Gut, 1992, 33, 164-169

Mucosal adaptation to indomethacin induced gastricdamage in man - studies on morphology, blood flow,and prostaglandin E2 metabolism

C J Shorrock, W DW Rees

AbstractThe effect of 28 days' continuous administra-tion of oral indomethacin on gastroduodenalmorphology, gastric mucosal blood flow, andgastric mucosal prostaglandin E2 (PGE2) meta-bolism in man was studied to define further themechanisms of mucosal injury induced byindomethacin. Indomethacin caused acutegastroduodenal damage in ali cases, which wasmaximal at 24 hours of administration. Withcontinued intake, mucosal adaptation occursresulting in resolution of endoscopic mucosaldamage. At the time of maximal mucosaldamage, gastric mucosal blood flow was

significantly reduced compared with valuesbefore treatment (p<0001 in fundus andp<0002 in antrum), with good correlationbetween the severity ofdamage and the magni-tude of the reduction in blood flow (r=076).Mucosal recovery was associated with a returnof the blood flow to normal. PGE2 in mucosalhomogenate was significantly reduced by indo-methacin in both the hfndus (p<0O01) andantrum (p<001) after 24 hours but there wasno correlation between the magnitude of thisreduction and the severity of mucosal damage(r=-034). Despite mucosal recovery by 28days, PGE2 values remained significantlybelow those before treatment in both thefundus (p<0.01) and antrum (p<0.01). ThePGE2 degredation capacity was not influencedby indomethacin. In conclusion, mucosaladaptation to acute damage by indomethacinoccurs in man and seems independent of localPGE2 metabolism.

University Department ofMedicine, QueenElizabeth Hospital,BirminghamC J Shorrock

Department ofGastroenterology, HopeHospital, SalfordWDW ReesCorrespondence to:Dr C J Shorrock, UniversityDepartment of Medicine,Queen Elizabeth Hospital,Edgbaston, BirminghamB15 2TH.Accepted for publication15 April 1991

It is well known that non-steroidal anti-inflammatory drugs (NSAIDs), extensively usedas simple analgesics and in rheumatic diseases,cause gastrointestinal injury, particularly in thestomach and proximal small bowel. 1-' Thedamage produced can result in serious complica-tions such as ulceration, bleeding, and perfora-tion, which can be life threatening.'How NSAIDs cause gastroduodenal damage

is unknown, although the depletion of mucosalprostaglandins, which reduces the competenceof the mucosal defence mechanisms, is thoughtto be of major importance.7 Support for thismechanism is provided by the findings thatNSAIDs reduce many apsects of mucosaldefence such as epithelial HCO3 secretion,'0mucus synthesis and secretion," 12 and surfacehydrophobicity'3 - effects largely prevented bytreatment beforehand with prostaglandin. Inaddition to their effects on mucosal defence,NSAIDs also reduce the integrity of the 'mucus

cap', produced in response to superficial injury,which is important in providing a suitableenvironment for epithelial repair to take place.'4The effects ofNSAIDs on mucosal blood flow,

a vital component ofmucosal defence and repair,remain controversial. Ashley et all' have shownthat asprin causes a reduction in gastric mucosalblood flow at the sites ofdamage, with augmenta-tion of blood flow elsewhere, and an overallincrease in gastric blood flow. Main andWhittle'6 observed reduced basal gastric mucosalblood flow in rats after doses of indomethacinthat inhibited prostaglandin formation, whileKaufman'7 found a decrease in basal gastricmucosal blood flow in conscious dogs givenindomethacin. Studies in man by Konturek'8have suggested that indomethacin reduces basalgastric mucosal blood flow. These and otherfindings suggest that endogenous prostaglandinscontribute to the maintenance of basal gastricmucosal blood flow in animals and man. 19The incidence of serious side effects with

NSAIDs, however, is low considering the largenumbers of these drugs prescribed.20 Since endo-scopic studies have shown acute gastric mucosaldamage in most subjects during the first week ofNSAID administration,' 21 these rather con-tradictory observations suggest that tolerancedevelops during the course of continued NSAIDintake. This tolerance or adaptation of the gastricmucosa to a variety of damaging agents has beenwell documented in animal studies,22 and morerecently with aspirin23 and indomethacin24 inman. The mechanisms whereby the mucosadevelops tolerance to damage, however, remainuncertain.The aims of our studies were thus threefold:

firstly, to document the morphological changesoccuring in gastroduodenal mucosa during28 days of treatment with the NSAID indo-methacin; secondly, to measure mucosal bloodflow in the gastric mucosa during this period;and thirdly, to measure gastric mucosal prosta-glandin E2 (PGE2) metabolism during this 28 dayperiod.

Methods

SUBJECTSStudies were carried out on 14 healthy volunteers(eight men and six women) with a mean age of 23years (range 19-26). Five volunteers weresmokers, smoking less than 20 cigarettes daily,and they did not change their smoking habitsduring the study. Alcohol was allowed duringthe study and subjects followed their normaldrinking habits (all subjects consumed less than

164

on October 7, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.33.2.164 on 1 F

ebruary 1992. Dow

nloaded from

Mucosal adaptation to indomethacin induced gastric damage in man - studies on morphology, bloodflow, andprostaglandin E2 metabolism

80 g of alcohol per week). No subject had aprevious history of gastrointestinal disease orhad taken aspirin or any other NSAID in theprevious three months.

All subjects gave written informed consent tothe studies and ethical approval was given bythe Salford Area Health Authority EthicalCommittee.

ENDOSCOPYStandard upper gastrointestinal endoscopy wasperformed by one investigator throughout(CJS). A 1% lignocaine hydrochloride throatspray was applied before introduction of theendoscope and no sedation was used. All endo-scopies were performed between 11 am and 12mid-day, subjects having had a very light break-fast at 7 am on the morning of endoscopy.

STUDY DESIGNSubjects underwent endoscopy at entry to thestudy. Oral indomethacin (50 mg three timesdaily) was then taken for 28 days continuously.The indomethacin was taken with meals andwith a light breakfast at 7 am on the morning ofendoscopy. Endoscopy was repeated at 24 hours,seven days, and 28 days while on the indo-methacin and was performed between 11 am and12 noon. At each endoscopy, mucosal damagewas graded and mucosal blood flow measured inthe fundus and antrum of the stomach using alaser Doppler technique. At the end of eachendoscopy three biopsy specimens were takenfrom normal looking areas of both the antrumand fundus of the stomach. These specimenswere immediately placed in polyethylene vials,frozen in liquid nitrogen, and stored at -700Cfor later measurement ofPGE2 metabolism.

Before each endoscopy, subjects werequestioned directly about gastrointestinal symp-toms and in addition kept a detailed writtenrecord of the duration, time of day, and severityof any such symptoms between endoscopies.Ten ml of blood were taken randomly from eachsubject during the study for later detection ofindomethacin to confirm compliance.

Three further subjects were used as controls.These subjects underwent endoscopy withmucosal grading, measurement ofmucosal bloodflow, and PGE2 metabolism as above over a 28day period but received empty indomethacincapsules. The endoscopist was blind to thecapsule content.

MORPHOLOGYAt endoscopy, gastric and duodenal integrity

TABLE I Variation in mucosal bloodflow measured by laserDopplerfowmetry in the antrum ofstomach in six subjectsmeasured over 10 seconds

Subject Bloodflow range at Mean (range) bloodflowno stngle point (n=3) (n=6, arbitrary units)

1 39-42 40(36-44)2 46-50 46 (44-48)3 41-43 39(37-42)4 38-43 42(40-44)5 46-49 51 (48-52)6 52-54 50(46-53)

was graded and recorded according to astandardised scoring system as previously des-cribed.24 The entire stomach and duodenal capwere examined in proximal to distal mannerbefore measurement of blood flow in order toeliminate any errors that might have been causedby a misinterpretation of artifacts from either theendoscope or the laser Doppler probe.

MEASUREMENT OF MUCOSAL BLOOD FLOWGastric mucosal blood flow was measured usingthe technique of laser Doppler flowmetry(LDF). The operating principal of LDF hasbeen fully described elsewhere25 but briefly isbased on the principal that light scattered bymoving red blood cells undergoes a shift in itsfrequency, the mean Doppler shift providing anestimate of blood flow. Mucosal blood flowmeasured by LDF correlates well with flowmeasured by other methods.26 27

In the laser Doppler flowmeter used in thepresent study (Periflux PF2, Perimed Ltd,Stockholm, Sweden) light from 2 mW He-Nelaser is transmitted down an optical fibre(diameter 0-7 mm) contained in a PF109 endo-scopic probe (Perimed Ltd, Sweden) of outsidediameter 2-5 mm. The probe is inserted downthe biopsy channel of an Olympus Q10 endo-scope (Keymed, Southend-on-sea, UK). Itbecame apparent in early experiments that thedistance between the probe tip and the end of theendoscope was critical as the magnitude of thelaser Doppler readings varied according to this.To obviate this problem the probe was marked insuch a way that it always protruded 3 cm fromthe end ofthe endoscope. After diffuse scatteringof the incident laser light, a portion of thebackscattered light is picked up by two furtherfibres (diameter 07 mm) contained in the endo-scopic probe and transmitted to the flowmeter.The signal is processed giving a low noise outputwhich corresponds to the tissue blood flowbeneath the probe. A continuous recording ofblood flow is obtained using a chart recorder witha paper speed of 1 mm/second. Throughout thestudies the upper limit of the signal processorwas set to 12 kHz with a constant gain of threeand a time constant of 15 seconds.Measurements of blood flow were made under

direct vision in the gastric fundus and antrum forareas that looked endoscopically normal, the tipof the probe being gently abutted against themucosa by advancing the whole endoscope andnot the probe alone. Particular care was takennot to dimple the mucosa. A valid reading wasone where a constant reading was obtained for atleast five seconds uninterrupted by any move-ment artifact of the probe head relative to thetissue or by any loss of optical coupling due toperistalsis. A minimum of three such readingswas taken at each site and the mean was calcu-lated and recorded.

Studies on the reproducibility of the laserDoppler technique were performed on six sub-jects attending for routine endoscopy. Thesesubjects had blood flow measured at six points inthe antrum and six in the fundus of the stomachover a 10 second period as described above. Inaddition, blood flow was measured at one par-

165

on October 7, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.33.2.164 on 1 F

ebruary 1992. Dow

nloaded from

Shorrock, Rees

ticular point over 10 seconds on three occasions,withdrawing the probe from the mucosa betweenmeasurements. Results are given in Table I andshow the good reproducibility of the techniquein each subject.

PGE2 METABOLISMBiopsy specimens were processed and PGE2content and degredation rate in mucosal homo-genates (calculated as the rate of formation of13,14-dihydro-15-keto-PGE2 from radiolabelledPGE2) were measured as previously described.28Briefly, biopsy specimens were homogenisedand filtered through a nylon mesh. The proteincontent of the homogenate was measured by themethod of Lowry et al29 and ranged between 04and 0-4 mg/ml. For the measurement of PGE2, a200 ,ul portion of the homogenate was immedi-ately added to 3 mol/l citric acid to prevent anypossible enzymatic conversion. To this wasadded 800 p1 of standardised buffer containing3H-PGE2 (7000 dpm) as a recovery marker. Thiswas followed by immediate extraction into 3 mldiethyl ether. Degredation was determined byincubating 200 ,ul of homogenate at 37°C with800 R1 of buffer containing 5 [tmol/l PGE2, 0418[tCi 3H-PGE2, and 2 5 mmol/l NAD+ as acofactor. Assays were stopped after 30 minuteswith 3 mol/l citric acid and were immediatelyextracted into 3 ml diethyl ether. Blanks weresimultaneously prepared in which citric acid wasadded before the tissue homogenate. All assaysand blanks were prepared in duplicate.

Extraction of PGE2 and the degredation pro-duct 13,14-dihydro-15-keto-PGE2 was by thinlayer chromatography. For each assay, 200 p1 ofthe ether phase were plated, dried, and run in theorganic phase of ethyl acetate/iso-octane/aceticacid/water (110:50:20:100). For assay ofdegredation, the region corresponding to 13,14-dihydro-15-keto-PGE2 was scraped off andcounted in phase contrast scintillation fluid. Theregion corresponding to PGE2 was similarlyscraped into glass vials and the silica washed toelute the PGE2, which was measured by radio-immunoassay (Dupont PGE2 [1251] RIA kit).Recovery of PGE2 was over 80%. Degredationrates were calculated as the difference betweenthe timed assay and the blank value. Results areexpressed as pmol PGE2/mg protein/minute.

STATISTICAL ANALYSISResults are calculated as mean (SEM) and differ-ences compared statistically using a Student's ttest for paired data.

Results

MORPHOLOGY (FIG 1)After 24 hours' indomethacin, all subjects hadevidence of mucosal damage that was similar inseverity in both the stomach and the duodenum(mean scores: 1-5 (0 25) in the stomach and 1 0(0 3) in the duodenum, n= 14, p=0 10 for differ-ence between gastric and duodenal damage(unpaired t test)). By day 7 the damage hadresolved significantly (mean scores: 0 7 (0.22) in

the stomach and 0-57 (0 2) in the duodenum,n=14, p<0 01 for improvement in gastricdamage and p<0 05 for improvement in duo-denal damage between 24 hours and day 7). Twosubjects with persistent headache and nauseawere withdrawn from the study. By day 28, in allbut two subjects who had persistent grade 2gastric or duodenal damage, macroscopicmucosal damage had resolved (n= 12). Nosubject experienced gastrointestinal symptomsduring the study and indomethacin was detectedin the serum of all subjects. At no time was thereany mucosal damage in the three control sub-jects. There was no relation between the endo-scopic severity of gastric damage and alcoholintake.

BLOOD FLOW (FIG 2)Fundal mucosal blood flow was significantlyhigher than that in the antrum (p<0 001).Gastric mucosal blood flow was significantly

Time 015

12

9

6

3

0

15

12

9

6U,

30

0

a)-0 15EZ 12

9

6

3

0

15

12

9

6

3

0

1 day

7 days

28 days

0 1 2 3 4

Endoscopic gradingFigure 1: The effect oforal indomethacin(50 mg three times daily) on gastric (open)and duodenal (hatched) morphology over 28days ofcontinuous administration in 14healthy volunteers (n= 12 at 28 days).

166

n on October 7, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.33.2.164 on 1 F

ebruary 1992. Dow

nloaded from

Mucosal adaptation to indomethacin induced gastric damage in man - studies on morphology, bloodflow, and prostaglandin E2 metabolism

20

TimeO0 day 7 days 28 days*.; --< .Time of endqe.ppy

Figure 2: Gastric mucosal bloodflow measured in fundus(hatched) and antrum (open) during 28 days' continuousadministration ofindomethacin (50 mg three times daily) in 14healthy volunteers (n= 12 at 28 days). *p<005, **p<0.01compared with baseline bloodflow.

reduced in the fundus (p<0-001) and antrum(p<0 002) of the stomach after 24 hours' indo-methacin intake - the time of maximal mucosaldamage. By day 7 blood flow had increased: itwas still significantly less than the value beforetreatment in the fundus (p<002) but not in theantrum (p=008). By day 28, the time of endo-scopic mucosal recovery, blood flow was notsignificantly different to that before treatment.There was no change in blood flow in the threecontrol subjects over the 28 days of the study.To determine the correlation between the

severity of mucosal damage and the change inblood flow, it was necessary to regrade theendoscopic damage. This was done by using theoriginal grading criteria but grading damageseparately for fundus and antrum on a scale of 0(normal) to 4 (severe damage±ulceration). Thisis because the original grading system did notdifferentiate between antral and fundal damage.There is a good correlation (r=0 76) between theseverity of mucosal damage and the percentagechange in mucosal blood flow after 24 hours ofindomethacin intake, the time of maximalmucosal damage (Fig 3).

PGE2 METABOLISM (FIG 4)PGE2 in mucosal homogenate was measured in10 of our volunteers. Total PGE2 content wassignificantly higher in the antrum than in thefundus ofthe stomach (antrum, mean (SEM): 83(21) and fundus: 36 (11) pmol/mg of protein,n= 10, p<0 05 (unpaired t)). After 24 hours'indomethacin, PGE2 content was significantlyreduced from values before treatment in both thefundus and antrum (antrum: 30 (9) and fundus:21 (7) pmol/mg of protein, n=10, p<0-01 forreduction in both fundus and antrum). By day 7,when endoscopic damage was improving, PGE2content remained significantly reduced com-pared with values before treatment (mean (SEM)43 (13) in antrum and 24 (8) in fundus, n= 10,p<001 for both fundus and antrum). In spite ofmucosal recovery by day 28, PGE2 contentremained significantly reduced (36 (16) in theantrum and 22 (12) in the fundus, n= 10, p<0 01for difference between fundal and antral valuesbetween values before treatment and 28 days).There was no correlation between the fall inPGE2 content and the severity of the mucosaldamage after 24 hours' indomethacin, the time ofmaximal damage (r= -0 34). In the three control

0

V-o00

CU-ca)

cJ

01 2 3 4

-20 1

-40 i

-60 1 Endoscopic grading

Figure 3: Correlation between severity ofmucosaldamage infundus and antrum ofstomach and %change in mucosal bloodflow after one day ofindomethacin treatment (50 mg three times daily) in 14healthy volunteers. r=0- 76.

subjects, there was no change in the PGE2content of mucosal homogenates during thestudy period.During the 28 days of the study there was no

significant change in mucosal PGE2 degredationcapacity (Table II).

DiscussionOur results confirm previous findings24 thatindomethacin (50 mg three times daily) producesacute gastroduodenal damage in all subjects andthat with continued administration endoscopicdamage resolves. This adaptation to damageappears at day 7 and is complete by day 28,although not all subjects exhibit completeresolution of injury. Two of our volunteers hadpersistent grade 2 mucosal damage at day 28,although in both of these it was less than thatafter 24 hours of drug administration. It isunclear why mucosal damage failed to resolve inthese two subjects, both were non-smokers,drank less than 40 g of alcohol per day, and hadno other features to differentiate them fromother volunteers.

Adaptation includes healing of the acuteinjury and enhancement of mucosal defence toprevent further damage during continuedexposure to the damaging agent and has beenwell described in a variety of animal models andto a range of injurious agents.22 This phenome-non was initially attributed to increased genera-tion of 'protective' mucosal prostaglandins,22although recently this has been questioned.30

Figure 4: Prostaglandin E2 (PGE2) content of biopsyspecimen homogenatefrom thefundus (hatched) and antrum(open) ofthe stomach during 28 days' continuousadministration ofindomethacin (50 mg three times daily) in 10healthy volunteers. *p<0 01 compared with baseline values.

167

on October 7, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.33.2.164 on 1 F

ebruary 1992. Dow

nloaded from

168 Shorrock, Rees

TABLE II Effect ofindomethacin (50 mg three times daily) ongastric prostaglandin E2 degredation capacity (pmollmglminute). (Values, mean (SEM))

Treatment days

TimeO 1 7 28

Fundus 325 (45) 346 (86) 483 (279) 354 (71)Antrum 511 (122) 543 (119) 646 (226) 758 (263)

n=10.

With NSAIDs this is clearly not the case.Because of the cyclo-oxygenase inhibitory actionof these drugs (responsible for their anti-inflammatory properties) mucosal prosta-glandins are reduced, as shown in our results andalso those of others.3-33 During treatment withindomethacin, PGE2 remains significantlyreduced from values beforehand, despite endo-scopic evidence of mucosal recovery. It wouldtherefore seem that adaptation occurs indepen-dently ofmucosal PGE2 metabolism. We did notmeasure PGE2 synthesis directly but clearly thereduction in PGE2 content of mucosal homo-genates with indomethacin is a result of reducedsynthesis, as PGE2 degredation capacity was notaffected by indomethacin in our studies. Thefinding of a lack of correlation between themagnitude of the reduction in PGE2 and mucosaldamage, coupled with the data showing mucosalrecovery despite continued reduction in PGE2provides further evidence of a cause and effectrelationship between inhibition of mucosalprostaglandins by NSAIDs and gastric mucosaldamage.31-33How indomethacin causes the initial damage

and the subsequent adaptation with continuedintake remain unknown. Previous work involv-ing histological studies has suggested that indo-methacin induced injury is focal, with little in theway of diffuse change.2434 We have shown areduction in the gastric mucosal blood flow withindomethacin, which confirms earlier findingsby Konturek et al. 18 In addition, we have found agood correlation between the severity of themucosal damage and the magnitude of the reduc-tion in mucosal blood flow. It has been suggestedthat endogenous prostaglandins contribute tothe maintenance of basal gastric mucosal bloodflow in animals and man.'9 Thus, reduction ofendogenous prostaglandins by indomethacin'scyclo-oxygenase inhibitory action may explainthe reduction in gastric mucosal blood flow seenin our subjects. It could be hypothesised thatendogenous prostaglandin reduction by oralindomethacin is not uniform, possibly because ofgreater inhibition of mucosal cyclo-oxygenase inareas with greater local concentrations of indo-methacin. This, coupled with more generalisedreductions in mucosal defence - for examplereduced bicarbonate secretion and reduced sur-face hydrophobicity - may result in the focaldamage described.

In our studies, mucosal recovery was associ-ated with a return of blood flow to normal despitea continued reduction in PGE2. How mucosalblood flow regulates itself in this situationremains unclear but an increase in blood flowwould enhance mucosal defence and repair ofinjury and contribute to the adaptive process. As

the mucosal blood flow was not increased aboveits values before treatment, however, it isdifficult to envisage its role in preventing furtherdamage during continued exposure to the drug.Clearly, other mechanisms play an importantrole in the adaptive process.The mechanisms of the initial damage by

indomethacin and subsequent adaptation to con-tinued intake therefore remain unknown. Theclear evidence of an adaptive response in mostsubjects should provide further stimulus fordefining the mucosal mechanisms responsiblefor it.

1 Lanza FL. Endoscopic studies of gastric and duodenal injuryafter the use of ibuprofen, aspirin, and other non-steroidalanti-inflammatory agents. AmJ Med 1984; 77: 19-24.

2 Caruso I, Bianchi-Porro G. Gastroscopic evaluation of anti-inflammatory agents. BMJ 1980; 280: 75-8.

3 Rees WDW, Turnberg LA. Reappraisal of the effects ofaspirin on the stomach. Lancet 1980; i: 410-3.

4 Somerville K, Faulkner G, Langman MJS. Non-steroidal anti-inflanmmatory drugs and bleeding peptic ulcer. Lancet 1986;i: 462-4.

5 Walt R, Katschinski B, Logan R, Ashley J, Langman MJS.Rising frequency of ulcer perforation in elderly people in theUnited Kingdom. Lancet 1986; i: 489-92.

6 Armstrong CP, Blower AL. Non-steroidal anti-inflammatorydrugs and life-threatening complications of peptic ulcera-tion. Gut 1987; 28: 527-32.

7 Vane JR. Inhibition of prostaglandin synthesis as a mechanismof action of aspirin-like drugs. Nature New Biol 1971; 231:232-5.

8 Whittle BJR, Higgs GA, Eakins KE, Moncada S, Vane JR.Selective inhibition of prostaglandin production in inflam-matory exudates and gastric mucosa. Nature 1980; 284:271-3.

9 Whittle BJR. Prostaglandin cyclo-oxygenase inhibition and itsrelationship to gastric damage. In: Harmon JW, ed. Basicmechanisms of gastrointestinal mucosal cell injury and protec-tion. Baltimore: Williams and Wilkins, 1981: 197-210.

10 Rees WDW, Gibbons LC, Turnberg LA. Effects of NSAID'sand prostaglandins on alkali secretion by rabbit fundus invitro. Gut 1983; 24: 784-9.

11 Rainsford KD. The effects of aspirin and other NSAID's ongastrointestinal mucus glycoprotein biosynthesis in vivo:relationship to ulcerogenic actions. Biochen Pharmacol1978; 27: 877-85.

12 Menguy R, Masters YF. Effects of aspirin on gastric mucussecretion. Surg Gynecol Obstet 1965; 120: 92-8.

13 Lichtenberger LM, Richards JE, Hills BA. Effect of 16,16dimethyl prostaglandin E2 on surface hydrophobicity ofaspirin treated canine gastric mucosa. Gastroenterology 1985;88: 308-14.

14 Wallace JL, McKnight GW. The mucoid cap over superficialgastric damage in the rat. A high pH microenvironmentdissipated by nonsteroidal anti-inflammatory drugs andendothelin. Gastroenterology 1990; 99: 295-304.

15 Ashley SW, Sonnenschein LA, Cheung LY. Focal gastricmucosa blood flow at the site of aspirin-induced ulceration.Amj Surg 1985; 149: 53-9.

16 Main IHM, Whittle BJR. Investigation of the vasodilator andantisecretory role of prostaglandins in the rat gastric mucosaby use of non-steroidal anti-inflammatory drugs. Br JPharmacol 1975; 53: 217-24.

17 Kaufman GL Jr, Aures D, Grossman MI. Indomethacindecreases basal gastric mucosal blood flow. Gastroenterology1979; 76: 1165.

18 Konturek SJ, Kwiecien N, Obtulowicz W, et al. Effect ofcarprofen and indomethacin on gastric function, mucosalintegrity and generation of prostaglandins in man. Hepato-gastroenterology 1982; 29: 267-70.

19 Guth PH, Leung FW. Physiology ofthe gastric circulation. In:Johnson LR, ed. Physiology of the gastrointestinal tract. 2nded. New York: Raven Press, 1987: 1031-53.

20 Langman MJS. Epidemiological evidence of the associationbetween peptic ulceration and anti-inflammatory drug use.Gastroenterology 1989; %: 640-6.

21 Lanza FL, Royer GL, Nelson RS, Chen TT, Seckman CE,Rack MF. The effects of ibuprofen, indomethacin, aspirin,naproxen and placebo on the gastric mucosa of normalvolunteers. Dig Dis Sci 1979; 24: 823-8.

22 Robert A. Cytoprotection by prostaglandins. Gastroenterology1979; 77: 761-7.

23 Graham DY, Smith JL, Splut HJ, Torres E. Gastric adapta-tion: studies in humans using continuous aspirin administra-tion. Gastroenterology 1988; 95: 327-33.

24 Shorrock CJ, Prescott RJ, Rees WDW. The effects ofindomethacin on gastroduodenal morphology and mucosalpH gradient in the healthy human stomach. Gastroenterology1990; 99: 334-9.

25 Bonner R, Nossal R. A model for laser Doppler measurementsof blood flow in tissue. Appl Opt 1981; 20: 2097-107.

26 Shepherd AP, Riedel GL. Continuous measurement ofintestinal mucosal blood flow by laser-Doppler velocimetry.Am3r Physiol 1982; 242: G668-72.

on October 7, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.33.2.164 on 1 F

ebruary 1992. Dow

nloaded from

Mucosal adaptation to indomethacin induced gastric damage in man - studies on morphology, bloodflow, andprostaglandin E2 metabolism 169

27 Kiel JW, Riedel GL, DiResta GR, Shepherd AP. Gastricmucosal blood flow measured by laser-Doppler velocimetry.AmJPhysiol 1985; 249: G539-45.

28 Crampton JR, Gibbons LC, Rees WDW. Simultaneousmeasurement of in vitro gastroduodenal prostaglandin E2synthesis and degredation in peptic ulcer disease. Scand JrGastroenterol 1987; 22: 425-30.

29 Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Proteinmeasurement by the Folin phenol reagent. J Biol Chem1951; 193: 265-75.

30 Hawkey CJ, Kemp RT, Walt RP, Bhaskar NK, Davies J,Filipowcz B. Evidence that adaptive cytoprotection in rats isnot mediated by prostaglandins. Gastroenterology 1988; 94:948-54.

31 Redfern SJ, Lee E, Feldman M. Effect of indomethacin ongastric mucosal prostaglandins in humans - correlation withmucosal damage. Gastroenterology 1987; 92: 969-77.

32 Levine RA, Petokas S, Nandi J, Enthoven D. Effects ofnonsteroidal antiinflammatory drugs on gastrointestinalinjury and prostanoid generation in health volunteers. DigDisSci 1988; 33: 660-6.

33 Ligumsky M, Sestieri M, Karmeli F, Zimmerman J, Okon E,Rachmilewitz D. Rectal administration of nonsteroidal anti-inflammatory drugs. Effect on rat gastric ulcerogenicity andprostaglandin E2 synthesis. Gastroenterology 1990; 98:1245-9.

34 Laine L, Marin-Sorensen M, Weinstein WM. The histology ofgastric erosions in patients taking NSAID's: a prospectivestudy. Gastroenterology 1988; 94: A247.

on October 7, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.33.2.164 on 1 F

ebruary 1992. Dow

nloaded from