514
`
`Gut, 1990,31,514-517
`
`Abolition by omeprazole of aspirin induced gastric
`mucosal injury in man
`
`T K Daneshmend, A G Stein, N K Bhaskar, C J Hawkey
`
`Abstract
`This study investigates whether aspirin injury
`to the human gastric mucosa can be prevented
`by profound acid suppression with omepra(cid:173)
`zole, in a randomised, double blind, crossover
`design according to latin square. It was con(cid:173)
`cluded that profound acid suppression can
`prevent aspirin induced gastric mucosal injury
`in normal subjects. This approach may pre(cid:173)
`vent the development of peptic ulcers and their
`complications in patients taking aspirin and
`other non-steroidal anti-inflammatory drugs.
`
`bd for the last 48 h; (c) omeprazole 20 mg each
`morning for seven days plus aspirin 900 mg bd
`for the last 48 h; and (d) omeprazole 40 mg bd for
`seven days plus aspirin 900 mg bd for the last
`48 h. The last doses of omeprazole and aspirin
`(or the corresponding placebo) were taken at
`07 00 hand 07 30 h- that is, 120 and 90 minutes,
`respectively, before. measurement of gastric
`blood loss. Each of the above regimens was
`separated by a seven day washout period. The
`order of treatments was randomised according to
`a latin square design and the study conducted in
`double blind manner.
`
`Aspirin and non-steroidal anti-inflammatory
`drugs are strongly associated with peptic ulcer
`complications in the elderly in Britain.,_, Assum(cid:173)
`ing that the association may be causative, a
`number of approaches have tried to reduce the
`extent of damage caused by these agents.•
`Although old people appear to be at greatest risk,
`it is difficult to evaluate possible therapeutic
`manoeuvres in this population. We have there(cid:173)
`fore investigated healthy adult volunteers whose
`acute responses appear to reflect those of older
`patients.' Our previous studies show that acid
`inhibition by ranitidine' and famotidine• results
`in a reduction in gastric mucosal damage, as
`quantified by the rate of gastric mucosal bleeding
`and endoscopic appearance, but aspirin induced
`damage was not totally abolished.
`In the present study we have examined the
`hypothesis that gastric acid is necessary for the
`occurrence of aspirin induced gastric damage.
`To achieve virtual gastric anacidity we used
`omeprazole, an irreversible inhibitor of the
`proton pump in gastric parietal cells. 10
`
`Methods
`
`SUBJECTS
`Sixteen healthy non-smoking adults (eight
`men; age range 19-25 years) were studied. They
`took no regular medication, except the contra(cid:173)
`ceptive pill. All had normal biochemical and
`haematological
`values,
`including
`platelet
`count, prothrombin time and activated partial
`thromboplastin time. The study was approved
`by the Nottingham Medical School Ethical
`Committee and subjects gave written informed
`consent.
`
`STUDY DESIGN
`Gastric blood loss was measured in each subject
`on four occasions, at the end of each of the
`following regimens: (a) placebo omeprazole plus
`placebo aspirin, both for seven days; (b) placebo
`omeprazole for seven days plus aspirin 900 mg
`
`SAMPLE COLLECTION
`Gastric blood loss was determined after an
`overnight fast, the final doses being taken at
`07 00 hand 07 30 h on the study day. At 09 00 h
`each subject swallowed a 16 French gauge Salem
`sump orogastric tube. After aspiration of resting
`gastric juice, the stomach was rinsed three times
`with distilled water (not a glucose solution) as
`originally described by Hunt.'' The first of three
`10 minute study periods then commenced. Half
`way during each period phenol red (2 mgin IS ml
`water) was introduced through the orogastric
`tube and dispersed around the stomach. After
`nine minutes, distilled water 100 ml was intro(cid:173)
`duced, dispersed and then aspirated by 10
`minutes. After two more rinses a second 10
`minute study period started, and after two
`further rinses there was a third 10 minute study
`period. The subjects were recumbent on their
`left side to reduce pyloric loss of gastric contents,
`except when liquids were introduced into the
`stomach when a standard series of manoeuvres
`was performed in order to ensure maximal
`contact with the gastric mucosa.
`
`ASSAYS
`The pH of resting gastric JUICe and gastric
`washings was measured immediately after collec(cid:173)
`tion using a glass electrode (Corning). The
`volume of blood in gastric aspirates was quanti(cid:173)
`fied by the peroxidase activity of haemoglobin,
`using the orthotolidine reaction. 11 " Briefly,
`samples in citrate buffer were mixed with
`orthotolidine. The rate of development of a blue
`colour was determined on a spectrophotometer
`at 640 nmol between 30 and 60 sec after addition
`of hydrogen peroxide. This was compared with
`values from a standard curve constructed using
`the subjects own blood. Gastric blood loss was
`expressed as ~tl of blood/10 min period after
`correction for phenol red recovery. The median
`value from the three 10 minute periods was used
`for analysis. Phenol red concentration in gastric
`aspirates obtained during the study periods was
`
`University Department r.~
`Therapeutics, University
`Hospital, Nottingham
`T K Daneshmend
`AGStein
`N KBhaskar
`Cj Hawkey
`Correspondence to:
`Dr T K Daneshmend, Dept of
`Medicine, Royal Devon and
`Exeter Hospi£al, Wonford
`Barrack Road, Exeter
`EXZ SOW.
`Accepted for publication
`10 August 1989
`
`

`
`Abolition by omeprazole of aspirin induced gastric mucosal injury in man
`
`SIS
`
`TABLE Gastric mucosal blood loss, pH of resting juice, pH of gastric washings and phenol red recovery in 16 subjects after
`(a) placebo only, (b) aspirin 900 mg bd only, (c) aspirin 900 mg bd plus omeprazole 20 mg mane, and (d) aspirin 900 mg bd plus
`omeprazole 40 mg bd
`
`Placebo
`
`Aspirin
`
`Aspirin plus
`Aspirin plus
`Omeprazole ZO mg mane Omeprazole 40 mg bd
`
`Gastric musocal blood loss f'VIO min
`geometric mean (95% CI)
`Initial pH median (lower-upper quartiles)
`Gastric aspirate pH median (lower-upper
`quartiles)
`Phenol red recovery (%) mean (SD)
`
`1·4 (0·8-2-4)
`3·12 (2 ·07- 2·93)
`
`16·1 (9·5-27·5)•
`3·12 (2·31-5·19)
`
`3-4 (1 ·4-5·2)t
`6·01 (4 ·25-7·32)§
`
`2·43 (1·94-2 ·53)
`54·6 (11·5)
`
`2· 26 (2·11-2·58)
`57·6(9·7)
`
`3·07 (2·66-6·52)
`63-4(13·1)
`
`2-4 (I· 3-4·4 ):j:
`6·60 (4·91-7·16)§
`
`6·65 (5·59-6·88)§
`61·1 (11·0)
`
`•p<O·Ol compared with placebo; tp<O·Ol compared with aspirin alone; :j:p<O·OOl compared with aspirin plus 0meprazole 20 mg!day,
`and NS compared with placebo; §p<O·Ol compared with placebo.
`
`measured spectrophotometrically at 560 nmol,
`after adjustment of pH to 10·5 with sodium
`hydroxide.
`
`STATISTICAL METHODS
`Analysis of variance was used to assess the
`influence of aspirin and of omeprazole on gastric
`mucosal bleeding. To approximate to a normal
`distribution, these data were logarithmically
`transformed before computation and results
`given as geometric mean with 95% confidence
`limits. Phenol red recovery data were analysed in
`similar manner, but without logarithmic trans(cid:173)
`formation. Results are expressed as mean and
`standard deviation (SD). Friedman two way
`analysis of variance by ranks was used to assess
`the significance of changes in pH values and
`results expressed as medians and interquartile
`ranges.
`
`was slightly higher than placebo, this difference
`was not significant (p=0·07).
`The effects of omeprazole on the pH of resting
`juice or of gastric washings are shown in the
`Table. The median initial pH was unaffected by
`aspirin (p=0·79), but increased significantly
`with omeprazole 20 mg/day and 40 mg bd
`(p=0·035 and p<O·OOl, respectively compared
`with placebo). The median initial pH was not
`significantly different between the two omepra(cid:173)
`zole dose regimens (p=0·4). The median pH of
`gastric washings is also given in the Table. The
`only notable difference was a lower pH with the
`lower omeprazole dose, while pH of washings
`remained at the initial value on the higher
`omeprazole dose, a difference which was signifi(cid:173)
`cant (p<O·Ol). Overall, the reduction in gastric
`mucosal bleeding rate was significantly corre(cid:173)
`lated with the pH of initial gastric aspirates
`(r=0·423, p<0·02).
`
`Results
`The rate of gastric mucosal bleeding after aspirin
`was over 10-fold greater than that after placebo
`(p<O·OOl) (Table and Figure). The value after
`aspirin plus omeprazole 20 mg/day was reduced
`significantly by 79% when compared with that
`after aspirin alone (p<O·Ol). Omeprazole 40 mg
`bd plus aspirin resulted in a gastric mucosal
`bleeding rate that was 85% less than after aspirin
`alone (p=O·OOl), and although the mean value
`
`Discussion
`Endoscopic observations implicating aspirin in
`the pathogenesis of iatrogenic gastric damage
`were made by Douthwaite ~nd Lintott half a
`century ago." The subsequent widespread use
`of aspirin and other non-steroidal anti(cid:173)
`inflammatory drugs has been implicated in
`peptic ulcer perforation rates
`in Britain.'
`Emergency admission because of bleeding from
`gastric and duodenal ulcers in the elderly is
`
`100
`
`90
`
`80
`
`70
`
`"
`~ 60
`'5. 50
`"
`0 40
`'C
`0
`
`.2 30
`
`20
`
`10
`
`Placebo
`
`Aspirin
`
`Aspirin 0 me 20 mg
`om
`Figure: Individual rates of gastric mucosal blood loss (flll 10 min) in 16 normal adults on placebo, aspirin 900 mg bd only,
`aspirin 900 mg bd plus omeprazole 20 mg each morning, and aspirin 900 mg bd plus omeprazole 40 mg bd.
`
`Aspiri~~me 40 mg
`
`

`
`516
`
`Daneshmend, Stein, Bhaskar, Hawkey
`
`associated with aspirin and other non-steroidal
`anti-inflammatory drug use in our hospital popu(cid:173)
`lation. ' ·' " Aspirin ingestion is associated with a
`relative risk of three, even for short periods of
`exposure.'" Moreover, aspirin can provoke
`gastric mucosal bleeding at doses of up to 75 mg
`taken daily for five days or less." It is evident that
`aspirin is probably responsible for a spectrum of
`damage, ranging from acute gastric erosions to
`peptic ulcer complications.
`In our present study aspirin induced gastric
`mucosal damage (as quantified by gastric mucosal
`blood loss) was abolished by omeprazole 40 mg
`bd, a dose that produces virtual anacidity. 16
`Quantification of gastric mucosal injury by
`measurement of gastric mucosal blood loss
`closely reflects direct endoscopic evidence of
`mucosal damage: we have previously shown
`gastric mucosal blood loss to correlate with the
`extent of petechial haemorrhage seen endo(cid:173)
`scopically! Our findings accord with a small
`endoscopic study recently reponed in abstract
`form which showed prevention by omeprazole of
`gastric injury after a single aspirin dose."
`The dissociation constant (pKa) of aspirin
`is 3·5. Thus at the levels of intragastric pH
`achieved with omeprazole in our study, aspirin
`ionisation is virtually complete. In this form
`passive absorption of aspirin into the gastric
`mucosa does not occur." In contrast, at normal
`is almost entirely
`intragastric pH aspirin
`unionised and able to diffuse passively into cells
`of the gastric epithelium where a neutral pH
`results in reionisation and intracellular trapping
`of salicylate in high concentrations. The conse(cid:173)
`quent topical toxicity of salicylates is well recog(cid:173)
`nised and results in impaired barrier function,
`reduced mucus and bicarbonate secretion, and
`capillary injury.'"" The underlying metabolic
`changes are not firmly established, but in the
`presence of acid aspirin may achieve intracellular
`levels sufficient to uncouple oxidative phos(cid:173)
`phorylation or interfere with carbohydrate
`metabolism." As most other non-steroidal anti(cid:173)
`inflammatory drugs are weak acids, similar con(cid:173)
`siderations are likely to apply although direct
`evidence is lacking.
`Apart from these specific benefits, acid inhibi(cid:173)
`tion may result in other non-specific advantages.
`Gastric acid enhances mucosal injury caused by a
`variety of stimuli" and damages the basal lamina
`resulting in impaired epithelial restitution.13 In
`addition, the activity of pepsin is pH dependent
`and is inhibited at high pH. Whatever the
`mechanism, the observations presented here
`strongly support the hypothesis that gastric acid
`is crucial in the genesis of aspirin- (and possibly
`other non-steroidal ami-inflammatory drug-)
`related gastroduodenal injury. Our data show
`that the reduction in gastric mucosal bleeding
`rate bears a close relationship to the intragastric
`pH achieved with omeprazole. In this context, it
`is notable that patients with pernicious anaemia
`and resultant achlorhydria are resistant to aspirin
`injury compared with healthy controls." Even
`with achlorhydria, however, there was slight
`injury, possibly because aspirin itself acted as a
`source of exogenous acid."
`It is possible, however, that oral omeprazole
`protects the gastric mucosa by additional acid
`
`independent mechanisms. In animals, omepra(cid:173)
`zole given by the oral route is much more
`effective than when given parenterally in pre(cid:173)
`venting aspirin induced gastric damage despite
`complete inhibition of gastric acid." " In addi(cid:173)
`tion, oral omeprazole protects against ethanol
`induced gastric damage if given between 15 to 60
`minutes before ethanol, there being no effect
`evident 3·5 h after the dose. This protective
`effect of omeprazole is not mediated through
`gastric mucosal prostaglandins," changes in
`gastric mucosal blood flow," or alterations in
`gastric mucosal bicarbonate secretion.19 It may
`be the result of a direct effect of omeprazole on
`the vascular endothelium," as omeprazole also
`protects human gastric epithelial cells in vitro
`from indomethacin induced damage."
`From these and previous data it is clear that
`aspirin remains injurious to the human gastric
`mucosa until pH values approaching neutrality
`are reached. Greater suppression of acid may be
`needed to protect the gastroduodenal mucosa
`against aspirin and other non-steroidal anti(cid:173)
`inflammatory drugs than is adequate for ulcer
`healing. Profound suppression of acid, however,
`may not be without risk. An increased incidence
`of enteric infection is recognised but this is
`nonetheless relatively uncommon. At neutral pH
`(however achieved) serum gastrin concentra(cid:173)
`tions are raised." Whether there is a real risk of
`gastric carcinoid tumours in man given omepra(cid:173)
`zole is much less clear. There is no evidence in
`man that the small rises in plasma gastrin seen
`with gastric antisecretory drugs" produce
`any sustained hyperplastic change in entero(cid:173)
`chromaffin like cells." For the frail elderly
`patient at high risk of developing aspirin or other
`non-steroidal ami-inflammatory drug associated
`ulceration, bleeding, perforation or death the
`benefits of profound acid suppression with
`omeprazole are likely to outweigh the risks of
`enterochromaffin like cell carcinoid tumour
`development. Although our study showed a
`statistically significant advantage of omeprazole
`40 mg bd over omeprazole 20 mg each morning
`in reducing gastric mucosal bleeding, the higher
`dose may not necessarily confer a clinical advan(cid:173)
`tage. A clinical trial of this approach, using
`omeprazole 20 mg once per day, would therefore
`be justified in patients at risk of gastric damage
`from aspirin and non-steroidal anti-inlammatory
`drugs. Recent prospective studies in patients
`have shown that ranitidine" and misoprostol"
`attenuate the damaging effects of non-steroidal
`ami-inflammatory drugs on the upper gastro(cid:173)
`intestinal tract. Ranitidine, however, has a pre(cid:173)
`ferential protective effect on the duodenum but
`did not protect against drug induced gastric
`ulceration," while the converse was true of
`rnisoprostol." In this context, it is possible that
`omeprazole may possess an advantage (as yet
`untested) because of its more potent acid inhibit(cid:173)
`ing action and be capable of attenuating non(cid:173)
`steroidal
`anti-inflammatory drug
`induced
`damage in both stomach and duodenum.
`
`These data were presented to the British Society of Gastro(cid:173)
`emerology Autumn Meeting: Sheffield September 14-16, 1988.
`We thank Dr Keith Gillon and Astra Clinical Research Unit,
`Edinburgh, for supply of study medication and financial support.
`
`

`
`Abolition by omeprazole of aspirin induced gastric mucosal injury in man
`
`517
`
`I Coggan D, Langman MJS, Spiegelhalter D. Aspirin, para(cid:173)
`cetamol and haematemesis and melaena. Gut 1982; 23:
`340-3.
`2 Collier OS, Pain JA. Non-steroidal antiinflammatory drugs
`and peptic ulcer perforation. Gut 1985; 26: 359-63.
`3 Somerville KW, Faulkner G, Langman MJS. Non-steroidal
`anti-inflammatory drugs and bleeding peptic uker. Lancet
`1986; i: 462-4.
`4 Wah RP, Katchinski B, Faulkner G, Logan RFA, Langman
`MJS. Non-steroidal anti-inflammatory drugs and bleeding
`peptic ulcer. Lancer 1986; i: 489-92.
`5 Armstrong CP, Blower AL. Peptic ulcer complications and
`non-steroidal antiinflammatory drugs [Abstract]. Gut 1986;
`27: A609.
`6 Hawkey CJ, Prichard PJ , Somerville KH. Strategies for
`preventing aspirin-induced gastric bleeding. Scand J
`Gastroenrero/ 1986; 21 [suppll25]: 170-3.
`7 johnson PC. Gastrointestinal consequences of treatment
`with drugs in elderly patients. Am Geriatr Soc 1982; 30:
`S52-5-57.
`8 Kitchingman GK, Prichard PJ , Daneshmend TK, Wah RP,
`Hawkey CJ. Human gastric mucosal bleeding induced by
`aspirin 300 mg and its prevention by ranitidine [Abstract].
`Guti988;29:A711.
`9 Daneshmend TK, Prichard PJ, Bhaskar NK, Millns PJ,
`Hawkey CJ. The use of microbleeding and an ultra-thin
`endoscope
`to assess gastric mucosal protection by
`famotidine. Gastroenterology. 1989; 97:944-9.
`10 Clissold SP, Campoli-Richards OM. Omeprazole: a pre(cid:173)
`liminary review of its pharmacodynamic and pharmaco(cid:173)
`kinetic properties, and therapeutic potential in peptic ulcer
`disease and Zollinger-EIIison syndrome. Drugs 1986; 32:
`15-47.
`•
`II Hunt JN. A procedure for measuring gastric bleeding caused
`by drugs. Dig Dis Sci 1977; 24: 525-8.
`12 Hawkey CJ, Simpson G, Somerville KW. Reduction by
`enprostil of aspirin-induced blood loss from human gastric
`mucosa. Am] Med 1986; 81 [suppi2A]: 50-3.
`13 Douthwaitc AH, Linton GAM . Gastroscopic observations of
`the effects of aspirin and certain other substances on the
`stomach. Lancet1938; ii: 1222-5.
`14 Prichard PJ, Somerville KW, Faulkner G, Langman MJS
`[Abstract]. Gutl987; 28: Al401.
`IS Prichard PJ, Kitchingman GK, Hawkey CJ. Gastric mucosal
`bleeding: what dose of aspirin is safe? [Abstract]. Gutl987;
`28: Al401.
`16 Hetzel DJ, Shearman DJC. Omeprazole inhibition of noctur(cid:173)
`nal gastric secretion in patients with duodenal ulcer. Br J
`Clin Plrarmacol1984; 18: 587-90.
`17 Bigard M-A, Isal J-P. Complete prevention by omeprazole
`of aspirin-induced gastric lesions
`in healthy subjects
`[Abstract]. Gutl988; 29: A712.
`18 Flower RJ, Moncada S, Vane JR. Analgesic-antipyretics and
`anti-inflammatory agents: drugs employed in the treatment
`of gout. In: GilmanAG, Goodman LS, Rail TW, Murad F,
`eds. The PharmacologiCal basis of therapeutics. 7th ed. New
`York: Macmillan, 1985: 674-715.
`
`19 Davenport HW. Gastric mucosal haemorrhage in dogs. Effects
`of acid aspirin and alcohol. Gastroenterology 1969; 56: 439-
`49.
`20 Hawkey CJ, Rampton DS. Prostaglandins and the gastro(cid:173)
`intestinal mucosa: are they important in its function, disease
`or treatment? Gastroenterology 1985; 89: 1161- 88.
`21 Spenney JG, Marshall G. Uptake and intracellular concentra(cid:173)
`tion of salicylate and p-hydroxybenzoate in in-vitro fundic
`gastric mucosa. Gastroenterology 1883; 84: 1318.
`22 Morris GP, Wallace JL. The roles of ethanol and of acid in the
`production of gastric mucosal erosions in rats. Virchows Arch
`(Cell Pathol) 1981; 32: H-8.
`23 Black BA, MorrisGP, Wallace JL. Effects of acid on the basal
`lamina of the rat stomach and duodenum. Virchows Arch
`(Cell Patho/) 1985; SO: 109-18.
`24 StJohn DJB, McDermott FT. Influence of achlorhydria on
`aspirin-induced occult gastrointestinal bleed loss: studies in
`pernicious anaemia. Br MedJ 1970; 2: 450-2 .
`25 Mausson H, Andersson K, Larsson H. Omeprazole provides
`protection against experimentally induced gastric mucosal
`lesions. Eur J Plrarmacol1983; 91: 111-4.
`26 Kollberg B, Isenberg JI, Johansson C. Oral omeprazole
`protects gastric mucosa against ethanol. Gastroenterology
`1983; 84: 1212.
`27 Komurek SJ , Brzozowski T, Radecki T. Protective action of
`omeprazole, a benzimidazole derivative, on gastric mucosal
`damage by aspirin and ethanol in rats. Digestion 1983; 27:
`159-64.
`28 Mansson H, Larsson H. Effect of omeprazole on gastric
`mucosal blood flow in the rat. Scand] Gastroenterol 1986; 21
`[supplll8]: 72-4.
`29 Mattson H, Carlsson K, Carlsson E. Omeprazole is devoid of
`effect on alkaline secretion in isolated guinea pig antral
`mucosa. In: Allen, Flemstrom, Garner, Sileo, Turnbrg, eds.
`Mechanisms of mucosal protection in tlu upper gastrm.ntestinal
`traer. New York: Raven Press, 1984: 141-6.
`30 Mattsson H. Protective effects of omeprazole in gastric
`mucosa. Scand] Gastroentero/1986; 2l(supplll8]: 86-8.
`31 Romano M, Razandi M, lvey KJ. Cimetidine and omeprazole
`directly protect human gastric epithelial cells in viuo.
`Gastroenterology 1987; 92: 1599.
`32 Lanzon-Miller S, Pounder RE, Hamilton MR, et al. Twenty(cid:173)
`four-hour intragastric acidity and plasma gastrin concentra(cid:173)
`tion in healthy subjects and patients with duodenal or gastric
`ulcer or pernicious anaemia. Aliment Pharmacol Ther 1987;
`1:225-37.
`33 Lamberts R, Creutzfeldt W, Stockmann F, Jacubaschke U,
`Mass S, Brunner G. Long-term omeprazole treatment in
`man: effects on gastric endocrine cell populations. Digestion
`1988; 39: 126-35.
`34 Ehsanullah RSB, Page MC, TildesleyG, Wood JR. Prevention
`of gastroduodenal damage induced by non-steroidal anti(cid:173)
`inflammatory drugs: controlled trial of ranitidine. Br MedJ
`1988; 297: 1017-21.
`35 Graham DY, Agrawal NM, Roth SH. Prevention of NSAID·
`induced gastric ulcer with misoprostol: multicentre, double·
`blind, placebo-controlled trial. Lancet1988; ii: 1277- 80.