Oesophageal manometry in patients with varices and following oesophageal transection

Comparison of postoperative results following terminal esophagoproximal gastrectomy and esophageal transection for esophageal varices

The results of 44 terminal esophagoproximal gastrectomies (TEPG) and 53 esophageal transections (ET) for esophageal varices, performed during the period between January, 1975 and March, 1989, were retrospectively compared. The results examined prognosis, recurrence of esophageal varices and late postoperative complications. The 5-year survival rates for patients who underwent selective or prophylactic surgery were 85.9% following TEPG and 81.6% following ET. However, the 10-year survival rate for the former group was significantly lower than that for the latter group at 59.3% versus 70.0% (P < 0.05) because of the number of deaths due to hemorrhage and liver failure caused by anastomotic ulcers. The respective 5-year recurrence rates of varices for the TEPG and ET groups were 18.4% and 26.4%, respectively, while the 10-year recurrence rate for the former group was again significantly lower than that for the latter group at 27.1% versus 53.7% (P < 0.01). As for postoperative late complications, reflux esophagitis and/or anastomotic ulcers were found twice as frequently after TEPG as after ET. Thus, although TEPG was more effective for preventing variceal recurrence it left the potential for an anastomotic ulcer to develop, which was the dominant cause of death more than 5 years after surgery.

Bleeding from staple line erosion after esophageal transection: effect of omeprazole

Esophageal staple transection effectively controls acute variceal bleeding, but up to 50% of these patients will have recurrent upper gastrointestinal bleeding. In our experience, most of these bleeding episodes are caused by total or partial circumferential ulceration at the level of the staple transection: staple line erosion. It caused rebleeding in 29 (40%) of our patients. Whereas the pathogenesis of this lesion is unknown, acid reflux is a consequence of transection surgery. Assuming that staple line erosion could be healed by acid suppression therapy, thereby preventing recurrent bleeding, an acid suppression regimen was evaluated prospectively in 24 patients. Only six (25%) healed with daily standard (300 mg) or high-dose (1,200 mg) ranitidine combined with sucralfate (4 gm). The remaining 18 (75%) healed after omeprazole administration (40 mg/day) for 1 mo. Maintenance ranitidine alone (300 mg/day) was introduced, but 11 (48%) had relapse of erosions. All 11 healed with omeprazole (40 mg/day) for 2 mo, but again on maintenance ranitidine, 10 relapsed. All healed with further omeprazole and healing persisted with long-term administration (20 mg/day). Fifteen rebleeding episodes occurred in eight patients on maintenance ranitidine. Whereas relapse of staple line erosions did occur in the absence of rebleeding, all rebleeding episodes were associated with the relapse of staple line erosion. Omeprazole is more effective than ranitidine alone and combined with sucralfate in healing staple line erosion. Omeprazole prevents rebleeding, which may enhance the long-term benefits of staple transection for acute variceal bleeding.

Staple-line erosion: a common source of recurrent bleeding following stapled oesophageal transection

Recurrent bleeding after stapled oesophageal transection was studied in 73 patients with cirrhosis transected for acute variceal bleeding. The most frequent source of bleeding was partial or total circumferential oesophageal erosion at the transection: staple-line erosion. This lesion occurred in 36 (49 per cent) patients and was the source of rebleeding in 29 (40 per cent) patients with 54 episodes. Rebleeding in 22 (30 per cent) patients was due to varices in nine (12 per cent), peptic ulcer in six (8 per cent), gastric erosions in two (3 per cent) and unknown sources in five (7 per cent), accounting for 33 episodes. The mean(s.e.m.) blood transfusion requirement for bleeding from staple-line erosions were 1.5 (0.25) units per bleed versus other sources, 6.5(1.0) units per bleed (P less than 0.001). Staple-line erosion was present at the first postoperative endoscopy in 11 (15 per cent) patients but the time to appearance varied widely. The lesion was more common in patients with Pugh's grade A liver disease at the time of transection, reflecting the increased survival rate of these patients. Staple-line erosion is a common source of minor recurrent bleeding following stapled oesophageal transection.

Transabdominal gastro-oesophageal devascularization and oesophageal transection for bleeding oesophageal varices

Fifty consecutive unselected patients presenting with acute haematemesis and melaena from endoscopically confirmed bleeding oesophageal varices were treated by one-stage transabdominal oesophageal transection with gastro-oesophageal devascularization and splenectomy. According to Child's classification, 15 were Grade A, 29 Grade B and six Grade C. The mortality rate was 10 per cent. Postoperative complications included gastric fundal leak (4 per cent), pleural effusion (4 per cent), subphrenic abscess (2 per cent), atelectasis (2 per cent) and pneumothorax (2 per cent). Although five patients (10 per cent) complained of transient dysphagia during follow-up, only one (2 per cent) presented evidence of stricture at the site of oesophageal transection. Clinical encephalopathy was not present in the surviving patients in the postoperative period. No recurrence of bleeding has been recorded in the surviving patients over a follow-up period of 2-3 years.

Prolonged ambulatory pH monitoring in patients following oesophageal transection and control subjects

Prolonged ambulatory oesophageal pH monitoring using a radiopill was carried out in 19 patients following oesophageal transection for varices. The results were compared with 14 asymptomatic control subjects. The duration of oesophageal pH less than five while erect was significantly greater in the post-transection group than in the control subjects. The duration of pH less than 4 while erect was also greater following transection but this did not quite reach statistical significance. The number of reflux episodes did not differ between the two groups. There was no significant difference in duration or number of reflux episodes while supine between the two groups.