Treatment of duodenal ulcer with antacids in combination with trimipramine or cimetidine.

Ninetyfive patients with endoscopically verified duodenal ulcer were randomly allocated in a double blind manner to 3 different treatments. Thirtyfive patients were treated with trimipramine 25 mg nocte, 32 patients with cimetidine 400 mg nocte and 26 patients with cimetidine 200 mg 3 times a day and 400 mg nocte (standard dose). In addition all patients got intensive antacid treatment with LinkR, 20 ml 1 and 3 hours after meals. Two patients were excluded from the trial. After 6 weeks of treatment re-endoscopy revealed healed ulcer in 35 patients (86%) in the trimipramine group, in 32 patients (85%) in the group treated with low dose of cimetidine and in 26 patients (100%) in the group treated with standard dose of cimetidine. The differences in healing rate were not significant. The time until complete relief of pain was significantly longer in the trimipramine group than in the cimetidine groups. The drugs were well tolerated but constipation was reported by 31% of the patients.

The human gastrointestinal secretory immune system in health and disease.

The main function of secretory IgA is to exert immune exclusion; that is, by intimate cooperation with innate non-specific defence mechanisms, it dampens down penetration of soluble antigens and inhibits epithelial colonisation of bacteria and viruses. Secretory IgM may exert a similar protective function in the gut as its local synthesis sometimes is markedly increased, especially in selective IgA deficiency. IgG should not be considered a secretory immunoglobulin because its external translocation depends on passive intercellular diffusion. By activating complement, antibodies of this isotype may cause increased mucosal permeability and tissue damage. IgG may thus contribute to persistent immunopathology in mucosal lesions. The same is true for IgE antibodies which, in atopic individuals, may be carried into the gut mucosa by mast cells and cause their degranulation with histamine release. Secretory IgA and secretory IgM are the products of two cell types: plasma cells synthesise IgA dimers and IgM pentamers which, by non-covalent association, become complexed with the secretory component (SC) which is synthesized by serous-type glandular cells. The adsorption of the Ig polymers to the SC-expressing epithelial cells depends on J chain-determined binding sites. This fact gives biological significance to the striking J chain expression shown by mucosal immunocytes regardless of the Ig class they produce. The immunocytes populating the gut mucosa apparently belong to relatively early memory B cell clones. The obvious functional goal of J chain expression at this stage of clonal differentiation is local generation of SC-binding IgA and IgM polymers. In various gut diseases, altered immune regulation results in a disproportionately increased number of J chain-negative IgG-producing cells in the mucosa. Such altered immunological homeostasis may contribute to perpetuation of inflammatory bowel diseases.

Terminally differentiated human intestinal B cells. IgA and IgG subclass-producing immunocytes in the distal ileum, including Peyer's patches, compared with lymph nodes and palatine tonsils.

The relative distribution of IgA and IgG subclass-producing immunocytes was examined by two-colour immunohistochemistry in normal human distal ileum including Peyer's patches (PP), regional mesenteric lymph nodes (MLN), and peripheral lymph nodes. IgA2 cells predominated slightly over IgA1 cells in the PP dome area. There was a decreasing median proportion of IgA2 cells in the order of PP (52%), distant ileal lamina propria (40%), MLN (32%), and peripheral lymph nodes (11%). The reverse was true for IgA1 cells in independent enumerations. These results support the notion that PP-derived B cells after stimulation are seeded mainly to the lamina propria of the distal gut, but that there is a substantial retention and terminal differentiation of this migrating population in regional MLN. The median subclass proportions of IgG-producing cells in the PP dome area were in independent determinations 68% IgG1, 23% IgG2, 8% IgG3, and 9% IgG4. This distribution was fairly similar to that seen in other tissue categories, except for a trend towards increased IgG1 and reduced IgG2 proportions in peripheral lymph nodes and reduced IgG1 along with increased IgG3 in normal palatine tonsils. The data suggested an association between the expression of IgG2 (and possibly IgG4) and IgA2 in intestinal mucosal immune responses.

Immunomorphological characteristics of human Peyer's patches.

Human follicle-associated epithelium (FAE) was found not to express the secretory component (SC) or polymeric immunoglobulin (pIg) receptor, and is therefore unable to transport pIgA to the gut lumen. However, human FAE (except the M cells) was positive for MHC class II (HLA-DR) determinants. It may therefore perform class II-restricted uptake and presentation to T cells of antigens that have been adequately processed, whereas the function of the M cells is perhaps limited to transport of particulate or undergraded luminal antigens to subjacent antigen-presenting cells (APC). A large number of APC of the inter-digitating cell phenotype (positive for HLA-DR and S-100 protein) and relatively few L1-positive reactive macrophages were found in PP between the follicles and beneath the FAE, compared with villous mucosa. Also, there were significantly more intra- and subepithelial T cells in PP than in distant villi, and the T cells were concentrated adjacent to the M cells. The proportion of the CD4+ subset (putative helper T cells) was much higher in FAE (40%) than in villous epithelium where the CD8+ subset (putative suppressor T cells) predominated strikingly (90%). This disparity might reflect differences in capacity for positive and negative immune regulation at the two sites. The relatively few B cells terminating with Ig production in PP apparently belonged to relatively mature memory clones as they showed a large proportion of IgG immunocytes and reduced J-chain expression. Conversely, both IgG and IgA immunocytes in distant mucosa showed a high percentage of J-chain positivity (80-100%); such expression was also considerable (45-60%) in mesenteric lymph nodes (MLN) in contrast to peripheral lymph nodes (PLN) and palatine tonsils. Moreover, there was a decreasing percentage of IgA2 immunocytes in the order of PP (52%), distant ileal mucosa (40%), MLN (32%), PLN (11%), tand tonsils (5%). These results support the notion that migration of relatively immature memory B-cell clones takes place from PP through MLN preferentially to distant intestinal mucosa.

Terminally differentiated human intestinal B cells. J chain expression of IgA and IgG subclass-producing immunocytes in the distal ileum compared with mesenteric and peripheral lymph nodes.

Two-colour immunofluorescence staining for intracellular J chain and IgA (or J chain and IgG) was performed on tissue sections of normal human ileal mucosa (eight adult kidney donors), mesenteric lymph nodes (MLN), peripheral lymph nodes, and palatine tonsils. The most prominent J chain positivity was seen for IgA (97.3%) and IgG (81.7%) immunocytes in the ileal lamina propria (LP). Moreover, the proportion of J chain-expressing extrafollicular immunocytes was significantly higher (P less than 0.05) in MLN than in peripheral lymph nodes for the IgA class (58.5% versus 25.6%); the same proportion for the IgG class was 45.9% versus 30.4%. In clinically normal palatine tonsils of adults, extrafollicular J chain expression was much lower than in peripheral lymph nodes; 14.2% for IgA cells and 5.5% for IgG cells. When related to subclass production, J chain expression was found to be higher for IgA2 than for IgA1 cells in all tissues examined (palatine tonsils excluded because of a small number of IgA2 cells), the difference being significant in MLN and ileal LP (P less than 0.05). The J chain positivity tended to be higher for all IgG subclasses in MLN than in peripheral lymph nodes; this difference was significant (P less than 0.05) for IgG2-producing immunocytes. Taking J chain expression as a marker of clonal immaturity, our results may reflect to some extent distribution of newly generated memory B cell clones from gut-associated lymphoid tissue to MLN, peripheral lymph nodes, and palatine tonsils in a strikingly decreasing order.

Lack of relation between expression of HLA-DR and secretory component (SC) in follicle-associated epithelium of human Peyer's patches.

Follicle-associated epithelium (FAE) of normal human Peyer's patches (PP) was studied with regard to expression of HLA-DR determinants and secretory component (SC); the latter acts as a receptor for polymeric immunoglobulins (pIg). Putative M cells were identified in FAE by lack of a brush border with alkaline phosphatase. These cells were virtually negative for HLA-DR whereas the remaining FAE was strongly positive like villous epithelium. Conversely, the complete FAE showed no SC expression and was negative for IgA. These findings suggested that the FAE (including the M cells) does not participate in SC-mediated transport of pIgA, which in the gut mainly takes place through columnar crypt cells. The FAE (excepting the M cells) may be involved in an MHC class II-restricted antigen-presenting function as recently suggested for villous epithelium. The role of M cells may hence be limited to uptake and transport of luminal antigens.

Immunoglobulin- and J chain-producing cells associated with lymphoid follicles in the human appendix, colon and ileum, including Peyer's patches.

Immunohistochemistry of B cells associated with normal human Peyer's patches and solitary lymphoid follicles of the ileum, colon and appendix mucosa showed that local accumulation of IgG-producing cells is a common feature of gut-associated lymphoid tissue (GALT). These immunocytes have strikingly down-regulated J-chain expression, indicating that they belong to mature memory clones. They are located mainly in the dome areas, alongside the follicles, and to a lesser extent in the germinal centres, and are accompanied by a much smaller number of J-chain negative IgA- and IgM-producing cells. It is concluded that B cells of mature memory clones are retained in GALT, whereas relatively early counterparts with a high J chain-expressing potential probably emigrate rapidly after stimulation and seed distant secretory sites where they undergo terminal differentiation to produce mainly J chain-containing dimeric IgA.

Human Peyer's patches: lympho-epithelial relationships and characteristics of immunoglobulin-producing cells.

Human Peyers patches (PP) were studied by immunohistochemistry to characterize functional properties of the follicle-associated epithelium (FAE) including the "membrane" (M) cells. The FAE had no transporting capacity for polymeric IgA (pIgA) because it did not express the secretory component (SC) which acts as a pIgA receptor. However, it expressed MHC class II (HLA-DR) determinants, except for the M cells (which were tentatively identified by absence of brush border alkaline phosphatase). It is possible, therefore, that the FAE generally performs class II-restricted transport and presentation to T cells of antigens which have been adequately processed in the gut lumen. The function of M cells may be limited to transport of particulate or undegraded antigens to subjacent macrophages for processing and subsequent presentation. There were significantly more intra- and subepithelial T cells in PP than in distant villi, and the T cells were concentrated adjacent to M cells. The proportion of the CD4+ phenotype (putative helper T cells) was much higher in FAE (approximately 40%) than in villous epithelium where the CD8+ (putative suppressor) phenotype predominated strikingly (approximately 90%). This disparity might reflect differences in capacity for positive and negative immune regulation at the two sites. The B cells terminating with Ig production in PP and adjacent to solitary lymphoid follicles apparently belonged to relatively mature memory clones as they showed a large proportion of IgG immunocytes and reduced J-chain expression. Conversely, both IgG and IgA immunocytes in lamina propria (LP) showed a high percentage of J-chain positivity (80-100%); such positivity was also considerable (45-60%) in mesenteric lymph nodes (MLN) in contrast to peripheral lymph nodes (PLN) and palatine tonsils (PT). Moreover, there was a decreasing percentage of IgA2 immunocytes in the order of PP (52%), distant ileal LP (40%), MLN (32%), PLN (11%), and PT (5%). Taken together, our results suggested that dissemination of relatively immature memory B-cell clones with high J-chain expression takes place from PP through MLN and that preferential settlement of such clones occurs in LP.

T cell distribution is different in follicle-associated epithelium of human Peyer's patches and villous epithelium.

Immunohistochemical analyses performed on specimens of normal human ileum showed a significantly raised number of T cells in the follicle-associated epithelium (FAE) of Peyer's patches compared with the epithelium of distant villi. The T cells tended to be clustered in all layers of the FAE and were significantly more numerous adjacent to interruptions of the brush border (revealed by lack of staining for alkaline phosphatase). Such interruptions were taken to indicate 'membrane' (M) cells. Our findings therefore suggested a spatial relationship between M cells and the aggregation of T cells. The ratio of CD4+ to CD8+ T cells (approximately 4:10) was significantly higher in the FAE than in the villous epithelium (approximately 0.6:10). This suggested that the FAE may be involved to a greater extent in induction of 'helper' T cell functions, perhaps depending on luminal antigens transported by M cells, whereas the villous epithelium may be more involved in stimulation of 'suppressor' T cell functions as indicated by recent studies in vitro.

Cimetidine treatment of recurrent ulcer after proximal gastric vagotomy.

Twenty-one patients with endoscopically confirmed recurrent ulceration after proximal gastric vagotomy entered an open trial of treatment with cimetidine, 1.0 g/day. Twenty patients completed 6 weeks' treatment, and repeat endoscopy showed ulcer healing in 18 of 20 patients. One patient's ulcer was found to be healed on X-ray examination, and the other patient had a healed ulcer after an additional 2 weeks' treatment. Eighteen of the patients with healed ulcer entered a maintenance trial of cimetidine, 400 mg at night. During the 1-year follow-up period the occurrence of symptoms led to re-endoscopy in 11 patients. Re-ulceration was confirmed in six patients (33%), and mean time to ulcer recurrence was 18.2 weeks. Ulcer recurrence was treated with an increased dose of cimetidine and antacids. All the ulcers then healed again, and the patients remained well on a maintenance dose of cimetidine, 800 mg/day. None of the patients had to be operated on during the trial. Two patients developed gynaecomastia during maintenance treatment with 400 mg cimetidine a day. No serious untoward signs or symptoms occurred that necessitated withdrawal from the trial. It seems as if recurrent ulcers after proximal gastric vagotomy respond to cimetidine treatment approximately as do peptic ulcers in unoperated patients.

Distribution of immunoglobulin producing cells is different in normal human appendix and colon mucosa.

The densities of IgG-, IgA-, IgM- and IgD-producing immunocytes were determined by paired immunofluorescence staining and morphometric analysis in the lamina propria of normal appendix specimens. Normal colon specimens were used as reference material, mostly paired from individual subjects. The density (median of cells/mm2 lamina propria area) of IgA immunocytes tended to be slightly higher in the appendix than in the colon (1259 vs 962) and the same held true for IgM cells (71 vs 55). Conversely, the overall density of IgG immunocytes was much higher in the appendix than in the colon (95 vs 38). A striking feature was the fact that almost 50% of all immunocytes were of the IgG isotype adjacent to lymphoid follicles. It seemed justified to conclude, therefore, that the abundance of such follicles explains the overall enrichment of IgG-producing cells in normal appendix mucosa. These immunocytes most likely represent follicle derived B cells that have reached terminal maturation locally, whereas precursors generated from less mature memory clones probably emigrate and home ubiquitously to distant sites of the gut lamina propria where they develop into IgA-producing immunocytes.

Effect of cimetidine on pentagastrin-stimulated gastric secretion before and after proximal gastric vagotomy for duodenal ulcer.

Nine patients with duodenal ulcer were studied before and 2--3 months after proximal gastric vagotomy (PGV). Infustion or cimetidine, 1.2 mg.kg-1h-1, reduced mean gastric acid output, in response to infusion of 1.5 microgram.kg-1h-1 of pentagastrin, by, on an average, 79.4% before and 79.1% after vagotomy. The corresponding values for pepsin output were 66.5% before and 77.0% after the operation. The values were not statistically different. Thus, in terms of per cent inhibition, cimetidine was similarly effective before and after PGV. No correlation was found between per cent reduction of acic output by vagotomy and by cimetidine. The effect of the drug was added to that of the vagotomy. Patients with relapse ulcer after vagotomy are therefore interesting candidates for cimetidine treatment.

Relapse of duodenal ulcer after treatment with trimipramine/antacids or cimetidine/antacids.

Ninety-three patients with duodenal ulcer were treated with trimipramine, 25 mg at night; cimetidine, 400 mg at night; or cimetidine, 1000 mg/day. In addition, all patients were given 20 ml antacids 1 and 3 h after meals. The healing rates after 6 weeks' treatment were 86%, 85%, and 100% in the three groups, respectively (differences not significant). After healing of the ulcers 88 of the patients were checked monthly during 1 year to study relapse rate. Endoscopy was performed on recurrence of symptoms suggesting relapse and after 6 and 12 months in all asymptomatic patients. Relapse was diagnosed in 64% of the patients in the trimipramine-treated group, in 48% in the 400 mg cimetidine group, and in 54% in th 1000 mg cimetidine group (differences not significant). The results suggest that ulcer recurrence is not commoner after treatment with standard dose of cimetidine than after a low dosage of cimetidine or trimipramine.

Distribution of macrophages and granulocytes expressing L1 protein (calprotectin) in human Peyer's patches compared with normal ileal lamina propria and mesenteric lymph nodes.

Antibodies to the cytosolic leucocyte L1 protein (or calprotectin) were examined for reactivity with macrophages, neutrophils, and eosinophils identified by paired immunofluorescence staining in sections of normal human ileal mucosa, including Peyer's patches. Macrophages were recognised by expression of the myelomonocytic antigen CD68 (monoclonal antibody KP1). Neutrophilic granulocytes were identified by their content of neutrophil elastase, and eosinophilic granulocytes by monoclonal antibody EG2. Virtually all CD68+ macrophages in normal lamina propria and Peyer's patches were L1- and the same was true for most extravasated macrophages in normal peripheral lymph nodes. Some mesenteric lymph nodes, however, and all peripheral lymph nodes with overt pathological processes (malignant lymphoma) contained many CD68+L1+ macrophages. Numerous L1+ cells were also localised to the crypt region and to some extent beneath the villous epithelium in normal lamina propria, but they were mainly identified as EG2+ eosinophils. Such cells were remarkably scarce or absent beneath the follicle associated epithelium in the dome region of Peyer's patches, where CD68+L1- macrophages were abundant. Also subepithelial and interfollicular CD68- interdigitating dendritic cells in Peyer's patches (recognised by antibody to S-100 protein) were usually unreactive with L1 antibody. The L1 protein shows a broad spectrum of antimicrobial activities in vitro, and its putative antiproliferative properties are interesting in relation to the immunosuppression postulated to take place in lamina propria. The virtual absence of L1 producing cells beneath the follicle associated epithelium in Peyer's patches may support the immunostimulatory function of these macrophage rich structures, which are held to be crucial for induction of specific mucosal immunity.

Production and secretion of immunoglobulins in the gastrointestinal tract.

Two decades ago it was shown that the major immunoglobulin (Ig) present in human secretions is a dimeric IgA covalently bound to an epithelial glycoprotein of about 80 kD, now called the secretory component (SC). Pentameric IgM is likewise actively enriched in most exocrine fluids and is associated with SC, although not in a covalently stabilized complex. Three findings explain the selective translocation of polymeric Ig (pIg) into exocrine fluids: (1) preferential local production; (2) J-chain-expressing capacity of pIg-producing immunocytes; and (3) SC-mediated epithelial transport. Human hepatocytes lack SC and the human liver, therefore, cannot act as an efficient "IgA pump". This is in contrast to the rat liver which shows a remarkable capacity for transport of dimeric IgA from blood into the bile. The J chain of pIg and the epithelial SC represent the "lock and key" in the glandular transport of secretory IgA (SIgA) and SIgM. It has recently been shown that SC is synthesized as a transmembrane protein of about 95 kD and constitutes the actual pIg surface receptor. Complexing between ligand and receptor in the plasma membrane is followed by endocytosis. The completed SIgA and SIgM molecules are then translocated in cytoplasmic vesicles through the epithelial cell to the gland lumen along with an excess of free SC. The main function of SIgA is to exert immune exclusion; that is, by intimate cooperation with innate nonspecific defense factors it decreases penetration of soluble antigens and inhibits epithelial colonization of bacteria and viruses. Especially in selective IgA deficiency, SIgM may exert a similar protective function since its synthesis is markedly increased in the intestinal mucosa. Leakage of IgG into exocrine fluids is enhanced by mucosal irritation. Although IgG should not be considered as a SIg, it may contribute to immune exclusion. This is seen especially in the respiratory tract where IgG is less easily subjected to proteolytic degradation than in the intestinal juice. In contrast, by activating complement, IgG antibodies may at the same time be phlogistic and accelerate mucosal penetration of antigens. IgG may thus contribute to persistent immunopathology in mucosal disease. The same is true for IgE antibodies which may be carried into mucous membranes and secretions by mast cells and cause their degranulation with local histamine release. Traces of IgD may likewise be found in the secretions but without obvious biologic significance. Regulation of secretory immunity takes place both in organized lymphoepithelial structures, such as the Peyer's patches, and adjacent to the glands in the lamina propria.(ABSTRACT TRUNCATED AT 400 WORDS)