ALIMENTARY TRACT PEPSIN SECRETION, PEPSINOGEN, AND GASTRIN IN "BARRETT'S ESOPHAGUS" Clinical and morphological characteristics. Material and Methods

Transcription

1 GASTROENTEROLOGY 70: , 1976 Copyright 1976 by The Williams & Wilkins Co. Vol. 70, No.5 Printed in U.S.A. ALIMENTARY TRACT PEPSIN SECRETION, PEPSINOGEN, AND GASTRIN IN "BARRETT'S ESOPHAGUS" Clinical and morphological characteristics JAGDISH C. MANGLA, M.D., F.R.C.P.(C), F.A.C.P., ERIC A. SCHENK, M.D., LAURENT DESBAILLETS, M.D., GILDA GUARASCI, NORMAN P. KUBASIK, PH.D., AND MICHAEL D. TURNER, M.D., PH.D. Gastroenterology Unit, Monroe Community Hospital, and Department of Pathology and Isaac Gordon Laboratory for Gastrointestinal Research, University of Rochester School of Medicine and Dentistry, Rochester, New York Four cases of Barrett's esophagus are presented. Three cases presented with significant esophageal bleeding and one case presented with high esophageal stricture. Gastrointestinal panendoscopy was done in each case and multiple biopsies were taken. The biopsies were utilized for histomorphology, pepsinogen agar gel electrophoresis, and tissue gastrin assays. Tissue gastrin levels in esophageal mucosa were elevated in 2 cases when compared to controls with and without hiatus hernia. Pepsin and acid secretory studies were done by isolating the esophagus. Barrett's esophagus was shown to produce pepsin by both chemical studies (2 cases) and agar gel electrophoresis at ph 5.7 (3 cases), and was also shown to produce acid. The mucosa contained either cathepsin or cathepsin and pepsinogens in all cases. Nissen's fundoplication was performed in all of the patients. Of 3 patients who were bleeding, 2 who consented for this operation stopped bleeding after the operation. It is to be noted that the usual clinical treatment of antacids, bedrest, and raising the head end of the bed failed in all of the patients. The follow-up of 9 months to 3 years postoperatively has shown persistence of Barrett's mucosa with no evidence for any reversion to normal esophageal type. An esophageal mucosa composed of columnar epithelium was first described in 1950 by Barrett. 1 It is not always simple to recognize the occurrence of Barrett's epithelium, and many techniques,2 such as barium X-ray studies, manometry, measurement of mucosal electrical potentials, pertechnetate scintigraphy (Reference 3 and J. C. Mangla and M. Brown, submitted for publication), and endoscopy (Reference 2 and L. Desbaillets and J. C. Mangla, submitted for publication) have been used for this purpose. Definitive diagnosis is possible only with multiple serial esophageal biopsies. The presence of oxyntic cells has been shown in many studies 2,.-7 and two 5, 6 have demonstrated the formation of acid by such mucosa. The previous report7 was the first to show the presence of pepsinogens and chief cells in Barrett's mucosa. Received July 15, Accepted November 3, Address requests for reprints to Dr. Jagdish C. Mangla, Department of Medicine, Gastroenterology Unit, Monroe Community Hospital, 435 East Henrietta Road, Rochester, New York Dr. Desbaillet's current address is: 41 Chemin des S les, 1212 Grand Laney, Glmeve, Switzerland. 669 This paper will describe further morphological characteristics of this type of mucosa. It demonstrates for the first time that this mucosa may contain gastrin and that it secretes pepsins under basal conditions. Material and Methods Panendoscopy with an Olympus GIF D2 was done in each patient and in control subjects. Serial biopsies of the mucosa were taken from the lower end of the esophagus to approximately 2 cm distal to the upper esophageal sphincter. The biopsies were kept moist in gauze soaked in saline and placed on ice. Biopsies of the duodenum and gastric antrum and fundus were also obtained. Biopsies were submitted for histopathological examination, for gastrin assay, and for pepsinogen analysis by agar gel electrophoresis. Biopsies obtained from normal esophageal mucosa as described previously7 served as controls. Agar gel electrophoresis. This was carried out as previously described 8-1o both at ph 8.3 and at ph 5.7. Esophageal and gastric secretions were electrophoresed only at ph 5.7 because at ph 8.3 pepsins are destroyed. Tissue gastrin assay. The biopsy specimens were blotted, weighed, and homogenized as rapidly as possible in a Potter

2 670 MANGLA ET AL. Vol. 70, No.5 Elvehjem homogenizer in 1 ml of cold 0.02 M barbital buffer at ph Adjacent biopsies were sent for histopathology to confirm that the tissue for gastrin assay was coming from Barrett's epithelium. Gastrin assays were performed by radioimmunoassayll using a "Gastrin Immunotype Kit" (E. R. Squibb and Sons, Inc., New York). Appropriate dilutions of the homogenates were prepared and assayed in triplicate. Esophageal biopsies were obtained from the mid esophagus from 4 patients undergoing gastrointestinal endoscopy for other reasons. These 4 patients served as controls and did not show any evidence of Barrett's epithelium. Two of the controls had hiatus hernia and 2 had nonspecific upper gastrointestinal symptoms. Biopsies were also obtained from the mucosa of the gastric antrum and duodenum and assayed for gastrin only as further check on the gastrin assay. Acid and pepsin secretory studies. In order to determine whether the columnar epithelium of the Barrett's esophagus secreted acid or pepsin, separate aspirates were obtained from the esophagus and compared with aspirates obtained from the stomach. Special measures were taken to avoid reflux of gastric contents in the esophagus. A Baker's tube, which has a construction similar to a Foley catheter with a balloon at the end, was used for this purpose. Three lateral holes were cut in the terminal 5 em of a nasogastric tube (fig. 1) which was attached to the side of a Baker's tube in such a way that its lower end was 3 cm above the balloon of the Baker's tube. After passage of the combined tubes, the balloon was inflated in the stomach with 20 cc of air and pulled up to fit snugly against the cardia to prevent reflux. Blue dextran was injected into the stomach after inflation of the balloon and esophageal aspirates were examined for dye to detect any reflux of gastric contents into the esophagus. The patient was instructed to expectorate his saliva and not to swallow it. Secretions were aspirated for 1 hr before and for 1 hr after an injection of {1-aminoethyl pyrazole hydrochloride (Histalog) (1.5 mg per kg of body weight). Samples were collected from the esophagus and stomach and placed on crushed ice to preserve pepsins. The volume, ph, and proteolytic activity of the esophageal samples were measured and each was checked for the presence of blue dextran. Agar gel electrophoresis of pepsins was done as described above using ph 5.7 buffer for esophageal and gastric secretions. Proteolytic activity of esophageal samples. This was measured by a modification 12 of the procedure of Anson and -STOMACH FIG. 1. Diagrammatic representation of the arrangement employed to isolate the esophagus from the stomach. A nasogastric tube was attached to a Baker's tube as shown in this diagram. Mirsky.'3 One unit of enzyme activity was defined as that amount of enzyme in 1 ml of solution which released an average of 1 ~ m o(181 l e ~ g of ) tyrosine-like substance per min from 5 ml of 2% acid hemoglobin substrate during a I-hr incubation at 37 C. Results Brief Case Histories of Four Patients with Barrett's Esophagus Case 1. J. M., a 53-year-old white male school custodian, was admitted to the hospital with upper gastrointestinal bleeding. The diagnosis of Barrett's esophagus was made and pepsinogens were documented for the first time in this mucosa. 7 After the failure of conservative treatment for reflux, he underwent a Nissen's fundoplication. Within a few days the patient was relieved of his symptoms and has been well since. Three years after fundoplication the endoscopy was repeated. There was a sma1l5-mm mucosal erosion 5 cm above the lower end of the esophagus, but the large ulcer at the upper end had healed. Three further biopsies were taken from the esophagus at the same level as before, and on a separate occasion esophageal and gastric secretions were collected as described above. Case 2. E. K., a 51-year-old white male, presented in February, 1974 with a history of four episodes of hematemesis since He complained of pain high in the epigastrium radiating to the back. Panendoscopy of the upper gastrointestinal tract revealed the presence of an ulcer 5 cm above the lower end of the esophagus and showed the typical changes of Barrett's epithelium (L. Desbaillets and J. C. Mangla, submitted for publication) in the lower two-thirds of the esophagus. Appropriate studies described in the Material and Methods section were done. > The patient was treated for 4 weeks with bed rest, antacids, and elevation of the head end of the bed without success. The ulcer was, therefore, excised surgically, and a vagotomy, pyloroplasty, and Nissen's fundoplication were done. Nine months later esophagogastroscopy revealed remarkable healing of the esophageal mucosa. Multiple serial biopsies were done at this time. Case 3. J. B., a 75-year-old white male, developed a quadriplegia 'at the age of 66 and had many episodes of upper gastrointestinal hemorrhage, the cause of which was obscure. Esophagogastroscopy after an episode of bleeding revealed a hemorrhagic, friable, edematous esophageal mucosa with a pseudomembrane typical of Barrett's epithelium. Multiple biopsies of the esophageal mucosa were obtained and assayed as described in cases 1 and 2. The patient refused surgery, continued to bleed, and required some 20 to 25 units of blood to maintain his hematocrit. Finally after 5 months, he underwent a Nissen's fundoplication and died 2 weeks later as a result of renal complications. Case 4. E. H., a 63-year-old white female, suffered from heartburn and intermittent dysphagia for 20 years. Esophagoscopy revealed a narrowing at 6 cm below the cricopharyngeal sphincter which would not admit the

3 GIF D2 Olympus gastrofibroscope. The stricture was dilated to size 38 French, with relief of symptoms, although they returned when dilation was suspended for 2 to 3 weeks. Esophagoscopy after dilation revealed a normal mucosa above the stricture, but below the stricture as far as the lower end of the esophagus there was marked hyperemia with a patchy pseudomembrane. Multiple mucosal biopsies were obtained and examined TABLE Patient (case no.) J.M. (1) E. K. (2) J.B. (3) E.H.(4) a 671 BARRETT'S ESOPHAGUS May 1976 Cm (from incisors) Squamous as described above. The biopsy at 20 cm was above the stricture and showed normal esophageal mucosa on histopathology. As the dilation was insufficient to relieve the sumptoms completely and as the mucosa below the stricture was not healing, a Nissen's fundoplication was done. The patient has remained asymptomatic since without further dilations. An esophagosccipy 9 months later showed 1. Types of mucosal cellsa Columnar Goblet Argentaffin Parietal Chief Secretory N onsecretory Acute ulcer (aortic arch) (5 em above cardia), present in biopsy; -, absent. FIG. 2. Patient case 2. Biopsy at 25 em shows junctional zone between stratified squamous epithelium (left) and irregularly arranged columnar epithelium (right). H & E, x 250.

4 672 MANGLA ET AL. no stricture and a markedly improved esophageal mucosa with no "pseudomembrane" and much less hyperemia. Morphological Findings Biopsies were obtained from the esophagus at 20 to 35 cm from the incisors, as well as from the stomach in all patients. The various types of mucosal, epithelial, and other cells present in these biopsies obtained at different levels are summarized in table 1. In all cases, the most proximal biopsy (20 cm) showed squamous epithelium. A histologically distinct squamocolumnar junction (fig. 2) was seen in the biopsies from two cases. The distal esophageal biopsies from all cases showed a glandular mucosa, the luminal surface of which consisted of irregular sized and shaped villous projections (fig. 3), whereas the deeper portion contained irregular tortuous crypt-like glands. Both the surface and glandular epithelium consisted predominantly of tall columnar cells with basally located nuclei. Follow-up biopsies obtained in three cases (cases 1, 2, and 4) from 9 months to 3 years after fundoplication still revealed Barrett's epithelium along with persistence of parietal and chief cells in the 2 cases in which they were present initially (figs. 3 and 4). The results of agar gel electrophoresis of esophageal mucosa, acid and pepsin studies, and esophageal mucosal gastrins are summarized in table 2. Some details of these results follow. Agar Gel Electrophoresis of Esophageal Mucosa and Esophageal Secretions There are two major groups of proteolytic bands in gastric mucosa on agar gel electrophoresis at ph 8.3 buffer.8-10 There is still a controversy on the number of bands in each group. The fastest-moving proteases are destroyed by sequential acidification and alkalinization and are, therefore, regarded as pepsinogens. The activity of the slowest-moving band 9 10 is not affected by acidification and neutralization and is thought to represent a cathepsin. The body and fundus of the stomach may show all of the bands, whereas only cathepsin and two slow-moving pepsinogens are found in the distal antrum and proximal duodenum. Normal esophageal mucosa from 5 patients undergoing esophagogastroscopy for other reasons 7 did not reveal any proteolytic bands on agar gel electrophoresis. Similarly, no proteolytic activity could be demonstrated in esophageal secretions in normal individuals when care was taken to avoid gastroesophageal reflux. Case 1. The electrophoresis, at ph 8.3 buffer, of a homogenate of a gastric mucosal biopsy was compared with that made from a midesophageal mucosal biopsy. In this patient gastric mucosa lacked the cathepsin band. The esophageal mucosal pattern (lower slot, fig. 5A) showed two bands. The first band after the slot appeared to be cathepsin, which was very prominent. Esophageal secretions were insufficient for proteolytic assay; however, agar gel electrophoresis at ph 5.7 buffer could be performed. On electrophoresis two bands (fig. 5B, upper slot) similar to those seen in the mucosal biopsy were obtained. This suggests that there is some secretion of pepsin and cathepsin from this esophageal mucosa. It is interesting to note that the gastric secretions (fig. 5B, lower slot) did not show a cathepsin band and in this regard resembled the gastric mucosa (fig. 5A). FIG. 3. Patient case 1. Follow-up biopsy 3 years after fundal plication shows persistence of villiform mucosa and glandular epithelium. There is a prominent chronic inflammatory cell infiltrate and some fibrosis in the lamina propria. H & E, x 200. Vol. 70, No.5 The marked differences between the gastric and esophageal pepsin patterns confirm that the findings did not result from reflux of gastric secretions into the esophagus. Case 2. The homogenate from the gastric mucosal biopsy (upper slot, fig. 6A) and that from the esophageal mucosal biopsy (lower slot, fig. 6A) show completely different patterns, confirming that the biopsy was not taken from a herniated portion of the stomach. The ph of the basal esophageal secretion was 1.60, and this fell to 1.0 after administration of Histalog. There was also an increase of total acid output from 0.17 meq per hr to 0.34 meq per hr. The patient secreted 78,765 units of pepsin in a 1-hr basal sample. There was no evidence of any stimulation of pepsin activity seen after Histalog administration (table 2). The decrease in activ-

5 May 1976 BARRETT'S ESOPHAGUS 673 FIG. 4. Patient case 4. Follow-up biopsy shows deep mucosal glands lined by triangular shaped parietal cells and by irregularly shaped vacuolated cells. H & E, x 400. TABLE 2. Results of agar gel electrophoresis Patient (case no.) Gastrins (done in 4 normals, 0-8.8pg/mg wet wt of mucosa) Pepsin in Pepsinogen (zymogram in Esophageal secretions Acid esophageal esophageal mucosa compared (pepsin pattern compared secretion to gastric mucosa) to gastric secretion) pg/mg Ulhr J. M. (1) 87 2 bands completely different Similar 2 bands (one band pattern from stomach; cathepsin) present; very cathepsin different pattern when compared to gastric juice E.K. (2) 5 Basal ph 1.60; 0.17 meq/hr in 43 ml secretion 78,765 Stimulated ph 1.0; 0.24 meq/hr in 49.4 ml secretion 40, Many bands completely different pattern from stomach; no cathepsin Pattern of esophageal secretions completely different from gastric juice J. B. (3) E.H.(4) 9 40 Basal ph 2.45; 0.02 meq/hr in 17ml Stimulated ph 2.75; meq/hr in 35 ml 19,456.4 Only cathepsin present No secretions 4, Completely different pat- Esophageal secretions pat- tern from stomach; tern contain cathepsin cathepsin band; very different from gastric juice or activated gastric mucosa ity of pepsin after Histalog may be a dilutional phenomenon. A gastric mucosal homogenate was activated at ph 2 by the addition of 0.2 N HCI and was allowed to stand at 37 C for 30 min. Activation converts the pepsinogens in the gastric mucosa to pepsins. The upper slot of figure 6B contained activated gastric mucosal homogeante, the middle slot contained gastric juice and the bottom slot contained esophageal secretions. The pepsin pattern of the esophageal secretion is different from the other two. Case 3. This patient revealed only the histological features with minimal evidence of other characteristics of Barrett's mucosa seen in the other 3 patients. On agar gel electrophoresis (fig. 7), gastric mucosal homogenate (upper slot) was compared with esophageal mucosal samples taken at 20 and 30 cm from the incisor teeth. The esophageal mucosa showed only a cathepsin band. Even after two attempts, no esophageal secretions could be obtained for study.

6 674 MANGLA ET AL. Vol. 70, No.5 thology (table 1). Antral samples in 4 patients showed an average of 15,009 pg per mg, whereas duodenal mucosal samples showed an average of 648 pg per mg wet weight of tissue. Discussion Previous studies revealed the presence of chief cells and pepsinogens in the Barrett's mucosa, but no attempts were made to determine whether the mucosa is capable of secreting pepsins into the lumen of the esophagus. This study used Baker's tubes with an intragastric balloon to prevent reflux, stimulated the patients with Histalog, and obtained esophageal secretions from 3 of 4 patients. These 3 patients did indeed show pepsin in their esophageal secretions. In 2 patients (cases 2 and 4, table 2) the esophageal secretions were sufficient to measure pepsin by chemical assay. In one (case 2) there was an apparent slight increase in acid production after administration of the stimulant, and neither case showed any increase in pepsin output. Throughout the procedure, the stomach was kept empty by continuous and manual sunction and no evidence of blue dextran marker was found in esophageal samples. The esophageal secretion revealed a pattern of proteolytic enzymes on agar gel electrophoresis different from that of gastric secretion. It is of interest to note that in case 1 the esophageal secretion pattern was very similar to esophageal mucosal patterns (fig. 5, A and B). 7 FIG. 5. A, case 1. Agar gel electrophoresis at ph 8.3 of gastric and esophageal mucosal biopsy homogenates. Upper slot is gastric mucosa. Lower slot is esophageal mucosa showing the prominent cathepsin band lacking in gastric mucosa (upper slot). B, case 1. Agar gel electrophoresis at ph 5.7 of esophageal and gastric secretions. The Upper slot of esophageal secretion shows two bands corresponding to that seen in the esophageal mucosal biopsy (fig. 5A, lower slot). The lower slot shows the absence of a corresponding cathepsin band in gastric secretion in the same way as shown in gastric mucosa (fig. 5A, upper slot). Case 4. The gastric mucosal biopsy homogenate pattern (fig. 8A, upper slot) was different from that of the esophageal biopsy (lower slot). The esophageal biopsy contains a cathepsin band missing from the gastric biopsy. There are missing bands in the esophageal sample. Esophageal secretions were collected free of gastric contamination. Although acid and pepsin are present in both basal and stimulated secretions, no actual rise in values was observed on Histalog stimulation (table 2). Esophageal secretion pattern (upper slot, fig. 8E) is different from gastric secretions and activated gastric mucosa. Note the presence of the cathepsin band in esophageal secretion; it goes hand in hand with the cathepsin band seen in the esophageal mucosal biopsy (fig. 8A, lower slot). These distinctive patterns are some evidence that reflux of gastric juice was successfully avoided in these experiments. Tissue Gastrins Midesophageal mucosal samples between 27 and 30 cm were examined for tissue gastrins. The results in 4 controls were 1.7, 0, 3.4, and 8.8 pg per mg of net wet weight of tissue, with an average of 3.5 pg per mg. The results in 4 patients are shown in table 2 with an average of 35.2 pg per mg. Adjacent biopsies from these 4 patients showed "Barrett's epithelium" on histopa- FIG. 6. A, case 2. Agar gel electrophoresis at ph 8.3 of gastric (upper slot) and esophageal mucosa (lower slot) showing completely different pattern. B, case 2. Agar gel electrophoresis at ph 5.7 of activated gastric mucosal homogenate to ph 2.0 (upper slot), gastric juice (middle slot), and esophageal secretions (bottom slot). Note that the pattern of esophageal secretions is different from the upper two slots.

7 May 1976 BARRETT'S ESOPHAGUS 675 FIG. 7. Case 3. Agar gel electrophoresis at ph 8.3 of gastric mucosal homogenate (upper slot) compared with esophageal biopsy homogenates. Note the presence of only cathepsin band in lower two slots. FIG. 8. A, case 4. Agar gel electrophoresis at ph 8.3 of gastric mucosa (upper slot) and esophageal mucosa (lower slot). Lower slot contains cathepsin band and a somewhat different pattern of bands from that seen in the gastric mucosa. B, case 4. Agar gel electrophoresis at ph 5.7 of esophageal secretions (upper slot), gastric secretion (middle slot), and activated gastric mucosal homogenate to ph 2.0 (bottom slot). Note the presence of cathepsin band in esophageal secretion, which compares well with the esophageal mucosal sample in figure 8A. Another patient, case 4, also illustrates the different patterns between esophageal and gastric secretions (fig. 8, A and B). If there were reflux, the pattern of gastric juice should be present in the esophageal secretions. Similarly, a herniated portion of the stomach would be expected to have a completely different pattern from the esophageal one observed in these patients. This study also reveals that Barrett's mucosa has even more of the functional characteristics of gastric mucosa. It was shown in 2 cases (cases 1 and 4) that esophageal mucosal gastrin is raised 5 to 10 times above the level seen in controls. These authors, unfortunately, did not have frozen tissue available for demonstration of gastrincontaining cells by immunofluorescent techniques. Barrett's mucosa is also important from many clinical aspects. It is very well known that ectopic gastric mucosa, as in Meckel's diverticulum, is likely to bleed. 14, 15 Of 4 patients described here, 3 presented with significant upper gastrointestinal bleeding, and the diagnosis was not established by the usual investigations. On endoscopy one can easily observe the bleeding from the raw, friable, hemorrhagic and pseudomembrane-covered mucosa (L. Desbaillets and J. C. Mangla, submitted for publication). Sometimes bleeding may be from an esophageal ulcer. It is extremely important to make the diagnosis of Barrett's esophagus, because after Nissen's fundoplication the bleeding stopped in all of the patients. One patient who repeatedly refused surgery continued to bleed for 5 months. The operation of fundoplication has been used before 16 for this condition, with the rationale that cessation of acid and pepsin reflux takes away the added insult to this abnormal mucosa and leads to healing. All patients in the present study showed remarkable healing of esophageal mucosa on endoscopy. Case 4 had a high esophageal stricture which needed many dilations to keep it open, but after fundoplication no more dilation was needed. In spite of the dramatic reversion of symptoms there was no reversion of the mucosa to normal stratified squamous epithelium, and no significant change in esophageal enzyme patterns was seen in a follow-up of 9 months to 3 years. The observations on multiple biopsies obtained from different levels in 4 patients indicate that certain types of epithelial cellular elements are constantly present in all biopsies, whereas other cellular elements are variable. The constant cellular elements are the tall columnar epithelial cells and goblet cells, both of which cover the luminal surface of villiform projections and line cryptlike glands within the mucosa. Usually a small number of Argentaffin cells are present in the deeper mucosal glands. The inconstant cell types found in the distal esophageal mucosa are parietal and chief cells, Two of the 4 cases showed such cells; in one case they were found only in one biopsy, in the other case in biopsies obtained at two levels. In the former case, the biopsy showed an area which consisted of typical gastric fundic mucosa with mucous neck glands and well formed crypts of Lieberkiihn with pareital and chief cells. In the second case small nests of parietal cells and fewer and less distinct chief cells were present in mucosa which consisted predominantly of columnar epithelial lined glands. The inconstancy of the parietal and chief cells in Barrett's mucosa is reflected by previous reports. 2,5, As with all biopsy procedures, interpretation of findings must be related to the adequacy of sampling, especially when structural features are nonuniform. This point is emphasized by the findings of parietal and chief cells at only one level in one patient, and the presence of esophageal acid secretion in another patient in whom multiple biopsies showed no parietal cells. REFERENCES 1. Barrett NR: Chronic peptic ulcer of the oesophagus and-oesophagitis. Br J Surg 38: , Burgess IN, Payne WS, Andersen HA, et al: Barrett's esophagus:

8 676 MANGLA ET AL. Vol. 70, No.5 The columnar-epithelial-lined lower esophagus. Mayo Clin Proc 46: , Berquist TH, Nolan NG, Carlson HC, et al: Diagnosis of Barrett's esophagus by pertechnetate scintigraphy. Mayo Clin Proc 48: , Jones FA, Gummer JWP: Clinical Gastroenterology. Springfield, Illinois, Charles C Thomas, 1960, p Ustach TJ, Tobon F, Schuster MM: Demonstration of acid secretion from esophageal mucosa in Barrett's ulcer. Gastrointest Endosc 16:90-100, Hershfield NG, Lind JF, Hildes JA, et al: Secretory function of Barrett's epithelium. Gut 6: , Mangla JC, Kim Y, Guarasci G, et al: Pepsinogens in epithelium of Barrett's esophagus. Gastroenterology 65: , Turner MD, Mangla JC, Samloff 1M, et al: Studies on the heterogeneity of human gastric zymogens. Biochem J 116: , Samloff 1M: Slow moving protease and the seven pepsinogens: electrophoretic demonstration of the existence of eight proteolytic fractions in human gastric mucosa. Gastroenterology 57: , M angla JC, Guarasci G, Turner MD: Simultaneous electrophoresis of pepsinogens and pepsins. Am J Dig Dis 18: , Yalow RS, Berson SA: Radioimmunoassay of gastrin. Gastroenterology 58:1-14, Seijffers MJ, Segal HL, Miller LL: Separation of pepsinogen I, pepsinogen II, and pepsinogen III from human gastric mucosa. Am J Physiol 205: , Anson ML, Mirsky AE: The estimation of pepsin with haemoglobin. J Gen Physiol 16:59-63, Jewett TC Jr, Duszynski DO, Allen JE: The visualization of Meckel's diverticulum with mtc-pertechnetate. Surgery 68: , Berquist TH, Nolan NG, Adson MA, et al: Diagnosis of Meckel's diverticulum by radioisotope scanning. Mayo Clin Proc 48:98-102, Jordan PH, Longhi EH: Diagnosis and treatment of an esophageal stricture (ring) in a patient with Barrett's epithelium. Ann Surg 3: , Allison PR, Johnstone AS: The oesophagus lined with gastric mucous membrane. Thorax 8:87-101, Hershfield NB, Lind JF, Hildes JA, et al: Secretory function of Barrett's epithelium. Gut 6: , Abrams L, Heath D: Lower esophagus lined with intestinal and gastric epithelia. Thorax 20:66-72, Adler RH: The lower esophagus lined by columnar epithelium: its association with hiatal hernia, ulcer, stricture, and tumor. J Thorac Cardiovasc Surg 45:13-34, Trier JS: Morphology of the epithelium of the distal esophagus in patients with midesophageal peptic strictures. Gastroenterology 58: , Goldman MC, Beckman RC: Barrett syndrome: case report with discussion about concepts of pathogenesis. Gastroenterology 39: , Mossberg SM: The columnar-lined esophagus (Barrett syndrome): an acquired condition? Gastroenterology 50: , 1965