Multiple esophageal suction biopsies were obtained from 5 patients with midesophageal peptic strictures and studied by light and electron microscopy. All biopsies from above the stricture were lined by squamous epithelium, and all from below were lined by columnar epithelium, with the latter containing three cell types. The most abundant was a columnar epithelial cell which contained many glycoprotein granules, and in which the apical surface consisted of a brush border composed of many microvilli. Many mucus-secreting goblet cells and a few chromaffin cells were also observed but no cells resembling gastric parietal or chief cells. Biopsies from the gastroesophageal junction from 2 normal subjects differed from the distal esophageal epithelium of stricture patients in that surface cells contained only sparse microvilli and more abundant secretory granules and intestinal type goblet cells were ab

A method for continuous monitoring of esophageal pH for periods of up to 18 hr was used to record the timing and duration of reflux in 15 patients with symptomatic sliding hiatal hernia, and in 11 control subjects. A gastrointestinal glass pH electrode was positioned to record at the junction of the middle and lower thirds of the esophagus and reflux periods were considered to have occurred if the pH was below 4. In 15 patients, maximal acid output (mEq/hr) was determined by use of 1.v. histamine infusion, 0.04 mg/kg/hr, or 1.m. pentagastrin 6g/kg. Mean 18-hr reflux periods were significantly longer in the patients with hiatal hernia (266.9 min. ± 282.3), than in the control group (41.1 min. ± 17.0) and reflux occurred (on the average) during 25% of the 18-hr recording in the hiatal hernia patients. Nocturnal reflux occurred in the majority of patients with hiatal hernia, but in none of the control subjects. In five control subjects with duodenal ulcer, there was a significant positive correlation between maximal acid output and the duration of reflux, but no such correlation was present in patients with hiatal hernia.

8 yr following surgery after having experienced slow hepatic deterioration and late neurological disease, and 1 patient is well 2 yr postoperatively. Of the 9 patients receiving splenectomies and spleno renal shunts, 5 patients died within 2 yr while 4 are still alive. Splenectomy in 1 patient was followed by neurological deterioration, with so recurrence of bleeding and the patient is still alive 6 yr after surgery. One patient with gastric transection and another with porta-azygos disconnection died of recurrent hemorrhage. Of 4 patients who received no surgical treatment, 1 is alive, 2 died of recurrent hemorrhage and 1 died of progressive hepatic and neurological deterioration. Treatment with penicillamine or dimercaprol prevented neither initial bleeding nor postoperative neurological or psychiatric deterioration. Incidence and severity of neurological and emotional deterioration in patients with surgery are both significantly greater than in medically treated patients with Wilson's disease, but without esophageal hemor rhage. Surgical decompression of portal hypertension should seemingly be reserved for patients in whom recurrent and uncontrollable hemorrhage threatens life.

3338

HIDDEN GASTRIC AUTOANTIBODIES TO INTRINSIC FACTOR IN PERNICIOUS ANEMIA. (E.) Goldberg, L. S. (U. California Sch. Med., Los Angeles) and R. Bluestone. J Lab Clin Med 75(3): 449-456, 1970.

Sera and gastric juices were analyzed for presence of antibody to intrinsic factor (IF) in 12 patients with pernicious anemia (including one with coexistent agammaglobulinemia). IF content of the eluted peaks was greater than that of concentrated gastric juices, when 5 of gastric juice samples were fractioned by acidic gel filtration. IF was not detected in 8 of 12 unconcentrated gastric juice samples from subjects with pernicious anemia. Four specimens contained 74, 80, 54 and 40 pg. of IF per 0.1 ml resp and when concentrated samples of gastric juice were tested, larger amounts of IF (383 to 660 pg. per 0.1 ml) were present in the 4 specimens but not in the other 8. Blocking antibodies were detected in 6 of 12 concentrated samples of whole gastric juice (875 to 4,840 pg. per 0.1 ml); one of these also contained the binding antibody. Blocking antibody was found in 4 additional specimens only after subjecting concentrated gastric juice to acidic gel filtration. Gastric antibodies were predominantly IgG immunoglobulins, except for one which appeared to be a secretory IgA antibody. Gastric juices of patients with pernicious anemia apparently frequently contain autoantibodies to IF as well as IF and blocking autoantibodies and the IF seemingly exist as an antigen-antibody complex.

Mild atrophic lesions (thinning of the mucosa, decrease of the mucosal volume and significant reduction of parietal cell mass), without signs of inflammation or hypersensitivity, were produced in gastric mucosa of rats injected for 6 to 8 weeks with immunoglobin G processed from sera of patients with atrophic gastritis and pernicious anemia containing circulating parietal cell antibody. The parallel between the reduction of the parietal cell mass and decrease of the mucosal volume, as well as decrease in the intrinsic factor activity of the gastric juice of rats treated for 8 weeks with parietal cell antibodies containing immunoglobulin G, suggests concomitant reduction of peptic cell mass under these circumstances. Rats treated with parietal cell antibodies also demonstrated a profound decrease of the hydrochloric acid output, which was more pronounced than was warranted by the coexisting reduction of the parietal cell mass. Parietal cell antibodies may play a contributory role in the natural history of gastric atrophic lesions in man by producing the reduction of cell mass and cell output.

Carcinogenic activity of air pollutant particles was observed in stomachs of CFW mice fed (in Purina chov) thermal carbon and lamp black, fiber glass filter and particulate matter obtained from the air in the vicinity of petrochemical plants. There was no sig nificant weight change in mice fed various types of hydrocarbons. Papillomas and squamous cell carcinomas occurred in the squamous portion of the stomach with numbers ranging from 1-3 and small in size, but neoplasms were not present in remainder of gastrointestinal tract. Similar gastric tumors occurred in mice fed a control carbon diet. Carbon pigment was usually present in the gastrointestinal tract of mice fed carbon and lamp blacks but no inflammatory reaction was observed. Pulmonary adenomas were present in experimental and control groups (no significant difference in number between lung tumors in treated and control groups). Papillomas and papillary carcinomas readily occur in squamous por tion of stomach of mice fed food pellets with benzo (a) pyrene added. The experimental model for carcinogen study in air pollutants would seemingly be useful in clinical and epidemiologic study of gastric cancer in air polluted areas.

3342

HISTOCHEMICAL STUDY OF THE MUCOSUBSTANCES IN THE CANINE STOMACH: III. THE EFFECT OF CORTICOSTEROIDS. (E.) Lev, R. (New York Med. Coll., New York), H. I. Siegel and G. B. Jerzy Glass. Gastroenterology 58(4):495-508, 1970.

Prednisone (1.5 and 3.0 mg/kg for up to 8 weeks) was administered p.o. to 3 dogs with total gastric fistulas and mild chronic gastritis prior to treatment. Prednisone induced hyperplasia of mucous cells in the fundic and antral surface epithelium and increased synthesis of mucus by these cells and the extent of the hyperplastic response depended upon the dose of the drug and the length of time that it was administered. Prolonged treatment with prednisone resulted in erosions and one ulcer, and there was concomitant histological evidence of widespread mucosal damage such as the

MAY 1970

355

Microbial population and bile salt metabolism at different levels of the alimentary tract were studied in 15 patients with disorders of the stomach and small intestine (4 with polya partial gastrectomy, 3 with single duodenal diverticulum, 4 with multiple jejunal diverticulosis, 2 with lesions of the distal small intestine, and 2 with resection of the distal small intestine). Although many patients had an extensive growth of microorganisms such as coliforms and lactobacilli throughout the small bowel irrespective of the causative lesion, the presence of anerobic microorganisms and free bile acids was related to local anatomical defects (areas of stagnation). The micro-environment (low oxidation-reduction potential) necessary for growth of microorganisms of the bacteroides group appears to be available in areas of intestinal stasis and this requirement is fulfilled in the colon or terminal ileum of normal subjects and in the stagnant area of small bowel in patients with diverticula, strictures, or blind loops. Treatment of 2 of these patients with lincomycin (specifically effective in eradicating the anerobic flora) led to a reduction in steatorrhea and disappearance of free bile acids from the intestinal fluid. Vitamin B12 malabsorption appeared to be related to the total number of bacteria colonizing the small bowel rather than to any specific type. In 1 patient, the broad-spectrum antibiotic tetracycline was effective in eradicating an abnormal bacterial flora.