Annals ofthe Rheumatic Diseases 1992; 51: 1089-1093 1089 Ann Rheum Dis: first published as 10.1136/ard.51.9.1089 on 1 September 1992. Downloaded from REVIEW in the prevention of gastroduodenal damage in rheumatology

A B Ballinger, P J Kumar, D L Scott

Abstract suggested there is no benefit to be gained from Patients receiving non-steroidal anti-inflam- long term prophylaxis. matory drugs (NSAIDs) are at an increased When should misoprostol be used? To risk of gastroduodenal erosions, ulcers, and resolve this issue we have reconsidered the the associated complications of haemorrhage, incidence, pathogenesis, treatment, and pre- perforation, and death. Many NSAID asso- vention of NSAID induced gastroduodenal ciated ulcers that bleed or perforate have been damage, focusing on the role of misoprostol and asymptomatic until the time of presentation its advantages and drawbacks. and conversely many patients with dyspepsia do not have ulcers. Symptoms are a poor guide to the presence of an ulcer. During Incidence of NSAID induced gastroduodenal continued treatment with NSAIDs misoprostol damage is the best choice for NSAID induced gastro- Although a subject's risk from NSAIDs is not duodenal damage; it achieves higher rates of high, their widespread use makes gastro- healing than other drugs in these circum- duodenal damage due to NSAIDs a major stances. Misoprostol is superior to other problem in the population as a whole. Extrap- drugs in the prevention of gastric damage but olation from 'record linkage' data indicates misoprostol and H2 antagonists are of similar that NSAID use is associated with 30 000 benefit in the duodenum. Prophylactic studies serious gastrointestinal events each year in the have ali used endoscopic damage as an United Kingdom4 with an associated mortality endpoint, and much larger studies will be of 10%. The prevalence of ulcers in the needed to show an effect of misoprostol on and duodenum of long term users of NSAIDs the incidence of ulcer complications. There ranges from 10 to 30%"7 and there is an are no as to increased clear guidelines which patients occurrence of upper gastrointestinal http://ard.bmj.com/ should receive prophylactic treatment with haemorrhage and perforation. Case control misoprostol but those particularly at risk of studies show a clear relation between upper ulcer complications-that is, those with gastrointestinal bleeding and NSAIDs. Holvoet previous peptic ulceration, the elderly, et al 8 studied 161 patients admitted to hospital medically unfit, patients receiving large doses with upper gastrointestinal bleeding who had of NSAIDs, and those patients receiving oesophagitis, gastric erosions, and gastric or steroids in addition to NSAIDs-should be duodenal ulcers at endoscopy. Using age and considered. sex matched hospital inpatients as controls they on September 25, 2021 by guest. Protected copyright. found a highly significant difference between (Ann Rheum Dis 1992; 51: 1089-1093) patients and controls in the use of non-aspirin NSAIDs and aspirin. Previous NSAID use was Non-steroidal anti-inflammatory drugs (NSAIDs) a risk factor for bleeding from gastric and are widelyprescribed and are effective for treating duodenal ulceration but not oesophagitis or inflammatory arthritis. Their therapeutic gastric erosions. A similar study of Somerville et benefits are accompanied by clinically significant al ' found patients with upper gastrointestinal gastrointestinal side effects including dyspepsia, bleeding used NSAIDs three times more often gastroduodenal ulcers, and the associated than did subjects in the community or hospital Department of complications of and case Gastroenterology, haemorrhage perforation. control groups. Such control studies have St Bartholomew's Fries et al i have shown in a large number of potential for bias as patients who believe Hospital, patients with computerised clinical records that NSAIDs cause ulcers are more likely to report Charterhouse Square, thereis a relatively unfavourablerisk/benefit ratio their use. In the infarction London ECIM 6BQ, aspirin myocardial United Kingdom for the long term use of NSAIDs due to study'0 where patients and investigators were A B Ballinger gastrointestinal toxicity. There is, however, the blinded, 4524 subjects received aspirin or P J Kumar potential to prevent some of this toxicity by placebo; over three years the relative risk for Department of co-prescribing the synthetic admission to hospital from duodenal ulcer Rheumatology, St Bartholomew's misoprostol. Overviews have given conflicting disease was 10-7 times higher in subjects treated Hospital, advice about the value of misoprostol for the with aspirin than placebo treated subjects. Charterhouse Square, prophylaxis of NSAID induced gastroduodenal Forty per cent of patients treated with aspirin London ECIM 6BQ, damage. Soil et al 2 the balance of to ulcer disease United Kingdom thought admitted hospital for peptic had D L Scott evidence was in favour of prophylaxis with gastrointestinal bleeding. Many NSAID asso- in some at risk are Correspondence to: misoprostol especially groups high ciated ulcers that bleed or perforate clinically Dr Scott. ofgastroduodenal damage. By contrast Barrison3 silent with little or no antecedent ulcer-type 1090 Ballinger, Kumar, Scott

pain before the onset of complications. " gastric ulcers after four weeks of treatment Ann Rheum Dis: first published as 10.1136/ard.51.9.1089 on 1 September 1992. Downloaded from Conversely many patients receiving NSAIDs compared with 75% healing in placebo treated have dyspepsia but do not have an ulcer or patients.25 Although there have been no com- erosions at endoscopy. parative trials of misoprostol and H2 blockers, in a similar study achieved only 71% healing after four weeks of treatment.26 Pathogenesis of ulcers associated with When it is clinically preferable to continue NSAIDs the treatment with NSAIDs, misoprostol is The mucosal lining of the stomach resists injury the treatment of choice to achieve ulcer by several mechanisms. Secretion of mucus and healing. Roth et al showed that misoprostol hydrogen carbonate allows the apical surface of coadministered with aspirin is effective in cells to be relatively alkaline compared with the healing aspirin associated gastroduodenal acidic environment of the stomach lumen. damage in patients with rheumatoid arthritis27; Epithelial cells migrate along the basement 238 patients received a baseline endoscopy and membrane to fill in gaps which form in the were then randomised to receive either miso- epithelial cell lining after mucosal damage. prostol or placebo while continuing treatment Mucosal blood flow provides continuous with aspirin. After eight weeks healing of nutrients, 02 and hydrogen carbonate, and gastric mucosal injury had occurred in 70% of disposes of diffused . Endogenous prosta- the misoprostol treated group compared with glandins play an important part and bind 25% of the placebo treated subjects. Gastric directly to parietal cells to inhibit acid sec- ulceration was reduced in the misoprostol retion,12 stimulate secretion of mucus and treated group; by eight weeks the healing rates hydrogen carbonate,'3 and maintain mucosal were 62% in misoprostol treated subjects com- blood flow. pared with 32% receiving placebo. Duodenal NSAIDs damage the gastric mucosa by an injury healed in 86% of the misoprostol treated irritative topical effect and an indirect systemic group and 53% receiving placebo. Of these only effect owing to the inhibition of production of 18 patients had duodenal ulcers on the baseline . Enteric coating of NSAIDs endoscopy; by eight weeks healing was seen in reduces damage following acute administration 90% receiving misoprostol and in 50% of but is not effective in preventing ulceration.'4 controls. Sulindac, a prodrug which is not active until Most studies have shown no benefit of H2 after absorption and hepatic metabolism, is antagonists over placebo in healing NSAID associated with peptic ulceration"' and paren- induced ulcers and damage during continued teral administration ofNSAIDs produces gastric administration ofNSAIDs.2>30 In patients with ulceration.'6 Levi et al " showed that NSAIDs classical peptic ulceration misoprostol (200 [tg inhibit proliferation of mucosal cells at the edge four times daily) is as effective as cimetidine of ulcers, a process important for ulcer healing, (300 mg four times daily) in healing gastric and

and this effect was reversed by misoprostol. duodenal ulceration at four and eight weeks.3' 32 http://ard.bmj.com/ It has been suggested that with continued There have been no long term studies directly administration of NSAIDs gastric mucosal comparing misoprostol with H2 antagonists or injury lessens and may resolve, a process termed other antiulcer drugs in the healing of NSAID adaptation that is well documented in healthy induced damage during continued admini- volunteers.'8 19 The situation may be different stration of NSAIDs but the data which are in patients, however. Two studies20 21 have available suggest that misoprostol is superior. shown that the risk of from Sucralfate is not useful in the of complications healing on September 25, 2021 by guest. Protected copyright. NSAID induced peptic ulcers is greatest in the gastric erosions and ulcers during continued first month of treatment. This finding is con- treatment with NSAIDs. Over four weeks 1 g sistent with the development of mucosal four times daily is no better than placebo or adaptation with long term administration of cimetidine.33 34 Sucralfate and have NSAIDs but could also result from early been compared in the healing ofNSAID induced discontinuation of these drugs among patients gastric and duodenal ulcers. Omeprazole (20 who are intolerant of side effects. Other studies mg/day) produced better healing than sucralfate have shown no relation between duration of (1 g four times daily) and healed 100% of peptic treatment and development of ulcer complica- ulcers at four weeks.35 In a comparative trial tions.8 In one study22 78 patients with arthritis omeprazole was significantly better than rani- who had been receiving long term treatment tidine; healing of ulcers with 40 mg omeprazole with NSAIDs had non-ulcer damage at an was 81% at four weeks compared with 32% with initial screening endoscopy; after two further .36 weeks of treatment with NSAIDs 19% had In conclusion, for patients with a peptic ulcer developed peptic ulcers. Serial endoscopies do seen at endoscopy misoprostol is the current not show a decrease in mucosal injury with treatment of choice to achieve ulcer healing continued administration of NSAIDs for up to during continued treatment with NSAIDs. one year.23 24 Dyspepsia is a poor guide to the presence of ulceration and starting treatment cannot be recommended on the basis of symptoms alone. Treatment of NSAID associated ulcers When patients develop gastrointestinal ulcera- tion while receiving NSAIDs it is best to stop Prevention of NSAID associated treatment wherever possible and heal the ulcer. gastroduodenal damage In this situation misoprostol will heal 95% of Misoprostol is superior to placebo in preventing Misoprostol in the prevention ofgastroduodenal damage in rheumatology 1091

gastroduodenal damage induced by NSAIDs in In patients H2 blockers have been ineffective in Ann Rheum Dis: first published as 10.1136/ard.51.9.1089 on 1 September 1992. Downloaded from healthy volunteers and patients. Table 1 preventing gastric injury but are significantly summarises recent studies. Gastroduodenal better than placebo in preventing duodenal damage often occurs early (in two weeks) during damage.45 In comparative trials misoprostol is treatment with NSAIDs and this is prevented superior to ranitidine in the prevention of by coadministration with misoprostol.22 Longer gastric ulceration but both H2 blockers and term studies have shown that the efficacy of misoprostol have produced similar results in the misoprostol is maintained. Graham et al 5 prevention of duodenal injury' (figure). Fewer recruited patients with osteoarthritis receiving trials have been conducted with either sucralfate NSAIDs who at an initial endoscopy were free or omeprazole; sucralfate is inferior to miso- of gastric ulceration. Patients were randomly prostol42 and is no better than placebo in assigned to misoprostol (100-200 pg four times preventing gastroduodenal erosions or ulcera- daily) or placebo during continued treatment tion." In volunteer studies omeprazole does not with NSAIDs; repeat endoscopy was performed prevent gastric damage but prevents duodenal at one, two, and three months. The cumulative damage and in this respect provides similar three month prevalence of gastric ulceration was protection to ranitidine.48 22% in the placebo group, 6% in the 100 pg In conclusion misoprostol is significantly misoprostol group, and only 1% in patients better than placebo and standard antiulcer receiving 200 [tg misoprostol. The two doses of treatment in the prevention of NSAID induced misoprostol were significantly better than gastric injury. Misoprostol and H2 blockers placebo. In a 12 month study the cumulative provide similar protection against duodenal prevalence of gastric ulceration was 12-5% in damage. When considering prophylactic treat- the group of patients receiving misoprostol ment, however, the drug chosen must be able to (600-800 R,g/day), compared with 30% in the provide gastric and duodenal protection and group receiving placebo (p<0r05).23 therefore misoprostol is the current treatment of Studies with H2 antagonists have produced choice. In clinical studies ofthe coadministration variable results in placebo controlled trials. In ofmisoprostol with NSAIDs there is no evidence normal volunteers cimetidine prevented acute that the-efficacy of the NSAID is decreased. gastric mucosal injury induced by a single dose Synovial fluid concentrations of prostaglandins of aspirin43 but was not effective in reducing are unchanged by misoprostol and the con- gastroduodenal erosions or ulcers in a one week centration of thromboxane B2 is actually trial during coadministration with naproxen.4 decreased,49 suggesting that misoprostol exerts an anti-inflammatory effect rather than pro- inflammatory which may be expected by Table I Controlled studies showing prevention of non-steroidal anti-inflammatory drug damage by misoprostol. exogenous administration of prostaglandins. Author Year Number Duration Site of Signifcancet of cases (weeks) damage*

Which patients should receive antiulcer http://ard.bmj.com/ Bardhan et al 37 1991 277 2 D/G 0 005 De Rossi et al 38 1990 95 4 D/G <0 05 prophylaxis? Schimke et al 3 1990 82 2-4 D/G 0-002 Fries et al have examined a large cohort of Saggioro et al 4 1988 153 4 D/G <0-01 with and defined patients Graham et al 41 1991 374 12 D/G <0-01 patients arthritis Geis et al 24 1990 244 12-48 D/G 0 005 within that group who are particularly at risk Graham et al 1988 420 12 G <0 001 risk patients Agrawal et al 42 1990 259 12 G <0 001 for ulcer complications.' High Elliott et al 23 1990 83 12-48 G <0 05 were older, had previously stopped NSAIDs

often on September 25, 2021 by guest. Protected copyright. *D/G=duodenal or gastric. because of upper abdominal pain and were tCompared with placebo or other drug used. receiving corticosteroids. Their findings concur with other studies9 50 which have also identified high doses ofNSAIDs20 and degree offunctional disability as risk factors in patients with arthritis. i Misoprostol (Cytotec) Table 2 shows potential indications for co- 200 1Lg four times a day (n = 230) administration of misoprostol with NSAIDs. Ranitidine 150 mg twice daily (n = 235) 8F 7-2 Unanswered questions 7 The most important question is whether 6 5.9 coprescribing misoprostol with NSAIDs reduces the occurrence of serious life threatening 5 complications of gastroduodenal damage. The 40 4 logical extrapolation from the endoscopic

0~ studies is that it is likely to be so, but there is no 3 2-6 direct evidence that this is true. The principal 2 '13 1-3 1±3 Table 2 Recognised indications for coadministration of misoprostol with non-steroidal anti-inflammatory drugs 0 (NSAIDs) Duodenal Gastric Gastroduodenal ulceration ulceration ulceration Elderly patients NS p = 0-01 p = 0.03 Medically unfit who may not withstand an ulcer complication Previous peptic ulceration High doses of NSAIDs Intent to treat analysis ofulcer rates in patients receiving non-steroidal anti-inflammatory Patients receiving steroids drugs across eight week study period7. p Values were obtained using Fisher's exact test. 1092 Ballinger, Kumar, Scott Ann Rheum Dis: first published as 10.1136/ard.51.9.1089 on 1 September 1992. Downloaded from reason is that serious complications are un- 15 Lanza F I, Nelson R S, Ruck M F. A controlled endoscopic study comparing the toxic effects of sulindac, naproxen, common and that prospective studies would aspirin and placebo on the gastric mucosa of healthy have to be large to have sufficient power to show volunteers. J Clin Pharmacol 1984; 24: 89-95. 16 Lanza F L, Karlin D K, Yeo J P. A double-blind placebo a reduction in life threatening problems resulting controlled endoscopic study comparing the mucosal injury from NSAID induced gastroduodenal ulcers. seen with orally and parenterally administered non-steroidal analgesic ketorolac tromethanine at therapeutic and sub- Does misoprostol reduce the incidence of all therapeutic doses [abstract]. Am J Gastroenterol 1987; 82: NSAID related ulcers, including those likely to 939. 17 Levi S, Goodlad R A, Lee C Y, et al. Inhibitory effects of bleed or perforate, or does it decrease the non-steroidal anti-inflammatory drugs on mucosal cell frequency only of the more benign ulcers? The proliferation associated with gastric ulcer healing. Lancet 1990; 336: 840-3. reasonable approach is to consider the first 18 Graham D Y, Smith J L, Spiut H S, Tolles E. Gastric proposition most likely, but to continue seeking adaption. Studies in humans during continuous aspirin administration. Gastroenterology 1988; 95: 327-33. confirmatory evidence. 19 Shorrock C J, Prescroft R J, Rees W D W. The effects of A second question is whether misoprostol indomethacin on gastroduodenal morphology and mucosal pH gradient in the healthy human subject. Gastroenterology also has beneficial effects lower down the 1990; 99: 334-9. which are clinically signi- 20 Griffin M R, Piper J M, Daugherty J R, Snowden M, Ray ficant. There is some evidence5' 52 that this may W A. Non-steroidal anti-inflammatory drug use and increased risk for in elderly patients. be so and this would swing the pendulum in Ann Intern Med 1991; 114: 257-63. favour of coprescribing misoprostol more often. 21 Carson J L, Strom B L, Soper R A, West S L. The dissociation of non-steroidal anti-inflammatory drugs and A final point is where to draw the line in upper gastrointestinal bleeding. Arch Intern Med 1987; 147: coprescribing misoprostol. Subjects in whom 85-8. 22 Scott D L, Bardhan K D, Bjarnason I, Griffin W M, Fenn there is a high risk of complications from G C, Shield M J. Misoprostol and NSAID-associated ulcers are a gastro-duodenal damage. BrJ Rheumatol 1991; 30 (suppl 2): gastroduodenal minority. Many 9. serious adverse events due to NSAIDs in the 23 Elliott S L, Yeomans N O, Buchanan RR C, et al. Longterm upper tract will affect effects of misoprostol on gastropathy induced by non- gastrointestinal patients steroidal anti-inflammatory drugs. Clin Exp Rheumatol without other risk factors. At what point is it 1990; 8 (suppl 4). reasonable not to use misoprostol? This issue 24 Geis G S, Stead H. Prevention of diclofenac induced gastroduodenal damage by cytotec. Clin Exp Rheumatol should be resolved from the viewpoint that the 1990; 8 (suppl 4). less taken the less the risk of 25 Graham D, Jaszewski R, Stromatt S, Brown J. Treatment of drugs problems, NSAID induced gastric ulcer with misoprostol. Gastro- but it is as yet unanswerable and must remain enterology 1990; 98: A52. within the remit of the of doctors. 26 Lancaster Smith M J, Jaderberg M E, Jackson D A. judgment Ranitidine in the treatment of non-steroidal anti-inflam- matory drug associated gastric and duodenal ulcer. Gut We are grateful to the Arthritis and Rheumatism Council for 1991; 32: 252-5. their support. Dr Ballinger is a research fellow supported by an 27 Roth S H, Agrawal N M, Mahawald M, et al. Misoprostol Aylwen Bursary, St Bartholomew's Hospital. Dr Scott is Muir heals gastroduodenal injury in patients with rheumatoid Hambro Fellow of the Royal College of Physicians. arthritis receiving aspirin. Arch Intern Med 1989; 149: 775-9. 28 Roth S H, Bennett R E, Mitchell C S. Cimetidine in non-steroidal anti-inflammatory drug gastropathy. Double- 1 Fries J F, Miller S R, Spitz P W, Williams C A, Hubert H B. blind by long-term examination. Arch Intern Med 1987; Bloch D A. Toward an epidemiology of gastropathy 147: 1798-1801.

associated with nonsteroidal anti-inflammatory drug use. 29 O'Laughlin J C, Silvoso G K, Ivey K J. Resistance to medical http://ard.bmj.com/ Gastroenterology 1989; %: 647-55. therapy of gastric ulcer in rheumatic disease patients taking 2 Soll A H, Weinstein W M, Kurata J, McCarthy D. aspirin. Dig Dis Sci 1982; 27: 976-80. Nonsteroidal anti-inflammatory drugs and peptic ulcer 30 Wallin M D, Frank W 0, Fox M J. The effects of cimetidine disease. Ann Intern Med 1991; 114: 307-19. drug on the healing of non-steroidal anti-inflammatory 3 Barrison I. Prophylaxis against nonsteroidal induced upper drug-induced gastroduodenal damage while continuing gastrointestinal side effects. Ann Rheum Dis 1991; 50: NSAID therapy. AmJI Gastroenterol 1988; 83: 1076. 207-9. 31 Rachmilewitz D, Chapman J W, Nicholson P A. A multi- 4 Beardon P H G, Brown S Y, McDevitt D G. Gastrointestinal centre international controlled comparison of two dosage events in patients prescribed nonsteroidal anti-inflammatory regimes of misoprostol with cimetidine in treatment of drugs: a controlled study using record linkage in Tayside. gastric ulcer in outpatients. Dig Dis Sci 1986; 31: 755-805. Q Jf Med 1989; 71: 497-505. 32 Nicholson P A. A multicenter international controlled on September 25, 2021 by guest. Protected copyright. 5 Graham D Y, Agrawal N M, Roth S H. Prevention of comparison of two dosage regimes of misoprostol and NSAID-induced gastric ulcer with misoprostol: multi- cimetidine in the treatment of duodenal ulceration in centre, double blind, placebo-controlled trial. Lancet 1988; outpatients. Dig Dis Sci 1985; 30: 171-7. ii: 1277-80. 33 Caldwell J R, Roth S H, Wu W C, et al. Sucralfate treatment 6 Morris A D, Holt S D, Silvoso G R, et al. Effect of anti- of nonsteroidal anti-inflammatory drug-induced gastro- inflammatory drug administration in patients with rheuma- intestinal symptoms and mucosal damage. AmJ' Med 1987; toid arthritis. An endoscopic assessment. Scand Gastro- 83 (suppl 3B): 74-82. enterol 1981; 18: 131-5. 34 Shepherd H A, Fine D, Hillier K. Effect of sucralfate and 7 Silvoso G R, Ivey K J, Butt J H, et al. Incidence of gastric cimetidine on rheumatoid patients with active gastroduo- lesions in patients with rheumatoid disease on chronic denal lesions who are taking non steroidal anti-inflam- aspirin therapy. Ann Intern Med 1979; 91: 517-20. matory drugs. Am J Med 1989; 86: 49-54. .3 Holvoet J, Terriere L, Van Hee W, Verbist L, Fierens E, 35 Bianchi Porro G, Santalucia F, Petrillo M. Omeprazole Hautekeefe M L. Relation of upper gastrointestinal versus sucralfate in the treatment of nonsteroidal anti- bleeding to non-steroidal anti-inflammatory drugs and inflammatory drug induced gastric and duodenal ulcer. Gut aspirin: a case-control study. Gut 1991; 32: 730-4. 1990; 31: Al 175. 9 Somerville K, Faulkner G, Langman M. Non-steroidal anti- 36 Walan A, Bader J P, Classen M, et al. Effect of omeprazole inflammatory drugs and bleeding peptic ulcer. Lancet 1986; and ranitidine on ulcer healing and relapse rates in patients i: 462-4. with benign gastric ulcer. N Engl3J Med 1989; 320: 69-75. 10 Kurata J H, Abbey D E. The effect of chronic aspirin use on 37 Bardhan K D, Griffin W M, Fenn G C, Shield M J, duodenal and gastric ulcer hospitalizations. Clin Gastro- Bjarnason I T. Prevention and reversal of NSAID damage enterol 1990; 12: 260-6. by misoprostol. Gut 1991; 32: A571. 11 Skander M P, Ryan F P. Nonsteroidal anti-inflammatory 38 De Rossi A, Brudenfeldt R. Prevention of NSAID induced drugs and pain free peptic ulceration in the elderly. BMJ GI symptoms and lesions by a low daily dose of misoprostol. 1988; 297: 833-4. Clin Exp Rheumatol 1990; 8 (suppl 4): 59. 12 Chen M C, Amurian D A, Toomey M, Sanders M J, Soll 39 Schimke K, Bolten W. Misoprostol prevents NSAID- A H. Prostanoid'inhibition ofcanine parietal cells: mediation induced gastrointestinal symptoms and lesions. A double- by the inhibitory guanosine triphosphate binding of blind placebo controlled studyin patients with osteoarthritis. adenylate cyclase. Gastroenterology 1988; 94: 1121-9. Clin Exp Rheumatol 1990; 8 (suppl 4): 59. 13 Isenberg J I, Smedforth B, Johannsson C. Effect of graded 40 Saggioro A, Vaiami G. Misoprostol in the prevention of doses of intraluminal hydrogen ions. Prostaglandin E2 and NSAID associated gastrointestinal lesions and symptoms: inhibition of endogenous prostaglandin synthesis on results of a double-blind placebo controlled study con- proximal duodenal bicarbonate secretion in unanaesthetized ducted by an Italian multicentre study group. Endoscopy rats. Gastroenterology 1985; 88: 303-7. 1988; 20 (suppl 2): 83. 14 Lanza F L, Royer G L, Nelson R S. Endoscopic evaluation of 41 Graham D Y, Stromatt S C, Jaszewski, et al. Prevention of the effects of aspirin, buffered aspirin and enteric coated duodenal ulcer in arthritics who are chronic NSAID users. aspirin on gastric and duodenal mucosa. N Engl Med A multicentric trial of the role of misoprostol. Gastro- 1980; 303: 136-8. enterology 1991; 100: A75. Misoprostol in theprevention ofgastroduodenal damage in rheumatology 1093 Ann Rheum Dis: first published as 10.1136/ard.51.9.1089 on 1 September 1992. Downloaded from 42 Agrawal N M, Roth S, Graham D Y, et al. Misoprostol study of the elt-cacy and safety of misoprostol and compared with sucralfate in the prevention of nonsteroidal ranitidine in the prevention of NSAID induced gastric anti-inflammatory drug-induced gastric ulcer. Ann Intern ulcers and upper GI symptoms: preliminary findings. Med 1991; 115: 195-200. Digestion 1991; 49 (suppl 1): 50-1. 43 Kimmey M B, Silverstein F E. Role of H2 receptor blockers 48 Oddsson E, Gudjonsson -, Thiodleifsson B. Protective effect in the prevention of gastric injury resulting from non- of omeprazole or ranitidine against naproxen induced steroidal anti-inflammatory agents. Am Med 1988; 84 damage to the human gastroduodenal mucosa. World (suppl 2A): 49-52. Congress of Gastroenterology 1990: FP22. 44 Lanza F L, Graham D Y, Davis R E, Rack M F. Endoscopic 49 Doube A, Davies J, Notarianni L, Holgate K, Fenn G C. comparison of cimetidine and sucralfate for prevention of Effect of misoprostol on concentrations of prostaglandins in naproxen-induced acute gastroduodenal injury. Effect of synovial fluid. Ann Rheum Dis 1991; 50: 797-9. scoring method. Dig Dis Sci 1990; 35: 1494-9. 50 Piper J M, Ray W A, Daugherty J R, Griffin M R. Corti- 45 Ehsanullah R S B, Page M C, Tildesley G, Wood J R. costeroid use and peptic ulcer disease: role of nonsteroidal Prevention of gastroduodenal damage induced by non- anti-inflammatory drugs. Ann Intern Med 1991; 114: steroidal anti-inflammatory drugs: controlled trial of 735-40. ranitidine. BMJ 1988; 297: 1017-21. 51 Bjarnason I, Smethurst P, Fenn C G, Lee C E, Menzies I S, 46 Lanza F L, AspinallR L, Swabb E A, Davis R E, Rack M F, Levi A J. Misoprostol reduces indomethacin-induced Rubin A. Double-blind placebo-controlled, endoscopic changes in human small intestinal permeability. Dig Dis Sci comparison of the mucosal protective effects of misoprostol 1989; 34: 407-11. versus cimetidine on tolmetin-induced mucosal injury to 52 Morris A J, Madhok R J, Capell H A, Sturrock R D, the stomach and duodenum. Gastroenterology 1988; 95: MacKenzie J F. The effect of misoprostol on enteroscopy- 289-94. diagnosed NSAID small bowel enteroscopy-a retrospective 47 Raskin J, White R, Jaszewski R. Double-blind comparative study. Gut. In press. http://ard.bmj.com/ on September 25, 2021 by guest. Protected copyright.