Reflux esophagitis in patients with Zollinger-Ellison syndrome

GASTROENTEROLOGY 1990;98:341-346

Reflux Esophagitis in Patients With Zollinger-Ellison Syndrome L. S. MILLER,

R. VINAYEK,

H. FRUCHT,

J. D. GARDNER,

R. T. JENSEN, and P. N. MATON Digestive Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland

The incidence of ulcers of the stomach and duodenum and their response to medical therapy, in patients with Zollinger-Ellison syndrome is well described. However, reflux esophagitis is less well recognized. In this study we determined the frequency of reflux esophagitis in 122 patients with Zollinger-Ellison syndrome and examined their response to medical therapy. Esophageal symptoms, endoscopic abnormalities, or both were present in 61% of patients. Forty-five percent of patients had esophageal symptoms consisting of heartburn, dysphagia, or both. Forty-three percent of patients had endoscopic abnormalities of the esophagus, and 23 % demonstrated moderate or severe disease. When sufficient antisecretory medication was administered to lower gastric acid secretion to ~10 mEq/h in the last hour before the next dose of drug, 67% of the patients with reflux esophagitis responded with complete disappearance of symptoms and normalization of the endoscopic abnormalities. The other 33% of patients required an increase in medication to lower acid output to ~5 mEq/h in 7% and tl mEq/h in the other 26% to resolve symptoms and signs completely. We conclude that reflux esophagitis occurs in the majority of patients with Zollinger-Ellison syndrome and responds well to medical therapy, although one third of patients require intensive antisecretory medication.

in 3% (2-6). However, 2 small series totalling 32 patients in which esophageal disease was sought specifically found that one third of the patients with ZES had evidence of reflux with esophageal symptoms, endoscopic abnormalities or both (7,B). We have shown previously that medical therapy with histamine H,-receptor antagonists (91 or omeprazole (10,ll) is safe and effective for producing symptomatic relief and healing of mucosal disease of the stomach and duodenum in patients with ZES. In the majority of patients, reduction of acid to 40 mEq/h is sufficient (12). However, patients with a previous partial gastrectomy require antisecretory drug sufficient to reduce gastric output to t5 mEq/h to heal all mucosal disease (13). In this study we examined the frequency of esophageal symptoms and endoscopic abnormalities in patients with ZES. Furthermore, we examined the response of esophageal disease to the reduction of gastric acid secretion by antisecretory agents.

Patients and Methods Patients

ollinger-Ellison syndrome (ZES) is characterized by gastric acid hypersecretion leading to ulceration of the upper gastrointestinal mucosa (1). In early

One hundred twenty-two consecutive patients (77 male and 45 female) with ZES were studied. Patients were referred to the National Institutes of Health (NIH) between 1976 and May 1988. Patients ranged in age from 19-68 yr, with a mean age of 45 yr. Nearly all patients referred to the NIH were already taking histamine Hz-receptor antagonists alone or in combination with an anticholinergic drug, either isopropamide (Darbid; Smith, Kline and French Laboratories, Philadelphia, Pa.) or propantheline bromide (ProBanthine; Searle Laboratories, Chicago, Ill.).

descriptions of the syndrome, emphasis was given to the findings of multiple duodenal ulcers and ulcers in unusual sites such as the jejunum and distal duodenum (1). Reflux esophagitis complicating ZES was thought to be rare. In 5 series of approximately 350 patients with ZES, esophageal involvement occurred

Abbreviotions used in this paper: NIH, National Institutes of Health; ZE!3,Zollinger-Ellison syndrome. 0 1990 by the American Gastroenterological Association 0016-5065/90/$3.00

Z

342

MILLER

ET AL.

GASTROENTEROLOGY

The criteria used for the diagnosis of ZES were a basal acid output >l5 mEq/h in unoperated patients or ~5 mEq/h in patients with previous gastric surgery and a fasting serum gastrin concentration >lOO pg/ml (normal, 400 pg/ml) together with a positive secretin test, calcium infusion test, or histological evidence of gastrinoma, or a combination of these (14). All patients were part of an ongoing study of the management of ZES that had been approved by the Clinical Research Committee of the National Institute of Diabetes, Digestive and Kidney Diseases of the NIH.

Methods

P 8

z

60 -

P

60-

1

40-

3

s

‘S

2

20-

ocI,

Heartburn

Results of Reflux

Esophagitis

at Initial

Assessment Of 122 patients, 74 (61%) had esophageal symptoms, endoscopic abnormalities of the esophagus, or both at the time of entry into the study. As illustrated in Figure 1,55 (45%) of 122 patients reported esophageal symptoms, either dysphagia or heartburn. In no case

Endoscopic Abnormalities

IOO-

o-

A detailed history was taken from each patient to elicit information on symptoms compatible with esophageal disease. Specifically, all patients were asked about symptoms of heartburn and dysphagia. Heartburn was defined as substernal burning pain that might radiate to the midchest, was brought on by meals or postural changes, and was promptly relieved by antacids. Dysphagia was defined as a sensation of food sticking when swallowed. Patients were also questioned about bronchitis, nocturnal cough, and wheezing. All patients underwent upper intestinal endoscopy. Specific findings of esophageal disease were noted, and esophagitis was categorized into one of 3 grades (15): grade 1, mild-erythema with friability; grade 2, moderate-mucosal erosion; and grade 3, severe-frank ulceration. Stricture was defined as an obvious narrowing of the esophagus associated with resistance to the passage of an adult endoscope. Patients received sufficient histamine Hz-receptor antagonist to reduce gastric acid secretion to t10 mEq/h in the last hour before the next dose of drug and were reassessed within z wk. If symptoms or signs did not resolve completely with the initial dose of medication, the dose was increased to reduce acid output to t5 mEq/h in the last hour before the next dose of drug and the patient was reassessed within 2 wk. If symptoms or signs persisted, patients were switched to omeprazole administered every 24 h or, if necessary, every 12 h to reduce acid output to tl mEq/h. The patients were then reassessed. Each patient was readmitted to the NIH every 6 to 12 mo to reassess symptoms, to check adequacy of control of acid secretion, and for upper intestinal endoscopy. Drugs used in this study were cimetidine (Smith, Kline and French), isopropamide, probanthine, ranitidine (Glaxo, Research Triangle Park, N.C.) famotidine (Merck Sharp & Dohme, Rahway, N.J.), and omeprazole [Hassle, Molndal, Sweden).

Incidence

Symptoms

Vol. 98, No. 2

Dysphagia

Esophagitis

Stricture

Figure 1. Frequency of esophageal symptoms and endoscopic abnormalities of the esophagus in patients with Zollinger-Ellison syndrome at the time of their initial assessment.

were these the sole symptoms of which the patient complained. All patients with esophageal symptoms also had pain suggestive of duodenal ulceration, diarrhea, or both. Of the 122 patients, 52 (43%) had heartburn and 15 (12%) had dysphagia. No patient reported a history of pulmonary symptoms that could be attributed to esophageal disease. Endoscopic

Abnormalities

As illustrated in Figure 1, 51 (42%) of the 122 patients had abnormalities of the esophagus at endoscopy. Fifty-two patients (42%) had esophagitis and 10 (8%) had an esophageal stricture. Of the 52 patients with esophagitis, 24 (46%) had grade 1 esophagitis, 7 (13%) had grade 2 esophagitis, and 21(41%) had grade 3 esophagitis. Thus, 28 patients (23% of the total) had either moderate or severe esophagitis. Nine of the 52 patients with esophagitis (17%) had typical Barrett’s esophagus. Response of Reflux Esophagitis to Medical Therapy Reduction of gastric acid output to ~10 mEq/h. When patients were given histamine Hz-receptor antagonists in doses sufficient to lower gastric acid secretion to ~10 mEq/h, complete resolution of gastric and duodenal disease occurred in all patients who had not had a previous partial gastrectomy. However, as shown in Figure 2, esophageal symptoms and mucosal abnormalities resolved in 55 of the 75 patients (73%) with esophageal disease. Of the 52 patients with heartburn, symptoms resolved completely in 40 (77%). Of the 15 patients with dysphagia, symptoms resolved completely in 12 (80%). Of the 52 patients with endoscopic evidence of esophagitis, the mucosa became entirely normal in 40

February 1990

ESOPHAGITIS

(77%). The mucosa became normal in 21 of 24 patients with mild esophagitis (86%). Of the 7 patients with moderate esophagitis, 4 (57%) had complete healing of erosions. In the 21 patients with severe esophagitis, 17 (81%) had complete healing of ulcers. Stricture was present in 10 patients and resolved in 6 (60% of patients with stricture) after initial therapy with esophageal dilatation. None of these 6 patients whose strictures resolved have had subsequent stricture formation in 9-92 mo of follow-up. Thus, the response of esophagitis to reduction of acid output to ~10 mEq/h did not depend on the severity of the mucosal disease. Reduction of gastric acid output to ~5 mEq/h in patients who did not respond to reduction of acid output to tl0 mEq/h. Twenty patients had partial improvement but not complete disappearance of esophageal symptoms, signs, or both after reduction of acid output to ~10 mEq/h. Nine had heartburn only, 4 had esophagitis only, 3 had heartburn and esophagitis, and 4 had esophagitis, dysphagia, and stricture. Nineteen of these patients received increased doses of histamine HZ-receptor antagonists to reduce acid output to t5 mEq/h. The other patient who had esophagitis and stricture with dysphagia received omeprazole in a dose sufficient to reduce acid output to ~1 mEq/h because of the severity of mucosal disease (see below). The responses of symptoms and endoscopic abnormalities in the 19 patients who received increased doses of histamine HZ-receptor antagonists are illustrated in Figure 3. In only 4 of the 19 patients did the esophageal symptoms or endoscopic abnormalities resolve when acid output was reduced to t5 mEq/h. Of the 12 patients with heartburn, only 3 had resolution of this symptom. In none of the 3 patients with dysphagia did the symptom resolve. Of the 10 patients with persistent endoscopic abnormalities of the esophagus after reduction of acid output to ~10 mEq/h,

Endoacopic Abnormalities

-

-

-

-

Heartburn

Dysphagia

Esophagitis

I -

Stricture

Figure 2. Response rate of esophageal symptoms and endoscopic abnormalities of the esophagus to reduction of gastric acid to ~10 n&q/h with histamine Hz-receptor antagonists in patients with Zollinger-Ellison syndrome.

IN ZOLLINGER-ELLISON

SYNDROME

343

122 5 'Z a"

loEt-

B

6-

k

4-

E 1

2OHeartburn

Dysphagia

Esophagitis

Stricture

Figure 3. Patients with Zollinger-Ellison syndrome with esophageal symptoms or endoscopic abnormalities despite reduction of acid output to ~10 mEq/b with histamine Hz-receptor antagonists. The cross-hatched areas indicate the numbers of these patients whose esophageal symptoms or signs resolved when gastric acid output was reduced to ~5 mEq/h by increased doses of histamine Hz-receptor antagonists.

abnormalities resolved in only one after further reduction of acid output to t5 mEq/h. This patient had esophageal erosions after reduction of acid output to 40 mEq/h that resolved after reduction of acid output to ~5 mEq/h. Three patients with grade 1 esophagitis, 2 patients with grade 2 esophagitis, and all 3 patients with esophageal ulcers continued to have some degree of disease when acid output was reduced to ~5 mEq/h. All 3 patients with esophageal strictures continued to require repeated esophageal dilatation at intervals of 2-4 wk. Reduction of gastric output to tl mEq/h with omeprazole therapy in patients who did not respond to reduction of acid output to t5 mEq/h. Twelve patients in whom reduction of acid output to ~5 mEq/h and 1 patient in whom reduction of acid output to ~10 mEq/h (see above) had failed to resolve esophageal symptoms, endoscopic abnormalities, or both were given sufficient omeprazole to reduce gastric acid output to tl mEq/h. Three other patients with heartburn after reduction of acid output to t5 mEq/h continued to receive histamine HZ-receptor antagonists because they regarded their symptoms as acceptable. The response of symptoms and endoscopic abnormalities to reduction of acid output to ~1 mEq/h in 12 patients after failure of other therapy to resolve completely esophageal disease is illustrated in Figure 4. Three patients had heartburn only, 2 had esophagitis only, 3 had esophagitis and heartburn, and 4 had esophagitis, stricture, and dysphagia. Of the 6 patients with heartburn, symptoms resolved completely in 2; in the other 4 there was considerable improvement in the symptoms. In 8 of the 9 patients with esophagitis there was complete healing of the mucosa. In one patient with esophageal erosions, the erosions healed but the patient continued to have grade 1 esophagitis of the distal 5 cm of the esophagus. All 4 patients with dysphagia and stricture formation prior to omeprazole therapy improved. These patients

344

MILLER

ET AL.

GASTROENTEROLOGY

Symptoms

Endoscopic Abnormalltk

oHeartburn

Dysphagia

Esophagitis

Stricture

Figure 4. Patients with Zollinger-Ellison syndrome whose esophageal symptoms or endoscopic abnormalities persisted despite reduction of acid output to c5 mEq/h with histamine Hz-receptor antagonists. The cross-botched oreos indicate the number of these patients whose esophageal disease resolved completely when gastric acid output was reduced to tl mJIq/h with omeprazole. The stippled areas indicate the patients whose esophageal disease markedly improved during omeprazole therapy.

had required dilatations every 2-4 wk while taking histamine Hz-receptor antagonists, but after reduction of acid output to 4 mEq/h with omeprazole, dilatation was required only once every 4-6 mo. Discussion This study establishes the frequency, type, and severity of reflux esophagitis and response to medical therapy in a large group of patients with ZES. It demonstrates that esophageal symptoms, mucosal abnormalities, or both are common manifestations of ZES. We found that 61% of patients with ZES had one or more manifestations of reflux esophagitis. The frequency of both esophageal symptoms and endoscopic signs of esophageal disease were much higher than in previous series of patients with ZES. A review of the previously published series that report the frequency of clinical symptoms and signs in patients with ZES before 1979 gives the impression that esophageal disease in ZES was either rare or nonexistent (2-6). In 260 cases of ZES reported by Ellison and Wilson (21, only 2 ulcers of the esophagus were reported. There was no mention of heartburn or dysphagia. In two series from the Mayo Clinic (3,4), 1 reporting 40 patients and 1 reporting 26 patients with ZES, only 1 esophageal ulcer was observed, and although peptic ulcer symptoms were described, pyrosis, heartburn, and dysphagia were not reported. Bonfils and Bader (6) reported a 7% frequency of esophageal ulcer in 27 patients with ZES, and Stage and Stadil (5) reported dysphagia and esophageal stricture in 2 of 34 patients with ZES. The frequency of other mucosal abnormalities of the esophagus and the incidence of heartburn were not reported. Two more recent studies of patients with ZES have sueeested that reflux esoohaaitis is more common than

Vol. 98, No. 2

described originally. McCallum and Walsh (7) noted that of 17 patients with ZES, 5 (29%) reported chronic heartburn, and 2 of these had esophageal strictures requiring dilatation. In a prospective study, Richter et al. (8) found that 5 of 15 patients (33%) with ZES had reflux esophagitis. At the time of the study, 6 of 15 patients had heartburn and 2 had dysphagia. Overall, 9 of 15 (66%) patients had evidence of esophagitis as judged by a Bernstein test, endoscopy, or biopsy. It is likely that in the past the incidence and importance of reflux esophagitis were underestimated because of the overwhelming problems with perforation and bleeding of duodenal ulcers. Nevertheless, we believe that in the present study we have underestimated the frequency and severity of esophageal disease. Nearly all patients referred to the NIH for evaluation were referred after having started histamine Hz-receptor antagonists, with or without anticholinergic medication. Therefore, it is likely that some esophageal lesions had time to heal before the patients’ initial endoscopic evaluation at the NIH. Furthermore, biopsy specimens of the esophagus were not taken in every case. Because visual inspection is less sensitive than histological analysis for the diagnosis of mild esophagitis (161, we probably underestimated the frequency of mild esophagitis patients with ZES at initial evaluation.

present

in our

At one time, when it was discovered that pharmacological doses of gastrin increased lower esophageal pressure, it was suggested that the hypergastrinemia in patients with ZES might protect them from gastroesophageal reflux. The data in this report demonstrate clearly that this is not the case. Previous studies by us have demonstrated that reduction of acid output to t10 mEq/h during the last hour before the next dose of medication in patients who have not had a partial gastrectomy (12,14) or to ~5 mEq/h in patients with previous gastric surgery (13) eliminates symptoms and heals all mucosal disease in the stomach and duodenum. However, in the present

study complete resolution of reflux esophagitis occurred in only 73% of patients in whom gastric acid secretion was decreased to t10 mEq/h, indicating that in some patients with ZES, reflux esophagitis is more refractory to the effects of decreasing gastric acid hypersecretion than gastric or duodenal disease. Furthermore, the majority of patients with reflux esophagitis that did not resolve completely when acid output was reduced to ~10 mEq/h required intensive antisecretory therapy with omeprazole to decrease gastric acid output to ~1 mEq/h to produce resolution or near resolution of all esophageal disease. These studies in patients with ZES showing that reflux esophagitis is more difficult to resolve than gastric or duodenal disease are similar to results of treatment of reflux in

February 1996

patients with idiopathic reflux ,esophagitis, who have acid outputs in the normal range. Richter (17)reviewed the results of 10 short-term trials of cimetidine and 5 studies of ranitidine treatment of idiopathic gastroesophageal reflux disease. The dose of cimetidine was 300-400 mg with meals and at bedtime, and the dose of ranitidine was 150 mg b.i.d. Overall approximately 60% of patients with esophagitis healed or markedly improved during treatment, whereas healing rates of duodenal or gastric ulceration with histamine Hz-receptor antagonists are 85%-95% (12,181. Furthermore, as in the present study, omeprazole is significantly superior to the histamine H,receptor antagonists in the relief of heartburn and the healing of idiopathic reflux esophagitis (19-251, and omeprazole is capable of healing the mucosa in nearly all patients with idiopathic reflux esophagitis who fail to respond to histamine Hz-receptor antagonists (26,27). As with the response to therapy in patients with idiopathic reflux esophagitis, some patients with ZES and esophagitis will require sufficient antisecretory medication to render them virtually achlorhydric. Although such therapy is safe in short-term and midterm treatment, studies of the long-term risks of complete suppression of gastric acid production remain to be determined (27). The potential long-term problems include the development of gastric carcinoid, gastric cancer, and increased risk of bacterial infection in the gut (28).

References 1.

2.

3.

4.

5.

6.

7.

8

9.

Wolfe MM, Jensen RT. Zollinger-Ellison syndrome: current concepts in diagnosis and management. N Engl J Med 1987;317: 1200-9. Ellison EH, Wilson SD. The Zollinger-Ellison syndrome. Reappraisal and evaluation of 280 registered cases. Ann Surg 1964;160: 512-30. Regan PT. Malagelada JR. A reappraisal of clinical, roentgenographic, and endoscopic features of the Zollinger-Ellison syndrome. Mayo Clin Proc 1978;53:19-23. Cameron AJ. Hoffman HN. Zollinger-Ellison syndrome. Clinical features and long-term follow-up. Mayo Clin Proc 1974;49: 44-51. Stage JG, Stadil F. The clinical diagnosis of the ZollingerEllison syndrome. Stand J Gastroenterol1979;14(Suppl. 53):7991. Bonfils S, Bader JP. The diagnosis of Zollinger-Ellison syndrome with special reference to the multiple endocrine adenomas. In Glass GBJ. Progress in gastroenterology. Volume 2. Philadelphia: Grune & Stratton, 1978332-55. McCallum RW, Walsh JH. Relationship between lower esophageal spincter pressure and serum gastrin concentration in Zollinger-Ellison syndrome other clinical settings. Gastroenterology 1979;76:76-81. Richter JE, Pandol SJ, Caste11 DO, McCarthy DM. Gastroesophageal reflux disease in the Zollinger-Ellison syndrome. Ann Intern Med 1981;95;37-43. Howard JM, Chremos AN, Collen MJ, McArthur KE, Cherner JA, Maton PN, Ciarlegio CA, Cornelius MJ, Gardner JD, Jensen

ESOPHAGITIS

10.

11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

IN ZOLLINGER-ELLISON

SYNDROME

345

RT. Famotidine, a new, potent, long-acting histamine H,receptor antagonist: comparison with cimetidine and ranitidine in the treatment of Zollinger-Ellison syndrome. Gastroenterology 1985;88:1026-33. Maton PN, Lack E, McArthur KE. Collen MJ, Frucht H, Miller LS, Saeed ZA, Stark HA, Gardner JD, Jensen RT. Effects of long term omeprazole on gastric acid secretion and mucosal histology in patients with Zollinger-Ellison syndrome (ZES) (abstract). Gastroenterology 1988;94:A288. McArthur KE, Collen MJ, Maton PN, Cherner JA, Howard JM, Ciarleglio CA, Cornelius MJ. Jensen RT, Gardner JD. Omeprazole: effective, convenient therapy for Zollinger-Ellison syndrome. Gastroenterology 1985;88:939-44. Raufman JP, Collins SM. Pandol SJ, Korman LY, Collen MJ, Cornelius MJ, Feld MK, McCarthy DM, Gardner JD, Jensen RT. Reliability of symptoms in assessing control of gastric acid secretion in patients with Zollinger-Ellison syndrome. Gastroenterology 1983;84:108-13. Maton PN, Frucht H, Vinayek R, Wank SA, Gardner JD, Jensen RT. Medical management of patients with Zollinger-Ellison syndrome who have had previous gastric surgery: a prospective study. Gastroenterology 1988;94:294-99. Jensen RT, Gardner JD, Raufman J-P, Pandol SJ, Doppman JL, Collen MJ. Zollinger-Ellison syndrome. NIH combined clinical conference. Ann Intern Med 1983;96:59-75. Caste11 DO, Wu WC, Ott DJ. Gastroesophageal reflux disease. Mount Kisco, N.Y.: Futura, 1985. Ismail-Beigi F, Pope CE II. Distribution of the histological changes of gastroesophageal reflux in the distal esophagus of man. Gastroenterology 1974;66:1109. Richter JE. A critical review of current medical therapy for gastroesophageal reflux disease. J Clin Gastroenterol 1986; 8(Supp1.1):72-80. Brogden RN, Carmine AA, Heel RC, Speight TM, Avery GS. Ranitidine: a review of its pharmacology and therapeutic use in peptide ulcer disease and other allied diseases. Drugs 1982;24: 267-303. Havelund T, Laursen LS, Skoubo-Kristensen E, Andersen BN, Pedersen SA, Jensen KB, Fenger C. Hanberg-Sorensen F, Lauritsen K. Omeprazole and ranitidine in treatment of reflux oesophagitis-a double-blind comparative trial. Lancet 1988;296: 89-92. Klinkenberg-Knol EC, Jensen MBJ, Festen HM, Meuwissen SGM, Lamers CBHW. Double-blind multicentre comparison of omeprazole and ranitidine in the treatment of reflux oesophagitis. Lancet 1967;1:349-51. Dammann HG, Blum AL, Lux G, Rehner M, Riecken EO, Schiessel R, Wienbeck M, Witzel L, Berger J. Differences in healing tendency of reflux oesophagitis with omeprazole and ranitidine. Results of an Austrian-German-Swiss multicenter trial. Deut Med Woch 1986:111:123-8. Dehn TCB, Shepherd HA, Colin-Jones D, Kettlewell MGW. Double-blind comparative study of omeprazole (46 mg OD) vs cimetidine (466 mg qds] in the treatment of erosive reflux esophagitis. Gut 1988;29:A1440. Zeitoon P, Isbal JP. Omeprazole versus ranitidine in erosive esophagitis: results of a French-Belgian multicentre study. Gut 1987;29:A710. VanTrappen G, Rutgeerts L. Schurmans P, Coenegrachts JL. Omeprazole (40 mg) is superior to ranitidine in short-term treatment of ulcerative reflux esophagitis. Dig Dis Sci 1988;33: 523-9. Sandmark S, Carlsson R. Lundell L, Fausa 0. Omeprazole or ranitidine in the treatment of reflux oesophagitis. Results from a double-blind, randomised, Scandinavian multicentre study. Gut 1987;28:Al375.

346 MILLER ET AL

26. Klinkenberg-Knoll EC, lansen JBMJ, DeBrayne ]W, Nehs GF, Festen HPM, Snel P, Meuwissen SGM. Long term efficacy and safety of omeprazole (OME) on healing and prevention of resistant reflux esophagitis. Gastroenterology 1088;94:A230. 27. Clissold SP, Campoli-Richards DM. Omeprazole: a preliminary of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in peptic ulcer disease and ZollingerEllison syndrome. Drugs 1986;32:15-47.

GASTROENTEROLOGY

Vol. 98, No. 2

28. Howden CW, Hunt RH. Relationship between gastric secretion and infection. Gut 1987;28:96-107.

Received March 24,1989. Accepted July 24.1989. Address requests for reprints to: P. N. Maton, M.D., Building 10, Room 9C-103, National Institutes of Health, Bethesda, Maryland 20892.