Portal hypertension: an investigation into its clinical features, with especial reference to chronic disease of the liver

1. Haematemesis and melaena were the principal symptoms in 60 per cent of the cases of portal hypertension treated in the Professorial Surgical Unit of the Royal Infirmary, Edinburgh. The source of the bleeding was the lower end of the oesophagus and the cardia and fundus of the stomach, where the subepithelial and subglandular plexuses form part of one important anastomotic link between the portal and the systemic venous circulations when there is obstruction to the normal flow of portal blood through the liver. Because it is separated only by epithelium from the lumen of the oesophagus, the subepithelial plexus has been classified as a dangerous or vulnerable type of collateral circulation, whereas the anastomoses which occur in the falciform ligament and retro - peritoneally lie deep in the tissues and are entirely beneficial.2. The presence of such oesophageal varices is regarded as the major indication for surgical intervention, but statistical analysis does not show that the survival rate after operation is significantly better than after medical treatment alone. Possible reasons for this paradox have been discussed, and particular stress has been laid upon the selection of patients for operation and the selection of the time for operation in individual cases.3. The other indications for operation in portal hypertension are a depression of the numbers of circulating granulocytes or platelets sufficiently severe to be causing symptoms or interfering with recovery from intercurrent illness, and splenomegaly which is already painful or is so large as to be a potential source of danger to the patient. There is some evidence to suggest that once the pattern of the collateral circulation is determined in the individual case, it does not change. Hence, if such symptoms occur in a patient in whom there is no history of haematemesis and no demonstrable oesophageal varices, splenectomy may be performed as the operation of deliberate choice. That splenectomy alone is not the elective operation in patients who have been bleeding before operation or who have demonstrable varices in the oesophagus is indicated by the high percentage of such cases in which haematemesis recurs.4. Experience with limited oesophago-gastrectomy in the treatment of portal hypertension is still small, and the time which has elapsed since surgery is too short to permit other than tentative opinions about its usefulness. It may have a more adverse effect upon the functional efficiency of the cirrhotic liver than portal -systemic anastomosis, especially when it is followed by difficulty in swallowing or regurgitation of food. At present it appears to be indicated in children, in young people and in any patient who has had recurrence of bleeding after other forms of surgical !intervention.5. In cases with hepatic cirrhosis, operation is always followed by deterioration in hepatic function, the degree and duration of the changes in the biochemical tests depending principally upon the pre- operative condition of the liver. If this period of functional depression is successfully weathered, tests done at intervals of a year or more after operation indicate that the functional efficiency of the liver at that time is not significantly different from its preoperative state, irrespective of the type of operation performed. It is because of the immediate impact of surgery upon the decompensated liver that there is so high a mortality after operation in cases with ascites, for ascites is typicallyi a sign of hepatic failure. Such cases should be treated by medical measures and should be submitted to portal -systemic venous anastomosis only when no further benefit can be expected from medical treatment alone and when there is the additional risk of haemorrhage from oesophageal varices.6. The propriety of any form of surgery in portal hypertension associated with hepatic cirrhosis has been discussed. The two outstanding causes of death in these cases are gastrooesophageal haemorrhage and hepatic failure. Haemorrhage is the principal cause of death in nearly 50 per cent of cases not submitted to surgery. Frequently the amount of blood lost is small compared, for example, with the loss from a bleeding peptic ulcer, but the effect of even temporary hypotension and hypoxia on the already damaged liver may be an extensive and fatal parenchymal necrosis. The prognosis in cirrhosis is evidently worse where the possibility of haemorrhage from oesophageal varices exists, and this is the feature which it is the prime aim of surgery to relieve.7. Further improvement in the results of treatment of Portal Hypertension depends less upon the elaboration of surgical techniques than upon better assessment of the functional efficiency of the liver and the development of quicker and more effective methods of preparing for operation the patient in whom severe hepatic dysfunction exists along with clinical evidence of portal hypertension