IN VITRO STUDIES OF ULCERATIVE ILEOJEJUNITIS

GASTROENTEROLOGY 08:572-582, 1975 Copyright 1975 by The Williams & Wilkins Co. Vol. 68. No.3 Printed in U.S.A. IN VITRO STUDIES OF ULCERATIVE ILEOJEJUNITIS HAYDEN L. KLAEVEMAN, M.D., ROGER L. GEBHARD, M.D., CLEMENTINE SESSOMS, AND WARREN STROBER, M.D. Section on Gastroenterology, Digestive Diseases Branch, National Institute of Arthritis, Metabolism and Digestive Diseases; and Immunophysiology Section, Metabolism Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland A patient with ulcerative ileojejunitis was studied by determination of HL-A phenotype, by measurement of jejunal IgA synthesis using a labeled amino acid incorporation technique, and by in vitro organ culture. The patient carried the HL-A8 phenotype in common with 87.5% of patients with gluten-sensitive enteropathy. During a period of exposure to dietary gluten, jejunal tissue from tht! patient exhibited a high IgA synthetic rate. In organ culture of the jejunal biopsy specimen there was an increase in alkaline phosphatase activity in the absence, but not in the presence, of gluten peptides. The synthetic rate value and the organ culture behavior are similar to those observed in patients with gluten-sensitive enteropathy in exacerbation. During a period in which the patient was not exposed to dietary gluten, including a prolonged period of intravenous alimentation, jejunal IgA synthesis and organ culture behavior were studied repeatedly. In contrast to patients with gluten-sensitive enteropathy in remission (i.e., on a gluten-free diet), jejunal IgA synthesis did not decline. However, in organ culture in the patient's tissue now behaved like that of other patients with gluten-sensitive enteropathy in remission: alkaline phosphatase activity increased during culture in the presence and absence of gluten peptides. These studies support the concept that ulcerative ileojejunitis is a complication of gluten-sensitive enteropathy in which escape from control by gluten restriction has occurred. They suggest that ulcerative ileojejunitis is due to a supervening pathological process which is, at least in part, immunological in nature. Chronic ulcerative ileojejunitis (UIJ) is a rare illness of undetermined etiology, with certain similarities to gluten-sensitive enteropathy (GSE, celiac sprue) and with a high morbidity and mortality. Since the first report in 1949 by Nyman I there have Received April 23, 1974. Accepted September 4, 1974. Address requests for reprints to: Dr. Warren Strober, Building 10, Room 4N1l4, National Institutes of Health, Bethesda, Maryland 20014. The authors wish to thank Dr. Robert A. Fischer for referring this patient for study, Dr. John H. Hardley of Johns Hopkins School of Medicine for reviewing t he biopsy material, and Mrs. Blanche Fors for her assistai)ce preparing this manuscript. 572 been 27 patients reported in the English literature, of whom 19 are known to be dead,2-4 The major features of UIJ are generalized malabsorption, usually severe and associated with profuse diarrhea,5 histology of the small intestinal mucosa characterized by villous atrophy with an intense mononuclear infiltrate, 5-8 and diffuse mucosal ulceration of the small bowel The majority of the patients with UIJ had a favorable response to a gluten-free diet, although often this response was not sustained,8 In 14 patients, jejunal biopsies had histological findings characteristic of GSE but the lesion had a diffuse distribution throughout the small bowel, in con-

March 1975 CASE REPORTS 573 trast to the proximal location of the mucosal abnormality in GSE. 3-5, 8-10 Patients with active BSE have increased local mucosal immunoglobulin synthesis 11 of which a large fraction is specific for gliadin, the toxic agent in gluten-containing foods. 12 Additionally, these patients have an increased likelihood of carrying the histocompatibility antigen HL-A8. 13 In organ culture, jejunal biopsy specimens from patients with GSE show gluten-induced alterations in brush border enzyme activities which are not seen in biopsy specimens from control subjects. 1 In this report we have investigated the behavior in organ culture as well as local mucosal immunoglobulin responses of tissue from a patient with UIJ in order to gain further insight into the cause of the disease. These studies support the concept that UIJ is related pathogenetically to GSE, but in UIJ an additional pathological process supervenes which leads to diffuse small bowel disease in the absence of dietary gluten. Case Report The patient, a 42-year-old Jewish male of Northern European extraction (patient N. G., National Institutes of Health No. 08-97-42-5), was referred to the National Institutes of Health in May 1972 for inclusion in a study of glutensensitive enteropathy. The patient had been healthy as a child and young adult. During a visit to Israel in 1967 he indulged in his mother's pastries. This was followed by a brief period of diarrhea, abdominal bloating, and mild weight loss. His stools, which rarely exceeded three per day, were loose, watery, foul, and floating, and were associated with mucus but no blood. In 1968 a similar episode occurred after another trip to Israel. In December 1970, the patient returned from a third trip to Israel with more fulminant diarrhea and a 10-kg weight loss which led to admission to a local hospital in early 1971. On a regular hospital diet he continued to have diarrhea and lost an additional 5 kg. An upper gastrointestinal series showed a postbulbar ulcer, thickened duodenal folds, and dilution, flocculation, and segmentation of barium in the small bowel. A barium enema was normal. An intestinal biopsy was consistent with a diagnosis of GSE. The patient was placed on a gluten-free diet and had a prompt remission of his symptoms with a weight gain of 15 kg over 2 months. In July 1971, while in Israel, the patient abandoned the gluten-free diet and suffered a recurrence of the abdominal symptoms. Over the next 6 months he followed his diet erratically. In late December 1971, the patient developed partial small bowel obstruction necessitating laparotomy. At surgery he was found to have three areas of scarring and stenosis in the midjejunum. An enteroenterostomy was constructed bypassing the area of obstruction. No biopsies were taken. Postoperatively he did well on a strict gluten-free diet and returned to his normal weight of 68 kg. In May 1972, the patient was seen at NIH for the first time. He reported that diarrhea and abdominal distention had recurred. He gave no family history of celiac disease, and denied ingesting enteric-coated potassium, corticosteroids, azulfidine, or aspirin. There was no known exposure to tuberculosis. The only abnormalities on physical examination were digital clubbing and trace pedal edema. His weight was 60.9 kg. Laboratory values included a normal hemoglobin and hematocrit, a white cell count of 9000 with a differential that included 56% mature polymorphonuclear leukocytes, 5% band forms, 26% lymphocytes, 2% e o s i n o p h i l and 3% basophils. The urine was normal. Stools contained no occult blood, parasites, or pathogenic organisms. Other tests included serum cholesterol of 140 mg per 100 ml (normal, 150 to 250 mg per 100 ml), albumin of 3.2 g per 100 ml (normal, 3.1 to 5.4 g per 100 ml), serum carotene of 2.2 g per 100 ml (normal greater thlid 50 Ilg per 100 ml), fecal fat excretion of 44.5 g per day on 100-g fat diet (normal, less than 5 g per day), and D-xylose excretion of 2.5 g in a 5-hr collection of urine after a 25-g oral dose (normal, greater than 5 g). Serum electrolytes, urea nitrogen, creatinine clearance, prothrombin time, triiodothyronine, tetraiodothyronine, calcium, magnesium, and serum immunoglobulins were normal. Serum vitamin B12 was 325 pg per ml (normal, 200 to 900 pg per ml), and folate was 7 pg per ml (normal, 5 to 21 pg per ml). Chest X-ray, sigmoidoscopy, barium enema, and electrocardiogram were normal. An upper gastrointestinal series showed no abnormality of the esophagus, stomach, or duodenum. Small bowel series showed an abnormal loop of bowel in the left upper quadrant with nodular indentations along its margin. A peroral jejunal biopsy specimen showed complete villous atrophy with cuboidal pyknotic epithelial cells. There was a dense infiltrate of plasma cells, eosinophils, and histiocytes (fig. 1).

574 CASE REPORTS Vol. 68, No.3 B FIG. 1. A, low power view of jejunal biopsy obtained during May 1972 admission showing villous atrophy with marked inflammatory cell infiltrate (x 140). B, high power view of same biopsy showing pyknotic cuboidal epithelial cells and small round cell infiltrate (x 350). At the end of the challenge period, he had no increase in stool frequency and a repeat intestinal biopsy was unchanged from that obtained prior to challenge (which already showed maximal villous flattening). However, the fecal fat excretion was now 68.7 g per day and the o-xylose excretion was now 1.8 g per 5 hr. The patient was admitted 3 1 /2 weeks later for

March 1975 CASE REPORTS 575 further evaluation. In the interim on a glutenfree diet, he had lost 2 V2 kg and his diarrhea persisted. There was no change in his fecal fat excretion, o-xylose excretion, or serum carotene. A 2-hr gastric analysis revealed low basal acid secretion but normal stimulation with Histalog. In mid-august 1972, because of continued gastrointestinal symptomatology, the patient was begun on tetracycline in order to treat possible bacterial overgrowth secondary to a surgically induced blind loop. On the antibiotic, his bowel movements became formed but he did not gain weight. Consequently, in mid-october 1972, he was admitted for the third time for further evaluation. Physical examination was unchanged. Malabsorption studies included a carotene of 34.9 f.lg per 100 ml, o-xylose of 4.5 g, and fecal fat excretion of 13.2 g per day. Serum albumin had declined to 2.1 g per 100 ml. Serum vitamin B I level was 115 pg per ml, and - he excreted only 1.7% of an oral dose of 57CO B I., given with intrinsic factor, in 24 hr (normal, > 15%). Duodenal aspirate grew Klebsiella resistant to tetracycline. The combination of malabsorption of vitamin B I with symptomatic improvement on tetracycline, as well as improvement in serum carotene, D-xylose, and fecal fat excretion, suggested that some component of his illness at this point was due to bacterial overgrowth in the surgical blind loop. The patient returned to NIH in early November 1972, complaining of anorexia, left upper quadrant pain, abdominal distention, nausea, vomiting, and occasional oral temperatures of 38.6 C in the evenings. On admission his weight was 56.5 kg, he was afebrile, and his physical examination was normal. Laboratory values included a white blood cell count of 9100 with a differential of 23% polys, 35% band forms, 24% lymphocytes, 16% monocytes, and 2% eosinophils. The stool contained no occult blood or parasites, and on culture grew Escherichia coli and Klebsiella. A duodenal aspirate had no anaerobic growth and only normal throat flora aerobically. A gastrointestinal series showed marked changes from the previous study. The third portion of the duodenum was thickened. The proximal jejunum, beginning at the ligament of Treitz and extending for several feet, was narrowed and effaced. The distal small bowel showed dilation, with hyperdilution of the barium. There were no radiographically visible ulcerations (fig. 2). As a result of this obvious deterioration in the patient's condition, an exploratory laparotomy was performed in early January 1973. At opera- FIG. 2. Small bowel series from November 1972. The 30-min film shows thickened proximal sma ll bowel with narrowing and effacement of the lumen. tion the jejunal mesentery was markedly thickened, measuring 4 to 5 cm. There was no fat seen on the antimesenteric border of the intestine, and there were no fissures or fistulae. The upper jejunum had several incomplete strictures. The area of the blind loop appeared normal externally, as did the ileum and colon. The blind loop with contiguous upper jejunum was excised as were several mesenteric lymph nodes, and a full thickness biopsy of the upper jejunum was taken. On examination of the tissue there were multiple small ulcerations in the blind loop and upper jejunum (fig. 3). All regions excised exhibited blunting, shortening, and fusion of the villi and a dense submucosal infiltrate of plasma cells. There were no granulomas seen, nor evidence of lymphoma. Mesenteric nodes showed only follicular hyperplasia. Postoperatively the patient was given intravenous hyperalimentation with 3000 cal per day for 1 month to determine whether complete absence of gluten would improve his mucosal histology. Repeat upper gastrointestinal series 3 1 / 2 weeks postoperatively showed no abnormality.of the proximal jejunum. Peroral jejunal biopsy in early February 1973, while the patient was still receiving intravenous hyperalimentation, showed severe flattening and blunting of the villi with a dense plasma cell infiltrate. There was no improvement in his serum caro-

576 CASE REPORTS Vol. 68, No.3 FIG. 3. Surgical specimen (January 1973) from upper jejunum showing margin of ulcer crater. Adjacent mucosa is atrophic (x 55). tene or o-xylose excretion. The patient was discharged on February 8, 1973 on a diet eliminating gluten, milk, and eggs. Over the next month, in spite of strict adherence to this diet, the patient lost 3 kg in weight coincident with the recurrence of anorexia, nausea, and intermittent diarrhea with abdominal cramps. By early March 1973, he was cachectic and had trace pedal edema. Upper gastrointestinal series showed marked deterioration from his posthyperalimentation study with disruption and distortion of the duodenal mucosa. In the proximal jejunum there was loss of the feathery mucosal pattern and the terminal ileum was abnormally dilated. He continued to have evidence of generalized malabsorption. In view of the patient's continued deterioration, in mid-march 1973, he was placed on oral prednisone, 60 mg per day for 2 weeks, followed by 60 mg every other day for 2 weeks. During this month, his weight did not change significantly. An upper gastrointestinal series in mid April 1973 showed no significant change; however, on repeat jejunal biopsy there was definitely less villous atrophy and inflammatory infiltrate, and the brush border appeared intact over the tips of the villi. Tests of malabsorption continued to be abnormal. In mid-april 1973, his prednisone was decreased to its present maintenance level of 30 mg every other day. In late April 1973, the patient developed acute partial small bowel obstruction which responded promptly to decompression with a Miller-Abbot tube. The obstruction was attributed to a surgical adhesion with kinking of the bowel rather than underlying disease, since no strictures were demonstrable on upper gastrointestinal series. After this episode he improved dramatically. Gastrointestinal complications subsided, and his weight, having reached a nadir of 44.8 kg in May, increased to 68 kg by

March 1975 CASE REPORTS 577 September 1973. Jejunal biopsy in October 1973 showed further improvement in villous architecture and no inflammatory infiltrate (fig. 4). Methods Clinical studies. Fecal fat excretion was evaluated by collection of a 72-hr stool sample while the patient was on a 100-g fat diet and measuring stool fat according to the method of van de Kamer et al. 15 o-xylose absorption was evaluated by administering 25 g of the pentose orally with subsequent measurement of xylose excretion by the method of Roe and Rice '6 in a 5-hr urine collection. Serum carotene was determined by extraction of serum with alcohol and petroleum ether.'7 HL-A typing. HL-A antigens were determined by Dr. G. N. Rogentine by the use of a lymphocyte microcytotoxicity method, as previously described. '8 Sera used to detect HL-A antigens were obtained from the Serum Bank maintained by the Transplantation and Immunology Branch of the National Institute of Allergy and Infectious Diseases. Special studies. Intestinal biopsies were obtained with a four-hole Rubin tube biopsy instrument positioned fluoroscopically at the ligament of Treitz.'9 The biopsy specimens were immediately placed into Krebs-Ringer bicarbonate buffer (ph 7.3) (4 C), rinsed three times in the same buffer, then cut into pieces approximately 1 to 2 mm in diameter. Organ culture of biopsy tissue was performed as outlined previously, H. 20 in the presence and absence of gluten peptides. Alkaline phosphatase activity was assayed by a modification H of the method of Bessey et al. 21 using p-nitrophenyl phosphate as a substrate. Assay of L- [HC Jleucine incorporation into JgA. The assay procedure used to measure L- [HC Jleucine incorporation has been described in detail previously." 12 Briefly, biopsy specimens were incubated with L- [HC Jleucine in a modified Krebs-Ringer buffer for 90 min. The tissue was homogenized, and the supernate from the homogenate was incubated for 3 hr at 37 C with anti-iga antibodies linked covalently to bromacetyl cellulose (BAC). The BAC antibody was then separated from the incubation mixture by filtration on fiberglass filters, washed, and placed into scintillation vials for counting. The extent of nonspecific binding of radioactivity to the BAC anti-iga was assessed with BAC anti-iga previously rendered immunologically inactive with excess nonradioactive immunoglobulin. Values obtained for nonspecific binding with inactivated BAC anti-iga were subtracted from the total binding value obtained with active BAC anti-iga. The final value was normalized by the amount of tissue protein contained in the specimen. Results Histocompatibility type. The patient had a 2/0 in the first segregant series indicating either a double dose of the A2 antigen or the presence of A2 plus an unknown antigen in the allelic locus. In the second segregant series he carried the 7/8 antigens. JgA synthesis by intestinal mucosa in vitro. Incorporation of L- [14C ]leucine into IgA by jejunal tissue obtained from 8 normal individuals was 5830 ± 3190 counts per min per mg of biopsy tissue protein. This value is similar to that obtained previously in normal individuals using similar techniques. 11. 22 In the patient, incorporation of L- [14C ]leucine into IgA by jejunal tissue obtained after the patient was on a gluten-containing diet for 7 days was 32,647 counts per min per mg of protein. Four additional studies while the patient was on a gluten-free diet showed similarly elevated values (34,000 ± 6,500 counts per min per mg) (mean ± 1 SD). However, after institution of steroid therapy, L- [ 14 C ]leucine incorporation values progressively decreased, and, after 2 months of therapy, reached a normal value of 6814 counts per min per mg of tissue protein (fig. 5). Organ culture. Specimens of jejunal biopsy tissue obtained from the patient were maintained for 48 hr in organ culture. Initial and 48-hr alkaline phosphatase values (units per mg tissue) were determined and are recorded in table 1. For comparison, previously obtained values of patients and controls are also listed in table 1.14 When the patient was on a gluten-containing diet, alkaline phosphatase in biopsy specimens cultured in the gluten peptide-free medium increased from an initial value of 108 U to a 48-hr value of 497 U; in contrast, in the presence of gluten peptides, this increase was inhibited and the initial value rose from 108 U to a final value of only 301 U. This behavior in

578 CASE REPORTS Vol. 68, No. 3 culture parallels that of patients with GSE in exacerbation in whom a similar inhibition of alkaline phosphatase was also observed. )4 A In three separate culture studies performed when the patient was on a glutenfree diet and prior to the institution of corticosteroid therapy, the initial alkaline B FIG. 4. A, jejunal biopsy obtained 6 months after institution of corticosteroids. Villi are present and the inflammatory cell infiltrate is reduced (x 225). B, same specimen as in A. Epithelial cells are columnar and the brush border is intact (x 350).

March 1975 CASE REPORTS 579 35 30 25,;;- Q ~ ~ 20 fo- 0 a: Cl. <!) :0 15... :0 Cl. u 10 5 ~ NORMALS (N=8) I.. GSE GLUTEN FREE DIET (N=7) 0 GSE GLUTEN CONTAINING DIET (N=7) FIG. 5. Incorporation of ["ell-leucine into IgA by intestinal mucosa from normals, patients with glutensensitive enteropathy (GSE) and patient N. G. Data are expressed as mean ± SEM. Data for normals, GSE in remission, and GSE on a gluten-containing diet have been reported previously. 11, 12 The number of patients studied is given in parentheses. Incorporation into IgA by jejunal biopsies obtained from the patient while he was on a gluten-free diet is represented by closed circles ; while on a gluten challenge, by the open circle; and after institution of corticosteroids, while on a gluten-free diet, by the closed triangle. phosphatase activity in the biopsy specimen increased from an initial value of 51 ± 34 U (mean ± 1 SD) to a final value of 262 ± 39 U; a similar increase was observed when gluten was added to the culture medium (final value 242 ± 9). The behavior in culture is again similar to that observed in specimens from patients with GSE on a gluten-free diet (or in normal individuals or individuals with nongluten-sensitive gastrointestinal disease) in whom no inhibitory effects are seen in culture by gluten peptides. 14 Discussion Ulceration of the small bowel is a feature of a variety of clinical states including regional enteritis, lymphoma, Zollinger Ellison syndrome, exposure to entericcoated potassium or corticosteroids, and a variety of infectious processes. 5 8 In addition, ulceration of the small bowel has been described in patients grouped under the term " chronic ileo-jejunitis." These patients most frequently have an antecedent history of gluten sensitivity8 and share many of the clinical findings of GSE, including diarrhea with malabsorption of D-xylose, fats, and carotene, as well as malabsorption of iron, folate, and Bu. 5 7, 8 As in GSE these patients have blunting and flattening of villi on small bowel biopsy which is associated with a dense infiltrate of lymphocytes and plasma cells in the lamina propria. 5 7, 8 However, unlike GSE patients this condition is marked by the presence of diffuse, widely disseminated small bowel ulcers with strictures. Most patients are not gluten-sensitive at the time ulcerations are found, and their clinical course is marked by progressive deterioration frequently ending in death from perforation or obstruction at the site of ulceration. 5, 7, 8 In this report we describe a patient who clearly fits into the UIJ category. The patient had malabsorption with weight loss initially responsive to a gluten-free diet, as well as villous flattening on small bowel biopsy. Later he developed intestinal obstruction associated with diffuse small bowel ulceration without evidence of lymphoma, inflammatory disease, or endocrinopathy. Gluten restriction in the ulcerative phase did not reverse the villous flattening or malabsorption. This included a 1 month period during which the patient was maintained on total parenteral nutrition. Therefore, during the fulminant phase of the disease this patient had become unresponsive to gluten restriction and a condition had developed that might be called "gluten escape."

580 CASE REPORTS Vol. 68, No.3 TABLE 1. Intestinal alkaline phosphatase activity during organ culture of jejunal biopsies in normals, patients with gluten-sensitive enteropathy (GSE) and patient N.C. Alkaline phosphatase activity Patients D-hr 48-hr 48-hr with gliadin, Normals (ll).... _ 0- o........... 384 ± 83 561 ± 151 578 ± 156 GSE Gluten-free diet (10)............... 181 ± 157 418 ± 156 354 ± 161 Gluten-containing diet. (9)................ 117 ± 79 399 ± 203 203 ± 93 Patient N. C. Gluten-free diet (3)............ 51 ± 34 262 ± 39 242 ± 9 " Gluten-containing diet (1)......... 108 497 301 Activity is expressed as micromoles of p-nitrophenyl phosphate formed per gram of tissue per minute. Results are expressed as mean ± 1 SD. The number of patients studied is given in parentheses. Data for normals and GSE patients have been reported elsewhere." A pathophysiological understanding of chronic ileojejunitis depends, to a great extent, on an understanding of the relationship of gluten sensitivity to UIJ. Previous reports, in pltrticular that by Bayless et al. 8 have emphasized that GSE and UIJ are chronologically related. In Bayless' review, 16 patients were assembled with preceding GSE or impressive presumptive evidence for GSE. On the other hand, the clinical differences between the two conditions, and in particular the unresponsiveness to gluten restriction in UIJ, have led to speculation that the two conditions are not pathophysiologic ally related. Lending support to this idea is the fact that 8 patients had no preceding history of GSE as defined by responsiveness to gluten restriction. However, 4 of these patients were never placed on a gluten-free diet, 1, 5, 7 and the other 4 presented with ulcerative, nongluten-sensitive disease, 3. 5 having perhaps already moved beyond the glutenresponsive phase. In order to further clarify the relationship between GSE and UIJ we addressed ourselves to two interrelated questions: (1) Is our patient with UIJ similar to patients with GSE by a number of newly developed laboratory criteria of GSE? and (2) assuming similarities between these two conditions are present, in what additional ways do these two conditions differ? These questions were investigated by means of determinations.of HL-A phenotype and gastrointestinal IgA synthesis in a patient with UIJ, as well as a study of the behavior in organ culture of jejunal biopsies obtained from a patient with UIJ. HL-A phenotype. We have reported that HL-A8 frequency in patients with GSE is strikingly increased. 13 This cell surface marker occurred in 87.5% of a large group of patients with GSE, whereas it occurs at a 21 to 22% frequency in normal populations. The relationship between GSE and HL-A8 had been confirmed in other patient populations 23 and is also found in patients with dermatitis herpetiform is, a disease with known association with gluten-sensitive enteropathy. 24, The UIJ patient under study was also phenotyped and found to carry the HL-A8 phenotype. This findi:j.g, although subject to the proviso that HL-A8 may occur in normals and that GSE may occur without HL-A8, lends credence to the idea that GSE and UIJ are related entities. JejunalIgA synthesis. In several studies we have observed that patients with GSE have increased jejunal immunoglobulin synthesis as determined by radioactive amino acid incorporation into immunoglobulin by jejunal biopsies in short term in vitro culture. 11 22 These abnormal local immunoglobulin synthetic rates correlated with the clinical state of the patient: in patients with GSE in remission, local immunoglobulin synthesis was normal or only moderately increased; in patients with

March 1975 CASE REPORTS 581 GSE in exacerbation, immunoglobulin synthesis was markedly increased (2- to 3-fold over the remission study, each patient serving as his own control). In this study we show that the patient with UIJ also had increased jejunal IgA synthesis. In this case, however, IgA synthesis was markedly elevated both after gluten challenge and after the strict gluten-free period, including a period of total parenteral nutrition. Thus, in regard to jejunal IgA synthesis, the patient with UIJ was both similar and dissimilar to ordinary GSE patients: similar in that elevated IgA synthesis was found in both instances, yet dissimilar in that in GSE patients, but not in the patient with UIJ, IgA synthesis was gluten-sensitive. Organ culture of jejunal biopsy tissue. We have previously presented evidence that when jejunal biopsies are placed into organ culture, they regularly undergo an increase in alkaline phosphatase activity. 14 As discussed, this most likely represents a cell maturation phenomenon which occurs in vivo as well. Biopsy specimens obtained from patients with GSE display a lower than normal initial alkaline phosphatase activity which, in common with specimens from normal individuals, increases markedly during organ culture. Biopsy specimens obtained from normal individuals are unaffected by the presence of gluten peptides in the culture medium-alkaline phosphatase activity increases in the presence and absence of gluten peptides. In contrast, specimens obtained from patients with GSE in exacerbation are markedly affected by gluten peptides in the culture medium. In these cases, alkaline phosphatase increases in the absence but not in the presence of gluten peptides. We have called this inhibition of alkaline phosphatase increase by gluten peptides an in vitro model of gluten-sensitive disease in that this effect appears to mimic the in vivo effect of gluten in susceptible patients. Interestingly, whereas gluten peptides affect the behavior of tissue in organ culture from patients with GSE in exacerbation, they have little or no effect on the increase in alkaline phosphatase from patients with GSE in well established remission. We have interpreted this as evidence for the concept that a period of exposure to dietary gluten is necessary to activate a local jejunal mechanism of tissue toxicity before in vivo or in vitro gluten toxicity is seen. To compare the behavior in organ culture of tissue from the patient with UIJ to that of patients with GSE in exacerbation, the patient was challenged with dietary gluten for a period of 7 days. It is noteworthy that this challenge resulted in deterioration of absorptive function as measured by stool fat excretion. In organ culture the jejunal biopsy specimen initially had a low alkaline phosphatase activity which increased in the absence of gluten peptides during a 48-hr period of culture. However, this increase was markedly inhibited when gluten peptides were present in the culture medium. It is clear, therefore, that in respect to both absorptive criteria and cultural criteria, the UIJ patient after gluten challenge was similar to GSE patients in exacerbation. To compare the behavior of tissue in organ culture from the patient with UIJ to that of patients with GSE in remission, tissue was obtained from the patient with UIJ and studied on three separate occasions, including one biopsy obtained after a period of total parenteral nutrition. On these occasions the alkaline phosphatase activity of tissue in organ culture behaved as other specimens obtained from patients with GSE in remission: the alkaline phosphatase activity of specimens increased during the culture period whether the specimens were cultured in the presence or absence of gluten peptides. Thus, the patient had clinically active disease, yet by a laboratory criterion of GSE, he was in remission. To conclude, the organ culture studies just discussed, taken in conjunction with the studies of local IgA synthesis, allow several inferences concerning the pathogenesis of UIJ. First, they suggest that UIJ is indeed a complication of an underlying glut en-sensitive enteropathy and that the gluten-sensitivity persists in spite of the apparent clinical nonresponsiveness to gluten restriction. This is inherent in the fact that alkaline phosphatase changes in the

582 CASE REPORTS VoI.68,No.3 cultured biopsies taken from the patient when on a gluten-free diet and while ingesting gluten were parallel to the enzyme changes in ordinary GSE patients in similar phases with respect to gluten exposure. Second, these studies indicate that the complication is related to a supervening, nongluten-dependent, pathological process which in some manner involves the local immunological system. This follows from the fact that jejunal IgA synthesis in the UIJ patient did not decrease after gluten restriction as it usually does in ordinary GSE. The precise role of this persistent jejunal IgA synthesis in the pathogenesis of UIJ is not presently discernible and is an appropriate subject of subsequent investigation. REFERENCES 1. Nyman E: Ulcerous jejuno-ileitis with symptomatic sprue. Acta Med Scand 134:275-283, 1949 2. Moritz M, Moran JM, Patterson JF: Chronic ulcerative jejunitis. Report of a case and discussion of classification. Gastroenterology 60:96-102, 1971 3. Jones PE, Gleeson MH: Mucosal ulceration and mesenteric lymphadenopathy in coeliac disease. Br Med J 3:212-213, 1973 4. Seliger G, Goldman AB, Firooznia H, et al: Ulceration of the small intestine complicating celiac disease. Am J Dig Dis 18:820-824, 1973 5. Jeffries GH, Steinberg H, Sleisenger MH: Chronic ulcerative (nongranulomatous) jejunitis. Am J Med 44:47-59, 1968 6. Collins JR: Small intestinal mucosal damage with villous atrophy. A review of the literature. Am J Clin Pathol 44:36-44, 1965 7. Goulston KJ, Skyring AP, McGovern VJ: Ulcerative jejunitis associated with malabsorption. Australas Ann Med 14:57-64, 1965 8. Bayless TM, Kapelowitz RF, Shelley WM, et al: Intestinal ulceration-a complication of celiac disease. N Engl J Med 276:996-1102, 1967 9. Case records of the Massachusetts General Hospital. N Engl J Med 28:885-895, 1969 10. Davidson AR: Recurrent benign ileal ulcer occurring with the coeliac syndrome. Br Med J 3:341, 1969 11. Falchuk ZM, Strober W: Increased jejunal immunoglobulin synthesis in patients with nontropical sprue as measured by a solid phase immunoadsorption technique. J Lab Clin Med 79:1004-1013, 1972 12. Falchuk ZM, Laster L, Strober W: Gluten-sensitive enteropathy: intestinal synthesis of antigluten antibody in vitro (abstr). Clin Res 19:390, 1971 13. Falchuk ZM, Rogentine GN, Strober W: Predominance of histocompatibility antigen HL-A8 in patients with gluten-sensitive enteropathy. J Clin Invest 51:1602-1605, 1972 14. Falchuk ZM, Gebhard RL, Sessoms C, et al: An in vitro model of gluten-sensitive enteropathy. Effect of gliadin on intestinal epithelial cells of patients with gluten-sensitive enteropathy in organ culture. J Clin Invest 53:487-500, 1974 15. van de Kamer JH, ten Bokkel Huinink H, Weyers HA: Rapid method for the determination of fat in feces. J Bioi Chern 177:347-355, 1949 16. Roe JH, Rice EW: A photometric method for the determination of free pentoses in animal tissues. J Bioi Chern 173:507-512, 1948 17. Varley H: Practical Clinical Biochemistry. Third edition, Heinemann, London, 1962, p 509-510 18. Mittal KK, Mickey MR, Singal DP, et al: Serotyping for homotransplantation. XVIII. Refinement of microdroplet lymphocyte cytotoxicity test. Transplantation 6:913-927, 1968 19. Brandborg LL, Rubin CE, Quinton WE: A multipurpose instrument for suction biopsy of the esophagus, stomach, small bowel, and colon. Gastroenterology 37:1-16, 1959 20. Browning TH, Trier JS: Organ culture of mucosal biopsies of human small intestine. J Clin Invest 48: 1423-1432, 1969 21. Bessey OA, Lowry OH, Brock MJ: A method for the rapid determination of alkaline phosphatase with five cubic millimeters of serum. J Bioi Chern 164:321-329, 1946 22. Loeb PM, Strober W, Falchuk ZM, et al: Incorporation of L-leucine- 14 C into immunoglobulins by jejunal biopsies of patients with celiac sprue and other gastrointestinal diseases. J Clin Invest 50:559-569, 1971 23. Stokes PL, Asquith P, Holmes GKT, et al: Histocompatibility antigens associated with adult coeliac disease. Lancet 2: 162-164, 1972 24. Gebhard RL, Katz SI, Marks J, et al: HL-A antigen type and small intestinal disease in dermatitis herpetaformis. Lancet 2:760-762, 1973 25. Katz SI, Falchuk ZM, Dahl MV, et al: HL-A8: A genetic link between dermatitis herpetiformis and gluten-sensitive enteropathy. J Clin Invest 51:2977-2980, 1972