Release of somatostatin, neurotensin and vasoactive intestinal peptide upon inhibition of gastric acid secretion by duodenal acid and hyperosmolal solutions in the conscious rat.

The inhibitory effect of duodenal exposure to acid and hyperosmolal solutions on pentagastrin-stimulated gastric acid secretion was studied in conscious rats equipped with chronic gastric fistula and duodenal Thiry-Vella loop. The loop was challenged with saline, HCl or hyperosmolal polyethylene glycol. Gastric acid secretion was measured in samples from the gastric fistula. Gut peptide concentrations were measured in duodenal perfusates collected each 30 min, and in plasma samples collected both during stimulated acid secretion alone, and at the end of experiments in combination with luminal challenges of the loops. During pentagastrin-stimulated gastric acid secretion, luminal perfusion of the duodenal loop with acid caused inhibition of acid secretion (P < 0.001) and a prominent release of somatostatin both to the lumen (P < 0.001) and to the circulation (P < 0.05). Also, neurotensin (P < 0.01) and vasoactive intestinal peptide (P < 0.01) were released to the lumen, but not to the circulation. Upon perfusion of the duodenal loop with hyperosmolal polyethylene glycol, acid secretion was inhibited (P < 0.05) and somatostatin alone was released to the luminal side (P < 0.01). In conclusion, duodenal exposure to acid inhibits pentagastrin-stimulated gastric acid secretion and releases SOM to the circulation that may directly inhibit acid secretion. Concomitantly, somatostatin (SOM), neurotensin and vasoactive intestinal peptide are released to the lumen. Duodenal exposure to hyperosmolal polyethylene glycol inhibits acid secretion with a luminal release of SOM only. Thus, luminal acid and hyperosmolal solutions inhibit gastric acid secretion by separate mechanisms. After acid or hyperosmolal challenge, the release of SOM to the circulation indicates gastric acid inhibition in an endocrine manner, while a luminal release of gut peptides indicates a local peptide overflow that might be of importance via paracrine regulatory mechanisms in the intact animal.

Transforming growth factor-alpha and epidermal growth factor inhibit gastric acid secretion and stimulate release of somatostatin and neurotensin in the conscious rat.

The study compared inhibitory actions of transforming growth factor-alpha (TGF alpha) and epidermal growth factor (EGF) on gastric acid secretion and effects of these peptides on release of gut peptides considered important for acid inhibitory and gastrointestinal protective mechanisms. TGF alpha and EGF did not affect basal acid secretion, but inhibited pentagastrin-stimulated acid secretion in a dose-dependent manner from 0.10 to 1.7 nmol kg-1 h-1 i.v. by maximally 72% for TGF alpha (P < 0.001) and 76% for EGF (P < 0.001). At the highest doses, TGF alpha and EGF caused 194% and 698% increase of somatostatin-like immunoreactivity (SOM-LI) in plasma, respectively (each P < 0.05). Neurotensin-like immunoreactivity (NT-LI) increased 438% by EGF (P < 0.05), but the increase of 700% with TGF alpha did not reach statistical significance. The levels of vasoactive intestinal peptide-like immunoreactivity (VIP-LI) did not change. In gastric juice, SOM-LI increased 80% by TGF alpha i.v. (P < 0.05), but NT- and VIP-LI did not change. EGF i.v. had no effects on levels of SOM-, NT- or VIP-LI in luminal juice. Thus, TGF alpha and EGF inhibit acid secretion, but also promote the release of SOM and NT into the circulation and may be involved in the acid inhibitory effects of these growth factors.

Ten-year follow-up of a prospective, randomized trial of selective proximal vagotomy with ulcer excision and partial gastrectomy with gastroduodenostomy for treating corporeal gastric ulcer.

Between 1975 and 1980, 30 patients with type I corporeal gastric ulcer were randomly allocated to undergo selective proximal vagotomy with ulcer excision or partial gastrectomy with gastroduodenostomy. Sixteen patients underwent selective proximal vagotomy (1 was excluded from the follow-up since microscopic examination of the excised ulcer revealed an early gastric cancer) and 14 underwent partial gastrectomy. No significant differences in the clinical results were found 3 years after surgery. During a median follow-up of 10 years, ulcer recurred in 3 patients after selective proximal vagotomy and in 2 after partial gastrectomy. One patient in each group had recurrent ulcer without symptoms and received no treatment. Two selective proximal vagotomy patients and three partial gastrectomy patients had epigastric pain with or without ulcer. One patient with selective proximal vagotomy underwent a second operation because of epigastric pain and recurrent ulcer. Bowel habits remained unchanged in all but one patient in each group, and mild or moderate dumping was recorded for two patients in each group. Very good or good results (modified Visick scale) were recorded for 11 of 15 patients after selective proximal vagotomy and for 10 of 14 patients after partial gastrectomy. Except for one patient in each group who had moderate dumping, patients classified as Visick III or IV had no symptoms during treatment with antacids or H2-blockers, or had asymptomatic ulcers and needed no treatment. Selective proximal vagotomy reduced the median acid response to insulin hypoglycemia and to pentagastrin by 100% and 80%, respectively, for at least 3 to 5 years, and partial gastrectomy reduced the median acid response to pentagastrin by 97%. In our opinion, selective proximal vagotomy with ulcer excision is an alternative to partial gastrectomy for surgically treating type I gastric ulcer.

Gastric intestinal metaplasia eleven years after randomized selective proximal vagotomy for peptic ulcer.

The presence of intestinal metaplasia (IM) 11 years after selective proximal vagotomy (SPV), selective proximal vagotomy with pyloroplasty (SPV + PP) and selective vagotomy with pyloroplasty (SV + PP) was investigated in 38 consecutive patients. IM was significantly more frequent in SPV than in SV + PP, SPV + PP or in unoperated controls of matching ages. IM occurred more frequently both at an older age (> or = 60 years) in SPV and in a larger number of gastric areas than in the other group of patients. Reports in the literature indicate that vagotomy may increase the risk of gastric carcinoma and that IM may antedate malignant transformation. It would thus appear that patients previously operated with SPV (without pyloroplasty) having IM, should be the group of patients to be enrolled in endoscopical surveillance programs for detection of possible cancer development.

Six-year results of a prospective, randomized trial of selective proximal vagotomy with and without pyloroplasty in the treatment of duodenal, pyloric, and prepyloric ulcers.

In a consecutive series of patients with uncomplicated prepyloric, pyloric, or duodenal ulcer, 39 patients were randomly allocated to selective proximal vagotomy with pyloroplasty, and 40 patients to selective proximal vagotomy alone with no operative mortality. Before surgery, all patients had undergone H2-receptor antagonist treatment. No patient was lost for follow-up. At an average follow-up of 6 years, recurrent ulcer was recorded in 15% and 20%, respectively, after selective proximal vagotomy with and without pyloroplasty. Three of 14 recurrent ulcers were asymptomatic. Epigastric pain with or without ulcer was significantly less common after selective proximal vagotomy with (13%) than without pyloroplasty (40%). Mild diarrhea or mild dumping was recorded in a few patients. The overall results were very good or good (Visick I or II) in 77% and 55% (significant difference) after vagotomy with and without pyloroplasty, respectively, and in 82% and 58%, if asymptomatic ulcers were graded as Visick I or II results. Of the 27 patients with Visick III or IV results, three patients needed no treatment (asymptomatic ulcers), and 10 patients had no symptoms during medical treatment. Two patients with vagotomy and pyloroplasty and nine with vagotomy alone were reoperated. There were no deaths, and the results were graded as Visick I or II in 10 patients and as Visick III in one patient. The authors conclude that selective proximal vagotomy with pyloroplasty is superior to vagotomy alone for the treatment of prepyloric-pyloric and duodenal ulcer. Recurrent ulcer after vagotomy has a benign course and responds well to ranitidine treatment.

Twelve-year follow-up of a prospective, randomized trial of selective vagotomy with pyloroplasty and selective proximal vagotomy with and without pyloroplasty for the treatment of duodenal, pyloric, and prepyloric ulcers.

Between 1973 and 1981, 161 patients with prepyloric, pyloric, or duodenal ulcers were randomly allocated to selective vagotomy with pyloroplasty, selective proximal vagotomy with pyloroplasty, or selective proximal vagotomy alone. No significant differences in clinical results were found 3 years after surgery by Emås and Fernström (Am J Surg 1985; 149: 236-42). There was one postoperative death, and one patient lost to follow-up. Of 159 patients, 52 underwent selective vagotomy with pyloroplasty, 55 selective proximal vagotomy with pyloroplasty, and 52 selective proximal vagotomy alone. Fifteen patients did not undergo endoscopy, but they had no epigastric complaints. From 1 to 16 years after surgery, recurrent ulcer was detected in 13%, 18%, and 23%, respectively, after selective vagotomy with pyloroplasty, selective proximal vagotomy with pyloroplasty, or selective proximal vagotomy without pyloroplasty. Twenty-eight percent of the patients with recurrent ulcer had no symptoms and received no treatment. Sixteen patients died within 8 years after surgery of causes unrelated to the ulcer disease. At their final examination, 14 of the 16 patients had Visick I or II (modified Visick scale) results, and the disease that caused their deaths obscured evaluation in 2 patients. The remaining 143 patients were followed up for 8 to 16 years (average: 12 years). Epigastric pain with or without ulcer was recorded more often (significant) after selective proximal vagotomy alone (40%) than after selective vagotomy with pyloroplasty (17%) or selective proximal vagotomy with pyloroplasty (14%). Bowel habits were unchanged in 96% of patients who underwent selective vagotomy with pyloroplasty or selective proximal vagotomy with pyloroplasty and 100% of patients who had selective proximal vagotomy alone. Mild dumping tended to be more common after vagotomy with pyloroplasty but was a minor nuisance in only a few patients. Very good or good results (Visick I or II) were recorded in 75% of the patients after selective vagotomy with pyloroplasty or selective proximal vagotomy with pyloroplasty or selective proximal vagotomy with pyloroplasty and in 54% after selective proximal vagotomy alone (significant difference). Seventeen patients underwent reoperation with antrectomy and gastrojejunostomy Roux-en-Y (13 patients) or gastroduodenostomy (4 patients) with no mortality. The results of the reoperations were graded as Visick I or II results in all but one patient. The final grading, including the reoperations, were Visick I or II in 85% of patients after selective vagotomy with pyloroplasty and selective proximal vagotomy with pyloroplasty and in 55% after selective proximal vagotomy alone (significant difference).(ABSTRACT TRUNCATED AT 400 WORDS)

Fat inhibits pentagastrin-stimulated acid secretion from the duodenum but not from the proximal jejunum in chronic gastric fistula rats.

The effect of fat emulsion in the upper intestine on the maximal gastric acid response to pentagastrin was studied in chronic gastric fistula (GF) rats with a 4-cm blind loop of the duodenum anastomosed to the jejunum (Roux-en-Y). Fat emulsion in the loop inhibited the acid response by 85%. To localize the site of the inhibitory mechanism, GF rats were provided with Thirty-Vella loops of the duodenum (bile and pancreatic ducts transplanted to the proximal jejunum) or with Thirty-Vella loops of the proximal jejunum and a Roux-en-Y loop of the duodenum to prevent gastric juice from entering the duodenum. Perfusion of the duodenal loop with fat emulsion mixed with bile and pancreatic juice reduced the acid response by 49%, but perfusion of the proximal jejunal loop did not alter the response. It is concluded that the intestinal mechanism for inhibition of acid secretion by fat is located in the duodenum in rats.

Duodenogastric bile reflux before and after selective proximal vagotomy with and without pyloroplasty.

Duodenogastric bile reflux was determined by nasogastric aspiration in 60 patients with prepyloric, pyloric, or duodenal ulcer before and after selective proximal vagotomy (SPV) or SPV and pyloroplasty (PP). All patients underwent two aspiration tests before and 3 months, 1 year, and 3 years after operation. The highest 1-h output of bile acids was recorded, and 100 mumol x h-1 or more was considered a positive reflux test. Sixteen patients (27%) had positive tests before operation; seven of them had previously been cholecystectomized. Forty-four patients (73%) had negative tests preoperatively; one had been cholecystectomized. Of 43 non-reflux patients without previous cholecystectomy 21 were randomized to SPV and 22 to SPV + PP. Sixteen patients (76%) remained negative after SPV, and 19 of 22 (86%) after SPV + PP. The incidence of negative bile reflux before and after SPV or SPV + PP did not differ. Cholecystectomized patients had a high incidence of bile reflux before (seven of eight) and after (six of eight) SPV or SPV + PP. It is concluded that neither SPV nor PP causes bile reflux.

Duodenal acidification and jejunal hyperosmolality inhibit pentagastrin-stimulated acid secretion in chronic gastric fistula rats.

In chronic gastric fistula (GF) rats, HCl and a hyperosmolal solution of polyethylene glycol (PEG) in the upper intestine inhibit pentagastrin-stimulated gastric acid secretion by different mechanisms, but their anatomic sites have not yet been established. In the present study GF rats were provided with Thiry-Vella loops of the duodenum and the bile and pancreatic ducts transplanted to the proximal jejunum, or with Thiry-Vella loops of the proximal jejunum. In the latter rats the duodenum was anastomosed as a blind loop to the jejunum to prevent any gastric juice from entering the duodenum. Duodenal loop perfusion with 0.20 M HCl inhibited the acid response to pentagastrin by 62%, but perfusion with 1200 mOsmol x kg-1 of PEG solution did not alter the response. In contrast, acidification of the proximal jejunal loop did not alter but hyperosmolality inhibited the response by 41%. The study shows that the mechanism for inhibition by intestinal acidification is confined to the duodenum and that for inhibition by hyperosmolality is located in the proximal jejunum--but whether only to the proximal part is unknown.

Comparison of gastric aspiration and HIDA scintigraphy in detecting fasting duodenogastric bile reflux.

Gastric aspiration and HIDA scintigraphy have been compared to assess duodenogastric bile reflux. Gastric aspiration was performed on two separate occasions with a total examination time of 3 h. The highest 1-h output and the highest concentration of bile acids were recorded. HIDA scintigraphy was carried out for 90 min after an injection of 60 MBq 99mTc-dimethyl-iminodiacetic acid (HIDA). Forty-six patients with different gastrointestinal disorders were studied; 24 patients were positive and 13 negative in both tests. Accordingly, the methods agreed in 37 to 46 patients (80.4%). It is concluded that gastric aspiration is as reliable as HIDA scintigraphy to assess fasting bile reflux.

Acid in proximal and distal duodenum inhibits, but hyperosmolal solution does not inhibit pentagastrin-stimulated acid secretion in chronic gastric fistula rats.

In chronic gastric fistula (GF) rats, hyperosmolal 0.20 M HCl infused into a duodenal loop anastomosed to the jejunum (Roux-en-Y) produced a greater inhibition of the maximal acid response to pentagastrin than HCl or 1200 mosmol kg-1 solution of polyethylene glycol (PEG) alone, suggesting that HCl and hyperosmolal solution inhibit secretion by different mechanisms. In the present study on chronic GF rats with Thirty-Vella loops of the proximal or distal duodenum (bile and pancreatic ducts transplanted to the jejunum), perfusion of the proximal or distal loop with 0.20 M HCl significantly inhibited the maximal acid response to pentagastrin, but perfusion with hyperosmolal PEG solution did not alter the response. The results suggest different anatomical sites for the inhibitory mechanisms, sensitive to acid and hyperosmolal solution.

Intraduodenal instillation of acid and hyperosmolal solution suppresses pentagastrin-stimulated acid secretion but not histamine mobilization in the rat stomach.

In chronic gastric fistula rats provided with a duodenal loop anastomosed to the jejunum (Roux-en-Y), maximal stimulation of acid secretion by continuous intravenous infusion of pentagastrin produced a 15-fold increase of gastric histidine decarboxylase activity. Intraduodenal instillation of 0.20 M HCl or 1200 mOsm X kg-1 solution of polyethylene glycol (PEG) inhibited the maximal acid response to pentagastrin by 61% and hyperosmolal (1200 mOsm X kg-1 of PEG solution) 0.20 M HCl by 79% but did not suppress the pentagastrin-induced increase in enzyme activity. Hence, instillation of HCl and/or hyperosmolal PEG solution in the duodenum depresses the responsiveness of the parietal cells to pentagastrin without interfering with pentagastrin-induced activation of gastric histamine formation.

Antrum participates in the postprandial augmentation of the acid response to exogenous gastrin in conscious cats.

A humoral mechanism, potentiating the maximal acid, but not the pepsin response to exogenous gastrin in cats, is activated by protein-rich food in the stomach or duodenum, but not in the jejunum. In the present study, the effect of an oral meat meal on the maximal acid response to pentagastrin was investigated in Heidenhain pouch (HP) cats, and in antrectomized HP cats with duodenal exclusion and gastrojejunostomy Rouxen-Y. Antrectomy and duodenal exclusion abolished the postprandial HP acid response, and feeding did not potentiate the acid response to pentagastrin. The finding suggests that the gastrin-potentiating mechanism in the stomach is localized in the antrum, and it cannot be demonstrated in the jejunum. The basal plasma levels of gastrin and somatostatin did not differ in antrectomized and non-antrectomized cats. Antrectomy and duodenal exclusion abolished the postprandial gastrin and somatostatin responses. The plasma somatostatin increase during pentagastrin infusion persisted after antrectomy and duodenal exclusion. It is concluded that the antrum is not mandatory for the basal plasma levels of gastrin and somatostatin, but the postprandial gastrin response is of antral origin and may release somatostatin. Pentagastrin infusion releases extra-antral somatostatin in cats.

Gastric acid, plasma gastrin, and somatostatin responses to feeding and exogenous gastrin alone and in combination in conscious cats.

Meat in the stomach or duodenum potentiates pentagastrin-induced acid secretion in cats, presumably by a humoral mechanism. In the present study on cats, a meat meal significantly augmented the maximal acid response from a Heidenhain pouch (HP) to pentagastrin or to human synthetic gastrin I by 31 and 30%, respectively. The maximal HP acid response to pentagastrin was augmented also by peptone instilled into the stomach through a gastric fistula. Intravenous infusion of amino acids stimulated acid secretion but did not augment the maximal acid response to pentagastrin. The plasma concentrations of gastrin and somatostatin increased during infusion of pentagastrin and gastrin I and were not further altered by simultaneous feeding. The present results indicate that the mechanism for potentiation of gastrin-induced acid secretion is of physiological significance, since feeding augmented also the acid response to heptadecapeptide gastrin, the only gastrin secreted from the antrum and duodenum in cats. The potentiation of acid secretion is not dependent on the vagal excitation induced by oral feeding, since potentiation was demonstrated also by intragastric peptone instillation. The mechanism for the potentiation is not due to absorbed amino acids or a decrease of plasma somatostatin.

Medical principles for treatment of peptic ulcer.

The pathophysiology behind chronic gastric and duodenal ulcer is not fully understood, but may be explained as an imbalance between aggressive and defensive factors acting on or in the mucosa. Medical therapy may aim to reduce the aggressive factors acid and pepsin by compounds that neutralize the acid or inhibit the secretion of acid. Peptic activity is reduced by the high acid pH and also by the inhibition. Another group of compounds increases the mucosal defence by stimulating the secretion of mucus, bicarbonate, and growth of the mucosa, or by forming a protective layer on the ulcer crater. Some compounds affect both the aggressive and defensive factors. All modern compounds heal most ulcers in 4 to 6 weeks, but do not alter the natural history of ulcer disease, as indicated by the high rate of ulcer relapses after cessation of treatment.

Gastric acid response to pentagastrin and gastrin-releasing peptide in conscious cats.

The mammalian counterpart to bombesin, gastrin-releasing peptide (GRP), is considered to stimulate gastric acid secretion by release of endogenous gastrin. The present study was carried out to examine if GRP has a stimulatory action on acid secretion in addition to that produced by released gastrin. In 4 gastric fistula (GF) and Heidenhain pouch (HP) cats, 160 pmol X kg-1 X h-1 i.v. GRP produced a GF and HP acid response that amounted to 22 and 13%, respectively, of the maximal acid response to pentagastrin, with no rise in serum gastrin concentration. GRP significantly increased the maximal acid response to pentagastrin in the GF but not in the HP. These results suggest that GRP stimulates acid secretion in cats also by an action not related to the release of gastrin. This action is greater in the presence of vagal innervation.

Acid and pepsin responses to graded doses of pentagastrin in duodenal and corporeal gastric ulcer patients before and after selective proximal vagotomy.

The dose response curves for acid and pepsin output to increasing intravenous doses of pentagastrin (0.01, 0.02, 2, 8, and 16 micrograms/kg per hour) were determined in six male patients with duodenal ulcer and six with type 1 corporeal gastric ulcer before and 3 to 6 months after selective proximal vagotomy and excision of the gastric ulcer. The maximal secretory capacity (maximal response) of acid and pepsin was greater in the duodenal ulcer patients than in the corporeal gastric ulcer patients, but the sensitivity of the oxyntic and peptic cells to pentagastrin (the dose required for half the maximal response) was equal for the two ulcer groups. Selective proximal vagotomy reduced the acid response to insulin by 96 to 100 percent. The acid secretory capacity and the sensitivity of the oxyntic cells to pentagastrin was reduced by selective proximal vagotomy to the same extent in the duodenal ulcer patients and the corporeal gastric ulcer patients. Selective proximal vagotomy reduced the pepsin secretory capacity in the duodenal ulcer patients but did not reduce the already low capacity in the corporeal gastric ulcer patients. Selective proximal vagotomy decreased the sensitivity of the peptic cells in both ulcer groups. Similar results were obtained when the dose response curves were analyzed according to Michaelis-Menten kinetics. Our results justify clinical trials of selective proximal vagotomy with complete ulcer excision for treatment of type 1 corporeal gastric ulcer.

Acid and hyperosmolal solutions in the upper intestine of chronic gastric fistula rats inhibit gastric acid secretion by different mechanisms.

In chronic gastric fistula rats 0.20 M HCl and 1200 mOsm X kg-1 solution of polyethylene glycol (PEG) infused into a duodenal loop anastomosed to the jejunum (Roux-en-Y) produced maximal inhibition of pentagastrin-stimulated acid secretion, which amounted to 60% and 50%, respectively. In the present study in Roux-en-Y rats with gastric fistula, perfusion of the loop with hyperosmolal (1200 mOsm X kg-1 of PEG solution) 0.20 M HCl produced a greater reduction of the maximal response to pentagastrin (91% inhibition) than perfusion with 0.20 M HCl (64%), suggesting that HCl and hyperosmolal solution inhibit secretion by different mechanisms. The maximal acid response to histamine was more resistant to inhibition than that to pentagastrin; 0.20 M HCl inhibited secretion by 43%, 1200 mOsm X kg-1 of PEG solution by 42%, and hyperosmolal 0.20 M HCl by 60%. The results suggest that HCl and hyperosmolal solution also inhibit histamine-stimulated secretion by different mechanisms. The anatomical sites of the mechanisms remain to be established.

Effect of a protein meal in stomach and jejunum on gastric secretory response to pentagastrin in cats.

A humoral gastrin potentiating mechanism was previously demonstrated in the upper intestine of cats. The present study was carried out on 10 cats with a gastric fistula (GF) and a Heidenhain pouch (HP). In five of the cats the duodenum and proximal jejunum were excluded as a Roux-en-Y loop, and the other cats were provided with an isolated jejunal Thirty-Vella loop. In the Roux-en-Y loop cats, a protein-rich oral meal augmented the maximal HP acid response to pentagastrin but not the pepsin response. Protein-rich food in the Thirty-Vella loop did not stimulate acid secretion, nor did it augment the maximal GF or HP acid response to pentagastrin. Previous and present results suggest that an unknown humoral gastrin-potentiating mechanism is activated by a protein-rich meal in the stomach and the duodenum but not in the jejunum. The mechanism potentiates pentagastrin-induced acid, but not pepsin, secretion.

Prospective, randomized trial of selective vagotomy with pyloroplasty and selective proximal vagotomy with and without pyloroplasty in the treatment of duodenal, pyloric, and prepyloric ulcers.

In a prospective, randomized trial, 161 patients with duodenal, pyloric, or prepyloric ulcer underwent selective proximal vagotomy. Randomization was then performed to determine if the operation was finished (52 patients), if a pyloroplasty should be added (56 patients), or in addition, if the nerves of Latarjet should be divided (53 patients). Prepyloric and secondary gastric ulcers were excised for microscopy; all were benign. Sex, age, site of ulcer, and duration and incidence of complications of the ulcer disease were similar for the three groups. There was one operative death. The postoperative complications did not differ for the three groups. Four patients were lost to follow-up. The average follow-up for the 156 patients was 3 years (range 1 to 8 years). Recurrent ulcer was detected up to 5 years after surgery in 4 of 53 patients who had selective vagotomy with pyloroplasty, in 4 of 53 who had selective proximal vagotomy with pyloroplasty, and in 5 of 50 who had selective proximal vagotomy. Diarrhea was rare and mild or absent. Dumping was twice as common after selective vagotomy or selective proximal vagotomy with pyloroplasty than after selective proximal vagotomy only, but dumping resistant to treatment was recorded in only two or three patients in each group. The overall results (modified Visick scale) were unsatisfactory in 7 patients after selective vagotomy with pyloroplasty, in 4 after selective proximal vagotomy with pyloroplasty, and in 10 after selective proximal vagotomy, mainly because of epigastric pain with or without recurrent ulcer. We conclude that pyloroplasty may cause mild dumping without nuisance to the patient. The rates of recurrent ulcer in long-term follow-up trials are essential for final evaluation of the operations.