Association Between Helicobacter and Gastric Ulcer Disease of the Pars Esophagea in Swine

Transcription

1 GASTROENTEROLOGY 1996;111:19 27 Association Between Helicobacter and Gastric Ulcer Disease of the Pars Esophagea in Swine DULCIENE MARIA DE MAGALHÃES QUEIROZ,* GIFONE AGUIAR ROCHA,* EDILBERTO NOGUEIRA MENDES,* SÍLVIA BELEZA DE MOURA, ANDREIA MARIA ROCHA DE OLIVEIRA,* and DAIRTON MIRANDA *Laboratory of Research in Bacteriology, Department of Pathology, Faculdade de Medicina, and Department of Microbiology, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil See editorial on page 244. pathogenesis of the lesions, including infection, intoxication, nutrition, stress, gastric acidity, and hormonal and seasonal changes, their etiology remains un- Background & Aims: Swine present spontaneously pep- known. 2,7 Similarly, the answer to the peptic ulcer questic ulcer in the gastric pars esophagea and are fre- tion in humans was largely unknown until the spiral quently colonized by a spiral bacterium that is of the bacterium Helicobacter pylori was accepted as being the same species as Helicobacter heilmannii type 1. This major predisposing factor in the pathogenesis of the organism is also observed in the gastric mucosa of patients with gastric symptoms and who present with disease. Inasmuch as we have shown previously that gastritis at histology. The aim of this study was to swine can harbor in their stomachs the spiral bacterium investigate the association between the presence of Gastrospirillum suis, 8,9 it seems justified to assume that, H. heilmannii type 1 and lesions of the pars esophagea. in analogy with human ulcer disease, this bacterium Methods: H. heilmannii type 1 infection was investigated could be involved in the pathogenesis of the gastric by mouse inoculation, urease test, and carbol- peptic ulcer of swine. Furthermore, it was shown re- fuchsin stain in 20 stomachs with ulcer, 30 stomachs cently that a tightly spiral bacterium, morphologically with preulcer lesions, and 20 stomachs with a macro- similar to G. suis, was more frequently observed in hisscopically normal pars esophagea. Results: The micro- tological sections of gastric mucosa from swine with organism was more frequently found in the stomachs ulcer than in those sections from animals without any with ulcer (100%) and in those with preulcer lesions lesions of the pars esophagea. 10 The relevance of this (90%) than in stomachs with macroscopically normal pars esophagea (35%). Histological alterations were subject is increased by the fact that this organism is observed in the pars esophagea and in the glandular also a human pathogen candidate because it was shown regions of bacterium-positive stomachs. Conclusions: recently by 16S ribosomal RNA sequencing that it be- It has been shown that the microorganism is strongly longs to the Helicobacter genus, being one of the two associated with naturally occurring ulcer and preulcer species described of Helicobacter heilmannii, 11 spiral bac- lesions of the pars esophagea of swine, which raises teria that have also been observed in the gastric mucosa the possibility that the bacterium is an important factor of patients with gastric symptoms and presenting with in the pathogenesis of these lesions. gastritis at histology G. suis is 99.5% similar to H. heilmannii type 1 11 ; such a similarity allows us to wine frequently develop ulcer spontaneously in the conclude that H. heilmannii type 1 and G. suis are the Spars esophagea of the stomach under laboratory conas same microorganism. Therefore, we will refer to G. suis ditions and routine animal husbandry practices with a H. heilmannii type 1 throughout this report. prevalence rate reaching 96%. 1,2 Ulcer has been diag- The aim of the present study was to investigate the nosed with increased frequency since the late 1950s be- presence of H. heilmannii type 1 in the stomach of swine cause of intensification of pig management, with the presenting all stages of evolution of ulcer disease of the animals being maintained in confinement under crowded pars esophagea at the macroscopic and microscopic conditions. 3,4 The disease presents stages of evolution level, using a highly sensitive method that allows an that vary from mild epithelial changes to deep and broad accurate diagnosis of the presence of the bacterium. ulcer extending to the entire glandless region of the stomach. 5, by the American Gastroenterological Association Although several factors have been mentioned in the /96/$3.00

2 20 DE MAGALHÃES QUEIROZ ET AL. GASTROENTEROLOGY Vol. 111, No. 1 Materials and Methods Animals The stomachs of 70 well-nourished swine about 6 months old from an industrial abattoir were studied. After the animals were killed and the carcasses eviscerated, each stomach was separated from the rest of the alimentary tract and transported immediately to the laboratory in individual bags. The stomachs were cut open along the greater curvature from the pylorus to the diverticulum, and the contents were discarded. The stomachs were then washed gently in tap water, taking care to remove food particles only. The mucosal surfaces were examined and assessed for the presence of macroscopic lesions. We selected 20 stomachs in which the pars esophagea was macroscopically normal, presenting a white, smooth, and glistening surface (Figure 1A). The second group was formed by 30 stomachs in which the surface of the pars esophagea was roughened and irregular, varying in color from white to yellow or yellow-green and sometimes showing erosions (Figure 1B). Finally, we also studied 20 stomachs showing lesions that were round, oval, or linear in shape and at least 2 cm in size, devoid of epithelium and presenting the base with granular appearance and with a blood or brown-tinged exudate, that were considered to be ulcers (Figure 1C). Microbiology Two fragments were collected from the antral, cardiac, and body regions for microbiological evaluation. One sample was placed in a tube containing Christensen s urea agar and was left to stand at 37 C for up to 24 hours. The second sample was smeared on a glass slide, stained with 40% carbolfuchsin, and examined by light microscopy under an oil immersion lens. 8,20 Scrapings of gastric mucus were obtained from the antral, cardiac, and body regions of the stomach of each pig, homogenized in three parts of 0.85% saline in a vortex blender, and then used for mouse inoculation. 21 Two mice free of gastric bacteria were anesthetized with ether and inoculated via a stomach tube with 0.4 ml of mucus homogenate from each pig. Fifteen days after inoculation, the mice were killed by spinal dislocation; the stomachs were excised and opened along the greater curvature; and fragments obtained from the antral and body gastric regions were smeared onto a glass slide, stained with 40% carbolfuchsin, and examined under an oil immersion lens for the presence of tightly spiraled bacteria. Swine were considered to be H. heilmannii type 1 positive by mouse inoculation when the bacterium was observed in the antrum and/or corpus of at least one mouse. Histology of the Pars Esophagea Fragments of cm of the pars esophagea were taken from the lesions when present and from the left lateral side of the pars esophagea near its junction with the glandular epithelium of the cardia, 2 cm above the antrum and pars esophagea junction of all the stomachs, for histology. Fragments were fixed in Bouin s fluid for 24 hours, dehydrated in Figure 1. Gross view of the pars esophagea of pig stomachs showing (A) white, smooth, and glistening surface (normal); (B) roughened, irregular, and corrugated surface; and (C) ulcers and erosions.

3 July 1996 HELICOBACTER AND GASTRIC ULCER IN SWINE 21 Proportions were compared by using two-tailed x 2 analysis or Fisher s Exact Test. Intensity of the gastritis was compared by the Wilcoxon s signed rank test for paired samples and Mann Whitney test for unpaired data. Differences were considered significant when P values were õ0.05. an alcohol-xylene series, oriented on edge, and embedded in paraffin wax. Then, 4-mm-thick sections were obtained from each block and stained with H&E for histology. All slides were examined blindly by an observer who was unaware of the macroscopic alteration or the microbiological results. The mi- croscopic findings, including increased epithelial thickness, parakeratosis, keratinization, and presence of balloon cells in the epithelium and increased papillary depth, basal zone thickness, presence of infiltrate of mononuclear and polymorphonuclear cells, and neovascularization in the lamina propria, were investigated. Histology of the Glandular Region Fragments from the glandular gastric mucosa of 10 bacterium-positive animals with ulcer, 10 bacterium-positive swine with preulcer lesions, and 7 bacterium-positive and 10 bacterium-negative swine with macroscopically normal pars esophagea were also obtained for histopathologic evaluation. Eight fragments of 0.5 cm cm from the lesser curvature of the antrum at 4 cm from the torus piloricus and eight fragments from the greater curvature of the gastric body were fixed in Bouin s fluid and processed in the same manner as described above for the pars esophagea. Histology of the antral and oxyntic mucosa was classified as normal when just very few scattered mononuclear cells were observed in the lamina propria and when no change was present in the superficial and glandular epithelium (score of 0). The presence of mononuclear cell inflammatory infiltrate in the lamina propria diffusely distributed or present in focal areas was considered gastritis that was classified as very mild, mild, moderate, and severe (scores of 1, 2, 3, and 4, respectively), based on the degree of inflammatory infiltrate. The presence of lymphoid aggregates and lymphoid follicles was also evaluated. Statistical Analysis Results Microbiology Among the methods used for H. heilmannii type 1 diagnosis, the inoculation of gastric mucus in mice Figure 2. Smear of the antral mucosa of a mouse dosed with gastric mucus from a pig showing tightly spiraled bacteria (carbolfuchsin staining; original magnification 13751). Figure 3. Ulcer in the pars esophagea of a pig stomach. (A) Notice an overhanging margin (left) and a broad necrotic zone in the base of the ulcer (stained with H&E). (B) High-power magnification of A showing an exudate composed of polymorphonuclear and mononuclear cells and neovascularization (stained with H&E) (original magnification: A, 631; B, 3201).

4 22 DE MAGALHÃES QUEIROZ ET AL. GASTROENTEROLOGY Vol. 111, No. 1 Figure 4. (A) Photomicrograph of the pars esophagea of a pig stomach showing irregular surface with desquamation, parakeratosis, keratosis, balloon cells, and elongation of the rete pegs (stained with H&E). (B) Higher magnification of A showing parakeratosis and balloon cells (stained with H&E) (original magnification: A, 631; B, 3201). was the most sensitive and showed bacterial positivity and granulation tissue consisting of proliferating fibroblasts in 54 swine (Figure 2). The urease test gave a positive and capillaries (neovascularization). result in the gastric fragments of 14 pigs, and tightly Histological changes of different degrees were observed spiraled bacteria were observed in carbolfuchsin-stained in all pars esophagea with macroscopic epithelial smears of the antral, oxyntic, and/or cardiac mucosa of alterations, including increased epithelial thickness of the 17 swine but not in the smears of any pars esophagea. epithelium in 93.5%, elongation of papillae in 90.0%, H. heilmannii type 1 was detected in all stomachs with presence of parakeratosis in 90.0%, keratinization in ulcer of the pars esophagea (20 of 20), in 90% of the 86.6%, and balloon cells in 90.0% (Figure 4A and B). stomachs with macroscopic epithelial changes (27 of 30), In two stomachs in which small erosions were observed, and in 35% of the macroscopically normal stomachs (7 neovascularization and an intense mononuclear, neutro- of 20). The presence of macroscopic alterations of the phil, and eosinophil infiltrate were also present (Figure pars esophagea, both ulcer and macroscopic epithelial 5A and B). changes, was significantly more common in bacterium- The stomachs with a macroscopically normal pars esophagea positive pigs than in bacterium-negative pigs (P Å ), were divided into two groups, i.e., seven H. heilpositive but no difference was observed between stomachs with mannii type 1 positive and 13 H. heilmannii type 1 ulcer and with macroscopic epithelial changes in regard negative stomachs (Figure 6). Significant differences were to the presence of H. heilmannii type 1 (P Å 0.3). observed between the two groups in regards to epithelial Histology of the Pars Esophagea changes of the pars esophagea. Increased epithelial thickness (P Å 0.01), parakeratosis (P Å 0.02), and balloon The presence of peptic ulcer was confirmed histo- cells (P Å 0.05) were observed in the pars esophagea of logically in all stomachs with macroscopic ulcer (Figure H. heilmannii type 1 positive swine. On the other hand, 3A and B). Histologically, the ulcer extended through no difference was observed in these groups with macrothe epithelium and lamina propria into the muscularis scopically normal pars esophagea concerning any micromucosae. The lumen of the ulcer contained an exudate scopic findings in the lamina propria of the pars esophaconsisting of neutrophils, eosinophils, and mononuclear gea. In both groups, the presence of a few mononuclear cells. In the lamina propria, there were areas of necrosis cells scarcely distributed was the only finding observed.

5 July 1996 HELICOBACTER AND GASTRIC ULCER IN SWINE 23 Differences were also observed between bacterium- all bacterium-positive swine presented variable degrees positive and -negative groups when pars esophagea with of antral gastritis, ranging from very mild to intense macroscopic epithelial changes without ulcers and mac- mononuclear cell infiltration, irrespective of the concurroscopically normal were analyzed together in regard to rent presence of macroscopic alterations of the pars esophthe presence of parakeratosis (P õ ), keratiniza- agea (P õ ) (Table 1 and Figure 8A). tion (P õ 0.001), balloon cells (P õ ), increase Regarding the oxyntic mucosa, 14 of 27 bacteriumin the epithelial thickness (P õ ), and elonga- positive swine (51.9%) presented with a low degree of tion of papillae (P õ ). gastritis (12 with a score of 1 and 2 with a score of 2) Histology of the Glandular Region (P Å 0.006) (Figure 8B). Gastritis was significantly more intense in the antral None of the fragments of the antral and oxyntic than in the oxyntic mucosa of swine with ulcer (P Å mucosa of all 10 bacterium-negative animals presented ) and with preulcer lesions (P Å 0.002) and in signs of gastritis (Figure 7A and B). On the other hand, those bacterium-positive swine with macroscopically normal pars esophagea (P Å 0.01). However, the intensity of gastritis in each region was similar among the groups (P ú 0.05). Neither the bacterium-positive nor -negative swine presented atrophy or intestinal metaplasia. Lymphoid aggregates or lymphoid follicles were observed more frequently in the gastric mucosa of bacterium-positive pigs (77.8%) than in bacterium-negative pigs (30%) (P Å 0.016). The majority of the bacteriumpositive swine presented multiple lymphoid aggregates and large lymphoid follicles with germinal centers, sometimes extending through the full thickness of the antral mucosa. Two bacterium-negative animals presented with a total of one or two lymphoid follicles in the 16 fragments studied, and one presented with a small aggregate of lymphocytes. Figure 5. (A) Low-power view showing an erosion in the pars esophagea of the stomach of a pig. Notice an overhanging margin (left) (stained with H&E). (B) Higher-power magnification of an erosion with Figure 6. Normal microscopic appearance of the pars esophagea of infiltration of neutrophils, eosinophils, and remnants of epithelium the stomach of a bacterium-negative pig (stained with H&E; original (stained with H&E) (original magnification: A, 631; B, 3201). magnification 631).

6 24 DE MAGALHÃES QUEIROZ ET AL. GASTROENTEROLOGY Vol. 111, No. 1 Table 1. Histological Findings of the Antral Mucosa of H. 1995). These results confirmed for the first time the heilmannii Type 1 Positive and Negative Swine hypothesis, previously suggested by several au- Stomachs thors, 12,23,24 that H. heilmannii infection in humans is a Microscopic alterations zoonosis and indicated that pigs probably are a reservoir Present (scores) of H. heilmannii type 1 for human infection. Although H. heilmannii has been observed in the gas- Macroscopic alterations Absent tric mucosa of patients with acute 18,19 and chronic gastritis, Absent as well as with peptic ulcer 16,25 and esophagi- Bacterium negative (n Å 10) tis, 12,15 its pathogenic role is presently unclear in part Bacterium positive (n Å 7) because of the fact that the infection is rare among hu- Present (all bacterium positive) Preulcer (n Å 10) mans (õ1%) and the bacterium has not yet been culti- Ulcer (n Å 10) vated in vitro, a fact that precludes systematic studies. On the other hand, the bacterium is frequently observed NOTE. Scores of 1, 2, 3, and 4 mean very mild, mild, moderate, and severe gastritis, respectively. in the gastric mucosa of swine, with positivity rates of about 60% when mouse inoculation is used for diagnosis. Discussion 21 However, similar to the infection of human beings, there are few data concerning the pathogenicity of the G. suis is an uncultured tightly spiraled bacterium bacterium in swine. that colonizes the gastric mucosa of swine. Recently, by The present results show a strong association between analysis of 16S ribosomal RNA, it was shown that the the presence of H. heilmannii type 1 and the gastric ulcer bacterium belongs to the Helicobacter genus and is of the disease of swine, with the bacterium being observed in same species as H. heilmannii type 1. 11,22 Preliminary 47 of 50 stomachs with ulcer and preulcer lesions. A results of another study in progress in our laboratory similar H. pylori detection rate has been verified in pa- showed that the strains of G. suis of the present study, tients with duodenal ulcer when sensitive methods are analyzed by polymerase chain reaction using specific used for diagnosis. In this study, we also used a highly primers to detect H. heilmannii types 1 and 2, were found sensitive diagnostic method to detect H. heilmannii type to be H. heilmannii type 1 (unpublished data, December 1 in swine stomach. 21 In some stomachs, even when the Figure 7. Normal gastric mucosa of a bacterium-negative pig. (A) Antrum and (B) corpus (stained with H&E; original magnification 3201).

7 July 1996 HELICOBACTER AND GASTRIC ULCER IN SWINE 25 Figure 8. Gastric mucosa of a bacterium-positive pig showing an inflammatory reaction characterized by infiltration of mononuclear cells. (A) Antrum and (B) corpus (stained with H&E; original magnification 3201). possible to establish that swine free from the bacterium did not show any inflammatory reaction in their antral and oxyntic mucosa in contrast to infected animals that present different degrees of inflammatory reaction in these regions. Although we have observed that the presence of the microorganism is strongly associated with gastritis in swine, the bacterium does not seem to elicit active gastri- tis in this animal, but it does produce activity in humans. 12,14 Otherwise, H. pylori induces active gastritis in humans but only lymphocytic gastritis in gnotobiotic swine. 26 Thus, it seems that different hosts present differ- ent histopathologic responses to the same Helicobacter, and the identification of the factors involved in these different responses may provide insights into the mechanisms involved in the pathogenesis of the diseases in- duced by Helicobacters. bacterium was not detected by the urease test or on smears stained with carbolfuchsin from 15 or more fragments of different regions of the gastric mucosa, its presence was observed in mice inoculated with mucus obtained from the same stomachs. Inasmuch as the sensitivity of the method we used allowed an accurate diagnosis of H. heilmannii type 1 infection, it made it possible to study precisely the association between the presence of the microorganism and lesions of the pars esophagea and histopathologic alterations of the glandular region. In a previous study, because we did not have a good parameter for separating bacterium-negative swine from bacterium-positive swine, we observed gastritis in several animals that we considered bacterium negative, and we did not consider the presence of a discrete inflammatory infiltrate as significant in some bacteriumpositive animals. 9 In the present study, however, it was

8 26 DE MAGALHÃES QUEIROZ ET AL. GASTROENTEROLOGY Vol. 111, No. 1 It has been suggested that the ulcer disease of the pars significantly greater acid output than control animals. 27 esophagea of swine presents several stages of evolution Furthermore, it has been shown recently that gastrin- with the erosions probably originating from the desqua- mediated acid secretion is increased threefold in H. pylori positive mation of the epithelium as a result of the parakeratotic patients with no ulcer and increased sixfold alterations and with the ulcers representing advanced in H. pylori positive patients with duodenal ulcer. 28 Another erosive lesions. 5,6 If a bacterium from the Helicobacter hypothesis to explain the role of H. heilmannii type genus, similar to what occurs in human ulcer disease, is 1 in the pathogenesis of the ulcer disease of swine may an important etiopathogenic factor in swine ulcer, it be the production of noxious substances by the bacte- would be necessary to show the presence of the bacterium rium, such as cytotoxins, that may act directly or indirectly during all stages of the disease. In fact, the data of this on the pars esophagea. In regard to the cytotoxin, study further substantiate the association between the a close association between human ulcer disease and toxigenic bacterium and all stages of evolution of the ulcer from strains of H. pylori has been shown. 29 Furthermore, mild mucosa alterations to erosions and ulcers. Even Ghiara et al. 30 have shown recently, using a mouse model when macroscopic alterations were not observed but the of gastric disease, that H. pylori cytotoxin has an im- bacterium was present in the gastric mucosa, superficial portant role in the induction of gastric epithelial cell changes restricted to the epithelium of the pars esophagea lesions but not in eliciting inflammation. These are histo- were observed at histology, such as some degree of para- pathologic alterations that we have also observed in the keratosis, increased epithelial thickness, and presence of pars esophagea in the initial stages of swine ulcer disease. balloon cells. Thus, in the initial stage of the disease, In conclusion, our results indicate that H. heilmannii epithelial injury more than inflammatory cell infiltration type 1 is likely to play a prominent role in the pathogenewas observed in the pars esophagea. sis of ulcer disease of the pars esophagea of swine. Further There are some differences between the swine and hution evidence of the importance of H. heilmannii type 1 infecman peptic ulcer. In humans, ulcers occur in only a small in the development of swine ulcer would be provided fraction of H. pylori infected individuals and are more if animals experimentally infected with the bacterium frequently localized in the duodenum and in the gastric showed ulcers similar to those that occur spontaneously. antrum. On the other hand, in swine, the prevalence of Experiments on this subject are in progress in our laborathe disease is greater, and it localizes mainly in the pars tory. A model of infection using swine naturally or exper- esophagea of the stomach. However, the two diseases imentally infected with Helicobacter will be potentially present some similarities. The histopathologic findings useful to investigate the pathogenesis and treatment of of the swine peptic ulcer closely resemble those described ulcer disease of both human beings and swine, because in human peptic ulcer (layers of necrosis, numerous inflammatory no other comparable model of ulcerogenesis is available. cells, and granulation tissue and fibrosis), and our results have also shown that antral gastritis associated References with a Helicobacter infection is always present in swine 1. Sabec K, Schroder J. Feed quality and the incidence of gastric ulcer disease as reported for human peptic ulcer disease. ulcer in pigs. Dtsch Tierarztl Wochenschr 1970;77: O Brien JJ. Gastric ulceration (of the pars oesophagea) in the Based on the results of our study, the following quespig a review. Vet Bull 1969;39: tion arises: how could a bacterium that colonizes the 3. Kowalczyk T. Etiologic factors of gastric ulcers in swine. Am J glandular mucosa of the stomach promote ulceration of Vet Res 1969;30: the glandless region of the pig s stomach? Basically, the 4. Muggenburg BA, Reese N, Kowalczyk T, Grummer RH, Hoekstra WG. Survey of the prevalence of gastric ulcers in swine. Am J occurrence of lesions in the pars esophagea can be caused Vet Res 1964;25: by the fact that the epithelial membrane in this portion 5. Muggenburg BA, McNutt SH, Kowalczyk T. Pathology of gastric of the stomach is glandless and has no mucous barrier ulcers in swine. Am J Vet Res 1964;25: Kowalczyk T, Hoekstra WG, Puestow KL, Smith ID, Grummer RH. to protect it from the deleterious action of the gastric Stomach ulcers in swine. J Am Vet Med Assoc 1960;137:339 juice. Thus, hyperchlorhydria may be an important factor 344. in the etiopathogenesis of the ulcer disease of swine, as 7. Huber WG, Wallin RF. Pathogenesis of porcine gastric ulcers. Am suggested by some investigators. 2,7 Taking these facts J Vet Res 1967;28: Queiroz DMM, Rocha GA, Mendes EN, Lage AP, Carvalho ACT, and our results together, we may hypothesize that the Barbosa AJA. Spiral micro-organism in the stomach of pigs. Vet bacterium acts by altering the gastric secretion of infected Microbiol 1990;24: animals. This hypothesis is strengthened by the fact that 9. Mendes EN, Queiroz DMM, Rocha GA, Nogueira AMMF, Carvalho AST, Lage AP, Barbosa AJA. Histopathological study of porcine increased acid secretion has been shown in rhesus mongastric mucosa with and without a spiral bacterium ( Gastrospikeys, naturally infected with tightly spiraled bacteria rillum suis ). J Med Microbiol 1991;35: morphologically similar to Gastrospirillum, to present a 10. Barbosa AJA, Silva JCP, Nogueira AMMF, Paulino E Jr, Miranda

9 July 1996 HELICOBACTER AND GASTRIC ULCER IN SWINE 27 CR. Higher incidence of Gastrospirillum sp. in swine with gastric LS. An uncultured gastric spiral organism is a newly identified ulcer of the pars oesophagea. Vet Pathol 1995;32: Helicobacter in humans. J Infect Dis 1993;168: Mendes EN, Queiroz DMM, Dewhirst RE, Paster BJ, Rocha GA, 23. Lee A, Dent J, Hazell S, McNulty C. Origin of spiral organisms in Fox JG. Are pigs a reservoir host for human Helicobacter infection human gastric antrum. Lancet 1988;1: (abstr)? Am J Gastroenterol 1994;89: Stolte M, Wellens E, Bethke B, Ritter M, Eidt H. Helicobacter 12. McNulty CAM, Dent JC, Curry A, Uff JS, Ford GA, Gear MWL, heilmannii (formerly Gastrospirillum hominis) gastritis: an infec- Wilkinson SP. New spiral bacterium in gastric mucosa. J Clin tion transmitted by animals? Scand J Gastroenterol 1994;29: Pathol 1989;42: Lee A, Dick E, Eckstein RP, Kellow JE, Fevre DI. Non Campylo- 25. Borody TJ, George LL, Brandl S, Andrews P, Ostapowicz N, Hyland L, Devine M. Helicobacter pylori negative duodenal ulcer. Am J bacter pylori spiral organisms in the gastric antrum. Aust N Z J Gastroenterol 1991;8: Med 1989;19: Krakowka S, Morgan DR, Kraft WG, Lunk RD. Establishment of 14. Waring PM, Shilkin KB. Corkscrew-like bacteria associated gastric Campylobacter pylor infection in the neonatal gnotobiotic with gastritis. Histopathology 1989;15: piglet. Infect Immun 1987;55: Dye KR, Marshall BJ, Frierson Jr HF, Guerrant RL, McCallum RW. 27. Dubois A, Tarnawski A, Newell DG, Fiala N, Dabros W, Stachura J, Ultrastructure of another spiral organism associated with human Krivan H, Heman-Ackah LM. Gastric injury and invasion of parietal gastritis. Dig Dis Sci 1989;34: cells by spiral bacteria in Rhesus monkeys. Gastroenterology 16. Fléjou JF, Diomandé I, Molas G, Golkfain D, Rotenberg A, Florent 1991;100: M, Potet F. Gastrite chronique associée chez l homme à la prés- 28. El-Omar E, Penman I, Dorrian CA, Ardil JES, McColl KEL. Eradicatence de germes spiralés non Helicobacter pylori (Gastrospiril- ing Helicobacter pylori infection lowers gastrin-mediated acid selum hominis). Gastroenterol Clin Biol 1990;14: cretion by two-thirds in duodenal ulcer patients. Gut 1993;34: 17. Heilmann KL, Borchard F. Gastritis due to spiral shaped bacteria other than Helicobacter pylori: clinical, histological, and ultramanno 29. Figura N, Guglielmetti P, Russolini A, Barberi A, Cusi G, Musstructural findings. Gut 1991;32: RA, Russi M, Quaranta S. Cytotoxin production by Campy- 18. Wegmann W, Aschwanden M, Schaub N, Aenishanslin W, Gyr lobacter pylori strains isolated from patients with peptic ulcer K. Gastrospirillum hominis-assoziierte gastritis eine zoonose? and from patients with chronic gastritis only. J Clin Microbiol Schweiz Med Wochenschr 1991;121: ;27: Ghiara P, Marchetti M, Blaser MJ, Tummuru MKR, Cover TL, 19. Al-Himyary AJS, Zabaneh FI, Zabaneh SS, Barnett S. Gastrospiril- Segal ED, Tompkins LS, Rappuoli R. Role of the Helicobacter lum hominis in acute gastric erosion. South Med J 1994;87: pylori virulence factors vacuolating cytotoxin, caga, and urease in a mouse model of disease. Infect Immun 1995;63: Rocha GA, Quieroz DMM, Mendes EN, Lage AP, Barbosa AJA Simple carbolfuchin staining for showing C. pylori and other spiral bacteria in the gastric mucosa. J Clin Pathol 1989;42:1004 Received April 24, Accepted February 6, Address requests for reprints to: Dulciene Maria de Magalhães 21. Mendes EN, Queiroz DMM, Rocha GA, Moura SB, Barbosa MT, Queiroz, Avenida Alfredo Balena 190/464, , Belo Hori- Carvalhaes SM, Freitas MLP, Reis ACS. In vivo culture improves zonte, Minas Gerais, Brazil. Fax: (55) the diagnosis of Gastrospirillum suis infection (abstr). Ir J Med Supported by grants from CNPq (Brazil). Sci 1992;161:78. The authors thank FrigorıBfico São João Ltda. for providing the pig 22. Solnick JV, O Rourke J, Lee A, Paster BJ, Dewhirst FE, Tompkins stomachs.