Chinese Medical Journal

eee eae

PORTAL HYPERTENSION AT

revealed a large amount of non-coagulated blood accumulated in the peritoneal cavity, for which we could find no satisfactory explanation. In the second case a left curved subcostal incision was used and the spleen was removed without much difficulty; but signs of internal hemorrhage appeared about twenty-four hours later and the patient quickly went into shock. Exploration was carried out immediately and a spurting artery over the under surface of the diaphragm was found and controlled by transfixing ligatures. The postoperative course after the second operation was uneventful. While the bleeding in the first case seemed to be unrelated with the type of approach, the bleeding in the second case probably could have been avoided, if a thoraco-abdominal approach had been used. The thoraco-abdominal incision, we believe, gives the best exposure and facilitates the freeing of the spleen and control of bleeding. Vhen the curved subcostal incision is used, a careful search for bleeders, especially under the diaphragm, should be made and the abdomen should only be closed after adequate control of bleeding has been achieved.

As shown in Table 14, the infection rate among our patients was high. Patients with portal hypertension usually have poor reparative power and low general resistance, because of frequent liver damage. Hence, special precaution should be taken against infection before and after operation.

Complications that occurred less frequently included subdiaphragmatic abscess in 2 cases, epidemic parotitis in 3 and wound disruption in 1. Wound disruption was a rare complication in this series. We consider that retention sutures are in general not required unless marked impairment of liver function is shown before operation.

RESULTS OF OPERATIVE TREATMENT

Mediastinotomy. Mediastinotomy with packing of iodoform gauze around the esophagus was done in 2 cases. The purpose of this operation was to promote the formation of abundant collaterals between the coronary and azygos veins in order to reduce the portal pressure. No distinct effect was observed in either case immediately after operation. In one of these cases, the patient after two attacks before operation had two more attacks about two months after operation and was admitted to another hospital where splenectomy was done. He returned to work soon afterwards and remained symptomless until six months later, when he died of hepatic coma following another attack of hemorrhage. Since no favorable results were observed with mediastinotomy, we made no more trials.

Ligation of hepatic artery. Ligation of the hepatic artery was done in 2 cases of portal cirrhosis. Both cases had a history of repeated attacks of profuse hematemesis before operation. In the first case, ligation