Bismuth Subsalicylate Inthe Treatment of H2 Blocker Resistant Duodenal Ulcers: Role of Helicobacter Pylori

Gut, 1992, 33, 179-183 179 Bismuth subsalicylate in the treatment of H2 blocker

resistant duodenal ulcers: role ofHelicobacterpylori Gut: first published as 10.1136/gut.33.2.179 on 1 February 1992. Downloaded from

S Wagner, M Gebel, K Haruma, W Bar, P Lange, J Freise, U Gladziwa, F W Schmidt

Abstract patients with endoscopically proved H2 blocker Fifty nine patients with Helicobacter pylon resistant duodenal ulcers were recruited to the positive duodenal ulcers that failed to heal study. A resistant duodenal ulcer was defined as after a six week course of treatment with H2 one that failed to heal after at least six weeks' blockers were randomly assigned to one of the continuous treatment with cimetidine 800 mg following three regimens: (i) bismuth subsali- daily or ranitidine 300 mg daily. The largest cylate, 600 mg three times daily (n=19), (ii) ulcer diameter was not less than 5 mm. The ranitidine, 300 mg at night (n=20), (iii) bismuth patients did not have complications of peptic subsalicylate plus ranitidine (n=20). Cumula- ulcer disease, previous gastric surgery, concomi- tive ulcer healing rates after four and eight tant treatment with ulcerogenic drugs, anti- weeks respectively were as follows: bismuth coagulants, or antibiotics, or any serious chronic subsalicylate 74% (14/19) and 95% (18/19), disease. Only patients who had not had antibiotic ranitidine 40% (8/20) and 65% (13/20), bismuth treatment in the previous six months were subsalicylate plus ranitidine 80% (16/20) and included. 95% (19/20). Bismuth subsalicylate treatment At initial endoscopy the ulcer size was assessed was better than ranitidine at both four and at and antral biopsy specimens were taken for eight weeks (p<0*05). The clearance rates for histological examination and H pylori screening. H pylon after four weeks were: bismuth sub- All patients were H pylori positive. Detailed subsalicylate 58%, ranitidine 0%, bismuth sub- information on duration and age of onset of salicylate plus ranitidine 55%. After stopping dyspeptic symptoms, previous drug therapy, bismuth therapy bacterial recrudescence fre- previous ulcer complications, and social habits quently occurred. After bismuth treatment were recorded. 86% (19/22) ofulcers had healed ifHpylori had After giving informed verbal consent, patients been cleared, whereas only 65% (11/17) had were randomised by a nurse (to guarantee blind- healed if H pylon persisted (NS). This study ness of the investigators) using given regimens shows that bismuth subsalicylate is more effec- stratified for 63 subjects. Randomisation pro- tive in the treatment of resistant duodenal cedure was accomplished by a computer pro- http://gut.bmj.com/ ulcers than standard dose ranitidine. It may be gram. Patients were allocated to receive one of that suppression of H pylori by bismuth sub- the following three treatment regimens: (i) salicylate promotes ulcer healing. bismuth subsalicylate (Jatrox) 600 mg three times daily (two chewable tablets half an hour before the three meals); (ii) ranitidine (Zantac) Approximately 85-95% of duodenal ulcers heal 300 mg at night; (iii) bismuth subsalicylate 600 on September 26, 2021 by guest. Protected copyright. within six to eight weeks of treatment with mg three times daily plus ranitidine 300 mg at Departments of standard doses of H2 blockers. '2 The reasons for night. Treatment began within three days of the Gastroenterology and treatment failure remain unclear but inadequate initial endoscopy and was continued for four Hepatology and Clinical control of acid secretion may be a weeks. If healing had occurred after four weeks Microbiology, crucial Medizinische factor.'3 Recently, the identification of Helico- the treatment was stopped; if not the patient Hochschule Hannover, bacter pyloni and its association with duodenal continued on the same regimen for another four Germany ulcer disease led to the hypothesis that H pylon weeks. If at the end of eight weeks the ulcer had S Wagner M Gebel might be involved in H2 blocker resistant ulcera- not healed the patient was withdrawn from the W Bar tion.2 The successful treatment of H2 blocker trial and ranitidine 900 mg/day was given. No P Lange refractory ulcers by colloidal bismuth subcitrate, other medications were allowed during the study J Freise F W Schmidt which is able to eliminate H pylon, supports this period. theory.4 5 Unfortunately, the role of H pylon was Compliance was ascertained by counting the Department of Medicine, not investigated in these therapeutic trials. number of remaining tablets at the end of each Rheinisch-Westfalische treatment Technische Hochschule Bismuth salts exert bactericidal effects on period. Clinical symptoms were asses- Aachen, Germany H pylon and have cytoprotective properties as sed by recording the number and severity ofpain U Gladziwa well, thus it is not clear whether the beneficial episodes on a diary card. Endoscopy and Hpylon tests were First Department of effect of colloidal bismuth in refractory ulcers is screening repeated every four weeks Internal Medicine, directly related to the elimination ofH pylon. during the course of treatment and four weeks Hiroshima University This study aimed to investigate the efficacy of after the end of the trial. School of Medicine, bismuth in the treatment of refrac- Hiroshima, Japan subsalicylate K Haruma tory duodenal ulcers and to elucidate the role of Correspondence to: H pylon. ENDOSCOPY Dr S Wagner, Department of Endoscopies were performed by PL, SW and JF Gastroenterology and Hepatology, Medizinische who were not aware of the clinical data, the bac- Hochschule Hannover, Methods teriological findings, or the treatment regimen. D-3000 Hannover 61, Germany. At each endoscopy five biopsy specimens were Accepted for publication PATIENTS AND STUDY DESIGN taken with sterilised biopsy forceps from antral 7 May 1991 Between September 1987 and May 1990 out- mucosa 2 cm proximal to the pylorus. Two 180 Wagner, Gebel, Haruma, Bar, Lange, Freise, Gladziwa, Schmidt

specimens were placed in 2 ml phosphate STATISTICS buffered saline at 4°C for bacteriological exami- Statistical analysis was carried out by the X2 test nation, two were fixed in 10% formalin for for the evaluation of the healing rates and by the histopathology, and one specimen was used for a matched pairs Wilcoxon signed rank test for Gut: first published as 10.1136/gut.33.2.179 on 1 February 1992. Downloaded from rapid urease detecting test (CLO-test, Delta analysis of H pylon and gastritis scores. Differ- West Ltd, Australia). In addition, smears of ences with p values less than 0 05 were con- biopsy specimens were made for cytological sidered significant. examination using Giemsa staining. Results H PYLORI SCREENING Four ofthe 63 patients who entered the trial were H pylon status was assessed by bacterial culture, lost during follow up (bismuth subsalicylate, n= a modified Giemsa stain, and the CLO-test as 2, ranitidine, n= 1, bismuth subsalicylate plus described previously.6 Two biopsy specimens ranitidine, n= 1). Fifty nine patients completed were cultured under microaerobic conditions in the study and their characteristics are shown in blood agar base containing 5% horse blood and Table I. There were no major differences Skirrow selective supplement for seven days. between the three treatment groups. Median Cultures were considered positive for H pylori if age, sex ratio, mean duration of ulcer history, Gram negative, oxidase positive, catalase smoking habits, and previous H2 blocker positive, and urease positive spiral rods were therapy were comparable (Table I). All patients present. The degree ofcolonisation with Hpylori were H pylon positive. No relevant side effects was estimated by examination of sections and were observed in any treatment group. smears stained with Giemsa and was graded Figure 1 summarises the outcome of the semiquantitatively as follows: 0=no organism; patients under the different regimens. After four 1= occasional; 2 = moderate; 3= large numbers. weeks, ulcer healing had occurred in 74% (14/19) Hpylon status was regarded as positive ifculture of patients receiving bismuth subsalicylate, in was positive or ifthe urease test and Giemsa stain 40% (8/20) of those receiving ranitidine, and in were positive. 80% (16/20) of those treated with a combination The term 'clearance' was used ifHpylon status of both regimens (Table II). In patients with was negative immediately after stopping treat- incomplete ulcer healing the same therapeutic ment. 'Eradication' was used if H pylon was regimen was continued for another four weeks. absent four weeks after the end of treatment. After eight weeks' treatment the cumulative healing percentages were 95% on bismuth sub- HISTOPATHOLOGY salicylate, 65% on ranitidine, and 95% on bis- For histological examinations, formalin fixed muth ranitidine. The healing

subsalicylate plus http://gut.bmj.com/ biopsy samples were embedded in paraffin and rate on bismuth subsalicylate alone or in combi- sections (4 ,tm) were stained with haematoxylin nation with ranitidine was significantly higher and eosin and Giemsa. Each biopsy specimen compared with that on ranitidine, both at four was assessed for the presence, type, density and and at eight weeks (p<005). There was, how- localisation of the inflammatory infiltrate. The ever, no significant difference between bismuth degree ofactivity ofgastritis was graded (0-3) by subsalicylate as a single agent and in combination estimating the density of polymorphonuclear with ranitidine. After eight weeks there was no leukocyte infiltrates as described previously.6 ulcer relapse in patients whose ulcers had healed on September 26, 2021 by guest. Protected copyright. during a four week course of bismuth subsalicylate while two ulcers relapsed in the ranitidine group. TABLE I Patient characteristics and treatment schedules After four weeks H pylori was cleared in 58% of patients receiving bismuth subsalicylate and BSS Ranitidine BSS (3x600 mg) plus in 55% of those treated with bismuth subsalicy- (3 x 600 mg) (300 mg) ranitidine (300 mg) late plus ranitidine, while none of the patients in Patients (n) 19 20 20 the ranitidine group was Hpylon negative (Table Sex (male) 15 13 16 Median age (yrs) 46 47 43 II). At eight weeks only a small group ofpatients Duration of ulcer history (median (yrs)) 7 9 8 was still negative for Hpylon (bismuth subsalicy- Smokers (n) 14 13 11 Pain free (n) 4 6 6 late 21%, bismuth subsalicylate plus ranitidine H pylori positive (n) 19 20 20 20%). Pretreatment (6 wks): Cimetidine 800 mg/day (n) 14 12 16 The degree of mucosal infestation with Ranitidine 300 mg/day (n) 5 8 4 H pylorn was similar in all treatment groups at trial entry (Fig 2). At four weeks a noticeable BBS=bismuth subsalicylate. reduction in H pylon scores was observed in groups treated with bismuth subsalicylate, either alone or in combination with ranitidine. At eight TABLE II Ulcer healing rates and Helicobacter pylori status afterfour and eight weeks' weeks the density of colonisation with H pylori treatment increased again in the bismuth groups because treatment had been stopped in the majority of Ulcer healed H pylori negative patients whose ulcers had healed after four Treatment 4 Weeks 8 Weeks 4 Weeks 8 Weeks weeks. In a subgroup of patients who received BSS 14/19 (74%* 18/19 (95%)* 11/19 (58%) 4/19 (21%) bismuth subsalicylate for eight weeks H pylori Ranitidine 8/20 (40%) 13/20 (65%) 0/20 (0%) 0/20 (0%) scores resembled those at four weeks (data not BSS+ranitidine 16/20 (80%)* 19/20 (95%)* 11/20 (55%) 4/20 (20%) shown). Ranitidine monotherapy had no signifi- *p<0.05; BSS v ranitidine and BSS plus ranitidine v ranitidine. BSS=bismuth subsalicylate. cant effect on Hpylori scores at any time. Bismuth subsalicylate in the treatment ofH2 blocker resistant duodenal ulcers: role ofHelicobacter pylori 181

14 Healed

Bismuth 4 Healed Gut: first published as 10.1136/gut.33.2.179 on 1 February 1992. Downloaded from subsalicylate 19

8 Healed * 1 Not healed 7 Healed Ranitidine 20 12 _

16 Healed 5 Not healed Bismuth 3 Healed subsalicylate 20 Figure 1: Healing ofH2 + 4 > blocker resistant duodenal ranitidine ulcers with bismuth subsalicylate, ranitidine, or 1 Not healed both, atfour and at eight weeks. *Two relapsed at 0 4 8 12 eight weeks. Time (weeks)

In the biopsy specimens of the antrum taken subsalicylate in eradicating H pylori, patients before entry active chronic gastritis was found to were reinvestigated~~~~~~~~~~~~~~1four weeks after stopping a similar extent in all study groups (Fig 3). bismuth treatment. Thirty patients received bis- Gastritis scores fell after four weeks of bismuth muth subsalicylate for four weeks, and of these subsalicylate and were still somewhat lower at three were still Hpylon negative four weeks after eight weeks in these patients when compared completing treatment, resulting in an eradica- with basal values. In parallel to the H pylon tion rate of 10%. Nine patients were treated with scores, ranitidine monotherapy failed to improve bismuth subsalicylate for eight weeks, only one gastric inflammation at any time. patient (11%) eradicated H pylon in this sub- Table III shows ulcer healing in relation to group. clearance of H pylon in patients treated with At trial entry epigastric pain was recorded in bismuth subsalicylate alone or in combination 14/19, 16/20, and 14/20 respectively in treatment with ranitidine. Ulcer healing was higher after groups (i), (ii), and (iii). Most patients became clearing of H pylon in comparison with the asymptomatic after two weeks and there was no http://gut.bmj.com/ permanently infected patients, but this differ- difference in the proportion of patients with ence did not reach statistical significance (healing residual symptoms in the three treatment rate 86% v 65%; odds ratio 3 15; NS). Some 65% groups. (11/17) of ulcers healed during a four week course of bismuth (alone or combined with ranitidine) despite persistence of H pylon in Discussion gastric mucosa, even though there had been a There is no general aggreement on the definition on September 26, 2021 by guest. Protected copyright. reduction in the overall bacterial load. of H2 blocker resistant duodenal ulcers.2 Since In order to study the efficacy of bismuth most ulcers heal after six weeks' treatment with

-i 2-

Figure 2: Mean (SD) ofH pylori colonisation in antral biopsy specimens under different therapies before treatment and atfour and O:- 1., eight weeks. A significant o..... Before entry 4 Weeks..-- 8 Weeks decrease in H pylori colonisation in comparison with the preentry level is I[_Bismuth subsalicylate L Ranitidine - Bismuthsubsalicylate + ranitidine indicated by an asterisk. 182 Wagner, Gebel, Haruma, Bar, Lange, Freise, Gladziwa, Schmidt

3- Gut: first published as 10.1136/gut.33.2.179 on 1 February 1992. Downloaded from

2- .*1

Figure 3: Mean (SD) scores ofgastritis in antral biopsy specimens under different therapies before treatment 0- _. I-. i and atfour and eight weeks. fore 4 A significant decrease in the entry Weeks 8 Weeks gastritis score in comparison with the pre-entry level is | e Bismuthsubsalicyate [ Ranitidine Bismuth.subsalicylate + ranitidine indicated by an asterisk.

a standard dose of an H2 blocker, this time comparable with those achieved with tripo- criterion was chosen in the present study for the tassium dicitratobismuthate in the trials of Lam definition of a resistant ulcer.7 In our series a et al and Bianchi Porro et al. Thus, the efficacy of further four week course of treatment with bismuth salts in ulcer healing seems to be related ranitidine 300 mg was able to heal 40% of the solely to the bismuth component while the anion resistant duodenal ulcers, and 65% were healed may be of minor importance. after eight weeks. Therefore, only a small group Bismuth salts are site protective agents which ofpatients is truly resistant to H2 blockers, while have various cytoprotective properties."'` most represent slow ulcer healing. Additionally, they have recently been shown to shows that bismuth sub- Our controlled trial exert bactericidal effects on H pylori.' , The http://gut.bmj.com/ salicylate is superior to standard dose ranitidine present study is the first which attempts to in the treatment of resistant ulcers. Both at four elucidate the role of H pylori in the healing of and at eight weeks, the healing rate was signifi- resistant duodenal ulcers. None of the patients cantly higher with bismuth subsalicylate than treated with ranitidine cleared H pylon, nor did with ranitidine treatment. The combination of they show a decrease in the degree of bacterial bismuth subsalicylate and ranitidine resulted in infestation. Some 65% (11/17) of the duodenal healing rates similar to those with bismuth ulcers healed with bismuth subsalicylate despite on September 26, 2021 by guest. Protected copyright. subsalicylate alone. persistance of H pylori, although at a decreased While the efficacy of colloidal bismuth sub- number. Thus, bacterial elimination is not a citrate in the treatment ofduodenal ulcers is well prerequisite for healing of resistant duodenal established,8`10 only two reports have addressed ulcers. the role ofbismuth in the treatment ofrefractory After four weeks' treatment with bismuth, duodenal ulcers.45 Lam et al have shown that however, 86% (19/22) of ulcers had healed if the tripotassium dicitratobismuthate (120 mg four bacterium was cleared, whereas only 65% (11/17) times daily) heals cimetidine resistant ulcers had healed if H pylori persisted. This difference significantly better than high dose cimetidine did not reach statistical significance due to the (1-6 g/day) (healing rates 85% v 40%). Similar small number of patients but the finding may results have been obtained by Bianchi Porro et al, lend support to the hypothesis that bacterial who reported significantly higher healing rates clearance promotes ulcer healing. Recently, on tripotassium dicitratobismuthate (120 mg Marshall et al showed that duodenal ulcer heal- four times daily) than on cimetidine (1-2 g and 2 ing is significantly higher in patients in whom g). Our study shows for the first time that H pylori has been eliminated than in those with bismuth subsalicylate is effective in the treat- persistent bacterial infection. 16 In our study, ment of resistant duodenal ulcers. The healing 14% (3/22) of duodenal ulcers did not heal rates on bismuth subsalicylate in our study are despite of clearance of H pylon, suggesting that clearance of H pylori does not necessarily result in ulcer healing. It could be, however, that TABLE III Ulcer healing in relation to clearance ofH pylori bacterial eradication is a stronger promoter of in patients treated with bismuth subsalicylate alone or in ulcer healing than temporary clearance of combination with ranitidine H pylori. Ulcer healed Ulcer not healed Bismuth treatment reduced the density of at 4 weeks at 4 weeks Total bacterial colonisation of the gastric mucosa, but H pylonr-ve 19 3 22 there was only a transient clearance of H pylon. Hpylor +ve 11 6 17 Stopping treatment was frequently associated Total 30 9 39 with a recurrence of the organism in the gastric Bismuth subsalicylate in the treatment ofH2 blocker resistant duodenal ulcers: role ofHelicobacter pylori 183

mucosa showing that eradication of H pylon is a 4 Lam SK, Lee NW, Koo J, Hui WM, Fok KH, Ng M. Randomised crossover trial of tripotassium dicitrato rare event during bismuth treatment. In agree- bismuthate versus high dose cimetidine for duodenal ulcers ment with previous studies,'7"1 we found a close resistant to standard dose cimetidine. Gut 1984; 25: 703-6. 5 Bianchi Porro G, Parente F, Lazzaroni M. Tripotassium relation between the grade of chronic gastritis dicitrato bismuthate (TDB) versus two different dosages of Gut: first published as 10.1136/gut.33.2.179 on 1 February 1992. Downloaded from and the degree of infestation of gastric mucosa cimetidine in the treatment of resistant duodenal ulcers. Gut 1987; 28: 907-11. with H pylon. A reduction in the H pylon score 6 Wagner S, Freise J, Bar W, Fritsch S, Schmidt FW. was associated with an improvement in gastric Epidemiologie und Therapie der Campylobacter-pylori- Infektion. Dtsch Med Wochenschr 1989; 114: 407-13. inflammation. Therefore, the difference in ulcer 7 Delchier JC, Isal JP, Eriksson S, Soule JC. Double blind healing rates in patients treated with bismuth v multicentre comparison of omeprazole 20 mg once daily versus ranitidine 150 mg twice daily in the treatment of those treated with ranitidine may be partly a cimetidine or ranitidine resistant duodenal ulcers. Gut 1989; result of an improvement of histology. 30: 1173-8. 8 Lee FI, SamloffIM, Hardman M. Comparison of tripotassium The role of H pylon in acute ulcer healing dicitratobismuthate tablets with ranitidine in healing and seems to be of less importance than its role in the relapse ofduodenal ulcers. Lancet 1985; i: 1299-302. 9 Hamilton I, O'Connor HJ, Wood NC, Bradbury I, Axon recurrence of duodenal ulcers. A growing body ATR. Healing and recurrence of duodenal ulcer after of evidence has accumulated indicating that treatment with tripotassium dicitrato bismuthate (TDB) tablets or cimetidine. Gut 1986; 27: 106-10. eradication of H pylori is associated with a 10 Tytgat GNJ. Colloidal bismuth subcitrate in peptic ulcer - A significant reduction in ulcer relapses and may review. Digestion 1987; 37 (suppl 2): 31-41. 11 Konturek SJ, Radecki T, Piastucki I, Drozdowics D. even cure duodenal ulcer disease. 5 16 19 Advances in the understanding ofthe mechanism ofcytopro- The causes of H2 blocker ineffectiveness in tective action by colloidal bismuth subcitrate. Scand J Gastroenterol 1986; 21 (suppl 122): 6-10. acute ulcer healing have not been fully clarified. 12 Wagstaff AJ, Benfield P, Monk JP. Colloidal bismuth sub- Inadequate acid suppression has been frequently citrate. A review of its pharmacodynamic and pharmaco- kinetic properties, and its therapeutic use in peptic ulcer observed in H2 antagonist non-responders.2"23 In diseases. Drug 1988; 36: 132-57. addition, almost all H2 blocker resistant duo- 13 Hall DWR. Review ofthe modes ofaction ofcolloidal bismuth subcitrate. Scandj Gastroenterol 1989; 24 (suppl 157): 3-6. denal ulcers can be healed by the potent proton 14 Ericsson CD, Tannenbaum C, Charles TT. Antisecretory and pump inhibitor omeprazole.2425 Therefore, antiinflammatory properties of bismuth subsalicylate. Rev Infect Dis 1990; 12 (suppl 1): S11-5. inadequate control ofacid secretion by H2 block- 15 Coghlan JG, Humphries H, Dooley C, Keane C, Gilligan D, ers seems to be the most important factor in McKenna D, et al. Campylobacter pylon' and recurrence of duodenal ulcers - a 12-month follow-up study. Lancet 1987; treatment failures. ii: 1109-11. In conclusion, our study shows that bismuth 16 Marshall BJ, Goodwin CS, Warren JR, Murray R, Blincow ED, Blackbourn SJ, et al. Prospective double-blind trial of subsalicylate is effective in the treatment of H2 duodenal ulcer relapse after eradication of Campylobacter blocker resistant duodenal ulcers. In addition, pylori. Lancet 1988; ii: 1437-42. 17 Rauws EAJ, Langenberg W, Houthoff H, Zanen HC, Tytgat bismuth subsalicylate is superior to a prolonged GNJ. Campylobacter pyloridis-associated chronic active administration of standard dose H2 antagonists. antral gastritis. Gastroenterology 1988; 94: 33-40. 18 Stolte M, Eidt S, Ritter M, Bethke B. Campylobacterpylori und The beneficial effects of bismuth subsalicylate in gastritis - association oder induktion. Pathologe 1989; 10: the treatment of resistant ulcers may result from 21-6.

19 Rauws EAJ, Tytgat GNJ. Cure of duodenal ulcer associated http://gut.bmj.com/ suppression of H pylon and the associated with eradication of Helicobacter pylori. Lancet 1990; i: improvement of gastritis in addition to its well 1233-5. 20 Savarino V, Mela GS, Scalabrini P, Celle G. H2-receptor known cytoprotective properties antagonist non-responders. Lancet 1987; ii: 1281. 21 Wagner S, Freise J, Schmidt FW. H2-receptor antagonist Part of this work was presented at the 92nd Meeting of the non-responders. Lancet 1988; i: 128. American Gastroenterological Association in New Orleans, on 22 Johnston DA, Wormsley KG. The effects of fasting on 24-h May 20, 1991, and was published in abstract form: Gastro- gastric secretion ofpatients with duodenal ulcers resistant to enterology 1991; 100: A180. ranitidine. AlimentPharmacol Therap 1989; 3: 471-9. 23 Collen MJ, Stanczak VJ, Ciarleglio CA. Refractory duodenal ulcers (nonhealing duodenal ulcers with standard doses of on September 26, 2021 by guest. Protected copyright. 1 Wormsley KG. Duodenal ulcers which do not heal rapidly. antisecretory medication). Dig Dis Sci 1989; 34: 233-7. BMJ 1984; 289: 1095. 24 Tytgat GNJ, Lamers CBHW, Hameetman W, Jansen JMBJ, 2 Domschke W, Lam SK, Pounder RE, Andersen D. H2- Wilson JA. Omeprazole in peptic ulcers resistant to hista- blocker-resistant duodenal ulceration. Gastroenterol mine H2-receptor antagonists. Aliment Pharnacol Therap International 1989; 2: 85-91. 1987; 1: 31-8. 3 Gledhill T, Buck M, Hunt RH. The effect of no treatment, 25 Brunner G, Creutzfeldt W, Harke U, Lamberts R. Therapy cimetidine 1 g/d, cimetidine 2 g/d and cimetidine combined with omeprazole in patients with peptic ulcerations resistant with atropine on nocturnal gastric secretion in cimetidine to extended high-dose ranitidine treatment. Digestion 1988; non-responders. Gut 1984; 25: 1211-6. 39: 80-90.