E causes no pathological lung change in awake rabbits). FDP-D may contribute to various organ pathologies after trauma.

The Effect of Aspirin on the Fibrinolytic Activity of Viable Granulocytes R.C. Franz, W.J.C. Goetzee, B. Rotunno and R. Anderson Department of Surgery, University of Pretoria, RSA Although several influential authors have suggested that low dose (150 rag) Acetyl Salicylic Acid (ASA) represents a balanced daily antithrombotic regimen probably by both inactivating thromboxane A2 production and enhancing prostacyclin synthesis little is known about the effect of aspirin on the fibrinolytic activity of live granulocytes [1] . The present study was designed to evaluate this effect in vivo. Methods: Granulocytes from fasting samples of heparinized venous blood taken from male volunteers were separated from monomuclear cells and platelets by density gradient centrifugation (Ficoll: Sodium metrizoate). Viable granulocyte suspensions and plasma samples were placed as drops on a coin-

Before aspirin After aspirin P = [2] . The experiment was repeated 3 h after each subject had ingested 1.8g of aspirin. The results are summarized in Table 1 .

1. There appears to be a significant increase in granulocyte fibrinolytic activity 3 h after the ingestion of 1.Sg of aspirin. 2. This increment is insufficient to overcome the resting inhibitor potential of plasma on granulocyte fibrinolysis. 3. Aspirin does not evoke a significant increase in plasminogen activator-(urokinase) induced flbrinolysis in platelet free plasma or in the combined system of granulocyte-plasma mictures. 4. The optimal dosage of aspirin as a fibrinolytic agent requires further study. The terminal vascular bed of malignant tumors is characterized by a lack of organization, differentiation and sufficient developement of nutritional capillaries. As a result, malignant tumors reveal consistently small regions of low or even no perfusion. Pre-vious data in a melanoma indicate that due to the rarefication of capillaries, the full impact of tumor treatment ±st diminished by an elevated microvascu lar resistance, which could significantly affect the impact of tumor therapy. Since the improvement of the blood's fluidity has been shown as one therapeutic modality to increase significantly the capillary blood flow, it was assumed that this measure might enhance the accessibility of tumor tissue for bloodborne drugs. This study was aimed to investigate the effects of the improvement,of microcirculatory flow on tumor growth and tissue oxygenation. Moreover, the response of the melanoma to chemotherapy was evaluated when isovolemic hemodilution was employed in conjunction with chemotherapy.

A transparent chamber technique, intravital microscopy, a platinum multiwire electrode (local PO2 measurement) and quantitative television image analysis (capillary blood cell velocity and diameter) were employed to study the microvasculature in the amelanotic melanoma A-MEL-3 of 18 hamsters in the event of hemodilution without and in conjunction with chemotherapy (200 mg/m 2 DTIC, dimethyl-triazeno-imidazol-carboxamid). Permanent indwelling catheters in carotid artery and jugular vein served for measuring systemicpressures, heart rate, for withdrawing blood and the infusion of DTIC and/or Dextran 60. After inoculation of 4 x 104 cells of the amelanotic hamster melanoma A-Mel-3 into s.c. tissue in the preparation, this tumor re~ched a diameter of approx 3 mm within five days.

The reduction of systemic hematocrit from 0.45 to 0.31 (1.3 ±0.2 ml blood vs Dextran 60, 7 animals) at a tumor diameter of 3 mm increased the growth rate of the melanoma by about 30 % while enhancing significantly the volume flow through capillaries and the mean local PO2. Table 1 Control Hemodilution capillary velocity (ram/s) 0.39 _ 0.12 0.50 ± 0.12 capillary blood flow (ml/min x 10 -5) 5.4 ± 1.7 8.9 ± 1.2 mean local PO2 (mmHg) 9.4 (0 -28) 15.7 (0 -60)

The frequency distribution of local PO2 on the tumor's surface showed a distinct shift toward higher PO2 values with still some hypoxic regions remaining. Intravital microscopy, however, revealed petechial bleeding and localized, interstitial edema which compressed a small number of capillaries. By contrast, the tumor's diameter remained at app. 3 mm for a period of ten days with chemotherapy alone (3 animals). In one of the animals, a complete stop in the melanoma microcirculation was seen within four hours after infusion of DTIC followed by a significant decrease of tumor diameter. When chemotherapy was initiated in hemodiluted animals, neither retardation of tumor developement nor vascular obstruction was observed (5 animals). Conclusion: Capillary blood flow of the melanoma can be enhanced by hemodilution thus diminishing tissue hypoxia. This measure, however, was associat-ed with an increase in melanoma diameter of 30%. At the present, we investigate whether, in hemodiluted animals, a reduction of tumor size can be obtained with a higher dose of DTIC. ). However, the etiology of stress hyperglucagonemia in the immobilized rat is only poorly defined. Since during restraint stress, catecholamines (CA) are elevated and stimulation ofglucagon by CA is accepted (Woods SC D Jr [1974] Physiol Rev. 54: 596), we decided to study by surgical means the the relative contribution to glucagonemia of different sources of CA, i.e. peripheral sympathetic nervous system and adrenal medulla. Methods: Male Sprague-Dawley gastric fistula rats (n=74), weight approx. 250 g, were subjected to either sham-op or various sympathectomies [microsurgical splanchnicotomy = S-SX; chemical sympathectomy = C-SX (150 mg/kg 6-OH-dopamine ip two days prior to the experiment); adrenal demedullation = AMX; combinations: S-SX + AMX; C-SX + AMX]. Gastric acid secretory trials (duration 8 h), preceded by a 24 h fasting period were carried out 6-7 days following the operation. At the start of the experiment an intraperitoneal polyethylene tube, was inserted into the abdominal cavity of the rats, allowing a constant infusion of physiological saline (4 ml/ kg/h). In addition, stress was performed by pairwise restraint of the extremities and small electric shock waves applied by a tail electrode. At the end of the experiment, blood was drawn from the vena portae and the abdominal aorta for plasma and serum. Hormones (glucagon, insulin) were measured by radioimmunoassay, glucose enzymatically, volume was read to the next 0.1 ml, acidity by microtitration. Results (see table) : Volume and acidity are not changed by the various sympathectomies, when 179 compared to the sham group. The same is true for acid secretion, except in S-SX + AMX, where it is elevated. Glucagon in peripheral plasma is elevated in C-SX, AMX and C-SX + AMX. In the portal vein, glucagon is dccreased in S-SX + AMX (408___99 pg/ ml) and elevated in C-SX + AMX (2625___405 pg/ml) when compared with sham rats (930--.195 pg/ml). The portal/aortal glucagon ratio is significantly decreased only in C-SX and AMX (1.7--+0.9, 1.7--.1.0, resp.) when compared with sham (3.8--_ 2.1). Insulin is increased only in AMX, glucose decreased in AMX, S-SX + AMX and C-SX + AMX, insulin and glucose are unchanged in the other groups.

Conclusion: 1. Stress hyperglucagonemia in the rat is confirmed (levels during zero stress 30-180 pg/ml) and also the rise in insulin following removal ofadrenomedullary CA (but not other sympathectomy); 2. The blood glucose fall (AMX; S-SX + AMX; C-SX + AMX) is not uniformly paralleled by hyperglucagonemia, but in the case of AMX it may be secondary to relative stress hypoglycemia owing to removal of adrenal medullary CA or reactive insulin release; 3. Mechanism underlying the increased systemic glucagon despite partial (C-SX; AMX) or total (C-SX + AMX) sympathectomy are yet unknown. 4. During stress the enterogastrone component of hyperglucagonemia may be of minor importance. EVLW showed good agreement with gravimetric lung water determinations. Significant lung water accumulation was produced by pressure elevations over 20 mmHg. Reductions in plasma oncotic pressure significantly increased transvascular fluxes at each level of pressure elevation. However, fluid accumulation was not significantly greater during hypoproteinemia. We conclude that a 40-50% reduction in plasma oncotic pressure does not contribute to increased high pressure edema because the lymphatic safety factor is augmented. This phenomenon may explain the well tolerated state of hypoproteinemia in patients after hemorrhagic shock. Computerized gamma scintigraphy is a new technique for the analysis of albumin flux in the acute respiratory distress syndrome (ARDS). The objectives of this study were to obtain normal control values and to determine the method's validity in patients with cardiogenic vs. permeability pulmonary edema. Methods: Following 10 mCi99mTechnetium-human serum albumin, lung :heart radioactivity ratios were determined. This ratio remains constant unless there is a leak o falbumin, when a rising ratio is seen, called the ,,slope index" (SI). SI's were determined in 5 control individuals who had :> 50 % left ventricular ej ection fraction and < 7s pulmonary circulation. Thirtythree studies were obtained in 27 patients using a portable gamma camera. Fourteen patients had clinical evidence of ARDS. Results: Studies were considered positive if the SI was 2 S.D. > control mean ( -0.4___0.5 x 10-3 units/ min). Among 15 positive studies, all had diffuse air space disease on chest radiographs. Their average pulmonary capillary wedge pressure (PCWP) was 14.6_+4.6 mmHg. The average artertial: alveolar oxygen tension ratio (a/A)O2 was 0.31+--0.15 on 10.8 cm H20 PEEP, which were both significantly (p<0.01) different from patients with normal SI's. Positive SI's were present from 24 hours to 6 days following the apparent onset of ARDS in 6 patients. Recovery of gas exchange was associated with normal SI's on repeat studies in 4 patients. Of 6 patients with cardiogenic pulmonary edema, 5 had negative studies (18-29 mmHg PCWP) and I a positive study (40 mmHg PCWP). Conclusion: Gamma scintigraphy was a sensitive, non-invasive tool for the detection of a pathological increase in pulmonary protein flux, which was usually normal in cardiogenic pulmonary edema. Positive scintigraphy was associated with significantly impaired gas exchange. The method documented that the leak of albumin in ARDS may last for days but resolves with recovery.

Cancellous Bone Thermocoagulation Ph. Dumontier, R. Benichoux and A. Vidrequin Institut de Recheres Chirurgicales -C.H.U. de Brabois 54511 Vandoeuvre Les Nancy Cedex, France Electrocoagulation can not stop bleeding from the cancellous part of a sectioned bone. Therefore we tested the efficiency ofthermocoagulation by hot air.

The hot air generator delivers a flow, of non illtercd air, at 25-30 1/min at a fixed temperature of 300°C measured at the exit of a 3 mm diameter pipe and 60 °C at the site of bleeding. The generator sustains usual steam sterilization. 14 dogs were operated on both patellae, femoral short segments and iliac crests giving 53 different site of cancellous bone.

In vitro sterility studies: the conduit of hot air was applied at various distances, from 10 cm to 2 meters, above a Petri Plate containing culture material. Thus the turbulence in the atmosphere around the zone of thermocoagulation has been bacteriologically controlled and a particle counter used. In vivo thermocoagulation: three sites of cancellous bone were used in 14 anesthetized dogs, using a sterile procedure: the iliac crest, a small segment of the femoral bone and the divided patella. 1. Five iliac crests were divided and bleeding measured after thermocoagulation. 2. The 2 segmental femoral resections were thermocoagulated. 3. The patellae were vertically divided and each section submitted separately either to thermo or electrocoagulation. The pipe of thermo-181 coagulation was directed to the bleeding surface at a 2 cm distane, sweeping it during 5 to 6 seconds. The bleeding was compared by photography and the two fragments were approximated by a wire synthesis to provide a bone fusion. In few cases both sides were thermocoagulated.

Results: In vitro: No contamination was found in the atmosphere, up to a distance of 2 meters. There was a significant decrease of particles around the operating site (60 % less, at 80 cm and 30 % less, at 2 meters). In vivo: the bleeding was weighted around 50% less than with conventional coagulation. Thermocoagulation did not delay or disturb the healing of the patella after wire synthesis. The in vitro nucleation time of cholesterol crystals from gallbladder bile of patients with gallstones is more rapid than that from normal persons, (Holan RT [1979] Gastroenterology 77: 611). This study determined whether this was due to a gallbladder or liver defect and wether the defect was the addition of a nucleating factor or the deletion of an antinucleating factor. Hepatic and gallbladder bile were gathered at surgery in stone patients and gallbladder bile in patients with a normal biliary tract. After ultracentrifugation, the isotropic phase was observed daily by polarizing microscopy until cholesterol crystals appeared. In gallstone patients, the nucleation time of gallbladder bile was significantly more rapid, 2.5 days+0.9 SEM, than that of hepatic bile 7.9+2.1 days, although hepatic bile was significantly more saturated with cholesterol [cholesterol saturation index (CSI), 2.12_+0.36], than gallbladder bile, (CSI, 1.31_+ 0.12). Thus the characteristic short nucleation time of stone formers is due to an alteration in bile after it enters the gallbladder. To determine whether the gallbladder defect was due to addition of a nucleating factor or the deletion of an antinucleating factor, isotropic phases of normal gallbladder bile and that from stone formers were mixed and nucleation time determined. Mixtures of up to 95 % normal bile had pathological nucleation times demonstrating that the defect is the addition of a nucleating factor by the gallbladder, and that this factor is potent. The rate of formation ofgaUstone precursor crystals in bile, although faster in gallstone patients than in controls, is unrelated to the degree of cholesterol supersaturation [1] , implying that other factors are involved. Two competing factors seem likely; (a) secondary seed crystals in bile may trigger and accelerate gallstone crystal formation from supersaturated solution; (b) "Poisons" in bile may retard or inhibit crystal growth.

Because of the complexity of bile itself, experiments were performed in highly purified mixtures of bile salt, lecithin and cholesterol, in concentrations closely resembling those of gallbladder bile.

(a) Lipid solutions were seeded with calcium carbonate, hydroxy-apatite, calcium bilirubinate and biliary mucus, all of which are found in gallstones [2] . Cholesterol crystal formation was significantly faster in_ the presence of all of the seed compounds tested (x= 221.7/~g ml-lh-1) than in unseeded controls (128.0/ag ml-~h-1) (/7<0.05).

(b) Substances with "crystal-poisoning" properties included heparin, chondroitin sulphate and bile salt. These and changes in pH altered the quantitiy (20-80 % decrease) and rate of calcium carbonate and calcium phosphate crystal formation.

We suggest that gallstone precursor crystal formation may be affected by a subtle balance between crystal seeding and crystal growth-inhibition, both due to the presence of other compounds in bile. At the last Tripartite Meeting we reported experimental data on a new method to destroy concrements of the kidney in situ by shockwaves allowing for spontaneous excretion via the urinary tract [1] . The shockwaves are generated externally by underwater discharge of a condensor with sparking electrodes which are localized in a focus o fan elliptic cavity. For treatment the renal concrement must be positioned exactly in the second, virtual focus opposite to the elliptic cavity. Since then, altogether 126 patients were subjected to this form of treatment in our institute by the colleagues of the department of urology at the University of Munich. 90% of the patients got rid of their concrements within a few days, in 8.5 % small remnants remained in the renal pelvis, and in 1.5% (2 cases) additional surgery became necessary. Meanwhile this technique is employed on a routine basis in the department of urology of the University of Munich [2] .

The experimental as well as clinical results were considered encouraging enough to extend the technique for the treatment of biliary concrements. For this purpose, human gallbladder concrements of different composition (bilirubin, cholesterol) were implanted into the gallbladder of dogs for exposure to shockwave treatment after wound healing. Under in vitro-conditions the biliary concrements could be crushed into any size desired, irrespective of their composition, while only in 8 out of 10 experiments this was accomplished under in vivo-conditions. Blockage of the biliaiy duct after shockwave exposure was never observed. Concrements which were experimentally implanted into the bile duct could be visualized without difficulties by contrast medium. Here, destruction by shockwaves was accomplished as well. Currently experiments are conducted to dissolve remnants of biliary concrements after treatment by administration of desoxycholic acid. Precise positioning of the gallbladder concrements in the second virtual focus is a problem which has not been satisfactorily solved so far, because the concrements cannot be visualized by conventional X-ray techniques. Alternatively it is attempted to employ ultrasound, or visualization by retrograde injection of Xray contrast medium through a catheter. In experimental animals, we leave a T-formed drain in the gallbladder for injection of contrast medium. In animal experiments conducted so far, histological or clinical evidence for tissue damage has not been obtained, as is the case with shockwave treatment of kidney stones. We are convinced that treatment of biliary concrements by shockwave exposure can be employed under clinical conditions in the near future. Exploration of the common bile duct (CBD) for calculi, particularly in the presence of obstructive jaundice, is a procedure with considerable mortality and 183 morbidity. To avoid the problem of retained stones, choledochoduodenostomy and transduodenal sphincteroplasty have been recommended, but have their own complications. This morbidity might be reduced by removal of CBD calculi prior to surgery. Endoscopic sphincterotomy (ES) allows this. A review of 123 cases of ES performed for calculi indicated that this was a safe (complications 8% no deaths) and reliable procedure (85 % success rate). A study was conducted of patients with known CBD stones who had either preliminary ES followed by operation at a later date (Group I) or operation alone (Group II). This study showed a lower morbidity in Group I. A prospective randomised controlled study has begun on the basis of these findings and the data from both studies are shown in the Table. These results suggest that pre-operative endoscopic sphincterotomy my reduce the morbidity of CBD stones.

Group II n = 28 Two controversies regarding the physiology of the biliary sphincter (BS) concern its functional independence from the duodenum [1] and those aspects of its acitivity which control bile flow [2] . The rabbit was chosen as the experimental animal as it has an easily identifiable sphincter. During anaesthesia induced by intravenous pentobarbital sodium, recordings of the electrical and mechanical activity of the BS and duodenum were made from (a) starved, (b) fed and (c) starved animals during administration of cholecystokinin, pentagastrin, secretin and glucagon. Spike complexes (SC) were ordinarily associated with mechanical acitivity of the sphincter and duode-num. Of 2.172 sphincter SC, recorded in 18 animals, 775 (36%) were not associated with duodenal acitivity, whereas 1.397 of 1400 (99.8 %) duodenal SC were accompanied by synchronous BS activity. This supports the hypothesis that the rabbit's BS can contract independently of the duodenum but that duodenal contraction is usually accompanied by simultaneous contraction of the BS.

Sphincter SCs correspond to its phasic acitivity. Food and cholecystokinin increased the number of SC without altering the baseline pressure of the perfused common bile duct. Pentagastrin produced a transitory increase in sphincter activity whereas :secretin and glucagon were without effect. Phasic activity of the spincter may influence bile flow through the choledochoduodenal junction. Natural blood coagulation finally results in the formation of Fibrin, which is one of the most important components of hemostasis in the human organism and thus provides the basis of all reparative procedures that are part of wound healing. It stands to reason to utilize the properties of fibrin for hemostasis during surgery and for joining severed tissue. First attempts of this kind were made at the beginning of this century. But only after greater insight had been gained into the coagulation proc-185 ess and the manufacturing techniques of blood derivatives had become more sophisticated, the essential breakthrough was made. A biological adhesive system has been developed, which consists of highly concentrated fibrinogen, thrombin and clotting factor XIII. This tissue sealant is completely resorbable and of high adhesive property. Further advantages are elasticity of consistence and excellent tissue compatibility.

After extensive animal experimentation, first clinical experience was made in 1973. In the meantime the Fibrin-Adhesive-System (FAS) has been introduced into numerous surgical disciplines with excellent results. The outstanding properties are: atraumatic tissue synthesis; enhancement of fibroblast proliferation and promotion of rapid wound healing; obtaining of local hemostasis by sealing bleeding surfaces, which is of special importance in the treatment of patients suffering from hemophilia or during operations under heparinization.

The authors experience in using the FAS within the last 10 years is reported and a review over indications, techniques and advantages of this method is given. Bile salts have been shown to enhance the stability and prolong the activity of intraluminal pancreatic enzymes and may therefore influence the effects of impaired exocrine secretion in patients with pancreatitis [1] . Individual bile salts in the peak 15 min collection of duodenal fluid following CCK/Secretin administration have been quantitated by high performance liquid chromatography in 7 patients without pancreatic or hepatic impairment (Group C), 6 patients with acute pancreatitis (Group AP) and 8 patients with chronic pancreatitis (Group CP) all with functioning gallbladders, and 4 patients with gallstone related acute pancreatitis (Group GS). The peak total bile salt output in moles and the trihydroxy: dihydroxy (Tri:Di), primary:secondary (P:S) and glycine:taurine (G:T) bile salt ratios are shown below (mean___ SEM). The duodenal aspirates contained detectable amounts of taurine and glycine conjugates of cholate, chenodeoxycholate, deoxycholate and ursodeoxycholate but not lithocholate or free bile salts. The low total bile salt output in groups CP and GS were due to decreased levels of all the individual bile salts. Although the bile salt pattern in groups C and AP were similar, the relative proportions of trihydfoxy and secondary bile salts were higher in groups CP and GS respectively.

These results indicate that patients with chronic pancreatitis without obvious large bile duct obstruction have an impaired bile salt output into the duodenum and this may exacerbate the effects of pancreatic exocrine insufficiency. An elevated amylase creatinine clearance ratio (ACCR) was considered a specific test for the diagnosis of acute pancreatitis (AP). However, it has been found elevated in other diseases as well as after surgery. The aim of this study was to evaluate prospectively the ACCR levels in patients with AP and in several groups of surgical patients. We studied 197 subjects divided into 6 groups: Group A: Acute Pancreatitis (n=67). Group B: Patients undergoing biliary tract surgery (n=38). Group C: Peptid ulcer patients undergoing gastric surgery (n=25). Group E: Patients undergoing cardiac surgery under extracorporeal circulation (n = 16). Group F: Control group of healthy subjects (n = 40).

The ACCR was determinated using the Levitt method. In the surgical groups the ACCR was measured before and after the operation. Amylase was determinated by the Phadebas amylase test. AP was diagnosed on the basis of both clinical and radiological findings and the presence of high serum amylase levels. This diagnosis was confirmed through laparotomy in 18 cases (26%).

The ACCR was 1.96 + 0.8 % (mean___ SD) in group F, control group, and 7.5+5.7% in group A, acute pancreatitits p<0.01. ACCR values below 4 % were found in 21 cases of acute pancreatitis (31%). Among those patients whose AP diagnosis was confirmed through surgery the ACCR was 10.6+7.9% (mean_+SD), higher than in the rest of the AP patients, 6.4±4.2% (p<0,05). In groups B,C,D and E (surgical groups) the ACCR before the operation was 1.83±0.8%, 2.34±0.9%, 2.28+ 1.6% and 1.54+ 1.06% (mean___ SD), respectively. After the operation it was: 3.85+2.2%, 3.18+2%, 2.83± 1.5% and 2.05+ 1.6%, respectively.

On the average, we found an increase in the ACCR levels after the operation in the biliary tract group (p<0.01), but not in the other surgical groups. In 22 patients (24 %), the ACCR after operation was above the upper limit of normal. None of these patients had symptoms compatible with clinical pancreatitis. In conclusion: 1. The ACCR increase in acute pancreatitis (sensitivity: 69o/0) 2. The average ACCR increase after biliary tract surgery, but not after either gastric or thyroid surgery or after cardiac surgery under extracorporeal circulation. However, it is possible to find isolated cases with high ACCR after any type of surgery without any symptoms of pancreatitis, suggesting that an increase of ACCR may be an unspecific finding in postoperative patients which require further investigations. Out of 217 patients with acute pancreatitis 25 developed a fulminant type. 18 were males and 7 females, mean age 45 years (range 23-83 years). All 25 patients were primarily treated by peritoneal lavage applied at laparotomy. Indication for laparotomy was sudden deterioration with (20) or without (5) organ failure. A necrotic or hemorrhagic pancreas was found in every patient. The pancreas was exposed and 2 soft and large catheters were placed close to the pancreas. Mean duration of lavage (ll/h) was 9 days. Due to secondary deterioration a pancreas resection was performed 2-10 days later in nine patients. 7 patients with acute fulminant pancreatitis died, a mortality of 27.7%. All 192 patients with the mild type of acute pancreatitis survived, thus the overall mortality was 3.2%. None treated by peritoneal lavage only developed diabetes mellitus, whereas 4 out of 5 surviving patients with an additional pancreas resection had this complication. 19 patients with acute fulminant pancreatitis displayed 3 or more Ranson criteria [1] and six died. However, no less than 32 of those with a mild type of acute pancreatitis fulfilled 3 or more of these criteria and they all survived.

A laparotomy -not a laboratory test -is necessary to confirm the diagnosis of acute fulminant pancreatitis. The indication for laparotomy is mainly clinical and therefore such a patient should preferably be handled by surgeons or physicians experienced in this disease. After confirmation of a correct diagnosis peritoneal lavage is one of the methods by which the mortality of acute fulminant pancreatits -which by conservative means is 80-100% -can be reduced. Coincidences ofhyperparathyroidism and pancreatitis have been given up by different author varying between 1% and 19%. Frequently a causal relationship has been defended. Recently, however, any causality has been queried [1] .

We analysed our own series of patients with surgically and histopathologically confirmed primary hyperparathyroidism (PHPT) (n=686) and found a coincidence with a coexisting or prior pancreatitis of 1.5 % (n = 10). From this a causal relationship cannot be concluded. However, 3 out of these 10 patients (i.e. 0.4%) had an acute onset or exacerbation of a pancreatitis immediately following the parathyroidectomy, which is strikingly more than one would expect after an operation without any anatomical relation to the pancreas (<0.01%) [2] . None of these 3 patients had another cause of the pancreatitis such as cholelithiasis or alcohol abuse. Hence a causal relationship cannot be excluded. It is imaginable that excessive amounts of parathyroid hormone are released during the surgical manipulation of the pathologically altered parathyroids. The postoperative pancreatitis may be caused by the acute elevation of parathyroid hormone levels in the presence of hypercalcemia.

In our series the parathyroidectomy was combined with a partial or total thyroidectomy in 27 patients. In another 7 patients a thyroid operation was performed prior to the parathyroidectomy. Although calcitonin has been employed as a possible therapy in patients with acute pancreatitis (AP), the normal levels of this substance in AP are not well documented and have not been correlated with PTH. Furthermore no definitive role in human calcium homeostasis is accepted for calcitonin, whereas high PTH levels have previously been correlated with hypocalcaemia [1] .

In 21 patients with AP the mean serum calcitonin on admission was 121 ng/1, and the peak level mean 173 lag/1 (upper limit of normal 75 ng/1). Calcitonin levels were higher in severe than mild AP, and 5 of 6 patients with levels >200 ng/1, were objectively graded as severe AP [1] . In the 3 hypocalcaemic patients, with corrected serum calcium <2.0 mmol/1 all recorded calcitonin levels >100 ng/1 and had elevated PTH. Nine of the 21 patients had significantly elevated PTH ('>700 rig/l) and of these only 1 did not have an associated elevation in calcitonin.

The high calcitonin levels in patients with AP suggest that supplementary calcitonin intended to inhibit pancreatic secretion is unnecessary. At present it is merely speculative to suggest a role for calcitonin in AP but intriguing to report a tendency to parallel the elevations of PTH. Nuclide labelled microspheres were used to measure pancreatic and other visceral blood flow in two groups of conscious dogs before and after intravenous alcohol infusion. Blood alcohol concentrations at the time of blood flow measurements were similar to those encountered in intoxicated humans. Thus 8 dogs (Groups A) were given a somewhat low dose of alcohol to produce a mean blood alcohol level of 0.16 gm d1-1, while 6 dogs (Group B), received substantially greater doses of alcohol, and reached a mean blood alcohol of 0.32 gm dl-1. Wilcoxon matched pairs signed rank test confirmed a biphasic, concentration-related, response of pancreatic blood flow after alcohol infusion; no such response was found in blood flow to other viscera. Moderate alcohol levels (Group A) were associated with a decrease in pancreatic blood flow (p<0.03), while high blood alcohol concentrations resulted in increased pancreatic blood flow (p'<0.02).

Colonic blood flow increased in both Groups A and B, but blood flow to the gallbladder, small intestine and the parotid gland increased in Group B only. Gastric, duodenal, renal, hepatic (arterial) and cerebral perfusion did not change.

In addition, direct observations of the surface of the pancreas showed occasional haemorrhagic areas and mottling. These findlings however could not be confirmed by objective attempts to measure blood flow in such discrete areas

In conclusion pancreatic blood flow shows a biphasic, concentration-related, response shortly after intoxication. This response appears to be peculiar to the pancreas and does not occur in other viscera. We recently showed that acute ethanol (E) and/or aspirin (A) ingestion increased the permeability of the pancreatic duct to large molecules, this suggested that pancreatic enzymes might leak from the duct into the parenchyma, causing pancreatic disease. This is a new concept in the study ot he pathophysiology of this organ (to be presented at AGA, May, 1982, Plenary Session). In the present experiments wie studied the effects of chronic E/A ingestion on pancreatic function in dogs. Methods: 6 dogs were fitted with duodenal and gastric cannulas. After recovery, baseline secretory sutdies were performed by cannulating the pancreatic duct and collecting pancreatic juice during secretin infusion (0.1 (submax) or 2.0 U/kg-hr. (maximal) IV). At least 2 studies were performed in each dog at each dose. After baseline studies, 3 dogs (Group E) were given daily intragastric ethanol (2 gm/kg-day). 3 dogs (Group A/E) were given E and A (100 mg/kgday). After 36-48 weeks, secretory studies were repeated. Results: All dogs gained weight (x = 1.28 kg). The pancreases appeared normal by light microscopy. Drug treatment increased volume and HCO3 output by 47 and 67%, respectively, in Group E, submax secretin, but decreased them by 29 and 34 % in Group A/E animals. (p=0.0001 vs. predrug values and Group E vs. A/E values). Drug treatment decreased volume and HCO3 output in both Groups by 5-10°/c (p=0.0001) after maximal secretin stimulus. (Manova test for all statistical analyses). Conclusions: Consumption of E alone increased volume and HCO 3 output after submaximal and decreased them after a maximal secretin stimulus. This confirms the work of Sarles. Consumption of E and A reduced volume an I-ICO3 output after secretin at both doses. Thus chronic A ingestion further impaired pancreatic function in these animals. Since only a small proportion of chronic alcoholics develop clinically significant pancreatic disease, an aggravating "cofactor" may be operating in this group. Chronic ASA ingestion, not uncommon in alcoholics, may represent such a cofactor. The development of diabetes mellitus in pancreatic cancer is well known but the standard oral glucose tolerance test is not recognised as a useful diagnostic indicator. Glucose homeostasis, insulin and C-peptide secretion in response to intravenous glucagon were studied prospectively in patients with suspected pancreatic cancer and assessed as to their diagnostic value. 34 fasting patients were given glucagon, 1 m.g., i.v., and serial measurements of blood glucose, plasma insulin and C-peptide concentrations made for 60 min. Subsequently it was shown that 12 patients had pancreatic cancer and the remainder constituted a control group. There was not significant difference in the rise in blood glucose between the groups after glucagon. The mean plasma insulin concentrations rose rapidly in both groups peaking between 2 and 5 rain but the values were sginificantly lower in the pancreatic cancer group (P<0.01 at 5 min: P<0.001 at 15 min: P<0.01 at 60 min). A similar pattern was observed with C-peptide. In patients with obstructive jaundice the plasma insulin response was a better discriminator of pancreatic cancer. We conclude that abnormal pancreatic beta cell funktion exists in patients with pancreatic carcinoma, detectable before any change in glucose homeostasis, particularly in patients with obstructive jaundice. The glucagon stimulation test may have a useful role in the diagnosis of pancreatic cancer. T-suppressor cells (T~) have previously been implicated as mediators of graft surival in baboons tolerant to their renal grafts [1] . In addition, it seems that T-helper cells (Th) are also affected by totallymphoid irradiation in that they are unable to provide help in mitogen-induced T-cell responses or pokeweed mitogen induced immunoglobulin synthesis in Bcells. In this study the evolution of T~ and Th is followed in baboons undergoing graft rejection at different rates. Peripheral blood was collected from the animals at weekly intervals after renal transplantation, defibrinated and the lymphoid cells isolated by flotation on Ficol Hypaque. The T-lymphocyte fraction was purified by filtration through a column packed with nylon wool. Th, Ts and total T-cells were enumerated using monoclonal antibodies OKT4, 8 and 11 respectively (Ortho). The Th/Ts rations reported were taken immediately before a rapid rise in serum creatinine occurred. In longlived baboons who maintained normal creatinine values, a mean of the ratios over the last month was used. B rats are Produced by sublethal (750 Rad) X-irradiation of thymectomized animals, reconstituted with bone marrow from syngeneic, thymectomized, thoracic duct drained donors. In these Lew rats, (Lew x BN)F1 cardiac allografts survice indefinitely (>100 days); unmodified Lew rats acutely reject such allografts (8-+ 1 days). In this study, we have tried to restore the processes of acute rejection in B recipients. Graft survival appeared independent of blocking factors or suppressor cells, as transfer of serum or lymphocytes from B recipients into syngeneic normal animals failed to increase survival of test allografts, placed subsequently. Similarly, immunogenicity of long surviving grafts was unchanged; such grafts functioning >100 days and retransplanted into normal animals were rejected acutely. Adoptive transfer of 108 unseparated spleen cells (SL) from nonimmune syngeneic animals produced slow rejection (28--+5 days) in B rats; 108 sensitized SLwas somewhat more effective (22_+ 1 day). Transfer of 2 x 107 syngeneic peritoneal exudate (PE) cells plus 108 sensitized SL caused acute rejection in 50 % orB recipients (11 _ 1 days), the remainder experiencing rejection at c 3 weeks. PE cells harvested from rats injected ip with thioglycollate 3 days previously, were primarily macrophages/adherent cells, as the cells used were that fraction sticking to plastic dishes and removed with lidocaine (purity >90 %). In vitro, B rat macrophages were abnormal, having only 50 % of capacity of normal macrophages to promote production of interleukin 2 (IL2) when co-cultured with purified T lymphocytes. However, B recipients experienced acute graft rejection (10-+ 1 days) after transfer of 108 sensitized SL plus semipurified IL2, thus bypassing the above defect. Addition oflL2 to the T cell (purity >95°/0) equivalent of 108 SL (purified over Degalan bead columns coated with rabbit antirat lgG, nonadherent fraction) failed to reestablish acute rejection (17-+7 days), while further addition of B lymphocytes (Degalan bead adherent fraction, purity >90%) or macrophages was uninfluential (16___2 days and 16___3 days, respectively). Transfer of IL-2 alone never produced rejection. Acute rejection can be re-established, however, by increasing the number ofT lymphocytes (108, transferred concomitantly with IL2). Thus, the state of unresponsiveness in B rats can be reversed in vivo by adoptive transfer of particular cellular elements in the presence of growth factors; increased graft survival seems dependent ultimately upon IL-2 production by sensitized T cells, presumably T helper cells. The relative inability of B rat macrophages to promote production of IL-2 by T cells may be primarily responsible for the immunological deficit of the B rat. One of the most intriguing findings ofCyclosporin A (CyA) immunosuppression is that in some species a short course of treatment will produce very prolonged allograft survival. We have tested the ability of CyA to prolong the survival ofvascularized heart, kidney and pancreas allografts by direct comparison in a DA (RT1 a) to LEW (RT1 ~) rat allograft model. Accessory abdominal heart and orthotopic left kidney transplantation were performed using standard microsurgical techniques. In renal transplantation the left kidney was removed at the time of transplantation, the remaining right kidney 5 days thereaf- ter. Streptozotocin-diabetic animals received ductligated pancreas whole organ grafts isolated on the portal vein and a segment of the aorta giving offthe coeliac axis and the superior mesenteric artery. Rejection was taken as complete stop of palpable pulsations in heart transplantation, the day of death in renal transplantation and recurrance of hyperglycemia above 14 mmol/1 in pancreas transplantation, respectively. Cyclosporin A 15 mg/kg body weight, dissolved in olive oil, was administered intramuscularly for 14 days starting with the day of transplantation. In all instances functional demonstration of rejection was confirmed by histological examination. Cyclosporin A is effective to prolong the survival of vascularized heart, kidney and pancreas allografts. While CyA is administered none of the grafts has been rejected. However, following withdrawal of the drug pancreas grafts are rejected within 14 days and heart grafts within 33 days. None of the kidney grafts has been rejected so far.

The differential susceptibility of vascularized heart, kidney and pancreas allografts to CyA immunosuppression may be caused by differences in immunogenicity due to organ specific alloantigens or a differential representation of spezialized antigen presenting cells. It may also reflect different patterns of rejection of the various organs. During CyA administration all rejection processes are effectively suppressed. In the maintaince phase after withdrawal of CyA such immune responses may prevail and ultimately lead to rejection of pancreas and to a lesser degree of heart allografts. The venous allograft still remains an attractive alternative for the reconstruction of small caliber vessels. However, when the venous graft is introduced to a non-histocompatible host, rejection and early occlusion is the rule. This study evaluates the use of Cyclosporin A (CyA) as a graft pretreatment, or systemic immunosuppressant for venous allografts. In addi-tion, cryopreservation techniques for pretreated venous allografts was investigated. 45 adult mongrel dogs, weighing between 12 and 24 kg, were used as recipients for donor jugular vein segments (5-6 cm) which had been excised and flushed with 100 cc of plasma protein fraction (PPF) at 4°C. These venous allografts were anastamosed end-to-end into a carotid artery of the recipients. The animals were divided into five groups as follows: Group I (n=6) received untreated venous allografts without subsequent immunosuppression, Group II (n = 5) was the same as Group I with minimal immunosuppression (azathioprine 2.5 mg/kg/day). In Group III (n= 10) the animals were transplanted with venous grafts stored in 25 cc of plasma protein fraction (PPF) containing Cy A (50 mg/1) at 4 ° C for 24 hours, immunosuppression was as in Group II. In Group 1V (n=9) the animals received allografts that had been cryopreserved in a 15% DMSO solution at -196 ° C for 25-35 days and then the animals had azathioprine as in Group II. In Group V (n= 15) venous allografts recipients were treated with systemic CyA (20 mg/kg/day x 2 weeks, followed by 15 mg/kg/day x 2 weeks) as the only immunosuppression. The patency of the allografts was evaluated at 1, 2, 3, 4 and 8 weeks post transplantation. Patency results at one month showed that azathioprine alone failed to improve the patency rate (Gr. I and II = 0% patency). CyA graft pretreatment, however, significantly improved the one month patency (Gr. III = 57%). In addition, cryopreservation appeared to enhance the graft pretreatment effect of CyA (Gr. IV; one month patency = 85.7%) of the allografts. Finally, Systemic CyA proved very effective in preventing rejection and occlusion (Gr. V; one month patency = 100 %). Grafts that remained patent for a initial critical period of 4-6 weeks, all showed long term patency. The effect of CyA in preventing graft rejection was further documented by histiological studies of the allografts which showed a marked cellular infiltration and degenerative changes in all the grafts of the control group as compared to minimal or no cell infiltration in the patent grafts of the treatment groups. In summary, it appears that CyA used as a graft pretreatment with minimal immunosuppression of the recipient, in conjunction with cryopreservation or given systemically as the sole immunosuppressant can significantly improve the survival of venous allografts. In our previous reports, it was shown that isolated hepatocytes transplanted into the splenic parenchyma of syngeneic rats, proliferated markedly and recomposed the hepatic tissue. This experimental system provided a new model to elucidate the mechanism of hepatic regeneration which could not be obtained in in vitro cell culture experiments. In the present paper, fetal hepatic tissue instead of isolated adult rat hepatocytes were transplanted into the rat spleen. We document briefly long-term morphological observations on the transplanted fetal hepatic tissue with special reference to proliferation of the hepatocytes and bile ducts. Materials andMethods:Wistar rats were mated in our laboratory for a fetal liver source. Gestation day 0 was when a plug or sperm were observed in the vaginal smear. Fetuses used were of 18 to 20 days gestation. About ten fetal livers which were obtained from one maternal rat, were minced with scissors. The liver fragments were washed three times with saline solution. Transplantation was carried out by direct injection into the spleens of syngeneic adult rats using a 15 gauge needle. Half of the liver fragments obtained from one maternal rat were innoculated into the spleen of one animal. A total of approximately 30 rats with transplanted liver fragments were killed 1, 3, 7, 14 and 30 days and then every two to three months until one year after transplantation. The spleens removed were stained by H.E., PAS and silver nitrate for histological examination. Results: Fetal livers exhibited no lobular architecture or hepatic cord structure. The very sparse cytoplasm of the hepatocytes and many hemopoetic cells among the hepatocytes were characteristically found only in the fetuses. One week after transplantation, the survived hepatocytes revealed almost the same morphological features as in fetal liver except for the presence of several proliferated bile ducts around the hepatocytes. Two weeks later, the hepatocytes formed apparent hepatic cord structures and the extoplasm of each hepatocyte increased abunduntly and became acidophilic as seen in normal neonatal hepatocytes. Hemopoetic cells disappeared. Four weeks later, hepatocytes began to proliferate sporadically among the markedly proliferated bile ducts, Groups of survived hepatocytes with cord structure were very similar to a neonatal liver except for the lack of the Glisson's area. Two or three months later, proliferation of the hepatocytes became prominent. There seemed to be no interrelationship between proliferated hepatocytes and bile ducts. One year after transplantation, a white nodule was observed on the spleen macroscopically and it consisted of numerous bile ducts and hepatocytes with or without cord structure on histology. Summary: 1. Fetal hepatocytes transplanted into the spleen, differentiated to almost normal neonatal hepatocytes two weeks after transplantation. 2.

Hepatocytes began to proliferate about 4 weeks after transplantation. 3. Three days after transplantation, proliferation of bile ducts was already observed independent of the transplanted hepatic tissue. 4. When comparing the difference in proliferation between fetal hepatic tissue and isolated hepatocyte transplantation, marked proliferation of the bile ducts in fetal hepatic tissue was observed and fetal hepatocytes proliferated more rapidly, while there were no proliferated bile ducts in isolated hepatocyte transplantation. Pretransplant splenectomy (Sx) has been of disputed benefit since its introduction two decades ago. 220 of 315 patients with first cadaver transplants treated at our institution between Dec. 1968 and Dec. 1981 have had pretransplant Sx. At six monts, Sx patients had 10% better kidney survival, but this benefit was lost shortly after 1 year and by 5 and 10 years was 10% and 15% worse in Sx patients. Patient survival for Sx and no Sx was identical for the first year but was 10% and 20% worse by 5 and 10 years respectively in Sx patients. Thus, the early improvement in kidney survival was more than offset by a late high mortality. A rational basis for selecting patients who might benefit most from pretansplant splenectomy is urgently needed. Since July, 1978, 34 patients ages 16-45 have received first cadaver transplants after having been tested for reactivity to DNCB. Nine of 20 DNCB negative patients had splenectomy as did 6 of 14 DNCB positive patients. Kidney survival at 1 year for DNCB negative patients without Sx was 79%; for DNCB negative with Sx, 29%; for DNCB positive without Sx, 21%; for DNCB positive with Sx, 62%. Rejection was the sole cause for kidneyloss in DNCB positive patients without Sx. However, 5 of 9 DNCB negative patients with splenectomy died, primarily of septic complications. Since 1978 survival of Sx patients has been 60% compared to 95% in non Sx patients (p<0.04). Sx appears to be beneficial in DNCB positive patients but has an adverse effect in DNCB negative patients because of an increased susceptibility to fatal infections. Prior blood transfusion improves renal graft survival [1] . Plasma from uraemic patients suppresses the in vitro responses of normal lymphocytes to antigen (plasma suppressive activity, PSA) and this effect is mainly attributable to the plasma protein macroglobulin (a2M) [2] . The aims of the present study were: a) to identify changes in PSA and a2M concentration in uraemic subjects following primary blood transfusion. b) to correlate the PSA of transfused renal transplant recipients with subsequent graft survival. a) Ten potential transplant recipients were studied before and after their first blood transfusion. Following blood transfusions the PSA increased significantly (p<0.01) reaching a maximum at two months. There was no significant change in the plasma a2M concentration over the same period.

b) The plasma of 137 consecutive chronic renal failure patients was tested for PSA prior to renal transplantation and before institution of immunosuppressive therapy. All but two patients had received previous blood transfusions. After transplantation patients were followed for a minimum of 3 months and a maximum of 42 months. 16 grafts failed for non-immunological reasons and were excluded from the study group. Patients were divided into two groups according to the degree of suppressive activity of their plasma. A volume of5/t 1, producing a 50% inhibition of normal lymphocytes was used as a treshold to differentiate those with a high or low suppressive activity.

Graft survival in the first three months was significantly better, 92% (/7<0.05) for those recipients with a high PSA as compared to 74% for those with a low PSA.

We conclude that blood transfusion causes a significant increase in PSA although not a2M concentration and that patients with high PSA have a better graft survival. The effect of in vitro steroid on antibody dependent cellular cytotoxicity (ADCC) was studied in patients awaiting renal allotransplantation and the results were correlated with transplant outcome. 85 recipients of primary cadaveric allografts were classified as steroidsensitive or steroid-resistant from the degree of ADCC suppression induced in vitro by methylprednisolone, 36 patients being steroid-sensitive and 49 steroid-resistant. Following transplantation patients received azathioprine and prednisone, and rejection crises were treated with bolus doses of methyl-prednisolone.

Graft failure occured in 5 of the 36 steroid-sensitive patients, and in 34 of the 49 steroid-resistant patients. The observed one year graft survival rate was 57.6% for the whole group, 83.1% for the patients with steroid-sensitive ADCC and 36.9 % for those with steroid-resistant ADCC, the difference between the two groups being highly significant (XZ=23.6). A high incidence of early graft failure was seen in steroid-resistant ADCC patients, 49.7% of grafts being lost in the three months after transplantation, as compared with only 2 of 36 graft failures in the steroid-sensitive ADCC group in the same period. Analysis of HLA-A, HLA-B and HLA-DR incompatibilities showed no significant difference between the groups, and since all patients had received deliberate pregraft blood transfusion, the difference in survival rates between the two groups appears to be independent of these two variables.

These findings confirm our preliminary observation that pregraft assay of ADCC response to in vitro steroids identifies those patients who are unlikely to respond to steroid therapy in the treatment of rejection, and in whom alternative forms of therapy may be appropriate. Post-operative DXT, whilst not influencing survival, protected patients from loco-regional recurrence 07<0.001, hazard ratio (HR) = 3.1). Interestingly it was found to be most effective against axiallary node recurrence (p<0.001, HR = 4.0), reasonably effective against chest wall recurrence (/7<0.001, HR=2.1) but conferred no protection against supraclavicular node recurrence (HR = 1.2) in spite of a supraclavicular field being routinely employed in the radiotherapy technique.

With such large numbers involved, this trial has facilitated the study of the prognostic significance of sub-groups of patients with different patterns oflocoregional recurrence as first evidence of treatment failure (see table) .

Of those patients developing loco-regional recurrence who have since died (222 out of 356 in WP group; 73 out of 128 in DXT group) 63% in the WP group and 69% in the DXT group did so with evidence of persistent loco-reglonal disease. However, the incidence of uncontrolled local disease at death was higher in the WP group overall. Stress as well as dietary fatty acids have been shown to prolong allograft survival in rats [1] . Poly unsaturated fatty acids (linoleic acid, arachnoidic acid) have been reported to depress immune response [2] . Depressed immune response was suggested to correlate with a higher incidence of spontaneous tumor [3] as well as with an increased growth rate of inoculated tumors [4] . The objective of this study was to elucidate the effect of two environmental factors i.e. chronic stress (change in light/dark pattern) and diets low and high in linoleic acid on immune response and growth of transplantable tumors in BN rats. Immune response: Four experimental groups (n >10) were used in immune response studies. Group I: high linoleic acid dietl; group I1: low linoleic diet 2, group III: L/D shift weekly, normal diet and group IV: controls on normal diet, normal lighting. Seven weeks after the start of the experiment the immune response was measured. The results showed that corticosterone levels were slightly increased in all experimental groups, although only the high linoleic group showed statistic significant difference with the control group. Cellular immune response (Con A stimulation and Popliteal Node Assay) was decreased in all experimental groups compaired to controls. Transplantable tumors: 1 x 106 leukemia cells were injected i.v. and pieces of 1 mm 3 of an spontaneous adrenal cortical carcinoma, a urethral squamous cell carcinoma and a round cell cervix sarcoma were implanted subcutaneously. All tumors were inoculated in groups of 10 animals each. Spleen weight as a measure of leukemia growth was high in the control group and low in the experimental group. The same pattern was seen in the growth of the subcutaneously implanted adrenal cortical carcinomas. Both the urethral squamous cell carcinoma and the round cell cervix sarcoma, being non-immunogenic, did not show any difference in growth. So far, it can be concluded, that the immunosuppression as induced by mild chronic stress or dietary fatty acids does not lead to enhanced tumor growth. In contrary, the results of both leukemia and adrenal cortical carcinoma show a possible reserve effect. Little is known of the derivation or content of human breast cysts. Recent reports have shown wide variations in the content of steroid hormones, particularly dehydroepiandrosterone sulphate (DHAS) [1, 2] . No explanation for this is apparent. To confirm the large variation in DHAS concentrations and to further define the contents of cyst fluids, 100 cysts from 85 patients have been analysed for DHAS, sodium and potassium. DHAS concentrations ranged from 1.5-1155 pmol/1. Both sodium and potassium content also varied widely (sodium 30-200 pmol/1 and potassium 3-158 ~umol/1). There was a significant direct correlation between the content of potassium and DHAS in cyst fluid (p<0.001) and a significant negative correlation with sodium content (/7<0.001). Three separate subpopulations of cysts could be identified according to the sodium and potassium content and these were, predominantly potassium cysts (47), predominantly sodium cysts (44) and mixed cationic cysts (9). The median DHAS concentration of the potassium cysts was 215 pmol/1 similar to the levels found in human breast secretions [3]. In contrast the median concentration of DHAS in the predominantly sodium cysts was 14 pmol/1 and significantly different (p<0.0005), with many of these cysts having DHAS concentrations in the same range as those found in plasma. The remaining mixed cysts had a median DHAS concentration intermediate between the two main groups. It may be that the variation in cationic content and DHAS concentration in these two major subpopula-tions of human breast cysts represents either, derivation from two different sources, namely breast secretions and plasma or marked differences in the secretory activity of the epithelium lining these two groups of cysts. There is no uniform agreement on the correct management of patients with invasive lobular carcinoma (ILC). It is widely considered that in ILC there is an increased risk of developing a contra-lateral carcinoma and the major controversy surrounds the management of the second breast. The survival of patients with ILC was significantly better than that of IDC fp<:0.025). SIX patients had bilateral carcinomata at diagnosis and a further 14 developed a contra-lateral carcinoma during the period of follow-up (10 to 21 years). Survival data showed poor survival for patients with simultaneous bilateral disease, but no difference in survival for patients with metachronous bilateral or unilateral disease. This suggests that the later development of a second carcinoma does not necessarily reduce the probability of survival for patients with ILC.

The major factor predicting patients at risk of developing a contralateral carcinoma was histologi-195 cal type. Of 22 patients with a particular histological pattern of ILC [4] with a classical pattern of spread but showing nuclear pleomorphism and cellular cohesion, 10 developed a contralateral carcinome, compared with a further 10 in the remaining 81 patients (p<0.001). If bilateral mastectomy is justified it ought to be restricted to patients with this histological type of ILC. Both the anti-oestrogen Tamoxifen and cyclical combined chemotherapy will provide significant palliation in advanced breast cancer. The optimal use of these agents requires further evaluation and thus this trial was designed to compare a combination ofcytotoxic therapy and Tamoxifen, against cytotoxics alone in patients with advanced breast cancer. Post-menopausal patients presenting with metastatic breast cancer, locally advanced cancer extending beyond the breast and regional nodes, or with tumor recurrence following primary local treatment were allocated to the 3 treatment arms via sequential manner. Doxorubicin, Cyclophosphamide, 5-Fluouracil, and Vincristine were given intravenously once every 3 weeks. Tamoxifen was prescribed in a dose of 20 mg. b.d. On failure or relapse from one of the single modality arms, a crossover of those arms occurred. The combination consisted of both the above therapies.

Assessment of therapies was made in terms of objective response (UICC criteria), duration of response, and survival. We have previously reported that the combination results in a significantly greater response rate [1] . as a result of stenosis reducing flow or by platelet embolisation [1] . As neither aniography nor ultrasound can identify thrombotic activity we have evaluated gamma camera neck imaging using n 1indium platelets. Labelled platelets on endarterectomy specimens were also measured and the activities found were then examined in a theoretical model. Twentyfive patients with TIA received rain platelets and sequential gamma images were interpreted by two observers, Carotid endarterectomy in 11 patients allowed measurement of specimen radioactivities. Angiography and Doppler spectral analysis [2] were also performed. All endarterectomy specimens contained labelled platelet deposits with the most active equivalent to platelets from 1.8 ml of blood. This activity level was at the threshold of resolution in the theoretical model. Both observers agreed that 22 of the 50 carotid bifurcations showed platelet accumulation on imaging. Of the 12 atheromatous ulcers demonstrated by angiography 11 were visualised, but only 5 of 10 stenoses greater than 80 per cent were detectable~ Since ultrasound identified all stenoses only one angiographically diseased carotid was not detected by combining Doppler and platelet imaging.

Diseased carotids accumulate rain platelets with the more thrombogenic ulcerated plaques identified more frequently than stenoses. Long term follow-up is required to establish the clinical relevance of platelet deposition. Major problem in vascular endoscopy is the existence of blood which prevents clear visualization.

We devised a new technique using a combination of balloon catheter and slender fiberoptic endoscope, by which clear visualization was obtained experimentally and clinically. Three to four pairs of orifices of intercostal arteries were also visualized in one visual field. In some dogs, acute aortic dissection was experimentally created by means of Blanton's method. The entry, which was located at the descending aorta just distal to the left subclavian artery, was clearly identified. Complete occlusion of blood flow and clear visualization could be obtained when balloon pressure exceeded systemic blood pressure.

Clinical Study: In six patients requiring major vascular reconstruction of the aorta (abdominal aneurysm 4, Leriche's syndrome 1, dissecting aneurysm 1), vascular endoscopy was performed intraoperatively. In five patients, balloon catheter was introduced through the one of the limbs of Y graft after proximal anastomosis. In each case, orifices of the major abdominal aortic branches were clearly observed. Irregular orifices and atheromatous plaque of the aortic intima which were not expected from aortogram, were also identified in all patients. Intimal tears by vascular claps were more extensive than expected and anastomotic suture lines were able to be checked from inside. In a case of dissecting aneurysm, balloon catheter was advanced through the 11 mm graft which was sutured to the common femoral artery with finding the entry just above the left renal artery.

using fiberoptic endoscope and balloon catheter was useful to observe orifices of the major aortic branches, unexpected intimal tears by vascular clamps and atheromatus plaques. It was particularly usbful to check the anastomotic suture line from inside of the aorta and to identify the exact location of the entry in dissecting aneurysm. Vascular endoscopy could be one of the invaluable methods to examine, diagnose and treat the patients requiring aortic, caval and other major vascular surgery. (3) produced endothelial injury and a local increase in shear stress in 14 cynomolgus monkeys by suture plicating and constricting the aorta and then feeding an atherogenic diet for 6 months. Our findings reveal that carotid plaques localize on the outer wall of the internal carotid (plaque thickness 0.7--+0.1 mm) which is an area of low flow velocity ( -4___ 1 cm/s at RE 800) and shear stress (0-+2 dynes/cm 2) and not at the flow divider (thickness 0.1___0.1 mm, p<0.01) which is an area of high flow velocity (81---3 cm/s) and shear (161-+48 dynes/cm2). Distal to the carotid bulb, velocity and shear increased on the outer wall and little or no plaque was observed. In experimental coarctations, no endothelial damage was observed (SEM and TEM) within the high-shear coarct channel and the channel was noted to be free ofatherosclerotic plaque despite the development of extensive diet-induced lesions proximally and distally. Thus, high flow velocity and shear stress do not appear to produce endothelial damage in vivo. In addition, plaques were minimal in high shear areas in the human carotid bifurcation and high shear appears to have an inhibitory effect on experimental plaque formation. These data contradict previous investigations implicating high shear stress in plaque pathogenesis. In contrast, host aortic endothelium (AE) fails to cover large VP by pannus ingrowth even over much longer times. To see if IAES succeeds because of inherent differences in growth potential between AE and VE, we used AE to seed 8 cm x 6 mm diameter Dacron velour infrarenal VP in dogs. An average of 4 x 105 cells obtained by trypsin/collagenase digestion of the bypassed aortic segment was used to seed each VP by a 4 step preclotting method. The identity of AE was confirmed by stains for Factor VIII antigen. Viability of seeded AE was verified by growth of subaliquots in tissue culture. Six weeks after surgery central segments of AEseeded (n = 9) and control unseeded (n = 9) VP were compared by light and scanning electron microscopy using an endothelial coverage score range of -1 (for fibrin/platelet thrombi) to +1 (for confluent endothelial coverage). AE-seeded VP had a score of+0.89 ___ 0.33 (mean___ SD) versus. -0.67___ 0.70 for controls (P<0.001). In addition to endothelial coverage, the subluminal smooth muscle and intramural vasa vasorum previously reported in VE-seeded VP were also seen in AE-seeded VP. Since AE and VE seeding give identical results, the success of IAES with VE cannot be due to inherent biological differences in mitotic potential between AE and VE. IAES must instead achieve additional endothelial growth either through a) the action of the proteolytic enzymes used for cell harvest or b) mitogenic stimuli to nonconfluent cells at the edges of seeded cell clusters on the VP. Further improvement of the efficiency of IAES to allow use of less harvested vein per cm 2 of VP should come from enhancing one or both of these effects. Pyrolytic carbon is a crystalline form of carbon that has been extensively used in the construction of cardiac and bone prostheses. Since it has also been suggested that pyrolytic carbon will prevent thrombosis from occuring in vascular prostheses, the aim of the present study performed in dogs was to test the immediate blood compatibility of this material and to evaluate its biocompatibility when inserted as vascular substitute. After pryolysis of a gazeous hydrocarbon, the carbone crystalite was deposited on a knitted textile surface or tube. Its surface examined by scanning electron microscopy (SEM) was rough and porous to a depth of 40p. This material was tested 1 °) for immediate hemocompatibility as inserts within the vascular lumen (aorta and inferior vena cava). The specimens were examined sequentially by SEM and histology at 10, 20, 30, 180 s and 10 min after reestablishment of the blood flow, 2 ° ) for long term biocompatibility as vascular cylinders (7 mm ID) inserted either in the aorta or inferior vena cava or as intraatrial (left or right) implants.

Patency of vascular cylinders was tested during 2 postoperative month by Doppler ultrasound investigations, specimens were examined by histology, electron microscopy (scanning transmission) at 15, 30 and 60 days following implantation. Satellite lymph nodes were examined by histology. Already 10 s after establishement of the blood flow, platelet adhesion and limited fibrin mesh with few erythrocytes developed on the material. Platelet aggregates of limited extent were only observed on intravenous implants. Plasmatic protein deposition, an early event on polymeric vascular material was not observed. After 30 s a fibrino-erythrocytic membrane recovers completely the material. Except in the case of intravenous insert, no thrombosis developed at the contact of intraarterial or intracardiac implant. After 15 days it was completely recovered by a 5-7 fibrocellular layer consisting of large myofibroblasts with microfilaments, newly synthetized collagen and elastin. The blood interface was of fibrous nature. At one month by SEM, endotheliallike cells developed in a mosaic-like pattern, characterized by transmission E.M., by microvillous projections, numerous pinocytic vesicles and intercellular tight junctions. This endothelial-like cell lining was complete 2 months after implantation. Their immunocytochemical properties are now under investigation using specific anti-dog Factor VIII-RAG sera. Although preliminary, the present results suggest that among the numerous vascular biomaterials tested, pyrolytic carbon may represent a unique feature of rapid cell development and differentiation of endothelial lining at the blood material interface.

Department of Connective Tissue Biology, Institute of Anatomy, University of Aarhus, Aarhus, Denmark In the surgical clinic a significant number of patients report that their incision wound has burst, even though the scar appears to be intact. By mechanical testing of strips from skin wounds we have noticed a breaking pattern, in which the deepest layer of the wound ruptures earlier than the superficial part. Therefore, we have investigated the strength and extensibility of rat skin wounds at different levels (superficial-deep) of the epidermis (0.03 mm), dermis (1.8-2.2 mm) and m. panniculus carneous (0.6-1.5 mm), average thickness is indicated. 10, 20 and 60 day old standardized skin wounds from the dorsal region of rats have been used. Strips were punched out at right angle to the wound line and mounted in a materials testing machine. The strips were stretched until rupture and load-strain curves registered continuously. Simultaneously, the strips were transluminated and the breaking pattern was studied by taking photographs of the wound specimens during the mechanical testing (30-35 photographs of each specimen). The photographs were marked on the load-strain curve by means of a connection between camera and x-y-recorder. From the load-strain curves the maximum load and the failure energy were calculated.

The breaking patterns of 10, 20 and 60 day old wounds were found to be similar. The deep part of the wounds ruptured first. The force required to break the deep part was less than that required to break the superficial part of the wounds. The musculus panniculus carneous was very extensible and did not break. However, it possessed only minimal strength.

Quantitative measurements of the strength of the combined superficial-deep layers were performed on 2 mm wound strips. Specimens contained the superficial 0.5, 1.0 and 1.5 mm of the wound area and were produced by cutting off the deep layer parallel to the skin surface. Specimens containing the total wound area down to the musculus panniculus carneous were produced by cutting off the muscular tissue. These specimens were mechanically tested as described above.

The present studies demonstrate the mechanical inhomogeneity of incision wounds. A new method for testing the mechanical properties of the tissue of incision wounds at various levels (superficial-deep) is presented. The superficial layer of an incision wound contributes a major part of the strength of the wound and is more extensible than the deep layers. These results may explain the clinical observations. The effect on wound healing of different kinds of vitamins is worth investigating, since the efficiency of Vitamin C has been dearly demonstrated. The possible action of Vitamin Bs-whose trophic effect on skin is well known -has been experimentally studied on skin and aponeurosis healing after a standard laparotomy. Materials and Methods: Experiments were carried out on 42 five months old rabbits which were randomly divided into three groups: in group I, 18 animals served as controls (3 animals in sequence of 5 days from the 5th to the 30th post-operative day), group II, 12 animals injected with Vit B5 (25 mg/kg of body weight/24 h) and group III, 12 animals injected with a placebo (2 animals in sequence of 5 days for each group). In each case four samples were tested of skin and aponeurosis for determinating tensile strength, directly recorded with an original technic [1] : this new apparatus allowed us to obtain simultaneously two dynamic parameters, the healing tensile strength and stretching of the scar. Results: 1. No significant difference was found between controls (Group I) and the placebo group (Group III) both for resistance of skin and the aponeurosis. 2. As far as Vitamin B5 treated animals were concerned (Group II) there was no significant difference regarding skin resistance when compared with the other two groups. 3. Inversely aponeurosis resistance become significantly greater when measured on the 20th (p<0.050), 25th (/r<0.025) and 30th (p<0.050) post-operative day. In 17 mongrel dogs (7x 15 cm cranial based rectus abdominis) MC and corresponding RP flaps were raised. In group I (9 dogs) skin BF was determined from the clearance curve for ~33Xenon injected intradermally and measured with a computer-linked gamma camera. In group II (8 dogs) subcutaneous PrO2 was determined by a recently developed method using a silastic tonometer, subcutaneously implanted. The PrO 2 inside the tonometer was measured in infused saline, by a platinum oxygen needle electrode and a silver/silver Chloride reference electrode. B F and PtO2 were measured before and after the flaps were raised and on postoperative days (POD) 1, 3, 6 and 15. PtO2 were taken at 6 various inspiratoric oxygen levels (F~O2) ranging from 21% (air) to 100 % oxygen. Intact areas lateral to the flaps and in flap regions prior to surgery served as controls. Immediately after surgery BF in the MC increased while in the RP flaps was 24 %, 23 % and 31% of the flow in the MC flaps, in lateral intact area and in the preoperative areas (P<0.001). During POD 1-2 BF in the RP flaps increased to the preoperative level, but not to the increased levels found in the MC flaps and the lateral intact areas. By POD 6 there were no differences in BF between the two types of flaps and the lateral areas, but all were higher than corresponding preoperative values (/'<0.005). Tissue oxygen tension showed a dramatic fall POD 1, 3 and 6 in the RP flaps for all FIO2, and for all days the values were lower than the preoperative level (P<0.001). One RP developed POD 3 distal necrosis and the PrO 2 was then 0 even with a FIO2 of 100%. The MC flap showed an increased PrO2 on the operative day but at POD 3 the values were slightly lower than the preoperative level, but POD 1, 3 and 6 the values for all FIO2 were higher than for RP flaps (P<0,005). At POD 15 the PrO2 reached preoperative level for RP as well as MC flaps. Lateral intact areas showed comparable changes to that observed in the MC flaps.

It is concluded that the MC flap demonstrates superior BF as well as PrO2 when compared to the RP flap. Early postoperative PrO 2 in the distal part of the RP flaps is critically low despite of increasing F~O2 to 100 % and increasing BF. Differences in BF and PtO2 may be the biologic factors responsible for the superior healing characteristics of the MC flap. (1) ATP~ADP + Pi (inorganic phosphate)

(2) PCr + ADP~ATP The net result ofreaotjOxas 1 and 2 is a fall in PCr and a rise in Pi while ATP ~'emains relatively constant. All of the phosphorus metabolites are easily measured in gastrocnemiaas muscle using 31PNMR spectroscopy. Normal volunteers and patients with angiographically documented arterial occlusions were studied in a 101/4" bore Oxford Research Systems TMR-32 spectrometer at rest and after exercising each limb separately. Normal resting values ofPCr/P iwere >10 and the NMR Index = P/(P~+PCr) was 0.07_+0.03 (S.D.). Limbs with femoral arterial occlusions whose ankle systolic pressure index was <0.5 had NMR Index which was significantly elevated above norreals (0.18 + 0.09p<0.01) indicating a failure of metabolic compensation for reduced bloodflow and oxygen delivery, although ATP concentration was norreal. Exercise produced a five-fold rise in NMR Index in both normal and diseased legs. Spectra were taken over one minute intervals during the recovery period and in normal limbs returned to resting values within 2 rain. The recovery period was considerably slower in the diseased limbs indicating abnormal mitochondrial oxygen delivery and impaired mitochondrial formation of ATP. These data demonstrate the feasibility of using 31PNMR to non-invasively probe the biochemical abnormalities of energy metabolism in patients with peripheral vascular disease. The incidence of urinary calculous disease (UCD) in the South African Black population is very low in comparison with the White population group. No biochemical differences in serum nor urine account for this discrepancy and no other measurable parameters have demonstrated any difference between the two groups. Urinary particulate activity measurements have demonstrated differences between normal persons and those with UCD who are otherwise biochemically similar, and it would therefore seem rational to expect such measurements to demonstrate differences between the two population groups. Urinary particulate activity was measured in the urin of 100 normal Whites and 100 normal Blacks, the two groups being matched for age, height and weight, and monitored under normal dietary hydrational and environmental conditions. The three parameters of particulate nucleation, growth and aggregation were measured and the two groups compared.

Particulate nucleation demonstrated the most significant contrast between the two groups with the production of new particles through nucleation being far greater in the White group than that which occurred in the Black group (p<0.001). Particulate growth occurred at similar rates in the two groups although at slightly higher rates in the White group. Particulate aggregation occurred at a greater rate in the White group but the difference between the two groups was not statistically significant. The differences between the two groups are shown to occur as a consequence of differing rates of particulate nucleation although the rates of particulate growth and aggregation are parallel. Whilst the factors responsible for the low nucleation rate in Black person remain unknown their effect can now be measured quantitatively through the parameters of urinary particulate activity. Blood levels of ketone bodies appear to determine skeletal muscle amino acid release; high levels conserve protein and attenuate gluconeogenesis. Starvation indt/Ced ketosis is suppressed by infection [1] .

To determine if the relative hypoketonaemia following sepsi s in turn contributes to increased glucogenesis, arterial substrates and glucose production (constant infusion 6-3H(N)-glucose) were measured before and after infusion of Na-DL-./3-hydroxybutyrate (/3OH) to raise levels three-to fourfold in fed (n= 11), in fasted (n = 8) and in fasted-infected (n= 9) animals.

In fasted-infected animals before infusion ketosis did not occur (/3OH 0.81±0.1 mm/1 fasted; 0.45 ±0.55 fasted-infected) and basal glucose turnover was increased (5.05±0.18 /tm/kg/min fasted; 9.5±0.83 fastedinfected). With infusion of glucose and alanine concentrations decreased as expected in fed and fasted animals but not in fasted-infected (glucose 2.33±0.17 ram/1 befor; 2.44+0.11 mm/1 after). Glucose production also fell significantly in the fed (10.11±1.33 /~m/kg/min before; 8.44±1.05 after) and fasted (5.05±0.28 v. 4.11±0.33) groups but was unaffected by infusion in the fasted-infected group (9.5+-1.83 v. 9.11+1.44).

The accelerated rate of gluconeogenesis in infection is thus not a consequence of hypoketonaemia. The usual reciprocal relationship between glucose and ketone utilisation during feeding and fasting has not been demonstrated~in sepsis. Preliminary experiments in a hindlimb model support the hypothesis 201 that during infection amino acid release from muscle is not affected by ketone levels. We have developed a technique for measuring the total body carbon of the living subject which is suitable for measuring the critically-ill as well as the ambulatory patient. By combining this measurement with that of total body nitrogen and calcium [1] an estimate of total body fat is derived. Measurement at the beginning and end of a given period enables the changes in total body protein and fat to be obtained, as well as the patient's energy expenditure if energy intake is also known. The method is a radiation technique. The supine patient is irradiated laterally with a horizontal beam of fast neutrons and the resulting gamma rays from the body are detected by a radiation detector placed unterneath the subject. The nuclear reaction employed is the inelastic scattering of fast neutrons by the carbon nuclei of the body with the emitted gamma rays having an energy of 4.43MeVI In the initial application, measures of total body fat obtained using the technique were compared with those derived from skinfold thicknesses in six volunteers: there was no significant differences between the two measurements, (see table) . The method is being employed in studying the changes in total body protein and fat, and the energy requirements of surgical patients receiving nutrition. In order to investigate the mechanism of this effect, normal monocytes were incubated at 37°C for 30 min (with intralipid 40/A/ml) and their function assessed by three different techniques (chemotaxis, phagocytosis and chemiluminescence). All three methods showed impairment of function following exposure to intralipid. In order to try and prevent this potentially damaging effect, heparin was added to the various in vitro tests and found to cause marked impairment of phagocytosis. (P<0.01) To assess its effect in vivo, 11 volunteers were given 5,000 units of subcutaneous heparin 2 h prior to intravenous intralipid (as above). Although the use of heparin did not affect either immunological function, it completely prevented the fall of monocyte chemotaxis following intralipid alone. These findings suggest that monocyte function may be impaired by the presence ofintracellular lipid particles. The use of S.C. heparin may help to alleviate this problem and could, therefore, be beneficial to ill and often septic patients requiring intravenous nutrition. To investigate the effect of elevated glucocorticoids of stress and trauma on peripheral glutamine metabolism, 0.44 mg/kg BW Dexamethasone was injected daily intramuscularely in adult mongroel dogs over a period of 2 weeks. At least 3 weeks prior to the experiments catheters were placed into the animal's abdominal aorta (2) and caval vein (1) in order to measure A-V differences and hindquarter blood flow. During Dexamethasone treatment nitrogen balances were negative, -7.1-4-1 g N per day, whereas slightly positive N-balances were observed during the control period (3.1 +__ 1.7 g N/day). Muscle glutarnine concentrations declined constantly from 22.1 +2.3 mmol/1 intracellular water to 10.0-4-1.2 by 56% within two weeks. Whole blood arterial and venous plasma concentrations remained constant. To test the hypothesis of increased peripheral glutamine utilisation or decreased glutamine formation, the activities of Glutaminase and Glutamine Synthetase were measured in a muscle homogenate obtained before and 8 days after Dexamethasone treatment. Both enzyme activities were found to be unchanged. Hindquarter glutamine efflux increased from 20.3+22.6 pmol/min in the control state to 80.0+24.1 during Dexamethasone treatment indicating a 4 fold muscle glutamine output. This increased glutamine output was enirely due to increased A-V differences and despite decreased hindquaarter blood flow during Dexamethasone. It is concluded that Dexamethasone reproduces the metabolic response of trauma and sepsis in terms of negative nitrogen balance and muscle glutamine depletion. Muscle glutamine is shifted from peripheral tissues to visceral organs with muscle compensating for visceral demands rather than skeletal muscle being the primary target of corticoid action. It has been suggested that there is abnormal glucose utilisation in malnourished patients and that this may explain the adverse clinical sequelae of high rates of glucose infusion during intravenous feeding. We have investigated the hypothesis that there is a depression of the key enzymes of glucose oxidation in the muscle of malnourished patients which is due to an alteration of muscle fibre type proportions. 14 malnourished patients (P) (9M,5F 56+12 yrs) Our results demonstrate that there is a positive correlation between preoperative CP and stage of cancer 0' =34.4+-3.88x; r=0.62;/9<0.001). Nevertheless before surgery there is no difference between CP values in the two groups considered (g.c.= 46.2___9.7 mg %;p.u. =44.4----9.0 mg%), but after surgical trauma CP presents a positive response in patients with p.u. (mean increase +14.9 %), whereas it acts as a negative AP protein in g.c. patients (mean decrease 5-60/0) (P<0.02). Moreover malnourished g.c. patients present a reduction of CP values ( -10 %) which is greater than g.c. patients with albumin >3.5 g% ( -1.5%); this difference is not statistically significant. Cancer patients undergoing palliative (n=10) or radical surgical procedure (n=31) show parallel decrease of preoperative CP ( -6%), the first group presenting higher preoperative values in relation to the tumor diffusion.

In conclusion our results demonstrate that CP is not only a positive AP protein, but in some circumstances it may act as a negative PA protein depending on the underlying disease and the preoperative nutritional status. In this study the free AA concentration in liver tissue of 4 non septic patients (cholecystectomy) were compared with those of 7 septic patients (abdominal sepsis). The liver specimens were taken intraoperatively..The nature and possible risk involved in this study were explained to the patients and their consent obtained. The data presented in this abstract are part of a metabolic screening program of septic patients including the determination of AA (plasma, muscle), hormones (insulin, glucagon, cortisol), nutritional parameters (prealbumin, retinol-binding protein, transferrin), and of energy metabolism (ATP, ADP, glucose, free fatty acids).

For the determination of the free AA the intra-and extracellular water content (chlorid method) and of fat content of the liver specimen were analysed. A membrane potential o f -45mV was assumed. The AA analysis were performed with an automatic AA analyser (Kontron, Svitzerland) by means of an ionexchange resin (Durrum DC-4) and a lithium buffer system (Durrum-Pico buffers). Conclusions: 1. This study reveals decreased concentrations of nearly all AA in liver tissue of septic patients (exception: Phenylalanine, Tyrosine, Cystathionine). 2. The significantly decreased concentrations of the gluconeogenetic AA (Thr, Ser, Ala) indicate that the gluconeogenetic capacity of the liver is not exhausted through an increased uptake of those AA as shown earlier by Wilmore et al. [4] 3. An increased administration of gluconeogenetic and basic AA (Lys, His) may normalise the AA pattern in the liver of septic patients. The liver is being increasingly recognized as a critical organ in postoperative multiple organ failure. The principle factors precipitating postoperative multiple organ failure were sepsis, hypotension and injury to the liver. Previous studies from our laboratory have shown that hepatic failure, which has a high mortality rate, is linked to the marked decrease in energy charge. In order to evaluate the possible presence of metabolic blocks, the changes in the ratio of acetoacetate to fl-hydroxybutyrate (ketone body ratio), which reflects the hepatic mitochondrial redox potential, were analyzed in relation to energy charge in hepatectomized, jaundiced, hemorrhagic-shokked and septic animals, as well as patients with postoperative multiple organ failure. Experimental: 1. In hepatectomized rabbits, mitochondrial phosphorylative activity increased to 170 % of the control and the energy charge level decreased from the normal level of 0.87 to 0.76 at 24 h after hepatectomy (/7<0.001). Afterward, these values returned to preoperative levels within a week. The ketone body ratio in arterial blood was positively correlated with hepatic energy charge (r=0.696, p~0.01).

2. In jaundiced rabbits, the hepatic energy charge decreased rapidly after the bile duct ligation along with the decrease of mitochondrial pbospborylative activity. The hepatic energy charge fell from 0.87 to 0.73 at 48 h postoperatively with a maximum incidence of mortality. Moreover, changes in the blood ketone body ratio were positively correlated with the hepatic energy charge (r= 0.844,/~0.01). The decrease in the blood ketone body ratio was attributed to the restricted mitochondrial reoxidation of NADH due to an inhibition of oxidative phosphorylation. 3. In hemorrhagic-shocked rabbit with a mean arterial blood pressure of 40mmHg, the changes in the blood ketone body ratio were correlated with hepatic energy charge (r=0.772, p<0.01). 4. In septic pigs subjected to the ligation and perforation of the cecum, the hepatic energy charge level decreased gradually from 0.86 to 0.80 and the mitochondrial phosphorylative activity was enhanced to 150% of controls in the hyperdynamic state. In the hypodynamic state, the hepatic energy charge level fell drastically 0.69 concomitant with the decrease in mitochondrial phosphorylative activity and blood ketone body ratio. From these results, the blood ketone body ratio may be regarded as a reliable indicator for assessing the degree of decreased energy charge. Clinical: Changes in the blood ketone body ratio were measured in 55 patients who underwent major surgery such as hepatectomy. These patients were classified into 4 groups according to the postoperative changes in blood ketone body ratio: Group A without decrease to below 0.7, Group B with transient decrease to 0.4, Group C with progressive decrease to below 0.4 and Group D with terminal decrease to below 0.25. All 35 Group A and B patients tolerated the operation well. By contrast, the 20 Group C patients showed multiple organ failure with 85 % mortality rate, which involved pulmonary failure (70%), hepatic failure (70%), gastrointestinal bleeding (30%), renal failure (600/0), cerebral failure (70%) and coagulopathy (65%). All patients who transitioned to the terminal stage of Group D died of cardiogenic decompensation. In 5 patients of Group C, the decreased blood ketone body ratio was restored with the amelioration of clinical symptoms after ex vivo pig or baboon liver crosshemodialysis and 3 patients of them were later discharged. Evidence presented indicates that the decreased blood ketone body ratio has a direct bearing on multiple organ failure.

Conclusion: Sepsis, hypotension or injury to the liver are a metabolic burden to the liver mitochondria which can result in mitochondrial impairment leading to a marked decrease in hepatic energy charge. Such impairment ultimately leads to multiple organ failure as a result of the critically decreased energy and substrate store and the reduced protein synthesis relative to demand in the various organs.

110

In interferon-treated cells, the 2'-5'A synthetase, activated by double stranded RNA, polymerizes ATP into a series of oligonucleotides characterized by 2'-5' phosphodiester bonds and collectively designated 2'-5'A. These activate an endoribonuclease which cleaves RNA. Other regulatory functions of this enzyme may be expected because of its wide occurence in mammalian cells (untreated with interferon), where its activity appears, in vitro, to be dependent on the growth conditions, hormone responses Regenerating liver after partial hepatectomy is often used as a model for the study of control growth and cell proliferation in vivo. In order to evaluate the role of the 2'-5'A synthetase in the processes leading to initiation of cell division, we measured this enzymatic activity in the rat liver during the first 40 h after partial hepatectomy.

Partial hepatectomy (50 rats) was performed under neuroleptanalgesia by removing the median and the left lateral lobes of the liver according to the method of Higgins and Anderson. Control animals (3 rats) were subjected to a sham operation.

After selected time intervals, the animals were sacrificed and the enzymatic activity in the regenerated liver was measured. The 2'-5'A synthetase activity present in the two first removed lobes was defined as 100%.

We observe a very rapid decrease of enzymatic activity which reaches 50 % already 10 h after partial hepatectomy. The lower level of enzymatic activity (25 %) is measured between 20 and 24 h after partial hepatectomy. This minimum is followed by a slow restoration of the activity (at 40 h: 60 %). During this early phase of liver regeneration, a maximal incorporation of tritiated thymidine in DNA takes place 24 h after surgery.

So well differentiated liver cells have elevated levels of2'-5'A synthetase. But, after partial hepatec-tomy, the 2'-5'A synthetase activity decreases dramatically before the first wave of cell mitosis.

These observations clearly illustrate the relationship between 2'-5'A synthetase activity and the growth status. Moreover, this drop of enzymatic activity may be a trigger for the initiation of cell division. The primary event or events setting in motion the process of liver regeneration after partial hepatectomy (PH) remain unsettled. Regarding the so called hepatotrophic factors, i.e. insulin, glucagon and recently EGF, present evidence suggests that they would play mainly a promoting rather than a initiating role. Early changes such as glycogen breakdown, fat infiltration and changes in adenine nucleotides and mitochondrial phosphorylative activity are usually, at least the first two, ascribed to metabolic overload of the remaining liver (Bucher et al (1969) Johns Hopkins Med, J, 125: 250). So far, however, little attention has been paid to a possible involvement of this phenomenon in the initiation of liver regeneration.

Attempts were therefore made to modify metabolic overload through early changes in energy metabolism in order to study their influence on the pattern of DNA synthesis.

Fed or 16 h fffsting male Wistar rats weighing 200+20 g were examined after PH at 4, 12, 18, 24, 30 and 36 h for adenine nucleotides, oxidative phosphorylation and DNA synthesis, based on the rate of~H thymidine incorporation. Fasting animals received continuous infusion of 20 % dextrose for 24 h at the rate of 0.45 ml/100 g/h. In addition to the above mentionned parameters hepatic glycogen and fatty acids were measured.

Within 12 h partial hepatectomy caused a decrease in hepatic ATP which was maximal at 3 h (from 2.99___0.13 to 2.36+0.07 /~ moles/g p<0.001); in energy charge (ATP + 0.5 ADP/ATP + ADP + AMP) from 0.851-+0.01 to 0.816___0.03 (p<0.01) and increase in phosphorylation potential which was maximal at 3 h (from 76+3 to 106_2p<0.001). DNA synthesis began at 18 h reaching a peak by 24 h. Glucose infusion to PH rats suppressed the decrease of hepatic ATP, 2.97___ 0.07/1 mole/g at 3 h vs 2.99_ 0.13 in the control group, prevented glycogen depletion (histochemical estimation) and the increase in fatty acids (two folds increase in FFA and Triglycerides (TG) at 24 h vs 10 folds in FFA and 8 folds in TG in the control group),with little effect on mitochondrial activity. The initiation of DNA synthesis was delayed and the whole pattern was considerably modified. Cessation of glucose infusion restored the usual rate of 3H thymidine incorporation after a late fall of hepatic ATP.

In conclusion, glucose infusion was shown by one of us to modify the hormonal response to PH, but insulin and glucagon administration to glucose treated animals failed to normalize the pattern of DNA synthesis. It is suggested that metabolic overload as estimated on the basis of early changes in energy matabolism may account for one of the events involved in the initiation of DNA synthesis after PH.

Infusion on Liver Cell Regeneration after Partial Hepatectomy in the Rat B. de Hemptinne, J.F. Ngala and L. Lambotte University of Louvain, Laboratory of Experimental Surgery UCL 5570, 1200 Brussels, Belgium

After partial hepatectomy (PH) the portal and the peripheral blood serum concentration of immunoreactive insulin suffers a drastic fall and levels ofglucagon show a rapid increase which is maximal 4 h after the liver resection.

As these changes appear closely correlated to the blood glucose levels which show a 30 % decrease at 4 h and progressive restoration towards normal values up to 24 h, attempts have been made to alter the insulin/glucagon ratio by glucose infusion after PH and study its relation to liver regeneration. The purpose of this work is thus to determine after PH and hypertonic (20%) glucose infusion:

1. The effect of glucose on insulin and glucagon blood levels over 24 h; 2. the repercussion of the insulin/glucagon modified ratio on DNA synthesis;

3. the possible improvement of DNA synthesis by extensive glucagon infusion.

Male Fisher rats underwent a standard 70 % PH. A 20 % glucose solution or isotonic salin was infused at a constant rate of 0.9 ml/h through a cannula placed in the iliac vein. The rats were sacrified at 0, 4, 8, 12, 22 and 24 h. (3H) thymidyne (27..3/~ci/mmol) was injected 2 h prior to sacrifice and the liver caudate lobes removed for analysis of (3H) thymidine uptake into nuclear DNA. Blood samples were withdrawn from the portal vein, the inferior vena cava and the aorta for glucose, insulin and glucagon assays.

Compared to the salin treated group, the infusion of glucose while keeping a normal steady blood glycemia was responsible of a marked increase of insulin (0.38--+0.01 ng vs 2.24_+0.4 ng,/7<0.01) and decrease of glucagon (0.615-+0.085 ng vs 0.335+0.037 ng, p<0.01), with a major switch of the insulin/glucagon ratio at 4 h after PH (from 6.68 to 0.61).

DNA synthesis started at 22 h in both series, but was very significantly impaired at 24 h in the glucose infused group (12800-+2280 cpm/mg DNA vs 27450+2050 cpm/mg DNA, p<0.01).

Infusion of increasing doses of glucagon (from 0.01 to 2 mg/kg/day could not restore the impaired DNA synthesis. Only a slight improvement was recorded at 0.5 mg/kg/day as the insulin/glucagon ratio tended to approach that of the control group. Fractionation of various doses of glucagon over the 24 h perfusion time, in such a way that the changes in concentration of glucagon after partial hepatectomy was imitated, remained unsuccessful to improve thymidine incorporation.

In conclusion: 1. The infusion of hypertonic glucose which impairs DNA synthesis after PH, modifies markedly the insulin and glucagon secretion. 2. If a specific insulin/glucagon ratio after PH is important to sustain normal regeneration, its modification does not seem to be the major factor contributing to the blunted DNA synthesis response in the hypertonic glucose infusion model. Operative mortality of emergency shunt operations or esophageal transection during acute hemorrhage from ruptured esophageal varices in cirrhotic patients (Child's category C) has been intolerably high. Hence, the emergency operations in these patients should be avoided when the bleeding could be stopped by non-operative measures. When the emergency operation eventually becomes inevitable, the operative procedures should be simple and of short duration.

In 1976, we have introduced the endoscopic balloon tamponade method for the management of esophageal variceal bleeding. The new balloon tube used in this method has essentially the similar structure as the Sengstaken-Blakemore tube, but is made of translucent plastic materials, and has a larger internal diameter so that a small caliber optic fiberscope (ex. bronchofiberscope) can be passed through the tube. By this method, the endoscopic observation of the esophagus is possible through the translucent balloon tube during tamponade of the ruptured varices. It is possible to know directly whether the bleeding from varices has been successfully stopped or not, which seems to be an advantage over the blind tamponade method using the Sengstaken-Blakemore tube.

This endoscopic tamponade method has been used for emergency hemostasis of 166 acute bleeding from ruptured esophageal varices in 66 patients. The mean initial tamponade pressure by the esophageal balloon necessary to stop bleeding was 30_+12 mrnHg and the average duration of tamponade was 50 h. The location of the ruptured v~ices observed by this method was in the lower esophagus within 10 cm of the esophagocardiac junction in 86% of the cases. The bleeding has been stopped in 156 occasions (93.9 %) by this method. No significant complications other than atelectasis in 2 patients have been observed.

In 4 patients, the bleeding was initially stopped by the new translucent balloon tube, but recurred within 24 h after decompression of the esophageal balloon. Tamponade was repeated for several times since these patients were in the category C of the Child's classification, however, the emergency operation eventually became necessary. Transthoracic esophageal transection was performed using autosu-ture apparatus, EEA, in these patients. One patient died of liver failure in the 6th postoperative day, but others survived the operation and recovered. The use of EEA in transthoracic esophageal transection simplifies the esophageal anastomosis, and shortens the duration of operation by 40 rain.

We have so far used the autosuture apparatus, EEA (28 mm cartridge) in elective esophageal transection of 33 patients. Neither anastomotic bleeding or insufficiency has been encountered.

Acute variceal bleeding in category C patients should be treated initially by endoscopic balloon tamponade, when emergency operation is inevitable, transthoracic esophageal transection using EEA is the operation of choice. Arterialisation of the portal vein in conjunction with an end-to-side portacaval shunt has been proposed as a method of improving survival following shunting [1] . However, there is little experimental evidence to support this suggestion and so we have examined the effects of arterialisation of the portal stump in cirrhotic rats.

24 rats with dimenthylnitrosamine-induced cirrhosis were used in this study. 8 of the rats received an end-to-side portacaval shunt, 8 had a portacaval shunt and the portal stump arterialised with the left gastric artery, eight were sham-operated. Liver blood flow (LBF) and wedged hepatic venous pressure (WHVP) were measured before and after shunting. Three weeks after surgery LBF and WHVP measurements were repeated, the animals bled and the liver removed and weighed.

An end-to-side portacaval shunt led to an immediate fall in LBF (31.3_+5.6 to 15.5_+4.3 mls/min/ 100g-~) and WHVP (7.6_+0.8 to 3.1-+0.5 mmHg). However, if the portal stump was arterialised LBF (34.6-+2.9 to 45.3+5.7) and WHVP (7.9_+1.0 to 7.3_+1.3) did not change significantly. No further changes in LBF and WHVP were observed three weeks after operation in any group of animals. Arterialisation of the portal stump prevented the loss in body weight, loss in liver weight deterioration of liver function tests and hyperammonemia observed in animals with a portacaval shunt alone.

These findings suggest that arterialisation of the portal stump may prevent some of the deleterious effects after shunting. Departments of Surgery and Pathology, University of Virginia Hospital, Charlottesville, Virginia 22908, USA

We have hadexperience with operative restoration ofhepatopedal portal blood flow in five patients intolerant of total splanchnic shunting. Hepatopedal flow was reestablished by takedown of the total shunt and construction of a selective, distal splenorenal shunt, or by isolation and arterialization of the hepatic limb of the shunted portal vein. In two patients, shunt revision was undertaken electively for chronic encephalopathy, unresponsive to low protein diet, intestinal antibiosis and oral lactulose. Each individual had been hospitalized more than eight times for encephalopathy or coma. Nine and 34 months postoperatively, both patients have had no encephalopathy on unrestricted protein intake, and work. actively as homemakers. Serial liver needle biopsies have shown bilobed nuclei and enhanced mitotic activity suggesting hepatocyte regeneration.

In three patients, shunt conversion or arterialization was undertaken in desperate circumstances, characterized by liver failure (bilirubin >10 mg/dl, albumin <2.5 girl, prothrombin time >1"6 s)~ coma and respirator dependency. Although two patients showed immediate, marked improvement in mentation, all three died of intraabdominal hemorrhage in the first few postoperative days in spite of prolonged attempts to achieve hemostasis and maximum blood product support. Three conclusions can be drawn from this limited experience: 1. Total shunt procedures which are anatomically suitable for subsequent conversion to a selective configuration or for hepatic limb arterialization should be favored over those not offering such potential; 2. at a time of election, restoration of hepatopedal portal flow can be accomplished in patients with side-to-side portacaval or hemodynamically equivalent shunts with considerable benefit; and 3. similar procedures in patients with fulminant liver failure are unlikely to succeed. Peritoneal-venous (LeVeen) shunts are associated with a significant incidence of disseminated intravascular coagulation (D.I.C.). This study identifies a site of origin, a pathogenic mechanism, and the hemostatic pathway which accounts for the thrombogenicity of human ascites. 500 ml of peritoneal fluid were removed percutaneously from individuals with malignant and cirrhotic ascites. Ficoll-Hypaque column chromatography and ultracentrifugation were utilized to prepare four fractions: cellular; a low speed cell-free fluid; a high speed supernatant; and the precipitate from the high speed centrifugation. The cellular fraction from both types demonstrated an ability to shorten a one stage clotting time by 60 % relative to saline and endotoxin controls. Similarly, low speed cell-free fluid shortened the clotting time of pooled normal plasma by 61%; was also effective in factor VIII (required for intrinsic pathway of coagulation) deficient plasma; but had no effect on factor VII (required for extrinsic pathway of coagulation) deficient plasma or platelet aggregation and release. The high speed supernatant was demonstrably less thrombogenic. The resuspended precipitate shortened the clotting time of pooled normal plasman by 67 % and of factor VIII deficient plasma from infinity to 99 s. In contrast, this material was ineffective on factor VII deficient plasma. We conclude that the thrombotic potential of human ascites derives from peritoneal cells, either leucocytes or malignant cells. Consistently, the thrombotic factor exists in suspension and is thromboplastin-like in its behavior, operating through the extrinsic pathway of coagulation. Thus, a site of origin and a pathway of activity for the thrombotic agent in human ascites is identified.

The Effect of Somatostatin in Hepatic Haemodynamics in the Cirrhotic Rat S. Jenkins, P. Devitt and R. Shields Department of Surgery, University of Liverpool, Liverpool, U.K.

Although somatostatin has been suggested as an alternative treatment to vasopressin in the emergency control ofbleeding oesophageal varices [1] , recent studies on its effect on wedged hepatic venous pressure in cirrhotic patients have provided conflicting results [2, 3] . Therefore we have examined the effect of somatostatin on hepatic heamodynamics in cirrhotic rats. Rats with dimethylnitrosamine-induced cirrhosis received a bolus injection of 2, 4 or 8 pg/kg body weight of somatostatin followed by a 30 min infusion of either 2, 4 or 8 pg/h/kg body weight. Control rats received saline only. Portal venous flow (PVP) portal pressure (PP), liver blood flow (LBF) and wedged hepatic venous pressure (WHVP) were measured before, during and after somatostatin infusion.

At the lowest rate of somatostatin administration (bolus injection of 2 pg/kg body weight followed by an infusion of 2 pg/h/kg body weight)there was a rapid decrease in PP (9.8+0.5 to 8.0--+0.4 mmHg), WHVP (7.5___0.4 to 5.8---0.3 mmHg) and PVF (31.2 ___ 4.9 to 21.0___ 3.8 ml/min). 30 rain after the start of the infusion PP, WHVP and PVF were still signaflcantly lower than preinfusion levels. At higher rates of somatostatin infusion there was no furhter decrease in PP, WHVP and PVF. LBF was significantly reduced at all the doses of somatostatin. The highest rate of infusion of somatostatin (8 pg) produced a significantly greater reduction in LBF than either 2 or 4 pg.

These data suggest that somatostatin may have a role in the management of portal hypertension, but that higher doses may have a detrimental effect on LBF. ER positive breast cancers preferentially metastasize to bone (1) Prostaglandin E2 (PgEz) is synthesized by breast cancers and potentiates bone resorption in vitro (2) . This study investigates the relationship between ER status and PgE2 synthesis in breast cancer cells. PgE2 was measured by radioimmunoassay in 75 primary breast cancers: a) After ethanol inhibition of further prostaglandin (Pg) production -,,Basal Pg" b) After stimulating Pg production with excess Arachidonic Acid -,,Total Pg". ,,Total" minus ,,Basal" = ,,Synthesized Pg". In order to relate Pg production to breast cancer cells, measured PgE2 values have been corrected for the epithelial cellularity of each tumour. PgE2 x 100 Corrected PgE2 = Actual (%) Cellularity Cellularity was evaluated by a proportional count of cancer cells expressed as a percentage against non cellular material in all fields of 3-5 histological sections at 63x magnification. We conclude that ER positive tumour cells synthesize greater amonts PgE2 than ER negative cells. This may account for the greater tendency of ER positive cancers to recur in bone. Oestrogen receptor activity (ER) is of prognostic value in breast cancer [1] . However, the chosen cutoff between ER-negative and -positive is arbitrary and likely to affect its prognostic value.

In 201 patients with invasive breast cancer treated by mastectomy, tumour ER was determined [2] and the patients were followed up until first recurrence. Using a computer program to perform repeated logrank analysis, the effect of varying the cut-off on prognostic value of ERwas studied between 1 and 25 fmol/mg protein. ER levels were higher in postmenopausal patients (0-1301, median 72, n=121) than in pre-menopausal patients (0-201, median 25, n=80). For the whole group, optimal prognostic discrimination was achieved with a cut-off 6 fmol/mg protein. In premenopausal patients, the effect of varying cut-off was complex, leading to an optimum of 17 fmol/mg protein, whilst in post-menopausal patients the optimum was 3 fmol/mg protein. These findings were unexpected and may relate to the relationship of ER to tumour cellularity [3] .

It is concluded that: 1. The failure of some centres to relate ER to prognosis may be due to the inappropriate cut-off point chosen; 2. in our patients, the optimal cut-off was 6 fmol, which agrees with the level suggested by the British Breast Group [4], though it differed between pre-and post-menopausal patients; 3. optimal cut-off for prognosis may differ from that for predicting response to endocrine therapy. Previous studies have reported that the use of adjuvant chemotherapy improves relapse free survival following mastectomy. The effect on overall survival is less certain. 290 patients with histologically confirmed axillary node metastases were randomised to receive postoperative radiotherapy (RT), chemoterapy (CMF) or radiotherapy plus chemotherapy (RT + CMF). 244 patients have now been followed for between 6 months and 5 years. Patients initially treated with RT received chemotherapy on failure where possible.

Of 79 patients receiving RT alone, 28 have developed distant recurrence and 20 have died. Of 78 receiving CMF alone 22 have developed distant  recurrence and 15 have died (see table) .

Although the use of adjuvant chemotherapy significantly prolongs the relaps free interval there is no corresponding improvement in overall survival. A sample of 89 patients with histologically proven prostatic carcinoma underwent transrectal find needle aspiration at six month intervals, for a mean follow-up of 18 months, as an out-patient procedure, without significant side effects. 49 patients were followed from time of initial diagnosis. The others had been on treatment for a mean period of 36 months before the study commenced.

In 25 of the new patients, cytology was positive at the time of histological diagnosis, and correlated with the histological grade. There were no false positives. Repeat cytology on 15 patients after 12 months showed 6 positive and in 5 of these the disease was progressing, and in one it was stable. Of the 9 who were negative, 8 were stable and one progressing.

16 of the 48 patients already on therapy had positive cytology at their initial aspiration -9 showing progressive disease and 7 being stable. 5 patients whose cytology was initially negative became positive 12 months later and all had disease progression at this time_ 19 patients had negative cytology on 3 or more occasions and in only 2 cases was there evidence of disease progression.

Cytology showed evidence of squamous metaplasia on follow-up in 13 of the patients whose condition was stable.

The prognosis for all patients was assessed on the basis of Gleeson's score and aspiration biopsy has been shown to be a safe and useful procedure for monitoring disease activity and response to treatment. The optimum method of restoring the ability to swallow in patients with unresectable carcinoma of the oesophagus remains controversial. This study evaluates the palliative potential of pulsion intubation versus retrostemal gastric bypass of the excluded oesophagus in unresectable carcinoma of the upper thoracic segment (20-32 cm from the incisor teeth). Patients and Methods: 71 patients were prospectively randomised for treatment by pulsion intubation (37 patients) or gastric bypass (34 patients). Non-resectability was indicated by (1) tumour lengths greater than 6 cm, (2) tracheal or bronchial invasion, (3) disrant dissemination, (4) mediastinal invasion. The operative mortality, morbidity, palliation of dysphagia and postoperative nutritional status was compared in the two groups. Results: Mortality: Intubation resulted in 2 deaths, a mortality rate of 5,4%. Gastric bypass resulted in 3 deaths, a mortality rate of 8,8 %. Complications: Intubation was complicated by respiratory infection (8), tube migration (1), respiratory obstruction (1), oesophageal perforation (1), bleeding (1). Gastric bypass was complicated by chest infection (9), pneumothorax (3), wound infection (5), subphrenic abscess (1), anastomotic leak (4), pulmonary embolism (1), purulent neck discharge (3). Ability to swallow: Palliation of dysphagia was achieved in 92 % of patients following intubation and 91% of patients following bypass. Postoperative nutritional status: Improvement in nutritional status was more rapid following intubation. Conclusion: Pulsion intubation is the preferred palliative procedure because of fewer complications and lesser degree of postoperative catabolism. The current technique for investigating the response of vascular prosthetic materials to infection is by challenge with a sub-lethal dose of bacteria, usually an intravenous infusion of 108 organisms in an animal model. This large bacterial innoculum, however, obscures any difference in the infectibility of prostheses that may be inherent in the material, its incorporation into host tissues, or its resistance to infection. We have developed a sensitive method for determining the susceptibility to infection of vascular prostheses based on calculation of the number of bacteria required to infect a specific prosthesis in 50% of trials (IDs0). Following implantation of the prosthesis to be tested in the canine infra-renal aorta, a dose-response curve was generated by the intravenous injection of known innocula of S. aureus (at log intervals from 102 to 108 bacteria). At six weeks the prosthesis was harvested and cultured to document infection with S. aureus. A characteristic sigmoid curve resulted from which the IDs0 was determined. To test this method, a comparison was made between commercial human umbilical vein grafts (HUVG) and HUVG impregnated with silver sulfadiazine in 35 dogs. Although both prostheses were infected by the standard 108 innoculum, a greater than tenfold increase in the resistance to infection by the treated HUVG (<102 to 101) was demonstrated in the dose-response curves. Since the number of bacteria in a postimplant bacteremia rarely exceeds 102 organisms/ml, such differences in infectibility are clinically significant. IDs0 determination provides a sensitive, reproducible method for quantitating resistance to infection in vascular protheses. Prosthetic infection following reconstructive vascular operations is an infrequent but often fatal complication which generally persists until the graft is removed. It is accepted that infection arises from operative contamination, bacteraemic seeding and abscesses or viscus eroding into the graft. This study investigates the role played by distal sepsis on groin grafts.

Ten dogs had a specific strain of staphyloccocus aureus inoculated onto a surgical wound in the right foot pad. Five days later interposition 6 mm Dacron grafts were implanted into the ipsilateral and contralateral groin in continuity with the superficial femoral artery. Ten days following this the grafts were removed for bacteriological and histological examination. Blood cultures and lymphnode cultures were also taken at this time.

In seven dogs the specific staph, aureus, was grown from both grafts. Two dogs failed to grow the specific staph, aureus from either graft. These results are significant at the 1% level using Fischer's exact test.

Blood cultures grew staph, aureus from only one dog. Ipsilateral lymphnode cultures yielded the specific staph, aureus in seven dogs.

We believe that distal sepsis plays an important role in the estabishment of graft sepsis. The mechanism of spread would appear to be via the lymphatics.

The Influence of Tumor Growth on Wound Healing P.W. de Graaf and A. Zwaveling Laboratory Experimental Surgery, State University, Leyden, The Netherlands

Leakage of a colonic anastomosis is a complication each surgeon fears, because it usually leads to a prolonged hospital stay and is accompanied by a high mortality. Carcinoma as an indication for operation and preoperative malnutrition are considered to be high risk factors. The subject of this study was to measure the influence of malignant tumor growth at a distant site on woundhealing in colon and skin, and to find a correlation between the influence of the nutritional state of the experimental animal, and the influence of malignant tumorgrowth on woundhealing. We also tried to find ways to compensate the possible unfavourable influence a malignant tumor might have on woundhealing. The study was performed in rats, the tumor used was a rhabdomyosarcoma, transplanted subcutaneously in the rats right flank. Parameters for woundhealing were tensile strength of the skin-incision, and bursting pressure of a colonic anastomosis. Six groups of rats were studied, each rat undergoing a standardized colonic operation after two or four weeks. The groups consisted of: control animals (20); tumorbearing animals (20), without tumor removal; malnutrltioned animals (20); tumorbearing animals receiving intravenous hyperalimentation (20), without tumor removal; animals undergoing tumor removal and colonic operation in one session (10); animals undergoing the colonic operation two weeks after tumor removal (10). Woundhealing was negatively influenced in an early stage of tumorgrowth (F<0.05). Four weeks of tumor growth did not impair woundhealing to a greater degree than two weeks. Laboratory determinations showed no deviations. Malnutrition, leading to less weight gain than in control and tumor bearing animals took much longer to influence woundhealing than tumor pearing. This is in accordance with publications by Irvin and Hunt (1974), Devereux et al (1979) and Garattini and Guaitani (1981) , which led us to the conclusion that the negative effect of tumor bearing on woundhealing was not solely the result of anorexia. Hyperalimentation in tumorbearing rats resulted in undisturbed woundhealing, despite the fact that hyperalimentation as practised by us (with amino-acids and carbohydrates) stimulated tumorgrowth. We concluded from these experiments that tumorgrowth caused a metabolic chaos in the animal, which in turn led to a disturbance in woundhealing. Intravenous hyperalimentation could apparently compensate this. Woundhealing in rats operated in one session was significantly more disturbed than woundhealing in rats operated in one session was significantly more disturbed than woundhealing in rats operated in two sessions (p<0.05). Irvin an Hunt (1974) have demonstrated however that extraabdominal trauma had no influence on colonic healing. This led to the conclusion that the diminished woundhealing found in rats operated in one session was an effect of the tumor. Obviously this influence is not instantly removed with tumor excision. This study offers a theoretical foundation for the clinical observation that in operations for colonic carcinoma with a substantial operative trauma, the anastomosis needs extra protection and/or needs to be performed in a second session. The aorta of the Blotchy mouse undergoes dilatation during adult life, resulting in the formation of fusiform aortic aneurysms. The mutation is X-linked, so only the males are affected. We have found a shift in the fluorescent spectrum of insoluble, hydrolyzed, dermal collagen in these mice; suggesting the presence of an abnormal crosslinkage compound. The purpose of this report is to describe additional studies carried out on the soluble collagen fraction in these animals, which lend support to the hypothesis of a crosslinkage abnormality.

Dermal collagen was prepared from one-monthold mutant mice and their normal male littermates by sequential extractions in salt solution, acetic acid, and urea. Amino acid analyis and SDS gel electrophoreses of the salt-soluble fractions revealed similar profiles of amino acids and collagen-chain types, suggesting that there is no abnormality of the basic building blocks of collagen in the mutants. Fluorescent spectroscopy of the acid hydrolysates of these fractions also demonstrated similar absorption and emission peaks. However, reduction with sodium porohydride abolished fluorescence in the collagen fraction from the mutants. Results (2emission m a x ± SEM):

Normal Blotchy Before reduction 424.3 ± 3.0 420.7 -----2.9 After reduction 420.0 + 2.1 None Sodium borohydride is used experimentally to stabilize labile collagen cross-linkages. The present studies indicate that the salt-soluble collagen from the Blotchy mice is borohydride-reactive and is thus chemically ,,unstable". These studies suggest a rational for developing probes based on these fluorescent croperties of investigate collagen from human subjects affected with aneurysms. One third of testes are removed after torsion (1, 2) because of delay in presentation and diagnosis. In a retrospective study of 89 patients with torsion our salvage rate was 64%. Nineteen percent were misdiagnosed as epididymitis by a junior doctor and 7 % by a surgical registrar. Doctors could not distinguish between torsion of the testis and its appendages. In the literature, eight factors have been reported to be of discriminant value (Table 1) .

A computer program, using these parameters and Bayes' theorem, was written to calculate the most likely diagnosis. Data from the 89 patients with torsion and 96 patients with acute epididymitis were analysed by this program ( Table 2) .

The program correctly diagnosed all 96 cases of epididymitis. Of 75 patients with torsion of the testis . In this study we determined the critical ischemic time for the development of an arrhythmogenic potential. EGG and bipolar electrograms were recorded from the His bundle, ventricular endocardium and epicardium in twelve anesthetized dogs. Short bursts ofventricular pacing (3 beats; 240-420 beats/min) were used to induce ventricular arrhythmias before and after lidocaine administration (2, 4, 6 mg/kg). Previously, lidocaine has been shown to exacerbate conduction delay in ischemic myocardium leading to reentrant ventricular arrhythmias. In the control state, ventricular pacing with or without lidocaine induced either no response or single or repetitive ventricular responses. Sustained ventricular tachycardia was never evoked in the control studies. In six dogs, after 20 min of left anterior descending coronary artery ligation and reperfusion, ventricular pacing with and without lidocaine produced sustained ventricular tachycardia in one animal. In another six dogs, after 30 rain ofligation and reperfusion, one dog showed sustained vertricular tachycardia in response to ventricular pacing alone. The other 5 showed sustaine~l vertricular tachycardia after ventricular pacing with lidocaine. Sustained ventricular tachycardia averaged 400/min and degenerated into ventricular fibrillation within 1-2 rain. In conclusion 20-30 min of ischemia appears to be the critical period for development of malignant arrhythmogenic potential in the canine heart. Administration of lidocaine during this critical period enhances the inducibility of cardiac arrhythmias, which may have clinical relevance in the management of acute myocardial infarction. Untreated coronary artery air embolism in the experiments without ECC (group I) resulted in a transmural myocardial ischemia with a mortality rate of 28 % In contrast there were no death in the experiment when extracorporeal circulation was used (group II to V). The best therapeutic effect with regard to the reversibility of temporary myocard ischemia following coronary artery air embolism was achieved by increasc of the postembolic perfusion pressure (group V), a finding being significantly different (p<0.05) to groups I through IV. Also volume depletion of the left ventricle on ECC (group II) resulted in a faster regression of the left ventricular hypothermic area compared to group I (p<:0.05). The application of Dipyridamole (group III) and St.Thomas cardioplegic solution (group IV) was not able to produce a significant therapeutic effect.

The increase of perfusion pressure following coronary artery air embolism has demonstrated to be the most effective procedure to achieve reversibility of myocardial ischemia. To transfer these experimental findings into clinical application is suggested because of its practicability and convenience. Transmural biopsies were obtained before and 5, 15, 30, 60, 90, 120 min after ACC for biochemical and structural analysis. Myocardial temperature (T) ph (MpH), and pCO2) were measured continously with a thermistor, a new pH electrode, and a mass spectrometer respectively. In Group I (n=8) ACC was under normothermia. In Group II mean T was reduced to 19°C by the administration of cold potassium cardioplegia immediately after ACC and consecutively every 30 min. MpH at the end of ACC reached 5.39-t-0.08 Group I) and 6.49___0.13 (Group II) (/7<0.001); PmCO2 rose from 41___4 to 234_+13 mmHg in Group I and did not change in Group II. Tissue content of ATP decreases by 51% Group I) and by 14°/0 (Group II) (F<0.001); tissue creatine phosphate fell by 98 % (Group I) and by 48 % (Group II) (p<0.001). Changes in ATP and creatine phosphate as well as in tissue glucose and glycogen related ton concomitant changes in pH. The degree of ischemic damage assessed histologically by a mean ischemic score was 2.5+--0.06 in Group I and 0.75_+ 0.19 in Grup II (p<0.01). Irreversible structural demage assessed by electron microscopy occurred in Group I 60-90 min after ACC and was associated with a MpH below 6.2. No such damage was observed in Group II. We conclude: 1. irreversible damage during normothermic arrest occurs considerably later than has been reported for regional ischemia in the beating heart; 2. during 2 h of ACC cold potassium cardioplegia does not completely protect against ID when mean T is 19°C; and 3. on-line oaeasurement of MpH reflects the inadequacy of' MP more accurately than do either PmCO2 or T and thus, provides a useful potential method for intraoperative monitoring of MP. In rabbits, infusion cardioplegia was installed at 37, 27, and 17°C respectively. The infusate contained Na 27, K 5, 30, 120, 160 or 240, Ca 0, 1 or 5 mmol/1, and varying amounts of glucose. Measured osmolality varied from 305 to 500 mosmol/kg. The hearts were excised at the end of a 5 rain infusion period; one half was immediately frozen, the other half was incubated for varying periods of time. Specimens of left ventricular myocardium were analyzed for metabolite and electrolyte contents.

At termination of the cardiplegic perfusion, total adenine nucleotides (TAN) and total creatine (TCr) were within control ranges under all conditions. However, ATP and ECP = (ATP + 0.5 ADP)/TAN as well as creatine phosphate (CrP) and the ratio CrP/TCr decreased significantly with increasing dosages of K and Ca and higher temperatures. There were corresponding increases in ADP, AMP, and free creatine. At 17°C, there were no such changes except in hearts perfused with 240 retool/1 K and 5 mmol/1 Ca. On the contrary, CrP and the ratio CrP/ TCr were elevated above the control values in hearts perfused with Ca-free solutions containing between 30 and 120 mmol/1 K.

The decrease in ATP served as the parameter of the ischemia-induced metabolic alterations and the energy deficit developing during cardiac arrest. At 37°C, it averaged 1.67 and 2.55/zmol/g (p<0.02) after 10 min of arrest induced by 30 and 240 mmol/l K respectively. Ca, 1 or 5 retool/1 respectively, significantly increased the ATP-decay to 2.41 and more than 3.63 pmol/g/lO min respectively (/7<0.02 and p<0.05). Hypothermia of 27°C reduced the rate of metabolic deterioration and suspended the influence of the K-dosage. However, the Ca-effect was still visible: Induction of cardiac arrest by 30 and 240 mmol/l K without/with Ca decreased ATP by 2.07/ 2.58 pmol/g (p<0.025) and 2.07/2.41 pmol/g (n.s.) respectively within 20 rain. At 17°C, the rate of metabolic alterations was further reduced. The effects of K-dosage and of Ca were completely abolished. The ATP-decreased 2.2 pmol/g during 40 min of arrest.

Although higher concentrated K-solutions for infusion cardioplegia do not enhance the ischemiainduced myocardial metabolic alterations in deep hypothermia, they are not recommended for practi-Despite the advantages of CBK carrdioplegia, the severely impaired myocardium and/or long ischemia time continue to be a challenge.

Because of the association of Ca ~ with cell injury and death the use of Ca ~ channel blockers is logical. Investigation of CBD revealied no advantages over CBK. The combination of K and D is appropriate because of their different mechanisms of action.

Ten dogs had one hour of myocardial ischemia (MI) with topical ice (temp 7___2°C) after coronary perfusion with 200 ml CB (5___ l___C) containing K, 30 meq/L and D, 400 pg/kg. Eight dogs had two hours of MI after perfusion with 200 ml CB containing K, 30 meq/L and D, 200/zg/kg. Six dogs received the same treatment as the previous group except that D was increased to 1600/zg/kg for all four perfusions. Baseline studies were repeated after 60 min ofreperfusion without the use of Ca ~ or inotropic agents. Heart rate, peak systolic pressure, Vc¢, V~,~x, peak + dp/dt, peak -dp/dt, dp/dt over common peak isovolumic pressure, left ventricular compliance, stiffness and elasticity and great water were unchanged from control. Coronary vascular resistance was unchanged in groups one and two but declined in group 3. Creatine phosphate returned to control but ATP, ADP and the adenosine pool were depressed. Ultrastructure was well preserved. In 16/24 dogs defibrillation was not required whereas 48/48 dogs with CBK and 13/13 with CBD required defibrillation. These data suggest that D is a worthwhile addition to CBK. Early one-stage repair of some congenital vascular anomalies might be accompleshed readily if the material used for grafts grew along with the reconstructed tissues. We have evaluated the capacity of tubular grafts constructed from anterior rectus sheath to serve as growing segmental replacements of the descending thoracic aorta of puppies (age 9 weeks). Grafts measuring 3 cm in length were placed in 17 animals; 10 were implanted with attention to tissue preservation (live grafts) while 7 received grafts subjected to prior freezing and thawing in order to kill the donor tissue cells (devitalized grafts). Grafts were measured initially and at sacrifice 1, 2 or 3 months later. Each months 3 animals with live grafts and 2 with devitalized grafts were sacrificed. No aneurysmal dilatations or grafts ruptures were observed in any of the specimens. By 3 months, the live grafts had increased 26.9-1-1% in length and 27.3___ 3 % in diameter, while length and diameter of the devitalized grafts showed minimal growth (2.1___20/o and 5.3___5o/o respectively). During the same period the length of the thoracic aorta increased 4.5___5% in the live graft group and 41.1___ 8 % in the devitalized graft group. Grafts were lined by endothelium and organized into 3 transmural zones. The thickness of each zone in the live grafts remained static from 1 to 3 months, whereas thickness of the middle and outer zones in the devitalized grafts increased 4 to 6 fold. In living grafts the layers consisted principally of newly formed fibrocellular tissue; the rectus sheath was reduced to a atrophic fibrous band. By contrast, the rectus sheath remained prominent in the devitalized graft specimens. Cellularity (cells/field) of the live grafts was much greater than that of the devitalized grafts (ratio 15.9___3 : 6.0-----2). A newly formed, fibrocellular layered structure replaces the rectus sheath in the live grafts. These findings indicate that rectus sheath grafts are capable of growth in the young animal, providing that at implantation the tissue is well preserved. It is well known that durability of Valvular Bioprostheses (VBP) depends mainly on the structure of constituent biomaterial and long-term preservation of the Collagen network. To our knowledge, no studies on ultrastructure of tissues currently used for VBP has been reported so far after a long time of chemical preservation. The presentation of a new anysotropic tissue, Xenogenic Cervical Duramater (XCD), with a Collagen Tridimensional Network, and a comparative study of Scanning (SEM) and Transmission (TEM) Electron Microscopy of Bovine Pericardium (BP), Human Duramater (HD) and Procine Aortic-Cusp (PAo) represent the purpose of the present communication.

Samples of the tissues were obtained in semisterile fashion, rinsed and fixed in Karnovsky medium at 4°C for 24 h. Materials were then minced on paraphine plates and washed with 0.2M cachodylate buffer for 1 h and dehydrated in increasing concentrations of Ethanol (30 %-100 %) and intermediate exposure to 1% Uranyl Acetate for the in-bloc contrasts. were conducted, and results are exposed in Table 1 .

As a general rule, histologic evaluation was by far, more satisfactory for tissues preserved with 0.5 % glutaraldehyde (XCD, PAp) than those with 98 % glycerol (HD) and 4 % formaldehyde (BP) in combination. It is concluded that the new anysotropic tissue herein presented, XCD, can be appropriately preserved with 0.5 % glutaraldehyde as collagen preservation is concerned, and it's anysotropic and ultrastructural features, maintained for as long as 12 months as assessed by SEM and TEM. Albeit, clinical use of this new biomaterial is subjected to further experimental and animal trials. Tetralogy of Fallot and absent pulmonary valve (TAPV) is associated with massively dilated pulmonary arteries which cause tracheobronchial compression in the newborn and heart failure and cyanosis in older patients. Corrective operations have been attended by mortality rates exceeding 40% due to pulmonary insufficiency causing right heart failure (RHF) and pulmonary complications. Pulmonic valve insertion (PVI) with complete repair has resulted in improved survival. The purpose of this paper is to assess the results of total correction and PVI in ten patients.

During the last five years, 152 patients with tetralogy were corrected. Of these, ten patients (ages 51 days to 34 years) had absent pulmonary valve. One patient (51 days) present with severe RHF and pul-monary insufficiency and nine patients presented with mild RHF and cyanosis. Chest roentgenographs showed increased cardiothoracic ratio and pulmonary prominence in all. Arteriography revealed massively enlarged pulmonary arteries with a mean right pulmonary artery to aorta size ratio of 2.7 to 1. Associated pulmonic stenosis and insufficiency was present in all.

Seven patients underwent closure of ventricular septal defect (VSD) and PVI. Of these, three had PVI (2 tissue and 1 prosthetic) with outflow patch and four had right ventricle to pulmonary artery (RV-PA) tissue valved conduits. Two patients had repair without PVI, and one had repair with a monocusp pericardial valve patch.

Nine patients have done well with no episodes of thromboembolism or infection. Death occurred in a 51 day old infant who had VSD closure and relief of pulmonic stenosis. Pulmonary valve insertion seems indicated in these patients at it lowers peak pulmonary artery pressure and thus reduces compression effects on the trachea and bronchi. As well, when PVI was used right heart failure was not noted postoperatively. Although in experimental acute myocardial ischemia intraaortic balloon pumping (IABP) appears to increase regional contractility of ischemic segments of the left ventricle by increasing the collateral flow, evidence for this effect of IABP in patients with refreactory myocardial ischemia is not available. This study was done to analyze the effect of LABP on global and segmental left ventricular performance in patients with refractory, acute myocardial ischemia using gated blood pool scanning with Tc-99 albium tracer. Eight patients were studied on-and-off IABP prior to and following coronary bypass surgery. Left ventricular (LV) wall motion analysis was undertaken and both cardiac output (CO) and left ventricular filling pressure (LVFP) were determined from thermistor-tipped pulmonary artery catheters. When placed on IABP, mean preoperative global ejection fraction (GEF) increased (0.264-0.05 (OFF) to 0.354-0.05 (ON); P<0.01) abut no changes in CO or LVFP were observed. When comparisons were made praend postoperatively, CO increased after operation (P<0.05), due chiefly to an increase in heart rate (824-6 to 1074-8 beats/rain; P<0.01), but GEF was unchanged whether or not IABP was on. Only segmental wall motion analysis pre-and postoperatively revealed that IABP increased regional ejection fraction (REF) and contraction velocity by 50% (mean) in the angiographically determined regions of myocardial ischemia and these regional changes were maximal during the last one-third of the LV contraction cycle. It appears that REF is a more sensitive indicator of changes in LV performance than CO and GEF in these patients. Further, this study indicates that IABP does increase the regional wall contracility in the ischemic areas of the myocardium in man. The use of aspirin (ASA) to inhibit platelet thromboxane synthesis (TBX2) in patients undergoing coronary artery bypass grafting (CABG) has been advocated to reduce the potential for graft occlusion. However, ASA in conventional doses also inhibits the venous synthesis of prostacyclin (PGI2) a potent anti-thrombotic factor, and may contribute to excessive surgical bleeding. To investigate the relationship between ASA dose, blood loss and inhibition of venous synthesis of PGI2, 27 patients undergoing CABG were studied. Control patients (no ASA) were compared to those receiving 80 or 325 mg. ASA as a single dose 8-16 hours prior to surgery. Production of PGI2 by saphenous vein specimens removed at the time of surgery was measured by radioimmunoassay (RIA) of 6-keto PGF~a following incubation with 25 uM Na arachidonate. Blood loss was assessed by chest tube drainage and transfusion requirement in the perioperative period. Serum TBX2 was measured by RIA following ASA ingestion. Transfusion (units), Drainage (cc), Vein PGI2 (pg/mg tissue) and serum TBX2 (ng/ml) were (mean 4-SEM):

Therefore, ASA 325 mg or 80 mg did not increase blood loss during CABG. However, ASA 325 mg, significantly reduced saphenous vein PGI2 synthesis (/7<0.01) while the lower dose ASA 80 mg spared the production of PGI2. ASA is both doses inhibited TBX2 (p<0.001) and blocked platelet aggregation. Low dose ASA deserves further investigation as an anti-thrombotic agent since it does not appear to increase operative bleeding or significantly inhibit venous PGI2 production.

Transfusion Postoperatively, IgG increased progressively in patients in group A, reaching preoperative levels on the 6th or 7th day, whereas in patients in group B, IgG remained at lower than preoperative levels during the first week. The number of patients with abnormally low lavels of IgG during the whole PO period was significantly higher in group B (P"~.01). No significant differences were found in the others studied variables.

One patient in group A had infection. Three patients in group B had infections. Conclusion: High doses of Fab2IgG administered during OHS and the first PO days are effective in lessening the PO decrease oflgG and thus may be beneficial in preventing PO infection. A water insoluble but very hygroscopic polyvinyl alcohol powder, Zy-15044", is capable of significantly stimulating the cicatrization of open, contaminated skin wounds in rats. When it is compared to other substances in clinical use such as powders containing antibiotics, antiseptics, amino-acids, enzymes or a dextranomer, no other agent tested was capable of producing a similar beneficial effect.

These excellent experimental result justified pilot clinical trials on 30 patients: 18 varicose vein ulcers, 4 atherosclerotic leg ulcers, 4 infected traumatic wounds and 4 infected postoperative wounds. Whether in rats or in patients, Zy-15044 powder removed secretion, bacteria and tissue debris; it produced an early and increased proliferation of fibroblasts with the appearance of dense granulation tissue; it reduced wound size after less than three applications permitting very early grafting of the wound of eventually, cicatrization advanced to complete epithellzation. Regardless of whether the treatments were applied daily in the hospital bed or every other day in outpatients, there were not complaints of unpleasant sensations such as itching, burning or pain. Few studies have dealt with long term healing in stomach and none is available for duodenum. This study evaluates morphology and development of mechanical strength and collagen concentration in and adjacent to wounds after 20 to ]g0 days of healing.

Incisions were made in the non-glandular (Rumen) and in the glandular oxyntic (Corpus) part of the stomach and in duodenum of 64 male Wistar rats, 23 intact rat served as controls. The wounds were dosed with 6-0 polypropylene sutures using a single sutur technique. After 20, 40 and 80 days of healing complete load deformation curves were determined on strips perpendicular to the incision line after the sutures were cut. Collagen distribution was measured as hydroxyproline in strips parallel to the incision line. Wound tissue continued to gain strength up to 80 days of healing (breaking energy significantly increased in all tissues from 20 to 80 days of healing). No such increase was found for wounds tested togetherwith adjacent tissue, because the "point of maximum weakness" had moved laterally from the incision line with healing time. Wound collagen concentration increased in corpus and duodenum between 20 and 40 days of healing, but was unchanged between 40 and 80 days. The biochemical active zone around the incision line was unchanged 4-5 mm on each side from day 20 to 80. Conclusion: While the wound tissue itself continues to gain strength during the 80 days ofheaiing studied, the increase seen after 20 days does hot,enhance the functional properties of the organ as a'whole. The "point of maximum weakness" has at that time moved to the borderline of the biochemically active zone, which lies outside the tissue area enclosed by sutures. Myocutaneous flaps based on either the inferior or superior epigastric artery may be used to cover extensive soft tissue defects from the mid thigh to the clavicle. We report our experience with 33 rectus abdominis flaps, 29 based on the superior epigastric artery and 4 based on the ingerior epigastric artery. The flap is easily elevated, no skin loss has occurred and primary closure of the donor defect has been achieved in every case. In a follow-up period of between 1 month and 15 months no patient has developed an incisional hernia.

The flaps have been used to cover extensive chest wall resection for recurrent carcinoma of the breast (6 patients), extensive radionecrosis of the chest wall (1 case) as part of a primary breast reconstructive procedure (22 patients) and to replace soft tissue overlying the femoral vessels after radical dissection of the groin for metastatic tumour (4 patients). Of the 22 patients undergoing primary reconstruction 20 required additional subpectoral prosthesis but in 2 adequate bulk was provided by the flap itself. Primary healing occurred in all cases.

This easily raised and reliable flap will shorten hospital stay after major ablative surgery for recurrent disease and provides a useful addition to methods of breast reconstruction.

A. Watson Department of Surgery, Royal Lancester In girmary, Lancaster, U.K.

Occult gastro-intenstinal bleeding which defies readiological and endoscopic diagnosis provides one of the greatest diagnostic challenges in surgery. Lesions producing this clinical situation are often very small by definition and not infrequently mucosal angiodysplastic lesions, localisation of which may be extremely difficult pre-operatively and indeed at laparotomy.

Various methods have been described in an attempt to improve pre-operative localisation of such lesions. String tests an dthe detection of SICr labelled red cells in the faeces are notoriously inaccurate. Arterioghraphy is invasive and usually necessitates a bleeding rate in excess of 500 ml per day. Scintiscanning after red cell tagging with 99mTc gastro-intestinal blood loss, but localisation is difficult.

A method oflocalisation of such lesions using SlCr labelled red cells and intestinal intubation is described. After labelling of the red cells, a Miller-Abbot intestinal tube with the balloon inflated and rendered radio-opaque is passed and samples of gastrointestinal secretions aspirated at each 5 cm during passage down the gastro-intestinal tract. Samples are counted on a well scintillator counter and when a sample of high radioactivity is obtained, localisation is assessed both by the length of tube passed and by instilling dilute barium down the Miller-Abbot tube underfluoroscopic control. This method has been used successfully in five patients which were subsequently treated surgically with localised resection resulting in cessation of bleeding. Case histories together with photographs of the method and respected specimens will be presented in poster form. Postoperative sepsis is tlae most frequent complication of surgery and is the commponest cause ofprolungation of hospital stay. Purpose of the study is to prospectively evaluate incidence predisposing factors, bacteriology and costs of postoperative infections.

758 consecutive surgical patients admitted to our institute from May 79 to July 81 were studied. Patients undergoing minor surgical procedures (wound less than 2 cm) were excluded from the study. Patients were evaluated daily during hospital stay for onset of infection and results recorded in a data sheet. Hemocultures in septic patients and free plasma, activated partial thromboplastin time, prothrombin time, and thrombin time, ristocetin test for soluble monomer complexes and fibrin degradation products (FDP, Welco test) euglobulin lysis time (ELT) and platelet counts. Conclusion: 1. Thrombocytopenia is not a typical feature of boomslang coagulopathy. 2. The demonstration of soluble monomer complexes is blood samples (positive ristocetin test) appears to be an early and consistent indication of intravascular coagulation. 3. Despite unequivocal evidence of advanced consumption, platelet function seems to be maintained, thus preventing complete heamostatic failure. 4. Support for this view is found in the clinical observation that bleeding is essentially mild and late in onset. 5. Thrombelastographic hyperretractility the "spinning top" appearance is a constant feature and is probably mediated via the platelets. 6. Although D. (vpus venom is extremely potent and often fatal, the antivenom is rapidly effective despite advanced consumption. This mechanism which may involve the platelet release reaction has not been fully elucidated and deserves detailed study. In most cases artificial ventricles are driven pneumatically with the advantage of easy handling and the disadvantages of regulation problems, high weight and volume of the driving units and the danger of air embolism in the case of membrane rupture. The driving unit developed at Innsbruck Universiy consists of a microprocessor-controlled electromagnet driving a pump that functions as a safety chamber (SK). Force transmission to the artificial ventricle (ellipsoid heart-EH) is effected hydraulically. Thus there is a fixed connection between armature stroke and membrane movement in the EH. The armature position is measured by the microprocessor by means of a position sensor with programmable switch-over points determining the change from systole to diastole so that Idling and beat volumes, respectively, of the EH can be programmed.

The SK is a newly designed double rolling membrane pump for pressure and suction with very low compliance. Pressure and suction are determined by the armature force which is proportional to the microprocessor-controlled current in the solenoids. Thus a very positive control of the pumping action is possible.

In mock circulation experiments with the hydraulic safety driving unit the cardiac output (CO) was between 21/min and 121/min at beat frequencies of 40 to 130 bpm. With constant piston-stroke a direct relation between frequency and CO was observed. The influence of preload (Starling's Law) and afterload decreases with increasing armature force (decreasing periods of systole and diastole, respectively) which is due to friction in the hydraulic transmission system. Improvements in the geometry of the next system should reduce these losses.

In vivo experiments confirmed the hemodynamic efficiency of the system. Applied as LVAD the system proved to relieve the beating heart considerably. In the case of heart fibrilation the circulation could be maintained by the safety driving unit. The variable height differences between driving unit and EH as they occur in long-term experiments (animal lying or standing) affect only the ratio systole/diastole periods with the CO remaining constant. Mechanically assisted circulation is indicated in patients with cardiac failure after open heart operations. The clinical experiments show that left ventricular assist devices are capable only in a few cases to maintain full circulation. The limiting factor is the additional right heart failure syndrome. In patients with postoperative cardiac failure, a distinction must be made between isolated left heart failure and total heart failure. In patients with total heart failurebased on diffuse coronary sclerosis or on an increased pulmonary resistance -left ventricular assistance alone is not effective and right ventricular support is desired. Therefore, a simple, quick and safe biventricular assist device is necessary. With the biventricular bypass it is possible to maintain complete circulation in cases of cardiac insufficiency. For a successful outcome in patients with low cardiac output after cardiopulmonary bypass, the E-BVAD was evaluated in animal experiments (with 9 calves in 4 acute and 5 survival experiments) in cardiac failure situations. The cannulas for the inflow are put into the left and right ventricle, the outflow tracts into the descending aorta and the pulmonary artery. Both parts are connected with the Ellipsoid hearts.

With the E-BVAD it is possible to have a replacement of the heart -a total heart assistance similar to the total artificial heart. In cases of clinical emergency this device can be recommended because of the satisfactory hemodynamic effects achieved and the small degree of traumatic hemolysis (1,8 mg% free hemoglobin). It represents an easy and quick implantable system for total functional heart replacement. The realtionship between acute pancreatitis (AP) and the enzyme and hormone level in blood and gastric and duodenal juices is the factor that had driven us to do this experiment that would complete the study of the physiopathology of this illness. Material and Methods. We used eight 5-year-old dogs. A) Production of two antiperistaltic fistula in the Mann and Bullman way; one gastric and the other duodenal (first part). 1. Blockage of the gastric and duodenal compartments: by the gastric fstula we introduced one Foley's bucket with closed point and inflatable globe to block the pylorus. By means of the duodenal fistula we introduced another Foley's bucket into the duodenum. Once inflated, the ball blocks the first duodenal portion. 2. Juice extraction: By adjacent openings in the gastric bucket blocking the pylorum, we extracted the gastric juice separately. With another thin Foley's bucket, that was introduced by the duodenal fistula near the already blocked one, to block the third part of duodenum when the globes inflates. B) Production of AP: 15 days after the first surgical operation, by the Morandeira method with intraparenchymal injection of a mixture of autologous bile and olive oil. Results. One of the animals died on the fifth day after the first operation. + No animal died spontaneously after the second operation. They were killed on the 4th day. In each one acute necrotizing pancreatitis could be verified. + The radiograph showed that the gastric and duodenal compartments were sealed. -It was easy to separately extract blood and gastric and duodenal juices. Conclusion. We believe that this method is complete, simple, with good results and very useful for the study of the physiopathology of AP. Liver samples were taken form 18 patients operated for cholelithiasis and common bile duct obstruction by gallstones or pancreatic tumour. Early evident histoenzymatic deficiency was found even without jaundice or liver structural lesions. In the bilio-obstructive jaundice, the histochemical methods revealed more marked liver impairment, as compared to the histological picture. The degree of the enzymatic depression could explain the occurence of the postoperative liver failure in some patients.

Experimentally, the effect of common bile duct ligation was studied in 80 rats. Group I = 10 healthy controls. Group II = Aspartate (1 ml 5% sol.) was injected i.p. in 10 rats daily, for a week. Group III = the common bile duct was ligated under other anes-thesia, in 20 rats 48 h before killing, and group IV=in another 20 rats 7 days before killing. Group IV =in 20 rats aspartate was administered 3 h before an twice after choledocus ligation, and group VI = in another 20 rats aspartate was injected daily for a week after surgery.

The liver slices frozen in liquid nitrogen were cut in a Slee-type cryostat and the histochemical reactions were performed according to Arnold, Chayen-Bitensky and Lojda-Gossrau-Schiebler's methods. The biochemical reactions were performed using Merckotests and Merz-Dade test-kits.

After 48 h, and still more after 7 days, very severe structural, histochemical and enzymatic lesions developed. The liver cell glycogen and RNA, as well as the "marker" enzymes of infrastructures markedly diminished: NACH2-tetrazolium reductase -21.33%, glucose-6-phosphatase -48.33%, alphanaphthylacetat esterase -31.33%, ATP-ase -71.0%, 5'-nucleotidase -38.67%, PAS-reaction -56.0%, MGP-staining -30.67%. Aspartate treatment seems to exert a protecting effect against the noxious action of retained bilirubin and conjugated bile salts upon the liver cells: NADH2-tetrazolium reductase +17.37%, GsP-ase +40.64%, esterase + 24.27%, ATP-ase +67.81%, 5'N-ase +41.85%, PAS + 25.05%, MGP +29.32% and lipids -33.12%. But it does not influence the biochemical signs ofcholestasis: bilirubinemia, alkaline phosphatase, leucinearylamidase, gamma-glutamyltransferase, lactate dehydrogenase. Aspartate prevented partially but significantly only the increase of cytolysis enzymes: ASAT -40.48% and ALAT -52.97%. The histochemical and enzymatic results are in agreement with the severity ofmorphologic changes. Therefore, aspartate treatment might be adequate for the preoperative improvement of the liver function in cases of obstructive-jaundice, in order to reduce the incidence of liver failure, without influencing cholestasis.

Oral Glucose Tolerance Tests (OGTT) Y. Yamoaka, A. Sugitani, IC Kimura, IC Ozawa and Y. Tobe Department of Surgery, Kyoto University Medical School, Sakyo-ku, Kyoto, 606 Japan Two patients with cholecystographic evidence of septate gallbladder underwent dynamic hepatobiliary radionuclide scanning with 99mTc-EHIDA. When a steady state of emission had been reached from both gallbladder lobes, cholecystokinin was infused incrementally in four doses (0.005, 0.01, 0.03 and 0.06 Ivy-Dog-Units kg-lmin -1) and counts from the gallbladder analysed by computer.

In each an obvious functional difference was revealed between the lobes: both distal lobes ceased emptying abruptly during the third hormone infusion whereas the proximal lobes continued to empty predictably [2] .

In one case, histology revealed the septum to be a well-developed smooth muscle ring with cholecystitis glandularis proliferans confined to the distal lobe; the other case awaits surgery. The cause of the difference in response was probably contraction of the muscle in the septum acting as a sphincter.

This study provides indirect confirmation that pain in septate gallbladder is due to septum contraction and raised distal intralobe pressure. It is known that the gallbladder stone passes into the common bile duct. Furthermore, common bile duct stones pass to the duodenum. These gallstone movements were analysed in 334 cases during the past six years. The phenomenon in which gallbladder stones pass through the cystic duct down to the common bile duct are seen in cases where stones are small and numerous, the cystic duct is wide and connected with the common bile duct angularly or in parallel.

In 13.75% of our cases common bile duct stones passed into the duodenum. There were neither inflammatory stenosis of the terminal choledouchous, high grade pappilitis nor severe obstruction of the common bile duct in these cases.

It appears that the movement of gall stone is related to size and number of stone and morphological variations in the biliary system. The study was conducted in order to investigate whether intraoperative mano~etry of the bile ducts could be of additional valtie in diagnosis of afflictions of biliary tract or of the papilla. In contrast to current methods of water manometry it seemed important to the authors to use a simple and safe method which could deliver precise results. Method: Electromanometrics studies were carried out with a cannula placed in the common duct. Three manometric parameters were obtained for evaluation. 1. Resting pressure (RP); 2. pressure increase after injection of 10 ml saline (lml/s) (PI); 3. time in seconds needed for return of initial pressure (bile flow) (TR). For pressure measurements a pressure transducer with a recorder was used. Results: (Table) Manometry results in patients with calculi in the common duct or with organic Stenosis of the papillawere significantly changed as compared with normal findings in the bile ducts. False positive results (proven by cholangiography and biopsies of the papilla) were found in 3.5% of cases, false negatives in 6 % 97 % of all patients with patho-histological changes in the papilla had pathological bile pres-. sures. Although false positive and negative results were mainly caused by methodical errors, the described method delivers a high percentage of correct results in agreement with the final diagnosis. It is avery sensitive method in indicating impaired and reduced bile flow and discovers early pathological changes in the papilla as proven by histological examinations.

A functional anhepatic state in experimental animals for biochemical studies and as a model for treatment of acute liver failure can be achieved by different surgical procedures like portocaval shunt and complete arterial devascularization or by replacement of the liver with vascular prosthesis and protocaval shunt.

We developed a simple hepatectomy procedure using t/4 inch silastic tubing, carbon-Y-connector and specially designed "vascular spirales" to restore normal portal and caval blood flow.

This model was used for auxiliary liver perfusion studies in heparinized animals; the operation takes 15 min, requires no surgical skill, vascular occlusion is less than 1 min, no signs ofsplanchic hypertension are present and blood loss in fully heparinized animals over 24 h is insignificant. One of the main problems in atypical and anatomical liver resections consists of achieving appropriate hemostasis. Also because bile capillaries are opened when liver tissue is separated there is the danger of infection and subphrenic or parahepatic abscesses.

The authors have used the human Fibrinogen clue to seal the surface of the resected liver in 7 cases (3 hemihepatectomies and 4 resections). Even in anatomical resections exists, in the majority of cases, a diffuse bleeding or oozing of the resected area. This can be completely controlled by sealing the surface with the Fihrinogen adhesive. Also one of the main advantages of the Fibrinogen consists of having the possibility to avoid the insertion of numerous tubes for drainage. All our cases were drained by a single Silicon tube.

The postoperative courses were uneventful in all cases no major complications were observed. Liver regeneration is thought to be stimulated by changes in portal blood constituents, or flow or by a substance in the hepatocytes. Rice et al. have documented increases in total hepatic perfusion in rats during liver regeneration. This study has measured the portal and arterial components sequentially in large animals.

Young pigs (6-8 weeks) were subjected to shamoperation or 50% partial hepatectomy (PH). Blood flow in the portal vein or hepatic artery were measured pre-operatively and at intervals of 24 h postoperatively using cuff probes of an S.E.M. electromagnetic flowmeter; these were positioned daily under light anaesthesia. Systemic arterial and portal venous pressures were measured via indwelling catheters.

Previous studies have shown that the peak of regenerative response occurs in pigs 3 to 4 days after PH. The mean pre-operative total liver blood flow increased from 1.02+0.36 (s.e.m.) ml.g-lmin-1 to a peak of 226+ 33 % on day 2. Thereafter flow declined towards normal by day 6. The pre-operative portal component was 71 ___3 %; this increased to 85+2.5 % on day 2 and returned to normal level within 4 to 6 days.

It seems that both total hepatic flow and the portal component increase markedly just preceding the peak of regeneration. This response may stimulate regeneration, or merely accompany it, or may be proyoked by similar stimulators. The analysis of individual bile acids is relevant to several clinical investigations although established techniques have a number of disadvantages. To overcome these we have developed a simple HPLC method using a Waters Associates RCM100 Radial Compression Module with a Radial-PAK C18, 10,u, 8mm ID, reverse-phase cartridge (column). Bile acids were eluted from the column at a flow of 2.0 ml min-1 with a mobile phase of methanol:water (75:25 v/v) containing 2.5 % (v/v) acetic acid and adjusted to pH 5.25 with 10 M NaOH. A mixture of standards of cholic, chenodeoxycholic, deoxycholic, lithocholic and ursodeoxycholic acids and their glycine and taurine conjugates were resolved, while the 10 conjugates alone were completely separated in under 20 rain. The technique has been applied to the study of human bile from the common duct, gallbladder, tube and duodenal fluid, and serum from patients with hepatobiliary disorders. Bile acids in these samples were rapidly extracted using a SEP-PAK C18 cartridge before being applied (10/A to 200pl) to the HPLC system, although bile could be analysed directly without extraction. Quantitation (~mol ml-1 sample) of separated bile acids was achieved by comparison with standards using an on-line integrating computer and less than 5 nmol could be detected using a refractive index detector.

Acute Hepatic Failure Induced by Total Liver Devascularization in Pigs -Amino Acid Uptake by Combined Charcoal -Resin Hemoperfusion Support System where losses averaged 38 %. JIB patients had progressively greater losses with increasing preoperative weight (/250 lbs-30 %, 250-299 lbs-35%, 300-349 lbs-38 %, 350-399 lbs-43 %, 1400 lbs-50 %). Weight loss in patients dbs was statistically and clinically greater with JIB.

2. Diabetics, especially insulin-dependent, were rapidly cured after JIB. Normal plasma glucoses, serum insulins, and oral glucose tolerance curves were usually seen within l postoperative mouth, unrelated to weight loss. All 13 patients requiring insulin or oral medication preoperatively could discontinue mediation usuallywithin a weekpostoperatively. Improvement after GB was gradual, appeared related to weight loss, and was often incomplete.

3. Hypercholesterolemic patients had 51% decreases, from 275 to 134 mg/dl, after JIB, but only 19 % decreases after GB. All serum cholesterols were normal after JIB, but 23 % remained elevated after GB.

Because of complications after JIB, some needed reoperation and conversion to GB. Fortunately, most benefits were retained after conversion to GB. We suggest considering JIB for "superobese", diabetic and hypercholesterolemic patients. In the past 9 years 140jenunoileal bypass procedures were performed. The most common complications and side effects were frequent abdominal pain, or discomfort and flatulence in almost half of the cases. In addition kidney stones, blind loop invagination, electrolyte and liver dysfunction problems could be observed. The over all complication rate was 20%, the postoperative mortality rate 2 %. To eliminate the blind loop of the small intestine, to maintain enterohepatic circulation of bile, and to diminish undesirable side effects in the conventional jejunoileal bypass, biliointestinal bypass was introduced in 1980.

Twenty patients with a mean weight of 128 kg were subjected to primary biliointestinal bypass within the last two and a half years. Three additional patients had secondary biliointestinal bypass due to side effects, especially diarrhea and flatulence, of jejunoileal bypass, performed two to four years previously.

The surgical procedure entailed establishment of an end-to-side jejunoileostomy. 25 cm of the jejunum and 25 cm of the ileum were left in the continuity. The blind loop of the jejunum was anastomosed to a functioning gallbladder.

So far the above mentioned complications of conventional jejunoileal bypass could be remarkebly diminished. There have been no deaths in this material and no metabolic side effects. Frequent diarrhea is avoided by reduced bile spill-over to the colon. The weight reduction has been satisfactory. In summary: Due to a lesser complication rate biliointestinal bypass seems to be superior to the simple jejunoileostomy in the treatment of morbid obesity.

Hyperplasia or Neoplasia G.W. Geelhoed, C.J. Schaeffer and P. Daudu Department of Surgery, George Washington University Medical Center, Washington, USA Patients with various thyroid disorders who were undergoing thyroid operation were studied for the presence of absence of estradiol and progesteronebinding proteins in thyroid tissue. Steroid receptor assays were carried out using similar techniques and standards as those routinely employed for study of breast cancer specimens. Quantitative data were collected by coded specimen number by an observer unaware of the patients' clinical diagnoses, and diagnostic correltions were drawn following de-coding.

There was no correlation with age or sex and the presence or quantitative value of steroid-binding site. Specimens with the histologic diagnosis of follicular adenoma had estradiol binding sites, and had them at high levels (14.19 to 143.83) femtomoles/mg cystosol protein). Patients with the histologic diagnosis of thyroid hyperplasia had marginally positive estradiol-binding and lacked progesterone binding. Patients with adenocarcinoma of the thyroid exhibited neither estradiol nor progesterone-binding in significant quantity.

Presence or absence of steroid binding sites in thyroid tissue appears independent of sex or age, but correlates with benign neoplasia of the thyroid, suggesting a possible etiologic association for thyroid adenomas. 77 patients with esophageal cancer underwent resection and primary reconstruction of the esophagus by posterior invagination esophagogastrostomy which we devised. The anastomosis was made in the cervical level in 58 cases and in the left thoracic cavity in 19 cases. Functions of stomach placed in the posterior mediastinum were examined in 7 patients surviving more than 5 years and in 13 patients surviving less than 3 years. Within one year postoperatively, the absorption ofVitamine B12 decreased markedly. One and a half year after operation, however, the secretion of Castle's instrinsic factor recovered and it showed normal values in patients surviving over 5 years. In secretion of gastric acid by tetragastrin, both total acidity and free hydrochloric acid increased in the progress of the postoperative period and showed normal values in all 4 patients surviving over 9 years. This fact was also confirmed by endoscopic observation of color development by Congo-red in the gastric mucosa; its coloring area was scattered in 2 patients surviving 5 years, but extended to the whole surface in patients surviving over 9 years. Roentgenographycally in head-down position, the surgically created fornix with a sharp angle of His was effective in preventing gastroesophageal reflux completely. the strain combination of donor and recipients. The nature of the mechanisms determing the different fate of allografts remains obscure and all interpretations of the above findings must be speculative. However, the long-term acceptance of hepatic allografts in the rat appears to be similar to that in other species and emphasizes the role of the liver as an immunologically favoured organ. Footnote: Part of the work (intrahepatic bile duct proliferation!) contained in this abstract will be presented at the XVII ESSR meeting 1982.

Peoples' Friendship University, Tamojenii pr. 4, Department of Operative Surgery, Moscow, USSR A transplanted kidney, besides tissue incompatibility reaction, is influenced by non-specific injuring factors, including decentralisation. For normal functioning of a kidney transplant over a long period of time we carried out an experimental study. The purpose of the study was the investigation of operative reinnervation possibilities and its influence upon the functional state of a transplanted kidney. For the kidney reinnervation Kirpatovski's method for suturing perivascular fascial tissue flaps of anastomosed vessels with branches of vegetative plexus nervosus on them was used. Experiments were carried out on 175 mongrel dogs, both male and female. In the first group of experiments a kidney was transplanted into the pelvis while the opposite kidney was preserved or ablated. In the second group the kidney autotransplanted into pelvis was reirmerved by the described method with preserving the contralateral kidney in part of the animals and ablating the kidney in the rest of them. The third, fourth and fifth were control groups. In the third group the kidney was denerved, in the fourth it was denerved and reinnerved by the mentioned method, in the fifth group unilateral nephrectomy was performed. After the operation the animals, in different periods of time (from 3 days to 2 years), were studied by isotopic renography and by scanning kidney's by JBl-hippuran and neohydrin-Hg 2°3 intravenous injections. Results of the isotopic renography were estimated qualitatively and quantitatively by universal renographic index (Hirakawa-Corcoran, 1963). Results of the experiments indicate that autotransplantation and denervation of a kidney without its reinnervation were followed by lowering of a kidney vascularisation and its secretory and excretory functions. During the first 2 or 3 months after the operation a tendency to improvement of the kidney function took place; later those functions gradually oppressed and the renographic index lowered down to 13.9+__2.1, 16.6+--.5.5 (p<0.001) respectively according to the groups. After operative reirmervation of a kidney autotransplant and denervation of a kidney without its transplantation gradual improvement of separate functions of the kidney took place: in 1 or 2 months the kidney circulation was restored; in 2 or 4, and in 4 or 6 months respectively, secretory and excretory functions of the kidney restored; renographic index reached its norm after 4 or 6 months and remained stable during 2 years (according to the groups 54.9+___5.5, 52.3-1-4.3, p>0.05 respectively). Thus, the operative reirmervation of a kidney prevented depression of its functions and provided its normal functioning during a long period of time. Introduction of fine needle aspiration cytology in renal transplantation gives the possibility not only to distinguish acute tabular necrosis, arterial and venous obstruction and viral infection from acute rejection, FNC also seems to allow judgement of the effectiveness of various types ofimmunosuppressive therapy during rejection episodes. Method: After sterile fine needle biopsy the cytological evaluation of the aspirate was performed on cytocentrifuged smears after staining the cells by the method of May-Grtinwald-Giemsa. Thus normal and "activated" mononuclear cell populations as well as parenchymal cells such as tabular and endothelial cells and their pathological changes could easily be recognized. The white blood cell types in FNC were compared with those of the peripheral blood. The difference in number due to the graft invading cells was expressed as "increment". the findings of FNC were correlated with clinical signs and serum creatinine values.

Results: FNC allows to judge the in situ inflammation in the rejecting kidneys within 2 h. Non-rejection grafts show cell counts comparable to peripheral blood. With onset of rejection (grade 1) T-and Bblasts and monocytes appear in the graft. With sever-ity of the rejection (grade 2) monocytes and lymphocytes as well as blood cells increase in numbers. Acute rejection (grade 3) is heralded by high numbers of monocytes and macrophages. Acute tubular necrosis (ATN) and virus infections can be distinguished from rejection episodes to be treated. In all cortisone and azathioprine resistant cases where ALG was used in order to suppress inflammation during rejection episodes it reduced not only the peripheral white blood cells, but also within 12 h after start of ALG therapy the number of in situ cells. Beside reduction of inflammatory cells typical changes in shape of the nucleus and density of chromatine in up to 10 % of the lymphocytes and granulocytes could be detected. These changes look closest to what has been described as apoptoses.

Conclusion: Cortisone and azathioprine resistant rejection episodes could be monitored within 12-24 h using fine needle aspiration cytology. This method is safe, cheep, and it provides good information about the in situ situation of the graft, it allows to distinguish acute rejection from ATN and other situations where higher immuno-suppressive therapy, especially with cortisone, could only be harmful for the patient. Criteria of the border of normothermic ischemic tolerance in dog kidneys are first normal PAH-and exogenic creatinine-clearances compared with unilateral nephrectomized dogs in neuroleptic analgesia under excessive water and sodium load and second perfectly normal kidneys after two weeks in pathological examination. The protective solution HTK ® by Bretschneider has a superior buffering capacity compared to Euro-Collins-Solution ® , which, by itself, enhances anaerobic glycolysis by substrate stimulation (glucose) especially between 15 ° and 35°C. Without preservation in normothermia oligo-anuric ARF is observed at the border of ischemic tolerance (15-20 min; 36°C). The obligo-anuric ARF is dependend on ischemia which will result in necrosis of the kidney after 45 rain and 36°C ischemic temperature. The border ofischemic tolerance corresponds to an intrarenal pH of less than 6.2 (outer stripe of medulla), and medullary lactate levels of 60 to 80 pmol/gdw.

Substrate stimulation of anaerobic glycolysis limits the effectiveness of Euro-Collins-Solution ® above 15°C earlier than anaerobic glycolysis in pure ischemia. After warm ischemia (>--__..27°C; 60 min) all kidneys protected with Euro-Collins-Solution ® show anuric ARF, whereas all HTK®-protected kidneys (60 to 240 min of warm ischemia; 32°C) are having polyuric ARF. Under these conditions, renal pH will not increase above 250 nmol/1 H+-concentra -tion. The border of ischemic tolerance for HTK ®protection ofkidneys is 120rain-31°C and seems to be limited by a transitory secondary postischemic oliguria (3-48 h). In summary: Intrarenal anaerobic glycolysis limits tolerance of pure warm ischemia by inducing anuric ARF. In contrast polyuric renal failure is observed at the border ofischemic tolerance by using the protective procedure with the HTK®-Solution mentioned above. At least 60 % of unrelated pigs respond to liver allografting with a rejection episode which they overcome without immunosuppression; kidney transplants are rejected uniformly within 9 to 14 days. In this study, MLC responses were measured at weekly intervals in such animals.

Non-litter mate pigs were subjected to autograft, exchange liver allograft, or exchange renal allograft. Blood from MLC responses were taken from unaesthetised pigs before and at weekly intervals after operemained the same for each pair of transplants. In 5 of 8 pairs of liver allografts, MLC responses were depressed up to 10 fold for 3 weeks post-operatively. In all these pigs, the MLC responses were initially high. In 2 pairs where the MLC response was initially low, there was no change after the transplant, and in one pair with a low pre-operative response, there was a marked increase post-operatively. In 3 pairs of kidney allografted pigs, there was a similar depression in the single post-operative sample available. In 3 liver autografted pigs, there was a slight post-operative rise in MLC response.

It appears that liver and kidney transplantation but not hepatic autografting, markedly depress sequential post-operative MLC responses. The process of ischaemic kidney degeneration was estimated by measuring the adenine nucleotide (ATP)levels and the effect of Inosine and Naftidrofuryl (Praxilene) was assessed. 50 min ischaemia was induced on both dissected kidneys of wistar male rats. Then, either both clamps were removed and the right kidney was excised after I0 rain reperfusion The role of mononuclear phagocytes in various disease states has been extensively studied by phagocytosis, Fc and complement receptor sites, and enzyme content. Chemotaxis of peripheral blood mononuclear phagocytes, however, has not been studied to any great extent. We have therefore investigated a method for quantifying chemotaxis by mononuclear phagocytes. Mononuclear phagocytes were purified from peripheral blood by centrifugation over Ficoll-Hypaque and a discontinuous percol gradient. Chemotaxis was quantified by the 'under-agarose method'. Mononuclear phagocytes were allowed to migrate towards the chemotactic agent zymosan activated serum (ZAS), and the control non-chemotactic agent Eagles medium. The distance of cell migration was measured after 20 hours incubation at 37°C in 5 % CO2 with the aid of a microscope eyepiece graftcule.

Using non-sepcific esterase staining the purity of mononuclear phagocytes obtained was 85%___ 10% and the viability by trypan blue dye exclusion was at least 99 % Preliminary results have shown selective migration by purified human peripheral mononuclear phagocytes towards ZAS. This method may therefore represent a means of investigating an important role of these cells in disease states. The intrathoracic pressure rises when a person exhales into a manometer to such an extent that finally the venous blood-flow to the heart stops. It is supposed that this venous stop flow pressure (VSFP), measured by an ultrasound device, is equal to the central venous pressure. In a prospective clinical trial by two independant examiners in 97 % of the cases (n= 100) the correlation was within 5 cm H20. Therefore, the central venous pressure (CVP) can be measured non-invasively with an ultrasound device. We have previously reported that opiate receptor blockade with naloxone (NAL) significantly improves cardiovascular function and survival in canine hemorrhagic shock [1, 2] . Since species differences do exist, we investigated the effects of NAL in cynomulgus monkeys anesthetized with N20/O2. Blood was withdrawn to achieve a mean arterial pressure (MAP) of 45 mmHG (t---0) which was maintained until t= 1 h when the reservoir was clamped and the animals treated with either NAL at 2 mg/kg bolus plus 2 mg/kg.h infusion I.V. or 0.9% NaC1 in equivalent volumes. There were no significant differences between NAL (n=5) and CON (n=5) in MAP, left ventricular contractility (LV dp/dt max, mmHg.10Vs), and survival; the NAL animals, however, were acidotic (PHa 7.24-----0.09) and colder (37.95-0.3°C) than CON (7.42+__0.02, 38.5-+-0.1). MAP responses to NAL were proportional to pH a Complications of vascular bypass grafts, especiallyin the femoro-popliteal region are common. The most frequent of these by far is occlusion. Other late complications iclude leaking and false aneurysm formation at the anastomotic site, dilatation of the graft material and stenosis ofanastomotic sites. The most commonly employed non-invasive studies render little information of value in the diagnosis of these complications.

We have recently undertaken a study to evaluate arterial grafts at various intervals after implantation with the use of a Duplex ultrasonic scanner. Imaging of graft material with this system ist excellent and patency of the lumen can easily be established. Graft diameter can be measured and accumulation of material within the lumen can be measured and quantitated. Although anastomoses are not always clearly visualised they are often seen satisfactorily for diagnostic purposes. Graft dilatation, false aneurysm formation and occlusions can be accurately diagnosed with this method. We have also studied the behaviour of femoro-popliteal grafts across the knee joint using the Duplex scanner.

As a non-invasive technique scanning with a Duplex system has many advantages enabling frequent investigations and therefore early recognition of complications where synthetic material vascular grafts have been used for arterial bypass. An experimental model for dissecting aortic aneurysm has been designed to obtain much better understanding of the disease, leading to more effective methods of surgical treatment.

Dissection of the descending thoracic aorta was successfully performed by modified Blanton's prodedure in 85 % of more than 100 mongrel dogs.

Fals lumens ruptured distally into true lumens to form a doublebarreled aorta in 70% of the dogs.

The method of surgical treatment performed were (1) closure of entry by direct suture, (2) closure of entry by inserting Dacron vascular prosthesis with a stainless steel ring, and (3) bypass grafting with vascular prosthesis and ligation of the thoracic aorta.

In 10 dogs treated by methods (1) or (2) at the time of performing the dissection, cine angiography revealed that false lumens hat thrombosed within 1 month and dissection was found to have completely healed on autopsy.

In 10 chronic cases treated by methods (1) or (3), false lumens were all patent with various degrees of formation of mural thrombi at observation of 3 to 6 months after surgical treatment.

These results indicate that closure of entry for acute dissecting aneurysm should be a curative procedure. Studies are now continuing on chronic dissecting aneurysm. The lack of controlled trials not rarely resulted in the incorrect acceptance o fan ineffective operation as an effective one 0igation of internal mammary arteries for relief of angina pectoris, for example). This failure also explains the non-stop controversies over the appropriate operation (and whether the question operate at all) for various myocardial revascularization procedures, over the thymectomy for myasthenia gravis, the limited vs radical mastectomy for breast cancer etc. Several clinicians and surgeons assert that controlled clinical trials for new operations are unrealistic and naive because of some important differences between operations and drugs. Compared with medical trials surgical ones are often really more difficult to be carded out. They are subject to more restrictions, ethical, statistical and practical, and may require longer or even remarkably long times. In any case, patients must nevertheless be sheltered from harmful and ineffective surgical operations since luckily not all the undoubted difficulties are unsurmountable, and prospective controlled trials of surgery, including random allocation of patients, are quite feasible.

Comparison between surgical and non-surgical treatment is a really particularly suitable field for these studies. Comparison of effectiveness of two operations of the same type or of different types can also be carried out, but only on certain conditions and using suitable and appropriate expedients. Also the doubleblind approach may be feasible, although remarkable restrictions and adequate cooperation are necessary. Acceptance in trials of surgery as placebo is objectively very difficult, and obviously it has to be considered only when an effective operation for the considered disease does not exist. The ethical sound of this delicate aspect of clinical research can be minimized enough only if (1) there are substantial doubts about the effectiveness of the intended operation and (2) requested placebo-surgery is of very slight importance.

Posters for Scientific Meetings Mary Evans and A.V. Pollock Scarborough Hospital, North Yorkshire Y012 6QL1, U.K.

The display of information in posters antecedes even the,art of writing. They have been used to advertise, to educate and to inform and their function is to disseminate information to a wide audience quickly, simply and effectively. In medicine they have been used since the thirteenth century for the promotion of health education.

In recent years poster sessions have been introduced into scientific meetings as an alternative to the spoken paper for the presentation of original work.

Following this experimental procedure hepatic coma supervenes in 6-8 h and animals die after ca. 24 h. Five male pigs, aged 3-4 months, weighing 30-35 kg were studied ca. 20 h after operation. The detoxifying system consisted of activated charcoal (Norit ¢ (lmm x 2ram) and ionexchange resin (Dowex lx2), subsequently inserted in an extracorporeal circulation. Plasma amino acids were determined serially (3, 20, 40, 60 rain) in and out the detoxifying system. Plasma extraction of individual amino acids (means_ SE in/zmol/min) is reported in the table A remarkable extraction was present in the first 40 rain and chiefly involved aromatic amino acids and tryptophan which did not further increase. During hemoperfusion, the molar ratios between neutral amino acids improved. The ratio of branched-chain to aromatic amino acids increased from 1

Late Hepatic Effects of Small Intestinal Bypass (SIBP) for Obesity

Plasmafl-endorphin (/3-END, pg/ml, by RIA) was related to T:fl-END = 334 (T-37) + 597, r=0.67, p<0.05. When acid-base balance and temperature were maintained, NAL (n=6) significantly (*p<0.05) improved MAP and LV dp

pg/ml in NAL and 969 + 94 pg/ml in CON). r-END is released in and contributes to the cardiovascular depression of primate hemorrhagic shock. NAL reverses this depression and improves survival but only when arterial pH and core temperature

samples of exudates at site of infection were taken wherever possible for aerobic and anaerobic cultures. In following infections were recorded: wound, respiratory tract, thrombophlebitis, indwelling intravenous catheter, unidentified origin fever (>38 °C lasting more than 3 days) (FUO) and miscellaneous. Incidences: 310 patients (41%) had postoperative septic complications. Wound infection was recorded in 145 patients (19.1%), respiratory tract infection in 105 (13 %), urinary tract infection in 100 (13 %), FUO in 55 (7.2%) thrombophlebitis in 18 (2.3 %) and miscellaneous infections in 49 (6.4o/0). Predisposing factors: Wound infections were 53/413 (12.8 %) in clean operations, 41/215 (22.7 %) in potentially contaminated, 19/65 (29%) in contaminated and 24/65 (36.9o/0) in dirty.Wound infections were diagnosed later (mean 9th day) in clean than in dirty operations (mean 5th day) (P<0.05). A statistically significant correlation was found between wound infection and leght of preoperative hospital stay: from 5.4% in patients operated on within 0-6 days to 42.8% for patients operated after 28 days or more (P<0.001). No correlation was found between wound infection and age. Urinary infections were more frequent when the patients were catheterized at least once in the postoperative period (34 % vs 11%). A statistically significant correlation was found between the incidence of respiratory infections and duration of anaesthesia (3.2 %----<60 min, 120/0>60<120 min, 270/0>120 min;/~0.05). Bacterology: 181 out of 295 cultures gave positive results (61%): aerobes were isolated in 129 samples (490/0); mixed aerobes and anaerobe in 29 (9.8o/0); anaerobes in 23 (7.7 %). Bacteroides Fr. was the commonest isolated anaerobe in all types of sample except hemocultures were propionibacterium aches was the most frequent. A statistically significant correlation was found between the incidence of recvery of anaerobes and intraoperative contamination (7.4 % in clean operations, 13.5 % in potentially contaminated, 12.5% in contaminated and 20.8% in dirty; P<0.05).In wound infections the most frequent aerobes were staphylococcus in clean and E. coli in contaiminated operations. Costs: mean postoperative hospital stay of patients with septic complications was 15 days, whereas patients with no postoperative sepsis were discharged after an average of 10 days (P<0.05). Mean daily cost in our hospital was extimated $100; accordingly, the mean postoperative hospital stay of patients with sepsis costs $1500 vs $1000 of that o~ the patients without postoperative infections. Overwhelming postsplenectomy sepsis/infection has been accepted the highest risk with more than 50 percent mortality due to Pneumococcus species. However, E. coli, Pseudomonas and Staph. aureus have been reported a deadful threat to the splenectomized individual also. Pneumococcal challenge after splenectomy in animal experiment and after partial salvage has been well examined. Staph. aureus-challenge has not been tried after splenic repair, so far. Material and method: A total of 150 NMRI-mice (18-20 g) have been subjected to: 1 sham-operation, 2 splenectomy, 3 peritoneal splenosis, 4 controls as non-operated group.The technique was similar to the one performed in 90 rabbits, prior reported. 6 weeks postoperatively, the mice were exposed to intraperitoneal injection of Staph.aureus concentrations of 101°c/ml down to 104c/ml. The peritoneal cavity of mice has a high resistance to Staph.aureus, as only No. 36835 proved lethal. Results: Whereas the macroscopic view post mortem seemed promissing, showing almost total salvage of the splenic particles 6 weeks postoperatively bacterial challenge with Staph.aureus was contradictory: 3 x 10 s c/ml was survived by all splenectomized mice, by only 25percent of the splenosis micecand by none of the sham and nonoperated mice. Due to the small number of only 48 mice within this mice-pathologenic Staph. aureus species, we may not draw definite conclusions. As for one, the intraperitoneal application may not be the natural way of septicemia. On the other hand, splenic remnants may not be as effective in the protection of Staph.aureus-sepsis as in a pneumococcal challenge. Therefore, further studies with Staph.aureus species in other animals and by other application route will be needed! Studies of the coagulant effects of boomslang (Dispholidus typus) venom have indicated that the coagulant effect was mainly due to its ability to activate prothrombin. It also activates prethrombin I, factor X and possible factor IX as well [1] . Although boomslang envenomation is said to represent a classic model of disseminated intravascular coagulation, certain features are worthy of critical thought. The coagulation profile of a nine year old lad, who was admitted to the H.F. Verwoerd-Hospital, 48 h after a reputed boomslang bite, provided the impetus to explore the in vivo effect ofcurde. D. typusvenom on the thrombelastographic and other haemostatic changes in the chacma beboon. Methods: 0.0005 mg, 0.0015rag and 0.05mg respectively of crude. D. typus venom were injected subcutaneously in 3 adult baboons. Serial determinations of the flow parameters were done over 3-6 days. Thrombelastography on whole blood an platelet This report presents the ten year survey of liver mitochondria analized in 385 patients.1. Two mechanisms were of major importance in the augmentation of mitochondrial ability to synthesize ATP: a) the enhancement of ATP-generating capacity per unit of respiratory assemblies and b) the increase in respiratory enzyme contents. Those compensatory mechanisms were functional within a certain range of contents in cytochrome A (0.5-1.5 nmol/mg protein, normal; 0.8). Hepatic insufficiency was observed in patients with cytochrome A contents less than 0.5 or more than 1.5.2. In all patients with cytochrome A of 0.7-1.0 nmol/mg, the blood glucose level after an oral glucose load (50g) returned toward normal within 2 h (parabolic Pattern). In 25 patients with cytochrome A of more than 1.5, the blood glucose level did not return toward normal within 2 h (linear pattern). Parabolic and linear patterns were intermingled in the patients with cytochrome A form 1.0 to 1.5.In this report, we will emphasize the importance of preoperative OGTT and measurement of cytochrome A contents during operation for the prediction of operative prognosis, better than any other currently used index of liver function, in the patients receiving major surgery such as hepatectomy or operation for esophageal varices in severe cirrhosis. The reticulo-endothelial system (RES) has largely been ignored in studies of liver regeneration. In this study, the RES was "blocked" with colloidal carbon, or was stimulated with olive oil or with glucan. Indices of liver regeneration were the Thymidine kinase acitivtiy (TK) and the number of mitotic figures (MI) in liver biopsies. Fibronectin was measured as a putative index of Kupffer cell function.Four animals in each group were sacrified at 6, 12, 24, 36, 48, 72 and 96 h after sham-operation or 68% partial hepatectomy (PH). Group 1: carbon suspension (Pelikan Ink, Wagner) 16 mg/100 g rat single injection pre-operative; Group 2: olive oil (10 % in 5 % dextrose water + 0.1% Tween 20) 0.25 ml/20 g rat single injection pre-operative; Group 3: as group 2 given at intervals of 24 h; Group 4: glucan 1.0 rag/ 100 g rat pre-op, and at intervals of 24 h.Results: Glucan or a single dose of oil increased TK in PH and sham-operated rats but did not influence M.I. Carbon inhibited MI but did not influence TK. Fibronectin levels were increased in sham-operated and PH rats after daily oil or glucan.In conclusion, no administration enhanced both TK and Mi but stimulation of the RES with glucan or oil did increase fibronectin levels in both sham and PH rats. It appears that the RES does not have a role in liver regeneration as assessed above, but fibronectin appears to have been an indicator of RES activity. Anatomical abnormalities of the gallbladder include multiseptate and bilobed organs, and the more common transverse septate variant. Patients with the latter type often have typical biliary symptoms which are thought to be caused initially by raised intracystic pressure and subsequently by inflammation and calculi formation in the distal lobe [1] .This study evaluate~ the long term effects of SIBP on hepatic lipid and fibrous tissue accumulation. Liver was obtained at the time of abdominal operation from 6 normal weight patients, 18 morbidly obese patients (mean body weight 163+21 kg) and 27 patients 4 or more years (mean 69-+ 19 months) postoperative from SIBP. From wedge liver biopsies hepatic total lipid, triglyceride, cholesterol, phospholipid and protein contents were determined as well as the activities of acetyl CoA carboxylase and fatty acid synthetase. Histologic evaluation of needle biopsies of the liver was used to quantitatively estimate the amount of hepatic steatosis and fibrosis present.There was also a significant histopathologic increase in hepatic fibrosis present in the liver tissue from SIBP patients when compared to hepatic fibrosis in liver from obese patients as well as when compared to liver specimens obtained from the SIBP patients at the time of their SIBP. The effect of reversal of the SIBP on hepatic fat content and fibrosis was determined by transcutaneous liver biopsy 6-22 mos. following take down of SIBP. Post-reversal histologic quantitation of hepatic fibrosis and inflammation was significantly decreased from pre-reversal values and the improvement was greater in patients who lost weight. Patients year after SIBP have persistent hepatic steatosis with hepatic fibrosis and inflammation. Improvement in these parameters may be anticipated following reconstruction of the intestinal tract particularly if weight loss is maintained. The aim of this paper is to study the metabolic effects of gastric bypass in dogs. We used dogs weighing between 21 and 25 kg. The animals were divided into three groups: A) five dogs without surgical operation (control); B) six dogs with a fundoj ejunal bypass, and C) six dogs with a gastroplasty. In group B and C the exclusion was 9/10 of the stomach. In all of the animals and groups were determined: cholesterol , live r function tests, calcium, magnesium-and electrolyte levels, the body weight evolution and the Apparent-Digestibility Coefficient (ADC) of fat, protein, minerals and glucosides for a type of diet. The composition of this diet in dry substance was: fat 35.4%, protein 35.2%, glucosides 26.5% and minerals 2.9%. These analyses were determined five times during the twelve months of control. The animals were kept in individual metabolism cages which allow feces and urine to be gathered separately as well as the food consumed to be controlled. The cages were housed in a room thermoregulated at 18 °C±2. The cholesterol, calcium, magnesium and electrolyte levels were normal during the twelve months of control in the three groups. The liver function tests were normal in all groups with the exception of serum aspartate transaminase initially deteriored in group B. The body weight evolution was significantly diminished in groups B and C compared with the control group (p<0.002 and p<0.002). The results of ADC expressed in % ± SD were.By 15roviding a functional gastric capacity of less than 100 ml and consequently a forced reduction in food intake the gastric bypass produces an effective loss of weigth. Our results suggest that the gastric bypass can be considered as an effective and safe alternative to intestinal bypass for treatment of morbidly obese subjects who have failed nonsurgical treatment; however, a great and deeper research is necessary to discover the possible side effects of 227 gastric bypass surgery; then its ultimate benefits will be fully understood. There was no operative mortality. The percent of the excess weight lost following the two operations is shown and compared with weight lost following gastric bypass. It is concluded that: 1. Complications occured more often following gastroplasty, 25.5 % compared with 10.4 % of the patients treated by gastrogastrostomy. 2. The amount of excess weith lost was greater after gastroplasty, a mean of 61% for 30 patients followed for two years, than after gastrogastrostomy, a mean Of 41% for 13 patients followed for two years. 3. Because of the incidence of stomal obstruction requiring reoperation following gastroplasty, 14.8 % compared with 1.4% after gastrogastrostomy, is our preferred operation. 4. To improve weight loss after gastrogastrostomy the TA-55 is now used instead of the TA-90 stapling instrument. The proximal gastric pouch is reduced in size to 20-25 ml and the stoma made smaller to 9 ram. An artifical esophagus was made of silicone rubber tube covered with a Dacron mesh. A segment of thoracic esophagus of 16 dogs was replaced with this graft using three different types ofamastomosis, i.e., overlayer end-to-end anastomosis, two layer end-toend anastomosis both using a flanged tube, and monolayer end-to-end anastomosis with noflange tube. Seven of 16 dogs (44%) survived more than 12 months without complications, and 4 of them more than 6 years. In 6 of 7 of the prolonged survivors, extrusion of the graft was recognized in the 3rd to 6th month after operation. Esophageal stenosis increased slightly up to the 6th month after extrusion of the graft, but it did not further advance until sacrifice. In these dogs, mucosal regeneration of the neoesophagus was complete with muscle layers and mucous glands in the submucosa recognized microscopically. Proximal esophagus from the replaced portion was apparently dilatated more than that of the distal portion. There was no definite difference between the anastomotic techniques with regard to complication or prognosis. These results suggest a possibility for clinical trials.

Paper withdrawn 171 Attempts to reduce nephrovascular hypertension by surgical techniques deviating renal venous blood and renin directly to the hepatic filter are at present discouraging. Experimental data with portocaval transposition are contradictory, due mostly to the use of heterogeneous biological models (mongrel dogs) and collateral circulation developement. Renal to portal vein end-to-side (ETS) and end-to-end (ETE) anastomoses in the rat were therefore tested as possible experimental models. We observed that even after right kidney decapsulation, collateral vein circulation develops through the periureteral plexus, particularly after ETS shunts, one month after surgery. Nevertheless, collateral cirulation in the male rat can be prevented by ligature and cutting offthis plexus near the kidney. In the female rat instead, collateral circulation is still possible. In fact, newly formed pericapsular veins join the right kidney to the right ovaric vein and therefore preventive ligature and eradication of these vessels is also necessary. In conclusion, both models (ETE-ETS shunt) appear feasible in the rat, and reliable for studies dealing with nephrovascular hypertension and hepatic metabolism of renin. The handling of the exocrine system is one of the main problems inpancreas transplantation. One trial to overcome this problem consists of the occlusion of the pancreatic duct system. A the theoretical disadvantage is the induction of a secretion oedema which may lead to blood flow disturbance followed by venous thrombosis. The aim of the present experiments was to study the effect of an anti-oedematous drug (cumarin and rutin sulphate (Venalot ®) on the blood flow of autologous segmental pancreatic transplantation was tested by chosing the cervical region of the dog as the site of grafting. Material and Methods: 10 dogs weighting 20-30 kg were used. These dogs were divided into two groups (control group, n=5; treated group, n=5, Venalot ® dose: 1 rag, cumarin/kg/b.w.). The body and the left limb of the pancreas was removed, perfusion with Eurocollins was started wia the splenic artery and the duct system was injected with prolamine. A distal a.

v. fistula of the splenic vessels was performed. This conditioned graft was transplanted at the neck of the same dog, performing the anastomoses between the carotid and the splenic vessels. The blood flow of the pancreatic graft was measured with radioactively labelled microsperes of 15/~m immediately after, as well as 4 days after grafting. The developing oedema of the graft was estimated by weight gain, histological and electron microscopic investigation. Residual exocrine function of the graft was measured by determining the serum lipase and amylase levels.Results: The autotransplantation of an segmental pancreatic graft to the neck of a dog is a feasable and technically easy surgical procedure, without major Venalot -treated dogs local complications. In the ® there was a higher weight increase of the graft in comparison to the controls. As tested so far there was no difference in the behaviour of the I~lood flow of the grafts in both experimental groups. There was a significant increase of serum amylase and lipase values in the Venalot®-trea'ted group. Conclusion: The technique of cervical segmental pancreatic transplantation in dogs is recommended in cases where immunological monitoring (aspiration cytology, biopsies) is the aim of the study (good access to needles). The technique of microspheres injection is a useful method for the examination of the blood flow in segmental pancreatic grafts in dogs. The prolamine-induced secretion-oedema of ductoccluded pancreatic grafts is resistant to Venalot ®treatment. Graft pretreatment has been used in various organs to prolong aUograft survival. We have recently demonstrated that graft pretreatment of canine renal allografts with Cyclosporin A (Ca A) led to improved animal and graft survival. Our present study assesses the effect ofCy A as a graft pretreatment on pancreatic islet cell allografts. Pancreases were removed from unrelated donor mongrel dogs and placed in iced saline (4°C) after the collagenase digestion. The tissue fragments were washed once more and then another 5 mg Cy A was added to the preparation prior to intrasplenic injection of the islet cell allografts. Three groups were studied: Group I (n= 15) served as pancreatectomized non-transplanted controls, Group II (n = 15) received a non-pretreated islet cell allograft after total pancreatectomy and Group III (n= 10) received a Cy A graft pretreated islet cell transplant after total pancreatectomy. All animals were given minimal immunosuppression with azathioprine (5 rng/kg/day x 3, followed by 2,5 mg/kg/ day). Blood glucose values were monitored to determine engraftment and subsequent rejection. Hyperglycemia was considered when plasma glucose values rose above 150 mg/dl. All pancreatectomized controls (Group I) became hyperglycemic by the first post-operative day. Non-pretreated islet cell allografts in Group II had variable function and became hyperglycemic between one and nine days after transplantation. Only five of ten animals in Group III, receiving Cy a pretreated islet cells, became hyperglycemic levels greater than 90 days and one died of unknown causes immediately after transplantation. The following table presents the animal survival data for the 90 day follow-up period.Although the exact mechanism of action is not known, this study indicates that Cy A graft pretreatmerit can be beneficial in prolonging pancreatic islet cell allograft survival, Further studies will optimize the use ofCy A in this application and hopefully contribute to the improvement of pancreatic islet cell transplantation. Grafting of vascularised organs has become a standard procedure in surgical research. However, only few data of the fate oforthotopicliver transplantation in the rat are available, probably because of the difficult ~iargery involved. This communication gives an account of the outcome of 40 liver allografts and 16 liver isografts in four inbred strains of rat. The following groups of allo and isografts are formed; the survival time is given in days postoperatively.The technical details of the operative procedure of orthotopic liver transplantation in the rat are described elsewhere (Eur Surg Res 13:236 (1981). The distribution of death among allallograft groups show 3 blocks of survivors. The first block includes animals surviving up to 30 days. Acute rejection and infection are the diagnosed causes of death in this compartment. The second block includes animals surviving up to 110 days. In the grafts, which belong to this compartment a intrahepatic bile duct proliferation is a frequent and dominant histological feature. Because of the development of identical lesions in isograft surviving the same periode and the induction of this lesions by common bile duct ligation experiments, chronic rejection is excludet as the cause of death. The third block includes all so called long-term survivors, in which the structure of the grafts are remarkably well preserved and very few abnormalities are present. The occurence of fatal acute rejection and long-term survival in each allograft group demonstrates, that the fate of the orthotopically transplanted livers in not dependent upon