Ranitidine, 300 mg daily, was given to 92 patients with duodenal ulcer (DU), 38 with prepyloric ulcer (PPU), and 21 with gastric corporeal ulcer (GCU). The healing rates at 4 weeks differed for the different types of ulcers (P less than 0.01), being 91% for DU, 68% for PPU, and 81% for GCU. After established ulcer healing, maintenance treatment with either ranitidine, 100 mg twice daily or 150 mg at night, or placebo was given for 1 year or until ulcer relapse in a total of 108 patients--71 with DU, 24 with PPU, and 13 with GCU. There were no significant differences in relapse rates between the two groups treated with active drug or between the three ulcer groups. However, the overall relapse rate in the active drug groups was 16%, against 72% in the placebo group (P less than 0.001).

In 93 patients with cirrhosis of the liver and portal venous hypertension the main tributaries of the portal vein were examined by percutaneous transhepatic catheterization. The appearance and degree of porto-systemic collaterals were analysed. Esophageal varices were demonstrated in 82 patients. No correlation was found between the portal venous pressure and the extent of porto-systemic communications.

Percutaneous transhepatic cholangiography (PTC) with subsequent external bile drainage by nonsurgically established percutaneous transhepatic intubation of bile ducts was performed in 105 patients with obstructive jaundice. Recovery of liver function and improvement in the patients' general condition prior to radical or palliative surgery, nonsurgical palliation in advanced cases of malignancy as well as relief of postoperative leakage from a biliodigestive anastomosis are the indications for the bile drainage technique used in the present study. Clinical aspects such as optimal period of preoperative drainage, frequency of catheter dislodgement, and rate of complications such as cholangitis, bile leakage to the abdominal cavity and risk for peritoneal hemorrhage are discussed. Two deaths occurred within this series.

Portacaval anastomosis in rats elevated the brain serotonin content by about 30%, the greatest increase being noted in the brain stem. A minor increase in norepinephrine content but no change in dopamine concentration was also noted. Low dose L-dopa treatment reversed the serotonin changes but did not affect catecholamine concentrations. Protein load tended to lower most transmittor concentrations without clinically affecting the animals. The formation of 5-HT from tryptophane in vitro in rat cortical slices was reduced by about 35% whereas the formation of 5-HIAA in the same slices was increased with about 60%. This in vitro study indicates a change in the activity of serotonin neurons caused by the protacaval shunt. The findings suggest that changes in brain serotonin metabolism and brain serotonin neurons may be an explanation of hepatic encephalopathy.

The predictive value of 15 pre- and peroperative parameters upon survival after portacaval shunt was analyzed in a retrospective investigation of 134 elective operations. A multiple linear regression model was used. Survival was measured at three different points of time: one month, one year and five years after the operation. Survival at one month was influenced by the parameters bilirubin/s and ascites only. Survival at one year was influenced by albumin/s, sex, bilirubin/s, BSP, heart disease history and ascites. Survival at five years was influenced by albumin/s, alkaline phosphatase/ s, history of alcohol abuse, and globulin/s. These findings indicate that prediction of survival after portacaval shunt is an intricate process and that considerable improvement of the Child criteria is possible.

56 consecutive patients with portal hypertension were studied with percutaneous transhepatic portography and the results were correlated to clinical findings and the number of upper gastrointestinal haemorrhages and the size of the individual bleeding. An abundance of collateral paths was noted in most patients. No regularity in development of these collaterals was found. It was not correlated to liver disease etiology, sex or liver function parameters. Portal pressure was not correlated to the size or amount of collaterals. In four patients with liver cirrhosis hepato-fugal flow in one segment of the liver was noted proving that portal flow through the liver is not uniform in this disease. The size of the haemorrhages was only correlated to presence of hepato-fugal flow in the main stem of the portal vein. It was not correlated to the estimated size of the oesophageal varices or to portal pressure. Percutaneous transhepatic portography seems to be of little help in selecting "high risk bleeders" in portal hypertension. Other factors may be of greater help in this task as indicated by the findings in this investigation that patients with alcohol cirrhosis had larger haemorrhages than those with cirrhosis of another etiology and that patients with none or few bleeding episodes had higher thrombocyte count than those with several haemorrhages.

16 patients with cutaneous or subcutaneous melanoma recurrence on an extremity were treated with regional perfusion with Melphalan. 18 perfusions were performed on 15 patients with stage II disease, that is with tumor growth restricted to an extremity including possible regional node metastases. All patients except two had new recurrences within the observation time. However, many of the patients had been treated surgically for recurrences once or several times previously. By comparing the length of the recurrence-free period following surgery alone with that following surgery plus perfusion in the same patients it was shown that perfusion treatment gave a significant extension of the recurrence-free time. Four perfusions were performed on patients in stage III, that is those with distant metastases. These perfusions gave a moderate or good temporary palliation as regards to tumor growths on the extremity. The traditional treatment for melanoma recurrences on an extremity has been surgical excision or less often amputation. An analysis of the literature shows that perfusion, usually combined with excision, seems to give definitely better results than surgical excision alone. There is evidence to suggest that perfusion treatment is even superior to amputation as regards survival; if so an immunological mechanism might be responsible for this effect.

Endogenous production of carbon monoxide (VCO), total and direct reacting serum bilirubin (TSB, DRB) were determined in 26 patients with liver cirrhosis and portal hypertension to evaluate the effect of various shunt operations on total heme catabolism. The material was divided into 3 groups. In group I, 11 patients not operated upon, mean VCO (+/- S.D.) was 18.4 +/- 6.0 micronmol/mmol total body heme per day (reference value 12.6 +/- 2.9). In group tii, 7 patients operated upon with subcutaneous transposition and a subtotal resection of the spleen, mean VCO (14.4 +/- 4.7) was not significantly raised. In group III, 8 patients operated upon with a modified distal splenorenal shunt, the highest mean VCO (26.1 +/- 9.0) was found. Mean TSB in the three groups was 34.8 +/- 29.2, 11.2 +/- 3.0, and 46.4 +/- 41.0 micronmol/l, respectively, and mean DRB 18.2 +/- 20.8, 3.7 +/- 1.0, and 26.8 +/- 34.1 micronmol/l, respectively. Estimated from preoperative laboratory values there was no difference in liver function between the three groups. The conclusion drawn is that heme catabolism, increased by 50% in liver cirrhosis complicated by portal hypertension probably due to a slight decrease in erythrocyte survival, tends to normalize after subcutaneous transposition and subtotal resection of the spleen. After spleno-renal shunting, on the other hand, a further increase in heme catabolism is seen. And so the increase in serum bilirubin often seen after the latter type of surgery is mainly related to a raised bilirubin production and not to a further decrease in liver function.

Twenty-five patients were treated with a distal spleno-renal shunt modified after that of Warren. The operative mortality was 4/25. One patient had an early thrombosis. All post-operative angiography otherwise showed patent shunts. After a median observation time of 43 months, 10/20 patients included in the followup were dead. The chief cause of death was liver failure. Encephalopathy has been common although generally of minor degree Hypersplenism, judged by thrombocyte count, was not significantly affected by the operation. Six of 21 patients have had gastrointestinal hemorrhage after the operation but no hemorrhage proved fatal. Postoperatively esophageal varices size was considerably diminished in most cases as judged by contrast x-ray. Ascites has not been a problem in this series. Postoperative angiography showed a marked and rapid reduction of portal blood flow to the liver with progressively more blood deviated through the coronary vein towards the shunt. This reduction in portal flow is a possible explanation of the high frequency of postoperative liver failure. This version of the distal spleno-renal shunt has probably no advantages over the portacaval shunt.

Percutaneous transhepatic catheterization of the portal vein was performed in 21 patients with liver cirrhosis and esophageal varices. Coronary and short gastric veins were selectively catheterized and obliterated. The examination was performed to stop bleeding in 6 patients and was successful in 5. Follow-up examination showed recanalization of previously obliterated veins in 13 of 16 patients.

Psychometric tests were performed in 41 patients with cirrhosis of the liver and suspected hepatic encephalopathy and compared with EEG-examinations and clinical investigations. Marked intellectual impairment was noted frequently even when the clinical investigation was normal. This difference was mainly due to the preserved verbal ability of the patient. The etiology of the cirrhosis did not influence the test results. Male cirrhotic patients with and without alcoholism showed significantly more intellectual impairment then alcoholics without cirrhosis. Patients Patients with constructed porto-systemic shunts showed only slightly reduced intellectual ability compared to those without shunts.

The histochemical enzyme pattern in normal, extra-hepatic bile ducts from rabbits was studied. A difference between the duct epithelium and the crypt epithelium was noted mainly in the activity of gamma-glutamyl-transpeptidase and alkaline phosphatase which only stained positively in the crypts. No difference from the normal enzyme pattern was noted after diversion of the bile flow for up to 30 days.

The healing process of longitudinal and transverse incisions of the common bile duct in rabbits was studied by identification of 10 different enzymes in the duct wall at various times after the operation. Bile was deviated during the healing. An almost complete disappearance of the subepithelial alkaline phosphatase activity was noted early after longitudinal incisions in contrast to unchanged alkaline phosphatase activity after transverse incisions. Transversely incised ducts almost uniformly developed stricutre formation. The size of the lumen after longitudinal incisions was generally normal. The difference in enzyme pattern following the two types of lesion might reflect a different response of the tissues to the two different types of trauma.

Distal splenorenal shunts were made in 25 patients with cirrhosis of the liver and portal hypertension. Angiograms were obtained in 16 patients at one week, three and six months, and at one year postoperatively. Portal vein flow had diminished in all patients compared with flow seen in preoperative angiograms. Hepatofugal flow developed in 6 patients during the follow-up period, and in 2 patients only one week postoperatively. The shunt remained patent in all of the patients but one.