Outside Japan portosystemic shunts have been favored as the surgical procedure of choice for the management of portal hypertension of noncirrhotic etiology. Devascularization procedures have resulted in high rebleed rates probably owing to a limited extent of devascularization. We performed this study to assess the efficacy of our modification of Sugiura's procedure for long-term control of variceal bleeding in patients with noncirrhotic portal hypertension. Forty-six patients with extrahepatic portal venous obstruction (EHPVO) and 22 with noncirrhotic portal fibrosis (NCPF) were subjected to transabdominal extensive esophagogastric devascularization with esophageal or gastric stapled transection (modified Sugiura's procedure), 38 in an emergency situation and 30 electively. Follow-up endoscopies were performed every 6 months. Operative mortality, morbidity, variceal status, and causes of recurrent bleeding were evaluated. The postoperative mortality was 4%. Early procedure-related complications were seen in 6%, and esophageal strictures formed in 7 of 45 survivors undergoing esophageal transection (15%). Over a mean +/- SD follow-up of 53 +/- 34 months, 95% of patients were free of varices. Seven survivors (11%) had a rebleed, but only 5% were due to varices (two esophageal, one gastric). Six (9%) patients developed gastropathy. The 5-year survival was 88%. The modified Sugiura's procedure is safe and effective for long-term control of variceal bleeding especially in the emergency setting and in patients with anatomy unsuitable for shunt surgery or if surgical expertise for a shunt operation is not available.

Transabdominal extensive oesophagogastric devascularization combined with gastrooesophageal stapling is an effective and safe procedure for control of acute variceal haemorrhage with satisfactory long-term control, especially in patients without cirrhosis and low-risk patients with cirrhosis.

Transabdominal extensive oesophagogastric devascularization combined with gastrooesophageal stapling is an effective and safe procedure for control of acute variceal haemorrhage with satisfactory long-term control, especially in patients without cirrhosis and low-risk patients with cirrhosis.

Of 104 patients with portal hypertension who were subjected to oesophageal variceal sclerotherapy, gastric varices were seen in 81 (78 per cent) at endoscopy and 69 (74 per cent) at splenoportography. In 50 (48 per cent) patients gastric varices were seen at the initial endoscopic examination and in 31 they developed during follow-up at intervals varying from 1 to 56 weeks. Gastric varices were seen significantly more often along the lesser curvature than in the gastric fundus and the left gastric vein was the main feeding vessel in 75 per cent of cases. Varices bled in nine of 81 patients and bleeding was seen significantly more often from fundal varices (30 per cent) than from lesser curve varices (5 per cent) (P less than 0.02). The incidence of gastric varices is high, and contrary to popular belief they are more often located along the lesser curvature of the stomach than in the gastric fundus.

No abstract