Pii: s0140673684918166

Saturday 16 June 1984 UNIDENTIFIED CURVED Patients and Methods All patients referred for gastroscopy on clinical grounds were BARRY J. MARSHALL eligible for the study which continued until there were 100participants who gave informed consent and in whom biopsy was Departments of Gastroenterology and Pathology, considered to be safe. The study was approved by our hospital's Royal Perth Hospital, Perth, Western Australia human rights committee.
Biopsy specimens were taken from intact areas of antral mucosa in 100 consecutive Where possible patients completed a clinical questionnaire consenting patients presenting for gastroscopy. Spiral or designed to detect a source of infection or show any relationship with "known" were demonstrated in specimens from 58 gastritis or Campylobacter infection, rather patients. Bacilli cultured from 11 of these biopsies give a detailed account of each patient's history. The emphasis negative, flagellate, and microaerophilic and appeared animal contact, travel, diet, dental hygiene, and drugs, rather than symptoms.
species related to the genus Campylobacter. The bacteria were present in almost all patients with active chronic gastritis, duodenal ulcer, or gastric ulcer and thus may be an The gastroscopies were done by colleagues at the Royal Perth important factor in the aetiology of these diseases.
Hospital. Participants fasted for at least 4 h before endoscopy. AnOlympus GIF-K fibreoptic gastroduodenoscope was used. Routine biopsies were done when indicated. For the study two extra GASTRIC spiral bacteria have been repeatedly observed, specimens were taken from an area of intact antral mucosa, at adistance from reported, and then forgotten for at least 45 years. 1-3 In 1940 any focal lesion such as an antral ulcer. When the mucosa appeared inflamed the specimens were taken from a red Freedburg and Barron stated that "spirochaetes" could be area, otherwise any part of the antrum was used. One biopsy was found in up to 37% of gastrectomy specimens,4- but immediately fixed in phosphate-buffered formalin for histological examination of gastric suction biopsy material failed to examination, the other was placed in chilled anaerobic transport confirm these findings.5 The advent of fibreoptic biopsy medium and taken to the microbiology laboratory within 1 h. In a techniques permitted biopsy of the antrum, and in 1975 Steer few cases an extra specimen was taken for ultrastructural and Colin-Jones observed gram-negative bacilli in 80% of patients with gastric ulcer. The curved bacilli they The gastroenterologist dictated his report soon after the illustrated were said to be Pseudomonas, possibly a endoscopy. We had not planned to analyse these reports so astandard contaminant, and the bacteria terminology was not used and no special attention was paid to minor endoscopic lesions. Findings of doubtful clinical repeated demonstration of these bacteria in inflamed gastric significance, such as mild endoscopic gastritis or duodenogastric antral mucosa7 prompted us to do a pilot study in twenty bile reflux, may thus have been under-reported. (Hereafter the term patients. Typical curved bacilli were present in over half the "gastritis" refers to a histological grade of chonic gastritis unless biopsy specimens and the number of bacteria was closely stated otherwise.) Before we analysed the data, the endoscopy related to the severity of the gastritis. The present study was reports were coded for the major diagnoses.
designed to confirm the association between antral gastritisand the bacteria, to discover associated gastrointestinal diseases, to culture and identify the bacteria, and to find Sections were stained with haematoxylin and eosin (H & E) and factors predisposing to infection.
graded for gastritis (by J. R. W.) as 0 (normal), inflammatory cellsrarely seen; 1 (normal), lymphoid cells present but within normal paper read at Second International Workshop on Campylobacter no other evidence of inflammation (see below); 2 Infections (Brussels, 1983).
(chronic), chronic gastritis; or 3 (active), active chronic gastritis.
Gradings were based solely on the type of inflammatory cells.
TABLE II-ASSOCIATION OF BACTERIA WITH ENDOSCOPIC DIAGNOSES Other types of mucosal change, such as gland atrophy or intestinalmetaplasia, were noted separately, but were not used as evidence ofinflammation. "Chronic gastritis" indicated inflammation with noincrease in polymorphonuclear leucocytes (PMNs). There wereeither increased numbers of lymphoid cells or normal cell numberswith other evidence of inflammation such as oedema, congestion, orcell damage. The term "active" was used to indicate an increase in PMNs.8 The gastritis was considered active if a few PMNsinfiltrated one gland neck or pit, if occasional PMNs were scatteredthroughout the superficial epithelium, or if there was an obviousincrease in PMNs in the lamina propria.
Later, sections stained with Warthin-Starry silver stain were examined for small curved bacilli on the surface epithelium.
*More than one description applies to several patients (eg, 4 patients had both Numbers of bacteria were graded as 0, no characteristic bacteria; 1, gastric and duodenal ulcers).
occasional spiral bacteria found after searching; 2, scattered bacteria tRefers to endoscopic appearance, not histological inflammation.
in most high-power fields or occasional groups of numerous TABLE III-HISTOLOGICAL GRADING OF GASTRITIS AND bacteria; or 3, numerous bacteria in most high-power fields.
Tissue smears were Gram stained and examined for curved bacilli resembling Campylobacter. The remaining tissue was minced,plated on non-selective blood and chocolate agar, and cultured at37°C under microaerophilic conditions as used for Campylobacterisolation.9 At first plates were discarded after 2 days but when the first positive plate was noted after it had been left in the incubator for 6 days during the Easter holiday, cultures were done for 4 days.
*Gastritis grades 0 and 1 normal.
t case showed bacteria on gram stained smear.
Analysis of Results TABLE IV-RELATION BETWEEN Questionnaires, gastroscopy reports, and histopathology and PATIENTS WITHOUT microbiology results were coded independently in separatedepartments. Complete results for individual patients were notknown until the statistician had received all the data. The findingswere tested for positive correlation with the presence of eitherbacteria or gastritis, by the chi-squared method. Fisher's exact testof significance was used for all the 2 x 2 tables in this paper.
In 12 weeks 184 patients were examined by the gastroenterology unit. Of the 84 patients excluded, 5 refused Gastritis could usually be graded with confidence at low consent, 4 had contraindications to biopsy, and 75 patients, magnification. There was some difficulty with about 25 cases mostly unbooked cases, could not be invited to participate.
where the changes were mild or the specimens were small, These patients closely matched the study group for age, sex, superficial, or distorted. To ensure that gradings were and incidence of peptic ulcers (table I).
reliable, single H & E sections from the last 40 cases wereexamined "blind" by another pathologist who agreed with the presence or absence of gastritis in 36 cases (90%), and gave 99 patients completed the questionnaires. The only an identical grading in 32.
symptom which correlated with gastritis or bacteria was Gradings for bacteria by silver staining were more "burping" which was more common in patients with bacteria straightforward. The bacteria stained well and were easily 03) or gastritis (p = 0 007). This association remained differentiated from contaminant bacteria or debris. Silver when patients with peptic ulcer were excluded. None of the staining was the most sensitive method of detecting the spiral other questionnaire responses showed any relationship to the bacteria. Silver stained sections and Gram stained smears presence of gastric bacteria or gastritis.
both done in 96 cases and spiral bacteria were seen in 56 them; 32 with both stains, 23 with silver alone, and 1 case with the Gram stain alone.
There was a very close correlation between both gastric The correlation between gastritis and bacteria, defined by ulcer and duodenal ulcer and the presence of the bacteria Gram and/or by silver staining, was remarkable (table III).
(table II). Most patients with peptic ulcer also had gastritis Gastritis was present in 55/57 biopsy specimens with bacteria (29/31; p=0-0002). (p=2X 10'). When the 31 patients with peptic ulcer were excluded, the correlation persisted, implying that the TABLE I-COMPARISON OF PARTICIPANTS WITH EXCLUDED PATIENTS presence ofbacteria was not secondary to an ulcer crater (table Specimens for culture were received from 96 patients and 11 were culture positive, all being seen with Gram and silver staining also. No spiral bacteria were grown from the first 34cases, probably because the cultures were discarded too soon.
Electron micrograph from a mucosal biopsy with active chronic gastritis.
Upper: many profiles of sectioned pylonc campylobacter are located on the lumlllal aspect of mucus- secreting epithelial cells; plasma membranes are intact, but indented and almost devoid ofmicrovilli Lower: at higher magnification groups of transversely and longitudinally cut sheathed flagella are visible (arrows; bar= 100 nm).
The bacteria were S-shaped or curved gram-negative rods, material was sent (to J. R. W.) with study biopsies, mainly 5 /Am, with up to 11/a wavelengths. In electron from cases of gastric ulcer. However, an independent blind micrographs they had smooth coats and there were usually gastritis in 40 cases matched the study results four sheathed flagella arising from one end of the cell. They grew best in a microaerophilic atmosphere at 37 °C, a campylobacter gas generating kit was sufficient (OxoidBR56). Moist chocolate or blood agar was the preferred spiral bacteria of the human gastric antrum have never was evident in 3 days as 1 mm before, and their association with active non-pigmented colonies. In artificial media the bacteria gastritis has not been described. They are a new usually larger and less curved than those seen on Gram stains species closely resembling campylobacters morphologically of fresh tissue. They formed coccoid bodies in old cultures.
and in respect of atmospheric requirements and DNA base The bacteria were oxidase +, catalase +, HzS +, indole composition, but their flagellar morphology is not that of the urease -, nitrate -, and did not ferment glucose. They were genus Campylobacter.9 Campylobacters have a single erythromycin, kanamycin, flagellum at one or both ends of the cell whereas gentamicin and penicillin, and resistant to nalidixic acid.
organism has four sheathed flagella at one end .7,10 If DNA base analysis gave a guanine+ cytosine content of 36 premature to talk of "Campylobacter pyloridis"ll perhaps a value in the range for campylobacters.
"pyloric campylobacter" will do to define the site where these organisms are commonly found and to indicate the similarity to known Campylobacter spp.
The patient sample was from a defined population with There was no well-defined clinical syndrome associated gastric symptoms expected to have some gastroenterological pyloric campylobacter. Only "burping" was abnormality. The biopsy tissue studied was from apparently significantly associated. Others have described this symptom intact mucosa-ie, not the sort of specimen a pathologist in patients with non-ulcer dyspepsia and PMN usually sees. We attempted to limit bias by making the study the antrum is also common in such patients. 12,13 We consecutive and blind, and were partly successful. The study abdominal pain to correlate with pyloric campylobacter or was not strictly consecutive since 84 patients had to be gastritis, but it did not. Perhaps, since most patients excluded. However, gastroscopy reports and laboratory undergoing gastroscopy have pain (75% in our study) the investigations were completed serially and usually question "Do you have abdominal pain-yes or no?" was too independently ("blind") except that clinically relevant Much of the questionnaire was designed to select likely patients. The failure of the H2 receptor antagonists to prevent sources or causes of pyloric campylobacter infection. For ulcer relapse is attributed to an underlying ulcer diathesis example, bacteria might have colonised patients who already which is unaffected by therapy. A bacterial aetiology, with had gastritis and were taking antacids, milk, or cimetidine, continuing gastritis, could be the explanation. The diathesis thus impairing their "gastric acid barrier" and predisposing may be a myth. Of ulcer-healing agents the only one them to infection.14 Animal contact and carious teeth were thought to improve relapse rates is tripotassium dicitrato- also considered as sources of infection. Campylobacters are bismuthate.29 This compound is bactericidal to pyloric commensals of domestic and farm animals (C coli, C jejunz), campylobacter and in patients treated with it the gastritis and they also inhabit the human mouth (C sputorum ss improved and the bacteria disappeared. 30 sputorum).15 We found no evidence that any of these factors The aetiology of peptic ulceration is unknown but until predisposed to the infection.
a bacterial cause has not really been considered. We The absence of a relation between "known causes" of found colonisation of the gastric antrum with pyloric gastritis and the presence of histological gastritis has been campylobacter in over half of a series of cases at routine noted by others. For example, analgesic abusers often have no endoscopy. The bacteria were present almost exclusively in gastritis, even when a gastric ulcer is present;16 alcohol patients with chronic antral gastritis and were also common consumption is not clearly related to gastritis;17 the quantity in those with peptic ulceration of the stomach or duodenum.
of bile in the stomach (duodenogastric reflux) is not obviously Although cause-and-effect cannot be proved in a study of this related to the state of gastric mucosa; 18 autoimmune disease is kind, we believe that pyloric campylobacter is aetiologically unlikely cause, since gastric autoantibodies are uncommon related to chronic antral gastritis and, probably, to peptic except in pernicious anaemia, where the main histological ulceration also.
changes are in the body of the stomach, not the antrum.19 thank Dr T. E. Waters, Dr C. R. Sanderson, and the gastroenterology Gastric ulcer seems an unlikely primary cause of antral unit staff for the biopsies, Miss Helen Royce and Dr D. I. Annear for the gastritis because the gastritis remains after successful microbiological studies, Mr Peter Rogers and Dr L. Sly for supplying the G C data, Dr J. A. Armstrong for the electron the ulcer with cimetidine microscopy, Dr R. Glancy for or carbenoxolone, and reviewing slides, Miss Joan Bot for the silver stains, Mrs Rose Rendell ofRaine gastritis is just as common in patients with duodenal ulcer as Medical Statistics Unit UWA, and Ms Maureen Humphries, secretary, and, with gastric ulcer.', 20-23 Thus, the aetiology of chronic for travel support, Fremantle Hospital.
gastritis remains uncertain.
Correspondence should be addressed to: B. M., Department of We have found a close association between pyloric Microbiology, Fremantle Hospital, PO Box 480, Fremantle 6160, Western campylobacter and antral gastritis. When PMN infiltrated the mucosa the bacteria were almost always present (38/40).
In the absence of inflammation they were rare (2/31),suggesting that they are not commensals. The bacteria were Doenges JL. Spirochaetes in gastric glands of macacus rhesus and humans without definite history of related disease. Proc Soc Exp Biol Med 1938; 38: 536-38.
not cultured unless the patient had histological evidence of 2. Ito S. Anatomic structure of the gastric mucosa. In: Heidel US, Cody CF, eds.
both gastritis and pyloric campylobacter. We know of no Handbook of physiology, section 6: Alimentary canal, vol II: secretion Washington,DC: American Physiological Society, 1967 705-41 state where, in the absence of complicating 3. Fung WP, Papadimitriou JM, Matz LR. Endoscopic, histological and ultrustructural factors such as ulceration (table IV), bacteria and PMNs are so correlations in chronic gastritis. Am J Gasiroenterol 1979, 71: 269-79.
intimately related without the bacteria being pathogenic.
4. Freedburg AS, Barron LE. The presence of spirochaetes in human gastric mucosa. Am J Dig Dis 1940, 7: 443-45.
How does pyloric campylobacter survive? The bacteria 5. Palmer ED Investigation of the gastric Spirochaetes of the human. Gastroenterology were usually in close contact with the mucosa, often in 1954; 27: 218-20.
6. Steer HW, Colin-Jones DG Mucosal changes in gastric ulceration and their response grooves between cells, within acinus-like infoldings of the to carbenoxolone sodium. Gut 1975; 16: 590-97.
epithelium or within the mucosal pits (figure). The surface 7. Warren JR, Marshall B. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet 1983; i: 1273-75 coating was superficial to the bacteria and any foreign 8. Whitehead R, Truelove SC, Gear MWL. histological diagnosis of chronic gastritis material or organisms from the oral flora were present above in fibreoptic gastroscope biopsy specimens. J Clin Pathol 1972; 25: 1-11 mucus, rarely mixed with it, and not beneath it: the Kaplan RL. Campylobacter. In: Lenette E, Balows A, Hausler WJ, Truant JP, eds.
Manual of clinical microbiology, 3rd ed. Washington, DC: American Society for appeared to form a stable layer over the spiral bacteria. The Microbiology, 1980: 235-41.
antrum secretes mainly mucus, and the deeper levels of the 10. Pead PJ. Electron microscopy of Campylobacter jejuni J Med Microbiol 1979; 12: surface mucus coating are slightly alkaline.24 Thus pyloric 11. Skirrow MB Taxonomy and biotyping. Morphological aspects. In: Pearson AD, campylobacter grows in a near-neutral environment, in close Skirrow MB, Rowe B, Davies JR, Jones DM, eds. Campylobacter II: Proceedings ofthe Second International contact with the Workshop on Campylobacter Infections. London. Public mucosa and protected from the bactericidal Health Laboratory Service, 1983: 36.
gastric juice. The absence of these bacteria from past reports 12. Crean GP, Card WI, Beattie AD, Holden RJ, James WB, Knill-Jones RP, Lucas RW, of gastric microbiology may be because only gastric juice was Spiegelhalter D. Ulcer-like dyspepsia. Scand J Gastroenterol 1982; 17 (suppl 79):9-15.
cultured.25,26 Even salmonellae cannot survive the low 13. Greenlaw R, Sheahan DG, Deluca V, Miller D, Myerson D, Myerson P intragastric pH for more than a few minutes.14 Where gastric Gastroduodenitis: a broader concept of peptic ulcer disease. Dig Dis Set 1980; 25:660-72.
biopsy material has been cultured,6,27,28 microaerophilic 14. Giannela RA, Broitman SA, Zamcheck N. Gastric acid barrier to ingested techniques were not used and pyloric campylobacter did not microorganisms in man: Studies in vivo and in vitro. Gut 1972; 13: 251-56.
15. Blaser MJ, Reller LB. Campylobacter enteritis. N Engl J Med 1981; 305: 1444-52.
16 MacDonald WC Correlation of mucosal histology and aspirin intake in chronic gastric Peptic ulcer was the only endoscopic finding associated- ulcer. Gastroenterology 1973; 65: 381-89.
with histological gastritis and pyloric campylobacter. This 17. Wolff G. Does alcohol cause chronic gastritis? Scand J Gastroenterol 1970; 5: 289-91 18. Goldner FH, Boyce HW. Relationship of bile in the stomach to gastritis. Gastrointest was surprising since the bacteria were not prominent on Endosc 1976; 22: 197-99.
gastric ulcer borders and in duodenal ulcer 19. Whitehead R. Mucosal biopsy of the gastrointestinal tract. In: Bennington JL, ed.
Ma)or problems in pathology: Vol III, 2nd ed. Philadelphia: WB Saunders, 1979: would be expected. Perhaps the mucus coating is deficient or unstable near ulcer borders, thus allowing damage to the 20. Gilmore HM, Forrest JAH, Pettes MR, Logan RFA, Heading RC Effect of short and long term cimetidine on histological duodenitis and gastritis. Gut 1978; 19: 981.
as well as the mucosa. Within a few millimetres of an 21. Mclntrye RLE, Piris J. Truelove SC. Effect of cimetidine on chronic gastritis in gastric ulcer, both pyloric campylobacter and gastritis were usually ulcer patients Aust NZ J Med 1982; 12: 106.
22. Schrager J, Spink R, Mitra S. The antrum in present. Other studies have shown continuing gastritis after patients with duodenal and gastric ulcers.
Gut 1967; 8: 497-508.
ulcer healing with cimetidine and we have observed thepersistence of pyloric campylobacter colonisation in such gentamicin in the treatment of serious urinary tract infections in patients requiring parenteral therapy.
Patients and Methods Comparative Study JEFFREY S. LOMBARD PETER C. APPELBAUM Adult patients with a presumptive diagnosis of urinary tract infection who required systemic antibiotic therapy were eligible forstudy. Minimum criteria for enrolment included: fever 37' Division of Infectious Diseases and Epidemiology, Department of signs and symptoms of urinary tract infection; > 10 leucocytes Medicine, Division of Urology, Department of Surgery, and high power field of urinary sediment, and microscopic evidence of Department of Pathology, Pennsylvania State University College of bacteriuria (1 gram-negative rod/oil immersion field in fresh Medicine, Hershey, Pennsylvania USA uncentrifuged urine or bacteria too numerous to count per highpower field in unstained sediment of fresh urine collected by clean 52 patients with serious urinary tract infec- catch or catheterisation). Patients who had had an indwelling Foley were randomised to receive either nephrostomy tube in the preceding 48 h were excluded.
All patients gave informed consent.
aztreonam (35) or gentamicin (17). In the aztreonam group 23 pharmacist assigned the enrolled patients to receive aztreonam patients had unqualified cures, 6 cures with relapse, and 6 (I g every 8 h) or gentamicin (1 mg/kg every 8 h) in a 2:1 manner cures with reinfection; the comparable numbers in the according to a table of random numbers. If bacteraemia was gentamicin group were 9, 1, and 4. There were no failures suspected the doses were increased to 2 g and 1' 7 mg/kg, with aztreonam and 3 with gentamicin. The most important respectively. In patients with creatinine clearances below determinant of outcome was the presence or absence of 30 ml/min, the dose of aztreonam was halved and that of gentamicin urological abnormalities. 11 further patients, with renal was changed according to serum drug levels. Both drugs were given failure or gentamicin-resistant isolates, treated with by intravenous infusion over 20-30 min or intramuscularly if there were all cured. Toxic effects were limited to inadequate venous access; the two routes of administration give similar areas under the serum concentration curve and the liver-function-test proportions in the two groups receiving the drugs intramuscularly aztreonam, whereas deterioration in renal function occurred Treatment was continued for 5-10 days in in 4 gentamicin-treated subjects. Urinary colonisation with uncomplicated cases. Patients with bacteraemia were treated for streptococci occurred in 14 of 46 aztreonam-treated 10-14 days. Treatment was discontinued if pretreatment urine patients (1 required treatment) compared with only 1 of 17 cultures did not grow >105 gentamicin-treated patients. 97% of 309 consecutive gram-negative urinary isolates tested, including 50 Pseudomonas aeruginosa, were susceptible in vitro to aztreonam and 91% to Other patients were treated with aztreonam in an open, gentamicin. Aztreonam may prove an effective and safe unrandomised way if their infecting organisms were known to be alternative to the aminoglycosides.
resistant to gentamicin or if they had renal failure (creatinineclearance <50 ml/min). Enrolment, treatment, and follow-up evaluation were otherwise carried out as in the comparative study.
AZTREONAM is the first of a class of synthetic antimicro- Laboratory Studies bials called monobactams. These are monocyclic 3-lactam Urine was collected for culture and sensitivity testing before drugs that lack the two-ring configuration of penicillin and treatment, after 2-4 days of treatment, and at the end. Test-of-cure cephalosporin molecules. In vitro aztreonam inhibits the cultures were obtained 5-9 days and 4-6 weeks after the last day of growth of most Enterobacteriaceae, including multiply drug treatment. Blood samples for cultures were collected before resistant strains of Serratia marcescens, at concentrations of treatment and, if positive, after 24-48 h of treatment, and 24 h after 2 g/ml or less. 1-3 Moreover, more than 90% of Pseudomonas the completion of treatment. Haematology and chemistry test aeruginosa isolates are inhibited by concentrations of 16 lAg/ml results were monitored for drug-related toxic reactions. These tests and less.I-3 In human beings, 1 were done before the or 2 g given intravenously study, every 3-5 days during treatment, and day of treatment.
concentrations of at least 100-200 J.lg/ml. 4,5 In addition, the serum half-life of approximately 2 h issuitable for a dose interval of 8 or 1'2 h.4,s These in-vitro and Definitions of Outcome pharmacological properties suggest that aztreonam has the Unqualified cure.-Urine cultures were sterile at the completion of potential to replace the aminoglycosides for therapy of gram- treatment and 5-9 days and 4-6 weeks later.
negative infections. We have compared aztreonam with Cure with relapse.-Urine cultures at the end of treatment and 5-9 days later were sterile but those at 4-6 weeks grew ≥105 colony- forming units/ml (cfu/ml) of the original infecting organism.
Cure with reinfection.-Urine cultures at the end of treatment and 5-9 days later were sterile but those at 4-6 weeks Magnus HA. Gastritis. In: Jones FA, ed. Modern trends in gastroenterology. London Butterworth, 1952: 323-51.
of an organism different from the original infecting isolate.
24. Allen A, Garner G. Mucus and bicarbonate secretion in the stomach and their possible Failure.-Urine culture during treatment or 5-9 days later grew role in mucosal protection. Gut 1980; 21: 249-62.
1O5 cfulml of the original infecting organism.
25. Draser BS, Shiner M, McLeod GM. Studies on the intestinal flora I: The of the gastrointestinal Superinfection.-Urine culture during therapy or in healthy and achlorhydric persons. Gastroenterology 1969; 56: 71-79.
after treatment grew an organism different from the original 26. Enander LK, Nilsson F, Ryden AC, Schwan A The aerobic and anaerobic flora of the infecting organism and corresponding gram stain of uncentrifuged gastric remnant more than 15 years after Billroth II resection. Scand J Gastroenterol 1982; 17: 715-20.
per oil immersion field.
27. Mackay IR, Hislop IG. Chronic gastritis and gastric ulcer. Gut 1966; 7: 228-33.
28. Rollason TP, Stone J, Rhodes JM. Spiral organisms in endoscopic biopsies of the human stomach. J Clin Pathol 1984; 37: 23-26.
29. Martin DF, May SJ, Tweedle DE, Hollanders D, Ravenscroft MM, Miller JP.
Comparative Study Difference in relapse rates of duodenal ulcer after healing with cimetidine or tripotassium di-citrato bismuthate. Lancet 1980; i: 7-10.
From July 1, 1982, to Feb 28, 1983, 60 patients were 30. Marshall B, Hislop I, Glancy R, Armstrong J. Histological improvement of active gastritis in patients treated with De-Nol. Aust NZ J Med (in press) (abstr).
comparative study. 8 patients (6 in the

Source: http://www.monetazione.it/forum/immagini/R_123624111_1.pdf